Peter Bing, The Scroll and The Marble

Contents
List of Figures ix
List of Abbreviations xi
Introduction 1
pa rt 1 . Reader and Voice in Callimachus and Hellenistic Poetry

chapter 1. The Unruly Tongue: Philitas of Cos as Scholar
and Poet 11
chapter 2. Impersonation of Voice in Callimachus Hymn
to Apollo 33
chapter 3. Callimachus and the Hymn to Demeter 49
chapter 4. Reconstructing Berenikes Lock 65
pa rt 2 . Epigram and Its Audiences

chapter 5. Ergnzungsspiel in the Epigrams of Callimachus 85
chapter 6. Text or Performance / Text and Performance:
Alan Camerons Callimachus and His Critics 106
chapter 7. The Un-Read Muse? Inscribed Epigram and Its Readers
in Antiquity 116
chapter 8. Allusion from the Broad, Well-Trodden Street:
The Odyssey in Inscribed and Literary Epigram 147
viii Contents
pa rt 3 . Inscription and Bookroll in Posidippus

chapter 9. Reimagining Posidippus 177
chapter 10. Between Literature and the Monuments 194
chapter 11. Posidippus Iamatika 217
chapter 12. Posidippus and the Admiral: Kallikrates of Samos
in the Epigrams of the Milan Posidippus Papyrus
(P. Mil. Vogl. VIII 309) 234
chapter 13. The Politics and Poetics of Geography in the
Milan Posidippus Section One, on Stones 120 AB 253
Bibliography 273
Index of Ancient Passages Cited 293
Subject Index 302
List of Figures

Fig. 1. R. Pfeiffer, Callimachus, vol. 1. Oxford: 1949, 11415 7677

Fig. 2. Jimmy Stewart as Jefferson Smith, at the Lincoln
Memorial, in Frank Capras Mr. Smith Goes to Washington 117
Fig. 3. Youth reading stele. Tondo of Attic red-gure kylix, Ancona
Painter, 500480 b.c. Corpus Vasorum Antiquorum 16.2
no. 62 123
Fig. 4. Pensive Athena, votive relief from the Acropolis, Athens,
ca. 470450 b.c. 124
Fig. 5. J. H. W. Tischbein, Goethe in the Campagna, 178687,
Frankfurt a. M., Stdelsches Kunstinstitut 148
Abbreviations

AB C. Austin and G. Bastianini. Posidippi Pellaei quae supersunt omnia.

Milan, 2002.
AP Anthologia Palatina.
BG G. Bastianini and C. Gallazzi. Papiri dell Universit degli Studi di
Milano. Vol. VIII, Posidippo di Pella: Epigrammi (P. Mil. Vogl. VIII 309).
Milan, 2001.
Bernand E. Bernand. Inscriptions mtriques de lgypte grco-romaine.
Paris, 1969.
CEG P. A. Hansen. Carmina Epigraphica Graeca. 2 vols. Berlin,
198389.
CGFP C. Austin. Comicorum Graecorum Fragmenta in Papyris Reperta.
Berlin, 1973.
Clairmont C. W. Clairmont. Gravestone and Epigram. Mainz, 1970.
Durrbach, Choix F. Durrbach. Choix dinscriptions de Dlos. Paris,
192122.
FGE D. L. Page, ed. Further Greek Epigrams. Cambridge, 1981.
FH P. Friedlnder and H. B. Hofeit. Epigrammata: Greek Inscriptions
in Verse from the Beginnings to the Persian Wars. Berkeley, 1948.
GG W. Peek. Griechische Grabgedichte. Darmstadt, 1960.
GLP D. L. Page. Greek Literary Papyri. Cambridge, MA, 1962.
xii Abbreviations
GP A. S. F. Gow and D. L. Page. The Greek Anthology: Hellenistic

Epigrams. Cambridge, 1965.
GV W. Peek. Griechische Vers-Inschriften. Vol. I, Grab-Epigramme.
Berlin, 1955.
ID F. Durrbach. Inscriptions de Dlos. Paris, 192629.
IG Inscriptiones Graecae. Vol. 1ff. Berlin, 1873.
Kaibel G. Kaibel. Epigrammata Graeca ex Lapidibus Conlecta. Berlin,
1878.
LIMC Lexicon Iconographicum Mythologiae Classicae. Zurich, 1981.
L-P E. Lobel and D. Page. Poetarum Lesbiorum Fragmenta. Oxford,
1955.
MS R. Merkelbach and J. Stauber. Steinepigramme aus dem griechischen
Osten. 5 vols. Stuttgart, 19982004.
Nilsson, GGR M. P. Nilsson. Geschichte der griechischen Religion. 2 vols.
Munich, 1967.
OGIS W. Dittenberger. Orientis Graeci Inscriptiones Selectae. 2 vols.
Leipzig, 19035.
Pack R. Pack. The Greek and Latin Literary Texts from Graeco-Roman
Egypt. 2nd ed. Ann Arbor, 1965.
PCG R. Kassel and C. Austin. Poetae Comici Graeci. Berlin, 1983.
Pf. R. Pfeiffer. Callimachus. 2 vols. Oxford, 194953.
PGR A. Giannini. Paradoxographorum Graecorum reliquiae. Milan, 1966.
PMG D. L. Page. Poetae Melici Graeci. Oxford, 1962.
Powell J. U. Powell. Collectanea Alexandrina. Oxford, 1925.
SB F. Preisigke, et al. (ed.). Sammelbuch griechisches Urkunden aus
gypten. 1915.
SEG Supplementum Epigraphicum Graecum. Vol. 1ff., Leiden/
Amsterdam, etc. 1923.
SH H. Lloyd-Jones and P. Parsons. Supplementum Hellenisticum. Berlin,
1983.
Snell-Maehler B. Snell and H. Maehler. Pindari carmina cum
fragmentis. 2 vol. Leipzig, 198789.
Introduction

D ionysius of Halicarnassus recounts an anecdote that he describes as

well known to all those who are fond of learning. It tells of how, fol-
lowing the death of Plato, the philosophers writing tablets were found. On
them, it was discovered, he had jotted down the rst sentence of his Repub-
lic in numerous versions, each time adjusting the word order in a different
way (
, ! -
" #
( + ,
/, Dionys. Hal. De comp. 25, p. 208 Reiske: cf. Diog. Laert.
3.37; Quint. 8.6.64). Dionysius goes on to describe Platos , that
is, his painstaking, even fussy manner with his texts composition; and he
does so wittily, in terms taken from hairstyling: The philosopher is pic-
tured as the writerly equivalent of some elderly dandy standing before a
mirror, primping and making sure that each last strand of hair is perfectly
in place, for even when he turned eighty, Plato never stopped combing
his dialogues, dressing them out and braiding them in every conceivable
way ( 1 2
" 4 5).
Of course, this passage dwells on the moment of composition, the pro-
cess whereby the writer creates his text. But at the same time, it implies a
manner of readinginitially that of the author himself as the rst link in
his own reception. Here the reader carefully scans what stands before him
on the page (or, as in this case, lies etched in the wax of a writing tablet),
2 The Scroll and the Marble
mulls over every clause, weighs each word on the tongue to measure its
impact in relation to every other. This anecdote presumes a particularly
intense engagement with a text, one that transpires not in the context of
oral performance, as one might have expected given Socrates emphasis on
the primacy of the spoken word in Platos Phaedrus, but, rather, through a
written medium. That medium facilitates not only the play of permutation
in the act of composition but also a multifaceted, reective, and unhurried
encounter with the text in the act of reading. Here the reader may pause
over a given sentence, look back to another, juxtapose it with one else-
where in the text, and reread them all if so inclined. Precisely such a scene
of readerly reception is dramatized in Platos Phaedrus when Socrates re-
quests multiple re-readings of the start of Lysias speech (262de, 263e
264a), each followed by a nuanced interpretation. Contrast a works recep-
tion in performance (as with a play at a theater, an epic or choral lyric at a
festival), which unfolds serially in time and so offers far less scope for ex-
acting, word-by-word contemplation.
In the Hellenistic age, poets writing for the social elite became con-
cerned as never before with the act of reading itself and with the impact of
the written word, as artifact and medium, on the reception of their work
(see, e.g., Meyer 2005). Their textsinasmuch as they translate onto the
written page traditional poetic genres earlier experienced mainly as part of
large civic occasions, such as religious festivals (thus epic, choral lyric,
tragedy, and comedy), or at smaller private occasions, such as symposia
(the home of such genres as lyric monody and elegy)invite readers to
ponder their experience of a given poem and to ask how it differs from
performance-oriented reception. The essays gathered together in this vol-
ume take up that invitation and focus largely on the question, What is
entailed in the act of reading? More specically, in what ways do the texts
themselves reect a new awareness of their written form and, indeed, of
the diverse types of written media available to authors? How do the poets
construct their readership in relation to each of these media? Do the texts
envision various sorts of readers, accommodate different modes of readerly
experience?
Though numerous kinds of material object could serve as the vehicle of
poetic communication (e.g., writing tablets, or deltoi, as in the earlier ex-
ample of Plato; papyrus letters containing just one or two poems; wooden
tablets, or pinakes), this books title draws attention to two in particular, the
chief media in which the writtenness of poetry was experienced at this time:
the papyrus scroll and inscriptions in stone. Two poets of the third century
b.c. may be seen as representing, respectively, these two media and are at
Introduction 3
the same time the clearest embodiments of the new aesthetic of reception
by means of reading. I am referring to Callimachus of Cyrene and Posidip-
pus of Pella. These authors, while not necessarily of equal quality, form the
twin poles that frame this collection of essays: Callimachus, famously work-
ing in and drawing inspiration from the Library of Alexandria, is a poet of
exquisite polish and erudition; Posidippus, besides being the author of
purely literary epigrams, is the rst poet expressly referred toand publicly
honoredas a writer of inscribed verse, or 6 (though as
we now see from the Milan papyrus, he also conceived of his epigrams as
destined for collection in the scroll). Both were favorites of the Ptolemaic
court. The title The Scroll and the Marble plays on and pays homage to
J. Svenbros La parole et le marbre (1976), a work dealing with the develop-
ment from oral to written poetic production in archaic and classical Greece.
While that book has to do mainly with composition and performance, mine
throws the spotlight on reading and reception. In its focus on Hellenistic
poetry (rather than that of the Archaic period, which is Svenbros concern),
my book explores an era for which the text as material objectthe scroll
and the marble, as well as other potential mediaassumed a new level of
importance.
The thirteen essays in this volume, written over the course of some f-
teen years, largely recur to and persistently circle such questions as those
enumerated earlier in this introduction. The book is divided into three
parts. The rst, Reader and Voice in Callimachus and Hellenistic Po-
etry, begins with a study of Philitas of Cos, a seminal gure at the dawn of
the Hellenistic age and an important inuence on Callimachus and other
poets of the next generation. Philitas is the rst author described as being
simultaneously a poet and a critic, and this essay examines his famous lexi-
cographical work, Ataktoi Glossai, or Disorderly Words. That works learned
interest in and playful approach to rare words are consistent with what we
can see in the (meager) fragments of Philitas poetry and also point ahead
to the bookish delight that the scholar-poets who succeeded him displayed
for lexicographical rarities.
The following three essays deal with poems by Callimachus. The rst,
Impersonation of Voice in Callimachus Hymn to Apollo, examines the
theme of how the poet constructs multiple audiences. Indeed, we nd here
a hierarchy of audiences in which the community of readers is gured as a
cultic community present at Apollos rite. This rite, including a choral per-
formance of Apollos paean, is represented mimetically within the hymn, in
such a way that the text itself comes to embody the gods epiphany. Readers
must then decide whether their experience of the poem is so involving and
vivid that they ultimately lose their sense of detachment from the text and
actually see the god (thus counting among the elect, or 6, in the
poems hierarchical terminology) or whether their engagement is of a less
intense variety (thus placing them among the poorer celebrants, or ).
The next essay, on Callimachus Hymn to Demeter, again suggests different
possible readerships, including an emphatically bookish one, eager to play
literary detective and to uncover recondite allusions to the hymns archaic
model. Finally, part 1 ends with a study of a modern counterpart to such lit-
erary detective work, namely attempts to reconstruct Callimachus Lock of
Berenike from its meager remains in the papyri. In this undertaking even
prudent scholars have thrown aside their customary reserve in the face of a
fragmentary text and, spurred by the existence of Catullus seemingly faith-
ful translation, have succumbed to the lure of the empty space, that is, the
urge to ll in the gaps. Here one sees with particular clarity how such liter-
ary activity arises from a readers encounter with the text as object, in this
case a patchwork of multiple, lacunose papyri, whose transcribed remains
can be physically juxtaposed to the Latin translation. To be sure, that read-
erly urge to ll in the gap here results from an accident of preservation, but
it also points toward a possibility discussed in the next section of my book,
to wit, that this same impulse can be deliberately triggered by a poet for
aesthetic ends.
Part 2, Epigram and Its Audiences, focuses on the only ancient
Greek poetic genre that, from its inception, was experienced in writing
(rather than aurally) and that expressly calls attention to the act of reading
and to the relationship between reader and text. Epigrams were originally
short poems inscribed on objects of various kinds: they could appear as
epitaphs on tombstones, as dedications on votive objects, and as honoric
inscriptions on political or personal monuments. While such poems con-
tinued to be inscribed throughout antiquity, they became popular in the
early Hellenistic era as pure literature, written for the scroll, and were
favored by the leading poets of the day. In addition, poems originally com-
posed for inscription were also now collected for inclusion in epigram
books. The impact of that migration from monument to scroll on the
readers experience of the poem is the subject of the rst essay of part 2,
Ergnzungsspiel in the Epigrams of Callimachus, where I trace how, even
in the scroll, the genre retained its traditional deixis, that is, its reference to
a monument or geographical location and (in the case of sepulchral epi-
grams) to the corpse interred beneath it. In its new setting, however, such
deixis no longer pointed to anything real, whether actual artifact or land-
scape. The resulting shortfall or gap was, however, seized on by the poets
Introduction 5
of the age and put to productive use, for with it they could stimulate the
reader to a particularly lively interaction with the text, an Ergnzungsspiel,
or imaginative game of supplementation. The urge to ll in the gap is here
the productive counterpart to that potentially dangerous inclination we
saw earlier in the example of modern scholars attempts to ll in the lacu-
nae of the fragmentary text of Callimachus Lock of Berenike.
The following essay, Text or Performance/Text and Performance, is a
polemical reaction to Alan Camerons Callimachus and His Critics, in which
he argues for the persistence and ongoing primacy of traditional poetic per-
formance in the Hellenistic era and against any signicant development to-
ward a more self-consciously literate type of verse and toward its reception
within a culture of reading. In my essay, I suggest that while many of the
texts by elite Hellenistic poets can accommodate performance and were pos-
sibly performed, what evidence we have points, rather, to an audience that,
although possibly just a small part of the larger poetry-consuming public,
demonstrably engaged with the text through its visual manifestation on
the page. After this comes an essay, The Un-Read Muse? Inscribed Epi-
gram and Its Readers in Antiquity, that explores the odd disconnect
between how inscribed epigrams pervasively envision scenes of encounter
between passersby and monuments and how sources outside of epigram
nevertheless suggest that it was actually rare for people to stop and read
them. Those who did so were exceptional personsamong them the Hel-
lenistic poets who absorbed the conventions of the inscribed genre, re-
moved it from its functional setting, and transformed it into literature.
Their literary epigrams may in turn have spurred new interest in the in-
scribed variety.
The nal essay of part 2, Allusion from the Broad, Well-Trodden
Street: The Odyssey in Inscribed and Literary Epigram, examines how the
authors of inscribed epigrams tried, by keeping the poetry simple, both to
address the uncertainty of nding readers willing to pause and contem-
plate a monument and to accommodate those who might just cast a glance
at their chiseled texts in passing. If and when they incorporated literary
allusions, these tended to play on the best-known parts from the most
famous works of literature. In particular, my essay examines allusions to
the Odyssey in inscribed epigram, comparing them with those found in
their literary counterparts of the Hellenistic era. The poems reveal a wide
range of readerly expectation, from the most basic and clear-cut to the
most sophisticated and challenging. Not surprisingly, because its medium
allows for a more leisurely and sustained encounter with the text, the allu-
sive artistry of literary epigram puts far greater demands on its readers
than that of inscribed epigram. That becomes especially apparent in the

epigram of Callimachus with which the essay ends, which gures allusion
as a kind of larceny and makes discovering the allusion a sort of cat-and-
mouse game of detection.
Part 3, Inscription and Bookroll in Posidippus, steps back from the
literary sophistication of Callimachus to explore the work of Posidippus,
the one poet of the third century b.c. known certainly to have written for
inscription as well as for the scroll. The rst essay of this section, Reimag-
ining Posidippus, suggests that the image Posidippus projects of himself in
his so-called seal poem (SH 705 AB 118) pointedly embraces and melds
together both media in which he worked, showing that he conceived of
himself as operating on two tracks within the single genre of epigram. That
self-image is borne out both by the evidence for his work on monuments
and by the new Milan Posidippus papyrus (P. Mil. Vogl. VIII 309). The sec-
ond essay, Between Literature and the Monuments, begins from an ex-
amination of Posidippus poem on the dedication of the statue of Zeus
Soter atop the Pharos Lighthouse by Sostratus of Cnidus and goes on to
explore how the aesthetic impact of epigram on a monument differs from
that of epigram on a scroll. The themes of this essay hark back to the essays
in part 2, where Ergnzungsspiel in the Epigrams of Callimachus likewise
looks at what happens when epigram migrates between media.
The next three essays focus on different sections of the Milan papyrus.
The rst concerns the Iamatika, the section on miraculous cures, possibly
written at the behest of a prominent Ptolemaic courtier, the doctor
Medeios of Olynthos. It attempts to show how Posidippus drew for these
poems on a previously unmined epigraphic genre, the miracle tales as-
sembled for collective inscription by temple authorities at such shrines as
the sanctuary of Asclepius at Epidaurus. Though he makes knowing use
of the epigraphic genre here, Posidippus deliberately sets it at a remove
from its functional context, thereby opening it to a wider range of mean-
ings, including ironic ones. The topic of the next essay is Posidippus
poems for another eminent Ptolemaic courtier, the admiral Kallikrates of
Samos. Posidippus wrote an epigram to accompany a statue commemo-
rating the admirals victory in the chariot race at Pytho and also wrote a
whole series of epigrams celebrating Kallikrates foundation of the shrine
of his divinized sovereign, Arsinoe/Aphrodite at Cape Zephyrium outside
Alexandria. Through these epigrams, Posidippus lent the aid of his art to
Kallikrates policy of bridging the cultures of old Greece and his adoptive
country of Egypt. A nal essay looks at the rst section of the Milan Po-
sidippus papyrus, the , an artfully organized series of twenty poems
Introduction 7
about an array of remarkable stones (notably precious gems, but also a mag-
net, a huge boulder, etc.). Frequently dwelling on their exotic origins and
how they came, through the skill of expert craftsmen, into the hands of
(Ptolemaic) Greeks as jewelry, Posidippus describes the stones in this sec-
tion in such a way as to collectively embody, in their geographic scope, a
map of the world that Alexander the Great had conquered. It was the world
that the Ptolemies aspired to control as the true heirs to his power, and the
poems of this section jointly express that ambition. Thus this last essay re-
turns us to that other aspect of Posidippus endeavor as a poet of literary
epigram, one keenly aware of the aesthetic potential not just of the marble
but of the scroll.
In the essays here reissued, I have taken the opportunity to update bib-
liography, revise certain sections where my thinking has changed, delete
some material that seemed repetitive in the framework of a collection, and
make additions when these seemed appropriate. By way of supplement to
the previously published work, the volume includes two new contributions
(Allusion from the Broad, Well-Trodden Street and Reimagining Po-
sidippus), which extend its themes and reect my most current ideas.
Together, I hope that these essays will continue to stimulate the thinking
of students and scholars of ancient literature and of Hellenistic poetry in
particular.
chapter 1
The Unruly Tongue

PHILITAS OF COS AS SCHOLAR AND POET
P hilitas of Cos stands as a gray eminence at the start of Alexandrian

scholarship and literature.1 Described as simultaneously a poet and a
critic (Strabo 14.2.19.657c), he was picked by Ptolemy I Soter to be tutor
to his son Philadelphus and is said to have taught Zenodotus of Ephesus,
rst librarian of the Library of Alexandria (Suda s.v. Philitas). His com-
ments on epic vocabulary in his pioneering lexical study, Ataktoi Glossai,
caused Aristarchus, the great Homeric scholar and Alexandrian librarian,
who lived more than a century later, to write a work entitled Against Phili-
tas (schol. A ad Il. 1.524). As to his verse, its artistry was celebrated in pro-
grammatic poems by Callimachus (Aetia 1.910, with the Florentine Scho-
lia) and Theocritus (7.3941), eminent poets of the next generation
working in Alexandria; and Roman poets cite him as an authoritative
This is an updated version of an article that appeared in CP 98 (2003) 33048. 2003 by The
University of Chicago. I wish to thank Professors Richard Hunter and J. Lee for their in-
sightful and stimulating comments while I was writing this essay.
1. I cite Philitas poetry according to the editions of Powell (1925), Sbardella (2000), and
Spanoudakis (2002); his scholarly work, the Ataktoi Glossai, according to the editions of
Kuchenmller (1928), Dettori (2000a) and Spanoudakis (2002). I am persuaded by C. W.
Mller (1990) that our earliest sources (including Propertius) speak mainly for the name hav-
ing been spelled Philitas and that Philetas is the result of later etacism. Cf. also Sbardella
2000: 37; Spanoudakis 2002: 1923. To Mllers evidence, add now Posidippus epigram on
the statue of Philitas in the new Milan papyrus (col. 10.1625 BG 63 AB), cited and dis-
cussed shortly in text.
11
model for elegy (Propertius 3.1.16, 3.5152, 9.4344; Ovid Ars am.
3.32948; Ovid Rem. am. 75960; Statius Silv. 1.2.252).
As all this suggests, Philitas activity and impact loom largein the
work of others. Regrettably, he says very little to us in his own voice, since
his oeuvre has mostly vanished, surviving only in brief citations and frag-
ments. I want to turn to these, however, so as to explore in what sense
Philitas may in his age have served as a model of a poeta doctus.
In one of his surviving poems (10, p. 92 Powell 12 Sbardella 25
Spanoudakis), a female speaker discriminates between ignorant rustics
and those versed in song; the latter stand out by virtue of their laborious,
hard-won knowledge: No benighted rustic from the mountains / will
take me . . . toting his mattock, / but only an expert in songs ordered
verses, who through much toil / knows the way of every kind of tale.
Here poetry is considered the product of toil, of diligent skill and learning
as much as of inspiration.2 This attitude toward poetry was embraced by
subsequent poets in Ptolemaic Alexandria.
Philitas erudition gave rise to an entirely new comic type: the spindle-
thin professor so engrossed in study that he quite forgets to eat and drink,
being himself rather consumed by his researches until he becomes a feeble
shadow of a man.3 This new comic type seems to have appeared speci-
cally in response to the personality and interests of Philitas, whose partic-
ular obsession was with words. Already his younger contemporary, the
elegist Hermesianax, describes how there was a statue of him set up by the
people of Cos, in which he was portrayed as frail with all the glosses, all
the forms of speech ( / 8
, fr. 7.7778, p. 100 Powell). A similar picture emerges in one of
4
the newly published epigrams of Posidippus, a near contemporary (col.

