Standard Languages and Language Standards - Greek, Past and Present-Ashgate (2009) PDF

Standard Languages
and Language Standards:

Greek , Past and Present
Edited by
Alexandra Georgakopoulou
and
Michael Silk
STANDARD LANGUAGES
AND LANGUAGE STANDARDS:
GREEK, PAST AND PRESENT
Centre For Hellenic Studies
King’s College London
Publications 12
Standard Languages
And Language Standards:
Greek, Past and Present
edited by
Alexandra Georgakopoulou and Michael Silk
Copyright © 2009 Alexandra Georgakopoulou and Michael Silk
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, photocopied, recorded, or otherwise without the
prior permission of the publisher.
Published by Ashgate Publishing Limited
Wey Court East, Union Road.
Farnham, Surrey GU9 7PT
United Kingdom
Ashgate Publishing Company
Suite 420
101 Cherry Street
Burlington, VT 05401–4405
USA
The editors have asserted their moral right under the Copyright, Designs and Patents Act,
1988, to be identified as the editors of this work.
Ashgate website http://www.ashgate.com
British Library Cataloguing in Publication Data

Standard Languages and Language Standards: Greek, Past and Present. – (Publications of the
Centre for Hellenic Studies, King’s College London)
1. Greek language – History. 2. Greek language - Standardization.
I. Series II. Georgakopoulou, Alexandra. III. Silk, M.S.
480.9-dc22
US Library of Congress Cataloging-in-Publication Data
Standard Languages and Language Standards: Greek, Past and Present / edited by Alexandra
Georgakopoulou and Michael Silk.
p. cm. – (Publications of the Centre for Hellenic Studies, King’s College
London)
Includes index.
1. Greek language – History. I. Georgakopoulou, Alexandra. II. Silk, M.S.
PA227.S73 2009
480’.9-dc22 2009008259
Typeset by W.M. Pank, King’s College London.
THE CENTRE FOR HELLENIC STUDIES,

KING’S COLLEGE LONDON, PUBLICATIONS 12
Contents
Foreword vii
Contributors ix
Introduction: Greek Language-Standardizing, Past, Present and Future xi

Part I Establishing a Standard

1. The Invention of Greek: Macedonians, Poets and Others 3
Michael Silk
2. The Greek Koine and the Logic of a Standard Language 33
Stephen Colvin
3. Primary Education in a Non-Standard Language as a Tool of Social
and National Integration: The Case of Vernacular Greek, 1923–30 47
George Kritikos
4. Greek With No Models, History or Standard: Muslim Pontic Greek 67
Pietro Bortone
Part II Standardization Practices

5. The Lexica of the Second Sophistic: Safeguarding Atticism 93
Claudia Strobel
6. Grammatical Metaphor and the Function of Participles in
High-Register Versions of the Life of Aesop 109
Chrystalla Thoma
7. Orthographic Standardization of the Modern Greek Classics:
Gain and Loss 131
David Ricks
8. Correcting the Courtroom Cat: Editorial Assaults on Cavafy’s Poetry 149
Anthony Hirst
9. Modern Greek Dictionaries and the Ideology of Standardization 167
Assimakis Tseronis and Anna Iordanidou
10. Greek in Cyprus: Identity Oscillations and Language Planning 187
Dimitra Karoulla-Vrikki
11. ‘Greeklish’: Transliteration Practice and Discourse in the Context of
Computer-Mediated Digraphia 221
Jannis Androutsopoulos
v
vi Contents
Part III Ideologies and Contestations

12. A Tradition of Anomaly: Towards the Regularization of the Greek
Language 253
Emmanuel Kriaras
13. Mother and Daughters, Roots and Branches: Modern Greek
Perceptions of the Relationship between the Ancient and Modern
Languages 259
Peter Mackridge
14. Constructing a Science of Language: Linguistics and Politics
in Twentieth-Century Greece 277
Effi Gazi
15. ‘Language Issues’ after the ‘Language Question’: On the Modern
Standards of Standard Modern Greek 293
Spiros Moschonas
16. Competing Ideologies and Post-Diglossia Greek: Analysing the
Discourse of Contemporary ‘Myth-Breakers’ 321
Dionysis Goutsos
17. Korais and the Second Sophistic: The Hellenistic Novel as Paradigm
for a Modern Literary Language 341
Roderick Beaton
Index of names 355
Subject index 361

Foreword
As is made clear in A.G.’s Introduction (pp. xi–xxviii below), and as will be

apparent from the seventeen chapters that follow it, this is a volume of unusual
scope. In disciplinary and methodological terms, it ranges from systemic functional
linguistics to historical geography, from literary criticism to sociolinguistics and
discourse analysis; in terms of chronological spread, the points of reference cover
three millennia, and more. On an intellectual level, this breadth has been at the
heart of our enterprise. On the purely practical level, it has inevitably given rise to
specific challenges, which the editors have done their best to meet.
Our hope is that interested readers will seek enlightenment from the volume
as a whole, and not merely from particular chapters that appeal most obviously to
their own specialist perspectives or agendas. Accordingly, we have tried to make
sure that each chapter, whatever its disciplinary standpoint, is self-explanatory,
so far as possible. In the same spirit, we have, for instance, imposed conventions
of referencing and footnoting more usual in the world of the intelligent ‘general
reader’ than in much technical scholarship, while (more simply) we have ensured
that all quoted Greek (Ancient, Medieval or Modern) is accompanied by an English
translation. All this, in the expectation of making what we believe are important,
and mutually illuminating, discussions more accessible.
In the vexed matter of English transliteration of Greek names, however, we have
not sought to impose any uniform system on all chapters, though we have, of course,
insisted on self-consistency within each chapter, as appropriate. In a volume like
this one, where standardization (and the problematizing of standardization) is a
substantive concern, it would indeed have been arbitrary, not to say incongruous,
to override authorial decisions, whether these reflect ideological conviction, a
sense of historical propriety, or whatever. The upshot is: Korais is usually ‘Korais’,
but sometimes ‘Koraes’; ‘Frangoudaki’ and ‘Fragoudaki’ are in complementary
distribution; and Heliodorus is ‘Heliodorus’ in one chapter, and ‘Heliodoros’ in
another. In the index of proper names (pp. 355–9 below), Greek names are cited
under the (most) usual form, but with alternative spellings in brackets.
Abbreviations pose practical problems of their own. In the interests of
accessibility, once again, we have sought to ensure that, so far as is practicable, these
are explained, or made self-explanatory, within each chapter, with any exceptions
strictly confined to references in footnotes. Most of these exceptions concern
From Standard Languages and Language Standards: Greek, Past and Present, ed. Alexandra
Georgakopoulou and Michael Silk. Copyright © 2009 by Alexandra Georgakopoulou
and Michael Silk. Published by Ashgate Publishing Ltd, Wey Court East, Union Road,
Farnham, Surrey, GU9 7PT, UK.
vii
viii FOREWORD
material from the ancient world; and in chapters dealing with such material, any
unexplained abbreviations (of authors, works, collections of source material) will
broadly follow those in either LSJ (Liddell–Scott–Jones, Greek-English Lexicon, 9th
edn, Oxford, 1925–40) or OCD (The Oxford Classical Dictionary, ed. Hornblower
and Spawforth, 3rd edn revised, Oxford, 2003).
The seventeen chapters, as submitted and as edited, abound in interconnections,
and on various different levels. We could have added numerous cross-references
between chapters, but these (we have judged) would have been so numerous, in
many cases, as to make the practice self-defeating. Instead, we have supplied such
cross-references only in special cases, or of course where our authors opted for
them. The main substantive connections between individual chapters are, however,
flagged by A.G. in her Introduction.
Beyond this Foreword (short and essentially practical), our volume contains no
joint statement by the two editors, but rather A.G.’s Introduction and a separate
Chapter 1 by M.S.S. To avoid any misunderstanding, however, it should be noted
that the collection as a whole reflects the editorial efforts of both editors, who take
joint responsibility for the entire volume.
In conclusion, we wish to thank our contributors for their patience and tolerance
of what may, at times, have felt like undue editorial interference. We gratefully
acknowledge the timely assistance of Korina Giaxoglou with the translation of
an original Greek text for Chapter 12, and of Anna Charalambidou and Irene
Theodoropoulou with the subject index, the help and support of Charlotte
Roueché, from the time of the Logos conference to the production of this volume,
and the skill and patience of Wendy Pank, who prepared the electronic copy from
which the volume has been printed.
A.G.
M.S.S.
Contributors
Alexandra Georgakopoulou is Reader in Modern Greek Language and Linguistics

at King’s College London.
Michael Silk is Professor of Classical and Comparative Literature at King’s College

London.
Jannis Androutsopoulos is Reader in Sociolinguistics and Media Discourse at King’s

College London.
Roderick Beaton is Koraes Professor of Modern Greek and Byzantine History,

Language and Literature at King’s College London.
Pietro Bortone is Assistant Professor in Modern Greek Studies at the University of

Illinois at Chicago.
Stephen Colvin is Reader in Classics and Comparative Philology at University

College London.
Effi Gazi is Assistant Professor of History and Theory of Historiography at the

University of Thessaly.
Dionysis Goutsos is Assistant Professor of Text Linguistics at the University of

Athens.
Anthony Hirst is Lecturer in Modern Greek at Queen’s University Belfast.
Anna Iordanidou is Associate Professor of Modern Greek Language at the

University of Patras.
Dimitra Karoulla-Vrikki is Assistant Professor of Linguistics at the European

University Cyprus.
Emmanuel Kriaras is Honorary Professor of Medieval Philology at the Aristotle

University of Salonica.
ix
x Contributors
George Kritikos is Associate Professor in Historical Geography at Harokopion

University, Athens.
Peter Mackridge is Emeritus Professor of Modern Greek, and an Emeritus Fellow of

St Cross College, at the University of Oxford.
Spiros Moschonas is Senior Lecturer in Linguistics and Philosophy of Language at

the University of Athens.
David Ricks is Professor of Modern Greek and Comparative Literature at King’s

College London.
Claudia Strobel is a doctoral student at Corpus Christi College, Oxford.
Chrystalla Thoma received her doctorate from the University of Bremen; she
currently works at the European University Cyprus.
Assimakis Tseronis is a Junior Researcher at the University of Leiden.

Introduction:
Greek Language-Standardizing,
Past, Present and Future
This volume grew out of the Logos Conference, which took place in London
on 9–11 September 2004, jointly organized by the Centre for Hellenic Studies,
King’s College London, and the Society for the Promotion of Hellenic Studies.
The publicity for the conference stressed the exploratory nature of the event: it
sought to bring together scholars with a professional interest in the Greek language
from different perspectives and, above all, with expertise in very different periods.
The goal was to exchange ideas and concerns and engage in the sort of dialogue
that disciplinary and chronological boundaries rarely allow.
Many of those involved in the conference were aware then (as the editors
certainly have been, throughout the preparation of this volume) that a narrative
of exceptionalism, combined with an ideologically charged and often overstated
argument in favour of the continuity of the Greek language, has (arguably)
encouraged many scholars to shy away from the project of putting the language’s
past and present under scrutiny on the basis of a single focal concern. Chronological
boundaries have in this case served as demarcation lines par excellence. In opposition
to this tendency, our aim in the Logos conference, and by extension in this volume,
has been to foreground thematic and analytical affinities on the broad topic of
language standardization and standards, precisely at the expense of an approach
that would privilege a linear chronological perspective. In this spirit, the volume is
organized thematically into three parts – Establishing a Standard, Standardization
Practices, Ideologies and Contestations – each of which covers a wide chronological
span.
The inclusion of chapters in the volume has been decided on the basis of their
representative and illustrative value, and on the understanding that, far from aiming
at an exhaustive or monologic account,1 our intention is to provide ‘snapshots’
from the long and complex history of Greek. In fact, our contributors tend to
detect, not seamless transitions or deceptively homogeneous representations, but
1
Two recent histories of the Greek language in English are notable: a definitive overview by
Horrocks (1997) and the monumental edited volume by Christidis (2007), which ostensibly goes up to
late antiquity, but in fact ventures much further.
Farnham, Surrey GU9 7PT, UK.
xi
xii ALEXANDRA GEORGAKOPOULOU
disjunctures, fragmentations and discontinuities in a never-ending and multi-

faceted story of standardization, which subsumes many moves and counter-moves.
In addition, none of the analyses in our collection falls into the trap of equating
standardization choices with one particular agenda only or with clearly delimited
boundaries of ‘before’ and ‘after’; nor have they set out to highlight uniqueness
and lack of comparability with other languages. Rather, the truism that the Greek
language, from antiquity to the present day, has had a long and distinctive history
(of standardization too) serves here as the impetus for documenting the subtleties
of standardization processes and the often elusive discourses (official and lay) that
have actively shaped and been shaped by them. In this sense, all the chapters do
something to uncover the importance of socio-cultural macro-forces (as against
hard linguistic ‘facts’) for the construction of ideas of correctness and, more
broadly, value, in respect of the Greek language, at various points in its history. It
is intriguing to see in this process comparabilities emerging with diverse languages,
sociolects and genres, from modern standard Arabic to British popular songs. It is
also instructive to see which interrelationships each analyst seeks to establish and
how these are relatable to (and arguably motivated by) their various disciplinary
standpoints.
Standardization, practices of control and perceptions of correctness in Greek
have been intimately linked with a long history of diglossia and have been – and
continue to be – implicated in ideological and political projects that many of our
analysts themselves have been part of. More than three decades after the official
resolution of the so-called ‘language question’, as the chapters on the Modern
Greek situation make amply clear, diglossia has not gone away from our analysts’
interpretative narratives nor, for that matter, from Greek speakers’ repertoire of
devices for making sense of their language and their socio-cultural reality. However,
what this volume makes clear is that the time is ripe for the language question to be
tackled with dispassionate and reflexive accounts that allow the researcher’s own
role in the analysis to be problematized and that thereby move the discussion of
standardization in Greek away from the increasingly unrewarding polarizations of
the past.
Correlatively, meanwhile, another thread that runs through our collection is a
theoretical and empirical commitment to situated accounts that accept from the
outset that language is contextualizing and contextualized, and that any study
of language should shed light on its interrelationships with the local and socio-
cultural context of its occurrence. Most of the chapters, certainly, are focused on
‘the context of prescription – who prescribes for whom and for what purposes’2
– and not on earlier dichotomous views of language description as ‘good’ and
prescription as ‘bad’.
2
Cameron (1995) 11.
INTRODUCTION xiii
Part 1: Establishing a Standard

The authors of the chapters in Part I broadly subscribe to the idea that standardization
is a never-ending and gradual process and that crystallizing moments or cases for
analysis serve the purpose of affording us glimpses of standardization at work. They
would also, no doubt, agree with the familiar claim that a language is a dialect
with an army: standardization is mostly the outcome of socio-political and cultural
considerations, sometimes accidents as well, and in that process strictly defined
linguistic criteria tend to have little impact. In this first set of chapters, too, we may
detect operative concepts and themes that will figure prominently throughout the
volume as a whole. Among much else, these opening discussions offer us a historical
paradigm for the emergence of diglossia by documenting the close links between
language standardization and identity building and identification. At various
points in the history of Greek, this has involved the routine standardization choice
of reviving the language’s past, of looking back for current models of language use.
Finally, here as elsewhere in the volume, our analysts are reflexively aware of the
fact that it is not only language standardization that is value-laden and ideologically
charged; any language-focused analysis of it is bound to be so too.
On this note, Michael Silk asks us as analysts to take seriously the concept of
value in language choice and language use, and not to shy away from judgements
about different social, or other, varieties. Contrasting the situation of dialects and
literary standards in early Greece with later developments, his chapter deplores the
fact that the Attic-based koine, institutionalized by the Macedonian kings Philip
and Alexander from the latter part of the fourth century BC, was never a literary
medium. Put differently, standardization happened in this case (as in many others
too) in the direction of a non-literate model that served administrative functions,
a development which Silk reads as a loss. Practical considerations always play a
role in the choice of a language variety to be standardized, and the very process
of standardization, as we shall see in other chapters (Strobel’s, for one), involves
a gradual codification and elaboration of the functions of that variety. What is
important in this chapter is to see the links between the standardization of the
koine, Athens’ political and cultural dominance in the fifth century, and identity-
building: this last centred, in the first place, on the linguistic articulation of a
collective sense of identification as Hellenes, defined oppositionally (us versus
them) after the Persian Wars. Standardizing (we may agree) is interwoven with
‘acts of identity’3 that are not always collective or consensual.
In his review of the earlier situation of the Greek dialects, Silk searches for
parallels in the Americanizing practices of British popular song-performance in the
1930s, whereas the subsequent use of Attic as the language of the Macedonian
elite is seen to be comparable to the use of French in the European courts of a
later age. Silk’s term ‘generic dialectalization’ brings to the fore a phenomenon
that has been under-represented in sociolinguistic discussion of standardization,
3
Le Page and Tabouret-Keller (1985).
xiv ALEXANDRA GEORGAKOPOULOU
perhaps as a result of the serious problematizing of the terms language and dialect.
Generic dialectalization involves a close association between a particular genre
and a particular dialect, such that a non-standardized language variety can assume
(covert) prestige and be authenticated through its conventional association with
certain ways of telling and acting in certain settings. At the same time, in the
Greek case, the ‘particular’ dialect is almost always a highly generalized dialect,
rather than a close equivalent of any one speech variety. Generic dialectalization
can be seen at work most strikingly in the case of early Greek lyric poetry, which
was generally (but not always) in some version of Doric accommodated to the
composite dialect of epic. This model of generic dialectalizing is in tune with the
view of standardization as presenting degrees and shades rather than being an all-
or-nothing matter; different language varieties (to use a more neutral term) may be
implicated in standardization in varying ways.
Stephen Colvin’s complementary chapter, more sociolinguistic in character,
scrutinizes the factors that led to the rise of the Attic-based koine. Colvin is up-
front in acknowledging that sociolinguistic study of the past, unlike any synchronic
analysis, is bound to involve imaginative reconstruction on the part of the analyst;
up-front, too, in making it clear that the particular model of language change and
variation the analyst subscribes to will ultimately shape the results of the analysis.
Colvin goes on to critique (albeit in broad terms) conventional (Western) classical
scholarship on the history of Ancient Greek, and on several grounds: it prioritizes
a chronological reading of standardization as a neat and linear process; it over-
privileges literary texts; and it assumes the propriety of ‘reading back’ from the
peculiar sociolinguistic contexts of Western Europe and North America, where
the prevalent language model is that of nation states with colonial histories and
standardized national languages. In striking contrast, Colvin looks to modern
standard Arabic for a model to understand the Greek koine: in this light, the
koine is seen to have constituted a written standard to which no spoken variety
corresponded exactly, while (crucially) adherence to this standard was not
necessarily, or not at all, a superimposed coercive process but followed on from the
feelings of the speakers in question about their linguistic identity.
Put somewhat differently, Colvin’s discussion presents the koine as the inevitable
outcome of the development of writing, and what emerges from it is a notion of
diversity and pluralization in the concept of koine itself: from this point of view,
we can talk about literary koinai, where Silk for instance would talk about generic
dialectalizing. More significantly still, Colvin invites us to include in the analysis
of the formation of the koinai texts and sources that have hitherto been under-
represented; he provides a prime example of this opening-up with his focus on
inscriptions and their phonological or morphological differentiae, along with their
public or private status. It is instructive to see the way that, before the ultimate
dominance of the Attic-based koine, differences between regional epigraphic
standards were deliberately maintained to create and reaffirm distinctiveness.
As we shall see in Androutsopoulos’ chapter in Part II, this pattern is strikingly
INTRODUCTION xv
reminiscent of the way that different contemporary communities of e-mail users

choose from the available transliteration options for Greek in Latin characters;
they do so not haphazardly, but in ways that index group-belonging.
George Kritikos turns his attention to attempts at standardizing the Greek
vernacular of the time through the educational reforms of 1929, when demotic
became the language of teaching in all years of the elementary schools for the first
time in Greek history. In the historical context of standardization in the Modern
Greek nation-state, it was a somewhat unorthodox choice to standardize in
the direction of the ‘low’ variety. The decision is linked by Kritikos to the aim
of integrating foreign-speaking and refugee populations and of strengthening a
national sentiment amongst them through a language that (by comparison with
katharevousa) they had more chance of understanding. Once again, we see that
the institutional choice of a language variety as ‘standard’, which in this case really
means ‘official’, is an act of identity-building interwoven with narratives of nation-
building and specific socio-political conditions.
In all three chapters so far, we have also seen how standardization and education
go hand in hand; and, as in ancient Greece the koine served the purposes of (and
ultimately galvanized) an educated elite, in the case explored by Kritikos an uneasy
co-existence of demotic with katharevousa (which was still dominant in secondary
education) failed to produce any upward mobility for the populations which the
reforms were designed to serve in the first place. It went as far as allowing them
to enter the labour market but, as Kritikos’ chapter shows, it ultimately served
as a mechanism for the (re)affirmation of social and economic inequality. In this
chapter too, standardization emerges as a gradient notion, with different degrees
and projects of standardization linked to different – and sometimes competing –
social and cultural processes and serving different social and cultural groups in
different ways.
The last chapter in Part I by Pietro Bortone looks at a remarkable case: a
language variety completely lacking in standardization which at the same time is
in need of standardization as a means of linguistic revival – or survival. Bortone’s
focus is on isolated villages in the easternmost part of northern Turkey, where
a fast disappearing form of Greek (Romayka) is still spoken. Romayka has had
no exposure to Standard Greek or to Greek policies of purism, standardization
and archaization, despite preserving numerous archaic elements in its own right,
precisely because of its historical isolation. The speakers, of essentially Turkish
and Muslim identity, descend in part from Greeks who converted to Islam over 300
years ago. It is instructive to see this dying variety contrasted with the related Pontic
dialect. Ever since Pontian Greeks arrived in Greece in 1923 (with the population
exchange), their version of Pontic Greek has undergone standardization in the
direction of Standard Modern Greek. At the same time, almost as a counter-move,
it has undergone a largely hegemonic institutionalization and monumentalization
of the regional: a process which we have seen in operation in the last two decades
in many other instances outside Greece.
xvi ALEXANDRA GEORGAKOPOULOU
The other noteworthy fact about Romayka is of course that it presents ‘no
models, no history, no standard’ (as Bortone stresses), if, but only if, we consider
it from the point of view of Greek. In fact, Romayka has been decisively shaped
by Turkish in ways that may well remind us of Arvanitika, and in particular the
way that Arvanitika too has been shown to be a dying variety with substantial
interference from Greek.4 In a chilling reminder of how little impact linguistic
factors are liable to have in any decisions and projects of language standardization,
Bortone shows how, although on the basis of linguistic criteria (quantitative,
functional, communicative) Romayka could easily be classified as a dialect, and
a case for standardization could be made, there is simply no such chance on
the basis of socio-political criteria. Bortone’s discussion also neatly exposes the
arbitrariness involved in which language variety is labelled as a ‘dialect’, and which
as a ‘language’. We are forcibly reminded of the way that the notion of standard
became one of the significant apparatuses of the modern nation-state for the
creation of boundaries between the ‘national’ and the ‘sub-national’.5
Part II: Standardization Practices

In one of the most eloquent accounts of communicative practices, Hanks defines
them as essentially socio-cultural moments of synthesis for linguistic forms with
the relatively stabilized form of social activity in which they occur and with the
ideology that serves as a system of evaluation for those activities.6 The chapters
in Part II document this kind of synthesis with regard to standardization. They
bring to the fore the relationships between the significant variables: what gets
standardized? how? in what type of activity or social arena? and with what kinds
of ideological project? We are shown that it is in everyday regimented practices
that macro-processes and social forces are articulated. The small – for instance, the
standardization of orthographical minutiae in a literary work (see Ricks and Hirst)
– constitutes the big.
Claudia Strobel looks at the lexical Atticism of the Second Sophistic, in
particular the developments of the second century AD, as an early but archetypal
example of standardization involving codification.7 As a counter-movement to the
koine (the formation of which is debated by Silk and Colvin), Atticism sought to
imitate and revive the Attic dialect of the classical period. As we shall see from
subsequent contributors too (Tseronis and Iordanidou, among others), going back
4
Tsitsipis (1998).
5
Where ‘sub-national’ subsumes regional, provincial, dialectal, ‘native’, colonial, ‘common’, etc.:
cf. Blommaert (2009).
6
Hanks (1996) 230.
7
Haugen (1966) specified that the route to standardization of a selected variety involves four
processes: codification (graphemization and grammatical definition), elaboration (development of
vocabulary and stylistic variants), endorsement and implementation by the state (political support), and
finally endorsement and implementation by the community (public support). The linear conception of
Haugen’s sequence has been subjected to significant problematization in more recent studies, but what
is important to note is the unquestionable role of codification in any standardization process.
INTRODUCTION xvii
to older forms of language (archaizing) is in one form or another a leitmotif of the

history of Greek. It is characteristic that such archaizing is closely associated with
the educated or bourgeois elite (Thoma’s discussion is of special interest here),
and it is characteristic too that the archaizing project is wrapped in a narrative of
identity-building. The premise of all such attempts is the intimate relationship that
all languages develop with specific social meanings: reviving a language of the past
is thus aimed at iconically restoring power and prestige, political and cultural.
As Strobel shows, these were precisely the roots of the Second Sophistic: the
Greek elite sought to maintain political independence as far as was possible under
Roman rule, and in that endeavour language became a way of demonstrating
Greekness or cultural identity. Strobel discusses three lexicographers, Phrynichus,
Moeris and Pollux. She shows how, in addition to seeing themselves as ‘guardians of
language’ who reinforced standards of correctness, these lexicographers recognised
the importance of organizing their dictionaries well and themselves producing
new ‘designs’ that were not only practical, but aesthetically beautiful in their own
right. It may seem unsurprising that these early lexicographical attempts were not
informed by any fully articulated linguistic theory, as tends to be the case today
(compare Tseronis and Iordanidou’s chapter). Nevertheless, lexicography both
then and now is ultimately shaped by the socio-cultural contexts and intellectual
milieux in which the lexicographers operate. Which discourses about language
they subscribe to, and which standards of correctness – it will probably be widely
circulating discourses and standards – we may trace in their individual choices;
but, as these processes are mutually constitutive, their work will (re)shape those
discourses and standards in turn.
In Chrystalla Thoma’s chapter, attention shifts to how the use of specific
linguistic devices can be related more or less directly to certain aspects of context.
Her framework of analysis, systemic functional linguistics, is in fact premised on
the idea that the choice of specific language forms is not accidental but rather
fulfils specific functions and communicative purposes in the genre and context
of occurrence. In similar vein, the frequency of language forms is an important
indicator of the kind of genre and the purposes that the genre routinely fulfils. As
in the case of Silk’s ‘generic dialectalizing’, but actually in the opposite direction of
‘generic standardization’, Thoma too shows us how genre becomes conventionally
associated with a specific kind of language. Her focus is on two versions of the Life
of Aesop from the fourteenth century, versions in classicizing Greek written for the
educated few. The notion of genre here is intimately linked with that of register,
which, within systemic functional linguistics,8 is seen as the primary concept for
explaining linguistic variation and as the composite of three interacting semiotic
properties of the situational context of language. In essence, these properties are: field
(what the language is about); mode (the channel through which communication
is carried out); and tenor (the language users and the level of formality in their
relationship). The interactions between these aspects of context define a register as
8
See e.g. Halliday and Matthiessen (2004).
xviii ALEXANDRA GEORGAKOPOULOU
high or low: in this case, as Thoma shows, high. She reaches this conclusion on the
basis of the increased frequency of participles and ‘grammatical metaphor’ in the
Lives of Aesop under review: an increased frequency which, in her terms, creates
a ‘lexically packed and dense style’ and in turn indexes ‘authority’.
Thoma’s analysis reveals an interesting shift, in certain instances of fourteenth-
century narrative, from the experiential and largely oral-based style that is
normally associated with narrative to a more developed ‘textuality’ (in the sense of
literacy-based style) and, in particular, to a style that is currently associated with
the scientific register. These findings are in tune with the impressive (and well-
documented) co-existence of oral and literate style that is especially characteristic
of narrative genres in Greek literary history, and which is mainly attributable to the
long-standing tradition of diglossia.9
In David Ricks’s chapter, more recent, and more canonical, specimens of
literature come under scrutiny. Ricks takes the example of some ‘Modern Greek
classics’, and looks at their orthographic standardization by editors as a set of
complex decisions that involve both gain and loss. Standardizing as harmonization
to modern norms involves the opposite of what we have seen in previous chapters.
Through a situated account, Ricks examines the way that each of the literary works
under review – works by Makriyannis, Papadiamandis, Cavafy and Solomos –
have presented editors with a different set of orthographical choices, and notes the
relevance of the distinctive publication and reception histories of each. With Ricks
(as with Strobel), we see both authors and editors as agentive social actors whose
attitudes to language use and correctness are individual and biographical on the
one hand and, on the other, inextricably linked to the wider contexts in which they
operate. It is also instructive to see how different authors have displayed awareness
of, or resistance to, the textual control of their work, in varying degrees and in
different ways. In particular, Cavafy (as Hirst’s chapter will confirm), is shown to
be a case par excellence of a creator who goes to great lengths to police his creation
against any editorial intervention.
Throughout our volume, we find standardization involving delicate and
complex gauges of calculation regarding decisions that are too often reduced to
a simple choice. The issue, again and again, is whether to embrace and legitimize
contemporary and (by and large) vernacular uses of language, or, conversely, to
‘monumentalize’ language (one of the metaphors that Mackridge and Gazi discuss
in Part III) and, in that respect, to keep the traces of history intact, including the
supposed etymological transparency of words. Ricks’s chapter does not offer any
easy answers to this dilemma. What he does is critically assess the editorial choices
made in each case, while himself leaning towards respect for the individual, even
idiosyncratic, style of texts, which have after all succeeded in gaining a place in
the Modern Greek literary canon. What Ricks’s chapter also does is mitigate
the exceptionalism that usually surrounds such issues in discussions of Greek
9
See e.g. Tziovas (1989), Georgakopoulou (1997).
INTRODUCTION xix
by showing, for instance, that any editor of Thomas Hardy, a contemporary of

Papadiamandis, is faced with similar questions of orthographic standardization.
The issues and the instances presented by Anthony Hirst in his discussion of
Cavafy are closely related to those confronted by Ricks. Hirst begins his discussion
of ‘editorial assaults’ on Cavafy’s work with the observation that Cavafy left very
little for editors to do. His intentions as a poet were clear, and as Hirst’s painstaking
analysis shows, his creative uses and manipulations of the small things in language
that may have maximum stylistic effect (diacritics, elisions, accents) were frequent
and strategic. In this perspective, any editorial standardization in the posthumous
editions is bound to be seen as a loss, an intervention and a form of gratuitous
control. In effect, Hirst’s fine-grained analysis constitutes a defence of the admitted
eccentricities of Cavafy’s style against over-conventionalizing editorial decisions:
it shows how non-standard choices (like spellings) in his poems, far from being
accidental (or mis-spellings), may be taken as indexes of the poet’s sensitivity to
the sound (and the meaning) of his poetry.
In the three chapters of Silk, Ricks and Hirst, we are shown how standardization
is anything but a one-size-fits-all process. Different language varieties have the
power to invoke and create different social meanings, styles, registers and genres;
and the promotion of one language standard, regardless of conventional associations
and contextual subtleties, may both stifle creativity and result in linguistic
impoverishment.
Assimakis Tseronis and Anna Iordanidou propose an analysis and evaluation of
the four most recent and authoritative monolingual dictionaries of Modern Greek,
as texts that have contributed to the ideology of standardization. Even the titles
of the dictionaries are revealing of the allegiances of each and also the marketing
options that they imply. For example, two carry the well-known names of the
actual lexicographers, both of them linguists and professors, Kriaras and Babiniotis.
By contrast, another is named after the research foundation associated with it, the
Triandafyllidis Institute, which in turn was named after the demoticist author of a
celebrated grammar of Modern Greek (1941).
Tseronis and Iordanidou place each of their lexicographers in a narrative of
standardization and in a context of attitudes to language and professional engagement
with language, thereby proposing relationships between their culturally mediated
biographical projects and the compilation and circulation of their dictionaries.
The analysis focuses on symptomatic choices such as the inclusion or exclusion
of entries, particularly with regard to neologisms, etymological information
provided, and spelling. Orthography, in this chapter too, emerges as a major aspect
of standardization, and the choices available are, here again, broadly classifiable
into the two familiar positions: they either lean towards the monumentalization
and crystallization of language, by opting for spellings that relate Modern Greek
words to their Ancient Greek cognates, or they strive towards modernization and
simplification. How far back the actual etymology for each entry is traced is also
seen as an important indicator of the lexicographer’s views on standardization.
xx ALEXANDRA GEORGAKOPOULOU
So too is the treatment of semantic borrowings and re-borrowings (notably

Greek-origin foreign terms that were actually coined in a foreign language and
later borrowed back into Greek through the learned tradition) and, in particular,
the extent to which their Greek origins are stressed. Once again, the history of
diglossia and the disjuncture it engendered between the learned and vernacular
traditions is drawn upon by the authors as the main interpretative framework for
the lexicographers’ individual choices. A revealing comparison between the four
dictionaries on the basis of the extent to which they prescribe and seek to improve
standards finds Babiniotis and Kriaras to be the most explicitly prescriptive, even if
for different reasons and in different ways.
As already indicated, this volume has set out to raise the profile of reflexive
approaches to the history of standardization in Greek, and a prime illustration
of this aim is the dual status of Kriaras within it, as both researcher and
researched, analyser and analysed. In Tseronis’ and Iordanidou’s chapter, Kriaras’
standardization practices are put into context and critically placed, whereas in Part
III the roles are reversed: Kriaras discusses his own dictionary and others in relation
to standardization and the aftermath of diglossia.
If the Romayka speakers in Bortone’s chapter are not aware of the links between
their language and Greek or Greece, the Greek speakers of Cyprus are all too
poignantly aware of those links, as Dimitra Karoulla-Vrikki’s chapter shows us.
She traces two competing official discourses that also circulate widely as lay meta-
representations within Cypriot society and which have been instrumental in shaping
language planning: Cyprocentrism, which is based on civic nationalist approaches,
and advocates the promotion of the language(s) of the state as the symbol of political
and economic supremacy; and Hellenocentrism, which involves the promotion of
Standard Modern Greek as the ethno-culturally marked language that contributes
to the formation of national consciousness. Karoulla-Vrikki uncovers an on-
going pattern of shifts and fluctuations between the two poles of a continuum of
positions that present numerous overlaps rather than a simple dichotomy. She
does this by charting the prevalence of one or other pole in the civil service, in
the law-courts, and in education. In each of these domains, language planning
has followed a distinct trajectory from Cyprocentrism to Hellenocentrism, with
Standard Modern Greek becoming gradually more dominant and, in the case of
the courts and the civil service, slowly replacing English. It is noteworthy that
English, albeit a language associated with the island’s colonial past, has frequently
served Cyprocentric interests in the name of practical convenience.
Part II concludes with Jannis Androutsopoulos’ distinctively focused
examination of Latin-alphabet Greek, or ‘Greeklish’, on the internet. From a
sociolinguistic and discourse-analytic perspective, Androutsopoulos analyses
what are demonstrably practices of emergent standardization from below. A brief
historical overview of Greeklish shows how the internet provides new instances
and opportunities for an old choice that, though initially motivated by practical
considerations, has now succeeded in serving as a ‘new literacy’ system which
INTRODUCTION xxi
a majority of young people are familiar with. Androutsopoulos coins the term
‘computer-mediated digraphia’ to refer to the simultaneous use of both the native
Greek and the Latin script in computer-mediated interaction. He goes on to flesh
out the main characteristics of this digraphia, scrutinizing transliteration practices,
the metalinguistic discourses surrounding their use, and the shifting patterns of use
and evaluation.
Androutsopoulos’ discussion shows that, despite the lack of a widely shared
transliteration standard, internet users create consistent transliteration styles
by orienting to either a ‘phonetic’ or an ‘orthographic’ transliteration scheme.
In addition, local norms of Latinized spelling emerge among individuals who
regularly interact with a user group and wish to identify with that group, through
well-attested sociolinguistic processes of convergence and ‘focusing’. Like Ricks’s
and Hirst’s chapters, Androutsopoulos’ demonstrates that orthography is not
a neutral technology for the representation of spoken language, but rather a set
of social practices rooted in specific social and cultural contexts and associated
with a multiplicity of symbolic and aesthetic meanings. Furthermore, exactly as
with some of the lexicographical choices reviewed by Tseronis and Iordanidou, an
imperative to preserve the etymological transparency of words seems to underlie
the ideology attached to the orthographic representation. However, in this case, as
Androutsopoulos stresses, the practices and ideologies of script choice cannot be
fully understood without taking the development and social spread of technology
into account.
As we shall see in Part III, and particularly in the chapters by Moschonas
and Goutsos, language ideologies very often articulate themselves in linguistic
and metalinguistic practices through a process of iconicity,10 whereby language
choices are ‘naturalized’ and become transparently emblematic of social, political,
intellectual or moral character. In the case examined by Androutsopoulos, the
orthographic representation of words in Greeklish aims at an ideal of visual literacy
that presents iconicity with the Greek alphabet. This is standardization from
below that is nonetheless informed and shaped by standardization ideologies and
practices from above. As Androutsopoulos shows, the participating individuals’
rationalization of orthographic representation has been shaped by public
discourses and outbreaks of ‘moral panic’ in Greece, which present Greeklish as
a threat to the language. As we shall see in Moschonas’ chapter, these cases of
moral panic are manifestations of a new form of purism that has surfaced since
the official resolution of diglossia in Greece in 1976, one which involves invoking
‘exterior’ threats to the language from other languages, particularly English, within
the framework of globalization. As Androutsopoulos indicates, however, with the
increasing linguistic localization of the web, the use of Greeklish is actually on the
decline.
10
See Woolard (1998) 3.
xxii ALEXANDRA GEORGAKOPOULOU
Part III: Ideologies and Contestations

Language ideologies are and have been at the heart of standardization practices,
and not just in relation to Greek. Language ideologies refer to ‘representations that
construe the intersections of forms of talk with forms of social life, that is, link
language differences with social meanings’.11 The importance of such ideologies
for standardization practices and attitudes to correctness in Greek – and the
importance of uncovering what tends to be registered in language use or language
planning in opaque and subtle ways – is the starting-point for the chapters in
Part III. The multi-faceted ways of encoding ideologies range from the use of
specific vocabulary and lexical associations to images, depictions and metaphors,
as well as metalinguistic, rationalizing accounts. They can form part of official
or lay discourses alike. They can be invoked as ‘shared’ cultural background or
naturalized as ‘universal’ principles. Overall, language ideologies serve as systems
of signification that can be conceptualized as interpretative repertoires:12 available
resources for making evaluations and constructing versions of self, action and social
structure in language.
The chapter by Emmanuel Kriaras provides a historical overview of the
language question, which serves as an interpretative repertoire for many of his
fellow-contributors. Kriaras’ aim is to assess the contribution of individuals who
acted as regulating agents at various landmark moments in the history of diglossia
in Greece. Among much else, Kriaras deplores the lack of intervention and official
planning in the aftermath of the official resolution of diglossia in 1976, particularly
with regard to two highly contentious language issues: the teaching of Ancient
Greek at school and the spelling reform (compare and contrast Moschonas, later
in Part III). Orthography once again becomes a fertile site for competing language
ideologies, and Kriaras urges the case for further simplification of the current
standardized orthography, which largely reflects a historical orthography model. In
this spirit he subjects three primary sources to a critical discussion in respect of their
orthographic choices: these are, in fact, three of the four dictionaries that Tseronis
and Iordanidou examine, including Kriaras’ own. As in this case the analysed is
also the analyser, the choices in his dictionary are presented, uncompromisingly, as
the model to be adhered to. In a similar vein to Tseronis and Iordanidou, though,
Kriaras reports an extensive use of etymological orthographic forms in Babiniotis’
dictionary. The fact that the historical-orthography model still holds some sway in
standardization practices in Greece is an indication of the power of the ideological
narrative of language continuity for Greek-speakers, which Mackridge now
discusses in more detail.
Peter Mackridge’s chapter offers an analysis of the metaphors that have been used
by participants in the Greek language controversy (mainly scholars and language
professionals) to express the relationship between the Ancient and Modern Greek
languages. Mackridge seeks to explicate the ideological assumptions behind their
11
Ibid.
12
Potter and Wetherell (1995) 85.
INTRODUCTION xxiii
use in the period 1750–2000, but with special reference to the first 150 years. His
discussion traces a path through a plethora of kinship-related metaphors (mother/
daughter) and family tree metaphors (roots/branches), along with the metaphor of
language as edifice or monument that needs to be preserved (compare Gazi in a
later chapter); he thereby highlights the difficulties faced by Greeks in defining the
characteristics of their contemporary variety of Greek in its own terms. At the same
time, as Gazi too will confirm, such metaphors are seen to have been intimately
linked with ideas that enjoyed some currency in the linguistics scholarship of the
time. Mackridge rightly associates attitudes to language with issues of self-definition
and identity, and scrutinizes the changing use of ‘labels’ for the ‘Modern Greeks’
(‘Romans’?, ‘Hellenes’?), tellingly using the term ‘Modern Greeks’ himself, where
other scholars would perhaps have chosen a different one. Not least, like Beaton
in the final chapter, Mackridge discusses the views of Korais, as an agent with
prescriptive power. He points to the way that Korais’ thinking ultimately opened
up to purism, with the undoubted subtleties of his argument (notably that the
spoken variety of the time should serve, in a re-Hellenized form, as the language of
education) lost in the process. In this chapter, as in Gazi’s and Beaton’s, we have a
focus on individuals (in effect, the ‘usual suspects’) and the ways they have shaped
the history of standardization; but these are far from accounts of ‘great men’ that
take their words and deeds at face value. The contextualization of the analyses
offered serves to make familiar distinctions between purists and vernacularists
seem crude and monolithic, shedding a necessary light on the shades of competing
arguments and views. Above all, the perceived recurrence over time of particular
conceptual metaphors and modes of describing and prescribing language-use
points to the gradual construction of interpretative repertoires, which are not easily
attributable to any single scholar or intellectual, and indeed are not confined to
official or scholarly discourses but become part of the national imaginary.
Effi Gazi reviews the career of Georgios Hatzidakis (1848–1941), Professor
of Linguistics at the University of Athens, whose views on language not only
influenced the politics of standardization, but also shaped the development
of modern linguistics in Greece. Specifically, Gazi’s analysis shows how the
institutionalization of a distinct science of language, and therefore of the scientific
study of the Greek language itself, was seriously implicated in the politics of
katharevousa by privileging written linguistic forms and by linking them to class
and gender perceptions. As Goutsos will argue in a later chapter in connection
with the discourse of the anti-purists, the scientific paradigm, and particularly its
ideals of objectivity, rationality and truth, is readily appropriated for a particular
kind of language-political cause – above all because scientific terminology tends
to conceal ideology or make it opaque. In this way, the findings of historical
linguistics regarding rule-bound language-change were drawn upon by Hatzidakis
to provide legitimation for the cause of unity and continuity that was prevalent
in the Greek national ideology of his day. At the same time, Gazi shows how
Hatzidakis could adjust his scientific views so as to align them with the politics
xxiv ALEXANDRA GEORGAKOPOULOU
of katharevousa – concealing, but ultimately revealing, the usual ideological

apparatuses at work. A notable example is his association of correct language
with educated women (iconicity again), an ingenious act of re-appropriation of
the demoticists’ naturalization of spoken language as mothers’ language. In turn,
this association was projected onto explicit value judgements about different forms
of spoken Greek according to the social standing of their speakers.13 Through a
critical discourse analysis that draws on the social theorizing (by Bourdieu in
particular) of power relations and the processes of institutionalizing discourses,
Gazi sheds new light on the interrelationships between the development of Greek
linguistics as a scientific field and forms of nationalistic politics structured around
issues of linguistic distinctiveness and exclusivity.
In Spiros Moschonas’ chapter, as in Mackridge’s, perceptions and constructions
about the language end up being much more important than linguistic facts.
Moschonas examines a range of public, metalinguistic representations and attitudes
expressed in the Greek print media and propagated by language professionals,
folk ideologists and journalists, in the period following the language reform of
1976 when an official standard based on the demotic was finally established. His
discussion shows how post-diglossia language issues form a coherent and collective
media narrative – in his terms, a ‘regime ideology’ – which gradually moves away
from the long-standing debate on demotic-versus-katharevousa to new forms of
purism. In a genre-based analysis, Moschonas creates a typology of these issues,
distinguishing for instance between those that provoked a moral panic (like the
issue of teaching Ancient Greek in secondary schools) and routine or recurrent
issues which never resulted in intense public debate (like the adoption of loan
words, mainly from English). He goes on to examine their occurrence in, and their
relation to, different genres in newspapers such as features or reportages. In this
way, his discussion demonstrates a case of ‘double indexing’: language issues are
intertwined with given types of publication, while at the same time subtly and
indirectly encoding language ideologies.
Moschonas documents a clear shift from the predominant language issues of
the 1980s (the teaching of Ancient Greek in secondary education, the presumed
language impoverishment of the younger generation, the influence of English on
Modern Greek) to issues of territoriality, language contact, and hegemony, in the
1990s. He visualizes this shift as a move away from a battle in the interior realm of
the language (one that has to do with high registers versus low) to a battle between
the interior and the exterior (one that has to do with, for instance, threats to the
Greek language from alien languages like English). As in Goutsos’ and Mackridge’s
chapters, we see how easily language ideologies can be semantically organized on
the basis of conceptual metaphors: in this case, the territorial metaphor. Another
common thread between Moschonas and Goutsos is a historical reading of the
language issues raised in Greece after the 1976 resolution of diglossia: these would
13
On such recursivity, cf. Woolard (1998).
INTRODUCTION xxv
be inconceivable without the ideology of purism that marked the history of

diglossia before that date.
Contemporary meta-linguistic accounts in the post-diglossic period also
form the subject of Dionysis Goutsos’ chapter, but in his case, the focus is on
professional scholars, in the shape of Greek professors of linguistics, like the late
Anastasios Christidis. Bringing the works of a number of such scholars together,
Goutsos offers a critical analysis of what he sees as a characteristic anti-purist
discourse. This discourse emerged in recent times as a counter-move to the ‘more
dispersed and less localized’ arguments (compare Gazi and Moschonas) that have
characterized ethnocentric and purist theorizing on Greek, and which (as we have
seen in Mackridge’s chapter, and elsewhere) have historically sought to emphasise
and (if possible) ‘restore’ the continuity of the language. Through a corpus-based
analysis of this discourse, Goutsos identifies its main themes as myth and ideology,
and proposes that these themes are strategically used in argumentative moves so as
to delegitimize the opposing position. Myth and ideology are articulated through
lexical items belonging to their respective word-families, but are primarily deployed
as underspecified, all-encompassing and derogatory descriptors of the purist
discourse. Put differently, the claim is that these descriptors become iconic of a
lack of scientific method and rationality. Goutsos argues that this anti-ethnocentric
and anti-purist discourse is grounded in liberal and essentialist positions and,
furthermore, conceals specific aspirations such as the goal of integration into the
European Union.
In general terms, the thrust of Goutsos’ critique is that no discourse or argument
is devoid of an ideological standpoint, with the practical corollary that the highly
politicized and polarized debate in post-diglossic Greece between purist and
anti-purist scholars has ultimately hindered a dispassionate and socio-historically
informed view of the language. All such situated approaches to language issues,
Goutsos affirms, should become reflexive and enter a serious dialogue with
lay understandings and discourses, including the often-derided discourse of
language continuity: like any others, such a discourse acts as cultural resource and
interpretative framework. But for any progress of this kind, as Goutsos stresses in
conclusion, ‘we must first clear the conceptual ground of the old polarities’ and, in
a critical spirit, ‘search for continuities and ruptures both in the diachrony and the
synchrony of Greek’.
Roderick Beaton’s concluding chapter takes us back to the end of the eighteenth
century, and specifically to Adamantios Korais (1748–1833), who became
synonymous with attempts to reform the Modern Greek language in the nineteenth
and twentieth centuries. Like Mackridge, Beaton argues that the widely held belief
in a close association of Korais’ views with the later development of katharevousa
is to some extent a misconception. Anticipating, in part, the preoccupations of the
Neogrammarians, and their erstwhile disciple Hatzidakis (cf. Gazi), Korais was in
favour of the ‘common language’ as the language that Greek-speakers learnt from
their mothers, but at the same time he advocated a position of classic moderation.
xxvi ALEXANDRA GEORGAKOPOULOU
Beaton fleshes out what he acknowledges were neither systematic nor consistent
views on language by focusing on a largely unexplored source, Korais’ preface to
his edition of Heliodoros’ Aithiopika (1804). Close analysis of Korais’ ideas for
reforming the Modern Greek language in this preface reveals their metonymic
relationship with the Aithiopika. Heliodoros, a distinctive representative of the
Second Sophistic (Strobel’s discussion has a special relevance here), is put forward
by Korais as a model for correct usage in the new age. This project is seen to
involve a creative selection from Ancient Greek but not a mindless imitation.
It is in his Aithiopika preface that Korais coins the first ever generic term
for the novel in Greek, a term still in wide currency in a slightly modified form
(μυθιστόρημα instead of the original μυθιστορία). Beaton’s argument is that, in
so doing, Korais stakes a claim to Greek ownership of the present-day European
genre and aims to pave the way for future achievements in the genre by the revived
Greek nation. In this sense, Korais is prescribing both a linguistic and a literary
model, and here Beaton detects the crucial influence of Romanticism. Language,
literature and the emergent idea of the nation are, for Korais, interconnected (as
they never quite were in the days of the Attic-based koine, discussed by Colvin
and Silk) – and the interconnection involves genre too. Time and again in this
volume, the close links between genre and standardization have come to the fore. In
Beaton’s chapter we can locate them at the fundamental level of generic labelling:
to construct and offer a genre as a standard, in effect to prescribe it (or prescribe
with it), you need to name it first.
In our collection overall, a wide variety of language phenomena from different

periods are exposed to critical analysis from a wide variety of viewpoints and
disciplinary perspectives. It is our conviction that these multifaceted elucidations
of language standards and standard languages in the Greek example (but standards
and languages, always, in pluralized form) will be seen to shed further light on the
concepts of a ‘standard’ language and a language ‘standard’ tout court.
Throughout the volume, the aspects and instances analysed and debated by
our contributors are documented as elements of complex histories, mediated by
socio-cultural constraints. Furthermore, the protagonists of these stories are both
contextualized in relation to the dominant discourses of their time and seen as
contextualizing and contributing to those discourses. The term ‘discourses’,
however elusive and variously used in many chapters, seems to capture the
constructionist aspects of standardization processes and practices in the history of
Greek. Again and again, correspondence to linguistic and historical facts is seen to
be less important and ultimately less influential than language ideologization and
the processes of misrecognition and erasure14 that this involves. In this process, we
see official and lay attitudes mutually informing one another; we also see the long
14
On erasure, cf. Woolard, ibid.
INTRODUCTION xxvii
history of the Greek language ultimately (re)constructed and interwoven into a

cultural narrative of continuity and exceptionalism, but also contestations to that
exceptionalist premise.
Problematizing and resisting the polarization of the past around master-narrative
and counter-narratives of what constitutes good and bad language, the chapters in
this volume have unravelled some of the complex ways in which both of these have
shaped language ideologies and practices of standardization in Greek. In so doing,
the contributors have broadly, and sometimes explicitly, taken the position that
any academic treatment of prescriptivism in any language cannot be undertaken
as a ‘neutral’ descriptive project. The focus on literary uses of language in several
chapters, meanwhile, serves to give this point a special force. The values – the
‘standards’ – that creative literature and its creative language evoke resist any
attempt at scholarly neutrality. They too properly call for both a critical and a self-
critical response.
As far as the contemporary situation of the Greek language is concerned, this
reflexivity – itself implicated in late modern and constructionist views of language
– is here brought to bear on modernist conceptualizations of language standards
in Greek and (closely related to them) ideas of nation-building, which, as several
chapters have shown, are still resonant in Greece today. At a time when processes
of de-territorialization and language contact are taking hold of contemporary
societies and reshaping their linguistic and socio-cultural landscapes, a continuing
determination to interrogate the relevance and (re)formation of long-standing
views about the Greek language represents one of the main desiderata for future
studies. It is hoped that our volume will make a distinctive contribution to this
emerging agenda.
References
Blommaert, J. (2009), ‘On delicacies: postscript’, Pragmatics: Special Issue on Language,
Discourse and Identities: Snapshots from Greek Contexts, eds. A. Georgakopoulou and V.
Lytra.
Cameron, D. (1995), Verbal Hygiene, London.
Christidis, A-F. (ed.) (2007), A History of Ancient Greek: From the Beginnings to Late
Antiquity, Cambridge.
Georgakopoulou, A. (1997), Narrative Performances. A Study of Modern Greek Storytelling,
Amsterdam and Philadelphia.
Halliday, M.A.K. and Matthiessen, C.M.I.M. (2004), An Introduction to Functional
Grammar, 3rd edn, London.
Hanks, W.F. (1996), Language and Communicative Practices, Colorado and Oxford.
Haugen, E. (1966), ‘Dialect, language, nation’, American Anthropologist, 68: 922–35.
Horrocks, G. (1997), Greek: A History of the Language and its Speakers, London.
xxviii ALEXANDRA GEORGAKOPOULOU
Le Page, R. and Tabouret-Keller, A. (1985), Acts of Identity: Creole-Based Approaches to

Language and Ethnicity, Cambridge and New York.
Potter, J. and Wetherell, M. (1995), ‘Discourse analysis’, in J.A. Smith, R. Harré and L. Van
Langenhove (eds.), Rethinking Methods in Psychology, London: 80–92.
Tsitsipis, L. (1998), A Linguistic Anthropology of Praxis and Language Shift: Arvanitika
(Albanian) and Greek in Contact, Oxford.
Tziovas, D. (1989), ‘Residual orality and belated textuality in Modern Greek literature and
culture’, Journal of Modern Greek Studies, 7: 321–35.
Woolard, K.A. (1998), ‘Introduction: language ideology as a field of inquiry’, in B.
Schieffelin, K.A. Woolard and P.V. Kroskrity (eds.), Language Ideologies: Practice and
Theory, Oxford and New York: 3–47.
Part I
Establishing a Standard
This page has been left blank intentionally
1
The Invention of Greek:

Macedonians, Poets and Others*
Michael Silk
The Macedonians of my title are Philip and Alexander, along with their elite
supporters and successors, who institutionalized a standard language, the Attic-
based koine, for the new imperial Greek world, from the latter part of the fourth
century BC. The poets are the likes of Homer and Hesiod, Solon and Pindar, in
the earlier age of Greece, whose composite literary dialects, arguably, constitute
a different kind of collective idiom.1 And the others, who are more numerous,
range from ancients to moderns: those ancients, both before and after Macedonian
rule, who helped to ensure the (as I suppose) damaging ascendancy of Attic
over (as I suppose) less damaging and less damaged forms of Greek; and those
moderns (sociolinguists, in particular) who would deny me the right to make any
such judgement of value as I just made – on the grounds that it is ‘prescriptivist’,
‘unscientific’, ‘subjective’. Judgements of value about versions of a language (I
would insist) are matters of rational debate; they are also important matters, and
matters of special importance to anyone who supposes, as I suppose, that languages
and their literatures stand in a special and privileged relationship.
‘It is well known’, writes Peter Trudgill, ‘that linguistics is a descriptive, rather
than prescriptive, science’ – and such, indeed, is the common perception of
linguistics by linguists.2 One well-known corollary is a kind of pragmatic relativism,
on the part of linguists, vis-à-vis alternatives within, or between, languages. Thus,
Trudgill insists, any one language is as good and as adequate as any other; and he
quotes Halliday with approval: ‘all languages are equally capable of being developed
* This chapter represents an expanded version of the paper read to the Logos conference in 2004;
adjusted versions were subsequently delivered to audiences at Boston University in 2005 and the
University of North Carolina, Chapel Hill, in 2007. I am grateful to members of all these audiences
for helpful comments, and also, and especially, to Stephen Colvin and Alexandra Georgakopoulou for
further thoughts on the written version.
1
Not in itself a new or controversial idea, at least in respect of Homer: ‘Homère a été le premier à
fournir à la Grèce une langue commune, bien que littèraire’ (Lopez-Eire (1993) 45).
2
Trudgill (1983) 201.
3
4 Michael Silk
for all purposes.’3 Trudgill himself goes on to offer a series of arguments against the
application of value judgements to languages, whether in respect of correctness, or
the superior adequacy of particular language varieties, or their ‘aesthetic’ qualities.4
He urges fellow linguists, furthermore, to combat popular notions of (for instance)
correctness, and to do so for ‘educational and social’ reasons: ‘many children, in
many different language communities, are still discriminated against . . . for using
non-standard dialects and low-status accents.’5
This familiar sociolinguistic position is surely open to serious objections. These
forbidden evaluations (Trudgill warns) are usually ‘social judgements about
the status of speakers who use particular forms, rather than objective linguistic
judgements about the correctness of the forms themselves’.6 Well yes, one’s
tempted to say, ‘social judgements’: what else is ‘correct’, in the context of human
behaviour, likely to mean? But why, necessarily, are they judgements about ‘the
status of speakers’? And in what sense are such ‘social judgements’ necessarily any
less ‘objective’ than linguistic ones? To take the most obviously relevant example:
it is surely an objective matter, usually, to decide whether, in a given case, there
is (‘for better or worse’) a standard dialect or not, and (if there is) which it is and
what it involves. But even if the judgements are, indeed, not ‘objective’ or less
‘objective’, that does not in itself invalidate them or remove the need for them,
any more than it would do with judgements in the realms of ethics, politics or
aesthetics. Then again, in making a cautionary point about discrimination, the
linguist himself is making a ‘social’ value judgement’, in favour of egalitarian goals.
Why is the linguist allowed to make this sort of value judgement, if Joe Public is not
allowed to make any?
A quite different kind of objection is that many linguists seem not to take value
seriously. In the socio-linguistic handbooks, under the heading of ‘aesthetic’ values,
for instance, we are usually asked to smile knowingly at Aunt-Sallyish propositions
like ‘Birmingham accents are ugly’ or ‘Italian is a pretty language because of its
open vowels’ (both of which, as they stand, amount, indeed, to nothing more than
grossly trivial, take-it-or-leave-it prejudices),7 instead of considering things that
matter, like George Orwell’s 1946 critique of politically suspect language – or, as I
suppose, the regrettable victory of Attic Greek.8
3
Ibid. 206; Halliday et al. (1964) 100.
4
Trudgill (1983) 202–14.
5
Ibid. 202–3.
6
Ibid. 205.
7
Birmingham, England (not Alabama). In Greek antiquity we meet similar prejudices. Thus
Pausanias (3.15.2) pronounces Laconian the ‘ugliest’ dialect (τῶν Λακώνων ἡ γλῶσσα, ἥκιστα
παρεχομένη τὸ εὔφωνον), and a character in Theocritus (15.88) likewise finds the broad Doric
alpha ‘tiring’, whereas, conversely, Aristides Quintilianus (De Mus. 2.13) thinks it sounds ‘manly’.
There can, of course, be contexts in which even this kind of judgement may be less trivial: e.g. the issue
of open vowels in Italian is not entirely inconsequential for the composer of Italian opera.
8
Orwell, ‘Politics and the English Language’: Davison (1998) 421–32. Compare, more recently,
Fairclough (2000), from the sociolinguistic sphere itself. Contemporary sociolinguistics is, no doubt, a
good deal more heterogeneous than my reflections make it sound, and a fuller discussion would need
The invention of Greek 5
The most obvious objection of all, though, to this empiricist disdain for value
is its fundamental self-contradiction. Everything that language speakers do in
their everyday practice is beyond criticism – except for their everyday practice of
criticizing (or praising) the usage of others, and thereby reading linguistic activities
on a scale of value. Awareness of this contradiction is part of what lies behind a
striking challenge to the traditional linguistic view mounted by Deborah Cameron.
In her book, Verbal Hygiene,9 Cameron begins by noting how puzzled her fellow
linguists are by value:
Most everyday discourse on language is above all evaluative discourse . . . This
overriding concern with value is the most significant characteristic that separates lay
discourse on language from the expert discourse of linguists. As scientists, professional
linguists aspire to objectivity and not to moral or aesthetic judgement.
But (she adds):

linguists not only disapprove of the forms that popular interest in language typically
takes; they find the whole phenomenon somewhat bewildering – much as a chemist
might be puzzled by laypeople forming an association devoted to the merits of the
inert gases.10
In a revisionist spirit, Cameron argues forcefully, and surely rightly, that

metalinguistic value judgements are in principle entirely reputable, but in practice
often inconsequential or confused: ‘the standard of public discourse on linguistic
topics is lamentably low.’11 And she appeals, admirably, to ‘critical discourse about
literature, architecture or painting (or for that matter football)’ to remind her linguist
colleagues that one can and should get beyond ‘polarised and dogmatic statements
like “x is just wrong” versus “everything is relative”’.12
All this is admirable, and no less admirably, Cameron goes on to suggest: that
language and linguistics are implicated in value, whether their expert practitioners
admit it or not; that the ‘negative connotation’ carried by the term ‘prescriptivism’
within linguistics needs to be challenged; and that the issue is not, for instance,
‘whether there ought to be [language] norms, but which norms they ought to be.’13
On the positive side, Cameron herself speaks up for ‘standards of excellence in
the use of language’, as opposed to ‘mere superficial “correctness”’:14 the kind
of superficial correctness, a classicist might add, that derives directly from later
to confront, for instance, the ethnopoetic perspectives associated with Dell Hymes. For my purposes,
though, the central issue is the willingness to contemplate evaluation of a language, or versions of a
language, as such – to which competing traditions of sociolinguistics are united in opposition.
9
Cameron (1995).
10
Ibid.: Preface, x.
11
Ibid. xiii.
12
Ibid. 224–5.
13
Ibid. 8–11, 164, 3, 229–36.
14
Ibid. 115.
6 Michael Silk
antiquity’s obsession with what the Greeks called Hellenismos and the Romans
Latinitas – interpreted as a matter of linguistic purity.15
Cameron’s own specific concerns with ‘evaluating language’ lie in the area of
political agendas. Her discussion ranges from ‘the politics of style’ in newspaper
editing to issues of ‘political correctness’ in feminist-sensitive usage.16 With rather
less urgency, she touches on issues of value and valuation as between dialects or
languages, citing in passing William Labov’s celebrated defence (in 1972) of ‘the
logic of non-standard English’. In her words: ‘Labov seeks to defend the non-
standard vernacular of Harlem adolescents by contrasting the logical and lively
utterance of a vernacular speaker with the standard, but on inspection merely
verbose and confusing, utterance of a middle-class black informant.’17 Here,
she might well have cited (but doesn’t cite) the conservative counter-valuation,
epitomized by the reaction of the linguistic historian and lexicographer Robert
Burchfield to just such vernaculars some years later:
Black English . . . is potently political in its animosity towards the structured patterns
of Received American, colourful, fancy, animated, and subversive. If it is possible to
see a variety of English as a threat to the acceptability of the language handed down
to white Americans from the seventeenth century onwards, this is it. Its dislocation
of normal syntax, its patterned formulas showing disregard for the traditional shape
of sentences, make it at once deeply impressive and overtly threatening to currently
agreed standards.18
Yes: if ‘public discourse on linguistic topics’ is to aspire to the level of (for instance)
‘critical discourse about literature’, then it needs to develop beyond isolated
pronouncements, however provoking, towards true critical dialogue and debate.
And here – once we confront Labov with Burchfield – we begin to see a significant
debate in prospect, with, even, the possibility of a meeting of minds at the end of
it.19 In this particular debate, among much else, one is gratified to observe the terms
of reference – extended, as they are, beyond the usual obsession with phonological
or grammatical minutiae to more fundamental issues like ‘the traditional shape
of sentences’. Gratified too to find technical linguistic considerations merging
seamlessly into literary-critical issues. The vocabulary of this debate-in-embyo is
revealing. We get a ‘non-standard vernacular’, which is ‘logical and lively’, versus
a standard which is ultimately ‘verbose and confusing’. On the other side we get
15
Albeit (to be fair) formulated in the first instance largely as rules for clarity: see Arist. Rhet. 3.5
on τὸ ἑλληνίζειν.
16
Cameron (1995) 32, 33–77, 116–211.
17
Ibid. 100: Labov (1972) 201–40.
18
Burchfield (1985) 164.
19
In the sphere of ‘critical discourse about literature’, compare: ‘You cannot point to the poem;
it is “there” only in the re-creative response of individual minds to the black marks on the page. But
– a necessary faith – it is something in which minds can meet . . . The implicit form of a [critical]
judgement is: this is so, isn’t it? The question is an appeal for confimation that the thing is so; implicitly
that, though expecting, characteristically, an answer in the form, “yes, but – ”, the “but” standing for
qualifications, reserves, corrections’: Leavis (1962) 28 (my italics).
‘patterned formulas’, which are ‘fancy, animated, and subversive’: both ‘deeply
impressive’ and ‘overtly threatening’.
Irrespective of one’s own sympathies (mine, on balance, would be with Labov,
where Cameron seems evasive), the methodological implications of this kind of
debate are large – and I am not sure that Cameron herself sees them. At this point,
one might well glance back at that curious analogy she offers between linguists and
chemists, in association with that flat antithesis between scientists and ‘laypeople’:
Linguists not only disapprove of the forms that popular interest in language typically
takes; they find the whole phenomenon somewhat bewildering – much as a chemist
might be puzzled by laypeople forming an association devoted to the merits of the
inert gases.
Curious indeed, and perhaps unthinking – and, if so, especially revealing.

Linguistics may be, may aspire to be, a science – but a ‘hard’ experimental science,
like chemistry? An alternative, and surely more defensible, and more productive,
analogy might be: ‘much as a botanist might be puzzled by gardeners arguing
over the relative merits of the lily and the rose.’ But then, the analogy only has
to be restated in this way to prompt the thought: ah, but a botanist wouldn’t be
puzzled, would she? Not even by a critical comparison of lily and rose; let alone
by arguments in favour of fragrant roses (or lilies), or healthy roses (or lilies); and
indeed, not puzzled either by the familiar, indispensable, socialized, distinctions
between flowers, wild flowers, and weeds – or even by the gardener’s familiar
distinction between weeds and bad weeds.20
A language, complex human creation that it is, is like – a garden? There
are surely worse analogies, and not only because this one makes it clear how
straightforwardly implicated in value language might be said to be. It also serves
to lay convenient stress on actuality, and especially achieved actuality: ‘making the
green one red’, as a gardener might, and as Shakespeare’s Macbeth, in a paradigm
of literary actualization, actually did:
Will all great Neptune’s ocean wash this blood
Clean from my hand? No, this my hand will rather
The multitudinous seas incarnadine,
Making the green one red.21
The passage is rightly famous, not only for its actualizing properties, but for the way
it achieves its actualization by having the learned, distancing idiom of polysyllabic
Latinity (‘multitudinous’, ‘incarnadine’) compete with, and collapse into, the raw
immediacy of the ‘Saxon’: ‘The . . . seas . . . , | Making the green one red.’
We are dealing, suddenly, with issues of achieved value; and one big, relevant
problem that linguists seem to have in dealing with, or even just acknowledging,
these issues, and thereby advancing the discussion, is that whenever questions of
20
Cf., pertinently, the argument about literature and weeds in Ellis (1974) 24–103.
21
Macbeth, II. 2. The analogy between a language and a garden is not uncommon – for a recent
example, see Burridge (2005) – but tends to be invoked ironically (as by Burridge, ibid.).
8 Michael Silk
achieved value come into view, their understanding is complicated and (frankly)
confused by a prior egalitarian commitment to potential, irrespective of actual
achievement. What was that Halliday principle, quoted and endorsed by Trudgill?
– ‘all languages are equally capable of being developed for all purposes.’22 The
principle is endorsed, it seems, by linguists at large, even the most heterodox. ‘There
is nothing intrinsic to the nature of languages,’ says Roy Harris, ‘which makes one
“inferior” to another, or incapable of adaptation to whatever new purposes its users
may require of it.’23 The claim only makes any sense if languages are conceived
of, not as actualities whose versions are known and documented and experienced,
but (remarkably, for ‘scientists’) as hypostatic fictions whose properties and
parameters can only be imagined – and then are always imagined in comfortingly
Rousseauesque-egalitarian terms: ‘everybody has won, and all must have prizes.’24
All languages are equally capable . . .: the proposition is rather like crediting all
mammals with the potential of flight on the grounds (perhaps) that some mammals
(like bats) already can fly, and one day (who knows?) pigs might. But ‘Lord, we
know what we are, but know not what we may be’:25 Ophelia, Hamlet, again
Shakespeare, whose capacity to fly with language is – as yet – unsurpassed. ‘As
yet’: a language might, sentimentally, be said to subsume what it may become;
what it surely must be said to subsume is what it has been and is.
More precisely: what a language should be seen to subsume is what has been done
in it, and with it, and especially what is now being done in and with it. A language
subsumes its uses; and uses subsume, not least, literary uses like Shakespeare’s,
the sorts of uses for which value judgement – critical judgement – is known to
be central. And here the suggested relationship between linguistic analysis and
‘critical discourse about literature’ is seen to take on a new and larger significance,
and especially if one accepts that literature, and especially poetry, instantiates,
innovates, activates, the highest uses of language. This principle, sadly, is one that
has been lost on some of the most influential of those who have sought to relate
linguistics and literature in modern times. Witness, above all, Roman Jakobson,
whose still fashionable schema of ‘linguistic functions’ absurdly contrived to put
the ‘poetic function’ of language on a par with – among others – the ‘emotive
function’. So (in one pan), as it might be, the works of Shakespeare and (in the
other) a set of interjections: oh! shit! tut tut!26
Against Jakobson, any theorist who means to take value seriously, and the uses
of language seriously, should take the ‘poetic function’ seriously. This means, I
would suggest, being prepared to apply the same habits of critical comparison
to and between versions of languages as literary critics are used to deploying to
22
Above, pp. 3–4.
23
Harris (1980) 129–30.
24
Lewis Carroll, Alice’s Adventures in Wonderland, ch. 3.
25
Hamlet, IV. 5.
26
Jakobson (1960) 353–7: the example ‘tut, tut!’ comes on p. 354. Jakobson was equally helpless
when confronted with issues of value within poetry itself: see e.g. his almost embarrassing attempt
(1987) to find mechanical reasons for approving the first, inferior version of Yeats’s ‘Sorrow of Love’.
and between versions of literature, and being prepared to do so on the basis of

perceived achievement. And among the provisional working assumptions behind
such comparisons, one would certainly include the following: that wherever a
language lacks a significant literature, that language (irrespective of its potential)
is, at the given stage, correspondingly deficient; that (say) English without
Shakespeare would not be the same language; that (more pointedly) a state of the
English language that helps to make Shakespeare possible must be, to that extent,
a desirable state, where a state of the English language in which such an outcome
seems inconceivable is, to that extent, not; and that (still more pointedly) a state of
the English language in which we have our Shakespeare, along with other figures
of stature, behind us, but only behind us, and no glimmering of comparable figures
in prospect, is a symptom of crisis: crisis, not just in literature, but in language.27
The critical-comparative attitude under discussion seems to me a prerequisite
for an adequate response to the fate of the Greek language in antiquity28 – and not
least to the remarkable series of issues arising from a scrutiny of the relationship
between the Greek language in its several phases and Greek literature, in its. In a
straightforward sense, of course, before the koine any talk of ‘the’ Greek language
is problematic, or else refers only to a hypothetical, early period: in the historic
period there is no single standard, but only a series of versions. In the archaic and
classical ages (say, from the late-eighth century BC to the fifth), modern analysts
detect a pattern of four dialect groupings: West Greek (chiefly, but not entirely, the
Doric dialects), along with the two East Greek groupings, Attic-Ionic and Arcado-
Cypriot, and the fourth grouping – Aeolic – seemingly intermediate between West
and East. Before that, in the latter part of the second millennium, Mycenaean
Greek (known from the Linear B tablets) already shows some of the East Greek
innovations, from which it follows that the pattern of archaic and classical dialect
differentiation has already begun to emerge; before that, we have the presumed
arrival in Greece of the first Greek speakers (late third millennium?); and before
that (earlier in the third millennium?), the formation of proto-Greek, hypothetical
27
See more fully Silk (forthcoming, b): my Macbeth discussion (above, p. 7) is a miniature paradigm
here. In this connection, I note the thoughts of the great nineteenth-century linguist Jakob Grimm in
a speech of 1851, in the days when linguists were more comfortable with value. Praising the English
language for its ‘characteristic power of expression’, he traces the ‘power’ to ‘a surprisingly intimate
alliance of . . . the Germanic and Romance’, with the Germanic supplying ‘the material groundwork’ and
the Romance ‘the higher mental conceptions’, and adds: ‘Indeed, it is not for nothing that the English
language has produced and supported the greatest, the most prominent, of all modern poets (I allude, of
course, to Shakespeare).’ The passage is quoted (in this translation) by the linguist Bailey (1991: 109),
who concludes his own survey of such worthwhile claims (along with others, less worthwhile) with
the (frankly) inane dismissal: ‘English is, after all, a language much like all the others’ (ibid. 287). It is
surprising that such woolly-mindedness has not been challenged by the post-modern consensus that
language constructs thought (no-one – ‘after all’ – supposes that all thoughts, or systems of thought, are
‘much like all the others’) or by the wealth of evidence about the language-consciousness of writers, on
which see e.g. Lesser (2005).
28
The most convenient overview is Horrocks (1997) 3–127.
10 Michael Silk
common ancestor of the historic dialects and independent branch of the Indo-
European language family. All this is relatively uncontroversial.
Regarding the place of the eventually dominant Attic dialect within these
groupings, it is, again, uncontroversial that Attic is to be seen as a distinctive member
of the Attic-Ionic group (distinctive, chiefly, as it seems, because of influence from
the neighbouring Boeotian dialect); but that without (and even with) the Boeotian
influence, Attic belongs with Ionic, and is not some independent branch of Greek,
as so many later-ancient dialectological formulae were to claim. For most of the
ancient analysts, Attic is itself one of four dialects – Attic, Doric, Aeolic, Ionic29 –
albeit one learned tradition does see a special relationship between Attic and Ionic
within the fourfold set.30
Within the modern understanding of the Greek language and its earlier
history, various uncertainties remain – like the degree of dialect differentiation in
the Mycenaean age and the precise place of Aeolic within the emergent dialect
groupings of the age that followed – but, again, what is not in doubt is the practical
autonomy of the individual dialects in earlier Greece. In the archaic period,
certainly, there is no overall standard, and the individual dialects are the Greek
language. In this early period, which for us effectively comes to life at the end
of the eighth century with the poetry of Homer and Hesiod, the dialect groups
and their numerous local versions seem to have had something approaching equal
status as dialects of Greek. Between the well-attested vernaculars, at least, there
is no evidence for any dialect hierarchy before the end of the fifth century BC.
There is no hint, for instance, of anything like the dismissal of Pittacus’ Lesbian as
a ‘barbarian language’ that we meet in Plato’s Protagoras.31 Revealingly, perhaps,
as Stephen Colvin has shown, in the Attic comedies of Aristophanes, a generation
before Plato, there is no condescension towards non-Attic dialects or dialect usages
– irrespective of current Athenian loyalties or animosities towards the speakers of
those dialects in their polis instantiations.32 It is, however, in the fifth century, in the
age of Aristophanes and the generation or two before him, that dialect differentiation
begins to erode, most obviously across the new Athenian empire. And then, it is
in the fourth century, not long after the death of Plato, that the new standard, the
Attic-based koine, is institutionalized across the Greek world, under Macedonian
rule, and ‘the dialects’ – the other, non-Attic, dialects – begin to disappear, at least
from public view.33
29
This set of four is already attested in the third century BC: ps.-Dicaearchus fr. 61 (= Heraclides
Criticus) FGH II pp. 263–4 Müller. At the opposite extreme from modern linguistic principles, the set is
skewed by an overvaluation of the literary: the four (dialects or dialect groups) are the four represented
in canonical literature. On Greek views of Greek dialectology, see Hainsworth (1967), Morpurgo
Davies (2002), Versteegh (1987).
30
First attested in Strabo 8.1.2. There was, in particular, an awareness that (as D.Hal. Thuc. 23 puts
it) ‘old Attic differed only slightly from Ionic’.
31
Prot. 341c Λέσβιος ὢν καὶ ἐν φωνῇ βαρβάρῳ τεθραμμένοι.
32
Colvin (1999) 305–7.
33
Horrocks (1997) 37–41; cf. Bubenik (2007) 482–5.
Broadly, then, as between the age that follows Homer and Hesiod and the age that
follows Plato, Greek can be said to have been reoriented from one language situation
to another. In the first situation, each vernacular dialect has local authority and
none has overriding authority. Each vernacular dialect community acknowledges
its own distinctive dialect, so that, on the epigraphic level, as Bubenik puts it, ‘each
city state employed its own dialect, not only in private and public documents of
internal concern, but also in those of a more external or interstate character.’34
In the second situation, amidst a multiplicity, still, of local dialects, one adjusted
version becomes a prestige version and is regarded as the standard dialect, or not as
a dialect at all, but as ‘the’ normative language, of which the others (sooner or later
in decline) are seen as eccentric, non-standard versions. This second situation then
continues for several hundred years, and is only modified when, and in the special
sense that, superior normativity is attached to an older and ‘purer’ version of the
normative version itself.
During the momentous shift from the first to the second situation, and for some
time afterwards, contemporary scholarship and contemporary constructions are
seen to lag behind reality: from Heraclides Criticus in the mid-third century BC to
Strabo, two centuries later, we continue to hear about the supposed four dialects
– Attic, Doric, Aeolic, Ionic.35 In subsequent formulations, however, the koine is
added to the list, and given a distinctive weight. ‘The Greeks say they have five
dialects, Attic, Ionic, Doric, Aeolic, and fifth the koine’: this, from Clement of
Alexandria, around AD 200, with the koine evidently regarded as either a special
dialect or not actually a dialect at all, but the new (or, by now, old) standard and
norm.36
The overall shift, from dialects to standard language, is unmistakable, but there
are different ways of reading it. One way has been proposed recently by Jonathan
Hall in his book, Hellenicity (2002). In this important (but problematic) study,
Hall uses the shift, along with a good deal of other evidence, to argue that until
the conflict with the Persians at the start of the fifth century BC, and the rise of
Athenian power and prestige associated with it, the Greeks, subjectively, have no
clear Greek identity: they construct their ‘Hellenicity’ only now. Among much
else, there is (this is not in dispute) no extant reference to the Greek language
(nor, obviously, to the foregrounding of the Greek language as central to Hellenic
identity) before Herodotus in the later fifth century, nor to Ἑλλάς to mean the
whole Greek-speaking world before Xenophanes in the late sixth, nor to Ἕλληνες
to mean ‘the Greeks’ before the same period (on pottery from Naukratis), nor to
the foundational myth of the eponymous Hellen before the Hesiodic catalogue
(possibly around 600, though Hall takes it to be later).37
34
Bubenik (1989) 288.
35
Above, nn. 29–30.
36
Clem. Al. Strom. 1.142
37
ἡ Ἑλλὰς γλῶσσα, Hdt. 2.154.2; τὸ Ἑλληνικόν, ἐὸν . . . ὁμόγλωσσον, Hdt. 8.144.2 (cf.
7.9β.2); Xenoph. 6 West; on Naukratis pottery, see Hall (2002) 130 n.22; on Hesiodic ‘Hellen’, see
Fowler (1998) 9–16, Hall (2002) 131, 238–9.
12 Michael Silk
In all this, Hall plays down (maybe rightly) the widely assumed significance of
eighth-century colonization as an index of oppositional Greek self-assertion, and
likewise the establishment, in the same early period, of the all-Greek institutions
of the Olympic games and of Delphi as a sacred centre; but more important, and
certainly more relevant to us, he fails to take full account of the issues of language
in archaic and early classical Greece. Yes, he is surely right in positing a new and
sharply defined ‘oppositional’ Hellenism with, and after, the Persian wars, whereas
(as he himself acknowledges) there was indeed a sense of ‘aggregative’ Hellenicity
long before: ‘in the archaic period, Hellenic self-definition was “aggregative”. That
is to say, it was constructed by evoking similarities with peer groups . . . Now [sc. in
the period of confrontation with Persia] Hellenicity was defined “oppositionally”
through differential comparison with a barbarian outgroup.’38 And (though Hall
himself does not foreground the point) this new confrontational identity is readily
articulated in terms of the linguistic opposition between – in Herodotus’ words –
‘barbarian speech and Greek’.39 But – as Hall seems not to appreciate – the earlier,
‘aggregative’ Hellenicity has its own decisive focus in, precisely, the sphere of
language in general, and literary language in particular.
In a rewarding chapter of his book On Dialect, Peter Trudgill analyses what he
calls ‘acts of conflicting identity’ in respect of ‘the sociolinguistics of British pop-
song pronunciation’.40 He discusses, with due precision, the Americanizing touches
characteristic of the vocalizing of British popular music from the late 1950s. In
particular, he draws attention to the way British singers use some kind of sub-
American voiced ‘d’ [d̯] intervocalically, in place of a native British voiceless ‘t’
(so, to a British ear, ‘bedder’, not ‘better’), and likewise the way the ‘longer’ British
‘a’ (in words like ‘dance’) is replaced by the stereotypically ‘shorter’ American
equivalent ([æ] for [a:]). Trudgill discusses the phenomenon as reflecting a sense
of ‘appropriateness’: accommodation on the part of British singers to what is, in
general perception as in historical fact, an American genre. In effect, the British
singer is deferring to American authority in this domain.41
Oddly, Trudgill finds it ‘difficult to think of precise parallels of cultural
domination in fields other than music’.42 And this is odd, not only because there is,
precisely, such a parallel to hand, but because the field in question – both the field
of the parallel and of Trudgill’s case study – is not exactly, or exclusively, ‘music’ at
all, but rather the lyrics of song or, in broader terms, performance poetry; compare,
then, the various traditions of dialectally standardized poetry (lyric and other) in
archaic and early classical Greece.
In an aside, Trudgill suggests that the Americanizing pattern ‘probably’ goes
back to the 1920s. The very late ’20s, I would say, because the pattern only
38
Hall (2002) 179.
39
βαρβάρου γλώσσης ἀντὶ Ἑλλάδος, Hdt. 8.135.3.
40
Trudgill (1983) 141–60.
41
Ibid. 141–4.
42
Ibid. 145.
becomes general in the early ’30s;43 and the recorded evidence from that period
(I would also say) repays attention, partly because it contains more impressive
examples of ‘lyric poetry’,44 as also more audible renditions, and partly because it
serves to pinpoint the coherent logic of the process as it first develops. A revealing
case in point is provided by the work of the once famous singer, Al Bowlly, featured
vocalist with Roy Fox, Ray Noble and other English dance bands between the two
wars. Bowlly himself was actually half-Greek and half-Lebanese and brought up
in South Africa, but once in England (from the late ’20s) showed himself able to
perform, as appropriate, with an irreproachable ‘Received Pronunciation’ British
accent. What is particularly instructive is the way Bowlly preserves the impression
of RP English on English-sounding material,45 but gestures towards American,
both on American songs, and also on songs American in spirit (so to speak), but
actually very British in point of fact.
A prime instance is the now classic standard, ‘The Very Thought of You’, written
by the British bandleader Ray Noble. This song was composed and published in
1934, at a time when Al Bowlly was singing with Noble’s band in London, and
Bowlly himself records it, hot off the press, in the spring of that year, with piano
accompaniment – and suitable Americanizing features. Thus, in the opening
words of the verse of the song, ‘I don’t need your photograph’, we get the word
‘photograph’ with distinct intervocalic voiced ‘d’ [-d̯-] and sub-American ‘short a’
[æ]. Likewise, in the chorus, the title line gets a very audible voicing in the stop at
the end of ‘thought’ (intervocalic in context): ‘The very thought [-d̯] of you.’
What is going on here with Al Bowlly, and what was going on in archaic Greece,
are closely related artistic procedures, centred on what I propose to call generic
dialectalization. The salient differences would seem to be: first, that (contrary
to current scholarly prejudices about the supposedly distinctive orality of ancient
Greek culture) the British Americanizing in popular songs is not written down,
whereas the Greek equivalent (even in the archaic period) is; then, secondly,
43
In the late 1920s one can sometimes find the converse: an archaizing practice whereby American
singers ‘Britishize’ on American songs. A good example is to be found on a recorded version of the
Gershwin song, ‘The Man I Love’, by Sophie Tucker. Recorded in March 1928, in New York, when
this very American singer was at the peak of her career and the very American brothers Gershwin at the
peak of theirs, the version contains both natural Americisms (‘a little while’ has the [d̯] etc.) and at least
one striking Britishism on the words, ‘maybe Tuesday will be my good news day’, where ‘Tue-’ has
American [tu:], but ‘new-’ has British [nju:]: the sequence comes twice, and both times is realized the
same way. The Britishizing, no doubt, reflects a long-standing deference to the the supposed primacy
of British traditions in the wider field of light music (Gilbert and Sullivan et al.), which effectively
disappears from American singing around this time, in recognition of the growing authority of the
distinctively American popular song.
44
Pace Christopher Ricks on Bob Dylan: Ricks (2003). Note Furia (1992).
45
As on a version of ‘Glorious Devon’ recorded, with piano accompaniment, in December
1931. For this very English song (composed in 1905: words by Harold Boulton, music by Edward
German), Bowlly does not abandon, but does perceptibly tone down, his customary Americanizing.
Straightforward British-English, and non-American, features include the RP ‘long a’ [a:] in the words
‘planted’ and ‘Raleigh’.
14 Michael Silk
that the Greek genre in question was of higher cultural status, while the literary
achievement it represents (though not necessarily the musical achievement) is
clearly superior. Among the points in common, however, are crucial issues of what
Trudgill rightly calls identity. The British Americanizing may be momentary, even
fugitive, but implies the sense of a collective (albeit not the sense of a single nation
or, strictly, a single people); and such a sense is inferential, on a larger scale, in the
Greek case too.
In the popular songs, the Americanizing gesture unmistakably assumes a notional,
therefore in fact dialect-composite, American, without regional specifics: it is a
conventionalized and generalized notion of an ‘American accent’ that the singers
aim at. Likewise, it is characteristic, and perhaps even invariable, for the ancient
Greek pattern to involve, not specifiable vernaculars, but more or less artificial
composites. The Greek pattern is in any case more complicated: it involves an
elaborate series of correlations between identifiable literary genres, or sub-genres,
and identifiable, albeit generalized, dialects. The genre or sub-genre is often defined
by performative context, and usually by formal or other features of a given literary
repertoire; and the dialect is readily identifiable, however generalized – that is,
however notional and composite. The correlations between dialect and genre are
not mechanical, but, from our earliest extant literature, even – from Homer and
Hesiod – the pattern is unmistakable, and, as with British Americanizing song, it
presupposes an association between generic dialectalizing and authority.46
In the light of this argument, ‘generic dialectalizing’ (I suggest) is seen to be
an especially appropriate label, because it serves to evoke both the correlation of
dialect and genre and the fact that the ‘dialect’ is itself usually less than specific.
For instance, Greek lyric poetry (meaning the words of Greek song) is usually in
some version of Doric, or at least has a Doric colouring; but it is quite exceptional
for the Doric to be specific, as it is in the poetry of Alcman (late seventh-
century), which is widely taken to have been close to, though not identical with,
contemporary Laconian.47 The resulting spectrum of more or less generalized
dialect forms and gestures becomes such a familiar feature of Greek poetry, in lyric
and elsewhere, that in the fourth century Aristotle (who knows no poetry outside
Greek poetry) actually reads dialect gesturing as a poetic universal, subsuming it
(very misleadingly, it has to be said) under the heading of elevational exoticism (τὰ
ξενικάv).48
46
This crucial aspect of authority is underplayed in many formulations of the principle, e.g.: ‘The
Greeks . . . did not suffer from linguistic insecurity; it was accepted without worry that certain types
of dialect were appropriate for certain types of literary composition. The reason for this, according
to popular belief, was that certain genres were developed in particular regions (or by individuals
from particular regions) and that as a result they were associated with the dialect of the region or the
individual concerned’: Colvin (1999) 307. Fine, but the ideological implications surely need adding.
47
Calame (1983) Introduction, xxiv–xxxiv; Risch (1954); Buck (1955) 344–5. Arguably, the
language of archaic lyric is never a purely local dialect (Ruijgh 1980: 174), though some instances (like
Alcman) are much closer to it than others (like Pindar).
48
Arist. Poet. 22, 1458a18–23, where the list of ξενικάv begins with the γλῶττα, defined earlier
The earliest attested exemplification of the principle, the most cut and dried,
the most influential, and possibly also the original model for all the rest, is the
composite dialect of epic. This composite – widely discussed under the heading
of the Homeric Kunstsprache – is an archaic version of eastern Ionic (the Ionic
spoken in the localities where, presumably, the tradition arose), with Aeolic and
(possibly) other elements, together with unclassifiable archaisms of various kinds.49
From a vernacular point of view, the outcome is strikingly composite, because
it embodies both mutually exclusive dialect features and different stages of the
dialects in question. Paradigmatically, then, we get, on the one hand, hybrid forms
like κύνεσσιν (‘dogs’, in the dative plural), with Aeolic double sigma, but Ionic
movable nu;50 then, on the other, we have the alternation between older forms with
digamma and newer ones without, between archaic genitive -οιο and modern -ου,
and so on.
It is often supposed that the specifics of the composite and even the fact of
the composite itself reflect the hypothetical development of the oral-epic poetic
tradition; that may be. It is also widely assumed that the early oral transmission
of epic poems, or epic traditions, across the Greek world involved minor dialectal
adjustment in the direction of the vernacular language habits of the locality within
which a version of a given poem was performed. ‘Homeric’ performance, then,
would (within metrical constraints) be more Aeolic in Aeolic-speaking areas,
more Ionic in Ionia (and so on), than such performance elsewhere. In our Homer,
quirkily, the most apparent indicator of this process is an Attic veneer, reflecting
performance (and subsequent transmission) in Athens, and yielding such forms
as οὖν (‘therefore’), in place of Ionic ὦν, and μείζων (‘greater’), in place of Ionic
μέζων.51
Much more important, though, is the overall maintenance of the composite, and
its remarkable diffusion, across all known Greek domains. The Homeric Iliad and
Odyssey and the so-called Homeric Hymns are all presumably Ionian in provenance,
and certainly they embody one and the same Ionic-based composite, give or take
minor details. But so too does the work of the Boeotian poet Hesiod (native dialect,
a form of Aeolic); and so too (as surviving fragments and attributions suggest) does
the work of other, seventh-century, epic poets throughout the Greek-speaking
regions of the Mediterranean, like Stasinus from Cyprus and Cinaethon from
Sparta.52 And, no less significant: with or without minor dialectal adjustment, the
(21, 1457b3–4) as ‘a name which others [sc. other Greek-speaking communities] use’, i.e. a linguistic
usage current in someone else’s dialect: see the helpful note by Janko (1987) 129–31 (on 57b6 and
57b33–58a7), and see further Excursus, pp. 28–9 below..
49
Horrocks (1996).
50
Albeit ν-movable in dative plurals is also attested sporadically in other dialects, including once in
Thessalian Aeolic, IG IX. ii. 257: Thumb-Scherer (1959) 66; Buck (1955) 84.
51
Cf. Janko (1992) 35, who, however, believes in a fixed, early, written text of Homer (ibid. 37)
and therefore discusses the Atticizing phenomenon purely in terms of (a) textual transmission, rather
than (b) performance or (c) performance and transmission.
52
For Hesiod, see Edwards (1971), especially his conclusions about the ‘epic koine’ and the
16 Michael Silk
epic dialect composite is rapidly standardized in other verse forms in the same
dactylic metre – from Delphic oracles and local verse inscriptions of the seventh
century to the hexameters of the Sicilian philosopher Empedocles in the early
fifth53 – while also determining, in modified form, the dialect of other metrically
related genres, notably elegy.
By the end of the seventh century, in any event, the epic composite is a pan-
Hellenic possession and exercises pan-Hellenic authority. This is, no doubt,
especially obvious with Homer’s Iliad, traditionally taken to be first (as well as
greatest) of the attested epics, and articulating, as it does, a pan-Hellenic enterprise
(one indeed that subsumes Trojans as well as Achaeans, since both are represented
as equally Greek, in language, proper names, society and religion). But the pan-
Hellenic principle is (so to speak) inscribed in the dissemination of the epic
composite itself. Arguably, that is to say, the dissemination is promoted by the
very compositeness of the composite: by the fact that (in Geoffrey Horrocks’s
words) ‘[its] dialect was not that of a particular region but a “poetic” variety which
. . . transcended the parochialism of local and even official varieties’.54 For this
composite idiom to exercise its huge influence, however, the towering quality of
the Homeric epics was itself, surely, a prerequisite. As Dio Chrysostom was to
remark, with pardonable exaggeration, around the end of the first century AD,
Homer ‘not only combined all the dialects: he also spoke like Zeus’.55
Given the way that the epic dialect comes to be acknowledged and used across
the Greek world, it is obvious that it is indeed a pan-Hellenic point of reference, and
from an early period. And this phenomenon, palpably, implies a de facto perception
of Greek as a common language from, say, the early seventh century.56 It follows
that, in emphasis at least, Jonathan Hall is wrong. Yes, it is only in Herodotus that
the Greeks are first described as ‘speaking the same language’ (ὁμόγλωσσοι);57
but the principle is long since implicit in the generic dialectology of the epic.
Furthermore, the unifying legacy of the epic composite extends far beyond the
bounds of epic itself. In the first place, the epic dialect influences the language mix
‘predominantly Ionic appearance’ of Hesiodic verse on pp. 201–2. Stasinus was credited with the epic
Cypria and Cinaethon with (among other things) the Little Iliad: Bernabé (1987) 36–8, 115–16.
53
Thus CEG I. 143 (Corcyra, seventh century) has local (Doric) σᾶμα, but epic κασιγνήτοιο etc.
Many early hexametric inscriptions reproduce epic phraseolgy with mechanical dialectal adjustment.
Thus CEG I. 326 (Boeotia, 700–675) reproduces the Homeric verse ending χαρίεσσαν ἀμοιβήν (Od.
3.58) in the local form χαρί¸ετταν ἀμοιβάν. The same principle underlies Milman Parry’s rule, that
in traditional phraseology oral epic composers ‘use the most recent form which still keeps the same
metrical shape’ (Janko (1992) 17–18), irrespective of dialectal (as well as diachronic) consistency.
54
Horrocks (1997) 19.
55
Dio Chr. 11. 23.
56
Morpurgo Davies (2002) 157–8 even gives the point an ‘oppositional’ significance: ‘The dialect
switching practised by poets and writers must have contributed to the contemporary feeling that the
various Greek dialects were joined by a special relationship which separated them from other non-Greek
speech varieties’ (my italics). Her focus, however, is effectively the fifth century, when that oppositional
sense is in any case explicit.
57
Above, n. 37.
of most subsequent poetry, even if that mix does not have the same Ionic basis.
For instance, much melic poetry, up to and including Pindar’s victory odes and
the choral lyrics of Attic tragedy, is actually composed in an adjusted version of
epic usage and phraseology, or at least a mode of Greek that encompasses epic
usage, despite the ubiquitous Doric colouring – which is strong in Pindar’s case,
and weak in the case of the tragic choruses. Take a typical example from Pindar,
the description of Philoctetes as Ποίαντος υἱὸν τοξόταν· | ὃς Πριάμοιο πόλιν
πέρσεν, τελεύτασέν τε πόνους Δαναοῖς (‘Poeas’ archer son, who destroyed
Priam’s city and brought to an end the Danaans’ troubles’: Pythian 1. 53–4). Doric
colouring is indeed present and correct: note the familiar long alpha in τοξόταν
and τελεύτασεν. But positively and negatively, the whole cast of the sequence
(as here befits the Trojan War subject) is palpably Homeric: positively, in the -οιο
genitive and the shape of a phrase like Πριάμοιο πόλιν πέρσεν;58 negatively, in
the pervasive elimination of syllabic augments (πέρσεν, τελεύτασεν) and definite
articles. It is usual to discuss Pindar’s Greek as a version of Doric. Much of the
time it is actually epic composite, with (so to speak) a Doric accent.59
Then again, the epic composite sets the pattern for other genres: they have
their own, new, composites in turn; and the melic attachment to Doric typifies
one of them. In melic poetry itself, as the difference between Pindar and the tragic
choruses suggests, there is in fact no single specifiable composite mix, as there is in
epic. Even the Doric gesture is not invariable, witness the poetry of Sappho; but
then, that counts (if only for this very reason) as a separate poetic tradition – which
takes us to the heart of the matter. Generic dialectalizing defers to poetic tradition,
to the originary, just as Al Bowlly defers to the American origins of the popular
song. We do not always know what determines the originary association, but we
can assume with confidence that there always was one. With melic poetry, where
our earliest instances, probably, are Alcman’s more-or-less Laconian Doric, we can
assume there already was a pull towards Doricizing before Alcman, if only for the
paradoxical reason that his own Laconian Doric is not one hundred per cent pure.
The originary principle is clear with Greek prose too. Prose originates, as the
medium of intellectual record and exploration, in the Ionian regions of Asia Minor;
and Ionic dialect, or colouring, characterizes much early prose, irrespective of its
provenance. Take the collection of medical texts (late-fifth and fourth centuries BC,
along with a few later items) that we know as the Hippocratic corpus. The corpus
is composed in a (more or less consistent) conventionalized Ionic, notwithstanding
the fact that the fabled Hippocrates himself (who is probably not the author of any
of the texts in question) comes from Doric-speaking Cos.60 In the same way, but on a
much smaller scale, even the Attic of Thucydides’ very Athenian histories gestures
58
ἐκπέρσαι Πριάμοιο πόλιν Il. 1. 69; πόλιν πέρσεν Il. 6. 415.
59
In other words, a version, albeit a more complex version, of the phenomenon discussed in
connection with early verse inscriptions, n. 53 above. See further Silk (forthcoming, a).
60
The salient points of the situation are well summed up by Dover (1987) 84–5. Cole (1991) 102
argues that, with some technical prose, choice of dialect may have been tailored to the operational needs
of the readership, but this hardly explains why (e.g.) the Hippocratic corpus should have been in Ionic.
18 Michael Silk
minutely towards dialect authority, with its Ionic -σσ- for Attic -ττ- (θάλασσα
not θάλαττα) the representative instance. Eventually, though, the principle and
practice of generic dialectalization breaks down, and the breakdown begins in the
fifth century with the political and cultural ascendancy of Athens and Attic itself.
By the end of the fifth century, and the turn of the fourth, Athens is the
unquestioned cultural centre of Greece; much of the literary output of Greece
comes out of Athens; and if we take a snapshot of that output, genre by genre,
we see a quite new, Athenocentric, principle in operation. In historiography, yes,
Thucydides gestures towards Ionic still – but his successor Xenophon hardly does
so at all, in effect dismissing the Ionic ancestry of the genre.61 Athenian orators
(Lysias and others) go about their business in good Attic – but the Athenians did not
invent oratory. Athenian philosophy (like Plato’s) is articulated in irreproachable
Attic – but the Athenians certainly did not invent philosophy (even if the dialogue
form is their own, Plato’s own, distinctive achievement). Tragedy is a more
interesting, and a more complex, case. In their lyrics, the tragedians (including,
still, the experimentalist Euripides) preserve the tradition both of Doricizing and
of epicizing idiom, while tragic dialogue (though indeed an Athenian invention) is
actually executed in a finely constructed new composite, mostly and basically Attic,
but with a spectrum of archaizing features (not all strictly Attic) and innovatory
vernacular elements as well.62 In Old Comedy, though, the gloves are (almost) off.
Apart from parodies, paratragedies and representations of non-Athenian speakers,
Aristophanes, along with his predecessors and contemporaries, offers a purely
Attic dialogue – irrespective of the Dorian claim to have invented the form, and
the extant evidence of Sicilian-Doric Epicharmus to corroborate it – whereas
Aristophanic lyric sometimes, still, defers modestly to the Doricizing (and even
the epicizing) habit, but sometimes comes out in straight Attic.63 Overall, then,
the pattern is not entirely homogeneous; nevertheless, in most of these cases, and
above all in prose, what we have is Athenian linguistic imperialism in operation, as
part of, or corollary of, a wider imperialist movement, political and cultural.
61
Notwithstanding his occasional use of dialectal features for ad hoc reasons like characterization of
speakers (Dover (1987) 110–12) and his occasional anticipations of koine Greek (Gautier (1911) 66–
74); one notes, for instance, that for Plutarch it is an assumption that readers will turn to Xenophon, as
to Plato, for his Attic purity (καθαρόν τε καὶ Ἀττικόν: Moral. 79d). Conversely, a few later specimens
of revisionist dialectalizing are recorded: as late as the fourth century AD, an Athenian, Praxagoras,
wrote a history of Constantine in Ionic.
62
On tragic dialogue, see the summary in Horrocks (1997) 20–1. On tragic lyric, Colvin (1999)
263 remarks: ‘the strongest non-Attic elements . . . are in fact Homeric and Ionic (often difficult to
distinguish)’. In terms of authority, this means, ‘in fact, Homeric’.
63
In broad terms, Doricizing, where it occurs, occurs in high-style lyric, Atticizing in low(er) lyric
sequences. Thus in Clouds: high-style Doric alpha in the flamboyant parodos (ἀρδομέναν, 282), Attic
eta in more idiomatic passages (κλεινοτάτην, 1025). As a caveat, one should acknowledge both that
textual transmission may have disrupted the pattern and that identification of the pattern is to some
extent (though only to some extent) circular: it is partly because of Doricizing features that a sequence
is seen to be high style in the first place. On Dorian comic origins, see Csapo–Slater (1995) 89–101,
412–13.
The extraordinary course of Athenian cultural imperialism (duly discussed by

Jonathan Hall) was presented definitively by Nicole Loraux in her classic study,
The Invention of Athens, to which my own title of course alludes. In this book,
which focuses on the distinctively Athenian funeral oration, Loraux expounds the
distinctive Athenocentricity of Athens’ view of Greece, documenting the way (for
instance) that Athenian writers appropriate for Athens the pan-Hellenic Persian
Wars, just as they naturalize as Athenian mythological figures like Ion, eponymous
founder of everything Ionian and Ionic.64
By way of enforcing Loraux’s argument and extending it to the linguistic sphere
(on which she in fact has little to say), one might ponder the phenomenon of the
poet Solon, around the beginning of the sixth century BC. Solon is the earliest
literary figure from, or associated with, Athens, and in his elegiac and iambic verse
Athens is presented as a land of tradition (established by gods, θεόκτιτον) – but
then also as a wretched site of anti-cultural dissolution, bemoaning its exiles who
(apparently) no longer even ‘speak Attic’ (γλῶσσαν οὐκέτ᾽ Ἀττικὴν | ἱέντας).65
Aptly enough, as befits the then ruling principle of generic dialectalization, Solon
tells us all this in a literary Ionic composite – and, though Athenian, he feels himself,
as an Athenian, to be part of the Ionic world. Athens is, as he puts it, the oldest
land in Ionia; and he articulates that revered name in hybrid epic-Ionic (’Ιαονίη)
too.66
Jump forward two centuries to Thucydides (with his attenuated, but still visible,
Ionic gestures), and note Pericles’ funeral speech, where Athens now claims to be
provider and embodiment of culture for all Greece (τῆς Ἑλλάδος παίδευσις):67
compare and contrast Solon. And now jump forward again, only a few decades
more, to the culmination of this Athenocentric logic in the unalloyed Attic orations
of Isocrates. In his Panegyricus (380 BC), Isocrates proclaims that ‘everyone
agrees that our city is the oldest and the greatest, and amongst all mankind the
most renowned’, because ‘our city has surpassed the rest of mankind in ideas
(τὸ φρονεῖν) and words (τὸ λέγειν), so much so that her pupils have become
teachers for all the rest, and Athens has ensured that the name “Hellenes” no
longer suggests a race (γένος) but a cast of mind (διάνοια), and that the name is
applied to those who share our culture (παίδευσις), rather than to those who share
our kin (φύσις).’68 Furthermore, if Athens is so central and all-encompassing, and
if ‘our city’ is so superior ‘in words’, and not even only ‘in ideas’, then inevitably
the Attic dialect is too. In some unexplained way, it seems, Attic itself can claim
a kind of superior inclusiveness; and later in his prolific career, Isocrates can duly
appeal to ‘the communality’ of Athenian Greek: ἡ τῆς φωνῆς κοινότης.69 To put
it bluntly, in the process of imperializing, Athenian spokesmen have translated
64
Loraux (1986: French original, 1981) 133, 84.
65
Solon 36. 8, 10–11 West.
66
Solon 4a West.
67
Thucydides 2. 41.
68
Isoc. Paneg. 23, 50.
69
Isoc. Antid. (353 BC) 296.
20 Michael Silk
Athenocentricity into a radical rereading of Greek dialect relations, whereby the

Attic dialect too is at the centre.
Yet this Athenian construction of the Attic dialect is more than fictive
construction: it is also reality. Athens’ political and cultural emergence in the fifth
century is matched by the emergence of a modified form of Attic – in modern
scholarship often known as ‘Great Attic’70 – as the administrative language, and
eventually the vernacular, including the educated vernacular, of a largely Ionic
empire. This version of Attic is more innovatory than the Attic of an Isocrates,
and in particular is marked by certain limited Ionic usages; and it spreads even
beyond Ionic territory, as dialect inscriptions attest.71 There is a striking moment
in the tirade by the Old Oligarch, anonymous author of a late-fifth-century, anti-
democratic pamphlet, on the impurity of current Attic usage:
Hearing every kind of language [through trade and empire], [the Athenians] have
taken something from each. Greeks prefer to use their own – language, manners,
dress. Athenians use a mish-mash from all Greeks and non-Greeks.72
This fastidious listener, no doubt, is listening to Great Attic, and not what he would
regard as Attic proper. And it is this Great Attic that is the basis for the koine,
soon to be institutionalized as standard Greek by Macedon under Philip II and his
successors. Central to the logic of that choice, however, is the cultural prestige of
pure Attic, as it were a notch or two higher up the linguistic-cultural scale.
Until the late fifth century and the reign of King Archelaus, patron of arts and
artists (Euripides among them), Macedonia was, to say the least, a marginal player
in the Greek world and the world of Greek culture, notwithstanding the heroic
pretensions – as well as heroic names – of its ruling dynasty. George Cawkwell
sums it up nicely:
Linguistically, as geographically, Macedonia was remote from the main stream of
Greek life. King Alexander ‘the Philhellene’ had been allowed to compete in the
Olympic Games only after his claim to being Greek had been fortified by the claim
that the Macedonian royal house had originated in Argos in the Peloponnese, which
really conceded that those who sneered at Macedonia as ‘barbarian’ were right. The
sneers went on. The sophist Thrasymachus at the end of the fifth century referred even
to King Archelaus as a ‘barbarian’. Isocrates in the fourth, no less than Demosthenes,
spoke of the Macedonians as ‘barbarians’. The truth was that Macedon was as
culturally backward as it was linguistically remote, and even the exact Thucydides
classed it as ‘barbarian’. Archelaus began to change all this.73
If ‘barbarian’ is taken to mean the opposite of ‘Greek-speaking’, we do not actually

know if the Macedonians were barbarians. Given the paucity of evidence available,
70
So called by Albert Thumb (‘Grossattisch): Thumb (1906) 261.
71
Horrocks (1997) 27–31, 33–6; Bubenik (1989) 7–8, 175–82; Schlageter (1906) and (1912).
72
Ps.-Xen. Ath. Pol. 2. 7-8: φωνὴν πᾶσαν ἀκούοντες, ἐξελέξαντο τοῦτο μὲν ἐκ τῆς, τοῦτο δὲ
ἐκ τῆς· καὶ οἱ μὲν Ἕλληνες ἰδίᾳ μᾶλλον καὶ φωνῇ καὶ διαίτῃ καὶ σχήματι χρῶνται, Ἀθηναῖοι δὲ
κεκραμένῃ ἐξ ἁπάντων τῶν Ἑλλήνων καὶ βαρβάρων.
73
Cawkwell (1978) 22.
it remains unclear what the Macedonian language was: whether an aberrant Greek
dialect, or an Indo-European relative of Greek, or perhaps a relative subjected to
Greek influence and therefore presenting the appearance of a hybrid (like Germanic
English, transformed by Norman French and then by Latinate borrowings).74
What is clear is that the Macedonian court had cultural aspirations, which meant
not just Hellenizing, but Atticizing, aspirations; that by the end of the fifth century,
it conducted itself in one or other version of Attic (rather as some European courts,
in a later age, were to use French); and that Attic (in whichever version) was soon
standardized as the language of the Macedonian elite as a whole.75 And once the
cultural pretensions of Macedonia are associated with political power, once Philip
unifies Greece and Alexander confirms and extends the new political entity beyond
Greece, the linguistic consequences are predictable: a version of Attic – the koine,
or ‘collective’ dialect, in fact a modified version of Great Attic – is standardized as
a unified form of Greek on Greece and the other conquered territories.
For the Macedonian overlords, the koine has several apparent advantages. In the
first place, it is Attic enough to offer continuity with their own earlier Atticizing.
Secondly, it is, again, Attic enough to count as a suitably high-prestige lingua
franca for a new world-empire. And thirdly, as regards the existing Greek-speaking
communities, many of them now familiar with Great Attic, it represents the best
available approximation to a single national version of Greek. The Attic basis of the
koine is central to these considerations, albeit obscured both by the name, koine,
itself and by the intermittent modern habit of calling it ‘Attic-Ionic’.76 This Attic
basis must have been obvious at the time, and a fragment of a third-century (BC)
comedy suggests that, indeed, it was.77 The fact of the matter, in any case, is that,
whether seen essentially as a vernacular norm or as a practical medium (it clearly
operated as both), the koine is not significantly more Ionicized than Great Attic
was – nor indeed much more Ionicized than the otherwise very different language
of tragic iambic poetry (with which, however, it actually shares a few features, like
the use of λαός, ‘people’, for Attic-proper λεώς).
74
For a summary of the evidence about Macedonian, see Brixhe and Panayotou (1994).
75
See succinctly Horrocks (1997) 33. In retrospect, Philip and Alexander could be seen as
defenders of the Athenian/Attic faith: ‘Plutarch presents Alexander as the paradigm of paideia . . .
[which] was still defined in predominantly Athenocentric terms’ (Hall (2002) 221).
76
For a recent example, see Hall (2002) 221. Greek-language specialists are not exempt: cf.
e.g. Bubenik (1989) 175. Contrast the thrust of a recent summary by Brixhe (2007) 486: ‘Attic . . .
became a “common language”.’ The word κοινήv itself, one should always remember, means, not
‘mixed’, but ‘collective’ (albeit also, by the relevant period, ‘ordinary’: LSJ s.v. κοινός A.III.1). That
‘the fundamental basis of koine was . . . Great Attic’ has been clear since the work of Thumb in the
early twentieth century: Bubenik (1989) 181–2. The special connection with Attic is occasionally
acknowledged even by ancient grammarians: εἰς τὰ πολλὰ γὰρ ἡ κοινὴ διάλεκτος τοῖς Ἀττικοῖς
ἕπεται (Choeroboscus I. 201. 22–3).
77
Posidippus fr. 30 Kassel–Austin appears to have a Thessalian speaker objecting to the Athenian
habit of seemingly equating ἑλληνίζειν (‘talk Greek’) with ἀττικίζειν (‘talk Attic’) (but cf., rather
differently, Kassel–Austin ad loc. and Morpurgo Davies (2002) 167–8).
22 Michael Silk
Vernacular norm, yes; practical medium, yes; but few features of the koine itself
would strike any observer of earlier Greek as ‘poetic’, either per se or by association
with earlier poetic usage; and here is a major, and under-discussed, aspect of the
new linguistic regime. The koine is never used as – thus, is not – a thorough-going
literary medium. Its literary uses, such as they are, are in fairly prosaic prose:
Polybius’ history, Plutarch’s essays, and then the special, and momentous, but
tangential, cases of the Septuagint and the Greek New Testament. The influence
of the koine on Hellenistic verse (for instance, on Menander and Callimachus) is
mostly early, limited and indecisive.78 This represents a strange new turn for the
Greek language – towards a non-literary standard – and a loss for its users; and the
extent of the loss becomes fully apparent in the second and third centuries AD.
The koine, as Horrocks euphemistically puts it, was ‘periodically subject to
influences from belletristic classical Attic’,79 and the most concentrated of these –
influential, in the end, for the whole future of the Greek language – is the problematic
phenomenon we know as Atticism. The phenomenon is represented at its best by
the witty essays of Lucian and the innovative narrative fiction of Achilles Tatius
and Heliodorus, among others; and at its worst by the shrill and shallow effusions
of Aelius Aristides. In Aristides’ Panathenaicus (mid-second century AD), the
notion of Athenocentricity takes on an awesomely literal force, and, with it, the
rhetoric of Athenian cultural imperialism sinks to a new low. For Aristides, Athens
is the middle of Attica, Attica is the middle of Greece, and Greece is the middle of
the world.80 Athens, it seems, gave birth to mankind: ‘this city’ is ‘the mother, the
common nurse, the starting-point of nature’.81 And among much else, ‘all the cities
and races of mankind have turned to you and your way of life and speech.’82 All
men everywhere have come to acknowledge Attic as the ‘one common language’
of the human race.83 Through Athens ‘the whole inhabited world has come to
share the same language’, and (Aristides gloats) Spartans and other Greeks have
‘abandoned their native dialects’.84 The world has identified Attic as ‘the mark of
education’ – and this (he adds) ‘is what I call the real Athenian “empire”’.85 And
on internal grounds, it seems, Attic is fully deserving of its high status: it is specially
endowed with ‘dignity’ and ‘charm’, both of these qualities writ large in Attic poetry
78
In a few marginal cases it seems to have supplanted verse altogether: thus by Plutarch’s day, in the
first century AD, even oracles might be ‘in ordinary prose’ (καταλογάδην καὶ διὰ τῶν ἐπιτυχόντων
ὀνομάτων, Plut. Moral. 397d). For a representative instance of koine influence on Callimachus (the
Fifth Hymn), see Silk (2004) 366. On the more complicated issue of Menander’s Greek, see the
survey by Willi (2002) 21–3. For the sake of completeness, meanwhile, one should acknowledge the
occasional use of verse words (like δρυμός at Polyb. 2.15.2), besides the regular use of what, from an
Attic standpoint, would have felt like verse words (λαός, in particular).
79
Horrocks (1997) 37.
80
Aristid. Panath. 15.
81
Ibid. 25.
82
Ibid. 322: πρὸς . . . τὴν ὑμετέραν δίαιταν καὶ φωνὴν ἀπέκλινεν.
83
Ibid. 324: μίαν φωνὴν κοινήν.
84
Ibid. 325–6: τὰς μὲν πατρίους φωνὰς ἐκλελοίπασιν.
85
Ibid. 326–7.
(presumably in tragedy and New Comedy)86 – and now, as if suddenly mindful of

the existence of Greek literature outside Attic, and, in particular, of the dignity and
charm of the Greek epic, Aristides shamelessly contrives to annex Homer to the
Athenian cause: after all, ‘his dialect is obviously from here.’87
Attic, in one or other version, as the new lingua franca of the Greek world, and
the correlative decline of the other dialects, in prestige and incipiently in daily
usage: all this is already the situation in the Hellenistic age; but for Aristides and
his contemporaries the requisite ‘Attic’ has to be a pre-Hellenistic variety. The
Atticism of the second- and third-century ‘Second Sophistic’ represents an attempt
to adjust the koine and reinstate the ‘pure’ Attic of five to six hundred years before,
as a medium for literary prose and educated discourse of a wider kind. The project
was not to everyone’s taste. The growing fad for ‘reproduction Greek’ (as Donald
Russell has called it) was already criticized by Plutarch around the end of the first
century AD; and the developed Atticist movement is satirized by Lucian, albeit
himself implicated in it.88 And, at this distance, it is not easy to assess the success or
failure of the movement in its own terms. Clearly, the unchanging antique Attic it
promoted never supplanted the developing vernacular koine, because that survived
to be the eventual ancestor of Modern Greek. Nevertheless, its consequences were
far-reaching: the establishment of the idea that a language of the past could be
revived as a current reality; the effective institutionalizing of diglossia as not just a
staple within the Greek language (as was the case with the koine), but as an ideal;
and a significant skewing of the canon of approved ancient authors, and therefore
of the survival of ancient texts – in favour (most obviously) of Plato and Attic
oratory, at the expense of the poetic masterpieces and prose explorations of ‘dialect’
Greece before the Attic ascendancy.
Whatever its more positive side, any movement like Atticism must be inimical
to the creation of a dynamic literary culture, because it denies the intimate relation,
the necessary continuity, between literature and language that is lived and felt:
the continuity summed up, for poetry, by the Wordsworthian aspiration to ‘the
language really spoken by men’;89 a continuity actually maintained in classical
poetry (most obviously in Attic tragedy) against (or even through) all the evident
elevation of the poetic composites. More precisely, this cultivation of (in Horrocks’s
words) ‘Atticized Greek as a semi-living language by the educated classes’ ignores
and frustrates the symbiotic relationship between spoken language and literature,
which is at its most sensitive in the case of poetry:90 literature, especially poetry,
86
Ibid. 327: σεμνότητα καὶ χάριν.
87
Ibid. 328: ἡ φωνὴ σαφῶς ἐνθένδε.
88
Russell (1991) xxi; Plut. Moral. 42d–e. Lucian’s satirical response is best represented in his
Lexiphanes and Pseudologista.
89
Preface to Lyrical Ballads, 1802 version: Brett and Jones (1965) 254.
90
Horrocks (1997) 86, at the end of a good survey of the linguistic premises of the movement (ibid.
79–86; cf. Kazazis (2007) 1200–12); for a complementary summary of its ideological implications,
see Whitmarsh (2005) 41–56. Stephen Scully has suggested to me that there is a degree of parallelism
between the epic Kunstsprache at the beginning of classical Greece and Atticism towards the end: ‘both
24 Michael Silk
needs the vitality and immediacy of lived speech, just as speech itself is enhanced
by the proximity of literature, past and present. Productive diglossia is possible
– but any situation in which evolving speech is simply at odds with a backward-
looking literature can only be damaging for both.
As far as the diglossia represented by Atticism is concerned, in retrospect the
establishment of the koine is surely the root of the problem. Either because the
koine was imposed as a top-down language, or because that language at its top was
too fraught with administrative and mercantile associations to invite the full range
of literary use,91 Hellenistic literature, and Hellenistic poetry in particular, had no
natural linguistic medium. The outcome, in this period, is first a largely unrelated
set of – admittedly remarkable – academic experiments in archaizing dialect forms,
which we know as Alexandrian poetry: experiments which come to exercise a huge
and positive influence on Roman, and subsequent, literatures; but experiments
which do nothing to solve the immediate language problem. There is, then and
thereafter, an effective vacuum which, in a sense, Atticism – with its genuine
literary associations – seeks, however fruitlessly, to fill. The institutionalizing of the
koine, then, has the effect of institutionalizing a gulf between literary languages and
others – beteween the functional and the artistic – which Atticism seeks to bridge,
but actually makes wider still.
In an important sense, though, the problem with the koine is not how Attic it
wasn’t, but how Attic it actually was. This so-called ‘Attic-Ionic’ koine: if only it had
been more Ionic. Notwithstanding the stature and achievement of Athenian drama,
the triumph of Attic literature is, in a long perspective, the triumph of prose92 – and
not just prose, but prose abstraction. The line that runs from the amazing linguistic
manoeuvres of Thucydides (which so bemused Dionysius of Halicarnassus in a
later age)93 to the magnificent, if sometimes strained, articulations of Aristotle, is
the line of abstraction writ large, typified by the development, or fetishization,
of hypotactic sentence structure, expansive abstract nouns or nominal phrases,
multi-prefixed verbs or derivatives. In most versions of developed Attic prose, such
features abound, even when the subject seems not to demand them – when the
matter is physical and, so to speak, close at hand. Here, almost at random, is a
are artificial, literary and culturally dominant.’ It will be apparent from my argument why I am resistant
to this suggestion. The epic Kunstsprache is a composite evolved from miscellaneous linguistic usages,
current as well as obsolete, as a poetic medium in the context of dialect equality; it neither denies
current speech patterns as such nor imposes any politically sensitive dialectal normativity. Atticism
does both; it is simply an attempted reinstatement of a past and supposedly pure state of a given dialect
as normative for the speech and (therefore) literary prose of an educated class.
91
In a longer perspective, of course, things are different: the koine-derived vernacular Greek does,
in due course, become a language of fresh literary achievement – and (or but) Atticism surely helps to
delay the process, and by centuries. On the role of the koine in stunting the development of the literary
genres in the short term, cf. Zgusta (1980) 127.
92
This is essentially the premise of Plutarch’s essay, Bellove an Pace Clariores Fuerint Athenienses
(Moral. 345c–51b: see especially 348b–c and 350b). Among those who have confronted this thought
more recently, compare Headlam (1902) 439.
93
D. Hal. De Thuc. and Amm. 2.
sentence from Aristotle, on what happens to blood during sleep. Converted into
‘acceptable’ English, it runs:
Sleep occurs after the absorption of food, because the blood becomes less
differentiated, until the purer part of the blood does separate upwards, and its thicker
part downwards.
Aristotle’s Greek, however, is more abrasively abstract, especially at the

beginning:
διὰ δὲ τὸ γίνεσθαι ἀδιακριτώτερον τὸ αἷμα μετὰ τὴν τῆς τροφῆς προσφορὰν ὁ
ὕπνος γίνεται, ἕως ἂν διακριθῇ τοῦ αἵματος τὸ μὲν καθαρώτερον εἰς τὰ ἄνω, τὸ
δὲ θολερώτερον εἰς τὰ κάτω. 94
‘On account of the becoming more undivided. . .’ Any such remorseless concern
for precise distinctions is poor soil for poetry, unless (as it might be) a Greek Dryden
or perhaps a Greek Eliot might have taken root in it:
They came, at a predetermined moment, a moment in time and out of time,
A moment not out of time, but in time, in what we call history:
transecting, bisecting the world of time, a moment in time
but not like a moment of time . . .95
In all probability, though, the abstract idiom of Attic prose is too remote from
traditional Greek casts of thought and expression to make such speculations
imaginable.
With the long and momentous tradition of Attic abstractness, compare and
contrast the world of Ionic prose. Here the treatises of the Hippocratic corpus, in
their practical and sometimes quaint composite Ionic, are an object lesson. No-
one now (as indeed, no doubt, no-one then) reads them ‘for their language’; but
if one does, one finds, again and again, both a refeshing independence from the
new Attic-based modes of expression and, more specifically, a cast of idiom –
concrete, even sensuous – in touch with the language of (even) Homer.96 Where
Aristotle’s account of blood and sleep finds its natural expression in elaborate noun
phrases and the rest, the Hippocratic sense of things and words produces a single,
self-contained, and wholly representative sentence (this, from the fourth-century
Epidemics):
τὸ αἷμα ἐν ὕπνῳ ἔσω μᾶλλον φεύγει.
In sleep the blood slips further inside.97
94
Arist. Somn. Vig. 458a21–5.
95
T.S. Eliot, ‘Choruses from The Rock’, VII.
96
Two qualifications: (a) some Hippocratic treatises, like De Arte and De Flatibus are sophistic in
character and ambitious of rhetorical/literary effect (cf. Silk (1974) 84); (b) other treatises may have
isolated stylistic flurries, especially at the beginning, as, famously, does Aphorisms I. My concern here is
not with either, but with the language of the ‘ordinary’ treatises or parts of treatises.
97
Hp. Epid. 6.5.15: ‘inside’ = ‘away from the surface of the skin’. This and the following Hippocratic
examples belong to books of varying dates, ranging from the late-fifth to the later-fourth centuries BC:
see Silk (1974) 84.
26 Michael Silk
The sensuous brevity of expression is characteristic of the corpus, and what is

especially noteworthy is the φεύγει, ‘slips’, not an archaism (though it might seem
so to a Attic-minded contemporary), and not a metaphor (though ditto), but a
vivid life-centred idiom (the escaping ‘inside’ is how it is to you, on the outside),
with residual animating connotations (the word also means ‘flees’), and a usage
which maintains the linguistic habits of earlier Ionic, as in Homer,
Νέστορα δ᾿ ἐκ χειρῶν φύγον ἡνία σιγαλόεντα
The gleaming reins slipped from Nestor’s hands98
and which is attested elsewhere in the corpus, as here of a fractured bone:

ὅταν τρωθῇ, φεύγει ἐς τὸ ἄνω μέρος τὸ πρὸς τῷ στήθει προσεχόμενον, καὶ οὐ
μάλα ἐς τὸ κάτω μέρος ἀναγκάζεσθαι ἐθέλει.
When fractured, the part attached to the breast[bone] slips upwards, and just will not
be forced downwards.99
In this second passage, the crypto-animating verb φεύγει (‘slips’ [‘flees’]) is in

parallel structure with another, ἐθέλει (‘will’ [‘wants’]) – again to be regarded,
not as archaism or literary trope, but as standard Ionic usage, and again in use in
Homer:
ζέε δ᾿ ὕδωρ | οὐδ᾽ ἔθελε προρέειν
and the water boiled: | it would not flow on . . .100
Or consider a rather different kind of case from the corpus. This is a summary
description of fever symptoms:
ἱδρῶτες· πᾶσι πάντοθεν πολὺς πλάδος
Sweatings: in all [cases], much moisture from all [the body].101
The penchant for concrete concision (no connecting particles, no articles, no copula)
is as apparent as the instinct to shape and enforce the miniature syntactic unit by a
decisive alliteration, associated with the choice of the relatively concrete πλάδος
for ‘moisture’, in preference to possible alternatives, whether used elsewhere in
the corpus or not.102 The sentence is not ‘poetry’, though in Jakobson’s terms, no
doubt, it counts as ‘poetic’.103 What it exemplifies is a proto-stylization, in touch,
98
Il. 8. 137. ‘Not a metaphor’: for the principle and methodology here, see Silk (1974) 27–56.
99
Hp. Art. 14.
100
Il. 21. 365–6; cf. the usages cited in LSJ s.v. ἐθέλω, I.2 and II.2 (which include a very few Attic
instances – against the abstract grain). The φεύγει . . . ἐθέλει sequence in the Hippocratic Art. passage,
though indeed involving no ‘literary trope’, resembles poetic instances of crypto-imagery discussed
under the perhaps infelicitous heading of ‘matching clichés’ in Silk (1974) 243–4.
101
Hp. Epid. 1.5.
102
E.g. νοτίς, used of sweat at Hp. Epid. 5.60.1 and 7.32. Symptomatically, the favoured
expressions for ‘moisture’ used in the roughly contemporary non-Ionic prose of Theophrastus, in his
short treatise On Perspiration, are the relative abstractions ὑγρότης and τὸ ὑγρόν (both of which are
also used in the Hippocratic writings elsewhere: Kühn and Fleischer (1986) s.vv.).
103
Jakobson (1960) 357–9.
again, with the distinctive poetry of the earlier classical past: Πριάμοιο πόλιν
πέρσεν: σκιᾶς ὄναρ ἄνθρωπος.104
Or consider – no less revealing – the kind of idiom represented at its most basic by
instances of the uncompounded verb βαίνω, ‘come’. In the corpus, we find usages
like βέβηκεν ἰητρική (medicine has ‘arrived’: is established), or, more simply
still, usages like πυρετοὶ βεβῶτες (fevers are ‘coming’) – as we might in Homer
(even), where we do find, for instance, ἰκμὰς ἔβη (the moisture ‘comes’).105 In
innumerable such ways, the largely unpretentious Ionic of the corpus is, seemingly,
both in tune with living Greek and pragmatically and perceptibly related to the
usages of traditional Greek poetry, as its Attic counterparts usually are not; and –
to anticipate one possible objection – the reflection that some sophisticated Attic
prose (Plato’s, above all) is sometimes itself ‘poetic’, in one or other sense, has no
bearing on the case. It comes as no surprise, indeed (albeit modern scholarship
barely acknowledges the fact), to find that in the heyday of Atticism, in the
second century AD, and perhaps much earlier too, Ionic Greek was liable to strike
sophisticated Attic-prose-lovers as ‘poetic’ per se.106
The standardization of the koine had the incidental, but damaging, effect of
promoting a version of Greek, which, both by association and by virtue of its
inherent tendencies, institutionalizes and even intensifies the otherness of the
great poetic traditions of the past: it ensures that they are no longer presences, but
are past. In a straightforward sense, the koine is the first Greek national language
and, as such, marks the invention of Greek. In a subtler sense, that title belongs to
the composite literary dialects from epic onwards. The one presents a poignantly
instructive contrast with the other.
104
Pind. Pyth. 1. 54 (‘destroyed Priam’s city’: above, p. 17) and 8. 95–6 (‘Shadow’s dream, man’),
the latter peculiarly distinctive as a three-word sentence (see Silk (2001) 30–6). In the Hippocratic
corpus, short assonantal sentences seem especially characteristic of the Epidemics: γλῶσσαι χλωραὶ
χολώδεις (‘green tongues [are] full of bile’: Epid. 6. 5. 8) is one of various striking examples.
105
Hp. Loc. Hom. 46, Prog. 20; Il. 17.392. Compare and contrast the largely fossilized usages
(εὖ βεβηκέναι etc.) cited in LSJ s.v. βαίνω, A.I.2.a.
106
See Excursus, below.
28 Michael Silk
Excursus: Ancient Evidence for the View that ‘Ionic is Poetic’

1. The clearest evidence comes from, or in the wake of, the Second Sophistic. The most
explicit testimony is H, De Ideis 2.319 (p. 336 Rabe): ἡ γὰρ Ἰὰς οὖσα ποιητικὴ φύσει.
In the Attic-based grammatical traditions of late antiquity, we also find related evidence
in the form of implicit equations of poetry and Ionic dialect, with usages ascribed to οἱ
Ἴωνες καὶ οἱ ποιηταί or characterized as ποιητικάv . . . Ἰωνικάv etc.: thus e.g. Pollux
2.76; Herodian, Techn. Rel. p. 313 Lentz; Choeroboscus, Schol. in Theodos. Canon. p. 110
Hilgard. Other evidence, especially earlier evidence, is suggestive rather than decisive,
but, given these later testimonies, may be read as corroboration that the underlying
association of ‘Ionic’ and ‘poetic’ was probably current earlier. Of the material cited, the
evidence from D. Hal. (5, below) is the most compelling: that would take the association
back to the Augustan age.
2. Aristotle on γλῶτται. Before Aristotle, γλ. is used to refer to exotic words in (e.g.)
Homer (Ar. fr. 233 Kassel–Austin, Democrit. fr. 20a Diels–Kranz). After Aristotle, γλ.
is understood to mean obsolete words, especially those found in poetry – and probably
always means this in later antiquity (τὰς ποιητικὰς φωνὰς γλώττας ἐκάλουν, Pollux
2. 109; ἡ γλῶττα παλαιόν ἐστιν ὄνομα τῆς συνηθείας ἐκπεπτωκός, Galen, Lex.
Hippocr. 19. 65–6 Kühn). As with some other important literary-analytical words,
however, Aristotle idiosyncratically reinterprets γλ., in this case to mean, effectively,
usages from another dialect that a poet uses in his own (Poet. 21: 1457b4–6 – cf. e.g. his
ubiquitous use of µῦθος, ‘story’, to mean ‘plot-structure’). It is still clear, meanwhile,
that γλ. are a feature of poetry (see Poet. 25: 1460b11–13, and also Rhet. 3.2: 1404b10–33
and 3.3: 1406b3). Evidently, then, given that an ‘own’ dialect is assumed to be an Attic
or Attic-based version of Greek, non-Attic dialect usage is somehow taken to be ‘poetic’
per se, at least when re-used by users of Attic. Nothing is said about Ionic in particular,
however.
3. Aristides Quintilianus (? third century AD) associates Ionic dialect usage with femininity,
Doric with masculinity (De Mus. 2. 13), but also femininity with the decorative (ibid. 2.
8). Behind these correlations, seemingly, there lies a pair of loose Platonic associations
– between poetry and the decorative (as Rep. 10. 601) and poetry and the feminine (as
Rep. 10. 605). An association between Ionian (people or behaviour) and the feminine,
in the sense of effeminacy, goes back to the fifth century BC (see Ar. Ach. 104, and Olson
ad loc.); but in this particular tradition (vel sim.) no explicit connection between Ionic
dialect and ‘poetic’ is attested.
4. Demetrius (? first century BC), Eloc. 112, associates Herodotus, as opposed to
Thucydides, with τὸ ποιητικὸν ἐν λόγοις, albeit disapprovingly, while at Eloc. 177 he
floats the thought that there is something ‘popular’ about Attic usage (ἡ γὰρ Ἀττικὴ
γλῶσσα συνεστραμμένον τι ἔχει καὶ δημοτικόν). The possible implication that it is
Herodotus’ Ionic dialect which is itself ‘poetic’, in contradistinction to the inherently
‘popular’ character of Attic, is only a step away, but the step (which would presumably
have implications for the stylistic value of sophisticated Attic usage) is not taken.
5. Dionysius of Halicarnassus (end of first century BC), Pomp. 3, offers what is
effectively a poetry/prose opposition, correlated with the Herodotus/Thucydides
opposition, with Herodotus explicitly the model of Ionic writing, as Thucydides
is of Attic: in particular, Herodotus is said to be τῆς Ἰάδος ἄριστος κάνων, with
the virtues of ἡδονή, πειθώ, τέρψις καὶ αἱ ὁμοιογενεῖς ἀρεταί. At Thuc. 23
a very similar characterization of Herodotus omits the ‘Ionic’ but adds ‘poetic’:
παρεσκεύασε τῇ κρατίστῃ ποιήσει τὴν πεζὴν φράσιν ὁμοίαν γενέσθαι πειθοῦς

τε καὶ χαρίτων καὶ τῆς εἰς ἄκρον ἡκούσης ἡδονῆς ἕνεκα. At Thuc. 5 the early
historiographers like Hecataeus – designated as ‘those who wrote in the same dialect’
(sc. Ionic) – are likewise credited with a characteristic χάρις. Here, the ‘Ionic’/‘poetic’
correlation is pretty well implicit.
6. Various modern scholars seem to assume or imply that within Attic prose any Ionic usage
would indeed have been felt as ‘poetic’ (so Dover (1987) 225 and Finley (1967) 63), but
without reference to any of the evidence cited above. Conversely, Bers (1984) 7 cites
Hermogenes’ claim (1, above), only to query it, while Denniston (1952) 16 ascribes a
similar-sounding claim to the Suda (‘all Ionic is poetical’), which I cannot verify and
assume to be a slip (no reference or other documentation is provided). No-one, to my
knowledge, has assembled, or confronted, the range of evidence set out above.
References
Bailey, R.W. (1991), Images of English: A Cultural History of the Language, London.
Bernabé, A. (1987), Poetae Epici Graeci, I, Leipzig.
Bers, V. (1984), Greek Poetic Syntax in the Classical Age, New Haven.
Brett, R.L. and Jones, A.R. (1965), Wordsworth and Coleridge, Lyrical Ballads, 2nd edn,
London.
Brixhe, C. (ed.) (1993), La Koine grecque antique, I: Une langue introuvable, Nancy.
Brixhe, C. (2007), ‘A modern approach to the ancient Greek dialects’, in Christidis (2007)
486–99.
Brixhe, C. and Panayotou, A. (1994), ‘Le macédonien’, in F. Bader (ed.), Langues indo-
européennes, Paris: 205–20.
Bubenik, V. (1989), Hellenistic and Roman Greece as a Sociolinguistic Area, Amsterdam.
Bubenik, V. (2007), ‘The decline of the ancient dialects’, in Christidis (2007) 482–5.
Buck, C.D. (1955), The Greek Dialects, 2nd edn, Chicago.
Burchfield, R. (1985), The English Language, Oxford.
Burridge, K. (2005), Weeds in the Garden of Words: Further Observations on the Tangled
History of the English Language, Cambridge.
Calame, C. (1983), Alcman, Rome.
Cawkwell, G.L. (1978), Philip of Macedon, London.
Christidis, A.-F. (ed.) (2007), A History of Ancient Greek: From the Beginnings to Late
Antiquity, Cambridge.
Cole, T. (1991), The Origins of Rhetoric in Ancient Greece, Baltimore.
Colvin, S. (1999), Dialect in Aristophanes, Oxford.
Csapo, E. and Slater, W. J. (1995), The Context of Ancient Drama, Ann Arbor.
Davison, P. (ed.) (1998), George Orwell, I Belong to the Left, 1945, London.
Denniston, J.D. (1952), Greek Prose Style, Oxford.
Dover, K.J. (1987), Greek and the Greeks, Oxford.
Edwards, G.P. (1971), The Language of Hesiod in its Traditional Context, Oxford.
30 Michael Silk
Ellis, J.M. (1974), The Theory of Literary Criticism: A Logical Analysis, Berkeley.
Fairclough, N. (2000), New Labour, New Language?, London.
Finley, J.H., Jr (1967), Three Essays on Thucydides, Cambridge, Mass.
Fowler, R.L. (1998), ‘Genealogical thinking, Hesiod’s Catalogue, and the creation of the
Hellenes’, Proceedings of the Cambridge Philological Society, 44: 1–19.
Furia, P. (1992), The Poets of Tin Pan Alley, New York.
Gautier, L. (1911), La Langue de Xénophon, Geneva.
Hainsworth, J.B. (1967), ‘Greek views of Greek dialectology’, Transactions of the Philological
Society: 62–76.
Hall, J.M. (2002), Hellenicity, Chicago.
Halliday, M.A.K., McIntosh, A. and Strevens, P.D. (1964), The Linguistic Sciences and
Language Teaching, London.
Harris, R. (1980), The Language-Makers, London.
Headlam, W. (1902), ‘Metaphor, with a note on transference of epithets’, Classical Review,
16: 434–42.
Horrocks, G.C. (1996), ‘Homer’s dialect’, in I. Morris and B. Powell (eds), A New Companion
to Homer, Leiden: 193–217.
Horrocks, G.C. (1997), Greek: A History of the Language and its Speakers, London.
Jakobson, R. (1960), ‘Closing statement: linguistics and poetics’, in T.A. Sebeok (ed.), Style
in Language, Cambridge, Mass.: 350–77.
Jakobson, R. (with S. Rudy) (1987), ‘Yeats’ “Sorrow of Love” through the years’, in
K. Pomorska and S. Rudy, Roman Jakobson, Language in Literature (1987), Cambridge,
Mass.: 216–49.
Janko, R. (tr.) (1987), Aristotle, Poetics, Indianapolis.
Janko, R. (1992), The Iliad: A Commentary, Volume IV: Books 13–16, Cambridge.
Kühn, J.-H. and Fleischer, U. (1986), Index Hippocraticus, Göttingen.
Labov, W. (1972), Language in the Inner City: Studies in the Black English Vernacular,
Philadelphia.
Leavis, F.R. (1962), Two Cultures? The Significance of C. P. Snow, London.
Lesser, W. (ed.) (2005), The Genius of Language: Fifteen Writers Reflect on Their Mother
Tongue, New York.
López Eire, A. (1993), ‘De l’attique à la koiné’, in Brixhe (1993): 41–57.
Loraux, N. (1986), The Invention of Athens, tr. A. Sheridan, Cambridge, Mass.
Milroy, J. and L. (1999), Authority in Language: Investigating Standard English, 3rd edn,
London.
Morpurgo Davies, A. (2002), ‘The Greek notion of dialect’, in T. Harrison (ed.), Greeks and
Barbarians, Edinburgh: 153–71.
Ricks, C.B. (2003), Dylan’s Visions of Sin, London.
Risch, E. (1954), ‘Die Sprache Alkmans’, Museum Helveticum, 11: 20–37.
Ruijgh, C.J. (1980), ‘De ontwikkeling van de lyrische kunsttaal, met name van het literaire
dialect van de koorlyriek’, Lampas, 13: 416–35.
Russell, D.A. (1991), An Anthology of Greek Prose, Oxford.
Schlageter, J. (1906), Zur Laut- und Formenlehre der ausserhalb Attikas gefundenen attischen
Inschriften, Freiburg im Breisgau.
Schlageter, J. (1912), Der Wortschatz der ausserhalb Attikas gefundenen attischen Inschriften.
Ein Beitrag zur Entstehung der Koine, Strassburg.
Silk, M.S. (1974), Interaction in Poetic Imagery, Cambridge.
Silk, M.S. (2001), ‘Pindar meets Plato’, in S.J. Harrison (ed.), Texts, Ideas, and the Classics,
Oxford: 26–45.
Silk, M.S. (2004), ‘Alexandrian poetry from Callimachus to Eliot’, in A. Hirst and M.S.
Silk (eds), Alexandria, Real and Imagined, Aldershot: 353–72.
Silk, M.S. (forthcoming, a), ‘The language of Greek lyric poetry’, in E. Bakker (ed.), A
Companion to the Ancient Greek Language, Oxford.
Silk, M.S. (forthcoming, b), Poetic Language in Theory and Practice, Oxford.
Thumb, A. (1901), Die griechische Sprache im Zeitalter der Hellenismus, Strassburg.
Thumb, A. (1906), ‘Prinzipienfrage der Koine-Forschung’, Neue Jahrbücher für das
klassische Altertum, 17: 246–63.
Thumb, A. (1959), Handbuch der griechischen Dialekte, II, 2nd edn by A. Scherer,
Heidelberg.
Trudgill, P. (1983), On Dialect: Social and Geographical Perspectives, New York.
Versteegh, K. (1987), ‘Latinitas, Hellenismos, ‘Arabiyya’, in D. J. Taylor (ed.), The History
of Linguistics in the Classical Period, Amsterdam: 251–74.
Whitmarsh, T. (2005), The Second Sophistic, Oxford.
Willi, A. (ed.) (2002), The Language of Greek Comedy, Oxford.
Zgusta, L. (1980), ‘Die Rolle des Griechischen im römischen Reich der Kaiserzeit’, in
G. Neumann and J. Untermann (eds.), Die Sprachen im römischen Reich der Kaiserzeit,
Cologne.
2
The Greek Koine

and the Logic of a Standard Language
Stephen Colvin
My purpose in this chapter is to look at the factors that led to the rise of the koine,
and by so doing give a sketch of what, in my view, a definition of the koine might
look like. I do not mean that I shall be looking at the external realia of the expansion
of the koine (Macedonian imperialism, and so on), though this has its place in the
study of any language. Rather, taking as a starting point the view that language is in
an important sense a cultural product (an approach that is at least as old as Edward
Sapir), I want to investigate how we imagine a community shifts from thinking
about language as a bundle of overlapping resemblances (Wittgenstein) to thinking
of it as essence with variation (Plato).
It is part of my premise that the first model reflects a view of language that
obtained in the Greek world at the time (say) of the Persian wars. The reference
to Wittgenstein is a short-hand reference to his critique in the Philosophical
Investigations of an ancient, and still prevalent, view of definition:
Consider for example the proceedings that we call ‘games’. I mean board-games,
card-games, ball-games, Olympic games, and so on. What is common to them all? –
Don’t say: ‘There must be something common, or they would not be called “games”’
– but look and see whether there is anything common to all. – For if you look at them
you will not see something that is common to all, but similarities, relationships, and
a whole series of them at that . . . And the result of this examination is: we see a
complicated network of similarities overlapping and criss-crossing.1
Wittgenstein here questions an approach (the second model) which is captured by

Plato in an amusing passage of the Meno. Socrates, trying to manoeuvre Meno into
defining the essence of virtue, turns to the word ‘bee’ to illustrate his point:
I seem to be in luck. I wanted one virtue and I find that you have a whole swarm of
virtues to offer. But seriously, to carry on this metaphor of the swarm, suppose I asked
you what a bee is, what is its essential nature [εἴ μου ἐρομένου μελίττης περὶ οὐσίας
ὅτι ποτ᾽ ἐστίν], and you replied that bees were of many different kinds, what would
1
Wittgenstein (1953) §66.
33
34 stephen colvin
you say if I went on to ask: ‘And is it in being bees that they are many and various and
different from one another? Or would you agree that it is not in this respect that they
differ, but in something else, some other quality like size or beauty?’2
The second model represents not only a view of definition which is still widespread,
but also a view of the relationship between language and dialect (or, more simply,
of linguistic variation). That is to say, just as in the ancient grammatical tradition
nominal inflection was conceived as a citation form with πτῶσις (Latin declinatio),
falling away from the standard,3 so by the time of the Alexandrian grammarians
linguistic variation was regarded in terms of a standard which underwent mutation
or modification. The phrase for ‘citation form’ (ὀρθὴ πτῶσις ~ Latin casus rectus)
is echoed in the term ὀρθότης ‘uprightness, correctness’, used of language by
Gorgias, and picked up by Aristophanes in a reference to Sophistic discourse.4
The slippage between correctness as a property of locution (a matter of style and
rhetorical theory) and correctness as a property of language itself (a concern of
dialectology) was as easy in antiquity as it is in the modern national context. At
approximately what period this slippage started in the Greek world, and how
it relates to the development of the koine, is an important issue in the study of
how a notion of standard language emerged. The focus of this chapter will be the
willingness or otherwise of speakers in the Greek world to modify the way they
wrote more or less before 400 BC.
Western classical scholarship has not, on the whole, dealt in a satisfactory way
with the notions of linguistic diversity and standard language in the ancient world,
no doubt because the glasses through which we look at ancient views on language
are inherited from the classical tradition itself, to borrow another image from
Wittgenstein).5 The interpretation of linguistic variety as essence and variation
(mostly conceived as corruption) which emerged in the complex sociolinguistic
milieu of Hellenistic and Roman Greece was easily translated into a Latin context
by Roman grammarians, and spread with equal ease into medieval and modern
European thought. There was no obvious external challenge to this way of thinking
about language. The closest neighbour and intellectual rival of the Greco-Roman
world (subsequently medieval and Byzantine Europe) was the Islamic Arab
civilization along its southern border. Here by coincidence a similar model (and
an analogous grammatical tradition) emerged, owing to the canonization of the
language of the Qur’ân as ‘Arabic’ tout simple, and (as in Greece) the subservience
of grammatical activity to textual exegesis.6
2
Meno 72a–b, tr. Guthrie.
3
First at Aristotle, Poetics 1457a18.
4
So also in Plato; useful discussion by Dover (1993) 29–31; Gorgias fr. 6.2 (Epitaphios: probably
last quarter of the fifth century); Ar. Frogs 1181 τῶν ἐπῶν.
5
Wittgenstein (1953) §104.
6
Unlike the Greeks, the Arab grammarians refer to notional (idealized) native speakers (in general,
the Bedouin) whose language is by definition ‘correct’ (the Roman attitude is perhaps closer to that
of the Arabs in this respect). The critical similarity, however, is as follows: ‘Through a process of
idealization, not uncommon in a speech community in which there is diglossia, this language [sc. the
The Greek koine and the logic of a standard language 35
The questions that need to be answered are: (i) where exactly do we find
standard and variation in the epichoric period (Greece before the Macedonian
hegemony) and what is the relationship between them? (ii) how does this relate to
the so-called Hellenistic koine? In Greek studies linguistic diversity as a fact has of
course been faced, since it is a troublesome feature of ancient texts which cannot
be ignored. The dominant paradigms in interpreting the data have, it seems to
me, been as follows: to handle linguistic diversity as a literary device, and to view
linguistic homogeneity (the rapid rise of the koine) as analogous to the imposition
or spread of a modern colonial language such as English or Spanish.
In this context modern scholarship has on the whole found it convenient to divide
the Greek data into three distinct categories, corresponding to three chronological
stages. Firstly, the language of oral epic (Homer and Hesiod), which is a special
case and is explained by appeal to the literary term Kunstsprache. This is a vague
notion which denotes a language marked by forms belonging to different dialects
and different periods: it could never have been spoken at any particular time, and is
therefore ‘artificial’.7 The position taken here is that the difference between the epic
Kunstsprache and other literary languages of the Greek world (poetry and prose)
is not so qualitatively significant, given that every genre of ancient Greek literary
output was marked by a characteristic (and, on a simplistic view, ‘artificial’) mixture
of dialect forms. Secondly, we find in the so-called archaic and classical periods
a period of acknowledged diversity where across the Greek world people spoke
and wrote in various local dialects, except that when composing high literature
they often used a ‘foreign’ dialect, or at least a dialect marked by ‘foreign’ forms.
This can be interpreted as an artistic-literary device, connected with the classical
notion of ‘first inventor’ (πρῶτος εὑρετής): the genre reflects the dialect of the
person or group most associated with the development of that genre.8 Finally the
dialects were squashed by the koine, an idiom spread by the Macedonian empire
and clearly an expanded form of Attic.
None of this is unreasonable as part of a literary analysis of individual texts,
or indeed of a particular poetic idiom (Homeric language, or the language of
Pindar). The problem is that the relationship between diversity and standard has
hardly been explored beyond the most basic level. What is needed is a framework
within which to make sense of the (quite substantial) data that we have. This is a
correct variety] is also regarded as the mother tongue of all members of the community who have
received an education and make an effort to speak correctly’: Versteegh (1997) 42.
7
Since the work of Parry it has been accepted that the linguistic mix of the poetry owes its genesis
to the input of a number of different dialects in oral composition over a long period. A useful overview
in Palmer (1980) 83–101.
8
Attic comedy and forensic rhetoric are generally (and no doubt rightly) thought to be the literary
idioms closest to ‘real’ Attic. This of course begs the question, which Attic? Presumably one linguistic
variety (and not necessarily the Umgangssprache) of a particular socio-economic group. In any case,
it is clear that comedy – being written in verse – contains literary forms (such as the disyllabic dative
plural) which are by convention ‘filtered out’ of dialectal analysis. Analogous processes may have been
at work even in naturalistic orators such as Lysias. See in general Dover (1997) 96–130, and for comedy
Willi (2002).
36 stephen colvin
sociolinguistic rather than a literary issue, and for answers we can turn to models
from modern linguistic investigation.
We can start by asking: was there no notion of standard in the epichoric period?
How, for example, do we imagine the widespread familiarity with epic across the
Greek world influenced the linguistic culture? How do we imagine the dramatic
transition from dialect diversity to koine took place?
An accident of historical terminology on the Greek side has prevented the soul-
searching which Latinists have been unable to avoid: that is to say, the question of
when Latin turned into Italian (or Spanish, or Romanian) has not been an issue for
Western Hellenists, simply because we speak of Ancient and Modern Greek, but
not Ancient and Modern Latin. The question (when did Latin turn into Italian)
is badly put and misleading, and we shall return to it: but it forced Romance
philologists to question notions of ‘standard’ versus ‘vernacular’, in fact to sharpen
their notions of what constitutes a koine. In the Greek world, on the other hand,
one might think that people spoke and wrote as they pleased until the business-
like imperialism of Macedon enforced a standard. The notion of koine, however,
once unpackaged, becomes difficult to contain within its traditional chronological
boundaries. One problem that has dogged discussion of standard, variation
(regional and social dialect), and koine is that the disciplines these terms pertain to
(classics and linguistics) developed in an unusual sociolinguistic context, namely
Western Europe and North America; and the language model that is in some sense
built into them reflects their origin (nation states with peculiar colonial histories
and standardized national languages). This is true also of the term diglossia, which
was introduced into academic linguistic discourse (by Ferguson in 1959) in an
effort to describe a situation which is essentially alien to Western thought about
language: linguists have used the term ever since while arguing about what it
means and criticizing Ferguson’s first attempt to apply it.9 The problem, however,
lies in the underlying language model, rather than in this or that nuance tacked
onto the term: to talk of a continuum is hardly more helpful than talking of high
and low varieties if one fails to distinguish between the language that speakers are
producing on each occasion, and the language they imagine they are producing.
We have already noted some of the similarities between the Greek and the
Arabic language communities; and if we look to Arabic for a model to understand
the Greek koine,10 we are immediately tempted by a new working definition: on
the analogy of modern standard Arabic we can say that for our purposes the koine
constitutes a written standard to which no spoken variety corresponds exactly.
It is an abstraction which arguably corresponds to the feeling of speakers about
their linguistic identity; adherence to the ‘standard’ is, on this model, a positive
statement, not the result of coercion. We are then forced to question whether there
could ever be a written language which corresponded precisely to a vernacular
9
Criticism by El-Hassan (1977), among others. Ferguson (1959) did not in fact coin the term: see
Shalev (2006).
10
Following Versteegh (2002).
(spoken variety). If not, it might turn out that this is not an accidental but an
essential property of written language: and that koine starts in Greek with (a) the
development of writing, and (b) the development of a sense that there existed a
body of canonical ‘texts’.11
Jumping back for a moment to the other side of the dark ages, the decipherment of
Linear B started a perplexed debate among Mycenologists: how could tablets from
across the Greek world (from Crete to Thebes) be so linguistically homogeneous?
The answer seems to be a common scribal language (another koine), bundled up
with a largely uniform script, which does not reflect local dialect differences.12
Mycenaean script disappeared, of course, with Mycenaean civilization, and
the Greek dialects flourished and diverged for two or three relatively unsettled
centuries, free from the checks that a writing system and associated cultural
paraphernalia interject.13 If any sense of common ‘Greek’ identity survived these
centuries it must have been tied up with cultural artefacts such as poetry. One thin
thread of continuity which emerged from the so-called Dark Age was the tradition
of heroic song which the Greeks, in common with many other Indo-European
peoples, had maintained from an earlier era. The designation of epic language
and other poetic idioms as koinai is well-established,14 and implies, not that the
poets working within the tradition use an identical idiom, but that they refer back
to a common idiom which their own production both instantiates and expands.
The poetic idiom is a variety of language which is identifiable by being subject
to certain norms, and the reasons for accepting these norms are cultural: a speech
community accepts constraints (such as foreign dialectal forms) for the sake of a
perceived benefit (the location of the ‘text’ in a particular space). The poetic koinai
are not so different from the political koine which is most associated with the term.
In this connection we may recall Lord’s study of Homeric composition, where he
draws an analogy between the bardic appropriation of the poetic tradition and the
speaker’s mastery of her native tongue:
When we speak a language, our native language, we do not repeat words and phrases
that we have memorized consciously, but the words and sentences emerge from
habitual usage. This is true of the singer of tales working in his specialized grammar.
He does not ‘memorize’ formulas, any more than we as children ‘memorize’ language.
He learns them by hearing them in other singers’ songs . . . The learning of an oral
poetic language follows the same principles as the learning of language itself.15
The notion of a literary koine is useful, therefore, because it implies a package

11
As Homeric and Hesiodic epic gradually supplanted rivals. The social tensions that a body of
classical literature can produce is a favourite theme in Aristophanes: see Dover (1993) 24–37.
12
So Bartoněk (1966), Palmer (1980) 53; cf. Duhoux (1985) 38–9. The term ‘Mycenaean koine’
is also used by archaeologists to refer to the material culture of the region.
13
See Morpurgo Davies (1987), on the development of ancient notions of ‘language’ versus
‘dialect’, and Ruijgh (1995) on the fluctuating date of the Greek alphabet (recent work has tended to
put it back, even as early as the tenth century BC).
14
For a good collection of essays on the subject, see Hodot (2001).
15
Lord (1960) 36.
38 stephen colvin
of cultural behaviour greater than a mere literary dialect or Kunstsprache. It is

widely, if vaguely, accepted that Homeric epic did not have to be ‘translated’ from
one dialect to another: this is a statement partly about the pan-Hellenic nature
of the texts, and partly about the mutual intelligibility of the Greek dialects. It
is of course true that many peculiarities of the epic tradition will not have been
subject to dialectal alteration: features of phonology, morphology or lexicon which
were either integrated into specific formulas, or constituted part of the resources
of the ‘artificial’ bardic language.16 Nevertheless, Parry’s analysis of bardic
method gives every reason to believe that epic language (especially perhaps at the
phonological level) was adapted to local dialect within these parameters.17 There
is indeed evidence from the dialects for regional varieties of epic diction: Boeotian
inscriptions, for example, show that there must have been a native Homeric
tradition which was fundamentally the same as the tradition familiar to us from the
vulgate of Homer, but which had made itself at home in the Greek of Boeotia.18 The
notion of normativity, then, that epic language carries with it is not tied narrowly to
dialect; but it does imply that genre is associated with a specific linguistic variety
which is likely to contain alien elements. There is an interesting analogy with the
local alphabets of Greece: while it seems likely that the Greek states individualized
their own varieties of the script (to achieve distinctiveness),19 they are nevertheless
all variations on a single pattern (recognisably the same script).
It is notoriously uncertain when the Homeric poems were written down: but
the spread of Homeric epic, and in general the whole pan-Hellenic consciousness
that has been connected with the later Geometric period (eighth century BC),20
seems to coincide also with the spread of the new alphabet in the Greek world.21
These two developments (which I am assuming to be unrelated) must have had an
impact on Greek linguistic consciousness. In particular, it seems likely that they
played an important role in the rise of three (related) ideas: (a) genre, specifically
the connection between form and a peculiar variety of language; (b) the notion of
a standard, against which everyday speech could be compared and judged; and (c)
the connection between text and (national) identity.22
16
Apart from the technical reasons, there are also literary or ideological reasons for maintaining
exotic and archaic elements in a special register such as epic: witness the use of this medium by the
Delphic oracle.
17
So rightly Horrocks (1997) 18: ‘Although such an artificial language could never have been
the spoken dialect of any region, it should be emphasized that the fundamentals of epic grammar and
diction were subject to regular modernization, broadly in line with the contemporary spoken Greek of
the localities where epic bards were working, albeit with archaic and “foreign” dialectal retentions at
each stage.’ Cf. Parry (1932) 9–10, 17–21.
18
See the study by Vottéro (1996).
19
So Luraghi (forthcoming).
20
So Snodgrass (1971); cf. Nagy (1979) 7 and (1990) 52–115.
21
The relationship between writing and Homeric epic has been treated at length by Powell (1996),
whose conclusion (not widely accepted) is that Phoenician script was adapted for the purpose of writing
down epic.
22
Cf. Meillet (1929) 138: ‘Au moment où la littérature est apparue et s’est développée, sans doute
So (returning to Plato’s bee) if Greeks had a sense that they were Greeks by
reference to a common national property such as epic, then one can see how there
might be a subtle shift in the linguistic culture, as bundled-up with epic comes
the idea of epic language. At the same time, writing leads to a number of regional
standards, to which we can now turn.
The political structure of the ancient Greek world meant that there was no
standard language corresponding to Latin in Roman Italy, or a modern standard
such as English, French or Italian. It is hard to think of a parallel, ancient or
modern, for this situation: a collection of small states speaking closely related
dialects, with a loose sense of political and ethnic affiliation, and each state using
its own written standard (and indeed its own variety of the alphabet). Even within
the Greek world, however, there were exceptions to the principle of unchecked
diversity: the larger Greek city-states (πόλεις) such as Attica and Laconia must
presumably have contained numerous ‘sub-dialects’ (social and regional) for
which there was no written form; and in Ionia the Ionian states adopted a written
standard based on Miletus at such an early stage that there is very little evidence
for the diversity which Herodotus records and general dialectology would in any
case predict.23 There were, however, distinctive cultural attitudes towards language
use and literacy across the Greek world: we can contrast the Ionian practice with
(say) that of Laconia or Boeotia, where differences in orthographic culture grew,
presumably, out of differing degrees of interest in language. The Boeotians made
great efforts to record changes in their language, as it raced ahead of all the other
dialects towards Modern Greek, while the laconic Spartans seem to have been
relatively uninterested in orthographic consistency.24
The area of the Greek world for which we have the most evidence is of
course Attica, which, as we noted above, is unlikely to have been linguistically
homogeneous. Clearly there is orthographic standardization: but the phonology
and morphology have also been standardized, and, since we are used to believing
that Greeks were unworried by linguistic diversity, we need to ask why. I have
argued elsewhere that there is evidence in the fifth century for a prestige variety
within Attica, rather than the mere recognition that different social groups speak
in different ways.25 The literary evidence points to this: Old Comedy refers to
politicians who, it is alleged, could hardly speak proper Attic.26 This is generally
the result of one or both of the following deficiencies: barbarian blood (the link
with ethnicity), and low social status (stemming from the absence of an appropriate
paideia – to speak good Attic one needs an education). Another reference to what
avec rapidité, le monde hellénique, tout divers qu’il était, sentait son unité.’ See further Silk, pp. 13–18
above.
23
Herodotus 1. 142, γλῶσσαν δὲ οὐ τὴν αὐτὴν οὗτοι νενομίκασι, ἀλλὰ τρόπους τέσσερας
παραγωγέων.
24
Cf. Bourguet (1927) 8): ‘Je crois bien que nulle part n’est attesté un usage aussi peu tyrannique
qu’à Sparte. Ce fait est dans doute fort inattendu dans un pays de discipline.’
25
Colvin (2004a).
26
Colvin (1999) 282, 292.
40 stephen colvin
we might call social dialect is the ranting of the Old Oligarch against the incursions
of the new ‘Piraeus Greek’;27 this, no doubt, was the expanded international Attic
which lay at the vernacular end of the continuum later known as the Hellenistic
koine.
There is also evidence from epigraphy: the contrast between public and private
inscriptions is an obvious place to look. We know that the Ionic alphabet was in
widespread use in fifth-century Attica; Threatte puts the main period of transition
at 480-30, and concludes that ‘Ionic script was employed by most persons for
private purposes by the last quarter of the century.’28 Public inscriptions, on the
other hand, are written almost exclusively in Attic script until 403/2 BC. By the
middle of the fifth century a stoichedon Attic chancellery style was established,
which Immerwahr describes as a ‘purist Attic alphabet that tended to resist the
influx of Ionic letters’.29 Now, the first page of any linguistic textbook warns you
not to confuse script with language; but a written standard plays an enormously
important role in the formation of a social consciousness of a ‘language’,30 and we
shall need to return to this.
For further evidence of standardization we can turn from script to language
proper, and here we do indeed find differences between private and public
inscriptions at the phonological and morphological levels. They have been well-
documented, and include the retention of the a-stem dative plural ending -ασι/
-ησι in public inscriptions until around 420, when (it must have been) by official
decision they were replaced by the -αις that one finds in literary texts and private
inscriptions.31 (The distribution of o-stem dative plurals is similar, although -οισι
gives way to -οις by the middle of the century.) Phonological variation is notoriously
difficult to detect, owing to the standardized orthography that we have already
considered. To find traces of colloquial varieties in a corpus language the best one
can do is to look at graffiti, curse tablets, and a variety of private inscriptions in the
hope of finding an orthographic mistake: that is, a spelling which gives an insight
into the pronunciation of a (relatively unlettered) writer. The recent publication
of an ostrakon bearing the text τὸν λιμὸν ὀστρακίδ<δ>ω indicates (in my view)
the likelihood of a social dialect of Attic which shared the δδ (< *dy) reflex with
Boeotian in place of standard Attic ζ.32
For cultural and political reasons Boeotian was a dialect the Athenians may have
had particular reasons to distance themselves from (we can speculate that this may
27
Ps.-Xenophon, Ath. Pol. 2. 7–8.
28
Threatte (1980) 33.
29
Immerwahr (1990) 121.
30
Cf. Meillet (1929) 121: ‘Il y a de l’hypocrisie dans le dédain des linguistes pour les langues
littéraires.’
31
Dover (1981), who also points out that Ar. Thesmo. 431 (τὰ δ᾽ ἄλλα μετὰ τῆς γραμματέως
συγγράψομαι, ‘the rest I shall get written down with the secretary’) implies the existence of a
chancellery style.
32
Kerameikos ostrakon published by Brenne (2002) 97–100, and dated to c. 471. Discussion of
linguistic implications by Colvin (2004a).
be a reason why the chancellery language took such a long time to let go of the
disyllabic dative plural). We have already noted the tendency of the Boeotians to
innovate orthographically as their dialect changed. Now, Luraghi has argued that
Greek states with similar dialects manipulated their graphic systems to achieve
distinctiveness from their neighbours, and the difference between the Boeotian
and Attic practice is intriguing in this light:33 the implication is that differences in
regional epigraphic standards were deliberately maintained. However, Athenian
conservatism may also be partly due to the large body of literary material in Attica
which was written down by the late fifth century, a tradition of writing which
could be traced back to epic itself (we may imagine that the Athenians regarded
themselves as heirs to the Ionic literary tradition). Constant metacharakterismos
(transcription) would be very inconvenient in a culture which was coming to
regard itself as the centre of Greek literary production; and we remind ourselves
that this is another area in which the Athenians differed from the pitiful Boeotians,
who were probably without a tradition of written high literature at this date.34
Here it may be helpful to consider the question we posed above regarding the
change from Latin to Romance, where the link between linguistic consciousness
and written standard seems to have played a central role. In the last two decades
it has become accepted in Latin studies that one must distinguish between
linguistic change (which may be gradual, and generally operates at a level below
the consciousness of the language speakers), and change in linguistic terminology
(including the designation by name of languages, dialects, and other varieties). To
name a linguistic variety is to make an ideological choice which is likely to have
social or political implications; it need not be the immediate result of linguistic
change (and conversely, linguistic change need not result in a change in language
name). Latin turned into Italian when speakers stopped calling it Latin, shortly
after Dante established a new written standard. Language naming seems always
to have been intimately connected with the creation of a written variety: the
Italian discovery of Italiano follows a similar development in Gaul, namely the
creation of written Old French.35 Dante had called Latin Grammatica, and Italian
Latino: he ‘did not regard Latin as the origin of the popular languages, but rather he
apprehended it as a common way of writing, unaffected by dialectal differences’.36
There is a parallel between Latin in the linguistic diversity of ‘pre-Romance’
Europe and modern standard Arabic: the contemporary difference between the
two language areas is the result of European nationalism in the early modern period
33
Luraghi (forthcoming).
34
The effect of metacharakterismos in Attica has been overestimated: it was formerly imagined that
literary production was in the Attic script until 403/2, when a wholesale translation into Ionic script
took place which is likely to have introduced errors into the manuscript traditions of earlier texts. It
seems clear, however, that most literary production made use of the Ionic script by the second half of
the fifth century: see Colvin (1999) 92–103 and D’Angour (1999).
35
Wright (1991b).
36
Janson (2002) 123, a sketch of the position given more fully in Janson (1991). See also Lloyd
(1991) and Wright (1991b).
42 stephen colvin
which led to the creation of a number of regional standards (the development is still
in progress in post-Franco Spain and elsewhere).
To return to the Greek world: the beautiful stoichedon inscriptions erected by
the Athenian state must have been influential in leading to a notion of Attic as a
theoretical entity which defined a political-ethnic group (just as the Homeric text
may have contributed to the concept of pan-Hellenism some centuries earlier).
The critical period in which the groundwork was laid for a new political koine was
the time between the Persian wars and the Macedonian hegemony – precisely,
in fact, the period which has traditionally been designated ‘classical’ in the West.
The Persian wars reinforced a sense of common identity among the Greeks, and
were followed by a period of prosperity and self-confidence which saw an increase
in epigraphic and literary production. Education became increasingly common,
and was to a certain extent institutionalized; and this obsession with education
was a development which transcended the individual city-states owing to the pan-
Hellenic nature of the sophistic movement.
This is the context which made the emergence of a new koine natural and,
indeed, inevitable. This was the koine of paideia, which covered education, rhetoric
and (therefore) political discourse. It was, as a result, the koine of literary prose, and
in this sense was the true heir to Ionic. No doubt the Athenian empire led to the
emergence of an expanded ‘international’ Attic(-Ionic) vernacular, but without the
underpinning of the koine this would have been just one more lingua franca that
perished when the conditions which gave rise to it changed. No doubt also the
Ionic flavour of the epic vulgate contributed to the sense that the Hellenistic koine,
with its Ionic flavour, was a pan-Hellenic dialect; but there is a complex nexus of
connections here, rather than the immediate causal link that has been supposed.
The koine did not take hold in the Greek world because it was imposed by
the Macedonian regime. It was the natural idiom for the political and cultural
structures of the new Greek world, as the language of government (decrees, laws,
letters) and education. Clearly, the Hellenistic world provided the bureaucratic
and institutional framework for a prose koine, which was imposed only in the sense
that ‘Homer’ imposed epic language onto heroic verse in archaic Greece (both
koinai are positive statements of identity and cultural loyalty). And there is no
reason to suppose that local dialects, or even local languages (for example, in Asia
Minor),37 ceased to be spoken as a result of the integration of the koine into the
structures of government and elite education. The koine was the ideal, and the
symbol of Greek history, culture, and identity. In an analogous manner, Arabic
vernaculars persist, and Romance vernaculars also – until their political conversion
into languages. In any case (as with Arabic and Latin), phonological changes which
are already detectable in the late-classical period would have made many features
of the educated standard highly ambiguous in a spoken context.38
37
I have suggested elsewhere that Lycian continued to be spoken long after its disappearance from
the epigraphic record (Colvin 2004b).
38
In a ritual context, of course, there are likely to be strategies employed to preserve important
To return to a question we posed at the beginning of the chapter: the transition

from the Greek idea of Greek as a family of overlapping idioms to the idea of Greek
as an ideal with approximations will (on this view) have been in progress in the
classical period itself; we can imagine the koine bubbling to the surface towards
the end of this period like an underground stream which has been there all along,
rather than simply appearing by decree in the new world of Alexander. Indeed,
more widely, it might be argued that Western classicists have always tended to
overestimate the break between the classical and the Hellenistic worlds, and to
overlook connections at the literary or political level. New Comedy and Hellenistic
poetry would undoubtedly look less new if we had more fourth-century poetry and
drama.
We might well question what we are talking about when we assign a name to
this idiom, which lasted a millennium or more. The present chapter has taken the
position that it is more helpful to see a koine as an abstract norm based on a written
tradition than as something likely to emerge from the mouth of a particular speaker.
The development of chancellery language in imperial Athens systematized
a conceptual framework whose genesis we can see, if we choose, in the very
beginnings of Greek literacy and pan-Hellenic identity. The literary prestige of
the Ionic dialect is unlikely to have had an impact on the spoken language in the
Hellenistic period: languages generally change in the direction of the lowest social
variety, not the highest;39 and this is in fact what spoken Greek (like spoken English,
in a later age) did. The koine is an idiom which implies that the speakers know who
they are (Greeks); they know which language they are speaking (Greek); and they
know that they will be able to communicate with anyone who shares the paideia
which has become the defining feature of their language community.
morphophonemic or lexical distinctions in a liturgical language (at least for an educated elite); this may
offer a clue to understanding how declamations of classical or classicizing material were conducted in
late antiquity (Libanius, for example, in the fifth century AD).
39
See e.g. Milroy (1992), who comments (on 149): ‘the difficulty in explaining why linguistic
changes do not usually move in the direction of the prestige norm (as used by elite groups) is a familiar
one.’
44 stephen colvin
References
Bartoněk, A. (1966), ‘Mycenaean koine reconsidered’, in L.R. Palmer and J. Chadwick
(eds.), Proceedings of the Cambridge Colloquium on Mycenaean Studies, Cambridge: 95–
103.
Brenne, S. (2002), ‘Die Ostraka (487 – ca. 416 v. Chr.) als Testimonien’, in P. Siewert (ed.),
Ostrakismos–Testimonien, vol. 1, Stuttgart: 36–166.
Bourguet, E. (1927), Le dialecte laconien, Paris.
Colvin, S.C. (1999), Dialect in Aristophanes: The Politics of Language in Ancient Greek
Literature, Oxford.
Colvin, S.C. (2004a), ‘Social dialect in Attica’, in J.H.W. Penney (ed.), Indo-European
Perspectives. Studies in Honour of Anna Morpurgo Davies, Oxford: 95–108.
Colvin, S.C. (2004b), ‘Names in Hellenistic and Roman Lycia’, Yale Classical Studies, 31
(The Greco-Roman East): 44–84.
D’Angour, A. (1999), ‘Archinus, Eucleides, and the reform of the Athenian alphabet’,
Bulletin of the Institute of Classical Studies, 43:109–30.
Dover, K. J. (1981), ‘The language of classical Attic documentary inscriptions’, Transactions
of the Philological Society: 1–14 [repr. in id., Greek and the Greeks (Oxford 1987), 31–
41].
Dover, K.J. (1993), Aristophanes, Frogs, Oxford.
Dover, K.J. (1997), The Evolution of Greek Prose Style, Oxford.
Duhoux, Y. (1985), ‘Mycénien et écriture grecque’, in A. Morpurgo Davies and Y. Duhoux
(eds.), Linear B: A 1984 Survey (Louvain-La-Neuve): 7–74.
El-Hassan, S.A. (1977), ‘Educated spoken Arabic in Egypt and the Levant: a critical review
of diglossia and related concepts’, Archivum Linguisticum 8: 112–32.
Ferguson, C.A. (1959), ‘Diglossia’, Word, 15: 325–40
Hodot, R. (ed.) (2001), La koiné grecque antique IV: les koinés littéraires, Nancy.
Immerwahr, H.I. (1990), Attic Script: A Survey, Oxford.
Janson, T. (1991), ‘Language change and metalinguistic change: Latin to Romance and
other cases’, in Wright (1991a): 19–28.
Janson, T. (2002), Speak: A Short History of Language, Oxford.
Lloyd, P.M. (1991), ‘On the names of languages (and other things)’, in Wright (1991a):
9–18
Lord, A.B. (1960), The Singer of Tales, Cambridge, Mass..
Luraghi, N. (forthcoming), ‘Local scripts from nature to culture’, in P. Haarer (ed.),
Alphabetic Responses to Western Semitic Writing, Oxford.
Meillet, A. (1929), Aperçu d’une histoire de la langue grecque, 3rd edn, Paris.
Milroy, J. (1992), ‘Social networks and prestige arguments in sociolinguistics’, in K. Bolton
and H. Kwok (eds.), Sociolinguistics Today: International Perspectives, London: 146–62.
Morpurgo Davies, A. (1987), ‘The Greek notion of dialect’, Verbum, 10: 7–28 [repr. in T.
Harrison (ed.), Greeks and Barbarians (London 2002): 153–71].
Nagy, G. (1979), The Best of the Achaeans: Concepts of the Hero in Archaic Greek Poetry,
Baltimore.
Nagy, G. (1990), Pindar’s Homer: The Lyric Possession of an Epic Past, Baltimore.
Palmer, L.R. (1980), The Greek Language, London.
Parry, M. (1932), ‘Studies in the epic technique of oral verse-making, II: the Homeric
language as the language of an oral poetry’, Harvard Studies in Classical Philology, 43:1–
50 [repr. in A. Parry (ed.) The Making of Homeric Verse (Oxford 1971): 325–64].
Powell, B.B. (1996), Homer and the Origin of the Greek Alphabet, Cambridge.
Ruijgh, C.J. (1995), ‘D’Homère aux origines proto-mycéniennes de la tradition épique:
analyse dialectologique du langage homérique, avec un excursus sur la création de
l’alphabet grec’, in J.P. Crielaard (ed.), Homeric Questions: Essays in Philology, Ancient
History and Archaeology, Amsterdam:1–96.
Shalev, D. (2006), ‘Heliodorus’ speakers: multiculturalism and literary innovation in
conventions for framing speech’, Bulletin of the Institute of Classical Studies, 49: 175–
201.
Snodgrass, A. (1971), The Dark Age of Greece, Edinburgh.
Threatte, L. (1980), The Grammar of Attic Inscriptions, vol. 1, Berlin and New York.
Versteegh, K. (1997), Landmarks in Linguistic Thought III: The Arabic Tradition, London.
Versteegh, K. (2002), ‘Alive or dead ? The status of the standard language’, in J.N. Adams
et al. (eds.), Bilingualism in Ancient Society, Oxford: 52–74.
Vottéro, G. (1996), ‘Koinès et koinas en Béotie à l’époque dialectal (7e – 2e s. av. J.C.)’, in
C. Brixhe (ed.), La koiné grecque antique II: la concurrence, Nancy: 43–92.
Willi, A. (ed.) (2002), The Language of Greek Comedy, Oxford.
Wittgenstein, L. (1953), Philosophical Investigations, Oxford.
Wright, R. (ed.) (1991a), Latin and the Romance Language in the Early Midde Ages,
London.
Wright, R. (1991b), ‘The conceptual distinction between Latin and Romance: invention or
evolution?’, in Wright (1991a):10–13
3
Primary Education in a Non-Standard Language

as a Tool of Social and National Integration:
The Case of Vernacular Greek, 1923–30
George Kritikos
The Lausanne Peace Treaty of 1923, which ended the Greco-Turkish war, called
for something new in world history: the first obligatory exchange of populations
based exclusively on the criterion of religion. By April 1923, some 355,000 of the
Muslim population of Greece had been moved from Greece to Turkey, while the
census of 1928 recorded that 1,221,555 refugees from Turkey had been finally
settled in Greece.1
That settlement took place in a context of linguistic contestation within the
educational and social life of the country, provoked by the respective propagators
of the ‘purist’ language (katharevousa) and the vernacular (demotic). During
Eleftherios Venizelos’ governments in the 1910s, the education reforms of 1913
and 1917 introduced demotic in the first four years of the primary schools.
Venizelos also appointed supporters of demotic – Alexandros Delmouzos, Dimitris
Glynos, Manolis Triantafyllidis and Dimitrios Dingas – to important positions in
the ministry of education. In the reform legislation, emphasis was placed on the
‘education of the people’, and specifically on the strengthening of the vocational-
technical sector of the system as well as on the introduction of demotic in the
primary schools.2 All the demoticist reforms met with strong opposition from the
conservatives. It was conservative governments that had introduced rules for the
protection of katharevousa as official state language into the Constitution of 1911
and subsequently designated any attempt to change its status as a punishable
offence. The conservatives likewise rejected the institutionalization of demotic
and abolished all the demoticist reforms of 1913, and especially those of the period
1917–20.
1
Statistical Annual of Greece (1931) 31.
2
Dimaras (2006) 336.
47
48 george kritikos
Nonetheless, the education reform of 1929 provided for the teaching of demotic
in all years of elementary school for the first time in Greek history. Moreover, the
reform brought to an end a troubled period for Greek education: between 1920
and 1929 there were thirty-four changes of government and twenty-five Ministers
of Education. The reform, indeed, changed the administrative framework of Greek
education for the first time in seventy-two years, and established a new one, which
lasted until 1964.3
Not surprisingly, historians of these events agree in considering the year 1929 as
a turning point in the education policy of the Greek state. However, scholars have
failed to analyse the reasons why demotic was institutionalized only at primary
school level. In the event, the official language of the state, the language of the
press and the church, as well as the language of secondary and higher education,
all remained katharevousa. The then Minister of Education, Georgios Papandreou,
remarked that although in principle he favoured the general use of demotic in
secondary education, as also in the administrative and academic spheres, the
Greek vernacular was as yet a non-standard and non-codified language in terms of
grammar and syntax. He concluded that the institutionalization of demotic would
be a ‘turning point’ from which Greece was still ‘a long way away’.4 There were,
however (as I shall try to show), other reasons, besides its lack of standardization,
for not teaching the Greek vernacular at the levels of secondary or higher education.
These who have studied the education system and its reforms tend to ignore
the role of the Asia Minor refugees in the language reform and, conversely, the role
of this reform in the social integration of the refugees with the native population.
There are many who subscribe to a ‘Hellenization’ theory, whereby it is supposed
that all the Greek-Orthodox populations that came from Asia Minor carried a
homogeneous Greek culture with them and ‘Hellenized’ all the areas in which
they were established. The advocates of this theory seem to forget that since
religion was the exclusive criterion of exchangeability, there were many refugees
who found themselves in virtual exile among their co-religionists who spoke a
different language.5 I propose a different interpretation that will shed some light
on the link between the language reform of 1929 and the refugee settlement. My
main focus will be on the policy of teaching the non-standard Greek vernacular
only in primary schools, and the question whether this served the integration of the
refugee and native populations in national as well as socio-economic terms.
My argument is organized into two different sections. The first addresses the
question of national integration by focusing on the linguistic composition of the
refugees who fled to Greece and the linguistic map of the country after 1923. The
second explores the role of language in social integration.
3
Fragoudaki (1992) 59.
4
Efimeris ton Syzitiseon, Fourth Session (9 July 1929) 94.
5
Lewis (1965) 75–8.
primary education in a non-standard language 49
Integration in national terms

Schooling in Asia Minor was developed in the multi-lingual and multi-cultural
environment of the Ottoman Empire, where the millet religious units functioned as
the main administrative and ethnic entities.6 That education system delegated the
responsibility for providing education to local authorities.
From the mid-nineteenth century, there was an expansion of the Greek
school network, which soon grew rapidly in size and importance. The impetus
for this restructuring came from the prosperous commercial class, whose growing
economic power now began to express itself culturally at a local level. Across the
empire, an entire school network was developed at community level, because ‘to
found a school or hospital in his native town is the honourable ambition of every
Greek merchant’.7 These transactions in Ottoman society were backed up by
legislative measures which provided for a reform programme of secularization and
Westernization within the Empire under the name of Tanzimat.8
According to a map found in the Greek literary and historical archive (ELIA) in
Athens, in 1912 there were 1,633 Greek schools with 165,381 pupils in operation
in Asia Minor alone (excluding Adana, Dodecanese and the areas dependent on
the Patriarchate of Antioch and Constantinople).9 The proliferation of the Greek
education system throughout the empire is also apparent in the increased ratios of
enrolment at primary school (10.23%), which were now almost as high as those in
Greece itself (10.40%).10
Nevertheless, one may question whether these high levels of school enrolment
represent an equally high level of language dissemination. As Karl Dieterich
remarked, ‘The Turkish language was taught as a foreign language, occupying the
same place in the curriculum of the Greek schools as foreign languages hold in
European or American schools’.11 Indeed, all school subjects were taught in Greek,
and no Turkish was taught, until 1895, when the Turkish government issued special
regulations for the teaching of the Turkish language in non-Muslim schools.12
However, an article in the refugee newspaper Prosfygikos Kosmos (‘Refugee
World’) drew the attention of its readers to the fact that:
The great majority of the inhabitants of Asia Minor, particularly in its interior, were
Turkish-speaking. However, the education movement, which began in the last
quarter of the nineteenth century, and promoted from that time, had contributed to
the learning of the Greek language by younger generations. Even so, the Turkish
language was still used in conversation.13
6
Hertz (1945) 142.
7
Wilson (1905) 70.
8
Lapidus (1990) 599.
9
ELIA, File 61: 10.
10
Tsoukalas (1987) 395.
11
Dieterich (1918) 15.
12
Kazamias (1966) 95.
13
Prosfygikos Kosmos, 5 February 1933: 5.
50 george kritikos
In the event, the dissemination of the Greek language was directly related to the
urbanization process of the Ottoman Empire.14 In 1907, ‘the more we move away
from the city of Smyrna, the more the number of Turkish-speaking villages increases
sharply’.15 Myrsini Kapsali, a refugee from Balukeser (a town 148 kilometres from
Smyrna),16 recalls: ‘The school was in the courtyard of the church. We learned
Greek, but we spoke it nowhere. We always spoke in Turkish; at home and in the
street’.17 On the other hand, Theodoros Loukidis, who came from a village (Coban
Isa)18 only 48 kilometres from Smyrna, argued that ‘he learned a little Turkish at
the primary school of the village, where they happened to have a Turkish-educated
teacher’.19 In many ways, the flourishing and highly urbanized commercial centre
of Smyrna gave the impression of a Greek cultural capital, whereas the periphery
was full of local idioms, accents and Greek dialects ‘native to Asia, or . . . at least
. . . there since pre-Turkish times’.20
It can therefore be argued that the enrolment rates of those undergoing a
communal school education in Asia Minor do not represent the real number of
Greek-speakers. These data do not really uphold the view sometimes proposed
by statistically minded historians (Georgios Chassiotis,21 Christos Soldatos,22
among others) that these schools had far-reaching effects on the rate of linguistic
Hellenization.
Any attempt to determine the linguistic composition of the different waves of
refugees and their impact on Greece must acknowledge various complications. In
the first place, it is difficult to categorize the refugees on the basis of their exact
origins (towns, villages, and so on) or to ascertain the level of dissemination of
the Greek language among them. A significant source during the period under
survey is the 1928 population census. This presented Greece as a country with the
following linguistic and religious minorities:
14
Tsoukalas (1987) 393–7.
15
Antonopoulos (1907) 160, quoted in Anagnostopoulou (1977) 195.
16
According to data from the Asia Minor Studies Centre, this town was 148 km north-east of Izmir
and had a population of 36,000 (2,000 Greeks, 2,000 Armenians, and the rest Turks) prior to 1922.
17
Apostolopoulos (1980) 261.
18
An agricultural village with a population of 2,000 (all Greeks), 45 km north-east of Izmir.
19
Apostolopoulos (1980) 128.
20
Dawkins (1916) 5.
21
Chassiotis (1881).
22
Soldatos (1989).
Table 1
Linguistic and religious minorities in Greece in 192823
Minority Language Religion Number %
1. Turks Turkish Moslem 86,506 1.39
2. Slavo-Macedonians Slavo-Macedonian Orthodox 81,844 1.32
3. Chams Albanian Moslem 18,598 0.30
4. Sephardic Jews Ladino Jewish 63,000 1.02
5. Armenians Armenian Orthodox 31,038 0.50
6. Koutsovlachs Koutsovlach Orthodox 19,679 0.32
7. Pomaks Bulgarian Moslem 16,755 0.27
8. Greek Catholics Greek Catholic 27,747 0.45
9. Greek Jews Greek Jewish 9,090 0.15
10. Others * Miscellaneous 30,685 0.49
TOTAL MINORITIES 384,942 6.20
Orthodox Greeks§ 5,819,742 93.80
TOTAL POPULATION 6,204,684 100.00
* Includes all other groups defined by some combination of religion and language,
none exceeding 4,000.
§
Includes 103,642 Turkish-speaking refugees.
The census data indicate that there were 384,942 Greek citizens who made up the
percentage of linguistic minorities (6.2%) in the total population. However, this
figure includes minorities defined by religion rather than language, such as 9,090
Greek-speaking Jews and 27,747 Greek-speaking Catholics. Subtracting their
number from the total number of 384,942, one would conclude that there were
in fact 348,105 Greek citizens among the linguistic minorities of the Greek state.
Yet these data also point to the presence of 103,642 Turkish-speaking Christian
refugees who were exchanged under the criterion of religion. It appears, then, that
the number of linguistic-minority citizens in Greece was 451,747.
It is worth noting that among the Turkish-speaking refugees was a group of
Greek-Orthodox Turkish-speakers, about 50,000 in number, called Karamanlis.24
What was striking about them was that, although they spoke Turkish and knew no
Greek, they used the Greek alphabet to write Turkish. Finally uprooted in 1923,
they were scattered throughout the settlements of Athens and Piraeus.25 They
often encountered a considerable degree of prejudice on the part of the native
population. Nikolaos Markoglou spoke with bitterness about their experiences:
The citizens of Old Greece, the royalists, did not treat us well. They disliked us
because we were Venizelists. They called us Tourkosporous [‘Turkish spawn’] and
23
Elaboration of data from the census of 1928, in Mavrogordatos (1983) 227.
24
Great Britain, Parliamentary Papers (1923) 208.
25
Prosfygiki Phoni, 9 May 1926: 1.
52 george kritikos
laughed at us. As Turkish-speakers we found it hard to speak in Greek, and it did

not occur to them to think about Menidi or other villages in Attica where, a few steps
away from Athens, they still speak Arvanitika [‘dialect Albanian’].26
To compensate for the alienation of these Turkish-speaking Christian

populations, the daily Greek newspaper Panprosfygiki (‘United Refugee’) was
published in Turkish with Greek characters (karamanlidika), at least for its first
193 editions. In 1926, the newspaper changed its title to Prosfygiki Phoni (‘Refugee
Voice’) with only its last two pages translated into karamanlidika under the Turkish
title Muchadjir Dedasi (‘Refugee Voice’).27
One may well suppose that all these Turkish-speaking refugees, along with
the other foreign-speaking refugees and natives included in the 1928 census,
exacerbated the problem of linguistic unification in Greece, as did a further group
of those speaking Greek dialects.
As we have pointed out, there are inadequate data regarding the exact origins of
refugees. Accordingly, their linguistic composition can be inferred only on the basis
of the recorded categories of generalized geographical origin (Asia Minor, Thrace,
Pontus, and so on). According to the census, almost 627,000 people came from
Asia Minor. It can safely be assumed that they were not all from the urban areas of
Smyrna with its standardized Greek. The second biggest refugee group (256,635
people) came from Eastern Thrace. According to League of Nations officials, they
were all agricultural workers.28 Bearing in mind that the percentage of the Anatolian
populations who spoke Greek was related to levels of urbanization, one can infer
that the majority of this rural population had limited Greek or none at all.
The third biggest refugee group was that from Pontus. According to the census
almost 240,695 Pontic refugees, from Russia (11,435), from Anatolia (182,169)
and from the Caucasus (47,091), were established in Greece.29 Scholars who have
analysed the Pontic dialect30 have traced it back to the fourth century BC, as
an independent branch of the Hellenistic koine.31 From this perspective, spoken
Pontic Greek was far removed not only from the ‘purist’ language of the Greek
schools, but also from the spoken language of the indigenous population of Greece.
Oral histories show how proudly Pontic speakers preserved their dialect. In 1959,
a Pontic refugee from Gole in Argyroupolis, Pantelis Vassiliadis, recalled that in his
settlement in the Ptolemaida suburb:
Turks and Slav-speakers were living here. The Turks left but the Slavs remained. . . .
Our language has not changed at all, because we were all speaking the same language
26
Mourelos (1982) 43–4.
27
Prosfygiki Phoni, 9 May 1926: 1
28
LNA, R 1761, 48/24722/24337, report by Dr. Nansen on the refugee situation in Greece, part
2, 28 November 1922: 9.
29
30
Mackridge (1991) 335–9.
31
Dieterich (1918) 31.
in Pontus. Our children, however, speak in Greek, but they certainly know the Pontus
dialect because they speak it with us.32
If we add the numbers of those speaking a foreign language to the refugees who
spoke a Greek dialect, we must conclude that almost one million of the Greek
population had a command of a language or dialect almost unintelligible to the
natives and, of course, far removed from the archaizing standards of official ‘purist’
Greek. There is no doubt that foreign languages or dialects were traditionally
considered a serious challenge to the homogeneity favoured by the state in linguistic
and cultural terms. In the event, the different readings of, and the proposed solutions
to, the problem of linguistic unification represented two conflicting ideological
traits in Greek society.
On the one hand, the supporters of demotic Greek perceived the vernacular as a
powerful mechanism for integrating peripheral or centrifugal powers that threatened
the national cohesion of the state. In 1912–13, the Balkan wars had increased
the territory of the Greek state by 68% and its population from approximately
2,700,000 to 4,400,000.33 The expansion of the Greek state led to the inclusion of
numerous linguistic minorities within its borders, such as Slavs, Albanians, Spanish-
speaking Jews and Vlachs; these populations had lived in situ, since the days of
the Ottoman, or even the Byzantine, Empire; along with them many speakers of
local dialects (Chiots, Epirots, Cretans, among others) were incorporated too. In
1915, the ‘National Language Society’ published a pamphlet that included an
article of Manolis Triantafyllidis under the title ‘Our language in the schools of
Macedonia’. The society made their view clear: the big weapon of the nation was
the school; it was essential to teach in demotic, particularly in these northern areas
where ‘Bulgarian-speaking or Vlach-speaking Greeks [ . . . ] are lost to the Greek
nation, since the school is unable to turn them into Greek speakers, because it
teaches them a “dead” language’.34 On similar lines, the demoticist Triantafyllidis
himself maintained that only by teaching in demotic could linguistic integration or
assimilation be implemented ‘in the schools of Macedonia or anywhere that foreign
speakers exist’.35
Thus, in 1923, the reconstitution of the uprooted communities took place in
what was already a linguistic amalgam. The Asia Minor refugees brought their
dialects into contact with a variety of Greek vernaculars and indigenous languages:
Turkish, Greek, Judaeo-Spanish, various Slavonic dialects, Albanian, and the rest.36
The problem of linguistic amalgamation was exacerbated above all in the northern
provinces of the country, where settlement of 750,000 refugees had taken place.37
32
Centre of Asia Minor Studies, Oral Evidence: File, Attica (from Aryroupolis, Goli); Researcher,
A. Ioakimidis, 14 June 1959.
33
Dakin (1966) 472.
34
Triantafyllidis (1915) 32–3, quoted from Fragoudaki (1977) 68.
35
Triantafyllidis (1915), quoted in Tziovas (1994) 111.
36
Mackridge and Yiannakis (1997) 5.
37
League of Nations, Plenary Meetings 1923–6, 21 September 1926: 3.
54 george kritikos
The 1928 census recorded that almost half the inhabitants of Macedonia (45.18%)
were of refugee origin.38 In these northern areas there also remained almost 82,000
Bulgarians,39 the Muslims of Thrace who had been exempted from the otherwise
obligatory exchange, along with Spanish-speaking Jews and Vlach or Albanian-
speaking populations. These disparate groups made up a new map of linguistic
confusion. In Parliament, in 1924, it was noted that:
Even since the withdrawal of the Turks from Macedonia, the ethnic character of the
population has not been completely changed. During the process of re-establishment
of refugees, it has happened that Turkish-speaking refugees are settled in villages
with Bulgarian-speaking populations; therefore, Greek residents of the Kingdom of
Greece have been learning two foreign languages [it was made explicit that these two
were katharevousa Greek and Bulgarian], without knowing [demotic] Greek, while
Cyrcasians and Armenians have been settled at the border.40
The point was firmly made that, in order not to condemn non-Greek-speaking
refugees to double illiteracy by teaching them two forms of Greek, katharevousa
(purist Greek) should not be taught in the foreign-speaking villages, ‘since it is a
type of language that they do not hear on a daily basis’.41 Even the refugee press
declared that ‘the demotic language is considered to be the unique vehicle of
national expression’.42
On the other hand, the settlement of refugees in the northern Greek provinces
terrified not only the supporters of demoticism, but also the supporters of purist
Greek. The latter supported katharevousa as the touchstone of patriotism in the
hope that the state would move to prohibit the corruption of its purist official
language and the degradation of its traditions. They stated emphatically that ‘after
the Asia Minor disaster there is an urgent need to return to our ancestral and family
values’.43 They argued that the only way the dangers for Greek culture and society
created by the influx of the Asia Minor refugees could be overcome was by means
of an educational model imbued with the values of the nation and its classical
orientation.44 The same conservative opposition also declared that those who were
identified with demoticist ideals were to be held responsible for the destruction of
the link between the Greek nation and ancient Hellenism.45 These conservative
voices proclaimed that Greece was in danger through the abolition of katharevousa
as a school language: the result would be that almost half a million pupils would be
subjected to ‘pseudo-demoticism’, which was equivalent to communism.46
38
39
Pallis’ Archive, File 10, Pallis (1949) 8.
40
Gazette of Debates, 28 June 1924: 520–1.
41
Ibid.
42
Prosfygikos Kosmos, 230, 17 October 1933: 3.
43
Gazette of Debates, 7 December 1927: 252, 260.
44
Ibid.
45
Ibid. 21 March 1927: 3, 11.
46
Ibid. 7 December 1927: 205.
In 1926, the officials of Pangalos’ dictatorship came down heavily on the side
of the conservative position. They encouraged ‘purist’ education with the aim
of ‘combating the enemies and corrupters of religion, language, family, property,
morality, national consciousness and fatherland’.47 The dictator himself, during his
time, boasted that all artistic and intellectual pursuits were promoted, in the hope
that the country would ‘once again give birth to new [incarnations of] Pheidias,
Praxiteles, Aeschylus and other demi-gods’.48 His ideas, indeed, still figured
prominently in the agenda of parliamentary debate even after the dictatorship had
ended. In this context, the anti-Venizelist spokesman Georgios Athanasiadis-Novas
made reference to Plato and Fichte and claimed that ‘the renaissance of national
education should be the basis and the keystone of every reform proposal’.49
For all the high-sounding declarations in their education bills, the 1929
reform fell short of many demoticists’ high expectations. While the education
reform designated demotic as the language of instruction, it was only for primary
school grades. Katharevousa remained the language of instruction for secondary
school and, in the top two classes of primary school, it was taught in parallel with
demotic. In essence, katharevousa remained, in Eric Hobsbawm’s words, ‘the
official or culture-language of rulers and elite’ as well as ‘the official language of
the Greek state via public secondary or higher education and other administrative
mechanisms’.50
One can understand that in a parliament which often equated demoticism
with communism or vulgarism, any attempt to implement radical linguistic reform
would have met with strong opposition. However, language reforms should be
interpreted not only in a national but also in a socio-economic context. Let us turn
our attention to the way the teaching of demotic in primary education served the
cause of integration, on the socio-economic level.
Socio-economic integration
During the 1920s, Greek economy and society were subject to two overwhelming
pressures. In the first place, the state had to integrate hundreds of thousands of
destitute refugees who had fled to the country after the end of the Asia Minor war.
Economic development called for a cheap, skilled work force, both in industry and
agriculture. From 1923 until 1928, Greek industry employed 114,512 refugees out
of 429,831 workers, and 83.8% of its new workers were refugees. At the same time
the sharp increase in the population of the urban centres offered many employers
the opportunity to reduce the daily wage by 20% in 1923 and by 50% in 1927.51
The rural colonization of Greece was set as the main priority of the Refugee
Settlement Commission. Implicit in the policy was a change in agricultural
47
Dimaras (1974) 143–4.
48
Pangalos’ Archive (1971) 167.
49
50
Hobsbawm (1990) 62.
51
Kritikos (2000) 201, 203
56 george kritikos
production methods along with the re-education of agricultural workers.52 As the

Commission’s report of 1930 noted, a switch ‘to intensive cultivation’ was called
for,53 and in turn ‘it has been necessary to re-educate the refugees, amongst whom
even experienced farm-workers were not prepared for this system of exploitation;
the cultivator had to be converted into a specialist’.54 In response to this demand for
skilled labour, the development of vocational and rural training in secondary public
education was cited as one of the goals of educational reform by the Minister of
Education in 1929.55
Then again, the Greek state had to cope with thousands of unemployed native
graduates of the secondary education system. The problem and the paradox was
that the Greek state had allowed the proliferation of secondary schools out of all
proportion to their primary equivalents. By the end of the nineteenth century,
indeed, Greece scored high for enrolment rates in secondary education even by
comparison with western European countries,56 perhaps because, unlike their
western counterparts, public high schools in Greece were entirely free. In 1855, the
total number of male secondary students for the whole Greek population had been
less than 5,000,57 which was the same rate of enrolment as for the vast Russian state
but only a quarter of the equivalent rate for France (where the total of enrolments
was twenty times higher than in Greece).58 At that time, too, the enrolment of girls
was far behind the male enrolment level.59
By 1890 the situation had changed. The following figures illustrate the
proportion of Greek students in secondary education compared with various other
nations:
Table 2
Students in the higher grades of secondary education, 1890 (per 1,000 residents)60
Greece 27
France 26
Belgium 25
Italy 22
Romania 15
USA 11
52
Kritikos (2005) 321–46.
53
Statistical Annual of Greece (1931) 89, 91.
54
League of Nations, Official Journal (1930) 1476–7.
55
E. Venizelos’ Archive, File 115.
56
Mulhall (1892) 231; Webb (1911) 219, quoted from Tsoukalas (1987) 396.
57
Moraitis (1953) 290.
58
Hobsbawm (1962) 166–8.
59
Chassiotis (1881) 496–526.
60
Bickford-Smith (1893) 179, quoted from Tsoukalas (1987) 398.
As the table suggests, the Greek secondary system was extensive. Konstantinos
Tsoukalas argues that education in nineteenth-century Greece was the key
to upward social mobility for the rural populations and produced ‘a large-scale
movement from the small peasantry to the urban bourgeoisie’.61 This open
secondary system and the mobility it produced masked the conflict between city
and countryside, as well as forging organic bonds between the new small-holders
and the old, who could afford a conscious investment in education.62
In contrast to the secondary system, primary education was not fully
institutionalized until the end of the nineteenth century. This imbalance coexisted
with a situation in which katharevousa was the official language of the Greek
state, and secondary education was based on ‘purist’ and classicizing ideals. The
education system did, in a special sense, facilitate the social advance of some of
the lower classes. They graduated from the secondary system and thus felt that
they were socialized with or incorporated into the upper classes. Their knowledge
of katharevousa gave them access to the language of the church and the press as
well as to the administrative mechanisms of the state. At the same time, this newly
educated elite served to reinforce the social hierarchy of Greek society, as against
the overwhelming majority of the population which was illiterate. The illiteracy
figures are set out in Table 3:
Table 3
Percentage of illiterate population in the Greek state63
1870 82.2%
1879 80.67%
1907 66.27%
1920 58.22%
1928 50.2%
However, notwithstanding the prestige given by a classical education or a com-

mand of katharevousa, it is undeniable that this classical-oriented system produced
school leavers with none of the skills necessary for their employment as technical
personnel. Speeches in parliament also make it apparent that there was, in any
case, growing pressure for a reduction in the number of those lacking practical
skills. In 1929, the Greek MP (and ex-Prime Minister) Alexandros Papanastasiou
denounced the existing education system on the grounds that it created a mass of
job-chasers who had graduated from high school and now sought appointment to
the public services.64 In the same context, the MP Georgios Kokkinakis suggested
that ‘the Minister of Education should be aware that with the classical high schools
61
Tsoukalas (1987) 381.
62
Mavrogordatos (1983) 170.
63
64
Gazette of Debates, 4–9 July 1929: 17.
58 george kritikos
as they are, we do not have anything else but position-seekers’,65 and argued that
‘many of these high schools must to be turned into vocational or technical training
schools’.66 The Minister of Education (Theodoros Nikoloudis) replied that he was
in complete agreement on this point.
As late as 1933, the MP Alexandros Michalopoulos warned that the mass of job-
seekers and unemployed school-leavers represented a source of political instability.
He pointed to certain large villages in his constituency, where there were:
129 school-leavers who were not appointed to any position and thus considered
themselves unfairly dealt with by society, without, however, deigning to help their
father in his agricultural labour. They spend their day in the coffee-houses reading
and ‘ill-digesting’ books, and in the end they become communists, not because of
ideology – which ideology would want half-educated people anyway? – but because
of disappointment about their failure to find a job. We have created miserable people,
unhappy parents, miserable young people, who have nothing else to do but spread
unconstructive and anarchic germs.67
The official attitude of the Greek government was expressed by Konstantinos

B. Gontikas who introduced most of the education reform bills of 1929. In the
preamble to the legislation, the Minister acknowledged that ‘our education system
has neglected the education of the masses and has turned our culture into an
oligarchy. It has also given our patterns of learning an obvious bias in favour of the
theoretical, to a neglect of practical education, and to the detriment of the wealth-
producing sectors.’68 Before 1920 there had been only one agricultural school in
Greece, whereas by 1930 there were a hundred agricultural, technical and public
or private commercial schools with almost 10,000 students.69
At the same time, many politicians also favoured limiting access to secondary
education. They proposed stringent examinations as a condition of access to high
schools. They noted that ‘these examinations will make primary school pupils
understand that high school is not a game, and is not for everyone who decides
to go on to secondary education’.70 Before the elections of 1928, the newspaper
Eleftheron Vima published a speech of Eleftherios Venizelos, in which he stressed
that his future government would:
support the social regime more effectively through a direct and radical reform of the
education system. As long as hundreds of schools release thousands of young leavers
every year who are deficient in education and incapable of any productive work, our
future cannot be anything else but dark, and the state is obviously exerting itself to
prepare a future army of social subversion. I am a supporter of classical education only
65
Ibid. 4 December 1928: 701.
66
Ibid.
67
68
General preamble and legislative decrees submitted by the Minister of Education (K.B. Gondikas)
to Parliament during the session of 2 April 1929: 2.
69
70
Ibid.
for a small percentage of the young people at school who are going to constitute the
leaders of tomorrow.71
In Venizelos’ view, this large category included ‘poor but excellent students who
will be given state scholarships to continue their studies’.72
The enactment of a law limiting access to secondary education showed that a
meritocratic spirit was likewise shared by the Ministry of Education.73 The reform
reduced the number of high schools and their leavers. In the event, the Greek
government closed 295 secondary schools and reduced the number of their students
by 41% by 1932: in 1928 there were 96,214 students in these schools, whereas
their number went down to 57,225 in 1932.74 In the late 1920s, correspondingly,
the tally of those completing higher education presents a gradual decline, as Table
4 shows:
Table 4
School leavers from secondary schools75
1927–28 18,578
1928–29 15,921
1929–30 10,070
With immigrants and refugees, however, education and occupation are crucial
factors in social advancement and thus in the eradication of social inequality. One
must bear in mind that:
populations who migrate have on the whole higher educational and occupational
aspirations than indigenous groups, both when they are in a majority as well as a
minority, and are more determined to use education as a strategy for upward social
mobility than non-immigrants of comparable class background. Immigrant parents
and children assume that education can enhance opportunities to compete for jobs.76
The relation of cause and effect between schooling policy and the integration of
foreigners seeking to adapt to a society is exemplified by the evidence of asylum-
seekers and refugees throughout history. First-generation immigrants have very
specific needs and, in particular, must be able to compete in the labour market.
Knowledge of the language of the host society raises their self-esteem and facilitates
their integration. In France, when immigrants came from former colonies, language
courses were necessary to enable them to participate in vocational training.77 In
Sweden, a comprehensive programme of occupational training is required for
71
Eleftheron Vima, 22 July 1928: 1.
72
Ibid.
73
Sifnaios (1929) 1826.
74
Petridis (1998) 298.
75
76
Gibson (1988) 174.
77
Verbunt (1985) 151.
60 george kritikos
first-generation immigrants with their instruction even more specialized than any
provided for their children.78 In the event, the various refugee groups attempt to
ensure that their children acquire an education that will ensure their success in the
new society that has given them shelter.79 In Great Britain, on some indices, ethnic
minorities are more likely than their white counterparts to stay in education after
the school-leaving age of 16.80
The Greek Orthodox refugees of the 1920s likewise perceived education as a
key to their own integration and to upward mobility for their children. This is
shown by the reports of Refugee Settlement Commission officials, who noted that
‘the refugees did not want to settle in places where there was no teacher or priest’.81
And indeed, no sooner had the refugees arrived in Greece than they petitioned for
the establishment of churches and schools in their settlements, on the grounds that
schools and churches were ‘the best way to bring up their children’.82
It was also the view of the Greek Minister of Education, Georgios Chatzikyriakou,
that schooling could function as an important mechanism of assimilation. As he
argued in 1924, education could improve the ability of refugees to adapt themselves
to their new environment by improving their social as well as their cognitive skills,
so that the effective application of the education system, along with the juridical
and administrative systems, was a prerequisite for any solution to the refugee
problem.83 He stressed that ‘the refugee question is closely related to the issue of
education. Along with housing and feeding them, we have to provide for their [the
refugees’] education.’ Without education, he insisted, and without a programme
designed to help them ‘become Greek’, any prosperity they might achieve would
‘get them nowhere’.84
However, the children of refugee families in 1920s Greece were often seen as
disadvantaged. The school inspectors of northern Greece described ‘1,500 refugee
children’ as ‘roaming through the streets of Cavalla’, while the pupils who attended
regularly were ‘stacked into overcrowded classrooms’.85 In 1924, the Minister of
Education denounced existing policy on the grounds that it left the refugee children
in the streets, doomed to ignorance and countless moral dangers. For the same
reasons, the MP Alexandros Pappas declared that it would be ideal to establish
kindergartens everywhere, but mainly in places occupied by the labouring classes
and in refugee settlements, ‘where the children are out on the streets and need a
shelter while their mothers are at work’.86 On the very eve of the education reform
of 1929, he claimed that, though compulsory education had been decreed by the
78
Hammar (1985) 38.
79
Jones and Rutter (1998) 3.
80
Jones (1996) 33.
81
Prosfygiki Phoni, 238, 9 May 1926: 1.
82
I.A.Y.E., 15 January 1925, Protocol No. 3926: 4.
83
Gazette of Debates, 3 July 1924: 637.
84
Ibid. 8 April 1924: 940.
85
Ibid. 941.
86
Ibid. 2 July 1924: 13
legislation of 1834 and this provision had been in force ever since, ‘according to the
statistics’, 200,000 children were ‘outside the education system’.87
Arguably, though, the nature of the secondary education on offer remained a
more serious problem. Like many native Greeks who had no family care, or who
could not afford to continue their studies in the classical secondary school, the
majority of the refugee children were to drop out of school and be used as a cheap,
skilled labour force. For those children who had the luxury of parental care, the
increased attention to secondary technical or agricultural education did indeed
meet the need for a skilled work force, and (as was argued in the Greek parliament)
also showed that the policy-makers were not interested in making high or classical
culture widely available ‘for those children whose parents did not want to send
them to school’.88
For those refugee women who were employed in the urban (and gender-biased)
labour market without any knowledge of Greek, a solution to the problem of
communication was provided by domestic carpet-weaving or other household
occupations, or even by employment in factories, where a high level of literacy was
not a job prerequisite. 89 As the experience of other European countries has shown,
‘in the early stages of factory production literacy is not necessary’.90
In the event, this system of employment proved to be very effective in reproduc-
ing existing inequalities in the labour market. The data of the Annual Statistics of
Greece demonstrate that there was no real social mobility, in terms of occupational
change, throughout this period. Despite the fact that almost 1,300,000 refugees
were integrated into the social and financial fabric of the nation, the division of
labour remained the same as it had been before the influx of the Asia Minor refu-
gees. The number of non-manual, higher-status jobs was not growing. In the short
term, the division of labour, and accordingly the high status of the relatively few
non-manual jobs, was unchanged.91
At the same time, the use of ‘purist’ Greek as the language of instruction pre-
cluded access to secondary school and closed the prospects of medium-term social
levelling through education. A basic level of literacy and technical competence, in
Gellner’s terms, simply allowed the refugees to be properly employed in the labour
market, if necessary by following the manuals and instructions of a new activity or
occupation.92
On the other side, classical education remained the predominant element in the
secondary school curriculum. In 1931, the total length of time assigned weekly in
the secondary schools to teaching classical Greek was over twice as much as the
time allotted to Modern Greek, which covered both demotic and katharevousa.93
87
Ibid. 2 July 1929: 19.
88
Ibid. 9 July 1929: 91.
89
Kritikos (2000) 189–206.
90
Pollard (1968) 180.
91
92
Gellner (1983) 35.
93
Psomas (1977) 194.
62 george kritikos
In the end, the education system had not changed from the model that the MP
Nikolaos Konstantopoulos once denounced as still centred on an ideal of ancient
Greece, pointing out that ‘for a hundred years’ the Greek nation had been ‘staring
at ancestral shadows’.94
Conclusions
Literacy in vernacular Greek certainly offered a means of integration and
communication between native and refugee citizens of the Greek state. As we have
seen, the linguistic integration of its northern provinces was exacerbated after the
settlement of refugees who spoke different languages or dialects. In this context,
teaching children in the purist language (katharevousa) implied a double illiteracy
for them. Without any common medium of communication they were left to reject
a demanding archaic language, which was not spoken at home, as well as to come
to terms with the various languages that were the means of socialization in the
northern provinces, where the majority of the inhabitants belonged to a linguistic
minority. The policy-makers and the Ministry of Education realized that other
tools of school teaching were needed to integrate the foreign-language students.
Within this framework, schooling in the vernacular acquired a two-fold purpose:
it was meant not only to minimize alleged foreign propaganda and to allay the
anxieties of the indigenous population, but also to strengthen national sentiment
among the foreign-language refugee and native populations. Teaching vernacular
Greek was perceived as a tool of cultural homogeneity and national integration that
would not permit the conflicting linguistic output of local native units, in Gellner’s
terms,95 to act as centrifugal forces and to bring about the decomposition of the
nation-state after the settlement of the refugees.
In the event, the settlement of refugees played a crucial role in the
institutionalization of demotic at primary school. This aspect of the reform has
not hitherto been examined. The refugee settlement nourished fears of national
disintegration and contributed decisively to the defeat of katharevousa in its battle
with the non-standard vernacular, as a medium of instruction at primary school.
At the same time, the continuation of katharevousa as the language of the state,
as well as of secondary and higher instruction, fulfilled the definition of education
as something which ‘creates discontinuities’96 between the literate and illiterate. It
replicated the traditional social roles and structures of power in Greek society, even
while the teaching of demotic at primary school provided a measure of integration
within the social and economic fabric of the state.
The teaching of vernacular Greek at primary school level only, and the highly
selective education system that was instituted after the 1929 reform, along with the
practical reorientation of the secondary education system, all functioned to limit
upward mobility. Instead, these factors promoted the idea of the ‘integration of
94
Gazette of Debates, 4–9 July 1929: 83.
95
Gellner (1983) 38.
96
Goody (1968) 58.
the working-class child into society: those who are “bright” are helped to prepare
their escape from their working-class condition: the rest are helped to accept their
subordination’.97 Along with the elite that wanted to preserve their racial kinship
with ancient and Christian Greece, these ‘bright’ children were to defend classicism
and the ‘purist’ language, and none more so than ‘those of a humble background
who derived their prestige and hard-won status from the use of katharevousa’.98
Within this framework, primary education in demotic only reproduced social
inequalities at the level of literacy and occupation. In a country that pursued in-
dustrialization as well as the integration of the unemployed in its labour market,
education in the non-standard vernacular at primary school implied a functional
literacy. It provided the necessary basis for the integration of linguistic minorities
and speakers of peripheral dialects into the national centre, as also for the training
of workers. It was not used as the great equalizer of men.
References
Unpublished Primary Sources
Centre of Asia Minor Studies, Oral Evidence: File, Attica (from Aryroupolis, Goli);
Rsearcher, A. Ioakimidis, 14 June 1959.
ELIA (Greek Literary and Historical Archive), File 61:10.
I.A.Y.E (Istoriko Archio Ypourgiou Exoterikon) [Historical Archive of Greek Foreign
Office], File: G/68, XX, 1925, Memorandum of the Common Trusteeship of the
Displaced Greeks, 15 January 1925, Protocol No. 3926: 4.
LNA (League of Nations Archives), R 1761, 48/24722/24337, report by Dr. Nansen on the
refugee situation in Greece, part 2, 28 November 1922: 9.
Pallis’ Archive (in National Centre of Research, Athens), File 10.
E. Venizelos’ Archive (in Benaki Museum, Athens), File 115: brief account of the industrial
policy that must be implemented by the Greek State [Law: 1948].
Published Primary Sources

Efimeris ton Syzitiseon [Gazette of Debates]: 1924 – 1933, Library of the Greek Parliament
in Athens.
Eleftheron Vima, Athens, 22 July 1928: 1.
General preamble and legislative decrees submitted by the Minister of Education (K.B.
Gondikas) to Parliament during the Session of 2 April 1929: 2.
Great Britain, Parliamentary Papers: 1923, Turkey, No.1, ‘Lausanne Conference on Near
East Affairs, 1922-23’, Proceedings, London.
League of Nations, Official Journal, November 1930, Official No C.444.M.202.1930, II.
Annex: 1235, [Twenty-seventh quarterly report of the Refugee Settlement Commission,
Athens, 21 August 1930]: 1476–7.
97
Miliband (1968) 243.
98
Mavrogordatos (1983) 170.
64 george kritikos
League of Nations, Records of Plenary Meetings 1923–6, Verbatim Record of the Seventh
Ordinary Session of the Assembly of the League of Nations, Held on Tuesday 21
September 1926: Work Report of the Second Committee on the Work of the Greek
Refugee Settlement Commission (Chairman of the RSC, Charles P. Howland): 3.
Pangalos’ Archive (1971), vol. 2, Athens: 167.
Prosfygiki Phoni [Refugee Voice], 238, Athens, 9 May 1926: 1.
Prosfygikos Kosmos [Refugee World], 203, Athens, 5 February 1933.
Prosfygikos Kosmos [Refugee World], 230, Athens, 17 October 1933: 3.
Sifnaios, C., Pandektis Neon Nomon kai Nomoschedion [Digest of New Laws and Bills], vol. 4,
1929, Law No. 4373, 13/20 August 1929. General Preamble of the Bill Regarding the
Organization of Secondary Schools (Article 3).
Statistical Annual of Greece, Athens, 1931.
Statistical Annual of Greece, Athens, 1938.
Secondary Literature
Anagnostopoulou, S. (1977), Mikra Asia, 19os ai.–1919, Oi Ellinorthodoxes Koinotites, Apo to
Millet ton Romion sto Elliniko Ethnos [Asia Minor, Nineteenth Century–1919, The Greek-
Orthodox Communities: From the Millet of Romios to the Greek Nation], Athens.
Antonopoulos, S. (1907), I Mikra Asia [Asia Minor], Athens.
Apostolopoulos, F. D. (ed.) (1980), I Exodos: Martyries apo tis Eparchies ton Dytikon Paralion
tis Mikras Asias [Exodus: Evidence from the Provinces of the West Coast of Asia Minor], vol.
1, Athens.
Bickford-Smith, R.A.M. (1893), Greece Under King George, London.
Chassiotis, G. (1881), L’ instruction publique chez les Grecs depuis la prise de Constantinople
par les Turcs jusqu’à nos jours, Paris.
Dakin, D. (1966), The Greek Struggle in Macedonia (1897–1913), Thessaloniki.
Dawkins, R.M. (1916), Modern Greek in Asia Minor: A Study of the Dialects of Silli,
Cappadocia and Pharasa, Cambridge.
Dieterich, K. (1918), Hellenism in Asia Minor, New York.
Dimaras, A. (ed.) (1974), I metarythmisi pou den egine [The Reformation that Never
Happened], vol. 2, Athens.
Dimaras, A. (2006), ‘Modernisation and reaction in Greek education during the Venizelos
era, in P. Kitromilidis (ed.), Eleftherios Venizelos: The Trials of Statemanship, Edinburgh.
Fragoudaki, A. (1977), Educational Demoticism and the Linguistic Reconciliation of 1911,
Ioannina.
Fragoudaki, A. (1992), Ekpaideftiki metarythmisi kai Fileleftheri Dianoumeni [Education
Reform and Liberal Intellectuals], Athens.
Gellner, E. (1983), Nations and Nationalism, New York.
Gibson, M.A. (1988), Accommodation Without Assimilation: Sikh Immigrants in an American
High School, Ithaca, NY.
Goody, (ed.) (1968), Literacy in Traditional Societies, Cambridge.
Hammar, T. (ed.) (1985), European Immigration Policy: A Comparative Study, Cambridge.
Hertz, F. (1945), Nationality in History and Politics, London.
Hobsbawm E.J. (1962), The Age of Revolution, New York.
Hobsbawm E.J. (1990), Nations and Nationalism since 1780: Programme, Myth, Reality,
Cambridge.
Jones, C. and Rutter, J. (1998), ‘Mapping the field: current issues in refugee education’, in J.
Rutter and C. Jones (eds.), Refugee Education: Mapping the Field, London: 1–12.
Jones, T. (1996), Britain’s Ethnic Minorities, London.
Kazamias, A.M. (1966), Education and Quest for Modernity in Turkey, London.
Kritikos, G. (2000), ‘State policy and urban employment of refugees: the Greek case (1923–
30)’, European Review of History, 7: 189–206.
Kritikos, G. (2005), ‘The agricultural settlement of refugees: a source of productive work
and stability in Greece, 1923–1930’, Journal of Agricultural History, 79: 321–46.
Lapidus, I.M. (1990), A History of Islamic Societies, Cambridge.
Lewis, G. (1965), History of Modern Turkey, London.
Mackridge, P. and Yiannakis, E. (eds.) (1997), Ourselves and Others, The Development of a
Greek Macedonia: Cultural Identity Since 1912, Oxford.
Mackridge, P. (1991), ‘The Pontic dialect: a corrupt version of Ancient Greek?’ in Journal
of Refugee Studies, 4: 335–9.
Mavrogordatos, G.T. (1983), Stillborn Republic: Social Coalitions and Party Strategies in
Greece, 1922–1936, Berkeley.
Miliband, R. (1968), The State in Capitalist Society; An Analysis of the Western System of
Power, London.
Moraitis, D. (1953), Historia tis Paidagogikis [History of Teaching], Athens.
Mourelos, G. (ed.) (1982), I Exodos: Martyries apo tis Eparchies tis Kentrikis kai Notias
Mikras Asias, [Exodus: Evidence from the Provinces of Central and South Asia Minor], vol.
2, Athens.
Mulhall, M.G. (1892), The Dictionary of Statistics, London.
Pallis, A.A. (1949), Macedonia and the Macedonians: A Historical Study, London.
Petridis, P. (ed.) (1998), To Ergo tis Kyverniseos Venizelou kata tin tetraetia 1928–1932. Ti
ypeschethi proeklogikos kai ti epragmatopiise [The Work of Venizelos’ Governments During
the Period 1928–1932: What He Promised and What He Delivered], Athens.
Pollard, S. (1968), The Genesis of Modern Management, London.
Psomas, A.I. (1977), The Nation, the State and the International System: The Case of Modern
Greece, Athens.
Soldatos, Ch.P. (1989), I Ekpaideftiki kai Pnevmatiki Kinisi tou Ellinismou tis Mikras Asias,
1820–1922, [The Educational and Intellectual Movement of Asia Minor Hellenism, 1820–
1922], vol. 3, Athens.
Triantafyllidis, M. (1915), ‘I glossa mas sta scholeia tis Makedonias’, [‘Our language in the
schools of Macedonia’], Deltion Ekpedeftikou Omilou [Educational Society Bulletin], 5:
32–3.
Tsoukalas, K. (1987), Exartisi kai Anaparagogi – O koinonikos rolos ton ekpaideftikon
mechanismon stin Ellada (1830–1922) [Dependence and Reproduction – The Social Role of
the Educational Mechanisms in Greece (1830–1922)], Athens.
Tziovas, D. (1994), ‘Heteroglossia and the defeat of regionalism in Greece’, Cambos:
Cambridge Papers in Modern Greek, 2: 95–120.
Verbunt, G. (1985), ‘France’, in T. Hammar (ed.), European Immigration Policy: A
Comparative Study, Cambridge: 127–64.
Webb, A.D. (1911), The New Dictionary of Statistics, London.
Wilson, C.W. (ed.) (1905), Murray’s Hand-Book for Travellers in Aia Minor, London.
4
Greek With No Models, History or Standard:

Muslim Pontic Greek*
Pietro Bortone
In memory of Tasos Christidis
Background
In north-eastern Turkey, in the area known to the Greeks as Πόντος, there used
to be a large, high-profile Greek community, which was forcibly expelled en masse
in 1923 in the infamous ‘Exchange of Populations’ between Greece and Turkey.
It is less well known that there are still a few rather isolated villages in the eastern
corner, in the Trabzon area, especially near Of, where the locals speak varieties of a
‘dialect’ that is in fact Greek – akin to the Pontic Greek dialects once spoken there
by Greeks.1 The speakers, of essentially Turkish and Muslim identity, descend in
part from Greeks who converted to Islam over 300 years ago, developed a separate
identity, came to be regarded as Turks, and were thus able to remain in the area to
this day.
Standard Greek amongst Pontian Greeks

Until 1976, the linguistic situation of Greece was the textbook example of diglossia,
with demotic (vernacular ‘Modern’ Greek) as the spoken variety, and katharevousa
(archaizing, puristic Greek) as the normal non-literary written medium. When
Pontian Greeks arrived in Greece in 1923, they found themselves in a country
where their dialect had low status, where linguistic uniformity was seen as essential
for political unity, where refugees coming from Turkey faced prejudice, and where
* I wish to thank Peter Mackridge for encouraging me to work in this area, as well as for comments
on an earlier draft of this article. I also gratefully acknowledge the support given to me for my research,
through scholarships and fellowships, by The Wingate Foundation, The Alexander S. Onassis Public
Benefit Foundation, The Program in Hellenic Studies of Princeton University and The Institute for
the Humanities of the University of Illinois at Chicago.
1
With the exception of Ioannis Parharidis in the 1870s, Peter Mackridge was the first scholar to
research this dialect (see p. 88 below, under References).
67
68 PIETRO BORTONE
the wider community used the variety of Greek that even in Pontus was regarded
as the proper form. The most prominent Pontian scholars soon declared that their
dialect was doomed.2
The Pontic dialect, nevertheless, still survives in some of the areas of Greece to
which it has been transplanted – spoken, according to some estimates, by as many
as 300,000 people.3 Standard Greek, naturally, has exerted a discernible influence
on it – but this had already been happening even when the speakers were still in
Turkey: Greek-identified (that is, Christian Orthodox) Pontians always maintained
very strong cultural ties with Greece, and even had Greek schools, usually under
the control of the Church, from kindergarten to secondary level. The schools were
a focal point for the community, who led an independent life from the surrounding
Muslims and often lived in separate villages.
Greek schools in Pontus taught Standard Greek prescriptively. In some, Standard
Greek was even spoken, and the pupils’ parents were also actively encouraged to
use it at home. Curricula and syllabi were decided in Greece, and the books and
school manuals came from Greece. School books had an explicit nationalistic
outlook, and prominence was given to the classics, to Ancient Greek grammar, to
religious texts and other books written in the archaic style favoured by the church.
The teachers were often Orthodox clerics, many of whom had been trained in
Greece. In the larger cities there were also Greek libraries, and Greek newspapers
circulated. Thus the same literary models, linguistic values and language variety
that were promoted in Greece were promoted among Greek-identified Pontians as
well.
Pontian Greeks educated in the Greek schools of Pontus assumed that the very
function of schools was to teach a language far removed from the vernacular; when
they first moved to Greece, they ‘trouvaient tout à fait bizarre, voire anormal,
qu’une langue que les gens connaissaient déjà puisse être enseignée à l’école’.4
Their surprise is not unjustified: across the world, it is rare for children to go to
school already knowing the language variety that the school wants them to know;5
even where the language used at school is not totally different from the one spoken
by the pupils, the schools’ aim is usually proficiency in the Schriftsprache.
No models, history or standard

Muslim Pontic Greek speakers, on the other hand, did not regard themselves as
in any way Greek. They therefore had no contact with Greeks from Greece, and
no exposure to the language of Greece. To this day they have never seen Modern
Greek literature, have never heard Biblical Greek, have never studied classical
Greek, have never learnt any Standard Greek (not even the Greek alphabet),
2
E.g. Papadopoulos (1953) 84 n. 1, and Lampsidis (1959) 204.
3
Drettas (1999) 15.
4
‘They found it completely bizarre, indeed abnormal, that a language that people knew already
could be taught at school’: Drettas (1998) 82.
5
Spolsky (2004) 46.
GREEK WITH NO MODELS, HISTORY OR STANDARD 69
have not heard Greek radio or TV, nor any form of the Greek language other than
their own – and have not been touched by the strict Greek policies of language
standardization, archaization and purism. In other words, their Greek has had no
external models for centuries. Furthermore, it is not written, printed, or broadcast.
So it has no recorded local tradition and therefore no internal models to refer back
to either.
In Greece, Pontic studies has developed into a recognised field, with respected
journals, known publishers, well-stocked archives and active research institutes.
The pervasive diachronic slant and the heavy emphasis on history which is typical
of traditional Greek scholarship has also been applied to Pontic studies. Greeks
of Pontic origin have been writing extensively about the philological history of
their dialect and about the historical background of their community, highlighting
its classical past and its Greek connections. Muslim Pontic, conversely, has no
history, especially for its speakers: not only do they have no written records, but
many in their speech community do not even know that the language they speak
has anything to do with Greek. Some do not know which parts of what they say
are Turkish and which are their local ‘other language’. Many call that language
lázika or laziká, confusing it with Laz, a quite different and unrelated language
also spoken in northeastern Turkey. Many call it Romayka, but never Pondiaká
(the standard term in Greece, of learned origin), and never Eliniká either.
Romayka is not formally taught anywhere, and no norm for it has been
established, maintained or promoted. Accordingly, it has no standard of any kind:
its characteristics have not been planned or fixed, there is no official variety, no
prescriptive rules, no concept of correctness, no anguish about purity. All this is,
of course, amazing if we consider that Romayka is, after all, a variety of Greek –
one of the languages with the longest and most vexed histories of prescriptivism,
politicization, artificial intervention, linguistic self-consciousness, angst over
correctness, and battles over choice of a standard that have even led to people being
killed in the streets.6 Indeed, Romayka seems to provide the textbook example of a
language in its natural state as dreamt of by sociolinguists and anthropologists:
To see language in its ‘natural’ state, one must find a variety which is neither a standard
language, nor a dialect subordinate to a standard (since these too show pathological
features, notably the difficulty of making judgements in terms of the non-standard
dialect without being influenced by the standard one).7
Greek standards
The Christian, Greek-identified Pontians, being Greek-educated, had firm opinions
on linguistic correctness, purity and standards. Those who used to live in the Of
area occasionally heard the Greek dialect spoken in the nearby Muslim villages
and, given their Greek background, had clear views about that too – although they
6
The εὐαγγελικά riots in Athens (8 November 1901) following the serialization, in the newspaper
Ἀκρόπολις, of A. Pallis’ translation of Matthew’s Gospel into demotic.
7
Hudson (1996) 34, one of the main textbooks in sociolinguistics.
70 PIETRO BORTONE
were the only local Greek-speakers who judged the ‘Muslim’ dialect according
to a standard. Researching historical archives in Athens, in several handwritten
interviews with 1923 Pontian Greek refugees, I found that many reported with
bemusement that there were τουρκικὰ χωριὰ ποὺ µιλούσανε ἑλληνικὰ
καλύτερα ἀπό µας – ‘Turkish villages that spoke Greek better than we did’.8
In recent years, a few Greeks of Pontian origin have travelled to eastern Turkey
on holiday, where they heard Muslim Pontic Greek. They regarded it simply as
another variety of Pontic, and they told me that the western varieties of the Muslim
dialects, which are the ones that most closely resemble ‘Christian’ Pontic, are ‘τα
πιό καθαρά ποντιακά’, the purest Pontic. This shows that they classify varieties
of Pontic as good or bad with implicit reference to an ideal norm – even if that
remains vague, unofficial and probably variable9 – and they do so by taking the
Pontic spoken in Greece (the variety spoken by the overwhelming majority of
Pontic speakers) as their automatic standard. When they assess the Pontic dialects
of Greece, on the other hand, they often label as ‘purest Pontic’ the varieties least
influenced by Standard Greek,10 treating the degree of difference from the national
language as an index of dialectal ‘purity’.
Indeed, now that Greek Pontic has been relocated to Greece, and is used in
parallel with Standard Greek,11 its speakers have come to feel that it needs a standard
form of its own. A fairly standardized orthography, after lengthy discussions in
specialized journals, has been developed: it uses the Greek alphabet as a matter
of course (with diacritics for the sounds that Greek lacks), and restores historical
spelling – for instance, with diphthongs where pronunciation has monophthongs.
Small details, such as the use of apostrophes to mark vowels ‘missing’ in comparison
with the national language (e.g. ἐγροίκ’σα), show how Standard Greek is treated as
a yardstick. A supra-regional variety of Pontic (incorporating elements of Standard
Greek) has been emerging,12 and Greek linguists refer to it as κοινὴ νεοποντιακὴ,
‘Common Modern Pontic’. As the name indicates, it is conceptualized in the same
terms as the standard form of Greek (ἡ κοινὴ νεοελληνική, ‘Common Modern
Greek’), and this is probably because it is the most common written form of Greek
Pontic – even if no form of Pontic is officially standard.13 But the perception that
there must be a consensus variety is not only felt when writing: actors in the
Pontic theatre, even if not always fluent in Pontic, strive to maintain a ‘correct’
pronunciation.14
8
Archives of the Centre for Asia Minor Studies, Athens: interview catalogued as PO129, 29. I
consider what they meant by ‘better’ below, pp. 82–3.
9
Drettas (1997) 21.
10
Mackridge (personal communication).
11
Tombaidis (1996) 230.
12
Drettas (1999) 17.
13
Cf. Drettas (1997) 19.
14
Drettas (1998) 84.
The ‘language-or-dialect’ issue

The wish to emphasize both the distinctiveness of Pontic and its Greek lineage
has been fuelling an intense debate about whether Pontic should be classified as a
dialect or a language. The former view is favoured by Greek scholars, while foreign
(or foreign-based) scholars often support the latter view. Thus, for Andriotis and
Kondosopoulos, Pontic is a dialect,15 while Tombaídis adds bluntly:
για όσους ἑχουν έστω και στοιχειώδεις γνώσεις της ιστορίας της γλώσσας µας
δεν υπάρχει θέµα . . . Οπωσδήποτε, δεν µπορούµε να θεωρήσουµε γλώσσα την
ποντιακή.
For those who have even just an elementary knowledge of the history of our language
there is no question . . . We cannot possibly regard Pontic as a language.16
On the other hand, Dawkins wrote that Pontic is ‘almost’ a language, while
Shirokof described Pontic, Tsakonian and the varieties of Greek spoken in the
Ukraine, Italy and Corsica as ‘self-standing languages’ (самостоятельные языки).
This view is echoed by Drettas, who talks of ‘langue pontique’ and of ‘langues
néo-grecques’.17
Linguists have been pointing out for a long time that there is no scientific
distinction between dialect and language18 – but the distinction is still commonly
made. In Greek, a further complication is the currency of a third term, ἰδίωµα,
that indicates a variety only slightly diverging from the norm – and this may be
another reason why Greeks may designate with the term ‘dialect’ (διάλεκτος)
some less intelligible varieties that in English we might label ‘languages’. One often
comes across articles that argue, with appeals to various criteria, for or against the
classification of a speech variety as a language or as a dialect. Pontic is only one
example. The criteria most commonly invoked are:
(a) quantitative: number of speakers;
(b) political: legal and social status;
(c) functional: suitability for multiple uses, especially depending on whether a written
form exists;
(d) communicative: intelligibility, also related to structural dissimilarity to a related
standard language.
All these criteria are open to debate – clearly so if applied to Pontic:

(a) The fact that Pontic has relatively few speakers does not prove that it is a
dialect, as there are many languages with only a handful of speakers.
(b) It is true that linguistic varieties classified as ‘dialects’ often have no official
status and low social status – but few of the people keen to emphasize that Pontic
is a dialect of Greek would cite as evidence the fact that Pontic lacks legal
15
Andriotis (1995) 100–1; Kondosopoulos (1981) 10.
16
17
Dawkins (1937) 24; Shirokov (1972) 317; Drettas (1997) xxi–ii, 19, and (1998) 75.
18
Cf. e.g. Haugen (1966) 922–7.
72 PIETRO BORTONE
recognition and has lower prestige than Standard Greek. The term ‘dialect’ (like
the term ‘accent’) commonly has belittling undertones, and often suggests that
the speech variety in question is sub-standard, whereas the label of ‘language’ has
connotations of correctness and authority. But status cannot be the deciding factor
anyway, both because dialects can be (and often have been) raised to the rank
of languages – the essence of the process of language extension (Ausbau) – and
because many languages (like Romany in Greece), whose status is indeed poor
socially and non-existent legally, would never be classified as dialects given their
linguistic distance (Abstand) from the national language.
(c) It is also true that, in popular parlance, the term ‘dialects’ normally describes
the speech varieties for which no written form has been developed, either at all or
for literature.19 However, forms of Greek that are recognised as dialects, like Cretan,
have been written and have included influential and widely read literary texts –
and Cretan could easily have become the basis of the modern national standard.
Similarly, what we refer to as the ‘dialects’ of Ancient Greek are known to us
through their rich and largely literary written tradition. Indeed, the more general
idea that there is (or should be) one normative, unifying, standard language – an
assumption that has had vast linguistic, cultural, social and political consequences
in modern Greece – is alien to ancient Greece. In the classical period, no single
standard was recognised, probably because there was no single Greek state; and
different varieties of Greek had similar social status.
Moreover, if ‘dialect’ means ‘unwritten language’, Muslim Pontic would have to
be called a dialect, but Christian Pontic could not, because it has been extensively
written. The writing of Christian Pontic occurred mainly after 1923, when the
Christian Pontians resettled in Greece. In other countries where Greek-identified
speakers of Pontic live, varieties of Pontic have also had a written form. In Russia,
where Greeks, for a time, were recognised as an ethnicity and had their own
schools and press, Pontic and non-standard varieties of Greek started being written
and printed in the late 1920s. This entailed all the usual events associated with the
writing of ‘languages’; there were spelling reforms and lengthy debates about which
form of Greek should be selected as the standard – Pontic being one contender.
A general conference held in Moscow in 1926, during a period when Soviet
authorities encouraged minority language schools and publications, ruled that the
Greeks of Russia would no longer use katharevousa but rather demotic Greek with
a ‘phonetic’ spelling that abandoned historical orthography and the traditional
system of multiple accents and breathings.20 The Greeks of Greece were horrified.
The main Greek publishing house of the USSR, Κοµυνιςτις [sic] (‘Communist’),
which had a very popular newspaper of the same name, even published a Pontic
grammar. The newspaper was initially in simple katharevousa but then switched to
Pontic, developing it with Russian loanwords and words close to demotic Greek.
For years after the 1926 conference, it advocated Pontic as official language of the
19
Petyt (1980) 11; Chambers and Trudgill (1998) 3.
20
Dawkins (1937) 40.
community and of its schools.21 An editorial in 1928 revealed that the journalists
had felt forced to switch to Pontic to be comprehensible to their target readership;
their main worry in terms of linguistic models was that an archaic dialect like
Pontic would sound like ecclesiastical Greek22 – a very unfortunate situation for a
communist paper that was even printed in red ink.
(d) Structural similarity and partial intelligibility with Standard Greek are often
cited as reasons for classifying Pontic as a dialect. But these too are inconclusive
criteria, since Pontic differs more from Standard Greek than Norwegian does from
Danish, Serbian from Croatian, or Romanian from Moldavian. Conversely, the so-
called Chinese ‘dialects’ are mutually unintelligible, but are traditionally seen as part
of a single language because they share a written form. Greek has no close relatives
amongst official languages, and this may have reinforced the Greeks’ perception
that large linguistic differences (Abstand) are needed for ‘language’ status.23 The
problem with Abstand and intelligibility as criteria is that they are matters of degree,
and their assessment is therefore subjective. Pontic has been said to be a dialect by
virtue of claims that it has basically the same phonemic inventory, morphological
system and syntax as Standard Greek, and that the lexical influence of Turkish did
not alter the physiognomy of Pontic because the inflections remained Greek.24 All
this is questionable.
It is, therefore, not only the choice of criteria but also the interpretation of the
criteria that remains arbitrary. Opinions are affected by extra-linguistic factors too:
many Greeks object to classifying Pontic as a language because they feel that this
suggests that Pontic is not Greek, and that it therefore raises emotional issues about
the identity of the speakers and of the land they left behind. In Turkish, Romayka
is indeed referred to by a term (Rumca) unrelated to the one used for the Greek of
Greece (Yunanca), but this distinction is not unlike the one made in Greece – with
good reason – between Αλβανικά (Albanian) and Αρβανίτικα (a related dialect
spoken in Greece) or between Ρουµάνικα (Romanian) and Βλάχικα (a related
dialect spoken in Greece).
Turkish standards
Although there is no standard form of Greek recognised in Pontus today, it would
be very misleading to say that there is no linguistic norm that is recognised and
promoted amongst Muslim Pontic speakers. There is one, and that is Turkish.
Turkish, of course, is unrelated to Greek – it is a language of the Turkic branch of
the Altaic family, related to Uzbek, Tatar, Kazakh, Uighur, Kyrghyz, Azerbaijanian
and Turkmen. Turkish scholarship often classifies all these languages as dialects of
Turkish;25 but even if we disregard these, and count only the speakers of Turkey’s
21
Karpózilos (1988/9) 62.
22
Karpózilos (1996/7) 31.
23
Trudgill (2001) 23.
24
25
Cf. Hengirmen (1997) 496–510.
74 PIETRO BORTONE
Turkish (Türkiye Türkçesi), Turkish speakers number nearly seventy million. The
alphabet they use, which is an expanded form of the Latin alphabet, is the only one
known to Romayka speakers.
Turkish is officially a ‘language’; as such, it has – like Greek – received many
artificial interventions designed to standardize it, expand it and above all ‘purify’
it from the foreign elements it has adopted. The Turkish language reform – which
took off in the 1930s and is, to a degree, still operative – went to extreme lengths,
and is rightly known in Turkish history as nothing less than the dil devrimi
‘language revolution’. Although it is very common for languages with official
status and a written tradition to undergo artificial interventions (especially if they
belong to a country struggling to assert a separate or new national identity), few
languages have experienced such radical and ongoing tampering as Turkish and
Greek. However, Turkish prescriptivism, unlike its Greek counterpart, promoted
the adoption of new forms, and only rarely the revival of old ones.
In Pontus, Turkish is the language for all official purposes. Even place names
have been Turkicized. In 1964, by orders of the Ministry of the Interior, the
foreign (or seemingly foreign) names of the localities in the region were modified
or changed entirely. Thus, for instance, Αληθινός became Uzuntarla, Σαράχος
became Uzungöl, Κατωχώριν became Çaykara, Ζησινό became Bölümlü. Many
locals, nonetheless, still use the old Greek and Laz names of villages, and at times
are not sure about the new ones.
As a result of the status of Turkish as their only standard, speakers of Romayka –
although not sure about what ‘correct’ Romayka could be – have very clear views as
to what is and is not correct Turkish. They also feel that the Turkish dialect of their
area, which is also spoken by very many monolingual Turks, is not proper Turkish.
Interestingly, scholars have ascribed the differences between Black Sea Turkish
and the standard language, such as the striking violation of vowel harmony (the
fundamental and all-pervasive principle of Turkish phonology),26 to the substrate
influence of Greek.27
Turkish is taught in all schools in Pontus, and is the only language that is taught.
This has profound effects, since education in Turkey is free and, in principle,
mandatory for both sexes until the age of fourteen. Despite the fact that to this
day many Muslim Pontians report that they did not know Turkish until they
went to school (indicating that their Turkish was learnt rather than acquired: late
bilingualism in its technical sense),28 Turkish is also the only medium of school
instruction.
This attitude on the part of the schools is not peculiar to Turkey: in most, if not
all, countries, and long before the development of modern media, schools have
been an instrument of the state in the promotion of the official line on language use.
26
I have myself often heard forms such as gittuk, yukari, olmadi, güni for gittik, yukarı, olmadı,
günü.
27
Cf. Brendemoen (2002) 204–7.
28
Cf. Hoffmann (1991) 35.
The expectation is that pupils will adapt to the school system and not the other way
round – which does not surprise any of us, given our own prescriptivist schooling. In
Europe too, it has been the norm for schools to ignore or even to disallow the home
language of their pupils, and very seldom has this been due to practical problems
such as outside teachers lacking the knowledge of the community language. It
is rare for schools to use an unwritten local language to introduce a widely-used
language, despite recommendations by many educationalists who believe that this
would ‘ease in’ the pupils, help them achieve, and contribute to their self-esteem.29
It would also favour additive bilingualism.30 Many school systems across the world
do use the pupils’ home language as a medium to teach a target language, but this
happens primarily when both languages have similar status, such as when English
is taught in Greece. If the home language has less prestige and no written standard,
it is sometimes used at least initially (in the so-called ‘early exit programmes’) but,
at a later stage, the target language is employed for all purposes. Thus, for instance,
in northern Switzerland, Schwyzertüütch (the mother-tongue of most Swiss, but
with no written standard) is used as a medium in schools, but university lectures are
given only in German. The much-publicized suspicion of Turkey towards ‘ethnic’
languages is not exceptional either: it mirrors Europe’s and America’s long history
of mistrust of bilinguals, who were seen as less intelligent and as potentially disloyal
to their country.31
If we take the term ‘bilingual’ in its loose application32 and accept that it does
not necessarily presuppose total mastery of two languages,33 we must recognise that
virtually all speakers of Muslim Pontic are now bilingual. They use extensive code-
switching, and many are more fluent in Turkish than in Romayka. Their proficiency
in Turkish is subtractive bilingualism,34 as they are clearly transitionally bilingual:
Romayka is slowly being replaced by Turkish, which is already the dominant
language. Subtractive bilingualism is typical of communities where the recessive
language is of little advantage or prestige; additive bilingualism (such as the very
fluent English of most Scandinavians, or the French of the pre-revolutionary
Russian elite) would presuppose, conversely, a socio-economic utility.
There are other, largely macro-sociological, factors favouring the adoption of
Turkish by Romayka speakers. Urbanization, coupled with increased geographical
mobility, has dismantled many rural communities. The traditional lifestyle that
kept the speakers together and in situ is now unsustainable, and upward social and
economic mobility is more associated with speakers of Turkish. Males are also
required to do military service, for which full command of Turkish is necessary,
and during which they inevitably spend fifteen months in non-Romayka-speaking
parts of the country. Turkish is also the language of radio and TV, which have
29
Arguments and counterarguments are outlined in Appel and Muysken (1987) 61–3.
30
Several studies supporting this are cited in Landry and Allard (1992) 223.
31
Romaine (2000) 224–30.
32
Cf. Hoffmann (1991) 14: ‘bilingualism defies delimitation’.
33
Myers-Scotton (2006) 3.
34
In the sense of Lambert (1974).
76 PIETRO BORTONE
now become common, albeit belatedly, even in Romayka-speaking villages.