X.1625 BG 63 AB). The poem depicts another statue of Philitas, com-
missioned from the sculptor Hekataios by none other than Ptolemy
Philadelphus and perhaps intended for display in Alexandriaa devoted
pupils tribute to his distinguished tutor.5
2. For the varying interpretations of this fragment and its female voice, cf. Bing 1986;
Dettori 1999; the commentaries of Sbardella (2000) and Spanoudakis (2002).
3. As Yeats might have said, a tattered coat upon a stick (Sailing to Byzantium, v. 10).
4. The latter term, , later came to mean dialect; cf. Cairns 1979: 220.
5. The fact that the monarch is called both god and king ( / 9 ,
v. 9) demonstrates that Ptolemy Philadelphus is meant. For when the cult of the Theoi
Adelphoi was established in 272/1, he became the rst of his line to receive divine honors
while still alive. The founding of that cult thus constitutes the terminus post quem of our poem,
and the dedication of the statue must be seen as Philadelphus tribute to his distinguished
tutor Philitas. It is uncertain whether Philitas was still alive when this epigram was written
The Unruly Tongue 13
[] [:] {} ; []

] [] [(] , =
] [ ]
], ? " + ({} ,,
@ [ ]
] , 4 [(]
" ] / (, @ A,
(] , 65 C D
6 ] / E
/ 9
] {} G .
[Hekataios made this bronze like Philitas in every way,
accurate down to the tips of his toes
in size and frame alike describing this investigator [?]
on a human scale. He included nothing from the physique of heroes.
No, with the straightedge of truth, and all his skill he cast
the old man full of cares.
He seems like hes discoursinghow fully his features are
elaborated!
alive, though of bronze, this old man.
I stand here dedicated by Ptolemy, god and king at once,
for the sake of the Muses, the Coan man.]
As in Hermesianax image of the scholar frail with all the glosses,

Posidippus statue portrays this investigator (]/, v. 4) on a human
scale, as an emphatically old man (] , v. 6; , v. 8). He is,
moreover, extremely anxious, full of cares (, v. 5). The
adjective is otherwise unattested but aptly describes the absentminded in-
tellectual, engrossed in thought. According to the Hellenistic epigram-
matist Dionysius (1.3 GP AP 7.78.3), another well-known personage,
the scholar-poet Eratosthenes, died . In our poem,
words in - describing the sculptors painstaking realism] []
[(] , = (v. 2)suggest an afnity with the
. . . ] (vv. 56). Likewise the artists deployment of
all his skill (@ [ ] , v. 5) in the accomplishment of his
he is generally thought to have been born ca. 340. In any case, he is represented as an old man
(], v. 5; C , v. 8), so if he was dead, it may be that the image of him in his last
years was still vivid. For recent discussion of this epigram, cf. Angi 2002; Bernsdorff 2002:
esp. 1926; Hardie 2003; Scodel 2003. I adopt Scodels supplement at the start of v. 10. Sens
(2005: 20916), Prioux (2007b: 1974), Mnnlein-Robert (2007: 6066).
task seems appropriate to the creation of an image of the meticulous

scholar-poet.6 Finally, Philitas is not idealized in any way. On the con-
trary, he cuts a wholly unheroic gurea point Posidippus drives home
with the striking formulation in verse 4 that the sculptor blended in
nothing from the form of heroes (/ ? / " + (/{} ,), a
metaphor doubtless evoking the metallurgical act of making a com-
pound.7 There is in Philitas no alloy of heroism.8
This picture of Philitas is embellished in stories preserved for us in var-
ious sources. Aelian, for instance, relates the following (Varia Historia
9.14):
They say that Philitas grew extremely thin. Thus as the slightest thing
could easily send him sprawling, he put lead weights in his soles, so as
not to be blown over if there happened to be a stiff wind.
And in Athenaeus we learn of the tragicomic denouement when one of the

characters humorously admonishes the host of the great banquet at which
the work is set (9.401de):
Ulpian, you always refuse to take your share of food until youve
learned whether the word for that dish is ancient. Like Philitas of Cos,
therefore, . . . you risk withering away some day. For he became utterly
emaciated through these studies and died, as the epigram in front of his
memorial makes clear:
Stranger, I am Philitas. The deceiving word caused my death,
and the evenings thoughts extended deep into the night.
6. Signs in the poem point to Philitas being seen here as both scholar and poet. For in-
stance, we might compare " ] / ( in v. 7 with Asclepiades " / (
(43.3 GP Anth. Plan. (A) 120.3), where, as in most instances, the verb is used of spoken
utterance, not song, suggesting that Philitas statue represents him less in his function as poet
than as scholar. In Posidippus phrase {} G (v. 10), on the other hand, the Muses
may stand both for poetic and scholarly accomplishment, esp. liberal arts (LSJ s.v. ),
while the nal words of the poem, describing Philitas as , implicitly recall
Simonides description of Homer as (fr. 8 dub. West, IE2) and suggest a compari-
son of poets.
7. The epigram is suffused with both metaphorical and literal terms that have strong
artisanal resonance: [(] (v. 2), (v. 3), ] (v. 5), 4 [(]
(v. 6), (v. 7).
8. Perhaps, as Caldern Dorda (1990) rst suggested in reference to the biographical
tradition, the physical characterization hints at the poetic qualities that made Philitas verse
appeal to Callimachus and Theocritus, i.e., a Callimachean slenderness. Building on this,
cf. Sens 2002: 5; Sens 2005; Bernsdorff 2002: 1926; Prioux 2007b: 5156.
As Alan Cameron has recently noted, thin-jokes were a staple of ancient

comedy, the counterpart to our modern-day fat-jokes.9 Yet the emaci-
ated professor is something new. It is as yet unknown to Theophrastus
Characters. Nor is it the brainchild of Attic comedy.10 Rather, it seems to
emerge in elegy and epigrama genre for whose learned practitioners
Philitas was an important model.11
Philitas was evidently famous already within his lifetime for his research
on words. But what was the nature of these lexical researches, and in what
way did they gure in his poetry and in that of the scholar-poets who suc-
ceeded him? Much work on the erudite poets of Alexandria has focused on
their reception and creative reuse of the epic tradition, in particular of
Homer. That is not surprising given the wealth of Hellenistic hexameter
and elegiac poetry, where we can trace how these poets mined early epic for
the rare and atypical.12 It is tempting to suppose that Philitas did the same,
devoting his lexical interests mainly to Homer. After all, in a comic frag-
ment of Strato (Kassel-Austin PCG VII fr. 1), a contemporary of Philitas, a
master of the house sputters in exasperation as he describes being driven to
distraction because the cook he hired for a party possessed the peculiar and
hilarious tick of speaking almost exclusively in Homerisms: one would
have had to use the books of Philitas and look up every word to check its
meaning (( / / I
8.). The assumption here is that the books of Philitas
dealt with Homeric vocabulary (but cf. my appendix to the present essay).
9. Cameron 1995: 490, 493 with n. 25.

10. Contra Cameron 1995: 48891 (app. B, Thin Gentlemen). Characteristically,
Cameron seizes on precisely the genre in which Philitas is not depicted as emaciated and
asserts that it must be the source of that image. We have it in the elegy by Philitas younger
contemporary Hermesianax; it is implied in the new Posidippus epigram on the statue of
Philitas; thereafter it is much elaborated in the stories of Aelian and Athenaeus (with its epi-
gram, which might wellas Cameron suggestsbe contemporary with Philitas). The one
place it is absent is in Stratos comic fragment, where Philitas is merely the scholarthere is
nothing about his being thin. See the appendix to the present essay. For a critique of Cam-
erons general approach, cf. Bing 2000 chap. 6 below.
11. The gure of the scholar here becomes the mundane counterpart to that more cel-
ebrated image of the singer as the Muses devotee, whose preoccupation is not scholarship
but song; he, too, is so absorbed that he shrivels away and dies, ultimately becoming the
cicada, who needs no food and spends the livelong day in songan image rst elaborated in
Platos Phaedrus (259bd), then powerfully revived in Callimachus prologue to the Aetia (fr.
1.3235).
12. The bibliography is enormous. I would cite the pioneering works of von Jan (1893)
and Kuiper (189698) and, thereafter, Herter 1929 and the collected essays of Giangrande
(1980). More recently, see especially Hunter 1996, the studies of Rengakos (1992, 1993,
1994a, 1994b), and Bonanno 1995.
We recall as well that Aristarchus, the great Homeric scholar, wrote the
work Against Philitas.
Yet the surviving fragments of Philitas researches, and some of his
verse, point in another direction, toward a different area of intense learned
and poetic interest: namely, exotic diction and local customs. This is where
I believe he had his greatest impact on subsequent poetry. From what we
can tell from citations of his famous lexical work, the Ataktoi Glossai, Phili-
tas often took Homer as his starting point. But let us have a look at some
examples to see where he went from there.
Our rst is the word , which probably refers to Homers -
at Iliad 18.553. That term is a Homeric hapax legomenon,13 one of
ve treated by Philitas out of a total of twenty-ve surviving glosses (i.e.,
1/5). I am of course assuming that he chose to treat these words because
they were Homeric hapaxes, though none of the citations makes explicit ref-
erence to Homer. Nevertheless it is a reasonable assumption since Homeric
language overall was privileged, and such words in particular provoked
perennial discussion and so, for all their rarity, were culturally marked.14
The word occurs in the description of that part of
Achilles shield on which there is a royal precinct with reapers harvesting
the grain (Il. 18.55257).
/ / = 6 (
/ 6 6 .
/ /
6D " =
, 6 ,
+ D 1 / 6 J
( 2 6/ = .
[Some sheaves fell to the earth in a row, one after another,
some the amallodeteres bound with cords.
Three amallodeteres attended to the job, and behind them
children gathered the sheaves, and carried them in their arms,
and quickly brought them over. And silent beside them the king
along the swath stood holding his scepter, rejoicing at heart.]
Broken into its constituent parts, - means those who bind
the amalla. This substantivized binding is immediately repeated, and so
13. Strictly a dis legomenon: it comes up twice in as many lines.

14. For Homeric hapax legomena, cf. Kumpf 1984. For their reuse in Hellenistic poetry,
see, e.g., Rengakos 1992, 1993, 1994a, 1994b; Kyriakou 1995.
explained, in the verb and its object : those who bind the
amalla tie sheaves together. The resulting bundle, then, is the amalla.
Here we have a case of instant exegesisa nice example of the interpretive
maxim, often linked with Aristarchus, that Homer is his own best inter-
preter, or, in Greek, K 6 ; (literally, explaining
Homer through Homer).15 In the world of the shielda synoptic, pars
pro toto world, where homely scenes are crammed with universalizing
meaningno further explanation is required. We need know nothing
more about the amalla.
Yet Philitas commented on the word,16 as we hear in the following
citation in Hesychius (s.v. 46 Kuchenmller 18 Dettori 46
Spanoudakis):
D , . . .M ,
N, O P, + 6 N.
[Amalla. Sheaves, bundles of grain . . . a bunch; one hundred sheaves,

according to Istros, but Philitas says it consists of two hundred.]
Now where would Philitas have come by such a fact? Not in Homer, thats
for sure. Kuchenmllers hunch was that hed asked the farmers them-
selves, presumably on Cos (haec ab ipsis rusticis quaesivit Philetas). One
may balk, of course, at picturing the scholarespecially the Philitas of
biographical loreexiting his study, blundering down the road to a local
farm (attened several times, no doubt, by nasty gusts; hed left his
weights at home) to ask the farmers (whom no one could have mistaken,
for they looked beyond all like farmers), Just how many sheaves make up
an amalla?17 But stripped of the biographical caricature, the basic suppo-
sition that he learned from a source with rsthand knowledge is not far-
fetched. Perhaps it was a farmer, perhaps a treatise on farming, such as
those mentioned in Platos Minos (316e: farm manuals,
). In any case, what the gloss of Philitas is clearly not is an
15. Pfeiffer (1968: 22527) demonstrates that the attribution to Aristarchus is false (it
comes from Porphyry) but that it is in keeping with his practice.
16. It was later taken up by Callimachus in the Aetia (fr. 186.27: Hyperboreans).
17. Intriguing but ultimately speculative links have been found between Philitas and the
early history of bucolic, on the basis of the character of the old cowherd Philetas in Longus
Daphnis and Chloe (2.3), a singer and player of the syrinx. On this basis, it has been suggested,
moreover, that the gure of Lycidas in Theocritus Idyll 7 is either Philitas or a character
from one of his poems. Cf. Reitzenstein 1893; Cairns 1979: 2527; Hunter 1983: 7683;
Bowie 1985; Cameron 1995: 41820.
attempt at explaining Homer through Homer (K 6 ;

).18 It is not intended, I think, to explain Homer at all. Rather,
here, as elsewhere, Philitas reveals an interest in rustic life for its own
sake. The Homeric word is merely the cue. What appeals to him is the
specialized knowledge, the gure two hundred, which tells us something
about actual farming, not about the remote heroic world of epic.
Our next example likewise takes its cue from Homer but suggests, too,
that Homerfar from being just a springboardis a crucial foil against
which the scholar sets his gloss. The word in question is , another
hapax legomenon, whose one appearance in Homer comes at Iliad 16.642.
There it clearly means milk pail and is part of a rustic simile in which
soldiers swarming over the dead Sarpedon are likened to ies whirring
through a sheepfold, about the pails [] overowing with milk
[]. Here is how the word is glossed by Philitas (Athen. 9.495e
33 Kuchenmller 5 Dettori 33 Spanoudakis):
6 + 1
, , + D / 6 /
.
[Cleitarchus [of Aegina, rst century b.c.] in his Glosses says the Thes-
salians and Aeolians call a milk pail pelltr, but a drinking cup pella. Phili-
tas in his Ataktoi says a wine cup (kylix) is call pella by the Boeotians.]
Now milk and wine are virtual opposites in the Greek imagination. And the
shallow, broadly aring kylix, a wine cup, could scarcely be more different
from a milk pail. Yet can be used to signify either. What strikes us are
the widely (and wildly) divergent meanings of a single wordhomonyms
but in jarring antithesis. Add to that Philitas concern with local usage (in
this case, Boeotias): on the one hand, there is the culturally authoritative
Homeric meaning; on the other, the regional peculiarity. Signicantly,
Philitas appears to present the Boeotian meaning for its own sake, not to
illustrate that of the Homeric hapax. Was he interested in the lack of uni-
formity, in semantic dissonance itself?19
18. As I argued earlier, Homer did that sufciently himself in the very same verse.
19. A comparable interest in homonyms, though of a less jarring kind, occurs in Philitas
gloss of another Homeric hapax: :. This word appears at Od. 18.300 as one of the gifts
the suitors give to Penelope; it is usually taken to mean necklace. Though the text is corrupt,
it appears that Philitas explains the way in which : means both neck and necklace
by analogy with , which can signify both the top of the head and crown:
:D . A C
I should emphasize that that would be quite contrary to the practice of

the anonymous gloss writers, the Glossographoi, in the Homeric scholia.
While they too occasionally x on a dialect usage, they do so in order to
suggest that this is what Homer meant by a certain term in a given passage
under discussion.20 To be sure, the citations of Philitas provide us no con-
text; we cannot tell what he intended by noting that a Homeric hapax was
used in a certain way in a regional dialect. Yet nothing suggests it was
meant to illustrate the sense in Homer (can one even argue that can
be glossed as wine cup in the Iliadic simile?).21 On the contrary, what
leaps out at the reader of the gloss when set against the Homeric use is its
irreducible differencean opposition so neat as to suggest a deliberate
strategy and interest in semantic dissonance.
Another example reinforces that impression and adds a new dimension.
It is a further Homeric hapax, the word ,22 which occurs at Iliad
9.206 in the sense of butcher block or tray for cutting meat. There it is
described as , a great butcher block, big enough to hold a
sheeps chine and used by Patroclus and Achilles in preparing the meal for
the embassy of Greeks. Here is what Philitas says about it (Athen. 14.645d
, T /
/ 6 , 6/ " ,
. 4
+ . O 6 /.
[Krion is a at cake or loaf that the Argives bring from the bride to the
groom. It is baked on charcoal, and the friends are invited to partake of
it, served with honey. So says Philitas in the Ataktoi.]
Again the meaning is strikingly anomalous and local in origin. But this
time Philitas reveals his interest not just in regional words but in customs.

"J
"M
. (Athen. 15.677 41
Kuchenmller 13 Dettori 41 Spanoudakis; emended by Kuchenmller). The word was
also glossed by Simias of Rhodes; cf. Frnkel 1915: 113.
20. For their approach generally and deployment of dialectal evidence, cf. Dyck 1987:
esp. 123. Dyck writes: The basic aw . . . is their habit of tailoring their denitions of Ho-
meric words to . . . a handful of passages: they seldom undertook the laborious task of . . . at-
tempting to do justice to the totality of evidence, and the may have used . . .
an awareness of dialectal variations as license to posit wild semantic shifts [in Homer] (126).
21. See Dettori 2000b: esp. 186.
22. This is the Doric form of Homers . The latter is cited in the Etymologicum
Magnum in the same sense that Philitas assigns it: D C + 6
. Athenaeus reading, , would seem to be thrown into doubt by Hesychius
D D ,
but he also has D M, dened by
LSJ as cakes in shape of animals. For detailed discussion, see Kuchenmller 1928: ad loc.
Among other things, that custom may serve here to highlight the differ-
ence from the Homeric context. For while Homer describes the manly
preparation of meat for the feast of heroic friends in a council of war,23
Philitas recounts how the Argive bride bakes cakes for her groom and his
friends, to be served with sweet honey. The single word thus carries con-
notations of both marriage and butchery, love and war.24
Our nal instance, the word , again drives home the impres-
sion of deliberate dissonancethe more so as, far from being a hapax, this
word is fairly frequent in Homer, appearing ve times in both the Iliad and
Odyssey, invariably with the meaning goblet and often coupled with the
stately epithet goldena cup t for heroes, in other words.25 The fol-
lowing, however, is what we nd in the Ataktoi Glossai (Athen. 11.483a
+

6 .
[Philitas says the Syracusans call the crumbs of barley cake and bread
left on the table kupella.]
The referent has shifted from goblets made of gold to humble bread
crumbs, the debris left over from a meal. The dissonance could not be more
striking, verging almost on the paradoxical. It is as though Philitas had
wanted to see how dissimilar a usage he could nd, how far one could depart
from the culturally authoritative norm; and here, too, his source is local, this
time Syracusan. The effect resembles that of the Homeric simile: for like
the simile that reveals a life apart, beyond the sphere of heroic action, the
gloss throws open a door, exposing the epic term to a strange, incongru-
ous local meaning and so disclosing an unexpected world. But the gloss does
something more radical, too, perhaps: unlike the simile, which subordinates
its world to that of heroic narrative both formally and in the hierarchy of

23. N.b. the emphasis on friendship in Achilles welcome (9.19798): D V
, / G M /
D V 6.
24. Homer acknowledges and exploits the poignancy of this antithesis in critical mo-
ments, such as when Hector awaits Achilles in Il. 22.12628 and says in resignation, "
6 " / M 4, 9 V , /
V / 4 . Spanoudakis (2002: 362) interestingly observes,
Coans considered themselves as deriving their origin from the Argolid and it is an Argive
custom that P. describes here.
25. Thereafter it is a rarity, appearing only in poetry and mostly much later. It does come
up once each in Antimachus fr. 22.2, Ap. Rhod. 2.1271, and Lycoph. Alex. 1104in the for-
mer two with its Homeric epithet golden.
epic values, the gloss recongures that relationship by shifting the balance.
It takes for its text not merely Homer but the totality of the language.26 The
Iliad and Odyssey thus appear within the larger textual/linguistic matrix, a
touchstone still, no doubt, but no longer the alpha and omega. Instead the
gloss is part, now, of a complex, often surprising web of discourse.27
I believe that this preoccupation with anomalous usage helps illuminate
why Philitas called his glosses ataktoi, something no other glossary is
called.28 The title is often translated Miscellaneous Glosses; that is, the words
are interpreted as being , without order, in relation to each other
and as a group. Hence, most inuentially, Rudolf Pfeiffer supposed that
the book was not systematically arranged like the later collections made
by grammarians and that we may compare the name Miscellanea given by
the poet Politian to his various learned writings put together without
proper arrangement.29 But our examination suggests that one could also
take it to refer to words that are disorderly in themselves individually.
Put somewhat differently, that semantic deviation or dissonance of the
single word, which Philitas evidently found appealing, might be character-
ized as a kind of unruliness, hence Ataktoi Glossai might be translated Dis-
orderly Words, orsince glossa can also mean tongueeven Unruly
Tongues.30 It isI should stressinherently plausible and likely that this
student and lover of words, Philitas, would have been alive to such a play-
ful potential connotation in the title of his workeven if on some level it
referred to its organization as a miscellany.
26. In this, we see how Philitas anticipates and inuences Callimachus. Cf. Hollis de-
scription of that poets catholic interest in language (1990: 1114), as well as Selden 1998: esp.
374: Linguistic hybridization is the single most prominent characteristic of Callimachus
style . . . His texts are essentially constructions, vocabular mosaics assembled out of disjoint
verbal tesserae: anomalous and eccentric glosses, incompatible morphologic features, freak
syntactical constructions which have been culled from every dialect and genre, from poetry as
well as prose, from substandard speech and scientic treatises alike, from every stage of the
languages historical development.
27. The repositioning of the Homeric text within the larger discourse may recall the
concept of intertextuality as lucidly framed by Don Fowler (1997: here 16). He sees the in-
herently multiple nature of intertextual reference, where the notion of a hierarchy of refer-
ence itself becomes questionable. I believe that Philitas retains the hierarchy in somewhat
diluted form, even while questioning it.
28. As Kuchenmller (1928: 114) states, Nimirum Glossae extabant permultae,
P praeter Philetam nullae. For an overview of the many attempts to inter-
pret the title, cf. Dettori 2000a: 2122 n. 54; cf. also 27: Il signicato di
rimane, a mio parere, alquanto misterioso. See also Spanoudakis 2002: 38486.
29. Pfeiffer 1968: 90.
30. Selden (1998: 377) draws attention to the somewhat later Stoic interest in linguistic
phenomena that exploit a single signier to designate several different signieds, an effect of
semantic ambiguity which Chrysippus called amphibolia.
A measure of support for this unruly suggestion may come from the
evidence that Simonides wrote a work entitled Ataktoi Logoi (PMG 653)
containing speeches or stories that are ataktoi. We hear of it in a remark by
Alexander of Aphrodisias on Aristotle. In his Metaphysics (N 3.1091a5),
Aristotle says:
, " 2 ",
( 6 " , C . C
O C @ + #+ .
[All this is absurd, and conicts both with itself and with what is prob-
able, and it seems to be an instance of Simonides long story [
], for the long story comes about, like those that slaves tell,
when people have nothing sound to say.]
Pseudo-Alexander comments on this as follows (Comm. in Arist. Graeca

1.818.48 Hayduck):
C 6 , W / 6,
W , 6 6
6 "1 I 6; "1
, "+ + #+ X
, 6 6 M
D
E , .

[Simonides, in the stories that he entitles Ataktoi, imitates and re-

counts the stories that slaves are apt to tell when they have bungled
something and their masters are grilling them as to why: he has them
make long prattling excuses that have nothing sound or convincing
about them; rather, all their conclusions contradict what theyd previ-
ously asserted. Such speech, as it seems, is typical of a barbarian and
one without education.]
David Campbell translates Ataktoi Logoi as Miscellaneous Stories.31 But

the descriptions in Aristotle and Alexander suggest, rather, that the title
refers to the disorderly speech of the slave or barbarian, speech in dis-
array with regard to both its form and its content. Thus the logos is ataktos
not in relation to other logoi, and so miscellaneous, but in and of itself
(i.e., within its own constituent parts), and hence disorderly.
31. Campbell 1991: 507, on Simonides 653.

A still stronger piece of evidence for taking Ataktoi Glossai as unruly or

disorderly tongues comes from the second-century b.c. poet Nicander. In his
Theriaca (75658), Nicander tells us that are the result of the bite
of a venomous spider ().
+ @ (

, + ,
/ , + =.

[But for all their size around the troublesome bite of one [sc., the ven-
omous spider phalangion] blisters always rise, and the mind wanders and
is crazed; the tongue [glossa] shrieks disordered words [atakta] and the
eyes squint.] (trans. Gow)
Inasmuch as the collocation of glossa and atakta appears nowhere else in

Greek literature, I feel certain that Nicander was playfully alluding here to
the title of Philitas Ataktoi Glossai. Mad departures from familiar speech
that might be an apt, if comically exaggerated, way of describing a central
facet of Philitas Disorderly Words or Unruly Tongues. Did the spider bite
Philitas?
In proposing that the works title was Unruly Tongues, we may wish to
revisit Kuchenmllers controversial suggestion about the identity of Bat-
tis (or Bittis), the putative beloved of Philitas. This Battis is mentioned in
the Leontion, an elegy in three books by Philitas younger contemporary
and acquaintance Hermesianax of Colophon.32 Can she corroborate the
name Unruly Tongues? Perhaps. The elegy recounted tales of famous loves,
from the poetic tradition and elsewhere. In the long fragment that sur-
vives from book 3 (fr. 7, pp. 98105 Powell), Hermesianax describes the
loves of his poetic predecessors in chronological order and generic pair-
ings from Orpheus to Philitas.
However one may judge this poem in other respects, it displays an ap-
pealing zaniness in its catalog of poets loves. Elsewhere I have argued that
this was an ironizing response to the burgeoning interest in poets lives
among biographers,33 and I think any interpretation that does not ac-
knowledge the elegys antic humor fails to do it justice. Here, for instance,
we nd Homer madly in love with Penelope, leaving his native land so as
to be close to her in Ithaca (7.2934); we nd Anacreon vying with the far
earlier Alcaeus for the love of Sappho (7.4756)I think Hermesianax
32. See the discussion of this poem in Bing 1993a.

33. Bing 1993a.
was joking about this love affair, Athenaeus comments as he cites the
verses (13.599cd); most preposterously, we nd Hesiod courting a girl
named Ehoie and obsessively invoking her name at the start of each section
in his catalog poetrya droll conceit on the epic phrase X G with which
each heroine in Hesiods Ehoiai, or Catalog of Women, was introduced.
It was against the backdrop of such comic ights of fancy that Kuchen-
mller made his suggestion about Philitas beloved Battis, who appears in
the following verses of the poem (7.7578):
, + , T "
# M
M
,
8 .
[And you know that singer whom the Koan citizens
of Eurypylos raised in bronze beneath the plane tree,
Philitas, singing of nimble Bittis, when he was weak
with all the glosses, all the forms of speech.]
As we can see, the Greek text transmits the name Bittis. Ovid, however,
twice refers to her in company with Philitas as Battis (Tr. 1.6.13; Pont.
3.1.5758), the only further references to her in all of ancient literature.
With this as his cue, and noting the equivalence of and
, Kuchenmller argued that names in the - stem signied
loquaciousness and that Battis consequently meant chatterbox, like Hor-
aces Lalage.34 But more, he cleverly proposed that seen thus, Battis was
nothing other than the humorous personication of Philitas scholarly
passion, the gloss:35 Battissic enim corrigendum ex Ovidionihil aliud
signicat quam , quae quantum cordi fuerit Coo, nemo nescit.36
While Kuchenmllers proposal initially met with acclaim, in recent
years the tide has turned against it.37 Peter Knox, for instance, dismisses it
disdainfully: Kuchenmller supposed that Hermesianax was making a joke
and that there was no poem about Bittis. His highly implausible hypothesis
to explain these lines has won few supporters, but deserves none.38 Of
course, this is nothing but assertion. Knox is wedded to the notion that
34. Kuchenmller 1928: 27 n. 5.