Turkish is also used locally by Romayka speakers as a lingua franca to interact
with other Turks, some of whom use other unwritten languages at home: besides
Romayka, there are other languages spoken in the area – notably Laz – which are
not normally written or broadcast, and are, like Romayka, disregarded in schools.
Romayka fares poorly on most key indicators normally cited as crucial for linguistic
survival:35 it is demographically weak, and its speakers are culturally, religiously
and visually similar to the wider community; the wider community, on the other
hand, is culturally, politically, economically and demographically stronger, and
allows assimilation.
The language shift away from Pontic is not peculiar to its Turkish setting. In
Greece too, Pontic, like all other Greek dialects, is facing terminal attrition, and
is being replaced by Standard Greek. The same is happening in other countries
with a sizeable Pontic-speaking population, along the other (northern and eastern)
shores of the Black Sea. In Russia and the Ukraine, Pontic is threatened not only
by Russian and Ukrainian, but by Standard Modern Greek, which is promoted
there by Greece, and has been introduced as a foreign language at local schools.
Something similar takes place in other countries where Greek dialects are spoken:
in Apulia, southern Italy, the local Greek is officially recognised and partly taught,
but has to compete not only with Italian but with Standard Modern Greek, which
is also taught in local schools.
Expected effects of not having a Greek standard

A unified standard encourages, in principle, both continuity and homogeneity. It
promotes diachronic and synchronic uniformity. It is normally assumed that the
existence of a standard greatly favours the maintenance and survival of a language.
Conversely, the existence of a written and standardized form of the rival majority
language is considered a typical factor contributing to language shift.36 Many
linguistic communities have created an artificial mixed variety of their language
to be used as a common form, in the belief that this will make it more durable.
The standardization of other varieties of Greek spoken outside Greece, such as the
Grico of Apulia, in southern Italy, has been advocated by scholars who claim that
this is a priority if the speech community wants to save its language.37 The same
thing is happening with other Balkan languages; there is a campaign, for example,
for the standardization of Arbresh, the form of Albanian spoken in Sicily, in the
belief that this will halt its demise.38 Indeed, other endangered languages have been
standardized in the express hope of saving them: a case in point is the creation in
Switzerland of Rumantsch Grischun in 1982, as a ‘compromise language’ to unify
and preserve Romansh.
35
Appel and Muysken (1987) 38.
36
Paulston (2003) 402.
37
Profili (1999) 52.
38
Derhami (2002) 249.
To assess whether this could be done with Romayka, we need to consider

what the establishment of a particular language form as the standard commonly
entails:39
(a) Selection: identification of a favoured variety (‘status planning’),
(b) Codification: graphization and grammatical definition (‘corpus planning’),
(c) Elaboration: development of vocabulary and stylistic variants,
(d) Endorsement and implementation by the state (political support),
(e) Endorsement and implementation by the community (public support).
None of these factors exist for Romayka:

(a) Romayka has many equal varieties, differing significantly in vocabulary,
morphology and syntax.40 Lack of uniformity is a weakening factor for a language.
A standard, conversely, is more than an aid to interdialectal communication: it is
a partial protection from linguistic (and, according to some, social and political)
fragmentation.41 In Greece, this used to be pointed out with reference to spoken
Greek by those who supported katharevousa as the only unified and unifying
language of a nation that they perceived as being under threat.
(b) Except for the occasional word quoted in a text in another language, a couple
of internet sites, and sporadic and unsystematic use in private messages, Romayka
is not written at all. Therefore, it also lacks the most common material required for
the emergence of a standard: a written tradition, with an agreed canon of ‘best’
literature. Standardization is a practice essentially concerned with the written
language: in Greece too, the perennial battle between katharevousa and demotic
was about the selection of a written standard. Some hoped that katharevousa would
also gradually become the spoken language of Greece, but even the advocates
of katharevousa mostly used demotic when chatting. Their assumption, that an
artificial, archaizing, ‘purified’ form of written Greek could eventually replace the
vernacular, only shows the power attributed to texts: the very fact that a language
variety is written gives it an air of concreteness, permanence and authority. This
explicit graphocentrism has long held sway, not only in public perceptions of
language but also in the history of scholarship.42 The prestige of ‘dead’ languages like
classical Greek and Latin, seen as the absolute prototypes of perfection, correctness
and intellectual propriety, has reinforced the perception that linguistic models are
fixed, that languages have diachronically a best phase and synchronically a bon
usage, and that the written form is what matters and constitutes the ‘real’ language.
Demotic Greek was often thought to have no grammar and, like many languages
that had to coexist with a second, more prestigious language that had a long written
history (like Yiddish with Hebrew, Catalan with Spanish), was not thought to be a
language at all.
(c) The very few writers with a Romayka-speaking background, such
39
Cf. Haugen (1966) 933 and Edwards (1994) 173.
40
Cf. Mackridge (1987) 120.
41
Cf. Myers-Scotton (2006) 392.
42
Cf. Joseph (1987) 37 and Linell (2005) 11–12.
78 PIETRO BORTONE
as Muhammet Çakıral, even if writing about local people in a local setting,

understandably choose to write in Turkish, at most transcribing the occasional
Romayka word.43 When I asked Çakıral why he did not write his stories in
Romayka, his response was that Romayka does not even have a standard alphabet,
and that there is no grammar book for it – in other words, no established norm. But
Turkish, being officially ‘a language’, has also a wider audience and – crucially – a
wider vocabulary and stylistic repertoire. It is, after all, characteristic of standard
languages that they have minimal variation in form but maximal variation in
function:44 richer lexicon and multiple registers. It must be recognised that Turkish
provides Romayka speakers with a vast technical, formal and abstract vocabulary
that they otherwise would not have. Romayka is stylistically more limited, and
monostylism is considered another symptom of impending language death.45
(d) Romayka has no legal recognition. However, over ninety-five per cent of the
world’s languages are not officially recognised in the states where they are spoken;46
furthermore, institutional acknowledgement and support, such as official status
or school teaching, cannot compensate for the restricted use of an endangered
language by its speakers or for its limited intergenerational transmission.47 Formal
recognition of a language does not guarantee its survival, nor its status beyond
the law books: in countries that officially recognise more than one language
(Netherlands, Israel, Luxembourg, Singapore, Finland, India, Belgium, South
Africa, Canada, to name but a few) each language has a different social standing.
Legal provisions only give a weaker language a slightly better chance of survival,
and only in some cases do they enable its speakers to remain monolingual. Indeed,
many languages have official status in a country where they are nonetheless dying
(for instance, Romansh in Switzerland, Irish in Ireland); this includes ‘situational’
minority languages,48 which are very strong in another, even neighbouring, country
(like Swedish, which is disappearing in Finland but thriving in Sweden). School
teaching is, likewise, useful but not decisive. As has been wryly pointed out by
Joshua Fishman,49 schools cannot ensure active knowledge of algebra a few years
after graduation, unless there is a continuous interest and use afterwards; and the
same surely applies to language maintenance.
(e) Romayka speakers do not seek formal recognition of their language and do
not endeavour to spread its use. In global terms, people actively fighting to save their
language are usually very rare: a few activists guided by theoretical or sentimental
reasons which the broader community endorses in principle but ignores in practice.
Furthermore, Romayka, unlike languages such as Hebrew and Irish, lacks political
motivation: the speakers have no distinct ethnic and political identity, and no
43
Çakıral (2006a), (2006b).
44
Haugen (1966) 931.
45
Appel and Muysken (1987) 44.
46
Romaine (2002) 194.
47
Fishman (1997) 192–4.
48
So called by Trudgill (2001) 25.
49
Fishman (1992) 400.
separatist aspirations, which are often key factors in the development and retention
of a different language variety. Indeed, the few Romayka speakers who might want
to do something to save their language may be discouraged by the fear that their
aims may be misconstrued as being political.
As the continuing demise of Irish and the unexpected success of ‘Modern
Hebrew’ have shown, what really decides the fate of an endangered language is
not institutional support but the speakers’ own attitude. Becoming bilingual may
be a necessity, but abandoning the use of the community’s language in internal
communication is, to some degree, a choice. Romayka will probably disappear,
ultimately, because its speakers will not bother to keep it going.
Across the globe, the prospects for endangered languages are not encouraging.
It is estimated that, in the last five hundred years, half the world’s languages have
vanished,50 mainly with no trace, and that half of those left will become extinct
in this century, with eighty per cent of the remainder being close to extinction.51
Although we mourn the demise of Greek in Asia Minor (probably because we see
it happening), we ourselves easily forget that, in earlier times, the spread of Greek
in Asia Minor (like that of Turkish later) obliterated several pre-existing languages
too.
Greek and purism

Borrowing foreign words is the first and most common structural change resulting
from language contact, and is popularly identified as the beginning of the slippery
slope towards language death. The scholarly contribution to the establishment
and maintenance of a standard has often involved purging foreign words (or, less
often, phrases and syntactic constructions)52 by studiously producing new ‘native’
terms. It is worth noting that even scholars who decry a prescriptive approach to
languages nevertheless often favour measures to prevent one language from being
gradually replaced by another.
The artificial creation of new words from native roots is very common and often
successful. A language-engineering programme of this kind has been effected on
a vast scale in Israeli Modern Hebrew, and the substitution of borrowed elements
by native ones, often designed ad hoc, has been carried out in many languages,
such as Swedish, Hebrew, Tamil, Croatian, French, Icelandic, Welsh, German,
and Lithuanian.
In Greek and in Turkish, foreign words have been strongly objected to, and
have often been successfully expurgated and replaced. Puristic intervention in
both languages was a response to political rather than linguistic issues, as shown
by the fact that the words the Greek purists wanted to expunge because they
were Turkish (a few random examples: σοκάκι ‘[back] street’; κουσούρι ‘flaw’;
χαϊβάνι ‘animal, brute’; ντέρτι ‘tribulation’; τζάµι ‘glass’; λεκές ‘stain’) were
50
Sasse (1992) 7.
51
Hale (1998) 192.
52
Cf. Thomas (1991) 115.
80 PIETRO BORTONE
often the same words that Turkish purists wanted to expunge because they were
not Turkish (sokak, kusur, hayvan are from Arabic, dert, cam, leke from Persian).
Now the efforts of purists in both countries are focused, with little success, on
loanwords from English.
Today, Standard Modern Greek, unlike Romayka, may have all the words it
needs, but this is also because its lexicon is not only taken from demotic Greek: it
has been considerably (and some may say artificially) enriched by katharevousa.
Many missing words have been taken from the immense learned tradition (there is
indeed some truth to the cliché that ‘the Greeks have a word for it’). For modern
concepts, an ancient word has often been revived, given a novel semantic nuance
and a new lease on life, e.g. ὑπάλληλος, ‘subordinate > employee’; ὑπουργός,
‘assistant > minister’; βουλή, ‘council > parliament’; κράτος, ‘sovereignty > state’.
Other words have been made up from Ancient Greek roots, often in imitation of
the western European originals, classic examples being French réaliser becoming
πραγµατοποιῶ, German Weltanschauung becoming κοσµοθεωρία, and English
skyscraper becoming οὐρανοξύστης.53 This may seem contrived and cosmetic,
but it is effective. The same thing has been done in Turkish, where countless new
words and morphemes have been coined, often as replacements of foreign ones.
This type of artificial lexical enrichment not only aims at expanding or modernizing
a language’s vocabulary, but also at ‘protecting’ the language, at least superficially,
from foreign elements. In Greece this is still successfully done: even if English is
making inroads, numerous new Greek words are being created by calque, with
morphemes almost invariably taken from Ancient or Medieval Greek. Opening
a Greek web browser, one sees that a webpage is an ἰστοσελίδα, the navigation
toolbar is the ἐργαλειοθήκη πλοήγησης, the cache is the λανθάνουσα µνήµη,
and the blocking of pop-up windows is the ϕραγὴ ἀνδυοµένων παραθύρων.
The puristic tradition of lexical coinage from native roots, typical of katharevousa,
is alive and well.
Romayka and purism

The deliberate – and essentially learned – lexical engineering just discussed has not
happened and cannot happen in Romayka. One reason is that ‘it is incontrovertible
that puristic activity is usually associated with written, standard languages’.54
Then again, purism presupposes an awareness of the history of one’s language
which is unavailable to Romayka speakers. It is also unlikely that they would turn
to Standard Modern Greek as a source of borrowing – in the way Turkish has
looked to other Turkic languages, Estonian to Finnish, and Romanian to other
Romance languages – because Greek is the language associated with Greece and
with Christianity. Lexical and phrasal enrichment does take place in Romayka,
but the source normally is Turkish. In modern Greece, awareness of the history of
Greek made it clear that Western European words like telephone and photograph
53
Cf. Mackridge (1985) 312.
54
Thomas (1991) 2.
had been coined with Ancient Greek roots; accordingly, they were re-Hellenized
as τηλέφωνο and φωτογραφία. In Romayka, the Turkish terms have been
adopted, with minimal morphological and phonological adaptation: /tele’fᴐni/, /
fᴐtᴐ’ɣrafi/ < Turkish telefon and fotoğraf + Romayka -i.
In Romayka, borrowings from Turkish appear to have no limits: no restrictions
are imposed socially or formally, and even the constraints normally expected in
code-switching barely apply. The two languages are converging, as can be seen
from the use of intra-sentential switching:55 not only are content morphemes
(lexical items) of one language embedded into the morphosyntactic frame of the
other, but syntactic configurations and morphological patterns taken from both
languages are also blended together. It is becoming difficult to say that Turkish is
not a native language (in itself a very elusive concept) of Romayka speakers.
Why is there no purism in Romayka? Why do speakers accept the dominance
and the gradual take-over by Turkish? The reason is not only the standard status of
Turkish. It is true that any standard language is ‘the codified wishes of the socially
dominant’,56 and that the adoption of a foreign language by an entire community,
either as first or second language, is also usually due to the political, social, economic
or cultural clout (nationally or internationally) of its speakers. But this happens if
that privileged position is recognised and accepted by the speakers of other dialects
or other languages. A puristic attitude, on the other hand, is very often part of
a nationalist stance. Theoreticians of purism as a cross-linguistic phenomenon
see it as part of a dualistic perception of the world.57 This might be an adequate
description of the ‘Greek/non-Greek’ dichotomy in the world-view of ancient and,
mutatis mutandis, modern Greeks, but it is not the way that Romayka speakers
see other Turks. For Romayka speakers, Turkish is the language of their fellow-
nationals; in Pontus, therefore, the crusades to ‘de-Turkify’ the language that were
launched in Greece58 (as well as in other ex-Ottoman areas, such as Bulgaria)59
are neither feasible nor desired. Even the peoples who do fight against the foreign
elements found in their language often do so selectively: Romanians endeavoured
to expunge Slavic elements, but gladly adopted forms from French and Italian
because that was the direction in which they wished to take their identity.
Another key factor is that, since their language has not been codified,
systematized and given a model form, Romayka speakers have no notions of
‘purity’ vis-à-vis other languages – just as they have no notions of correctness to
be used against ad hoc grammatical or syntactic constructions, and no notion that
one local variety of Romayka is ‘higher’ than another. They seem free from the
classic linguistic fears that beset speakers of standardized languages. The history of
English also shows that anxiety about linguistic change and correctness peaked in
55
Cf. Myers-Scotton (2002) 105.
56
Edwards (1994) 7.
57
Thomas (1991) 37.
58
Cf. Dizikirikis’ popular book, Let Us De-Turkify Our Language – An Essay for the Liberation of
Greek from the Words that have Turkish Origin.
59
Grannes (1970) 11.
82 PIETRO BORTONE
the eighteenth century, when spelling became more standardized. Standardization,

after all, is essentially the suppression of optional variability (and so, in theory,
of diachronic change),60 with some flexibility allowed only in speech. Romayka
speakers have no standard (besides Turkish) and no worries: whenever I asked
them how one decides which Romayka form is, in their view, right or better, they
answered ‘serbest’, the Turkish for free(ly). This seems to contradict the general
perception that:
All users of language in all speech communities apparently hold evaluative attitudes
towards variant forms: some variants are regarded as ‘better’ or ‘more beautiful’ or
‘more appropriate’ or ‘more correct’ than others.61
Making up words, and even morphological and syntactic patterns, is a recognised

possibility for speakers of languages without a standard, or at least without a
standard known to the speakers.62 This does not mean that their language lacks
structure: it does not have fewer linguistic rules – only fewer linguistic norms.
Effective communication and competence (in the Chomskyan sense) certainly
requires a shared language, but we all have an idiolect and some room for creativity.
And this is particularly acceptable in Romayka. For instance, I have heard the same
(very fluent) Romayka speaker refer to the Greeks (of Greece) by four different but
perfectly transparent terms:
(a) /ela�ðᴐti/
Greek *Ελλαδ+ώτ+οι, as in Ἑλλάδ(α), (στρατι)ώτ(ες), (ἄνθρωπ)οι
(b) /e�leni/
Greek *Ελλήν+οι, as in Ἕλλην(ες), (῎Αγγλ)οι
(c) /ju�nani/
Turkish yunan ‘Greek’ + Greek -οι
(d) /junanistan�liðes/
Turkish yunanistan ‘Greece’ + Turkish -li or Greek -λή + Greek -δες,
as in Turkish Türkiye-li ‘from Turkey’, or Hellenized Turkish µερακ-λή-δες.
Unexpected effects of not having a Greek standard

What does a variety of Greek look like, if it is unexposed to the Greek classicizing
tradition and ἀρχαιολατρεία? The paradox is that, apart from the copious Turkish
elements (to be expected, since the speakers are in Turkey), Muslim Pontic Greek
has remained, in some respects, far more archaic than Modern Greek – even more
archaic than ‘Christian’ Pontic. A big paradox indeed, if we consider the extensive
and strenuous efforts made by the Greeks, for centuries, to make their Greek more
archaic. This is why Greek-educated Pontians felt that Muslim Pontic was the
‘best Greek’. In unpublished interviews, several of the Christian refugees who
moved from the Of area to Greece in the 1920s made relevant comments:
60
Milroy and Milroy (1985) 8.
61
Ferguson (1994) 18. Cf. an almost identical statement in Wald (1985) 123.
62
Cf. Le Page and Tabouret-Keller (1985) 11.
οἱ Τοῦρκοι τῆς Ζησινὸς γνωρίζανε καλὰ τὰ ἑλληνικά. Οἱ γριές καὶ οἱ γέροι

µιλοῦσανε τὰ καθαυτό ἑλληνικά, τὴν ἀρχαία γλῶσσα.63
The Turks of Zisinó [a village in the area] knew Greek well. The old women and men
spoke real Greek, the ancient language.
Αὐτοὶ µιλούσανε τὰ ἑλληνικὰ ὅπως οἱ ἀρχαίοι Ἕλληνες. Ἑµεῖς καλὰ καλὰ
δέν τοὺς καταλαβαίναµε . . . ἐπειδὴ αὐτοὶ δὲν ἐρχόταν τόσο σὲ ἐπαφὴ µὲ
Ἕλληνες ἄλλων περιφερειῶν τοῦ Πόντου, κράτησαν τὴν τοπική µας γλῶσσα
καλύτερα.64
They used to speak Greek like the ancient Greeks. We would barely understand
them . . . they preserved our local language better because they were not coming into
contact much with Greeks from other areas of Pontus.
ἀπὸ µᾶς πιὸ καθαρὰ µιλούσανε τὰ ἑλληνικὰ τ᾽ ἀρχαῖα.65
They spoke purer Greek than we did – Ancient Greek.
Scholars of Greek Pontic also report that Muslim Pontic dialects are regarded as
‘modèles de pureté archaïque’ that Greece has lost.66 In accordance with a common
principle of dialect geography, Greek peripheral dialects have archaic traits; but the
Greek of the Of region has traits lost everywhere else.
Within traditional Greek dialectology there has been a tendency to hunt single-
mindedly for archaic elements, and even to try to claim continuity between Modern
and Ancient Greek dialects,67 downplaying innovations and foreign elements. This
approach, with its obvious ideological underpinnings, highlights the low status and
limited interest afforded to dialects in themselves – except to the ancient ones,
which have the prestige of ‘languages’ thanks to their written tradition. While we
should endeavour to avoid such pitfalls, we would do well to emphasize the archaic
nature of Romayka, if only because of the implicit irony: its archaic character is due
to the very fact that Romayka has been isolated from the Greek tradition. Let us
review some archaic features found in Romayka.
(a) Pronunciation
(i) Romayka shows retention of initial unstressed vowels /ε/, /i/, /o/, /u/:
/εksεro/, /εpiɣa/, /εkliðosa/, /ospiti/ – Modern Greek usually ξέρω, πῆγα,
κλείδωσα, σπίτι etc. ‘Christian’ Pontic too would retain the initial vowel
but, like most northern Greek dialects, it would also delete the post-tonic /i/
and /u/: οσπίτ, etc.
(ii) There is no synizesis of final diphthongs /'ia/ /'eo/ /'io/: /mirᴐ'ðia/, /ki'lia/,
/pa'lεo/ – unlike Modern Greek µυρωδιά ‘smell’, κοιλιά ‘stomach’, παλιό
‘old’, etc.
(iii) Romayka, like Greek Pontic, has also an intriguing open /ε/ sound (for
63
(PO125) 24–5.
64
(PO129) 30, 33.
65
(PO133) 127.
66
Drettas (1998) 87.
67
Tzitzilis (2000) 15.
84 PIETRO BORTONE
Modern Greek /i/) where Greek spelling has an η: /εrθεn/, /psεno/, /

εkusa/, /εɣapεsa/ – representing Greek ἦρθεν, ψήνω, ἤκουσα, ἠγάπησα.
Greek scholars are certain that this is the ancient pronunciation preserved;68
it is an extremely seductive thought, though doubts have been raised.69
(iv) One remarkable feature, both philologically and sociolinguistically, is the
pronunciation of third-person personal pronouns. Pontian refugees in Greece
came to be derogatorily nicknamed αούτηδες to mock their pronunciation
of the pronoun αὐτός: in Modern Greek it is pronounced /aftᴐs/, but /autos/
in Pontic. The latter, however, is much closer to the classical pronunciation.
We may note that even the highly educated Greek archaizers never advocated
that a more ancient pronunciation should be revived. Essentially, they ignored
the profound changes in the phonetics of individual sounds from classical to
contemporary Greek. Indeed, Modern Greeks pronounce even Ancient Greek as
if it were Modern – just as, to a very large degree, they spell Modern Greek as if it
were Ancient.
(b) Morphology
(i) The ancient imperative in -(s)on survives: /akuson/, /suron/, /ðulεpson/,
/apson/– for Modern Greek άκουσε ‘listen!’, σύρε ‘pull!’, δούλεψε ‘work!’,
ἅψε ‘lit!’, corresponding to ancient ἄκουσον, etc.
(ii) One also hears several old verbal forms such as /εton/ ‘was’– for Modern
demotic Greek ἦταν but older ἦτον – or /εksεvεn/ ‘came out’ – for Modern
βγήκε but Ancient ἐξ-έβην.
(iii) The vocalic temporal augment is still used: /εɣapεsan/, /εkusa/ – for Modern
demotic Greek ἀγάπησαν, ἄκουσα.
(iv) Some old possessives are still in use, like /εmon/ – for Modern Greek µας,
but Ancient Greek ἡµῶν.
(v) The ancient aorist passive has not merged with the perfect: /εfovεθε/,
/εstaθε/, /εpsεθε/ – unlike Modern Greek φοβήθηκε ‘(s)he got scared’,
στάθηκε ‘(s)he stood’, ψήθηκε ‘it was cooked’, but like Ancient Greek
ἐφοβήθη, ἐστάθη, ἐψήθη. These ancient forms were highly recommended
and valued by purists, but all Greek scholars admitted that one could not hope
to hear them in spoken Greek, not even in the most educated conversations.
Tzartzanos, for instance, writing in the katharevousa years, acknowledged
that even a university professor conversing with another university professor
would never say, for ‘I slept’, ἐκοιµήθην (the classical form, which now would
be pronounced /εkimiθin/, corresponding to modern κοιµήθηκα)70. In
Romayka, however, even an illiterate speaker says that (s)he /εkimεθε/.71
68
Savvidis (1973) 36–9, Tombaidis (1996) 265, and Andriotis (1995) 96
69
Horrocks (1997) 312–13
70
Tzartzanos (1934) 17.
71
Easternmost subdialects palatalize velar stops before front vowels: εtʃimεθε, in this case.
(c) Syntax
(i) Weak pronouns always follow the verb: /eɣrikisa to/ ‘I understood it’. In
Standard Greek this happens only after imperatives and gerunds.
(ii) Some varieties of Romayka also form the future tense like Medieval Greek,
with νά: /na trᴐ:/ ‘I will eat’ (Standard Greek θα φάω). The particle θα,
used also in Christian Pontic, is known only to some Romayka speakers.
(iii) As first noted by Mackridge, the infinitive, which has vanished from Modern
Greek, from Christian Pontic, and from all dialects in Greece, is still used in
some Romayka-speaking villages after the past tenses of /pᴐrᴐ/ and /θεlᴐ/,
as in Medieval Greek.72 Romayka speakers say: /utʃ εpᴐrεsa tʃimεθinε/
‘I couldn’t sleep’, like Medieval Greek οὐκ ἐµπόρεσα κοιµηθῆναι, /utʃ
εθεlεsε εrθinε/ ‘he did not want to come’, like Ancient Greek οὐκ ἠθέλησε
ἐλθεῖν, blurring -εῖν and -ῆναι.
(d) Semantics
(i) /tεrᴐ/ still means ‘to look’, as it did in Ancient Greek, unlike Modern Greek
τηρῶ, which means (mainly) ‘to keep’.
(ii) /fεvɣo/ still means ‘to flee’, as it did in Ancient Greek, unlike Modern Greek
φεύγω ‘to leave’.
(e) Vocabulary
(i) /kaʎo/ ‘better’: in Modern Greek, normally, καλύτερα, but classical
κάλλιον.
(ii) /γrikᴐ/ ‘I understand’: in Modern Greek normally καταλαβαίνω, but
Medieval ἀγροικῶ.
(iii) /lihᴐ/ ‘I lick’: in Modern Greek normally γλείφω, but classical λείχω.
(iv) /mizᴐderi/ = ‘elders’: lit. µειζότεροι, like koine µείζονες, for Modern Greek
µεγαλύτεροι.
(v) /kruo/ = ‘I hit’: Modern Greek normally χτυπώ, classical κρούω, ‘Christian’
Pontic κρούγω.
(vi) /u(tʃ)/ = ‘not’, in the Of and Sürmene areas only: in Modern Greek δεν, but
Classical οὐκ.
(f) ‘Accidental’ purism

Some writers in Greece have suggested that Greek words they knew from Christian
Pontic should be officially taken over to replace foreign ones used in Greek – the
prospective change to be effected by inclusion in the authoritative Lexicon of the
Academy of Athens – and that more terms of this kind should be sought in Pontic
dialects. So, for instance, it has been recommended that instead of saying κουνιάδος
‘wife’s brother’ (from Venetian cugnado, unknown to Romayka speakers), Standard
Greek would do better (‘ασυγκρίτως καλλίτερα’), like Pontic, to use the native,
ancient (‘ελληνικώτατο’) term γυναικάδελφος. In the same vein, it has been
72
Mackridge (1995) 159 and (1999) 27.
86 PIETRO BORTONE
urged that Greek replace µπατζανάκης, ‘wife’s sister’s husband’ (from Turkish
bacanak), with the Ancient Greek σύγγαµβρος, a word still used in Pontic.73
While Romayka is innovative in many respects that would not have pleased
traditionalists, it would have delighted advocates of the lexical engineering of
katharevousa with its countless spontaneous creations such as:
(i) /pεrtʃεpaγo/ ‘to accompany’ (< παίρ[νω] καὶ πάγω ‘take and go’).
(ii) /ajinetʃiɣos/ ‘unmarried [male]’ (for *α-γυναίκ-ιγος, whereas in Modern
Greek one says, somewhat confusingly, αν-ύπ-ανδρος).
Greek purists would have been all the more thrilled with Romayka’s use of certain
native Greek terms (with occasional parallels in other dialects) instead of the
‘foreign’ ones used in Greek. Amongst these are:
(iii) /aθᴐγala/ ‘milk-cream’, instead of the polysemous Standard Greek κρέµα
(from Italian crema), or instead of Greek καϊµάκι, which is from Turkish
kaymak.
(iv) /lεftokari/ ‘hazel-nut’ (in katharevousa λεπτοκάρυον, and exceptionally, in
some forms of demotic Greek, λεφτόκαρο), rather than Greek φουντούκι
< Turkish fındık (although ultimately from Greek ποντικ-, possibly via
Arabic).
These terms highlight the paradox of Romayka. A Turkish word may be used by
Greeks in Greece, who, for instance, say γιαούρτι ‘yoghurt’; this is from Turkish
yoğurt, a Turkish word adopted by almost all national, standardized European
languages (despite their purism), as well as by many other languages, from Hebrew
to Japanese. Romayka speakers, on the other hand – although they make no attempt
to ‘purify’ their language, use Turkish every day, and live in Turkey – use a Greek
word for it: /ksinᴐɣala/.
References
Andriotis, N.P. (1995), Ἱστορία τῆς ἑλληνικῆς γλώσσας – τέσσερις µελέτες,
Thessaloniki.
Appel, R. and Muysken, P. (1987), Language Contact and Bilingualism, London.
Bazin, L. (1983), ‘La réforme linguistique en Turquie’, in I. Fodor and C. Hagège (eds.),
Language Reform: History and Future, vol. 1, Hamburg: 155–77.
Brendemoen, B. (2002), The Turkish Dialects of Trabzon, Their Phonology and Historical
Development, vol. 1, Wiesbaden.
Chambers, J.K. and Trudgill, P. (1998), Dialectology, 2nd edn, Cambridge.
Çakıral, M. (2006a), Şamiram’ın ustaları, Istanbul.
Çakıral, M. (2006b), Her zaman griydi Karadeniz, Istanbul.
Dawkins, R.M. (1937), ‘The Pontic dialect of Modern Greek in Asia Minor and Russia’,
Transactions of the Philological Society: 15–52.
73
Lavrentidis (1985) 14.
Derhemi, E. (2002), ‘The endangered Arbresh language and the importance of standardized
writing for its survival: the case of Piana degli Albanesi’, International Journal on
Multicultural Societies, 4: 248–69.
Dizikirikis, Y.S. (1975), Να ξετουρκέψουµε τη γλώσσα µας – δοκίµιο για της
απαλλαγή της νεοελληνικής απο τις λέξεις που έχουνε τουρκική προέλευση,
Athens.
Drettas, G. (1997), Aspects pontiques, Paris.
Drettas, G. (1998), ‘La langue pontique comme objet identitaire: questions de
représentations’, in M. Bruneau (ed.), Les Grecs pontiques – diaspora, identité, territoires,
Paris: 71–88.
Drettas, G. (1999), ‘Το ελληνο-ποντιακό διαλεκτικό σύνολο’, in Α.-Φ. Χριστίδης et al.
(eds.), Διαλεκτικοί θ�λακοι της ελληνικής γλώσσας, Athens: 15–24.
Edwards, J.R. (1994), Multilingualism, London.
Ferguson, C.A. (1994), ‘Dialect, register, and genre: working assumptions about
conventionalization’, in D. Biber and E. Finegan (eds.), Sociolinguistic Perspectives on
Register, Oxford: 15–30.
Fishman, J.A. (1992), ‘Conference summary’, in W. Fase, K. Jaspaert and S. Kroon (eds.),
Maintenance and Loss of Minority Languages, Amsterdam: 395–403.
Fishman J.A. (1997), ‘Maintaining languages – What works? What doesn’t?’, in G. Cantoni
(ed.), Stabilizing Indigenous Languages, Flagstaff, Ariz.: 186–98.
Grannes, A. (1970), Étude sur les turcismes en bulgare, Oslo.
Hale, K. (1998), ‘On endangered languages and the importance of linguistic diversity’, in
L.A. Grenoble and L.J. Whaley (eds.), Endangered Languages, Cambridge: 192–216.
Haugen, E. (1966), ‘Dialect, language, nation’, American Anthropologist, 68: 922–35.
Hengirmen, M. (1997), Türkçe dilbilgisi, Ankara.
Hoffmann, C. (1991), An Introduction to Bilingualism, London.
Hudson, R.A. (1996), Sociolinguistics, 2nd edn, Cambridge.
Joseph, J.E. (1987), Eloquence and Power: The Rise of Language Standards and Standard
Languages, London.
Karpózilos, A. and M. (1988/9), ‘Ελλήνο-ποντιακά βιβλία στη Σοβιετική ένωση’,
Αρχείον Πόντου, 42: 57–104.
Karpózilos, A. (1996/7), ‘The Greeks of Russia: pages from the political and cultural history
of Pontian and Mariupol Greeks in Southern Russia’, Αρχείον Πόντου, 47: 16–40.
Kondosopoulos, N. (1981), Δίαλεκτοι καὶ ἰδιώµατα τῆς νέας ἑλληνικῆς, Athens.
Labov, W. (1970), The Study of Nonstandard English, Champaign, Ill.
Lambert, W.E. (1974), ‘Culture and language as factors in learning and education’,
in F.E. Aboud and R.D. Meade (eds.), Cultural Factors in Learning and Education,
Washington.
Lampsidis, O. (1959), ‘Un dialecte qui se meurt: le dialecte grec du Pont-Euxin (Asie-
Mineure)’, Αρχείον Πόντου, 23: 199–205.
Landry, R. and Allard, R. (1992), ‘Ethnolinguistic vitality and the bilingual development
of minority and majority group students’, in W. Fase, K. Jaspaert and S. Kroon (eds.),
Maintenance and Loss of Minority Languages, Amsterdam: 223–51.
Lavrentidis, I. = Λαυρεντίδης Ι. (1985), ‘Πόντος, πόντιοι, και ποντιακή διαλεκτος’,
Ποντιακή Εστία, 61: 1–16.
88 PIETRO BORTONE
Le Page, R.B. and Tabouret-Keller, A. (1985), Acts of Identity: Creole-Based Approaches to

Language and Ethnicity, Cambridge.
Lewis, G. (1999), The Turkish Language Reform: A Catastrophic Success, Oxford.
Linell, P. (2005), The Written Language Bias in Linguistics: Its Nature, Origins, and
Transformations, London and New York.
Mackridge, P. (1985), The Modern Greek Language: A Descriptive Analysis of Standard
Modern Greek, Oxford.
Mackridge, P. (1987), ‘Greek-speaking Moslems of north-east Turkey: prolegomena to
a study of the ophitic sub-dialect of Pontic’, Byzantine and Modern Greek Studies, 11:
115–37.
Mackridge, P. (1995), ‘Τα ποντιακά στη σηµερινή Τουρκία: αρχαία στοιχεία στο ιδίωµα
του Οφη’, Αρχείον Πόντου, 46: 153–61.
Mackridge, P. (1997), ‘The Medieval Greek infinitive in the light of modern dialectal
evidence’, in C.N. Constantinides et al. (eds.), Φιλέλλην: Studies in Honour of Robert
Browning, Venice: 191–204.
Mackridge, P. (1999), ‘Η ελληνοφωνία στην περιοχή του Όφη (Πόντος)’, in Α.-Φ.
Χριστίδης et al. (eds.), Διαλεκτικοί θ�λακοι της ελληνικής γλώσσας, Athens:
25–30.
Milroy, J. and Milroy, L. (1985), Authority in Language: Investigating Language Prescription
and Standardization, London.
Myers-Scotton, C. (2002), Contact Linguistics: Bilingual Encounters and Grammatical
Outcomes, Oxford.
Myers-Scotton, C. (2006), Multiple Voices: An Introduction to Bilingualism. Malden, Mass.
Papadopoulos, A.A. (1953), ‘Χαρακτηριστικὰ τῆς Ποντικῆς [sic] διαλέκτου’, Αρχείον
Πόντου, 18: 83-93.
Paulston, C.B. (2003), ‘Linguistic minorities and language policies’, in C.B. Paulston and
G.R. Tucker, Sociolinguistics: The Essential Readings, Oxford: 394–407.
Petyt, K.M. (1980), The Study of Dialect: An Introduction to Dialectology, London.
Profili, O. (1999), ‘Η αναζωογόνηση της grico στην Grecía Salentina’, in Α.-Φ. Χριστίδης
et al. (eds.), Διαλεκτικοί θ�λακοι της ελληνικής γλώσσας, Athens: 47–54.
Romaine, S. (2000), Language in Society: An Introduction to Sociolinguistics, 2nd edn,
Oxford.
Romaine, S. (2002), ‘The impact of language policies on endangered languages’,
International Journal On Multicultural Societies 4: 194–212.
Sasse, H.-J. (1992), ‘Theory of language death’, in M. Brenzinger (ed.), Language Death:
Factual and Theoretical Explorations, with Special Reference to East Africa, Berlin: 7–30.
Savvidis, A.St. = Σαββίδης, Α.Στ. (1973), Ἀρχαϊκὰ φθογγολογικὰ στοιχεῖα τῆς
ποντικῆς [sic] διαλέκτου καὶ τῶν ἄλλων νεωτέρων, Athens.
Shirokov = Щироков, C. (1972), ‘Грeчecкий язык’, Прохоров, A. M., Большая Советская
Энциклопедия. Москва, 7: 317-18.
Spolsky, B. (2004), Language Policy, Cambridge.
Thomas, G. (1991), Linguistic Purism, London and NewYork.
Tombaidis, D.E. (1988), ‘Η ποντακή: γλώσσα ή διάλεκτος;’, Ο Ξενιτέας, 5: 19–22.
Tombaidis, D.E. (1996), Μελετήµατα Ποντιακῆς διαλέκτου, Thessaloniki.
Trudgill, P. (2001), ‘The Ausbau sociolinguistics of Greek as a minority and majority
language’, in A. Georgakopoulou and M. Spanaki (eds.), A Reader in Greek Sociolinguistics,
Oxford: 23–40.
Tzartzanos, A.A. = Τζάρτανος, Α.Α. (1934), Τὸ γλωσσικὸ µας πρόβληµα – πῶς

εµφανίζεται τώρα καὶ ποιὰ εἶναι ἡ ὀρθὴ λύσις του, Athens.
Tzitzilis, H. = Τζιτζιλής, Χ. (2000), ‘Νεοελληνικές διάλεκτοι και νεοελληνική
διαλεκτολογία’, in Α.-Φ. Χριστίδης (ed.), Η ελληνική γλώσσα και οι διάλεκτοί
της, Thessaloniki.
Wald, B. (1985), ‘Vernacular and Standard Swahili as seen by members of the Mombasa
Swahili speech community’, in N. Wolfson and J. Manes (eds.), Language of Inequality,
Berlin and New York: 123–43.
Part II
Standardization Practices
5
The Lexica of the Second Sophistic:

Safeguarding Atticism*
Claudia Strobel
Socio-political influences on language call for some form of prescription to regulate

linguistic change and development, which is then codified by dictionary-makers
and grammarians. The dictionary-makers continuously monitor this filter, and
subsequently grammarians describe and often prescribe the extent to which new
uses are treated as acceptable. An early and archetypal example of the process
is provided by the Atticism of the Second Sophistic, and in particular the
developments of the second century AD. This chapter will focus on the Atticist
lexicographers of that period and their lexicographical methods.
The phenomenon of Atticism needs to be understood against the background
of the linguistic and political developments of the centuries that preceded it. The
stronger and the more influential Athens became during the fifth century BC,
the more widespread the Attic dialect became. The Macedonian kings and their
successors accepted Attic as a public language, which helped to spread it even
further during the fourth century BC. During the Hellenistic period, however,
various dialect features were adopted (especially from Ionic), with the result that
certain typical Atticisms (like ττ, instead of Ionic, and most Greek, σσ) were
replaced by ‘easier’ or less idiosyncratic forms (so that ττ now gave way to σσ).
The product of these developments, the so-called koine, was both spoken and used
for written texts, as by Polybius and later by the writers of the New Testament.
Atticism was a counter-movement, in reaction to the koine, which got under way
in the late Hellenistic period and was at its height during the Second Sophistic.1
Under the Empire, however, the koine as a public medium largely succeeded in
superseding the individual dialects other than Attic itself.
* I owe great gratitude to Professor E.L. Bowie for his patience, suggestions and advice. Further
information on the Atticist lexicographers can be found in Strobel (2005).
1
On the development of Atticism (and its difference from the koine) see Swain (1996) 27–33 (on
the political and social meaning of Atticism), Horrocks (1997) 78–86, and Schmitz (1997) 67–83.
93
94 claudia strobel
Over these centuries, earlier sociolinguistic boundaries were torn down by the
koine, only to be re-erected by the renewed worship of Attic by the educated elite.
From the epoch of Alexander, the new ‘common dialect’ had been developed for
the purpose of government and administration.2 Palmer nicely characterizes its
rationale: ‘Intense communication tends to produce uniformity of language within
the limits of the social unit which uses it. On the other hand, differentiation of
speech results from barriers to, and interruptions of, such intercommunications.’3
Throughout the fourth century BC, and even more between the third and the
first centuries BC, the koine attained the status of a commonly recognised language
in the Greek world by adding elements from elsewhere to a base version of Greek
that was Attic both in syntax and vocabulary, with the major contribution coming
from Ionic:4 the verb form πράσσω, as a pragmatic compromise between Attic
πράττω and Ionic πρήττω, is representative here.5 It is the communis opinio that,
whereas the koine indeed derived mainly from Attic, the various dialects displayed
minor variations on the ‘common’ language and that these variant versions retained
some importance in smaller communities (for instance, in local administration and
in the law-courts).6 However, the ‘non-native’ Greek speakers of a unified empire
required a common language:7 a language into which colloquial expressions and
usages could and did find their way, both on the level of everyday speech and of
2
Cf. Meillet (1965) 241 on the problems of defining particular linguistic developments in the koine
as spoken language: ‘Les linguistes modernes, qui s’intéressent à la langue parlée plus qu’aux langues
littéraires, entendent volontiers par κοινή la langue parlée en Grèce, depuis l’époque d’Alexandre
environ, et qui était comprise partout où l’on parlait grec. Comme de toute langue parlée, on n’a pas
de témoignages directs de cette langue. Des textes écrits par des gens peu lettrés, notamment certains
papyrus trouvés en Égypte et la plus grande partie des ouvrages qui composent le Nouveau Testament,
en donnent une idée.’
3
Palmer (1980)174.
4
To all intents and purposes the vernacular koine is the later vernacular Attic as it developed
in the historical environment created by Alexander’s conquests. On this base there were deposited
varied influences from the other dialects, but not enough to change the essential Attic character of the
language: Schlageter (1910), Robertson (1934) 71. For Higgins (1945) 93, however, the koine seemed
to be purely Attic-Ionic without any influence from other dialects. Palmer (1980) 176 comments:
‘Profound linguistic consequences might have been expected from the adoption of what was basically
the Attic dialect by users of not merely non-Attic, but non-Greek speech. In fact the changes were
remarkably slight. In phonology certain tendencies already observable in the ancient dialects persisted.
In the morphology there was simplification and systematization which ironed out a number of Attic
idiosyncrasies. But the main result of the immense extension of Macedonian power and the use of the
language for all the purposes and occasions of life in a world empire . . . was a great enrichment of the
lexicon.’
5
Significant linguistic changes, as between Attic and the koine, included itacism, together with
‘the change from a pitch to a stress accent and the loss of the phonemic distinction of quantity in the
vowel system’: Palmer (1980) 177.
6
Cf. Anlauf (1960) 34: ‘Im griechischen Mutterlande war die Koine die Sprache der Gebildeten;
sie sollte aber auch den geschäftlichen Verkehr regeln, wobei sich allerdings zeigte, daß gerade lokale
Behörden am stärksten an den heimischen Dialekten festhielten. Das einfache Volk gebrauchte seine
Dialekte weiter.’
7
Cf. Anlauf (1960) 38: ‘Eine Gemeinsprache . . . muß in ihrer Substanz einheitlich sein.’
the lexica of the second sophistic 95
more formal written language;8 a medium more practical than academic, adjusted
to clarity rather than eloquence; grammatically simplified, and easier in sentence-
construction.9 In response, the Atticists sought to establish a new, ‘purified’,
intellectual language, primarily for writing, one that would distinguish (on the
one hand) the literature and communication acts of the elite from (on the other)
those which used the common koine, the dialect now established for supra-regional
commercial and administrative communication.10
What happened in the Atticism of the Second Sophistic was the imitation of
the Attic dialect of the classical period, with strict exclusion of elements from other
dialects. As such, Atticism was a straightforward expression of language purism11
– which is, indeed, how the Atticist lexicographers understood it. Among the side-
effects of the movement was a revival of interest in the ten canonized orators of the
late fifth and fourth centuries BC, because their Greek, both in language and style,
was considered to represent a desirable standard;12 and a prerequisite for the revival
of the Greek of those centuries was that anyone who wanted to ‘converse in the old
and proper way’ should observe ‘the following’ specified usages (ὅστις ἀρχαίως
καὶ δοκίµως ἐθέλει διαλέγεσθαι τάδ᾿ αὐτῷ ψυλακτέα), as Phrynichus puts
it in the introduction to the first entry in his Attic lexicon. The imitatio of classical
authors and their styles was pursued enthusiastically,13 and rules were formulated
for syntax, vocabulary, morphology and stylistic mechanisms that aimed at the
preservation of the Attic language and its grammatical rules, as it had been spoken
and written in the fourth century BC. Needless to say, this Atticism could never
produce a perfect copy of classical Attic Greek, which anyway had its own internal
variations.
Throughout the history of Greek literature a distinction had been drawn between
written and spoken Greek. Many genres, especially poetic genres, were rooted in
a regional tradition and, therefore, in regional dialects.14 Dialects were artificially
merged in a Kunstsprache peculiar to each genre of poetry with each genre having
a different blend of Kunstsprache. The choice of dialect in prose of the classical
period also relates to genre, although it is not always easy to explain. Ionian writers,
8
Changes in vocabulary are too numerous to list here: there are many shifts in the meaning of words
and in the frequency of their usage. Some examples are given by Gingrich (1954).
9
Robertson (1934) 71 states the basic differences between classical Greek and the koine succinctly.
The koine was the language of everyday life rather than literature, more practical than academic, putting
the stress on clarity rather than eloquence. Its grammar was simplified, exceptions were decreased and
generalized, inflections dropped or harmonized, and sentence-construction made easier.
10
A task for the future is the exploration of the language of significant figures like Pollux, an imperial
secretary and advocate of Atticism himself.
11
‘Language purism’ is used as equivalent to ‘Atticism’ by e.g. Swain (1996) 20.
12
Cf. the advocacy by Dionysius of Halicarnassus in the late first century BC, which calls for
further discussion elsewhere.
13
Frösén (1974) 113: ‘Imitation of classical models, µίµισις or imitatio, made the literature of the
classical period known among the educated classes.’
14
Or, as Anlauf (1960) 33 puts it, ‘Epos, Lyrik, Tragödie usw. haben ihre Kunstsprache’.
96 claudia strobel
however, had been the first to create an artificial language for prose writing,15 and
the obvious question arises: if Ionic was the literary language before Attic took over
that role at the end of the fifth century BC, why now go back to Attic and not to
Ionic? What made Attic more desirable as a medium of elite communication than
any other obsolete version of Greek?
In principle, the Greek chosen could well have been Ionic (one thinks of
Lucian’s De Dea Syria and De Astrologia in imitation of Herodotus),16 and one
needs to ask why Attic took over any elements of Ionic in the first place. Among
the few modern scholars to provide any detailed insight into these issues is Swain,
who comments: ‘Apart from population movements and general contacts, Ionic
Greek’s possession of an old and varied prestige literature in the arts and sciences
must have been a factor in its influence over Attic.’17
The Attic dialect had originally been the language of Athens at its greatest
moment of political power. During the Second Sophistic the Greek elite tried to
maintain political independence as far as was possible under Roman rule. Language
became a way of showing their ‘Greekness’ or ‘cultural identity’.18 By choosing the
Attic dialect, they were referring back to Athens at its political and cultural peak.
The Athens of the fifth and fourth centuries BC had produced some of the greatest
of all Greek writers and orators. By scrupulous imitation of their literary Attic,
educated Greeks of the Second Sophistic could evoke their achievements, along
with a tradition of Greek political power.19
The history of both the koine and Atticism suggests that a substantial proportion
of the educated Greek population embraced the view that certain usages were
wrong, and even brought discredit upon those who used them, and accepted Greek
in the form of the koine or classical Attic as a regulating ideal. Unfortunately, there
were no regulatory bodies or dictionaries for the koine, which makes it more difficult
for us to define its origins, influences and linguistic peculiarities. In default of any
formal bodies charged with responsibility for ‘correct Attic’, the main agencies
15
Used by Herodotus (himself from a Doric-speaking region), by Antiochus (from Syracuse), by
Hippocrates (another presumed Doric-speaker) etc.: cf. Anlauf (1960) 33. Dover (1997) 79–95 seeks
to show that early Greek prose is a largely Ionian phenomenon; Rutherford (1881) 3 showed long
ago, with copious examples, that ‘the basis of the language of Tragedy is the Attic of the time when
Tragedy sprang into life’, a time when the Attic and Ionic dialects were less sharply differentiated from
one another.
16
Cf. the insightful discussion by Lightfoot (2003).
17
Swain (1996) 18.
18
Whitmarsh (2004) 23–5 has a helpful discussion of ‘Greekness’ and its different degrees. He
correctly points out (ibid. 24) that it was possible for a barbarian to obtain a certain degree of ‘Greekness’
through paideia. The lexica of the Second Sophistic would surely have helped the non-native Greek
speaker to attain such knowledge of Attic.
19
Compare and contrast Dihle (1994) 56–7: ‘it was the general mood of the period which was
the most important reason for the success of the Atticist movement . . . Atticism as an educational
programme focusing on the image of an ancient, free Athens suited the traditional policy of the Romans
towards the Greek upper class and made Rome appear as supporting a return to the best cultural
traditions of the Greeks.’ This stress on the Greek desire to maintain that ‘image of an ancient free
Athens’ surely supports my own position.
for prescription (the observance of rules) and codification (the formal setting
out of rules) were teachers, text-books and reference books; and these included,
in particular, dictionaries and manuals of usage, composed by lexicographers
and grammarians. The Second Sophistic seems to have started or re-inforced a
distinctive version of Atticism with distinctively ideological aims and motivations,
which helped to make the case for Attic by stressing its new users’ Greekness and
by encouraging the search for a common cultural identity with their ancestors,
with imitation of the ten orators and other select Attic authors the characteristic
means. Although Atticism had been designed originally for ‘literary prose and
rhetorical instruction’,20 it had a wider influence on all prose forms of the written
language. In the process, Hellenistic Kunstprosa was slowly but surely replaced by
various degrees of imitatio of those Attic authors who were considered to be role
models for literary prose.
Here, it is important to differentiate grammatical Atticism, which was to
influence the written language throughout all the genres of prose, from a version of
Atticism that was seen as a stylistic position in rhetoric.21 Imitation of a particular
author as a role model was nothing new, but grammatical Atticism went one step
further in demanding preservation of features of the ancient Athenian authors even
‘on the most basic linguistic level’.22 Not only was the style of the ten great orators
to be imitated, but so too was their vocabulary, syntax, and grammar – with the
result that, for example, the now defunct optative had to be re-introduced.23
The rationale of the two kinds of Atticism was quite different: stylistic Atticism
was a reaction against Asianism and targeted bombastic Hellenistic eloquence,
whereas grammatical Atticism, as we have seen, was a linguistic movement with its
origin in socio-political change. And it is in response to the needs of this movement
that the first Atticist dictionaries were written in the first century BC, beginning
with the work of Caecilius of Caleacte. Such dictionaries accompany the entire
development of Atticism, from Harpocration to Phrynichus to Moeris. The ideals
of Atticism were not uncontested, as one can tell from such works as Lucian’s
Λεξιϕάνης24 and Δίκη ϕωνηέντων, from criticism by the Antiatticista, and
from Galen’s critical comments about Atticism. Not only was there disapproval
of the favouring of Attic over other dialects or over the koine; Atticism was also
only practised by the educated elite and probably not even by all its members.
As a consequence one must think of this period as a multilingual era, with several
versions of Greek running in parallel: the koine, revived Attic, and (as some scholars
argue) an intermediate version which the grammarian Moeris called ‘Hellenic’.25
20
Dihle (1994) 55.
21
Swain (1996) 22 points out the differences between ‘stylistic’ and ‘grammatical’ Atticism and
emphasizes the influence the latter had on the world of Greek rhetoric.
22
Dihle (1994) 53.
23
On the use of the optative in late Greek, cf. Higgins (1945) and Anlauf (1960).
24
See Lucian, Lex. 20, where Attic is treated as ‘tongue-gymnastics’.
25
So Anlauf (1960) 48, on Moeris’ threefold division into Ἀττικόν, Ἑλληνικόν, κοινόν.
98 claudia strobel
The arguments for and against Atticism varied in specificity and in kind. At one
extreme, Galen makes a point of not writing for those who Atticize, but for those
who understand the importance of clarity and a pure Greek,26 and he insists that
his medical lexicon has the purpose of explaining the senses of ‘Greek’ – not Attic –
words.27 At the other extreme, we find Aelius Aristides celebrating the Attic dialect
as part of his encomium of Athens.28 Why was Atticism a favourite with some and
not others? Arguably for the socio-political reasons already cited: Greek aristocrats
sought to use language as a means of political self-assertion within, and against, the
Roman empire.29 The koine, it may be, gave them the feeling of a loss of control
over their own language: if the koine was in some sense a ‘common’ form of all the
dialects, might it even be influenced by Latin?30 The attempt of many Greeks to
‘purify’ their language, even prospectively the spoken language, by going back to
its roots in Attic was essentially an attempt to recreate the language of Athens at its
moment of supreme authority and power.31
This linguistic background produced a market for Atticist lexica. And that is
where the lexicographers of the Second Sophistic come into play.
Phrynichus
Phrynichus,32 no doubt, would never have considered himself a lexicographer (and
certainly not a drudge).33 His profession was that of a grammarian with a particular
interest in words. Two pieces of his work have survived more or less intact: the
Ecloga and the Praeparatio Sophistica.
26
See Galen’s Third Commentary on Hippocrates’ Prognostic, cviii b, 307.14–308.3.
27
Galen, On the Order of His Own Books, xix. 60.11–61.20.
28
Cf. Anderson (1993) 87–8.
29
See the influential work of Schmitz (1997) on this topic.
30
In Adams, Janse and Swain (2003) 78–9, F. Biville notes that during the Roman era Romans spoke
Greek and Greeks spoke Latin, and (ibid. 100) points to a Greco-Latin lexical koine with reference to
Isid. Ed. 20. 8. 3. For the graphemic influence of Latin, cf. Adams (2003) 45, who also discusses the
difficulties of determining whether Latin influenced Greek or Greek Latin (ibid. 427).
31
On the problematic nature of the attempt to mimic Attic in everyday life, cf. Anderson (1993)
88–9. Swain (1996) 21, though essentially correct, simplifies the issues: ‘The importance of atticism in
language and literature is simply that language was the best way to reproduce the past in a culture that
placed such enormous value on the classical heritage and on oral communication.’
32
What is known about Phrynichus’ life is known through Photius (Bibl. Cod. 158). Phrynichus
lived in the reign of Marcus (161–80) and Commodus (180–92) at the end of the second century AD.
Phrynichus’ Ecloga is dedicated to Cornelianus, the imperial secretary, and the Praeparatio Sophistica
is dedicated to Commodus. In its original form the latter was the longer work (the Suda speaks of 36
Books).
33
Samuel Johnson defines the lexicographer in his preface to A Dictionary of the English Language
(1755): ‘I knew that the work in which I engaged is generally considered as drudgery for the blind, as
the proper toil of artless industry; a task that requires neither the light of learning, nor the activity of
genius, but may be successfully performed without any higher quality than that of bearing burdens with
dull patience, and beating the track of the alphabet with sluggish resolution.’ One of his classic entries
reads: ‘Lexicographer: a writer of dictionaries, a harmless drudge . . .’
The introductory dedication of the Ecloga to the imperial civil servant

Cornelianus states the aim of this work. After praising Cornelianus as an educated
man, who seems to have asked for a collection of all ‘improper usages of language’
(τὰς ἀδοκίµους τῶν ϕωνῶν), Phrynichus explains why such a collection would
be useful as a repository of usages that ‘have disappeared from the old language’
(ἀποπεπτωκότες τῆς ἀρχαίας ϕωνῆς). The purpose is defined briefly and,
seemingly, precisely: ‘we wish to concentrate . . . on the most proper elements
of the language as used by the ancients’ (ἡµεῖς δὲ . . . ἀϕορῶµεν . . . πρὸς τὰ
δοκιµώτατα τῶν ἀρχαίων).34 The modern scholar, though, will want a more
specific definition of οἱ ἀρχαῖοι (‘the ancients’) and the adjective δόκιµος
(‘proper’). Whom does Phrynichus count among ‘the ancients’? Which authors
qualify? What exactly does he consider to be ‘proper’ and ‘improper’ Greek?
Which ancient authors he prefers does become more obvious after a reading of the
whole Ecloga, but the definition of ‘proper’ seems to be essentially subjective, as a
measure of the language spoken by his contemporaries.
Greek, insists Phrynichus, must be used ἀρχαίως, ‘in the old way’, which
means ἀκριβῶς, ‘in the correct way’. In relation to the kind of language he is
promoting, these key words stand in striking contrast to the opposite and detested
sort of language, which he characterizes by νεοχµῶς and ἀµελῶς: in effect,
‘in an innovative, modern way’ and ‘carelessly’. The first of these two adverbs,
νεοχµῶς, is most often found in poetry and Ionic prose,35 but does not occur in
classical Attic prose, and a search of the literature of the second century AD in the
Thesaurus Linguae Graecae has not yielded any instance apart from this one passage
in Phrynichus (though there is a single, distantly comparable usage in the much
earlier works of the Ionic-dialect Hippocratic corpus).36 One might be tempted to
conclude that Phrynichus himself was an Atticist who made mistakes, especially
as there was not yet any formal authority on the Attic language in his period, and
Atticizing involved a measure of individual choice. Or maybe Phrynichus used this
word in order to distance himself from the new way of speaking by paradigmatically
relating the ‘new’ idiom to a ‘new’ word.
At the end of his introduction, at all events, Phrynichus justifies his undertaking
by giving the correct usage of language an aesthetic and almost moral point, in
the form of a brief and memorable statement: οὐ γάρ τις οὕτως ἄθλιος, ὡς τὸ
αἰσχρὸν τοῦ καλοῦ προτιθέναι (‘for no one is so wretched as to set the ugly
before the beautiful’). Anderson notes that Phrynichus ‘censures Favorinus for
using σύµπτωµα in the sense of συντυχία, while he himself acknowledges an
instance in Demosthenes’,37 and points to three other examples where Phrynichus’
34
Note his humorous criticism of Cornelianus in entry 203, from which we can infer that
Cornelianus was a less strict Atticist; cf. Rutherford (1881) 306. Here too we find the adjective
ἀδόκιµος for ‘improper’ (incorrect) usage.
35
The adjective can be found in the tragedians of the fifth century BC.
36
De mulierum affectibus, i.16.18. Here it means ‘again’ or ‘in a new way’ and not ‘in a modern/
innovative way’.
37
Anderson (1993) 92; Phrynichus, Ecl. 223. Anderson’s book is illuminating on the lives of
100 claudia strobel
allegations are unfounded, by way of illustrating the tension and rivalry between
the scholars of this period.38 We have indeed a surprisingly large number of lexica
from the second century AD, and it would be even more surprising if there were
no rivalries. Given that there was no authority for Atticism as such, each of these
lexicographers tries to establish such authority through his own work.
Compared with Pollux, for instance, Phrynichus was clearly an Atticist of the
strictest kind. His endeavour was to purify language by reviving Attic as it had been
written by Plato, Demosthenes, the other nine orators, Thucydides, Xenophon,
Aeschines Socraticus, Critias, Antisthenes, Aristophanes and other poets of the
Old Comedy and finally the three tragedians.39 Pollux, as we shall see, included
a wider variety of ancient authors, and thus seems to have represented a more
moderate form of Atticism.
The original shape and structure of the Ecloga will probably remain a mystery, and
the fact that our editio princeps is in alphabetical order confuses rather than clarifies
the issue. There are traces of alphabetical subordination within sequences,40 which
might indicate copying from other lexica.41 We find parts of the Ecloga quoted by
the Antiatticista where the quoted entries are arranged in alphabetical order – ε-µ
(307–20), α-θ (326–345), ρ-χ (348–70) – or once in reverse alphabetical order
(372–409).42 Nevertheless, it is noteworthy how many thematic areas Phrynichus
covers with his codified vocabulary. Word groups are detectable that are suggestive
of various sub-classifications: domestic, sympotic, comic, botanic and zoological,
medical and scientific. It is a very diverse collection of words.
Phrynichus’ plan for his Ecloga, however, is beyond detection, as the full text
has been lost: the structure that has survived resembles a notebook or draft for a
later composition, which is filled with ideas on the correct usage of language. But
even if the original structure could provide an alphabetical or thematic order for
us, Phrynichus does not seem to be aiming at either a grammar or a lexicon, but
rather at a selective combination of both. In its current state, it is hard to imagine
anyone using this book as a dictionary. Was the reader meant to plunge through
from the first page to the last? Or to cultivate selective reading or casual browsing?
Alphabetization would turn the work into a handbook, but even a thematic order
would help the reader to find his or her way around.
Phrynichus and Pollux, but only names Moeris (ibid. 91) without further comment.
38
Anderson (1993) 91–2.
39
Extremely condensed but precise observations on Phrynichus’ likes and dislikes in previous
authors and contemporaries can be found in Swain (1996) 53.
40
Cf. Strout and French (1941) 922: ‘Die von Phrynichus in der Anordnung der Artikel in der
ἐκλογή angewandte Methode festzustellen, ist nicht einfach. Es gibt Spuren einer alphabetischen,
einer sachlichen und einer Anordnung nach den von Phrynichus nachgeschlagenen Schriftstellern.’
Cf. Schöll (1893) 516 ff.
41
Rutherford (1881) distances himself even further than Lobeck (1820) from earlier editors in
rearranging the order of the different articles into subject-related groups.
42
See Fischer (1974) 46.
The Ecloga, one is left to assume, seeks to be easy on the reader (given its
apparently non-alphabetical, albeit perhaps thematic, order); to be prescriptive
in offering codified terms (Phrynichus has made a decision about what counts
as correct usage for the reader and, furthermore, offers explicit rules); and to be
etymological and encyclopedic (in that it offers more information on selective
words than is required for mere simplification or purification of usage).
In its original form, Phrynichus’ Praeparatio Sophistica was the longer of the
two surviving works,43 and is certainly the more elegant. In this text Phrynichus
advertises certain elegant, witty or especially descriptive words and phrases to the
rhetor of the Second Sophistic. He illustrates these words and phrases through
quotations taken from Attic writers of the fifth and fourth centuries BC. As well
as standard authors, he quotes Cantharus, Crates, Cratinus, Diphilus, Eupolis,
Pherecrates, Plato Comicus, Strattis and other comic poets, and (as noted by de
Borries) he frequently cites Homer.44 Thus we see a wide variety of authors chosen
for this work by comparison with the Ecloga.
The Praeparatio Sophistica contains several descriptive elements, as the book
sets out to show the correct usage of particular sententiae and phrases. However, it
is more prescriptive than descriptive, in that it helps with codification, but hardly
ever offers explanations for the usage prescribed. The book simply suggests phrases
for a sophist to use, but neither delivers general advice nor offers general rules on
how one might use such phrases. By choosing a body of authoritative material,
Phrynichus aids the process of standardization and consequent prescription. There
are many ways to compile a dictionary: ‘You may record words that are heard. You
may copy the words from other existing dictionaries. Or you may read, after which,
in the most painstaking way, you record all the words you have read, sort them, and
make them into a list.’45 It is impossible to say for sure which methods Phrynichus
might have used, but it seems likely that he read other dictionaries (as is implied
by the vestigial alphabetical order), that he included words he had read and found
interesting (an example is ἀσϕάραγος: ‘asparagus’), and that he included words
he had heard used in the wrong way, which he felt he ought to mention together
with the correct usage. As in the Ecloga one fails to find any principle of order.
Moeris
The case of Moeris is very different. His lexicon is hard to date, but I tend to agree
with Swain, who puts it in the early third century AD, as ‘it was apparently influenced
by the views of Phrynichus’.46 Moeris’ choice of whom to quote and whom to
avoid (which includes the tragedians) hints at a much more differentiated, or rather
more critical, understanding of Atticism, and would seem to imply a relatively late
43
All that survives is an epitome in a manuscript (Coisl. graec. 345) of the tenth century AD. This
epitome is in alphabetical order.
44
De Borries (1911) xxiv.
45
Winchester (1999) 83.
46
Swain (1996) 51.
102 claudia strobel
date within the Second Sophistic. Moeris made a significant contribution to the
codification and prescription of Attic usage. His work is an apparently random
selection of Attic words, and must have been designed to help writers who wanted
to look up individual lexemes. The alphabetical order according to the first letter
seems to be original, as Photius indicates (κατὰ στοιχεῖον δὲ καὶ τοῦτο τὸ
πονηµάτιον),47 but apart from this listing no other structure is recognisable. His
work should strictly be called a dictionary or a glossary, not a lexicon in the modern
sense, which implies something closer to an encyclopedia than a mere word list.
As with Phrynichus’ Ecloga, explanations of syntax, cases, declensions and the
like can be found in Moeris’ work, but certainly not in the same quantity or variety.
Unlike Phrynichus, Moeris does not criticize: rather he always puts the ancient
word first, followed by a more recent alternative.48 Throughout, the term ‘Greek
speakers’ (Ἕλληνες) is used in an extremely broad sense, serving as the opposite
to ‘Attic speakers’ (Ἀττικοί) but without a strong negative connotation that it has
in Phrynichus’ Ecloga, where it is implicit that Phrynichus has formed a negative
judgement of those who do not use Attic words.
Moeris quotes fewer authors than Phrynichus: Plato, Aristophanes, Thucydides,
Xenophon; the orators, Demosthenes, Antiphon (once), Hypereides (once) and
Isaeus (once); there is Old Comedy, but there are no tragedians. Moeris both knew
of Phrynichus’ works and used them for his own lexicon,49 but he does not copy
them.50 He aims at a different layout, that of a dictionary. He keeps his instructions
to the reader to a minimum. Like a true dictionary, his work is in alphabetical
order, and offers synonyms. By pointing out the current usage of a word and its
Attic equivalent, he gives the reader the clear impression of a prescriptive lexicon.
This is underlined by not only citing the Attic usage but by comparing it with
‘Hellenic’ usage as well. ‘Hellenic’ here must mean the language spoken, or maybe
even written, by the majority of Moeris’ contemporaries. This interpretation is
slightly complicated by his third category, the ‘common’ form (designated by
κοινόν or κοινῶς or the like). As Swain observes, these ‘common’ forms often
coincide with those of educated speakers – ‘hence “common” and “Hellenic”
could be identified with and, indeed, “common” alone (presumably signalling
“common” and “Hellenic”) could be contrasted with, “Attic”’.51 Swain adds: ‘in
referring to “common” forms, either in vocabulary or grammar, as distinct from
both “Hellenic” and “Attic”, Moeris is of course stigmatizing the language of those
who used them in respect of both competence and performance.’52
47
Phot. Bibl. Cod. 157.
48
Exceptions are 195, 23; 200, 18; 202, 25; 210, 17; 211, 24; 212, 5; 213, 15.
49
Cf. Fischer (1974) 39: ‘Moeris kannte und benutzte also beide Werke des Phrynichus, deren
Glossen er teilweise bis zur Unkenntlichkeit verkürzte.’
50
It would be interesting to compare Moeris’ lexicon with its sources (Aelius Dionysius, Philemon,
Pausanias, Diogenianus, Ammonius and Philetairus), but that is not possible here. One should note, at
least, that Moeris’ brevity is unique.
51
Swain (1996) 52.
52
Elsewhere I propose to discuss the difficulties of this view and of deciding what exactly Moeris
It might be said that in Moeris brevity is taken too far and has obscured clarity.
Or maybe it is only the modern scholar who finds it hard to see what is meant by
‘Hellenes’ and ‘users of Attic’, and precisely who is using the ‘common’ form. A
preface to Moeris’ short dictionary would have been of immense value.
Pollux
In his ten-book Onomasticon, Pollux, pupil of Hadrian of Tyre, uses yet another
procedure: he sets out a thematic list of words and their corresponding synonyms
with the occasional explanatory excursus. In some of his books, Pollux clearly has
certain texts in mind, and the detection of sources and quotations or allusions to
other texts can help the modern reader to get an insight into Pollux’s principles
of composition. Throughout the whole Onomasticon the influence of Phrynichus’
Ecloga is apparent.53 Book 4 has many words corresponding to those found in
Athenaeus’ books 4, 10 and 14. The fifth book is full of traces of Xenophon (in
the first half mainly the Cynegeticus, in the second half the Apology).54 The sixth
book covers the word-field of symposia, and unsurprisingly is filled with allusive
references to the dialogues of Plato. Book 7 seems to have influenced Hesychius, a
grammarian from Alexandria (c. fifth century AD), and other lexicographers, such
as Moeris. In Book 8 Pollux covers the classical Athenian political system, a field
that he obviously feels confident writing about: it contains hardly any citations
and only a few authors are mentioned by name (Plato, Solon, Demosthenes, et al.).
He did not need any other sources for his chosen words on politics and the judicial
system – πολιτικὰ ὀνόµατα Ἀττικά. Either his readership was acquainted with
the political terminology of the fifth and fourth centuries BC (perhaps through
familiarity with some of the political orations of that period) and did not require
more guidance in the form of references, or else he felt himself an expert in the field
who need not prove his authority by supplying citations. That Pollux was highly
educated and respected by his contemporaries is confirmed by Philostratus, who
informs us that he was awarded the chair of rhetoric in Athens.55
For all that, one might ask how useful these Attic terms of the fifth and fourth
century BC were for the second century AD. We can safely assume, at least, that
they were perceived as authoritative for exercises in declamation and literary
epideixis, though they might not have been so useful in real-life spoken contexts,
such as the courtrooms or council meetings of Pollux’s own day.
In his first six books, Pollux is generally very careful not to quote word for word.
Instead, he spices up his writing with allusions to passages and chosen instances of
meant when talking about ‘Attic’, ‘Hellenic’ and ‘common’.

53
Cf. Pollux 2.17 (παιδίσκη) and Phryn. Ecl. p. 239 Lobeck.
54
In his apparatus Bethe (1967) 266 notes: ‘sed Pollux Xenoph. sententias numquam ad verbum
exscripsit’ (‘Yet Pollux never quotes sentences from Xenophon word for word’).
55
Philostratus, VS 592–3: Pollux obtained the chair at Athens from Commodus through bewitching
the emperor; cf. Anderson (1993) 32. It may be, as has been conjectured, that Phrynichus was his rival,
but this is not confirmed by Philostratus.
104 claudia strobel
characteristic vocabulary, without in most cases citing in detail.56 His work is in

many ways a summary and compilation:57 he attempts to cover themes and areas
of life, declamation and general interest that might be useful for his readership. His
choice of words is a summary of existing terminology, although one has to admit
that he almost achieves completeness in giving so many choices of synonyms, while
the quotations are in a sense a summary of what he has read and what he expects
the reader to have read.
Pollux’s Onomasticon, though, is not only a highlight in lexicographical
scholarship. It is also a guide to the Second Sophistic, as the topics dealt with
shed light on the thematic preferences of those days, and in fact the whole layout
reflects the way of thinking of his time. The work is organized, but is not yet as
efficiently utilitarian as we would expect a modern lexicon to be: the reader is not
only to be informed but is also to be entertained and challenged. Atticist ideology
is apparent, but not determinative. Pollux’s selection of Attic authors is broader and
more liberal than Phrynichus’, first of all because he positions himself less strictly
as a scholarly Atticist, and secondly because his main aim in the Onomasticon
is not to give instruction in the correct usage of Attic. He promotes Atticism, of
course, but this is only part of his purpose. Phrynichus and Moeris write only to
promote Atticism, whereas Pollux manages to achieve a more rounded work of
lexicographical scholarship.
Conclusion
The forces examined in this chapter that favoured the restoration of ‘standard
Attic’ in preference either to some new artificial Kunstsprache, or to maintaining
the koine, were the same forces that helped to maintain the relative uniformity
of this standard in the educated Greek language of the second century AD.58 As
classical Attic had not been used in ‘real’ speech-situations for centuries, standards
of correctness had to be re-introduced in two ways: first, through a common, rule-
based grammar taught in schools (which presumably served not only children
whose native language was Greek, but also children in newly founded Greco-
Roman cities in non-Greek or only superficially Hellenized territories); and
secondly through dictionaries that embodied and reinforced such standards of
correctness.
The defining quality symbolized by the use of Attic during the Second Sophistic
is not social rank as such (though money buys education), but the educatedness
summed up by the Greek word paideia. Correct performance marks the user
as a responsible member of Greek elite society, because he makes an effort to
maintain a certain aspect of Greekness – whereas incorrect performance is viewed
as contributing to the decay of the language and an acquiescence in imperfect
56
Cf. Pollux 4.16–17 and Athen. 4.175e–f.
57
And this quite apart from the fact that the first books have only survived in epitome.
58
A similar phenomenon is observable in the twentieth-century lexicography of Modern Greek:
see Tseronis and Iordanidou, below, pp. 167–85.
Greekness. Although there was no official authority to regulate language in this

period, the educated elite seem to have sought such guidance and found it in the
rhetorical schools and the works of the grammarians and lexicographers:59 these
were indeed the guardians of the language,60 who recognised the importance of
language management and devoted their lives to language.
For the Greek language of the Second Sophistic, as for Greek and other
languages today, the mechanism of authority involves a filtering of usage acceptable
to the educated, and embodies the notion of prescription, which is then codified by
dictionary-makers and grammarians. The dictionary-makers continuously monitor
the filter, while the grammarians specify new uses and the extent to which they are
to be accepted .
Lexicographers of the Second Sophistic are not linguists in the modern sense.
Their objective is to create a beautiful, abstract, fitting design that they can put
forward as both intellectually coherent and practical. With these two goals in mind,
they recognise the importance of how best to organize a dictionary. The denotative
meaning of words becomes a matter of secondary importance; words, instead, are
measured in the light of psychological reality and impact – as a new era of language
seems to demand.
Many questions that meta-lexicographers have asked about the lexica of the
second century AD – questions about the ideology of standardization, the aims
and methods of any linguistic compilation, the objectivity of the lexicographer, and
the issue of prescriptivism versus descriptivism – are recurrent themes in linguistic
debate. Modern Greek dictionaries in particular have engendered debates on
similar topics, so that scrutiny of the second-century lexica may be considered as
one of two possible starting points for the discussion of meta-lexicographical issues
in the perspective of the Greek language over time.61 For Greek, as for the Western
tradition as a whole, purism and standardization can trace their origins to the revival
of Attic and the production of the Atticist lexica of the Second Sophistic.62
59
These lexicographers too, however, would have described themselves as grammatikoi.
Lexicographer is a modern term, and only in retrospect can we call such works as the Onomasticon
a ‘lexicon’. Linguistic awareness in the second century AD was vastly more limited than it is today;
and despite the fact that some Greeks obviously made a specific decision to use Attic and despite the
apparent effort of others to help by introducing handbooks and word lists, we cannot be sure how far
these grammarians/lexicographers could themselves identify the scope of their authority.
60
Already so named (‘Guardians of Language’) by R. Kaster in his book on the Roman grammarians
(1988).
61
The other starting point would be the consideration of Hellenistic lexica. The reasons for
composing lexica during the Hellenistic period, though, were very different, and included a concern
with the interpretion of poetic texts. Readers of poetry wanted to know more about the vocabulary used
in these texts, which helped to generate an interest in etymological and encyclopedic studies, and in the
late Hellenistic period a distinctive interest in dialects. The majority of titles and fragments of lexica
from this period imply specialist lexica that were guides to reading poetry.
62
The whole process of linguistic change covered several decades, during which there will have
been short-lived adjustments and compromises (one thinks of the way that different versions of the
Euro do and will coexist over a period of time). The Rechtschreibereform in Germany began a dozen
or so years ago, and users of German have still not found a consistent way of spelling certain words,
106 claudia strobel
References
Adams, J.N. (2003), Bilingualism and the Latin Language, Oxford.
Adams, J.N., Janse, M. and Swain, S. (eds.) (2003), Bilingualism in Ancient Society,
Oxford.
Anderson, G. (1993), The Second Sophistic, London.
Anlauf, G. (1960), Standard Late Greek oder Attizismus?, Cologne.
Anttila, R. (1973), ‘Linguistik und Philologie’, in R. Bartsch and T. Vennemann (eds.),
Linguistik und Nachbarwissenschaften, Kronberg im Taunus: 177–91.
Beyer, K. and Cherubim, D. (1973), ‘Linguistik und alte Sprachen. Eine Polemik?’,
Gymnasium, 80: 251–79.
Bethe, E. (1967), Pollucis Onomasticon, Stuttgart.
Bethe, E. (1937), ‘Zu Pollux’, Hermes, 72: 240.
Borries, I. de (1911), Phrynichos, Leipzig.
Brenous, J. (1895), De Phrynicho Atticista, Diss. Montpelier.
Brixhe, C. (ed.) (1993-2001), La Koiné grecque antique, 4 vols, Nancy.
Dihle, A. (1994), Greek and Latin Literature of the Roman Empire, tr. M. Malzahn, London.
Dover, K.J. (1997), The Evolution of Greek Prose Style, Oxford.
Erbse, H. (1950), ‘Untersuchungen zu den Attizistischen Lexika’, Abhandlungen der
Deutschen Akademie der Wissenschaften zu Berlin, Philosophisch-historische Klasse,
1950, vol. 2.
Fischer, E. (1974), Die Ekloge des Phrynichos, Berlin.
Frösén, J. (1974), Prolegomena to a Study of the Greek Language in the First Centuries AD,
Diss. Helsinki.
Georgacas, J. (1952), ‘A point of koine Greek lexicography’, Classical Philology, 47: 167–9.
Gingrich, F.W. (1954), ‘The Greek New Testament as a landmark in the course of semantic
change’, Journal of Biblical Literature, 73: 189-96.
Higgins, M.J. (1940-1), ‘Why another optative dissertation?’, Byzantion, 15: 443–8.
Higgins, M.J. (1945), ‘The Renaissance of the first century”, Traditio, 3: 49–100.
Kaibel, G. (1899), De Phrynicho sophista, Göttingen.
Kaster, R.A. (1988), Guardians of Language: The Grammarian and Society in Late Antiquity,
Berkeley and Los Angeles.
Koch, G.A. (1833), Observationes in Timaei Sophistae lexicon vocum Platonicarum et
Moeridis Atticistae lexicon Atticum a se nuper edita, Leipzig.
although the rules have been standardized in the Neue Duden. Then again, there will always be
differences recognisable between Gemeinsprache and Literatursprache, and it will also always be hard
to find common linguistic guidelines for both. Simon Swain, among others, has drawn a comparison
between the development of the koine, accompanied as it was by a desire for a more regulated written
language, and the fairly recent, and much-discussed, developments in Modern Greek: Swain (1996)
35–6. In modern Greece until recently, a somewhat standardized ‘demotic’ and a more ‘purified’
form of Greek, the katharevousa, have been competing for authority and acceptance. The project of
Adamantios Korais, to awaken in Greeks a sense of their ancient heritage, while, however, remodelling
the language on a spoken form of Greek (see Beaton, below, pp. 341–53), recalls (almost per contrarium)
the programme of the educated elite of the Second Sophistic. Neither Korais, however, nor that earlier
elite succeeded in changing the language overnight. Authorities for language had to be created through
a process of experiment and intellectual exchange, in which lexicography played a primary role. This
might help to explain why the lexica of the Second Sophistic show considerable differences even when
claiming to be propagating pure Attic.
Landau, S.I. (2001), Dictionaries: The Art and Craft of Lexicography, 2nd edn, Cambridge.
Lightfoot, J. (2003), Lucian: On the Syrian Goddess, Oxford.
Lobeck, C.A. (1820), Phrynichi Eclogae nominum et verborum Atticorum, Leipzig.
Maidhof, A. (1912), ‘Zur Begriffsbestimmung der Koine besonders auf Grund des Attizisten
Moeris’, in M. von Schantz (ed.), Beiträge zur historischen Syntax der griechischen
Sprache, Würzburg.
Matthews, P.H. (1997), Oxford Concise Dictionary of Linguistics, Oxford.
Meillet, A. (1965), Aperçu d’une histoire de la langue grecque, 7th edn, Paris.
Naechster, M. (1908), De Pollucis et Phrynichi controversiis, Leipzig.
Palmer, L.R. (1939), ‘Some late Greek ghost-words’, Classical Quarterly 33: 31–3.
Palmer, L.R. (1980), The Greek Language, London.
Pauly, A., Wissowa, G. et al. (1903–78), Paulys Realencyclopädie der classischen Altertums-
wissenschaft, Stuttgart and Munich.
Pfeiffer, R. (1968), History of Classical Scholarship: From the Beginnings to the End of the
Hellenistic Age, Oxford.
Robertson, A.T. (1934), A Grammar of the Greek New Testament in the Light of Historical
Research, Nashville, Tenn.
Rohde, E. (1870), De Julii Pollucis in apparatu scaenico enarrando fontibus, Leipzig.
Rutherford, W.G. (1881), The New Phrynichus, London.
Schlageter, J. (1910), Der Wortschatz der außerhalb Attikas gefundenen attischen Inschriften,
Freiburg.
Schmitz, T. (1997), Bildung und Macht, Munich.
Schöll, R. (1893), Die Ecloge des Atticisten Phrynichos, Sitzungsberichte der Königlichen
Bayerischen Akademie der Wissenschaften, Philosophisch-philologisch und historische
Klasse, 1893, vol. 2: 493–540.
Strobel, C. (2005), ‘The lexicographer of the Second Sophistic as the collector of words,
quotations and knowledge’, in R.M. Piccione and M. Perkams (eds.), Selecta Colligere II:
Beiträge zur Technik des Sammelns und Kompilierens griechischer Texte von der Antike bis
zum Humanismus, Alessandria: 131–57.
Strout, D. and French, R. (1941), ‘Phrynichos’, in Pauly–Wissowa, 20.1: 920–5.
Swain, S. (1996), Hellenism and Empire, Oxford.
Wendel, C. (1927), ‘Zu Moeris’, Philologische Wochenzeitschrift, 47: 1275–6.
Wendel, C. (1928), ‘Die Überlieferung des Attizisten Moeris’, Philologie, 84: 179–200.
Wendel, C. (1932), ‘Moeris’, in RE 15, 2: 2501–12.
Wentzel, G. (1895), ‘Beiträge zur Geschichte der griechischen Lexikographen’,
Sitzungsberichte der Königlichen Akademie, Berlin, 1895: 477–87.
Whitmarsh, T. (2004), Greek Literature and the Roman Empire, Oxford.
Winchester, S. (1999), The Surgeon of Crothorne, London.
6
Grammatical Metaphor and the Function

of Participles in High-Register Versions
of the Life of Aesop*
Chrystalla Thoma
Introduction
‘Grammatical metaphor’ is a term borrowed from systemic functional linguistics
(SFL), proposed by Halliday to cover certain constructions and mechanisms which
lead to information ‘packing’ in language. Grammatical metaphor is particularly
associated with scientific, expository discourse in a ‘syndrome’ called ‘things and
relations’, which is discussed by Halliday in a number of publications.1 Use of
grammatical metaphor is associated with questions of authority in text production,
since it contributes to producing highly ‘packed’, organized and ‘detached’ texts.
However, as this chapter will show, Greek scientific discourse has not always been
authoritative, while, quite surprisingly, narrative sometimes has. By contrasting the
formation of the Modern Greek scientific register and two ‘high’-register narratives
of the fourteenth century, it becomes clear that authority is a matter of public
and authorial expectations reflected in language choices, and that grammatical
metaphor in all its forms and guises is an effective means of measuring the authority
of a text regardless of discourse type and genre.
Here we focus on the use of perfective, mainly past-reference, participles as a
mechanism of grammatical metaphor, in the sense that they are atypical trans-
categorizations of verbal groups, aiming at lexical density, low grammatical
intricacy and effective managing of the distribution of new and contrastive
information – properties until now associated mainly with expository written texts.
* I would like to thank the Deutsche Forschungsgemeinschaft (DFG) for financial support through
the Research Centre 538: Multilingualism of the University of Hamburg. Special thanks also go to Prof.
Dr Hans Eideneier (University of Hamburg) and to Dr Notis Toufexis (University of Cambridge) for
their comments and help. My thanks also to Birsel Karakoc, Claudia Böttger und Julia Probst from
Research Centre 538 for fruitful discussion on the topic of infinite constructions.
1
Halliday (1994), (1997), (1998).
109
110 ChrystallA Thoma
In the present study we shall analyse and contrast late medieval and early
Modern Greek registers in their historical context; we shall also aim to place
grammatical metaphor in a new perspective, removing it for the first time from
scientific discourse and demonstrating its strong dependence on audience design
regardless of discourse type. Last but not least, we hope thereby to have contributed
to the placing of Greek in the wider frame of languages analysed in relation to
grammatical metaphor.2
Methodology and data

For the purposes of this study we rely on the framework of SFL. According to SFL,
language is represented as a system of choices, some of which are taken up and others
rejected with every instantiation of register and text. One especially interesting
implication of the system is that certain options are taken up systematically in
registers, offering the possibility of predictability.
In the framework of SFL, three metafunctions are seen as running parallel in
language: the experiential, the textual and the interpersonal.3 The experiential
metafunction construes human experience and is realized on the syntagmatic axis
through process, participant and circumstance, while on the paradigmatic axis it
is realized through clause types, further subdivided into process, participant and
circumstance types. We shall consider this function in greater detail below. The
textual metafunction creates discourse and is realized through information packaging
and managing (according to the distinction between given and contrastive/new)
and text structuring (according to the distinction between theme and rheme).
The interpersonal metafunction describes relations of hierarchy between author
and reader, the author’s attitude towards the subject matter, or relations between
characters in the text.
Here we focus on the experiential metafunction, although implications for the
textual and interpersonal metafunctions will also be discussed in connection with
the concept of grammatical metaphor. For the quantitative analysis we shall use
Halliday’s classification of process types, such as material (doing), mental (thinking/
feeling/perceiving), behavioural (behaving), verbal (saying) and relational (being/
having).4
The principal texts under analysis are two high-register versions of the Life
of Aesop from the fourteenth century, referred to here as ‘Aesop P’ and ‘Aesop
W’.5 They are marked by their use of vocabulary, morphology and syntax drawn
from classical Greek and the scriptures. These ‘high’ versions of Aesop’s fictive
2
For essential reading on grammatical metaphor see Halliday (1988), (1994), (1997), (1998);
Halliday and Martin (1993); Halliday and Matthiessen (1999); Ravelli (1985), (2003); Matthiessen
(1998); and, for Ancient Greek, Kappagoda (forthcoming).
3
For a comprehensive application of the three metafunctions to Modern Greek, see Thoma
(2006).
4
Halliday (1985).
5
Respectively, the Planudes version (in Eberhard (1872): see Karla (2001) 58–65) and the
Westermann version, Recensio II (= codd. S, B, P, Th: see ibid. 37) in Papathomopoulos (1999a).
grammatical metaphor and the function of participles 111
biography, a well-loved genre in the past, reflect their time of creation: they were
written for the educated few. In the Ottoman-occupied Greek-speaking areas,
between the thirteenth and fifteenth centuries, texts were almost exclusively
written in classicizing Greek, and few people could afford a classical Greek
education. Those who could formed a small elite, consisting mainly of members of
the upper clergy, who could actually read and understand texts written in the high
register. This situation resulted in the phenomenon of social diglossia: although the
language spoken by the common people was an early form of Modern Greek, the
written language was not based on it at all.6 The high register was an acquired and
cultivated language variety, a revived language and yet nobody’s native language,
the language of the upper social strata: a situation that brings to mind the well-
known diglossic situation – the ‘language question’ – of Greek in the nineteenth
and twentieth centuries.7
As a measure of comparison, three ‘low’ versions of the same texts from the
seventeenth century are also investigated: ‘Aesop D’, ‘Aesop I’ and ‘Aesop K’.
These texts are free translations (‘metaphrases’) of the high versions in a relatively
modern language.8 They belong to a period of change (sixteenth to eighteenth
centuries) when the reading public expands with the rise of the middle class: a public
with a basic, not a classical, education; a public which seeks out and purchases
texts written in a common Greek based on the spoken language of the time, and
consisting mostly of translations of earlier material, such as those under analysis
here.9 Comparison of these three low texts with the two high-register originals is
revealing.10 Although all five texts follow the same basic fictive biography of the
storyteller Aesop, versions D and I appear to be translations of version P, while
version K appears to be a translation of version W. In all versions there are slight
variations in the different episodes.11 Alongside these five versions, examples
from several early-modern Greek scientific texts are adduced here, by way of
comparison and contrast.12
6
See Toufexis (2008). ‘Diglossia is a relatively stable language situation in which, in addition to
the primary dialects of the language (which may include a standard or regional dialects), there is a very
divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large
and respected body of written literature, either of an earlier period or in another speech community,
which is learned largely by formal education and is used for most written and formal spoken purposes
but is not used by any sector of the community for ordinary conversation’: Ferguson (1959) 336.
7
Cf. Eideneier (1999), (2001).
8
Corresponding to versions Δ´, Γ´, Α´ in Papathomopoulos (1999b). For an analysis of both
registers, see Toufexis (forthcoming).
9
See Matschke and Tinnefeld (2001) and Eideneier (2001) for a description of the social changes
leading to the appearance of the low register.
10
For editions of the low texts, see Papathomopoulos (1999b) and Eideneier (forthcoming).
11
For an extended example of a corresponding text sequence in the five different versions, see the
Appendix (pp. 127–9 below).
12
Taken from online texts from the digital library Hellinomnimon (http://195.134.75.8/).
Grammatical metaphor and its syndromes

In order to understand the importance of grammatical metaphor, one needs
to introduce certain basic concepts from SFL. The experiential metafunction
encodes the world as the speaker sees it, and at its heart lies the ‘process’ which
is the dynamic part of the utterance, usually realized through the verbal group.
The process is commonly accompanied by participants: that is, entities which
bring about or are affected by a process. These are usually realized by nominal
groups. Participants are static but offer the possibility of grounding the utterance
in space and time. Two main categories of participants can be found: ‘things’ and
‘qualities’.13
In the same vein, Givón suggests that ‘linguistic categories like verb, noun
and adjective are conceptualizations of the varying time-stabilities of real-world
phenomena’.14 The category ‘noun’ thus lexicalizes concepts that persist for
relatively extended periods of time. The category ‘verb’, on the other hand, is a
lexicalization of phenomena relatively unstable in time: particularly ‘events’, the
prototypical ‘process as change’, as opposed to ‘states’.15 Verbs form the extreme
of the category ‘process’ and nouns of the category ‘participant’; and these two
extremes become blurred by means of grammatical metaphor, as we shall see.
‘Metaphor’ (as the name suggests in Greek) is taken to be a ‘transfer’ of some
kind. Grammatical metaphor is the replacement of one type of grammatical
construction (a ‘typical’ one) by another (an ‘atypical’) one. Halliday suggests
that participants (entities, things) are congruently realized as nouns, qualities
as adjectives, processes as verbs, properties as adjectives and logical relations
as conjunctions.16 A different realization of these constituents (in a narrative) is
incongruent and must result in grammatical metaphor. Grammatical metaphor,
therefore, is realized through a type of trans-categorization, mainly (according to
Halliday) the replacement of the ‘non-stable’ processes by ‘stable’ participants,
implicit in which is a tendency towards nominalization.17 Our first example,
(a), presents the typical, ‘congruent’ realization of the process, while (b) has a
metaphorical, ‘incongruent’ but nevertheless ‘agnate’ realization:18
(a) The engine failed.
(b) Engine failure.
The tendency towards nominalization, as in (b), is natural to language, since it

allows specification of participants and therefore classification, which is essential
for describing and making sense of the world.19 Expressions of paradigms exist on a
metaphorical cline; a realization is metaphorical to a lesser or a greater degree. But
13
Halliday and Matthiessen (1999) 184ff.
14
Givón (1979) 320–4; cf. Halliday (1998) 197, 201.
15
Cf. also Givón (2001) 52.
16
Halliday (1998) 208.
17
Halliday and Matthiessen (1999) 269.
18
Examples from Halliday and Matthiessen (1999) 268.
19
Halliday (1998) 197, 201.
although all grammatical metaphors share this important function, they are to be
found in different organic clusters or ‘syndromes’ connected to different registers.20
We offer a concise description of these syndromes below.
‘Process and range’

A very frequent grammatical metaphor syndrome is ‘process and range’, which is
sometimes treated in part under the heading ‘cognate objects’. In the case of this
syndrome, which is common in everyday, spoken language, the semantic load of
the utterance rests on the ‘range’, which is similar to a ‘patient’. A ‘patient’ is an
impacted participant, like ‘a house’ in ‘John built a house’; the difference is that
‘range’ is neither performing (as an actor is, like ‘John’ in ‘John built a house’)
nor impacted, but presents fuzzy boundaries with the category of ‘qualities’.21
‘Range’ is defined as specifying the scope or range of the process.22 In the following
example, the range is ‘a mistake’:
(a) to err;
(b) to make a mistake.
This syndrome represents the prototypical grammatical metaphor, which is

probably the ultimate source of ‘packed’ information structures with the same
tendency in written discourse, and from which arises the syndrome of ‘things and
relations’.
‘Things and relations’

In the scientific discourse of English, Modern Greek and other languages, we find
the syndrome of ‘things and relations’, a syndrome that Halliday has explored in
many of his writings.23 Halliday claims that this syndrome began with the written
work of the Greek philosophers.24 In the frame of this syndrome, processes are
replaced by nominal groups, nominal groups are transformed into qualities (genitives
and prepositional groups), and connectives (usually causative connectors) are
turned into a relational process, effectively turning the clause complex into a simple
clause with packed lexical information, a progress evident in the following set of
examples:
(a) CO2 increases, so temperature rises.
(b) Increases in CO2 cause temperature rises.
(c) CO2 increases are the cause of temperature rises.
20
21
Congruently expressed by means of adjectival groups.
22
See Thoma (2004) for a full description of the syndrome in Modern Greek.
23
Halliday (1985), (1994), (1997), (1998), (1999).
24
For an investigation of the origins of this syndrome in Ancient Greek, see Kappagoda
(forthcoming).
The importance of this syndrome lies in the power it confers on the author over
the general public, in that he or she is given absolute control over the management
of given/new information, creating ‘absolute truth’ statements and therefore ‘truth
myths’, which the untrained, non-specialized public will find difficult to detect or
judge. At the same time it targets an elite public which expects this difficulty and
finds it appropriate, because it maintains the status quo of one social group over
another by ‘restricting access’ to knowledge.25
We suggest that this has also been the function of participles in older Greek
texts, both narrative and expository; accordingly, we look at participles as ‘semantic
compromises’ in the linguistic system.
Participle as ‘semantic compromise’

Process represents change; ‘staying the same’ and ‘not changing’ becomes the
limiting case of a process.26 The process as expressed by means of a verbal group
forms the heart of the clause, as is to be expected, since it sets the tense and the
modal responsibility of the clause in conjunction with the participant.27 Participles,
which are relatively stable in time, are therefore ‘a kind of semantic compromise,
a means of having it both ways’, by construing an exchange of functions without
shifting class,28 as in:
(a) It was a terrible thing how/that he left her.
(b) It was a terrible thing, him/his leaving her.
Although Ancient Greek participles express voice (active, passive or middle),

they only express the indicative mood;29 they also, in most cases, lack tense30
(except for the future),31 instead making use of aspect (imperfective, perfective
or perfect). These features contribute to their relative time stability. It is easy to
imagine why the relative lack of tense should give the impression of time stability,
but the fixedness of the indicative mood is even more essential in giving this
impression. According to Fox:
25
26
27
Halliday (1985) 76.
28
Halliday and Matthiessen (1999) 261–2.
29
There are four moods in Ancient Greek: indicative, optative, subjunctive, imperative.
30
In Ancient Greek finites (as also in their Modern Greek counterparts), the present tense consists
of the base verb stem, which is imperfective. It has no tense marking; the aorist is formed with the
perfective aspect and the augment ‘ε’, which is not applicable to the ancient Greek participle (see
‘ἦλθον’ vs. ‘ἐλθών’, ‘ἔγραψον’ vs. ‘γράψας’): cf. Fox (1983) 30. The future tense is formed by
suffixing -s- to the verb stem. The perfect ‘tense’ is actually the perfect aspect which is formed with
reduplication of the verb stem and the addition of the suffix -k- to it: cf. Fox (1983) 24, 26.
31
See ibid. 26–7. ‘The non-finite forms lack tense (except the future), mood and agreement
markers. Moreover, of the three forms [i.e. the finite, the infinitive and the participle], only the participle
carries the inflection markers for gender, case and number’: ibid. 27.
Moods function to scale events according to the likelihood of their occurring.

Likelihood or probability is involved only when a change of affairs is possible (if there
is no change, the probability of the event occurring is 100%). The more time stable
the event described by the form, the less likely it is to show a change. We can expect,
therefore, that verbal forms that are used to describe durative, nonsequenced events
(like the participle and the infinitive) will tend not to be susceptible to changes in
mood.32
Participles are, indeed, more stable than verbs, in respect of both time and
probability.
According to Givón, there is a fuzzy area between the two extreme grammatical
entities, the nouns and the verbs.33 There are three possibilities: (a) mostly verb
plus some noun, (b) mostly noun plus some verb, and, (c) approximately half verb
and half noun. Building on this categorization, Fox argues that the Ancient Greek
participle belongs to the first category, being essentially verbal but with ‘nominal
trappings – case and gender’, which are indicative of a link to the time-stable
noun.34 The Ancient Greek infinitive is also a compromise, but ‘the participle is
a less independent, less typically verbal form than the infinitive’ and, of course,
than the finite verb.35 Participles are in fact akin to ‘qualities’, and this is not a
coincidence: ‘qualities’ are less stable in time than ‘things’,36 which are congruently
expressed by means of nominal groups. A complete nominalization of a narrative
would inevitably result in a list of ‘things’ (representing events and states) instead
of a dynamic sequence of actions, which is the prototypical narrative.
Fox’s insights into the discourse function of the Ancient Greek participle are
illuminating.37 Fox tried to link the Ancient Greek participle mainly to stative verbs
and background information: a real possibility, since participles lack finiteness.
However, working with texts from a different era, the fourteenth century, we shall
demonstrate how the emergence and perfection of the high narrative register takes
the function of participles in narratives one step further. By using Ancient Greek
participles at an even higher frequency than ancient texts, in order to encode the
most important and abundant processes, these fourteenth-century texts, with their
‘insider’ vocabulary and dense structures of ‘packed’ information, clearly address
the expectations of an elite, intellectual public.
Grammatical metaphor and the evolution of an authoritative register

Grammatical metaphor, a nominalizing force in language, represents a shift from
the logical to the experiential: turning all phenomena into a more ‘classifying’ form.
32
Ibid. 31.
33
Givón (1979) 320–4.
34
Fox (1983) 27.
35
Ibid. 28.
36
37
Fox (1983).
Halliday has argued that, viewed historically, the development of the ‘things and
relations’ syndrome of grammatical metaphor may be said to involve four stages:
(1) a, so x;
(2) because a, x;
(3) a causes x;
(4) a is the cause of x.
The syndrome is expressed as a ‘knight’s move’ in grammar: down in rank,

sideways in class and function.38 As explained above, this means that processes are
replaced by nominal groups, nominal groups are replaced by qualities (genitives
or prepositional groups), connectives (causatives) are replaced by a relational
process, and clause complexes are replaced by lexically ‘packed’ clause simplexes.
In the following series,39 we see the process moving from congruent realization to
the ‘things and relations’ syndrome, not only in structure but also in vocabulary
choices:
(a) Look – wasn’t it good that we watered that philodendron? See how well it‘s
growing!
(b) Like watering plants: you water them, and they grow.
(c) You know you’ve got the right idea because you can do something and it works.
(d) The best way of telling that we know what’s happening is to see that what we do
is working.
(e) What proves that we know things accurately is the fact that we can act
effectively.
(f) The fact that our knowledge is accurate is best confirmed by the fact that our
actions are effective.
(g) The truest confirmation of the accuracy of our knowledge is the effectiveness of
our actions.
The use of grammatical metaphor signifies an increase in textual meaning

(backgrounded/foregrounded participants) and a more or less dramatic loss in
experiential meaning. For instance, configurational relations become inexplicit;
categories of experience become blurred; ultimately, all human participants
can be avoided (thus dehumanizing the text); and as a result the author shuns
accountability for what is said. These characteristics are valued in today’s scientific,
un-common-sense, authoritative discourse, rendering grammatical metaphor
an essential construction.40 This discourse is already almost in place in Modern
Greek as well. Here, for example, is an extract from a 1986 online doctoral thesis in
physics:41
38
Halliday (1998).
39
Example taken from Halliday (1998).
40
Cf. ibid. 228.
41
The extract (from a thesis by E.T. Zygouris) is taken, at random, from the digital library
Hellinomnimon (n. 12 above).
Η χρησιμοποίηση ενός ψηφιακού φίλτρου (στην επεξεργασία ενός αναλογικού

συστήματος) απαιτεί τη σύνδεση (πριν και μετά το φίλτρο) ενός μετατροπέα
αναλογικού σήματος (σε ψηφιακό (A/D) και ενός μετατροπέα ψηφιακού
σήματος σε αναλογικό (D/A) αντίστοιχα).
The use of a digital filter (in the processing of an analogical signal) demands the
connection (before and after the filter) of an analogical signal transformer (to
a digital one (A/D) and of a digital signal transformer to an analogical one (D/A)
respectively).
The grammatical metaphor here becomes more evident when we ‘unpack’ it,
following Halliday’s example:
(a) Η χρησιμοποίηση ενός ψηφιακού φίλτρου απαιτεί τη σύνδεση ενός
μετατροπέα.
The use of a digital filter requires the connection of a transformer.
(b) Επειδή/Όταν θέλεις να χρησιμοποιήσεις ένα ψηφιακό φίλτρο, πρέπει να

χρησιμοποιήσεις ένα μετατροπέα.
Because/when you want to use a digital filter, you must use a transformer.
(c) Θέλεις να χρησιμοποιήσεις ένα ψηφιακό φίλτρο. Γι᾿ αυτό χρειάζεσαι ένα
μετατροπέα.
You want to use a digital filter. For that, you need a transformer.
Behind such a specialized discourse one usually finds a period of development

and perfecting, as well as a suitable historical and social context.42 In three
examples taken from Greek texts from different periods (sixteenth and eighteenth/
nineteenth centuries) we can observe the development of the Modern Greek
scientific register and, in particular, the development of the grammatical metaphor
syndrome of ‘things and relations’.
In the oldest of the three texts, a treatise on arithmetic by Manuel Glytzounis
(1530–96), we find language that attests to a new, as yet non-authoritative, register.
The emphasis is on ‘giving instructions’. As indicated in the example given, the
text consists mainly of material processes (processes of ‘doing’, in the imperative
mood) and relational clauses (of ‘being’ or ‘having’). The nominalization, ‘trial of
multiplication’, can be interpreted as an attempt at grammatical metaphor: it may
be given in its agnate, congruent form as the process, ‘try to multiply’. However, the
nominalization tendency in this text is extremely low. Its register may be compared
with Chaucer’s 1391 Treatise as analysed by Halliday:43
Ἤξευρε ὅτι ἡ δοκιμὴ τοῦ πολλαπλασιασμοῦ γίνεται οὕτως·
Know that the trial of multiplication takes place thus:
κάμε ἕναν σταυρὸν ὡσὰν βλέπεις . . .
make a cross, as you see . . .
42
43
Halliday (1988).
In a text on physics by Demetrios Darvaris (1757–1833) we find a different

picture. The main theme can be said to consist in ‘descriptions of action’, either with
impersonal (third-person) or general (first-person plural) constructions. We find
material, mental and relational clauses, in a frequency which points to a narrative
register. In the example given, we see that the language is not as densely ‘packed’
as one would expect if the register were an authoritative one. It is in fact easy to
follow, with open-ended structures and everyday language, but there is clearly an
attempt at information packaging by means of a participle and the establishment of
relations by means of a comparison. This register may be compared with Newton’s
Treatise on Opticks (1704), as analysed by Halliday:44
Τὸ μουχλιασμένον ψωμὶ θεωρούμενον μὲ τὸ τηλεσκόπιον
Mouldy bread seen through the telescope [sic]
φαίνεται εἰς τὴν ἐπιφάνειαν ὅμοιον μὲ ἕνα πυκνὸ δάσος,
appears on its surface similar to a dense forest
εἰς τὸ ὁποῖον εὑρίσκονται πολυάριθμα δενδράκια.
in which there are several little trees.
In a roughly contemporary text on geography (Στοιχεία Γεωγραφίας) by

Theotokis Nikiforos (1731/6–1805), we find language that resembles today’s
conception of scientific discourse even more, in that it offers ‘statements of truth’. It
may be compared with Joseph Priestley’s The History and Present State of Electricity,
with Original Experiments (1767), as described by Halliday.45 The example given,
which is typical of the text as a whole, presents an interesting feature: not only
does it establish a relation between two things (‘Χάρτης’ on the one hand and its
definition on the other), thereby attempting a scientific grammatical metaphor of
‘things and relations’; it also presents a trait no longer found in modern scientific
discourse – frequent use of participles. In this short example we find six:
Καθολικὸς μὲν Χάρτης ὁ καὶ μάππα ἢ μάππαμόνδο προσαγορευόμενος
World map, called mappa or mappamundo,
ἐστὶν ὁ δύω κύκλους ἐφαπτομένους ἀλλήλων περιέχων
is the one containing two contiguous circles
ὑπὸ πολλῶν τετμημένους κύκλων,
cut by many circles,
καὶ ὅλην τὴν Γὴν ἐμφαίνων εἰς δύω ἡμισφαίρια διατετμημένην.
showing the whole of the earth cut into two hemispheres.
The use of participles in scientific discourse might seem surprising to us today,

but more surprising still is the fact that such discourses, now on their way to
becoming authoritative, gravitated slowly from personal, narrative-like registers to
impersonal, expository texts, in a trajectory similar to that proposed by Halliday
for English scientific discourse.46 We can thus argue that as the register is still
44
Ibid.
45
Ibid.
46
Ibid.
mid-way, these authors resort to participles as a means of organizing given and

new information, creating ‘packed’ discourse and moving away from everyday,
spoken language. We shall see this technique in operation in authoritative narrative
registers as well.
Register and context

With this (admittedly brief) survey of the evolution of grammatical metaphor in
Greek scientific discourse, we can see a break in tradition from ancient Greece
to today. Ancient Greek, the language of philosophy, is replaced by the modern
language in its written form, and new registers appear. The new start can be
located around the beginning of the sixteenth century. A process similar to the
evolution of English scientific discourse takes place for early Modern Greek as
well. It is interesting to note, at least for Greek, that this new register does not seem
to be based on the Ancient Greek register of authoritative philosophical texts, but
rather follows an evolution of its own. Not only does this register evolve in time
and is therefore not institutionalized as a fixed register; it is also a new one. Its
early forms are personal narratives and direct instructions, an indication that it
was not an authoritative register at the beginning, but addressed an audience that
did not consider itself to be an elite. In time, the register evolved into today’s fully
authoritative register that makes ample use of the grammatical metaphor syndrome
of ‘things and relations’. It may seem surprising to contemplate the possibility of a
narrative register in such a prestigious position; however, it was a child of its time.
The Life of Aesop: the narrative ‘high’ register

The Life of Aesop is a typical example of Byzantine literary production at its
classicizing height. The high frequency of participles in the fourteenth-century
versions of this text is our focus of interest in this study; accordingly, we shall begin
by considering the function of participles in narrative discourse.
Participles: description and function in narrative discourse

The Ancient Greek participle carries aspect (and also occasionally tense),47 as well
as gender, case and number. These allow it not only a feature of finiteness (aspect)
but also features of definiteness (gender, case and number). Fox cites three types of
Ancient Greek participles:
(a) adjectival (attributive) participle (acts as an adjective or noun, as in οἱ κρατοῦντες:
‘the conquerors’);
(b) verb complement (supplementary) participle (usually replacing an infinitive, as in
παύεται λέγων: ‘he stops speaking’);
(c) adverbial (circumstantial) participle (expresses temporal, causal, conditional and
circumstantial relationships with another clause).48
47
See p. 114 above.
48
Fox (1983).
The participles that we are investigating here, the type already investigated by
Fox in Ancient Greek, belong to the third category, that of ‘adverbial participles’.
These, as Fox explains, are the most ‘unstable’ in time and the most similar in
function to finite verbs; they are therefore more likely to be used in narrative
sequences. They have a preference for active voice, an indication of their tendency
to encode ‘events’: they are therefore less stable in time than the other types of
participle.49
Givón proposes a differentiation among adverbial participles according to their
position in the clause complex:50 post-posed participles are usually understood as
having more local, semantic connections with their main clause and thus as more
closely integrated into their semantic structure. They are seen as backgrounded,
simultaneous with the main clause, and attached to it. In contrast, pre-posed
participles are understood as having more global pragmatic relations to the discourse
context and as being less integrated into the semantic structure of the main clause.
They are therefore seen as foregrounded, detached and sequential (occurring in
different time from the main clause). The participles in our corpus are mainly pre-
posed.
The participles in our texts encode sequences of events and replace the
prototypical event realizations (that is, finites). If we take our participles to be
sequential, then they have a different function in narrative from the participles
found in Herodotus and analysed by Fox, who states: ‘it thus seems that we can
define backgrounded discourse as that portion which contains nonsequentially
ordered events’.51
We agree that the adverbial participle is probably the least backgrounded of
the three types of participle, but we shall, nevertheless, separate the categories
of sequentiality and backgrounding, as we do not believe they always coincide.
We must therefore ask the question whether the sequenced participles in our
texts are backgrounded at all. They are pre-posed and are for the most part
perfective participles. Their definiteness gives the impression of ‘givenness’, and
‘accessibility’ from the previous discourse, the same impression as is given by their
perfectiveness. Furthermore, their perfectiveness, in addition to the obvious past
tense of the narrative, grounded by finites in past or historical present, also gives the
impression of past temporal reference, an impression which depends purely on any
temporal markers (including the finite) in the clause to which they belong.52 These
two features, definiteness and perfectiveness, allow long participial concatenations
with dense cohesive chains.53 Fox suggests that ‘a characteristic of backgrounding
49
Ibid. 37.
50
Givón (2001) 343.
51
Fox (1983) 29 (emphasis in the original).
52
Cf. Dahl (1985) for an extensive discussion of aspect; Hopper (1979) and Fleischmann (1985)
for its importance in the context of narrative; Lindvall (1997) for Modern Greek narratives. Discussion
of aspect falls beyond the scope of this chapter.
53
Correlation of definiteness and perfectiveness is typical of Modern Greek narratives as well:
Lindvall (1998) 208.
seems to be dependence on a more prominent form’.54 The participles in our

texts are indeed dependent on the main clause, and in this sense we can say that
the participles in the Life of Aesop are presented as sequential, but also as clearly
backgrounded. Only the finite verb is foregrounded, and it comes towards the
end of the clause complex, the typical position for new or contrastive information
(‘focus’), both in Greek55 and in general.56
If we now integrate these conclusions into the SFL framework and take a look
at the phenomenon in terms of textual metafunction, we see how useful this kind of
construction is for texture. The participles form a backgrounded complex of given
information, with the finite rounding up the utterance as new information. This
‘allows for maximum texturing and maximum streamlining, since information that
can be inferred from the main verb is not repeated on the backgrounded form’.57
As an effect, this use of participles creates what Aristotle in his Rhetoric called
lexis katestrammene (‘periodic style’, with sentences organized towards an end) as
opposed to lexis eiromene (‘loose style’, with phrases simply ‘strung together’).58
This opposition deserves our attention, since it demonstrates that ancient writers
were conscious of this choice in creating a highly ‘packed’, dense discourse. For
Aristotle, lexis katestrammene creates ‘perfect’ discourse, in the sense of building
up expectations (in effect, management of given/new information) and of giving
the feeling of completeness by creating ‘closed’ sentences, and rounded episodes,
as opposed to simple clauses. Interestingly, this is exactly the type of discourse used
nowadays for science, arguably the authoritative register of our time.
If the lexis katestrammene created by means of sequential adverbial participles
functions as a device for creating ‘packed’, authoritative discourse, we would
expect the lexis eiromene to be the norm in a ‘spoken’ register,59 which is easier
to understand and aims at a wider public. Accordingly, in the following example
(a), from high-register version P of the Life of Aesop, most events are realized as
participles. By contrast, passage (b), from ‘Aesop D’, a seventeenth-century
translation of version P, already offers us the ‘unpacked’ version of the phrase, with
congruent encoding of all processes as finites in the form of a prototypical narrative.
Unpacked, sequenced discourse with congruent realizations commonly addresses
a wide, general audience with no special knowledge. In the two passages quoted,
participles are printed in italics, finites in bold:
(a) ‘Aesop P’
Μεθ᾿ ἡμέρας πάλιν εἰς τὸ βαλανεῖον ἐλθόντος τοῦ Ξάνθου
Some days later, Xanthos, coming again to the bath,
54
Fox (1983) 30.
55
Cf. Holton et al. (1999) 438–9.
56
Cf. Halliday (1985) 298–9 and Givón (2001) passim.
57
Fox (1983) 31.
58
Rhet. III. 9.
59
Coincidentally, perhaps, εἰροµένη (‘joined together in a row’: LSJ, εἴρω A) could also, in itself,
be a passive of the epic-poetic εἴρω, ‘speak’ (LSJ, εἴρω B).
καί τισιν ἐντυχόντος ἐκεῖ τῶν φίλων,

and finding there some friends,
καὶ πρὸς τὸν Αἴσωπον εἰρηκότος εἴς τε τὴν οἰκίαν προδραμεῖν
and telling Aesop to run home
καὶ φακῆν εἰς τὴν χύτραν ἐμβαλόντα ἑψῆσαι,
and having put lentils in the pot to cook them,
ἐκεῖνος ἀπελθὼν κόκκον ἕνα φακῆς εἰς τὴν χύτραν ἕψει βαλών.
(Aesop) leaving (them), having put one lentil in the pot, cooked (it).
(b) ‘Aesop D’
Μετὰ δὲ ἡμέρας πάλιν ἐπῆγεν ὁ Ξάνθος εἰς τὸ λουτρόν.
Some days later again, Xanthos went to the bath.
Καὶ ἐκεῖ εὗρε τινὰς ἀπὸ τοὺς φίλους του.
And there he found some friends.
Καὶ λέγει πρὸς τὸν Αἴσωπον ὅτι νὰ πάγη νὰ βάλη φακῆ εἰς τὸ τσυκάλιν, νὰ τὴν
μαγερέψη.
And he tells Aesop to go and put lentils in the pot and cook them.
Καὶ ἐκεῖνος ἐπῆγε καὶ ἔβαλε ἕνα κουκκὶν φακῆν καὶ ἐμαγέρευσεν.
And he (Aesop) went and put one lentil (in the pot) and cooked (it).
A similar idea, from a very different starting-point, is proposed by Hoyle with

reference to the Greek New Testament. He argues that participles can represent a
stereotypical scenario for which the finite verb serves as a ‘title’.60
As Halliday has shown, typical characteristics of today’s authoritative scientific
register are lexical density and technical vocabulary.61 Similarly, the high register
under consideration here is characterized by what would have been felt to be
‘technical’ vocabulary at the time, in the shape of ancient lexical and grammatical
items, drawn from as far back as Homer.62 The similarities between the authoritative
Life of Aesop and today’s authoritative scientific register become clearer once we
realize that both registers target a similar, educated, narrow public which shapes
them both by its expectations and by its need to keep knowledge away from the
masses.
Findings from quantitative analysis

The results of our investigation are tabulated in Table 1, with reference to the
two high-register versions of the Life of Aesop (P and W), along with the three
low-register versions, I, D and K. This initial quantitative analysis reveals a clear
difference in frequency between the two registers. In the first two columns, the
figures for the high versions show a high frequency in the use of participles, which
appear more often than finite verbs. Conversely, the last three columns, with
figures for the low versions, show a very low frequency in the use of participles and
a high frequency for finite verbs. The last three columns imply a strong tendency
60
Hoyle (2001).
61
Halliday (1988), (1998); see also Ravelli (1985), (2003).
62
Cf. Toufexis (forthcoming).
towards a prototypical oral narrative register.63 We may assume that, since the low
versions are translations of their two high counterparts, the participles in the former
represent ‘contamination’ from the source texts:
Table 1
Life of Aesop: process encoding (participles and finites) in the five versions,
high (P, W) and low (I, D, K)
P W I D K
Participles 54.1% (576) 55.9% (297) 4.0% (48) 11.2% (127) 5.2% (79)
Finites 45.9% (297) 44.1% (234) 96.0% (1138) 88.8% (1004) 94.8% (1438)
Let us now turn to the two high versions. In Tables 2 (version P) and 3 (version
W), we find the correlation between process encoding (finite or participle) and
process types:
Table 2
Process encoding (by process type) in ‘Aesop P’
Material Verbal Mental Behavioural Relational
Participles 61.4% (205) 31.0% (225) 77.8% (30) 72.4% (8) 52.3% (21)
Finites 38.6% (326) 69.0% (101) 22.2% (105) 27.6% (21) 47.7% (23)
Table 3
Process encoding (by process type) in ‘Aesop W’
Participles 56.3% (324) 10.0% (46) 75.9% (123) 70.3% (26) 54.8% (17)
Finites 43.8% (252) 90.0% (416) 24.1% (39) 29.7% (11) 45.2% (14)
From these figures it becomes clear that the finites in the high register encode
predominantly verbal processes (processes of saying), and to a lesser degree relational
processes (processes of being) and material processes (processes of doing).
By contrast, in the low-register versions of the same texts, as is clear from
Tables 4, 5 and 6, we see only a modest tendency to encode mental and relational
processes by means of participles; this is not surprising, insofar as these participles
mostly provide backgrounded information for the story and do not advance the
plot. The majority of processes are congruently expressed by means of finites:
Table 4
Process encoding (by process type) in ‘Aesop I’
Participles 1.6% (8) 3.0% (14) 9.1% (11) 0.0% (0) 20.0% (15)
Finites 98.4% (479) 97.0% (458) 90.9% (110) 100.0% (31) 80.0% (60)
63
See Thoma (forthcoming) for oral narrative strategies in low-register narrative.
Table 5
Process encoding (by process type) in ‘Aesop D’
Participles 15.2% (75) 1.8% (8) 30.0% (33) 17.5% (7) 10.5% (4)
Finites 84.8% (417) 98.2% (443) 70.0% (77) 82.5% (33) 89.5% (34)
Table 6
Process encoding (by process type) in ‘Aesop K’
Participles 4.6% (32) 1.0% (6) 22.0% (28) 15.5% (9) 8.7% (4)
Finites 95.4% (660) 99.0% (588) 78.0% (99) 84.5% (49) 91.3% (42)
The supremacy of audience design

The results of the quantitative analysis provide ample evidence for the discourse
function of participles in the two high versions of the Life of Aesop, P and W.
Participles here are more common than finites and encode most processes in
the narrative, with the exception of verbal processes, which are more commonly
encoded as finites.
From the qualitative analysis, we can now see that a typical phrase in these two
high versions presents information as backgrounded, as part of a nominal phrase
which includes the actor as head, and a finite verb, usually a process of saying,
as the nexus or head of the clause. The finite is the focus of the clause. In the
case of a finite encoding a process other than verbal (say, a material or relational
process), the verbal process is often encoded as an imperfective participle (in
traditional grammar, the so-called ‘present-tense’ participle) at the end of the
clause, immediately followed by a quotation, as in this example from ‘Aesop P’:
τότε ὁ Αἴσωπος ἀποσχὼν τῆς στήλης βήματα τέτταρα καὶ ὀρύξας,
then Aesop, having moved four steps from the pillar and having dug,
ἀνέλαβέ τε τὸν θησαυρὸν καὶ ἤνεγκε τῷ δεσπότῃ, λέγων· . . .
(he) took the treasure and brought (it) to his master, saying. . . .
In such sequences, the participles usually appear towards the beginning of the clause
complex, with the finite coming towards the end, in order to present information as
given. The management of information in these texts is at least as sophisticated as
in the most technical scientific papers written today.
Conclusions
Within the framework of SFL, the grammatical metaphor syndrome of ‘things
and relations’ is considered to be a characteristic of scientific discourse as the
authoritative register of our time. In this chapter we have argued first that, in the
evolution of the Modern Greek scientific register, grammatical metaphor began
as a tendency to use participles which form a ‘semantic compromise’ between
process and participant. In older narrative registers, this tendency constitutes the
pinnacle of authoritative narrative writing, taking narrative action to the very limit
of nominalization, in a syndrome that, following Aristotle, we might call the lexis
katestrammene syndrome.
By a comparative examination of the characteristics and discourse functions of
Ancient Greek participles in two versions of the Life of Aesop from the fourteenth
century, we have argued that not only the form, but also the placement of the
participles in the clause complex, plays an important role, with implications for
their functions. Clearly, the participles in these texts have a sequential function
which is, however, backgrounded in the interests of good management and
streamlining of given/new information. This latter function is aided by the fact that
most of the participles carry the perfective aspect and the markers of definiteness
(gender, case and number), which allows for long, but very cohesive, participial
chains. Nevertheless, this realization of processes through participles is not the
prototypical one. Their (usual) lack of tense and mood, along with their tendency
to avoid any mention of the doer of the action at the beginning of the clause
complex, creates abstract, ‘packed’ discourse: ‘a mode of discourse for establishing
prestige and authority of an elite whose message is “this is all too hard for you to
understand; so leave the decision-making to us”’.64 That quotation refers to today’s
scientific discourse. But it is also applicable to the high narrative register in the
diglossic situation of the Greek Middle Ages and early modern times.
References
Dahl, Ö. (1985), Tense and Aspect Systems, Oxford.
Eberhard, A. (1872), Fabulae Romanenses Graece Conscriptae, vol. 1, Leipzig.
Eideneier, H. (1999), Von Rhapsodie zu Rap: Aspekte der griechischen Sprachgeschichte von
Homer bis heute, Tübingen.
Eideneier, H. (2001), ‘Die Metaphrase als Wechsel der Stilstufe in byzantinischen und
postbyzantinischen Texten’, Göttinger Beiträge zur Byzantinischen und Neugriechischen
Philologie, 1: 27–45.
Eideneier, H. (ed.) (forthcoming), Der Äsoproman [electronic text].
Ferguson, C.A. (1959), ‘Diglossia’, Word, 15: 325–40.
Fleischman, S. (1985), ‘Discourse functions of tense-aspect oppositions in narrative: toward
a theory of grounding’, Linguistics, 23: 851–82.
Fox, B. (1983), ‘The discourse function of the participle in ancient Greek’, in F. Klein-
Andreu (ed.), Discourse Perspectives on Syntax, New York: 23–41.
Givón, T. (1979), On Understanding Grammar, New York.
Givón, T. (2001), Syntax. An Introduction, vol. 1, Amsterdam.
Halliday, M.A.K. (1967), ‘Aspects of varieties differentiation’, Journal of Linguistics, 3:
177–274.
64
Halliday and Matthiessen (1999) 272; cf. Lemke (1990).
Halliday, M.A.K. (1985), An Introduction to Functional Grammar, London.