35. Cf. Shakespeares Much Ado about Nothing, act 2, scene 1, where Benedik refers to
Beatrice as my lady Tongue.
36. Kuchenmller 1928: 27.
37. For the state of the question, cf. Sbardella 2000: 5360; Spanoudakis 2002: 2934.
38. P. E. Knox 1993: 66 n. 25.
Ovid, whose elegy is addressed to his wife, cites Philitas love of Battis, as he
does that of Antimachus for Lyde in the line before, because she was his
wife.39 On this view, Ovid is referring to earlier works in that genre, that
is, elegies addressed to wives.40 It is unthinkable for Knox, therefore, that
Battis should be merely a gure for Philitas scholarly obsession.
Another scholar, Joachim Latacz, attacks Kuchenmllers proposal in a
more sustained analysis.41 Without going into the details, his conclusion is
that the Leontion celebrates Loves ineluctable power and that the catalog
of poets loves serves to justify, exalt, and even consecrate (ein Kata-
log . . . der . . . rechtfertigen, erhhen, ja konsekrieren soll) the authors
own existence as love poet. No doubt he had to invent loves for those who
had none by traditionand Latacz acknowledges that he did so lightheart-
edly for Hesiod and Homer. But for Philitas, the authors contemporary
and acquaintance, that would have been neither necessary nor appropri-
ate (weder ntig noch angebracht).42 Make no mistake, Latacz is referring
to what would have been socially, not poetically, appropriate. But propri-
ety itself is surely an inappropriate criterion, since it is just too solemn for
such a poet as Hermesianax, whose work abounds in cockeyed humor. The
critiques of Knox and Latacz are suspect a limine and ultimately do not
carry conviction since they do not do justice to the poems essential wit.
Kuchenmller, I suspect, was right. But perhaps we can improve on
him. It appears he thought that Battis name played only on the concept
, without its qualier ; consequently he looked to its lin-
guistic root solely for the notion of chattiness. But the name Battis is better
explained, like its masculine counterpart Battos, from the stem meaning
stutter (as in and ).43 That sense accords far better
with the disorderly words of Philitas title. A stammer, after all, is the
ataktos glossa, the unruly tongue par excellence. If, as Peter Knox sug-
gests, the adjective , with which Hermesianax describes Battis, means
volatile (i.e., ckle), that, too, would t the linguistic instability that we
have seen appealed to Philitas.44
39. Knox even claims (ibid., 66) that Leontion is Hermesianax wife, though with no
more evidence than there was for Battis.
40. Knox 1993: 6667.
41. Latacz 1985.
42. Quotations from Latacz 1985: 90 and 91, respectively.
43. Cf. Masson 1976.
44. P. E. Knox 1993: 66. LSJ does not go this far. But cf. Knox commonsense remarks
about the meaning of (email message to author, 28 January 2000): Speed in a woman is
not an attribute otherwise admired by poets, speed suggests running in general and running
away in particular, and in love poetry a woman who runs away is typically represented as
Anomolous meanings, words from the margins, and exotic customs stand
at the heart of Philitas glosses.45 I have examined theseone-sidedly per-
hapsagainst the normative language of Homer. But I should stress that
Philitas aim was not simply to blaze new trails for poetry. There was some-
thing more at stake. At a time when there was growing pressure toward lin-
guistic conformity through the spread of Koine Greek as the language of
political administration46a pressure to which people were more suscep-
tible as they were pulled away from previously insular linguistic commu-
nities by the centrifugal forces of the agea sense arose that dialect was a
precious marker of identity that might be lost, should be studied, needed
preservation.47 Against a trend to uniformity, unruly tongues might prove
a potent antidote. One gets an inkling of their power when, in the very dif-
ferent context of his Acontius and Cydippe, Callimachus says: Erudition
is bad in one who cant keep his tonguehis gloss!in line. Truly, hes like
a child with a knife (V , @ / D
E 6 @ (, fr. 75.89 Pf.).48
I want to close with a glance at the poetry of Philitas. Unfortunately,
none of the twenty-ve surviving words from the Ataktoi Glossai appears in
the paltry remains of Philitas poetry, itself not more than fty verses,
some of them partial. Yet I believe we can get a taste of how his scholarly
interests found expression in his poetry from the following example (Phili-
tas fr. 16 Powell, Coll. Alex. 93 18 Sbardella 20 Spanoudakis):
+ 4
4 D
[The deer can sing when it has lost its life
if it avoids the prick of the sharp kaktos.]
This enigmatic couplet is cited and explained by the late third-century b.c.
paradoxographer Antigonus of Carystus with reference to the wondrous
ckle . . . hence volatile. Sbardella (2000: 5659) summarizes attempts to explain this term
and endorses Alfonsis (1943: 16068, esp. 163) proposal that should be taken adver-
bially with , singing quickly (i.e., briey), on the model of its adverbial use at
Od. 8.38. The reference is to the deliberate brevity of Philitas poems.
45. In Kuchenmller (1928), the index lists words from eight dialects (Aeolic, Argive,
Boeotian, Cyrenaean, Lesbic, Megarian, Sicyonian, Syracusan), as well as several rustic words.
46. Cf. the useful discussion by Horrocks (1997).
47. Cf. Hunter 1996: 3145, esp. 32: In the face of the advance of the koine, third century
inscriptions reveal an impressive survival of both weak and strong Doric forms across wide
geographic areas; linguistic difference was a living issue for Theocritus ancient readers.
48. Note that the word for knife, , is glossed by Eustathius as an Aeolic term (cf.
Pfeiffer 1932: ad loc.).
properties of the kaktos, a thistleor artichoke-like plantthat Theo-

phrastus (Hist. Pl. 6.4.10; cf. Athen. 2.83) had said grows only in Sicily
(Antig. Car. Mir. 8 34.40 Giannini 1966).
" ? + , +
Y D , ! @ (
6 J
J, 4
6J "1
:. @ C 6 " :D
[No less marvelous than this, indeed proverbial, is the thorny plant
from Sicily called the kaktos. When a deer steps on it and is pricked,
its bones remain soundless and unusable for utes. For that reason
Philitas spoke of it.]
No doubt the kaktos pricked Philitas interest in recherch words. It was

not just that the plant was exotic and exclusively Sicilian, as Theophrastus
had pointed out; the word itself conrms its regional pedigree in its earli-
est appearances. They occur in the poet Epicharmus, a Sicilian (frs. 159,
160, 161), and thereafter in his countryman Theocritus (10.4). We recall
from the earlier example of kupella that Philitas was curious about Sicilian
vocabulary. Beyond this we note that while the tradition of the kaktos
prick apparently belongs to Sicilian lore, the ute made from deer bone
was (according to Athen. 4.80) a Theban invention. Do we glimpse here
Philitas interest not just in obscure traditions but in the changes wrought
on them as they shift from one locality to anotherthat is, in the different
meanings that accrue to them in different places, just as with the unruly
words examined before?
Philitas scholarly and poetic interests doubtless shaped the tastes of his
pupil Ptolemy II Philadelphus. We now know that he honored his tutors
memory with a bronze statue. But this king left a far greater monument to
his mentor by championing the rapid growth of the great Library of
Alexandria. It was a scholarly tool such as his learned teacher could only
have dreamt of. But for the scholar-poets of the next generation, who ad-
mired Philitas and followed his example, it was the means to build on his
modelCallimachus with his fascination for exotic lore and recherch
words, Theocritus in his foregrounding of rural customs and language,
Apollonius in his epic that teems with learned details from periegetic and
ethnographic sources. It was also the instrument that facilitated the emer-
gence of a privileged circle of learned readersa tiny elite, to be sure,
within a broader audience of diverse educational background, yet one best
able to appreciate the polish, erudition, and, indeed, unruliness of the new
poetry.49
Appendix: What Does Stratos Phoinikides

Tell Us about Philitas?
We are fortunate to have a contemporary reference to Philitas Ataktoi

Glossai in New Comedy. It comes in a scene from the Phoinikides by Strato
(ca. 300 b.c.), known to us from Athenaeus (9.382c383b) as well as from a
late third-century b.c. papyrus from the Fayum (CGFP 219 GLP 57, pp.
26069). In it a master of the house describes his confrontation with a
cook who speaks almost exclusively in Homerisms. Since the nature of the
cooks epic vocabulary directly affects how we interpret the reference to
Philitas, I print the entire passage here.50
/ /, " , , ,
:/D Z "+ I, 1 ,
[ D 8
E
D E , ,
" / 6 D 5
6 ; .
65 6 ; .
1 + ; 8
"/ 5 " @; 11
\ , , , 13
C /, C D / =/ 6D
49. See Morgan 1998: esp. 9495, on the great range of literate attainment: Ones im-
pression on reading Quintilian, and even more on reading other elite writers, is that they rec-
ommend virtually the whole of Greek and Latin literature . . . We know of some litterati
whose learning was as wide as these accounts suggest, but it beggars belief that most pupils
learnt anything like as much as this. When we look at the papyri we shall see that below the
elite, the evidence is that most people did not read a fraction of it. Cf. also Morgans concept
of elite litterati (10910): literate education in provincial Egypt is quite different from its
counterpart among elite litterati at Rome or Alexandria or Athens, despite the fact that it
looks supercially similar. Some litterati knew, or claimed to know whole works, and even
whole authors, by heart, and they had learnt the critical appreciation of literature along with
the texts themselves. In the papyri there is no direct evidence of the reading of whole books,
except very occasionally in summary. Hellenistic poets were aware of this range of audiences
and gured them in their poetry. Cf. Bing 1993b next chapter; 1994.
50. The text reproduced here is that of Kassel-Austin PCG VII fr. 1. I am indebted to
Pages translation (GLP 57) for many particulars.
" V 6 " "+ . 15

C / V/ O V 17
@ " D .
"/ 8/; _, (, 6.
"; " , .

; / 65 + _D 21
D /; " ,/. (, 23
, ", "+ D
,/, O/ Z . 25
" . / 6 ; 34
. " , , ; 35
; ; " ,
6 / T ;
/ ,, , . 9 D
/ (/ C , .
D (, ( I 40
/ ` "+ ,
, , /, 4D O/ (

I 8.
/ " A 5 45
. / "
( ? 5 / ".
8
5 6 Z

, / ; 8. 50
[Its a male sphinx, not a cook, that Ive taken into my house; by the
gods, I simply cannot grasp a thing he says. Hes come laden with new-
fangled words; for as soon as he came in, he looked at me and loudly
asked, How great a multitude of articulates have you asked to supper,
say?Me?! Ask Articulates to supper? Youre crazy! Do you think I
know these Articulates?Then shall there be no rationer at all?
Philinos is coming, Moschion, Nikeratos, this guy and that, I ran
through the names: there was not a single Rationer that I could see. He
got annoyed, as if hed been insulted that I hadnt asked a Rationer. Very
strange, believe me. Then are you sacricing no earth-breaker?
Not me, I said.A broad-browed ox?Im not sacricing oxen,
ass!Ewes shall be sacriced, then?Me?! By Zeus, no way!
Ewes are sheep.Ewes sheep? Cook, said I, I dont know, and I
dont want to know, anything about it. Im just too countried, so talk to
me in plain language.Bring hither the sacricial groats.And
what are they?The barley.Then why say it so complicatedly,
you cripple?Is brine on hand?Brine? Why dont you suck my
cock,51 and say what you want to say more plainly?You are rash in
word, aged sir, he replied. Bring me the saltthat is brine; show me
where it lies. The lustral basin was ready. He sacriced, spoke another
thousand words such as no one, by Earth, could understand: dicings,
lots, double-folds, spitstill one would have had to use the books of
Philitas and look up every word to check its meaning. Changing my
tone, I begged him to say something like a human being. But Persuasion
herself, if she had been right there, could not have persuaded him. I
think the rogue had been the slave of some rhapsode-type from child-
hood and so was stuffed with Homers words.]
To be sure, this cook is stuffed with Homers words and performs his
culinary rites as though he had stepped off the scroll of a heroic epic.52 But
in reality one needs little more than a rudimentary knowledge of Homer
to appreciate the scene, for the epicisms of this cook are mostly far from
rare.53 Indeed, even schoolchildren were expected to know these words:
the third century b.c. papyrus that preserves our fragment was a school-
masters book with texts for use in schools.54
We may conrm this by looking a bit closer at the cooks vocabulary.
51. For the translation of this verb, cf. Jocelyn 1980.

52. The idea of a Homeric cook is actually a thoroughgoing anachronism. There is no
professional specialist for cooking in the epics of Homer; the heroes do it themselves. The
one instance that may point in the direction of professional specialization is the reference to
heralds cooking in the shield of Achilles (Il. 18.55859). According to the fourth-century
atthidographer Kleidemos, this had long been the heralds role (FGrH 323 F 5a). Cf., gener-
ally, Berthiaume 1982: 67.
53. With unwitting perception, the master speculates that his cook grew up as the slave
of some rhapsode type (vv. 4849), i.e., in the service of a not creative but merely reproduc-
tive bard. This is someone who serves up the familiar, not the uncommon.
54. In their editio princeps, Gurard and Jouguet (1938) argued that the scroll was a
manual for schoolchildren. More recently, however, it has been seen as a schoolmasters text,
from which the schoolmaster could draw for classroom use (cf. Cribiore 1996: 269, no. 379,
with pp. 12128). Besides our fragment, the papyrus contains a syllabary illustrating vowel
usage with various consonants; lists of numbers, gods, rivers; brief excerpts from tragedy, epic
(Odyssey), epigram; three scenes from New Comedy, all involving a cook; and nally a math-
ematical section. Parsons (1977: 5) speculates that a papyrus containing a text of a very differ-
ent quality, Callimachus Victoria Berenices, was the private preparation of a Fayumic school-
master, but if he is suggesting that that poem was used in schools, it must have been for
another level and quality of student altogether.
The word (articulate people, v. 6), though not appearing as a sub-

stantive in early epic (cf. rst Aeschylus Cho. 1018) is common in both
Homer and Hesiod as an epithet of men (e.g., Il. 1.250, 2.285; Od. 20.49;
Hes. Op. 109). Likewise, , rationer, banqueter (v. 11), is well at-
tested in the Odyssey as well as classical prose (e.g., Od. 4.621, 22.12; Hdt.
1.73; Plato Resp. 345c). The earth-breaker, 8 (v. 19), is absent
from earlier poetry; yet the meaning of the compound is sufciently clear
that we know that some large farm animal is intended. Broad-browed
oxen (v. 20) are familiar gures from epic (cf., e.g., Il. 20.495; Od. 11.289;
Hes. Theog. 291), as are sheep (v. 21; cf., e.g., Il. 18.524; Od. 9.45). And
anyone acquainted with Homeric descriptions of sacrice will recall the
barley groats, " (v. 34; cf. Il. 1.449 etc.). The one truly obscure
word is , brine, that is, salt derived from the sea (v. 36). Hereafter,
however, we are back on familiar ground with the cooks description of the
master as rash, (v. 38; cf. Od. 4.693 etc.), as well as with the
description of the sacrice, with its dicings, (per Page ad loc.,
plural of , as if that were a neuter noun: in fact the cook had
used as 1st pers. sing. imperf. of the verb ); lots,
(Il. 8.470 etc.); double-folds, (Il. 1.461 etc.); and spits,
4 (Il. 1.465 etc.).
Thus the schoolchildren to whom this excerpt might have been as-
signed would not have been overly taxed. On the contrary, they were prob-
ably meant to chuckle at the doltish master who, in blissful navet, reveals
to the better instructed audience his own ignorance of epic idiom, which,
ironically, he describes as newfangled (, v. 3). , common,
would describe it better (though its use by a cook is novel indeedand
amusing).
The quality of audience appreciation assumed by the poet is very telling.
How different it is from that on which Aristophanes could count in his
plays, where the humor required a high level of literary awareness (think
only of the extended parodies of contemporary and earlier authors)! In our
scene, by contrast, there are no irksome allusions to be dealt with; a school-
boys grasp of epic is to generate the happy smile of superior knowledge.
The reference to the glossary of Philitas should not mislead us. When
Stratos distressed employer is nearly driven to consult this learned work,
we cannot imagine that many spectators would have laughed because they
had used it themselves. The overwhelming majority would probably only
have heard of it, nding it exotic and, in this context, funny, an extravagant
remedy for something so commonplace: for the thought of consulting it
comes just when the cooks preliminary arrangements culminatenot in an
action puzzling to a minimally cultured Greek, but in a conventional Ho-

meric sacrice (vv. 4042) such as virtually anyone would know from nu-
merous instances in the Iliad or Odyssey, a typical scene, as Homerists
would call it (cf., e.g., Il. 1.45868, 2.42131; Od. 3.45463).
Thus, while Stratos fragment does indeed testify to the notoriety of
Philitas Ataktoi Glossai, it is hardly grounds for imagining that it was eine
Art lexikalischer Bestseller, as J. Latacz proposed.55 Certainly it suggests
that Homeric words comprised some part of the Ataktoi Glossai, that the
public may even have pigeonholed the book on the basis of this particular
aspect. But it mainly suggests that a patina of difference from everyday
usage allowed Homeric diction to stand, corporately, for language that
needed to be explained. Thus, to the question in the title of this appen-
dixwhat does Stratos Phoinikides tell us about Philitas?we must sadly
answer, Not a great deal.
55. Latacz 1985: 78. To be fair, he included the qualication wenn wir der Komdie
trauen drfen.
chapter 2
Impersonation of Voice in
Callimachus Hymn to Apollo

T oward the end of the Delian section of the Homeric Hymn to Apollo, as
the climactic event in his description of the great Ionian festival, the
blind singer of Chios presents what he calls a great wonder, whose fame
shall never perish ( , @ _/ 4, v. 156). This
wonder is the chorus of Delian maidens, who sing hymns to Apollo, Leto,
and Artemis, followed by songs about men and women of old. What is
truly wondrous about their performance, however, is their ability to
mimic the voices and sounds of all men. Each man, claims the poet, would
say that he himself is speaking. So closely tted [i.e., in its verisimilitude] is
their beautiful song ( / 1 /
/ :D " I / /D a
, vv. 16264).1 The Deliades simulation is certainly an
awe-inspiring feat. How, one might ask, could an individual listener nd his
voice reected in the collective voice of the chorusmoreover, a speaking
voice in that of a singing voice? Further, how could a multitude of listeners
each think that he was speaking when the chorus sang its song?
The relationship between these voicesindividuals and chorus, spec-
tators and performerswas, I think, to intrigue Callimachus some two
This is an updated version of an article that appeared in TAPA 123 (1993) 18198. 1993.
The American Philological Association.
1. The difculties that these verses present to interpretation are notorious. See my ap-
pendix to the present essay for a discussion.
33
and a half centuries later.2 The Hellenistic poet certainly recurred to this
Homeric hymn with remarkable insistence, mining it as a source.3 And the
multiplication of voice, with its concomitant effects on the listenerthat
, which brought the Deliades undying fameis as striking a
component of Callimachus Hymn to Apollo as it was in its archaic counter-
part. Callimachus seems to have taken the traditional monologic speech of
hexameter hymn and invested it here with a veritable chorus of voices. In
his hands it becomes a nexus of overlapping identities (cf. Winkler 1985:
203), whose components cannot easily be disentangled or even held dis-
tinct. Perhaps that of the Deliades challenged the poet to pro-
duce one of his own, a contemporary equivalent that might likewise secure
him undying fame.
With his six hymns, Callimachus was apparently the rst poet to make
extensive use of the Homeric hymns and to revive them as a genre. He did
so, I think, because, rst of all, they suited his aesthetic program: they
were pleasing in their limited size and lack of epic bombast, yet they could
be viewed as genuinely Homeric.4 Their use as a model would permit
Callimachus to turn the Homeric tradition to productive use without try-
ing to rival it, for here he would nd those aspects that were less known,
atypical, unfaded. Hans Herter has aptly called this the desire to be, in
the footsteps of Homer, as un-Homeric as possible (1929: 50 1975:
371). What is more, Callimachus conspicuously uncoupled the genre
from its original task and, in so doing, made what I consider a program-
matic gesture. For whereas the evidence, both internal and external, sug-
gests that such hymns previously functioned as prooimia (or preludes) to
epic recitation and were unthinkable apart from epic performance (signal-
ing in fact that this would shortly begin),5 the Hellenistic poetcounter
to those expectationssimply dropped the epic sequel as inconsistent
with his aesthetic goals and made the hymn stand on its own, a self-
contained and independent genreas un-Homeric as possible, though
still in the footsteps of Homer. Second, and even more important for
the question of voice, these hymns provided the only Homeric model
2. For the date of the Homeric hymn, cf. Burkert 1979: 5962; on that of Callimachus
hymn, cf. Williams 1978: 2, 36 ad v. 26.
3. For its inuence on, e.g., Callimachus hymns to Artemis and Delos, cf. Bing 1988:
chap. 3 passim; Bing and Uhrmeister 1994.
4. Thucydides (3.102) and Pindar (Pae. 7b fr. 52h Snell-Maehler) considered the Ho-
meric Hymn to Apollo to be genuinely Homeric (cf. Bing 1988: 1045), and similar hymns were
evidently part of the rhapsodic tradition of the Homeridai, whoas their name suggests
situated themselves squarely in the Homeric tradition as if they were his heirs.
5. On the Homeric hymns as prooimia, cf. Richardson 1974: 34; Janko 1981.
Impersonation of Voice in Callimachus Hymn to Apollo 35
that permitted the unmediated involvement of the poets persona apart

from the formulaic rst person of the opening and closing of the poem (cf.
the blind singer of Chios, h.Ap. 165ff. and Hesiod in his hymn to the Muses
at the start of the Theogony).
The possibilities of rst-person voice in the hexameter hymn were
clearly congenial to Callimachus; and his Hymn to Apollo puts them to
breathtaking, if deliberately provocative, use.6 The scene in this hymn is a
temple precinct of Apollo; the occasion, a religious festival. But questions
at once assail us. Which of the many festivals of Apollo are we to imagine?
And where in the world is this temple set? For the time being, we are
simply left in the dark. The poem begins as follows:
C 6 @,
/ @ D 2 2 @ .
M D
" CY; 6 C ?
6, C + 6 V 5
" ,
" + D C " D
+ 6 6.
[How Apollos laurel branch is shaking

how the whole house is. Away, away, you who are impious.
Phoebus must now be kicking the doors with his beautiful foot.
What! Dont you see? All of a sudden the Delian palm trees nodding
in joy, in the air the swans singing sweetly.
Fly apart yourselves, you bars of the gates,
you bolts as well. The god is no longer far.
Begin your singing and dancing, boys.]
We are plunged medias in res, into an atmosphere of breathless anticipa-

tion. No device of third-person narration distances us from the event.
There is a documentary (you are there) perspective: A speaker exclaims at
the signs of the imminent arrival of the god and issues orders to those who
are present. The breathlessness of his excitement is conveyed by the massed
h sounds in the rst two lines (hoion ho . . . horpex; hoia d holon; hekas
hekas hostis). Paradoxically, the Hellenistic shift toward reading and away
from live performance as the primary means of encountering literature
6. On rst-person role playing in Greek poetry generally, cf. the pioneering discussion
of Dover (1964: 20612).
actually heightens the vividness of the experience. For even as this age put
unparalleled emphasis on the written aspect of the text (Bing 1988: chap. 1),
it simultaneously maintained an appreciative ear for the play of sounds on
the page. Consequently, active readers (as opposed to those merely listen-
ing) would themselves experience the very symptoms of breathlessness as,
to a greater or lesser extent, they reproduced (and relished) those sounds in
the act of reading.7 We see, then, how, right from the start, the audience is
unwittingly drawn out of its detached sense of self and into the scene and
how the speakers voice insinuates itselfphysiologically!into that of the
reader. This process, an unobtrusive penetration of the barriers between
voices, is all the more astonishing for emanating from an inanimate object,
the papyrus scroll containing the hymn.8
Miraculous signs abound. The laurel shakes; the palm tree nods; the
doors are rattling. The urgent question " CY; (What! Dont you see?)
is obviously addressed to an unspecied bystander, a fellow celebrant in
the ritual (and one who, by his very presence, is clearly not , impi-
ous; cf. v. 2). But a reader might well do a double take at this question,
glancing uneasily over his shoulder as though to ask, Who, me? For the
style of the hymn clearly coaxes the reader into the role of one of the wor-
shipers, into identifying with the ritual community (Koster 1983: 19;
Albert 1988: 66 n. 189). But though the intensity of the voice is almost
palpable and though details of the scene seem to spring to life before our
eyesdue, in part, to the extraordinary accumulation of denite articles in
this section of the hymn (a conventional sign that something is actually
present)9we nonetheless observe that the wondrous happenings are no-
tably generic.10 The laurel, the palm, the swan, the shaking of the temple
and eerie rattling of the doors, the bolts and bars that are to spring open by
themselves at the gods approachall these are traditional features of
temple lore, miracle stories that will make the scene seem familiar, while
excluding no one with all-too-specic details. And who is the speaker? In-
7. I do not mean to return to the old view that silent reading did not exist in antiquity.
It clearly did, as Bernard M. W. Knox showed (1968). It is just as clear, however, that reading
out loud (whether to oneself or to others) remained entirely normal, even though the poets of
the Hellenistic age stressed the purely visual aspects of the text as never before and elevated
the book to the status of literary theme. On the general impact of reading aloud in Greek
culture, cf. Svenbro 1993.
8. I will show later in this discussion how another voice, too, that of the chorus, insinu-
ates itself into that of the reader.
9. Cf. Svensson 1937: 6063 with Williams 1978: ad vv. 7, 28, 32. For this and other
means of conveying the vividness of the experience, cf. Harder 1992: 389.
10. Seeck (1975: 201ff.) draws a similar contrast between intensely realistic details and
what is generic in Theocritus Idyll 7.
dications are remarkably vague. Nevertheless, the imperatives in lines 2, 6,

and 8 suggest, I think, a person in a position of authority. Perhaps, as in
hymns 5 and 6, we are meant to think of someone ofciating: a master of
ceremonies or priest perhaps.11
With the following lines, the speaker introduces a new and important
distinction within the ritual community.
E " , / @ 6D
, T " :, 6.
@ :J, # 10

4/, b ;, 6/ _ .
[Apollo doesnt appear to all, but only the noble [@ 6];

he who sees him is great, he who hasnt seen, poor.
O god who works from afar, we will see you and we will never be
poor.]
As we recall, the impious had already been warned to keep away (v. 2). But
even among those remaining (i.e., even among the pious), not all will see
Apollo. Those so favored are the elite.12 And the prospect of such favor
raises the speakers feelings to such a pitch that it prompts, at verse 11, his
rst direct address of the god (b ;) as well as the rst instance of
rst-person speech. Signicantly, perhaps, it is rst-person plural. Who is
the speaker including thereby? Surely, after the startling " CY;
(What! Dont you see?) of verse 4, we are justied in suspecting that this is
not just the conventional poetic plural: we may be meant; Apollo may ap-
pear to us.
The speakers attention now swings back to the chorus of youths. With
Apollo at hand, he says, they must not keep silent. And evidently, the
youths comply, for the speaker exclaims, Well done, boys, since the lyre is
no longer still (V 1 , 6 "/ , v. 16).
With music in the air, the worshipers are commanded " (v. 17),
the traditional order for reverent silence during the ritual, here during the
11. Cf. Koster 1983: 18ff. I. Petrovic (2007: 13436) revives the suggestion of Cairns
(1979: 12126) that the speaker is the chorus leader. The speakers orders to the group of cel-
ebrants as a whole, however, far exceed what one would expect of someone in that position.
12. Unlike the 6 who sees the god and so becomes great (), a worshiper
who does not see Apollo will be poor ()a term denoting C , ac-
cording to the Etymologicum Magnum, i.e., here precisely a commoner among the celebrants,
as opposed to one of the elect to whom the god will appear. Being thus poor, one may not
have the distinction within the group of an 6, but that is a far cry from being impious
(). The can evidently not be equated with the , who were banished from
the scene altogether in v. 2.
performance of the paean, Apollos song.13 But at this point, we notice a

telling detail, one whose full signicance has not been grasped: even the
sea, notes the speaker, maintains such silence when singers celebrate
Apollos implements, the lyre and the bow (" , @ -
/ X X , ( , vv. 1819). In his
commentary on these lines, Frederick Williams correctly points to the
long-standing belief that the natural world observes ritual silence at the
gods epiphany (1978: ad 18). And that is indeed what was traditionally
thought to occur. But Williams misses the crucial point. For here the sea
observes silence not for an epiphany in any conventional sense but for songs,
for songs about the god: that is, the evocation of the god in song is sufcient
to produce the effects normally associated with epiphany; or differently, the
song itself apparently realizes that epiphany.
The implications of such a view for the interpretation of this hymn are
substantial. It explains, I think, the vivid dramatic pose. For the choice of
that pose accentuates the sense that the literary work is itself the sacrament,
the hymn itself an epiphany in the process of being accomplished.14 Conse-
quently, the ritual community will actually embrace the community of read-
ers, and theseor at least the 6 (the noble)may indeed be able to
experience not just a vivid representation, but the epiphany itself. Will we be
among them? The blurring of the lines between the audience in the poem
and that outside it, then, is not mere play (though its certainly that as well).
We hear now how the speaker commands the chorus to being their
song. Sing Hie hie, he tells them ( , v. 25), the ritual cry
that served as the paeans refrain.15 Such exhortations commonly signal the
start of a cult song.16 Our expectations, therefore, are primed. Yet with
13. That the song for Apollo, which the worshipers are to listen to in silence
("/ 6/ / J ), will indeed be a paean is rst suggested by the
mythical comparison of Thetis in vv. 2021, who stills her lament for Achilles C/

J, and is conrmed in v. 25 when the speaker calls for precisely this
genres ritual cry, .
14. Depew (1992: 32829) perceptively discusses Callimachus simulation of performa-
tive context in his Iambi as a function of their essential textuality (329). See also Hunter and
Fuhrer 2002: 155: this hymn is itself a manifestation of the god . . . The reception of the
poem is itself the presence ( 6) of the god.
15. This may, of course, also be addressed to the celebrants at large. It would thus be an
exhortation to join in the refrain of a paean thatone must presumeis already in progress
or just beginning. The paean as a genre displays fair exibility in its use of voice: though
mainly sung by choruses (male or female), it could also be performed solo, and the refrain
could be joined by members of the ritual community. Cf. Kppel 1992: 8081.
16. Cf. Nisbet and Hubbard 1970: ad Hor. C.1.21.1. Cf. also Hopkinson 1984: ad
H. Dem. 1: The opening hortatory imperative to choir or participants is standard in
cult-hymns.
these words, the stage is set for a new ambiguity of voice. There can be no
doubt that a choral song occurs in the subsequent course of the hymn,17
that it probably begins soon after the speakers command, and that it is
evidently still in progress over seventy lines later at verse 97, when the
speaker says that he hears the refrain ( ). But in be-
tween these two more or less xed pointsthat is, between the injunction
to sing in verse 25 and the reference to hearing the song at verse 97all is
uncertainty. There are no introductory or closing formulae to set off re-
ported speech; there is nothing comparable to quotation marks that could
help us distinguish the voices.18 Still, there is clearly a shift away from the
poems vivid mimetic frame to a more conventional hymnic narrative, and
that may well be due to the chorus song.
Concerning the precise location of this shift, however, there are con-
icting signals. Verses 2829 seem to suggest that the song is already in
progress, since we hear that Apollo will honor the chorus because it is
singing (n.b. the tense) in a pleasing fashion: E, @
, / . With regard to content, however, the shift to
the song would seem to occur at line 32, where the gods appearance be-
comes the theme. Some critics have accordingly settled on this as the start
of the hymn proper.19 Stylistically, however, the section is still of a piece
with that preceding it. There is still, as there had been from the start, that
extraordinary accumulation of denite articlesas we recall, a conven-
tional sign, not characteristic of hymnic style, for something actually pres-
ent before ones eyes. Perhaps, therefore, we are still in the mimetic frame,
17. It had already been anticipated in v. 8 ( + 6 6)

and vv. 1213 ( / : / 1 (
6 ). By v. 16, as we have seen, the instrumental accompaniment, at least, has
begun (V 1 , 6 "/ ). A sequence of events is thus drama-
tized, which moves from inactivity to action on the part of the chorus of youthsn.b. that
they are praised at v. 16 because the lyre is no longer inactive ("/ ). This makes
very unlikely Cairns assertion (1979: 121) that hymn 2 as a whole is a choric hymn imagined
by Callimachus as sung at the Karneia by a chorus of boys or young men and that these are
instances of choric self-address and self-fullling injunction.
18. On this Hellenistic affectation, cf. Bing 1988: 76 n. 42; McLennan 1977: ad v. 7 and
app. II, Direct Speech in the Hymns of Callimachus (pp. 14647).
19. Thus Williams (1978: 3 and ad vv. 3296). Similarly, e.g., Wilamowitz (1924: II 78,
83, 85) and Erbse (1955: 422 Skiadas 1975: 291). Somewhat more differentiated is the view
of Albert (1988: 6870): he, too, sees the start of the eigentlicher Hymnus (70) at v. 32 but
would place a shift in speaker at v. 17, taking vv. 116 as the words of the chorus leader, vv.
1731 as the Vorgesang of the chorus itself, and vv. 32ff. as its hymn. Harder (1992: 391)
thinks that the voice in vv. 32ff. might be either the speakers or the chorus but that in both
cases . . . it is suggested by 30f. that the contents of the narrative/descriptive part are taken
from a large range of existing songs and stories.
with the description referring to a cult statue, as argued by Williams (1978:

ad 32). Other critics have placed the shift at line 47, the start of a section
concerned with Apollos various functions in the country and in the city
(e.g., Wifstrand 1933: 25n; Svensson 1937: 61). Here, the incidence of the
denite article drops precipitously until the clear return of the ritual frame
near the end of the poem.20 Other scholars, such as Malten (1911: 45) and
Cahen in his Bud edition (19493: 221), have located the shift at still other
places. And it may be that this confusion of critical voices reects a deliber-
ate ambiguity. We cannot nd the seam; perhaps we were never meant to.21
We must still determine, however, how to construe the voicing of the
verses falling roughly between the command to the chorus at verse 25 and
the return to the frame at verse 97. Are they (1) the text of the chorus
song, (2) the speakers perception of that song, or (3) simply the discourse
of the speaker independent of the chorus?22 This last possibility is in any
case unlikely. For if, as already argued, verses 2829 suggest that the cho-
rus is already singing, the next verses (3031) invitethough they do not
compelus to view what follows as a reection of that song, for they
promise the chorus continued celebration of Phoebus, who (we are told)
is a bounteous theme. And since the subsequent sections of the hymn ac-
tually go on to celebrate diverse aspects of the god, it appears that this
promise is fullled.23 Further, the presence of the refrain
at verse 80not ltered through an external source, but simply
20. From vv. 47104, Wifstrand (1933) calculates three instances in fty-eight lines. In
the fty-ve lines of vv. 146 and 10513, by contrast, he nds no fewer than thirty-three. Cf.
Williams 1978: ad v. 7.
21. In connection with the indeterminate nature of the voice, cf. Roland Barthes de-
scription (1974: 41) of the plural text: The more indeterminate the origin of the state-
ment, the more plural the text. In modern texts, the voices are so treated that any reference
is possible: the discourse, or better, the language, speaks: nothing more. By contrast, in the
classic text the majority of utterances are assigned an origin, we can identify their parentage,
who is speaking . . . ; however, it may happen that in the classic text, always haunted by the
appropriation of speech, the voice gets lost, as though it had leaked out through a hole in the
discourse.
22. Wilamowitz (1924: II 78) pinpoints the problem: Man kann nicht anders denken,
als dass es das Lied auf Apollon ist, das die Knaben 17 anstimmen wollten. Aber das scheint
nicht mglich, denn in diesem Hymnus spricht ja der Dichter aus eigener Person. Da sitzt
die Schwierigkeit.
23. Here again, however, there is ambiguity. For the interrogative pronoun in the
question [ " 8 ; (31), which follows immediately upon the promise of
the chorus continued celebration of Phoebus in 3031, may refer either to the chorus (the
subject of the previous verse: "/ C 6/ d V ) or to a given
poet, such aspresumablythe speaker of our hymn. In other words, the subsequent sec-
tions may be viewed as instantiating either the ease with which the chorus can celebrate the
god or that with which an individual may do so.
integrated into the narrative (as indeed it would be in a paean)clearly

suggests a link in this part of the poem between the spoken hexameter dis-
course of the hymn and the chorus song ( pace Erbse).
Of the other two possibilities, one must say of the rst that if this is the
text of the chorus song, it is in any case at one remove from its source. For
it is translated out of the singing meters conventionally used for the choral
paean and into an individuals spoken versethe form already identied in
this poem with the voice of the speaker. Consequently, the simplest solu-
tion is the second: namely, that we have before us the speakers perception
of the choral song.24
The melody will, of course, have been lost. But that loss enables the
speaker to adopt the choral voice and t it to his ownand not only to his
own. That extended ritual communityeach individual, wherever he may
be in the inevitable diaspora of readershipcan, in the private enact-
ment of the poem through reading, experience the same.25 Each man, to
quote the Homeric hymn, might say that he himself is speaking. So
closely tted is their beautiful song (vv. 16364). May we suggest that this
tting of the choral voice to that of others in Callimachus specically
reects his interpretation of the verb , closely tted, in the
phrase a at verse 164 of the Homeric
hymn?26 The great wonder of the Delian maidens is perhaps the more
wondrous here for no longer being limited to those actually present at the
performance on Delos. Perhaps Callimachus has capped his model by ex-
ploiting the unparalleled possibilities of dissemination inherent in the
written text as opposed to traditional live performance.
24. Thus already Wilamowitz (1924: II 79), followed by Erbse (1955: 422 Skiadas
1975: 291). A similar interpretation is proposed by Hurst (1994).
25. Speaking the poem out loud or to themselves, readers will tend to identify their
voices to some extent with those within the text. Compared with earlier times, of course, the
relatively greater prominence of silent, more purely visual reading in the Hellenistic period
offered readers greater latitude in deciding their degree of involvement with a text. For in
their silence, they might initially approach a text quite casually and noncommittally, just try-
ing its perspective on for size, while actually still withholding identication. In the more oral
cultures of archaic and classical periods, by contrast, one may well speak, with Jesper Sven-
bro, of the reading voice having to submit to the written word and of a reader dispossessed
of his own voice, a reader who in these circumstances . . . has but one means of resistance:
he can refuse to read (1993: 47).
26. Another striking instance of Callimachus elaboration of a brief (and difcult) pas-
sage from a Homeric hymn into a major element in one of his own occurs in his Hymn to
Demeter. Here Callimachus takes a passing reference to a woodcutter in the Homeric Hymn to
Demeter (v. 229) and expands it into the story of the impious woodcutter Erysichthon. For a
detailed analysis of this elaboration, see the essay Callimachus and the Hymn to Demeter in
the present volume.
It is the written form of the hymn that blurs the line between the voices
in the poem and those of its audience. And it may be worth stressing here
that it is also the written aspect of the poem that, as I am sure Callimachus
was aware, occasions many of the ambiguities of voice with which we have
been dealing thus far. For if we were actually present as spectators at a ritual
celebration, it would be clear who was speaking. We would have no doubt as
to whether a chorus was singing or an individual reciting. We would see to
whom the question " CY; was addressed, just as we would be able to
situate the speaking voices within a specic setting. We would know what
tieif anybound us to the ritual community, that is, whether we were
passionately involved or adopting the attitude wait and see.
Translate that ritual celebration onto the roll, however, and distinctions
of voice (as well as those of setting and audience composition) are for the
most part erased.27 This is an experience not unlike what Hellenistic schol-
ars, Callimachus among them, must have had when they dealt with the
occasional poetry of archaic and classical times. For this poetry arrived in
the Hellenistic era largely in written form, severed from its original Sitz im
Leben in performance and stripped of its musical and choreographic dimen-
sion (cf. Herington 1985: 4245). Not surprisingly, there was little clarity
among these scholars about the manner in which such poems would origi-
nally have been performed.28 Could it be that this scholarly encounter with
works in which the source of voice was not easy to pin down piqued Calli-
machus interest in developing such indeterminacy for aesthetic effect?29
27. This will be the case unless, of course, one supplements the text descriptively as in
reported speech (e.g., standing next to the temple, where the chorus was dancing, I heard
them sing the following . . .) and so lls out those elements that in performance would be
self-evident to the viewer.
28. Herington 1985: 2728 with n. 68. The recent controversy about whether the epini-
cians of Pindar and Bacchylides were choral or solo is just a modern instance of the problem
caused by the unadapted translation of performative poetry onto the page. For arguments
advocating solo epinician, cf. Lefkowitz 1985: 3364; Lefkowitz 1988; Heath 1988; Lef-
kowitz and Heath 1991. For arguments advocating choral epinician, cf. Carey 1989: 54566;
1991: 192200; Burnett 1989: 28394; Bremer 1990: 4158.
29. Hunter and Fuhrer (2002: 156) suggest something quite similar, though in connec-
tion with another model for Callimachus hymn, Pythian 5the Pindaric ode at the heart of
the arguments about solo versus choral performance mentioned in the previous note: It
is . . . not improbable that it was precisely the ambiguous identity of the singers of Pythian 5,
a matter discussed in antiquity as well as (endlessly) by modern scholars, from which Calli-
machus developed the apparently shifting location of the speaking voice in his Hymn to
Apollo. As so often, he goes one step beyond his models. His reworking highlights by exag-
geration the problems that arise when a performative text, such as Pythian 5, is read away
from performance; it is the read and written text that offers the limit case of the text as
script.
Moving on in our examination of the hymn, we must now contend with

the addition of one nal layer of voice. In the section of the poem immedi-
ately preceding the return to the ritual frame, the focus shifts from
Apollos role in the foundation of cities generally (vv. 5564) to his partic-
ular connection with the city of Cyrene. And this connection is expressed
in a startlingly intimate and partisan way.
6 ( M 65
6 ? M
,, e
? D / _ /.
e, ,
+ , _ D 70
" 65 D 6 a.
[It was Phoebus, too, who told Battus of my fertile city,

and, as a raven, he led the people when they journeyed to Libya,
a favorable sign for the founder, and he swore hed give walls
to our kings; and Apollo always keeps his word.
Oh Apollo, many men call you Boedromios,
many men Klarios; in every place you have many names.
But I call you Karneios, since that is the way of my ancestors.]
After the striking lack of personal and geographical detail in the opening
sections of the hymn, we are now confronted with two important particu-
lars. First of all, it would seem that we may now plausibly locate the dra-
matic setting of the hymn in Cyrene.30 Secondly, the speaker is suddenly
revealed as being from Cyrene. But whose voice is it? Students of Greek
lyric will point out that the use of the rst-person singular in no way pre-
cludes it from being the collective voice of the chorus (which, we recall, is
just in the midst of its performance). Another possibility will of course be
the speaker. But the speakers identity itself acquires an added twist with
30. A passing reference at v. 15 to walls on ancient foundations visible at the site of the
ritual ( 6/ ) had already suggested that the location was a city.
Now the specic reference to Apollos promise of walls to the Cyrenaean kings (e
/ ? , vv. 6768), as well as the Cyrenaean focus of the entire section
(vv. 6596), strongly suggest that location. The gods role here in construction is productively
analyzed by Calame (1993: esp. the section New Semantic Directions). Calame sees this
theme as a leitmotif of the narrative section of the hymn, which proceeds from the uncon-
structed space of precivilized nomadic life in the Admetus episode (vv. 4754); to the
Delian Keraton, the rst civilized structure on a virgin territory (vv. 5564); to the con-
struction of the city of Cyrene itself (vv. 65ff.).
the information about Cyrene. For Callimachus was from Cyrene (Ep. 21
Pf. 29 GP), and what is more, he was a descendant of Battus, the citys
founder mentioned at the start of the passage just quoted (Ep. 35 Pf. 30
GP).31 Since this whole Cyrenaean section (vv. 6596) also concludes with
the statement that the descendants of Battus honor no god more than
Apollo ("+ + " / (,
vv. 9596), we may wonderas a third-century b.c. audience certainly
would havewhether these references are deliberately meant to tantalize,
whether they are in fact biographical clues.32 Though the hymn stub-
bornly resists an answer, we are impelled to ask what the relationship is
between the rst-person voice and the poet? To what extent, if at all, has
Callimachus been speaking all this time? And if there are now hints of an
identication of sorts, of a retreat into the pose of author, what then are
we who had been coaxed into adopting the role of celebrantsin part be-
cause there had previously been no local or personal detail to exclude us?
Perhaps we are readers.
But then, without preparation or explanation, comes the ending.
C / 6/ _ ,D 105
" T "/ @ .
E / A E / (D
/ 8,
6/ a I.
/ " a , 110
/ \
6 4 .
, D C + , G/ C , ( .
[Envy speaks furtively into Apollos ear:
I dont like that poet who does not even sing as much as the sea.
Apollo drives Envy away with a kick and speaks as follows:
31. The labilit between the voices of speaker, chorus, and poet has been well de-
scribed by Falivene (1990: esp. 11617). Cf. also Calame 1993.
)
32. In this light, it is possible to read the reference to our kings (?
at v. 68 as meaning the kings of our family (thus Williams 1978: ad loc.). The lives of the
Greek poets became, in this period, the subject matter not only of scholarly biography but of
poetry as well. For a discussion of some of this material, see Bing 1993a. Cf. also, e.g., Cal-
limachus fr. 64, Pf. on the tomb of Simonides; Sotades 15 (p. 243 Powell) on the deaths of
great poets; Hermesianax on the beloveds of great poets (p. 98 Powell), Euphorion Hesiod
frs. 22, 22b (p. 34 Powell); and the great wealth of biographical material in the epigrams col-
lected by Gabathuler (1937). On Dichterbiographie as literary theme in Roman poetry,
see Krasser (1995).
Great is the stream of the Assyrian river, but much

sludge of the earth and much rubbish it drags along on its waters.
The bees dont bring water to Deo from everywhere,
but that which gurgles up pure and unsullied
from the sacred spring, the tiny spray, the nest blossom.
Fairwell, lord. And may Blame go to the same place as Envy.]
It appears that an epiphany has occurred before us after all. And what does
Apollo do when he appears? He takes, to our surprise, a stand on poetry.
Responding to Envys complaint that the hymn isnt long enough, he
counters rst with a vigorous kick, then with a pair of metaphors that illus-
trate the Callimachean ideal of the diminutive art work: though perhaps
lacking in size, it avoids the usual trash associated with longer forms, such
as epic (Homer excluded, of course). For Callimachean poetry is not just
small, it is pure.33
In this epiphany, the play of voice that has spanned the poem seems to
crystallize, even as the cultic and purely literary aspects merge. True to the
speakers fervent certainty at the start (v. 11), we have seen Apollo and
proven ourselves 6. Or some of us have, at least. For each reader must
appraise his own experience of the poem and decide whether he may now
answer yes to the question " CY; The verbs , (v. 105), A and
( (v. 107) permit us either response. If the poems agitated you are
there immediacy appeals to us, if the speakers breathless excitement at
Apollos imminent arrival has swept us along, we may construe the forms as
instantaneous aorists: the speaker is describing what has just that moment
taken place before his eyes34 and, within the hymns dramatic frame, what
has taken place before our own as well. We see the god. Alternatively, if our
imaginative involvement does not go that far, we construe them, with
greater detachment, as historical aorists.35 The text studiously avoids
33. Asper (1997: 10925) usefully surveys the varying interpretations of the metaphors
in the hymns nal section.
34. As, for example, in the aorists of 6 (v. 1), 6 (v. 4), and V (v.
16). Thus Williams (1978; ad ,, v. 105), following Erbse (1955: 423 1975: 293).
35. Thus Wilamowitz (1924: II 86), Koster (1983: 1718), and Albert (1988: 72 n. 211).
In the same direction, Richard Hunter acknowledges that the hymn seeks to envision nar-
rative through a powerful mode of enargeia, but to the question " CY; he feels we are
compelled to answer well, no (1992b: 1213). A decade later, Hunter and Fuhrer (2002:
155) seem to envision readers far more actively implicated in the hymns epiphany: As this
hymn is itself a manifestation of the god, it demands our active response of praise; it cannot
simply be received as a narrative. The reception of the poem is itself the presence (
6) of the god. We must respond. This Callimachus has ensured by the mimetic
mode in which he has constructed his poem; our response is choreographed by the response
privileging the one or the other reading.36 That poetics of exclusion as

Karen Bassi (1989) has called it, which is applied with such force when the
speaker bars the impious as Apollo expels Envy, turns out, in our case at
least, to be self-selecting. Whether we see the god or not depends on us.37
And if we do see the god, do we then belong to a religious or a literary
elite? Our dual role as celebrants and readers and that of the speaker as
cultic supervisor and author arise from this uncertainty. For, as we saw, the
evocation of the god in song was sufcient to produce the effects normally
associated with epiphany; the religious encounter has been identied with
the literary encounter. But that would be only natural at a time when the
written word assumed primacy in the experience of verse. With perfor-
mance no longer a given, the performative setting, as well as all its atten-
dant voices, could be collapsed into the connes of a microcosm: the
bookroll.38 Perhaps the sound, at once perplexing and alluring, of so many
previously discrete voices condensed within that space prompted the skill-
ful polyphonal play that we have seen in Callimachus. And it may be, as I
have suggested, that the seeds of such play were discovered by the poet in
the scholars encounter with earlier poems as written texts, remote from
their original site in performance.
I think we may assume that such deft use of narrative pluralism and
equivocation would have been appreciated in Callimachus working envi-
ronment, that is, among those in or near the ruling circles of the Ptolemaic
court. Here, too, communicative acts must have been closely scrutinized as
of the choir. This interpretation, however, in which readers are virtually compelled to react
to the gods epiphanyit demands our active response; it cannot simply be received as a
narrative; We must respondseems to underestimate the dynamics of readerly involve-
ment, which can range from the detached to the actively engaged (and all gradations in
between). Indeed, Hunters response in 1992 vis--vis that of 2002 nicely denes the respon-
sive parameters.
36. Likewise, although for the rst time in the hymn there are conventional formulae in-
troducing direct speech in the here and now (vv. 105, 107; the formula introducing the speech
at v. 102 is different inasmuch as it is part of an aition dealing with an event in the distant past:
Apollos slaughter of the Python dragon), all indication of space is conspicuously absent.
Envy (as so often) appears out of nowhere, as does Apollo. Nor do we know precisely where
to imagine them in relation to the speaker or other celebrants. On both counts, however, I
think that this lack of certainty can be taken as contributing to the mystery and force of the
apparition.
37. Callimachus plays with the possibilities of varying degrees of readerly involvement
elsewhere in his works as well. For a discussion of this topic, see Ergnzungsspiel in the Epi-
grams of Callimachus in the present volume.
38. Similarly Hunter and Fuhrer (2002: 157): When reading becomes a, if not the, stan-
dard mode of reception, poets must accommodate a potentially very wide plurality of sites of
reception. There is no longer a performative context which allows the unspoken to be under-
stood by a collective audience. Ritual is thus inscribed within the text.
to the identity, reliability, and self-awareness of the narrative voice. And the
Ptolemies themselves, we may recall, adopted a plural stance in using two
quite distinct voices to address their subjects: on the one hand, that of tra-
ditional Egyptian pharaohs in dealing with the native majority; among
compatriots, on the other hand, that of Hellenic monarchs for whom the
language of sovereign authority was impeccably Greek. In this century,
scholars have debated the relationship of these voices with a passion.39
Appendix: The Song of the Delian Maidens
I have translated verses 16264 of the Homeric Hymn to Apollo as follows:

They know how to mimic the voices and sounds of all men. Each man
would say that he himself is speaking. So closely tted [i.e., in its verisimil-
itude] is their beautiful song ( / -
1 / / :D " I / /D a
). It will be clear from this translation that I take the
accomplishment of the Deliades not to be merely that of singing in di-
alect, as Allen, Halliday, and Sikes thought (1936: ad v. 163). Their view
has been accepted by many recent critics (cf. Martino 1981: 41; West 1988:
ad Od. 4.279; Miller 1986: 5960 with n. 147; Gentili 1988: 51). Though
could have this meaning by the time of Aeschylus, earlier instances
are dubious at best (cf. Garvie 1986: ad 56364; Hutchinson 1985: ad 170).
The linguistic milieu is, in fact, quite limited. The festival is explicitly
Ionian (v. 147); so, too, are the individual listeners who, upon hearing the
chorus, would say that they themselves were speaking. Their dialect,
therefore, is, broadly speaking, identicalthat is to say, Ionic. And simi-
larly, the claim to imitate / , though ostensibly
sweeping, will come from an Ionian perspective and apply to Ionianswe,
too, speak of the World Champion Toronto Blue Jays, though fully
aware that this excludes the Yokahama Giants (not to mention the Havana
All-Stars). This is opposed to Thalmann (1984: 208 n. 107), who enter-
tains the possibility that lines 16264 . . . imply an international gathering
on Delos, and Clay (1989: 50 n. 102), who asserts that the festival de-
scribed in the poem is a Panhellenic institution. Conversely, Gentili
39. Cf. Bagnall 1988: 2127 with bibliography; Zanker 1989: 83103 (on this topic, esp.
9199). Recent interpretations have been increasingly receptive to hearing the Egyptian as
well as the Greek voice in Hellenistic poetry: cf. Selden 1998: esp. 309404, on Egyptian un-
dercurrents in the Hymn to Apollo; Stephens 2003; Schroeder 2009; Posidippus and the Ad-
miral in the present volume.
imagines the Deliades song as an imitation of the unintelligible tongues

of menthat is, as an artistic re-creation of the voices and dialects of oth-
ers before a foreign public, the assembly of Ionians. But if that was so, how
would members of an Ionian audience think that they themselves were
speaking? Did they all suddenly believe they had become uent in unfa-
miliar languages or dialects? Is it, therefore, merely a question of the Deli-
ades mastering various shades of Ionic dialect, so that each member of the
audience would recognize his local patois? What, one might ask, would be
so wondrous about that?
In order not simply to rationalize the great wonder in this way, trivial-
izing it, I think we must accept some sort of prodigy like that described in
the fourth book of the Odyssey (278f.; cf. West 1988: ad loc.). There, at a
banquet for Telemachus, Menelaus tells how, in an effort to make the
Greeks betray themselves, Helen called to the men within the Trojan horse
by name, treacherously tempting each one of them by mimicking the voices
of their various wives: 6 / 4 4 , /
/ :/ . Typically, the ancient commen-
taries (or scholia) react incredulously: How did she know them all so as to
be able to imitate their voices? The imitation of the voices is utterly ridicu-
lous and impossible. And how could the men believe that their wives were
there? ( @ A J , G " ; +
? . / [ 6, @
" ;). Yet it is precisely characteristic of the mirac-
ulous that it strains belief. And so it is with the Homeric hymn; we should
not dilute the just to make it more easily believable.
Incidentally, the meaning of the hapax is obscure. The
are castanets (cf. Athen. 3.2, 2.137.7; PMG 955), and
describes making a clacking sound with castanets. Accordingly, LSJ (s.v.)
renders as rattling with castanets, to give the time in danc-
ing. But this makes little sense with what immediately follows:
" I / /. It is worth noting that (from
, castanets) and related terms could be used metaphorically of
persons in the sense of rattler or chatterer (Seaford 1984: ad 104). And
it may be that can bear a metaphorical interpretation as
well, in which case it will mean chatter (I very much doubt M. Lefkowitz
suggestion that it could mean humming an accompaniment [1985: 49]).
The variant reading , likewise a hapax, would amount to
much the same thing, as it comes from , meaning to chatter with
the teeth (Il. 10.375) or simply to stammer (Bion 9.9 Gow). (On abstract
nouns in -, cf. Khner and Blass 189092, II 272 no. 28.)
chapter 3
Callimachus and the

Hymn to Demeter

T he Hellenistic era saw revived interest in the Homeric hymns, and

among these the Homeric Hymn to Demeter is no exception.1 One
need look no further than book 4 of Apollonius of Rhodes Argonautica to
see that this hymn was not just available to the poets of the age; it was care-
fully read and appreciated.2 Yet its impact on Callimachus is less easy to
trace. Was this poet, likewise, a careful reader of the hymn? And does his
work reveal comparable inuence?
Though various Hellenistic authors composed hymns to Demeter,
Callimachus is the only one in this period from whom we have a hexame-
ter hymn, that is, one directly in the tradition of the Homeric hymns. In
light of Callimachus learned, ubiquitous allusions to Homeric hymns
elsewhere in his own, one might anticipate that his Hymn to Demeter would
make especially heavy use of its Homeric precursor. But this is not the
This is an updated version of the essay that appeared in Syllecta Classica 6 (1995) 2942.
The University of Iowa, 1996. My thanks to James Clauss and Mary Depew for their helpful
comments on earlier drafts of this essay and also to my colleague Garth Tissol for a careful
reading and good suggestions.
1. Cf. Richardson 1974: 68: The inuence of the Hymn, certain or probable, may be de-
tected or suspected in many places in Greek literature. But its popularity was clearly greatest
in the Hellenistic period. For the hymns inuence in modern times, cf. Foley 1994: 15369.
2. As numerous scholars have noted, Apollonius makes detailed and extensive use of the
Demophoon episode: this is his model for the narrative of Thetis vain attempt (at 4.868ff.) to
render Achilles immortal by bathing him nightly in ames. Cf., e.g., Livrea 1973: ad v. 868.
49
case. It is telling that Hopkinson (in his commentary on the hymn),

though occasionally pointing to the Homeric hymn in individual notes,
omits it entirely from both his Index of Subjects and his Index of Pas-
sages Discussed. 3 The same has been true of other scholars,4 though two
have begun to redress the matter in recent essays.5
Echoes of the archaic hymn are in fact audible, assuming one listens
carefully. For these echoes are not always distinct or obvious. Almost no
diction is adopted intact, or when it is, there are differences in application.
The same holds true for content and voice, which appear transformed in
Callimachus treatmentthough not, I should stress, beyond recognition.
For one recent critic, the tale of Persephones rape as told in Callimachus
illustrates a fundamental point about Alexandrian poetry, viz. its cultiva-
tion of obliquity for its own sake.6 This point may be applied more gen-
erally to how the Hellenistic poet sets himself in relation to his archaic
predecessor. That relationship is certainly oblique. But whether he culti-
vates that obliquity for its own sake remains an open question, and one
we must ask.
In what follows, I will attempt to trace the inuence of the Homeric
Hymn to Demeter on Callimachus and show that its echoes form a useful
counterpoint in reading his hymn. While that counterpoint may not lead
to a comprehensive view of the poem, I hope that it will at least illuminate
important aspects. We will see that the Hellenistic poets deployment of
this model entails deliberate distancing. I will examine that distancing in
two steps: (1) in connection with diction and content;7 (2) with respect to
voice. In a third and nal section, I will try to ask whether this twofold
distancing tells us something about the meaning of the poem. I must say
from the start, however, that we will be left with far more questions than
answers.
3. Hopkinson 1984.
4. Fantuzzi (1993), while (rightly) arguing the importance of religious choral lyric as a
model, actually precludes any inuence whatsoever from the Homeric hymns on Calli-
machus hymns 5 and 6 (inni solo di nome): certo che linnodia esametrica omerica
non rappresentava in nessun modo un modello, n per i contenuti e la morfologia comp-
lessiva, n per i singoli elementi strutturali (932). Bulloch (1977: esp. 99101) does indeed
set hymn 6 against the structural backdrop of a Homeric hymn, but it is that to Dionysus (no.
7), not Demeter.
5. Hunter 1992b: esp. 911; Haslam 1993: 119 n. 14.
6. This fundamental point is slyly tucked away in a footnote by Hinds (1987: 151
n. 11).
7. Several of the instances discussed in section I were detected by Hunter (1992b) and
Haslam (1993).
Callimachus and the Hymn to Demeter 51
I. Eludingand Alluding toa Literary Paradigm
The most immediately striking echoes of the Homeric Hymn to Demeter

come where Callimachus tells the same story as the archaic poem: namely,
Demeters desperate search for Kore. In the Homeric hymn, we hear
twicethat, in her grief, Demeter refuses to eat, drink, or wash: " /
? / / , "+
(neither did she touch ambrosia and sweet-tasting nec-
tar in her grief, nor did she bathe her skin, vv. 4950); /
6 V+ / ? (but never laughing, touching nei-
ther food nor drink, she sat, vv. 200201). In Callimachus, this likewise
appears twice: " _/ / ( "+ (you nei-
ther drank nor did you eat during that time nor did you wash, v. 12);
6 . . . / " " "+ (you sat . . .
parched without drink, and you neither ate nor did you wash, vv. 1516).
As close as these passages are, Callimachus typically varies every word: for
the Homeric hymns "+ . . . / and , he substitutes "
_/ / ( and " ; for "+ ,
he compresses to "+ . The Homeric hymn is consistent in men-
tioning Demeters abstinence from food before that from drink; in Calli-
machus, it is the reverse: he is consistent in mentioning her abstinence
from drink before that from food. In both hymns, moreover, fasting ap-
pears in parallel contexts: rst in connection with Demeters desperate
search (Hom. Hymn vv. 4950 Callim. v. 12), then when she is seated
(Hom. Hymn vv. 200201 Callim. vv. 1516).8 Incidentally, Callimachus
does adopt one word from the hymns description of fasting, but he char-
acteristically dislodges it from its original context. At verse 200 of the
Homeric hymn, the goddess herself is 6 V+ ; in
Callimachus, it is her fasting worshipers who are (v. 6).9
Where the themes of the poems diverge, such transposition will be a
fortiori more likely. This is the case in the myth of Erysichthon, which the
8. Callimachus combines this with a separate incident from the archaic poem, namely,
sitting by a well. In the archaic poem, it is the Parthenion (Hom. Hymn vv. 9899: I . . . /
M ); in Callimachus, the Callichoron (v. 15: 6 M 6
; cf. also fr. 611). This latter well is also mentioned in the Homeric hymn (v. 272), in
connection with the site of Demeters future temple. Is Callimachus correcting his archaic
predecessor?
9. The celebrants behavior, of course, itself echoes that of the goddess and so is a model
within the poem for the sort of imitation, with its inevitable dislocations and transformations,
that characterizes the relationship between Callimachus and his literary predecessors.
speaker now tells. In the Homeric hymn, for instance, the goddess briey
reveals her divinity on entering the house of Celeus: /
(her head reached the roof, vv. 188189). In Callimachus, this
same verb is used of Demeters sacred tree: , (a
great tree reaching to the sky, v. 37). Note the deliberate variation in the
use of the participle and the different (and unusual) case of the object.10
When, somewhat later in the hymn, Callimachus presents the epiphany of
the goddess herself, he tops the Homeric hymn by having her head reach
Olympus (: + , 9/ /, her feet
touched the ground, but her head reached Olympus, v. 58, as opposed to
the Homeric hymns ? / / 6/ " ( 8 /
, she stepped on the threshold with her feet, and her head reached
the roof, vv. 18889). In either case, the epiphany inspires fear ( /
, ,+ , awe, wonder, and pale fear
seized her, Hom. Hymn 2.190 + / ?, 6
,, / 6 , they, half dead when they saw the goddess,
rushed away, Callim. vv. 5960).
The goddess tree islike Kore in the Homeric hymnan object of as-
sault,11 for Erysichthon, his heart set on timber for a banquet hall, rushes
into the sacred grove (/, Callim. v. 33; cf. Hades in Hom. Hymn v. 17,
=), with the object of despoiling it.12 He sinks his axe into the tree.
The tree cries out ( : , she screamed a sad song to
the others, Callim. v. 39), just as Kore did at her rape (, / / =
J, she screamed with piercing voice, Hom. Hymn v. 20).13 In either
case, Demeter hears the cry of distress (Y @
, Demeter heard that her sacred tree was in pain, Callim. v. 40 =/
/ , / O , I heard her voice
through the barren air, as though she were suffering violence, Hom. Hymn
vv. 6768; cf. v. 39); and in either case, she reacts with anger (, +
, Callim. v. 41 , Hom. Hymn v. 91).
10. Bulloch (1977: 11617) notes, in other authors the genitive case would have been
used: Callimachus is in fact extending the Homeric / (Homeric Hymn to
Demeter 188f., Homeric Hymn to Aphrodite 173f.), rening the common construction into a
much rarer usage.
11. Note that Kore is compared with plants ( J, v. 8; , v.
66). The similarity in situation is observed also by G. Baudy (1991: 2526), though without
detailed thematic or structural analysis.
12. The inviolate grove, with its burgeoning fruit trees (vv. 2728), bubbling spring (vv.
2829), and frolicking nymphs (v. 38), is, of course, a standard emblem of virginity.
13. It is noteworthy that tree nymphs also make an appearancea suggestively brief
onein the Homeric hymn: the 6 of v. 23 (cf. Richardson 1974). These
nymphs, however, are endowed not with voice (as in Callimachus) but with hearing, though
it is decient. They fail to hear Kores cry when she is raped.
Thereafter, as in the Homeric hymn, Demeter impersonates a nurtur-

ing older female (Callim. vv. 42ff. Hom. Hymn vv. 101ff.), and in both
hymns, her rst word in this guise is , child (Callim. v. 46 ,
Hom. Hymn v. 119). And just as she addresses Erysichthon, the kings son,
as (child much prayed for by your parents,
Callim. v. 47), so, with similar paronomasia, she addresses the royal daugh-
ters of Celeus: [] . . . / E 6 (may
the gods give you children to bear, as parents wish for, Hom. Hymn vv.
13637).14 It is striking, too, that in either poem, the goddess addressee is
the subject of an extended epic-style simile (Hom. Hymn vv. 17478 Cal-
lim. vv. 5052), a feature that, as I have noted elsewhere, almost never oc-
curs in the Homeric hymns (that in the Homeric Hymn to Demeter is the
longest example) and is rare in the hymns of Callimachus.15 Finally, the
goddess expresses her wrath in either poem by inicting (hunger,
Callim. v. 66 Hom. Hymn v. 311)though in markedly different ways.
These examples will, I hope, sufce to show that Callimachus knew the
Homeric hymn and drew on itmore than is generally admitted. But the
very fact that it frequently lurks in the background makes Callimachus
relatively sparing use of it all the more striking. It suggests a deliberate dis-
tancing. Indeed, Callimachus expressly rejects the subject matter of the
Homeric hymn ( ` , Dont,
dont let us tell those things that brought tears to Deo, v. 17), abandons its
Attic/Eleusinian locale in favor of one equally meaningful to the goddess
( / 6 M / @ /, the goddess was wild about
the place, as much as about Eleusis, vv. 2930), and pointedly coats its
epic diction in a literary Doric veneer. As though to stress the gap between
how much he knows and could refer to and how little he chooses to, Calli-
machus most striking allusion to a Homeric hymn to Demeter is the ex-
tended, verbatim quotation from the last verse of the recondite and tiny
Homeric hymn 13, to Demeter (, , , Hail,
goddess, and keep this city safe, Hom. Hymn 13.3 , ,
, Callim. v. 134).
Let us look at one last possible allusion, which may serve to point up
just how radically Callimachus sets himself off from the Homeric hymn
even if he allows us to trace the root of his departure to that self-same
hymn. I am referring to his very different choice of myth, namely, that of
14. Of course, the hapax is closer yet in form and meaning to Metaneiras
statement that her son Demophoon is (Hom. Hymn v. 165, likewise a hapax) and
(Hom. Hymn v. 220).
15. Cf. Bing 1988: 12324. On Callimachus simile, cf. my discussion in the following
section of this essay.
Erysichthon.16 Scholars have often debated whether Callimachus was the

rst to specify that Erysichthons crime was cutting wood from the god-
dess grove. What seems clear, in any case, is that prior attestation is dubi-
ous and that this version of the myth is quite recherch.17 One expects no
less from Callimachus. But near the middle of the Homeric hymn is a baf-
ing passage that has deed precise understanding. It is verses 22730,
where Demeter agrees to take care of the infant Demophoon.
, _ ( J
_/ / 6 _/ #D
, #,
, / 6 6 6. 230
[I will nurse him. And not due to his nurses foolishness,
I think, will an attack or a cutter [?] do harm to him.
For I know an antidote far stronger than the woodcutter;
I know a good shield for a painful attack.]
While a critical consensus has developed that #, , and

# all refer to the same thing, namely the cutting of herbs for
magical purposes,18 there is much that is still unclear.19 In particular, crit-
ics have suggested that # must mean the same as 8, that is,
one who cuts herbs for magic or medicinal uses. But # never
appears in this sense. From Homer on down, whether as adjective or noun
or in the verbal form #, it refers to cutting wood. The term re-
mains a , then. I suspect that the same held true in Hellenistic
times.20 Could it be that this bafing woodcutter here intrigued Calli-
machus, pricked his fancy, gave him a nudge in shaping the myth? If so,
what a distance he put between himself and his source! While the core
myth of the Homeric hymn is expressly excluded by Callimachus, that pe-
culiar woodcutterwho receives the merest mention in the archaic
16. On this myth in Callimachus and Ovid and also on the comparable tale of Paraibios
father in Ap. Rh. 2.46890, see J. Murray 2004.
17. On this debate, cf. C. W. Mller 1987: 6576.
18. Thus Richardson (1974: ad v. 228).
19. The word #, for instance, is unique; no entirely plausible explanation for it
has been found. As Richardson points out, the context does not suggest a specic plant. And
a neuter participle in the sense of cutting does not accord with Homeric usage. Richardson
concludes, The form of the word remains unexplained (ibid.).
20. Of course, one cannot rule out the possibility that these were popular terms, current
in colloquial usage, and that anyone on the street would have known at once what they
meant.
poem, only to be dismissed as a threat to Demophoon by the goddess

looms up as the central gure of the myth in Callimachus.
II. The Feminized Voice
By far the most striking departure Callimachus makes from the Homeric
hymn is the voice he chooses to speak in. It is that of a womanone pres-
ent at a religious festival whose events unfold and engage us imaginatively
with their vivid immediacy as the poem proceeds.21 The festival can in all
probability be identied as the exclusively female Thesmophoria; and the
time can be pinpointed at the transition between the second daya day of
fasting, the and the nal day, the , with its hope for
fertility.22 At the close of a long day in which they have had neither food nor
drink (vv. 67), the celebrants look impatiently (v. 7) toward the coming of
the goddess (basket, vv. 12). This actually arrives in the course
of the poem, carried on a four-horse chariot and signaling the reestablish-
ment of fruitful and normal life.23 The speakers gender is established with
certainty at verse 124, where she says, we walk through the town without
, the headbands typically worn by women (
). A sense of strong collective gender identity is conveyed by the
fact that the speaker characteristically says we (i.e., uses the rst-person
plural)24 and addresses only women.25
21. As Wilamowitz (1924: II 25) says, Callimachus erreicht einen Grad von 6,
wie es sonst kaum mglich war, denn es wird uns nicht erzhlt, was die glubige Menge tut
und sagt und empndet, sondern alles spricht uns unmittelbar an.
22. Cf. Hopkinson 1984: 3536; C. W. Mller 1987: 8588. Both scholars point out how
those aspects singled out in vv. 1821 as appropriate themes virtually dene Demeter as
Thesmophoros (especially v. 18, ). Mller is especially good on
the assimilation to the Thesmophoria of details from mystery ritual.
23. Cf. Albert 1988: 6366, esp. 6566, for an analysis of the Szenerievernderung (to
use Alberts term) portrayed in the poem.
24. Cf. vv. 6, 17, 121, 124, 125, 127; rst-person singular only in the prayers at vv. 11617
and 138. Falivene (1990: 10328, here 12224), likewise notes the speakers identication with
the group, without, however, dealing with the implications of its gender. Henrichs (1993: 131)
would distinguish the rst-person singular at vv. 11617 as the voice of the poet, which al-
legedly draws attention to itself . . . at this juncture, while acting as an intermediary between
the imaginary ritual occasion and the mythical past. But nothing in the sentiment of the
prayer suggests the poet rather than a participant in the ritual. A more likely candidate for
the poets voice is the rst-person singular in the last line of the poem (G , v. 138),
where the weight of generic convention allows forthough it does not enforcethat possibil-
ity. We should note, however, that even here poetic self-reference is not explicit, as it is at the
close of the Homeric Hymn to Demeter ( / b / 4, v. 494), of
M
other Homeric hymns, or even of Callimachus hymns (1.92ff., 2.106ff., 3.268).
25. Cf. vv. 1, 5, 1016 (Demeter), 83, 11617 (Demeter), 118, 12833, 13438 (Demeter).
Callimachus seems to have taken pains to characterize this womans

voice with a subtle but unmistakably female accent, that is to say, to fem-
inize it.26 Its specifically feminine inflection is worth listening for in a
few instances, since it has not been noticed until now.27 Let us consider it
first in two passages containing similes. Since similes characteristically
link remote mythical happenings with the familiar conceptual world of
the audience to make them more vivid, they may tell us something about
the horizon of expectations shared by the speaker and her listeners in
the poem.
At verses 5052, to illustrate just how savagely Erysichthon glared at
Demeter, the speaker uses a simile in which the youth is gendered female:
he is compared to a lioness who has just given birth ( / / #
V+ / e 6 # /
b, =, glaring at her more
ercely than a lioness with newborn cubs glares at a hunter in the moun-
tains of Tmaros, whose look they say is the most terrible of all). Many
scholars have noted how closely the simile resembles one in Euripides
Medea where the heroines look is likewise compared with that of a mother
lion ( / , @ /
CJ, she glares at her servants with the look of
a lioness who has just given birth, whenever one comes near her with a mes-
sage, vv. 18789). None, however, has remarked on how the choice of gen-
dera lioness rather than a lionhelps delineate the character of speaker
and audience in both passages. It can be no coincidence that in both the
tragedy and the hymn, the speaker is a woman addressing herself to an
audience of women. I propose that the Hellenistic poets choice of a lioness
to describe Erysichthona choice even more striking than in Euripides for
its crossing of gendersis a deliberate touch of ethopoeia: it suggests that his
26. The feminized voice in Callimachus has recently been examined by Gutzwiller
(1992a: esp. 37378) regarding Berenikes Lock. While I nd her approach very helpful, the
question of the locks gender in that poem is problematic. For a strong case contra Gutzwiller
in favor of the locks masculinity, cf. Koenen 1993: especially 9495 with n. 164. Though
feminine characterization in our hymn is not amboyant or obvious, it will become clear that
I nd the speaking voice less nebulous and uncharacterized than is supposed by Hopkinson
(1984) or Wilamowitz (1924: II 2425).
On the appropriation of the feminine voice to male discourse, cf. Halperin 1990: 11351,
esp. 142ff.; Skinner 1993.
27. One might be tempted to construe the oft-remarked folksiness and navet of the
voice (cf., e.g., C. W. Mller 1987: 25 n. 64, 30) as a function of gender. But while these qual-
ities are not incompatible with a womans voice and indeed help characterize a kind of
womans voice here in the hymn, they could equally well be found in a man. We must there-
fore look elsewhere for evidence of the specically feminine.
speaker knows, as a woman, that a mother lions erceness in protecting her