Halliday, M.A.K. (1988), ‘On the language of physical science’, in M. Ghadessy (ed.),
Registers in Written English, London: 162–78.
Halliday, M.A.K. (1994), ‘The construction of knowledge and value in the grammar of
scientific discourse, with reference to Charles Darwin’s The Origin of Species’, in M.
Coulthard (ed.), Advances in Written Text Analysis, London: 136–56.
Halliday, M.A.K. (1997), ‘Linguistics as metaphor’, in A.M. Simon-Vandenbergen, K.
Davids and D. Noel (eds.), Reconnecting Language: Morphology and Syntax in Functional
Perspectives, Amsterdam: 3–27.
Halliday, M.A.K. (1998), ‘Things and relations: regrammaticizing experience as technical
knowledge’, in J.R. Martin and R. Veel (eds.), Reading Science: Critical and Functional
Perspectives on Discourses of Science, London: 185–235.
Halliday, M.A.K. and Martin, J.R. (1993), Writing Science: Literacy and Discursive Power,
London.
Halliday, M.A.K. and Matthiessen, M.I.M.C. (1999), Construing Experience Through
Meaning: A Language-Based Approach to Cognition, New York.
Holton, D., Mackridge, P., and Philippaki-Warburton, I. (1999), Greek: A Comprehensive
Grammar of the Modern Language, London.
Hopper, P.J. (1979), ‘Aspect and foregrounding in discourse’, in Givón (1979) 213–41.
Hoyle, R.A. (2001), Scenarios, Discourse and Translation: the scenario theory of cognitive
linguistics, its relevance for analysing New Testament Greek and modern Parkari texts, and
its implications for translation theory, PhD thesis, Roehampton University.
Kappagoda, A. (forthcoming), ‘Happening things: grammatical metaphor in Ancient Greek’
(paper delivered to the Australian Society for Classical Studies, 24th Conference, 2002).
Karla, G.A. (2001), Vita Aesopi, Wiesbaden.
Lemke, J.L. (1990), ‘Technical discourse and technocratic ideology’, in M.A.K. Halliday,
J. Gibbons and H. Nicolas (eds.), Learning, Keeping and Using Language: Selected Papers
from the 8th World Congress of Applied Linguistics, vol. 2, Amsterdam: 435–60.
Lindvall, A. (1997), ‘Aspect marking and situation types in Greek, Polish and Swedish’,
Working Papers (University of Lund), 46: 197–216.
Lindvall, A. (1998), Transitivity in Discourse. A Comparison of Greek, Polish and Swedish,
Lund.
Matschke, K. and Tinnefeld, F.H. (2001), Die Gesellschaft im späten Byzanz: Gruppen,
Strukturen und Lebensformen, Cologne.
Matthiessen, M.I.M.C. (1998), ‘Construing processes of consciousness: from the
commonsense model to the uncommonsense model of cognitive science’, in J.R. Martin
and R. Veel (eds.), Reading Science. Critical and Functional Perspectives on Discourses of
Science, London: 327–56.
Papathomopoulos, M. (ed.) (1999a), Ὁ Βίος τοῦ Αἰσώπου, Ἡ παραλλαγὴ W, Athens.
Papathomopoulos, M. (ed.) (1999b), Πέντε δηµώδεις µεταφράσεις του Βίου του Αισώπου,
Athens.
Ravelli, L. (1985), Metaphor, Mode, and Complexity: An Exploration of Co-Varying Patterns,
BA thesis, Department of Linguistics, University of Sydney.
Ravelli, L. (2003), ‘Renewal of connection: integrating theory and practice in an
understanding of grammatical metaphor’, in A. Simon-Vandenbergen, M. Taverniers, and
L.J. Ravelli (eds.), Grammatical Metaphor: Views from Systemic Functional Linguistics,
Amsterdam: 37–64.
Thoma, C. (2004), ‘The grammatical metaphor of “process and range” in Modern Greek’,
in Online Proceedings of the 6th International Conference of Greek Linguistics in Rethymno
– Greece, at http://www.philology.uoc.gr/conferences/6thICGL/default.htm
Thoma, C. (2006), Combining Functional Linguistics and Translation Theories: A Case Study
of Greek Cypriot and British Folktales, Frankfurt.
Thoma, C. (forthcoming), ‘Oral strategies in written narrative texts: towards a definition of
the “low” register’.
Toufexis, N. (2008), ‘Diglossia and register variation in Medieval Greek’, Byzantine and
Modern Greek Studies, 32: 203–17.
Toufexis, N. (forthcoming), ‘Defining “high” and “low” from a typological perspective: a
case-study of the Early Modern Greek intralingual translations of the Life of Aesop’.
Wills, L.M. (1997), The Quest of the Historical Gospel: Mark, John, and the Origins of the
Gospel Genre, London.
Appendix
(i) ‘Aesop P’ (Eberhard (1872) 251)
τῶν δὲ εἰξάντων καὶ πρὸς τὴν οἰκίαν ἀφικομένων,
Having left and arriving at the house,
ὁ Ξάνθος φησί· «δὸς ἡμῖν ἀπὸ λουτροῦ πιεῖν, Αἴσωπε.»
Xanthos said: ‘give us to drink from the bath, Aesop.’
τοῦ δ᾿ ἐκ τῆς ἀπορροίας τοῦ λουτροῦ λαβόντος καὶ ἐπιδεδωκότος,
Him having taken and having given from the emanation of the bath,
ὁ Ξάνθος τῆς δυσωδίας ἀναπλησθείς, «φεῦ, τί τοῦτο», φησίν «Αἴσωπε;»
Xanthos, filled with the stench, ‘alas, what is this’, (he) said, ‘Aesop’?
καὶ ὅς· «ἀπὸ λουτροῦ, ὡς ἐκέλευσας.»
And he: ‘from the bath, as you ordered’.
τοῦ δὲ Ξάνθου τῇ παρουσίᾳ τῶν φίλων τὴν ὀργὴν ἐπισχόντος καὶ λεκάνην αὐτῷ
παρατεθῆναι κελεύσαντος,
Xanthos having put aside his anger due to the presence of his friends and having ordered the
basin to be placed before him,
Αἴσωπος τὴν λεκάνην θεὶς ἵστατο. καὶ ὁ Ξάνθος· «οὐ νίπτεις;» κἀκεῖνος·
Aesop having placed the basin stood. And Xanthos, ‘are you not washing?’, he (said).
(ii) ‘Aesop I’ (Papathomopoulos (1999b) 70)

Καὶ ἐκεῖνοι ἀκολούθησαν καὶ ἦλθαν εἰς τὸ σπίτιν μὲ τὸν Ξάνθον.
And they followed and came to the house with Xanthos.
Λέγει τοῦ Αἰσώπου ὁ Ξάνθος: «Φέρ᾿ μας νὰ πιοῦμεν τώρα ἀπὸ τὸ λουτρόν.»
Xanthos tells Aesop: ‘Bring us to drink now from the bath.’
῾Ο δὲ Αἴσωπος ἐπῆγεν καὶ ἐμάζωξεν τὰ ἀποπλύματα τοῦ λουτροῦ καὶ ἔφερέν τα.
Aesop went and gathered the dirty water from the bath and brought it.
῾Ο δὲ Ξάνθος σὰν ἤκουσεν τὸν βρόμον, λέγει: «᾿Αλίμονον εἰς ἐμένα, τί εἶναι ἐτοῦτο,
Αἴσωπε;»
Xanthos, when he smelled the stench, says: ‘woe to me, what is this, Aesop?’
᾿Αποκρίθην: «᾿Απὸ τὸ λουτρόν, ὡς καθὼς ὥρισες.»
(He) answered: ‘From the bath, as you decreed.’
῞Ομως ἐντράπην ὁ Ξάνθος πολλά, πλὴν διὰ τοὺς φίλους εἶχεν ὑπομονήν.
Still, Xanthos was very ashamed, yet, for his friends, he had patience.
Καὶ μετὰ ὀλίγην ὥραν λέγει του: «Φέρε τὴν λεκάνην.»
And after a little time he tells him: ‘Bring the basin!’
Καὶ ὁ Αἴσωπος ἔφερέν την. Λέγει του ὁ Ξάνθος . . .
And Aesop brought it. Xanthos tells him . . .
(iii) ‘Aesop D’ (Papathomopoulos (1999b) 102)

Καὶ ὡς τοῦ ἤκουσαν οἱ φίλοι καὶ ἦλθαν μαζίν του εἰς τὸ σπίτιόν του,
And when the friends heard him and came with him to his house,
λέγει ὁ Ξάνθος πρὸς τὸν Αἴσωπον:
Xanthos says to Aesop:
«Δός μοι νὰ πίω ἀπὸ λουτροῦ, Αἴσωπε.»
‘Give me to drink from the bath, Aesop!’
Τοῦ δὲ Αἰσώπου λαβόντος ἀπὸ στραγγίσματα τοῦ λουτροῦ,
Aesop, taking from the dirty water of the bath,
ἔδωκε τῷ Ξάνθῳ.
gave to Xanthos.
Τοῦ δὲ Ξάνθου ὡς τοῦ ἔδωκεν ἡ βρόμα,
Xanthos, when he smelled the stench,
ἐταράχθη καὶ λέγει: «Φεῦ, Αἴσωπε, τί ἐστι τοῦτο;»
became upset and says: ‘Alas, Aesop, what is this?’
Καὶ ὁ Αἴσωπος εἶπε: «᾿Απὸ λουτρὸν μὲ ὥρισες νὰ σοῦ φέρω, καὶ ἀπὸ λουτρὸν σὲ
ἔφερα.»
And Aesop said: ‘From the bath you told me to bring you (water) and from the bath I
brought (it)’.
Καὶ ὁ Ξάνθος διὰ τοὺς φίλους ἐκράτησε τὴν ὀργήν του.
And Xanthos, because of his friends, held back his wrath.
Καὶ πάλιν ὥρισέν τον νὰ παραθέσῃ λεκάνην.
And again he told him to bring a basin.
Καὶ ὁ Αἴσωπος παρευθὺς ἔθεσε τὴν λεκάνην καὶ αὐτὸς ἐστέκετο.
And Aesop immediately placed the basin and he was standing.
Καὶ λέγει του ὁ Ξάνθος . . .
And Xanthos tells him . . .
(iv) ‘Aesop W’ (Papathomopoulos (1999a) 165)

καὶ ἀγαγὼν αὐτοὺς εἰς τὴν οἰκίαν φησίν “Αἴσωπε, δὸς ἡμῖν πιεῖν ἀπὸ βαλανείου.”
And taking them to the house he says, ‘Aesop, give us to drink from the bath.’
ὁ δὲ ξέστην λαβὼν καὶ δραμὼν ἐν τῷ πριβάτῳ
And he, taking a basin and entering the bath
πλήσας τε ἐκ τῆς ἀπορροίας καὶ κεράσας

and filling it to the top with dirty water,
ἐπέδωκε τῷ Ξάνθῳ.
gave it to Xanthos.
ὁ δὲ τῆς δυσωδίας πλησθεὶς φησὶν “ψῶ, τί τοῦτο; Αἴσωπε.”
He, filled with the stench, says, ‘Phew, what is this, Aesop?’
ὁ δὲ φησὶν “ἀπὸ βαλανείου, ὡς ἔφης.”
He says, ‘from the bath, as you said’.
ὁ δὲ Ξάνθος ἐνεὸς γενόμενος φησί “παράθες μοι λεκάνην.”
And Xanthos, having become angered, says, ‘place the basin before me.’
(v) ‘Aesop K’ (Papathomopoulos (1999b) 16)

᾿Επῆρε γοῦν τοὺς φιλοσόφους εἰς τὸ σπίτι ὁ Ξάνθος,
So Xanthos took the philosophers to his house,
καὶ ἐκάθισαν νὰ φᾶν.
and they sat down to eat.
Καὶ λέγει ὁ Ξάνθος: «Αἴσωπε, δός μοι νὰ πίω ἀπὸ λουτροῦ», ἤγουν πὼς
ἦλθα ἀπὸ τὸν λουτρόν, καὶ δός μοι νὰ πίω.
And Xanthos says: ‘Aesop, give me to drink from the bath’, meaning I came
from the bath, give me to drink.
῾Ο δὲ Αἴσωπος, ἁρπάσας ἀγγεῖον καὶ ἀπὸ τὰ ἀποπλύματα τοῦ λουτροῦ ὁποὺ
ἐστράγγιζαν, ἐγέμισε τὸ ἀγγεῖον
And Aesop, having taken a pot from the dirty water draining from the bath, (he) filled the
pot
καὶ ἔδωσε τὸν Ξάνθον καὶ ἔπιεν.
and gave it to Xanthos and he drank.
῾Ως δὲ ἔπιεν ὁ Ξάνθος καὶ ἔδωκέν τον ἡ βρόμα,
When Xanthos drank and smelled the stench,
γυρίζει πρὸς τὸν Αἴσωπον καὶ λέγει: «᾿Αλίμονον, Αἴσωπε, τί ἦτον αὐτὸ ὁποὺ μὲ
ἔδωκες, καὶ ἐταράχθηκα ὅλος ἀπὸ τὴν βρόμαν;»
he turns to Aesop and says: ‘Alas, Aesop, what is this you gave me, and I am all shaken by
the stench?’
῾Ο δὲ Αἴσωπος λέγει: «᾿Απὸ λουτροῦ, ὡς μὲ ὥρισες.»
And Aesop says: ‘From the bath, as you told me.’
῾Ο δὲ Ξάνθος ἔμεινεν ὡσεὶ νεκρὸς καὶ λέγει τον: «Φέρε με ὧδε λεκάνην.»
But Xanthos remained as if dead and tells him: ‘Bring me the basin here!’
῾Ο δὲ Αἴσωπος ἤφερε τὴν λεκάνην καὶ ἔστησέ την ἔμπροσθεν τοῦ Ξάνθου.
And Aesop brought the basin and placed it before Xanthos.
Καὶ ὁ Ξάνθος λέγει τὸν Αἴσωπον . . .
And Xanthos tells Aesop . . .
7
Orthographic Standardization
of the Modern Greek Classics: Gain and Loss
David Ricks
‘Don’t dispute about texts: buy a good text’: Benjamin Jowett’s coat-trailing advice.1
Most will gladly acquiesce when it comes to Homer or Theocritus; but it would be
unwise to do so in the case of the texts – the verse texts especially – which make
up the Modern Greek canon. This is not simply because of the dearth of adequate
critical, let alone critical and annotated, editions of Modern Greek authors.2
Indeed, the dearth of such editions seems to me to be yet another symptom of the
problems in standardizing Modern Greek: when a temerarious editor is smitten
hip and thigh by editors manqués, much of the venom will customarily concern
orthographical decisions.3
In referring to modern texts, moreover, it is reasonable to leave on one side the
distinctively thorny problems of the medieval texts written in a form of Greek
readily comprehensible to the modern reader, and in that sense ‘Modern Greek’.4
Such texts are typically anonymous and undated, and each recension will vary
linguistically according to the (varying) region of its production, its (widely varying)
period, and the (inevitably varying) individual choices of the scribe, redactor, or – if
you will – author.5 Instead, I shall look at some samples from the modern tradition,
since Independence, in which authorship, period and (perhaps more contentiously)
1
Madan (1980) 61.
2
Two examples, of greatly varying quality, for modern Greece’s founding national poets are
Solomos (1994) and Kalvos (n.d.).
3
For an agonizingly detailed account of some of the complexities see the editor’s preface to
Papadiamantis (1997) κδʹ–κεʹ.
4
The question when ‘Modern Greek literature’ begins is itself a contentious one: for a sample of
different answers see Panayotakis (1993) 37–73. The process of standardization of course depends on
there being a standard grammar and historical dictionary: for Modern Greek we still lack the latter. Here
the sine qua non of future study of medieval (for this purpose 1071–1669) texts will be the completion of
the grammar of medieval Greek in preparation in Cambridge.
5
To discussions of such issues, and to editions of such texts – see most recently and fully, Vejleskov
(2005), an edition of Ἀπόκοπος, the first printed Modern Greek text (1509) – the series Neograeca
Medii Aevi has been devoted.
131
132 david ricks
literary context are not a matter of dispute, but where our existing editions expose
the problems of orthographic harmonization to modern norms, when those norms
are both differently standardized and more highly standardized than any norms
prevailing during the authors’ lifetimes. In concentrating on questions of ‘mere’
orthography – as opposed to more drastic interventions on the part of editors – it is
my aim to point to some of the literary implications.
To begin with, one should frankly acknowledge that, for most purposes,
orthographic modernization is a gain.6 It is hard to see, for example, that the reader
of the Cretan Renaissance comedy Φορτουνάτος (completed by 1655) should
be obliged to read the text in the Roman characters in which Markos Antonios
Foskolos wrote it (as was common in the cultural milieu of the period) – even if to
assimilate such a text orthographically to the mainstream does carry with it certain
assumptions about the ‘Greekness’ of this work deriving from a Veneto-Cretan
environment.7 At the other extreme, no-one (I think) would argue that Cavafy
should only be read in facsimiles of his various fascicles, as opposed to (something
like) the form his collected poems took in the posthumous edition of 1935 – yet
any new edition of Cavafy so radically ‘modernized’ as to replace the traditional
accents and breathings with the single-stress-mark ‘monotonic system’ would
provoke well grounded resistance from many, perhaps most, readers.8
In warning of Scylla and Charybdis, without necessarily daring to make my
own way between them, I shall look at some issues of orthographic standardization
that arise from the work of a small number of writers of undisputed classic status.
This is in the belief that hard cases (such as writers who are zealots for spelling
reform) make bad law but also allow for easy editorial resolution.9 Yet, as we shall
see, it also reflects my reservations about the degree of standardization which the
conferral of (modern) Greek classic status can bring in its train.
It may well be thought that the problems posed by Modern Greek verse classics
are particularly acute, and indeed they are; but I would like to begin with a glance
6
Pace G.M. Young, who jocularly opined that it was as absurd to accentuate ancient Greek (in
effect, a modernization, however) as it would be to write it on papyrus (which would be a regression):
Madan (1980) 79.
7
‘Veneto-Cretan’ on the model of ‘Anglo-Irish’ – but I concede that this term is not in wide
currency: a fact with its own implications. A sample from Foskolos’ autograph manuscript is printed in
A. L. Vincent’s edition (Foskolos (1980) οηʹ). We can see from the following lines spoken by Bozikis
that the Roman version is not too difficult to read: ‘Afedimu sibathismu chie i pinamu na sisso/sto
modho tuto me came aprepa na miglisso.’ It is important to stress how necessary scholarly familiarity
with this writing system is, given that e.g. the best manuscript of the Renaissance tragedy Ἐρωφίλη is
so written (Chortatsis (1988) 17).
8
On unobtrusive, but by no means merely orthographical, interventions by editors in the Cavafy
text, see Hirst in the present volume (pp. 149–66).
9
Much of the point of Jean Psichari’s Τὸ Ταξίδι µου (1888) (cf. p. 254 below) would disappear if
its aggressively reformed spelling was standardized. Psichari’s novel spelling system is as ostentatiously
provocative to the unsuspecting reader as the substance of his arguments is to the linguistic specialist
with traditional assumptions about the nature of the Modern Greek language’s relation to Ancient
Greek. Such a work is at once as unreadable as ever and safe from standardization.
orthographic standardization of the modern greek classics 133
at two monuments of Modern Greek prose which are not textually problematic, but
for which very different problems of orthographic standardization arise. The first
is the Memoirs of the Revolutionary hero, General Makriyannis (1798–1864): an
unusual classic, perhaps, but generally regarded as one – most influentially through
a celebrated essay by George Seferis – since its posthumous publication in 1907.10
That edition was made from an exceptionally rebarbative manuscript, the work
of a semi-literate who could write phonetically but not read, or not at any rate
read with any fluency. Because of the General’s bitter opposition to King Otto,
who was removed only in 1863, the work remained long hidden away in a trunk
in the garden; but the family’s later decision to publish it had a great influence on
Greek literary history. Now, apart from the immense labour of transcribing such a
manuscript into a modern standard orthography, what were the choices facing its
first editor, Yannis Vlachoyannis? And what do they have to tell us about how a
work of this character might – or should – be orthographically standardized today?
Vlachoyannis’ summary of the issues is a model of clarity and responsibility. He
begins by describing Makriyannis’ method of writing:
lacking any knowledge of spelling (ὀρθογραφία), and writing exactly as he heard,
the author renders through his writing system (γραϕή) images of words behind which
the standard form of the word [literally, ‘the form current in the customary writing
system’: ὁ ἐν τῇ συνήθη γραϕῇ ὑπάρχων τύπος] is scarcely recognisable.11
This assessment is fully borne out in the transcription of a page from the text with
which Vlachoyannis concludes his long introduction: the diplomatic transcription,
with its essentially (but not consistently) phonetic spelling and unconventional
word-breaks is very taxing to read.12
Clearly, a degree of editorial rigour which insisted on a diplomatic transcription
(even with the concession of spacing between words and minimal punctuation)
would rule out the attainment, let alone the maintenance, of classic status for this
work: it would simply be too hard to read (perhaps as hard to read as a photograph
of an ancient inscription is, even for someone who can read a conventional
transcription with ease). Yet Vlachoyannis is unflinching in acknowledging the
losses that transcription into a standard orthography (ἡ συνήθης γραϕή) entails:
10
The status of Makriyannis’ work has been disputed in various ways: (i) historians are circumspect
about the work’s value as a historical document, given its strongly partisan character; (ii) critics
committed to suspicion about the canonization process such as Lambropoulos (1988) consider the
Memoirs’ classic status to have been ideologically manufactured by a 1930s neo-nationalism; (iii) an
old and tenacious urban myth (e.g. Gorpas (1981) 119) holds that the work was in fact written by its
first editor.
11
In Makriyannis (n.d.) 82 (my translation, as with all texts cited here). Of course, the standard
writing system of 1907 was not exactly that of the mid-nineteenth century, though the more radical
simplications of Triantafyllidis (1941) (see this volume, pp. 255–7) were in the future. Bernal (2007) is
a helpful discussion of radical orthographic proposals over the period in question.
12
In Makriyannis (n.d.) 86–7.
134 david ricks
It is true on the one hand that the phonetics and phonology (ϕθογγολογία) of the
language are deprived of a curious linguistic monument, although notes and also the
samples given immediately below do something to supply the lack; but, on the other
hand history and the study of literature (ϕιλολογία) enjoy no small gain. And the
nation itself is the gainer, since it will be able to read with ease the valuable work
by General Makriyannis, which would otherwise be inaccessible to most people.
We indeed have the example of Ioannis Vilaras’ orthographical system, which has
rendered some of the works of this excellent man of letters and poet hard to read and
has thus kept the Greek people unfamiliar with them. Let the reader also bear in mind
that Makriyannis did not write in some unknown dialect or one of limited scope, but
almost in the common spoken language of the time (ἐν τῇ κοινῇ τότε λαλουµένῇ),
with but a number of dialect variants from Central Greece, where he came from . . .
And even had the editor wished out of strict scholarly rigour to present to the public
the manuscript of General Makriyannis as it is, he would have been unable to do so in
the face of the express wish and assessment of the family, with whose view, as we have
said, he is in any case in agreement.13
Vlachoyannis squarely acknowledges here the compromises involved in the

conferral of classic status (which he correctly prophesies for Makriyannis) on
texts written in an unstandardized form of the language. And he gives enough
examples, even for those who look no further than his introduction, for us to
see that his standardization itself comes at a considerable cost, in terms both of
Roumeliot flavour (itself not without ideological importance, given the dominance
of Peloponnesians, and their dominant form of Greek, in the new state) and of
probable development in Makriyannis’ idiom over the years (for example, by the
assimilation of modern constitutional terms). As it is, we can detect important shifts
in Makriyannis’ use of terms such as ‘Hellene’: it would have been nice to know
how far, if at all, these were associated with orthographical changes reflecting, say,
a more ‘learned’ morphology.14
Yet transcribing the Memoirs exactly would, without doubt, make them very
taxing to the reader. The easiest sort of case would be readily comprehensible
examples of Roumeliot accent (like τοϕεκι for τουϕέκι, ‘rifle’), but these merge
gradually into equally justifiable cases that would puzzle the reader (like πενου for
µπαίνω, ‘go in’, πατο for παντοῦ`, ‘everywhere’). And although Makriyannis’
unconventional word divisions – such as the attached enclitic in πεθαµοτου
(‘hisdeath’ for ‘his death’) or, more deviantly from standard written practice, the
merged article and epithet of τοµεγαλο (‘thebig’ for ‘the big’) – reflect a sense
of the modern language’s structure which has been the subject of discussion in
relation to the function of the ‘metrical word’ in the Modern Greek versification
system,15 it would certainly fatigue the reader who is going to Makriyannis for his
historical or ideological content.
13
In Makriyannis (n.d.) 82–3.
14
On Makriyannis’ ideology, see Holton (1984–5).
15
Mackridge (1990).
The most compelling rationale for persisting with some form of standardization
(and for doing so even if we still possessed the manuscript) is that the eccentricities
of Makriyannis’ writing practice are not in a relevant sense self-conscious.
The General casts a withering gaze on his Greek foes and has a keen eye to
posterity; he is painfully conscious of his lack of education (his preface relates
how he learned ‘letters’ only from the age of thirty-two). Yet he does not have a
considered anti-standardizing impulse when it comes to the Greek language. (It
might have been interesting, however, to hear his views about the rival memoirs
for which revolutionaries enrolled literary ghost writers using a standard form of
Greek.16) This unselfconsciousness stands in stark contrast with Makriyannis’
exact contemporary John Clare (1793–1864), who makes a useful yardstick of
comparison.
Take Clare’s militant statement:
do I write intelligable I am genneraly understood tho I do not use that awkward
squad of pointings called commas colons semicolons etc and for the very reason that
although they are drilled hourly daily and weekly by every boarding school Miss who
pretends to gossip in correspondence they do not know their proper exercise for they
even set gramarians at loggerheads.17
In the light of this class antagonism, an editor receives due warning not to perform
on Clare the same intervention that his poems (and especially The Shepherd’s
Calendar) suffered in his lifetime; and it is not surprising that Eric Robinson and
Geoffrey Summerfield leave Clare’s spelling and punctuation intact – with the
single wise concession of leaving an extra space between sentences in the prose. Yet
the case of Clare stands out as one for which the relatively much more standardized
English literary language renders the presentation of the poetry editorially arduous
but not fundamentally problematic.
There is one respect, however, in which the precise form of standardization
imposed on Makriyannis’ text could be radically altered from Vlachoyannis’
conventions, and that is in the imposition of the modern ‘monotonic’ system, on
which this volume has much to say. For if any writer stands less in need of the
traditional orthographic paraphernalia, it is he; and a popular reprint of this classic
could happily be so presented.
Now Makriyannis is such an unusual case in literary history that plainer sailing
might reasonably be expected elsewhere. That it can’t necessarily be is clear from
a brief consideration of the writer generally regarded as the greatest modern
Greek artist in prose fiction, Alexandros Papadiamantis (1851–1911). And we
can conveniently trace some of the challenges of the orthographical issues in
the successive editorial notes to N.D. Triantafyllopoulos’ magisterial edition of
1981– 8.
16
See, notably, Kolokotronis (n.d.).
17
Clare (1967) xxii.
136 david ricks
Triantafyllopoulos candidly admits that ‘the orthography in the edition has

been my permanent nightmare’. His initial position ran as follows:
The critical edition . . . of Papadiamantis would have been realized some time ago,
were it not for the difficult, not to say agonizing, problem of the orthography. In a time
of rapid orthographic normalization [ἁπλοποίησις: literally, ‘simplification’], how is
an editor to present the orthography of the Papadiamantis text, in such a way that its
lineaments (ϕυσιογνωµία) do not become unrecognisable, while also not rendering
it rebarbative to the reader?18
Reacting against those who had argued that any normalization destroys this
writer’s style and any sense of his cultural allegiances and historical context,
Triantafyllopoulos admits the possibility but fairly points out that, in a great many
matters, Papadiamantis’ orthography consisted of accidentals, in the shape of
the conventions of the many newspapers and periodicals in which his stories first
appeared. Nor does Papadiamantis’ own spelling in his letters appear to have been
at all consistent: Triantafyllopoulos cites three different forms of one very common
word, ‘octopus’ (ὀκταπόδι, ὀχταπόδι, χταπόδι) in as many pages.19

Let me take just one example which shows the dangers of harmonizing (though
in this case, such harmonizing has not occurred in our standard editions): the
celebrated short story, ‘The Seal’s Lament’ (Τὸ µυρολόγι τῆς ϕώκιας, 1908).
Papadiamantis’ spelling of the title word µυρολόγι, based on an etymology from
µύροµαι (‘melt into tears’) is certainly non-standard today, the form µοιρολόγι,
from µοῖρα (‘fate’), being overwhelmingly more common. Without wishing to take
a position on the etymological probabilities, one may still consider it important to
preserve Papadiamantis’ chosen spelling for the implications it weaves in this short,
dense and ambivalent text. In particular, the figure of the (grammatically feminine)
seal which hovers around the dead child at the story’s end is invested with some of
the qualities of the myrrh-bearing women (µυροφόραι) at the tomb of Christ, and
the seal’s ‘song’ at the end is at least as much restorative as it is fatalistic.20
Such a line of thought might be deemed by the more ruthless standardizers
to be whimsical; and, more moderately, it could reasonably be retorted that, if
the search for particularly literary ambiguity is at stake, then spelling is no bar
to speculation. So stern an anti-impressionistic view neglects the fact that, in
defiance of etymology and often with no reason of substance, certain associations
of particular spellings do take on a life of their own and become, though hardly
normative, yet suggestive. Consider an English contemporary of Papadiamantis
with whom he has much in common, Thomas Hardy, and in particular his famous
poem ‘Neutral Tones’ (written c. 1867):
We stood by a pond that winter day,
18
In Papadiamantis (1997) 4.368, 1.κδʹ. The initial orthographic table alone takes up five pages
(1.λζʹ–λθʹ).
19
In Papadiamantis (1997) 1.κζʹ.
20
Papadiamantis (1997) 4.297.
And the sun was white, as though chidden by God,

And a few leaves lay on the starving sod;

– They had fallen from an ash, and were gray.21
Editors working with living authors are entitled to try to impose standardization,
and if we consult the Oxford Style Manual we will indeed find ‘grey’ mandated
and ‘gray’ marked as ‘US’. What harm would it do Hardy’s poem if we quietly
standardized on this point? The answer is, we would occlude the fact that the
‘grey’/’gray’ distinction, however scientific, represented, in Hardy’s own time,
something of a modest cause célèbre. What we now call the OED, a work to which
Hardy was considerably indebted, has a lengthy headnote to its definitions which
begins with an acknowledgement that ‘Each of the current spellings has some
analogical support.’ It continues, in illuminating detail, as follows:
With regard to the question of usage, an inquiry by Dr. Murray in Nov. 1893 elicited
a large number of replies . . . Many correspondents said that they used the two forms
with a difference of meaning or application: the distinction most generally recognized
being that grey denotes a more delicate or a lighter tint than gray. Others considered
the difference to be that grey is a ‘warmer’ colour, or that it has a mixture of red or
brown.22
It seems to me that such considerations can hardly be irrelevant to Hardy’s poem,

with the very word ‘tones’ in its title; and this perhaps primitive exercise in socio-
linguistics is a warning not to take orthographical matters as black and white or cut
and dried.
As it happens, the number of spelling variations of this kind in Greek (for
instance, κοιτάζω, ‘look’, versus the now uncommon κυττάζω) is nowadays
greatly diminished, compared, say, with the position a century ago; but we ignore
the possibilities of orthographically generated word-play at our peril. And I would
add that the greater number of homophones in Modern Greek compared with,
say, English, only reinforces the point that close attention to, short of fixation on,
orthographical minutiae is a sine qua non of literary readings.23
We can go further: there is surely no reading constituency for which a diluted
orthography in Papadiamantis editions would have any appeal. Symptomatic here
is that Papadiamantis, along with E.D. Roïdis, is a katharevousa writer for whom
there has long been a market in demotic translation. No-one is likely to argue that
it would be possible or desirable ever to print Roïdis’ novel Pope Joan (1866) in the
monotonic system. A work written in so highly learned a form of the language,
which (jokingly subtitled, ‘A Study in the Middle Ages’) trades so heavily on a
21
Hardy (1978) 12; his prose fiction is the subject of a comparison with that of Papadiamantis in
Ricks (1988).
22
Ritter (2003) 743; OED (1933) s.v. ‘grey’.
23
Contrast the case of English: Hollander (1985) 123.
138 david ricks
parade of learning, footnotes, endnotes and all, and which quotes extensively from
classical and Biblical Greek, would be not so much diluted as mutilated by reduction
to the monotonic system.24 At any rate, there could hardly be a clearer indication
than the existence of translations of Papadiamantis and Roïdis that in such cases
the losses of standardization to modern (and, if you like, ‘thinner’) orthographical
norms are compensated for by no gains. Does this hold for the poetic classics?
George Seferis memorably wrote of ‘our three great dead poets who didn’t know
Greek’: Andreas Kalvos (Zante 1792 – Louth 1869), Count Dionysios Solomos
(Zante 1798 – Corfu 1857), and C.P. Cavafy (Alexandria 1863 – Alexandria
1933).25 For all three, lifelong residents outside the Greek state, the writing of
Greek was a conscious, in some respects exacting, and apparently self-sacrificial
decision to eschew a language (Italian in the first two cases, English in the third)
whose standardization rendered it more tractable for literary purposes – at least for
literary purposes less unusual than these poets came to adopt. This fact does not
come near to accounting for the editorial difficulties that arise for all three poets;
but even the seemingly mundane question of orthographic standardization still
poses challenges with respect to all of them.
Perhaps the two basic questions which need to be borne in mind are these. First:
with respect to these three of the four poets (Seferis is the fourth) who have done
most to influence the texture of written and even spoken Greek since, and who in
that sense operate as a standard, how tenable is it to set them over against some
other notional standard in accordance with which they may be standardized?
And secondly: are the problems relating to these Greek poets, with their eccentric
publication histories and orthographies, best taken singly as individual challenges
in literary history and to the editor’s craft, or are they to be seen as symptomatic of a
wider problem? Neither question, certainly, admits of a simple answer on the basis
of the old allegiances in the language question: demotic versus katharevousa.
Setting aside Kalvos for another occasion, let us first take the more tractable of
the two remaining cases, that of Cavafy. Here we have both a very unusual chapter
in literary history – because of the poet’s unconventional method of publication,
and the fact that the collected poems appeared only posthumously in 1935 – and
a broadly unproblematic one. That there are some persisting stumbling-blocks,
however, is clear enough from Hirst’s chapter in this volume.26 As if Cavafy’s care
with spelling were not already evident from his poems, we have a prose squib of
24
In fact the current modern edition by Alkis Angelou – Roïdis (1993) – already makes one
concession too many to modern standardization: the removal of the rough breathing over rho (especially
when Ῥώµη, ‘Rome’, is of course one of the commonest words in the book).
25
Seferis (1984) 63.
26
See below, pp. 149–66. In support of his resistance to the standardizing of the form ‘µές + σ
+ article’ (‘in’), it should be noted that the form ‘µές + bare article’, which Cavafy uses to particular
euphonic effect, is common in other poets of the period.
1901 about the correct spelling of the common name Chrestos – on which, as it
happens, the poet’s view has never prevailed.27
Hirst’s points about the modest but real damage editorial intervention has done
must be taken seriously. Fortunately, even the imperfect Cavafy editions we possess
do have a normative status that means that his idiosyncrasies run little risk of being
normalized.28 The danger to Cavafy lies elsewhere – and it is already visible in some
of the secondary literature. Readers have come to respect Cavafy’s tenaciously
held eccentricities (often on the grounds that they reflect a putative Greek current
in Alexandria): publishers are sometime inclined to economize when it comes to
citing the poet within a critical text written (quite reasonably) in the monotonic
system. Of course, literary presses and journals operate their own conventions. But
who are we to economize when we are dealing with the poetry of Cavafy?
The reasons not to do so are, it seems to me, two. In the first place, Cavafy’s
Greek, like that of Papadiamantis, is a Kunstsprache which embraces, sometimes
flamboyantly, sometimes almost unnoticed, all the layers of the language back to
Homer. (Papadiamantis, for example, very commonly uses, not the spoken word
for a drystone wall, ϕράχτης, but the ancient equivalent αἱµασιά.) If we reject
the idea of printing Ancient Greek or the koine in a modern system, then we must
by extension reject the employment of any orthographical system which makes the
layers of Greek in Cavafy’s poem separate like oil and water – not least because
where they would separate is often deliberately unclear; and such unclarity may
be as crucial to a poem as an ambiguity between direct and free indirect speech.29
Since it was Cavafy’s justifiable boast to be a continuator of the Greek Anthology,
the question is of some consequence. And how odd it would be not to accord to
the full-dress orthography of Cavafy’s poems the same respect we give his lineation
and punctuation.30
27
Kavafis (2003) 107–8. An even more trivial orthographical dispute is reported in Malanos (n.d.)
269.
28
Take, for example, the eccentric but metrically necessary spelling πιαίνει (‘goes’), as disyllable
to maintain the iambic metre; or, for the same reason, the older stress on ἀτµοσϕαῖρα (‘atmosphere’)
on the third, not the second, syllable: Cavafy (2007) 6, 70. In the first of these cases, the orthographical
distance from the standard trisyllabic spelling πηγαίνει is considerable, and the resulting closeness
to Ancient Greek πιαίνω (‘fatten’) perhaps unfortunate; but Cavafy has clearly sought to generate a
word-play with πιάνω, ‘take, grab’, in order to convey the panicky piety of the mother whose son is
missing at sea (an impression enhanced by the compression of the normal word into two syllables).
29
In a recently published anthology, this point emerges clearly in the printing of a celebrated
Cavafy poem which features such layers: Garantoudis (2008) 36. Similar perils are exhibited by
Wenham (1965) vii–viii, a New Testament manual – which, while purportedly making the language of
the NT more accessible by shearing it of learned apparatus (here, the accents), in fact makes it dead.
This is because to omit the accents is to cut off the student’s awareness that the NT koine is the ancestor
of the living language, as of course also of the language of the Orthodox liturgy which cannot reasonably
be described as dead.
30
Take the – for Cavafy, very rare – use of spaced out (ἀραιά) letters at the emotional climax of his
longest and most agonized poem: Cavafy (2007) 190. This still widely used typographical convention
for emphasis has its origins (and not for Greek only) in the limitations of hot-type technology; but it has,
it could be argued, a starkly different appearance and poetic impact from italics.
140 david ricks
The second consideration which should give us pause before licensing a

monotonic Cavafy for any use whatever is a more sweeping one; and, while it
would have been clearly understood by Cavafy, it has application to Solomos and
other poets too. To put the point with due bluntness, it is this: given the gulf of
cultural context and temperament that exists between the post-1982 monotonic
system and the more conservative orthography of the older period, one may
venture that the cultural dislocation involved in transcribing classic texts into
the monotonic is not trivial. It might readily be assumed – indeed it is implicit
in some versions of the militantly pro-monotonic argument – that orthographic
updating is of a merely typographical nature; that, to take an English analogy, the
banishing of the traditional diacritical marks is on the lines of the standardizing
of ‘j’–‘i’, ‘v’–‘u’, ‘∫’–‘s’ in English, customary in editions even by those otherwise
committed to ‘old’ spelling. Two analogies come to mind, neither exactly fitting
the Modern Greek case, but suggestive nonetheless of the considerations here. One
is the printing of texts of British English origin (most obviously Shakespeare) in
modernized American spelling for an American or world audience (‘Love’s Labor’s
Lost’). It is hard to see how, if one concedes the principle of modernizing at all, that
modernizing to an American standard is particularly objectionable – unless one is
of an incorrigibly Atticizing temperament.31
But one might ask, conversely (to take a second analogy), whether the
transcription into a radically simplified monotonic system of texts long
accompanied by their full orthographic presentation in Greek, isn’t half-way to
the employment of the Roman alphabet (so-called ϕραγκοχιώτικα); or whether
it isn’t half-way to transcribing literary texts by Serbian writers into the Roman
letters used by their Croat colleagues. However much one may resist the sacralizing
of Greek orthographic conventions in the outlook of such conservatives as the
late Archbishop Chrysostomos of Athens, their perspective on this matter is not
without force if we are to read literary texts for their literariness, or in some cases
comprehend them at all.
Here a radical difference from the problems pertaining to the standardizing of
English literary texts makes itself evident. Consider the third component of F.W.
Bateson’s original rubric for what has since become a major series, the Longman
Annotated English Poets:
Since the reader in any English-speaking country will tend to pronounce an English
poet of the past (at any rate [back] to Chaucer) as if he was a contemporary, whatever
impedes the reader’s sympathetic response that is implicit in that fact – whether of
spelling, punctuation or initial capitals – must be regarded as undesirable. A modern
pronounciation [sic] demands a modern presentation.32
31
Quite separable is the issue of unhappy modernizations, of which ‘Ardennes’ for ‘Arden’ in the
Norton As You Like It (Shakespeare (1997) 1602), with its combination of cultural assumptions and
metrical incorrectness, is an outstanding example.
32
Bateson (1970).
However suspicious one may be about the universal applicability of such principles,
it is still easy, for our purposes here, to draw a broad distinction between this and
the Greek case: it is clearly possible for a modern English reader to be attentive
to individual cases of older pronunciation that prove the rule – especially where
rhyme or word-play are in question – and adjust accordingly. With Greek, however,
where the Greek reader’s (scarcely repressible) instinct is to extrapolate the modern
pronunciation as far back as possible – customarily to Homer – semantic confusions
will abound unless a traditional orthography is available.
Despite the poor editorial standards of many existing editions when it comes to
the normalizing of Greek (what is still the most comprehensive collection of Modern
Greek literature, Βασική Βιβλιοϑήκη, is a shocking example), recent scholars have
shown greater attention to such questions.33 Take Kostis Palamas (1859–1943), a
major poet and controversialist on the subject of the Greek language: the 1950s
‘sole authorized edition’ of his collected works by the Kostis Palamas Foundation
is haphazardly standardized to mid-twentieth-century norms, prompting a more
recent editor of some uncollected poems for the same Foundation to make the
following clarification:
In the present edition the orthography and punctuation of first publication has been
maintained, for the clearest possible depiction of the orthographic decisions, as also of
the customs regarding publication and typography in the period, these being part of
the ‘historicity’ of the texts themselves.34
There follows just one concession to modern norms which is perhaps

questionable:
The only element not to have been maintained is the inverted iota with circumflex . . .
indicating synizesis, mainly for technical reasons, but also because it does not always
appear where required in the published forms of the poems.35
The omission is, yes, understandable (though one might ask why the twenty-first
century can’t run to a typographical facility that the nineteenth century could), but
it raises an important literary-historical comparison: no-one would think of printing
Gerard Manley Hopkins’s poems without the extra markings indicating the sprung
rhythm in which they are to be read (so, in ‘Spelt from Sybil’s Leaves’: ‘self ín self
steepèd and páshed — qúite / Disremembering, dismémbering.’). Admittedly, this
eccentricity (and Hopkins’s largely private publication history) is very far from
Palamas’ case; further, say, than William Barnes’s Dorset dialect poems. On the
other hand, it could be argued that the metrical innovations of the poets of the so-
33
Dimaras (1954) is an example of how orthographically poor such an edition can be, even when
the work of a major scholar.
34
In Palamas (2004) 51.
35
Ibid. Palamas for a while used this symbol as a prompt to the reader to read his lines with a
synizesis in such cases, and not, as a previous generation’s instinct would have been, to assume that
a hiatus came between adjacent vowels. Cavafy’s πιαίνει, cited in n. 28 above, would be just such a
case where the special symbol Palamas uses would make it clear that the word was a disyllable, not a
trisyllable.
142 david ricks
called Generation of 1880, over which Palamas presided, were so far-reaching that
the typographical gambits to which they resorted in trying to create the taste by
which they were to be read were integral to their entire project.36
Where do these thoughts leave us in relation to Solomos? I must stress that I am

anxious to separate the wider and still burning issues associated with the national
poet’s textual instability from the specific matter of his orthography. What is a
classic? One subsidiary, pragmatic definition may be added to the many proposed,
and it bears keenly on the Greek case: a classic is a work which lies open to
orthographic standardization. Here, as in other respects, the editing of Solomos is
very much a matter of ownership.
If we are thinking historically – or if we entertain an excessive degree of suspicion
about editors (modern, or any) – we could simply decide to read Solomos in the
form that conferred on him classic status posthumously: Polylas’ edition of 1859.
(For plenty of the lesser poems, and for the sole masterpiece he published, the
Lambros extract, this is still quite adequate.37) If, by contrast, we were zealots on
the topic of textual instability, we could insist that Solomos may only be read in the
manuscripts or perhaps their diplomatic transcription (1964) – even though such
a procedure means the frequent realignment of a large volume on the desk and the
incessant turning back and forth of its many disordered and macaronic pages.38 But
it must be somewhere between these two extremes that a workable method lies,
even when it comes to orthography.
What kind of investment in the older affiliations of the modern language seems
to mandate a conservative orthography? In the light of the linguistic views set
out by Solomos in his Dialogue (1824), there is little reason to suppose that he is
(like Cavafy or Papadiamantis) attempting to ring the changes on the historical
depth of Greek: his search is rather for some pure, ineffable, edenic language.39
We do not find in Solomos’ mature work embedded elements of Ancient Greek
language (as opposed to literary allusion of a less textually anchored kind), and his
frequent recourse to the language of the Bible and liturgy is all within the sphere
of what could be recollected orally.40 For this reason, there seems to be no reason
not to follow a modern normalization, even one in monotonic – though in the latter
36
Palamas in particular sought with considerable artistic success and subsequent influence to
eliminate from his use of the iambic fifteen-syllable line the hiatus which an earlier generation’s taste
had found acceptable, and to cultivate the effective use of synizesis. Once the campaign for synizesis
was won, of course, the typographical conventions which had been used in the battle withered away.
37
Solomos (1859).
38
Solomos (1964). Renata Lavagnini, in Kavafis (1994) 23 n. 30, comes close to saying that
Solomos should only be read in this edition.
39
On the origins and implications of Solomos’ bilingualism, see Mackridge (1994); on his utopian
poetic, see Dimoula (2009).
40
Contrast e.g. Kalvos: see Ricks (1996).
case with especial care.41 We can go further, and ask if there is not an artificiality
in printing, with the full standard orthography of the mid-nineteenth century,
works of Solomos which the manuscripts reveal to have been written more or less
phonetically – in fact, half-way to φραγκοχιώτικα, perhaps even bordering on
dyslexia.42
There is clearly much work to be done here, and Mackridge has made a strong case
for recourse to the Solomos manuscripts in order to refine the sound and sense of the
poems familiar from Polylas’ and later editions.43 It is indeed the case that Solomos
still needs rescuing here and there from Polylas’ orthographical interventions,
much as Emily Dickinson, another reclusive poet of great metaphysical sweep,
has been progressively emancipated from standardizing interventions. As her most
recent editor reports,
in 1862 she acknowledged to [Thomas Wentworth] Higginson [who first put some
of her poems in the public domain] that, in the copy of ‘Of tribulation these are they’
. . . that she was sending to him, ‘I spelled Ankle - wrong’ (as ‘Ancle’ – her usual
form), without . . . effecting a reformation in her bright orthography. She was aware
of external standards but did not strive to adhere to them, only slowly altering some
spellings. Extasy was her form until 1873, Bethleem until some time after 1874 . . . ,
opon till 1880, etherial, Febuary, retrodade and others until the end.44
The case of Solomos, however, is editorially much thornier. We have a scrupulous

but practical-minded discussion of editorial problems by Peter Mackridge.45
Without repeating his assessment, let me pluck out one or two important features
to which he draws attention in the matter of orthography, and set them within the
wider discussion here.
In Mackridge’s own selection from Solomos’ work (which includes significant
editorial advances), he sets out his policy:
Contrary to the editorial tradition, I have printed Solomos’ Greek in the monotonic
(single-accent) system and in modern orthography . . . Solomos himself took no heed
of the Greek orthographic conventions of his time.46
It would be hard to dissent from this view, or from its broad implications for how
Solomos is to be edited. In the case of a poet who, to the chagrin of his circle, and the
mordant criticism of sceptics, scarcely deigned to commit his later verse to paper,
let alone in any systematic (or even monolingual) way, and whose poetic treats the
written signs for words as but pale copies of their all-but-ineffable, uninscribable,
41
See the caveats expressed by Stylianos Alexiou, in Solomos (1994) 41; likewise in Chortatsis
(1988) 18–19.
42
Alexiou (1997) 10 n. 1.
43
Mackridge (2001). Here the issue of capitalizing the first lines of Solomos arises, especially in
the third version of The Free Besieged, where so much weight is placed on each line as a unit of sense
and sensibility.
44
R.W. Franklin in Dickinson (1999) 9–10.
45
Mackridge (2001).
46
Mackridge in Solomos (2000) xxxiv.
144 david ricks
heavenly Forms, any given orthography, whether the more conservative one of
Polylas, or Mackridge’s modernized one (or Linos Politis’ in between), will be in
more than the usual sense conventional.47 The wide variation of forms of words
as common as ἀ(γ)έρας (‘wind’) or πέλα(γ)ος (‘sea’) encourages the view that
orthographical standardization – as opposed to more radical interventions, for
instance to restore the non-standard form Πολιορκισµένοι (for Πολιορκηµένοι)
to the title of Solomos’ most famous work, The Free Besieged – is not central to
the editorial endeavour.48 But Solomos’ case – we need hardly remind ourselves
– is a very unusual one in terms of publication history – let alone in terms of the
publication history of a poet who, for a congeries of not always compatible reasons,
came to be the national poet. It is a paradox that the poet for whom (any) historical
orthography has least meaning in the presentation of his work has come to be so
central to Modern Greek literary history, including the history of the language
question.

*
Each of our four canonical writers, then, presents us with a different set of
orthographical choices, each set related to a distinctive publication history and
an often vexed reception-history. In Makriyannis’ case, the question is how we
harmonize with the conventions of a literate reading public a work which stands
between oral and literate cultures: a work, moreover, which, despite its residual
dialect colour and local animus, was addressed to a pan-Hellenic posterity and, on
occasion, on that public’s behalf to the Great Powers. As it happens, the issue has
essentially been resolved for us by the loss of the manuscript, and the one decision
for us today is whether we impose the monotonic system, as we can largely without
cost.
In the case of Papadiamantis, by contrast, we cannot hope to reduce his texts to
complete orthographic harmony with authorial warrant, because of the scattered
organs to which Papadiamantis, one of the first professional authors in Greece,
consigned his stories. But nor can we without grievous loss carry out significant
orthographic modernization (in practice, this – invariably? – means simplification)
on texts which trail so much history, including history of the language. In the case
of Cavafy, though, we have ample indications of the orthographic preferences of a
poet who refused to surrender textual control in every jot and tittle to commercial
publication; and it is only an insidious tendency to mild standardization that has
issued in any (admittedly few and minor) problems. With Solomos, finally, no
generally agreed resolution to the orthographical question is to be found – yet these
problems pale in the face of a wider textual instability.
47
On Solomos’ influence within the Heptanesian tradition from which he sprang, see Garantoudis
(2001).
48
See Mackridge (2001).
What then is to be done, if, say, we plan to construct a high-school reader

embracing all the writers we have mentioned, and more? One safe course seems
to me to print all the texts in the orthography required by the most conservative
of the texts – except where such an orthography is clearly more conservative than
is warranted by clear manuscript authority – and to do so with the inclusion of
eccentric spellings.49 A glance at ancient papyri or inscriptions reveals that users of
Greek have been mis-spelling the language for many centuries; but that does not
mean that we need not maintain periodically adjusted standards.
Alternatively, we might introduce young readers to the manuscript or first
publication in all its glory by maintaining the original orthography for each text.
This would stand against the expository method adopted – not unjustifiably – by
Horrocks’s history of Greek with his concentration on progressive development.50
One example would be Alexander Pallis’ demotic translation of the Iliad in its elegant
and accentless capital letters – a strategic swipe at the dominant orthography.51
Perhaps this policy would have greater pedagogic value, as helping readers better
to understand the complexity of the processes that go into producing a classic.
Issues about orthographic standardization are not, of course, confined to modern
Greek texts, but the history of language standards in Greek – a history inseparable
from the history of literature – makes them particularly vexed. In using English as
a comparison, I have not just reached for a comparison serviceable to users of this
volume (and within my own reach): I have also aimed to take some of the wind out
of the sails of those who read the Greek language question as a uniquely traumatic
experience for writers and readers alike.52 My aim has been to ask how far the sort
of orthographic standardization entailed by the adoption of, say, Triantafyllidis’
principles, let alone modern monotonic conventions, runs the danger of imposing
a monochrome colour on even the central texts of a culturally heterogeneous
Modern Greek tradition.53 The widespread adoption of the monotonic system as
49
This would include eccentric spellings such as Cavafy’s, as well as significant mis-spellings. For
example, Seferis gives a rough breathing to the word ἀλακάτιν (Cypriot dialect word for ‘well-wheel’,
derived from Ancient Greek ἠλακάτη, ‘distaff’) in his poem ‘Λεπτοµέρειες στην Κύπρο’ (‘Details on
Cyprus’: Seferis (1982) 235). With respect to an anti-British poem of the time of the Cyprus rising, it is
legitimate to point out that the claim to linguistic and cultural authenticity through the preservation of
the Homeric ἠλακάτη in Cypriot dialect is weakened by the slip.
50
Horrocks 1997 (298), with respect to the Ottoman period onwards: ‘Since we are now dealing
indisputably with Modern Greek, the monotonic system of accentuation (actually adopted in 1982)
will be employed henceforth.’ Isn’t the ‘actually’ here just a little casual?
51
Pallis (1917).
52
Crisp revisionist accounts of the Modern Greek ideological considerations here, by Elli Skopetea
and Antonis Liakos, are to be found in Christidis (2007) 1280–95.
53
The goalposts are always moving: Mackridge’s descriptive study (1985: viii) continues to use
the polytonic, while his grammar with Holton and Philippaki-Warburton (1997: xxi) has conformed
to the monotonic.
146 david ricks
the norm in the printing of classic literary texts – unlikely as that possibility still
seems – would, I believe, raise problems greater still.54
References
Alexiou, S. (1997), Σολωµιστές καὶ Σολωµός, Athens.
Bateson, F.W. (1970), ‘Note by General Editor’, in J. Keats, The Complete Poems, ed. M.
Allott, London: xiii.
Bernal, J.M. (2007), ‘Spelling and script debates in interwar Greece’, Byzantine and Modern
Greek Studies, 31: 170–90.
Cavafy, C.P. (2007), The Collected Poems: A New Translation by Evangelos Sachperoglou with
Parallel Greek Text, ed. A. Hirst, Oxford.
Chortatsis, G. (1988), Ἐρωϕίλη, ed. S. Alexiou, Athens.
Christidis, A.-F. (ed.) (2007), A History of Ancient Greek, Cambridge.
Clare, J. (1967), Selected Poems, ed. E. Robinson and G. Summerfield, Oxford.
Dickinson, E. (1999), Poems, Reading Edition, ed. R.W. Franklin, Cambridge, Mass.
Dimaras, C.Th. (ed.) (1954), Ποιηταὶ τοῦ 19οῦ αἰῶνος [Βασική Βιβλιοϑήκη], Athens.
Dimoula, V. (2009), ‘The nation between utopia and art: canonizing Dionysios Solomos as
the “national poet” of Modern Greece’, in R. Beaton and D. Ricks (eds.), The Making of
Modern Greece, Aldershot: 201–12.
Foskolos, M.A. (1980), Φορτουνάτος , ed. A.L. Vincent, Herakleion.
Garantoudis, E. (2001), Οι Επτανήσιοι και ο Σολωµός, Athens.
Garantoudis, E. (ed.) (2008), Η ελληνική ποίηση τοῦ 20ού αιώνα, µια συγχρονική
ανϑολογία, Athens.
Gorpas, Th. (1981), Περιπετειῶδες, κοινωνικὸ και µαῦρο νεοελληνικὸ ἀϕήγηµα, vol. 1,
Athens.
Hardy, T. (1978), The Complete Poems, ed. J. Gibson, London.
Hollander, J. (1985), Vision and Resonance, 2nd edn, New Haven, Conn.
Holton, D. (1984-5), ‘Ethnic identity and patriotic idealism in the writings of General
Makriyannis’, Byzantine and Modern Greek Studies, 9: 133–60.
Holton, D., Mackridge, P. and Philippaki-Warburton, I. (1997), Greek: A Comprehensive
Grammar of the Modern Language, Cambrige.
Kalvos, A. (n.d.), Ὠδαί 1–20, ed. M.G. Meraklis, Athens.
Kavafis, K.P. (1994), Ἀτελῆ ποιήµατα, ed. R. Lavagnini, Athens.
Kavafis, K.P. (2003) Τα πεζά, ed. M. Pieris, Athens.
Kolokotronis, Th. (n.d.), Ἀποµνηµονεύµατα, ed. T. Vournas, Athens.
Lambropoulos, V. (1988), Literature as National Institution, Princeton.
Mackridge, P. (1985), The Modern Greek Language, Oxford.
54
A particularly regrettable example, in relation to what has already become a modern classic, is
the reprint of Thanasis Valtinos’ Στοιχεῖα για τη δεκαετία τοῦ 60 (first published in 1989): Valtinos
(2001). Here alternating pages of demotic and katharevousa glare at each other in such a way as to show
the cultural cleavages of post-war Greece; the effect is much reduced by printing the purist passages in
an anachronistic monotonic orthography.
Mackridge, P. (1990), ‘The metrical structure of the oral decapentasyllable’, Byzantine and
Mackridge, P. (1994), ‘Dionisio Salamon/Dionysios Solomos: poetry as a dialogue between
languages’, Dialogos, 1: 59–76.
Mackridge, P. (2001), ‘Οι µεταµορϕώσεις των Ελεύϑερων Πολιορκηµένων· από τα
χειρόγραϕα στις εκδόσεις’, Ελληνικά, 51: 109–39.
Madan, G. (1980), Notebooks, Oxford.
Makriyannis, Y. (n.d.), Ἀποµνηµονεύµατα, ed. Y. Vlachoyannis, Athens.
Malanos, T. (n.d.), Ὁ ποιητής Κ.Π. Καβάϕης , 3rd edn, Athens.
OED: Oxford English Dictionary (1933), Oxford.
Panayotakis, N. (ed.) (1993), Origini della letteratura neogreca, vol. 1, Venice.
Palamas, K. (2004), Ποιήµατα στον Ραµπαγᾶ και το Μη Χάνεσαι, ed. G. Andreiomenos,
Athens.
Pallis, A. (1917), Ἡ Ιλιάδα, 3rd edn, Liverpool.
Papadiamantis, A. (1997), Ἅπαντα, ed. N.D. Triantafyllopoulos, 5 vols, Athens.
Ricks, D. (1988), ‘Alexandros Papadiamantis and Thomas Hardy’, in R. Beaton (ed.), The
Greek Novel and its Influence, A.D. 1–1985, London: 23–30.
Ricks, D. (1996), ‘The progress of poesy: Kalvos, Gray and the revival of ancient literary
language’, Modern Greek Studies (Australia and New Zealand), 4: 111–32.
Ritter, R.M. (2003), Oxford Style Manual, Oxford.
Roïdis, E.D. (1993), Ἡ Πάπισσα Ἰωάννα, ed. A. Angelou, Athens.
Seferis, G. (1982), Ποιήµατα, Athens.
Seferis, G. (1984), Δοκιµές 1 (1936–1947), Athens.
Shakespeare, W. (1997), The Norton Shakespeare, ed. S. Greenblatt et al., New York.
Solomos, D. (1859), Τα εὑρισκόµενα, ed. I. Polilas, Corfu.
Solomos, D. (1964), Αὐτόγραϕα ἔργα, ed. L. Politis, 2 vols, Thessaloniki.
Solomos, D. (1994), Ποιήµατα και πεζά, ed. S. Alexiou, Athens.
Solomos, D. (2000), The Free Besieged and Other Poems, ed. P. Mackridge, Beeston, Notts.
Triantafyllidis, M. (1941), Νεοελληνική γραµµατική (τῆς δηµοτικῆς), Athens.
Valtinos, Th. (2001), Στοιχεία για τή δεκαετία του 60, Athens.
Vejleskov, P. (ed.) (2005), Ἀπόκοπος, Cologne.
Wenham, J.W. (1965), The Elements of New Testament Greek, Cambridge.
8
Correcting the Courtroom Cat:

Editorial Assaults on Cavafy’s Poetry*
Anthony Hirst
On 6 July 2003, on one of the main British television channels, there was a
programme about the poet Philip Larkin1, in which the verb in Larkin’s famous
line, ‘They fuck you up, your mum and dad’,2 was ‘bleeped out’. Objecting to
this in The Guardian the following day, the television critic Nancy Banks-Smith,
invoking Coleridge’s ‘homely’ definition of poetry,3 wrote as follows:
A great poem must not be touched, tweaked or buggered about. It is already the best
words in the best order. On the subject of language, a poet is the ultimate arbiter, the
judge, the jury and the courtroom cat.4
The principle rather colourfully enunciated here – that the author is the final
arbiter in matters of language – is one widely accepted among literary editors in
the English-speaking world. Where authors no longer living are concerned, it is
often expressed in terms of the primacy of the author’s last discernible intention for
a text. Catherine Phillips, for example, referring to her edition of Gerard Manley
Hopkins, based primarily on manuscripts, states that ‘in all cases the version which
I believe to be that last written has been taken for text’, and goes on to describe
* The research on which this chapter is based was funded by the Leverhulme Trust through a
Special Research Fellowship (2002–4) and the Research and Regional Services of Queen’s University
Belfast (2002–6), and aided by a Friends of Princeton University Library Short-Term Visiting
Fellowship (2002). I am grateful to the funding bodies concerned and to the librarians and archivists of
institutions too numerous to mention here, and also to Manos Haritatos, director of the Greek Literary
and Historical Archive (ELIA) in Athens, who gave me access to materials from his personal collection
as well as to the resources of ELIA. Some of the material in this paper has appeared in different form in
Hirst (2002) and in my ‘Note on the Greek text’ in Cavafy (2007) xxxiv–xxxix.
1
‘Philip Larkin: love and death in Hull’, dir. Ian MacMillan, Channel 4.
2
Larkin (1988) 180: the first line of ‘This be the verse’.
3
‘I wish our clever young poets would remember my homely definitions of prose and poetry; that
is, prose = words in their best order; – poetry = the best words in the best order’: Coleridge (1990), vol.
2, 68; cf. vol. 1, 90.
4
Banks-Smith (2003) 22.
149
150 anthony hirst
her painstaking attempts to distinguish added punctuation marks in ink made

by Robert Bridges (an earlier editor of Hopkins) from those also in ink made by
Hopkins himself.5 Similarly, Samuel Hynes, after outlining the extraordinarily
complex history of the texts of Thomas Hardy’s poetry,6 states that ‘the copy texts
for [his] edition are the first editions of Hardy’s first seven volumes of poetry’,
along with the holograph of the last (posthumous) volume, ‘emended according to
the following principles’:
I have considered substantive revisions that reached print to be expressions of Hardy’s
fixed intention, and have incorporated these revisions into the established text. Where
there is more than one revised version of a line or a word, I have taken the latest one
where chronology can be determined with certainty (as is usually the case); in those
instances where the chronology is uncertain, I have relied on my critical judgement
and my sense of what is characteristic of Hardy’s mature style.7
In the ‘Preface’ to his edition of Tennyson’s In Memoriam, Robert H. Ross

identifies the Eversley Edition8 as ‘the most reliable text’, and adds, indicating the
extent of his respect for the courtroom cat:
With the exception of changing the section headings from Roman to Arabic numerals,
I have allowed the Eversley text to stand. [. . .] I have not, for instance, modernized
Tennyson’s spelling (e.g. “though” for “tho’”), nor have I supplied the e for the poet’s
apostrophe in such words as heav’n. I have tried to retain as nearly as possible the
visual and aural qualities of the nineteenth-century text.9
And it is not only editors of poetry who aspire to such fidelity, as the following
quotation from the prefatory note to the Penguin edition of the novels of Jane
Austen indicates:
The editorial policy is one of minimum intervention: no attempt has been made to
modernize the spelling or punctuation, or to render spellings consistent so long as the
variant spellings were acceptable in the period.10
If only Cavafy’s Greek editors could have said the same! But such meticulous
respect for the author’s text has not been common among editors of modern Greek
literary texts. As we shall see, Cavafy left very little for his editors to do. There are no
problems of textual history and very little room for doubt about his final intentions
for most of the poems. But his editors have not been content to do only the very
little that he left them; instead they have done a great deal that did not need to
be done and, in terms of the principles exemplified above, ought not to have been
done, progressively standardizing, ‘correcting’ and modernizing his orthography.
5
In Hopkins (1986) xli.
6
In Hardy (1982) xii–xxiii.
7
In Hardy (1982) xxiv.
8
Tennyson (1909) 39–184.
9
In Tennyson (1973) vii.
10
Claire Lamont, Textual Adviser, in a prefatory note entitled, ‘The Penguin edition of the novels
of Jane Austen’, appearing on pages vii–viii of each volume in that series.
correcting the courtroom cat 151
Here I speak only of the editors of the principal posthumous editions, and not of
those periodical and anthology editors among his contemporaries, whose far worse
crimes had, mercifully, less far-reaching and less enduring effects.
Κ.Π. Καβάφης, or C.P. Cavafy as he styled himself when writing in English,
was born in Alexandria on 29 April 1863 and died, also in Alexandria, on his
seventieth birthday in 1933. As a poet he was a late developer and may be regarded
as essentially a twentieth-century poet, since 153 of the 154 short poems on which
his now very great reputation primarily rests were first published in the period
1897 to 1932, while the 154th was published only after his death. A twentieth-
century poet, and the first Greek modernist: a modernist in his subject matter, in
his diction, and in the freedom with which he adapts traditional metrics (though
without ever approaching free verse).
The first commercial edition of Cavafy’s poetry appeared only in 1935,
two years after his death; self-publication had been Cavafy’s chosen means of
disseminating his work. Many Greek poets, George Seferis among them, began
with self-publication but were glad to find commercial publishers as soon as their
work began to be known. Cavafy, though, never moved in that direction. And
unlike Seferis’ self-published volumes, Cavafy’s were never placed in bookshops,
indeed were never offered for sale at all.11 They were printed at his own expense
and distributed by him, directly or through trusted friends, to individuals and, less
often, institutions.12 Beginning with a series of five leaflets, each containing a single
poem, or, in one case, two poems, in the period 1891–1904, followed by two small
volumes of poetry, Poems 1904 (1905) and an enlarged version, Poems 1910 (1910),
the practice which Cavafy gradually evolved was to circulate his poems in two types
of collection: bound booklets in which the poems were arranged thematically, and
folders of unbound broadsheets of the more recent poems, arranged (and pinned
together) in chronological order by publication date.13 Some of the later poems
were only ever printed once; but some poems first published around 1911 were
printed and reprinted twelve, thirteen or fourteen times over a period of about
twenty years.14
It is a reasonable, even a safe, perhaps an unchallengeable assumption that,
apart from printers’ errors, Cavafy’s last printing of each poem represents his latest
11
Seferis’ Στέρνα (1932) printed in 50 copies was marked ‘Not for sale’, but of the 200 copies of
the earlier Στροφὴ (1931) half were placed in bookshops; only 20 were sold in the first five months, but
a total of 90 by 1939: Beaton (2003) 97, 115.
12
Cavafy’s distribution lists, which run from 1905 to 1933, have been published in Savidis (1966)
215–83. The lists are extensive, recording the recipients of almost 2,500 items, but it is clear that Cavafy
sometimes forgot to record the donation or despatch of a collection, since I have seen a number of
collections with autograph dedications which cannot be matched to entries in his distribution lists.
13
For full details of Cavafy’s self-publication practices, see Savidis (1966), and for a summary in
English, see Hirst (1995) 33–7, 40–3, 90–3.
14
Savidis’ list of Cavafy’s printings (Savidis (1966) 299–323) does not show any poem printed
more than thirteen times, but I have identified a number of printings which do not appear in his lists,
including one additional printing each of «Ἡ σατραπεία» and «Στὴν ἐκκλησία» and two additional
printings of «Ἰθάκη», bringing the total in each case to fourteen.
152 anthony hirst
intentions for that poem, modified only by any autograph revisions in copies of
the last printing. The earliest sixteen poems of the Cavafy canon, the poems first
published in the period 1897–1904, present a particular problem, since they were
never reprinted by Cavafy after 1910; and the earliest poem «Τείχη» (‘Walls’) was
only ever printed once by Cavafy, in leaflet form, in 1897. The other fourteen
poems (together with seven later poems) were printed in the second of the author’s
own editions, Poems 1910. Some of these poems were revised by Cavafy after 1910
and republished in periodicals. There is an autograph manuscript version of Poems
1910 (with the addition of «Τείχη») known as ‘The Sengopoulos Notebook’, dated
by Savidis to 1927 and published in facsimile15 in 1968. In theory this manuscript
should provide us with the definitive texts of the first sixteen poems. However, not
only is the manuscript text full of accidental minor errors (especially the omission
of diacritics and punctuation marks), it also contains variants which suggest that
Cavafy was, in some cases, copying from relatively early versions of the poems
which had been superseded by his own later revisions. In addition to Poems 1910
and the 1927 manuscript, any editor of the earliest sixteen poems of the canon
needs to take account of the versions of those poems republished, with the author’s
approval, in various issues of Νέα Τέχνη (1924–5) and Ἀλεξανδρινὴ Τέχνη
(1926–9), and especially those in the latter journal, of which Cavafy was himself in
effect an editor, in conjunction with the named editor, Rika Sengopoulou.
The last poem of the canon, «Εἰς τὰ περίχωρα τῆς Ἀντιοχείας» (‘In the
outskirts of Antioch’) exists only in manuscript and, though prepared for the
printers, it was never printed or published in Cavafy’s lifetime.16
Leaving aside the special problems posed by the earliest 16 and the last of
the 154 poems of the canon, we can say that as far as the other 137 poems are
concerned, all an editor needed to do was to reproduce the poet’s last printing, only
correcting any obvious typographical errors (of which there are very, very few) and
taking account of any corrections made by Cavafy himself. This, though, has not
happened until very recently.17
Cavafy has had two principal editors. The first was Rika Sengopoulou, already
mentioned as the editor of the journal Ἀλεξανδρινὴ Τέχνη, who was the wife
of Alekos Sengopoulous, Cavafy’s younger friend and neighbour to whom he
left most of his money, his property, his books and his papers (the papers now
known as the Cavafy Archive). It was Rika Sengopoulou who in 1935 produced
15
Cavafy (1968).
16
A manuscript (perhaps the only extant manuscript) of this poem was published in facsimile in
Νέα Ἑστία 872 (1 November 1963): 1486–7.
17
This is, in principle, what I have done in my edition of the Greek text for a dual-language volume
in the Oxford World’s Classics series, Cavafy (2007), though I did, with reluctance, agree to some
minor elements of standardization requested by the Oxford University Press (the number of points
in ellipses standardized as three, for example). With the earliest sixteen poems, I used in all cases (for
simplicity) the 1927 ‘Sengopoulos Notebook’ as the copy text, but in the critical edition which I am
preparing a decision on the copy text will be taken (and defended) individually for each of these sixteen
poems.
the first commercial edition of Cavafy’s collected poetry, the 154 poems of the
Cavafy canon; it was in fact this edition which defined the canon. The second
principal editor was the late George Savidis, who acquired the Cavafy Archive
from Sengopoulous in the 1960s. Savidis produced two editions: the first in 1963,
with numerous reprints, some including minor corrections; the second in 1991,
again with later reprints and minor corrections. Reprints of Savidis’ 1991 edition
(inferior, I would argue, in several ways to that of 1963)18 constitute the current
standard edition of the poet’s work. Between Sengopoulou and Savidis stand the
three editions, from the Ikaros publishing house, of 1948, 1952 and 1958. No
editor is named, and these editions, described on their title pages as the ‘second’,
‘third’ and ‘fourth’ editions, are based on Sengopoulou’s 1935 edition (the implied
‘first’ edition), with some systematic and many accidental changes.
Savidis stated in the introduction to his 1963 edition that he had based the text of
his edition primarily on Cavafy’s last printing of each poem, though acknowledging
that he had also consulted earlier printings and selected versions published in
periodicals.19 It is also evident that Rika Sengopoulou based her edition primarily
on Cavafy’s latest printings; but it is painfully obvious, when one begins to make
detailed comparisons, that neither Sengopoulou nor Savidis regarded the poet’s
latest version of the poems as imposing any strict obligation upon them as editors,
and they have often preferred readings from earlier printings or even from versions
in periodicals. For them the poet was not ‘the ultimate arbiter, the judge, the jury
and courtroom cat’. The roles of judge and jury they took upon themselves; but
I like to think of Cavafy as the courtroom cat, looking down on them from the
gallery with a mildly offended and distinctly haughty expression.
In all these editions there has been much editorial alteration of Cavafy’s use of
accents, even though Cavafy was certainly not alone in the early decades of the
twentieth century in any of his accentual practices: they would no doubt have
passed the test of historical acceptability with Jane Austen’s current editors.20
Let us take the two little words τὶ and ποῦ. Τὶ (meaning ‘what’, interrogative or
exclamatory) is, after 1910, almost invariably found with a grave accent in Cavafy’s
printings.21 This use of τὶ with a grave was the practice of some publishing houses
and literary periodicals in Cavafy’s day, but it was rejected by Sengopoulou and
Savidis, whose editions invariably print it with an acute.
Ποῦ is a word with several meanings. The convention, more firmly established
after Cavafy’s death than it was in his lifetime, was to distinguish by means of
18
See Ekdawi and Hirst (1999) 90–3.
19
In Cavafy (1963a) vol. 1, 12.
20
In the following discussion of linguistic details, I am grateful to Peter Mackridge for his careful
reading of an earlier version of the material and for his correction of a number of errors in my descriptions.
Any remaining errors are my responsibility and not his.
21
The four instances of τί (with an acute) in Cavafy’s final printings can be confidently treated as
printers’ errors, since stacked against them are over fifty instances of τὶ (with a grave). In one case both
forms occur in the same line in parallel phrases: Τὶ ὡραῖο παιδί· τί θεῖο μεσημέρι τὸ ἔχει / παρμένο
(«Ζωγραφισμένα», lines 9–10, seventh and final printing, 1930).
154 anthony hirst
accents two forms of the word: ποὺ (with a grave) was used for the all-purpose
relative pronoun and for the conjunction meaning ‘that’; while ποῦ (with a
circumflex) was reserved for the less frequent interrogative adverb meaning
‘where’. Cavafy, and many of his contemporaries, wrote ποῦ with a circumflex
whatever the meaning; and this is what we find in his printings.22 In this case,
Sengopoulou went along with Cavafy’s usage in her edition, but the Ikaros editors
and then Savidis imposed the distinction which Cavafy had chosen not to make.
There are a number of other pairs of homonyms which Cavafy did not distinguish
by means of diacritics. He never distinguished between γιατὶ (‘because’) and γιατί
(‘why’), or between the conjunction ὅτι (‘that’) and the pronoun ὅ,τι (‘what’),
or between the conjunction ὡς (‘as’) and the preposition ὣς (‘until’, ‘up to’, ‘as
far as’), or between the exclamatory or interrogative adverb πῶς (‘how’) and the
conjunction πὼς (‘that’); in each case he used the first of the two forms to the
exclusion of the other, whatever the semantic or syntactical context. Γιατί and ὅ,τι
were introduced by Sengopoulou where the former is interrogative or the latter is
a pronoun, and these forms were reproduced in the three Ikaros editions and in all
the editions of Savidis; πὼς (like ποὺ) was not introduced by Sengopoulou but is
found in all later editions, while ὣς does not appear until Savidis’ 1963 edition.
Minor variations and not worth worrying about? Perhaps; and it could be argued
that, however inauthentic, these editorial interventions add greater precision to the
text. If the standardization and modernization of Cavafy’s texts had gone no further
we would have no great cause for complaint. More unfortunate, to my mind, is a
development which appears for the first time in Savidis’ 1991 edition. This is a
systematic change in the representation of elision where it occurs between either
the conjunction νὰ (introducing subjunctives) or the particle θὰ (which marks
future and conditional tenses) or one of the neuter object pronouns, τὸ and τὰ, and
a following verb beginning with a stressed vowel. Cavafy’s preference, and that
of most of his contemporaries, was to combine νὰ, θὰ, τὸ or τὰ with the verb to
form a single word; and in these cases, as can be seen in the examples in the middle
column of the table below, the initial vowel of the verb is elided, and displaced by
the vowel of the preceding word, though the latter retains the diacrictics proper to
the displaced initial vowel of the verb. The following table shows a representative
sample of the more than sixty instances of this phenomenon:
22
The two instances of ποὺ (with a grave) in the final printings are clearly printers’ errors – there
are over 300 instances of ποῦ (with a circumflex), in all meanings, but predominantly in those meanings
which for later editors would require ποὺ.
Full Form Cavafy Savidis 1991

νὰ ἔπαιρνεν νἄπαιρνεν νά ᾿παιρνεν
νὰ ἔρθουν νἄρθουν νά ᾿ρθουν
νὰ εἶναι νἆναι νά ᾿ναι
θὰ ἤθελα θἄθελα θά ᾿θελα
τὰ ἔνοιωσαν τἄνοιωσαν τά ᾿νοιωσαν
τὸ ἔχουν τὄχουν τό ᾿χουν
τὸ ἔφθανε τὤφθανε τό ᾿φθανε
The last of the examples above presents a slightly different case. In Cavafy’s
poem «Ὁ ἥλιος τοῦ ἀπογεύματος» (‘The sun in the afternoon’) lines 15–16
read:
Πλάϊ στὸ παράθυρο ἦταν τὸ κρεββάτι·
ὁ ἥλιος τοῦ ἀπογεύματος τὤφθανε ὡς τὰ μισά.
Beside the window was the bed;

the sun in the afternoon would reach halfway across.
[literally: would reach it up to the middle]
Here the object prounoun τὸ has been combined with the verb ἔφθανε, but
instead of the omicron of τὸ displacing the epsilon of ἔφθανε (as in the preceding
example, τὄχουν), the two vowels have combined to form a third, omega, though
this formation does not follow the normal rules of vowel contraction, where omicron
+ epsilon = omicron-upsilon (and even if Cavafy had the indirect object pronoun τοῦ
in mind instead of τὸ, the result should still have been omicron-upsilon). What is
happening here looks more like what classical philologists call crasis (contraction)
rather than elision, and Savidis’ change in the orthography is thus, in this case,
a little more radical than simply representing elision by a different orthographic
convention.
There is something unsightly about Savidis’ preferred ‘modern’ forms with their
headless verbs which begin with an apostrophe and carry no accent. These forms
are arguably irrational, certainly anachronistic in relation to Cavafy’s poetry, and –
and this is what really matters – were never used by Cavafy himself. Nevertheless,
this is, in most cases, only substituting one orthographic convention for another
and (even with the substitution of omicron for omega) the pronunciation is not
affected.
I move on now to a more interesting case of editorial intervention where
pronunciation is, just possibly, affected. The adverb μέσα, meaning ‘inside’,
combines with the preposition σὲ to form a prepositional phrase μέσα σὲ (‘in’,
‘inside’, ‘within’, ‘among’). When σὲ is followed by a definite article the epsilon is
dropped and the sigma is added to the article, as in a phrase such as μέσα στὴν
καρδιά μου (‘within my heart’).23 In such phrases, μέσα is often abbreviated
23
The στ- forms may be explained in another way, as deriving from katharevousa εἰς τ- (σὲ being
the demotic equivalent of εἰς). Clearly some editors and publishers saw it this way and printed the
abbreviated form as ᾿ςτ-. See, for example, discrepancies 9 and 22 in Skokos’ version of ‘Since nine
156 anthony hirst
to μὲς (μὲς στὴν καρδιά μου). There are dozens of such phrases in Cavafy’s
printings of his poems. But there are also three poems whose printings show a
further abbreviation: the omission of the sigma derived from σὲ (μὲς τὴν καρδιά
μου). Indeed the very phrase μὲς τὴν καρδιά μου (in this form) occurs in line 2
of both of Cavafy’s printings (1922 and 1926) of his poem «Πρὸς τὸν Ἀντίοχον
Ἐπιφανῆ» (‘To Antiochus Epiphanes’). Such usage is unusual, though not
unprecedented,24 and, from a strict grammarian’s point of view, plain wrong, since
the adverb (μὲς for μέσα) is now connected directly to the noun phrase as though
it was itself a preposition. Could these few instances in Cavafy’s poems be printers’
errors? A brief survey of the evidence will show that they are not. Here is the
text of the short poem «Ὅταν διεγείρονται» (‘Whenever they’re aroused’), as it
appears in Cavafy’s sixth and final printing of the poem (1930):
Προσπάθησε νὰ τὰ φυλάξεις, ποιητή,
ὅσο κι ἂν εἶναι λίγα αὐτὰ ποῦ σταματιοῦνται.
Τοῦ ἐρωτισμοῦ σου τὰ ὁράματα.
Βάλ᾿ τα, μισοκρυμένα, μὲς τὲς φράσεις σου.
Προσπάθησε νὰ τὰ κρατήσεις, ποιητή,
ὅταν διεγείρονται μὲς τὸ μυαλό σου
τὴν νύχτα ἢ μὲς τὴν λάμψι τοῦ μεσημεριοῦ.
Strive, poet, to preserve them,

however few are those that still remain.
Your own erotic dreams.
Set them, half-hidden, among your phrases.
Strive, poet, to keep hold of them,
whenever they’re aroused within your brain
at night or in the glare of noon.25
The three phrases in bold of the type μὲς τ– in fact occur in all the previous
printings of this poem, and the first printing of 1916 constitutes the earliest
appearance of this phenomenon in Cavafy’s poetry. It is inconceivable that three
printers’ errors so close together could have survived through six printings of the
poem spread over a period of fifteen years without Cavafy spotting them. The form
of the phrases is considered and deliberate. And there are further examples.
The first three printings (1916, 1917 and 1919) of «Μανουὴλ Κομνηνὸς»
(‘Manuel Comnenus’) have, in the last line, the phrase μὲς στὴν πίστι των (‘in
their faith’) but in all subsequent printings, from the fourth (1921) to the seventh
and last (1930), the ironic comment that concludes the poem reads as follows:
o’clock —’ below.
24
As Peter Mackridge pointed out when this paper was first delivered at the Logos Conference,
examples can be found in the writings of many of Cavafy’s contemporaries. The Proïa dictionary
(Zevgolis 1933–4) records both μὲς τ- and μὲς στ- constructions in its entry for μέσα. I am grateful to
Peter Mackridge for bringing this dictionary to my attention.
25
Translations of Cavafy’s poetry in this chapter are my own, taken from Cavafy (forthcoming).
Εὐτυχισμένοι ὅλοι ποῦ πιστεύουν,

καὶ σὰν τὸν βασιλέα κὺρ Μανουὴλ τελειώνουν
ντυμένοι μὲς τὴν πίστι των σεμνότατα.
How fortunate are all those who believe

and, like the emperor Lord Manuel, meet their end
dressed in their faith most reverently.26
In the final example, the question of printers’ errors does not arise, as no printers
were involved. «Φωνὲς» is one of the poems which Cavafy never printed after
1910. The 1910 version of the poem contains the phrases μὲς στὰ ὄνειρά μας
and μὲς στὴν σκέψι (lines 4 and 5), but in the autograph ‘Sengopoulos Notebook’
(1927) we find, referring to the ‘voices [. . .] of those who have died or of those who
are, / for us, lost like the dead’:
Κάποτε μὲς τὰ ὄνειρά μας ὁμιλοῦνε·
κάποτε μὲς τὴν σκέψι τὲς ἀκούει τὸ μυαλό.
Sometimes within our dreams they speak;

sometimes, immersed in thought, the mind may hear them.
There are, then, four poems for which the author’s clearly discernible latest
intention is that they should contain phrases of the μὲς τ– type. And yet, apart
from the odd republication of one of these poems in periodicals, none of these
μὲς τ– phrases have been reproduced by editors until my own recent edition.27
Sengopoulou, the unnamed Ikaros editors, and Savidis (in both 1963 and 1991)
restored in all cases the sigma which Cavafy had clearly chosen to omit. Did they
think that μὲς τ– was a mistake (on the part of Cavafy or his printers) in every
case? The evidence shows that these cannot have been mistakes. Did they think
that μὲς τ– was too embarrassingly wrong to be reproduced? If, as can hardly be
denied, Cavafy deliberately chose, originally or in later revision, the form μὲς τ– in
four poems, while using or keeping the standard form μὲς στ– in some thirty-nine
other poems (fifty-six instances), it must have some significance. Even if they are
not sure what that significance is, it is not the job of editors to deprive the reader
of the evidence.
Two possible aspects of the significance of Cavafy’s occasional preference
for μὲς τ– over μὲς στ– may be suggested. He may want in those four poems to
give the visual impression of a more colloquial or intimate discourse by using a
mildly transgressive written form. Instead, or as well, the phenomenon may be
an indicator (there are others) of Cavafy’s increasing sensitivity to the sound of
his poems and increasing interest in the possibility of conveying acoustic nuances
by graphic means. A phrases such as μὲς στὴν καρδιά μου would normally, in
ordinary conversation, be pronounced μὲς τὴν καρδιά μου, with no prolongation
26
For a justification of translating σεμνότατα as ‘most reverently’ (or ‘most piously’), see Hirst
(2000) 48–55, especially 54–5.
27
See Cavafy (2007) 2, 64, 96, 138.
158 anthony hirst
or double articulation of the sibilant. Perhaps what Cavafy wants to indicate is

that, even in the most deliberate recitation, the two words (μὲς and [σ]τὴν in this
example) should be run together without the slightest sense of a gap. Whatever
the explanation, the point is that there is an enigma here – an enigma of potential
if minor poetic significance – which Cavafy’s principal editors have taken upon
themselves to hide from his readers.
Other examples of Cavafy’s increasing sensitivity to the sound of his poetry can
be found in many of his very late minor corrections, corrections which appear for
the first time in the last or next-to-last printings of the poems. Among these late
corrections there are several instances of the removal or addition of a comma, or of
the removal or addition of an optional final nu. It is surprising how many of these
late changes Savidis did not accept, especially where the deletion of commas was
concerned. Let us take as an example the opening lines of one of Cavafy’s best
known poems, «Ἀπολείπειν ὁ θεὸς Ἀντώνιον» (‘The god abandoning Antony’):
Σὰν ἔξαφνα* ὥρα μεσάνυχτ᾿* ἀκουσθεῖ
ἀόρατος θίασος νὰ περνᾶ
μὲ μουσικὲς ἐξαίσιες, μὲ φωνὲς —
τὴν τύχη σου ποῦ ἐνδίδει πιά, τὰ ἔργα σου
ποῦ ἀπέτυχαν, τὰ σχέδια τῆς ζωῆς σου
ποῦ βγῆκαν ὅλα πλάνες* μὴ ἀνοφέλετα θρηνήσεις.
When suddenly* at midnight* there is heard

an unseen band of revellers pass by
with their exquisite music and their cries —
your luck which here runs out, your exploits
which failed, and your life’s plans all proved
delusions now* don’t mourn to no avail.
Here, in the thirteenth and final printing (1930), in the places marked by
asterisks, Cavafy has removed three commas which were present in most or all
earlier printings (1911–1926),28 two from the first line and one from the sixth.
Sengopoulou followed Cavafy’s last printing here (and the Ikaros editors followed
Sengopoulou), but Savidis, in all of his editions, restored the commas found in the
penultimate and other earlier printings. Did he think the omission of the commas
in the 1930 version were printers’ errors? Or did he just prefer the earlier reading?
Three such identical errors within so close a space are unlikely, and I am sure that
Sengopoulou was right to accept the deletions.
The comma after μεσάνυχτ᾿ in line 1 was always odd, implying a pause before
ἀκουσθεῖ, while the elision of the final alpha of μεσάνυχτα implies that the two
words should be run together. If that comma goes, the other comma of the pair (after
ἔξαφνα) must go too, though here there is a necessary (and natural) hiatus at that
point between the final alpha of ἔξαφνα and the omega of ὥρα, since the metre will
28
I have not seen the second and third printings, but it is unlikely that they differed from the first
and fourth in this respect.
not allow synizesis at this point (an iambic line with stresses on the second, fifth,
eighth and twelfth syllables). Lines 4–6 consist of three complex direct objects of
the imperative which is withheld until the end of line 6. The three objects (each
consisting of a noun phrase and a relative clause) are separated from each other by
commas, and the third was, in all printings but the last, separated from the verb
phrase by another comma. The use of a comma to separate a complex object, or a
complex subject, from its verb when the verb follows, is often met with in Greek
(and was once a common practice in English). Cavafy’s removal of the comma in
the final printing is unlikely to have been in the interests of a more grammatical
punctuation, since the original use of the comma followed a well-established
convention of punctuation. It is more likely to be an indication of the acoustics of
the line as he now heard it: there is to be no significant pause between the last word
of the third complex object (πλάνες) and the first word of the verb phrase (μὴν);
this line, like the first, is to flow seamlessly. The remaining punctuation marks are
now given greater acoustic prominence: the dash at the end of line 3, marking an
apparent disjunction in the syntax (only resolved in line 6), and the three commas
in lines 3, 4 and 5, which separate grammatically equivalent items.
Perhaps the most interesting aspect of Cavafy’s poetry that his editors have
concealed from his readers is his occasional creative mis-spelling, or choice of non-
standard spelling. I will give the two most interesting examples.
The common verb παίρνω, whose basic meaning is ‘take’, occurs, in forms based
on the present stem παιρν-, in seven of Cavafy’s collected poems. In six of these
it is spelt in the normal way, with alpha-iota in the stem.29 But in «Ἐπέστρεφε»
(‘Return’), the imperfective imperative is spelt πέρνε (with epsilon), not once, but
three times, as one can see below in the text of the ninth and final printing of the
poem:
Ἐπέστρεφε συχνὰ καὶ πέρνε με,
ἀγαπημένη αἴσθησις ἐπέστρεφε καὶ πέρνε με —
ὅταν ξυπνᾶ τοῦ σώματος ἡ μνήμη,
κ᾿ ἐπιθυμία παληὰ ξαναπερνᾶ στὸ αἷμα·
ὅταν τὰ χείλη καὶ τὸ δέρμα ἐνθυμοῦνται,
κ᾿ αἰσθάνονται τὰ χέρια σὰν ν᾿ ἀγγίζουν πάλι.
Ἐπέστρεφε συχνὰ καὶ πέρνε με τὴν νύχτα,

ὅταν τὰ χείλη καὶ τὸ δέρμα ἐνθυμοῦνται . . . .
Return often and take me,

beloved sensation, return and take me —
when the body’s memory awakens,
and old desire runs in the blood once more;
29
See Cavafy (2007) 32 («Τελειωμένα», line 11), 40 («Τρῶες», line 4), 48 («Ἡ δυσαρέσκεια
τοῦ Σελευκίδου», line 10), 156 («Ἡ ἀρρώστια τοῦ Κλείτου», line 16), 178 («Ἐν μεγάλῃ Ἑλληνικῇ
ἀποικίᾳ», line 28), and 182 («Κίμων Λεάρχου», line 10).
160 anthony hirst
when lips and skin remember

and hands feel as though they touch again.
Return often and take me in the night,
when lips and skin remember . . . .
This spelling occurs not just in the last printing, but in all nine of Cavafy’s printings
of this poem, from 1914 to 1930. And yet, every edition of Cavafy’s collected
poems, from Sengopoulou’s in 1935 to the latest reprints of Savidis’ 1991 edition
(the only exception being my own more recent edition),30 have substituted the
normal spelling παῖρνε. Have the editors, in their wisdom, treated Cavafy as some
stupid schoolboy who suddenly forgot how to spell one of the commonest verbs in
the language? This unnecessary supposed ‘correction’ is an assault on the integrity
of the poem (and also an insult to the intelligence of Cavafy’s readers). The spelling
of the verb πέρνω with epsilon is not unprecedented.31 It is noted in Dimitrakos’
dictionary as an ‘erroneous demotic spelling’,32 though the Proïa dictionary simply
notes it as an alternative,33 and it was used by other literary writers of Cavafy’s day.
Why, in this poem alone, did Cavafy choose what some considered a ‘erroneous’
form of the verb? The basic answer is: to enrich the poem. And this it does in
three ways: visually, indirectly in terms of sound, and semantically. Πέρνε, with
two epsilons, makes a neat visual partner for the other imperative with which it is
paired, ἐπέστρεφε, with four epsilons.34 This visual affinity of the two imperatives,
using no vowels but epsilon, draws attention to the auditory quality of the lines in
which they occur, where the short e sound dominates, and also to the fact that,
taken with the words preceding and following, πέρνε forms a phrase, καὶ πέρνε
με, which is four-syllable and proparoxytone like ἐπέστρεφε, and with the same
four-fold repetition of the same vowel sound. But the unusual spelling does more
than this. Πέρνε with two epsilons is a hybrid form, midway between the standard
form, παῖρνε, and πέρνα, the corresponding imperative from the verb περνῶ,
meaning ‘pass’, ‘pass by’ or ‘pass through’. This hybridity subtly suggests that the
speaker calls upon the beloved feeling not only to take possession of him, but also to
pass through him (or even ‘penetrate’ him), providing a parallel to the fourth line of
the poem, where we find a compound form of περνῶ: ‘old desire runs in the blood
once more’ [more literally: ‘passes again (ξαναπερνᾶ) in the blood’]. The poem
has certainly been diminished by the editors’ ‘correction’ of the spelling.
30
Cavafy (2007) 72.
31
As I mistakenly implied in Hirst (2002) and in the original conference paper. I am grateful, again,
to Peter Mackridge for correcting me on this point.
32
Dimitrakos (1933–52), vol. 7 (1951): δημ[οτικὴ] ἐσφ[αλμένη] γρ[αφή]. Essentially this
dictionary belongs to the 1930s. The first six volumes were published by 1939; the publication of the
last three volumes was evidently delayed by World War II and the Civil War.
33
Zevgolis (1932 and 1933–4).
34
Ἐπέστρεφε is technically incorrect. Strictly it should be ἐπίστρεφε, but the retention of the
internal augment in imperatives was common practice in certain parts of the Greek-speaking world in
Cavafy’s day; and, happily, no editor has been concerned to ‘correct’ it.
My second example of creative mis-spelling comes from the poem ‘In a town in
Osroene’. In the opening lines, the speaker relates how his friend Remon had been
brought back wounded from a tavern brawl the previous day around midnight,
and how the moon shone on his beautiful body through the window which they
had left wide open. The word for ‘midnight’ (μεσάνυχτα) is rhymed with the
word for ‘wide open’ (ὁλάνοιχτα), and in the first four printings ὁλάνοιχτα is
spelt in the normal way with omicron-iota in the third syllable. However, in the
last two printings it is mis-spelt ὁλάνυχτα, so that it conforms to the spelling of
νύχτα, meaning ‘night’, playfully suggesting the double meaning ‘wide open’
and ‘all night long’ – evoking, through the suggested but non-existent adverb
ὁλάνυχτα, the real adverb ὁλανυχτίς. This adventurous, modernist gesture has
been overruled by all previous editors. Although it might be thought conceivable
that the mis-spelling arose from a printers’ error which Cavafy failed to notice or
correct, there is no clear evidence for such a prosaic conclusion, and in my edition
I have followed the final printings on the grounds that it is ‘at the very least possible
that the mis-spelling represents a deliberate amendment on Cavafy’s part’ and that
readers should be enabled to ‘consider the matter for themselves’.35
I referred earlier to the much worse crimes of some of the periodical and anthology
editors of Cavafy’s own day. I will now offer two examples of their handiwork.
In 1930 the highly respectable literary journal Νέα Ἑστία featured, as the first
item in the first issue of the new year, Cavafy’s poem ‘Young men of Sidon (400
AD)’.36 Νέα Ἑστία was generally hostile to Cavafy at this period, and this is, in
fact, the only one of his poems which they published before his death. The hostility
shows in an introductory note by the editorial committee, which describes this
as ‘perhaps not one of his most “artistic” but certainly one of his most thoughtful
[poems]’, and, after giving an interpretation of the poem, concludes that ‘if Cavafy
had the opposite view in mind, then he has not handled the subject well’. A
dateline below the poem indicates that the editors had taken the text from Cavafy’s
first printing of the poem, dated June 1920, although there had been two further
printings in 1923 and 1927. The Νέα Ἑστία version diverges at many points from
the first (and both subsequent) printings. There are many changes of punctuation,
including the addition of two exclamation marks.37 There is the omission of a break
between verse paragraphs. But what I want to draw attention to is the removal of
no less than twelve optional final nus from the last paragraph of the poem, that
is, from the speech of the young Sidonian who objects so strongly to the epitaph
Aeschylus is supposed to have written for his own tomb, which refers to his part in
the battle of Marathon, but says nothing of his dramatic works. Here is that speech
(lines 14–26) from Cavafy’s final printing of 1927, which in these lines differs from
the 1920 printing only in the correction of two errors which are also corrected in
35
Cavafy (2007) xxxviii. For the text of the poem see ibid. 92.
36
Νέα Ἑστία 4.73 (1 January 1930): 3–4.
37
At the end of the first and last lines of the excerpt given.
162 anthony hirst
Νέα Ἑστία;38 the asterisks indicate the points at which nu has been removed by the
editors of Νέα Ἑστία:
«Ἆ δὲν μ᾿ ἀρέσει τὸ τετράστιχον αὐτό.
Ἐκφράσεις τοιούτου εἴδους μοιάζουν κάπως σὰν* λειποψυχίες.
Δόσε — κηρύττω — στὸ ἔργον* σου ὅλην τὴν* δύναμί σου,
ὅλην τὴν* μέριμνα, καὶ πάλι τὸ ἔργον* σου θυμήσου
μὲς στὴν* δοκιμασίαν*, ἢ ὅταν ἡ ὥρα σου πιὰ γέρνει.
Ἔτσι ἀπὸ σένα περιμένω κι ἀπαιτῶ.
Κι ὄχι ἀπ᾿ τὸν* νοῦ σου ὁλότελα νὰ βγάλεις
τῆς Τραγωδίας τὸν* Λόγο τὸν* λαμπρὸ —
τὶ Ἀγαμέμνονα, τὶ Προμηθέα θαυμαστό,
τὶ Ὀρέστου, τὶ Κασσάνδρας παρουσίες,
τὶ Ἑπτὰ ἐπὶ Θήβας — καὶ γιὰ μνήμη σου νὰ βάλεις
μ ό ν ο ποῦ μὲς στῶν στρατιωτῶν τὲς τάξεις, τὸν* σωρὸ
πολέμησες καὶ σὺ τὸν*Δάτι καὶ τὸν Ἀρταφέρνη».
‘Ah, that quatrain I do not like.

Expressions of that sort suggest small-mindedness.
Give to your work — say I — all of your strength,
all of your care, and call to mind your work again
in time of trial or when at length your day declines.
That is what I expect, what I require of you.
And not that you should put completely from your mind
the glorious Discourse of Tragedy —
that Agamemnon, that marvellous Prometheus,
and those depictions of Orestes and Cassandra,
that Seven against Thebes — and put for your memorial
merely that in the ranks — the mass — of soldiers
you too had fought Datis and Artaphernes.’
This young man of 400 AD is clearly speaking modern demotic Greek, but the
presence of almost every possible optional final nu (not normally a feature of Cavafy’s
style) gives a rather archaic feel to his speech; it gives it an appropriate rhetorical
edge. All this is lost in the Νέα Ἑστία version. The editors had presumably failed
to see the point of Cavafy’s style here, and applied a prescriptive grammarian’s rule
about when optional nu should be dropped and when retained, as though the poem
were some school exercise in need of correction. And the loss is compounded by
the removal of the one item of katharevousa vocabulary in the young man’s speech.
Faced with the adjective τοιοῦτος (‘such’) in the phrase in the second line above,
Ἐκφράσεις τοιούτου εἴδους (‘Expression of such a kind’), the editors simply
substituted the demotic synonym τέτοιος: Ἐκφράσεις τέτοιου εἴδους.
My second example of serious editorial crime is from Konstantinos Skokos’
appallingly badly edited Modern Greek Anthology of 1923, issued as a series of
38
In the fifth line of the excerpt (line 18) ὧρα has been corrected to ὥρα, and in the seventh line
(line 20) τὰ has been corrected to νὰ. The accentuation of Δάτι in the last line appears in all three
printings, but is questionable, and has been amended to Δᾶτι in Cavafy (1963a), (1991) and (2007).
booklets. Below is Cavafy’s fourth printing (1922) of ‘Since nine o’clock—’, with
the corresponding line of Skokos’ version,39 in oblique type and offset to the right,
below each line of Cavafy’s text, with the discrepancies numbered.
ΑΠ᾿ ΤΕΣ ΕΝΝΙΑ —
Ἀπ᾿ τὶς [1] ἐννιὰ [2]
Δώδεκα καὶ μισή. Γρήγορα πέρασεν ἡ ὥρα

Δώδεκα καὶ μισή. Γρήγορα πέρασεν ἡ ὥρα
ἀπ᾿ τὲς ἐννιὰ ποῦ ἄναψα τὴν λάμπα,
ἀπ᾿ τὴς [3] ἐννιὰ ποὺ [4] ἄναψα τὴ [5] λάμπα,
καὶ κάθισα ἐδῶ. Κάθουμουν χωρὶς νὰ διαβάζω,
καὶ κάθησα [6] ἐδῶ. Κάθουμουν χωρὶς τὰ [7] διαβάζω [8]
καὶ χωρὶς νὰ μιλῶ. Μὲ ποιόνα νὰ μιλήσω
καὶ χωρὶς νὰ μιλῶ. Μὲ ποιόνα νὰ μιλήσω
κατάμονος μέσα στὸ σπίτι αὐτό.
κατάμονος μέσα ᾿ςτὸ [9] σπίτι αὐτό, [10]
Τὸ εἴδωλον τοῦ νέου σώματός μου,
Τὸ εἴδωλο [11] τοῦ νέου σώματός μου,
ἀπ᾿ τὲς ἐννιὰ ποῦ ἄναψα τὴν λάμπα,
ἀπ᾿ τὴς [12] ἐννιὰ ποὺ [13] ἄναψα τὴ [14] λάμπα [15]
ἦλθε καὶ μὲ ηὗρε καὶ μὲ θύμησε
ἦλθε καὶ μὲ ηὗρε καὶ μὲ θύμισε [16]
κλειστὲς κάμαρες ἀρωματισμένες,
κλειστὲς κάμαρες ἀρωματισμένες [17]
καὶ περασμένην ἡδονὴ — τὶ τολμηρὴ ἡδονή!
καὶ περασμένη [18] ἡδονή. [19] — Τί [20] τολμηρὴ ἡδονη! [21]
Κ᾿ ἐπίσης μ᾿ ἔφερε στὰ μάτια ἐμπρός,
Κ᾿ ἐπίσης μ᾿ ἔφερε ᾿ςτὰ [22] μάτια ἐμπρός [23]
δρόμους ποῦ τώρα ἔγιναν ἀγνώριστοι,
δρόμους ποὺ [24] τώρα ἔγιναν ἀγνώριστοι,
κέντρα γεμάτα κίνησι ποῦ τέλεψαν,
κέντρα γεμᾶτα [25] κίνηση [26] ποὺ [27] τέλεψαν [28]
καὶ θέατρα καὶ καφενεῖα ποῦ ἦσαν μιὰ φορά.
καὶ θέατρα καὶ καφενεῖα ποὺ [29] ἦσαν μιὰ φορά.
Τὸ εἴδωλον τοῦ νέου σώματός μου

Τὸ εἴδωλον τοῦ μαύρου [30] σώματός μου
ἦλθε καὶ μ᾿ ἔφερε καὶ τὰ λυπητερά·
ἧλθε [31] καὶ μ᾿ ἔφερε καὶ τὰ λυπητερὰ [32]
πένθη τῆς οἰκογένειας, χωρισμοί,
πένθη τῆς οἰκογένειας), [33] χωρισμοί,
αἰσθήματα δικῶν μου, αἰσθήματα
αἰσθήματα δικῶν μου, αἰσθήματα
τῶν πεθαμένων τόσο λίγο ἐκτιμηθέντα.
τῶν πεθαμμένων [34] τόσο λίγο ἐκτιμηθέντα.
39
Skokos (1923) 107–8.
164 anthony hirst
Δώδεκα καὶ μισή. Πῶς πέρασεν ἡ ὥρα.