newborn cubs will strike a chord of recognition in her female audience.28 In
other words, within the context of the all-female festival, the poet naturally
enough portrays her as speaking woman to woman.29
A female sensibility also emerges, I think, in the briefer similes of verse
91. Here Erysichthon is said to be melting away like snow on Mimas, like
a doll in the sun (E + , E M ( ). The rst part
of the comparison has Homeric precedent (Od. 19.2058), but the second is
strikingly un-Homeric. Does the reference to a childs wax doll reect a pe-
culiarly female perspective? One cannot be sure. Still it is highly suggestive
that precisely these two comparisons are likewise used by a woman to a fe-
male audience in the second poem of Theocritus: I went all colder than
snow, says Simaitha addressing the lady moon (v. 106), and my shapely
body froze like a doll (, v. 110). The two passages may in fact play off
each other deliberately, since, as Gow notes (ad v. 110), they reverse each
other: the frozen snow and doll in Theocritus are portrayed as melting in
Callimachus.30 In either case, the comparisons seem the appropriate ex-
pression of a woman speaking in a feminine setting.
Moving beyond similes, there is striking evidence that a womans out-
look colors the telling of the myth when, at verse 83, the speaker unex-
pectedly breaks into the narrative with an apostrophe. Turning directly to
Erysichthons mother to commiserate with her plight at having to make
excuses for her sons absence from normal social life, the speaker ex-
claims, Wretched mother, what lies did you not tell for love of your
child? ( , / " 6, ;). This is the only
28. The choice of the lions gender appears all the more deliberate when we note that in
HomerCallimachus chief paradigm for hexameter dictionlions are always male (the
feminine rst occurs in tragedy and Herodotus). Thus at Il. 17.13336, a simile ad-
duced by Hopkinson (1984: ad vv. 5052) as one model for Callimachus, Aias is likened to a
male lion defending its young. For sexual inversion in Homeric similes, cf. Foley 1978.
29. If the feminine aspect of Callimachus simile is illuminated by reference to its model
in the Medea, its novelty in a hymnic framework may be brought out by contrast with its
counterpart in the Homeric hymn. As previously noted, one of the very rare instances of an
extended simile in the Homeric hymnsthe longest there isoccurs at Homeric Hymn to
Demeter vv. 17478, and it comes in a context strikingly parallel to that in Callimachus. The
Iliadic pedigree of that simile (Il. 15.263ff.) is carefully analyzed by Richardson in his com-
mentary (1974: ad vv. 17078; cf. also p. 32 of his introduction): its voicing and style are stu-
diously epic. If Callimachus is indeed recalling that simile in his overall context even as he
draws on the very different model of Euripides, that recollection serves to highlight the
transformation undergone by the traditional hymnic narrative voice.
30. Cf. Deubners (1921: 37678) suggestion that Theocritus II was the inspiration and
model for Callimachus mimetic hymns 2, 5, and 6: Ein Stilprinzip hellenistischer
Dichtkunst.
such interruption in the myth. And its motivation clearly lies in the fellow
feeling of one woman for another. The compassionate apostrophe recalls
another earlier in the poem that was likewise addressed to a mother dis-
traught because of her child. I refer to the sympathetic questions posed to
Demeter about her search for Kore in verses 1016, leading to the emo-
tion-charged words ` (Dont,
dont let us tell those things that brought tears to Deo, v. 17).31 In retro-
spect, it seems likely that, here too, the poet invites us to understand the
apostrophe and the emotional coloring of the passage as prompted by fe-
male fellow feeling.32
Worthy of mention in this context is also the metonymous description
of Erysichthons nurse at verse 95 as the breast that nursed him (b
(). D. Fowler has suggested that the phrase represents
Erysichthons point of view: As a child he saw the Nurse as a breast, but
now too we catch him looking at her in his hunger in the same way.33
This can scarcely be correct, since it detaches the breast from its verb. In
verses 9495, we hear that Erysichthons mother wept and that his two
sisters and the breast that nursed him and the many tens of slave girls all
groaned heavily: + Z , 1 / ( / /
b ( . As far as the nurse is
concerned, clearly these groans come from her point of view, perhaps
with specic reference to her breast. For if we allow ( its imper-
fective force (the breast that was nursing him), rather than taking it
with Hopkinson as the loose epic use of the imperfect (ad v. 95), we
might picture the poor nurse pressed back into service by her ravenous
nurslingreason enough for her to groan or, rather, for Callimachus to
present the sympathetic female speaker and her listeners imagining her
groans.
Verses 9495 remind us, moreover, thatapart from Erysichthons
father, Triopasthe audience lamenting the youths calamity within the
narrative is an exclusively female group. In its range of status and age, it
corresponds (deliberately, no doubt) to the audience within the ritual
31. Cf. Hopkinson 1984: 9596 for the excited, emphatic repetition of the negative,
whose tone is generally semi-colloquial.
32. A womans sympathy for a mother perhaps also motivates the speakers use of the
compassionate term to describe Erysichthon at v. 93 (M 8 4
(). We should note, too, that when the speaker strangely refers to Erysichthon
as C (v. 56) just at the moment when he is portrayed at his most savage, she is mimicking
the maternal with which the goddess addressed him in vv. 4647.
33. Fowler 1990: esp. 43.
frame, the group of female celebrants listening to the sorry tale (cf. vv. 36
and especially vv. 12833).34
III. Whose Voice Is It Anyway? Speaker, Reader, and Poet
Only in this Callimachean hymn is the speaker certainly female. By an act

of narrative transvestism, the (male) poet impersonates a womanone ap-
parently in a privileged position, since she instructs the women when to
utter the ritual cry (vv. 12); is apparently the rst to see the evening star,
signaling the goddess imminent arrival (v. 7); forbids the uninitiated to
look on the goddess (v. 3) and tells them they may follow only as
far as the prytaneion (v. 128). As for initiates, only those under sixtyshe
saysmay witness the mysteries (vv. 12930); pregnant or disabled women
should follow as far as they can (vv. 13032). But impersonation is not
restricted to Callimachusas Wilamowitz saw when he asked, Who is
giving [these instructions]? . . . Who is narrating? To realize the poets
boldness and art, he continued, no more is necessary than to recite the
poem in a sufciently lively manner to oneself or to others.35 For each
reader is invited to play the role, to imagine him or herself in the part of
this woman present at the ceremony.
To be sure, the degree of imaginative involvement depends on the atti-
tude of the readers, on their relative willingness or reluctance to commit
themselves to the part. As the previous sections of this essay have shown, the
poem itself pulls readers in two directions, allowing them to adopt diver-
gent roles. A learned, sophisticated reader may play the part of literary de-
tective, uncovering traces of the Homeric Hymn to Demeter in Callimachus
Hymn to Demeter and appreciating the remarkable transformations that that
model undergoes in this work. Allusion of this kind is not intended for the
audience in the poem. Nor is it part of the consciousness of the narrating
34. I wonder whether we may also hear a specically female quality of voice in the prayer
at the end of the hymn ( , , , , , v. 136). Hopkin-
son 1984 ad loc. has noted that the line bears a marked resemblance to the corrupt Sappho
fr. 104 (P = 6/ _ / =, ,,
), a hymenaeal fragment that Page (1955: 121 n. 1) suggested might be
spoken by girls, or by some other party which sympathized with the bride. It may be, then,
that Callimachus wanted to recall a genuine womans voice in this passage. N.b. that Hopkin-
son also considers a lyric source, such as Sappho fr. 104, for the anaphora of Hesperus in vv.
78 (cf. Catullus 62.1).
35. Wilamowitz 1925: 248 (wer gibt ihn [sc., den Befehl]? . . . Wer erzhlt?), 249
(Nichts ist weiter dazu notwendig, als dass man sich oder anderen das Gedicht mit der
entsprechenden Lebhaftigkeit vorliest).
persona, authoritative though she is in her context. Rather, it is a pastime

strictly between poet and reader, injecting ironic distance between their
perspective and that of the characters in the poem.36 Yet even those who
maintain a posture of detachment toward the text, who withhold identica-
tion with the speaking role, function almost as observers at the festival: the
ritual community embraces the community of readers.37 The poet draws us
into joining the celebration in our minds eye through the striking 6
conveyed by the feminized voice of the speaker. And the hymn itself helps
ease us into the part, since role playing is pregured in the poem and sug-
gests our own. Demeter impersonates Nikippe, whom the city appointed
as her public priestess (Y, / (-
, 6, v. 4243). She thus provides the model for the poets compa-
rable guise: the woman directing the goddess festival. Demeter abstains
from food and drink, thus serving as paradigm for the fasting worshipers:
these identify with and enact her ordeal. Our own assimilation to the role
of onlooker, then, simply implements a model offered in the hymnone
more impersonation in a series. So long as we dont refuse to read, we im-
plicate ourselves in the proceedings; we play the part of witness to the se-
cret female rite.
For us to do so will mean diverse things depending on who we areone
thing if we happen to be women, something different altogether if we hap-
pen to be men. Let us pause for a moment to consider if the former possi-
bility is likely. Might the feminine perspective of the hymn be conceived
with a female audience in mind?38 Gutzwiller has recently argued that we
36. For a good discussion of this double perspective, likewise between male poet and
female narrator, cf. Segal 1984.
Many scholars would see the same sort of dissonance between the speakers stated pur-
pose in telling the tale of Erysichthon (so that one may avoid transgression, G
# , v. 22), and what they perceive to be a burlesque undercutting of that
purpose in the actual narrative. Thus, e.g., Depew (1993: 72) claims that it is not in any way
part of Callimachus intentions to make a moral point or, for that matter, to praise Demeter,
that [w]hat was announced as an introductory exemplum has turned out to be a comedic por-
trayal of a bourgeois family and its socially embarrassing son.
37. Interestingly, as Falivene notes (1990: 127 with n. 75), explicit poetic self-reference
indeed, the very notion of song, together with its terminologyare completely absent from
hymns 5 and 6 (6 in v. 1 is not a chorus periodically recurring refrain in song but
a ritual cry of the participants responding to an actual event at the scene: the baskets return).
Does that absencebecause it removes any special qualication needed for the partenhance
the possibility of readerly identication with the role of the speaker or of the celebrants (the
latter since it erases a potential difference between the experience of the community of readers
and that of the ritual community in the poem)?
38. I set aside such views as that of Laronde (1987: 365), who believes that the poem was
intended for performance at the festival it describes and that geographical allusions show that
must see another poem of Callimachus, the Lock of Berenike, as reecting

above all the concerns of the queen and that her court was at pains to cre-
ate a public image of her royal marriage that had a special appeal to femi-
nine tastes.39 A comparable approach to the Hymn to Demeter, taking into
account the interests of the queen and her circle, might be plausibleespe-
cially if there were greater certainty that the Ptolemaic queens celebrated
by Callimachus were assimilated to the goddess Demeter in cult.40 A hymn
to this goddess would, in that case, have obvious attraction for the court.
Yet on balance we must reject such a view, since evidence for the queens
assimilation to Demeter is slimespecially when compared with the wealth
of material connecting Ptolemaic queens with Aphrodite, as in the Lock of
Berenike.41 Problematical, too, is that the hymn (unlike the Lock of Berenike)
makes no reference whatsoever to a royal personage or context: one would
have to read that between the lines. Finally, a link to the Ptolemaic court
begs the question of the poems date, which is quite uncertain but which
many scholars would place relatively early in Callimachus career, before he
had even come to Alexandria.42 In the absence, then, of clear pointers
suggesting that Callimachus intended his hymn for the sort of female audi-
ence previously considered, I think we must assume that the majority of its
this may be localized in Cyrene: nous avons l un hymne que les femmes de Cyrne pou-
vaient chanter lors des Thesmophories. Against such interpretations and for an understand-
ing of the hymn as a generalized literary evocation of a festival of Demeter, see, among many
others, Hopkinson 1984: 3537. As the excavator of the sanctuary of Demeter at Cyrene has
trenchantly put it, all efforts to pin down a specic site and festival for the hymn are vitiated
by the fact that the few topographical features actually named in connection with the hypo-
thetical setting are duplicated in practically any Hellenistic Greek city and thus are empty of
any signicance except as poetical images. He adds, It must be stressed that Callimachus
was . . . not [composing] the Blue Guide, (White 1984: 48). Even A. Cameron, who resusci-
tates E. Cahens notion of performance at festivals outside the formal framework of the
festival itself (Cahen 1929: 281), seems to acknowledge that the lack of geographic speci-
city in this hymn speaks against performance at a particular site (cf. Cameron 1995: 64 n.
258). A more sustained argument for the possible performance of Callimachus hymns is
made by I. Petrovic (2007: 11481, esp. 13234 for Callimachus Hymn to Demeter), but
though she thinks that there is no basis for thinking that the hymns were not performed, she
admits that she cannot prove that they were: Diese Frage betrachte ich als unlsbar (137).
39. Gutzwiller 1992a: 361.
40. For Arsinoe II commemorated in Alexandrian street names as Eleusinia and Kar-
pophoros, cf. Fraser (1972: I 35, 23738); Longega 1968: 106 n. 168. For her assimilation to
Isisalready identied with Demeter by Herodotus (2.59)cf. Fraser 1972: I 23943. Calli-
machus portrays the queens deceased sister Philotera as associated in cult with Demeter of
Enna in the Apotheosis of Arsinoe (fr. 228.4345 Pf.). For the assimilation of Berenike II to this
goddess, cf. Pantos 1987. For an overview of Demeters association with Ptolemaic queens,
cf. Tondriau 1948.
41. For a review of the evidence, cf. Gutzwiller 1992a: 36367.
42. The dating is reviewed by Bulloch (1985a: 3842).
readers would have been male.43 In that case, the readers implication in the
role of witness to a rite where only women are allowed acquires a different,
somewhat transgressive avor. But what are we to make of it? And how would
it play? Would the experience have been unsettling, exciting, transforming?
In this connection, it is worth recalling a legend linked to the celebra-
tion of the Thesmophoria in Callimachus hometown, Cyrene. In a frag-
ment of Aelian (fr. 44 Hercher), we learn that the citys founder, Battus
an ancestor of Callimachus, as the poet proudly mentions on several
occasionslonged to experience the mysteries of Demeter Thesmopho-
ros and used force to do so, exulting with lewd eyes ( C
6 ,
4 ; cf. also Suda s.v. Thesmophoros). The
priestesses tried to dissuade him, but Battus was unrelenting (
6 " C. +
. . .). He did not escape without a grisly punishment, however:
for the sacricial priestesses (), swords drawn and hands and faces
smeared with sacricial blood, fell on Battus at a prearranged signal and
emasculated him ( : # ,
( G (V + 6
), #/ 2 6 JV , G
" ( , ; cf. also Suda s.v. ).44
With a story like this in ones family background, it is no light thing to
impersonate a woman at the Thesmophoria! Assuming for a moment that
the legend does in fact antedate Callimachus, the question of how to inter-
pret the speaking voice in his poem becomes even more urgent. I should
stress that nothing suggests a prurient interest. The speaker does not ob-
serve the scene with lewd eyes ( 4) like Battus, nor does
the text encourage us to do so. This is playing the other with a differ-
ence: not like the travesty of comic male interlopers in Aristophanes Thes-
mophoriazousai, where the (male) theater audience plays voyeur with
ironic distance; nor like the horror of watching Pentheus clothe himself as
a woman to observe maenadic rites in the Bacchae.45 Callimachus impli-
cates his readers in the dramatic situation in a more afrming, partisan
way. But to what end?
43. Male literacy was still preponderant over female in Hellenistic times, even among
upper classes, though not to the same degree as before. Cf. Harris 1989: 13637, 13946.
44. For the legend, cf. Burkert 1985: 244. It is adduced in connection with Callimachus
by C. W. Mller (1987: 8788), who, however, interprets it as a parallel for Erysichthons
transgression and punishment and for the distinction at the Cyrenean Thesmophoria be-
tween initiates and uninitiated, rather than as a potential model for the poets impersonation
of the female voice in the hymn.
45. Cf. Zeitlin 1990.
Mller has arguedpersuasively, in my viewthat the myth in Calli-

machus hymn is a narrative metaphor, embodying the programmatic op-
positions found in his Aetia prologue and Hymn to Apollo.46 In both of these,
as Mller stresses, Demeter occupies a special place. She appears in the for-
mer as = (bountiful lawgiver), with the elegiac Demeter
by the Coan Philitas hailed by Callimachus as a positive model (fr. 1.10
Pf.).47 In the programmatic language of the Hymn to Apollo, the long and
lthy Euphrates is opposed to the pure droplets carried by , bees,
to Demeter from a sacred spring (Hymn to Apollo vv. 11012).
To identify with the role of a devotee of Demeter is thus, in Calli-
machean terms, to be aligned with a particular literary stance. To identify,
further, with an exclusive groupas in our hymnis typical of that stance.
Callimachus art is not one of heroic isolationdespite the injunction not
to drive his chariot in the same tracks as others (fr. 1.2627). His milieu is
the group. We sing, he says in the Aetia prologue, among those who
love the shrill voice of the cicada (vv. 2930). That elite, charmed circle of
insiders belongs to his self-denition.48
The group is where the hymn begins. Rather than invoke the goddess
name and titles at the very start, as normally in Homeric hymns and as in
verse 1 of the Homeric Hymn to Demeter ( / V
/ , Demeter of the lovely hair, the awesome goddess, I start to
sing), Callimachus waits until verse 2 (, ,
, Demeter, all hail, who nurture many, who bring many
bushels). Before that invocationin deliberate and pointed contrast to the
Homeric hymnhe interposes the group ( 6-
, D as the basket returns, cry out, women, v. 1).49 And, in
46. There is, on the one hand, Erysichthon, rash and murderous, his actions conspicu-
ously portrayed in language evoking martial epic (cf. Hopkinson 1984: 67), the genre Calli-
machus eschewed (Aetia fr. 1.35). Vast quantities of food and wine ow down his gullet as
though into the depths of the sea ( / 6 , v. 89). Recall how, in the Hymn
to Apollo, Envy condemns the poet who does not sing a song big as the sea (T "/ @
, v. 106). Yet Erysichthon still wants more. Even the lthy garbage thrown out from the
feast (( , v. 115) is meet for his indiscriminate maw (cf. the car-
ried along by the Assyrian stream in Hymn to Apollo v. 109). On the other hand, there is the ex-
clusive circle of Demeters devotees, untouched by impurity, restrained in their appetites.
47. For the inuence of Philitas in Callimachus hymn, see Heyworth 2004: 14653.
48. In the Hymn to Apollo, the cultic community again coincides with the community of
readers. See, in the present volume, Impersonation of Voice in Callimachus Hymn to Apollo.
49. The address to the group, rather than the deity, is also observed by Fantuzzi (1993:
933). Contra Fantuzzi, however, that address does not just replace the naming of Demeter.
Her name is still there; still holds its position at verse start (v. 2); still leads off an utterance.
Callimachus thus pointedly recalls standard hymnic invocation precisely so as to stress its
displacement through the address to the group. The pointedness of this shift becomes clear
when we contrast other poems in which a group is ordered to celebrate the deity but in
distinction to the archaic hymn, it is in the framework of this privileged

groupwith whom (more than in any other of Callimachus hymns) the
speaker conspicuously identies by her recurrent use of the rst-person
pluralthat the experience of the transcendent occurs.
The facts that the groups exclusivity is here dened largely by gender
and that readers are invited to play a womans part introduce an element
different from the usual play of insider/outsider. But what does it mean? Is
the female circle just another way for Callimachus to say my crowd? Is
it the semantic equivalent of the literary elite of the Aetia prologue? Or
does the play of gender signify? That play would be the more emphatic if
we assume, as seems plausible, that the six hymns were set out in their pres-
ent order by Callimachus himself.50 For if we read them in sequence, the
setting shifts from the predominantly male symposium at the start (
J, Hymn to Zeus v. 1) to the exclusively female festival at the close.
It is important to note that the gender contrast framing the hymns is not
just thematic (male deity vs. female deity), as Hopkinson seems to suggest.51
Rather, it involves the reader in a shift in sensibility, as he responds imagi-
natively to the shift from a masculine to a feminine setting. Do readers
who adopt the female vision here, then, participate in an aesthetic experi-
ence that sets them apart from the established (more masculine?) aesthetic
norm?52 Does this experience, exceptional and exclusive as it is, reect the
otherness to which Callimachus aspired in his poetic program?
which the deitys name still comes rst in the poem. Cf., e.g., the Eritrean Paean (p. 136
Powell): [ ] / [ K]; Epidaurian Paean of
Isyllus (p. 133 Powell): ,+ , / 6[] / /;
the Paean of Macedonius (p. 138 Powell): "[ 2] _ M /
"[, , , ,, g ,+ ,] / [] 6 [, ] []
( / /[, , ,, g ,+ .
50. Hopkinson 1984: 13; Hunter and Fuhrer 2002; Depew 2004.
51. Hopkinson 1984: 13.
52. This is the sort of experience that Skinner (1993: 13638) posits for the male recep-
tion of a genuine womans voice, such as Sapphos: [A]llusion to Sappho became an obvious
tactic for projecting metaphoric difference upon one of two antithetical male-structured
categories. . . . Yet the Sapphic texts still stayed in play as a locus of real differentiation. . . .
There is a dash of actual female subjectivity even in Diotima. . . . Platos audience would have
obtained its artistic impressions of female homoeroticism chiey from the poetry of Sappho.
Since, as we have seen, Callimachus feminizes his narrating persona in part by alluding to the
representation of female speech in male poets, such as Euripides and Theocritus (cf. the sim-
iles discussed early in this essay), it appears that his feminine characterization has already
passed through a masculine lter. Is it thereby closer to that essentially male construction of
the feminine, in which woman . . . turns out to be a trope, a gure of male speech, as
Halperin (1990: 15051) has argued?
chapter 4
Reconstructing Berenikes Lock

I n his Miscellanea (LXVIII and LXIX) of 1489, Angelo Poliziano, the rst
modern scholar to assemble fragments of Callimachus Lock of Berenike,
remarked on the extraordinary elegance with which Catullus translated
this poem. Still, he deplored corruptions introduced by ignorant scribes,1
and drew attention to two in particular 2 for which he proposed readings
that had one thing in common: they presumed that Catullus had adopted
the Greek form of a proper name in his translation.3
In the rst instance, the locksevered from its queen by an iron
bladecalls down a curse upon a certain ironworking people (66.48). Ac-
cording to Poliziano, some read Iuppiter, ut Telorum omne genus pereat,
while the vulgatissimi codices read celitum (sc., caelitum)neither of
which produces good sense in the context. For his part, Poliziano adduced
a quote from the Callimachean original, which he had found preserved in
the scholia to Apollonius Rhodius 2.375: X E . . .
In the light of this evidence, he argued, who could doubt that Catullus
verse must read Iuppiter, ut Chalybon omne genus pereat, and that Chalybon
This is an updated version of an article that appeared in Collecting Fragments, ed. G. W. Most,
7894. Aporematal. Gttingen. 1997. Vandenhoeck und Ruprecht. My heartfelt thanks to
Prof. Garth Tissol of Emory University for his helpful comments on an earlier draft of this
essay.
1. Elegiam Callimachi de crinibus Berenices inter sidera receptis mira elegantia vertit
in latinam linguam, nobilis poeta Catullus . . . pleraque sint in ea corrupta, mendosaque, et
temere scripta librariorum inscitia.
2. Cf. the excellent discussion of these two passages in Grafton 1983: 16972.
3. Cf. Grafton 1983: 33.
65
is rendered with the Greek form? 4 To be sure, a few scholars have argued
for the Latin genitive plural Chalybum, deciding they can live with the un-
usual hiatus that results.5 But on the whole, Chalybon has been accepted
and is now the most commonly printed text.6
In the second instance, Poliziano claimed that he was taking a stand
with both feet (as they say) against the perverse boldness of the ignorant.
His indignation was aimed at a suggestion made a few years earlier (in
1485) by Parthenius concerning the poems very last word, which names
the constellation Orion. The transmitted text has Oarion, the uncontracted
form of the name in the Greek, which probably bafed Parthenius because
it is not otherwise known in Latin. The standard spelling of Orion did not
ll out the end of the pentameter. Hence Partheniusthinking of the
Latin epithet for Orion, ensifer (or ensiger), sword-bearerproposed
Aorion. This unattested name he derived from the Greek :
enim signicat ensem: inde Aorion, quasi ensifer. Against this ingenious but
perverse coinage, Poliziano argued, Callimachi eiusdem autoritate, that
we must retain Oarion, since it belongs to Callimachus own usage (Hymn
to Artemis v. 265; cf. also Nicander Ther. 15; Pindar Isthm. 3.67). By thus
defending the transmitted text of Catullus, Poliziano suggests what in all
likelihood stood in the Greek.
Implicit in both of Polizianos readings is an assumption that Catullus
translation hews very close to its sourceso close, indeed, that he adopts
the Greek forms of proper names, notwithstanding that these do not
appear elsewhere in Latin. In the case of the Chalybes, moreover, the fact
that they stand at precisely the same metrical position as in Callimachus
makes the proposed reading all the more convincing.
The Renaissance scholar is the rst in a long line to suggest that Catul-
lus translated his model with exceptional delity: the translators precision
is assumed. Polizianos lead in this respect was to have enormous inuence.
Indeed, his brief observations would to a large extent mold critical percep-
tion of Catullus poem vis--vis Callimachus ever after. Still, there are
those who have taken a different stand. Rather than highlighting the accu-
racy of the rendering, they draw attention to the incongruities between the
4. Ex quo versiculum sic illum legendum quis dubitet? Iuppiter ut Chalybon omne
genus pereat, ut sit Chalybon graece dictum.
5. E.g., Kroll 1968: ad loc.
6. Poliziano was apparently quite proud of his suggestion and wanted to make sure that
he (rather than Pico della Mirandola, who had been spreading it around) would get credit
for it: Sed hanc scio nostram observationem iam pridem esse pervulgatam, quam tamen a
nobis ortam, vel ille ipse scit, qui vulgavit, libenterque etiam fatetur vir doctissimus unde-
cumque Picus.
Reconstructing Berenikes Lock 67
Greek and Latin, even insisting on the Roman poets extreme unreliability
as a guide to the Greek. These opposing tendencies have acquired sub-
stance and become more pronounced as fragments of Callimachus
poemand information about its context in his workhave come to light
in papyri since the 1920s.7 Whichever view critics adopt, it shapes their
treatment of either poem: whether for similarity or contrast, each text is
used to elucidate the other. Not surprisingly, however, it is upon the frag-
mentary Greek that assumptions about the relative accuracy of the transla-
tion have more profoundly left their markespecially on the constitution
of its text. The presence of the Latin translation denitely alters our
handling of the fragments, adding an interesting twist to the usual prob-
lems we confront in grappling with a fragmentary text.
Typically, fragments of ancient works involve questions of reconstruc-
tion, of reconstituting the whole.8 How, for instance, do we construe an
indistinct letter; what weight do we assign marginalia in so doing? How
should we supplement lacunose verses and, on a larger scale, imagine what
lled the gaps between fragments? In pursuing such questions, we respond
to the challenge of the empty space, at once daunting as an image of loss
inicted by time and beguiling in the scope for play it offers the intellect
earnest play, to be sure, since it also holds out the prospect of restoring a
text to its pristine state, even if such an outcome is rare. To take up the
challenge is perhaps to become an articer of fact, as Herbert Youtie de-
scribed the papyrologist9a tempting prospect, which has turned the head
of many a textual critic. Normally, of course, we limit our play of intellect
with the rigorous checks and constraints of scholarly method: all proposals
must be met with caution and a critical eye, for no one dealing with a dead
language can hope to acquire the cultural/linguistic competence that na-
tive speakers enjoy as their birthright. Whether the challenge of the empty
space makes us cautious or bold determines to some extent which of the
7. These are PSI 1092, edited by Vitelli (1929), which contains twenty verses from the
middle of the poem; the so-called Milan , edited by Norsa and Vitelli (1934), which
cites the previously unknown rst verse as the nal lemma before the subscriptio of the fourth
book of the Aetia; and nally P. Oxy. 2258, edited by Lobel (1952), which Pfeiffer had in-
cluded in his edition as P. Oxy. ined. C (cf. also the addenda in Pf. II 11416): this contained
fragments of the same verses as PSI 1092 but also some from an additional twenty verses later
in the poem.
8. In this respect, they differ fundamentally from that sense of fragment found in some
modern literature (e.g., that of Samuel Beckett), whose whole tendency is toward a nonexis-
tence that precludes reconstitution. Cf. Strauss 1995. In ancient literature, we nd deliberate
use of fragmentary contexts as a spur to the imagination and to supplementation. See
Ergnzungsspiel in the Epigrams of Callimachus in the present volume.
9. Youtie 1963 ( 1973).
opposing critical tendencies outlined in the previous paragraph we incline