Δώδεκα καὶ μισή. Πῶς πέρασε [35] ἡ ὥρα! [36]
Δώδεκα καὶ μισή. Πῶς πέρασαν τὰ χρόνια.
Δώδεκα καὶ μισή. Πῶς πέρασαν τὰ χρόνια! [37]
Since nine o’clock —

Half past twelve. Time has passed so quickly
since nine o’clock when I lit the lamp
and sat down here. I have been sitting without reading,
and without talking. But with whom could I talk,
all alone inside this house.
The spectre of my youthful body,
since nine o’clock when I lit the lamp,
has come and found me and reminded me
of shuttered, scented rooms,
and of past pleasure — what audacious pleasure!
And it has also brought before my eyes
streets which have now become unrecognizable,
places of entertainment, so full of life, now gone,
and theatres and cafés which existed once.
The spectre of my youthful body
has come and brought to me distressing things as well:
bereavements in the family, separations,
sentiments of my loved ones, sentiments
of those who’ve died, which were so little appreciated.
Half past twelve. How the time has passed.
Half past twelve. How the years have passed.
The main point to note is the sheer quantity of the discrepancies between
Skokos’ reprinting and Cavafy’s text: thirty-five in the text of a poem of twenty-
two lines, and a further two in the title. There are only three lines which are not
affected. There are some systematic changes, that is, the imposition of Skokos’ own
orthographic preferences (1, 4, 6, 9, 13, 16, 22, 24, 26, 27, 29), and there are some
straightforward errors (3, 7, 12, 25, 31, 34). Seven punctuation marks are omitted,
including the dash in the title (2, 8, 15, 17, 23, 28, 32). The last of these omissions
is clearly not accidental, but a deliberate intervention in the text, which alters the
meaning: λυπητερὰ (with the accent changed from acute to grave, consistent
with the omission of the colon) becomes adjectival, agreeing with πένθη in the
next line, so that ‘brought to me distressing things as well: / bereavements in the
family’ now reads ‘brought to me as well distressing / bereavements in the family’.
There are also some changes of punctuation, including (again!) the addition of two
exclamation marks replacing full stops (36, 37), as well as the introduction of a
full stop (19), the replacement of a full stop by a comma (10), and the arbitrary
insertion of a closing bracket with no corresponding opening bracket (33). There
is the removal (again) of several final nus (5, 11, 14, 18, 35), encouraging in two
cases synizesis which would be inimical to the metre (18, 35). And then there is
the inexplicable, inexcusable and utterly bizarre substitution, in the first line of the
third paragraph, of μαύρου (30) for νέου, so that ‘the image of my youthful body’
becomes ‘the image of my black [or dark] body’.
This example from Skokos is bad enough, but it not as bad as it gets. There are
other examples, from newspapers and periodicals, where an editor’s systematic or
arbitrary changes compounded by printers’ errors produce infidelities in practically
every word.
After his meeting with George Tsokopoulos in Athens in 1901, Cavafy ruefully
reported in the journal of his summer trip to Greece with his brother Alexander,
that ‘At 3 p.m. Tsocopoulo called. Stayed with me till 3.40. Talked mostly about
literature, and the enormous difficulty met by authors to make an edition sell.
Tsocopoulo says that it is considered quite an achievement to have been able
to publish a volume and realise not profit, but no loss from it.’40 This may have
done much to confirm his intention to publish his poems himself and to move
on from pamphlets containing one or two poems to the small bound collection
of fourteen poems, Poems 1904, printed probably at the beginning of 1905. By
the mid 1920s, if not earlier, Cavafy would have had no difficulty in securing
commercial publication of his poetry had he wanted it. His poetry was well known
by then and had been the subject of a lively debate in the Greek periodical press,
particularly in 1923–4.41 It has often been said that he preferred self-publication
because he could keep control of the text, in the sense of making corrections to his
texts whenever he wanted, adding the latest variants to printed texts before they
were sent out. However, significant post-publication changes to Cavafy’s poems
are few in number; and I suspect that the real reason for his persistence in self-
publication was that he did not trust editors or publishers. In his principal printers,
the Alexandrian firm of Kasimati & Iona,42 he had found reliable allies. They were
close by; he could speak to them in person. Their work was elegantly laid out,
the print quality generally of a high standard and largely free from typographical
errors. Those few errors that did occur Cavafy was able to correct by hand before
circulating the poems in question. He had seen too much of the cavalier treatment
of his texts in newspapers, periodicals and anthologies to entrust his precious life’s
work to others – until, that is, death left him no choice.
40
Cavafy (1963b) 267–8. The journal is written in English.
41
See Karaoglou (1985).
42
Τυπογραφεῖον [later: Τυπογραφικὰ Καταστήματα] Κασιμάτη & Ἰωνᾶ.
166 anthony hirst
References
Banks-Smith, N. (2003), ‘What not to swear’, The Guardian, G2, 7 July 2003: 22.
Beaton, R.M. (2003), George Seferis: Waiting for the Angel: A Biography, New Haven, Conn.,
and London.
Cavafy, C.P. (1935): Κ.Π. Καβάφης, Ποιήματα, ed. Ρίκα Σεγκοπούλου, artistic ed. Τάκης
Καλμοῦχος, Alexandria (post-1980 undated reprints from more than one publisher).
Cavafy, C.P. (1963a): Κ.Π. Καβάφης, Ποιήματα, 2 vols.: Αʹ (1896–1918), Βʹ (1919–1933),
ed. Γ.Π. Σαββίδης, Athens.
Cavafy, C.P. (1963b): Κ.Π. Καβάφης, Πεζά, ed. Γ. Παπουτσάκης, Athens.
Cavafy, C.P. (1968): Κ.Π. Καβάφης, Αὐτόγραφα ποιήματα (1896–1910): τὸ Τετράδιο
Σεγκοπούλου σὲ πανομοιότυπη ἔκδοση, ed. Γ. Π. Σαββίδης, Athens.
Cavafy, C.P. (1991): Κ.Π. Καβάφης, Τὰ ποιήματα: νέα ἔκδοση, 2 vols.: Αʹ (1897–1918), Βʹ
(1919–1933), ed. Γ.Π. Σαββίδης, Athens.
Cavafy, C.P. (2007), The Collected Poems, ed. A. Hirst, tr. E. Sachperoglou, intr. P.
Mackridge, Oxford.
Cavafy, C.P. (forthcoming): Κ.Π. Καβάφης / C.P. Cavafy, ed. S. Ilinskaya, tr. A. Hirst,
Βιβλιοθήκη Κλασσικῶν Νεοελλήνων Συγγραφέων 2, Granada.
Coleridge, S.T. (1990), Table Talk, 2 vols., Collected Works 14, Bollingen Series 75,
Princeton, NJ.
Dimitrakos, D.V. (1933–52): Δ.Β. Δημητράκος (ed.), Μέγα λεξικὸν τῆς ἑλληνικῆς
γλώσσης, 9 vols., Athens.
Ekdawi, S. and Hirst, A. (1999), ‘Left out, crossed out and pasted over: the editorial
implications of Cavafy’s own evaluations of his uncollected and unpublished poems’,
Modern Greek Studies (Australia and New Zealand), 5 (1997–99): 79–132.
Hardy, T. (1982), The Complete Poetical Works of Thomas Hardy, vol. 1, ed. S. Hynes,
Oxford.
Hirst, A. (1995), ‘Philosophical, historical and sensual: an examination of Cavafy’s thematic
collections’, Byzantine and Modern Greek Studies, 19: 33–93.
Hirst, A. (2000), ‘C.P. Cavafy: Byzantine historian?’, Κάμπος: Cambridge Papers in Modern
Greek, 8: 45–74.
Hirst, A. (2002), ‘Cavafy’s Cavafy versus Savidis’s Cavafy: the need to de-edit the
“acknowledged” poems’, in e-journal greekworks.com, 1 March 2002 (http://www.
greekworks.com/content/index.php/weblog/archives).
Hopkins, G.M. (1986), Gerard Manley Hopkins, ed. C. Phillips, Oxford and New York.
Karaoglou, C.L. (1985): Χ.Λ. Καραόγλου, Ἡ ἀθηναϊκὴ κριτικὴ καὶ ὁ Καβάφης, Athens.
Larkin, P. (1988), Collected Poems, ed. A. Thwaite, London and Boston, Mass.
Savidis, G.P. (1966), Γ.Π. Σαββίδης, Οἱ καβαφικὲς ἐκδόσεις (1891–1932), Athens (repr.
1991).
Skokos, K.P. (1923): Κ.Φ. Σκόκος, Νεοελληνικὴ ἀνθολογία ἀπὸ τὰ ἐκλεκτώτερα
ποιήματα τῆς νεώτερας Ἑλλάδος (μετ᾿ εἰκόνων καὶ βιογραφιῶν), vol. 2, no. 11,
Athens.
Tennyson, A. (1909), Enoch Arden and In Memoriam, ed. H. Tennyson, London.
Tennyson, A. (1973), In Memoriam: An Authoritative Text: Backgrounds and Sources:
Criticism, ed. R.H. Ross, New York and London.
Zevgolis, G. (1933–4): Γ. Ζευγώλης (ed.), Λεξικὸν τῆς ἑλληνικῆς γλώσσης, 4 vols.,
Athens.
9
Modern Greek Dictionaries

and the Ideology of Standardization*
Assimakis Tseronis and Anna Iordanidou
Introduction
As Kahane and Kahane observe: ‘The lexicon with its many facets is a mirror of
its time, a document to be understood in sociolinguistic terms’.1 In this chapter,
we propose a discourse analysis of dictionaries as texts produced by an identifiable
authority or institution, addressed to a certain public, at a given time and with a
specific goal in mind.
We have chosen four of the most recent and authoritative dictionaries of Modern
Greek, each illustrating a different approach to the lexicographical description of
the language: the Greek Dictionary by Tegopoulos–Fytrakis publishers (Ελληνικό
Λεξικό), the Modern Greek Dictionary of the Contemporary Demotic Language,
Written and Spoken by E. Kriaras (Νέο Ελληνικό Λεξικό της Σύγχρονης
Δημοτικής Γλώσσας, Γραπτής και Προφορικής), the Dictionary of the Modern
Greek Language by G. Babiniotis (Λεξικό της Νέας Ελληνικής Γλώσσας),
and the Dictionary of Common Modern Greek by the Aristotelian University of
Thessaloniki’s Triandaphyllidis Institute of Modern Greek Studies (Λεξικό της
Κοινής Νεοελληνικής).2 Despite their differences, all four dictionaries considered
together lay the foundations for a proper lexicographical treatment of the Greek
language.
We do not intend to evaluate the four dictionaries on the basis of technical
aspects of lexicography, even though a good deal has been said about this.3 Modern
Greek lexicography has only begun to develop in the last decade, and there are, as
* We would like to thank the editors of the volume for comments and suggestions that have improved
the quality of our text. All remaining errors are our own.
1 Kahane and Kahane (1992) 20.
2
Hereafter, we refer to these dictionaries as the Tegopoulos–Fytrakis Dictionary, Kriaras Dictionary,
Babiniotis Dictionary and University of Thessaloniki Dictionary.
3
See the papers by Petrounias (1985), Charalambakis (1990), Vavadzani (1997), and Anastassiadi-
Simeonidi (2000). Burke (1989) is a review of the first sample printing of the University of Thessaloniki
Dictionary.
167
168 Assimakis Tseronis and Anna Iordanidou
yet, no full-scale institutes for the publishing of dictionaries. Accordingly, it is too

early to judge the dictionaries published so far on strictly lexicographical criteria.
Despite their claims to the contrary, current lexicographical treatments of Modern
Greek cannot be considered to be committed wholly to the strictly scientific
lexicographical principles that dictionaries of French, English, German, Italian or
Dutch have been following for decades. Our aim is not to review the dictionaries in
question, something that has already been done on various occasions in academic
journals, at conferences and in the press.4 We do not aspire to illustrate the extent to
which each dictionary influences language use or to assess its impact on language
change either. Our claim is that we can demonstrate the profile of each dictionary
and uncover the lexicographer’s aspiration to contribute to the standardization
of Greek by comparing the choices made with regard to the essential steps and
decisions involved in the compilation and circulation of a dictionary. Whether
their attempts will be successful or not is to be decided and evaluated by the public
and by experts in due course.
We look at a dictionary both as a cultural monument and as a commodity. As
a cultural monument, a dictionary is the treasury of the language and enjoys a
certain authority and prestige among the members of a linguistic community. As a
commodity, a dictionary is an artefact produced and distributed within a linguistic
community and circulated in a publishing market that serves a particular purpose.5
A compiler of a dictionary of a language such as Greek, which has a long written
and spoken history, and which has undergone various phases of purist movements
and has a recent past of diglossia,6 is likely to face problems and will have to make
decisions regarding the inclusion, exclusion and overall representation of the
vocabulary of the language. In this decision-making process, the lexicographer and
the publisher must also take into consideration the public which they address, and
the particular purpose their dictionary aims to fulfil – educational, commercial,
general, or other.
We shall first place the four dictionaries in the historical and social contexts
surrounding their publication and circulation, and then look in detail at their
choices as regards the inclusion or exclusion of words and variants, the labels used,
the etymological information given, and the spellings favoured. By comparing the
choices that the lexicographers make and by relating them to the character of each
dictionary, we propose a textual analysis of the respective dictionaries as discourses
contributing to the ideology of standardization.7
4
See e.g. Alissandratos (1995), Kalioris (1998), Maravelias (1999), Goutsos (1999), Kexagioglou
(1999), Kriaras (2000), Iordanidou (2000) and Mackridge (2002).
5
See the discussion in Cameron (1995).
6
See Alexiou (1982), Browning (1983), Christidis (1996), Frangoudaki (2001) and Tseronis
(2002).
7
We follow Milroy and Milroy (1991) and Cameron (1995) in understanding the ideology
of standardization as the illusion of an idea of standard language, which the dictionary compilers as
‘language guardians’ wish to sustain or, in the case of Greek lexicography, seek to construct.
Modern Greek Dictionaries 169
The lexicographers, their dictionaries and their time

Of the four dictionaries discussed in this chapter, two were published under the
name of the linguist or lexicographer in charge of the project (Kriaras, Babiniotis),
one under the name of the publishing house (Tegopoulos–Fytrakis), and the other
under the name of the academic institution and the research institute affiliated to
it that undertook the work of compilation (University of Thessaloniki). In all four
cases, there is a group of people, trained lexicographers, linguists or philologists, who
have worked on the project, and whose names are mentioned and acknowledged
in the foreword.
The Greek Dictionary published by the publishing house of Tegopoulos and
Fytrakis was the first of the four dictionaries to appear, in 1988. It has been
reprinted many times since. In 1997 an extended version of the same dictionary
appeared, entitled Magnum Greek Dictionary (Μείζον Ελληνικό Λεξικό), by the
same publishers. There is hardly anything known about the status and scholarly
authority or theoretical stance of the compilers of this dictionary, except for the
little that can be deduced from the four-page introduction. The introductory
note lacks any reference to a general statement about the Greek language or the
compilers’ perspective on it, except for a section where they justify their choice to
include purist words or variants of the demotic entries:
Αυτό έγινε για δύο λόγους: πρώτον, για να γίνει φανερό στο χρήστη του λεξικού
πόσο αξεδιάλυτα είναι, συνήθως, τα όρια λόγιας και δημοτικής γλώσσας και
πόσο λανθασμένη είναι η αποφυγή λέξεων με φανερό σημασιολογικό πεδίο,
με την αιτιολογία της λόγιας προέλευσης, και δεύτερον, για να βοηθήσει ιδ.
τους μαθητές να αναγνωρίζουν τους λόγιους τύπους της λέξης, όταν τους
συναντούν σε παλιότερα κείμενα.
[We have included purist words and variants,] first in order to make it apparent to the
user of the dictionary how unclear the boundaries between the purist and demotic
language usually are, and how wrong it is to avoid words that belong to a particular
semantic field because of their purist origin. Secondly, [we have done so] in order to
help students in particular to recognise such items when they meet them in reading
older texts.8
Interestingly, they have chosen to call their work a ‘Greek’ Dictionary, in

contradistinction to a Dictionary of French, or English, thus evading the issue of
designating their object of description as Modern Greek, Demotic, Neohellenic, or
Common Modern Greek.
Emmanuel Kriaras, Professor Emeritus of Medieval Greek at the University
of Thessaloniki, has long been known in the Greek linguistic community for his
outspoken support for demotic Greek. In 1995, he published his Modern Greek
Dictionary of the Contemporary Demotic Language (Written and Spoken), on the
compilation of which he had worked with a group of experts in Greek philology.
He formulates the aim of his dictionary as follows:
8
The excerpts from the four dictionaries discussed in this chapter are translated by the authors.
Με τη δημοσίευση του λεξικού τούτου ελπίζω ότι παρέχεται στο ευρύτερο

κοινό βοήθημα που μπορεί και αυτό να συμβάλλει ώστε να αποκατασταθεί
υγιής γραπτός και προφορικός λόγος στον τόπο μας, που για μακρό χρονικό
διάστημα ταλαιπωρήθηκε με την ύπαρξη της γνωστής μας διγλωσσίας,
προβλήματος που ευτυχώς ξεπεράστηκε στις μέρες μας.
I hope that the publication of this dictionary provides the general public with a helpful
tool that can contribute to the restoration of a sound language, written and spoken,
in our land, something that has long been lacking on account of the well-known
diglossia, a problem that we have fortunately overcome.
According to Kriaras, language is to be valued and treated with care, attention

and due respect, even affection, as is also clear from the ‘General and Practical
Guidelines for the Proper Use of the Language’ that are included in the introductory
part of his dictionary.
In 1998, Georgios Babiniotis, Professor of Linguistics at the University of
Athens, published his Dictionary of the Modern Greek Language (with comments
regarding the proper use of words). This lexicographer became known to the
academic and general public through his engagement in the so-called language
question early in the 1980s and later on in the 1990s. Babiniotis is an advocate of a
historical approach to the Greek language that views it as a unified, centuries-old
system, which underwent only minor changes in its various phases, from ancient to
medieval to modern. The following quotation illustrates Babiniotis’ belief that his
dictionary duly serves this peculiarity of the language:
Αυτό που θεωρώ ξεχωριστή προσφορά του Λεξικού στην ελληνική γλώσσα
και στα πνευματικά μας πράγματα γενικότερα, είναι ότι μέσα από αυτό,
έτσι όπως έχει συνταχθεί με έμφαση στο βάθος, την έκταση και την ποικιλία
των σημασιών των λέξεων και των φράσεων, αναδεικνύεται ζωντανός και
ανάγλυφος ο πλούτος της σύγχρονης ελληνικής γλώσσας.
What I consider to be this Dictionary’s exceptional contribution to the Greek language

and to our intellectual affairs in general is the fact that through the Dictionary, thanks
to the way it has been compiled, namely by putting the emphasis on the depth, the
extent and the variety of the meanings of words and expressions, the wealth of present-
day Greek becomes salient and vivid.
Such an approach, however, obscures the social and political, as well as the
linguistic, roots of language change. There is a reluctance here to assess present-
day Greek on its merits, as a fully-fledged code with its own morphological
and phonological rules. Instead, we find a superficial identification of the purist
code with the use of language in written speech and of the vernacular code with
spoken discourse, in abstraction from all other social, political or communicative
considerations that may come into play when one is using one or the other code.9
The conception of language in general, and of the Greek language in particular,
9
See the discussions in Frangoudaki (1992), (1997), Christidis (1995), and the study by Setatos
(1973) of the phenomenology of katharevousa.
as a value that stands above the everyday need of individual language users to
communicate, has led Babiniotis at times to deplore the present state of Greek as
a language misused and abused by the media, by political parties and by young
people; also to warn against the massive invasion of foreign usages into the Greek
vocabulary, mainly from English, and to advocate the study of Ancient Greek and,
in general, of older periods of the language in an attempt to improve the present use
of Greek in everyday communication.10
Ten years after the publication of the Tegopoulos–Fytrakis Dictionary, Babiniotis
succeeded in making a dictionary the topic of public discussion and media publicity,
and in introducing a new conception of a dictionary as user-friendly and calculated
to appeal to a wide public regardless of their linguistic preferences and attitudes.
The dictionary was even distributed in exchange for coupons by an established
daily and Sunday newspaper, Το Βήμα.
The Dictionary of the Common Modern Greek Language was brought out later
in the year 1998 by the University of Thessaloniki and a lexicographical team
working in the Triandaphyllidis Institute of Modern Greek Studies. The Institute
had already assigned the planning for a compilation of a dictionary of Modern
Greek to a committee of linguists who had supported demotic back in 1968. After
the constitutional establishment of demotic Modern Greek in 1976, the Ministry
of Education officially commissioned the compilation of a dictionary, mainly for
school purposes, from the Institute, which took more than twenty years to complete
its assignment. It is noteworthy that the same Institute publishes the Grammar of
Modern Greek (Demotic) written by the linguist M. Triandaphyllidis in 1941.11
In the introduction, the lexicographer-in-chief describes the dictionary as a
general-purpose work addressed, in particular, to students and teachers of Greek,
native and foreign alike. Its stated aim is to describe present-day Greek as spoken
and written by the average educated user of the language in the big urban centres
of the country, without regionalisms, or extreme dialectal or sociolectal usages or
any scientific jargon. The publication of this dictionary was followed by the launch
of an electronic version in an educational portal hosted by the Centre for the Greek
Language.
The dictionaries and their lists of entries

The number of headwords included in a dictionary is not only a way for the publisher
to promote a dictionary in the market as more authoritative and comprehensive,
but also a way for the lexicographer to accommodate a language’s history, in which
sense the dictionary becomes a cultural monument in its own right.12 In the case of
Greek, the lexicographer needs to tackle the thorny issues regarding which lexical
entries of purist origin that abounded in dictionaries of the previous century should
10
See Babiniotis (1978), (1984) and the discussion in Pavlidou (1991).
11
An abridged and revised version of the Short Grammar, which is based on the original Grammar
of Modern Greek (Demotic), has been distributed in schools as the official school grammar since 1976.
12
Barnhart (1980).
still be included, and how, and which foreign loanwords and neologisms should
make it to the list of entries. A glance at the entries in the four dictionaries reveals
a good deal of variation in their respective choices.
Both Tegopoulos–Fytrakis and Babiniotis excel in including obsolete words that
could hardly be considered to belong to the treasury of present-day spoken or written
Modern Greek: αγεληδόν, ανήρ, άρουρα, γηθοσύνη, δαψιλής, εκπώμαστρον,
ελλύχνιον, εναβρύνομαι, έννους, ήμαρ, καλλίπυγος, κάττυμα, κισσοστεφής,
λευχειμονώ, μελανειμονώ, μήτηρ, μητρόθεν, ορώ, οτρηρός, πεφυσιωμένος,
ρινόμακτρον, ρίπτω, σεισοπυγίς, σίζω, σκόλοψ, συνωδά, τάλας, χαμαί, χθων,
χους. The Babiniotis Dictionary even includes the words αμήτωρ, θως, κλιτύς,
ολοσηρικός, πολυΐστωρ and σαρδεληδόν, which do not appear in Tegopoulos–
Fytrakis. The dictionaries by Kriaras and the University of Thessaloniki include
none of these words.
As far as neologisms and foreign loanwords are concerned, the entries
in Babiniotis outnumber those in the other three, including words such as:
γκλάμουρ, δημοσιοσχετίστας, ζαργκόν, ίματζ, ίματζ μέικερ, ιν, κιτσάτος,
λουκ, μουλτιμίντια, μουράτος, παγκοσμιοποίηση. The Tegopoulos–Fytrakis
Dictionary at the time of its publication was quite open to foreign loanwords such
as αντεργκράουντ, γκολκίπερ, ζουμάρω, ιντερβιού, κυριλέ, μόνιτορ, πανκ,
πάνελ, πρες κόνφερανς, σικάτος, τεκνατζού, τεκνό and φαστφουντάδικο,
which are also found in the other two dictionaries but not in the Kriaras Dictionary.
Indeed, Kriaras adopts a stricter and more purist policy when he says:
Αδικαίωτους νεολογισμούς που κυκλοφορούν στα γραπτά και στο στόμα μας
δεν καταχωρίσαμε. Ο νεολογισμός για να γίνει δεκτός σε ένα λεξικό πρέπει να
έχει ήδη συναντήσει κάποια γενικότερη αποδοχή, και αυτό δεν συμβαίνει για
πολλά αυθαίρετα σημερινά νεολογικά κατασκευάσματα.
Unjustified neologisms that circulate in written and spoken discourse have not been
included. For a neologism to be included in a dictionary, it needs to enjoy a certain
degree of recognition already, and this is not the case with a number of arbitrarily
constructed contemporary neologisms.
The issue here is who decides whether a language change or a new word is
sufficiently widespread to be considered or accepted as normal. On the face of
it, such a decision seems to imply a prescriptive and conservative practice quite
unlike the one assumed by Babiniotis, as far as neologisms are concerned. But only
on the face of it: the treatment of those words included in the Babiniotis Dictionary
involves just another form of prescriptivism and purism, with cross-references used
to suggest the ‘Greek’ equivalent.13
13
On prescriptivism and purist practices, see Joseph (1987) and Thomas (1991).
Morpho-phonological variants
The prolonged diglossia in the history of Greek has contributed to the preservation
of certain archaic consonant clusters and endings alongside those that have
a colloquial origin in the Greek verb system.14 The lack of any large-scale
sociolinguistic research that would show the distribution of purist and vernacular
variants across speakers of differing social and educational background or across
registers, coupled with the lack of a large and reliable corpus of spoken and written
Greek,15 have left it up to the lexicographer to decide which variants to include
and in what order. Once the decision is made to include both variants, the labelling
of one or the other or both, and then the presenting of one before the other, are
choices that reflect the lexicographer’s own preference about what should be the
norm, rather than a description of what the norm really is.
The Tegopoulos–Fytrakis Dictionary opts for the lexicographically unjustifiable
and confusing practice of including in parentheses next to the main entry the purist
form of a word and marking it with ‘K’ for katharevousa. The compilers justify their
decision with a seemingly linguistic reason, but also a practical and pedagogical
one, namely that the boundaries between the two forms are often unclear, and that
students should be aided in identifying the purist forms when they come across
them in texts. The dictionary thus includes entries with alternative forms like
βασιλεύς, γονεύς, ερωτιδεύς, ηγεμών, πόλις, πρεσβύωψ, σκαπανεύς, φλεψ,
as the katharevousa variants of the headwords βασιλιάς, γονέας, ερωτιδέας,
ηγεμόνας, πόλη, πρεσβύωπας, σκαπανέας, φλέβα. The Babiniotis Dictionary
also includes these variants under the demotic headword, but rightly chooses not
to distinguish between katharevousa and demotic. Instead, Babiniotis labels these
variant forms on the basis of registers of use, as ‘λόγιο’ (‘learned’) or ‘λογιότερο’
(‘more learned’), without, however, following any specific criterion for the
assignment of these labels, and without grounding such a labelling or the existence
of those forms in a corpus search. The Kriaras and University of Thessaloniki
dictionaries do not include the above forms at all.
As far as the past passive forms of verbs are concerned, the Tegopoulos–Fytrakis
Dictionary records only the demotic endings (with -τηκα), contrary to its practice
of including purist forms of nouns and adjectives. The Kriaras Dictionary does
not systematically include information about the morphology of verbs, but one
can deduce the compiler’s preference for the demotic forms in the examples cited.
The Babiniotis Dictionary gives the purist ending after the demotic one, labelling
it as ‘learned’ or ‘more learned’. Interestingly, it mentions the purist form first,
followed by the demotic one, which is labelled as ‘καθημερινό’ (‘colloquial’),
for the verbs ληστεύομαι, ταλαντεύομαι, παίζομαι, whereas it gives no purist
variants for the past passive of the verbs χρειάζομαι, ανακατεύομαι, ονομάζομαι
and γκρεμίζομαι. The University of Thessaloniki Dictionary refers the user to an
appendix that includes conjugation tables for nouns, adjectives and verbs, where
14
See Browning (1982) and Holton et al. (1997).
15
Iordanidou (1996), (1999), (2002).
the demotic forms as prescribed in the Triandaphyllidis Grammar are generally

followed.
Deciding whether to include, or not include, words (or morphological or
phonological variants) in the list of entries is not a matter that is exclusively
dependent on the lexicographer’s own preference and stance. It is also an issue
affected by the factors of the market and of time.16 Nonetheless, the way in which
words are treated once they are included in a dictionary’s list of entries, and the
information given about their use and history, becomes a matter of choice that is
dependent on personal ideologies and is affected by the lexicographer’s stance on
the fundamental issues of standardization.
Labelling
Assigning stylistic labels to the entries of a dictionary is a complicated task,
especially when there can be no general agreement about a systematic theoretical
or empirical basis for the number or distribution of such labels.17 Scrutiny of the
introductory notes of the four dictionaries regarding their labelling practice reveals
a variety of approaches and criteria – more or less systematic or clear – which the
respective lexicographers claim to have used in marking the stylistic and usage
status of their entries.
What strikes one as odd in the Tegopoulos–Fytrakis Dictionary is the lack of
any label distinguishing between uses in written and spoken, formal and informal,
discourse. The few labels mentioned in the list of abbreviations concern text type,
geographical variation, attitude and technical field. There is no labelling at all to
distinguish the context of use for any of the words of purist origin, the foreign loan
words or the purist morpho-phonological variables that the dictionary includes,
such as ιχθυόεις, καταιονητήρ, μελανειφορώ and γιαβρί.18 When faced with
the thorny task of assigning usage labels to words of present-day Greek, which
originate in either purist or demotic or even foreign traditions, the lexicographers
have chosen the easy way out: that is, not to commit themselves to anything, but
simply to record the purist forms of the entries they have chosen to include, marked
with ‘K’ for katharevousa.
The list of labels in the Kriaras Dictionary is no more elaborate either, but here
the lexicographer does make the effort to distinguish between his use of ‘λόγιος’
(‘learned’) and ‘λαϊκός’ (‘popular’) in the foreword to his dictionary:
Για μια λέξη σημειώνεται ότι είναι λόγια όχι αν απλώς προέρχεται από τη λόγια
παράδοση, αλλά αν χρησιμοποιώντας την έχομε το αίσθημα ότι πρόκειται για
λέξη όχι της κοινής γλώσσας, αλλά για λέξη που χρησιμοποιείται για την
ανάγκη της στιγμής. [. . .] Σημειώνονται ως ‘λαϊκές’ μόνο οι κατεξοχήν λαϊκές
λέξεις και όχι κάθε λέξη που σώζεται στη γλώσσα μας από την προφορική
παράδοση.
16
On language change see Aitchison (1991).
17
On labelling see Hausmann (1989) and Corbin (1989).
18
None of these words appears in this purist form in any of the other three dictionaries.
A word is labelled as ‘λόγια’ not simply because it originates in the learned tradition,
but because when using it one feels that it does not belong to the common language,
but is a word used out of the necessity of the moment. [. . .] The words labelled
as ‘λαϊκές’ are the proper ‘λαϊκές’ and not any word that survives in the language
through popular tradition.
The Babiniotis Dictionary makes use of a more elaborate labelling system for
the headwords, which informs the user about style (archaic, learned, colloquial,
familiar, etc.), context of use (dialectical, slang, literary, etc.) and attitude
(ironic, insulting, derogatory, emotional, etc.). It even distinguishes two degrees
of ‘λόγιος’ and ‘λαϊκός’ labels, namely ‘λόγιος’ (‘learned’) and ‘λογιότερος’
(‘more learned’), and ‘λαϊκός’ (‘popular’) and ‘λαϊκότερος’ (‘more popular’).
Of the four dictionaries, this is the only one that uses the label ‘αρχαιοπρεπής’
(‘archaized’). The label is assigned to words that appear in this dictionary alone,
such as αμήτωρ, ημείς, πολυΐστωρ, and to words like γεληδόν, γηθοσύνη, ήμαρ,
ίσταμαι, κάττυμα, κήδομαι, λείχω, μητρόθεν, ορώ, παιδιόθεν, πεφυσιωμένος,
χθων, χους, which also appear in Tegopoulos–Fytrakis, as well as to words like
βροχηδόν, καθαίρω, καθεύδω, κύων, μέλας, όμμα, πας, στέαρ, υπνώττω, which
appear in the University of Thessaloniki Dictionary too.19 The labels ‘λόγιος’ and
‘λογιότερος’ appear to be assigned in a rather arbitrary fashion, however. It seems
that by assigning the label ‘αρχαιοπρεπής’ to words that could hardly qualify
as being part of the Modern Greek vocabulary, the Babiniotis Dictionary seeks to
validate the presence of such obsolete words by attributing to them a distinct level
of use, that of an elaborate and sophisticated discourse.
The University of Thessaloniki Dictionary has made an effort to label every
headword and every reference within a headword by indicating the context in which
a word is most appropriately and effectively used. Given its primary pedagogical
aim, the dictionary presents in detail the different types of labels used, intending
such labels to guide the user in choosing the right word in a particular context.
It distinguishes uses as ‘formal’, ‘scientific’, ‘vernacular’, ‘popular’, ‘learned’,
‘literary’, ‘familiar’, ‘child language’, ‘outdated’, ‘spoken’ and ‘vulgar’. The lack
of a corpus or a large-scale sociolinguistic study, however, deprives these labels of
any empirical grounding that could make them an authoritative resource, or indeed
could provide practical advice, for those learning Greek. A number of words are
invariably labelled as ‘learned’, leaving the user to decide whether they would be
effective in a formal as well as a familiar context, and what their communicative
effect might be.20
As Petrounias remarks,21 it should be the job of a dictionary of Modern Greek
to clarify the confusion over the ‘learned or purist origin’ of a word and its use in
19
None of these words appears in the Kriaras Dictionary. Those that appear in the Tegopoulos–
Fytrakis Dictionary are not labelled, while those that appear in the University of Thessaloniki Dictionary
are labelled as ‘learned’.
20
See, for example, the entries άνευ, γεώμηλο, ενταύθα, κύων, οίκος, πατήρ, ύδωρ, υιός, χείρα.
21
Petrounias (1985) 398.
a ‘learned or formal context’. The origins of the words should be dealt with in the
etymology section of each entry, whereas the usage and stylistic status of the word
should be part of the main entry. This is a practice that the Kriaras Dictionary
tries to follow, whereas the Babiniotis Dictionary confuses the reader about the
frequency and context of certain of the words it includes, when a number of them
(like γηράσκω, γήρας, δρυς, νέαση, όστρεον, πίλος) are not labelled at all, and
others (like άρουρα, κισσοστεφής, οποσάκις, ουδαμώς, πατήρ, πτύω) receive
only an etymological indication, ‘ancient’ or ‘medieval’, in brackets.
Etymological information22
Information about the history of words included in the etymological part of an
entry enhances the sense of dictionaries as cultural monuments, within which clues
to the history of a particular culture and its relationship to others are provided.
Keeping the balance between unduly specialized etymological information and
information about the history of words that may concern the wider public can also
be an efficient way of targeting a larger readership for a dictionary.23 In his detailed
study of the etymologies in Modern Greek dictionaries, Petrounias24 concludes
that the morpho-semantic and etymological information they provide is deficient,
because it is based solely on the form and spelling of the words and invariably seeks
to trace the root of as many words as possible back to Ancient Greek, disregarding
broad internal and external borrowing practices, during the diglossia period in
particular.
In the Babiniotis Dictionary, it is clear from the extent of etymological
information, and the number of usage notes referring to the history of words, that
etymology plays a crucial part not only in the way the entries are organized but also
in the way they are spelled. As the compiler acknowledges in the introduction:
Σε μια γλώσσα δε όπως η Ελληνική, όπου ισχύει η ιστορική ή ετυμολογική
ορθογραφία των λέξεων [. . .], η ετυμολογία αποκτά πρόσθετη βαρύτητα, υπό
τον όρο, βεβαίως, ότι στηρίζεται στις αρχές της γλωσσικής επιστήμης και όχι
σε εμπειρικές ετυμολογήσεις ή παρετυμολογήσεις («λαϊκή ετυμολογία») των
λέξεων.
In a language like Greek, where historical or etymological spelling of words arises

[. . .], etymology has acquired an additional importance, on condition that it is based
on the principles of linguistic science and not on the folk etymology of words.
Here, the lexicographer has chosen to trace the history of words from their earliest
roots, with reference to Indo-European origins, through intermediate steps and
22
In this section, we focus on the discrepancies between the Babiniotis and the University of
Thessaloniki dictionaries alone, as far as the scope of the etymological information and the origin of
the words included in their lists are concerned. The Tegopoulos–Fytrakis Dictionary merely copies the
etymology of words already proposed in Andriotis’ Etymological Dictionary (1983), while the Kriaras
Dictionary refers to the ancient or foreign origin of words only occasionally.
23
Drysdale (1989).
24
Petrounias (1985).
significant cognates. The reason for this, according to the compiler, is practical,
utilitarian and educational, since it is not easy for the ordinary language-user to
have access to the specialized dictionaries needed for this, and it also satisfies the
ordinary user’s curiosity about historical information. Again, such a choice is not
immune to the lexicographer’s own position on the historical roots of the Greek
language and its uninterrupted continuity.
On the other hand, the etymological information in the University of Thessaloniki
Dictionary, compiled throughout by Petrounias, is programmatically restricted to
cover the history of words dating back to what is seen as the historic and linguistic
basis of Common Modern Greek, namely the Hellenistic koine as it developed
from the end of the fourth century BC onwards. The compiler chooses not to give
redundant morphological information about the compounding and derivation of
words of Ancient Greek origin, but rather to focus on a detailed description of the
history of more recent loanwords and neologisms. In addition, special attention has
been paid to grammatical entries such as prefixes and suffixes that illustrate how
the compounding and derivational system of Modern Greek works. Overall, the
University of Thessaloniki Dictionary makes use of a more detailed marking system
for describing the history of words, distinguishing between loan translations,
semantic borrowings and ‘Rückwanderungen’ (‘αντιδάνεια’: ‘reborrowings’).
Unlike other dictionaries of Modern Greek, both the Babiniotis and the University
of Thessaloniki dictionaries agree that words like ακουστικό, κοινωνιολογία,
ουρανοξύστης and πολιτισμός are loan translations of foreign words, originally
coined in English or French, and made up of Greek lexical items. Nonetheless,
whereas both dictionaries acknowledge that internationalisms such as αθλητισμός,
ανθρωπολογία, βιολογία, μικρόβιο, πραγματολογία, have a foreign origin, the
Babiniotis Dictionary describes them as ‘Greek-origin foreign terms’ (‘ελληνο-
γενής ξένος όρος’). Such a designation seeks to underline the fact that the form of
the words is drawn from Greek and to play down the consideration that such words
were originally coined in a foreign language and only later borrowed into Greek
through the learned tradition. The etymological information in the University of
Thessaloniki Dictionary, by contrast, indicates clearly that these words have come
from a foreign language into Greek through the learned tradition.25
There are also a number of other words, labelled as ‘semantic borrowings’
(‘μεταφραστικά δάνεια’)26 in the University of Thessaloniki Dictionary, which
the Babiniotis Dictionary relates directly to Ancient Greek cognates: words like
εκνευρίζω, ένθημα, επίθημα, εφημερίδα, καθήκον, κέντρο. Again, by focusing
entirely on the form of the words, which have an Ancient Greek appearance and
which can sometimes be related to words that actually existed in Ancient Greek,
25
To add to the confusion about the history of internationalisms, the Babiniotis Dictionary states
that words like βιταμίνη, κομουνισμός, σοσιαλισμός, υφολογία are the ‘rendering in Greek of a
foreign word’ (‘απόδοση/μεταφορά στην Ελληνική ξένου όρου’), and rightly so – but contrary to the
labelling for other internationalisms in the same dictionary.
26
On words coined in the purist tradition that make use of Ancient Greek forms, or re-use already
existing words in order to render the meaning of foreign loan words, see Petrounias (1997).
the Babiniotis Dictionary fails to make clear that a new meaning has been added
to the word in modern times in order to address the need for a newly imported
concept or referent.
The inclusion of etymological information has a clear ideological function, if
anything in a more telling way, in the case of ‘Rückwanderungen’ (‘reborrowings’).
Both dictionaries rightly trace the origin of such words as άρια, αφιόνι, γαζία,
διαμάντι, δράμι, κάμαρα, καναπές, παλάβρα, φιντάνι back to Ancient Greek.
The Babiniotis Dictionary, however, accepts as ‘Rückwanderungen’ a number of
other words whose etymological origin is either still disputed or clearly assigned
to a foreign origin in the University of Thessaloniki Dictionary: αμπάρι, βάρκα,
γκράφιτι, γκροτέσκος, ζαμπόν, μπράβος, ουτοπία, πέναλτι, τσόκαρο, φιστίκι.
Such insistence on not only the roots of present-day Greek words in a distant and
glorious past, but also on the impact that the classical Greek tradition has had on
other languages, as a source of inspiration for the invention of new words, originates
in the compiler’s own conviction that Greek has the potential to reinvent itself.27
The etymological treatment of the words included in the Babiniotis Dictionary
and the attention paid to the narrative of their long history stamps this particular
dictionary as – in aspiration, at least – a truly cultural monument and a direct
tribute to the Greek language as a treasure that Greeks should be proud of.
Spelling and orthography

The extent and accuracy of the information about spelling that a dictionary
provides constitutes one of the main reasons, perhaps the primary reason, that
dictionary users have for choosing between one dictionary and another and for
consulting it after making their choice.28 Where Greek dictionaries are concerned,
lexicographers are faced with the thorny issue of deciding how close to be to the
orthographic prescriptions of the state and school grammar by Triandaphyllidis,
and how consistent they need to be when proposing alternative spellings or more
simplified ones. Triandaphyllidis proposed guidelines for the spelling of words,
which sought to establish a simplified historic orthography that is easy to teach and
memorize without the need for recourse to Ancient Greek cognates of the words
in question.29 His orthographic standardization has been much debated, and the
variety of positions is reflected in the dictionaries under review.
Of our four dictionaries, the Tegopoulos–Fytrakis Dictionary is the one that
follows the orthographical prescriptions of the Triandaphyllidis Grammar to the
letter. This does not seem to be the outcome of a conscious theoretical stance
on this much-debated issue, but rather of commercial practicalities, in that the
dictionary is primarily addressed to students and office workers, in whose milieux
the Triandaphyllidis Grammar is the one officially recognized. The Babiniotis
Dictionary makes etymology the primary criterion in prescribing the correct
27
For a discussion see Frangoudaki (1992), (1996), (1997) and Christidis (1995), (1999).
28
Whitcut (1989).
29
Iordanidou (1997).
spelling of all words included. The compiler in effect overgeneralizes the validity of
the historical principle and applies it to the orthography of all words originating in
Ancient Greek or Latin. He thus writes καλοιακούδα, κολλήγος, κουκκί, μάννα,
φύσκα, instead of καλιακούδα, κολίγος, κουκί, μάνα, φίσκα. In the same way,
the dictionary proposes spelling ‘Rückwanderungen’ according to their original
(Ancient Greek) spelling, disregarding the fact that they have been introduced
to Modern Greek at a later stage through the medium of a foreign language, and
should thus follow a simplified orthography too – at least in the case of those
coming through the popular tradition, which actually form the majority. He thus
writes γαρύφαλλο, γόμμα, τόννος, τσαννάκι, τσηρώτο, φυντάνι. By contrast,
the University of Thessaloniki and the other dictionaries register the simplified
spellings for such words: γαρίφαλο, γόμα, τόνος, τσανάκι, τσιρότο, φιντάνι.
Up to a point, the Kriaras Dictionary appears to favour a conservative spelling for
words that are closer to their ancient cognates, neglecting the simplified orthographic
principles of the Triandaphyllidis Grammar: κάππα, λειανός, λειχούδης, λειώνω,
ξενιτειά, πασσαλείφω, ρωδάκινο, στρυμώχνω.30 The problem, however, is that
the same principle is not applied to other words, which makes the spelling practice
of the Kriaras Dictionary appear to be less consistent than the disputed spellings in
the Babiniotis Dictionary. What is most confusing and problematic about both the
Kriaras and the Babiniotis dictionaries is that neither gives any explicit statement
or explanation of the principles for recording a specific spelling: their practice is
to seek to establish an orthographic standard de facto without any prior discussion
or justification. The preference for spellings that relate Modern Greek words
directly to their ancient cognates pro forma goes hand in hand with the emphasis
on the history of the words as discussed above. This practice confirms the status of
both dictionaries as would-be cultural monuments, at times to the detriment of the
practical needs of users and without reference to established language use.
The ideology of standardization

The compilation and circulation of a dictionary is a social practice that contributes
to the processes of standardizing a language, regulating variability in the language
system and ultimately prescribing usage by means of a codification of language.31
Dictionary compilers, then, act as ‘language guardians’, who are burdened with
the task of cataloguing the great variety of forms and meanings encountered in
language use. The decisions they take when recording the various forms, meanings
and expressions reflect both their particular assumptions about what a standard
should look like and their own aspirations for the contribution that their dictionary
is to make in the linguistic community.
30
The Babiniotis Dictionary favours the same spelling, while the other two dictionaries follow the
simplified spelling of these words: κάπα, λιανός, λιχούδης, λιώνω, ξενιτιά, πασαλείφω, ροδάκινο,
στριμώχνω.
31
For a discussion about standardization and codification, see Milroy and Milroy (1991), Downes
(1998), Aitchison (1991) and Cameron (1995).
From the discussion so far, it is obvious how much the four dictionaries under
review differ from each other, as regards both the treatment of entries and the
presentation of information provided. The discrepancies should not be seen as
the result of an accurate description of different object-languages or codes, for
all four dictionaries claim to have described the Greek language as spoken and
written in present-day Greece. In addition, their almost simultaneous circulation
in the second half of the last century32 would have resulted in commercial failure
had their readerships supposed that all four dictionaries were presenting the same
description of the Greek language. Their differences lie not in what they describe
but in how they describe it. And this variability is determined not only by the
individual lexicographer’s reaction to the intellectual climate of his time, as Zgusta
puts it, but also – to a larger or smaller extent – by considerations of marketing and
‘brand image’, as Cameron suggests.33
The study of dictionary compilation and circulation as a mechanism and
symptom of standardization in a linguistic community relates directly to the
discussion of prescriptivism in language and invites a systematic scrutiny of the
phenomenon from a linguistic perspective, as Milroy and Milroy, and Cameron,
have convincingly argued.34 For Cameron, the question is not ‘should we
prescribe?’, but ‘who prescribes for whom, what they prescribe, how and for what
purposes’.35 Depending on the authority and status of the ‘language guardians’
and on the means they use to publicize their ideas, the prescribed spellings, forms,
uses and word narratives can be more or less convincing and can reach a broader
or less broad audience. At the same time, the effectiveness of such prescriptivism
depends on how well the ‘language guardians’ accommodate the needs of the
language users and the current linguistic situation, and on the willingness of the
community to endorse their proposals. In assessing the influence of dictionaries on
the standardization process, Zgusta observes:
There is no doubt that dictionaries do influence the linguistic behaviour of their users.
They cannot stop change, nor can they cause changes disapproved of, or not accepted
by, the user. They can and do stabilise the usage, particularly in the written language
and in the related formal styles of spoken language; they do clarify meanings and
make them more systematic.36
As far as the Greek linguistic community is concerned, ‘Standard Modern

Greek’ is not a code that is homogeneously used in formal and official registers
and from which deviations can readily be identified, as is arguably the case with
32
Note that, although the Tegopoulos–Fytrakis Dictionary was first published in 1988, its enlarged
edition entitled Magnum Greek Dictionary appeared in 1997, only two years after the publication of
the Kriaras Dictionary and just one year before the simultaneous publication of the Babiniotis and the
University of Thessaloniki dictionaries.
33
Zgusta (1992) 7; Cameron (1995) 49.
34
Milroy and Milroy (1991); Cameron (1995).
35
Cameron (1995) 11.
36
Zgusta (1989) 77. See also Gallardo (1980).
English, German, French or Arabic. The compilers of the four dictionaries under
review were faced with the task of choosing which words and forms to register
from those inherited from the purist tradition of the past and those borrowed from
foreign languages, and how to label them and record their history.37 They also had
to account for the fact that the public to which each dictionary was addressed
included people brought up and educated during the diglossic situation before 1976,
and people who had been taught Greek at school through the Triandaphyllidis
Grammar after that. In addition, the lack of a large and scientifically compiled
corpus of written and spoken Modern Greek left the lexicographers to decide how
best to reconcile the cultural function of a dictionary with its commercial aspect,
in the light of their own attitudes towards language and language change. In their
case, there was no standard language to merely describe and elaborate by means
of publishing their dictionary. Instead, the aim was to propose a standard language
and contribute towards the making of one.
According to Zgusta, linguistic communities that go through the slow and
somewhat generational process of overcoming diglossia, like the Greek linguistic
community at the beginning of the twentieth century, tend to show a pattern whereby
‘archaizing’ dictionaries are replaced by ‘standard-descriptive’ dictionaries. Among
the difficulties with the compilation of a ‘standard-descriptive’ dictionary, Zgusta
mentions the problem of distinguishing which words and usages fall under the
standard variety and which under the non-standard, and making a selection from
the vast vocabulary and the possible variants already in use, as well as the dilemma
about whether to include or not include lexical units from earlier phases of the
history of the language that are still in use.38 In terms of Zgusta’s typology,39 all four
dictionaries under review belong to the ‘standard-descriptive’ type as opposed to
those published during the period of diglossia, which were more of the ‘antiquating
or archaizing’ type.40 In each of the four, however, certain repercussions of the
historic type of dictionary can be seen.
Depending on their respective authority and aspirations, each dictionary
pursues a goal of controlling and regulating the linguistic behaviour of the Greek
community either from a historical perspective or else from a perspective of
standardization proper. In both cases, the acknowledged aim is to improve the
standards of linguistic behaviour in speech or writing. The Kriaras and Babiniotis
dictionaries seek to do that by applauding the vast richness and depth of the
Greek vocabulary and by being explicitly prescriptive. Both lexicographers value
language as a treasure that lies beyond the everyday use and communicative needs
of Greek speakers and should thus be treated with care. They both emphasize the
37
See the study by Kahane and Kahane (1967) of the problems that Greek lexicographers face as
a consequence of the diglossic past of the Greek language.
38
Zgusta (1989) 74–5.
39
Zgusta (1989) distinguishes the following four types of dictionaries that influence standardization:
‘standard-creating’, ‘modernizing’, ‘antiquating or archaizing’, and ‘standard-descriptive’.
40
For an overview of Greek dictionaries, see Alissandratos (1980). For a brief discussion of the
general monolingual dictionaries published in the period of diglossia, see also Tseronis (2002).
roots of the language, Kriaras from a more demotic standpoint, Babiniotis from a
more puristic standpoint. Of the two, the Babiniotis Dictionary is the one which
also addresses commercial needs, by providing a work of reference that is easy
to use and includes a significant number of foreign words and neologisms. The
Kriaras Dictionary remains devoted to the principles of demotic Modern Greek
that are advocated by its compiler.
The Tegopoulos–Fytrakis Dictionary is the most commercial of the four. It is
the first one to address the market need for a user-friendly and reliable dictionary,
exploiting the prestige associated with a language dictionary as a cultural
monument. The compilers and publishers of this dictionary have avoided taking
any theoretical stance on the issues of language change and standardization. Their
contribution is simply to record as many forms and variants as possible (whether of
purist, popular or foreign origin) and let time and the users decide on their use and
viability. The University of Thessaloniki Dictionary, by contrast, bears the marks
of the institutional and educational role it has been commissioned to play since
1976. In terms of prestige and authority within the Greek linguistic community,
the Babiniotis and University of Thessaloniki dictionaries are competing points of
reference: the one representing the private, commercial perspective of an individual
but established linguist; the other, the product of a collective, institutional project
that enjoys the status of a national dictionary. In these two dictionaries, Standard
Modern Greek is invested with the prestige of a prominent ‘language guardian’ and
with the authority of an institution, respectively.
Given the lack of an extended and reliable corpus of present-day spoken and
written Greek, there can be no objectively defined linguistic reality that the
dictionaries under review could claim to describe. What each of these dictionaries
has done, in a more or less convincing manner, is construct one possible version of
what the linguistic reality of Modern Greek is or should be, given the lexicographer’s
own stance on the matter and under the constraints of the market or the attitudes
of prospective users. This is not to suggest that there is some deplorable anarchy
within the Greek linguistic community, as some commentators would have it.41
The point is rather that the compilation of a dictionary is not a private individual’s
own business – even if it does indeed take an individual mind and commitment to
get a team of people working together. The compilation of a dictionary is a public
project that addresses a certain linguistic community and should thus be open and
sensitive to public debate about what is not to be included, as well as what is, and
how.
41
On language mythologies surviving in the Greek linguistic community, and the media coverage
of language-related issues in Greece after 1976, see Charis (2001) and Moschonas (2004).
References
Aitchison, J. (1991), Language Change: Process or Decay?, 2nd edn, Cambridge.
Alexiou, M. (1982), ‘Diglossia in Greece’, in W. Haas (ed.), Standard Languages: Spoken
and Written, Manchester: 156–92.
Alissandratos, G. (1980), ‘Τα νεοελληνικά λεξικά. Συνοπτικό διάγραμμα’, Διαβάζω,
32: 26–36 (Part 1) and Διαβάζω 34: 30–44 (Part 2).
Alissandratos, G. (1995), ‘Το Λεξικό της Σύγχρονης Δημοτικής Γλώσσας του Εμμ.
Κριαρά’, Αντί, 586: 54–7.
Anastassiadi-Simeonidi, A. (2000), ‘Το Λεξικό της Κοινής Νεοελληνικής: σχεδιασμός
– συμβολή – χρήση’, Teaching Modern Greek: Proceedings of the Fifth Annual Meeting,
Thessaloniki: 47–56.
Andriotis, N.P. (1983), Ετυμολογικό Λεξικό της Κοινής Νεοελληνικής, 3rd edn,
Thessaloniki.
Babiniotis, G. (1978), ‘Πέραν της καθαρευούσης και της δημοτικής’, in S. Panou (ed.),
Για τη Δημοτική Γλώσσα, Athens: 149–58.
Babiniotis, G. (1984), ‘Νεοελληνική γλώσσα: Μέριμνα, αμεριμνησία και
υπερπροστασία’, in Ελληνικός Γλωσσικός Όμιλος, Ελληνική γλώσσα, vol.1,
Athens: 138–61.
Babiniotis, G. (1998), Λεξικό της Νέας Ελληνικής Γλώσσας, Athens.
Barnhart, C.L. (1980), ‘What makes a dictionary authoritative’, in Zgusta (1980): 33–42.
Browning, R. (1982), ‘Greek diglossia yesterday and today’, International Journal of the
Sociology of Language, 35: 49–68.
Browning, R. (1983), Medieval and Modern Greek, 2nd edn, Cambridge.
Burke, J.B. (1989), ‘Lost for words: vocabulary and usage in Modern Greek and the
Dictionary of the Triandaphyllidis Institute’, International Journal of Lexicography, 2:
157–65.
Charalambakis, C. (1990), ‘Κριτήρια επιλογής σε ένα χρηστικό λεξικό. Παρατηρήσεις
στο υπό έκδοση Λεξικό της Νέας Ελληνικής Γλώσσας του Ιδρύματος Μ.
Τριανταφυλλίδη’, Studies in Greek Linguistics: Proceedings of the Tenth Annual Meeting,
Charis, G.I. (ed.) (2001), Δέκα Μύθοι για την Ελληνική Γλώσσα, Athens.
Christidis, A. (1995), ‘Γλωσσικές μυθολογίες: η περίπτωση της Ελληνικής’, Σύγχρονα
Θέματα, 54: 21–6.
Christidis, A. (1996), ‘The Modern Greek language and its history’, The Greek Language,
Athens: 71–4.
Christidis, A. (1999), Γλώσσα, Πολιτική, Πολιτισμός, Athens.
Corbin, P. (1989), ‘Les marques stylistiques / diastratiques dans le dictionnaire monolingue’,
in Hausmann et al. (1989–90): 673–80.
Downes, W. (1998), Language and Society, 2nd edn, Cambridge.
Drysdale, P.D. (1989), ‘Etymological information in the general monolingual dictionary’, in
Hausmann et al. (1989-90): 525–30.
Frangoudaki, A. (1992), ‘Diglossia and the present language situation in Greece: a
sociological approach to the interpretation of diglossia and some hypotheses on today’s
linguistic reality’, Language in Society, 21: 365–81.
Frangoudaki, A. (1996), ‘On Greek diglossia: the ideological determinants of a long-standing
social conflict over language’, in Frangoudaki, The Greek Language, Athens: 83–9.
Frangoudaki, A. (1997), ‘The metalinguistic prophecy on the decline of the Greek language:
its social function as the expression of a crisis in Greek national identity’, International
Journal of the Sociology of Language, 126: 63–82.
Frangoudaki, A. (2001), Η Γλώσσα και το Έθνος, 1880–1980: Εκατό Χρόνια Αγώνες για
την Αυθεντική Ελληνική Γλώσσα, Athens.
Gallardo, A. (1980), ‘Dictionaries and the standardization process’, in Zgusta (1980): 59–
69.
Goutsos, D. (1999), ‘George D. Babiniotis. Γεώργιος Δ. Μπαμπινιώτης, Λεξικό της νέας
ελληνικής γλώσσας’ (review), Journal of Modern Greek Studies, 17: 163–70.
Hausmann, F.J. (1989), ‘Die Markierung im allgemeinen einsprachigen Wörterbuch: eine
Übersicht’, in Hausmann et al. (1989–90): 649–57.
Hausmann, F.J., Reichmann, O., Wiegand, H.-E. and Zgusta, L. (eds.) (1989–90),
Wörterbücher – Dictionaries – Dictionnaires: An International Encyclopedia of Lexicography,
vol. 1, Berlin.
Holton, D., Mackridge, P. and Philippaki-Warburton, I. (1997), Greek: A Comprehensive
Grammar of the Modern Language, London.
Institute of Modern Greek Studies (Triandaphyllidis Foundation) (1998), Λεξικό της
Κοινής Νεοελληνικής, Thessaloniki.
Iordanidou, A. (1996), ‘“Standard’’ Κοινή Νεοελληνική: απόπειρα καθορισμού’, Strong
and Weak Languages in the EU: Aspects of Linguistic Hegemonism, Thessaloniki: 139–47.
Iordanidou, A. (1997), ‘Η ορθογραφία στα νεοελληνικά λεξικά’, Terminologie et
Traduction, 2: 190–210.
Iordanidou, A. (1999), ‘Ζητήματα τυποποίησης της σύγχρονης Νεοελληνικής’, Strong
and Weak Languages in the EU, vol. 2, Thessaloniki: 835–54.
Iordanidou, A. (2000), ‘Σύγκριση και αξιολόγηση των τεσσάρων λεξικών’, Η
Καθημερινή, 6 November 2000.
Iordanidou, A. (2002), ‘Η κοινή νεοελληνική σε σύγχρονα λεξικά και γραμματικές’,
Terminologie et Traduction, 3: 111–24.
Joseph, J.E. (1987), Eloquence and Power: The Rise of Language Standards and Standard
Languages, London.
Kahane, H. and Kahane, R. (1967), ‘Problems in Greek lexicography’, in F.W. Householder
and S. Saporta (eds.), Problems in Lexicography, Indiana: 249–62.
Kahane, H. and Kahane, R. (1992), ‘The dictionary as ideology: sixteen case studies’, in
Zgusta (1992): 19–76.
Kalioris, G.M. (1998), ‘Το λεξικό Μπαμπινιώτη’, Νέα Εστία, 1706: 1056–82.
Kexagioglou, G. (1999), ‘Ένα λεξικό διεθνών προδιαγραφών. Βιβλιοκρισία του Λεξικού
της Κοινής Νεοελληνικής’, Διαβάζω, 395: 30–2.
Kriaras, E. (1995), Νέο Ελληνικό Λεξικό της Σύγχρονης Δημοτικής Γλώσσας Γραπτής
και Προφορικής, Athens.
Kriaras, E. (2000), ‘Το Λεξικό Μπαμπινιώτη και η Δημοτική’, Η Καθημερινή, 6
November 2000.
Mackridge, P. (2002), ‘G. Babiniotis, Λεξικό της νέας ελληνικής γλώσσας, and Λεξικό
της κοινής νεοελληνικής’ (review), Journal of Greek Linguistics, 2: 254–9.
Maravelias, Ch.E. (1999), ‘Α μεν συνήκα πρόχειρα. Βιβλιοκρισία του Γ. Μπαμπινιώτη,
Λεξικό της νέας ελληνικής γλώσσας’, Διαβάζω, 395: 33–5.
Milroy, J. and Milroy, L. (1991), Authority in Language. Investigating Language Prescription
and Standardization, 2nd edn, London.
Moschonas, S.A. (2004), ‘Relativism in language ideology: on Greece’s latest language

issues’, Journal of Modern Greek Studies, 22: 173–206.
Pavlidou, T. (1991), ‘Linguistic nationalism and European Community: the case of
Greece’, in Coulmas, F. (ed.), A Language Policy for the European Community: Prospects
and Quandaries, Berlin: 279–89.
Petrounias, E. (1985), ‘Τα λεξικά της Νέας Ελληνικής, οι ετυμολογίες τους και
οι ετυμολογίες του λεξικού του Ιδρύματος Τριανταφυλλίδη’, Studies in Greek
Linguistics: Proceedings of the Third Annual Meeting, Thessaloniki: 307–416.
Petrounias, E. (1997), ‘Loan translations and the etymologies of Modern Greek’, Greek
Linguistics ’95: Proceedings of the Second International Conference on Greek Linguistics,
vol. 2, Salzburg: 791–801.
Setatos, M. (1973), ‘Φαινομενολογία της καθαρεύουσας’, Επιστημονική Επετηρίδα
Φιλσοφικής Σχολής Πανεπιστημίου Θεσσαλονίκης, 12: 43–80.
Tegopoulos–Fytrakis (eds.) (1988), Ελληνικό Λεξικό, Athens.
Tegopoulos–Fytrakis (eds.) (1997), Μείζον Ελληνικό Λεξικό, Athens.
Thomas, G. (1991), Linguistic Purism, London.
Triandaphyllidis, M. (1941), Νεοελληνική Γραμματική (της Δημοτικής), Athens.
Tseronis, A. (2002), ‘Diglossic past and present lexicographical practices: the case of two
Greek dictionaries’, Language Problems and Language Planning, 26: 219–52.
Vavadzani, P. (1997), ‘Ο πίνακας των παραγωγικών καταλήξεων στο Ελληνικό
Λεξικό’, Studies in Greek Linguistics: Proceedings of the Seventeenth Annual Meeting,
Whitcut, J. (1989), ‘The dictionary as a commodity’, in Hausmann et al. (1989–90): 88–
94.
Zgusta, L. (ed.) (1980), Theory and Method in Lexicography: Western and Non-Western
Perspectives, Columbia, SC.
Zgusta, L. (1989), ‘The role of dictionaries in the genesis and development of the standard’,
in Hausmann et al. (1989–90): 70–79.
Zgusta, L. (ed.) (1992), History, Languages, and Lexicographers, Tübingen.
10
Greek in Cyprus:
Identity Oscillations and Language Planning
Dimitra Karoulla-Vrikki
Linguist Sue Wright has drawn a distinction between language planning in civic
nations (which she calls state nations) and language planning in ethnic nations
(which she calls nation states). Language planning in civic nations involves the
promotion of the language of political and ecomomic supremacy as the language of
communication.1 It reflects the main features of the civic nation, which according
to the political scientist Anthony Smith is a political community based on territory
and is ‘subject to common laws and institutions’.2
On the other hand, language planning in ethnic nations, according to Wright,
involves the promotion of the ethnoculturally-asssociated language that contributes
to the formation and strengthening of national consciousness. Language planning
here thus reflects the main features of an ethnic nation, which according to Smith, is
a political community based on ancestry that places emphasis on the community’s
‘native culture’ and ‘common descent’, or rather its ‘presumed common descent’.3
These two models of nationhood derive from alternative concepts of nationalism,
civic and ethnic, which reflect different perceptions of nation building. However,
they do not necessarily imply a dichotomy. There is a ‘profound dualism at the
heart of every nationalism’ since ‘every nationalism contains civic and ethnic
elements in varying degrees and different forms’.4 In both civic and ethnic
nationalism, language is a symbol or boundary marker of national identity. What
differs is the emphasis on the function of language. According to Oakes, ‘while in
ethnic models of nation, a language unites those with the same mythical ancestry,
in civic models, the dissociation of language and ethnicity is seen as the best way
to unify an ethnically diverse society’.5
1
Wright (2004) 43–4.
2
Smith (1991) 12.
3
Ibid. 11–12.
4
Ibid. 13.
5
Oakes (2004) 546.
187
188 dimitra karoulla-vrikki
Cyprus
The case of Greek in Cyprus merits investigation because of its association
with either civic nationalist approaches (Cyprocentrism) or ethnic nationalist
approaches (Hellenocentrism) to the identity of Greek Cypriots. In this chapter,
I shall focus on the law courts, the civil service and the education system between
1960 and 1997 in order to investigate the promotion or relegation of Greek through
language-planning strategies aimed at promoting either a Cypriot state identity or
a Greek ethnic identity.6
Cyprus has been an independent, sovereign state since 1960. It has a population
of approximately 700,000, of whom 80% are Greek Cypriots, 18% Turkish
Cypriots, and 2% minority groups such as Armenians, Maronites and Roman
Catholics (in Cyprus, also called Latins).7 The Republic of Cyprus was established
after a four-year armed insurgency carried out by a Greek-Cypriot organization of
fighters (EOKA) against British colonial rule – a struggle that was supported by the
majority of Greek Cypriots. According to the Constitution of 1960, the Republic
was composed of two chief communities, Greek-Cypriot and Turkish-Cypriot. In
the wake of inter-communal problems, however, the Turkish-Cypriot leadership
withdrew from the newly-formed institutions – legislative, executive and judicial –
and in 1964 created their own.
In 1974 a coup staged by the mainland Greek Colonels’ regime was followed
by a Turkish military invasion that de facto divided the island. Greek Cypriots
were forced to move to the south, which covers about 60% of the island’s territory.
Turkish Cypriots moved to the Turkish-occupied north, which the internationally
recognized Republic of Cyprus could no longer control. Interminable rounds of
talks aiming at establishing a bi-zonal, bi-communal federation have so far been
futile. At a historic referendum on a United-Nations-backed solution to the Cyprus
problem in April 2004, the majority of the Greek Cypriots (76%) voted ‘no’ to the
plan proposed by the UN Secretary-General, Kofi Annan. As a result, all recent
efforts to re-unite the island have failed.
The official languages of the Republic of Cyprus are Standard Modern Greek
(SMG) and Standard Turkish, while many official documents such as passports,
currency notes and stamps are issued in Greek, Turkish and English. The linguistic
repertoire of Greek Cypriots comprises Cypriot Greek, SMG and (for the majority)
English. Greek Cypriots use Cypriot Greek for unofficial, oral purposes, including
conversations in the family or with friends and colleagues, (mainly) satirical radio
and TV programmes, political cartoon captions, poetry and folk literature. At
school, Greek Cypriots learn SMG, the native and official language of Greece, and
use it for official oral and written purposes: for administration and as a medium of
instruction in state-run educational institutions, for news broadcasts, newspaper
6
Earlier and shorter versions of the chapter were presented at the King’s College London Logos
Conference in 2004 and at the Cyprus Academic Forum Conference in Lefkosia in 2005.
7
These minority groups, the last two of which are defined by religion rather than ethnic origin,
opted to belong to the Greek community on the basis of the constitutional rights granted to them.
Greek in cyprus 189
editorials, political speeches, university lectures and sermons in church. English

is widely used in administration, tourism and as a medium of instruction in most
private secondary-education schools and tertiary-education colleges.
Cyprocentrism and Hellenocentrism

Since the establishment of the Republic of Cyprus, the political and historical
developments on the island have propelled Greek Cypriots onto a continuum
between Cyprocentrism and Hellenocentrism. The former considers Cypriot state
identity to be the primary identity among Greek Cypriots; the latter considers
Greek ethnic identity as primary. The Cyprocentrists feel primarily Cypriot and
promote Cypriot state identity, while the Hellenocentrists feel primarily Greek
and promote a Greek ethnic identity or consciousness. However, Hellenocentrism
and Cyprocentrism cannot be seen as a dichotomy. The difference between
Cyprocentrism and Hellenocentrism is a matter of emphasis.8 The boundaries
between the two poles of identity should be seen as blurred and ambiguous because
the two ideologies have co-existed and have often overlapped.
The Cyprocentrists have stressed common citizenship in the state of Cyprus
and have viewed the Cypriot political community as subject to common laws and
institutions; this has been the criterion for the definition of their identity as Greek
Cypriots. For this reason, Cyprocentrism is categorized here as the Greek Cypriots’
civic-nationalist understanding of their identity and language and as an ideology
reflecting a concept of Cyprus as a civic nation.9 Conversely, the Hellenocentrists
have stressed the primordial cultural, religious and linguistic commonalities
between Greek Cypriots and mainland Greeks, and have invoked a 3000-year
tradition of Greek population and language to justify their attachment to Greece;
this has been the criterion for the definition of their identity as Greek Cypriots.10
For this reason, Hellenocentrism is categorized here as the Greek Cypriots’ ethnic-
nationalist understanding of their identity and language and as an ideology that
encodes a concept of Cyprus as an ethnic nation.11
The Cyprocentrists invoke the pre-Greek past of Cyprus and what they perceive
as a multi-ethnic and multicultural population on the island throughout Cyprus’
history to justify a multi-ethnic present, suggesting in this way a civic understanding
of contemporary Cyprus. Their views contradict an exclusively Greek descent as
8
See Mavratsas (1998); Peristianis (1995); Karoulla-Vrikki (2005).
9
See Karoulla-Vrikki (2005).
10
Evidence of Greek civilization on the island dates back to the sixteenth century BC, while Cyprus
was later colonized by the Phoenicians (IX–IV BC), and was subjected to successive conquests by the
Assyrians (VIII–VII BC), the Egyptians (VI BC) and the Persians (VI–IV BC). The Cypriots served
as allies in the campaign of Alexander the Great in the Near and Middle East (IV BC), while the island
subsequently became a province of the Ptolemaic Empire (until I BC). Cyprus became a province of the
Roman and Byzantine Empires (I–XII AD) until 1192, when it fell under the Lusignan regime. It then
became a Venetian province (1473–1570). Between 1571 and 1878 the island was under Ottoman rule.
In 1878 Cyprus was ceded to Britain and was declared a British Crown Colony in 1925.
11
promoted by the Hellenocentrists. In essence, however, both positions suggest

an ethno-nationalist approach to the imagining of the past of the Greek Cypriot
community. The truth seems to be that an ideological dichotomy between the
approaches was created after 1960, but neither attachment to the ethnic centre
(Greece) nor attachment to the state centre (Cyprus) has ever become entirely
dominant.
Language planning in the law courts

Language planning in the domain of the law courts reveals two opposing
tendencies.12 Until 1988, language planning granted English a pre-eminence that
reflected Cyprocentric identity orientations associated with an emphasis on the
state of Cyprus and its former British colonial character. After 1988, however,
language planning favoured the replacement of English by Greek (SMG) and
reflected Hellenocentric identity orientations that enhanced a Greek ethnic
identity.
In particular, after 1960 English still enjoyed a dominant position in the law
courts, a fact that reflected the linguistic situation during colonial rule. In the
first place, the bulk of the Cypriot legal system continued to be based on English
Common Law and the Rules of Equity, which had been in use since Cyprus came
under British administration in 1878.13 Cypriot legislators seemed to prefer to
consolidate the English system and to be bound by it, instead of choosing to develop
their own legal system. According to Michalakis Kyprianou, Senior Counsel of
the Republic of Cyprus for twenty-three years, no high-ranking court official
expressed any wish to drop the English system and adopt the Greek one. There was
a subconscious prejudice against the Greek system of law and the administration of
justice in Greece.14 As a result, all laws, rules and regulations available to Cypriots
before 1960, which had been drafted in English, would be maintained until they
were amended by way of variation, addition or repeal, by any law made under the
Constitution (Article 188).
Secondly, while a transitional period of five years (that is, until 1965) was given
for the translation of the English Common Law (and of law reports and textbooks
relevant to law cases) into the official languages (Constitution, Article 189), the
translating proved impossible to accomplish within five years.15
Thirdly, legislation enacted after Independence permitted the preservation of
the use of English and overlooked the linguistic provisions of the Constitution.
According to these provisions, for instance, judicial proceedings had to be conducted
and judgments had to be drawn up in Greek if the parties were Greek, in Turkish
if the parties were Turkish, and in both if the parties were Greek and Turkish
12
Karoulla-Vrikki (2001).
13
The Courts of Justice Law of 1960, N.14/1960.
14
Personal communication, 21 June 2000.
15
Mr Justice Savvides in Typographiki v. Pavlou and O’Sullivan (1987).
Greek in cyprus 191
(Article 3.4). In fact, the Constitution secured the necessary institutional support
and status that could safeguard the ethnolinguistic vitality of each community.16
However, these constitutional provisions were not fully complied with. A new
law enacted in 1965 permitted the continuation of the use of ‘any language used
in the courts hitherto’.17 The word ‘any’ meant Greek or Turkish or English, but
while not excluding Greek or Turkish in theory, it meant in practice that English
could carry on being used. In fact, the law secured the continued use of English
in contrast to the constitutional legislators of 1960, who showed more respect for
the right of Cypriots to use their ethnic, standard, mother tongue (SMG) in their
courts than did the Greek Cypriot legislators five years later.18
In any case, holders of British law degrees were reluctant to replace English by
Greek. The British educational background of the majority of the counsels and
judges remained – unofficially – the best qualification for an advocate’s career
advancement.19 For them, competence in English constituted what Pierre Bourdieu
has defined as ‘linguistic capital’.20 This linguistic capital enabled counsels and judges
in Cyprus to enjoy the privileges deriving from their educational qualifications and
their prestigious social position, which already offered them ‘cultural capital’ and
‘symbolic capital’ respectively.
As a result, judicial proceedings were dominated by the use of English in all
aspects and at all levels (see Appendix 1).21 In particular, English was a specific
requirement for the judges as their only official means of communication. During
judicial proceedings, they would unofficially use Greek (Cypriot Greek or Standard
Greek, that is the variety of Greek used in Greece as the official Standard at the
given time),22 in circumstances such as the following: when they briefly exchanged
views in whispers amongst themselves for purposes of quick communication or
when they addressed a non-English-speaking litigant or witness. Since the judges
themselves did not adopt Greek but were instrumental in the continuing use of
English, it was difficult for those positioned lower in the court hierarchy, such as
the counsels and the litigants, to act as agents of language change by adopting and
diffusing any new arrangement.
Similarly, English was a specific requirement for the counsels. Counsels could
speak in Greek to their clients, but they had to switch to English when they
addressed the judge. If the counsels attempted to address the judge in Greek, even
16
On ethnolinguistic vitality, see Giles, Bourhis and Taylor (1977).
17
N.51/1965.
18
The law derived from the adoption of the doctrine of necessity in 1964, when an emergency was
created by the withdrawal of the Turkish Cypriots from the judicial, executive and legislative branches
of government (see Attorney General v. Mustafa (1964) 212). In the name of public interest, it allowed
the bi-communal requirements of the Constitution to be overlooked when compliance was impossible.
19
Loukaidis (1982) 15–16.
20
Bourdieu (1991).
21
Karoulla-Vrikki (2001).
22
Henceforth, the term Standard Greek will refer to the version of Greek that was standard in
Greece at the given time (katharevousa before the mid-’70s and demotic and SMG thereafter).
if the judges were Greek-speaking too, they would be interrupted by the judge
and requested to switch to English. Meanwhile, the dominance of English created
problems for those counsels who were graduates of Greek universities and did
not have an adequate command of English. They had to have recourse to their
English-speaking colleagues to understand statements made during the judicial
proceedings.23
Those litigants and witnesses who did not have any command of English used
Greek (Standard Greek or Cypriot Greek) when they addressed the judge.24
However, this was the only part of the proceedings that was conducted in
Greek. Interventions and parts of the proceedings conducted in Greek had to be
immediately put into English in order to be smoothly integrated into the English-
speaking judicial procedure. The use of Greek had an unofficial character, as
interpreters translated into English or, in the absence of interpreters, the judges
themselves interpreted, by repeating in English what was said in Greek.
As a result, witnesses and litigants who could not understand English attended
a procedure held in an incomprehensible language. The litigants had to wait until
their counsels informed them about the verdict.25 A striking example was the case
of a Greek Cypriot whose land was scheduled to be sold to the Department of
Antiquities, a plan to which the owner objected. Since the verdict favouring the
Department of Antiquities was in English, the owner did not understand its content
and held a gathering at his house to celebrate the securing of his property.26 The
violation of the linguistic and constitutional rights of the participants was never
taken into consideration. Moreover, such a situation was artificial insofar as English,
a formerly colonial language, was neither the native language of the Cypriots, nor
one of the official languages of the Republic, nor one of the official languages used
in education.
Twenty-eight years after Independence, the enactment of the 1988 Law on
the Official Languages of the Republic stipulated the implementation of the
constitutional provisions and the replacement of the English language in the courts
and the civil service by the two official languages, that is, Greek and Turkish.27
The establishment of the use of Greek in these contexts derived from the
Hellenocentric preferences that now prevailed over earlier Cyprocentric
orientations from the mid-1980s. According to Mavratsas, the ideological objective
23
Minutes of the House of Representatives (1988).
24
Interpersonal communication between litigant and counsel during judicial proceedings moved
on a language continuum, the two extremes of which were Cypriot Greek and the official Standard
Greek used in Greece at the time. The degree of intimacy between counsel and litigant, the degree of
the litigant’s competence in Standard Greek and the degree of the litigant’s emotional disturbance were
among the factors that determined the quantity of dialectal features in their speech: increased intimacy,
low competence in SMG and high emotional disturbance of the litigant resulted in an increased use
of dialectal features. In the same way, competence and emotional situation determined the quantity of
dialectal features used in the litigants’ and witnesses’ address to the judge.
25
26
Ibid. 1699.
27
N.67/1988.
Greek in cyprus 193
was to reaffirm the Greek identity of Cyprus as an independent state by placing

Cyprus ‘within a common Greek world’, in defence, politics and culture.28 In view
of the increasing number of advocates educated in Greece, Hellenocentric voices
favouring the replacement of English by Greek in the courts (and the civil service)
gained ground. For instance, in 1986 a renowned Cypriot advocate, Loukis
Papafilippou, demanded the ‘de-secretization’ of justice in Cyprus, that is, the end
of the use of English in the courts, so that the public would not feel that part of the
trial took place beyond their access. The advocate claimed that justice in Cyprus
was ‘in question’, since most stages of a trial were not conducted in Greek, which
was the language of the people.29
However, advocates of Greek in place of English did not limit their arguments
to the promotion of Greek. By claiming the Greek Cypriots’ right to their ethnic
mother tongue (SMG), they associated this variety of Greek with identity.
Their argument, in fact, was fuelled by the ideal of strengthening Greek ethnic
identity rather than Cypriot state identity. There was an effort to establish pride
and confidence in their language and its associated ethnicity, which reflected an
ethno-nationalist approach to language and identity. This effort contrasted with
the earlier choice of a de-ethnicized language variety, namely English, for use in
the law courts, which reflected a civic nationalist approach.
One particular MP, Efstathios Efstathiou, who submitted the bill to Parliament,
stressed the right of citizens to communicate with ‘the government machinery
in their mother tongue’ as a requirement for the ‘establishment of a feeling’ of
Greek ethnic ‘dignity’ and ‘identity’.30 The keystone of Efstathiou’s argument was
precisely an appeal to the ethnic, rather than the state, pride of Greek Cypriots,
in terms of the need to foster the Greek language for the purpose of fostering their
Greek identity. Efstathiou warned of a possible future loss of Greek ethnic identity,
with the continued use of English by a new ruling class. His bill proposed a
language-planning approach designed to implement a programme of Hellenization
for language use in the courts and government departments.
As a result, the language situation in the courts was reversed. First, Greek
became the language used in judicial proceedings, and the dominant language for
both lawyers and judges. Minutes began to be taken in Greek, too, and the use
of English was only allowed when a reference to, or a quotation from, an original
law, text or decision was required. Then again, the laws in English began to be
translated into Greek, so that the Greek version of law might be granted the status
of original law. Finally, Greek acquired superior status over English in verdicts and
judgments.31 For instance, appeals lodged on the grounds that writ of summons
had to be drafted in English were dismissed on the grounds that Greek was the
appropriate language here. Similarly, in the second half of the 1990s, the use of
28
Mavratsas (1998) 19–20.
29
Papafilippou (1986) 24.
30
Minutes of the House of Representatives (1988) 1693, 1695.
31
See Karoulla-Vrikki (2001), (2002).
English in imprisonment decrees, administrative decisions and laws resulted in an

unprecedented series of dropped charges, since only Greek (or Turkish) could now
be used. In a parallel development, the upgrading of Greek resulted in a landmark
verdict for the career prospects of any counsel who did not speak Greek. Advocate
Mr Stewart MacBright, a British citizen of Cyprus, who was registered with the
Register of Advocates of Cyprus and who had practised as a counsel and used the
English language exclusively for more than thirty years, had to learn Greek in order
to be allowed to continue practising.32
The reversal of the linguistic situation in the law courts also involved a difficulty
in the adoption of Greek law terms, which delayed the linguistic Hellenization of
the courts. As there were no committees set up to provide the equivalent terms in
Greek, it was up to individuals to find the right Greek term among the SMG law
terms. The problem of avoiding the wrong term in Greek was mainly faced by
counsel and judges who had mastered the legal-linguistic code in English better
than they had in Greek. As a result, a form of ‘false’ corpus planning was developed.
It involved the introduction of a false terminology, since the shift from English to
Greek often meant coining words that did not exist in SMG or else using the wrong
SMG term.33
Language planning in the civil service

In the domain of the civil service, according to the Constitution, only the two
official languages, Greek and Turkish, were to be used. However, Greek coexisted
with English, which became an unofficially official third language. After 1988 the
proponents of Hellenization began to turn the scales towards pro-Greek language
measures in all departments and aspects of the civil service. A series of government
decisions and enactments aimed at implementing the 1988 Law on the Official
Languages of the Republic and at replacing English with Greek.
However, the use of English remained. Cyprocentric features could not
miraculously vanish overnight, and there was resistance to the change. For
instance, until 1993 a doctor in a state hospital could still provide a patient with
a document bearing name, address, diagnosis and recommended treatment in
English (see Appendix 2). Richly symptomatic of the long-established dominance
of English, and the conflicting tendencies within the state itself, was the case of
Ms Thekla Kittou, a lawyer and film director. In 1984 she applied for a driving
licence, demanding that it be issued in Greek. Stating that she was determined
to ‘re-Hellenize’ the island,34 Kittou applied to the Supreme Court, so that in
1985, twenty-five years after the establishment of the Republic of Cyprus, the first
driving licence in Greek was issued.35 However, driving licences in Greek were
still not fully institutionalized. They were only being issued on request. Adequate
32
33
34
Personal communication, 25 July 2000.
35
Thekla Kittou v. The Republic of Cyprus (1985).
Greek in cyprus 195
evidence of the government’s determination to continue using English for driving

licences and other official purposes is offered by licences issued in English with a
rubber stamp in Greek eight years later (in 1993: see Appendix 3).
In another court case in 1988, Kittou requested a passport in Greek, arguing
that the constitutional provisions on the official languages were being violated.36
The court accepted the Migration Officer’s argument that her passport could
not be issued in Greek, or in Greek and English, for ‘obvious reasons’, as these
were called. According to the officer, the use of English in passports constituted a
‘standard practice’, followed hitherto, which furthered both Ms Kittou’s interests
and the public interest. He also argued that there was no specific legislative act
defining the type, form or language to be used in the text of the passports. In the
event, the court ruled that as long as passports were printed in three languages,
Greek, Turkish and English, the appellant should have been content even if the
authorities filled the dotted lines in the text in English instead of Greek. Six years
later the Full Bench of the Supreme Court issued a verdict, according to which
English was an international language that could not cause any problems if used
in passports.37 In response, Kittou complained that the court perceived her rights
‘from the point of view of material and utility’, and declared, ‘I still insist that
my passport should bear my name in my language, for I consider it a question
of ethnic dignity.’ She compared the Cypriots of her day to ‘disguised Britons’
who behaved like the ‘indigenous population of an old colony’, and criticized
the state for not granting the citizens what they demanded: their own language.38
The issue revealed what Karyolemou has defined as a ‘striking contrast between
the symbolic value of Ms Kittou’s request and the utilitarian arguments used by
the authorities’.39 The government’s approach to language presupposed civic
nationalism, because it considered English as a de-ethnicized language. However,
people who embraced an ethno-nationalist approach to language, like Ms Kittou,
held fast to the conviction that language was inextricably linked to the expression
and fostering of ethnic identity.
The official use of Greek in the civil service was secured by the 1994 Council
of Ministers’ decision, which resolved the conflict. The decision required that
official documents issued by all government departments and semi-government
organizations should be issued in the official languages of the Republic rather than
in English.40 This applied to Greek, though not to Turkish, since the decisions
could be implemented only in the areas controlled by the Republic and inhabited
by Greek Cypriots. In addition, instead of being in English, citizens’ names and
place-names in passports, driving licences, identity cards and street names were,
according to the decision, expected to be in the official language, that is, in Greek
36
Thekla Kittou v. The Republic of Cyprus (1988).
37
Thekla Kittou v. the Republic of Cyprus (1994).
38
Papadias (1994).
39
Karyolemou (2001) 37.
40
I glossa (1994); Council of Ministers’ Decision (1994).
in the case of the Greek Cypriots.41 As a result, the use of Greek was established on
a wide basis.
However, English still had a presence in official government documents. For
instance, envelopes sent by the Republic of Cyprus to civil servants in 1994 bore
the name and address of the Greek Cypriot addressee in English (see Appendix 4).
Similarly, the Department of Income Tax notified citizens in English (see Appendix
5). In 1995, the Electricity Authority of Cyprus (EAC), a semi-governmental
organization, used the English language for 60 % of the topics in an examination
for electrical engineers; in addition, oral instructions to candidates on the day of the
examination were given in English.42 During the following year, 1996, the Cyprus
Telecommunications Authority continued to publish invitations to tender for
contracts for fax machines in English and required that the tenders be submitted in
English.43 In the same year, the Cyprus Development Bank was still using English
in almost all its documents.44 In 1997 forms used in all Nicosia General Hospital
departments, either for official hospital use only or for patient use, were wholly
printed in English (see Appendix 6). Only through oral communication was the
patient briefed in Greek (SMG or Cypriot Greek) about the diagnosis and the
recommended treatment. Until the beginning of 1998 Medical Certificates of
Cause of Death were issued and handed to the relatives in English.45
Official documents such as forms, invoices, instructions and certificates offer
interesting evidence for the switching between English and SMG – a phenomenon
reflective of ambivalent loyalties about identity within the government. It was
possible for documents printed in Greek to be filled in partly in English and partly
in Greek by employees of different government or semi-governmental institutions.
For instance, on a 1991 patient’s appointment card from Nicosia General Hospital,
the names of the patient and the doctor were completed in English while the date
of the appointment was completed in Greek (see Appendix 7). Conversely, forms
printed in both English and Greek could be partly filled out in English and partly
in Greek. A 1996 cash invoice form from the Electricity Authority of Cyprus was
printed in both Greek and English, but bore the name and address of the customer
in Greek and other details in English (see Appendix 8). Similarly, the Electricity
Authority of Cyprus sent notices to citizens in English, with only the address (street
and town) in Greek (see Appendix 9). The choice of language may have seemed
random, but in 1997 the same authority used the correct combination of languages
when it sent notices of interruption of electricity printed in three languages, the
two official languages (Greek and Turkish), along with English as an international
language (see Appendix 10).
41
Epitelous (1994).
42
Christoforou (1995).
43
Cyprus Telecommunications (1996).
44
Makridis (1996).
45
Pros ton Proedro (1998).
Greek in cyprus 197
Despite the continued use of English, the Greek language was institutionalized
by parallel measures in the wake of the 1994 Council of Ministers’ decision. For
instance, after 1994 the Cyprus Tourist Organization (CTO) and the Ministry of
Commerce, Industry and Tourism repeatedly reminded senior inspectors of a 1991
decision that restaurant menus were approved by the organization only if they were
in Greek, or else in Greek and in another language.46 In 1996 the organization
informed its employees that, in accordance with the Council of Ministers’ decision,
they were now required to use Greek in their correspondence with Greek Cypriot
citizens of the Republic unless other ‘exceptional reasons’ obliged them to act
otherwise.47 Similarly, the Ministry of Commerce required that importers of new
or second-hand vehicles should translate from English into Greek all manuals and
instruction leaflets that were handed to customers.48
By the end of the 1990s, law N.67/1988 and the 1994 Council of Ministers
decision, along with the laws and decisions that they helped to generate, had
largely established Greek in place of English. None of these steps towards
the establishment of the use of Greek would have taken place without the
Hellenocentric tendencies that now prevailed over earlier Cyprocentric norms. In
retrospect, the linguistic situation before the mid-1980s was also interpreted as a
kind of linguistic liberalism, marked by the ‘neutrality and disinterest of Cypriots
over linguistic matters’, and ascribed perhaps to problems ‘of physical and political
survival’.49 The interpretation is valid, in the light of the practical, utilitarian
reasons for the maintenance of English in the civil service after 1960. It is surely
now apparent, however, that the government’s three-decade delay in enforcing
a language-planning strategy that would protect Greek from the hegemonic
dominance of English also implies a conscious language choice associated with
identity preferences.
Language planning in education

In the domain of education, Standard Greek was dominant across the primary and
secondary system, but the promotion or relegation of English and Cypriot Greek
derived, once again, from alternating Hellenocentric and Cyprocentric educational
policies.
After the proclamation of Cyprus as an independent state, the objective of the
Helleno-Christian educational policy was to foster Greek ethnic consciousness
and cultivate the idea of Union, that is the political unification of Cyprus with
Greece. This ethno-nationalist educational policy was reinforced by the fact that
46
See Cyprus Tourist Organization (1991); on memoranda, see Cyprus Tourist Organization
(1995), (2000).
47
Cyprus Tourist Organization (1996). The circular was preceded by a letter sent by the Minister
of Education and Culture, K. Angelidou, to all Ministers a few weeks before. Angelidou urged them
to inform all the departments in their Ministry that they had to implement the 1994 decision of the
Council of Ministers (Y.P. 104/94).
48
Fragkos (1998a), (1998b).
49
Karyolemou (2001) 40.
an independent republic had not been the aim of the Greek Cypriots who had
fought for Union with Greece during their anti-colonial struggle (1955–9).
The policy was headed by the Minister of Education, Dr Costantinos Spyridakis
(1960–70), who had extensive experience in educational affairs.50 Spyridakis was
adamant in his position that ‘Greek educational policy should be the educational
policy of Cyprus’. For him, a separate policy would ‘shake the faith of the Greeks
of Cyprus [the Greek Cypriots] in Greece’, for what linked the Greek Cypriots to
Greece was ‘common blood, common language, common culture’, in addition to
‘the common life’ that would link them after Union.51
Spyridakis opted for full identification of the educational system in Cyprus
with that of Greece. His policy aimed at the establishment of a ‘genuine Greek
educational character’ and at the abolition of the characteristics of the previous
colonial administration which, in his view, had favoured the ‘perversion of the
character’ of education and its ‘anglification and neutralization’.52
Within this ethno-nationalist framework, new language-planning measures were
designed to achieve Hellenization. They aimed at enhancing the status of Greek
and at relegating English and Cypriot Greek, on the grounds that the beauty of
language and the relationship between language and ancestral heritage constituted
salient factors in the formation and maintenance of identity.
Spyridakis praised Greek (both Standard Greek and Ancient Greek) and
promoted it as the quintessential symbol of Greek ethnic identity. He drew the
attention of Greek Cypriot students and teachers to the value, the beauty and
uniqueness of their ethnic language (Standard Greek) as a mother tongue, and to
their responsibility to protect it. According to Spyridakis, the Greeks of Cyprus
had to have adequate Greek ‘language grounding’ in order ‘to become an adequate
and creative unit for their ethnos’, that is for the Greeks in Greece and the rest of the
world.53 The study of Ancient Greek was another factor in the effort towards the
strengthening of Greek ethnic identity. Spyridakis exalted the beauty of Ancient
Greek and its value as the only link to a student’s ancestral Greek heritage. He
opposed the study of Ancient Greek texts in translation rather than in the original.54
Ancient Greek constituted the ancestral heritage that ‘had been nurturing’ the
entire world.55
In a parallel development, Helleno-Christian educational policy aimed at the
elimination of Cypriot Greek at school. The dialect was not felt to contribute to
Hellenization proper and was granted a low status, associated with poor education
and an ‘incorrect’ use of Greek. Students had to be able to understand and use the
50
Before 1960, Spyridakis was Chairman of the Greek Board of Education. Between 1960 and 1965
he served as Chairman of the Greek Communal Chamber. In 1965 the Greek Communal Chamber was
replaced by the Ministry of Education, and Spyridakis was appointed Minister.
51
Minutes of the House of Representatives (1967c) 863.
52
Spyridakis (1959) 204.
53
Ministry of Education (1973) 20.
54
Minutes of the House of Representatives (1967c) 866.
55
Spyridakis (1968b) 6.
Greek in cyprus 199
official standard used in the textbooks without switching to Cypriot Greek, their
mother tongue.
Language planning also aimed at relegating the status and functions of English in
Cyprus. First, Standard Greek replaced English as the language of administration
in the department of education and between educational authorities.56 Secondly,
the primary school teachers’ academy was Hellenized in 1959, and the language
of teaching, which had been English for the previous twenty years, was changed
to Standard Greek.57 Similarly, the name of the academy was changed from the
English ‘Teachers’ Training College’ to the Greek ‘Παιδαγωγική Ακαδημία’
(‘Pedagogical Academy’).58 Moreover, mixed classes, consisting of English, Greek
and Turkish students, were replaced by classes composed entirely of Greek students,
who received their training in Standard Greek rather than English. Furthermore,
Spyridakis refused to grant state recognition to the international private secondary
schools that provided education in languages other than Greek, especially English.
He considered them a hotbed of ‘neutralizing [Greek] consciousness’ and a source
of Cypriot consciousness instead.59 Finally, English was abolished as a subject
in primary schools, because it threatened to make English ‘a tool of thought and
expression equal to the mother tongue’.60
After 1974, educational policy and language planning both underwent a shift.
The involvement of Greece in the coup against the President of Cyprus, Archbishop
Makarios, in 1974, along with Greece’s failure to offer military support to the
Greek Cypriots during the Turkish invasion of Cyprus, left Greek Cypriots feeling
betrayed by their motherland.61 In line with the Greek Cypriots’ attitudes, a new
educational policy placed emphasis on Cypriot state identity. As a result, the earlier
conviction that the Greek Cypriots had to align their educational policy with that
of the metropolitan country for reasons of ethnic survival was now questioned. A
Cypriot-centred educational policy was sought to promote conditions that would
help all Cypriots to survive in a common pluralistic state.
In 1976 a civic-nationalist approach to education was adopted. The decisive
step towards the Cypriotization of education was taken by a young professional
educationist, Dr Chrysostomos Sofianos, who was appointed Minister of Education.
For the first time, the national flag of the Republic was placed next to the existing
Greek flag in the minister’s office to symbolize an emphasis on the entity of the
Republic of Cyprus.
The new educational policy abandoned the terms ‘Helleno-Christian education’
and ‘Helleno-Christian ideals.’ The terms were too reminiscent of the attempts
made by the Greek junta to label as anti-Hellenic and anti-Christian anyone who
dared question their policies and to criticize them for disorienting students from
56
57
58
Ibid. 240.
59
Spyridakis (1968a) 522, (1970) 386.
60
61
Attalides (1979); Papadakis (1993); Peristianis (1995).
their Greek ethnic identity.62 As Greek ethnic identity could not be rejected, a
combination of Greek ethnic and Cypriot state identity was sought, albeit with an
emphasis on the latter. According to Sofianos, the democratization of education
aimed at fostering in the hearts of students ‘the confidence that the state entity
of Cyprus as an independent and sovereign state’ in no way conflicted ‘with the
notion of the ethnos’, and the insistence that there was no confusion between the
two terms.63
Language planning now also aimed at Cypriotization. Sofianos went so far as
to upgrade the status of the Cypriot dialect in education, while Standard Modern
Greek remained the official medium for teaching. First, an anthology of Greek-
Cypriot Literature was published in Cyprus to be used during the ordinary
teaching of Modern Greek literature. The teaching of Cypriot literature in dialect
form as part of the ordinary teaching of Modern Greek literature was intended to
enhance the students’ ‘own’ literary tradition which was put, if not on a par with,
at least closer to, the esteemed literature of modern Greece.
In addition, the ‘parenthetical’ or ‘occasional’ use of the dialect in class by the
teacher, previously prohibited, was now permitted whenever further explanation
was necessary. According to Sofianos, this was expected to relieve Greek Cypriot
students of a feeling of ‘committing a sin’ or ‘embarrassment’ when using their
mother tongue.64
These measures were taken in parallel to the adoption of reforms initiated in
Greece. For instance, in 1978 Ancient Greek texts began to be taught in Modern
Greek translations, permitting emphasis on ancient Greek civilization rather than
on Ancient Greek itself. Similarly, Sofianos promoted demotic in the civil service
by having a booklet published by his ministry, a decision similar to one previously
made in Greece.65
As already noted, after the mid-1980s socio-political conditions favoured
Hellenocentrism. The repercussions were felt on language planning in education.
As in the period 1960–74, language acquired a primary role for the strengthening
of Greek ethnic identity in the interests of an ethno-nationalist ideology.
In the early 1990s, the Minister of Education, Mrs Klairi Angelidou (1993–7),
returned to Helleno-Christian educational policy. The minister expressed a wish
that she had been able to adopt Spyridakis’ policy in full, with its emphasis on
‘Helleno-Christian’ ideals, its sensitivity about the Greek language and its reaction
against the dominance of English. She regretted she could no longer speak of Union
with Greece, following the High Level Agreements of 1977 and 1979, which
provided for a bi-communal, bi-zonal federation.66 Angelidou made it evident that
within the identity of Greek Cypriots, Cypriotness came second to Greekness.
62
Filias (1979).
63
Ibid.
64
Personal communication, 22 October 1997.
65
As stated in Landsman (1989), the booklet published in mainland Greece was I Neoelliniki
(1977). It gave directions on the use of demotic and recommended avoidance of extreme forms.
66
Papastylianou (1994).
Greek in cyprus 201
She accepted the term ‘Greeks on the periphery’ instead of ‘Cypriots’ and, by way
of clarification, declared: ‘What does I am Cypriot mean? Cypriots are like the
Cretans and the other islanders in Greece who claim to be Greeks’.67
The tendency to focus on ethnic identity was also confirmed in her annual,
first of October, messages to students on the commemoration of the establishment
of the Republic of Cyprus between 1993 and 1996, even though on that day it
might be expected that the Cypriot state would be in the spotlight.68 Although
she tried to explain that the two identities, Greek ethnic and Cypriot state, could
co-exist without being in conflict, there was an evident stress on ethnic rather than
on state identity.69 For example, in 1994 she began and ended her message on the
commemoration of the establishment of the Republic of Cyprus by referring to
the importance of ethnic identity and to Cyprus as ‘our small Greek motherland’
which, as a state, would survive only if it safeguarded its Greekness.70 She saw in
the Greek Cypriots a duty to preserve Greek civilization intact rather than a duty,
perhaps, to work for the successful progress of their state.71 The last lines of the
message on 1 October 1993 illustrate the paradoxical priority the minister placed
on the survival of the Greeks in Cyprus rather than on the survival of the Republic
of Cyprus:
We cannot betray our history. This is what the preservation of our Greek identity has
proved so far. On this land we should live as citizens of the Republic of Cyprus. At
the same time, however, we should also preserve our ethnic consciousness because
we were born Greeks, we speak Greek and we are raised in the Greek way. In this
struggle, you are all, teachers and students, summoned to make your mark as pioneers.
Hellenism on this island is destined to survive and it will survive.72
Within the framework of Helleno-Christian educational policy, language

planning emphasized Greek – both Ancient Greek and SMG – in order to foster
Greek ethnic identity. The relationship between language and ethnic identity
in Cyprus was of primary importance, for Angelidou believed that the struggle
for language was equivalent to the struggle for freedom, and that the survival of
Cypriot Hellenism was primarily due to the endurance of the Greek language on
the island for three thousand years.73 The duty of the current generation was to
safeguard the Greek language, which maintained their cultural identity, because
‘no people who managed to impress their experiences onto language vanished from
the limelight of history’.74 First, and foremost, she credited the Greek language
with a special superiority. For the minister, native speakers of Greek should feel
67
Kypriaki (1995); K. Angelidou, personal communication, 22 October 1997.
68
Angelidou (1993b), (1994), (1995c), (1996).
69
Angelidou (1994).
70
Ibid.
71
Ibid.
72
Angelidou (1993b).
73
Angelidou (1995b), (1995a).
74
Angelidou (1997).
proud of their exceptional language, which had been used as a medium for ‘awe-
inspiring texts’ and had lent ‘terms and vocabulary roots to the European languages
of today’.75 Angelidou promoted SMG as a language with exclusive properties.
She encouraged students to pursue a good command of it as this would facilitate
their acquisition of foreign languages.76 Another property of Greek, she claimed,
was a ‘plasticity’ that enabled it to assimilate foreign loanwords in a ‘dynamic
and catalytic manner’ so that one could not feel they were foreign.77 She stressed
the importance of learning ‘correct’ Standard Modern Greek and urged teachers
to praise the qualities of the Greek language, such as its vividness, its beauty, its
structure and above all, its ‘excessive’ richness.78
At the same time, and in accordance with linguistic developments in Greece,
Ancient Greek texts began to be taught in the original and not in Modern Greek
translations, so that students would be aware of Ancient and Modern Greek as
one and the same language, as well as learn about Greek civilization, which was
necessary to preserve their ethnic identity.79
In parallel with these initiatives, the minister launched a campaign against the
dominance of English. She was reluctant to give the curricula of English-medium
educational institutions equivalent recognition to those of the corresponding
Greek-medium institutions.80 She condemned the use of unassimilated foreign
words within Modern Greek as a portent of the loss of Greek ethnic identity. She
urged students ‘to accept or tolerate no deterioration or corruption of the language’,
for once Greek was ‘contaminated’ with foreign words, it would lead to ‘internal
enslavement’ and to the end of ‘a centuries-old cultural history’.81
The conflict over language and identity can also be traced in the long-drawn-
out debate on the bill of law instituting the University of Cyprus. The question
was whether the university was going to be ‘national’ (that is, Greek ethnic),
strengthening Greek identity among Greek Cypriot students and expressing
Hellenocentric orientations, or a ‘state’ university, strengthening the students’
identity as citizens of the Republic of Cyprus and expressing Cyprocentric
orientations instead. The choice of language for the medium of instruction revealed
sharp political divisions, because it was loaded with the expression of the character
of the university. Various options were considered: a Greek-medium, mono-
communal national university; a Greek- and Turkish-medium, bi-communal
state university; a Greek-, Turkish- and English-medium, international state
university.82
75
Angelidou (1997).
76
Angelidou (1996).
77
Angelidou (1997).
78
Ibid.
79
Angelidou (1993a).
80
Papastylianou (1994).
81
Angelidou (1994).
82
It is worth noting that, up to 1997, the existence of state, English-medium, tertiary-education
institutions on the island had never been a matter of debate as regards ethnic or state identity. The
Greek in cyprus 203
The debate began in 1968 with a preference for a Greek university in all its
aspects. Greek was to be the language of instruction, which according to Spyridakis
would ‘invigorate’ the Helleno-Christian ideals of the Greek Cypriots and their
‘instinctive’ loyalty to the Greek ethnos.83 Spyridakis expressed the belief that a
compromise with a language of instruction other than Greek would be harmful.84
He visualized a university staffed by individuals ‘enthusiastic about everything
Greek and not by those who speak of Cyprus for the Cypriots’. He disregarded the
Turkish Cypriots: they were a minority, rather than a community.85
After 1974 there was a preference for a trilingual Cypriot state university,
prompted by the new political context. The felt need now was to emphasize the
Republic of Cyprus, to express Cypriot state identity and to attract the Turkish
Cypriots to work for a common state. In 1976 President Makarios appointed the
Inter-Ministerial Committee on the Establishment of a University (DEMIP) to
investigate the establishment of a Cypriot university in line with the policies of the
Republic. He visualized a university of an independent and international character
that would be open to all Cypriot and foreign students, irrespective of racial, ethnic
or religious identity, and would have Greek and English as the basic languages
of instruction.86 ‘Purely in the light of the life and circumstances’ in Cyprus, and
in response to ‘the political problem and the geographical position of the island’,
DEMIP suggested Greek, Turkish and English as the languages of instruction.87
The shift became apparent in references to Greek and to Cyprus. In the 1976
DEMIP report, the Greek language was not referred to as a factor strengthening
Greek ethnic identity, and Cyprus was no longer called an island inhabited by
Greeks; it was not referred to as an integral part of Greece, but as an island situated
in the geographical area of the Eastern Mediterranean, at the crossroads of three
continents.88 Indeed, the criteria for the choice of language lacked any ethnic
orientation, in marked contrast to the Greek-medium university of ethnic Greek
character that had been proposed before 1974: ethno-nationalism was giving way
to something closer to a civic-nationalist approach to language and identity. Within
this framework, in 1977, Sofianos announced that the university would be inter-
communal and multilingual, with Greek, Turkish and English as the languages of
instruction.89
In 1986, however, the Minister of Education, Dr Andreas Christofidis,
unexpectedly switched to a Hellenocentrist position. He argued in favour of
relevant institutions were: the Higher Technical Institute, the Forestry College, the School for Nurses,
the Institute for Hotel and Food Supply, the Mediterranean Institute of Management, the Police School,
the School for Hoteliers.
83
84
Ellinikos (1968).
85
Epivalletai (1971) 6.
86
Sofianos (1977) 51–2.
87
DEMIP (1976) 38.
88
Ibid. 38, 42.
89
Diakoinotiko (1977).
Greek as the fundamental language for the University.90 For Christofidis, Greek in
education was indispensable for the Greek Cypriots in the struggle to preserve their
Greek identity. The bonds between Cyprus and Greece were significant. A Greek
university with Greek as its language would comply with the common goals set by
the two Ministers of Education, as well as the President and the Prime Minister of
the two countries respectively. Furthermore, Christofidis sought to enhance the
status of Greek as a suitable medium for academic use. He praised it as the source
of the terminology of science, taking biology and medicine as examples, and as the
language used by prominent scientists in other countries. Above all, he stressed the
importance of the relationship between language and ethnic identity.
Invoking the significance of language for Greek ethnic identity, the Minister
used the term ‘a foreign language’, obviously meaning English, to warn against its
negative potential if adopted on an official basis. A foreign language at the university
would undermine the Cypriot state and ‘in the long run’ could ‘orient primary and
secondary school students towards that language’, pushing aside Greek, which was
significant for the preservation of Greek identity.91
The establishment law of the University in 1989 ended an acute debate.92 The
law reflected the provisions of the two-community, two-language, Constitution,
by itself providing for a bi-communal, Greek-medium and Turkish-medium state
University. The resolution was a compromise, which left the Hellenocentrists
largely satisfied. They had secured a Greek-speaking university, whose Greekness
was assured by the inability of Turkish Cypriots to attend, owing to the political
conditions of the time, as well as by the exclusion of English, which was perceived
as a threat to Greek language and identity. Meanwhile, the resolution offered them
the chance to show obedience to the provisions of the Constitution, according to
which Turkish was an official language of the Republic of Cyprus, alongside Greek.
At the same time, the compromise largely satisfied the Cyprocentrists. They could
now envisage a bi-communal Cypriot university for all Cypriots, which would
enhance the Republic of Cyprus, and would not have Greek only as the language
of instruction.
Conclusions
In the law courts, the civil service and the education system, language planning
in Cyprus has been associated with conflicts of identity, between the competing
claims of Hellenocentrism and Cyprocentrism, which have been inextricably linked
to the socio-political developments on the island. These identity preferences either
promoted or relegated Standard Greek (the variety of Greek that was standard
at the given time), while affecting the positions of English, Cypriot Greek and
90
Minutes of the House of Representatives (1986). For the debate on the language(s) of instruction
at the University of Cyprus in the House of Representatives in the 1980s, see Karyolemou (2002) and
Karoulla-Vrikki (2005), (2008).
91
92
N.144/1989.
Greek in cyprus 205
Turkish. The promotion of Standard Greek through language-planning strategies

has owed its impetus to feelings favouring Hellenocentrism and emphasizing Greek
ethnic identity on the ethnic nation state model. Conversely, the continuing use of
English has been tacitly indicative of Cyprocentric orientations and the ideal of a
Cypriot state identity on the civic nation model.
The dominance of Greek was not established at the same time in each domain.
In the law courts the use of Greek prevailed after 1988. In the civil service Greek
was largely institutionalized after 1994, but the co-existence of English and Greek
continued. In education, Greek was always placed in a dominant position. The
conflict between Greek ethnic identity and Cypriot state identity was revealed
in educational policies that affected the status of English, Turkish and Cypriot
Greek.93
The persistence of English may be attributed to factors like the practicalities
of using the language in use before Independence or the prestige of English as an
internationally accepted language used for administrative purposes. By contrast, this
chapter has attempted to show that the main factor in the delayed implementation
of a language policy that would secure the use of Greek in place of English has
been a conscious language choice associated with an emphasis on Cypriot state
identity.
Our investigation of identity and language-planning efforts carries a clear
message: the goal for the Greek Cypriots should be an equilibrium between an
ethnic nation and a civic nation. The recent accession of Cyprus to the EU, in
2004, and a future solution to the Cyprus problem hold out the hope for a new,
overarching Cypriot identity associated with language-planning strategies that
will protect the ethnic Greek loyalties of Greek Cypriots, but will at the same
time safeguard a co-existence of European, Greek and Turkish languages and
identities.
References
Angelidou, K. (1993a), Η Υπουργός Παιδείας μίλησε στο σεμινάριο για τη διδασκαλία
της αρχαίας ελληνικής γλώσσας και του πολιτισμού [‘The Minister of Education
spoke at the seminar about the teaching of Ancient Greek language and civilization’],
Press Release: Public Information Office, 11 May 1993.
Angelidou, K. (1993b), Μήνυμα προς τους μαθητές για την επέτειο της ίδρυσης της
Κυπριακής Δημοκρατίας, 1 Οκτ 1993 [‘Message to students on the occasion of the
commemoration of the establishment of the Republic of Cyprus, 1 October, 1993’]
(available from Ministry of Education), 17 September 1993.
Angelidou, K. (1994), Μήνυμα προς τους μαθητές για την επέτειο της ίδρυσης της
93
Cf. Karoulla-Vrikki (2005), (2007), (2008).
commemoration of the establishment of the Republic of Cyprus, 1 October 1994’]

(available from Ministry of Education), 1 October 1994.
Angelidou, K. (1995a), ‘Στη διαχρονική αντοχή της ελληνικής γλώσσας οφείλεται η
επιβίωση του Κυπριακού ελληνισμού’ [‘The survival of Cypriot Hellenism is due to
the endurance of the Greek language over time’], Ελευθεροτυπία, 7 February 1995: 2.
Angelidou, K. (1995b), ‘O αγώνας για τη γλώσσα είναι και αγώνας ελευθερίας’ [‘The
struggle for language is also a struggle for freedom’], Αγών, 9 May 1995: 9.
Angelidou, K. (1995c), Μήνυμα προς τους μαθητές για την επέτειο της ίδρυσης της
commemoration of the establishment of the Republic of Cyprus, 1 October 1995’]
(available from Ministry of Education).
Angelidou, K. (1996), Γράμμα από την υπουργό [‘Letter from the Minister’] (available
from Ministry of Education).
Angelidou, K. (1997), ‘Προβληματισμοί για τη γλώσσα’ [‘Thoughts about language’],
Φιλελεύθερος, 4 July 1997: 7.
Attalides, A.M. (1979), Cyprus: Nationalism and International Politics, Edinburgh.
Attorney-General of the Republic v. Mustafa Ibrahim and others, Criminal Appeals nos.
2729, 2734, 2735 (1964).
Bourdieu, P. (1991), Language and Symbolic Power, tr. G. Raymond and M. Adamson,
Cambridge.
Charalampidis, Ch. (1994), ‘Επίσημα έγγραφα στ᾿ αγγλικά στέλλονται ακόμα σε
πολίτες’ [‘Official documents in English are still sent to citizens’], Σημερινή, 12 April
1994: 2.
Christoforou, L. (1995), ‘Ελληνική γλώσσα και εξετάσεις στο δημόσιο’ [‘The Greek
language and the examinations to the civil service’: letter to the editor], Σημερινή, 31
May 1995: 9.
Council of Ministers Decision (1994), Χρησιμοποίηση της ελληνικής γλώσσας σε
κυβερνητικά έντυπα ή/και έγγραφα [‘The use of the Greek language in all government
forms and/or documents’], no. 40,540: 3 February 1994.
Cyprus Telecommunications Authority (1996), Specification for Fascimile Machines,
Nicosia.
Cyprus Tourist Organisation (1991), ‘Setting up menus in Recreation Centres’, February
1991.
Cyprus Tourist Organisation (1995), ‘Memoranda: file no. 716,729’, 22 November
1995.
Cyprus Tourist Organisation (1996). Circular 249, file no. 31, 7 October 1996.
Cyprus Tourist Organisation (2000), ‘Memoranda: file no. 716,729’, 21 February
2000.
DEMIP (1976), Report of the inter-ministerial committee on the establishment of a university
in Cyprus (confidential), Nicosia.
[Diakoinotiko] (1997), ‘Διακοινοτικό το πανεπιστήμιο’ [‘The University will be inter-
communal’], Φιλελεύθερος, 5 November 1977: 10.
[Ellinikos] (1968), ‘Ἑλληνικός πρέπει νὰ εἶναι ὁ χαρακτήρ τοῦ τυχόν ἱδρυσομένου
πανεπιστημίου τονίζει εἰς δήλωσιν του ὁ Δρ. Σπυριδάκης’ [‘The character of the
would-be University of Cyprus should be Greek, Dr Spyridakis stresses in a statement’],
Ἐλευθερία, 25 December 1968: 3.
Greek in cyprus 207
[Epitelous] (1994), ‘Επιτέλους η ελληνική γλώσσα σε όλα τα επίσημα έγγραφα’ [‘The

Greek language in all official documents at last’], Σημερινή, 4 February 1994: 1.
[Epivalletai] (1971), ‘Ἐπιβάλλεται ἡ ἵδρυσις πανεπιστημίου εἰς Κύπρον μὲ ἐθνικόν
προσανατολισμόν καὶ καθοδηγητικήν ἀποστολήν’ [‘The establishment of a university
of ethnic orientation and guiding mission is imperative in Cyprus’], Ἐλευθερία, 5 March
1971: 1, 6.
Filias, V. (1979), ‘Φυτώρια δημοκρατίας τα σχολεία στην Κύπρο’ [‘Schools in
Cyprus are breeding grounds of democracy’: interview with Chrysostomos Sofianos],
Ελευθεροτυπία (Athens), 10 October 1979: 12.
Fragkos, G. (1998a,), ‘Υποχρεωτικά και στα ελληνικά τα βιβλιάρια των αυτοκινήτων’
[‘The booklets for cars must be in Greek too’], Φιλελεύθερος, 31 January 1998: 1.
Fragkos, G. (1998b), ‘Αυτοκίνητα: Προχωρεί το θέμα οδηγιών χρήσης στα ελληνικά’
[‘Cars: the issue of instruction manuals in Greek proceeds’], Φιλελεύθερος, 23 February
1998: 4.
Giles, H., Bourhis Y.R. and Taylor D.M. (1977), ‘Towards a theory of language in ethnic
group relations’, in H. Giles (ed.), Language, Ethnicity and Intergroup Relations, London:
307–48.
[I glossa] (1994), ‘Η γλώσσα μας’ [‘Our language’], Σημερινή, 7 February 1994: 9.
[I Neoelliniki] (1977), ‘Η Νεοελληνική στη Διοίκηση’ [‘Modern Greek in administration’],
Ministry of the Presidency of the Government, Greece.
Karyolemou, M. (2001), ‘From liberalism to legal regulation: the Greek language in Cyprus’,
Language Problems and Language Planning, 25: 25–51.
Karyolemou, M. (2002), ‘When language policies change without changing: the University
of Cyprus’, Language Policy, 1: 213–36.
Karoulla-Vrikki, D. (2001), ‘English or Greek language? State or ethnic identity? The case
of the courts in Cyprus’, Language Problems and Language Planning, 25: 259–88.
Karoulla-Vrikki, D. (2002), ‘Η αγγλική ως κυρίαρχη γλώσσα στα δικαστήρια της
Κυπριακής Δημοκρατίας’ [‘English as the dominant language in the courts of the
Republic of Cyprus’], Μελέτες για την Ελληνική Γλώσσα, 22: 289–99.
Karoulla-Vrikki, D. (2005), ‘Language planning in Cyprus: a reflection of an identity
conflict’, unpublished PhD thesis, University of London.
Karoulla-Vrikki, D. (2007), ‘Education, language policy and identity in Cyprus: a diachronic
perspective (1960–1997)’, in A. Papapavlou and P. Pavlou (eds.), Sociolinguistic and
Pedagogical Dimensions of Dialects in Education, Cambridge: 80–100.
Karoulla-Vrikki, D. (2008), ‘Γλωσσική πολιτική για της ίδρυση του Πανεπιστηµίου
Κύπρου: Ζήτηµα γλώσσας και ταυτότητας’ [‘Language policy on the establishment
of the University of Cyprus: an issue of language and identity’], Μελέτες για την
Ελληνική Γλώσσα, 28: 185–93.
[Kypriaki] (1995), ‘“Κυπριακή” ή “ελληνική” λογοτεχνία’ [‘“Cypriot” or “Greek”
literature’], Φιλελεύθερος, 5 November 1995: 17.
Landsman, M.D. (1989), ‘Greeks’ sense of language and the 1976 linguistic reforms:
illusions and disappointments’, Byzantine and Modern Greek Studies, 13: 159–82.
Loukaidis, G.L. (1982), Θέματα κυπριακού δικαίου [‘Cypriot law issues’], vol. 1,
Nicosia.
Makridis, A. (1996), ‘Επιμένουν να παρανομούν με τη χρήση της αγγλικής’ [‘They
insist on infringing the law by the use of English’], Σημερινή, 26 August 1996: 1.
Mavratsas, V.C. (1998), Όψεις του ελληνικού εθνικισμού στην Κύπρο [‘Aspects of Greek
nationalism in Cyprus’], Athens.
Ministry of Education (1973), Ἡ ἐκπαίδευσις ἐν Κύπρῳ 1973 [‘Education in Cyprus
1973’], Nicosia.
Minutes of the House of Representatives (1967), session G, 30 March 1967.
Minutes of the House of Representatives (1986), parliamentary period E, session A, 6
November 1986.
Minutes of the House of Representatives (1988), parliamentary period E, session B, 19 May
1988: 1692–1704.
N.14/1960, Ὁ περὶ Δικαστηρίων Νόμος [‘The Courts of Justice Law’].
N.51/1965, Ὁ περὶ Νόμων καὶ Δικαστηρίων Νόμος (Κείμενο και Διαδικασία) [‘The Laws
and Courts (Text and Procedure) Law’].
N.67/1988, Ο περί των Επισήμων Γλωσσών της Δημοκρατίας Νόμος του 1988 [‘The
1988 Law on the Official Languages of the Republic’].
N.144/1989, O περί Πανεπιστημίου Νόμος [‘Law on the establishment of the
University’].
Oakes, L. (2004), ‘French: a language for everyone in Québec?’, Nations and Nationalism,
10: 539–58.
Papadakis, Y. (1993), ‘Perceptions of history and collective identity: study of contemporary
Greek-Cypriot and Turkish-Cypriot nationalism’, PhD dissertation, University of
Cambridge.
Papadias, Ch. (1994), ‘Επιμένει ελληνικά η Θέκλα . . .’ [‘Thekla insists on Greek . . .’],
Σημερινή, 1 February 1994: 2.
Papafilippou, L. (1986), ‘Γλώσσα και δικαιοσύνη’ [‘Language and justice’], Άμαξα, 13:
21–8.
Papastylianou, Ch. (1994), ‘Ο μυστικός γάμος της Κλαίρης’ [‘The secret wedding of
Klairi’: interview with Klairi Angelidou], Σημερινή, 31 January 1994: 5.
[Perifronisi] (1993), ‘Περιφρόνηση της γλώσσας μας’ [‘Contempt for our language’],
Σημερινή, 20 September 1993: 2.
Peristianis, N. (1995), ‘Δεξιά – αριστερά, Ελληνοκεντρισμός – Κυπροκεντρισμός: Το
εκρεμμές των συλλογικών ταυτίσεων μετά το 1974’ [‘Right-Left, Hellenocentrism-
Cyprocentrism: the pendulum of collective identifications after 1974’], in N. Peristianis
and G. Tsangaras (eds.), Ανατομία μιας μεταμόρφωσης. Η Κύπρος μετά το 1974 –
κοινωνία, οικονομία, πολιτική, πολιτισμός, Nicosia: 123–56.
[Pros ton Proedro] (1998), ‘Προς τον Πρόεδρο της Δημοκρατίας’ [‘To the President of the
Republic’], Σημερινή, 3 March 1998: 9.
[Sfines] (1994). ‘Σφήνες’ [‘Digs’], Σημερινή, 15 May 1994: 3.
Smith, D.A. (1991), National Identity, Reno, Nev.
Sofianos, A.Ch. (1977), ‘Basic issues of educational policy: address in the House of
Representatives’, Minutes of the House of Representatives, parliamentary period C,
session B, 8 December 1977, Nicosia: 311–46.
Spyridakis, C. (1959), ‘Ἐκπαιδευτικαί μεταρρυθμίσεις ἐν Κύπρῳ’ [‘Educational reforms
in Cyprus’], Κυπριακή Ἐκπαίδευσις 1: 5–17.
Spyridakis, C. (1960), ‘Ἔκϑεσις περὶ τοῦ ἔργου τοῦ Γραφείου Ἑλληνικῆς Παιδείας
κατὰ τὸ σχολικὸν ἔτος 1959–1960’ [‘Report on the work of the Greek Educational
Board during the academic year 1959–1960’], in Spyridakis (1974): 222–43.
Greek in cyprus 209
Spyridakis, C. (1968a), ‘Ἡ ἑλληνική ἐκπαίδευσις ἐν Κύπρῳ’ [‘Greek education in

Cyprus’], in Spyridakis (1974): 517–35.
Spyridakis, C. (1968b), ‘Ὁλίγα λόγια διὰ τὴν διδασκαλίαν τῶν ἀρχαίων ἑλληνικῶν
εἰς τὰ γυμνάσια’ [‘A few words on the teaching of Ancient Greek in High Schools’],
Κυπριακή Ἐκπαίδευσις, 8: 3–6.
Spyridakis, C. (1970), ‘Ἡ ἐν Κύπρῳ ἐκπαιδευτικὴ πολιτική’ [‘The educational policy in
Cyprus’], in Spyridakis (1974): 377–89.
Spyridakis, C. (1973), ‘Ἡ ἀναγκαιότης ἐν Κύπρῳ ἀνωτάτης ἐκπαιδεύσεως’ [‘The need
for tertiary education in Cyprus’], in Spyridakis (1974): 567–71.
Spyridakis, C. (1974), Μελέται, διαλέξεις, λόγοι, ἄρθρα [‘Studies, lectures, speeches,
articles’], vol. 2, part B, Nicosia.
Thekla Kittou v. The Republic of Cyprus through The Minister of Interior (1988), case no.
742/88.
Thekla Kittou v. The Republic of Cyprus through The Minister of Interior (1994),
Administrative Appeal no. 1056.
Thekla Kittou v. The Republic of Cyprus through The Registrar of Motor Vehicles (1985),
case no. 651/84.
Typographiki Ekdotiki Etereia [sic] ‘Proodos’ Ltd. v. Pavlou, P. and O’ Sullivan, M. (1987),
Civil Appeal no. 7139, 1 C.L.R, 529.
Wright, S. (2004), Language Policy and Language Planning, New York.
Appendix 1
Use of languages during judicial proceedings (1960–88)
Criminal Court
Judges
English English
English
Shorthand-Typist
(Minutes English) English
(or Greek)
Witness / Litigant
Interpreter English English
English
(or Greek) English
Greek (or Greek)
Accused
Counsel Prosecutor
for the accused (Counsel / Policeman)
Public
Greek in cyprus 211
Appendix 2 Appendix 2
Hospital document, 1993
[cf. [Perifronisi] (1993)]

[Perifronisi (1993)]
Appendix 3
Driving licence, 1993
Appendix 4
Greek in cyprus 213
Appendix 4
Official envelope, 1994
[Sfines (1994)]
(Sfines1994)
Appendix 5
Appendix 5
Income tax notification, 1994
[Charalampidis (1994)]
[see Charalampidis (1994)]

Greek in cyprus 215
Appendix 6
Appendix 6
Hospital form, 1997
Appendix 7
Appendix 7
Hospital appointment card, 1991
Greek in cyprus 217
Appendix 8
Appendix 8
Cash invoice for electricity charges, 1996
Appendix 9
Appendix 9
Non-payment notice for electricity, 1996
Greek in cyprus
Appendix 10 219
Appendix 10
Electricity supply interruption notice, 1997
11
‘Greeklish’:
Transliteration Practice and Discourse
in the Context of Computer-Mediated Digraphia
Jannis Androutsopoulos
Introduction
‘Greeklish’ or Latin-alphabet Greek – that is, the representation of the Greek
language with the Latin script – has been a feature of the Greek-speaking internet
from the start.1 ‘Greeklish’ became widely known in the 1990s, was read as a
‘danger’ to the Greek language at the turn of the century, but is still widely used
today, in transnational communication as well as within Greece, even though
technological developments have largely abolished the conditions that necessitated
its spread. Example 1, which offers our first sample of ‘Greeklish’ spelling, resonates
with findings on ‘ASCII-ized Arabic’2 in an important respect: script choice in
computer-mediated discourse may be turned from a technical constraint into a
symbol of the medium in which it occurs:
Example 1
Oi perissoteroi xrhstes sthn ellada exoun pleon th dynatothta na grafoun me ellhnikous
xarakthres, kanonika. Omws, elaxistoi einai ekeinoi pou exontas th dynatothta afth
thn aksiopoioun, kai oi perissoteroi akolouthoun thn palia methodo grafhs, xwris
na yparxei kapoios profanhs logos. [ . . . ] H dikh mou ekdoxh einai pws to e-mail
antimetwpizetai apo tous perissoterous ws ksexwristo meso epikoinwnias, me dika
tou symvola kai kwdikous, kai enas ap’ aftous einai h xrhsimopoihsh twn latinikwn
xarakthrwn.3
1
The terms ‘Latin/Latinization’ are preferred to ‘Roman/Romanization’ – cf. Coulmas (2003)
32 and Coulmas (1996) – because they are used in Greek linguistics and Greek public discourse: cf.
Moschonas (2004). ‘Greeklish’ and Latin-alphabet Greek (or LAG) are used interchangeably in this
chapter, the former usually in inverted commas to signal its non-technical origin.
2
Cf. Palfreyman and al Khalil (2003).
3
Personal communication, 1998.
221
222 Jannis Androutsopoulos
Most users in Greece have now the option of writing normally, with Greek characters.
But only few are taking advantage of this option, and most are following the old writing
method without an obvious reason. [ . . . ] My own view is that e-mail is viewed by
most users as a distinctive mode of communication with its own symbols and codes,
one of these being the use of Latin characters.
This chapter approaches Latin-alphabet Greek (henceforth also LAG) from

a sociolinguistic and discourse-analytic perspective. It offers an overview of its
development and contemporary use, and examines its linguistic hallmark, namely
spelling variation, and the way this is constructed in the discourse of internet
users and the wider Greek speech community. Moreover, it aims at drawing
wider implications for the sociolinguistics of orthography in the age of computer-
mediated communication. For the purposes of my argument, I shall draw on the
distinction between ‘autonomous’ and ‘ideological’ approaches to orthography,
on the sociolinguistic notions of digraphia and focusing, and on the distinction
between transliteration as an abstract system and as individual practice.
In the first place, the notion of digraphia4 will provide a conceptual frame for
this chapter. In particular, the term ‘computer-mediated digraphia’ is coined for
the simultaneous use of both the native Greek and the Latin script in computer-
mediated interaction. The characteristics of Greek computer-mediated digraphia
include: a lack of stability and societal agreement on the use of LAG; its persistence
in transnational communication; the lack of a widely known transliteration standard,
which results in a wide range of variability in transliteration practice; the emergence
of metalinguistic discourses among internet users as well as in nationwide media;
and the importance of technological developments for the shifting patterns of LAG
use and evaluation.
Within this frame, transliteration practices and discourses on ‘Greeklish’ will be
examined in terms of ‘ideological’ and ‘autonomous’ approaches to orthography.5
Briefly, an autonomous approach views orthography as a ‘neutral’ technology for
the representation of spoken language. In contrast, an ideological approach views
orthography as a set of social practices in specific social and cultural contexts. From
this perspective:
orthography can be seen as the site of potentially intense struggles over identity and
power, in which issues like the purpose of literacy and the status of languages are
central, and orthographic characters [ . . . ] may be imbued with a symbolic meaning
that makes their phonemic symbolism and learnability of secondary importance.6
Rather than being a mere matter of technological necessity, as an autonomous

approach would assume, ‘Greeklish’ is a rich site of aesthetic and ideological
conflict. This holds true for both its relationship to the Greek script and the
relationship between different versions of Greek-to-Latin transliteration. An
4
See Coulmas (2003), Grivelet (2001a).
5
Cf. Street (1984), Sebba (1998), (2000), (2003), (2007).
6
Sebba (1998) 20.
Greeklish 223
ideological approach to orthography allows us to understand how writers’ spelling

choices and metalinguistic assumptions are shaped by the symbolic and aesthetic
meanings they attach to alternative schemes of transliteration.7 On the other hand,
the distinction between ideological and autonomous approaches to orthography
will be used as an analytical tool in order to reveal that the opinions and arguments
expressed for and against ‘Greeklish’ are sometimes ‘autonomous’ and sometimes
‘ideological’ in nature.
The variability of ‘Greeklish’ spelling is due to the fact that Greek-to-Latin
transliteration is not acknowledged within the Greek educational system. With
transliteration standards hardly known outside expert circles, Greek internet users
have developed a range of informal transliteration schemes, appropriating the Latin
script in innovative ways. The distinction between transliteration as an abstract
system and individual style, along with the notion of focusing,8 will be used to
examine how individual regularity and ‘local’ norms emerge in a normative vacuum,
in which choices of Latinized spelling are neither prescribed nor sanctioned. I shall
reconstruct two main transliteration schemes, ‘phonetic’ and ‘orthographic’, and
argue that they are relevant orientation points for both transliteration practice and
discourse. Moreover, I shall draw on the notion of ‘focusing’ to examine how the
range of transliteration alternatives is reduced as individual spelling styles converge
towards ‘local’, group-related norms.
‘Greeklish’, past and present