to. Still, in the case of the Lock of Berenike, the scales are denitely tipped
toward more liberal supplementation. And the reason is that, ever since
Poliziano, scholars have favored the view that we possess an accurate guide
to the missing Greek: the translation of Catullus.
Critics could of course invoke Catullus himself to justify this view. For
the poet seems to arouse an expectation of delity when, in the Lock of
Berenices accompanying epistle to Hortalus (poem 65), he describes him-
self as incapable of creative work due to his grief at his brothers death (nec
potis est dulcis Musarum expromere fetus / mens animi, vv. 34) but willing to
send Hortalus expressa . . . carmina Battiadae, verses of Callimachus trans-
lated (v. 16). The term for translation here, exprimere, originally [meant]
to take the impression of a seal, [but] comes increasingly to be used of lit-
eral translation, often with verbum de (or e) verbo added.10 As such we
must distinguish it from vertere, which signies a far freer rendering.
It is not surprising, then, that long before the papyri appeared to pro-
vide some measure of corroboration, there were many who championed
Catullus accuracy. In this essay, I want to consider some of those who did
and look at how that position affected the way they treat the fragmentary
Greek.11 I will begin with the centuries immediately following Poliziano
and lead up to the work of Rudolf Pfeiffer, preeminent scholar of Calli-
machus in our own time and exemplary editor of fragments. From the very
start, these champions of accuracy spared no toil in trying to alter the few
fragments of the Greek preserved in citations so that it would conform
better to the Latin. What is striking about their attempts is that not one has
proved correct in the light of subsequent discoveries.
The assumption of delity became an almost irresistible spur to conjec-
ture. Certainly it inspired Muret, in his 1558 edition of Catullus, to emend
and supplement the verse about the Chalybes (v. 48). His emendation has a
brilliance and seeming necessity, which made it extremely popular among
later scholars. Accepting Polizianos Chalybon, Muret read the Latin as fol-
lows: Iuppiter, ut Chalybon omne genus pereat. It only required a small change
to bring the Greek into almost perfect alignment with the Latin. Where
the transmitted text read X E , Muret suggested E
X . Omne now had an analogue in . And
even the conjunction ut stood revealed as corresponding precisely to the
10. Wormell 1966: 198. Cf. also Fordyce 1961: ad 65.16; J. Reid 1885: ad 2.77. The con-
notation literal translation is called into doubt by Warden (2006: 98100).
11. In these considerations, I have been greatly helped by Marinones collection of the
evidence (1984).
deployment of E in the newly improved original. Genus, of course, was al-

ready perfectly matched with . All that was missing now was Jupiter,
and he was supplied with the etymologically identical Z so as to
produce the convincing pentameter Z , E X
. Murets conjecture presumes that, as in the case of the Chalybes,
Catullus sought to reproduce exactly the sense, the essential form, and even
the metrical position of the putative Z . Of course, Muret could
have no inkling that the papyri would show that nothing in Callimachus
text corresponds to this oath to Jupiter.12 That, as it turns out, was Catullus
innovation, a note of mock solemnity in a distinctly Roman key.13
The quotation in the scholia to Apollonius concerning the Chalybes
does not end with this verse. It extends to the next one, and they together
read as follows: X E / ,
, G ( (may the whole race of the Chalybes perish, / who
uncovered it, rising from the earth, an evil ower). Even without further
context, scholars noted that in Callimachus the locks malediction con-
cerns only the Chalybes, who are seen not just as workers of iron but as the
discoverers of the art. In Catullus, these functions are split between the
Chalybes and an anonymous # : Iuppiter, ut Chalybon omne
genus pereat / et qui principio sub terra quaerere venas / institit ac ferri stringere
duritiem (Jupiter, may the whole race of the Chalybes perish, / and he who
rst began to look for veins under ground, / and to forge hard iron.).14
This incongruity nettled the champions of accuracy. To get rid of it,
Joseph Scaliger, for instance, who in 1562 had retranslated the whole of
Catullus Lock of Berenice into Greek as a present for Muret,15 proposed
in his 1577 edition of Catullus (75) that the Roman poet had read not
the masculine participle , whose antecedent we now know is
two verses earlier, but, rather, , the neuter participle
modifying . The transmitted text, ,
12. The beginning of the pentameter in Callimachus turned out to have the verb from
the temporal clause in the previous line: 8, @/ _ [M /
:; X E .
13. Thus Wormell 1966: 198.
14. Cf. Pfeiffer 1932 Skiadas 1975: 10052, here 115: Kallimachos unterscheidet sich
von allen andern, indem er ein ganz bestimmtes Volk nennt: die Chalyber, die als erste das
Eisenerz ans Licht brachten und es zu schmieden lehrten. Catull hat das nicht einfach ber-
setzt, sondern er kontaminiert sein Vorbild mit jenem allgemeinen, bei den Rmern beson-
ders beliebten Topos: / @ usw. ( pereat, qui primus . . .); so kommt er dazu
(nach Kallimachos), die Chalyber zu verwnschen und den, der nach Eisenerz gegraben und
es dann gehrtet hat. Die Chalyber sind also bei ihm einfach , als die sie seit
Aeschylus gelten . . . , aber sie sind nicht die ersten.
15. The text may be found in Scaliger 1864: 21417.
, G (, thus became
G (, and the locks ire was therewith aimed at not just one
but two targets, as in Catullusthat is, not only at the Chalybes but at
those who discovered iron, that evil ower [ ] rising out of
the earth.
A hundred years later, this emendation was rened by A. Riese.16 He
proposed , j ( ([may . . . the Chalybes
perish,] and those who previously uncovered the evil ower rising from
the earth), somewhat unnecessarily assuring that the temporal relation-
ship between the Chalybes and these putative discoverers of iron would
be clearbut perhaps he thought Catullus intended Klangspiel between
and qui principio. Interestingly, he comments that both poets
sehen in ihnen [sc., the Chalybes] nur erst Bearbeiter des Eisens und
knpfen mit et () zu gleicher Verwnschung dann die ungenannten
Entdecker des Eisens und seiner Bearbeitung an. Andernfalls htte Catull
ungenau bersetzt. Finally, in his commentary on Catullus of 1885,
E. Baehrens pressed the Greek into yet greater conformity with the Latin
by restoring a singular discoverer of iron, corresponding to Catullus
qui principio . . . institit: , @ (
([may . . . the Chalybes perish,] and he who uncovered it, an evil ower
rising from the earth).
The publication of PSI 1092, however, took the wind out of the sails of
these industrious correctors. For the papyrus conrmed the transmitted
text in no uncertain terms. Catullus once again stood revealed as translating
more freely than some had suspected. We do not know why he introduced
an anonymous discoverer of iron and of ironworking in addition to the
Chalybes. It was surely not, as Wendell Clausen supposed, because he was
simply unable to render Callimachus Greek in any way that would be in-
telligible in Latin (1970: 89). As Pfeiffer put it, [Es] bleibt sein Geheim-
nis (in Skiadas 1975: 115).
This corroboration of the tradition in the papyrus prompted Pfeiffer to
issue a stern warning about our understanding of the one text in relation to
the other: Den Catulltext gewaltsam in Einklang mit dem Original brin-
gen zu wollen, davor warnen die Differenzen (in Skiadas 1975: 115). But
despite this warning and his many acute observations detailing the Dif-
ferenzen, Pfeiffer has been by far the most forceful advocate of Catullus
accuracy in this century. This is how he characterizes the translation:
16. Riese 1881.

Der Eindruck, den das zunchst in die Augen springende enge Sich-
Anschmiegen an die Vorlage macht, lsst sich vielleicht am besten
lateinisch bzw. ciceronisch wiedergeben: Catullus convertit non ut poeta,
sed ut interpres, und zwar horazisch ergnzt, ut dus interpres. Wie
Catull sein exprimere der carmina Battiadae (c. 65, 16) gemeint habe,
darber konnte man frher im Zweifel sein: die 20 neuen Verse
zeigen, dass er nicht nur Vers fr Ver seinem Original folgte, sondern
das Verhltnis von Satz- und Versschluss und das Zusammenfallen von
Satzteilen und Versgliedern zu wahren suchte. Ja er lie zur
besonderen Markierung den Eigennamen ihren Platz innerhalb des
Verses, mitunter sogar anderen Einzelworten. (in Skiadas 1975: 135)
But while the famous statements of Cicero (Opt. Gen. 14) and Horace (Ars
Poetica 13334) condemn overly literal, word-for-word translation, Pfeiffer
argues that such delity was essential to Catullus aim, which was to achieve
nothing less than perfection (Vervollkommnung) of a technical sort.
Es ist ein leidenschaftliches Ringen um jene Meisterschaft der Form,

die sich die groen Hellenisten erworben hatten: etwas grundstzlich
anderes, als was die frheren rmischen Dichtergenerationen von
Livius Andronicus ber Ennius und Plautus zu Terenz erstrebt hatten.
(in Skiadas 1975: 142)
Thus, though he dutifully acknowledges subtle differences in the Latin

and condemns the excesses of earlier defenders of Catullan accuracy, Pfeif-
fer believes in a dus interpres. His condence in Catullus faithfulness to
the Greek may, however, have made him liable to attempt reconstructions
that are similar in principleif subtler in practiceto those undertaken
by his predecessors.
To take a minor example, in the rst line of PSI 1092, Pfeiffer was look-
ing for something that would correspond to verse 44 in Catullus: progenies
Thiae clara supervehitur ([that mountainsc., Athoswhich] the shining
son of Thia goes by). In the papyrus, he found traces at the end of the line
consistent with the reconstruction of the verb: #][] [ .17 For
progenies Thiae, however, he adduced a striking analogue in a previously
unnoticed gloss in the Suda (Y 308 Adler): D C C
D + C D E ; [Th. 379] ,
17. These were modestly conrmed in the corresponding line of P. Oxy. 2258: ] [.
( : The wind, Boreas.

means descendant. For as Hesiod says [Th. 379], the winds are descendants
of Thia). Pfeiffer realized that, in conformity with Heckers law, all such
anonymous citations in the Suda derive from the Hekale.18 He argued,
therefore, that Callimachus here quoted himself. Most strikingly, he de-
duced from the text of Catullusder gerade die Eigennamen mglichst an
ihrem originalen Platze lsst (in Skiadas 1975: 110)that the Greek poet
reversed his own word order. Pfeiffer assures us that die ganz schwachen
Spuren widersprechen, wie mir Vitelli versichert, nicht, der erste Buch-
stabe nach der Lcke war sogar mit Wahrscheinlichkeit ein .19 But in his
nal edition, after having examined the papyrus himself, Pfeiffer no longer
recorded a trace of that , but only of the following two letters: . I must
confess that in examining the photographs, I was unable to see even these.
Was Pfeiffer perhaps unduly swayed by his belief in Catullus desire to re-
tain proper names in their original position (cf. the Chalybes)? The pa-
pyrus grants us no certainty. But the fact that Catullus changes the location
of names in the Coma almost as often as he keeps it should at the very least
make us cautious in our assumptions.20
In this last instance, the assumption of delity concerned the position
of proper names. There is a comparable conviction underlying the treat-
ment of longer sections too, however: namely, that Catullus sought to
match his original line for line. Let us examine its consequences for Pfeif-
fer and others before him in their treatment of the second extended pas-
sage (besides the one about the Chalybes) preserved in an ancient source.
The lines, quoted in Schol. Arat. 146, correspond to those starting at verse
7 in Catullus.
( 6 V
(
T
[Konon saw me in the sky, Berenikes

lock, whom she dedicated to all the gods]
18. Pfeiffer locates it in his edition among the fragments of the Hekale as fr. 338. And it
is taken up in Hollis 1990: fr. 87.
19. Thus in 1932 Pfeiffer printed [] [ .
20. Catullus does not retain the position of the name with Conon or Berenice in vv. 78,
Medi in v. 46 (but note the play with per medium in the same position), or Arsinoe in v. 54
(though he did with Locridos). He substitutes the name Ariadne for at vv. 59
60 (a deliberate simplication, argues Wormell [1966: 197]) and leaves out Berenice in v.
62. The location of the deitys name in Callimachus v. 71 was surely different from that in the
Latin text.
Even more than in the case of the verses about the Chalybes, modern
scholars have shamelessly tampered with this couplet. The reason lies in
the character of Catullus rendering.
idem me ille Conon caelesti in lumine vidit

e Bereniceo vertice caesariem
fulgentem clare, quam multis illa dearum
levia protendens bracchia pollicita est.
[that same Conon saw me among the lights of heaven,
the lock from Berenices head,
shining brightly, whom she vowed to many of the goddesses,
stretching out her delicate arms.]
Elements of Callimachus text are spread over two couplets in Catullus

(multis . . . dearum corresponding to . . . , pollicita est seemingly
to (), while others do not appear in Callimachus at all (fulgentem clare
and levia protendens bracchia). The response of some nineteenth-century
critics was to assumeof all things!compression: two Callimachean
couplets had mistakenly been merged into one by the scholia. This, it
seemed, would explain the discrepancy between pollicita est and (,
which properly refer to different ritual moments (vow and dedication, re-
spectively): Callimachus missing vow had presumably been squeezed out
by the careless scholia.
In discussing this fragment, Otto Schneider, the author of the standard
edition of Callimachus fragments prior to Pfeiffers,21 made the interest-
ing argument that it was unlikely that Catullus would deviate from his
source so close to the poems start, for it was here that translators tended to
be most accurate.22 His remedy was to restore the couplet to its original
four verses, corresponding to those in Catullus. This he accomplished (1)
by arguing that the transmitted pentameter was a distortion of what had
actually been the second hexameter, reconguring it as such; and (2) by
adding two entirely new pentameters of his own composition. This is what
he proposed:
21. Schneider 1873. The couplet is fr. 34 in his edition.

22. Ibid., 14748: Tantum autem abest, ut in carminis initio, ubi qui convertunt accu-
ratissimi esse solent, Catullum censeam quidpiam dixisse a Callimacheis vel plane diversum
vel certe minus accurate eis respondens, ut totum hic se ad Callimachum accommodasse eum
existimem.
,,23 / ( 6 V
/ / "
, @ / / 6 9
" Z /
@/ . . .24
[scanned [themsc., the astronomical phenomena of the previous six
verses in Catullus], but Konon saw me in the sky, Berenikes / lock
shining ever so brightly from the heavens, whom she dedicated to all
the gods, / vowing with her delicate arm upstretched, at that time
when . . .]
Wilamowitz inveighed against the Misshandlungen Schneiders here,

with their arge Verste gegen Kallimachos Verskunst.25 And half a
century later, Pfeiffer lamented that Schneiders fatale Entstellungen
dieses Distichons werden, trotzdem seit Jahrzehnten durch E. Maa die
berlieferung der Aratscholien eindeutig feststeht . . . , immer wieder
abgedruct (in Skiadas 1975: 102 n. 9).
While not indulging in comparable reconstructive fantasies, Pfeiffer
nonetheless revealed a similar trust in Catullus delity. Convinced that the
poet strove to render Callimachus line for line (Vers fr Vers),26 he ex-
plained the discrepancies between the two versions by assuminglike
Schneiderthe existence of a second couplet. But unlike Schneiders, his
merely followed the one we have, which he had no doubt was intact. Pfeif-
fer concluded: man [wird] an dem vllig einwandfreien kallimacheischen
23. Schneider argued (ibid., 148) that this verb referred to all the astronomical phenom-
ena of the previous six verses in Catullus and corresponded to the Latin dispexit in v. 1.
Though it was only with the publication of the Milan in 1934 that Callimachus
rst verse came to light and, with it, the knowledge that he had included in it a verb of seeing
(,), Schneiders argument that Catullus shifted the verb six verses ahead seems strangely
at odds with his contention that translators are most accurate at the start of a poem.
24. Ellis (1878: 335) proposed something similar, though without composing new verses.
He cites vv. 710 of Catullus, unde Scholiasten credo fragmenta in unum congessisse quat-
tuor versuum. Scripserat Callimachus hoc modo
(1) J / ( 6 V
(2)
(3) @ /
(4) (
Nam post exciderunt nonnulla quae Catullianis istis Caelesti numine Fulgentem
clare (8, 9) responderent.
25. Wilamowitz 1879. Cf., e.g., the violation of Naekes Law in the second hexameter.
26. At Skiadas 1975: 135, we hear Pfeiffer argue that Catullus Vers fr Vers seinem
Original folgte. Cf. also 102.
Distichon. . . . nicht ndern, sondern annehmen, dass noch ein weiteres

Distichon dazugehrte und diese 4 Verse den 4 catullischen entsprachen.27
Pfeiffers assumption of line-for-line translation is innocuous enough
when applied to this couplet. But its consequences are greatly magnied else-
where. I want to draw attention to one example in particular, which occurs
where few would expect it and of which fewer yet are actively awaredespite
the fact that it gives that assumption its most vivid expression. I mean the ex-
traordinary layout of the poem in Pfeiffers edition (fr. 110), with Greek and
Latin texts carefully disposed on facing pages andespecially signicant
with corresponding numeration (see pp. 7677). Pfeiffer had matched Calli-
machus to Catullus in this way in his E of 1932
( Skiadas 1975: 1069)though less ostentatiously, for he printed only
those twenty lines of Latin that correspond to the Greek of PSI 1092.28 The
decision to set the whole of Catullus poem against the fragments of the Greek
(each with its own critical apparatus)29 for his 1949 edition is more radical, a
striking act of interpretation, since it leads the reader to expect a precise lin-
ear counterpart to the parallel poem across the page, even for sections where
not a shred of the original survives.30 That expectation is reinforced by the ti-
tles at the top of each page. On either side, Latin as well as Greek, they read
Aetia IV, Fr. 110 and, more remarkably still, include even verse numbers, so
that on page 115 (for example) the Latin is labeled Aetia IV, Fr. 1103350,
while over the lacunose Greek on the left-hand page we see Aetia IV, Fr.
1104050. In other words, Pfeiffers edition treats verses 3339 of Catullus
66 as though they actually are Aetia IV, Fr. 1103339, though those verses
are nowhere to be found in the Greek.
27. In Skiadas 1975: 104. In his remarkable reconstruction of the poem in Greek, Barber
(1936: 349) gives a good approximation of how the two couplets might have looked:
V ( 6 V
(
T
D #

28. He did, however, adopt matching numeration, implying that precisely forty-three
verses preceded the fragment in PSI, exactly as in Catullus. Cf. his treatment of P. Oxy. 1793,
discussed later in this essay.
29. There is a continuous dialogue between the poems in their respective apparatuses.
Cf., e.g., ad fr. 110.69 (suppl. L. ad Catull.); ad Catull. 66.77 (desideratur vox quae voci
correspondeat).
30. Perhaps the way had been prepared for this decision by Barbers translation of Catul-
lus into Greek (1936), which likewise sets the two texts opposite each other. But while that
layout is entirely proper in the context of an academic translation, which proclaims its clev-
erness through the comparison and freely admits its temerity in fathering on Callimachus
some sixty lines (345), it is another matter entirely in an edition that openly acknowledges
none of its presuppositions.
Fig. 1. R. Pfeiffer, Callimachus, vol. 1. Oxford: 1949, 11415
The layout reects just how tenaciously and with ever-growing assur-
ance Pfeiffer held to his belief in Catullus line-for-line delitydespite
the fact that it had previously led him badly astray, as he realized by 1949.
For that belief is already there in his earlier edition, Callimachi Fragmenta
nuper reperta (Bonn, 1923), where it colored his treatment of P. Oxy. 1793,
columns iiii (there fr. 60, pp. 9394 now frs. 38587 Pf.).31 In the third
of these badly tattered columns (fr. 387.2 Pf.) were found the words
, previously known from a quotation in the scholia
[ ]M
to Aratus (Achill. Isag. exc. c. 14, p. 41, 22 Maass, Comm. in Arat.). A. F.
Naeke had linked them to Catullus Lock of Berenice verses 8082: non
prius . . . quam . . . mihi (Callimachi Hecale [Bonn, 1845], 162), and many
subsequent scholars followed him in attributing them to the Callimachean
original. Consequently in The Oxyrhynchus Papyri vol. XV (London, 1922),
A. S. Hunt very tentatively proposed that P. Oxy. 1793, column iii, might
contain the nal part of Callimachus Lock of Berenike. In his edition of the
following year, Pfeiffer developed this proposal, and he did soit is well to
recallbefore the appearance of any of the major papyri (the rst three
columns of P. Oxy. 1793 contain only meager bits of nine lines). Thus his
conception, which is remarkably comprehensive, was still formed mainly
on the basis of the old citations and, of course, Catullus. He argued that if
the beginning of column iii corresponds approximately to verse 81 and if
we calculate that each column contained about twenty-seven verses, then
column i could have held verses 126; column ii, verses 2754. Conse-
quently we must reckon with one column missing before column iii. Fol-
lowing from this calculation, he even proposed supplements for the exigu-
ous scraps in each column, based on their imagined counterparts in
Catullus. Thus in column i, ] . . . (fr. 385.1 Pf.) became [\ 4],32
corresponding to rapidi solis in Catullus verse 3; in column ii, .[. . .] (fr.
386.1 Pf.) suggested on the model of regium at Catullus verse 27;
in column iii, ..[. . . .]n. .[.] produced [ ]33 to
match nudantes reiecta veste papillas at Catullus verse 81. As Pfeiffer says in
his E (Skiadas 1975: 102), die Voraussetzung
dieser etwas waghalsigen Rechnung war fr mich, dass Catull mglichst
Vers fr Vers seinem Original folgte. With the publication of PSI 1092 six
years later in 1929, his judgment seemed to be conrmed. But by 1949, he
31. Cols. ivv may, or may not, come from another poem. Cols. vix contain the Victory
of Sosibios (fr. 384 Pf.).
32. Pfeiffer simply ignored the papyrus clear circumex in .
33. This did violence to the reading in the papyrus, as Pfeiffer himself admitted in 1949.
Cf. the next footnote.
had seen that his waghalsige Rechnung had been just that. It could not
be right, and he recanted in a note on fragment 387.
and its alleged