Popular terms such as Fragolevantinika and Fragochiotika allude to the use of
LAG in the nineteenth and early twentieth centuries, notably by Greek traders
from the island of Chios and (ethnically non-Greek) Levantine traders in Smyrna,
Asia Minor. Sporadic evidence suggests that the Latin script had already been
used in the Early Modern era, notbly for folk poetry or catechism, in Greek areas
under Venetian rule or with some other Catholic presence.9 These early, albeit
poorly documented, instances of script change clearly differ from contemporary
‘Greeklish’ in terms of their political context, social spread and communicative
purposes; however, they display a typical feature of digraphia, in that they emerge
in a situation of interethnic and intercultural contact.10
The intercultural and transnational dimension of script choice is also manifest
in proposals for Greek orthographic reform in the inter-war era of the twentieth
century. Its advocates argued for the simplification of the historical orthography of
Greek as a measure against ‘the plague of illiteracy’, and proposed the adoption of
the Latin script in order to avoid confusion between different spellings of the same
7
Cf. ibid. 36, 40.
8
Cf. Le Page and Tabouret-Keller (1985); Sebba (2003), (2007).
9
See the articles in Filintas (1980) and Zachos-Papazachariou (1999). Wikipedia (2006) points
out a wider tendency for ‘script to follow creed’ in the Balkans, which also encompasses the use of the
Greek script for the Turkish language: Clogg (1999).
10
Cf. Coulmas (1996) 130.
word in the historical and the proposed phonetic orthography.11 Script change
was also seen as a source of potential economic benefits (especially in the context
of typewriting, printing and publishing) and as a symbolic alignment with the
‘civilized nations’ of the West. The reformist voices of this time explicitly challenged
the implicit assumption that script change would amount to a loss of national
identity, pointing to the precedent of the Turkish script reform. Thus, although the
main reformist arguments were ‘autonomous’ in nature, their proponents clearly
recognized that the ideological dimension of Greek orthography, in the shape of its
national symbolic value, would be their main obstacle. Significantly, Latinization
and phonetic spelling were implicitly equated in this reformist discourse; what is
referred to below as ‘visual transliteration’ was presumably inconceivable at this
time.12
Latin-alphabet Greek remained a minor issue in the first post-war decades,
with its appearance restricted to telegrams to or from abroad, cash register receipts,
and, allegedly, early broadcasts by EMY, the national weather-report service.13 It
resurfaced with the emergence of the internet in the late 1960s. The early internet
operated on the seven-bit ASCII character encoding set (first published in 1967),
which provided for the encoding of 128 characters based on the English alphabet,
and therefore excluded the representation of languages with non-Latin script.14
‘Greeklish’ was presumably ‘reinvented’ in those early internet days, perhaps
among Greek-speaking students and researchers at universities in the USA. To
judge from sporadic references in mailing lists and on websites, it was apparently
already in use in ‘Arpanet’, the computer network that preceded the internet; the
‘visual’ transliteration scheme with its peculiar letter-to-number correspondences
seems to have been an innovation of that time.
Latinization was presumably the only option available to the few Greek-speaking
internet users, in Greece or abroad, throughout the 1980s. Since the early 1990s,
the gradual development of the unicode character-encoding standard has made
the representation of symbols from a wide variety of writing systems possible on
computer screens, and during the 1990s, as a result, the use of Greek script on the
internet became an increasing technical possibility. However, its actual availability
to individual users was still limited by their access to hardware and software
facilities. This gap between technical possibility and individual availability led
to the persistence of LAG as the lowest common denominator throughout the
1990s. In an e-mail survey I conducted in 1999,15 sixty-nine per cent of residents
of Greece and eighty per cent of residents abroad reported using LAG in more
11
See documentation in the volume Φωνητική Γραφή (‘Phonetic Writing’), which comprises
essays by linguists and intellectuals published between 1929 and 1931 in the journals Νέα Εστία and
Πρωτοπορία: Filindas et al. (1980).
12
Below, p. 232.
13
Cf. Zavras (n.d.), Wikipedia (2006).
14
Information on computer-related terms such as Unicode, ASCII and Arpanet is provided by the
respective Wikipedia entries (see links below, p. 249).
15
See pp. 229–30 below.
Greeklish 225
than half their e-mails. Data from that period suggest that LAG was used for a
variety of public purposes in Greece, such as commercial newsletters and even
formal announcements within universities:
Example 2
H Sygklhtos sthn ar. [000/000] synedriash ths, apofasise na sas parakalesei na
enhmerwsete ta melh DEP tou Tmhmatos sas oti, basei ths isxyousas nomo8esias, den
einai nomimh h ana8esh autodynamou didaktikou ergou se metaptyxiakous foithtes
`h ypokatastash tous apo tous en logw foithtes sta didaktika tous ka8hkonta.16
In its nr. [000/000] meeting, the Senate decided to ask you to inform members of
staff in your department that on the basis of current legislation, the assignment of
independent teaching to postgraduate students or the substitution of staff by such
students for teaching duties is not allowed.
While such usage is probably extinct by now, ‘Greeklish’ is still technically

necessary in a number of settings abroad, such as university or internet café computers
where Greek fonts are unavailable. In contexts of transnational communication,
such as mailing lists with worldwide-dispersed members, ‘Greeklish’ has ensured,
and still does ensure, that even the few users without access to the Greek script
will be able to participate. In sum, even though an increasing number of Greek
internet users had access to the Greek script by the late 1990s, LAG was so firmly
established among early adopters of computer-mediated communication that it
was referred to as the ‘old writing method’, as in Example 1. One might suspect that
it was in this transitional period, when both scripts were available to an increasing
number of users, that symbolic values of LAG such as the ‘code of the internet’ or
the ‘code of the e-mail’ (Example 1 again) were established.
With the spread of computer-mediated communication (CMC) across Greek
society, it was only a matter of time before ‘Greeklish’ emerged as a matter of public
discourse. In 1996, acro, a typography magazine from Thessalonica, published on
its back cover this question: etsi tha grafetai i glossa mas apo do ke bros? (‘Is this how
our language will be written from now on?’). This anticipated a wider debate in the
years to come, a debate based on the synecdochic relationship of language use in
CMC to ‘our language’ as a whole. Although press reports of ‘Greeklish’ have been
attested since 1995, they remained sporadic until the issue in January 2001 of an
open letter by the Athens Academy, which warned against a possible substitution
of the Greek by the Latin alphabet as a consequence of LAG use on the internet;
an excerpt is given in Example 3. Clearly based on an ideological approach to
orthography, in which the Greek script is viewed as a paramount national symbol,
the Academy open letter proclaimed a ‘phobia of Latinization’.17 It positioned
‘Greeklish’ as the descendant of earlier uses under foreign rule, and constructed
16
University circular mail, 1998.
17
Cf. Moschonas (2004).
it, notably through religious allusion and metaphor, as a foreign threat to national
identity:
Example 3
Θεωρούμε ανόσια αλλά και ανόητη κάθε προσπάθεια να αντικατασταθή η
ελληνική γραφή στο λίκνο της [. . .] Όπως και επί Ενετών, όταν αυτοί στα μέρη
που κυριαρχούσαν προσπάθησαν να αντικαταστήσουν στα ελληνικά κείμενα
τους ελληνικούς χαρακτήρες, έτσι και τώρα θα αντισταθούμε, καλώντας όλους
τους συνέλληνες να αντιδράσουν για την πρόρριζα εξαφάνιση των ανίερων
αυτών σχεδίων.18
We consider unholy, but also senseless, any attempt to replace the Greek script in its
own birth-place [. . .] Just as during Venetian rule, when the rulers attempted to replace
the Greek alphabet in Greek texts, we will resist now too, calling on all fellow Greeks
to respond and ensure that these unholy plans are destroyed, root and branch.
The Academy statement sparked a lively public debate, which displayed all the
signs of a ‘moral panic’, with a rapid build-up of public concern and a minor issue
identified as a threat to a community’s values.19 Aspects of this debate are critically
examined by Koutsogiannis and Mitsikopoulou who identify three main trends
in a corpus of fifty-six newspaper articles.20 The ‘retrospective trend’, developed
in more than half of their corpus, aligns itself with the phobia of Latinization that
the Academy open letter evoked. Equating LAG with a supposed national threat,
it draws on metaphors of military attack on the Greek language, and metaphors
of resistance to that attack. The ‘prospective trend’ challenges this line of thought
and adopts a positive stance towards technology, by, for instance, discussing the
software problems that led to the use of ‘Greeklish’. It also includes traces of a
sociolinguistic discourse, by identifying Greeklish as a ‘new jargon’ or ‘language
variety’. The ‘resistive trend’ combines the technology-friendly discourse of the
prospective trend with a linguistic critique of globalization, thereby foregrounding
the promotion of linguistic diversity in the information age. Koutsogiannis and
Mitsikopoulou suggest that these trends are reminiscent of past debates on the
Greek ‘language issue’: the retrospective trend is reminiscent of ideas once used in
support of katharevousa, whereas the prospective and resistive trends echo ideas
used in support of demotic.21
Even though ‘Greeklish’ is not an issue of great concern in the Greek media at
present, its use still persists, despite official protestations to the contrary.
18
Open letter by the Athens Academy, 2001.
19
Moschonas (2004).
20
Koutsogiannis and Mitsikopoulou (2003).
21
Ibid.
Greeklish 227
Computer-mediated digraphia
Digraphia is defined as the use of two different scripts for the representation of
the same language.22 Although the term is reminiscent of diglossia,23 theoretical
attempts to deduce features of digraphia from diglossia are fraught with difficulties.
For instance, while the scripts involved in a digraphic situation will usually display
a functional distribution to different societal domains, they are not necessarily
stratified in terms of prestige in a collectively accepted way. It is therefore more
useful to treat diglossia and digraphia as conceptually unrelated, and to elaborate
the notion of digraphia based on inductive examination of particular digraphic
situations.24
Specialist discussions of digraphia do not focus on CMC, but refer to it as a side
issue (notably in the case of Chinese).25 Nevertheless, the notion is used in passing
by Tseliga with respect to ‘Greeklish’ and discussed in more detail by Palfreyman
and al Khalil with respect to ‘ASCII-ized Arabic’.26 Building on these observations,
I propose that ‘Greeklish’ be examined as an instance of computer-mediated
digraphia, broadly defined as the simultaneous use of native and Latin script
for the same language in computer-mediated communication. As the preceding
discussion has shown, Greek has witnessed instances of digraphia in the past. As is
the case for other languages with a non-Latin script, certain uses of digraphia have
been institutionalized by the Modern Greek state for purposes of international
communication (with road signs and passports the most obvious examples).
However, what is peculiar to computer-mediated digraphia is the active use of
Latin-alphabet Greek by considerable portions of the population in conjunction
with an even wider passive awareness. Internet penetration in Greece reached one
million users in 2000, and in 2005 an estimated 3,800,000 users (33.7 per cent
of the population).27 We may therefore assume that the majority of the younger
population are actively or passively familiar with ‘Greeklish’, not least because of
its continuous presence on web discussion boards. Media reports, meanwhile, have
spread this awareness to many non-internet users.
Within CMC, Greek/Latin digraphia is generally restricted to contexts of
computer-mediated interaction.28 These might be further specified in terms of
residency and communication technologies. Latin-alphabet Greek is widely used
in transnational exchanges, both within the Greek diaspora and between Greece
and abroad. For instance, it is the default choice in a number of diaspora mailing
lists, newsgroups and discussion boards that I am personally aware of, whether
22
Cf. Grivelet (2001b); Coulmas (1996) 129–30, (2003) 231–4; Unseth (2005) 36–7.
23
See Romaine (1995), ch. 2, for a comprehensive discussion of diglossia, and Li (2000) for a
collection of influential early papers.
24
Cf. Grivelet (2001b); Coulmas (1996), (2003).
25
Cf. the papers in Grivelet (2001a).
26
Tseliga (2007); Palfreyman and al Khalil (2003).
27
See http://www.internetworldstats.com/eu/gr.htm, accessed on 22 February 2007.
28
LAG is generally not used for edited content (weblogs, websites), except as an emblem of IRC
culture (e.g. on irczone.gr).
because of technical constraints or other reasons, such as convenience, convention,

audience considerations or literacy competence.29 Within Greece, LAG is used by
default, owing to technical constraints, in internet Relay Chat (a separate platform
that preceded the web), and within Instant Messenger systems.30 However, it is
optional in web-based chat environments and on discussion boards.
In sum, although LAG has a functional specialization with respect to the
native Greek script in that it affects only one particular area of written language
(computer-mediated interaction, as opposed to edited website content), there is
a lack of domain exclusivity, as native and Latin scripts are both used in the same
environments. However, the domain specialization of LAG does not rule out its
occasional appearance in off-line public discourse directed at a Greek audience.
This resonates with findings by Palfreyman and al Khalil, who note the occasional
use of ‘ASCII-ized Arabic’ in private off-line literacy practices.31 I suggest that
such ‘domain transgression’ may be understood as metaphorical ‘script-switching’,
in which LAG evokes certain symbolic values of CMC, such as future orientation,
technological competence, and an international outlook, outside its ‘proper’
domain.
Two examples from the turn of the century will illustrate this point. The
first is the headline of a bank advertisement in the form of an e-mail address:
epithesistomellon@geniki.gr (that is, ‘attackingthefuture@geniki.gr’, geniki being
the bank’s name). Script choice is embedded here in genre choice: the advertisement
headline appropriates the e-mail format, which inevitably comes in Latin-alphabet
Greek. Both genre and script transfer the connotations of technological competence
and future orientation to the advertised bank. The second example comes from a
restaurant review in a lifestyle magazine. It includes a script-switch not only for
the English word Wallpaper, but for parts of the Greek text as well (italicized in
the translation): Οι Αθηναίοι έγιναν πολίτες του Cosmou. Trone sto Kreas, που
τους θυμίζει την (αν)αισθητική του Wallpaper (‘Athenians have become citizens
of the World. They dine at Kreas, which reminds them of the (non) aesthetics of
Wallpaper’). The switch is locally motivated by the restaurant’s own choice of the
Latin script for a common Greek word (Kreas, ‘meat’) and contextualizes, perhaps
not without a certain irony, the restaurant’s supposed cosmopolitan character. In
both examples, the cultural semantics of script choice are strengthened by lexical
choice (‘future’, ‘citizens of the world’).
These examples suggest that the awareness of LAG in Greek society is so
widespread that media discourse may occasionally exploit its symbolic potential
for particular purposes. Interestingly, both examples draw not on the ‘orthographic’
29
These issues must be examined separately for each environment. For instance, technical
constraints still hold true for a transnational mailing list for Greek draft evaders; but on discussion
boards for second-generation Greeks in Germany (see Androutsopoulos, 2007) LAG is the default
choice, even though it is technically possible to use the Greek script.
30
Cf. Wikipedia (2006).
31
Greeklish 229
transliteration that is the true innovation of the Greek-speaking web, but on a more
traditional ‘phonetic’ representation. It is transliteration schemes we now turn to.
Spelling variation in Latin-alphabet Greek: schemes, styles, norms and

attitudes
Spelling variation is the most noticeable linguistic feature of Latin-alphabet Greek.32
This is independently confirmed by Palfreyman and al Khalil, who point out that
‘“ASCI-ized” “orthographies” do not typically have the consistency characteristic
of other orthographic systems’.33 Against the backdrop of the preceding discussion,
we might say that ‘Greeklish’ is lacking in consistency, because it is neither acquired
through the normative mechanisms of the educational system nor controlled by
norm-enforcing authorities. Greek-to-Latin transliteration standards do exist,
of course, notably the ISO/ELOT standard, as well as a variety of philological
transliteration schemes.34 However, none of these are taught in primary or
secondary education.
Data
We turn now to vernacular responses to the transliteration problem.35 The findings
reported here are based on research that was carried out between 1997 and 2000,
involving three sources of data. First, a small, non-systematic sample of e-mails
from various sources was examined in order to assess the range of transliteration
variation and to reconstruct transliteration schemes. Secondly, a case study of
personal e-mails by six individuals examined the relationship between transliteration
schemes, transliteration styles, and socio-professional milieus. Thirdly, in early
1999 an e-mail questionnaire was sent to various mailing lists and individual users.
The questionnaire was returned by seventy-six participants (thirty-one female,
forty-five male), almost equally divided between those residing in Greece (fifty-
one per cent) and abroad (forty-nine per cent). In terms of age, thirty per cent of
the respondents were under twenty-four, forty-two per cent were between twenty-
five and thirty-four, and twenty-eight per cent were over thirty-five. In terms of
occupation, the sample included university students, language professionals such as
translators and journalists, linguists and other university staff, IT professionals and
members of other occupations. Most respondents were members of four mailing
lists: thirteen belonged to GreekWeb, twelve to EEXI (‘Hellenic Association of
Internet Users’), fourteen to Hellas, and sixteen to a mailing list for Greek students
at King’s College London. Members of GreekWeb and EEXI were mostly residents
of Greece, whereas Hellas attracted a diasporic audience. To these were added
32
See Androutsopoulos (1998).
33
Cf. Palfreyman and al Khalil (2003) 12.
34
See Table 1 on p. 231 below, as well as Coulmas (1996) and (2003) and Zikmund (1996).
35
The following account draws on findings published for the first time in English here, but
previously published in Greek: Androutsopoulos (1998), (1999), (2000), (2001).
five staff members of the English Department at the Aristotle University of

Thessalonica, and a miscellaneous group of sixteen individual respondents.
The first part of the questionnaire included questions on the use and evaluation
of LAG. In the second part, respondents were presented with two Latin-alphabet
variants of a Greek sentence and asked to estimate which one they would expect
to receive from different sorts of interlocutors. The third part consisted of a
transliteration task, in which respondents were asked to translate into Greek four
English sentences which were constructed in such a way that their expected
Greek versions included letters that in practice engender several alternative
Latin spellings.36 Premised on the assumption that users would follow the
transliteration style they used in everyday practice, this task elicited a controlled,
self-initiated sample of ‘Greeklish’ spelling that can be analysed in a language
variation framework. Specifically, Greek graphemes that received two or more
alternative Latin realizations were considered to be linguistic variables. A case
in point is the Greek grapheme <ω> omega, which may be represented by any
of the Latin graphemes <w>, <o> or <v>.37 These alternatives were not divided
into standard and vernacular variants, as would have been the case in traditional
variationist sociolinguistics, but were grouped together in the transliteration
schemes presented below. This procedure allowed me to examine the relationship
between transliteration schemes and individual transliteration styles, as well as the
relationship of these styles to users’ demographic characteristics and their responses
concerning language attitudes in the questionnaire.
Transliteration schemes
At first glance, Latin-alphabet Greek is extremely heterogeneous. While thirteen
graphemes of the Greek script are always transliterated with a single Latin grapheme
in my data, the remaining graphemes (including digraphs and diphthongs) receive
two, three or even more Latin equivalents.38 Some of these correspond to official or
academic transliteration norms; others diverge from these norms, but are widespread
in vernacular practice; still others are rare or idiosyncratic. The heterogeneity of
transliteration practice becomes obvious with particular words that are ‘difficult’ by
the standards of Greek orthography. For example, the transliteration task elicited
twenty-three different Latin-alphabet versions of the word διεύθυνση (‘address’),
36
The sentences, with their expected Greek equivalents, are: (1) ‘I love my wife Eleni like my
own life’ (‘Αγαπώ τη γυναίκα μου την Ελένη σαν την ίδια μου τη ζωή’), (2) ‘I will go there next
week’ (‘Θα πάω εκεί την επόμενη εβδομάδα’), (3) ‘Yesterday I woke up very early’ (‘Χθες ξύπνησα
πολύ νωρίς’), (4) ‘Sorry, I forgot his address’ (‘Συγνώμμη, ξέχασα τη διευθυνσή του’). In the Greek
versions, omega <ω> occurs e.g. in the words αγαπώ, όπως, ζωή, πάω, νωρίς, συγνώμμη; eta <η>
and upsilon <υ> occur in e.g. ξύπνησα, διεύθυνση; theta <θ> in θα, διεύθυνση, χθες; and so on.
37
By convention, orthographic representations are enclosed in angle brackets < >; phonemic
representations are enclosed in slashes / /. Greek graphemes are also represented by their Latin names.
38
These are: three vowels <η, υ, ω>; eight consonants, <β, γ, δ, ν, π, ρ, ξ, χ>; six digraphs <αι, ει,
οι, ου, μπ, ντ>; and the diphthongs <αυ, ευ>. The invariant graphemes are <α, ε, ζ, ι, κ, λ, μ, ο, σ, ς,
τ, φ, ψ>.
Greeklish 231
which differed in the representation of the Greek graphemes <ευ>, <θ>, <υ>, <ν>
and <η>. Only three of these were employed by seven or more users (diefthinsi,
diey8ynsh, dieuthinsi), while thirteen versions appeared only once each, including
forms such as dieu0unsh, dieu8uvsn, dievthinsi and dief8hnsh.
The case of διεύθυνση is indicative of the popular belief that Latin-alphabet
Greek has ‘no rules’ and that people transliterate ‘as they please’. However, my
findings suggest that notwithstanding any individual inconsistencies, users tend
towards either a ‘phonetic’ or an ‘orthographic’ transliteration scheme, which can
be reconstructed on the basis of inductive generalization and by taking into account
users’ metalinguistic awareness. Table 1 displays the schemes, illustrated by a
selection of Greek graphemes, and adds the ISO/ELOT standard for reference,39
while Examples 4-6 provide a typical instantiation of each scheme:40
Table 1
Transliteration schemes, illustrated by a grapheme selection
‘Greeklish’ transliteration
Greek phonemic ISO/ELOT phonetic orthographic
grapheme value standard
keyboard-based visual
η /i/ i i h n
υ /i/ y i y u
ει, οι /i/ ei, oi i ei, oi ei, oi
ω /o/ o o v w
ου /u/ ou u oy ou
β /v/ v v b b
θ /th/ th th u 8, 0, 9
ξ /ks/ x x j 3
χ /x/ ch ch, h x x
Example 4: phonetic transliteration

Ta siberasmata tou vivliou dialioun tin paramorfomeni ikona pou epikrati simera se
merida akadimaikon i opii sindeoun tin epifilaktiki stasi ton Arheon Ellinon apenanti
sto erota me tous sihronous provlimatismous mas shetika me to sex, ton erota ke tin
ikogenia.
Example 5: keyboard-based transliteration

Oi Kybernografoi, poy dianyoyn ton deytero kyklo ekdoshs toys, einai pleon h apolyth
phgh lifestyle plhroforishs sto Internet [ . . . ] to periodiko einai diathesimo dwrean,
ejyphretwntas to diafhmistiko apolyto, afoy ta eksoda toy kalyptwntai plhrws apo tis
diafhmiseis toy.
39
ISO 8859–7 / ELOT–928: cf. Wikipedia (2006).
40
Translations are omitted, since the propositional content of the examples is not relevant to the
discussion. Examples 4–6 belong, respectively, to a newsletter of 1998, a newsletter of 1997, and a
mailing list post of 1998.
Example 6: visual transliteration

Pote dev eipa sigoura dev 8a 3ava-agapisw
pote mou egw de distasa apo tnv arxn v’ arxisw
Kai va pou bgeika aln8ivos kai n tuxn ntav mazi mou
nr8es kai eida eutuxos v’ allazei n zwn mou.
Phonetic transliteration is based on correspondences between Greek phonemes

and Latin graphemes. It therefore includes elements of transcription,41 and inevitably
results in a simplification of historical Greek orthography. A consistent phonetic
transliteration will use Latin for all six Greek graphemes representing the /i/
sound, <ι, η, υ, οι, ει, υι>; it will also employ Latin <o> for both <o> omicron and
<ω> omega, and for the digraph <ου>. By contrast, the premise of orthographic
transliteration is the preservation of Greek orthography. The correspondence
between the two scripts is achieved in two different ways, represented here as
sub-cases of the orthographic scheme.42 In the keyboard-based scheme, users type
on their keyboard as though typing in Greek script; as a result, <η> eta becomes
<h>, <ξ> xi becomes <j> and <ω> omega becomes <v>. The visual scheme aims at
simulating the shape of Greek letters with Latin characters as closely as possible.
Widespread solutions include <w> for <ω> and the use of similar-looking numerals
for letters without a visually similar Latin grapheme, as with <8> for <θ> theta and
<3> for <ξ> xi. These numerals are ‘graphemicized’: they are treated as distinctive
units of visual transliteration. Less common visual variants in my data include <n>
for <π> pi, <v> for <ν> nu and for <ρ> rho (compare Example 6). Thus a
comprehensive visual transliteration eventually amounts to a radical restructuring
of the inherited graph-to-graph correspondences between the Greek and the Latin
script.
While the phonetic scheme is closer to conventional transliteration standards
and offers readability to non-native or even non-speakers of Greek, it still involves
a new orthography that must be learned separately. By contrast, both sub-cases
of orthographic transliteration radically diverge from conventional transliteration
standards. They are entirely dependent on the native script for decoding and are
therefore unreadable without prior knowledge of Greek orthography or keyboard
layout: compare and contrast phonetic zoi , keyboard-based zvh, and visual zwh
or zwn, for the word ζωή /zo'i/ ‘life’. Keyboard-based transliteration is the most
convenient from a user’s perspective, but occasionally results in word-forms that
are neither phonetically accurate nor visually similar: thus ξανθός (‘blond’) is
41
I maintain a distinction between transcription as a scientific procedure for the written
representation of spoken discourse and transliteration as the conversion of graphemes from one script to
another: cf. Coulmas (2003) 31.
42
Other researchers favour a tripartite classification, e.g. sound-, glyph- and keyboard-based
(Zavras, n.d.) or positional, visual and phonetic transliteration (Dimoliatis, 2000). I treat keyboard-
based and visual transliteration as sub-cases of the orthographic scheme because they both aim at
retaining the Greek orthography, and because the realization of a purely visual scheme seems quite rare
in practice; a keyboard-based style with some additional visual equivalents is much more common.
Greeklish 233
rendered as <januov>. Visual transliteration offers a maximum of iconicity, though

at the expense of convenient key correspondences. And while the keyboard-based
scheme provides a fixed set of graphemic correspondences, the visual scheme
promotes individual linguistic creativity and allows for a large number of variants,
as engaged users seek ever better visual equivalents. For instance, in early 2006
I came across a hitherto unnoticed variant for upper-case pi <Π>, formed from a
double Latin <T>: thus <TTap8evwv> for Παρθενών (‘Parthenon’). If we arrange
these schemes across a ‘globalness/localness’ continuum,43 phonetic transliteration
is clearly the ‘globally oriented’ solution, while visual transliteration has the most
distinctive ‘local feel’.
From schemes to styles

In the absence of institutional acquisition and control of transliteration norms, the
relationship between schemes and individual practice is not clear-cut, because
transliteration schemes are orientation models that allow for internal variation. For
example, under a phonetic orientation, individual variation may occur through
the use of certain homophone digraphs which encode grammatical distinctions.44
Visually oriented users, again, differ in terms of how radically they restructure
the traditional script correspondences: the spellings <Par8enwn>, <Par8evwv>
and <TTap8evwv>, for instance, are all based on the same visual logic. Other
users operate on a keyboard-based scheme, thereby adding variation between
a few visual equivalents. Even so, transliteration schemes remain determinative
for individual practice. The evidence for this is both ‘etic’ (available through a
linguistic reconstruction of users’ preferences) and ‘emic’ (available through the
users’ own awareness).
Emic evidence is provided by questionnaire comments such as Examples 7–10.
The authors of Examples 7 and 8 point out that they are trying to imitate the
historic orthography of Greek or to follow an ‘orthographic’ and not an ‘auditory’
representation. By contrast, the authors of Examples 9 and 10 declare they are
following ‘the sound’ and not ‘the look’. Example 10 reveals an awareness of
Greek-to-Latin grapheme correspondences in the phonetic scheme, and points to
the fact that transliteration schemes are acquired informally:45
43
Cf. Koutsogiannis and Mitsikopoulou (2003).
44
For instance, instead of phonetic one may use the homophone <oi> as plural masculine
determiner or noun marker, or <ei> as third-person, present-indicative marker.
45
Example 9 indicates another source of transliteration variance not discussed in this paper: a
user’s second or preferred foreign language. For example, <th> for the Greek voiced dental fricative <δ>
(delta) is used only by speakers whose second language is English.
Example 7
Akolou8w to “susthma” pou blepete th stigmh auth, prospa8wntas na mimh8w oso
ginetai kalutera thn ellhnikh istorikh or8ografia!
I follow the “system” you are looking at right now, trying to imitate historic Hellenic
orthography as well as possible!
Example 8
prospa8o na akoloy8o thn kata to dynato pisth or8ografikh kai oxi akoystikh
apodosh.
I try and follow the orthographic and not the auditory rendering as faithfully as
possible.
Example 9
Vasika grafo simfona me to pos akougonte ta ellinika stin aggliki. Den kitao toso to na
miazoun i lexis stis 2 glosses optikos.
Basically I write according to how Greek sounds in English. I am not so much
concerned about the words being visually similar in the two languages.
Example 10
Xmmm nai, xrhsimopoio kapoio sistima alla epeidi to ematha ... grafontas, mou einai
diskolo na to perigrapso me rules. Pantos:
- to psi einai ‘ps’ kai to theta ‘th’ (k.o.k.)
- akolouthoume tous ixous (fthoggous?)
- DEN akolouthoume to “pos fainetai” opos aftoi pou grafoun to theta me ‘8’ kai to
omega me ‘w’
- h orthgrafia aplopoieitai ... ligo (isa isa gia na diabazetai pio efkola to keimeno ... px
ta omega sinithos ginontai omikron) (ΕΧ11)
Hm, yes, I do use a system, but because I learnt it . . . by writing, it is difficult for me
to describe it with rules. At any rate:
- Psi is ‘ps’ and theta is ‘th’ (etc.)
- We follow the sounds (phthongs?)
- We do NOT follow ‘what it looks like’ like those who write theta as ‘8’ and omega
as ‘w’
- Orthography is simplified . . . a bit (just to make the text easier to read . . . e.g. omega
is usually turned into omicron)
Further evidence for the empirical relevance of transliteration schemes is

provided by frequency analyses based on the questionnaire’s transliteration task. In
a completely random situation, every single spelling alternative for a given Greek
grapheme would have the same chance of appearing. Table 2 suggests that this is not
the case. The numerous Latin variants are not equally important in actual usage:
eta and omega have two main variants each, which make up more than eighty-five
per cent of the respective totals, while other variants such as visual <n> for eta and
Greeklish 235
keyboard-based <v> for omega are quite rare, as is inconsistent transliteration of a

grapheme by the same user (see the variants separated by a slash in Table 2). Theta
has one main variant, <th>, followed by the numeral <8> and then other numerals
with smaller frequency. Upsilon has three main competing variants and a number
of less frequent alternatives.
Table 2
Latin variants for four Greek graphemes (based on transliteration task)
H % Ω % Θ % Y %
i 45.7 o 48.6 th 62.9 Y 35.7
h 41.4 w 47.1 8 22.9 I 24.3
n 4.3 v 1.4 0 5.7 U 22.9
i/h 8.6 o/w 1.4 Q,q 2.9 i/y 10
o/v 1.4 other* 5.6 u/y 5.7
i/u/y 1.4
[* Four variants with a frequency of 1.4 % each, partly distinguishing
between upper and lower case: <9>, <U, u>, <th, u>, <Q, 0>.]
Moreover, spellings from the same scheme systematically occur together in

practice. Table 3 displays the frequency of such co-occurrence between the Latin
variants of a Greek grapheme. The cross-tabulation of variants for omega and eta
suggests that most users who prefer a phonetic transliteration of eta also do so for
omega, and vice versa. More than eighty-four per cent of occurs together with
<o>; a similar frequency holds good for <h> and <w>; but combinations of <h> and
<o> occur much less often. In practice, then, people spell ‘I love Helen’ as Agapo tin
Eleni (phonetic) or Agapw thn Elenh (orthographic) but hardly as Agapw tin Eleni.
The cross-tabulation of omega and theta yields a similar picture. Most people who
choose <th> for theta also use <o> for omega. Those who choose the numeral <8>
for theta tend to combine it with <w> for omega, and all other numerals for theta are
combined only with <w> in the data.
Table 3
Combinatory occurrence of variants (based on transliteration task)*
Ω*Η Ω*Θ
H Θ
i h th 8 0 q 9
N 27 2 N 31 2
Ω o Ω o -- -- --
% 84.4 6.9 % 70.5 12.5
N 5 24 N 12 13 4 2 1
w w
% 15.6 82.8 % 27.3 81.3 100 100 100
[*Excluding values for <ω> : <v> and <η> : <n> as well as for intra-personal
variation.]
The transliteration of Greek graphemes such as <ω> omega, <η> eta and <ξ>
xi emerges from this analysis as a key diagnostic feature for a user’s preferred
transliteration scheme: a preference for <h> for will probably coincide with
<w>, and <3> for xi will probably occur together with other visual variants. But
graphemes such as <χ> and <ου> are less useful ‘predictors’, because their most
frequent Latin variants occur in both predominantly phonetic and predominantly
visual transliteration styles.
From styles to local norms

The discussion so far suggests that individual transliteration styles display scheme-
based consistency without categorically excluding scheme mixing and idiosyncrasy.
However, this tells us little about inter-user similarities; individual regularity does
not by itself amount to societal homogeneity. In fact, one could argue that such
homogeneity is by definition impossible in the absence of institutionally transmitted
and controlled orthographic norms. However, such a view cannot explain, for
example, the spread and popularity of visual transliteration, which must be seen as
part of an implicit norm when used in institutional e-mails (as in Example 2).
The emergence of vernacular transliteration norms is illuminated by the
sociolinguistic notion of focusing, originally developed by Le Page and specifically
applied to orthography by Sebba.46 Focusing refers to the process by which a
community of speakers ‘orient towards a linguistic norm’. Focusing is the effect
of individual ‘acts of identity’: speakers adopt a certain way of using language to
the extent that they identify with a group, a leader, or a point of view. The social
prerequisites for focusing include regular interaction with the members of the target
group and support from educational institutions and the mass media. Its linguistic
outcome is a reduction of variability: the range of linguistic variation that is used
and tolerated by the speakers is reduced. Importantly, focusing does not imply
standardization; in other words, the norm acknowledged by a community is not
necessarily a standard language variety nor, in our case, a standard orthography.
Focused non-standard norms are quite possible in theory and in practice.47
Applying the notion of focusing to ‘Greeklish’ suggests that transliteration norms
will emerge not at the level of the Greek internet-user population as a whole, but
in particular social networks. My findings suggest that such ‘local’ transliteration
focusing may develop in both socio-professional groups and ‘online communities’
– meaning by that phrase networks of computer-mediated communication that are
formed around a common interest or cause.
A small sample of personal e-mails was used to examine the transliteration styles
of members of two distinct professional groups, linguists and media professionals.48
46
Cf. Le Page and Tabouret-Keller (1985); Sebba (2003), (2007). The sources of Examples 7–10
are, respectively, 02.HE, 09.EX, 04.AU, 11.EX, in Androutsopoulos (2001); for the abbreviations, see
Table 5 (p. 238 below).
47
Cf. Sebba (2003).
48
See Androutsopoulos (1998).
Greeklish 237
The data consists of e-mails by three female linguists and three male media
professionals. Despite the obvious limitations of this sample in terms of size
and intervening social variables, the findings (shown in Table 4) are instructive
with respect to the relationship between transliteration schemes and individual
practice:
Table 4
Transliteration styles in two groups
female linguists male media professionals
A B C A B C
ω o o o w ο w, ο
η i i i h, i h, i i
ει i ei i ei i ei
υ i i i y y, i y
γ gh / y g gh g g g
χ kh / h ch ch / x x x x
Although no two users spell identically, each group displays an orientation towards
a particular scheme and a family resemblance within itself. The linguists clearly
follow a phonetic scheme. In vowels, they all (with one exception) simplify Greek
graphemes for the /i/ sound to and use <o> for omega. In consonants, two
out of three choose different Latin equivalents for Greek <γ> and <χ> in order to
represent allophones (separated by slashes in Table 4).49 Media professionals follow
an orthographic scheme, although there is more inconsistency in their spelling.
They all choose the same variants for chi <x> and gamma <g>, and two out of the
three use visual equivalents for vowels.
The next set of findings belongs to the transliteration task. Table 5 charts the
individual choices of fifty-seven users for five graphemes. The columns on the left
show phonetically oriented transliterations, those on the right visually oriented
transliterations. The label at the beginning of each row identifies the relevant user
group. On the left, there is a quite consistent occurrence of variants for omega, eta
and theta, with some variation for xi and chi. The right side displays an equally
consistent occurrence of variants for omega, eta and chi, with some variation for the
other two graphemes. More than half of these users prefer a numeral for theta, and
somewhat less than half for xi as well. Three users towards the bottom of the right
side (04.HE, 69.RE and 66.RE) use the rare visual variant <n> for <η> eta.
The bottom rows on both sides, separated from the main block by a blank row,
might be classified as ‘exceptions’, in the shape of combinations of phonetic and
orthographic variants. At the bottom left, three users add a <w> to their otherwise
49
For example, linguist A transliterates <γ> as <y> for the glide (e.g. in Yanni) and as <gh> for the
voiced velar fricative (e.g. aghia, ligha). For <χ>, linguist C uses <ch> for its allophonic realization as a
voiceless palatal fricative [ç] (e.g. echo, mechri), and <x> for the voiceless velar fricative [x] before front
vowel (e.g. xereto).
phonetic style, and one user combines <o> and with visual variants. At the
bottom right are placed two visual transliterators who select an for eta, and one
keyboard-style transliterator. However, these cases are few, compared to the users
who follow a scheme more consistently.
Table 5
Transliteration styles of 57 users (based on transliteration task)
User Ω Η Θ Ξ Χ User Ω Η Θ Ξ Χ
61.RE o i th x ch 51.KC w h th x x
06.HE o i th x h 31.EX w h th x ch
49.KC o i th x h 20.GW w h th ks x
50.KC o i th x h 26.GW w h th ks x
56.AU o i th x h 34.EX w h th ks x
59.AU o i th x h 55.KC w h th ks x
75.RE o i th x h 70.RE w h th ks x
76.RE o i th x h 18.GW w h th kc x
46.KC o i th ks h 13.HE w h 8 ks x
47.KC o i th ks h 21.GW w h 8 ks x
53.KC o i th ks h 43.KC w h 8 ks x
57.AU o i th ks h 45.KC w h 8 ks x
15.GW o i th ks x 52.KC w h 8 ks x
22.GW o i th ks x 07.HE w h q ks x
25.GW o i th ks x 12.HE w h q ks x
35.EX o i th ks x 02.HE w h 8 3 x
48.KC o i th ks x 03.HE w h 8 3 x
63.RE o i th ks x 05.HE w h 8 3 x
64.RE o i th ks x 40.KC w h 8 3 x
65.RE o i th ks x 62.RE w h 8 3 x
68.RE o i th ks x 09.HE w h 0 3 x
24.GW o i/h th ks x 14.HE w h 0 3 x
38.EX o i/h th ks x 10.HE w h 9 3 x
72.RE o i/h th ks x 04.HE w n 0 3 x
69.RE w n 0 3 x
41.KC w i th ks x 66.RE w n 8 3 x
42.KC w i th ks x
58.AU w i th ks x 74.RE w i 8 3 x
01.HE w i 8 x x
28.EX o i/h 8 3 x
37.EX v h u j x
[Key to user groups: AU = Staff members of English Department of AUTH; EX = EEXI
(Greek Internet Users Association) mailing list; GW= GreekWeb mailing list; ΗΕ = Hellas
mailing list; KC = Greek students at King‘s College London mailing list; RΕ = other
responses.]
Greeklish 239
On this basis, we may now consider whether members of the same user group
transliterate in a similar way. In fact, this is not the case for members of GreekWeb,
EEXI and the King’s College mailing list. However, three out of four English
Department staff members are situated on the left side, and most members of
the Hellas mailing list display impressively similar transliteration styles: eleven
out of twelve Hellas members charted here have a clear visual orientation (the
exception is located at the bottom right). Overall, seventy per cent of Hellas
members transliterate theta with a numeral, the mean value of this choice in the
sample being just thirty per cent; ninety-two per cent transliterate omega with
<w> (sample mean value: forty-seven per cent), and seventy-five per cent select
<h> for eta (sample mean value: forty-one per cent). Seven Hellas members use a
characteristic grapheme combination, which turns out to be a sort of ‘trademark’
for this group: <w> for omega, <h> for eta, a numeral <8, 9, 0> for theta, <3> for
xi, and <x> for chi. Questionnaire comments suggest that participants are quite
aware of this consistency. It is only Hellas members who, when asked whether they
follow a particular ‘transliteration system’, pointed to group norms, as in Examples
11 and 12 (from 03.HE and 14.HE, respectively):
Example 11
PROSPA8W NA XRHSIPOIW TOYS XARAKTHRES POY XRHSIMOPOIOYN
H PLEIOPSHFIA TWN AN8RWPWN POY GRAFOYN S’ AYTH TH LISTA
I try to use the characters used by the majority of people who contribute to this list
Example 12
ena mallon koino systhma opws exei diamorfw0ei sthn Hellas, me kana-dyo prwsopikes
diafores
a rather common system as shaped at Hellas, with a couple of personal peculiarities
These findings lend support to the hypothesis that transliteration norms of limited
range may emerge from individuals who regularly interact within a professional or
a ‘virtual’ community, who wish to identify with that community, and who adjust
their spelling to the community’s prevailing style. Among professional groups that
regularly use e-mail, adopting a particular transliteration style may become part of
the group’s professional habitus or, notably in the linguists’ case, may reflect such
a habitus. The case of Hellas suggests that focusing may occur even in the absence
of off-line interaction. The identification of users with this long-standing on-line
community leads to a convergence of their spelling styles, which is analytically
evident by their clustering. The fact that individual transliteration styles are not
completely identical is no contradiction. As normative sanctions are suspended,
users create idiosyncratic variants without giving up their common orientation.
Attitudes, aesthetics and activism

The awareness of focusing by the members of an online community is just one aspect
of internet users’ practice-based knowledge about Latin-alphabet Greek. This
knowledge, which may also extend to different transliteration schemes as well as to
the historical predecessors of ‘Greeklish’, is evident in the metalinguistic discourse
that unfolds in private and public settings of computer-mediated interaction. This
discourse presumably existed long before ‘Greeklish’ became a media issue in the
late 1990s and, not surprisingly, differs from the mass media agenda on the subject
in important respects.
Fragments of this practice-based knowledge and discourse were elicited in
individual correspondence with experienced internet users as well as through the
questionnaire, which included seven attitudinal statements on ‘Greeklish’ with a
binary ‘yes/no’ response option. The findings suggest a predominantly pragmatic
stance on the part of the users, and contradict the assumptions behind the ‘moral
panic’ of that time.50 In particular, eighty-two per cent of users agree with the
statement that ‘Greeklish’ is ‘just an instrument’; sixty-seven per cent consider it a
‘necessary evil’; fifty-three per cent agree it is ‘ugly, not elegant’; but only twenty-
four per cent view ‘Greeklish’ as a ‘problem’ or ‘threat’ to the Greek language.
However, twenty-eight per cent of respondents believe ‘Greeklish’ is ‘difficult
to read’, and forty-six per cent consider it ‘difficult to write’. Both the reported
reading and writing difficulties and the perception of ‘Greeklish’ as a ‘problem or
threat’ rise with age, the latter ranging from only fourteen per cent among those
less than twenty-four years old, to thirty per cent for those aged between thirty-
five and forty-four, and fifty per cent for the forty-five to fifty-four age group.
Responses also vary by gender, with forty per cent of female users responding that
‘Greeklish’ is ‘ugly’, as opposed to sixty per cent of male users, and seventeen per
cent of female users endorsing the ‘threat’ statement, as opposed to twenty-eight
per cent of male respondents.
Questionnaire data and subsequent correspondence with selected respondents
suggest that the preference for a particular transliteration scheme affects users’
aesthetic evaluations as well as their occasional linguistic activism. Thus some
visual transliterators called phonetic Greeklish ‘anglicized’ (angloprepi) or ‘misspelt’
(anorthografa) – a critique that reveals how lay notions of orthography go beyond
script choice. Some phonetic transliterators accused the visual transliterators of
a ‘vain attempt’ to replicate Greek orthography, while one of them called visual
Greeklish ‘monstrous’ (teratomorfa).
Lay orthographic activism by phonetic transliterators tends to follow an
‘autonomous’ approach to Latinized spelling, while that of visual transliterators
favours an ‘ideological’ approach. For instance, some visual transliterators clearly
attached an aesthetic value to attempts to maximize the visual similarity of
‘Greeklish’ to its Greek model. One user provided me with an example of what
50
This is discussed in detail in Androutsopoulos (2000).
Greeklish 241
he judged to be a ‘beautiful’ (omorfi) transliteration (Example 13). Another user

provided a complete transliteration scheme created by a friend of his (Example 14),
which he called To pio prosegmevo kai omorfo optiko protupo pou exw dei (‘the most
careful and beautiful visual standard I have ever seen’). Both examples include a
number of rare visual variants, including <n> for lower case pi <π>, the numeral <5>
for capital pi <Π>, Latin for rho <ρ> and <v> for lower case nu. The inherent
dilemma of such visual schemes is overtly expressed in the commentary: as their
potential users would have to learn a number of unusual grapheme matches, the
cost of their ‘beauty’ is a loss of processing ease.
Example 13
As npoc8ecw ki’ egw oti ta teleutaia duo xpovia nou ekava Xpictougevva cthv
Qeccalovikh ta mova naidia nou hp8av va mas nouv ta kallavta htav npocfugonoula,
kopitcia cuvh8ws, ano thv Gewpgia; h mhtepa mou, gevvhmevh h idia kovta cta
Bopeioavatolika napalia ths Mauphs Qalaccas, ecneuce va “enalh8eucei” thv katagwgh
tous ... kai egw thv dikia ths :-)
Let me add that, the last two years, I spent Christmas in Thessaloniki; the only kids
who came to sing me the Christmas songs were refugees, mostly girls, from Georgia;
my mother, who was born near the North-Eastern coast of the Black sea, was quick to
“verify” their origin . . . and I verified hers :-)
Example 14
ABGDEZH0IKLMN3O5PSTYFX4Wabgdezh8iklmvjonpctufxyws
Prosnlwsn stnv optikn omoiotnta, ola ta kefalaia diaforetika apo ta mikra, diakrisn sigma
(“c”) kai sigma telikou (“s”)!
Duskolo omws va to ma8ei kai va to suvn8isei kaveis, gi’ auto, av kai to ektimw, exw
suvn8isei se eva pio “sumbatiko” optiko sustnma.
Focus on visual similarity; all capitals are different from small letters, distinction between
sigma (‘c’) and final sigma (‘s’)! It is, however, difficult to learn and get used to; therefore,
although I appreciate it, I have got used to a more ‘conventional’ visual system.
By contrast, phonetic transliterators who actively engaged in ‘Greeklish’ discourse

at the time of my survey resisted the ‘chaos’ and ‘confusion’ of visual transliteration
and proposed ways to improve the efficiency of LAG as a communication system.
A case in point is Spyros M. who circulated his views though mailing lists and
personal e-mails. To the teratomorfa greeklish he opposed his own logiki kai
sugkrotimeni proseggisi pou sevetai tis rizes tis glossas mas (‘logical and organized
approach that respects the roots of our language’), in the form of a transliteration
scheme he called ‘Inter-Greek’, which was basically a slightly modified version of
the ISO/ELOT standard. A second case of ‘autonomous’ orthographic activism
is an academic (though not a linguist) who used a university Computer Centre
newsletter to propagate his views.51 He criticized the ‘deformation’ of Greek by

visual transliteration, which he considered to be a practice ‘without principles’,
riddled with the ‘inherited tyranny of the image of the letter’. To this he opposed
a phonemically based transliteration scheme. While neither of these proposals has
had any wider impact, the fact remains that attitudes towards ‘Greeklish’ at the
turn of the century were structured along a distinction that passed completely
unnoticed in the ‘moral panic’ of that time.
‘Greeklish’ revisited: digraphic literacy and discourse on contemporary web

discussion boards
In the public debate that followed the Academy of Athens statement in 2001, the
then Minister of Education, P. Efthymiou, was quoted as saying that ‘Greeklish’ was
now a thing of the past, because Greek internet users were able to use the characters
of the Greek language.52 The Minister’s statement reflects an ‘autonomous’ view
of orthography, one based on the tacit assumption that the mere existence of a
technological solution must by itself lead to the disappearance of script variation.
Yet more than five years later, literacy practices in Greek cyberspace do not
quite confirm that prediction. A cursory examination of about fifteen Greek web
discussion forums in the spring of 2006 lends up-to-date support for the persistence
of computer-mediated digraphia.
In a nutshell, LAG is present on discussion forums with constituencies as
diverse as military personnel, online gamers, e-chat culture, guitar players, hip-
hoppers and the leftist scene.53 Its typical pattern of occurrence is in mixed-script
discussion threads, in which some entries come in Greek script and others in Latin.
Latin-alphabet posts generally seem less frequent than Greek-alphabet, though
this varies by forum. Many users do not use LAG at all, some use it consistently,
and others alternate between scripts across posts. Script-switching within a post
is quite rare. Notable examples include quotations of media content in the Greek
script, the user’s commentary being in ‘Greeklish’, and metalinguistic discussions
of script choice. Moreover, the script choice of the initial post of a thread does
not seem to determine the script choice of subsequent posts, nor is script choice
used as a contextualization device in the manner of code-switching (for instance,
script-switching as a resource for underscoring disagreement to previous posts); an
exception to this is once again debates on script choice.
Whether these usage patterns are different today from those at the turn of the
century is impossible to determine without detailed diachronic comparisons that
go beyond the scope of this chapter and are, in any case, difficult to carry out, given
51
Dimoliatis (2000).
52
See Maronitis (2001).
53
Instances of ‘Greeklish’ were attested on 14 April 2006 on the following discussion boards,
randomly selected from the in.gr directory and/or the Greek Google: e-steki.com; remalia.com; forums.
gr; greekarmy.net; apn.gr; kithara.gr; irczone.gr; hiphop.gr; athens.indymedia.org. No ‘Greeklish’ was
found on gameplanet.gr or on e-magazino.gr.
Greeklish 243
that most contemporary Greek web discussion forums did not even exist at the turn
of the century, and a large proportion of their members, now in their late teens and
early twenties, were not yet online. What does seem to have changed, however, is
overt evaluations of and policies directed against LAG. Wikipedia suggests that in
2004 ‘a hostile movement against Greeklish’ was formed on some Greek discussion
boards, and a web search for the phrase όχι Greeklish (παρακαλώ), ‘no Greeklish
(please)’, indicates that the ‘movement’ has spread to other forums since then.54 To
be sure, the Greek web landscape is in a state of flux in terms of script policy and
choice. While an explicit ban on ‘Greeklish’ accompanies its thorough absence on
some boards, its declared prohibition does not restrict its use on others; on still other
boards, LAG produces no overt commentary. But my overall impression is that
LAG is increasingly stigmatized among internet users themselves, an impression
supported by the observation that using LAG ‘for business purposes of business
communication is considered as a lack of ability or respect, by some’.55
To illustrate these points, a lengthy thread from a musicians’ discussion board
will be examined in more detail.56 Entitled Ελληνικά και όχι greeklish! (‘Greek not
greeklish!’), it extends over more than two years (January 2004 to May 2006) and
comprises more than forty printed pages. These figures are telling with respect to
the relevance of the topic, which is also underscored by the fact that the discussion
thread is linked on the forum’s front page. Contributions to this thread offer ample
evidence that ‘Greeklish’ never ceased being in use, or being an issue;57 they
illustrate a wide range of arguments against and for ‘Greeklish’; they reveal that its
symbolic value as ‘code of the internet’ is still endorsed by some users who resist
attempts to ban Latin-alphabet posts.
Here, I shall limit the discussion to views for and against ‘Greeklish’, and use the
distinction between ‘autonomous’ and ‘ideological’ approaches to orthography to
sub-classify these views. Some arguments on both sides are instrumental (processing-
or technology-related); others are aesthetically driven or identity-related. The
most common instrumental objection to ‘Greeklish’ is readability: people complain
it is hard to read, while some even claim they do not read Latin-alphabet posts at
all. Others point out that Latinization impedes keyword search and that there is no
necessity for ‘Greeklish’ whatsoever, since the board is completely localized. The
technology-related counter-argument for ‘Greeklish’ – technical necessity – comes
only from users who log in from abroad, and is overtly respected as such. A more
54
See Wikipedia (2006). A Google search for the phrase OΧΙ Greeklish παρακαλώ yielded 111
entries from eight different boards, including awmn.net and adsl.gr, both mentioned in the Wikipedia
article, as well as e-pcmag.gr and gameplanet.gr. A search for OΧΙ Greeklish yielded a larger number of
entries, including the discussion thread that is examined in this section.
55
Wikipedia (2006).
56
http://forum.kithara.gr
57
For instance, some participants frame their contribution with comments such as: Το κλασσικο
προβλημα ολων των Forum (‘The classic problem on all forums’), Παλιο το θεμα αλλα παντα
επικαιρο. Και παντως για την ωρα αλυτο (‘Old topic but always timely, and still unsolved’), or Σε
ολα τα forum, το ιδιο ζητημα (‘Same old issue on all forums’).
commonly shared view, which is less an explicit argument for ‘Greeklish’ than a
guess on the part of its opponents, is that ‘Greeklish’ is easier to type. This view,
which suggests yet another parallel to ‘ASCII-ized’ Arabic, challenges my survey
findings from the turn of the century, but confirms Tseliga’s finding that ‘Greeklish
is considered more convenient, faster, easier and less demanding than Greek’.58
The counter-argument to this, repeatedly put forward by users who claim to have
used ‘Greeklish’ in the past, is that Greek orthography skills can be improved by
continuous practice.
At the ‘ideological’ end of the debate, arguments against ‘Greeklish’ resonate with
public discourse at the turn of the century. Some foreground aesthetic concerns,
for instance by considering LAG μια μορφή κακοποίησης της γλώσσας (‘a form
of mistreatment of the language’). Others endorse an essential, timeless relationship
between language and script: η ελληνική γλώσσα φτιάχτηκε με ελληνικό
αλφάβητο και αυτό πρέπει να χρησιμοποιούμε (‘the Greek language came into
being with the Greek alphabet, and this is the one we should be using’). Still others
appeal to national pride: Λίγη υπερηφάνεια για την γλώσσα μας!!!!!!!!!!!!!! (‘Some
pride in our language!!!!!!!!!!!!!!’). Even though these statements are not unanimously
shared, it seems that the ‘nationalization’ of script choice in public discourse has
had an impact on the language ideology of younger, more recent CMC users.
Ideological arguments for ‘Greeklish’, clearly in the minority in this thread,
emphasize its medium-related symbolic value. In Example 15, the writer points
out that his preference for ‘Greeklish’ is not instrumentally motivated (by speed
of typing), but originates in his long internet experience. In a subsequent post, the
same writer claims that using ‘Greeklish’ is enas tropos ekdhlwshs ths diaforetikothtas
ths proswpikothtas tou ka8enos (‘a way of expressing individuality’), a claim that is
strongly rejected by others. In Example 16, the board’s administrator uses script-
switching to challenge this view:
Example 15
den to kanw apo antidrash, apo synh8eia, h taxythta … alla epeidh mou fainetai pio
computer-related. osoi asxolountai arketa xronia me tous H/Y 8a katalaboun ti ennow ...
gia na synopsisw, gia mena internet => Greeklish
I am not doing it as a reaction, because I am used to it or for its speed . . . but because it
seems more computer-related. Those who have been using computers for several years
will understand what I mean . . . To summarize, to me internet => Greeklish
58
Palfreyman and al Khalil (2003) 23; Tseliga (2007) 17.
Greeklish 245
Example 16
Το οτι τα greeklish δεν είναι πια λύση ανάγκης αλλά έχουν γίνει τρόπος έκφρασης,
δεν το φανταζόμουν...
|30r0 & g0 |\|@ Ε|<FR@$T0 0P0$ 8E|_0?
I would never have envisaged greeklish not as a solution of exigency but as a mode of
expression . . .
May I too express myself as I please?
At the end of Example 16, the administrator strategically selects, not a usual LAG
transliteration scheme, but a ‘distorted’, hardly legible spelling that is conventionally
associated with the cracker and hacker subculture to reinforce the main instrumental
argument against Latinization, that it impedes readability and therefore effective
communication. Later on, the administrator again draws on script-switching to
challenge the argument about ease of processing, while at the same time devaluing
phonetic transliteration. His point is that Greeklish is not faster because ‘you are
hitting the same number of keys’, unless one is simplifying diphthongs, which he
illustrates like this: ine kapii pu grafun griklis kapos etsi, ke den mu aresi ka8olu!
(‘some are writing greeklish like this, and I don’t like it at all!’). Yet another user, a
native speaker of English, script-switches in the opposite direction, by using Greek
characters for English, to demonstrate that using a different script for convenience
goes against readability and, ultimately, against the purpose of the board.
Besides providing a forum for the exchange of views on Latin-alphabet Greek,
this thread is also a stage for the administrator’s script policy. Despite the fact
that several participants support a tolerant position, a software script is included
in the board software in order to filter out posts in ‘Greeklish’. This measure has
not met with general acceptance. ‘Greeklish’ advocates have demonstrated how
to ‘trick out’ the script, others have complained about poor usability and abuse of
administrative power, and a few active members have resigned from the board in
protest.
Conclusions
This chapter is a response to the lack of English-language publications on Latin-
alphabet Greek59 as much as to a wider lack of research on informal Latinization
on the internet. As such, its objective is twofold: to offer a comprehensive account
of an important issue in the sociolinguistics of written Greek today, and to use
the Greek case as a test-bed for the sociolinguistic theorizing of script choice and
informal Latinized spelling in an era of computer-mediated communication. I have
suggested that the notion of digraphia responds to the persistence and wide societal
awareness of ‘Greeklish’, and that it allows us to theorize its shifting range and
evaluations. I conclude by summarizing the characteristics of Greek computer-
mediated digraphia and by sketching out some possible future developments.
59
But see Tseliga (2007); Marinis, Papangeli and Tseliga (2005).
The computer-mediated digraphia of Greek emerged as a response to a

technology-induced necessity, to enable native language use on the internet. It is an
instance of unstable, synchronic digraphic literacy, with its historical predecessors
more important when discursively evoked than as guidelines for transliteration
practice. Its functional distribution is restricted to one particular domain of written
language use, in which a few stable niches of Latin script exclusivity are opposed
to a larger number of environments in which both scripts are in competing use.
At the level of script choice, the dimension of social variability that must be
emphasized is that of space. Transnational communication in the Greek diaspora
was and still is a stronghold of ‘Greeklish’, whereas its use in inner-Greek CMC
has declined in recent years, because technical necessity is hardly an issue any
more. Practice, policies and discourses in public CMC environments such as
discussion boards will be decisive for the future of computer-mediated digraphia
within Greece. But the possibility cannot be excluded that Greece and the Greek
diaspora will follow different paths in this respect, with LAG eventually persisting
as a specifically diasporic script choice.
At the level of spelling, the lack of institutional acquisition of transliteration
norms has given rise to a situation of high linguistic variability. In my analysis,
vernacular responses to the transliteration problem have been modelled on the basis
of the relationship between transliteration schemes, individual transliteration styles
and local norms. Depending on the scheme, these responses are seen to range from
similarity to conventional norms to innovative correspondences between native and
Latin scripts, as with the use of numerals; here there is a remarkable interlinguistic
similarity between the Greek case and that of Russian or Arabic.60 Transliteration
schemes respond to the cognitive need for consistency by reducing the available
range of variability. As a further step, they form the bottom line for processes of
inter-user convergence. The phonetic/visual dichotomy shapes both practice, in
the form of the patterns of consistency people create to cope with Latinization,
and ideology. The logic behind phonetic transliteration is ‘global’, ‘instrumental’
and standardization-friendly. By contrast, orthographic and (in particular) visual
transliteration ignores official and traditional representations of Greek by Latin
characters. It is loyal to the native script, but rejects any convention beyond that.
Digraphia in statu nascendi generates a range of different metalinguistic
discourses, from discussions among internet users to nationwide media reports.
These discourses include both ‘autonomous’ and ‘ideological’ understandings of
orthography, and are shaped by positions of control and resistance. In user discourse
at the turn of the century, phonetic transliterators resisted uncontrolled visual
variation and the ‘tyranny of the written word’, while visual transliterators resisted
the idea of a new, separate norm of Latinized Greek orthography and engaged
in a linguistic creativity that was aesthetically driven, its main aim being to add
iconicity to the Greek alphabet. The mainstream view of the issue has contrasted
‘Greeklish’ and ‘Greek’, using the former as a ‘stimulus’ for debate over ‘the shaping
60
Greeklish 247
of modern Greek identity’.61 Public discourse at the turn of the century controlled
LAG by homogenizing it. It ignored its richness and diversity, the emic relevance
of the dichotomy between phonetic and visual transliteration, the creativity or
even ‘beauty’ of individual spelling solutions, and the group affiliations that spelling
choices might indicate. This discourse has sometimes ‘technologized’ LAG,
ignoring or downplaying its instrumental advantages or symbolic values for some
users, or else has constructed it as a threatening ‘exterior’ within the ‘interior’ of
the Greek language.62 The ideological impact of public discourse seems evident in
contemporary user debates and practices, in which aesthetic and identity-related
arguments are used to legitimize attempts to ban LAG from discussion boards,
and are resisted by challenges to the normative, essentialist articulation of script,
language and national identity.
A lesson of the ‘Greeklish’ case is that in this era of computer-mediated
communication, practices and ideologies of spelling and script choice are shifting
along with technological developments, and indeed cannot be understood without
taking the evolution and social spread of technology into account. Although LAG
has been in use since the infancy of the internet, its medium-related symbolic
meaning is a product of the spread of new technologies and the increasing
availability of the native script to a growing user population since the 1990s. The
last years have witnessed a decrease in the environments and the overall frequency
of the Latin in favour of the native script, a development supported by the linguistic
localization of the Web and actively promoted by actors with computer-mediated
authority such as board administrators. The fact that many contemporary internet
novices evidently use the Greek script for most purposes, and that old users
change their script choice, does not deter ‘Greeklish’ followers from continuing
to endorse its medium-related symbolic value, and possibly profiting from its
alleged ease of processing. Whether the story will eventually result in the demise
of Greek computer-mediated digraphia or instead in the persistence of a few stable
digraphic niches, with LAG enjoying a new ‘resistance’ value that it lacked in the
past, cannot, of course, be predicted.
References
Androutsopoulos, J.K. (1998), ‘Ορθογραφική ποικιλότητα στο ελληνικό ηλεκτρονικό
ταχυδρομείο: μια πρώτη προσέγγιση’, Glossa, 46: 49–67.
Androutsopoulos, J.K. (1999), ‘Από τα Φραγκοχιώτικα στα greeklish’, To Vima,
5 September 1999 (available from http://tovima.dolnet.gr/print_article.php?e=B&f=
12688&m=B03&aa=1).
61
Koutsogiannis and Mitsikopoulou (2003).
62
Moschonas (2004).
Androutsopoulos, J.K. (2000), ‘Λατινο-ελληνική ορθογραφία στο ηλεκτρονικό

ταχυδρομείο: χρήση και στάσεις’, Studies in Greek Linguistics, 20: 75–86.
Androutsopoulos, J.K. (2001), ‘Από dieuthinsi σε diey8ynsh. Ορθογραφική ποικιλότητα
στη λατινική μεταγραφή των ελληνικών’, in Y. Agouraki et al. (eds.), Proceedings of
the Fourth International Conference on Greek Linguistics, Nicosia, September 17–19, 1999,
Androutsopoulos, J.K. (2007), ‘Language choice and code-switching in German-based
diasporic web forums’, in B. Danet and S.C. Herring (eds.), The Multilingual Internet,
Oxford: 340–61.
Clogg, R. (1999), ‘A millet within a millet: the Karamanlides’, in D. Gondicas and C. Issawi
(eds.), Ottoman Greeks in the Age of Nationalism, Princeton, NJ: 115–32.
Coulmas, F. (1996), The Blackwell Encyclopedia of Writing Systems, Oxford.
Coulmas, F. (2003), Writing Systems, Cambridge.
Dimoliatis, G. (2000). ‘Για έναν ενιαίο τρόπο μεταγραφής των εληνικών [sic] στον
ηλεκτρονικό ταχυδρομείο’, Πληροφορία [University of Ioannina, Computer Centre
newsletter], 12.
Filindas, M. et al. (1980), Φωνητική Γραφή, Athens.
Grivelet, S. (ed.) (2001a), Digraphia: Writing Systems and Society (Berlin and New York) =
International Journal of the Sociology of Language, 150.
Grivelet, S. (2001b), ‘Introduction’, International Journal of the Sociology of Language, 150:
1–10.
Koutsogiannis, D. and Mitsikopoulou, B. (2003), ‘Greeklish and Greekness: trends and
discourses of “glocalness”’, Journal of Computer Mediated Communication, 9:1 (available
from http://jcmc.indiana.edu/vol9/issue1/kouts_mits.html).
Le Page, R.B. and Tabouret-Keller, A. (1985), Acts of Identity, Cambridge.
Li, W. (ed.) (2000), The Bilingualism Reader, London.
Marinis, T., Papangeli, A. and Tseliga, D. (2005), ‘“Potizo” or “Potizw”? The influence
of morphology in the processing of Roman-alphabeted Greek’, in Proceedings of the 17th
International Symposium on Theoretical and Applied Linguistics, Thessaloniki, April 2005,
Thessaloniki.
Maronitis, D. (2001), ‘Χάσαμε τη θεία στοπ’, To Vima, 21 January 2001 (available from
http://tovima.dolnet.gr/print_article.php?e=B&f=13172&m=B02&aa=1).
Moschonas, S.A. (2004), ‘Relativism in language ideology: Greece’s latest “language
issues”’, Journal of Modern Greek Studies, 22: 173–206.
Palfreyman, D. and al Khalil, M. (2003), ‘“A funky language for teenz to use”: representing
Gulf Arabic in instant messaging’, Journal of Computer Mediated Communication 9:1
(available from http://jcmc.indiana.edu/vol9/issue1/palfreyman.html).
Romaine, S. (1995), Bilingualism, 2nd edn, Oxford.
Sebba, M. (1998), ‘Phonology meets ideology: the meaning of orthographic practices in
British Creole’, Language Problems and Language Planning 22: 19–47.
Sebba, M. (2000), ‘Orthography and ideology: issues in Sranan spelling’, Linguistics 38:
925–48.
Sebba, M. (2003), ‘Spelling rebellion’, in J.K. Androutsopoulos and A. Georgakopoulou
(eds.), Discourse Constructions of Youth Identities, Amsterdam and Philadelphia, 151–72.
Sebba, M. (2007), Orthography and Society: The Culture and Politics of Spelling Around the
World, Cambridge.
Street, B.V. (1984), Literacy in Theory and Practice, Cambridge.
Greeklish 249
Tseliga, T. (2007), ‘“It’s all Greeklish to me! ”: linguistic and sociocultural perspectives of
Greeklish (= Roman-alphabeted Greek) in asynchronous CMC’, in B. Danet and S.C.
Herring (eds.), The Multilingual Internet, Oxford: 116–41.
Unseth, P. (2005), ‘Sociolinguistic parallels between choosing scripts and languages’,
Written Language and Literacy, 8: 19–42.
Wikipedia (2006), ‘Greeklish’ (accessed on 29 April 2006 from http://en.wikipedia.org/
wiki/Greeklish).
Zachos-Papazachariou, E. (1999), ‘Βαλκανική Βαβέλ’, in K. Tsitselikis (ed.), Γλώσσες,
αλφάβητα και εθνική ιδεολογία στην Ελλάδα και τα Βαλκάνια, Athens: 17–95.
Zavras, A. (n.d.), ‘Dealing with Greek characters in computers’ (accessed on 29 April 2006
from http://users.softlab.ece.ntua.gr/~sivann/xgrk/greek-zvr.html).
Zikmund, H. (1996), ‘Transliteration’, in H. Günther and O. Ludwig (eds.), Writing and
its Uses: An Interdisciplinary Handbook of International Research, vol. 2, Berlin and New
York: 1591–1604.
Weblinks
http://en.wikipedia.org/wiki/ (keywords: Greeklish; Romanization; ISO 8859–7;
Transliteration of Greek)
Part III
Ideologies and Contestations

12
A Tradition of Anomaly:
Towards the Regularization
of the Greek Language
Emmanuel Kriaras
Translated from the Greek by Korina Giaxoglou
It is widely held that Byzantium inherited a condition of diglossia from the Atticist
culture of the post-Hellenistic age, with the specifics of that diglossia determined
by successive writers. Each of them obeyed an archaizing imperative, albeit to a
different extent, according to the level of each writer’s classical scholarship. In the
centuries that were to follow, such archaizing would be practised by the majority
of ecclesiastical writers.
On the other hand, writers who were less well known and less competent in
Ancient Greek employed a kind of composite vernacular (‘folk’) idiom containing
Ancient Greek linguistic features, again in varying degrees, depending on the
individual competence of the writer. From late antiquity, the language these authors
used had gradually evolved, until it reached a more definite form by around 1100,
when we can detect a first version of the Modern Greek language and what one
might describe as a proto-modern Greek literature; the outcome, in due course, is
the splendid poetry and the earliest form of Modern Greek prose in the Cypriot
idiom, as well as comparable poetry, along with a distinctive form of drama, in the
Cretan.1
During the Greek Enlightenment (the period of roughly fifty years leading up
to the Greek War of Independence), the language issue figures more prominently
– but in its new ramifications must be understood against the background of the
earlier history of the language as a whole. As early as the sixteenth century, Nikolaos
Sofianos had already compiled the first grammar of Modern Greek, though this
1
There is now an almost complete dictionary for the Greek of this literature (up to π): my Lexicon
to Medieval Greek Folk Literature, 1100–1669 (abridged in two volumes, published by the Centre for
Greek Language, Thessaloniki, and edited by I. Kazazis, T. Karanastasis et al.).
253
254 emmanuel kriaras
was only published much later. During the period of Ottoman rule, two traditions
of written language started to take shape, differentiated by different degrees of
archaism. There was a less archaizing form employed mainly by ecclesiastical
authors and best represented by Agapios Landos, Frangiskos Skoufos and Elias
Meniatis. On the other hand, Alexandros Elladios and Evgenios Voulgaris, among
others, employed a more archaizing language.
Around the end of the eighteenth century, in the heyday of the Greek
Enlightenment, the language issue begins to arouse learned debate. Dimitrios
Katartzis, adhering to the views and the example set by Iosipos Misiodakas, favours
the use of the vernacular, in contrast to the archaist Lambros Photiadis and others.
Adamantios Korais proposes ‘a middle way’, but one which involved an unrealistic
solution. The writers Grigorios Konstantas, Daniel Philippidis and Athanassios
Psalidas, as well as the more radical Ioannis Vilaras, all follow a demoticist line.
Dionysios Solomos and Athanassios Christopoulos also use demotic in their poetry.
In the years to come, Anthimos Gazis, Konstantinos Koumas, Neofytos Vamvas
and Nikeforos Theotokis will be using language in the spirit of Korais.
In the newly established Greek state, Korais’ ‘middle way’ was officially adopted,
but within a few decades was displaced by an aggressive archaizing which provoked
a lively reaction in the form of an intellectual movement led by Psycharis, and
marked by the publication of his historic work (disturbing the ‘stagnant waters’),
My Voyage (1888). This proclamation on behalf of the people’s authentic language
was for a time influential on creative literature, which was, arguably, rejuvenated
by the efforts of various distinguished writers. These years, in fact, may be seen
as a period of momentum, which carries on for, roughly, the first quarter of the
twentieth century.
Beyond the literary field, however, it would need a continuing, intense struggle
on the part of the demoticist movement to achieve a victory for the demotic
in all aspects of national life, public and private – education and scholarship
included. The achievement of the goal was fraught with difficulties, from the
defective organization of the state to a widespread lack of cultural understanding.
An unstable and anomalous language situation was perpetuated, while at times
positive developments were impeded by doubts and uncertainties. The lack of
any consistency in the official promotion of authentic Modern Greek only served
to exacerbate these uncertainties. The state was, in fact, indifferent to the issue,
distracted as it was by other issues which needed immediate attention. Nonetheless,
there were several significant moments on the journey towards the regularization
of the language. These included, not least, the linguistic-educational reforms of
the years 1917–20, while the reactionary interventions of 1922 had no permanent
or definitive results. A renewed attempt at regularization towards the end of the
1930s was not taken any further – but by this time a troubled, and deteriorating,
political situation was bound to be a serious obstacle to any real progress.
The Second World War and its damaging political consequences for Greece
could hardly solve the problem, but at least served to intensify the struggle for
a tradition of anomaly 255
a definitive solution. In 1964, a new attempt to commit the state to linguistic

normalization, and bring an end to linguistic anarchy, had no substantial effect.
Under the Colonels’ dictatorship, seven long years (1967–74) saw a backward
course in every area of significant national activity, and a total impasse, in
consequence. Yet it was perhaps this impasse that eventually helped to provoke a
radical treatment of the language issue by the state, in favour of the demotic.
After the overthrow of the Colonels, the restitution of parliamentary institutions
and moves towards a more democratic regulation of political life were followed
by the introduction of demotic within the public sector and, in particular, the
beginnings of regularization in the sphere of language and education, as also in
scholarship and science. Even though the demotic had already achieved important
success in the literary field and, to a lesser extent, in the scholarly-scientific arena,
the decision taken by the Karamanlis government to institutionalize demotic in
1976 represented a watershed in our educational and social systems; it signalled
the start of a journey towards the regularization and unification of our language.
What was the language situation in Greece immediately after the recognition
of the demotic? Electronic and other information media used a language which
could be described as demotic, even though it contained many archaizing features,
anomalously situated within speech. It took an individual effort on the part of
each and every writer, as well as leadership from all qualified authorities, for the
linguistically uninformed to be helped to write demotic Greek in a more regular
way. Unfortunately, state support in this domain has been negligible: offered only
at the start, and devoid of the requisite continuity or persistence.
It is noteworthy that there was no significant reaction on the part of the
archaizers. There were some intellectuals who, either in theory or practice or both,
sought to react and did so. They very soon realized, though, that they were fighting
a lost cause, and gave up. Some of them had asked for changes to the solution put
forward by the state, which had been based on Manolis Triantafyllidis’ Grammar
(1917–41); and their suggestions produced a compromise between Triantafyllidis’
solution and their own. However, the need for reliable guidance on a more correct
and more consistent use of demotic was soon felt. Yet not many competent advisors
could be found, and the interventions of the uninformed were forever muddying
the waters. Worse, the state remained – as it still remains – indifferent to all
appeals: there is still no support for relevant research, nor is there any plan for
the establishment of specialized centres devoted to the monitoring and study of
specific language issues.
Occasionally, issues come to the fore that challenge the successes and solutions
achieved so far. One such was the attempted reintroduction of Ancient Greek
language-teaching throughout the first three years of secondary school, which
since 1976 had been restricted to the final three. Such a backward move could
only have confused young students, who as a rule leave primary school in a state of
unpreparedness even for a deeper knowledge of the modern language. Fortunately,
the threat did not fully materialize, and did little harm in the event.
256 emmanuel kriaras
In 1982, a more radical government instituted accentual reform, which had

exercised many scholars in the past. It is a widely-held belief, however, that the
orthographic problem awaits its resolution. The issue had initially been settled
in the form of a solution suggested by Triantafyllidis, which gradually became
accepted by almost all demoticists, then eventually by the state itself, along with
the demoticist programme as a whole. Nevertheless, the indisputable orthographic
difficulties of the Greek language have not only led the linguistically uninformed
into sometimes monstrous howlers, but have also persuaded others that in the
near future it will be necessary to consider the case for a further orthographic
simplification. Nowadays, in the orthographic representation of demotic by the
linguistically informed, the historical orthography suggested by Triantafyllidis has
been to a certain extent modernized. Even so, the need to readdress the question is
imperative.
Even if we do not want to accept (and justifiably so) all the demoticisms
recognised and suggested by Triantafyllidis, there are quite a few cases of
orthography where we have to follow him. I agree that it would be more acceptable
and consistent with other comparable cases if we were to write:
κτίριο (‘building’), ίσκιος (‘shadow’), λιώνω (‘melt’), τραβώ (‘pull’), τσιπούρα
(‘sea bream’), γλιτώνω (‘get away with’/‘save’), ροδάκινο (‘peach’), αγόρι (‘boy’),
ατόφιος (‘solid’), τσιρότο (‘plaster’), γρικώ (‘listen’/‘hear’), καθίκι (‘bog’ [i.e.
‘chamber-pot’], also ‘slimebag’), στιφάδο (‘stew’), γρέγος (‘northeast wind’),
κουλός (‘lame’, lit. ‘armless’), φτιάνω (‘make’), κουλούρι (‘bun’), καπέλο (‘hat’),
κανέλα (‘cinnamon’), κορδέλα (‘ribbon’)
instead of:
κτήριο, οίσκιος, λειώνω, τραυώ, τσυπούρα, γλυτώνω, ρωδάκινο, αγώρι, ατόφυος,
τσηρώτο, γροικώ, καθοίκι, στυφάδο, γραίγος, κουλλός, φτειάνω, κουλλούρι,
καπέλλο, καννέλα, κορδέλλα.
I believe that we should not be satisfied with the fact that most of the
linguistically informed, who are of course scholars in varying degrees, can do a
reasonably competent job of representing the new language orthographically. It
is also essential that they do not commit errors which would not be forgiven by a
wider educated public. The issue is serious and should be treated without haste or
a casual response. What is needed is a balanced proposal as a basis for informed
discussion. I would only note that historical orthography must be respected by any
proponents of further simplification in the future.
Triantafyllidis’ 1941 Grammar has exerted a significant influence on the
formation of our modern written language, in the domain of orthography (as also in
morphology). So too have – and will continue to – the recent dictionaries of Modern
Greek by Kriaras (1995), Babiniotis (1998/2002) and the Manolis Triantafyllidis
Institute (1998), despite the fact that their suggested solutions do not always or
entirely converge. Thanks to such efforts, we have achieved a more substantial
unification of Modern Greek in written and oral contexts. Let us hope that in the
a tradition of anomaly 257
future we will find a better resolution of the questions that remain open, including
the question of a more systematic and simpler treatment of the orthographic issue.
If we compare the orthographical preferences of the three dictionaries cited, we
find that, in almost every case, Kriaras and the Triantafyllidis Institute converge.
Babiniotis’ dictionary diverges, and its divergences are due primarily to the
conservatism of the lexicographer, as well as to his more personal demoticist views.
Here (as also in his, more recent, short dictionary of 2004), we often find that the
orthographic forms suggested are far too ‘etymological’ to be acceptable.
Let me illustrate the point with a small sample of instances (several involving
alternatives already cited):
Babiniotis Kriaras/Triantafyllidis Institute

αντιπροσωπία αντιπροσωπεία (‘delegation’)
βεβαρημένος βεβαρυμένος (‘burdened’)
γαρύφαλλο γαρίφαλο (‘carnation’)
γλύκυσμα γλύκισμα (‘pudding’)
γόμμα γόμα (‘rubber’/‘eraser’)
γρασσίδι γρασίδι (‘grass’)
καλοιακούδα καλιακούδα (‘crow’)
καννέλα κανέλα (‘cinnamon’)
κολλήγος κολίγος (‘serf’)
κολοιός κολιός (‘mackerel’)
κουλλός κουλός (‘crippled’)
μάννα μάνα (‘mother’)
νοννός νονός (‘godfather’)
παππάς παπάς (‘priest’)
πατρυιός πατριός (‘step-father’)
ρέβω ρεύω (‘wear out’/‘burp’)
στείβω στύβω (‘squeeze’)
στυφάδο στιφάδο (‘stew’)
τσηρώτο τσιρότο (‘plaster’)
As a postscript, I note that in his short dictionary Babiniotis (‘well-meaning’,

no doubt) abandons many scholastic orthographic forms which he had favoured
and promoted in his more extended work. He now acknowledges to a large extent
what the other two dictionaries have already accepted. One only wishes that he
had accepted the orthographic forms now proposed in the first place.
In conclusion, I would like to add that there are other orthographic simplifications
suggested by Triantafyllidis, but not yet widely accepted, and that it is desirable
that they too should become standard. One example is the spelling, ‘αντικρίζω’
(‘to face’); another is the simplified comparative ending -ότερος. I firmly believe
that in due course we shall be able to proceed to a further simplifying of the
orthographic system, as part of the necessary process of linguistic regularization.
13
Mothers and Daughters, Roots and Branches:

Modern Greek Perceptions of the Relationship
between the Ancient and Modern Languages
Peter Mackridge
Dedicated to the memory of Kostas Kazazis and Tasos Christidis
Modern Greeks have often had difficulty in defining and expressing their relationship
with the ancient Greeks. The study of modern Greek attitudes to the relationship
between their own language and that of the ancient Greeks is illuminating, because
these attitudes are indissolubly linked with their attitudes towards their more
general cultural (and at times even racial) relationship with the ancient Greeks.
This chapter ranges over the period 1750–2000, focusing particularly on the first
150 years.
Both before and after the nationalist dogma of the ‘continuity of Hellenism’ was
formulated in the nineteenth century, the way Greeks have talked about the Greek
language has reflected a difficulty in conceptualizing the processes by which the
passage of time constantly transforms human cultures, and in particular the way
that spoken languages alter over time. The rhetorical question humorously posed
by Jean Aitchison in the title of her book, Language Change: Progess or Decay?,1
would have been answered unhesitatingly by most educated Greeks since the
late eighteenth century as ‘decay’; among a few others in the nineteenth century,
Emmanouil Roidis and perhaps Ioulios Typaldos and Nikolaos Konemenos would
have answered ‘progress’.2 Both responses equally miss the point that language
change is simply inevitable.
1
Aitchison (1981).
2
See for instance Roidis’ stated preference for analytical rather than synthetic languages: Roidis
(1893). For Konemenos (1993) 81n. [1873], languages are perfected as they develop over time (in these
notes, dates in square brackets are the dates of the first publication of the relevant text). For Typaldos,
see p. 268 below.
259
260 peter mackridge
In the late eighteenth and early nineteenth centuries, the terms used to denote
the Modern Greek language were diverse and subject to change. Filippos Iliou
counted more than seventy different formulations used for the Modern Greek
language during the period 1801–20 alone.3 Changes in the labels for the Modern
Greek language went hand-in-hand with changes in the labels for the modern
Greeks themselves. As long as the modern Greeks called themselves Ρωμιοί (or
more formally Ρωμαίοι, ‘Romans’) and their spoken language ρωμαίικα, the
term Έλληνες (‘Hellenes’) could be reserved for ancient Greeks and ελληνικά
(‘Hellenic’) for the ancient language. But things were not as simple as that.
The phrase το ρωμαϊκόν γένος (‘Roman nation’) was commonly used in the
eighteenth century to mean ‘the Greek nation’. (This is, of course, nation in the
pre-nationalist sense, which distinguished the group primarily by religion and
secondarily by ethnicity, but with no sense of the nation-state: one may compare
the Ottoman term Rum-i millet and the term used in Italy at the time, nazione
greca.) However, terms such as ‘η καθ᾿ ημάς ελληνική Διάλεκτος’ (‘our Hellenic
dialect’) and ‘το ελληνικόν γένος’ (‘the Hellenic nation’) were also coming to be
used by Greeks to refer to their own contemporary language and group identity.4
(Διάλεκτος continued to be used to mean ‘language’ from at least the seventeenth
until the early nineteenth century.) Moreover, the terms Ρωμιοί and Έλληνες
were used side by side with a third term, Γραικοί, which is as ancient as the term
Έλληνες itself and which was favoured especially by Greeks who had lived among
Western Europeans and were accustomed to hearing themselves called Grecs, Greci,
Greeks, and so on.5 It was used especially by Western-orientated enlighteners such
as Adamandios Korais who wished to uncouple the modern Greeks from the
Byzantines;6 for Korais, the term ‘Romios’ connoted the subjection of the Greeks
to the contemptible ‘Greco-Roman emperors’ of Byzantium. It is notable that
the sense of ‘ancient’ or ‘modern’ does not figure explicitly in any of these terms,
though terms such as ‘η καθ᾿ ημάς Διάλεκτος’ or ‘η ημετέρα Διάλεκτος’ (‘our
dialect’) imply the language currently in use.
It is only comparatively recently that Greeks have called themselves
Νεοέλληνες (New or Modern Hellenes) and their language Νεοελληνικά
or Νεοελληνική (New or Modern Hellenic). The term Νεοελληνική for the
modern language was originally a literal translation from European terms such
as neugriechisch and neogreco. Even though Panagiotis Kodrikas used the term
3
Iliou (1997) lxv; see also the entry ‘γλώσσα νεοελληνική’ in the index, ibid. 658.
4
Among the plethora of formulations for the Modern Greek language to be found in the titles
of Greek books in the 1790s, I single out ‘εις την κοινοτέραν των νυν Ελλήνων Διάλεκτον’ (‘in
the more common Dialect of the Hellenes of today’ [1795]: Ladas and Hatzidimou (1970) 343), and
‘εις την καθ᾽ ημάς ελληνικήν Διάλεκτον’ (‘in our Hellenic Dialect’ [1799]: Ladas and Hatzidimou
(1973) 213). Half a century later, however, the eminently sensible E.A. Sophocles (1842) iv claimed
that the term Hellenes ‘is used chiefly by the inhabitants of Bavarian Greece, who perhaps do not
constitute more than one fourth of the Greek nation’.
5
Versions of the same word for ‘Greek’ are also used in the Balkan languages.
6
This term was objected to by Kodrikas (e.g. (1818) 225–6).
mothers and daughters, roots and branches 261
‘Nέοι Έλληνες’ (sometimes in contrast to ‘οι Παλαιοί’) in 1818,7 the terms

Νεοέλληνες and Νεοελληνική did not catch on for some time, and they have not
found universal acceptance even today. An exceptional early user of Νεοέλληνες
and Νεοελληνικός was Miltiadis (Panagiotis) Sofianopoulos, a follower of Korais,
who in an attack on the vernacularist Ioannis Vilaras in 1815 criticized his use of the
terms Ρωμαίοι and ρωμαίικα and insisted on using Νεοέλληνες, Νεοελληνικόν
γένος and νεοελληνική γλώσσα for the modern Greeks, their ‘nation’ and their
‘language’, instead.8 Later, the grammarian Georgios Chrysovergis used the term
Νεοελληνική γλώσσα to refer to the modern (archaized) written language, in
contrast to the spoken, which he called ‘demotic’: Δημοτική γλώσσα.9 The use of
the terms Αρχαία Ελληνικά and Αρχαία Ελληνική is also comparatively recent.
One of the earliest uses of the phrase ‘αρχαία Eλληνική’ (‘Ancient Greek’) that
I know of is by Korais, who contrasted it simply with the ‘σημερινή [γλώσσα]’
(‘today’s [language]’).10 But even Korais did not use this formulation systematically,
and, like Νεοελληνική, it was slow to catch on.
In the period immediately preceding the foundation of the modern Greek state,
educated Greeks seem to have been aware of three chief varieties of Greek, for which
I shall use our modern terms: first, Ancient Greek (including Classical, Hellenistic
koine and liturgical Greek); secondly, colloquial Modern Greek (the dialects and
such common spoken varieties as may have existed); and, thirdly, formal Modern
Greek (used especially by the Church hierarchy but also in a number of educational
works). In the eighteenth and early nineteenth centuries, the term ‘ύφος’ (which
in today’s Greek means ‘style’) was commonly used to refer to what we would
call a particular language variety or register, whether Ancient Greek, or modern
formal, or modern colloquial.11 When it came to defining Modern Greek, educated
Greeks disagreed as to which variety was the ‘true’ or ‘genuine’ modern language.
It is instructive to look at the title pages of the many Greek books that explicitly
indicate that they are written in what we call Modern Greek (the absence of such
7
Kodrikas (1818) 241, like Korais, also talks about ‘παλαιά και νέα γλώσσα’.
8
Moschonas (1981) 221–2. These uses are not recorded by Koumanoudis (1900), s.v.
ΝεοEλληνικός, who gives the earliest reference as ‘P. Petridis 1818’. In fact, Platon Petridis had
already used the term ‘Νεοελληνικόν ιδίωμα’ in the subtitle to his translation of Samuel Johnson’s
novel Rasselas (Corfu, 1817).
9
Chrysovergis (1839b) 9.
10
Korais (1984) 130 [1805]. Korais, who was living in Paris, was no doubt translating the phrase
‘grec ancien’.
11
Iosipos Moisiodax (1761, preface) defends his use of ‘το κοινόν ύφος’ (‘the common style’),
though at the same time he expresses himself in favour of correcting it according to ‘Hellenic’, which
is its mother. Evgenios Voulgaris (1766, preface), writing in a close approximation to Ancient Greek,
attacks the ‘vulgar philosophers’ who use ‘την κοινήν διάλεκτον’ or ‘το κοινόν ύφος’. Dimitrios
Katartzis wrote: ‘άφησα το φυσικόν ύφος και συγγράφω από τους ͵αψϰαʹ κατά το ύφος των
σπουδαίων’ (‘I abandoned the natural style and since 1791 I have been writing according to the style
of the learned’: (1970) 332). Grigorios Konstantas (1804, preface) distinguishes ύφος, by which he
seems to mean orthography and morphology, from λέξις and σύνταξις (quoted in Sathas (1870) 187).
Other commonly used terms were φράσις and ιδίωμα.
262 peter mackridge
an indication often implied that the book was written in archaic language). When
the Cretan monk Agapios Landos (1641) published his best-selling manual of
righteous living, Αμαρτωλών σωτηρία (The Salvation of Sinners), he specified on
the title page that it was written ‘εις κοινήν των Γραικών διάλεκτον’ (‘in the
common dialect of the Greeks’), which signalled to the reader that the language
of the book would be close to the spoken tongue. Unlike the more literary (that
is, fictional) works produced by Cretans during the last decades of Venetian rule,
Agapios’ book is not written in local dialect.12 This indication is typical of the use of
the adjective ‘κοινή’ in the seventeenth, eighteenth and early nineteenth centuries
to refer to a variety of Greek that is not only definitely modern, but is readable by
those with very little education, even though it uses some vocabulary and especially
features of orthography and morphology that were based on conventions dating
back at least to Hellenistic times.13
Another non-pejorative adjective that was used for everyday Modern Greek,
besides ‘κοινή’, was ‘απλή’ (‘simple’). Again, while this adjective was used to
distinguish Modern from Ancient Greek, it contains no explicit reference to
modernity. A patriotic poem published by a Greek in Cambridge in 1823 bears a
characteristic bilingual title: Κανάρης, ποίημα απλοελληνικόν. Canares, a Poem
in Modern Greek by Nicholas Maniakes. Here we see clearly that the adjective
‘απλοελληνικός’ (literally ‘simple-Hellenic’) was considered to be the equivalent
of ‘Modern Greek’; that is, when it came to the Greek language, ‘simple’ implied
‘modern’.14 Contrast the title used by Vamvas for his grammar in 1835; now that
the modern Greek state had been founded, he was able to use the formulation
‘σημερινή ελληνική γλώσσα’ (‘today’s Hellenic language’).15
In addition, there were pejorative terms for colloquial Modern Greek, the
most prevalent of which was ‘χυδαίος’ (‘vulgar’),16 while the term ‘χυδαϊσταί’
(‘vulgarists’) became the favoured term of opprobrium used by archaists and
extreme purists such as Neofytos Doukas at the beginning of the nineteenth
12
Cf. the formulations used by the printers of the Cretan literary texts Ερωτόκριτος and Ερωφίλη
respectively to characterize their linguistic form: ‘εις την Κρητικήν τως φυσικήν Γλώσσαν’ (‘in their
natural Cretan Language’: Kornaros (1980) 5 [1713]); ‘εις την φυσικήν της γλώσσαν την κρητικήν’
(‘in its natural Cretan language’: Chortatsis (2001) 86 [1676]); note that γλώσσα here means ‘dialect’.
Αμαρτωλών σωτηρία went through twenty-eight editions between 1641 and 1820; Ερωτόκριτος had
seventeen editions in the period 1713–1819.
13
I doubt that authors were thinking of what we call Hellenistic koine when they used this adjective:
they simply meant the language commonly used in their day. The term ‘διάλεκτος’ was often used as
a synonym for ‘language’, and did not really take on pejorative connotations – dialect as opposed to
language – until the late nineteenth century.
14
In such cases, the more analytical nature of Modern Greek, as opposed to the more synthetic
character of Ancient Greek, has been mistaken for simplicity. Even Hatzidakis (1901) 243 [1890] saw
Modern Greek as a simplification of Ancient Greek according to a process that began to take place in
Hellenistic times. Another example of the ‘simple-modern’ equivalence is the bilingual title Λεξικόν
απλορωμαϊκόν/Neugriechisches [. . .] Wörterbuch (Leipzig 1796: Ladas and Hatzidimou (1973) 46).
15
Γραμματική της αρχαίας και της σημερινής ελληνικής γλώσσης: Vamvas (1835).
16
E.g. Kodrikas (1818), who attacks the ‘γλώσσα των χυδαίων’.
century and Georgios Mistriotis at the beginning of the twentieth to refer to those
people I prefer to call vernacularists: that is, those who argued that a version of the
modern spoken language should be adopted as the standard written language.17
Greek writers of the late eighteenth and nineteenth centuries were often
confused about whether Ancient Greek and Modern Greek were one and the
same language or two different languages. At the end of the eighteenth century
the vernacularist Dimitrios Katartzis insisted that, although the modern Greeks
were the descendants of the ancients, Ancient Greek and Modern Greek were two
distinct languages because they had different grammatical systems;18 Romaic was
derived from Hellenic yet distinct from it, wrote Katartzis, just as Italian, French,
Spanish and Romanian were derived from Latin yet distinct from it.19 (I should
point out that at this time the idea that the modern Greeks were descendants of the
ancient Greeks was more an assumption than an assertion in the writings of Greek
intellectuals.) Korais too tended to talk about Ancient Greek and Modern Greek
as different languages, and at least once (in 1805) went so far as to refer to Ancient
Greek as a ‘foreign language’;20 but Korais’ views were much more complex than
this, as we shall see.
Those who were less favourably disposed towards the use of the spoken
language for written purposes tended to claim that Greek was a single language
from antiquity to the present day. In 1818 the conservative Kodrikas, for instance,
talks about the ‘ακεραιότης’ (‘integrity’: in effect, unity) of the Greek language.21
17
Koumanoudis (1900) lists twenty-nine nineteenth-century neologisms beginning with the stem
χυδαι-, of which Doukas was responsible for χυδαϊκός (also Kodrikas (1818) 225) and χυδαϊσταί,
and Kodrikas for χυδαιολόγος. Cf. Spyridon Trikoupis (1853) 9–10, who claims that in his book he is
following ‘την μέσην οδόν, την μεταξύ [. . .] των χυδαϊζόντων και των ελληνιζόντων’ (‘the middle
way, between that of the vulgarizers and the Hellenizers’). Trikoupis, ibid. 10–11 and 356, also claims
that the ‘κοινή γλώσσα’ is not the daughter of the ‘παλαιά Ελληνική’ (‘old Hellenic’) but ‘είναι η
ιδία η παλαιά’ (‘is the old language itself’); he also expresses his belief in the ‘Aeolodoric theory’,
for which see pp. 264–5 below. As late as 1886, Psycharis, writing in French, felt the need to remind
his readers that the colloquial language spoken by contemporary Greeks was not ‘langue vulgaire’ or
‘langue populaire’, but ‘langue moderne’ (Psichari (1886) xviii, 32).
18
Katartzis (1970) 318, 332.
19
Ibid. 217, 332.
20
Korais (1984) 129.
21
Kodrikas (1818) 234. Kodrikas, writing in France, distinguished between the ‘Δημοτική
Διάλεκτος’ (which he saw as the equivalent of the French term grec vulgaire), the local colloquial
language that differed from region to region, and the ‘Κοινή Διάλεκτος’ (grec moderne), the language
written (and spoken?) by the Phanariot princes and the prelates of the Orthodox Church: thus, e.g.,
(1818) 132. As Papazoglou (1991) 28–9 has argued, this first appearance of the term ‘Δημοτική’ to
mean colloquial Modern Greek was probably influenced by the recent discovery and (imminent)
decipherment of the Rosetta stone, on which the Egyptian text was inscribed in both hieroglyphic and
‘demotic’ writing systems (it was Herodotus, 2.36, who first applied the Greek terms ‘δημοτικά’ and
‘ἱρά’ (‘sacred’) to these two systems). Apparently following Christopoulos, Kodrikas states that Modern
Greek (but which variety of Modern Greek?) is a mixture of features from various Ancient Greek
dialects: thus, e.g., (1818) 241. Later, in 1860, Dimitrios Mavrofrydis wrote that ‘η νεοελληνική’ is not
the daughter of Ancient Greek as Romance languages are daughters of Latin, but ‘η νεωτάτη φάσις
του ελληνισμού’ (‘the most recent phase of Hellenism’): (1871) 30–1. Similarly, G. N. Hatzidakis calls
264 peter mackridge
Kodrikas’ worship of the Ancient Greek language was more than equal to Korais’:
he calls it ‘η ουράνιος Ελληνική Γλώσσα, το αρχέτυπον εκείνο σύνθεμα
μιάς υπερφυσικής επινοίας’ (‘the celestial Hellenic Language, that archetypal
composition of supernatural invention’).22 Kodrikas’ quarrel with Korais was
articulated in terms of the allegation that the latter was presuming to ‘correct’
Modern Greek instead of simply adopting the language that had been inherited
and moulded by generations of Greek princes and prelates. As a supporter of the
self-styled aristocracy of the Phanariots, Kodrikas saw the vernacularists as being
less of a threat than the mercantile bourgeoisie represented by Korais – hence
his contemptuous dismissal of what he called the ‘πακκάλικον ύφος’ (‘grocers’
style’); in Kodrikas’ eyes the former merchant Korais was presumptuously
attempting to wrest cultural and moral authority over the Greeks from the hands of
the Patriarchate and the Phanariots.
An indication of the difficulty that Greeks had in conceptualizing the linguistic
changes caused by the passage of the millennia is the frequently expressed theory
that contemporary colloquial Greek (Romaic) is a dialect of Hellenic in the same
way that Attic, Doric, Aeolic and Ionic are. This implies that modern spoken
Greek is somehow contemporaneous with what we nowadays call Ancient Greek.
It further implies that Hellenic has never died out, and that the modern spoken
language consists of a dialect that has existed since ancient times but didn’t happen
to be recorded as such in the extant writings of the ancient Greeks. This idea
became current among supporters of the vernacular at the turn of the eighteenth
and nineteenth centuries. Thus in the 1780s Dimitrios Katartzis regularly calls
Romaic ‘a sixth dialect of Hellenic’, the fifth being the koine of Hellenistic times.23
Similarly, in their innovative book Γεωγραφία νεωτερική (1791), Katartzis’
disciples Daniil Filippidis and Grigorios Konstantas describe the contemporary
spoken language as ‘the fifth dialect of Hellenic’.24
An extension of this attitude is the so-called ‘Aeolodoric theory’ propounded by
the jurist and poet Athanasios Christopoulos in his Grammar of the modern language
in 1805: it is significant that the book’s full title refers to a contemporary Greek
language that nevertheless is not distinct from the ancient, and that it labels the
modern Greeks as Hellenes.25 Using evidence of Ancient Greek dialects recorded
in ancient texts, Christopoulos argued that ‘the present language of the Hellenes’
consists of a fusion of the ancient Aeolic and Doric dialects, and that therefore it is
as ancient as the highly esteemed Attic. In fact, Christopoulos’ ancient linguistic
νέα Ελληνική ‘η νεωτέρα εξέλιξις της Ελληνικής γλώσσης’ (‘the modern development of the Greek
language’): (1977) 358 [1926].
22
Kodrikas (1818) 204.
23
Katartzis (1970) 17, 94; cf. 217. At the same time, Katartzis calls Hellenic the mother of Romaic:
ibid. 17. About AD 200, Clement of Alexandria had already described the Hellenistic koine as the
‘πέμπτη διάλεκτος’ (‘fifth dialect’): cf. Silk, this volume, p. 11.
24
Filippidis and Konstantas (1791) 133.
25
Γραμματική της Αιολοδoρικής [sic], ήτοι της ομιλουμένης τωρινής των Ελλήνων γλόσσας
[sic].
examples are chosen totally haphazardly and their similarity with modern forms is
purely coincidental, but this did not prevent his theory from enjoying considerable
popularity until it was definitively demolished by G.N. Hatzidakis in the late
nineteenth century.26 Christopoulos’ theory seems to me to be closely related to a
Greek attitude – still widespread today – that mistakenly identifies Ancient Greek
as a whole with the Attic dialect.27
For some of its proponents, the Aeolodoric theory had the advantage that they
could exploit the tremendous prestige enjoyed by Ancient Greek in order to defend
the use of colloquial Modern Greek for written purposes. In 1817, for instance,
Daniil Filippidis characterizes the ‘Aeolodoric Greek dialect’ of his translation
as being neither Romaic nor γραικικά, but Hellenic.28 The disadvantage of the
Aeolodoric theory was that its proponents were defending the use of Modern
Greek not as an autonomous synchronic system but as a medley of ‘survivals’ from
a more glorious ancient past. They also tended to feel that they should confine the
Modern Greek material in their own writing to features that in their view were
similar to forms attested in ancient texts. Thus Ioannis Venthylos’ grammar (1832)
allows certain Modern Greek forms because they are ‘Aeolic’ or ‘Doric’, while he
condemns other vernacular forms as ‘barbarous’ or ‘vulgar’.29
The Aeolodoric theory led some vernacularists such as Christopoulos and
Filippidis (at least, at those times when the erratic Filippidis was supporting the
vernacular!) to claim that Modern Greek is the ‘sister’ of the other dialects of
Greek and therefore of equal value.30 But the most prevalent of the metaphors that
were used for expressing the relationship between Ancient Greek and Modern
Greek in the late eighteenth and nineteenth centuries was that of mother and
daughter.31 The ‘mother–daughter’ metaphor was appropriate for a time (before
the ground-breaking work of G.N. Hatzidakis) when Ancient Greek and Modern
26
The Aeolodoric theory had already been rejected by J.M. Heilmaier, who in 1834 argued
correctly that Modern Greek is descended from Hellenistic and Byzantine Greek: see Kapsomenos
(1985) 11–12 and Delveroudi (1996) 226. Christopoulos himself implicitly abandoned the Aeolodoric
theory in Christopoulos (1853), which was probably written in the 1840s.
27
See e.g. the title of Vernardakis (1884), which implies an allegation that, in his attempt to ‘purify’
and archaize katharevousa, Konstantinos Kontos was replaying the Atticism of the Second Sophistic
(II AD).
28
Filippidis (1817), epilogue, 1, 6.
29
Venthylos (1832) 59 (Doric), 92 (Aeolic), 60, 89–90, 96 (‘barbarous’).
30
Christopoulos (1805): ‘[η αιολοδωρική γλώσσα] δεν είναι θυγατέρα της Αττικής, καθώς η
Ιταλική, Γαλλική, Ισπανική, και Βλαχική της Λατινικής, αλλ᾽ αδελφή με το πρωτότυπον κάλλος’
(‘[the Aeolodoric language] is not a daughter of Attic, as Italian, French, Spanish and Romanian are
daughters of Latin, but a sister, with the original beauty’), quoted in Sathas (1870) 207; Filippidis
(1817), epilogue, 11: ‘H αιολοδωρική μας γλώσσα, η παλαιότερη ίσως της Ευρώπης μετά την
σλαβικήν και την κιμμερικήν [sic], είναι αδελφή της αττικής’ (‘our Aeolodoric language, perhaps
the oldest in Europe after Slavic and Cimmerian [sic], is a sister of Attic’). Despite his parallel use of
‘Aeolodoric’ and archaic Greek in his writings, Filippipidis was consistent in his opposition to Korais
and his confederates. Athanasios Psalidas too claimed that both the spoken language of Hellenistic times
and ‘the common language of the nation’ in his own day were Aeolodoric: Moschonas (1981) 92, 122.
31
For the ‘sister’ and ‘mother-daughter’ metaphors, see Delveroudi (2004).
266 peter mackridge
Greek were being compared and contrasted as two distinct but related languages,
without reference to the medieval developments that in a sense both separated and
connected them.
Korais (1748–1833) had a greater sense of the passage of time than most of the
participants in the language controversy, and indeed he sometimes called Modern
Greek ‘η σημερινή των Ελλήνων γλώσσα’32 (‘the present-day language of the
Hellenes’: a phrase reminiscent of Christopoulos’ title) and ‘η νέα γλώσσα των
Ελλήνων’ (‘the modern language of the Hellenes’).33 Yet although he uses the
phrase grec moderne in a text he wrote in French in 1803,34 he seems not to have
used the equivalent term νέα Ελληνική in Greek. Korais claimed to be defending
what he most frequently called the ‘κοινή γλώσσα’, that is, the modern spoken
language, against the archaists, urging that it replace Ancient Greek as the language
used for teaching in schools. At the same time he urged that Greek education be
removed from the hands of the clergy and organized on a secular basis. He used a
variety of metaphors to convey the relationship between the ‘κοινή γλώσσα’ and
Ancient Greek. At times, contrasting Ancient Greek, the ‘παλαιά Ελληνική’,
with the ‘νέα’ (tout court), he called the latter the daughter of the former.35 In an
optimistic moment after the end of the Greek War of Independence he wrote: ‘Η
σήμερον λαλουμένη δεν είναι ούτε Βάρβαρος ούτ᾿ Ελληνική, αλλά νέα νέου
έθνους γλώσσα, θυγάτηρ και κληρονόμος παλαιάς πλουσιωτάτης γλώσσης
της Ελληνικής’ (‘The language spoken today is neither Barbarous nor Hellenic,
but the modern [or ‘new’ or ‘young’] language of a modern nation, daughter and
heir of an old and most rich language, namely Hellenic’; here he seems to be using
the term ‘Barbarous’ in its ancient sense of ‘non-Greek’).36
A less optimistic manifestation of the mother–daughter metaphor is to be found
in the subtitle of Korais’ famous open letter to Alexandros Vasileiou, published
many years earlier, in 1804: ‘Περί της κατά μικρόν βαρβαρωθείσης Ελληνικής,
και της εξ αυτής γεννηθείσης των Γραικών γλώσσης’ (‘Concerning the
gradually barbarized Hellenic and the language of the Greeks that was born of
32
Korais (1984) 130 [1805]. In his later life Korais even occasionally used the phrase ‘οι νέοι
Έλληνες’, as opposed to ‘οι παλαιοί’: (1832) vii.
33
Korais (1984) 198 [1805].
34
See Papazoglou (1991) 21.
35
E.g. Korais (1984) 329 [1809]. Similarly, Konstantas (1804): ‘The Hellenic language is the
mother of ours’ (quoted in Sathas (1870) 187).
36
Korais (1832) vii. ‘Barbarous’, ‘μιξοβάρβαρα’ (literally ‘mixed-barbarous’) and ‘graecobarbara’
were terms of opprobrium frequently attached by Greeks and non-Greeks to Medieval and Modern
Greek before Korais’ time. Earlier Korais had written that the rich mother of ‘our’ language is ready to
give it what it lacks (Korais (1984) 884 [1805]). Similarly, Chrysovergis (1839a) 6 wrote that Modern
Greek possesses infinite richness because it has a mother protectress who is always ready to fill any gaps.
Chrysovergis’ contrastive grammar is a genuine attempt to describe spoken Modern Greek in contrast
to Ancient Greek: this is what Korais had in mind when he urged that the teaching of Greek be based
on the ‘παράθεσις’ (‘juxtaposition’) of the ancient and modern languages. In this, Chrysovergis is more
successful than Vamvas (1835), whose ‘juxtaposition’ presents most of his Modern Greek examples in
an already ‘corrected’ (Koraized) form.
it’).37 This suggests that Modern Greek is a distinct language ‘born’ of Ancient
Greek at a time when the latter had been ‘barbarized’, meaning ‘de-Hellenized’:
Korais wished to ‘re-Hellenize’ it.38 Recalling that Korais had studied medicine,
we can perceive traces of a theory of genetics and upbringing in this expression
of his views: the mother (Ancient Greek) has been enslaved, and in the process
she has been deprived of much of her culture and her morality; she has then given
birth to a daughter who has inherited this debased culture and morality. Korais
goes on to say that, nevertheless, Modern Greek should not be despised, because
it contains many λείψανα (‘relics’) of the ancient language;39 note the potentially
sacral connotations of the word ‘relics’.
Underlying Korais’ theories is a belief in human entropy: if man is left to his
natural propensities he will fall into moral decline, while the path towards moral
perfection necessitates hard labour in the form of correction and cultivation. As
one might expect (again) from someone who had studied medicine, Korais also
likens the ‘correction’ of a barbarized language to curing a physical illness.40
It is interesting to observe how Korais’ ideas were later developed by purists
who no doubt believed they were improving on his legacy. In 1852, for instance,
Ioannis Zampelios wrote a ‘Διατριβή περί της Νεο-ελληνικής γλώσσης’
(‘Dissertation on the Neo-Hellenic language’), in which he presented the tripartite
nature of the Greek language in the following way: ‘The literary [φιλολογική]
language of modern Greece [meaning what later came to be commonly called
katharevousa], intermediate [. . .] between the ancient language and the vulgar
language [αγοραίας, literally ‘of the market place’], has on her right, as mother
and mistress [δέσποινα], the ancient, from whence it certainly expects more
assistance, and on her left, as handmaiden and servant [θεράπαινα], the vulgar
language, because of the venerable relics which the vulgar language has still
preserved from the ancient; hence she affords an opportunity to some to accept
her and provide her with Hellenic garments, and to others to dress her in penurious
rags.’ This ‘literary language’, Zampelios continues, is ‘a girl still in her infancy,
hitherto wrapped in swaddling-clothes’.41 Here we may note Zampelios’ use of
Korais’ idea of the valuable ‘relics’ that Modern Greek has retained from Ancient
37
Korais (1984) 1.
38
Similarly, Andreas Andreopolos (1820) 215, a supporter of Korais (note the ‘corrected’ version
of his surname!), refers to the ‘εξελληνισμός της γλώσσης μας’ (‘Hellenization of our language’)
(quoted in Daskalakis (1966) 146).
39
Korais (1984) 37.
40
Ibid. 489, 503 [1812].
41
Zampelios (1860) 502. Korais (1966) 234 had already used the δέσποινα/θεράπαινα metaphor
when talking about Erotokritos: ‘Ομολογώ ότι δεν είναι νόστιμος διατριβή ν᾿ αναγινώσκη τις τον
Ερωτόκριτον και άλλα τοιαύτα εξαμβλώματα της ταλαιπώρου Ελλάδος· αλλ᾿ όστις αγαπά την
ευειδεστάτην δέσποιναν, δεν πρέπει ν᾿ αμελή να κολακεύη και την δυσειδή θεράπαιναν, εάν
η προς την δέσποιναν είσοδος ευκολύνεται οπωσδήποτε δι᾿ αυτής’ (‘I confess it is not a pleasant
enterprise to read the Erotokritos and other such abortions of our poor Greece; yet whoever loves the
comely mistress must not omit to flatter the uncomely servant if entry to the mistress is in some way
facilitated by her’).
268 peter mackridge
Greek; yet, whereas Korais was claiming to ‘correct’ Modern Greek according to
Ancient Greek rules, Zampelios seems now to be adhering as closely as possible to
Ancient Greek while allowing some features of Modern Greek to contribute to the
mixture. By the time Zampelios’ son Spyridon published his Βυζαντιναί μελέται
only five years later, in 1857, it was possible for him to argue that Modern Greek
had uniquely defied the laws of historical linguistics, in that it was both modern
and ancient at the same time; if one excepted ten or twenty Latin words and a few
Frankish and Turkish ones, argued Zampelios fils, the living language was pure
Ancient Greek.42
By contrast, Korais’ views were being challenged in his own time by his
vernacularist opponents, who sometimes used his own ‘mother–daughter’
metaphor as a way of undermining his ideas. Korais’ ‘defence’ of Modern Greek
was turned on its head when, in 1812, Ioannis Vilaras wrote that Hellenic ‘is an
old woman, certainly beautiful in her youth, but it would be useless to try and doll
her up now’. By contrast, her daughter, Romaic, is ‘still in her swaddling-clothes’
(compare I. Zampelios a few decades later!) but ‘will eventually attain maturity
and beauty’. Romaic may borrow words from Hellenic, Vilaras continues, ‘just as
a daughter may borrow her mother’s finery’, but she must ‘adapt them to suit her
own appearance’ if necessary;43 by this, Vilaras means that Ancient Greek words
need to be adapted to the phonological and morphological rules of the modern
spoken language. Almost half a century later, in 1856, the Cephallonian poet Ioulios
Typaldos, having stated that ‘η δημοτική γλώσσα’ (‘the demotic language’) is the
daughter of the ‘παλαιά’ (Ancient Greek), argues that language change, called
‘διαφθορά’ (‘corruption’) by others such as Korais, is in fact ‘η ανανέωσις της
γλώσσας’ (‘the renewal of language’).44 That the word ‘ανανέωσις’ could be
interpreted not only as ‘renewal’ but as ‘rejuvenation’ is borne out by Typaldos
himself when he goes on to say that a beautiful daughter shouldn’t be made up
to look like her old mother. In 1870, rather differently, another writer from the
Ionian Islands, Antonios Fatseas, claimed that renouncing the spoken language
(καθομιλουμένη) in favour of archaic Greek as the language of poetry was like
rejecting a beloved mother in favour of a rich and noble old woman in the hope of
being adopted by her.45
Later, in 1902, the demoticist grammarian Menos Filintas presented yet another
variation on the mother–daughter theme. Romaic, he wrote, is the legitimate
daughter of the noble (αρχοντοπούλα) Ancient Greek, while katharevousa (which
he punningly called ‘καταρρέουσα’ or ‘collapsing’) is the illegitimate daughter
who is trying to usurp her rights. Nevertheless, in a more serious vein, Filintas
claimed that Ancient Greek and Modern Greek were the same language, while
he acknowledged that they were pronounced differently – something accepted by
42
S. Zampelios (1857) 646.
43
Moschonas (1981) 148–50.
44
Typaldos (1953).
45
Fatseas (1952) 125.
other demoticists such as Katartzis and Psycharis but not by most of the supporters
of katharevousa, who viewed any mention of differences in pronunciation between
the ancient and modern languages as a treacherous attempt to split the timeless
Greek language (and therefore the timeless Greek nation) in two. Filintas also
pointed out that so-called grammars of Modern Greek were still using terminology
that was properly applicable only to Ancient Greek, classifying beta, gamma and
delta as ‘άφωνα’ (‘voiceless’) even though they are now voiced, and αι and οι as
‘δίφθογγοι’ (‘diphthongs’) even though each of these combinations (digraphs) has
been pronounced as a single sound for almost two millennia.46
As Tasos Christidis has noted, Triantafyllidis pointed out the unsatisfactory
nature of the ‘mother–daughter’ metaphor in 1938: ‘The expression “daughter
of the ancient language” with reference to the modern language is an ill-chosen
(άστοχη) and deceptive metaphor. [The modern language] is the ancient language
itself, which, spoken uninterruptedly by the Greek nation for thousands of years,
from mouth to mouth and from father to child, changed through being spoken
[. . .] until it took on the present form of the mother [sic] tongue, in itself a starting-
point for further development’.47
In 1853 Korais’ linguistic compromise was attacked by Panagiotis Soutsos.
Soutsos was keenly aware that 1853 was the four-hundredth anniversary of the
Fall of Constantinople; it also marked the beginning of the Crimean War, in
which the Ottoman Empire fought against Orthodox Russia and was eventually
joined by Britain and France in what was perceived by Greeks to be an unholy
alliance. Soutsos claimed to be raising ‘the Hellenic language of our ancestors,
formerly called κοινή, from the tomb’48 – something that Korais had said was
as impossible as raising the dead themselves.49 Soutsos attacked what he called
Korais’ ‘Νεογραικική γλώσσα’ (the phrase suggests a Frenchified ‘langue néo-
grecque’), claiming that this ‘Νεογραικική γλώσσα’ is to Ancient Hellenic what
the English plaster cast of the Erechtheum Caryatid is to the original marble one.50
This is an example of a different kind of metaphor, this time drawn from sculpture
and architecture. Korais himself had already compared language to a building:
‘The prudent householder weeps as he views from afar the inevitable destruction
of his home; then, after the dilapidated edifice has completely collapsed and the
dust has scattered, he approaches and salvages (συναθροίζει) whatever material
he can from the ruins, so as to build another house. The ardently desired time for
46
Filintas (1902) 8–29.
47
Christidis (2001) 9; Triantafyllidis (1938) 56.
48
Soutsos (1853) 5. Similarly, Skarlatos Vyzantios (1862) xiii wrote that ‘το γλυκότερον ημών
όνειρον’ is ‘η εντελής νεκρανάστασις της πατρώας ημών γλώσσης’ (‘our sweetest dream is the
complete resurrection of our paternal language from the dead’).
49
Interestingly, Hatzidakis (1901) 226 [1890] criticized the vernacularist Psycharis for trying to
‘resurrect’ the ‘άκρατον δημοτικήν’ (‘unmixed demotic’).
50
Soutsos (1853) 26. The reference is to the statue that was taken to England by Lord Elgin
and sold to the British Museum and to the plaster cast erected in its place in the Erechtheum on the
Acropolis at Athens.
270 peter mackridge
reconstruction has at last arrived.’51 In another passage, after talking about the need
for scholars to tour Greece with the aim of ‘συνάθροισις των λέξεων’ (‘salvaging
words’), Korais writes: ‘How are we to redress the fall of the language if we do not
first excavate its ruins?’52
In the late nineteenth and early twentieth centuries, language as edifice became
the chief metaphor used by purists to defend the Greek linguistic status quo, but
with a significantly different spin to it from the one used by Korais. In 1909 G.N.
Hatzidakis (1848–1941), who several times attacked the vernacularists for using
metaphors,53 wrote that the Greek language could be compared to a complex ancient
mansion that for millennia has been kept standing with some alterations, while its
original plan was preserved.54 The Greeks were used to this situation, Hatzidakis
wrote elsewhere, and preferred to make minor repairs and adjustments rather than
raze the whole edifice to the ground and build a new one.55 (Significantly, this is
the opposite of Korais’ view.) What Hatzidakis means by this ‘edifice’ is of course
the conservative written language that in the second half of the nineteenth century
came to be commonly known as katharevousa; and he is thinking not only of
the variety of written language that the Greeks of 1821 had inherited from their
immediate past, but also of the immense restoration work that had been carried
out on it since the establishment of the Greek state. Hatzidakis saw the success of
all these efforts being undermined by the attempts of the vernacularists to replace
katharevousa by demotic as the written language of modern Greece.56 Similarly, the
archaist Kleon Rangavis, who proudly boasted that in his dramatic poem Θεοδώρα
(1884) he had written four hundred pages without a single instance of the modern
particles να, θα, or δεν (or a single ancient infinitive either),57 also claimed that
the Greeks of the nineteenth century had rebuilt and cultivated their language just
as they had built ‘καλλιμάρμαρα μέγαρα’ (‘palaces of fine marble’) from the
ashes and cultivated the fallow fields of their devastated country.58 Here we find
an almost explicit parallel between the linguistic neo-classicism of katharevousa
and the nineteenth-century fashion for neo-classical architecture in Greek towns.
51
Korais (1984) 36 [1804].
52
Korais (1984) 503 [1812]. I assume that Korais’ ‘ν᾿ ανορθώσωμεν’ here is the equivalent of
modern να επανορθώσωμεν.
53
E.g. Hatzidakis (1901) 297 [1890].
54
Hatzidakis (1909) 34.
55
Hatzidakis (1901) 247–8.
56
Despite his insistence on the unity of the Greek language (Hatzidakis (1905) 795), Hatzidakis
does distinguish between αρχαία, μέση/μεσαιωνική and νέα/νεωτέρα Ελληνική, and even on
occasion characterizes the δημώδης as the natural heir of Ancient Greek (Hatzidakis (1901) 299–300).
By contrast, in 1908 G. Mistriotis stated that there is no such thing as ‘νεοελληνική γραμματική’
(‘Modern Greek grammar’). ‘When we create a new grammar we create a new nation,’ he insisted,
whereas there was no ‘νεοελληνικόν έθνος’ (‘modern Greek nation’), only an indivisible Greek nation
(quoted in Triantafyllidis (1963) 209). Such people interpreted demotic translations from Ancient
Greek as implying that Ancient Greek and Modern Greek are two distinct languages.
57
Rangavis (1884) vii.
58
Quoted in Hatzidakis (1901) 255 [1890].
Likewise, in his argument in 1911 that the official language should be protected
by law, Professor Georgios Mistriotis accused the ‘vulgarists’ of behaving like
the ancient Vandals, trying to smash what he called ‘the linguistic Parthenon’.59
This kind of visual and spatial metaphor betrays a view of language as an object, a
monument or an heirloom, to be either preserved and admired – or, alternatively,
destroyed and abandoned. Either way, language is conceived of as primarily written
and fixed rather than spoken and mutable.
The conservation analogy is closely related to the treatment of ancient
monuments such as the temples on the Acropolis, which in the nineteenth
and twentieth centuries were restored through the removal of post-classical
accretions and the reincorporation of fallen and scattered members – the process
that is nowadays often called ‘αναπαλαίωση’ or ‘reveteration’ (the converse of
‘ανανέωση’ or ‘renovation’). In the view of Korais and other purists, the process of
linguistic conservation had broken down in the Byzantine and Ottoman periods,
and it was the duty of learned modern Greeks to restore their language by undoing
the centuries of decay.
A number of other metaphors have been used to illustrate the relationship
between Ancient Greek and Modern Greek. One of these is drawn from coinage.
The vernacularist Vilaras wrote in 1815 that he appreciated the Ancient Greek
language in the same way as he valued his collection of ancient Greek coins – but,
he continued, it wouldn’t be in his best interest to go to the market to try and buy
merchandise with them.60 Here Vilaras is expressing a view of language as primarily
a practical tool for communication.61 The same idea of language as currency was
used a century later by Hatzidakis, but this time in support of the Greek linguistic
status quo. The words of the written language, he argues, ‘are like coins universally
used and universally known as regards both their shape and their value’, whereas
the ‘vulgar’ word, he says, quoting the ancient scholar Dionysius of Halicarnassus
(late first century BC), ‘is neither familiar nor precise’.62 Here Hatzidakis suggests
that words and coins are given value, not by the current market, but by the
denomination that was stamped on them when they were originally coined – a
suggestion whose falsity is demonstrated by the modern market value of ancient
coins.
Another set of metaphors is more organic. We have seen how Kleon Rangavis
talked about the nineteenth-century Greeks cultivating their language as they
cultivated fallow fields. As early as 1804, in his letter to Alexandros Vasileiou,
Korais had exhorted his readers to improve the Modern Greek language: ‘Root
out from the language the weeds of vulgarity [. . .]; sow Hellenic seeds in it’63 – as
59
Mistriotis (1911) 12.
60
Moschonas (1981) 169.
61
A similar argument had already been used by Katartzis (1970) 334. Nevertheless, Vilaras argued
that some pedants deny his assertion that Modern Greek is ελληνική (Moschonas (1981) 146): i.e. he
accuses such archaists of denying the continuity of the Greek language.
62
63
Korais (1984) 51 [1804].
272 peter mackridge
though the language were a field, once fertile, that has been abandoned and left
uncultivated, with the result that the crops have died out and weeds have taken
their place. On the other side, the vernacularist Athanasios Psalidas complained in
1815 that the archaists (whom he called νεκρόμυαλοι or ‘dead-brains’) wanted to
graft dead Attic branches onto the living tree of the spoken language64 – a complaint
that applies more appropriately to Korais than to those who supported the use of
Ancient Greek as such.
In our own day it is the metaphor of roots that is most commonly used by Greeks
who argue that Ancient Greek and Modern Greek form part of a single ‘diachronic
Greek language’. The same metaphor was favoured by those who campaigned
vociferously against the government’s decision in 1976 to remove the study of
Ancient Greek from the first stage of secondary education (the Gymnasio, at which
attendance is compulsory), and confine it to the second stage (the Lykeio), and
then only for pupils specializing in the arts (the Klasiko or Geniko Lykeio) rather
than the sciences (the Techniko Lykeio).
The ideology of the ‘unitary Greek language’ and the ‘diachronicity of Greek’,65
expressed in 1982 by the short-lived Ellinikos Glossikos Omilos (Greek Language
Society), was a backlash against the demoticist dogma that had prevailed in Greek
government educational policy after the fall of the Colonels’ regime in 1974, itself
a reaction against the Colonels’ overuse and abuse of katharevousa. The ideology
became widespread in the period from the late 1970s onwards, gaining currency
particularly through the journalistic articles and broadcasts of Professor George
Babiniotis of Athens University (a leading member of the Ellinikos Glossikos
Omilos) and eventually becoming the basis for official educational policy regarding
the teaching of the Greek language. According to this ideology, it is impossible
to speak and write Modern Greek properly without having some familiarity with
Ancient Greek.
This argument, which challenges the boundaries between Ancient Greek,
katharevousa and demotic,66 and claims that all varieties of Greek, throughout
its history, constitute a single language, appeals especially to the large number of
Greeks who like to see themselves, not simply as the descendants of the Ancient
Greeks, but as their representatives and spokesmen in the modern world. Moreover,
the argument that Ancient Greek and Modern Greek constitute a single language
entails that the Greek language – and the Greeks who speak it – are almost unique
among the languages and peoples of the world. The argument runs like this: if the
Greek language is cut off from its roots, namely the vocabulary and grammar of
Ancient Greek, one cannot use the modern language properly; indeed, the Greek
language is incapable of flourishing and is in danger of withering and dying unless
64
65
The sophistic use of this derivative of a trendy modern linguistic term is characteristic: cf.
Rangousis (1999), who uses the terms ‘linguistic competence’, ‘communicative method’ and ‘structural-
functional approach’ to refer to the ‘diachronic teaching’ of the Greek language.
66
Frangoudaki (1992) 372.
it is constantly nourished through its roots. The metaphor of the tree implies that
the whole of the Greek language since the classical period exists in a timeless
present; thus the diachronic is made to appear synchronic.67 Proponents of this
theory argued that the quality of Modern Greek had gone into a swift decline
since the abolition of Ancient Greek lessons at the Gymnasio, with young people
showing an alarming ignorance of learned vocabulary and an inability to adhere
to the morphological and syntactical rules of Modern Greek. The President of the
Republic himself, Christos Sartzetakis, became obsessed with the theory of the
decline of the Greek language. Eventually, in 1987, the Minister of Education,
Antonis Tritsis, announced that compulsory lessons in Ancient Greek would be
reintroduced into the Gymnasio, arguing that Greeks cannot speak their language
properly unless they possess a good knowledge of Ancient Greek. The minister
went on to affirm that, because Greek is a single language from antiquity to the
present, Ancient Greek is not dead.68
For some years after 1993, the Greek language was taught in the Gymnasio
‘diachronically’, with the use of texts from the medieval and modern periods
written in learned varieties of the language, as well as texts from ancient times, both
classical and Christian. Instead of being taught Ancient Greek as a synchronic
system, pupils were provided with a smattering of Ancient Greek vocabulary and
grammar sufficient to enable them to read the relevant passages with a significant
degree of comprehension. The linguistic commentary on the passages provided no
sense of the historical development of the Greek language, nor were the passages
contextualized in their historical period; the only differences acknowledged were
those between ‘Ancient’ and ‘Modern’ Greek. The comments failed to point out,
for instance, in what ways a New Testament text differs linguistically from one
written in Classical Attic. The result was that pupils emerged believing they had
studied Ancient Greek, while in fact they were unable to distinguish the language
and historical background of Xenophon from those of Korais. According to one
of the co-ordinators of this policy, nevertheless, its ulterior purpose was to enable
pupils ‘to attain a more essential knowledge and, as a consequence, a more efficient
use of Modern Greek’.69
If I may be permitted to indulge in some metaphors of my own, many Greeks
feel they are privileged to possess the Greek language as a sacred heritage and that
they have a duty to keep their ancestral language alive, just as the Vestal Virgins
kept alight the sacred flame in the Temple of Vesta in ancient Rome. Or, to use a
slightly different metaphor, Greeks seem to see their language as a torch that has
been passed down through time from generation to generation, just as the Olympic
torch is passed in relay from hand to hand during the course of its journey round
the world. At the same time, the commonly held Greek view that their modern
67
The tree/roots metaphor has been convincingly deconstructed by Frangoudaki, ibid. 373.
68
Ibid. 374.
69
Rangousis (1999) 364–5. The chief textbook that was introduced in 1993 was Ypourgeio
(1993).
274 peter mackridge
language cannot stand on its own but must be supported on ancient crutches
– indeed, that their own language has value only as a function of its relation to
Ancient Greek – is symptomatic of a lack of national self-confidence.
I would like to finish by referring to my own favourite metaphor illustrating
the relationship between Ancient and Modern Greek. In 1886 Yannis Psycharis
prepared a paper to be delivered at a conference in Constantinople in which he
likened a language to a river.70 The river metaphor, I think, appropriately expresses
the essential fluidity of language. As Heraclitus said, you can never step into the
same river twice, in the sense that while the general course of the river remains the
same, the water itself is constantly changing, constantly being renewed. Seeing
the historical development of a language as being like a river perhaps helps us
to understand the falsity of the dilemma between viewing Ancient and Modern
Greek as either ‘one and the same language’ or ‘two distinct languages’.
Nevertheless, as John Joseph argues,71 ‘A language isn’t a thing, and it makes little
sense to imagine one English language evolving over many centuries, rather than
different English languages existing at different stages.’ The same can be argued for
Greek. Yet the belief that Greek is a single language has greatly influenced its actual
historical development and its actual current use.
References
Aitchison, J. (1981), Language Change: Progess or Decay?, London.
Andreopolos, I. (1820), ‘Ποιητική’, Ερμής ο Λόγιος, 8 (15 April): 213–28.
Chortatsis, G. (2001), Ερωφίλη, ed. S. Alexiou, Athens.
Christidis, A.-F. (2001), ‘Εισαγωγή: Ιστορίες της ελληνικής γλώσσας’, in A.-F. Christidis
(ed.), Ιστορία της ελληνικής γλώσσας από τις αρχές έως την ύστερη αρχαιότητα,
Christopoulos, A. (1805), Γραμματική της Αιολοδoρικής [sic], ήτοι της ομιλουμένης
τωρινής των Ελλήνων γλόσσας [sic], Vienna.
Christopoulos, A. (1853), Ελληνικά αρχαιολογήματα, Athens.
Chrysovergis, G. (1839a), Γραμματική της καθ᾽ ημάς ελλην. [sic] γλώσσης κατά
παράθεσιν προς την αρχαίαν, Athens.
Chrysovergis, G. (1839b), Της γραμματικής της καθ᾽ ημάς γλώσσης το πρακτικόν μέρος,
ήτοι η Γραμματική της νεοελληνικής γλώσσης, Athens.
Daskalakis, A.V. (1966), Κοραής και Κοδρικάς, Athens.
Delveroudi, R. (1996), ‘La question de la langue et les dialectes du grec moderne (1880-
1910)’, Revue des études néo-helléniques, 5: 221–39.
70
Psycharis (1902) 161–2. Athanasios Psalidas had already used the Heraclitean image of a river as
an analogy for the historical development of language in a letter of 1812: see Moschonas (1981) 86.
71
Joseph (2006) 7.
Delveroudi, R. (2004), ‘Soeur ou fille? Points de vue sur la relation entre le grec ancien et le
grec moderne à la fin du XVIIIème siècle’, in Les Langues classiques: la gestion d’un capital
culturel, Athens and Thessaloniki: 89–96.
Fatseas, A. (1952), ‘Περί της γλώσσης και παιδείας των νεωτέρων Ελλήνων’, in
Xωριάτικες γραφές, ed. G. Valetas, Athens: 123–43.
Filintas, M. (1902), Γραμματική της ρωμαίικης γλώσσας. Μέρος Α΄. Φωνολογία και
γραφή, Athens.
Filippidis, D. (1817), Επιτομή των Φιλιππικών του Πομπηίου Τρόγου νυν πρώτον εκ του
λατινικού εις την αιολοδωρικήν ελληνικήν διάλεκτον, Leipzig [translation of Historiae
Philippicae by Pompeius Trogus].
Filippidis, D. and Konstantas, G. (1791), Γεωγραφία νεωτερική, Vienna.
Fotiadis, F. (1902), To γλωσσικόν μας ζήτημα κι η εκπαιδευτική μας αναγέννησις,
Athens.
Frangoudaki, A. (1992), ‘Diglossia and the language situation in Greece’, Language in
Society, 21: 365–81.
Hatzidakis, G.N. (1901), Γλωσσολογικαί μελέται, vol. 1. Athens.
Hatzidakis, G.N. (1905), Το πρόβλημα της γραφομένης ελληνικής, Athens.
Hatzidakis, G.N. (1909), ‘Περί της επιστημονικής ανάγκης ενός Λεξικού της
Ελληνικής γλώσσης’, in Τεσσαρακονταετηρίς της καθηγεσίας Κ. Κόντου, Athens:
7–34.
Hatzidakis, G.N. (1977), ‘Διατί είμαι μεν δημοτικιστής αλλά δεν γράφω την
δημοτικήν’, in Γλωσσολογικαί έρευναι, vol. 2, Athens: 358–70.
Iliou, F. (1997), Ελληνική βιβλιογραφία του 19ου αιώνα. Τόμος πρώτος. 1801–1818,
Athens.
Joseph, J.E. (2006), Language and Politics, Edinburgh
Kapsomenos, S. (1985), Από την ιστορία της ελληνικής γλώσσας, Thessaloniki.
Katartzis, D. (1970), Τα ευρισκόμενα, ed. K.T. Dimaras, Athens.
Kodrikas, P. (1818), Μελέτη της κοινής ελληνικής διαλέκτου, vol. 1 [only], Paris.
Konemenos, N. (1993), Το ζήτημα της γλώσσας, ed. R. Papatsaroucha-Missiou, Athens.
Konstantas, G. (1804), Στοιχεία της λογικής, μεταφυσικής, και ηθικής, Venice [translation
of book by Francesco Soave].
Korais, A. (1832), Άτακτα, vol. 4, Paris.
Korais, A. (1966), Αλληλογραφία. Τόμος δεύτερος. 1799–1809, Athens.
Korais, A. (1984), Προλεγόμενα στους αρχαίους έλληνες συγγραφείς, vol. 1, Athens.
Kornaros, V. (1980), Ερωτόκριτος (ed. S. Alexiou), Athens.
Koumanoudis, S. (1900), Συναγωγή νέων λέξεων, Athens.
Ladas, G.G. and Hatzidimou, A.D. (1970), Ελληνική βιβλιογραφία των ετών 1791–1795,
Athens.
Ladas, G.G. and Hatzidimou, A.D. (1973), Ελληνική βιβλιογραφία των ετών 1796–1799,
Athens.
Landos, A. (1641), Αμαρτωλών σωτηρία, Venice.
Mavrofrydis, D. (1871), Δοκίμιον ιστορίας της ελληνικής γλώσσης, Smyrna.
Mistriotis, G. (1911), Περί εννόμου αμύνης της εθνικής γλώσσης, Athens.
Moisiodax, I. (1761), Ηθική φιλοσοφία, Venice [translation of book by Lodovico Antonio
Muratori].
Moschonas, E.I. (1981), Βηλαράς, Χριστόπουλος κ.α. Η δημοτικιστική αντίθεση στην
κοραϊκή ‘μέση οδό’, Athens.
276 peter mackridge
Papazoglou, Ch. (1991), ‘“Démotique”: Δημοτική (γλώσσα) et Δημοτικά (τραγούδια)’,

Μολυβδο-κονδυλο-πελεκητής, 3: 15–29.
Psichari, J. (1886), Essais de grammaire historique néo-grecque, vol. 1, Paris.
Psycharis, Y. (1902), Ρόδα και μήλα, vol. 1, Athens.
Rangavis, K. (1884), Θεοδώρα, Leipzig.
Rangousis, N. (1999), ‘Η διαχρονική διδασκαλία της ελληνικής γλώσσας στο Γυμνάσιο
σήμερα με νέα μέθοδο και νέα βιβλία’, Συνέδριο για την ελληνική γλώσσα 1976–
1996, Athens: 361–72.
Roidis, E. (1893), Τα είδωλα, Athens.
Sathas, K.N. (1870), Νεοελληνικής φιλολογίας παράρτημα. Ιστορία του ζητήματος της
νεοελληνικής γλώσσης, Athens.
Sophocles, E.A. (1842), A Romaic Grammar, Hartford, Conn.
Soutsos, P. (1853), Νέα Σχολή του γραφομένου λόγου, Athens.
Triantafyllidis, M. (1938), Νεοελληνική γραμματική. Ιστορική εισαγωγή, Athens.
Triantafyllidis, M. (1963), ‘Δημοτικισμος: ένα γράμμα στους δασκάλους μας’, in
Άπαντα, vol. 5, Thessaloniki: 168–239.
Trikoupis, S. (1853), Ίστορία της ελληνικής επαναστάσεως, vol. 1, London.
Typaldos, I. (1953), ‘H γλώσσα’, in Άπαντα, ed. D. Konomos, Athens: 324–9.
Vamvas, N. (1835), Γραμματική της αρχαίας και της σημερινής ελληνικής γλώσσης,
Ermoupolis.
Venthylos, I. (1832), Γραμματική της νεωτέρας ελληνικής γλώσσης, Athens.
Vernardakis, D.N. (1884), [anon.] Ψευδαττικισμού έλεγχος, ήτοι Κ.Σ. Κόντου Γλωσσικών
παρατηρήσεων αναφερομένων εις την νέαν ελληνικήν γλώσσαν ανασκευή, Trieste.
Voulgaris, E. (1766), Η Λογική, Leipzig.
Vyzantios, S. (1862), H Kωνσταντινούπολις, vol. 2, Athens.
[Ypourgeio] (1993), Υπουργείο Εθνικής Παιδείας και Θρησκευμάτων, Η ελληνική
γλώσσα μέσα από κείμενα αρχαία, βυζαντινά και λόγια. Α΄ Γυμνασίου, 2nd edn,
Athens.
Zampelios, I. (1860), Τραγωδίαι, vol. 2, Zakynthos.
Zampelios, S. (1857), Βυζαντιναί μελέται, Athens.
14
Constructing a Science of Language:

Linguistics and Politics
in Twentieth-Century Greece*
Effi Gazi
Constructing a discipline
‘Linguistic nationalism’ is a term frequently employed to describe forms of
nationalistic politics structured around issues of linguistic distinctiveness and
exclusiveness.1 Although widely used, the term is arbitrary in the sense that the
selective nature of nationalism makes it possible for a variety of factors constituting
national communities to come into play; in consequence, the defining features
of nationalism vary: they may be linguistic, religious, cultural, political or civic.
Certainly language has played a central role in forming and defining a variety
of national communities by securing communicative spaces and strengthening
communal ties. Today, it seems so natural that language and nationality go together,
that it is hard to recall how recently they fused. Frederick the Great spoke French,
which was a normal thing for a German prince to do, while in 1863 ten per cent of
French schoolchildren went to school speaking only patois.2 Language emerged as
an important national, as well as social, institution during the nineteenth century,
and only recently have historians and social scientists become increasingly aware
of ‘the power of language as well as the involvement of language with other forms
of power’.3
Linguistics too has played an important role. From the seventeenth century
onwards, the gradual decline of Latin in Western Europe, together with emerging
interest in the propriety and antiquity of vernacular languages, led to burgeoning
production of vernacular dictionaries. Colonial conquests in the eighteenth
* Research for this article was completed with the aid of a travel grant from the European Science
Foundation (ESF) Research Project, ‘Representations of the Past: The Writing of National Histories in
Europe’.
1 Hobsbawm (1992) 96.
2
Steinberg (1994) 198.
3
Burke (1994) 1.
277
278 effi gazi
century brought more vernacular languages to the attention of the European

public, transferring language from the private to the public domain and gradually
strengthening the perception that language holds a central position in human
interaction. During the nineteenth century, under German leadership, the
theory and practice of linguistics was oriented towards the study of language in
history, promoting research on the historical evolution of language rather than on
synchronic aspects of linguistic forms. The stress on history reflected a nineteenth-
century fascination with origins. The goal was the discovery of the oldest, and
therefore source, language; a related search for laws governing linguistic change
was of central importance.4 The rhetoric of science enabled linguistics to develop a
discourse that avowed the concepts of objectivity and ultimate truth.
In the Greek context, these developments can best be seen in the work and
activities of Georgios N. Hatzidakis (1848–1941), the leading figure of Greek
linguistics from the late nineteenth century up to the Second World War. Hatzidakis
was the first Professor of Linguistics at the University of Athens. He was appointed
to the newly founded Chair of Linguistics in 1885 (he was a Reader between the
years 1883 and 1885), after studying in Athens, Leipzig, Jena and Berlin. Before
Hatzidakis, linguistics had not been an autonomous discipline and the Greek
language was studied within the field of philology. Hatzidakis was trained as a
linguist in Germany by such important figures as Karl Brugmann (1849–1919) and
Berthold Delbrück (1842–1922). He acknowledged the latter as ‘σεβαστόν και
αγαπητόν καθηγητήν’ (‘respected and dear professor’).5
Throughout this period, linguistics continued to focus on the comparative history
of the Indo-European languages. Brugmann and Delbrück were leading figures of
a third generation of linguists who succeeded Rasmus Rask (1787–1832), Franz
Bopp (1791–1867), Jacob Grimm (1785–1863) and August Schleicher (1821–
68), the founding figures of Indo-European historical and comparative linguistics.6
Schleicher, in particular, is recognized as the empiricist scholar who established
the methodological foundation of historical-comparative linguistics, on which the
vibrant group of Neogrammarians (Junggrammatiker), including Brugmann and
Delbrück, later built.7 They were convinced that languages change over time,
that is diachronically, and they sought the binding laws governing the diachronic
changes that had taken place within a particular language or between languages.
Darwin’s theory of evolution had deeply influenced Schleicher’s views, whose
‘“Glossology” . . . was presumably intended to echo biology’.8 Such biological
naturalism was rejected by the Neogrammarians, who upheld the principle of the
historical development of language and claimed their approach as scientific by
arguing that determinate, all-encompassing laws constitute the basis of language
4
Crowley (1996) 9.
5
Hatzidakis (1888) 117–73. For Hatzidakis’ studies, see also his article «Περί των εν τη νεωτέρα
Ελλάδι γλωσσικών μελετών»: Hatzidakis (1959–63).
6
Petersen (1959), esp. 277–310.
7
Koerner (1989) 198–204.
8
Crowley (1996) 11.
constructing a science of language 279
development. They focused on modern, as well as ancient, languages, on phonetics

and on the rigorous explanation of linguistic change. In addition to the phonology
and morphology of Indo-European languages, comparative study of their syntax
was at the centre of Brugmann’s and Delbrück’s work.
The Neogrammarians’ perception of linguistics as a historical science and
their empiricist methodology shaped Hatzidakis’ views. In the introduction to
his translation of the work of the American linguist, William Dwight Whitney,
Hatzidakis emphasized that:
Η γλωσσική επιστήμη είναι επιστήμη ιστορική . . . Μόνον δ᾽ επιπόλαιος
εξέτασις εζήτησε να χαρακτηρίση αυτήν ως φυσικήν επιστήμη. Έργον των
γλωσσοδιφών είναι τη βοηθεία πάντων των προσιτών αυτοίς ιστορικών
μαρτυρίων, να εξερευνήσωσι και παραστήσωσι την πορείαν της αναπτύξεως
των επι μέρους γλωσσών . . . όταν θεωρώσι την γλώσσαν ως τι ανελισσόμενον
εν τη ιστορία και αδιαλείπτως μεταμορφούμενον.9
Linguistic science is an historical science . . . It is only a superficial examination

that has sought to define it as a natural science. The task of linguists is to explore and
present the process of the evolution of each language with the help of all the historical
evidence that is accessible . . . as they look upon language as something evolving in
history and continuously transforming.
Early in his academic career, Hatzidakis made it clear that he had three main
goals: to establish the study of language as an independent discipline within the
Greek academy; to promote the scientific study of language; and to develop research
on the Greek language on the basis of new scholarly principles. Accordingly,
he initiated research on the historical evolution of the Greek language that was
inspired by the search for origins while also emphasizing the details of linguistic
transformation. This research had three interlinked consequences.
In the first place, Hatzidakis successfully challenged interpretations that claimed
that the origins of Modern Greek lay in the ancient Aeolic and Doric dialects,
effectively criticizing works such as Dimitrios Mavrofrydis’ Δοκίμιον Ιστορίας
της Ελληνικής Γλώσσης (Essay on the History of the Greek Language, 1871). His
predecessors and contemporaries insisted on the so-called ριζικήν μέθοδον, or
‘root method’: that is, they searched for the roots and origins of linguistic elements,
on the assumption that these origins provided evidence of the relations between
ancient and modern linguistic forms. Hatzidakis, on the contrary, insisted that:
Η μέθοδος αύτη των ριζών ανάγκη να θεωρηθή ως εντελώς ανεπιτυχής, δια τε
άλλα . . . και διότι πολλαί κατά τους διαφόρους αιώνας συνέβησαν περί τε τους
φθόγγους και την κλίσιν και την σημασίαν και την σύνταξιν μεταβολαί, και
πολλαί λέξεις ούτω μετεβλήθησαν, ώστε αποβαίνει δυσχερεστάτη η διάγνωσις
αυτών.10
9
Hatzidakis (1898) 8. See also Hatzidakis (1906).
10
Hatzidakis (1905a) 111.
280 effi gazi
It is necessary to view the root method as completely unsuccessful, because (among

other reasons . . .) many changes occurred throughout the centuries in sounds,
conjugation, meaning, and syntax, and many words were in this way modified so that
their diagnosis becomes extremely difficult.
Hatzidakis was thus opposed to any partial study of linguistic phenomena,

claiming instead that language should be approached in its entirety as a coherent
system subsuming sounds, conjugations, syntax and semantics.11 In one of his major
works, Μεσαιωνικά και Νέα Ελληνικά (Medieval and Modern Greek), published
between the years 1905 and 1907, he discussed the development of the Greek
language in various historical periods and insisted that Modern Greek was based
on the Hellenistic koine.12
Secondly, in his effort to refute demoticist arguments, Hatzidakis turned to
the study of medieval Greek in order to criticize Ioannis Psycharis’ reading of
the contestation between Ancient and Modern Greek. According to Psycharis,
written forms of medieval Greek coincided with the language spoken at the time
and represented the coexistence of a variety of elements struggling for dominance.13
Hatzidakis opposed this interpretation by arguing that medieval written Greek was
based on a mixture of ancient and modern forms that was distinct from the spoken
language: an argument which suited his convictions about the distinctiveness of
spoken and written Greek. Some of Hatzidakis’ assumptions about the diachronic
unity of the Greek language are currently being questioned; nevertheless, his
argument against Psycharis’ interpretation is valid mainly because he refused to
concentrate exclusively on morphology, but also concerned himself with phonology
and syntax,14 and because he did not restrict his study to a narrow time-span but
provided evidence from an extended chronological perspective.15
Thirdly, Hatzidakis subscribed to the notions of unity and continuity prevailing
in Greek national ideology at the time, and attempted to legitimize national claims
by using scientific terminology to emphasize the need for a historical study of the
transformations of the Greek language.16 Such linking of scholarship and ideology
was common in nineteenth-century Greece as in other European countries.17
Hatzidakis’ positions and the new discipline of linguistics did not arise out of some
ideological or political vacuum. They were, on the contrary, developed within
a wider intellectual field that also encompassed history and folklore studies, as
practised by Spyridon Lambros (1851–1919) and Nikolaos Politis (1852–1921)
respectively. This intellectual field, with its complementary intellectual positions,
11
Ibid. 125.
12
Ibid. esp. 1–51. See also Hatzidakis (1888) 117-–23.
13
Psichari (1886) x–xiv, 163–88. For an in-depth analysis of this controversy between Psycharis
and Hatzidakis, see Mackridge (1998) 49–61.
14
Hatzidakis (1905a), 356–9, 360–405.
15
Ibid. 406–81.
16
On the issues of linguistic unity and continuity, cf. Christidis (2007) and, with special reference
to Hatzidakis, Mackridge (2004) 69–87.
17
Joseph (1985–6).
functioned as a network of relationships as well as a space for distributing power.18

All the participants had their own set of attitudes, beliefs and strategies – a ‘habitus’
as Pierre Bourdieu would define it – which informed their position and activities;
but these were not simply private phenomena. They were, instead, shaped and
reshaped as participants interacted with the field, with its orthodoxy, principles
and beliefs. Of course, this form of analysis runs the risk of underestimating the
originality of individual views and statements, and we need to keep in mind the
fact that intellectual fields are not static: they change. Their structures are stable
only in relation to external agencies or individual positions within them.
At all events, the primary nucleus of the intellectual field was the University
of Athens and its ‘national disciplines’ (as one might call them), such as history,
linguistics and folklore studies as practised by their distinguished representatives
(Lambros, Hatzidakis and Politis). The University of Athens also contributed
to the elaboration, institutionalization and promotion of perceptions of the past
and national identity. Its central role at the end of the nineteenth century and the
early twentieth involved not only the formation of academic disciplines, but their
reformulation within the Greek intellectual context. The outcomes were various:
a flourishing rhetoric of ‘science’ (science of history, science of language); a
resurgence of interest in the medieval and Byzantine world, as well as the dedicated
Hellenization of these periods; the reinforcement of scholarly research according
to empiricist principles; and the popularization of such knowledge in a society that
was undergoing a process of self-identification.19
In addition to their research and teaching at the university, Lambros, Hatzidakis
and Politis were engaged in a variety of public activities and served on committees
of public interest, such as the Επιτροπή περί των Διδακτικών Βιβλίων (Committee
on School Textbooks) and the Επιτροπή Τοπωνυμίων (Committee on Place Names).
The latter was established in 1909 by royal decree to study place names within
the country’s borders and to identify why these names were used. A law on the
legal foundations for municipalities and communities, passed on 14 February 1914,
enabled local councils to propose changes to Hellenize place names. Since some of
the proposed names were inappropriate, however, the Committee of Place Names
was to consider any proposed change and, after discussion with the local council,
to submit a recommendation to the Ministry of Internal Affairs.
This committee was formed as part of an official policy in favour of national
symbols. Since the Greek peninsula had been inhabited for centuries by populations
of different ethnic origins and had also been under different dominations, multi-
ethnic and multi-lingual spatial organizations were clearly expressed in the variety
of place names of Greek, Albanian and Turkish origin. The Hellenizing of place
names was an attempt to reconstruct a national code of reference, which could
then be one of the basic elements of the symbolic construction of the community.
That most of the names the committee proposed or accepted were Ancient
18
Bourdieu (1985), (1993).
19
Gazi (2000) 103–18.
282 effi gazi
Greek was due to the feeling, even at the beginning of the twentieth century, that
classical antiquity was the most celebrated period of national history and the most
appropriate evidence of ethnic purity and continuity.
Lambros, Hatzidakis and Politis were also distinguished members of the
Ιστορική και Εθνολογική Εταιρεία της Ελλάδος (Historical and Ethnological
Society of Greece) which was founded in 1882 in order to collect
πάντων των μνημείων του ελληνικού βίου, κατά τους χρόνους της εκλείψεως
αυτού από του ελευθέρου κόσμου και καθ᾽ ας ημέρας ετέλει το ελληνικόν
Γένος δούλον ταύτη ή εκείνη τη ξένη κατακτήσει.20
all memorials of Greek life during the time of its eclipse from the free world and the
days when the Greek race was a slave to this or that foreign occupation.
In the nineteenth and twentieth centuries, the Historical and Ethnological Society
was one of the most important institutions in the promotion of national history and
identity. The active participation of established academics in this and other societies
– notably the Ελληνική Λαογραφική Εταιρεία (Hellenic Folklore Society), founded
by Politis in 1909, and the Γλωσσική Εταιρεία (Linguistic Society), founded by
Hatzidakis in 1902 – extended their role beyond the strict institutional framework
of the university. This is apparent from the constant presence of these academics
in the press. They sought to popularize a specific kind of knowledge that was
primarily related to the national past. Academics acquired authority as figures with
special access to the past reality of Greece, but they had to legitimize that authority
through interaction with the public. Addressing a wider public was not simply a
matter of passing on information on cultural issues; that is, the interaction between
professional academics and the public was not one-sided. The academics sought
public authority. They claimed a special connection between themselves and the
public, themselves and the realities of the past. Public endorsement established
their ‘scientific’ argument, making it an irreversible reality.
One of the major complications with Hatzidakis’ linguistic positioning was that
while his youthful claims about Modern Greek and the Hellenistic koine complied
with the ideal of continuity in Greek national history, they implicitly challenged
similar purist arguments undermining spoken Greek and demotic. Purists wanted
to prioritize the study of Ancient Greek. At the time, the Athens School of
Philosophy was going through a deeply purist phase which overemphasized the
importance of the ancient language, but the newly appointed young linguist
insisted, nevertheless, that Modern Greek was a scientifically legitimate field of
research. As early as 1881, he had argued that:
Άτοπος είναι ο λόγος ο πολλάκις προφερόμενος ότι τούτο ή εκείνο το
φαινόμενον είναι του λαού, είναι χυδαίον, επλάσθη υπ᾽ αμαθών, επομένως δεν
έχει λόγον . . . Δύναμαι να διαβεβαιώσω τοίς ταύτα προφέροντας, ότι πολλώ
μείζων κανονικότης και τάξις παρατηρείται υπό των επιμελών εξεταστών εν
τη γλώσση εκείνου του χωρικού όστις ουδέποτε ανέγνωσε βιβλίον ουδ᾽ εξήλθε
20
Δελτίον (1883), 1: 7.
εκ της κώμης αυτού, ή εν τη γλώσση τη εν Αθήναις υφ᾽ ημών λαλουμένη . . .

η γλώσσα του λαού δεν είναι άλογος και ακανόνιστος . . . η εξήγησις αυτής
πρέπει να ζητηθή.21
The argument that this or that phenomenon is of the people, that it is vulgar, that
it was created by the uneducated and there is no rationale to it, is mistaken . . . I
can assure those who develop these arguments that diligent researchers observe more
regularity and order in the language of a peasant who has never read a book or left his
village than in the language we speak in Athens . . . the language of the people is not
without sense or order . . . its elucidation is called for.
Even in his response to Karl Krumbacher, Hatzidakis insisted that Modern

Greek needed to be studied. He referred to various publications from the beginning
of his career (primarily in the years 1883–4) in which he had insisted on the need
to study Modern Greek:
Αι φράσεις βάρβαρον, διεφθαρμένον, ελεεινόν ιδίωμα ή διάλεκτος ή γλώσσα
λεγόμεναι επί της δημώδους, της απλής λαλουμένης ημών γλώσσης είναι και
ασεβείς και ημαρτημέναι. Είναι ασέβεια και άκρα καταισχύνη να ονομάζωμεν
το από δεκαπέντε και πλέον αιώνων πνευματικόν όργανον του έθνους ημών
βάρβαρον και διεφθαρμένον.22
Expressions like ‘barbarous’, ‘corrupt’, ‘wretched’ idiom or dialect or language,

with reference to our demotic, simple, spoken language, are both disrespectful and
erroneous. Calling the intellectual instrument of the nation, over more than fifteen
hundred years, barbarous and corrupt is disrespectful and deeply shameful.
Hatzidakis’ insistence that modern, spoken Greek needed to be studied was

based on his belief in the unity of language over time and his conviction that all
such study would eventually serve the national interest. His early claims – that
spoken Greek was not a ‘barbaric idiom’, that it was rather one of the phases in the
evolution of the long history of the language, that, accordingly, it fully deserved
scientific research and study – were especially unwelcome within the purist circles
of the Athens School of Philosophy. There, ardent advocates of linguistic archaism
like Konstantinos Kontos (1834–1909), Hatzidakis’ former teacher, and Georgios
Mistriotis (1840–1916) set the tone. Mistriotis did not hesitate to publicize his
disapproval in order to safeguard the unity of the purist cause within academia:
Μη εγείρης, φίλε κ. Χατζηδάκη, το ζήτημα της γραμματικής της νεώτερας
γλώσσης, διότι δημιουργείς άλλον λαόν παρά τον Ελληνικόν, εισάγης την
σύγχυσιν εις την όλην του Έθνους παίδευσιν, διακόπτεις το εθνοποιόν του
Κοραή έργον και παρέχεις όπλα τοις χυδαϊσταίς . . . Η γλωσσολογία παρέχει
σοι ευρύτατον στάδιον ερευνών, δι᾽ ών και υμάς αυτούς θέλετε τιμήσει και το
έθνος περιφανώς ωφελήσει . . . μη, προς Θεού, παράσχητε το εφ᾽ υμίν έρανον
21
Hatzidakis (1905a) 3.
22
Hatzidakis (1905b) 502–3.
284 effi gazi
προς το απαίσιον έργον, αλλά συνενώσατε τας υμετέρας δυνάμεις προς

σωτηρίαν λαού.23
My dear Mr Hatzidakis, do not raise the question of the grammar of the modern
language, because you are constructing a people different from the Hellenes, you
are bringing confusion to the nation’s education, you are interrupting Korais’ work
of nation building and you are providing weapons to the supporters of the vulgar
language . . . Linguistics offers you a wide field of research through which you can
bring honour to yourself and evident benefit to the nation . . . do not, for God’s sake,
provide assistance to a sinister enterprise, but join your forces to the salvation of the
people.
How would the linguist meet this challenge? Hatzidakis’ attempts to balance his
position defined his academic and public activities from the late nineteenth century
onwards. Similar criticism must have been voiced against him earlier, as is shown
by his constant struggle to reconcile his linguistic research with the demands of the
academic, intellectual and social elites that controlled the politics of language in
Greece. His individual case provides the opportunity for a deeper understanding
of the workings of linguistic politics – and their class and gender implications – at
the time.
Politicizing a discipline
In his attempt to reconcile his positions with the hegemonic politics of katharevousa,
within and beyond Greek academia and its strict hierarchies, Hatzidakis employed
new lines of argument that in certain cases became a dialectic of contradiction or
ambivalence. He gradually developed the position that the scholarly study of the
spoken language did not necessarily constitute an argument for its usage, while
also insisting that oral and written linguistic forms were not only different but also
socially distinct. His first move took place in 1884, when Dimitrios Bernardakis
(1834–1907) published his work Ψευδαττικισμού Έλεγχος (An Examination
of Pseudo-Atticism), in which he criticized Konstantinos Kontos’ linguistic
manifesto of 1882, Γλωσσικαί Παρατηρήσεις (Linguistic Observations), as well as
the fascination with the use of extreme archaisms in written Greek. As a former
student of Kontos, Hatzidakis undertook the mission to defend his old professor
and counter-attacked Bernardakis in Μελέτη επί της Νέας Ελληνικής (A Study
of Modern Greek). The linguist’s major re-orientation began at the beginning of the
twentieth century and did not confine itself to personal links and affiliations; on
the contrary, it was implicated in new ideological and political processes within
Greek society.
23
Mistriotis (1909) 148. Mistriotis also opposed Hatzidakis’ project for a ‘Historical Dictionary
of the Greek Language’. In 1908, the year when the project was launched (though it only reached
fruition in the 1930s), Mistriotis insisted bitterly that this linguist was promoting the vulgarization of
the language: ibid. 169–93.
Shortly after the Ευαγγελιακά and Ορεστειακά riots (1901 and 1903),24
the language debate exploded. The debate encompassed opposing ideological
and political agendas and distinctive socio-political issues. The already complex
relation between language and politics was re-directed towards conflicts between
social groups that engendered opposing linguistic conformations. The necessity of
a language programme for any social and political group aspiring to cultural and
political hegemony became apparent. Facing new challenges, Hatzidakis turned to
new arguments, while his dialectics of contradiction acquired new features. He had
earlier argued that spoken Greek was not a ‘barbaric idiom’ and that even peasant
language was of scholarly interest. Although he had opted for a distinctive written
Greek and even the written usage of katharevousa, he had not disregarded modern
spoken Greek. This complicated position was not to last.
It has been argued that the ‘relationships between written (scriptocentric)
and spoken (phonocentric) modes of communication are themselves of historic
interest’.25 Many themes can be illuminated by analysing particular communities of
discourse with regard to the relationship between the spoken and the written word.
Of course, historical investigation itself is primarily based upon written documents,
which makes any history of oral sociolinguistic conventions in the past a difficult
enterprise.26 Written records such as Hatzidakis’ various publications, however,
provide information about the rules of communication in the Greek socio-cultural
context. In the course of the twentieth century, class and gender perceptions of the
spoken word that evolved around new political agendas emerged in Hatzidakis’
epistemology. Within these agendas, Hatzidakis politicized his science of language,
localized it, formalized its concepts and strategies, and placed it at the heart of the
hegemonic political positions of the time. In short, he placed his linguistics in a
‘discursive regularity . . . in a whole field of discursive practices’,27 exploration of
which uncovers its ideological and political implications.
In 1907, Hatzidakis attacked Yannis Psycharis for employing linguistic forms
from a lower social milieu, arguing that,
αντί της γλώσσης ην ημείς ομιλούμεν εν ταις αρίσταις συναναστροφαίς
. . . αντί της γλώσσης των πόλεων ή των αιθουσών ημών . . . επειδή ζων εν
τη ξένη δεν εγίνωσκε ταύτην επαρκώς, ουδ᾽ ότι αύτη απέρρευσε κατά μέγα
μέρος εκ της γραπτής παραδόσεως, μετεχειρίσθη το ιδίωμα του όχλου
της Κωνσταντινουπόλεως και της Χίου, όπερ εκ της συνομιλίας μετά των
υπηρετριών αυτού είχε μάθει.28
instead of the language we speak in our noblest interactions . . . instead of the language
of urban centres or salons . . . which he did not know sufficiently, unaware, because he
24
The riots were associated with sensitive cases of modern translation (of the Gospels and
Aeschylus’ Oresteia, respectively).
25
Corfield (1991) 15.
26
Burke (1994) 9–10.
27
Foucault (1992) 185.
28
286 effi gazi
was living abroad, that it had largely derived from the written tradition, he used the
idiom of the mob of Constantinople and Chios which he had learnt from his maids.
This was one of the first times that Hatzidakis referred to ‘proper’ spoken Greek by
relating it to bourgeois educated elites; in subsequent years, he would constantly
repeat the argument about the social inferiority of particular forms of spoken Greek
by stressing the necessity for a γλώσσα των αιθουσών (‘language of the salon’).
Not all spoken Greek deserved particular attention and protection in Hatzidakis’
new interpretative schema, but only this ‘salon’ language and the γλώσσα των
κυριών (‘ladies’ language’). The linguist insisted that ‘η συναναστροφή μετά
της καλής κοινωνίας εν ταις επισημοτάταις οικίαις θα ηδύνατο να διδάξη
έκαστον άλλα τε πολλά και την καλήν, εθνικήν γλώσσαν’ (‘interaction with
high society in the most formal houses could teach anyone a variety of things, but
especially our fine, national language’).29
Moreover, Hatzidakis gradually adjusted his distinction between written and
spoken Greek by developing a new argument, according to which written Greek
was not distinct from spoken forms of the language, but was actually the medium of
oral communication spoken by bourgeois ladies that was subsequently transformed
into written discourse:
Ημείς βάσιν του γραπτού λόγου ημών εθήκαμεν την υφ᾽ ημών εν ταις
αιθούσαις ημών, την υπό των γυναικών ημών λαλουμένην προς αλλήλας και
προς ημάς γλώσσαν.30
The basis of written discourse for us is the language spoken by us in the salons, the one
our ladies use to talk to one another and to us.
Hatzidakis refashioned his previous arguments in two ways. First, while earlier
he had clearly endorsed the validity of all oral and written forms of the Greek
language and the need for all of them to be studied, he now made explicit value
judgments about different forms of spoken Greek according to the social standing
of their speakers. Secondly, he endeavoured to counter demoticist arguments that
spoken Greek was a mother tongue, was literally taught by mothers to children
(μητρική γλώσσα, μητρικό ιδίωμα, μητροδίδακτος), and on these grounds
was ‘natural’.31 In response, he took up the case of female speakers of the Greek
language himself. In 1888, Psycharis had pointed out that women could be of great
help in addressing the language question, insofar as it was mothers who first taught
the language to children; conversely, he complained that educated Greek women
competed with each other in vanity and tried to speak as archaically as possible.32
In 1890, Hatzidakis attempted to present a different reality:
29
Ibid. 20–1.
30
Hatzidakis (n.d., a) 1.
31
For a general overview and synthesis of demoticist arguments with reference to the ‘mother
tongue’, see Triantafyllidis (1993) 3–169.
32
Psycharis (1971) 169–70.
Και είναι όλως ανακριβές το του κ. Ψυχάρι, ότι πολλαί οικογένειαι καταβοώσι
κατά της χρήσεως της γραφομένης . . . τουλάχιστον εγώ ουδέποτε ήκουσα
τοιούτο τι. Αν δ᾽ επιτρέπηται να εικάση τις εκ της γλώσσης της Εφημερίδος των
Κυριών ή της Πηνελόπης, των μόνων υπό κυριών και χάριν κυριών εκδιδομένων
εφημερίδων, πείθεται τάχιστα ότι αι κυρίαι δεν διαφωνούσι προς ημάς περί
τούτου.33
Mr Psycharis’ claim that many families oppose the use of written language is
completely inaccurate . . . I have never heard anything like that. If one may draw
conclusions from the language used in the Ladies’ Gazette or Penelope, the only two
newspapers written by ladies and addressed to ladies, one is convinced that the ladies
do not disagree with us in that.
For Hatzidakis, female speakers were socially distinct. Language did not belong to
women in general; it belonged to ‘ladies’, that is women who came from a particular
social milieu. The ‘ladies’ language’ was the proper form of spoken Greek which
could also be used as the basis for written Greek. In this spirit, the demoticist
argument that related spoken Greek to women and hence to ‘nature’ was trumped
by a thesis relating katharevousa to culture and social status, again via women.
Hatzidakis’ line of argument is an important point of departure for analysing
how both the language controversy and, to a certain extent, linguistics itself became
gendered during the first half of the twentieth century in Greece. Scholarship
on gender has extensively explored the ways in which the medico-legal and
intellectual alliances in nineteenth-century Europe established elaborate regulatory
schemes of gender signification according to which ‘masculinity’ and ‘femininity’
were attributed specific, often opposing characteristics. Masculinity’s cultural
associations with the mind and culture itself, and femininity’s cultural associations
with the body and nature, were articulated within this signifying economy.34 The
early demoticist argument about the naturalness of demotic as spoken by women
and mothers relates directly to this schema, and so does Hatzidakis’ attempted
counter-argument. By relating proper language to educated and civilized middle-
class womanhood, he was in effect playing a part in re-structuring the discursive
order of gender and laying claim to a new economy of signification.
At approximately the same time, soon after the Ευαγγελιακά, and along with
a group of journalists, university professors, and military officers, Hatzidakis co-
founded the Εταιρεία των Πατρίων (Society for the Defence of Ancestral Values).
Prince Nikolaos was the honorary president of the society, while Major-General
Ioannis Papadiamantopoulos (1841–1909), commander of the royal military
branch, was its president. From 1902 onwards, the society published and
distributed without charge the weekly newspaper Πάτρια, which became the main
opponent of the demoticist journal Νουμάς. It addressed language issues primarily
and argued for the necessity of katharevousa. Although all articles were published
anonymously, many of them must have been written by Hatzidakis, as a prominent
33
34
Scott (1996); Butler (1990).
288 effi gazi
member of the editorial board, which also included Konstantinos Mitsopoulos

(1844–1911), Stavros Voutiras (1841–1923), Ioannis Damvergis (1862–1938)
and Ignatios Moschakis (1847–1903): his distinctive style is often apparent.35 For
more than a decade, the newspaper attempted to stigmatize demoticist politics
by association with so-called ‘materialistic ideas’ (from Darwinism to secularism,
socialism and feminism), and constantly warned about the threats to traditional
values and the social order that these new ideas presented. Πάτρια, which at least
initially had the support of the royal family, moralized the language question by
relating the politics of katharevousa to the preservation of traditional values and
principles, such as religion and family. The first issue of the paper declared:
η θρησκεία και η γλώσσα υπήρξαν πάντοτε οι φύλακες άγγελοι του Εθνους
και του Γένους ημών· αυτοί διεφύλαξαν την Ελληνικήν φυλήν από
κινδυνωδεστάτων επιδρομών και την διετήρησαν αναλλοίωτον επί μακρούς
αιώνας σκότους και καταιγίδων . . . Αλλά κατά τους τελευταίους τούτους
χρόνους διεσαλεύθη η προς την θρησκείαν και την γλώσσαν πίστις ημών . . .
Διαμαρτύρησις κατά της ολεθρίας ταύτης πλάνης είνε η ίδρυσις της ημετέρας
Εταιρείας, έν δε των μέσων, δι᾽ ων θα αμυνθή κατ᾽ αυτής, κηρύττουσα τον
αληθή λόγο της πίστεως και της πατρίδος, έσονται ΤΑ ΠΑΤΡΙΑ.36
Religion and language have always been the guardian angels of the nation [ethnos kai
genos]; they have protected the Greek race from the most dangerous raids and have
kept it intact throughout long centuries of darkness and storms . . . However, in recent
years, our faith in religion and language has been disturbed . . . The establishment
of our Society is a protest against this disastrous error, and Patria will be one of the
media through which we shall fight against it by preaching the true spirit of faith and
fatherland.
Πάτρια tied the language question to concerns about the preservation of moral
values. Language purism ‘has functioned in part as a moral term applied to actions
aimed at overcoming evil’,37 and the constant emphasis the newspaper put on
morality and protecting ‘ancestral values’ helped to articulate a new discursive
economy that put the opponents of demoticism in a position of moral superiority.
Fear for loss of traditional values combined with fear of new ‘materialistic’
ideas gradually paved the way for the association of demoticism with Marxism
and communism. Hatzidakis became deeply concerned about the possible
consequences of the demoticist school textbooks introduced by the educational
reforms of 1917–20, as well as the alleged links between the reformers and
communism. The Russian Revolution in 1917 had already provoked an intense
reaction across Europe. When the reforms were abandoned in 1920, Hatzidakis
explained his sense of alarm:
35
Bagiakakos (1977) 124–5.
36
Τα Πάτρια (13 March 1902), 1: 1.
37
Shapiro (1989) 22.
Ουδαμού εις το σχεδιάγραμμα τούτο φαίνεται η θρησκεία, φαίνεται ότι

κατηργήθη και εις την κοινότητα ή το Soviet των παιδίων τούτων, όπως και εις
τα των συντρόφων εν Ρωσία.38
Nowhere in this plan is there a place for religion; it seems it has been abolished in the
community or Soviet of these children in the same manner as it has been abolished
in Russia.
From the 1920s onwards, Hatzidakis’ language politics underwent a significant

transformation. In effect, it was subsumed within a broader fight against materialistic,
Marxist and communist ideas. His warnings were stark:
η άρνησις της ιδέας της θρησκείας, της πατρίδος, του παρελθόντος, της
οικογενείας, της ιδιοκτησίας, και επομένως η επάνοδος εις την κτηνώδην
εκείνην κατάστασιν εις την οποίαν ευρίσκετο το ανθρώπινον γένος προ
μυριάδων ετών.39
the denial of the idea of religion, of fatherland, of the past, of family, of property, and,
as a result, the return to that bestial condition in which humanity found itself tens of
thousands of years ago.
Well into the 1930s, the elderly professor of linguistics remained faithful to his
linguistic and political agenda until he finally withdrew from public life and died
in 1941. Hatzidakis’ work and activities provide ample material for analysing the
complex network of language, ideology and power in twentieth-century Greece.
More than that: both the politics of language and the politics of linguistics, as
developed by their most distinguished and learned twentieth-century representative,
provide us with the opportunity to reflect on the appropriation of a ‘scientific
paradigm’ within a particular historical framework and to relate the evolution of
a discipline to the institutionalization and promotion of a political agenda. I have
argued that there is a major shift within Hatzidakis’ scholarship at the end of the
nineteenth century. This shift relates to attempts to control language according to
the priorities of political agendas set by bourgeois and conservative ideologies. In
this respect, his individual statements, practices, strategies and intellectual relations
are not private phenomena, and they certainly cannot be reduced to a strictly
internalist reading of the history of his discipline or the history of language itself.
My conclusion is that Hatzidakis’ linguistics became a field of ‘situated’ knowledge
directly related to the formation and segmentation of hegemonic political positions
in the first half of the twentieth century. Both language and linguistics as a discipline
developed within a particular political context that contributed to the formation
of a bourgeois public sphere, also to conservative politics, in twentieth-century
Greece.
38
Cited in Dimaras (1990) 128.
39
Hatzidakis (n.d., b), line 4.
290 effi gazi
References
Primary Sources
Δελτίον της Ιστορικής και Εθνολογικής Εταιρείας της Ελλάδος [Bulletin of the Historical
and Ethnological Society of Greece], 1 (1883).
Τα Πάτρια. Όργανον της Εταιρείας της υπέρ των Πατρίων Αμύνης [Patria: A Publication
of the Society for the Defence of Ancestral Values], 1, 13 March 1902.
Bernardakis, D. (1884), Ψευδαττικισμού Έλεγχος [An Examination of Pseudo-Atticism],
Trieste.
Dimaras, A. (ed.) (1990), Η Μεταρρύθμιση που δεν έγινε. Τεκμήρια Ιστορίας [The
Reformation that Never Happened. Historical Documents], vol. 2 (1895–1967), Athens.
Hatzidakis, G.N. (1884), Μελέτη επί της Νέας Ελληνικής [A Study of Modern Greek],
Athens.
Hatzidakis, G.N. (1888), Συμβολή εις την ιστορίαν της μεσαιωνικής ημών γλώσσης.
Πεντηκονταετηρίς Πανεπιστημίου Αθηνών (1886-7) [Contribution to the History of our
Medieval Language: Fiftieth Anniversary of the University of Athens], ed. G. Karamitsas,
Athens.
Hatzidakis, G.N. (1890), «Περί του Γλωσσικού Ζητήματος εν Ελλάδι» [‘On the Language
Question in Greece’], Αθηνά, 2: 169–235.
Hatzidakis, G.N. (1898), Αναγνώσματα περί των γενικών αρχών της Συγκριτικής
Γλωσσικής υπό Whitney και Jolly μετερρυθμισμένα εις την Ελληνικήν [Readings on the
General Principles of Comparative Linguistics by Whitney and Jolly Converted into Greek],
Athens.
Hatzidakis, G.N. (1905a), Μεσαιωνικά και Νέα Ελληνικά [Medieval and Modern Greek],
vol. 1, Athens.
Hatzidakis, G.N. (1905b), Το πρόβλημα της νεωτέρας γραφομένης ελληνικής υπό K.
Krumbacher και Απάντησις εις αυτόν [The Question of Modern Written Greek by K.
Krumbacher and a Reply to Him], ed. A.N. Maltos, Athens.
Hatzidakis, G.N. (1906), Περί της ενεστώσης μεθόδου εν ταις γλωσσικαίς ερεύναις [On
the Current Method in Linguistic Research], Athens.
Hatzidakis, G.N. (1907), Και πάλιν περί του γλωσσικού ζητήματος [On the Language
Question Again], Athens.
Hatzidakis, G.N. (1959–63), «Περί των εν τη νεωτέρα Ελλάδι γλωσσικών μελετών»
[‘On linguistics in modern Greece’], Μεγάλη Ελληνική Εγκυκλοπαίδεια, 10, Athens:
111–12.
Hatzidakis, G.N. (n.d., a), Σημειώσεις πανεπιστημιακού μαθήματος του Γ. Χατζηδάκι
περί του γλωσσικού ζητήματος [G.N. Hatzidakis’ Notes for a University Lecture on
the Language Question], G.N. Hatzidakis Archive, Α/Α 295, file 38/line 13.
Hatzidakis, G.N. (n.d., b), Επιστολή [Letter], G.N. Hatzidakis’ Archive, A/A 286, file 38/
line 4.
Kontos, K. (1882), Γλωσσικαί Παρατηρήσεις [Linguistic Observations], Athens.
Mavrofrydis, D. (1871), Δοκίμιον Ιστορίας της Ελληνικής Γλώσσης [Essay on the History
of the Greek Language], Smyrna.
Mistriotis, G. (1909), «Επιστολή προς Γ. Χατζηδάκην περί Νέας Ελληνικής
Γραμματικής» [‘A Letter to G.N. Hatzidakis concerning Modern Greek Grammar’]
(«Αθήναι», 3 June 1908), Ρητορικοί Λόγοι [Rhetorical Speeches], vol. 3, Athens: 141–8.
Psichari, J. (1886), Essais de grammaire historique néo-grecque, vol. 1, Paris.
Psycharis, Y. (1971), Το Ταξίδι μου [My Voyage], ed. A. Angelou, Athens.
Triantafyllidis, M. (1993), Άπαντα [Complete Works], vol. 3, Thessaloniki.
Secondary Literature
Babiniotis, G. (1983), «Γ.Ν. Χατζηδάκις (1848–1941). Η συμβολή του στην
αποκατάσταση της γλωσσικής μας ταυτότητας και στην επιστημονική σπουδή
της ελληνικής γλώσσας» [‘G.N. Hatzidakis (1848–1941): his contribution to the
reconstitution of our linguistic identity and to the scientific study of the Greek language’],
Αριάδνη. Επιστημονική Επετηρίς Φιλοσοφικής Σχολής Πανεπιστημίου Κρήτης, 1:
294–307.
Bagiakakos, D. (1977), Γεώργιος Ν. Χατζηδάκις (1848–1941). Βίος και Έργον [Georgios
N. Hatzidakis (1848–1941: Life and Work], Athens.
Bourdieu, P. (1985), ‘The genesis of the concepts of Habitus and of Field’, Sociocriticism,
2: 11–24.
Bourdieu, P. (1993), The Field of Cultural Production. Essays on Art and Literature,
London.
Burke, P. (1994), ‘Introduction’, in P. Burke and R. Porter (eds.), The Social History of
Language, Cambridge: 1–20.
Butler, J. (1990), Gender Trouble: Feminism and the Subversion of Identity, London.
Christidis, A.-F. (2007), ‘General introduction: histories of the Greek language’, in Christidis
(ed.), A History of Ancient Greek: From the Beginning to Late Antiquity, Cambridge:
1–23.
Corfield, P. (1991), Language, History and Class, London.
Crowley, T. (1996), Language in History: Theories and Texts, London.
Foucault, M. (1992), The Archaeology of Knowledge, tr. A.M. Sheridan Smith, London.
Gazi, E. (2000), Scientific National History: The Greek Case in Comparative Perspective
(1850–1920), Frankfurt and New York.
Hobsbawm, E. (1992), Nations and Nationalism since 1780: Programme, Myth, Reality,
Cambridge.
Joseph, B. (1985–6), ‘European Hellenism and Greek nationalism: some effects of
ethnocentrism on Greek linguistic scholarship’, Journal of Modern Greek Studies, 3–4:
87–96.
Koerner, K. (ed.) (1989), Practicing Linguistic Historiography. Selected Essays, Amsterdam.
Mackridge, P. (1998), ‘Byzantium and the Greek language question’, in D. Ricks and P.
Magdalino (eds.), Byzantium and the Modern Greek Identity, Aldershot: 49-61.
Mackridge, P. (2004), ‘“Sie sprechen wie ein Buch”: G.N. Hatzidakis (1848–1941) and the
defence of Greek diglossia’, Kampos, 12: 69–87.
Petersen, H. (1959), The Discovery of Language: Linguistic Science in the Nineteenth Century,
Bloomington, Ind.
Shapiro, M. (1989), ‘A political approach to language purism’ in B.H. Jernudd and M.J.
Shapiro (eds.), The Politics of Language Purism, Berlin and New York: 21–9.
Scott, J. (1996), Feminism and History, Oxford and New York.
Steinberg, J. (1994), ‘The historian and the questione della lingua’, in P. Burke and R. Porter
(eds.), The Social History of Language, Cambridge: 198–209.
15
‘Language Issues’
after the ‘Language Question’:
On the Modern Standards
of Standard Modern Greek*
Spiros Moschonas
Language standards as ideologies

According to a widely held view in sociolinguistic research, a standard language
is neither a particular variety nor a well-defined process or practice. As James and
Lesley Milroy have argued:
it seems more appropriate to speak more abstractly of standardization as an ideology,
and a standard language as an idea in the mind rather than a reality – a set of abstract
norms to which actual usage may conform to a greater or lesser extent.1
A standard language, according to this view, is a mental construct, not an actual

linguistic state or process. A standard language is a standard rather than a language.
The obvious purpose of any set of standards is the ‘imposition of uniformity’
upon linguistic variety.2 Because absolute standardization can never be achieved,
language uniformity exists only as an ideological rationalization. A standard
language is a language perceived as standard.
I wish to inquire into the methodological and theoretical implications of this
approach to standardization. Since changes in a standard language are, by definition,
changes in ideological standards (that is, in the ways a language is perceived rather
than the ways a language actually is), it should always be possible, in principle
at least, to trace changes in standard languages through respective changes in
language standards.
* I am greatly indebted to the editors for corrections and valuable comments on this paper, research
for which was supported in part by a University of Athens research grant (70/4/4131).
1
Milroy and Milroy (1999) 19.
2
Milroy (2001).
293
294 spiros moschonas
I consider Modern Greek standardization as a case of language ideology. My

aim is to trace the changes in the standard language presumed to have occurred
since the language reform of 1976 as changes in the language standards. Such
changes in standards are evidenced in public, metalinguistic representations of a
language. I shall only be concerned here with the metalinguistic attitudes expressed
in the Greek print media in the period following the 1976 reform, when an official
standard, based on the demotic norm, was finally established. I shall provide
evidence for two interrelated claims:
(i) After 1976, several seemingly disparate ‘language issues’ were raised in the Greek
press and, to a lesser degree, in the electronic media. In time, these post-diglossia issues
formed a coherent communicative sequence, a collective media narrative, so to speak,
which developed in accordance with the presuppositions of a new ‘regime ideology’
of Standard Modern Greek. To the formation of this collective narrative, language
professionals, folk ideologists and journalists have all contributed generously.
(ii) According to this new ideology, the one and only language of the state, Modern
Greek, should itself be pictured as a state, a territory or a regime, which comprises
a pure and sacred interior that has to be kept intact by everything surrounding it.
This ‘regimentation’3 of the Modern Greek language also marks an important shift
in public concern, a shift from ‘internal’ issues of norm definition and elaboration to
‘external’ issues of language contact, maintenance and spread.
The ‘language regime change’ of 1976 (the γλωσσική μεταπολίτευση, as it was

prophetically called by Manolis Triandaphyllidis in 1938)4 marks an important
shift in the ideology of the standard language. A series of new ‘language issues’ has
gradually replaced the perennial Greek ‘language question’ (Γλωσσικό Ζήτημα).
My hypothesis is that the new issues are the markers of a new ideology of the
standard language.
I wish to argue that there are actually no ‘regimes of language’,5 no basic facts
about language ‘depending’, as it were, ‘simply upon the fact of their fact’.6 There
are only regimes of language ideology, which involve complex facts about both
language and public representations of language. All public representations of
language are ideological.7
3
Kroskrity (2000).
4
Triandaphyllidis (1981) 168.
5
Kroskrity (2000).
6
Poe (2004) 9.
7
Following Silverstein (1979) 193, I understand language ideologies as metalinguistic systems
of value and belief. Language ideologies function metapragmatically (Silverstein (1976) 48–51 and
(1979) 207–8), referring not to language itself but rather to ‘perceived language structure and use’:
Silverstein (1979) 193. Accordingly, in taking account of language ideologies, one has to shift the
emphasis one level up, from the presumed ‘regimes of language’ to the vast (and largely unaccounted
for) regimes of language ideology. One has to ascend from language to metalanguage. Ideologies are
often thought to have a rationalizing function. But since ideologies might very well be incoherent or
unsound systems of belief, a general definition of ideologies should not, on a priori grounds, single out
any one of their ‘ideological strategies’ (Eagleton (1991) 33–61) or ‘functions’ (Silverstein (1993) and
‘language issues’ after the ‘language question’ 295
Language issues
In what follows, I consider articles printed in the Greek newspapers and in wide-
circulation magazines as indexes of the issues raised and debated within the Greek
linguistic community at large, or, at least, within the community sector of active
‘ideology brokers’:8 intellectuals, educators, politicians, planners, journalists,
opinion makers and the like.
Indexing language issues through newspapers articles is, of course, a perilous
enterprise.9 In Greece, only a handful of trained linguists cultivate the peculiar
genre of linguistic journalism (‘writing about language issues’). However, columns
advising on proper usage are quite common. Debates, when they occur, are rarely
carried out in the same newspaper; they are carried over to different newspapers
and they unfold in front of separate audiences. Newspapers are not an interactive
medium. Active access, for both the experts and the public, is usually restricted to
writing letters to the editor. It is clear that public opinion should not be confused
with expressed opinion; but expressed opinion is certainly an index of assumed or
implicit opinion. After all, this is how an index works. An index is a perspectival
representation attached to a denotatum in such a way that the denotatum and its
indexed representation cannot be separated. Indexes form part of the situations
they index (in much the same way as a road sign forms part of a traffic situation,
which it also helps to define). ‘Being in the newspapers’ remains, after all, a decisive
criterion for what it is to count as an issue.
I do not adopt the ‘view from below’.10 Newspaper articles provide no such
perspective. As George Thomas has pointed out,11 language ideologies propagate
in waves, continuously expanding to outer concentric circles. Since there is no
ideology without believers, my preferred view is from the mid-circle of followers
and devotees. I find the discourse of the ‘intermediaries’ much more revealing than
the refined discourse of the originators (‘grand’ ideologists, ‘fathers’ of a standard
language, ‘visionaries of the nation’, linguists and lexicographers, literary experts).
It is an advantage for the researcher that journalistic ‘entextualization’12 is highly
stereotypical, and stereotyping accelerates the ‘re-contextualization’ process.
Mediating discourse thus manages to become public in ways that the higher
registers cannot achieve. It has not yet become folk-ideological (in the sense of
(1979) 205–8) at the expense of all the others. Ideologies are ‘unifying, action-oriented, rationalizing,
legitimating, universalizing and naturalizing’ (Eagleton (1991) 45). Ideologies are also ‘integrative’
(Geertz (1973); Ricoeur (1986), ch. 15): they facilitate the integration of individuals into social groups.
They are representative: the Subject interpellated by Ideology – Althusser (1971) – usually speaks for
and in the name of others. They are confrontational: ideologists are constantly engaging in battles. But
no matter what their other functions are, language ideologies are specifically metapragmatic.
8
Blommaert (1999).
9
See Blommaert and Verschueren (1998) 190–1, for some necessary methodological precautions.
10
Ibid. 189.
11
Thomas (1991), ch. 6.
12
Silverstein and Urban (1996).
Niedzielski and Preston),13 but it is an efficient means for conditioning public

reflexes.
My research is based on an archive of articles from 76 Greek newspapers
(including 31 national dailies) and about 100 wide-circulation magazines (dealing
with non-linguistic subject matter). The publications I consider cover more
than two decades (1976–2001), but only the period from November 1999 to
January 2002 has been exhaustively covered (3,706 entries in 2000 and 5,170
in 2001). The period of 1990–9 is covered by a representative sample from the
newspapers with the widest circulation (about 1,500 articles). However, even
in this underrepresented period, issues that roused intense public concern over a
continuous period of time (mainly issues (1)–(7) below) are considered in detail.
The earlier period from 1976 to 1989 is represented by a less coherent sample of
newspaper articles and letters to the editor. However, the issues of this period have
already been documented and commented on in the scholarly literature.14 Most of
the newspaper articles written by the protagonists from this early period have also
been reprinted.15 All relevant publications (more than 10,600 articles) have been
summarized and classified by author, place, date of publication, genre, topic, cross-
references and keywords.16
The main language issues raised in the Greek newspapers in the period from
1976 to 2001 are listed in Table 1, and may be divided into three main groups.
Issues (1)–(6) in the first group are the ones that have provoked a ‘moral panic’ (see
below). Issue (7) is actually a cluster of related issues, raised immediately after the
official resolution of diglossia; these issues form a coherent group in their own right,
but only the debate on whether Ancient Greek should be taught in secondary
education acquired the impetus of issues (1)–(6). Issue (7) is not the group I focus
on, but it could be considered a starting point for many of the issues raised in the
wake of the 1976 language reform. In the third group, (8)–(10), are issues that
never caused intense public concern, yet continue to recur; I call them ‘routine
issues’. Item (11), in contrast, is a category for issues that do not arise in the press:
they are avoided, if not literally censored. Item (12) is the usual catch-all category
(‘varia’). The order of listing in Table 1 is semi-chronological: issues (1)–(7) appear
in reverse chronological order, while issues (8)–(10), and possibly some of (12),
recur from time to time:
13
Niedzielski and Preston (2000).
14
Landsman (1989); Frangoudaki (1992), (1997).
15
E.g. Kriaras (1979), (1984), (1988), (1992); Babiniotis (1994a), (1994b), (1994c).
16
Owing to the large number of articles involved, I have opted for a piecemeal analysis based on
several small-scale case studies. Here I draw on previous publications of my own and of my students:
respectively, Moschonas (2001a), (2001b), (2001c), (2002a), (2002b), (2004), (2005a), (2005b), and
Moschonas (2004/5). The numerical results in Table 2 and Table 3 below are derived from a pilot
study, which focused on the ‘control’ period of November 1999–January 2000, a period with no moral
panics (see Moschonas (2001a) 92–3, 99).
Table 1
New ‘language issues’ (1976–2001)
(1) English as a ‘second official language’
(2) ‘Romanization’ of the alphabet
(3) Bulgarians versus Babiniotis
(4) ‘Macedonian’
(5) The ‘five-language regime’ in the EU
(6) ‘Word poverty’
(7) Post-diglossia issues [teaching Ancient Greek, ‘monotonic’ orthography,
the ‘language problem’]
(8) Foreign words, influence of English, purism
(9) ‘Monotonic’ versus ‘polytonic’ orthography
(10) ‘Greek abroad’, Greek as a second language
(11) Censored [minorities etc.]
(12) Miscellaneous [local issues etc.]
What follows is a very short description of each entry in Table 1:17

(1) In November 2001, the Greek Commissioner in the EU, Anna
Diamandopoulou, proposed having English institutionalized as the ‘second official
language’ of the Greek state. The proposal was widely criticized in the press and
the media; it was judged to be ‘outrageous’ or even ‘inconceivable’. Under pressure
from its critics, the proposal was immediately withdrawn. However, disapproving
articles and letters continued to appear in the press well into 2002.
(2) In January 2001, forty members of the Academy of Athens signed an open
letter calling for a crusade against the use of the Roman alphabet for transliterating
Greek texts, especially in new technologies (e-mails and the like). Most
journalists approved this call against ‘Romanization’, but specialists received it
with considerable reserve.18 The Academy’s open letter was, in a way, Greece’s
unscheduled launch event into ‘2001: European Year of Languages’.
(3) In a dictionary of Modern Greek published by George Babiniotis in 1998,
there was a sport-slang entry under the word Βούλγαρος (‘Bulgarian’), referring to
a fan of (or a player in) a sports team from northern Greece. The entry was judged
unworthy of inclusion in a dictionary: it was read as an insult to northerners and
‘a move that divides the nation’. The case was brought to court and the dictionary
was temporarily banned. The Supreme Court later annulled this decision (1998–9).
However, the entry disappeared from later editions of the dictionary.
(4) In the mid-1990s, several articles argued against the use of the name
‘Macedonia’ in reference to the Former Yugoslav Republic of Macedonia. The
political turmoil of this period encouraged state-sponsored demagogism against
fyrom, which was represented in the media as a ‘usurper of Greek symbols’. The
17
For details see Moschonas (2004) 180–5.
18
See e.g. Kriaras (2001); cf. Chartoulari (2001).
official position of Greek foreign policy still is that neither the newly formed state
nor its language should be given a ‘Greek name’.
(5) In December 1994 the media reacted strongly to a suggestion by the French
presidency of the EU to reduce to five the number of ‘working languages’ in
the European Parliament and in other EU representative organizations. Greek
politicians and ‘men of letters’ called for a crusade against such a ‘barbaric act’.
Their rhetoric was later stigmatized as ‘exaggerated’ and ‘hysterical’ by members
of the ‘linguistic opposition’.19
(6) In 1985, during the national Examinations for Admission to the Institutions
of Higher Education, the failure of the examinees in essay writing to recognise
the meaning of two learned words, ευδοκίμηση (‘prosperity’) and αρωγή
(‘assistance’), was considered by many to be an alarming indicator of the younger
generation’s growing λεξιπενία (‘word poverty’: that is, inadequate vocabulary).
Youth slang, ‘marred’ as it was with foreignisms, was offered as additional testimony
to ‘word poverty’. Ignorance of learned words and youth slang are often invoked as
factual premises in arguments favouring the ‘re-introduction’ of Ancient Greek in
secondary education.20
(7) Diglossia did not end with the official resolution of the ‘language question’
in 1976. The late demoticists’ standardization formula (Standard Modern Greek =
demotic + a few archaisms as necessary)21 allowed – even encouraged – a ‘residual
diglossia’, which persists to this day. The question how much katharevousa (of the
archaistic, puristic variety) is to be allowed is frequently debated in newspapers’
usage-columns and in popular language-guides.22 However, the ‘language question’
has lost its impetus. What was once capable of causing a moral panic has now
become a routine issue. Several post-diglossia versions of this issue may be singled
out:
(a) Ancient Greek still functions ideologically as a substitute for katharevousa.
This explains why the question whether Ancient Greek should be taught in
secondary education had such publicity in the 1980s. From 24 November 1986 to
1 June 1987, a single newspaper, Ελευθεροτυπία, published one to three articles
on this question almost every day. It is still presumed, even among linguists, that
learning Ancient Greek is a prerequisite for students ‘coming to know’ Modern
Greek or becoming fluent in it.
(b) A 1982 reform of the orthography made the use of the ‘monotonic’ (single-
accent) system official. However, a few magazines and a considerable number
of ‘high-register’ books (of poetry mostly) are still published in the traditional
‘polytonic’ (multi-accent) orthography. Maintaining the use of the polytonic
system may be interpreted, by implication, as a continuing challenge to the 1976
reform. Several writers believe that the monotonic system ‘alienates’ Greeks from
19
E.g. Maronitis (1995).
20
See Androutsopoulos and Iordanidou (1999).
21
Triandaphyllidis et al. (1978).
22
See Moschonas (2001b), (2005a).
traditional Greek literature.23 Some writers consider the new accentuation system
‘a step towards Romanization’.24 Still others hold that the simplified orthography
may cause learning disabilities and dyslexia, which could be cured by the use of
the complicated polytonic orthography.25
(c) During the 1980s, George Babiniotis, an influential conservative linguist,
argued insistently that a new ‘language problem’ had now taken the place of the
perennial language question: the language was diagnosed to be ‘in a bad state’
and ‘in need of correction’. To help solve this ‘problem’, a Language Association
was formed in 1985.26 At about the same period, several ‘progressive’ linguists
(mainly from the University of Thessaloniki) formed what could retrospectively
be described as a loosely organized ‘linguistic opposition’. The members of this
opposition argued strongly in favour of a ‘descriptivist’ attitude, their typical
conclusion being that ‘the Greek language is in as good a state as ever.’ The two
camps have obvious connections to the former proponents of katharevousa and
demotic respectively.27
The following three items are routine issues:
(8) A recurrent issue in the 1980s and the early 1990s was the adoption of
loan words, mainly from English.28 Around this period, linguistic purism ceased
to be diglossic and became biglossic: translating katharevousa into the demotic
was not seen as a problem any more; what was now considered the main problem
was the adoption and adaptation of English loanwords in Greek. It should be
stressed that during this period purism was practised on a massive scale, mainly
through translations (not only of literary works but also of technical books, such as
computer manuals). Thus the ‘spirit of katharevousa’ survived in the ‘continuing
calquing of words and expressions’.29 In the press, purism manifested itself as an
outward rejection of bilingualism and language contact.
(9) The ‘monotonic’ system has been another topic of occasional concern.
Defence of the traditional ‘polytonic’ system mainly stresses its symbolic
advantages, while the new orthography’s main advantage, simplicity, cannot be
overemphasized. It looks as if the dispute cannot be settled. As indicated already
23
E.g. Elephandis (1998) 384.
24
E.g. Droumboukis (n.d.); Vrahnias (1992); Tsikopoulou (1995); Gotsis (1997).
25
Tsengos, Papadaki and Vekiari (2005); see Moschonas (2006) for a critique of such absurdities.
26
Greek Language Association (1984), (1986).
27
The existence of a linguistic opposition underlines the fact that language ideologies are
confrontational: there is never one ideology only; there are at least two, and they are in opposition.
Bourdieu (1991) placed particular emphasis on the confrontational character of language ideologies
within fields (‘champs’) of practice. See also Kroskrity (2000) 12, and the essays collected in Blommaert
(1999). The confrontational character of language ideologies has an important methodological
consequence: when speaking of ideologies, one needs to specify first the general conceptual framework
within which both consent and dissent are exercised. Language ideologies are not the subject matter;
rather, they are the condition of ‘language ideological debates’: Blommaert (1999). They are the
conceptual minimum necessary for any general ‘folk-linguistic’ statements to make sense.
28
On this issue see Delveroudi and Moschonas (2003).
29
Horrocks (1997) 364.
(7b), the issue of orthographic reform has been overblown, mainly through its
association with other, ‘bigger’ issues (like the 1976 language reform and the
‘danger of Romanization’).
(10) Greek as a second language has become a new topic – but not a language
issue – in the newspapers, marked by a gradual increase in the number of relevant
publications during the past two decades. It is also a brand-new area of research in
Greek linguistics. Greek is valued as a second language mainly because it serves
the politics of assimilating immigrants, repatriated Greeks or minority groups
within Greece. In the wake of the success of Greek as a second language, a new
politics is slowly emerging: that of Greek as a foreign language; in other words, a
politics for ‘extending’ the use of Greek outside Greece or reinforcing its use in
what is called, in an interesting spatial metaphor, ‘dialect enclaves’ of the Greek
language.30 Language issues rarely make it to the front page, and for this reason
a recent front-page headline from the newspaper Η Καθημερινή (18 November
2005) could be read as a portent: ‘Η ελληνική γλώσσα περιζήτητη σε όλες
τις βαλκανικές χώρες’ (‘Great demand for the Greek language in all the Balkan
countries’).
In the interest of a more complete description, two more categories are added in
Table 1. Item (11) is an open category, which contains issues that are systematically
not raised in the press. Item (12) comprises miscellaneous issues that belong to the
classification in less systematic ways.
(11) The issues in this category are the ones ‘erased’31 by the language ideologists.
They are issues to be avoided, or to be handled with tact. There is, for example,
a notable scarcity of publications about language minorities in Greece. On this
issue, the press follows or ‘respects’ the state’s official ‘politics of silence’.32 It should
be stressed that the attitude of the press involves no compulsion: the press is not
censored; it chooses to be silenced.
(12) The last category in a classification (‘varia’) typically comprises whatever
cannot be reasonably assigned to the main categories. It is a taxonomist’s confession
of his miscategorizations. Among other things, the following could be placed in this
miscellaneous category:
(a) Issues raised only or mainly in local newspapers. In 2001, for example, more
than one hundred shop owners in the city of Volos were brought to court because,
in defiance of the law (1491/84), they used the Roman script exclusively in their
shop signs.33
(b) Issues raised only or mainly in the Greek Cypriot press; for instance, the
debate on whether the official language(s) of the University of Cyprus, which
was founded in 1992, would be English only, Greek only, or Greek and Turkish;34
30
Christidis (1999).
31
In the sense of Irvine and Gal (2000) 38–9.
32
Kostopoulos (2000) 352.
33
Moschonas (2001c).
34
Karyolemou (1994) 257–9.
likewise the debate prompted by the government’s efforts to standardize Greek

Cypriot place names on the phonetic norms of Standard Modern Greek.35
(c) Issues raised only or mainly by certain writers, on particular occasions or in
specific genres. For example, regular newspaper columns advising on usage should
be treated separately; such columns elaborate and propagate the norms of the
written language, providing interesting clues as to when and under what conditions
the norms change.36 Letters to the editor are representative of a smaller circle of
followers and devotees; they employ an unsophisticated discourse echoing the
much more elaborate discourse of the elite.37 On the other hand, articles written
by trained linguists – or by the ‘linguistic opposition’ – should not be treated only
as simplified accounts that attempt to communicate linguistic wisdom to a wider
public;38 such texts are testimony to the ideological involvement of the professionals,
whose presuppositions they betray.39 Sponsored campaigns, such as the EU
campaign, ‘2001: European Year of Languages’, should also be dealt with in the
same spirit.40 The merits of multilingualism should not deter us from considering
such events for what they really are: organized forms of linguistic propaganda.
Indexing language issues

If newspaper coverage is to be considered as an index of language issues, the
question to be asked is: in what ways are all these diverse publications dispersed?
The obvious answer is: publications are distributed across different newspapers,
journalistic genres, times and audiences. Since there are no readership surveys
of language-interested groups specifically, I omit audiences and examine the first
three points of reference in turn.
Newspapers
Table 2 lists the four Greek newspapers which published the largest number of
articles on language matters over a relatively short period of time (November
1999–January 2000), a period in which no major language issue was raised in the
newspapers or the electronic media. Not accidentally, these four newspapers are
the ones with the widest circulation. Two measures are used in the table. In the
35
Karyolemou (2000).
36
Moschonas (2001b), (2005a).
37
Moschonas (2002b).
38
As Johnson (2001) has implied.
39
As Silverstein (1979) 193, has stressed, ‘[i]f we compare such ideologies [about language] with
what goes on under the name of ‘scientific’ statements about language, we might find that in certain
areas the ideological beliefs do in fact match the scientific ones, though the two will, in general, be part
of divergent larger systems of discourse and enterprise’. It should be noted, however, that folk-linguistic
conceptions of language, in sharp contrast to their professional counterparts, are rarely, if ever, apolitical,
asocial or non-ideological. They may be erroneous or blatantly wrong, but this has not been much of
a problem for the non-specialist. In contrast, the ideological conceptions of the professionals are easily
identified as such by the very fact that they profess to be non-ideological.
40
See Moschonas (2002a).
first, all articles referring to linguistic matters, even en passant, are counted. In the
second, articles containing only passing references to such matters are omitted:
Table 2
Number of articles per newspaper (November 1999–January 2000
Newspaper N=364 (%)
ΕΛΕΥΘΕΡΟΤΥΠΙΑ 49 13.5
Η ΚΑΘΗΜΕΡΙΝΗ 39 10.7
ΤΟ ΒΗΜΑ 31 8.5
ΤΑ ΝΕΑ 30 8.2
40.9
N=284 (%)
ΕΛΕΥΘΕΡΟΤΥΠΙΑ 35 12.3
ΤΑ ΝΕΑ 23 8.1
ΤΟ ΒΗΜΑ 21 7.4
Η ΚΑΘΗΜΕΡΙΝΗ 18 6.3
34.1
This table shows that 41% (or, by the second count, 34%) of all the publications
on language matters occur in wide-circulation daily newspapers. One may well feel
unsure how to articulate this finding. Obviously, a large number of articles do appear
only in wide-circulation newspapers; equally, a large number of articles do not
appear in the wide-circulation newspapers (that is, they appear only in newspapers
with a modest readership, or in local weeklies, bi-weeklies or monthlies; linguistic
articles do not appear in magazines of general readership). Be that as it may, Table
2 provides sufficient reason to mistrust research approaches that take into account
only ‘mainstream publications’.41
Genres
The reason why two counts are used in Table 2 is that one might or might
not want to take into account references to language issues made in passing,
during an interview perhaps, or in an article of otherwise non-linguistic subject
matter. However, the significance of such en passant references should not be
underestimated. Interestingly enough, the proposal by the Greek Commissioner
in the EU to institutionalize English as a second official language – Table 1, Issue
(1) – was made during an interview about EU political issues; this means that
it qualifies as a reference en passant. The necessary inference is obvious: even
a passing reference, when it comes from the authorities, or when it fulfils other
conditions that need to be specified, can cause a major communication event (a
‘moral panic’). Passing references should not be disregarded. One could also put
into the en passant category references to ‘body language’, the ‘language of politics’,
41
Blommaert and Verschueren (1998) 190.
the ‘language of sports’ and the like. References to literary style (common in book
reviews) could also be considered to be en passant if the reviews are not dealing
with a book on language, or if a review focuses on the content rather than the form
of a literary work.
Most en passant references appear in wide-circulation newspapers. I cannot
offer precise figures, because they are difficult to calculate (it is easy for en passant
references to be overlooked during the archiving process), but my estimate is that
they amount to some forty per cent of the whole corpus. They are characteristic
of journalists’ ‘language awareness’, of their training in matters linguistic, or of a
newspaper’s propensity to account for such matters. One should in any case take
into consideration the fact that, in a sense, all references to language issues in
newspapers and non-linguistic magazines are, by definition, accidental. Newspapers
do not regularly cover language matters; and language matters, in sharp contrast to
all the other matters a newspaper does cover, are not easy to define.
Table 3, using two counts as before, shows how the relevant publications are
distributed across newspaper genres:
Table 3
Types of article (November 1999–January 2000)
Type N=364 (%) N=284 (%)
Opinion 70 19.2 70 24.6
News 70 19.2 70 24.6
Book reviews 34 9.3 34 12.0
Short comments 33 9.1 33 11.6
Features 24 6.6 24 8.5
Letters 23 6.3 23 8.1
News in brief 21 5.8 21 7.4
Interviews 9 2.5 9 3.2
Editorials 0 0.0 0 0.0
References en passant 80 23.0 – –
Features (or ‘reportages’) typically consist of a number of related articles; they
attempt ‘in-depth’ coverage of an issue; they are polyphonic; and they attract expert
opinion. They form a safe index of what is actually recognised as a language issue (in
contrast to what a researcher assumes a language issue to be). In my corpus, there
are features dedicated to all the issues in the categories (1)–(8) of Table 1, but no
features for categories (9), (10) or, of course, (11). ‘News in brief’ is a convenient (if
perhaps awkward) cover-term for references to event listings, announcements, and
notices of, for instance, public presentations of books on language or of language
survey results. ‘Short comments’ appear mainly in the so-called παραπολιτικές
στήλες (‘asides’). Book reviews, letters to the editor and interviews I consider to
be well-defined genres. So too are front-page editorials, but there are none during
this period. There are scarcely any in my corpus; editorials on language issues tend
to occur only in periods of ‘moral crusades’ (or, possibly, in August).
The remaining categories are the two major ones: opinion and news articles.
They cover the largest proportions of occurrences (19.2% or 24.6%, depending
on the measure employed), significantly higher than the figures for any of the
other categories. Although not at all homogeneous, these two genres are the most
comprehensive. News articles might also be taken to subsume ‘news in brief’
(yielding 25% in total) and possibly feature articles too (in which case the total
would be 31.6%). Opinion articles might likewise be taken to subsume ‘short
comments’ (28.3% in total) and book reviews (the total now would be 37.6%).
The question is how to define each genre and how to distinguish them from each
other.
Opinion articles are, I suggest, quite easy to define and isolate within a corpus; for
instance, ‘usage’ columns are clear cases of opinion articles. A free-standing essay
on (say) ‘The merits of our great language, Greek, the world’s most ancient tongue’
is also not difficult to place in the category of opinion articles. But then, what is a
‘linguistic piece of news’? We can all very easily provide examples of non-linguistic
news (9/11, election results, the referenda on the EU constitution, bird flu, a train
collision). Of course, there is always the possibility that a news item is not properly
classified or named (witness such terms as a ‘terrorist act’). But what is ‘news’
about language? And what does a linguistic ‘news item’ name, that is, on what
terms is it related, metalinguistically, to language?
A researcher may employ technical and operational criteria in order to identify
linguistic news items. Such criteria may be textual (involving, for instance, the use
of event sentences or past tenses, the formation of narrative or tense sequences);
alternatively, they may be communicative or discursive (the position of a text in a
newspaper’s overall layout, its relation to other texts in a sequence); or they may be
conceptual – that is, language-ideological – criteria, as suggested already. I propose
that linguistic news items and, more generally, language issues, cannot be defined
independently of language ideologies, therefore of conceptual criteria. As I wish
to show, neither communicative sequences nor even the dichotomy presupposed
between news and opinion make much sense outside the conceptual framework of
a language ideology.
Time
How are the issues distributed over time? What are the types of sequence by means
of which language issues are sustained?
Articles do not occur in isolation. They form communicative sequences,42
which exhibit thematic and intertextual relations. Of such sequences, linguists are
familiar mostly with those under the rubric of a ‘moral panic’.43
42
43
Cameron (1995), ch. 3.
For a ‘moral panic’ to impinge as such, it is not sufficient that the relevant articles
assume a moralizing or spiritualizing tone. We need to have a sudden increase
in the number of publications and the number of persons involved in the debate
(journalists, opinion-makers, audiences). According to Thompson, Goode and
Ben-Yehuda, and others,44 the normal communicative sequence for a moral panic
is the following: something or someone is first defined as a threat to a community’s
values or interests; in the media the threat is portrayed ‘disproportionally’, in
an oversimplified manner; those involved in a moral panic show ‘a high level of
concern’, and an ‘increased level of hostility’ towards the perceived threat; the
relevant articles assume a spiritual, moralizing tone; moral panics appear suddenly
but are short-lived; there is a rapid build-up of public concern, followed by a
response from authorities or opinion-makers, before the panic recedes, never to
appear again with the same intensity.
This is one type of communicative sequence. Routine issues, as I have already
called them, follow a very different pattern: the media do not necessarily respond
to any moves or initiatives perceived as threatening; coverage of routine issues
has a very low ‘news value’; debates on such issues draw excessively on a limited
repertoire of types of argument (bringing forth an issue seems more important than
trying to resolve it); the presentation of the issue is not addressed to the ‘general
public’ but rather to a smaller circle of experts, followers, devotees; finally, media
coverage is characterized by its low intensity and non-periodic recurrence. A moral
panic disappears. Routine issues persist.
The two types of communicative sequence can be, and often are, combined.
After the panic recedes, the problem that was perceived as a threat would normally
turn into a routine issue. It will be recalled but it will not reappear. In sharp contrast
to this pattern, there is only one case of a routine issue becoming a moral panic in my
corpus – ‘Romanization’ of the alphabet (Issue (2) in Table 1), and this happened
only when the issue was taken up by the authoritative Academy of Athens; even
then, it did not acquire the intensity of other issues. Moral panics are caused by
issues that need to be resolved. Routine issues, on the other hand, only need to be
reaffirmed; they cannot be resolved, because they touch on the most conventional
aspects of language (such as orthography); or they need not be resolved, because
there is an almost unanimous consensus about them (for instance, teaching Greek
as a second language and adapting English loan words to Greek).
44
Thompson (1998) 8–9; Goode and Ben-Yehuda (1994) 33–41; see also Johnson (1999) 2–3.
Figure 1. Two types of issue

Figure 1
Two types of issue
number Issue 1 (A)

of Issue 2 (A)
articles Issue 3 (B)
Issue 4 (B)
time
Figure 1 is an idealized graphical representation of the types of time sequence
associated with each type of issue. Type A sequences are characteristic of moral
panics; type B sequences characterize routine issues. Each type of sequence is
associated with a different graph. Type A graphs reach a peak (a maximum value)
very quickly, and then gradually decrease, never to recur with the same intensity.
Type B issues, on the other hand, have considerably lower maximum values, but
the issues recur persistently and form intermittent sequences. Otherwise, both type
A and type B graphs are irregular and non-periodic.
A general ideological framework

I have sketched several ways in which a corpus of newspaper articles could be used
to index particular language issues. We are now left with the task of correlating
language ideologies with the language issues that they index. My assumption has
been that public representations of language are always ideological. There are
no language issues unless there is a language ideology that informs, supports and
sustains them. According to my assumption, language ideologies function through
this double indexing, of language issues with types of publication (or perhaps more
generally with discourse types) and then of conceptual frames (or idealogies) with
language issues.45
45
I employ Sir William Hamilton’s period term (idealogies) in order to stress the fact that language
ideologies are idealizations or conceptualizations. Recent literature on language ideologies also stresses
their iconicity (Irvine and Gal (2000) 37). But, as I wish to show, iconicity should be captured by
iconic means, such as ‘mental maps’. Language representations are images, depictions, illustrations of
language, therefore representations of something that cannot be pictured – except, perhaps, in writing,
which might very well be the reason why language ideologies are so preoccupied with the written form
of a language. I understand ‘iconicity’ through the notion of ‘mental maps’ or ‘mental spaces’, which
are tools employed mostly in cognitive linguistics. But see Preston (1989) and Niedzielski and Preston
(2000) 45–96, for surveys of research on perceptual dialectology, which is perhaps the most interesting
application of mental maps to language ideology; see also Gärdenfors (2000) on what could be a general
If the indexes of the language issues are distributed over newspapers, genres
and periods of time, the issues themselves are distributed over a conceptual space
through which language is pictured in one way or another. In our case, language
is pictured as a territory, as a ‘regime’ or a state. The issues arise only within this
‘regime of language’, which is necessarily a regime of language ideology. The
regime ideology of the Modern Greek language is represented in Figure 2:
Figure 2
The conceptual map of a language ideology
an Interior Greek spreads

within the Interior
Exterior
an Exterior
within the
Exterior Greek threatened Interior
According to this regime ideology, language can be defined by associating an

Interior to an Exterior of the language. Interior and Exterior are relative terms,
but the Interior is pretty much like a ‘vantage point’46 from which this relationship
can be defined. The Interior of Greek is (say) Greek per se, or ‘pure Greek’. The
Exterior of Greek is ‘non-Greek’, consisting of (say) English or some brand of
‘mixed’ or ‘improper’ Greek (such as youth slang). Loans are typically regarded as
belonging to the Exterior of a language. The Interior is conceived as ‘surrounded’
by the Exterior. Their relationship is a dynamic one. Either one can expand or
contract under pressure from the other. Either one can also form a duplicate of itself
within the other, creating an Exterior-within-the-Interior or an Interior-within-
the-Exterior. In the first case, Greek is pictured as ‘threatened’; in the second
case, Greek is said to ‘spread’ or to have ‘conquered’ the Exterior. Generally, the
Exterior is pictured as a ‘threat’ to the Interior. The expansion or reduplication of
both the Interior and the Exterior can be either ‘real’ or ‘symbolic’. It can be as
‘real’ as the learning of the Greek language by foreigners; or it can be as ‘symbolic’
theory of conceptual spaces. The Interior/Exterior dialectic presented here was, in part, inspired by
Antoine Culioli’s theory of notional domains: Culioli (1995).
46
MacLaury (2002).
as an expansion of the language’s authority. Figure 2 represents the conceptual map

of this ideology of the Modern Greek language. It is the pictorial representation of
the idea of a language as a realm or regime. It is the picture of a picture.
With the help of Figure 2, one can now confront ‘the most difficult question of
all’: ‘why the panic?’,47 or, in more general terms, ‘why those issues?’, and why
do they become issues at all, worth putting on the busy agenda of the media? My
answer to this question is that language issues are defined as such only through
a language ideology. It is clear now that all the issues raised in Greece after the
resolution of the perennial diglossia situation can be defined in accordance with
this new regime ideology of the Modern Greek language. In Figure 3, all the
issues of Table 1 are placed on the conceptual map of this ideology. Issues can be
now classified in conceptual terms into four categories: (a) issues relating to the
Interior, (b) issues relating to the Exterior, (c) issues relating to an Exterior which
has intruded into the Interior, and (d) issues relating to the expansion of the Interior
into the Exterior. As we shall see shortly (Figures 4 and 5), the four categories are
also related chronologically.
Figure 3
Language issues placed on a conceptual map
Greek abroad (10) ‘word poverty’ (6)

in Cyprus (12) post-diglossia issues (7)
orthography (9)
usage (12)
Greek spreads
an Interior
within the Interior
Exterior
an Exterior
within the
Exterior Greek threatened Interior
‘Macedonian’ (4) bilingualism (1)

‘five-language diglossia (7)
regime’ (5) Romanization (2,9)
foreign words (8)
‘Bulgarians’ (3)
minorities (11)

47
Johnson (1999) 5.
Let us consider two straightforward examples, relatively easy ones: the names
‘Macedonian’ and ‘Bulgarians’. We have here two clear cases of moral panic
focusing, at least in part, on the metalinguistic issues raised by the use of these
names. In the way the issues have been presented and debated, the trained linguist
would probably uncover a primitive or mystical faith in nominal essentialism, an
unjustifiable belief in the non-arbitrariness of names. But the issues could not be so
easily dispensed with. They lasted for months. They were related to the national
interest, to the foreign and internal policies of the state, to issues of identity. In
both cases, the issue, to be sure, was not conventionalism. The issue was about
whatever all the other issues were about. Any issue is sustained as such only
through a uniform framework, through a conceptual map that relates it to all the
rest. If ‘our language’ is a uniform, pure language (an Interior) that has to be kept
intact by everything surrounding it (an Exterior), then a name of ‘our language’
(‘Macedonian’) cannot be given over to the enemy, because the name is ‘ours’. In
similar terms, we cannot allow a foreign name (‘Bulgarians’) to name us; such a
name cannot appear in a dictionary (an esteemed and authoritative work) whose
aim is to define our language. That would be equivalent to using our language (the
uniform Interior) in ways that divide it, and this cannot be allowed. ‘Macedonian’
cannot be a name in the Exterior. ‘Bulgarians’ cannot be a name in the Interior.
Clearly, both issues presuppose the conceptual map of Figure 2 and, with it, the
whole topology of the regime ideology.
That all the issues are somehow related is evident not only topologically but also
chronologically. The issues occur in a semi-chronological order, forming a more or
less coherent media narrative. Figure 4 displays the four phases of this narrative:
Figure 4
Language issues as a media narrative: Phase A
‘monotonic’ orthography
an Interior teaching Ancient Greek
within the Interior
Exterior the ‘language problem’
‘word poverty’
an Exterior
within the
Exterior Interior
foreign words – English

Language issues as a media narrative: Phase B
within the Interior
‘word poverty’
an Exterior
within the
Exterior Interior
‘five-language regime’ foreign words – English

‘Macedonian’
Language issues as a media narrative: Phase C
within the Interior
‘word poverty’
minorities – Greek as
a second
anlanguage
Exterior
Englishwithin
officialthe
language
Exterior ‘Romanization’
Interior
‘Bulgarians’
‘Macedonian’
Language issues as a media narrative: Phase D
within the Interior
Greek as a foreign language ‘word poverty’
minorities – Greek as
a second
anlanguage
Exterior
Englishwithin
officialthe
language
Exterior Interior
‘Romanization’
‘Bulgarians’
‘Macedonian’
Phase A mainly involves the immediate post-diglossia issues. They are all
related to several, often conflicting, efforts, after years of divided usage, to define a
uniform, standard language (an Interior). The post-diglossia issues do not so much
arbitrate between demotic and katharevousa or seek to decide which of the two
belongs to the Interior (that was the issue of the ‘language question’); rather, they
are concerned with the extent to which each of the two belongs to the Interior.
Immediate post-diglossia issues push the question of the synchronic uniformity of
the Standard Language to the background. The issue that is being foregrounded
now is whether the Interior is diachronically uniform. The ‘language problem’, the
debate about whether Ancient Greek should be taught in secondary education, the
reservations about orthographic reforms which ‘detach’ the present language from
its glorious past, are all manifestations of this new concern. Late in this period, the
Exterior is entering the conceptual horizon of the new regime ideology through the
issues of language contact and loanwords. English is identified as the enemy of the
state, of the Greek language in its fight to become uniform.
Phase B focuses still more on the threatening Exterior, which is now being
uncovered: in the Exterior lurk the usurpers of our names (fyrom) and the enemies
of our language (the EU).
This uncovering of the Exterior as an area occupied by the enemy gives rise to
several issues concerning the status of the Exterior within the Interior. Emphasis,
once again, shifts to the ‘internal front’.48 In Phase C, we first have the ‘Bulgarians’,
a threat that can divide the internal front, and then the danger of Romanization,
which should unite all against the common enemy, English. And then, suddenly,
comes the ‘inconceivable’: the proposal in favour of English as an official
language.
48
Delveroudi and Moschonas (2003) 9–12.
Together with the realization of an Exterior-within-the-Interior comes the

discovery of a new means of dealing with the Interior’s uniformity: Greek as a
second language. The assimilating power of Greek is coupled with the discovery
that Greek can also be taught to minorities and foreigners in order to expand the
borders of the state; Greek can itself become an Interior-within-the-Exterior. We
have reached the stage of Greek as a foreign language (Phase D).
One can only hypothesize about the prospect of a new phase in this developing
scenario and hope that the renewed discovery of Greek through the lenses of a
foreign language may lead, this time with significant support from the professionals,
to a more sophisticated redefinition of the ‘Interior’.
The four Phases, which largely coincide with the four conceptual categories of
issues marked in Figure 2, are summarized again in Figure 5. It should be observed
that the four phases form chains: <A B><C D><D ?>. Language ideological
(metalinguistic) stages are just like linguistic ones.49 They are not uniform.
Ideologies do not develop in a completely straightforward manner. At each stage in
their development there is variation and confrontation. New sets of motifs emerge
through new encounters and redefinitions of the older issues:
Figure 5
The four phases of the media narrative
D ?
an Interior
within the Interior
Exterior D
A
C an Exterior
B within the
Exterior Interior
B C
The conceptual map of Figure 2 can also help us redefine the problematic
dichotomy (problematic, that is, as far as the language issues are concerned)
between news and opinion. Language news is mostly ‘foreign news’: thus, the
‘relinquishing’ of a name with a symbolic status (‘Macedonia’) to a ‘foreign enemy’
will be seen as newsworthy; the ‘threats’ facing the Greek language in the political
49
Saussure (1983) 178–81.
environment of the EU will be reported as ‘news’; likewise, initiatives taken

for teaching Greek as a second language to foreigners, either in Greece (in the
Interior) or abroad (in the Exterior), will also be considered newsworthy. Opinion
articles, on the other hand, are mainly concerned with the Interior’s constitution,
its composition, and its form. All this is summed up in Figure 6:
Figure 6
The news–opinion dichotomy
Interior
an Interior NEWS
within the
Exterior OPINION
an Exterior
within the
Exterior NEWS Interior
Finally, we can now define the case of moral panic with much more precision,
with respect to a particular language-ideology conceptual framework. Moral
panics are also associated with foreign news. A moral panic, within this particular
ideological framework, can be identified as the communicative sequence provoked
by an item of linguistic news that reports on the expansion of the Exterior towards
the Interior (see Figure 7). A moral panic is the declaration of a linguistic war. This,
of course, is a metaphorical statement, and one which cannot stand as a general
definition of moral panics, nor even as a general definition of moral panics about
language. In different ideological frameworks, moral panics will be focused on
quite different threats.
Figure 7
The case of moral panic
Interior
an Interior
within the
Exterior
an Exterior
within the
Exterior Interior
MORAL PANIC
Conclusions
In this chapter I have sought to show that the ‘regime ideology’ of the Modern
Greek language affects the ways seemingly disparate ‘language issues’ are defined
and covered in the press. The conceptual topology of this new ideology offers a
general framework for the development of coherent communicative sequences
involving several actors with distinct stances on the issues raised. I have adopted a
research procedure based on multiple indexing. I have used media coverage (and,
in particular, articles in the print media) as an index of language issues. Language
issues, both explicit and implicit, are employed, in turn, to index a comprehensive
ideological framework, within which both consensus and dissent are exercised.
This conceptual ideological framework can then be assumed to affect or to control
certain linguistic practices pertaining to standardization, such as the collective
practice of purism.50 My research procedure is summarized in Figure 8:
50
On the affective aspect of language ideologies, see Silverstein (1979) 231–4; Silverstein (1985);
Woolard (1998) 10–11; Moschonas (2005a) 165–7.
Figure 8
Multiple indexing
Linguistic practices
Ideological framework
moral panics routine issues
erased issues
Language issues
media genres
newspapers
Media coverage
Once the issues are associated and combined, one can analyse their hidden
presuppositions, their stereotypes, their rhetorical strategies and their traces in
unguarded discourse. Language ideologies are riddled with clichés, prejudices,
myths and idées reçues about language,51 which can be analysed as presuppositions
shared by members of a linguistic community;52 ideologies are often implicit53
and are semantically organized on the basis of conceptual metaphors, such as
the territorial metaphor of language.54 To a large extent, deciphering a language
ideology is tantamount to exposing its surreptitious or unavowed stereotypes,
such as the ‘myth of the uniformity of language’, the understanding of synchronic
identity through historical continuity (evidenced in all etymology-based
orthographic systems), a belief in nominal essentialism, and the conception of
language contact in confrontational terms. One can also associate such widespread
stereotypes with social hierarchies, power, prestige and mechanisms of the state.
A less obvious correlation is the one between language standards and the specific
linguistic practices collectively exercised within a linguistic community.
The ‘regime’ conception of a standard language can be shown to affect, or to
have been affected by, a change in language standards. It is interesting that the
prescriptive standardization formula of late demoticism (‘Standard Modern Greek
= demotic + learned forms’) has now assumed the status of a descriptive principle
51
Yaguello (1988); Bauer and Trudgill (1998).
52
Preston (2004) 87.
53
Tsitsipis (2003).
54
For analyses of conceptual metaphors, see also the studies collected in Dirven, Hawkins and
Sandikcioglu (2000) and Dirven, Frank and Ilie (2001). Spitzmüller (2005) 204–49 provides an
excellent analysis of the metaphorical discourse on purism.
and prevails in reference works about Modern Greek. As I have argued elsewhere,55
proper usage is now prescribed by a new elite of ‘craft professionals’ (linguists,
writers, editors, proof-readers) who place emphasis on ‘syntax’, phraseology and
idiomaticity, irrespective of any morphological differences between demotic and
katharevousa. Finally, linguistic purism has long ceased to be diglossic and has
become bi-glossic; it is being practised on a massive scale, mainly against English
loanwords; and it is manifested in the press as an outward rejection of bilingualism
and language contact.
As many linguists have observed, katharevousa continues to maintain a
shadowy but by no means unimportant existence alongside demotic. Archaisms
or ‘learned forms’ appear in many registers of Standard Modern Greek. The ‘spirit
of katharevousa’ survives in purism. My presentation of the language issues that
have followed the language question suggests that katharevousa does not exist
merely on a linguistic plane: katharevousa has also shaped the metalanguage of
the new standard. In an important sense, katharevousa survives as an ideology of
the demotic. By this I do not mean only that the demotic (or, rather, its successor,
Standard Modern Greek) now adopts a puristic (katharevousa) attitude and
practice. I mean above all that the language issues raised in Greece after the
language question would be inconceivable outside an ideology of the standard
language unrelated to the ideology of katharevousa.
References
Althusser, L. (1971), ‘Ideology and ideological state apparatuses (notes towards an
investigation)’, in Lenin and Philosophy and Other Essays, tr. B. Brewster, London: 121–
73.
Androutsopoulos, J. and Iordanidou, A. (1999), ‘“Πήρανε τη γλώσσα στο . . . κρανίο”:
Στάσεις των ΜΜΕ απέναντι στη γλώσσα των νέων’ [‘“Language . . . gets on their
nerves”: media attitudes towards youth slang’], in A. Mozer (ed.), Greek Linguistics ’97:
Proceedings of the 3rd International Conference on the Greek Language, Athens: 586–95.
Babiniotis, G. (1994a), Ελληνική γλώσσα: Παρελθόν, παρόν, μέλλον [Greek Language:
Past, Present, Future], Athens.
Babiniotis, G. (1994b), Η γλώσσα ως αξία: Το παράδειγμα της ελληνικής [Language as
Value: The Example of the Greek Language], Athens.
Babiniotis, G. (1994c), Παιδεία, εκπαίδευση και γλώσσα: Εκτιμήσεις και προτάσεις
[Culture, Education and Language: Appraisals and Proposals], Athens.
Babiniotis, G. (1998), Λεξικό της Νέας Ελληνικής Γλώσσας [Modern Greek Dictionary],
Athens.
Bauer, L. and Trudgill, P. (eds.) (1998), Language Myths, London.
55
Moschonas (2001b).
Blommaert, J. and Verschueren, J. (1998), ‘The role of language in European nationalist

ideologies’, in B.B. Schieffelin, K.A. Woolard and P.V. Kroskrity (eds.), Language
Ideologies: Practice and Theory, New York and Oxford: 189–210.
Blommaert, J. (ed.) (1999), Language Ideological Debates, Berlin and New York.
Bourdieu, P. (1991), Language and Symbolic Power, ed. J.B. Thompson, tr. G. Raymond and
M. Adamson, Oxford.
Cameron, D. (1995), Verbal Hygiene, London and New York.
Chartoulari, M. (2001), ‘Τα “γρίκλις”, η σύγχυση, η διαμάχη και η αλήθεια’ [‘“Greeklish”,
the confusion, the debate, and the truth’], Τα Νέα, 16 January 2001.
Christidis, A.-F. (ed.) (1999), Dialect Enclaves of the Greek Language, Thessaloniki.
Culioli, A. (1995), Cognition and Representation in Linguistic Theory, ed. M. Liddle, tr. J.T.
Stonham, Amsterdam and Philadelphia.
Delveroudi, R. and Moschonas, S.A. (2003), ‘Le purisme de la langue et la langue du
purisme’, Philologie im Netz, 24: 1–26.
Dirven, R., Frank, R. and Ilie, C. (2001), Language and Ideology, vol. 2: Descriptive Cognitive
Approaches, Amsterdam and Philadelphia.
Dirven, R., Hawkins, B. and Sandikcioglu, E. (eds.) (2000), Language and Ideology, vol. 1:
Theoretical Cognitive Approaches, Amsterdam and Philadelphia.
Droumboukis, D. (n.d. [c. 1985]), Στόχος: Ελληνική Γλώσσα. Σχέδιο εν εξελίξει. Φως
στη συνωμοσία [Target: The Greek Language. A Plot in Progress. The Conspiracy Brought
to Light], Athens.
Eagleton, T. (1991), Ideology: An Introduction, London and New York.
Elephandis, A. (1998), ‘Η μονοτονική “επανάσταση”’ [orig. 1982], in Διά γυμνού
οφθαλμού [By the Naked Eye], Athens: 379–86.
Frangoudaki, A. (1997), ‘The metalinguistic prophesy on the decline of the Greek language:
its social function as the expression of a crisis in Greek national identity’, International
Journal of the Sociology of Language, 126: 63–82.
Gärdenfors, P. (2000), Conceptual Spaces: The Geometry of Thought, Cambridge, Mass.
Geertz, C. (1973), ‘Ideology as a cultural system’, in The Interpretation of Cultures, New
York: 193–233.
Goode, E. and Ben-Yehuda, N. (1994), Moral Panics: The Social Construction of Deviance,
Oxford.
Gotsis, A.E. (1997), Η ελληνική γλώσσα υπό διωγμόν [The Persecution of the Greek
Language], Athens.
Greek Language Association (1984), Ελληνική γλώσσα: Αναζητήσεις και συζητήσεις
[The Greek Language: Searches and Discussions], Athens.
Greek Language Association (1986), Ελληνική γλώσσα: Αναζητήσεις και συζητήσεις
[The Greek Language: Searches and Discussions], Athens.
Irvine, J.T. and Gal, S. (2000), ‘Language ideology and linguistic differentiation’, in P.V.
Kroskrity (ed.), Regimes of Language: Ideologies, Polities, and Identities, Santa Fe, NM,
and Oxford: 35–83.
Johnson, S. (1999), ‘From linguistic molehills to social mountains? Introducing moral panics
about language’, Lancaster University, Centre for Language in Social Life, Working
Paper 105 [http://www.ling.lancs.ac.uk/pubs/clsl/wpapers.htm].
Johnson, S. (2001), ‘Who’s misunderstanding whom? Sociolinguistics, public debate and
the media’, Journal of Sociolinguistics, 5: 591–610.
Karyolemou, M. (1994), ‘Linguistic attitudes and metalinguistic discourse: an investigation
in the Cypriot press’, in I. Philippaki-Warburton, K. Nicolaidis and M. Sifianou (eds.),
Themes in Greek Linguistics: Papers from the First International Conference on Greek
Linguistics, Reading, September 1993, Amsterdam and Philadelphia: 253–9.
Karyolemou, M. (2000), ‘“Ne touchez pas à mon dialecte’: Normalization des noms
géographiques et sailance des variables à Chypre’, Journal de l’Association Canadienne de
Linguistique Appliquée, 3: 91–106.
Kostopoulos, T. (2000), Η απαγορευμένη γλώσσα: Κρατική καταστολή των σλαβικών
διαλέκτων στην Ελληνική Μακεδονία [Forbidden Language: State Prohibition of the
Slavic Dialects in Greek Macedonia], Athens.
Kriaras, E. (1979), Άρθρα και σημειώματα ενός δημοτικιστή [Articles and Notes by a
Demoticist], Athens.
Kriaras, E. (1984), Η σημερινή μας γλώσσα: Μελετήματα και άρθρα [Our Language of
Today: Studies and Articles], Thessaloniki.
Kriaras, E. (1988), Τα πεντάλεπτα μου στην ΕΡΤ και άλλα γλωσσικά [My Five-Minute
Broadcasts on Television and Other Linguistic Essays], Thessaloniki.
Kriaras, E. (1992), Η γλώσσα μας: Παρελθόν και Παρόν [Our Language: Past and Present],
Thessaloniki.
Kriaras, E. (2001), ‘Η φοβία για το αλφάβητο’ [‘The phobia over the alphabet’], Η
Καθημερινή, 11 February 2001.
Kroskrity, P.V. (2000), ‘Regimenting languages’, in P.V. Kroskrity (ed.), Regimes of Language:
Ideologies, Polities, and Identities, Santa Fe, NM, and Oxford: 1–34.
Landsman, D.M. (1989), ‘The Greeks’ sense of language and the 1976 linguistic reforms:
illusions and disappointments’, Byzantine and Modern Greek Studies, 13: 159–82.
MacLaury, R.E. (2002), ‘Introducing vantage theory’, Language Sciences, 24: 493–596.
Maronitis, D.N. (1995), ‘Υστερία’ [‘Hysteria’], Το Βήμα, 8 January 1995.
Milroy, J. and Milroy, L. (1999). Authority in Language: Investigating Standard English, 3rd
edn, London.
Milroy, J. (2001), ‘Language ideologies and the consequences of standardization’, Journal
of Sociolinguistics, 5: 530–55.
Moschonas, S.A. (2001a), ‘Δημοσιεύματα του Τύπου για τη γλώσσα’ [‘Publications
about language in the Greek press’], in P. Boukalas and S.A. Moschonas (eds.),
Δημοσιογραφία και γλώσσα. Πρακτικά συνεδρίου (Αθήνα, 15–16 Απριλίου 2000)
[Journalism and Language: Conference Proceedings (Athens, April 15–16, 2000)], Athens:
85–116.
Moschonas, S.A. (2001b), «Οι διορθωτικές στήλες στον ελληνικό Τύπο» [‘Columns on
language usage in the Greek press’], Εφαρμοσμένη γλωσσολογία, 17: 49–68.
Moschonas, S.A. (2001c), ‘Οι ξενόγλωσσες ταμπέλες του Βόλου’ [‘Volos’ Roman-script
shop-signs’], Η Καθημερινή, 7 October 2001.
Moschonas, S.A. (2002a), ‘“Εθνική νίκη” της μονογλωσσίας;’ [‘A “national victory” for
monolingualism?’], Η Καθημερινή, 2 February 2002.
Moschonas, S.A. (2002b), ‘Τα γράμματα των αναγνωστών για τη γλώσσα’ [‘Letters to
the editor about language’], Η Καθημερινή, 7 July 2002.
Moschonas, S.A. (2004), “Relativism in language ideology: on Greece’s latest language
issues”, Journal of Modern Greek Studies, 22: 173–206.
Moschonas, S.A. (ed.) (2004/5), Όψεις της γλώσσας στον Τύπο [Aspects of Language in
the Greek Press], in Γλώσσα 59, 2004: 45–93 and Γλώσσα 60, 2005: 32–-88.
Moschonas, S.A. (2005a). «Διορθωτικές πρακτικές» [‘Corrective practices’], in Χρήσεις
της γλώσσας, Επιστημονικό συμπόσιο (3–5 Δεκεμβρίου 2004) [Uses of Language,
Scientific Symposium (3–5 December 2004)], Athens: 151–74.
Moschonas, S.A. (2005b), Ιδεολογία και γλώσσα [Ideology and Language], Athens.
Moschonas, S.A. (2006), «Υπεράσπιση του πολυτονικού» [‘A defence of the “polytonic”’],
Η Καθημερινή, 7 February 2006.
Niedzielski, N.A. and Preston, D.R. (2000), Folk Linguistics, Berlin and New York.
Poe, E.A. (2004) [orig.1848], Eureka: An Essay on the Material and Spiritual Universe, ed.
S. Levine and S.F. Levine, Chicago.
Preston, D.R. (1989), Perceptual Dialectology: Nonlinguists’ Views of Areal Linguistics,
Dordrecht.
Preston, D.R. (2004), ‘Folk metalanguage’, in A. Jaworski, N. Coupland and D. Galasiński
(eds.), Metalanguage: Social and Ideological Perspectives, Berlin and New York: 75–101.
Ricoeur, P. (1986), Lectures on Ideology and Utopia, ed. G.H. Taylor, New York.
Saussure, F. de (1983), Course in General Linguistics, tr. R. Harris. London.
Silverstein, M. (1976), ‘Shifters, linguistic categories, and cultural description’, in K.H.
Basso and H.A. Selby (eds.), Meaning in Anthropology, Albuquerque: 11–55.
Silverstein, M. (1979), ‘Language structure and linguistic ideology’, in P. Clyne, W. Hanks
and C. Hofbauer (eds.), The Elements: A Parasession on Linguistic Units, Chicago: 193–
247.
Silverstein, M. (1985), ‘Language and the culture of gender: at the intersection of structure,
usage and ideology’, in E. Mertz and R.J. Parmentier (eds.), Semiotic Mediation, Orlando,
Fla.: 219–59.
Silverstein, M. (1993), ‘Metapragmatic discourse and metapragmatic function’, in J.A. Lucy
(ed.), Reflexive Language: Reported Speech and Metapragmatics, Cambridge: 33–58.
Silverstein, M. and Urban, G. (1996), ‘The natural history of discourse’, in M. Silverstein
and G. Urban (eds.), Natural Histories of Discourse, Chicago and London: 1–17.
Spitzmüller, J. (2005), Metasprachdiskurse: Einstellungen zu Anglizismen und ihre
wissenschaftliche Rezeption, Berlin and New York.
Thomas, G. (1991), Linguistic Purism, London and New York.
Thompson, K. (1998), Moral Panics, London and New York.
Triandaphyllidis, M. (1981) [1st edn, 1938], Νεοελληνική Γραμματική. Ιστορική εισαγωγή
[Modern Greek Grammar: An Historical Introduction], Thessaloniki.
Triandaphyllidis, M. et al. (1978) [1st edn, 1941], Νεοελληνική Γραμματική (της
Δημοτικής) [Modern Greek Grammar (of the Demotic)], Thessaloniki..
Tsikopoulou, C. (1995), H μονοτονική γραφή, ο σύγχρονος Δούρειος Ίππος [Monotonic
Writing, Today’s Trojan Horse], Athens.
Tsitsipis, L.D. (2003), ‘Implicit linguistic ideology and the erasure of Arvanitika (Greek-
Albanian) discourse’, Journal of Pragmatics, 35: 539–58.
Tsengos, I.K., Papadakis, Th.N. and Vekiari, D. (2005), Η εκδίκηση των τόνων: Η επίδραση
των «Αρχαίων Ελληνικών» και του «Μονοτονικού» στην ψυχοεκπαιδευτική εξέλιξη
του παιδιού – Συγκριτική μελέτη [Revenge of the Accents: The Influence of ‘Ancient Greek’
and ‘Monotonic’ on the Psycho-Educational Development of the Child], Athens.
Vrahnias, K.K. (1992), Μονοτονικό: Η μία πράξη του γλωσσικού μας δράματος
[Monotonic: One Act in Our Linguistic Drama], Athens.
Woolard, K.A. (1998) ‘Introduction: language ideology as a field of inquiry’, in B.
Schieffelin, K.A. Woolard and P.V. Kroskrity (eds.), Language Ideologies: Practice and
Theory, Oxford and New York: 3–47
Yaguello, M. (1988), Catalogue des idées reçues sur la langue, Paris.
16
Competing Ideologies and Post-Diglossia Greek:

Analysing the Discourse
of Contemporary ‘Myth-Breakers’*
Dionysis Goutsos
This chapter focuses on the most recent episode in the long history of meta-
linguistic theorizing on Greek. It attempts an analysis of the discourse employed
in a body of work by a number of contemporary Greek scholars (mainly linguists),
with particular reference to the question of the continuity of the Greek language
and its ideological use and abuse. This question has assumed central importance in
Greek metalinguistic discourse since the official demise of diglossia in Greece after
1974, which called for a revisiting of traditional preoccupations and concerns. I
would like to suggest that the contestation of ideologies over post-diglossia Greek
depends on conventional polarities and dichotomies, which hinder a renewed
understanding of the language in its contemporary dimensions.
I have chosen to focus on one particular discourse that has served as a response
to the purist or ‘ethnocentric’ view of the Greek language: the view that allegedly
seeks to reinstate the discredited katharevousa through the linking of Modern
Greek with its celebrated ancestor, Ancient or Classical Greek. This anti-purist
discourse can be identified in a body of work, which includes: collections of papers;
monographs; public interventions; individual papers; book reviews; reference
works and internet sites.1 These works, in my view, constitute a coherent corpus,
not only because of their timeframe (from the early ’90s up to the present) and the
*
By a tragic twist of fate, Professor Christidis, whose work, among that of others, is analysed here,
passed away during the revision of this chapter. In what was to be his last published paper – Christidis
(2004) – he briefly comments on my views as expressed in my presentation at the London Logos
Conference. It is my belief that the best contribution to his memory is the critical engagement with his
work in public debate. It is a matter of regret that he is not amongst us to participate in this dialogue.
1
Respectively: Christidis (1999a), Charis (2000); Frangoudaki (2001); Christidis (1999b);
Frangoudaki (1992), Christidis (2001a), (2001b); Christidis (2003), Frangoudaki (2003); Christidis
(2001c).
321
322 dionysis goutsos
identity of their authors (who have assumed a particular public stance and have
taken related official posts), but also because of their main line of argumentation.
The anti-purists2 wish to discard an ethnocentric view of the continuity of Greek
and the implications of superiority for its classical phase, along with the belief in
the concomitant inferiority of its present state, which is felt as a state of decline
from the language’s glorious past. They also seek to reveal the hidden interests of
the ethnocentric prescriptivists, realized in the sphere of education through the
re-introduction of Ancient Greek classes in secondary schools. Finally, they relate
the purists’ views to their point of origin, a traditionalist or nationalist reaction to
progressive and modernist changes in Greek society.
The advantage of choosing to focus on this anti-ethnocentric pole is that, despite
the differences between the various scholars studied here, there is a much clearer
and more coherent position in their writings than in ethnocentric views, which
seem more dispersed, less localized and, possibly because of this, more widespread.
It would be difficult to delineate the purist pole with any precision, because it
attracts a wide spectrum of ‘commonsense’ and scholarly opinion, ranging from an
emphasis on the peculiarities of Greek to worries about its fate, and from common
concerns with spelling and foreign loanwords, as expressed in ‘letters to the editor’
columns, to metaphysical beliefs in the language’s superiority. In addition to this
matter of convenience, I would like to note that the self-adoption of a critical stance
by the anti-purists has been taken to imply that it is somehow beyond criticism,
to the point that it is seen as setting ‘the terms for the debate’.3 This assumption
needs to be challenged if we are to achieve a more sophisticated view of the issues
involved.
At the same time, my decision to focus on the anti-ethnocentric pole of the
debate does not imply the endorsement of ethnocentric or traditionalist views
against the anti-prescriptivists. It is obvious that a comprehensive study would
require an analysis of the discourse of both positions, as well as a broader discussion
of context. The latter would include issues of geographical and political alignment,4
as well as links with wider intellectual debates, for instance debates on the concept
of nation, its meaning and its functions.5 However, such a study falls outside the
scope of this chapter, which can focus only on some aspects of the wider debate.
A final word of caution: I cannot and do not claim here that I am offering an
‘objective’ or ‘distanced’ analysis. I do not believe that any analysis can be divested
of the analyst’s personal stance and conditions of work. At the same time, though,
and in full awareness of the thin ice of Greek intellectual debate, I consider that
2
Moschonas (2000) uses a similar label for this discourse.
3
Christidis (2004).
4
As is relevant to, for instance, the age-old controversy between the Universities of Athens and
Thessaloniki, summed up by the debate between Professors Maronitis and Babiniotis in the columns of
Το Βήμα in 1993: see Babiniotis (1999). Cf. Christidis’ (2004) emphasis on the University of Athens in
his comment on my views.
5
A public debate on this can be followed in the columns of Το Βήμα in April/May 2004.
competing ideologies and post-diglossia greek 323
it is indeed possible to invoke standards of reason, evidence, grounds and specific

criteria,6 as these are employed in the practice of discourse analysis.
In particular, I follow analysts such as Stubbs and de Beaugrande,7 by combining
corpus-linguistics tools with critical discourse analysis in order to identify the
dominant themes and arguments of the discourse studied. In the following two
sections I shall examine the two central notions of myth and ideology in this
discourse and then show how these may be related to views on language and
history, before concluding with a personal view on how the ground for debate can
be cleared.
«Παντού υπάρχει ένας μύθος»: ‘there’s a myth everywhere’

An examination of the means by which the main position of the anti-purists is
argued in the material examined reveals a wide variety of labels to characterize the
opposite, purist, point of view. These labels range from the neutrally descriptive
(‘description’, ‘formulation’, ‘view’) to the negatively evaluative (‘error’,
‘syndrome’, ‘phantasm’, ‘distortion’, ‘prejudice’).8 Although the labels are widely
employed, along with other evaluative descriptions, as we shall see below, they do
not seem to be more than dismissive one-offs, since – though certainly significant
– they lack any sense of a consistent position behind them: often it is the same
practice which is variously described as a matter of false judgment, erroneous
argumentation, wrong attitude, purposeful manipulation, deceptive metaphor, self-
deception or some combination thereof.
If there is a dominant topos of the discourse under consideration, then this
surely centres on the concept of μύθος (‘myth’), used to characterize the opposing
position. The use of lexical items belonging to the word-family of μύθος in the
data is particularly repetitive and persistent. In a cursory analysis of the data in my
corpus, I have spotted sixty-eight such uses, including the derivatives μυθολογία
(‘mythology’), μυθικός (‘mythical’) and μυθοποίηση (‘mythologization’).9 This
usage is also made prominent by its promotion to the title of a collection of articles,10
where, however, it is interesting that only a few contributors employ an item from
this word-family. Instead, Moschonas, for instance, speaks more cautiously of
6
Cf. Toolan (1998) 98.
7
Stubbs (1996); de Beaugrande (1999), (2001).
8
The full list includes περιγραφή (Frangoudaki (2001) 151); διατύπωση (Christidis (1999a) 52);
άποψη (ibid. 54); έννοιες (ibid. 40, 83); επιχείρημα (ibid. (1999a) 11, 56 and (2001a) 84); set of
arguments (Frangoudaki (1992) 373); σύμπλεγμα ιδεών (Christidis (1999a) 79); στάση (ibid. 44
and Christidis in Charis (2000) 19); χειρισμοί (Christidis (2001a) 15); μεταφορές (Christidis (1999a)
83); πλάνη (ibid. 51, 56–7); σύνδρομο (Christidis (2001a) 15); εμμονή (Christidis (1999a) 49);
φαντασίωση (ibid. 64, 68); φάντασμα (ibid. 64); στρέβλωση (Maronitis in Charis (2000) 14);
προκαταλήψεις (Christidis in Charis (2000) 18–19); κινδυνολογία (Christidis (1999a) 51, 54–5, 89);
σενάριο (Christidis 2003), παραφιλολογία (ibid.).
9
In particular, the following number of types for the corresponding lemmas was found:
ΜΥΘΟΣ <21>, ΜΥΘΟΛΟΓΙΑ <15>, ΜΥΘΟΠΟΙΗΣΗ <9>, ΜΥΘΙΚΟΣ <7>, ΜΥΘΟΠΟΙΩ <6>,
ΜΥΘΟΠΟΙΗΤΙΚΟΣ <4>, ΜΥΘΟΛΟΓΙΚΟΣ <3>, ΜΥΘΕΥΜΑ <3>.
10
Charis (2000).
πεποίθηση (‘belief’), while Kakridi-Ferrari refers to εξωγλωσσικές συμβολικές

επενδύσεις (‘extra-linguistic symbolic investments’).11
This use of ‘myth’ is clearly idiosyncratic, as one can see from further data.
For instance, in a total of 95,000 words from Christidis’ Encyclopaedic Guide to
Language,12 there are twenty-five instances of the word-family ‘myth’ (2.6 per
10,000 words), of which only four occur in texts by contributors other than
Christidis.13 It is interesting to compare these figures with wider usage in several
genres of the language, taken from data in the Corpus of Greek Texts (henceforth
CGT).14 Table 1 compares the findings from three samples of CGT with those from
the Encyclopaedic Guide. The CGT samples consist of 500,000 words each, and
come from three different genres (academic writings, commentaries in newspapers,
and parliamentary minutes). The table shows the total number of occurrences of
the word-family ‘myth’, both in raw figures and adjusted per 10,000 words, as well
as the percentage of evaluative uses of the items in question:
Table 1
Instances of the μύθος word-family in CGT and the Encyclopaedic Guide
Academic Newspaper Parliamentary Encyclopaedic
writings commentaries minutes Guide
Raw instances 82 28 5 25
Adjusted per 1.6 0.56 0.1 2.6
10,000 words
Evaluative uses 9 (11 %) 11 (39 %) 5 (100 %) 25 (100%)
Apart from the higher frequency of the items in the Encyclopaedic Guide, as
compared to the CGT sample, it is interesting to note the discrepancy between
evaluative and descriptive uses of the items. In particular, descriptive uses that refer
to myth in the sense of ‘ancient story’ or ‘legend’ predominate in academic texts,
while evaluative uses in the sense of ‘false belief’ are the only uses in parliamentary
speeches.15 Overall, it would seem that the uses of ‘myth’ in the material under
analysis are closer to the rhetorical usage in parliamentary data from CGT than
to any other CGT genre, though in terms of total frequency the Guide is closer to
academic writings.
A qualitative analysis of these usages is also enlightening. Most of the twenty-
five evaluative instances identified are nominal or nominalized, while of the six
11
In Charis (2000) 31 and 33, respectively.
12
Christidis (2001c).
13
In particular, he uses ΜΥΘΟΠΟΙΗΣΗ <11>, ΜΥΘΙΚΟΣ <4>, ΜΥΘΟΣ <2>, ΜΥΘΟΛΟΓΙΑ
<2>, ΜΥΘΟΠΟΙΩ <1>, ΜΥΘΟΠΟΙΗΤΙΚΟΣ <1>.
14
The Corpus of Greek Texts is a core collection of Modern Greek texts stored in electronic form,
which has been designed to be representative of basic genres in the language and will be available for
linguistic analysis and pedagogical applications. More details can be found in Goutsos (2003).
15
There are only two instances of the use of ‘myth’ in the former, descriptive sense in our corpus as
a whole, both in Christidis (2001b) 25–7.
verbal uses there is only one in the active (η αγορά [. . .] μυθοποιεί τα προϊόντα
της: ‘the market [. . .] mythologizes its products’). What this seems to imply is
that the word-family is used for characterizations that are taken for granted and
are not to be analysed in their constitutive functions and roles. This is supported
by a comparison with evaluative uses in the samples from CGT, where the most
frequent phrases used all refer to the process of myth-building or myth-breaking.16
These patterns are absent from our material, with the effect that agents and roles
are concealed and, furthermore, words are related to myth by mere derogatory
labelling, without further ado.
What information can we obtain from an analysis of this word-family in its
context? First of all, mythologization can be applied to all kinds of objects, from the
decline of Modern Greek, the particularities of the language, its linguistic past, the
ancient world and Greek culture, to the nation, technology and technologization,
multilingualism, the hegemony of English, ‘powerful’ foreign languages, market
commodities, language diversity and the ‘Other’. It is not clear how such diverse
items can equally be subjected to the same process of ‘mythologizing’ (or is it
‘mystification’ that is intended in some cases?).
In addition, myth is understood in a number of diverse, if not divergent, ways.
On the one hand, it stands for a haphazard erroneous belief in a ‘coherent ideology’
or ‘philosophy’: οι μύθοι αυτοί, όταν δεν αποτελούν συγκροτημένη ιδεολογία
(‘these myths, when they do not constitute a coherent ideology’);17 Ολόκληρη
η φιλοσοφία – ή, καλύτερα, η μυθολογία – του καθαρευουσιανισμού (‘The
entire philosophy – or rather, mythology – of affiliation with katharevousa’).18
On the other hand, it is stressed that there is an ‘inner logic’ or a ‘deeper’ and
‘concealed’ content (το βαθύτερο περιεχόμενο – και η βαθύτερη λογική –
του μύθου: ‘the deeper sense – and the deeper logic – of the myth’),19 which is
ultimately ideological.20
In addition, myth is sometimes seen to take its meaning from its opposition
to history, while, in particular, mythologizing is defined with respect to de-
mythologizing. This applies to both language and nation:
16
In particular: γκρεμίζει μύθους χαλκευμένους, καταρρίφθηκε ο μύθος, μυθολογία που
«σκαρώνουν» οι πολιτικοί, έτρεφε με μύθους (newspaper commentaries); αποτελούσαν μύθο,
καλλιεργήθηκε ο μύθος, η κατάρριψη του μύθου, οικοδομούν διάφορους μύθους (parliamentary
minutes); μυθοποιούνται και αποκτούν τα χαρακτηριστικά μιας αυτόνομης οντότητας,
μυθοποίηση και ωραιοποίηση, αποκρύπτει και μυθοποιεί, χρησιμοποιεί αρνητικά, μυθολογικά
(academic writings).
17
Charis in Charis (2000) 13.
18
Christidis (1999a) 63.
19
20
See also ibid. 83: το νόημα του καταστροφολογικού αυτού μύθου δεν αποκαλύπτεται από
το εμφανές του περιεχόμενο (αυτό ισχύει γενικά για τους μύθους) αλλά από το αφανές (‘the
meaning of this eschatological myth is not revealed in its manifest content (as is generally true of myths)
but in its latent one’). Cf. Maronitis in Charis (2000) 14.
Η μόνη πηγή για να γίνει κατανοητό το τεράστιο ιδεολογικό βάρος του έθνους,
ώστε να πάψει να είναι μύθος, είναι η ιστορία της δημιουργίας του.
The only source that can allow us to understand the huge ideological weight of nation,
so that it ceases to be a myth, is the history of its creation.21
Christidis also systematically opposes the two concepts: ιστορική και όχι μυθική
(‘historical and not mythical’).22 We shall return to the conception of history in
due course.
Finally, the mythical is opposed to the scientific, as a backward phase in the
latter’s development; thus ‘pre-scientific’ equals ‘mythological’ (στον χώρο
της προ-επιστημονικής γλωσσικής μυθολογίας: ‘in the area of pre-scientific
linguistic mythology’).23 The explanatory framework here seems to be the story of
a triumphant progression from myth to science in the age of modernism:
η νεότερη γλωσσολογία, όπως αναδύθηκε από τον αιώνα της επιστήμης [. . .]
κατεδάφισε παλαιότερες γλωσσικές μυθολογίες και απέσπασε έτσι σημαντικές
όψεις της γλώσσας από την επικράτεια του μύθου.
modern linguistics, as it emerged in the era of science [. . .], demolished earlier

linguistic mythologies and thus extricated important aspects of language from the
prevalence of myth.24
According to Frangoudaki, likewise, the rules of science can and should be applied
in our age to diminish the lingering power and attraction of mythology.25
In sum, the role of myth in the data studied is not made explicit; the notion
seems to be used, instead, as a blanket derogatory term or as shorthand for
everything the anti-purists are opposed to. The origins of this usage are a matter
for further investigation, since they seem to lie both in Greek linguistics (in the
work of Triantafyllidis, as suggested by Moschonas) and in critical theory (Roland
Barthes’s Mythologies comes to mind).26 What is profoundly problematic, however,
is the assumption that the invocation of myth can stand in for a detailed analysis
of such a supposedly objectionable phenomenon as the prestige of classical Greek.
Does myth offer an appropriate conceptual handle for this phenomenon, and,
if so, does it apply equally well (and in the same manner) to other comparable
phenomena such as the international hegemony of English?
The use of a host of additional labels also confuses the issue. Frangoudaki, for
instance, speaks of ‘a metalinguistic theory [. . .] that the Greek language is currently
21
22
Christidis (1999a) 106. See also ibid. 13, 42, 95.
23
Ibid. 80; cf. ibid. 81, 83.
24
Ibid. 3.
25
26
Moschonas (2000). In this respect, Charis (2000) can be usefully compared to Bauer and Trudgill
(1999), who, however, deal with common misconceptions about language such as ‘Children can’t speak
or write properly any more’ or ‘Italian is beautiful, German is ugly’.
in decline’, which ‘has taken hold of large segments of society’.27 Are theory and
myth interchangeable, as seems to be implied by the rest of her discussion?28 Is
there a necessary ‘rational kernel within the mythical shell’, as Eagleton would put
it,29 and, if so, what is it? In a different context, Christidis suggests that the main
problem in the confrontation with the hegemonic ‘Other’ (English in this case) is its
‘non-systemic’ quality (Η αντιπαράθεση με τους ισχυρούς «άλλους» δεν έχει
«συστημικό» χαρακτήρα),30 which reflects an insufficient recognition of socio-
historical parameters in favour of some cultural opposition (English versus Greek,
West versus East). In other words, he would argue, cultural opposition is a false
antithesis, which should be reduced to the systemic progress of science and history.
For Christidis, the ultimate success of science and history would be guaranteed, if
only there were no ideological obstacles here as well.
Ideology, or what the other person has

We have thus reached the second basic topos of the anti-purist discourse, ideology.
Two of the related labels used for the purist position, which were not mentioned
above, are ιδεολόγημα and προϊδεασμοί.31 What seems to be important in both is
not so much their specific meaning (or their exact translation into English, for that
matter) as their general association with ideology. The word-family of ιδεολογία
(‘ideology’), though considerably smaller than that of μύθος (‘myth’), is also
prominent in the writings of the anti-ethnocentric scholars.32
With regard to the word-family’s uses, we may note that most of the collocations
involved are negative. For instance, we find the negative labels προκατάληψη
(‘prejudice’), πλάνες (‘fallacies’), παρεξήγηση (‘misunderstanding’), alongside
the more neutral στράτευση (‘enlistment’), περιπέτειες (‘adventures’), έρεισμα
(‘foothold’), βάρος (‘weight’). However, what we find, above all, are adjectives
and adverbs of negative evaluation: φορτισμένο (‘loaded’), ιδιαίτερα επικίνδυνα
(‘particularly dangerous’), υπόρρητη (‘insidious’), ακραία (‘extremely’), βίαια
(‘violently’). This constellation of words sets up a predominantly negative semantic
prosody for the term ‘ideology’: the term is invested with negative associations and
connotations, as reflective of negative assumptions.33
The evidence would suggest that the notion of ideology in the anti-purist
discourse is not far removed from the view of the ‘person-in-the street’, as
Eagleton puts it, which equates ideology with an attitude imposed ‘schematically,
stereotypically, and perhaps with the faintest hint of fanaticism’.34 The obvious
27
28
For instance, ibid. 374: ‘the decline myth’.
29
Eagleton (1991) 200.
30
31
Respectively: Christidis (1999b), Maronitis in Charis (2000) 14; Christidis (2001a) 15.
32
Thus in the material studied we find: ΙΔΕΟΛΟΓΙΚΟΣ <20>, ΙΔΕΟΛΟΓΙΑ <7>, ΙΔΕΟΛΟΓΗΜΑ
<4>.
33
Stubbs (1996) 172; Partington (1998) 65–6.
34
Eagleton (1991) 3.
problem with this view, to cite Eagleton’s analysis again, is that ideology ‘like
halitosis, is in this sense what the other person has’.35 There is further evidence for
this in some particular examples which contrast an ‘ideological’ view with ‘how
things are’:
Η κινδυνολογία, ασύστατη γλωσσολογικά, έχει μονάχα ιδεολογικό
περιεχόμενο.
This talk about dangers, which has no foundation in linguistics, has only an ideological
content.36
Οι εφιάλτες της επερχόμενης γλωσσικής Αποκάλυψης [. . .] διεγείρονται από

λόγους καθαρά ιδεολογικούς και δεν έχουν κανένα έρεισμα τόσο στη γλωσσική
πραγματικότητα όσο και στην επιστήμη που τη μελετά – τη γλωσσολογία.
The nightmares of an impending linguistic Apocalypse [. . .] are aroused on purely

ideological grounds and have no basis either in linguistic reality or in the science that
studies it, linguistics.37
In these examples, ideology, like myth, is opposed to science (linguistics). Whoever

subscribes to the latter has privileged access to true understanding, by contrast
with the preconceived frameworks of ideology. In Eagleton’s words, ‘I view
things as they really are; you squint at them through a tunnel vision imposed by
some extraneous system of doctrine’.38 And in this case, ‘how things really are’ is
identified with linguistic phenomena per se:
[Οι ιδέες αυτές] αντλούν τη λογική τους, και τη δύναμή τους, όχι από τα
γλωσσικά φαινόμενα καθεαυτά, αλλά από άλλους χώρους – τους χώρους
της πολιτικής, της ιδεολογίας, της ψυχολογίας – και γι᾽ αυτό έχουν ένα
χαρακτηριστικά μυθικό χαρακτήρα.
[These ideas] draw their logic and their power, not from linguistic phenomena per se,
but from other areas – those of politics, ideology, psychology – and that is why they
have a typically mythological character.39
As can be seen from this example, there is an obvious concern with the delimitation
of space, the segregation of the valid scientific observation from the impure realm
(the ‘other areas’) of the political and the ideological. According to Christidis,
υπάρχει ένας κατακτημένος χώρος γνώσης και δεοντολογίας, που δεν είναι – ή
δεν είναι πια – συνάρτηση επιστημονικών ή ιδεολογικών διαφορών.
there is a subjugated territory of knowledge and ethics, which is not – or no longer
is – connected with scholarly or ideological differences.40
35
Ibid. 2.
36
37
Ibid. 60, cf. ibid. 40.
38
Eagleton (1991) 3.
39
40
Ibid. 111.
In the last resort, this pure space also takes its meaning from its opposition to the
impure space of mythology.41
In sum: as with the invocation of myth, this particular understanding of ideology
is related to the privileging of one’s own position, through its binary opposition to
a negatively evaluated stance. The positive pole of the opposition is identified with
science as a space that gives access to ‘some unequivocally correct way of viewing
the world’.42 This positivist belief in the role of science is evident in a number
of examples which talk about ‘linguistic knowledge’, ‘elementary principles of
linguistics’, ‘acquired theoretical knowledge’,43 ‘scientific substantiation’ and
‘modern scientific products’.44
Ideology is thus invoked to withdraw legitimacy from a position and to
legitimize, instead, the opposite pole, which relies on pure science rather than
mere ideology. It would again be interesting to trace the genealogy of this usage in
earlier structuralist versions of critical theory (for instance, Althusser’s) and ponder
their reverberations in Greek scholarship.45 What is more relevant is to examine
how this conception of ideology is incorporated into argumentative moves of self-
legitimization, as it is in one of the few qualifying statements by Christidis that
concedes some ground to the opposite view:
Για να αποφευχθεί η οποιαδήποτε παρεξήγηση, θα πρέπει να δηλώσουμε ότι
προφανώς αναγνωρίζουμε και σεβόμαστε το δικαίωμα του [. . .] να έχει τις
απόψεις του [. . .] Το ζήτημα βρίσκεται αλλού: στην αντίληψη ότι ένα είδος
λόγου με το οποίο διαφωνούμε [. . .] απειλεί τη νεοελληνική γλώσσα σαν
σύνολο.
Είναι ακριβώς αυτό το ανιστόρητο άλμα που μετατρέπει τη θεμιτή διαφωνία

σε αθέμιτη κινδυνολογία [. . .] δεν έχει βέβαια κανένα έρεισμα στα γλωσσικά
πράγματα [. . .] προδίδει [. . .] ένα γνωστό πρωτόγονο «αντανακλαστικό» που
λειτουργεί στις ιδεολογικές αντιπαραθέσεις.
To avoid any misunderstanding, we must state that obviously we recognise and

respect [his] right to have his views. [. . .] The point lies elsewhere: in the opinion that
a particular kind of discourse which one disagrees with [. . .] is threatening the Greek
language as a whole.
It is precisely this unhistorical leap that turns a legitimate disagreement into illegitimate
scare talk [. . .] it has no basis in linguistic facts [and] betrays a well-known primitive
‘reflex’ which functions in ideological debates.46
41
Ibid. 112.
42
Eagleton (1991) 11.
43
Christidis (1999a) 86, 114.
44
Frangoudaki (2001) 184, 160. See also Frangoudaki (2003).
45
For instance, Iliou (1989).
46
In a nutshell, history is the ground of legitimacy, while the lack of historical

perspective and correspondence with linguistic ‘facts’ is what turns the opposite
view into a primitive spontaneous reflex, fit only for ideological confrontation.
This is the principal argumentative move of the discourse under consideration:
the ‘camera obscura of ideology’47 is juxtaposed to science, which, like the broad
light of the day (or the Enlightenment), is taken to be beyond doubt or challenge.
In Frangoudaki’s words,
όλα τα παραπάνω δεν τα αμφισβητεί σήμερα κανείς απολύτως.
Θεωρούνται επιστημονικοί κοινοί τόποι και αντίλογος δεν υπάρχει.
nowadays no one doubts any of the above. They are considered scientific common-
places and there is no counter-argument.48
There is an extraordinary lack of self-awareness in this position, as well as a

remarkable indifference to the post-modern critique of science’s privileged access
to the truth.
One outcome is that ideology is also found in the thin line that separates old-
fashioned prescriptivism from institutional regulation of language, as envisaged
by the anti-purist view. For instance, the problem with Tzartzanos and his
contemporaries is, according to Frangoudaki, his normative intention and his
interventionist position.49 On the other hand, there is no question that demotic
Greek is the ‘standard Greek, the regulated and codified choice’ (ρυθμισμένη
και κωδικοποιημένη επιλογή),50 which accepts no deviations or violations of its
rules.
Conversely, Christidis does not exclude prescriptive regulation, so long as this is
conducted in ‘real terms’ (μια συζήτηση με πραγματικούς όρους).51 As he points
out explicitly, he is not in principle against ‘regulation’, since language is regulated
by a number of institutions.52 The problem is how to apply regulation correctly:
να διδάσκεται σωστά η ιστορία της γλώσσας, χωρίς υπόρρητες ή ρητές
αξιολογήσεις και επιλογές (‘to teach the history of the language in the right way,
without any implicit or explicit evaluations and selections’).53 Similarly, the task
is to interpret and not to evaluate the continuity of the language.54 Although the
difference between interpretation and evaluation would seem difficult to sustain,
47
48
Frangoudaki in Charis (2000) 21.
49
50
Ibid. 57. Cf. ibid. 98, 110.
51
52
Christidis (1999a) 56. See also Christidis (1999b) 298, where there is talk of ‘institutional
planning’ (θεσμικός σχεδιασμός). Note the difference in these extracts between κανονίζω,
παρεμβαίνω, ρυθμίζω and σχεδιάζω: the first two are negatively evaluated, the others are positively
accepted.
53
Christidis (1999b) 300.
54
Whereas, in the purist discourse, η ενότητα δεν ερμηνεύεται αλλά αξιολογείται: Christidis
(2001a) 11.
the implication is again obvious: it is possible to teach and interpret without

selecting and evaluating, by putting ideology aside and distancing ourselves from
the ideological ‘womb which gives birth to fallacies’.55
Grounds and aporias

Let us summarize and co-ordinate our analysis so far, so that we can clarify
the presuppositions of the discourse under analysis. We can safely agree with
Moschonas’ observation that the anti-purist discourse is articulated as a response
to its perceived opposite, the discourse of Greek purity and unity.56 As such, it is
primarily an oppositional discourse built on a purposeful critique of the ‘opponent’.
Christidis aptly expresses this stance:
Σε αυτά τα περιεχόμενα απαντά κανείς μ᾽ έναν αντίπαλο λόγο που εκφράζει
τα αντίπαλα περιεχόμενα.
To this content one responds with an opposing discourse which expresses the opposite
content.57
In this mode of oppositional discourse there is no way of escaping polarization.

The basic moves through which opposition is articulated are the invocation
of myth and ideology, which characterize the discourse under attack, and the
conscious appropriation of the opposite pole, represented by science and history.
Both myth and ideology are self-evidently wrong, the former because it is pre-
scientific and non-theoretical and the latter because it is committed and thus not
disinterested. The only remaining position is that of a rational view of the linguistic
data per se, a view that would allow for a regulatory, but non-evaluative and non-
manipulative, intervention in linguistic affairs. It is, in truth, a view that assumes a
serene, enlightened detachment from the violent, value-laden, highly-charged and
covert engagement of its opponent.
While the discursive moves of anti-purism revolve around myth and ideology,
it is its views on the central notions of language and history that constitute the
rationale for the entire discourse. Here, the main proponents seem to diverge, only
to reach similar viewpoints. Frangoudaki seems to start from an unambiguous and
self-evident view of what a language is. For instance, she comments on demotic
and katharevousa that:
Από ποικιλίες του επιπέδου ύφους που (αν μελετηθεί από τους γλωσσολόγους
θα φανεί ότι) ήταν μέχρι τις αρχές του 20ού αιώνα [. . .] μετατρέπονται σε
γλώσσες προσανατολισμένες στο χρήστη.
55
Christidis (1999a) 52. According to a related topos, the problem is with managing or manipulating
data. For Christidis, a particular subject such as linguistic variation takes on illegitimate dimensions from
the moment it becomes an ‘object of management’ in science (αντικείμενο διαχείρισης: Christidis
(2001a) 11; see also (1999a) 157 and (2001a) 5, 9, 12). However, it is not clear how it is possible to
study data or linguistic phenomena without ‘managing’ them.
56
Moschonas (2000).
57
From varieties of register which (if studied by linguists will be shown that) they
constituted up to the beginning of the twentieth century [. . .] they turn into languages
oriented towards the user [sic]. 58
Note here the way the claim is supported by a contorted structure based on a
hypothetical clause. The same peculiar modality is found at another crucial
point in her argumentation, where again a clear-cut distinction between the two
‘languages’ is assumed:
μια γλώσσα που η γλωσσολογική της μελέτη θα έδειχνε ότι καθόλου δεν
είναι «πολυμορφία» από τις ελληνικές γλώσσες «χιλιάδων ετών» αλλά απλώς
δημοτική με ένθετα στοιχεία από την καθαρεύουσα.
a language, the linguistic study of which would show that it is not at all a [product
of] ‘multiversity’ from the Greek languages over ‘thousands of years’, but is simply
demotic with interspersed katharevousa items.59
What is significant here is that, despite the author’s perfunctory analysis of diglossia,60
she is always clear that what we are dealing with here is ‘two Greek languages’,61
of which there is no question which is the privileged one. Thus katharevousa is the
‘written code’, while demotic is the ‘codified and normalized form of the “natural”
language, the standardized variety that corresponds to the fundamental features
of the Greek language’.62 Similarly, while critical of the demoticists’ appeal to an
‘authentic’ language, she does not refrain from using this word, without inverted
commas, in the subtitle of her book («Εκατό χρόνια αγώνες για την αυθεντική
γλώσσα»: ‘A hundred years of struggle for the authentic language’)63 – while
those, like Tzartzanos, who would like to pay attention to all linguistic elements,
even if they originate in katharevousa, are charged with clumsiness for not clearly
perceiving the dividing line.64
Christidis’ view is more nuanced. For instance, he notes that the contrast
between language and dialect, and the attitudes towards that contrast, depend on
historical parameters, which may be wrongly perceived as ‘natural’;65 the basic
charge against the purists is precisely their negation of history and historical
change in the study of language.66 There are several remarks that can be made
about Christidis’ view of history. First, it can be argued that he does not himself
apply the principle of historical parameters in his analysis of the opposing discourse,
as when he claims that Triantafyllidis expresses the same opinion as Mavrofrydis
58
59
Ibid. 117.
60
Ibid. 53–6.
61
Frangoudaki (1992) 366 (and passim).
62
Ibid.
63
Frangoudaki (2001).
64
Ibid. 71.
65
66
Ibid. 10–11.
sixty years earlier, or even Kodrikas and Elladios 200 years earlier;67 or when he
charges his contemporaries with following the same ancient linguistic mythology,
albeit reformulated in its new context.68
On a more serious point, perhaps, Christidis’ view of the historicity of language
might strike one as peculiarly un-historical, since it is related to the ‘deeper nature’
of language. This is claimed, for instance, in the prologue to his book, which is
characterized as:
ένα βιβλίο που επιχειρεί [. . .] να δει τη γλώσσα και τον πολιτισμό χωρίς
μυθοποιήσεις – τόσο στο «ύψος» τους όσο και στα «έγκατά» τους, που συχνά
αγγίζουν τα όρια της «τυφλής φύσης». Και αυτή είναι η απώτερη – και βαθύτερη
– ιστορικότητά τους
a book that attempts [. . .] to view language and culture without any mythologizing –
both in their ‘heights’ and their ‘depths’, which often touch upon the limits of ‘blind
nature’. And this is their ultimate – and most profound – historicity.69
There are similar moments that seem to indicate a distinction between a ‘deep’
historical dimension and the ‘surface’ dimension of historical context: αυτή η
αξιακή διάσταση της γλώσσας συνδέεται με τη βαθύτερη φύση της – τη
βαθύτερη ιστορική φύση της (‘this value dimension of language is related to its
deepest nature – its deepest historical nature’).70
Even if we managed to give a solid definition to ‘value’ as used in the relevant
extracts, we would still have the problem of where to draw the line in this contrast,
in order to assign each feature of language appropriately. For instance, why are
the unity and continuity of Greek not parts of this deep historical dimension?
The answer presumably has to do with the unity of each language with all others,
ultimately based on the ‘unity of human intelligence’.71 In a purposeful oxymoron,
Christidis states that:
η απώτερη «πατρίδα» της γλώσσας, της κάθε γλώσσας, είναι η μία και ενιαία
ανθρώπινη νόηση που δεν γνωρίζει εθνικά σύνορα και, με αυτή την ευρύτερη
έννοια, είναι «άπατρις».
Language’s ultimate ‘homeland’ – every language’s – is the one, unified human

cognition which does not know of national boundaries and, in this wider sense, is
‘homeless’.72
67
Ibid. 9.
68
Cf. Christidis (1999a) 70, 89 and (2003), along with Frangoudaki (1992) 373: ‘the entire set of
arguments [. . .] are essentially the same as those used in the late nineteenth century by the supporters
of K.’
69
70
Ibid. 22; cf. ibid. 29 and (2001b) passim.
71
72
Ibid. 156.
Human cognition is the ultimate ground for this strange conception of history,
which relates to the ‘essential neutrality’ of language.73 This is a truly paradoxical
statement for a supposedly critical view of language, and difficult to reconcile with
a committed socio-historical perspective.
What is found at the heart of the anti-purist discourse is an essentialist view,
which allows Christidis to speak of the ‘phenomenon’ of language and its ‘nature’.74
There is no doubt that the origins of his ‘essential’ neutrality lie in Chomsky’s
universalism. In Christidis’ view, critiques of Chomsky have not touched the
postulate of universality itself but have concerned its ‘historically abstract [. . .]
expression’.75 Cognitive linguistics can provide a useful antidote to Chomsky’s
rationalist conception of universality by pointing out the mixture of ‘linguistic
and pre-linguistic, rational and pre-rational’.76 Thus, we come full circle from a
supposedly anti-mythical, rational, historical view of language to the acceptance of
an essentialist, universal core, which is historical only to the extent that it includes
the non-rational.
Our final point has to do with the political grounding of anti-ethnocentric
discourse. For Frangoudaki, Greece’s future is clearly identified with the need for
participation in the European Union. To this effect, it is necessary to historicize
the notions of nation and language and to place Greece and the Greek language
in their proper – that is, historical – dimensions.77 The dilemma for her is between
two ‘untimely’ positions: the well-known and discredited conservative view and
the nineteenth-century demoticists’ views.78 The implied solution is obvious:
a modernization of demoticist ideas. For Christidis, the dilemma is rather
between the discredited ethnocentric view and the supra-national approach of
multiculturalism.79 The answer is a call for a new humanism that will steer between
the Scylla of nation and the Charybdis of the free market.80
It is implicit that anti-ethnocentric discourse has a clear liberalist orientation,
as has already been pointed out by Moschonas.81 This is evidenced not only in its
essentialist humanism but also in the specific aspirations it projects, such as the goal
of integration within the European Union.82 There is no question, for instance, of
73
Ibid. 94: Έχει, λοιπόν, η γλώσσα μια συστατική ουδετερότητα.
74
Respectively: Christidis (1999a) 57, 160, 227, Christidis in Charis (2000) 19, and Christidis
(1999a) 40, 79 and (2001b).
75
Christidis (1999a) 180–1.
76
Ibid. 182; cf. Christidis (2001b). A similar correction to Chomsky, linking intuition with
social status, is suggested by Frangoudaki in her view that ‘a particular use of the notion of the native
speaker’s intuition, considered without regard to social status, is made in order to support a pro-K [sc.
katharevousa] position’ (1992) 377.
77
Frangoudaki (2001) 153, 157.
78
Ibid. 161.
79
Christidis (1999a) 14, 41, 70, 93, (1999b) 298, (2001a) 16, (2003).
80
Christidis (1999a) 164: ένας νέος ουμανισμός.
81
Moschonas (2000).
82
challenging the role of the market,83 or of involving the notions of class, gender and
so on.84
The liberalist orientation goes hand in hand with a fundamentally idealist
position which foregrounds the conceptual conditions for change. This is
Christidis’ view: δεν είναι, τελικά, η γλώσσα [. . .] που ενώνει ή διαχωρίζει
αλλά οι ιδέες (‘it is not language, in the end, that unites or separates but ideas’).85
Compare Frangoudaki’s assertion: Αλλάζουν οι γλώσσες, επειδή αλλάζουν οι
ιδέες (‘Languages change, because ideas change’).86 And that is why, according
to this discourse, it is sufficient to provide information about the ‘real’ linguistic
situation of Greek to undermine the ethnocentric position.87 If only people could be
convinced that their view was wrong, if only they had adequate information, they
would immediately abandon their position in favour of enlightened modernism.
Towards a socio-historical view of Greek

I have attempted a close analysis of the thematic motifs, the argumentative moves
and the theoretical underpinnings of the contemporary discourse in Greek linguistics
which opposes the ‘purist’ and ‘ethnocentric’ view of the Greek language and its
prescriptivism. I have argued that this discourse sets up an oppositional polarization
of myth and ideology, on the one hand, and history and science, on the other, which
it equates with a ‘them’ and ‘us’ contrast.88 The ‘them’ position is pre-scientific,
non-theoretical, committed and evaluative. The ‘us’ position relies on its scientific
neutrality, which aims at myth-breaking and ultimately finds its foundation in the
unity of human thought. The implication is that this position has no interests and
contradictions of its own, but occupies a view from nowhere, aiming at describing
language ‘in itself’. It is easy to conclude that anti-purist discourse is a kind of
negative critique, sharing with its real or imaginary forbearers the ‘patrician gloom
of the late Frankfurt School’.89 Its blind spots thus relate to its own liberalist and
essentialist presuppositions, coupled with its idealist view of change.
Is there any means of achieving a historically and sociologically informed view
of Greek that overcomes these blind spots? Is it possible to attempt a positive rather
than negative socio-historical critique? For instance, can we study how Greek has
83
According to Christidis (1999b) 298, while the market ‘mystifies its products [. . .], organized
society has every obligation to evaluate these products’. Again, the tendency to concede social goods
to the market ‘has to be controlled somehow’ – but this ‘does not mean disappearance of the market but
delimitation of the market’ (ibid. 300).
84
The closest to this is the recognition of ‘groups’. ‘The language is always and everywhere a set of
uses that is, inter alia, defined by groups’: Christidis (1999a) 82.
85
Ibid. 185.
86
Frangoudaki in Charis (2000) 20.
87
See e.g. Frangoudaki (2001) 159–60; note the emphasis on ‘the significance of the lack of
information’ in Frangoudaki (1992) 377 and Christidis’ belief that their opponents ‘must [. . .] be
convinced that the relative arguments constitute fallacies’ (Christidis (1999a) 56).
88
Cf. Dijk’s (1985) ‘inner’ and ‘outer’ group.
89
Eagleton (1991) 36.
been ‘more than a tool of communication’ by achieving the status of a cultural

asset throughout its long history? Can we assess the importance of written Greek
for its speech communities at large and the development of the language? Can
we work out a coherent view of the successive strata that co-exist synchronically
in its lexicon? What are the current attitudes of its speakers towards the issues of
continuity and particularity, and are these constitutive of their definition of the
language? Do text types bear the marks of diglossia, or, rather, is generic variation
in contemporary Greek dependent on other factors?
To attempt a reply to these questions, we must first clear the conceptual ground
of the old polarities that seem to hinder a more sophisticated socio-historical view
of Greek and its history. For this purpose, as has been implicit in the perspective
adopted in my analysis, certain prerequisites must be met, clearly related to the
aporias of the discourse under consideration. First of all, in my view, any starting
point has to assume that language is not value-free or neutral but value-laden.90
This implies that linguists should let go of the ‘Olympian point of view of the
unbiased observer’, as Coulmas puts it,91 and get involved in the investigation of
language as a cultural artefact – which is implicated in questions of value – rather
than language as a ‘natural’ phenomenon. It is obvious that some linguistic theories
may be more suitable for this than others, insofar as they may explicitly include
such a perspective in their goals, which can be usefully drawn on in this effort.
Secondly, we have to acknowledge that all language ideologies, including our
own, are situated in specific fields of production and circulation. We cannot safely
remove ourselves from specific ‘interpretive frameworks, interests and goals [. . .]
positioned within the political, economic and cultural fields that are implicated
in particular contexts of practice’.92 The discourse of language experts occupies a
particular position in the public sphere. That is why we need to develop theories
and models with explicit criteria and coherent principles that can be put into
argumentation and criticism, as well as opening the space for debate without falling
into the trap of endless critique.
Thirdly, if we are committed to an ideological analysis of language, we have to
accept that ideology is ‘an active material force which must have at least enough
cognitive content to help organize the practical lives of human beings’.93 In other
words, ideological positions are developed and elaborated both on their intrinsic
coherence and on the material basis of their interrelation with people’s lives. For
instance, as Coulmas has pointed out with respect to diglossia, ‘to know that certain
linguistic norms stem from social privileges and from using language as a socially
dividing rather than unifying force does not necessarily diminish their potency’.94
Sociolinguistic facts do not change because we may want to wish them away, just as
90
Toolan (1998) 97.
91
Coulmas (1989) 177.
92
Fennigen (2003) 453; cf. Eagleton (1991) 51. For a review of the problem and solutions in
discourse analysis, see Phillips and Jørgensen (2002) 180–211.
93
Eagleton (1991) 26.
94
Coulmas (1989) 180.
language planning and intervention may have long-term effects that have nothing
to do with the ‘fundamental’, ‘natural’ or ‘authentic’ trends of a language.
Finally, given our non-privileged access to truth about language and given the
material basis of beliefs about language, linguists must focus their attention on
the metalinguistic attitudes of language speakers and treat them as folk theories
about language.95 For instance, if continuity and history are relevant for speakers of
Greek, they cannot be simply dismissed as scientifically trivial or as mere remnants
of nationalism. Perhaps we should start listening more carefully to what Greeks
say and think about their language before pronouncing our verdict. As Hymes has
pertinently suggested, ‘if the community’s own theory of linguistic repertoire and
speech is considered (as it must be in any serious ethnographic account), matters
become all the more complex and interesting.’96
To overcome – to subsume and surpass – old dichotomies would thus mean
looking at Greek, and metalinguistic discourse about Greek, from a new perspective.
For instance, the pivotal issue of diglossia can be studied both as a linguistic ‘fact’
of a certain period of Greek and as a social construct, forming an integral part of
speakers’ attitudes and a means of controlling variation. The demotic/katharevousa
distinction can thus be seen, retrospectively, as both a feature of the linguistic
system of Greek and a metalinguistic notion, around which ideological positions
have evolved. At the same time, metalinguistic attitudes, theories and ideologies,
as well as other reflexivity practices on language, feed back into the linguistic
system or process – and it could not be otherwise with Greek and its speakers, for
whom literacy and orality span millennia of development. The simplifying view
of polar oppositions can thus give way to a picture of increased complexity, where
agents or forces of change and continuity can be identified. Rather than natural
processes, what we are dealing with are territorializing and deterritorializing forces
in language,97 such as standardization, canonization, language education versus
language contact, linguistic performance, new technologies. In this sense, more
interesting than the question of continuity and purity or their opposites is the
search for continuities and ruptures both in the diachrony and the synchrony of
Greek.
95
Niedzielski and Preston (2003).
96
Hymes (1972) 39.
97
In the sense of Deleuze and Guattari (1980).
References
Babiniotis: Μπαμπινιώτης Γ. (1999), ‘Αξιολογική προσέγγιση της γλώσσας.
Γλωσσολογία και ιδεολογήματα’, in Η Γλώσσα ως Αξία, Athens: 295–306, 329–32.
Bauer, L. and Trudgill, P. (eds.) (1999), Language Myths, London.
Beaugrande, R. de (1999), ‘Discourse studies and ideology: on “liberalism” and
“liberalisation” in three large corpora of English’, Discourse Studies, 1: 259–95.
Beaugrande, R. de (2001), ‘Interpreting the discourse of H.G. Widdowson: a corpus-based
critical discourse analysis’, Applied Linguistics, 22: 104–21.
Charis: Χάρης, Γ. (ed.) (2000), ‘Δέκα μύθοι για την ελληνική γλώσσα’, Τα Νέα –
Πρόσωπα 21ος αιώνας, 80 (16 September 2000): repr. as Γ. Χάρης (ed.) (2001), Δέκα
μύθοι για την ελληνική γλώσσα, Athens.
Christidis: Χριστίδης, Α.-Φ. (1999a), Γλώσσα, Πολιτική, Πολιτισμός, Athens.
Christidis: Χριστίδης, Α.-Φ. (1999b), ‘Συζήτηση’, in Συνέδριο για την Ελληνική Γλώσσα,
1976–1996 (Αθήνα, 29 Νοεμβρίου–1 Δεκεμβρίου 1996), Athens: 297–300.
Christidis: Χριστίδης, Α.-Φ. (2001a), ‘Ιστορίες της ελληνικής γλώσσας’, in Christidis,
A.-Ph. (ed.), Ιστορία της ελληνικής γλώσσας. Από τις αρχές έως την ύστερη
αρχαιότητα, Thessaloniki: 3–17.
Christidis: Χριστίδης, Α.-Φ. (2001b), ‘Η φύση της γλώσσας’, in Christidis, A.-Ph.
(ed.), Ιστορία της ελληνικής γλώσσας. Από τις αρχές έως την ύστερη αρχαιότητα,
Christidis: Χριστίδης, Α.-Φ. (ed.) (2001c), Εγκυκλοπαιδικός οδηγός για τη γλώσσα,
Thessaloniki. [Also available at the internet address: http://www.komvos.edu.gr.]
Christidis: Χριστίδης, Α.-Φ. (2003), ‘Στην επικράτεια του λόγου’(review of Γ. Χάρης, H
γλώσσα, τα λάθη και τα πάθη), Το Βήμα, 21 December 2003.
Christidis: Χριστίδης, Α.-Φ. (2004), ‘Χρήσεις της γλώσσας: Οι όροι μιας συζήτησης’, Ο
Πολίτης, 127: 16–22.
Coulmas, F. (1989), ‘Democracy and the crisis in normative linguistics’, in Coulmas (ed.),
Language Adaptation, Cambridge: 177–93.
Deleuze, G. and Guattari, F. (1980), Mille plateaux, Paris.
Dijk, T.A. van (1985), Discourse and Communication: New Approaches to the Analysis of Mass
Media Discourse and Communication, Berlin.
Eagleton, T. (1991), Ideology: An Introduction, London.
Fennigen, J. (2003), ‘Introduction’, Pragmatics, 13: 453–6.
Frangoudaki: Φραγκουδάκη, Ά. (2001), Η γλώσσα και το έθνος, 1880-1980. Εκατό
χρόνια αγώνες για την αυθεντική γλώσσα, Athens.
Frangoudaki: Φραγκουδάκη, Ά. (2003), ‘Της γλώσσας τα καμώματα’ (review of Γ.
Χάρης, H γλώσσα, τα λάθη και τα πάθη), Τα Νέα, 8/9 November 2003.
Goutsos: Γούτσος, Δ. (2003), ‘Σώμα Ελληνικών Κειμένων: Σχεδιασμός και υλοποίηση’,
Πρακτικά του 6ου Διεθνούς Συνεδρίου Ελληνικής Γλωσσολογίας, Πανεπιστήμιο
Κρήτης, 18–21 Σεπτεμβρίου 2003: CD-ROM publication.
Hymes, D. (1972), ‘Models of the interaction of language and social life’, in J.J. Gumperz
and D. Hymes (eds.), Directions in Sociolinguistics: The Ethnography of Communication,
New York: 35–71.
Iliou: Ηλιού, Φ. (1989), Ιδεολογικές χρήσεις του Κοραϊσμού στον 20ό αιώνα, Athens.
Moschonas: Μοσχονάς, Σ.Α. (2000), ‘Ιδεολογικός αντίλογος για τη γλώσσα’ (review of

Α.-Φ. Χριστίδης, Γλώσσα, Πολιτική, Πολιτισµός), Η Καθηµερινή, 8 February 2000.
Niedzielski, N.A. and Preston, D.R. (2003), Folk Linguistics, Berlin and New York.
Partington, A. (1998), Patterns and Meanings: Using Corpora for English Language Research
and Teaching, Amsterdam and Philadelphia.
Phillips, L. and Jørgensen, M.W. (2002), Discourse Analysis as Theory and Method, London.
Stubbs, M. (1996), Text and Corpus Analysis. Computer-Assisted Studies of Language and
Culture, Oxford.
Toolan, M. (1998), ‘Don’t leave your language alone’ (review of D. Cameron, Verbal
Hygiene), Language and Communication, 18: 87–99.
17
Korais and the Second Sophistic:

The Hellenistic Novel
as Paradigm for a Modern Literary Language
Roderick Beaton
The name of Adamantios Korais has become almost synonymous with attempts
to reform the Modern Greek language in the nineteenth and twentieth centuries.
Born in Smyrna in 1748, Korais spent the greater part of his life in Paris, where
he died in 1833. Under the name of Coray, he was a reputed classical scholar in
France; under various pseudonyms he wrote patriotic tracts and verses in Greek;
he also left a rich correspondence in that language. Probably Korais’ most lasting
legacy is the series of prefaces, written in his own preferred, ‘corrected’ form of
Modern Greek, to editions of Ancient Greek texts in the original. Through this
venture, Korais sought to raise the educational level of the Greeks of the Ottoman
empire, and it seems that, thanks to inexpensive subsidized editions, these were
widely read and respected. By the late nineteenth century, Korais was regularly
hailed among Greeks as ‘father of the nation’. One consequence and proof of
this esteem was that when a new chair in Modern Greek and Byzantine History,
Language and Literature was established at King’s College London in 1918 by
a triumvirate consisting of Eleftherios Venizelos, the Greek Minister in London
Ioannes Gennadius, and Ronald Burrows, Principal of King’s, it was named after
Adamantios Korais.1
Korais’ ideas and his example are often supposed to have laid the foundations for
katharevousa, the artificial hybrid that did duty as a written standard throughout the
nineteenth century, and which for much of the twentieth was officially sanctioned
by the Greek constitution. Korais himself never used that term; nor did he ever
organize the scattered, repetitive, usually trenchant, and sometimes contradictory
opinions that he expressed in his many scattered writings into anything that could
be called a system. Even the language in which he himself wrote is very different
1
On the establishment of the Koraes Chair at KCL, inaugurated in 1919, see Clogg (1986); on
Korais’ life and career see Beaton (1991); Chaconas (1968); Dimaras (1953); Henderson (1970) 142–
58.
341
342 roderick beaton
from the florid archaism cultivated, say, by Emmanuel Roidis half a century later,
though rather less so from that of the Constantinopolitan newspapers whose
gallicisms provoked the righteous indignation of Psycharis (Jean Psichari) in the
1880s.2 But no one doubts Korais’ central place in the ideological battleground
that opened up in print early in the nineteenth century, and remained the field of
frequent skirmishing, down to at least the 1970s.3
Korais’ earliest public statement of his ideas for reforming the Modern Greek
language is to be found in the letter to his friend Alexandros Vasiliou, in Vienna,
that was published as the preface to his edition of Heliodoros’ Aithiopika, in 1804.4
This is a key text for other reasons as well. First of all, this was the edition with
which Korais chose to inaugurate the series, already mentioned, to which he would
later give the collective title, ‘Greek Library’, whose publication would continue
throughout the following ten years. Then again, it was in this same text that Korais
proposed a definition of the novel that would later be implicitly adopted by the
first novelists of the newly founded Greek state, in the 1830s – and here, too, that
the first ever generic term for the novel came to be coined in Greek. This coinage,
only slightly modified, remains standard today. Finally, and for none of these
reasons, this preface has earned praise as Korais’ most elaborate exercise in literary
criticism.5
Curiously, for such an influential text, this one seems never to have been examined
critically in its entirety. Those who have drawn attention to it in recent years,
and here I include myself, have done so piecemeal, to highlight its contribution
either to the language question, or to the definition of the novel, or even to read
it, independently of either, as ‘an essay in method’.6 What I wish to argue here is
that, in his characteristically digressive way, Korais arrives at the point of being
able to formulate his substantive ideas on language reform as the conclusion to a
disquisition on literature, and specifically on the literature of what today we would
call the ‘Second Sophistic.’
I begin with the concluding part, on language reform, which is also the best known.
Korais was in no doubt that the desired standard must be based on the ‘common
2
See, respectively, Roidis (1988) (first published 1866); Psycharis (1971) 69–78.
3
On the Greek language question, including the role of Korais, see Beaton (1999) 296–346;
Browning (1983) 100–18; Horrocks (1997) 344–65; Liakos (1996); Mackridge (1985) 6–11;
Mackridge (2009) (now the fullest treament). See also the essays by Mackridge and Van Dyck in
Beaton and Ricks (2009).
4
Korais (1984).
5
Moullas (1991).
6
On the language question in this connection: Moschonas (1981) xxix–xxxiii; cf. Beaton (1999)
300–3. On the Greek novel in this connection: Chrysanthopoulos (2004); Roilos (2003) 66–7; Tonnet
(1994); Tonnet (1996) 83–102; Tonnet (2005); cf. Beaton (1999) 54–6; Beaton (2007). ‘Essay in
method’ (ένα δοκίμιο μεθόδου): Moullas (1991) 126.
korais and the second sophistic 343
language’ (koine glossa), as it was spoken in his time.7 This language, ‘which we
have imbibed with our mother’s milk’,8 is ‘one of the most inalienable possessions
of the nation’.9 Korais already understood what would become a central tenet of
the Neogrammarians later in the century, one that would be eagerly espoused by
his arch-rival Psycharis: even the harshest tyrant ‘cannot change [the citizen’s]
language . . . Only time has the power to alter the dialects of nations, as it alters
nations themselves.’10 The revival of Ancient Greek, Korais argued, was neither
possible (‘miracles do not happen every day,’ he warned drily), nor desirable:
‘Whoever writes in Ancient Greek [today] within a few years (and often, within
a few days) will be forgotten together with his writings.’11 In any case, the ancient
language had itself undergone many changes between the time of Plato and that of
‘Julian and Libanius’: which form ought one to use? Here Korais cites approvingly
the resolution of the Académie Française that the national dictionary should be
revised every century.12
Korais also delivers a dire warning – ironically enough, in view of what was
often later perpetrated in his name – against the kind of hotchpotch of ancient and
modern linguistic elements that would later come to be known as katharevousa:
those whom . . . forgive me if I call mongrel-Hellenizers, or if you prefer, mongrel-
barbarians. The aim of such people is certainly praiseworthy, inasmuch as they try to
bring the modern language as close as possible to its mother, Ancient Greek.13
But the results he castigates as ‘obscure, harsh, disgusting to the ear, monstrous and
truly hermaphrodite’, ‘born of neither ancient nor modern Greece’.14 Were it not
for the language in which they are expressed, these sentiments could equally well
have come from the pen of Solomos or Psycharis, implacable opponents of Korais’
programme of reform.
7
Korais (1984) 46–52.
8
Ibid. 37: την κοινήν ταύτην γλώσσαν, την οποίαν εθηλάσαμεν με το μητρικόν γάλα. Cf.
ibid. 44: εις την φυσικήν αυτού διάλεκτον, ήγουν εις εκείνην, την οποίαν εθήλασε με το γάλα,
και λαλεί καθ᾽ ημέραν . . .
9
Ibid. 49: η γλώσσα είναι έν από τα πλεόν αναπαλλιοτρίωτα του έθνους κτήματα.
10
Ibid. 50: . . . το οποίον ουδ᾽ ο σκληρότατος τύραννος είναι καλός να κατορθώση. Γυμνόνει
από τα υπάρχοντά του τον πολίτην ο τύραννος . . . αλλά δεν μπορεί να του αλλάξη την
γλώσσαν· . . . Μόνος ο καιρός έχει την εξουσίαν να μεταβάλλη των εθνών τας διαλέκτους,
καθώς μεταβάλλει τα έθνη.
11
Ibid. 42: θαύματα όμως δεν γίνονται καθ᾽ ημέραν. Ibid. 43: Όστις γράφει Ελληνιστί,
μετ᾽ ολίγους χρόνους (και συχνά, μετ᾽ ολίγας ημέρας) θέλει λησμονηθήν και αυτός και τα
συγγράματά του.
12
Ibid. 45–6 and n.1.
13
Ibid. 47: τους οποίους . . . συγχώρησον να ονομάσω Μιξελληνίζοντας, ή, αν αγαπάς
κάλλιον, Μιξοβαρβάρους. Ο σκοπός των τοιούτων είναι βέβαια επαινετός, επειδή προθυμούνται
να σιμώσωσιν, όσον δυνατόν εγγύτερον, την σημερινήν γλώσσαν εις την μητέρα της την
Ελληνικήν.
14
Ibid. 47–8: . . . συνθέτουν . . . σκοτεινόν, τραχύν, αηδή εις την ακοήν, τερατώδη και
αληθώς Ερμαφρόδιτον λόγου χαρακτήρα. Ibid. 49: . . . ούτε της παλαιάς ούτε της νέας Ελλάδος
γέννημα.
344 roderick beaton
Having gone so far, Korais even concedes that he risks being accused of
contradiction, in demanding ‘that we must correct and embellish our common
[that is, modern, spoken] language’.15 But, committed democrat though he was,
and citizen of post-revolutionary France, Korais had a horror of the vulgar mob:
When I say that the nation partakes of its language with democratic equality, I do not
mean that we must leave the development and creativity of the language to the mob-
rule imagination of the vulgar.16
The process of reform that he advocates is to be carried out with classic moderation
(‘a language moderately embellished’); the ‘seeds’ of Ancient Greek are to be sown
‘by the handful, not by the sackful’.17 A little later he returns to the political analogy,
further refining the unstable balance between authoritarian prescription and
fundamental respect for spoken usage, which is characteristic of Korais’ position:
The nation’s men of learning are naturally the legislators of the language spoken by
the nation; but they are (I repeat) legislating for a republic. To them belongs the
correction of the language, but the language is the possession of the whole nation, and
a sacred one.18
It remains, now, to trace the circuitous route by which Korais has arrived at this
point. The context is after all the preface to an ancient literary text, the Aithiopika
of Heliodoros. As one might expect of an editor, Korais begins with his text. He
does so boldly, indeed patriotically: the literary genre known to Europeans by
the French term roman and its cognates had once upon a time been invented by
Greeks. (He ignores the fact that Heliodoros describes himself, in the last sentence
of the Aithiopika, as ‘a Phoenician from Emesa’.)19 In urging this claim, Korais is
undeterred either by how far the modern French genre has diverged from its origins
15
Ibid. 50: Ίσως ήθελε τις νομίσειν ότι αντιφάσκω αυτός εις εαυτόν, επειδή προ ολίγου
έλεγον, ότι πρέπει να διορθώσωμεν και να καλλύνωμεν την κοινήν ημών γλώσσαν.
16
Ibid. 50: Όταν λέγω, ότι από την γλώσσαν μετέχει το έθνος όλον με δημοκρατικήν ισότητα,
δεν νοώ ότι πρέπει ν᾽ αφήσωμεν την μόρφωσην και δημιουργίαν αυτής εις την οχλοκρατικήν
φαντασίαν των χυδαίων.
17
Ibid. 51: . . . εις γλώσσαν μετρίως καλλωπισμένην . . . [Σ]πείρε . . . τα Ελληνικά σπέρματα,
αλλά . . . με την χείρα και όχι με τον σάκκον. The expression, ‘sackful’, proverbial already in Ancient
Greek, goes back to Plutarch (Moralia 348a).
18
Korais (1984) 52: Οι λόγιοι άνδρες του έθνους είναι φυσικά οι νομοθέται της γλώσσης,
την οποίαν λαλεί το έθνος· αλλ᾽ είναι (πάλιν το λέγω) νομοθέται δημοκρατικού πράγματος [I
translate this phrase as a Greek calque of res publica]. Εις αυτούς ανήκει η διόρθωσις της γλώσσης,
αλλ᾽ η γλώσσα είναι κτήμα όλου του έθνους, και κτήμα ιερόν.
19
According to Chrysanthopoulos (2004) 65: Και ο ελληνόφωνος Ηλιόδωρος από τη Φοινίκη
(Έμεσα) και τα Αιθιοπικά με την «μιμητική» και όχι «διηγητική» . . . αφηγηματική τεχνική τους
. . . ταιριάζουν απόλυτα στο σχήμα του διαφωτιστή νομοθέτη Αδαμαντίου Κοραή. Γιατί εδώ πια
το πρόβλημα της απόδοσης του γαλλικού όρου Roman συμπυκνώνει τη σχέση αρχαίου και νέου
ελληνικού με την ευρωπαϊκή διαμεσολάβηση. I am not convinced that Korais had even considered
the implications of the fact that Heliodoros was not, himself, Greek.
(generally for the worse, as he believes), or by the moral and linguistic deterioration
he detects in the late antique and medieval Greek texts available to him. Taking as
his basis the definition of the novel (or ‘romance’) by Daniel Huet back in 1670,
Korais produces a slight modification, which he argues is adapted to the practice
of writers in Greek between, as he believed, the fourth century and the twelfth:20
A fictional but plausible story of adventures in love, written with artistry and
dramatically, for the most part in prose, for the instruction and pleasure of readers.21
It now falls to us, he goes on, the modern Greeks [Graikoi], to baptize this genre for
which no ancient generic term exists:
nor is it right that we should confer the barbaric name of roman upon a kind of writing
which the Europeans have taken from the [ancient] Greeks ... it remains for us to
find a suitable name for it, especially now that the regeneration of Hellas, which has
already begun, promises for the future more works of this, as well as of every other,
kind.22
This was the occasion for Korais to coin the term mythistoria, which would soon
be adopted by the first novelists of independent Greece and which, in the slightly
modified form of mythistorema, introduced in the 1860s, has remained standard
ever since.23 From its context, it can be seen that there was more to this than a
philologically tidy attempt to fill a terminological gap left by late antiquity;24 the
previously missing generic term is necessary, simultaneously to reinforce the claim
to Greek ownership of the present-day European genre, and to pave the way for
future achievements in the genre by the revived Greek nation.
A little later, it is again clear that Korais, in this preface, is not merely describing
an ancient literary text, but prescribing for future Greek practitioners, whom he
daringly designates by another coinage, which has remained unchanged ever
20
Korais (1984) 5–15. Korais was mistaken about the relative chronology of the ancient novels,
believing the Aithiopika to have been the earliest of those preserved, whereas we now know it to have
been the latest. However, his dating of this text to the fourth century AD has not been overturned,
although a plausible case now exists for identifying Heliodoros, instead, with a historical figure of the
early third; for a summary of the evidence and bibliography, see Morgan (1997). Korais knew and
correctly dates the three twelfth-century novels which have been preserved entire: (1984) 13–15.
21
Ibid. 3: Πλαστήν, αλλά πιθανήν ιστορίαν ερωτικών παθημάτων, γραμμένην εντέχνως και
δραματικώς, ως επί το πλείστον εις πεζόν λόγον, προς ωφέλειαν και ηδονήν των αναγινωσκόντων
(the final phrase supplied from the previous page – where Korais translates his French source – and
not repeated, presumably as not requiring modification). Cf. Huet (1966) 4: ‘Ce que l’on appelle
proprement Romans sont des fictions d’aventures amoureuses, écrites en Prose avec art, pour le plaisir
& l’instruction des Lecteurs.’
22
Korais (1984) 5: Εις ημάς, φίλε μου Αλέξανδρε, τους Γραικούς, επειδή όνομα ακόμη δεν
έλαβον τα τοιαύτα, ουδ᾽ είναι δίκαιον να δώσωμεν την βάρβαρον ονομασίαν του Ρωμανού εις
είδος συγγράμματος, το οποίον έλαβον οι Ευρωπαίοι από τους Έλληνες, . . . μένει να εύρωμεν
όνομα κατάλληλον εις αυτό, τώρα μάλιστα, όταν η αρχομένη της Ελλάδος αναγέννησις
επαγγέλλεται και τοιαύτα . . . , καθώς και παντός άλλου είδους συγγράμματα.
23
Ibid. 5; cf. Beaton (1999) 55.
24
On the absence of an ancient or Byzantine generic term for the novel, see e.g. Hägg (1983) 3;
Beaton (1996) 62 and 239 n. 57.
346 roderick beaton
since:
It is not only, my friend Alexandros, these [Aristotelian] rules which the novelist
[mythistoriographos] must observe. The very name Mythistoria sufficiently informs us,
that the action is not to be confined to the myth alone, but partakes also of history. It
is surely forgivable for a writer to invent, as he thinks fit, the dramatis personae . . . but
he may not be forgiven if he writes falsehoods either about the geographical place, in
which he invents the scene of the action, or in narrating the manners of its people.25
In this way, and thanks to Korais, the Modern Greek novel came into existence
in name, and prescriptively, thirty years before it did in fact. It was during the
years 1833–4, and seemingly as an immediate and self-conscious response to the
establishment of the Greek Kingdom with the arrival of its first king, Otto, in
February 1833, that three writers more or less simultaneously set out to produce
the first Modern Greek novels. The brothers Panayotis and Alexandros Soutsos
and Iakovos Pitsipios not only used Korais’ term, mythistoria, to refer to their work
but also, with varying degrees of faithfulness, devised their plots in conformity
with his definition, and in varying degrees also with the precedent of the ancient
novel.26
‘Please do not be bored, Alexandros my friend’, writes Korais, ‘by my long

digressions.’27 Moullas is probably right in his assessment that in this text Korais
more than achieves his multiple and interlocking aims, thanks, in part, to the ‘form
of a letter-cum-preface [which] allows him to move with ease, overcoming its
limitations’.28 Whether by design or not, this manner of exposition easily obscures
the connections between one part of the preface and the next. Having discussed
25
Korais (1984) 24: Δεν είναι, φίλε μου Αλέξανδρε, μόνοι οι κανόνες ούτοι, τους οποίους
χρεωστεί να φυλάσση ο Μυθιστοριογράφος. Αυτό της Μυθιστορίας τόνομα ικανώς διδάσκει,
ότι το δράμα δεν περιορίζεται εις μόνο τον μύθον, αλλ᾽ έχει εξ ανάγκης και μέρος τι ιστορικόν.
Συγχωρείται βέβαια εις τον συγγραφέα, να πλάση, ως τον δόξει, τα πρόσωπα του δράματος . . .
[Δ]εν είναι συγχωρημένον εις αυτόν να ψεύδεται μήτ᾽ εις την γεωγραφίαν του τόπου, μήτ᾽ εις
την διήγησιν των ηθών του έθνους, όπου πλάσσει την σκηνήν του δράματος.
26
Promisingly, and surely correctly, Henri Tonnet has observed: ‘on peut supposer que Coray
propose l’oeuvre d’Héliodore comme modèle aux futurs romanciers grecs.’ But although this scholar
has done more than anyone to highlight the importance of the ancient novel as a template for at least one
of the first practitioners of the genre in the Greek state in the 1830s, he is unwilling to generalize beyond
the example of Pitsipios’ novel The Orphan-Girl of Chios, as he continues: ‘Ceux-ci n’ont en général
pas suivi le programme néo-classique de Coray. Ils ont préféré imiter Le Sage, Goethe ou Eugène Sue’
(Tonnet (2005) 101). I believe that here Tonnet seriously understates his own case. For the ancient
novel as a template for the novels of the Soutsos brothers, in the 1830s, see Beaton (2006); as a point of
departure for Palaiologos’ The Polypath (1839), see Chrysanthopoulos (2004) 65–9.
27
Korais (1984) 37: Μη βαρυνθής, Αλέξανδρε φίλε, τας μακράς μου παρεκβάσεις.
28
Moullas (1991) 126: Αν οι στόχοι του υπερκαλύπτονται, είναι γιατί ο Κοραής . . . έχει ν᾽
αποστείλει ουσιώδη, αλληλένδετα και πολλαπλά μηνύματα, ενώ η προλογική-επιστολική
μορφή τού επιτρέπει να κινείται με άνεση ξεπερνώντας τα όριά του.
the Aithiopika in literary and generic terms, Korais now moves to a consideration
of its language; the shift of focus has begun, that by the end of the preface will have
propelled the reader all the way from the definition of an ancient literary genre to a
programme for reforming the Modern Greek language:
In Heliodoros’ time and for some centuries before it, the Greek language had been
in decline; which is why those who wrote in Greek then did nothing but imitate the
writers of old.29
In the account that follows, Korais did not, of course, have the benefit of the
exhaustive linguistic studies of the late nineteenth century, which in effect defined
our modern understanding of the term ‘Second Sophistic’ (originally coined by
Philostratos in the third century). Although Korais did not use that term, it is, I think,
uncontroversial today to situate Heliodoros’ novel squarely within that literary and
linguistic movement, and indeed as one of its most elaborate achievements.30
Korais now proceeds to divide these ‘imitators’ into two groups. The first, and
‘less inspired’, consists of those who ‘ape’ a specific chosen author;31 the second,
those whose imitative powers range more widely. Of this second group, there
are those who mix everything up and make fools of themselves;32 only a few are
endowed with the ‘genius . . . to sip, like bees, at the flowers of older texts, and find
the art to combine things foreign to one another, so as to make them their own’. It
is to this latter subdivision that ‘our Heliodoros’ belongs,33 and earns praise from
Korais, in a rare moment of lyricism, for
imitation so artful that what is stolen seems to be born from the head of Heliodoros;
. . . a garment woven from various pieces, but woven together with such dexterity, that
only in a few places can one see the stitches and threads, so that it rather appears as a
brilliant garment, woven in a single piece, from top to bottom.34
Even in the case of Heliodoros, Korais goes on, ‘all is not golden’; sometimes
even his ‘imitations’ are ‘cold and wrong-headed’.35 (Again, modern judgements
29
Korais (1984) 26: Εις του Ηλιοδώρου τους χρόνους και τινάς εκατονταετηρίδας πρότερον, η
Ελληνική γλώσσα ήτον εις την παρακμήν της· όθεν όσοι συνέγραφον Ελληνιστί τότε, άλλο δεν
έκαμναν παρά να μιμώνται τους παλαιούς συγγραφείς.
30
See, indicatively, Anderson (1993) 86–100 and 263 n. 4. On the relation of the Aithiopika to the
Second Sophistic, despite its allegedly late date in the third quarter of the fourth century, see ibid. 158,
166–8; Reardon (1989) 8–10; Morgan (1997) xxi.
31
Korais (1984) 26–7.
32
Ibid. 27–8.
33
Ibid. 28: Ολίγοι τινές άνδρες ευφυείς . . . , καθώς ο ημέτερος Ηλιόδωρος, απανθίσαντες,
ως μέλισσαι, από των παλαιών την ανάγνωσιν, ευρήκασιν και τέχνην να συγκεράσωσι τα ξένα,
ώστε να τα κάμωσιν ιδικά των.
34
Ibid. 26: μίμησις τόσον έντεχνος, ώστε τα κλέμματα φαίνονται γεννήματα της κεφαλής
του Ηλιοδώρου· . . . ήγουν ένδυμα συρραμένον όμως τόσον επιδέξια, ώστε εις ολίγα μέρη βλέπει
τις τα ράμματα και τους κλωστήρας, αλλά φαίνεται μονοσώματον υφαντόν άνωθεν έως κάτω
λαμπρόν ένδυμα.
35
Ibid. 28: Μ᾽ όλον τούτο, ουδ᾽ εις αυτούς, καθώς ουδ᾽ εις τον Ηλιόδωρον, όλα είναι χρυσός.
Όστις αναγινώσκει με προσοχήν, ευρίσκει και εις αυτούς μιμήματα ψυχρά και κακόζηλα . . .
348 roderick beaton
since the late nineteenth century have often not been very different.) But at least
Heliodoros comes out far ahead of Korais’ third category of imitators, the later
Byzantine authors for whom he has already had harsh words, and whom he now
dismisses as ‘imitators of other imitators’.36
Heliodoros, we are told, lived at a time when ‘the vulgarity of the Greek
race [genos] was increasing’;37 faced with the linguistic choices available then, a
millennium and a half earlier, a novelist of ‘genius’ could come up with solutions
that would profoundly influence all his successors, and go on to found a lasting
genre that would produce some of the most admired works of the European
Renaissance.38 By now it is becoming clear that Heliodoros’ novel is being held up
as a model to be emulated in future, not only as an example of its literary genre, but
also for its linguistic choices.
The present edition, Korais goes on to inform his readers, includes a partial
commentary on his text. Here, he announces that he has been inspired by the
example of Eustathios of Thessaloniki, the twelfth-century bishop and commentator
on Homer, to ‘identify parallels between the ancient Hellenic language and this
new [language], which we speak today’.39 In following Eustathios’ practice,
Korais at the same time draws a clear distinction between the times in which
the Byzantine bishop had lived and his own. The parallels drawn by Eustathios
between the language of Homer and the spoken language of the twelfth century,
Korais explains, could have been ‘of very little help in the explication of the Poet,
and almost none at all in the correction of the common tongue, whose barbarism
was increasing more and more during those years’.40
From this it may be deduced that the parallels to be drawn by Korais, in his turn,
between the language of Heliodoros and that of his own time, are intended to be of
use in ‘correcting’ the latter. Korais has now, in this elliptical way, introduced into
his discourse a term that will shortly assume key importance, as he moves to outline
his own programme for linguistic reform. This term is ‘correction’ (diorthosis);
Korais defines it like this:
By ‘correction’ I mean not just the reformation of various barbarously formed words
and constructions, but also the preservation of many others, that people are too hasty
to expel from the language, as barbarous, who have not paid sufficient attention to the
nature of language.
‘No such correction was possible in Eustathios’ time’,41 Korais continues, convinced
36
Ibid.: άλλων μιμητών μιμητάς.
37
Ibid.: καθ᾽ όσον ηύξανεν η χυδαιότης του Ελληνικού γένους.
38
Ibid. 28–9.
39
Ibid. 35: τον παραλληλισμόν της αρχαίας Ελληνικής γλώσσης, και της νέας ταύτης, την
οποίαν λαλούμεν την σήμερον.
40
Ibid. 36: πολλά ολίγην βοήθειαν έλαβεν απ᾽ αυτόν εις την εξήγησιν του Ποιητού, και
ουδεμίαν σχεδόν εις την διόρθωσιν της κοινής γλώσσης, της οποίας ο βαρβαρισμός ηύξανεν επί
μάλλον και μάλλον εις τους χρόνους του.
41
Ibid.: Διόρθωσιν ονομάζω της γλώσσης, όχι μόνον τον μετασχηματισμόν διαφόρων
βαρβαρομόρφων λέξεων και συντάξεων, αλλά και την φυλακήν πολλών άλλων, τας οποίας ως
that the twelfth-century decline of Byzantium had been inherently irreversible

– and blithely conflating political decline with linguistic change. Unsurprisingly,
there is no mention of such notable twelfth- and thirteenth-century Atticizers as
Anna Komnene or Niketas Choniates; but of course Korais was not interested in
Byzantium. The contrast he is at pains to draw is this:
The long-awaited time of rebuilding is at hand; and just as day by day the race [genos]
is enriched by people like Eustathios, so it is liberated from others like Theodore
[Prodromos] and Niketas [Eugenianos: both authors of twelfth-century novels].
These new Eustathioi, by drawing parallels in the way that I have shown, will both
regularize the language that is common to all, and make easier and more effective the
task of learning Ancient Greek.42
Korais’ far from transparent strategy turns upon this point. What he is urging is
not just the relatively abstract ‘correction’ and ‘embellishment’ of which he speaks
in the frequently quoted closing pages. Rather, this programme depends upon
a systematic comparison between the ancient language on the one hand – at a
particular point in its history and exemplified in a specific literary text – and, on
the other, the language in common use in Korais’ own time.
Seen in this light, Korais’ programme comes to seem less dogmatically
prescriptive than it has often appeared. Its goal, as emerges from a revealing aside,
is the point already reached by other nations, which
are called enlightened, only when they perfect their languages. The Italians, the
French, the English, truly began to liberate themselves from barbarism, at the moment
when their learned men, few at first, set themselves the task of writing in their common
tongue.43
This is scarcely to be distinguished from the view of Solomos, who twenty years
later would bitterly oppose Korais’ views in his ‘Dialogue’: ‘All the wise nations
have settled after all for writing in the language of the people.’44
Implicit in the whole strategy of his preface, read in this way, is the recognition
that was already widespread in Europe before this time, that the way to a national
standard language lay through the nation’s literature. In his ‘Dialogue’, Solomos,
βαρβάρους σπουδάζουν να εξορίσωσιν από την γλώσσαν, όσοι μετά προσοχής δεν ερεύνησαν
την φύσιν της γλώσσης. Τοιαύτη διόρθωσις εις του Ευσταθίου τους χρόνους ήτον αδύνατος.
42
Ibid. 36–7 (my emphasis): Έφθασεν τέλος πάντων ο τόσον επιθυμητός καιρός ούτος
της ανοικοδομής· και το γένος καθ᾽ όσον πλουτίζεται ημέραν εξ ημέρας από Ευσταθίους,
ελευθερούται από Θεοδώρους και Νικήτας. Οι νέοι ούτοι Ευστάθιοι, μεταχειριζόμενοι τον
παραλληλισμόν, του οποίου δίδω το παράδειγμα, και την κοινήν εις όλους γλώσσαν θέλουν
κανονίσειν, και την μάθησιν της αρχαίας Ελληνικής ευκολωτέραν και εντελεστέραν θέλουν
καταστήσειν.
43
Ibid. 41–2: τότε μόνον ονομάζονται τα έθνη φωτισμένα, όταν φέρωσι την γλώσσαν αυτών
εις τελειότητα. Οι Ιταλοί, οι Γάλλοι, οι Άγγλοι, τότε αληθώς άρχισαν να ελευθερούνται από την
βαρβαρότητα, όταν οι κατ’ αρχάς ολίγοι λόγιοι αυτών άνδρες κατεχέρησαν να γράφωσιν εις την
κοινήν αυτών γλώσσαν.
44
Solomos (1986) 13: Ησύχησαν τέλος πάντων γράφοντας τη γλώσσα του λαού τους τα
σοφά έθνη.
350 roderick beaton
steeped in the literary history of Italian, the language of his formal education, placed
literature and the figure of the ‘Poet’ centre-stage. Central to the establishment
and recognition of a national standard in many parts of Europe, particularly in
the eighteenth and nineteenth centuries, was the then up-and-coming genre of
the novel.45 In his elliptical way, it appears that Korais was advocating precisely
this process for the new Greece: what Cervantes had done for Spanish, Goethe
for German, the writers of the Enlightenment for French, the eighteenth-century
English novelists for English, and what Manzoni would shortly do for Italian,
had already been done for Greek – fifteen hundred years earlier by Heliodoros of
Emesa.
If by this Korais had meant that his countrymen ought to learn to write in the
language of the Aithiopika, he would simply have joined the ranks of the archaizers
of his time, and would not have exercised the undoubted influence that he did.
This is why the remainder of the preface first of all turns vehemently against
precisely those archaizers, and then settles down to outline the programme of
moderate ‘correction’ with which it will conclude. Heliodoros, according to
Korais, is to be admired and upheld as a model, not literally for imitation (Korais is
hard on imitators),46 but rather for his judicious exercise of linguistic choice within
the possibilities available to him at a specific moment in history. It is Heliodoros’
bee-like versatility in harvesting the ‘flowers’ of the past that earns him Korais’
admiration; and implicitly, this is what his successors are to be encouraged to
emulate.
It is in order to flesh out this understanding, and to explain fully what he
understands by Heliodoros’ linguistic practice, that Korais moves finally into the
last, and best-known, part of his preface, in which he addresses the modern language
question directly. What Korais calls for, in the concluding pages, which have usually
been cited out of context, is a modern language that will not be identical to that of
any earlier period, but rather the equivalent of the ‘brilliant garment, woven in a
single piece, from top to bottom,’ that had been artfully created, a millennium and
a half before, by Heliodoros, in laying the foundations of the subsequent European
novel.
It has sometimes been suggested that modern Greek diglossia goes all the way
back to the Second Sophistic.47 But this is to simplify a complex and continuously
evolving situation that existed through the intervening centuries. Diglossia itself
certainly cannot be laid at the door of Korais. Indeed, strictly understood as
the formal separation of a language into two, and not more than two, registers
appropriate to specific kinds of usage, the existence of diglossia in Greek can
45
See, indicatively, several of the essays in Mander (2007).
46
Korais (1984) 26–8.
47
See e.g. Browning (1983) 4 and 44–50; Alexiou (2002) 19–42.
be very precisely dated to the period 1911–76, when successive constitutions

prescribed katharevousa, although without naming it, as the official language of
the Greek state.48
The reading of the influential preface to the Aithiopika attempted here suggests
that Korais’ conservative programme for reform of the modern written language
was directly inspired by the example of Heliodoros. Korais credited the author of
the Aithiopika with the inauguration of a literary genre that had gained enormously
in prestige over the centuries, and in his own day seemed to promise much for the
future, not least for the Greeks as a modern nation, as conceived by Korais from
the perspective of post-Revolutionary Paris. In presenting this text to his readers,
in the original language, and with a linguistic commentary in which he compared
and contrasted that language with contemporary usage, Korais at the same time
proposed a model for the literary language of the future Greek state. That model
was to be found not in the specific linguistic choices made by Heliodoros in the
fourth century, but rather in that author’s judicious eclecticism and profound
familiarity with the Greek of earlier periods.
In this way, Korais’ prescriptions for the Greek language have to be seen, not
as the arbitrary imposition of the self-appointed apostle of the Enlightenment,
but as grounded in a conviction of the essential interconnectedness of language,
literature and the emergent idea of the nation, which is an essential component of
Romanticism.
In attempting to formulate, in his own self-taught way, his own version of the
emerging Romantic quest for a national language and a national literature, Korais
had recourse, not to the classical age or its generic theory, which had already
become fossilized by neoclassical conventions, but to the ‘new’ genre of the novel,
and the linguistic practice of its most accomplished practitioners, at the time of the
Second Sophistic.
References
Alexiou, M. (2002), After Antiquity: Greek Language, Myth, and Metaphor, Ithaca, NY.
Anderson, G. (1993), The Second Sophistic: A Cultural Phenomenon in the Roman Empire,
London.
Beaton, R. (1991), ‘Koraes, Toynbee and the Modern Greek heritage’, Byzantine and
Beaton, R. (1996), The Medieval Greek Romance, 2nd edn, London (1st edn, Cambridge,
1989).
Beaton, R. (1999), An Introduction to Modern Greek Literature, 2nd edn, Oxford.
48
Beaton (1999) 318–19 and 337–8.
352 roderick beaton
Beaton, R. (2006), ‘Ήρωες με πολιτικά φρονήματα. Οι αδελφοί Σούτσοι, η

«μυθοπλασία» και το νεοσύστατο ελληνικό βασίλειο’, in P. Voutouris and G. Georgis
(eds.), Ο ελληνισμός στον 19ο αιώνα. Ιδεολογικές και αισθητικές αναζητήσεις,
Athens: 108–13.
Beaton, R. (2007), ‘The Greek novel and the rise of the European genre’, in Mander (2007):
225–33.
Beaton, R. and Ricks, D. (eds.) (2009), The Making of Modern Greece: Nationalism,
Romanticism, and the Uses of the Past (1797–1896), Aldershot.
Browning, R. (1983), Medieval and Modern Greek, 2nd edn, Cambridge (1st edn, London,
1969).
Chaconas, S.G. (1968), Adamantios Korais: A Study in Greek Nationalism, New York (first
published 1942).
Chrysanthopoulos, M. (2004), ‘Η πολυπαθής μυθιστορία: ορολογία, φιλολογία,
λογοτεχνία’, in Μνήμη Άλκη Αγγέλου: τα άφθονα σχήματα του παρελθόντος,
Clogg, R. (1986), Politics and the Academy: Arnold Toynbee and the Koraes Chair, London
(also published as Middle Eastern Studies, 21: 1–117).
Dimaras, K. (ed.) (1953), Ο Κοραής και η εποχή του, Athens.
Hägg, T. (1983), The Novel in Antiquity, Oxford.
Henderson, G.P. (1970), The Revival of Greek Thought: 1620-1830, Albany, NY.
Horrocks, G. (1997), Greek: A History of the Language and Its Speakers, London.
Huet, P.-D. (1966), Traité de l’origine des romans, Stuttgart (facsimile reprint of 1st edn,
Paris, 1670).
Korais, A. (1984), ‘Τα εις την έκδοσιν (1804) των Αιθιωπικών του Ηλιοδώρου
Προλεγόμενα’, in A. Korais, Προλεγόμενα στους αρχαίους Έλληνες συγγραφείς,
vol. 1, Athens: 1–56.
Liakos, A. (1996), ‘Language and history in modern Greece’, in A.-P. Christides (ed.), The
Greek Language, Athens: 75–82.
Mackridge, P. (1985), The Modern Greek Language, Oxford.
Mackridge, P. (2009), Language and National Identity in Greece, 1766–1976, Oxford.
Mander, J. (ed.) (2007), Remapping the Rise of the European Novel, Oxford.
Morgan, J.R. (1997), ‘Introduction,’ in Heliodorus, Ethiopian Story, London.
Moschonas, E. (1981), ‘Αγώνας για μια χαμένη υπόθεση’, introduction in I. Vilaras,
D. Psalidas, A. Christopoulos et al., Η δημοτικιστική αντίθεση στην Κοραϊκή «Μέση
Οδό», Athens.
Moullas, P. (1991), ‘Ο Κοραής και ο αυτοσχέδιος κριτικός στοχασμός του’,
Μολυβδο-κονδυλο-πελεκητής, 3: 125–37.
Psycharis, J. (1971), Το Ταξίδι μου, ed. A. Angelou, Athens (first published 1888).
Reardon, B.P. (1989), ‘Introduction’, in B.P. Reardon (ed.), Collected Ancient Greek Novels,
Berkeley: 1–16.
Roidis, E. (1988), Η Πάπισσα Ιωάννα, ed. A. Angelou, Athens (first published 1866).
Roilos, P. (2003), ‘The poetics of mimicry: Pitzipios’ “Ο Πίθηκος Ξουθ” and the
beginnings of the Modern Greek novel’, in G. Nagy et al. (eds.), Modern Greek Literature:
Critical Essays, New York: 62–78.
Solomos, D. (1986), Διονυσίου Σολωμού, Άπαντα, ed. L. Politis, vol. 2, Athens.
Tonnet, H. (1994), ‘Roman grec ancien, roman grec moderne. Le cas de l’Orpheline de Chio
de J. Pitsipios (1839)’, Revue des Etudes Néo-Helléniques, 3: 23–39 (reprinted in idem,
Etudes sur la nouvelle et le roman grecs modernes, Paris and Athens, 2002: 29–45).
Tonnet, H. (1996), Histoire du roman grec (des origines à 1960), Paris.
Tonnet, H. (2005), ‘Lieux antiques et modernes dans l’Orpheline de Chio ou la Triomphe de
la Vertu (1839) de Jacques Pitsipios’, in B. Pouderon (ed.), Lieux, décors et paysages de
l’ancien roman des origines à Byzance, Lyons: 101–10.
Index of Names
In the index that follows, modern names are listed with abbreviated forenames
(‘Aitchison, J.’), except for writers, and the like (‘Shakespeare, William’), for whom
mechanical abbreviation of forenames would be artificial. Premodern names are cited
in normal form (‘Achilles Tatius’).
Transliteration of Greek names: except in a few cases, where alternative forms are
given, the form cited is the only one (or usual one) used in the volume. Cf. p. vii above.
Achilles Tatius, 22 Barnes, William, 141

Aelius Dionysius, 102 Bateson, F.W., 140
Aeschines Socraticus, 100 Beaugrande, R. de, 323
Aeschylus, 161 Blommaert, J., xvi, 295, 299
Aesop, see Life of Aesop Bourdieu, P., xxiv, 281, 291, 299
Aitchison, J., 259 Bowlly, Al, 13–14, 17
Alcman, 14, 17, 29 Bridges, Robert, 150
Ammonius, 102 Burchfield, R., 6–7
Andreopolos, A., 267
Angelou, A., 138, 147 Caecilius, 97
Antiatticista, 97, 100 Callimachus, 22
Antiochus, 96 Cameron, D., xii, 5–7, 168, 180, 304
Antiphon, 102 Cantharus, 101
Antisthenes, 100 Cavafy, Constantine (Kavafis, K.P.),
Aristides, Aelius, 22–3, 98 xviii–xix, 132, 138–42, 144–5,
Aristides Quintilianus, 4, 28 149–66
Aristophanes, 10, 18, 34, 37, 100, 102 Cervantes, Miguel de, 350
Aristotle, 14, 24–5, 28, 121, 125 Charis, G., 326
Austen, Jane, 150, 153 Choeroboscus, 28
Choniates, Niketas, 349
Babiniotis, G., xix–xx, xxii, 167–82, Christidis, A.-Ph. (A.-F. / A.), xi, xxv,
256–7, 272, 297, 299, 322 321–35
Bailey, R.W., 9 Christopoulos, A., 254, 263–6
Banks-Smith, N., 149 Chrysostomos, Archbishop, 140
355
356 INDEX of names
Chrysovergis, G., 261, 266 Fatseas, A., 268

Cinaethon, 15–16 Favorinus, 99
Clare, John, 135 Ferguson, C.A., 36, 82, 111
Clement of Alexandria, 11, 264 Filintas, M., 268–9
Coleridge, Samuel Taylor, 149 Filippidis, D., 264–5
Cornelianus, 98–9 Foskolos, Markos Antonios, 132
Coulmas, F., 336 Foucault, M., 285
Crates, 101 Frangoudaki (Fragoudaki), A., vii,
Cratinus, 101 273, 321, 323, 325–7, 329–35
Critias, 100
Galen, 28, 97–8
Dante Alighieri, 41 Gazis, A., 254
Darvaris, D., 118 Gershwin, George and Ira, 13
Delmouzos, A., 47 Givón, T., 112, 115, 120
Demetrius, 28 Glynos, D., 47
Demosthenes, 20, 99–100, 102–3 Glytzounis, M., 117
Dickinson, Emily, 143 Goethe, Johann Wolfgang von, 346,
Dingas, D., 47 350
Dio Chrysostom, 16 Grimm, J., 9, 278
Diogenianus, 102
Dionysius of Halicarnassus, 24, 28, Hadrian of Tyre, 103
95, 271 Hall, J.M., 11–12, 16, 19
Diphilus, 101 Halliday, M.A.K., 3, 8, 109–10, 112–
Doukas, N., 262–3 13, 116–18, 122
Dryden, John, 25 Hardy, Thomas, 136–7, 150
Harpocration, 97
Eagleton, T., 294–5, 327–9, 335 Harris, R., 8
Eliot, T.S., 25 Hatzidakis, G.N., xxiii–xxv, 263–5,
Elladios, Alexandros, 254, 333 269–71, 275, 278–91
Empedocles, 16 Hecataeus, 29
Epicharmus, 18 Heilmaier, J.M., 265
Erotokritos, 267 Heliodoros (Heliodorus), vii, xxvi, 22,
Eugenianos, Niketas, 349 342–51
Eupolis, 101 Heraclides Criticus, 10–11
Euripides, 18, 20 Hermogenes, 28–9
Eustathios, 348–9 Herodian, 28
Herodotus, 11–12, 16, 28, 39, 96, 120
Index of names 357
Hesiod, 3, 10–11, 14–15, 29, 35 Labov, W., 6–7

Hesychius, 103 Landos, Agapios, 254, 262
Hippocrates / Hippocratic corpus, 17, Larkin, Philip, 149
25, 27, 96, 99 Le Page, R.B., 236
Homer, 3, 10–11, 14–16, 23–8, 35, Libanius, 43, 343
38, 42, 101, 122, 131, 139, 141, Life of Aesop, xvii, 109–29
348 Loraux, N., 19
Hopkins, Gerard Manley, 141, 149–50 Lord, A.B., 37
Horrocks, G., xi, 16, 22–4, 38, 145 Lucian, 22–3, 96–7
Huet, D., 345 Luraghi, N., 41
Hymes, D., 5, 337 Lysias, 18, 35
Hynes, S., 150
Hypereides, 102 Makriyannis, General Y., xviii, 133–5,
144
Isaeus, 102 Maniakes, N., 262
Isocrates, 19–20 Manzoni, Alessandro, 350
Mavrofrydis, D., 263, 279, 332
Jakobson, R., 8, 26–7 Meillet, A., 38, 40, 94
Johnson, Samuel, 98, 261 Menander, 22
Joseph, J.E., 274 Meniatis, Elias, 254
Milroy, J. and L., 43, 168, 180, 293
Kalvos, Andreas, 131, 138, 142 Mistriotis, G., 263, 270–1, 283–4
Katartzis, D., 254, 261, 263–4, 269, Moeris, xvii, 97, 100–4
271 Moisiodax (Misiodakas), I., 254, 261
Kodrikas, P., 260–4, 333 Morpurgo Davies, A., 16, 37
Komnene, Anna, 349 Moullas, P., 346
Konemenos, N., 259
Konstantas, G., 254, 261, 264, 266 Nikiforos, T., 118
Kontos, K., 265, 283–4
Korais (Koraes), Adamantios ‘Old Oligarch’, 20, 40
(Adamandios), vii, xxiii, xxv–xxvi, Orwell, George, 4
106, 254, 260–1, 263–72, 284,
341–53 Palaiologos, Grigorios, 346
Koumas, K., 254 Palamas, Kostis, 141–2
Kriaras, E., xix–xx, xxii, 167–85, Pallis, A., 69, 145
253–8 Palmer, L.R., 94
358 INDEX of names
Papadiamantis, Alexandros, 131, Sapir, E., 33

135–9, 142, 144 Sappho, 17
Parry, M., 16, 35, 38 Savidis, G.P., 151–5, 157–8, 160
Pausanias, 4, 102 Sebba, M., 236
Petridis, P., 261 Seferis, George, 133, 138, 145, 151
Petrounias, E., 175–7 Sengopoulou, R., 152–4, 157–8, 160
Pherecrates, 101 Shakespeare, William, 7–9, 140
Philemon, 102 Silverstein, M., 294–5, 301
Philetairus, 102 Skokos, K., 155–6, 162–5
Philippidis, D., 254 Skoufos, Frangiskos, 254
Phillips, C., 149–50 Sofianopoulos, M., 261
Philostratos (Philostratus), 103, 347 Sofianos, Nikolaos, 253
Photiadis, L., 254 Solomos, Dionysios, xviii, 131, 138,
Phrynichus, xvii, 95, 97–104 140, 142–4, 254, 343, 349
Pindar, 3, 14, 17, 27, 35 Solon, 3, 19, 103
Pitsipios, Iakovos, 346 Sophocles, E.A., 260
Plato, 10–11, 18, 23, 27, 33–4, 39, 55, Soutsos, Alexandros, 346
100, 102–3, 343 Soutsos, Panayotis, 269, 346
Plato Comicus, 101 Spyridakis, C., 198–200, 203
Plutarch, 18, 21–4, 344 Stasinus, 15–16
Politis, L., 144 Strabo, 10–11
Pollux, xvii, 95, 100, 103–4 Strattis, 101
Polybius, 22, 93 Stubbs, M., 323
Polylas, I., 142–4 Suda, 29, 98
Posidippus, 21 Swain, S., 95, 97–8, 101–2
Praxagoras, 18
Psalidas, A., 254, 265, 272, 274 Tennyson, Alfred, 150
Psycharis (Psichari / Psychares), Theocritus, 4, 131
I / J./ Y., 132, 254, 263, 269, 274, Theodore Prodromos, 349
280, 285–7, 342–3 Theophrastus, 26
Theotokis, N., 254
Rangavis, K., 270–1, 276 Threatte, L., 40
Rangousis, N., 272–3 Thucydides, 17–19, 20, 24, 28, 100,
Roidis (Roïdis), E.D., 137–8, 259, 342 102
Ross, R.H., 150 Tonnet, H., 346
Index of names 359
Triantafyllidis (Triandafyllides / Vilaras, I., 134, 254, 261, 268, 271

Triandaphyllides), M., 47, 53, 133, Vlachoyannis, I. / Y., 133–5
145, 255–7, 269, 326, 332 Voulgaris, Evgenios, 254, 261
Triantafyllopoulos, N.D., 135–6 Vyzantios, S., 269
Trikoupis, S., 263, 276
Tritsis, A., 273 Whitmarsh, T., 96
Trudgill, P., 3–4, 8, 12, 14, 326 Wittgenstein, Ludwig, 33–4
Tucker, Sophie, 13 Wordsworth, William 22–3, 29
Typaldos, Ioulios, 259, 268
Tzartzanos, A.A., 84, 330, 332 Xenophanes, 11
Xenophon, 18, 100, 102–3, 273
Valtinos, Thanasis, 146
Vamvas, N., 254, 262, 266 Zampelios, I., 267–8
Vasiliou (Vasileiou), A., 266, 271, 342 Zampelios, S., 268
Venthylos, I., 265 Zgusta, L., 180–1
Vernardakis, D.N., 265
Subject Index
Abstract/concrete, see language Atticism, xvi, 22–4, 27, 93–107, 265

Académie Française, 343 Grammatical/stylistic Atticism, 97
Accentuation, see orthography
Aeolic, 9, 11, 15, 264–5, 279 Behaviour, see linguistic behaviour
Aeolodoric, 263–5 Bible, 68, 138 (see also New
Aggregative/oppositional, xiii, 11–12 Testament, Old Testament)
Albanian, 51, 73, 76 Bilingualism, 74–6, 79, 299, 308, 316
Alphabet, xxi, 37–8, 40, 51, 68, 70, Boeotian, 10, 15, 38–41
74, 78, 132, 221–49, 297, 299– Bulgarian, 51
300, 303, 305, 307, 309–11 (see Byzantine (language), see Greek
also Latin-alphabet Greek)
American (English)/Americanizing, Calque, 80, 299, 344
xiii, 6–7, 12–14, 17, 140 Canon/canonization, xviii, 23, 34, 77,
Ancient Greek, see Greek 133, 337
Anti-purism/anti-purist, see purism/ Catalan, 77
purist Chancellery (language, style), 40–1,
Arabic, xii, xiv, 34–6, 41–2, 80, 86, 43
221, 227–8, 244, 246 Chinese, 73
Arbresh, 76 Civic nationalism, see nationalism
Arcado-Cypriot, 9 Codification, xiii, xvi, 77, 97, 101–2,
Archaism/archaizing, xv, xvii, 179
15–16, 18, 24, 26, 53, 67, 69, 77, Colonial, see language
82–6, 162, 173, 175, 181, 261–72, Comedy, 16, 21, 23–4, 35, 39, 43,
283–4, 298, 316, 341, 350 100–2, 132
Armenian, 51 Community, see linguistic community
Arvanitika, xvi, 73 Concrete, see abstract
Athenocentricity, 18–23 Continuity, xi, xxii, xxv, xxvii, 76, 83,
Attic, xiii–xiv, xvi, 3, 10–11, 15, 177, 259–76, 330, 333, 336–7
18–27, 35, 39–42, 93–107, 264–5, Corpus/corpus linguistics, xxv, 77,
272–3 (see also Great Attic) 194, 226, 303–6, 321, 323–4
Attic-Ionic, 9–10, 21–2, 24, 42, 94 Cretan, 53, 72, 262
361
362 Subject INDEX
Croatian, 73, 79, 140 Doric, xiv, 4, 10–11, 14, 16–18, 96,
Cypriot Greek, xx, 145, 187–219, 301 264–5, 279
Cyprocentrism, xx, 188–205
Editors/editing practices, xviii–xix,
Danish, 73, 170 131–47, 149–66, 316
Demotic/katharevousa, xv, xix, xxiii– Education, see schools
xxv, 47–8, 53–5, 61–3, 67, 69, 72, Education reform, 47–9, 55, 56–60,
77, 80, 84, 86, 137–8, 145, 155, 62, 200
160, 162, 169, 171, 173–4, 182, Elite, 55, 57, 63, 75, 301, 316
191, 200, 226, 261–73, 284–5, E-mail, xv, 221–49, 297
287, 294, 298–9, 311, 315–16, English (language), xx–xxi, xxiv, 6–9,
321, 325, 330–2, 334, 337, 343, 12–14, 21, 35, 39, 43, 71, 75, 80–1,
350–1 (see also language question, 135, 138, 140, 145, 188–210, 274,
vernacular) 299–300, 302, 307, 310–11, 349–
Demoticism/demoticist, xix, xxiv, 50 (see also American, ‘Received
53–5, 254, 256–7, 268, 288, 298, Pronunciation’, ‘Saxon’)
315, 332 (see also vernacularism/ Epic poetry/dialect, xiv, 14–19, 23–7,
vernacularist) 35–9, 41–2
Density, lexical, 109, 122 Epigraphy, xiv, 11, 16–17, 40–1, 43,
Dialect, xiii–xvi, 3–4, 6, 9–11, 34–6, 133
50, 52–3, 62–3, 67–89, 134, 141, Erasure, xxvi, 300
171, 175, 198, 200, 260–5, 300, Estonian, 80
332, 343 Ethnicity, 54, 60, 260
Dialect composites, 3, 14–19, 25–7 Ethnocentricity, xxv, 72, 187–205,
Dialects, Ancient Greek, xiii–xiv, 321–2, 327, 334–5 (see also ethnic
9–29, 34–41, 72, 83, 93–105, nationalism)
263–5 (see also Aeolic etc.) Etymology, xviii–xix, xxi-xxii, 101,
Dictionary, see lexicography 105, 136, 176, 178, 315
Diglossia, xii–xiii, xviii, xx–xxii, xxiv– Exceptionalism, xi, xviii, xxvii, 268,
xxv, 23–4, 34, 36, 67, 168, 170, 272–3
173, 176, 181, 227, 294–8, 308,
311, 321–39, 350 Feminism, see gender
Digraphia, xxi, 221–49 Finnish, 80
Discourse/discourse analysis, vii, xx, Foreign words, see loanwords
xxv–xxvi, 221–6, 228, 232, 240–2, Formal/informal language, 78, 223,
244, 246–7, 321–3, 327, 329–32, 225, 261
334–7
Subject Index 363
French, xiii, 21, 39, 41, 75, 79–81, 344–5, 347–9

263, 265, 268–9, 349–50 Modern, xv, xix–xxvii, 23, 36,
47–65, 67–86, 109–11, 113–14,
Gender, xxiii–xxiv, 6 117, 119–20, 124, 131–85, 188,
Generic dialectalization, xiii–xiv, xvii, 200, 202, 221–49, 253–353 (see
13–19 also Standard Modern Greek)
Genre, xii–xiv, xvii, xix, xxiv, 14, Greeklish, xx–xxi, 221–49
18, 24, 35, 38, 95–6, 228, 295–6, Grico, 76
301–4, 307, 315, 324 (see also
comedy etc.) Hebrew, 77–9, 86
German/Germanic, 9, 21, 75, 79–80, Hellenes/Hellenic/Hellenicity, xiii,
105, 326, 350 (see also ‘Saxon’ xxiii, 6, 11–12, 67, 97, 102–3, 134,
English) 197–201, 203, 229, 234, 259–69,
Glotta, 14–15, 28 271
Graikoi, 260 Hellenocentrism, xx, 187–205
Grammarians, 93, 105, 131, 135 High/low, xv, xviii, xxiv, 18, 36,
Ancient grammarians, 10–11, 21, 61, 67, 71, 81, 83, 111, 122–3,
28–9, 34, 93–105 198, 246, 264, 295, 298 (see also
Medieval grammarians, 34 register)
Modern grammarians, 5–6, 68, 78, Historiography/history (writing),
145, 266, 268–70 18–19, 22, 24, 28–9
Neogrammarians, xxv, 278, 343
Grammatical metaphor, xviii, 109–29 Icelandic, 79
Great Attic, 20–22, 40 Iconicity, xvii, xxi, xxiv–xxv, 233,
Greek: 246, 306
Ancient/Classical, xiii–xiv, xix, Identity, xiii–xv, xvii, xxiii, 11–12,
xxii, xxiv, xxvi, 3–45, 65, 68, 14, 36, 38, 41–3, 67, 73–4, 78, 81,
72, 77, 80–1, 83–6, 93–107, 96–8, 187–219, 222, 224, 226,
110, 113–15, 119–20, 125, 132, 236, 243, 247, 260, 281–2, 309,
138–9, 142, 145, 170–1, 176–9, 315, 322
198, 200–2, 205, 259–74, Ideology, xi, xiii, xvi, xix, xxi–xxvii,
296–8, 309–11, 322, 341–53 23, 41, 58, 105, 134, 145, 167–85,
(see also dialect) 189, 200, 244, 246, 272, 293–5,
Byzantine/Medieval/Early 299, 304, 306–9, 311, 313–16,
Modern, xvii–xviii, 80, 85, 321–37, 342
110–29, 131, 137, 142, 170, Illiteracy, see literacy
176, 260, 265–6, 271, 273, 280, Imitatio/imitation, 80, 95, 97, 347–8,
364 Subject INDEX
350 Language contact, xxiv, xxvii, 79,

Indexing, xxiv, 295–315 294, 299, 311, 315, 337
Indo-European (languages), 237, Language death/extinction, xv–xvi,
278–9 68, 76–9
Informal, see formal Language, exterior/interior, xxiv, 295,
Inscriptions, see epigraphy 307–14
Internet, xx–xxi, 77, 221–49, 321 Language guardians, 105
Ionic, 10–11, 15–20, 22, 24–9, 39, Language issues, xxii, 79, 293–320
41–3, 93–6, 99, 264 Language, literary, see literature
(see also Attic-Ionic) Language manuals, 97, 299
Ionic as ‘poetic’, 27–9 Language metaphors, xxii–xxiv, 226,
Irish, 78–9 259–74, 300, 315, 323
Italian, 4, 36, 39, 41, 76, 81, 85–6, Language naming, 41–3, 259–76
138, 263, 265, 326, 350 Language of the people, 193, 233
Language, official, 47–8, 54–5, 57,
Japanese, 86 69–70, 72–4, 76, 78, 188, 190–6,
204, 271, 294, 297–8, 300, 302,
Katharevousa, see demotic 310–11
Koine (Hellenistic), xiii–xiv, xvi, 3, Language planning, xx, xxii, 187–219,
9–11, 15, 18, 20–4, 33–45, 52, 85, 337
93–105, 139, 261–4, 280, 282, 348 Language question, xii, xxii, 111, 144,
koinai, xiv, 37, 42 170, 253–7, 266, 288, 293–320,
Kunstsprache, 15, 23–4, 35, 38, 95, 342, 350 (see also demotic)
139 Language reform, xxii, xxiv, 48, 55,
72, 74, 253–7, 294, 296, 300, 342,
Laconian, 4, 14, 17, 39 347–51 (see also reform)
Ladino, 51 Language, spoken/written, xiv, xxiv,
Language, abstract/concrete, 24–7, 78 37, 39, 52, 63–4, 68–86, 94–5, 113,
Language as essence, 33–4, 43 119, 168, 171–3, 180–2, 221–49,
Language as garden, 7 259–73, 283, 286–8, 296, 301,
Language as monument/edifice, 306, 332, 336 (see also vernacular)
xxiii, 134, 269–71 (see also Language, unwritten, 72, 76
monumentalization of language) Latin/Latinity, 6–7, 21, 34, 36, 39,
Language as possession, 343–4 41–2, 74, 77, 98, 260–1, 263, 265,
Language as treasure, 178, 181 268
Language, colonial, 25–6, 35–6, 192 Latin-alphabet Greek (LAG), xx, 132,
140, 143, 221–49, 300
Subject Index 365
Law-courts, xx, 59, 78, 187–95, 197, Macedonian, 20–1, 51, 297, 308–11
202, 204–5, 210 Media, 74, 171, 176, 182, 222,
Laz, 69, 76 226–8, 236–7, 240, 242, 294–8,
Lesbian, 10 301, 305, 308–15 (see also internet,
Lexical density, see density newspapers)
Lexicographers/lexicography, xvii, Medieval Greek, see Greek
xix–xx, xxii, 6, 93–107, 131, 137, Medical writing, see science
156, 160, 167–85, 253–7, 295, Metalanguage/metalinguistic, xxi–
297, 309, 316, 343 xxii, xxiv–xxv, 5, 222–3, 231, 240,
Lexicon, see vocabulary 242, 246, 294, 304, 309, 312, 316,
Lingua franca, 21, 23, 42, 76 321, 326, 337
Linguistic behaviour, 180–1 Metaphor, see grammatical metaphor,
Linguistic community, 168–9, language metaphor
179–80, 182, 239, 312, 315 Modern Greek, see Greek
Linguistic nationalism, see nationalism Moldavian, 73
Linguistics, modern, xvii, xix, xxiii, Monotonic/polytonic, see orthography
xxv, 3–10, 94, 221, 268, 272, Monumentalization (of language),
277–91, 300, 306, 316, 321–39 xviii–xix
(see also sociolinguistics, systemic Moral panic, xxi, xxiv, 226, 240, 242,
functional linguistics) 296, 298, 302, 304–15
Literacy/illiteracy, xviii, xx-xxi, 39, Multilingualism, 301, 325
43, 54–7, 61–3, 133, 175, 223, Mycenaean, 9–10, 37
228, 246, 337 Myth, xxv, 321–39
Literature/literary language, xiii–xiv, Mythistorema/mythistoria, 342, 345–6
xvi–xix, xxvii, 3–31, 34, 67–8, 72,
77, 110–25, 135, 140, 157–62, National identity, see identity
175, 180, 188, 200, 223, 267, 295, Nationalism, civic, xx, 187–205
299, 303, 341–53 (see also poetry) Nationalism, ethnic, xx, 187–205 (see
Literature, oral, xviii, 13, 15, 35, also ethnocentricity)
Lithuanian, 79 Nationalism, linguistic, 277
Loanwords, xx, xxiv, 72, 79–81, 86, Neogrammarians, see grammarians
172, 174, 177, 202, 297–9, 305, Neologism, xix, 172, 177, 182, 263
307–11, 316, 322 Newspapers, xxiv, 6, 49, 52, 58, 68–9,
Low, see high 72–3, 136, 165, 188, 295–303,
Lyric (poetry), xiv, 12–14, 16–19 307, 315, 324, 342
Lycian, 42
366 Subject INDEX
New Testament, 22, 68–9, 93–4, 122, Prescription/prescriptivism/

139, 273 prescriptivist, xii, xxvii, 3, 5, 68–9,
Norms, 73, 82, 197, 223, 229–30, 74–5, 79, 93–4, 97, 101–2, 105,
233, 236, 239, 246, 293–4, 301, 178, 315, 322, 330, 335, 344, 351
336 Prose, see poetry
Norwegian, 73 Purism/purist/anti-purism/anti-purist,
Novel: xx–xxi, xxiii–xxv, 5–6, 11, 47,
Ancient/Hellenistic, 22, 341–53 52–5, 57, 61–3, 67, 69, 79–82,
Modern, 261, 342, 346, 350–1 84–6, 95, 98, 100, 105, 168–75,
177, 181–2, 262, 267, 270–1, 282,
Official language, see language 297–9, 314–16, 321–3, 326–7,
Old Testament, 22 330–1, 334–5
Oppositional, see aggregative
Oracles, 16, 22 Re-borrowings, xx, 177–8
Oral/orality, see literature, oral; ‘Received Pronunciation’ (RP)
language, spoken English, 13
Oratory, 18–19, 23, 35, 95–7, 100, Rechtschreibereform, 105–6
102–3 Reflexivity, 337
Orthography, xvi, xviii–xxii, 37–40, Reform, see language reform,
70, 72, 82, 84, 131–47, 150–65, orthography, Rechtschreibereform
168, 176, 178–80, 221–49, 261, Refugees, 47–65, 67, 70, 82, 84
263, 297–300, 305, 308–11, 322 Register, xvii–xviii, 78, 109–29, 173,
179–81, 261, 298, 316, 332 (see
Paideia, 19, 21, 39, 43, 96, 104 also high/low)
Pan-Hellenic (consciousness), 16, 144 Rhetoric/rhetorical theory, 25, 34–5,
Participles, xviii, 109–29 103, 298, 315
Persian, 80 Romance (languages), 9, 41–2, 80,
Philosophy, 18, 23–4 263 (see also French etc.)
Poetry/poets, xiv, xviii–xix, 3–31, Romanian, 36, 73, 80–1, 263, 265
37–9, 105, 131–47, 149–66, 188, Romansh, 76, 78
223, 264, 268, 270, 298 Romany, 72
Poetry and prose, 17–29, 35, 43, Romayka, xv, xx, 69–86
93–7, 99, 133–8, 149 Romaic/Romaioi/Romioi, xxiii,
Polytonic, see orthography 260–1, 264–5, 268
Pontic, xv, 52–3, 67–89 Russian, 72, 76, 246
Popular song, xii–xiii, 12–14, 17
Subject Index 367
‘Saxon’ English, 7 Transliteration, xv, xxi, 221–49

Schools, xv, xx, xxii–xxiv, 47–65, 68, Tsakonian, 71
72–6, 78, 104–5, 135, 145, 168, Turkish, xv–xvi, 49–54, 69–76,
171, 173, 177, 181–2, 189, 199, 78–82, 86, 188–205, 223–4, 268,
223, 229, 236, 242, 261–2, 266, 300
272–3, 296, 298, 311, 320, 322,
337 Ukrainian, 76
Schwyzertüüch, 75
Science, language of, 17, 25–7, Value, xii–xiii, xxiv, 3–9, 54, 68, 82,
109–29 171, 198, 239–41, 265, 271, 274,
Second Sophistic, xvi–xvii, xxvi, 23, 331, 333, 336
93–107, 265, 341–53 Variant/variation, 77–8, 82, 222,
Septuagint, see Old Testament 229–42, 246, 268, 312, 331, 336–7
Serbian, 73, 140 Vernacular, xv, xviii, xx, 6–7, 10–11,
Slavic/Slavonic, 81, 265 (see also 14–15, 18, 20–4, 36, 40, 42, 47–8,
Bulgarian etc.) 53, 62–3, 67–86, 94, 170, 173, 175,
Sociolinguistic(s), vii, xiii–xiv, xx, 229–30, 236, 246, 264–5, 277–8
3–5, 12, 36, 69, 84, 167, 173, 175, (see also demotic)
222, 230, 245, 285, 293 Vernacularism/vernacularist,
Spanish, 35–6, 42, 53–4, 77, 263, 261, 263–5, 268–72 (see also
265, 350 demoticism/demoticist)
Spelling, see orthography
Standard Modern Greek (SMG), 76, Vlach, 51, 73
80, 188–205, 293, 298–320 Vocabulary/lexicon, xxii, 38, 77–8,
Style, see literary language 80, 85, 94–104, 115–16, 122,
Swedish, 78–9 167–8, 171, 175, 181, 202, 262,
Systemic Functional Linguistics 272–3, 298
(SFL), vii, xvii, 109–12, 121, 124
Welsh, 79
Tamil, 79 Written language, see language
Technical prose, see science
Thessalian, 15, 21 Yiddish, 77
Tragedy, 18, 21–4, 96, 99–102, 132
Centre For Hellenic Studies
King’s College London
Publications 12

Standard Languages and Language Standards - Greek, Past and Present-Ashgate (2009) PDF

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Standard Languages and Language Standards - Greek, Past and Present-Ashgate (2009) PDF

Caricato da

Copyright:

Formati disponibili

Standard Languages

and Language Standards:

Ashgate website http://www.ashgate.com

British Library Cataloguing in Publication Data

US Library of Congress Cataloging-in-Publication Data

Typeset by W.M. Pank, King’s College London.

THE CENTRE FOR HELLENIC STUDIES,

Introduction: Greek Language-Standardizing, Past, Present and Future xi

Part I Establishing a Standard

Part II Standardization Practices

Part III Ideologies and Contestations

Index of names 355

Subject index 361

As is made clear in A.G.’s Introduction (pp. xi–xxviii below), and as will be

Alexandra Georgakopoulou is Reader in Modern Greek Language and Linguistics

Michael Silk is Professor of Classical and Comparative Literature at King’s College

Jannis Androutsopoulos is Reader in Sociolinguistics and Media Discourse at King’s

Roderick Beaton is Koraes Professor of Modern Greek and Byzantine History,

Pietro Bortone is Assistant Professor in Modern Greek Studies at the University of

Stephen Colvin is Reader in Classics and Comparative Philology at University

Effi Gazi is Assistant Professor of History and Theory of Historiography at the

Dionysis Goutsos is Assistant Professor of Text Linguistics at the University of

Anthony Hirst is Lecturer in Modern Greek at Queen’s University Belfast.

Anna Iordanidou is Associate Professor of Modern Greek Language at the

Dimitra Karoulla-Vrikki is Assistant Professor of Linguistics at the European

Emmanuel Kriaras is Honorary Professor of Medieval Philology at the Aristotle

George Kritikos is Associate Professor in Historical Geography at Harokopion

Peter Mackridge is Emeritus Professor of Modern Greek, and an Emeritus Fellow of

Spiros Moschonas is Senior Lecturer in Linguistics and Philosophy of Language at

David Ricks is Professor of Modern Greek and Comparative Literature at King’s

Claudia Strobel is a doctoral student at Corpus Christi College, Oxford.

Assimakis Tseronis is a Junior Researcher at the University of Leiden.

disjunctures, fragmentations and discontinuities in a never-ending and multi-

Part 1: Establishing a Standard

reminiscent of the way that different contemporary communities of e-mail users

Part II: Standardization Practices

to older forms of language (archaizing) is in one form or another a leitmotif of the

by showing, for instance, that any editor of Thomas Hardy, a contemporary of

So too is the treatment of semantic borrowings and re-borrowings (notably

Part III: Ideologies and Contestations

of katharevousa – concealing, but ultimately revealing, the usual ideological

be inconceivable without the ideology of purism that marked the history of

In our collection overall, a wide variety of language phenomena from different

history of the Greek language ultimately (re)constructed and interwoven into a

Le Page, R. and Tabouret-Keller, A. (1985), Acts of Identity: Creole-Based Approaches to

The Invention of Greek:

But (she adds):

In a revisionist spirit, Cameron argues forcefully, and surely rightly, that

Curious indeed, and perhaps unthinking – and, if so, especially revealing.

and between versions of literature, and being prepared to do so on the basis of

The extraordinary course of Athenian cultural imperialism (duly discussed by

Athenocentricity into a radical rereading of Greek dialect relations, whereby the

If ‘barbarian’ is taken to mean the opposite of ‘Greek-speaking’, we do not actually

(presumably in tragedy and New Comedy)86 – and now, as if suddenly mindful of

Aristotle’s Greek, however, is more abrasively abstract, especially at the

The sensuous brevity of expression is characteristic of the corpus, and what is

and which is attested elsewhere in the corpus, as here of a fractured bone:

In this second passage, the crypto-animating verb φεύγει (‘slips’ [‘flees’]) is in

Excursus: Ancient Evidence for the View that ‘Ionic is Poetic’

παρεσκεύασε τῇ κρατίστῃ ποιήσει τὴν πεζὴν φράσιν ὁμοίαν γενέσθαι πειθοῦς

The Greek Koine

Wittgenstein here questions an approach (the second model) which is captured by

The notion of a literary koine is useful, therefore, because it implies a package