Propter haec verba [sc., [ ]M
correspondance with non prius . . . quam mihi] Hunt, etsi valde
dubitanter, coniecit in hac col. iii nem Comae scriptum fuisse posse;
deinde nimis audacter e Catullo nonnulla suppl. Pf.1 in col. iiii (nunc
fr. 3857). Haec omnia falsa esse existimo; columnas 21 versuum
fuisse, non 27, ut computaveram, e fr. 384 satis certum est.34
In the light of this chastening experience, it is remarkable that Pfeiffer re-

mained so sure of his dus interpres. Indeed, as the layout of texts in his
1949 edition attests, his certainty had, if anything, grown stronger.
Pfeiffers treatment of P. Oxy. 1793 brings us back to a more general
point about fragments and the special implications of dealing with one
from a text that seems to survive in translation. Precisely because it is a
fragmentand a highly fragmentary one at thatP. Oxy. 1793 elicits an
urge to make it whole. That urge is something we observed already with
fragments preserved in quotations. But in the case of a tattered papyrus, it
is, if anything, heightened. The visibly frayed edges and missing parts be-
yond provide space for the imagination, challenging editors ingenuity and
historical mastery of the language, their ability to see the forest even in a
few broken twigs. Scholars often compare the challenge and attendant
pleasures of working on such fragments with that of doing jigsaw puzzles
or crosswords. As we have noted, of course, the urge to make whole is nor-
mally tempered by caution born of sheer ignorance about whatbeyond a
certain pointmight have stood in the gap. Such caution is reinforced by
rules we recognize as prudent and agree to abide by, for instance, Youties
Law: iuxta lacunam ne mutaveris. In the case of the Lock of Berenike, how-
ever, there is an added temptation to conjecture, to venture into the empty
34. Of his proposed supplements, only one is even referred to in his 1949 edition; cf. ad
fr. 387.1: quae olim supplevi (ad Cat. 66.81, v. supra fr. 110) in litterarum et accentuum ves-
tigia quadrare non videntur ( ]). His proposals did have consequences,
however: Barber (1936), working closely from Pfeiffers 1923 edition and 1932 article, at-
tempted to incorporate (v. 80) and the other two supplements mentioned
here into his translation of Catullus (i.e., \ 4 [cf. his n. ad vv. 34] and , v.
27). Recently Hollis (1922) has revived the idea that P. Oxy. 1793 does gure in Catullus
translation, arguing that col. iii through the beginning of col. vi (perhaps also the earlier
columns) could have come from an elegy celebrating the wedding of Berenike, upon which
Catullus drew for his vv. 7988, which are absent from the papyrus containing the latter part
of the Lock of Berenike. In other words, he would like to see those verses of Catullus as an
instance of contaminatio. For these verses, cf. the discussion following in text.
spaces beyond the text, since we seem to have a template, like the picture
on a jigsaw-puzzle box, to guide our steps. Catullus translation appears
(and had appeared to many ever since the Renaissance) to be just that. We
are faced with a text, then, that excites our imagination, as fragments often
do, while simultaneously disarming the vigilance that fragments always
require. Under these circumstances, three badly mutilated columns yield
up a whole poem in outline and even concrete detailsthough they turn
out to be a mirage.
I want to close by discussing one last instance of Pfeiffers belief in his
dus interpres that has become a well-known in Callimachean schol-
arship; no discussion of the issues would be adequate without some men-
tion of it. I am referring to the notorious problem that nothing in P. Oxy.
2258 corresponds to ten lines of Catullus (vv. 7988) containing a digres-
sion about a marriage rite. Complicating the picture, the papyrus adds two
verses (vv. 94ab Pf.) that come after that corresponding to the nal verse in
Catullus and that appear to contain a farewell to Arsinoe Philadelphus,
mother of the royal couple (cf. Pf. II 116 ad fr. 110.94a), and perhaps to
the king and queen as well (cf. Marinone 1984: 67). For Pfeiffer, assuming
the translators faithfulness to his original, there was little doubt about the
ultimate source of the ten Catullan lines missing from the Greek. In his
commentary on verses 7988, he states hunc ritum nuptialem . . . Catul-
lum de suo addidisse veri dissimillimum est. Instead, he proposes that Cal-
limachus published two versions of the Lock (ibid.).35 On this view, the
poem was originally composed for a specic occasion, that is, to celebrate
the catasterism of the lock dedicated by Queen Berenike on the successful
return from war of her husband Ptolemy Eurgetes in September 245 b.c.36
This is the version preserved in the Oxyrhynchus papyrus, where the Lock
appears together with another occasional poem, the Victory of Sosibios. The
second version was that which Callimachus revised for inclusion in the
Aetia, where, according to the Milan , it came as the nal poem of
book 4. In this setting, Callimachus could dispense with the additional lines
of farewell, since the epilogue, which closes the book, mentions the queen
in a reverential context (fr. 112.2 Pf.) and prays for the well-being of the
whole royal house (fr. 112.8 Pf.).37 According to the thesis followed by
35. For an excellent summary and evaluation of the thesis, cf. Eichgrn 1961: 5559.
36. On the date, cf. Koenen 1993: 90 with n. 151.
37. Thus Eichgrn 1961: 59. However, Cameron (1995: 15462) rightly cautions that
the epilogue is not mentioned in the Milan , and he even proposes that it is mistak-
enly placed before the subscriptio Aetia Bk IV in P. Oxy. 1011. In that case, he argues, it
could be unrelated to an edition of Aetia 4, serving, rather, as prologue to the Iambi or as the
Pfeiffer, then, it was this nal version from the Aetia on which Catullus
based his translation and that he faithfully followed.
We see that the assumption of verse-for-verse translation by Catullus
has momentous consequences. Together with the conicting evidence of-
fered by P. Oxy. 2258 and the Milan as to the context in which
the poem was published, it serves as one of the major props for the thesis
that there were two versions of the poem and that the second was specially
revised so as to crown the last book of Callimachus most important work.
It is possible, however, to construe the evidence in other ways.
Lobel drew a very different conclusion from Pfeiffer when he pub-
lished P. Oxy. 2258: Some fresh light is thrown by the new part of the
on the relation of Catullus translation to the origi-
nal. I should judge that it is now evident that it is impossible to depend on
the Latin, which too often . . . recedes far from the Greek (1952: 70).
Such skepticism about the faithfulness of the translation is shared by other
scholars as well, who have shed welcome light on the differences between
the Latin and the Greek.38 In particular, no one has been able to explain
why Callimachus would have wanted to add ten verses about the aition of a
nuptial riteeasily separable and . . . gladly to be dispensed with (Lobel
1952: 98)when the poems aition is clearly the creation of a new constel-
lation. As Stephanie West puts it, I cannot see that this . . . does much
credit to his judgement.39
In fact, several scholars have argued that these lines, like the other less
extensive changes in the poem, are Catullus addition.40 The diction closely
resembles that found elsewhere in Catullus (though it is hardly surprising
that he would have translated the original into his own poetic idiom). And
the themes of adultery and delity, separation and harmonious unity, are
crucial in the rest of his work.41 In this connection, Hutchinson does well to
remind us that the mixture of close translation and independent inser-
tion, such as one might presume here in Catullus, has good precedent in
original epilogue to Aetia 12, in which a future project, the Iambi, is announced. If we accept
Camerons arguments, the epilogue (fr. 112) would no longer explain Catullus omission of
the last two verses of the Lock of Berenike (vv. 94ab).
38. Cf., e.g., Putnam 1960; West 1985b; Hutchinson 1988: 32224; Warden (2006).
39. West 1985b: 65 n. 24. According to Hutchinson (1988: 323 n. 91), the disputed
lines destroy an elegant connection, and make the queens offering a strange appendage, the
nal wish an abrupt resumption. Cf. also Warden (2006: 13234).
40. This was rst proposed by F. Della Corte (1951: 33) and elaborated by Putnam
(1960) and Hutchinson (1988).
41. This argument cuts both ways, however, as Marinone rightly warns us (1984: 6162).
For Catullus might have found Callimachus poem appealing precisely because it contained
themes that matched his own concerns.
Roman literaturefor instance, in the Aratea of Germanicus or in Plautus

(1988: 323 n. 91). Indeed, poem 51 of Catullus, his translation of Sappho 31
(Lobel-Page), is a shining example. The latest variant on this approach is
that of A. S. Hollis (1992), who defends the Callimachean pedigree of the
verses but shifts responsibility for their awkwardness in this setting onto
Catullus by proposing that the Latin poet translated them from a different
elegy on Berenike (which he nds in frs. 38588 Pf.). We would thus have
an instance of contaminatioa fascinating possibility, though highly specu-
lative, as Hollis himself forthrightly admits (1988: 28). His admission is
one that, if we are to be similarly forthright, we must extend to all attempts
to explain the absence of these verses from Callimachus.
What, then, may we conclude? Because of its astonishing accuracy in
many places, Catullus translation seems to offer a reliable guide to Calli-
machus poem both as a whole and in particulars. Yet we have seen that
reconstructions based on its authority have repeatedly proven badly mis-
taken. Such a claim as Wormells (1966: 197) that the Greek prototype and
Latin version are so close that each can be used as a check on the textual
tradition of the other must be treated with utmost caution. We nd, in
fact, that any reconstruction of the Greek on the basis of Catullus is likely
to be wrong.42 For the appearance of reliability may be illusory, adding not-
so-subtle bias to our dealings with the Greek. This can make us less wary,
more apt to succumb to the lure of the empty spaceand so to conjec-
ture. Since the Catullan whole may not always reect the Callimachean
parts, the existence of the translation may thus actually heighten the dan-
gers that normally attend the interpretation of fragments. That possibility
looms even larger when we recall that, despite the self-characterization as
expressa . . . carmina Battiadae, this is verse translation, composed by a
major poet in his own right, whose aims will not always have been the
same as those of his model. That is something the meticulous Pfeiffer was
perhaps not always ready to admit but of which we must constantly re-
mind ourselves, verse for verse and word for word, in interpreting the
fragmentary Greek.
42. Conversely, emendations of the Latin from Callimachus stand a slightly better
chance of hitting the mark. Certainly Polizianos Chalybon was right on target. But such a
clear instance is exceptional. Perhaps another is Lobels correction of Catullus v. 78, milia
multa bibi, to vilia multa bibi on the basis of Callimachus vv. 7778: / .
Milia has, however, found its defenders; cf. Putnam 1960: 88; Warden 2006: 115 n. 74.
chapter 5
Ergnzungsspiel in the
Epigrams of Callimachus

In memoriam Thomas G. McKeen
I n one of the Marvelous Things Heard by pseudo-Aristotle (Mirab.

auscult. 131), we learn of the following very dramatic incident: the Athe-
nians are in the process of building the shrine of Demeter at Eleusis when,
all at once, they make an exciting and mysterious discovery. Something has
been brought to light, closed in among the rocks. It is a stelebut a most
unusual one, made of bronze. This immediately suggests an artifact of
great antiquity, perhaps from the Heroic Age. The stele, moreover, is in-
scribed with an epigram, in the original sense of the wordthat is, with a
metrical inscription. Its text reads as follows: , This is
the tomb of Deiope. The question arises as to the identity of this Deiope.
Some argue that it is most likely the wife of the legendary singer Musaios;
others, however, claim that it must be the mother of Triptolemos.1
This thrilling tale offers us a clear illustration of how epigrams were
This is an updated version of an article that appeared in A&A 41 (1995) 11531, 1995 de
Gruyter. Earlier versions of this essay were delivered as talks at the Universities of Tbingen,
Cambridge, London, and Heidelberg. I gratefully acknowledge the criticism and suggestions
of Profs. R. Kannicht, E. A. Schmidt, G. W. Most, and Drs. H. Krasser, E. Krummen, K.-H.
Stanzel, R. L. Hunter, N. Hopkinson, and Peter Allen Hansen.
1. , / 6 /
# , 6/ ? 6D . ! +
, + + .
85
originally tied to a particular location. Inscriptions generally refer to spe-

cic objects or monuments in a given place, where people set them up at a
given historical moment, and for a certain purpose. Further, since such in-
scriptions are usually public, we may assume that they belong to the con-
ceptual framework of their given community, share its assumptions and
traditions, and may be readily understood by its members.
The stele of Deiope, by contrastinasmuch as it is buried amid the
rockshas apparently been dislodged from its intended place. More im-
portant, people evidently no longer know who Deiope is. This prompts
them to speculate: assuming that the tomb was not originally here, but
somewhere else (they might say to themselves), it was still probably linked
with Eleusis; and since its bronze material suggests great age, Deiope
probably belongs to the mythical prehistory of the site. In other words, the
readers sift through the clues they nd in the inscription and ll these out
with elements of their own knowledge, so as to form a plausible whole.
This activity might be called a process of supplementation. In Hellenis-
tic times, as epigram became increasingly literary, that process underwent
a shift. No doubt the epigram retained most conventions of votive and
sepulchral inscription, including that extraordinary concision that, as we
shall see, became ever more expressive. But while the epitaph of Deiope
belongs to a monument whose lack of context was accidental, Hellenistic
epigram was often deliberately severed from its object or monument and set
in the as-yet-uncharted landscape of the book. Here, poets came to exploit
and play with this process of supplementation in a deliberate and artful way.
Indeed, it became a favored and self-conscious device. As such, I would like
to call it Ergnzungsspiel.2
Before examining this development, I would like to consider how it
might have come about. In what circumstances would readers have experi-
enced the aesthetic appeal and poetic possibilities of an epigram thus de-
tached from its pragmatic context? If we had any certainty at all that inscrip-
tional epigrams were transcribed early on and available in collections, we
would have a plausible source for such an experience. Scholars have spilled
endless quantities of ink arguing the early existence of epigram books,3 es-
2. I have opted for this term rather than Ingardens conceptually related concretization
(1973: 5063 and passim) because the latter omits that self-conscious manipulation of and
(above all) play with supplementation that is crucial both to creation and reception of many
epigrams in the Hellenistic period. Unfortunately, I could not nd a comparably pithy term
in English to convey both this playfulness and the endeavor to make a thing whole ( ganz).
3. Reitzenstein (1893: 115), for instance, weighs the possibility of such a collection in the
second half of the fth century, within his whole discussion of book epigram and epigram col-
lections (10420). Cf. further L. Weber 1917: esp. 540 n. 1; Geffcken 1918; Beckby 1957: 68;
Sider 2007; A. Petrovic 2007b.
Ergnzungsspiel in the Epigrams of Callimachus 87
pecially a sylloge Simonidea.4 But beguiling as the thought of such early col-
lections is, there is no hard evidence that they existed before the end of the
fourth century, that is, before the start of the Hellenistic age. It is only for
this period that sources start to mention such gures as Philochoros, who
collected inscribed epigrams (Suda s.v. Philochoros F 441 A d1 FGrHist
328 T 1), orsomewhat later, at the transition from the third to the second
centurythe periegete Polemon of Ilion, who traveled through the Greek

cities transcribing inscriptions so as to write his book
6 (Athen. 10.442e).5
Perhaps a more productive line of inquiry will emerge from the follow-
ing question: how do authors of the fth and fourth century quote inscribed
epigrams?6 For the fth century, we can look to the numerous examples
from Herodotus and Thucydides. Herodotus 4.88, which concerns the
bridging of the Bosporus, is a good example.7
+ ? J J " -
6 . / E

, M

6 J "
, 6 K, 6
D
,
KJ ,
+ , +
#M ,
6 .
4. Cf., e.g., Gentili 1968: 4142. But cf. the welcome skepticism of Page (1981: 12023,
20710), who concludes, there is no evidence that any particular authors epigrams were
collected and published before the Hellenistic period; and Simonides is no exception to the
rule (120).
5. Cf. Deichgrber 1952: 1314.
6. The examples are conveniently collected in Preger 1891 and are here preceded by
Pregers bracketed numbers. For the fth century, see [109] Hdt. 4.88.2 AP 6.341; [79]
Hdt. 5.59 AP 6.6; [80] Hdt. 5.60 AP 6.7 8; [72] Hdt. 5.77.2 AP 6.343; [20] Hdt.
7.228 AP 7.677; [21] Hdt. 7.228 Lycurg. In Leocr. 109 AP 7.249; [200] Hdt. 7.228
AP 7.248; [84] Thuc. 1.132 Demosth. In Neaeram 97 AP 6.197; [71] Thuc. 6.54; [31]
Thuc. 6.59 Aristot. Rhet. 1.9.1367b. For the fourth century, see [233] Plato Phdr. 264c
AP 7.153; [208] Plato Charm. 165a; [197] Plato Hipparch. 228d; [144] Aristot. Rhet. 1.7.1365a
(cf. 1.9.1367b); [209] Aristot. Eth. Nic. 1.8.14.1099a27 Eth. Eud. A1.1214a5 Thgn. 255
56; [249] Aristot. fr. 565 R; [19] Aristot. fr. 565 R AP 7.54; [74] Aristot. De Ath. rep. 7; [17]
[Aristot.] Mirab. auscult. 131 843b3; [66] [Aristot.] Mirab. auscult. 58 (59); [95] [Aristot.]
Mirab. auscult. 133; [153] Aeschin. 3.184; [154] Aeschin. 3.187; [271] Demosth. De cor.
18.289; [99] ps.-Demosth.7.39 (prob. 3rd cent.) AP 9.786; [199] Lycurg. In Leocr. 109; [68]
Theopompos FGrH II 115 fr. 285; [73] Philochoros Anon. 102 FGE p. 406ff. For a recent
survey of many of these inscriptions, see A. Petrovic 2007a.
7. On this and other inscriptions in Herodotus, cf. West 1985a: esp. 28283.
[Dareios was pleased with the pontoon and gave the builder Mandrokles
a tenfold gift. From this, as an offering of rstfruits, Mandrokles com-
missioned a painting, in which the bridging of the Bosporus was shown.
King Dareios is seen sitting on a seat of honor, while his army makes its
way across. He dedicated the painting in the temple of Hera and wrote
on it the following inscription:
Mandrokles, who bridged the shy Bosporus,
dedicated to Hera a commemoration of his pontoon.
For himself he won a crown, for the Samians glory,
since he nished the task as King Dareios wished.]
What is typical here, both for Herodotus and Thucydides, is the attempt
to construct a context for the inscription. This allows us to experience it
almost as though we were there: historical background is provided; the
temple of Hera (apparently in Samos) is mentioned as the site of the dedi-
cation; the painting to which the inscription belongs is described in detail.
It is striking that the epigram itself, however, gives no hint as to what the
of verse 2 is. That is, there is no mention of a picture, let
alone details of the armys crossing or where the king is sitting, for any
viewer would see that immediately, picture and text form an indivisible
unity.8 Herodotus enables us to experience that unity.
The situation is quite different when we look at how, in the fourth cen-
tury, Plato and Aristotle quote inscriptional epigrams. Let us take as an ex-
ample Platos quotation of the Midas epigram in the Phaedrus (264c). Here
Socrates faults the speech of Lysias, which Phaedrus had just read him
with great enthusiasm, because of its arbitrary construction.
2 , . . . #

6 "+ , T Y M
6 . . .
X ,, / 6 .
=/ [ a J J,
" J 6 ,
@ J .

@ / "+ " X a , 6
.
, E 6M
8. . . . das sieht jeder Beschauer sofort. . . . Bild und Spruch stellen also eine un-
trennbare Einheit dar (Geffcken 1918: 99 with n. 4).
Look at the speech of your friend, . . . and you will nd that it is no dif-
ferent from that epigram that some say was written for the Phrygian
Midas . . .
I am a bronze maiden, and I lie on the tomb of Midas;
so long as water runs and tall trees put out leaves,
not moving from this spot, atop the much lamented grave,
I shall declare to those who pass that Midas is buried here.
You see, I imagine, that it makes no difference whether any verse of it is
spoken rst or last.]
The fact that Socrates tells us that some say ( ) that this epi-
gram was written for the Phrygian Midas suggests that the poem was cir-
culating freelywhether orally or in writingdetached from its site and
monument.9 Now if we focus purely on this epigram, without taking ac-
count of Socrates intent in citing it, and if we try to picture what it refers
to, we step into a void. The repeated deictic pronouns (" J, v. 3;
J, v. 4) have no point of reference. Where are we to imagine this
tomb?10 Platos text does not give us the means to decide. Of course, this
did not stop Leo Weber (1917: 538), in an oft-cited essay, from conjuring
up a detailed image of the tombs immediate environment as a rustic idyll.
The epigram is referring not just to any old trees and streams, he says,
but to those that adorn the area around the tomb, and protect it; trees
nourished by the water of a nearby spring. Should it ever dry up and grow
parched, the tomb itself would be left to fall into ruin.11 Even if we cannot
share this fantasyfor natures persistence, as described in the poem (as
long as water runs and tall trees put out leaves), is a topos, its reference
not concreteWebers impulse to deal seriously with the hints that the
poem gives and to supply the missing information is right on the mark.
9. If Simonides (PMG 581) refers directly to the Midas epigram, we would have to
imagine the poems free circulation at a much earlier date. That is not unlikely in the case of
individual metrical inscriptions, especially if they were linked with particularly famous gures
(Simonides connects the poem with Kleoboulos of Lindos, one of the Seven Sages). For a de-
tailed discussion of the Midas epigram and related problems, cf. Markwald 1986: 3483.
10. The ps.-Herodotean Life of Homer (5.24) evidently locates the tomb in Ionian Kyme
6
and says that it is still there ( 6 ( . . .
6). Dio Chrysostom, however, reports that he searched for the tomb and was un-
,
able to nd it (37 p. 120 R vol. II p. 304 Dind.): /, b + ,
+ + " a "+ .
11. Nicht beliebige Bume und Gewsser meint das Epigramm . . . sondern die Bume,
die um das Grabmal zum Schmuck und Schutze herumstehen und von dem Wasser einer
nahen Quelle getrnkt werden. Ihr Versiegen und Verdorren wrde auch das Grabmal selbst
der Zerstrung preisgeben.
How, then, are we to imagine the who lies on the

tomb? Could she be a Siren, a Sphinx, or even a death-bringing Ker? All
three suggestions have been made in modern discussions of the poem,12
and all are within the bounds of ancient tomb iconography. Each of these
mythic gures may be described as .
The kind of citation that we nd in Plato opens the door to experienc-
ing epigram in a manner quite different from an actual encounter with an
inscribed monument or even from the mediated encounter we had in
Herodotus. As quoted here out of context, its effect on us is altered. Its
concisionthat traditional marker of the genreacquires a wholly differ-
ent force: it virtually demands that we act, that we use our heads to supply
what is missing. And if we are willing to do this, we may just nd that the
action is pleasurable. We may, in other words, sense the potential appeal if
an epigram were to incite us to such a process deliberately.
Plato was, of course, not primarily (if at all) interested in offering us such
an experience. Socrates use of the quote aims in another direction, and
there is nothing to suggest that Phaedrus pondered how to ll out the de-
tails that the epigram evokes. It seems important to me, therefore, that we
the readers do not experience the poem (as Phaedrus does) while engaged in
a dialoguethat is, that we do not experience it orally, in circumstances in
which we could easily be distracted, but, rather, that the text is xed on the
page (or better, the scroll), where we can imagine it undisturbed and (if we
are so inclined) reect on it. The epigram fully reveals the aesthetic poten-
tial of its dislocation only when it is xed in writing (and that need not be in
a collection; quotations sufce): only then is it likely to prompt one to con-
sider what it would be like if one were to sever an epigram from its setting
deliberately and so spark the process of supplementation. To the extent that
this process came to be exploited regularly and self-consciously in the ever
more literary epigram of the Hellenistic age, I call it Ergnzungsspiel.13
12. For an overview, cf. Markwald 1986: 7980 n. 113.

13. The concept would have had some currency at the start of this era. Demetrius (De
eloc. 222), in describing the Plain Style in rhetoric, tells us that Theophrastus says: One
should not elaborate everything in detail but should leave some things for the hearer to com-
prehend and infer on his own. For when he understands what you have ommitted, he becomes
not just your hearer but your witness, and a friendlier one at the same time. For he thinks him-
self intelligent due to you, who have given him the possibility to show his intelligence. To spell
out everything, as though to a fool, is like despising your hearer (" 6/
, / ( M J 6 #D
6+ # " 9
"D 2M +, "M D
+ E M ( ). What is interesting for
our purposes is that this strategy from rhetoric came to be applied in poetry, precisely in the
imaginative space opened up for the reader by epigrams shift from monument to scroll.
Let us now take a look at Ergnzungsspiel in a few Hellenistic examples,

specically in several epigrams by Callimachus. I want to stress from the
start, however, that I am not proposing a general strategy for interpreting
Hellenistic epigram: though Ergnzungsspiel is common, Hellenistic poets
have many bows in their quiver. Let us start with a simple and inconspicu-
ous distichon, which largely retains the traditional form of a votive inscrip-
tion. It is Callimachus epigram 33 Pf. ( 21 GP AP 6.347).
D
P, / G J
1 + , , + .
[Artemis, for you Phileratis here set up this offering.

Accept it, lady, and keep her safe.]
This poem is so straightforward, so modest and plain, that it initially

makes us unsure of our critical faculties. Why do we even dignify it with
our attention? Would we do so if it had not come down to us under the
name of the great Callimachus? The feeling is like that which one some-
times has in museums when one drifts by a painting without giving it
a thought (for it is not particularly striking), then suddenly notices out of
the corner of ones eye that it is labeled Rembrandt, whereupon one
snaps to attention, examines it with care and interest, and might go so far
as to call it a masterpiece. On honest reection, however, it is hard to
shake the feeling that one has been a victim ofindeed, perhaps has
helped perpetuatea swindle. Returning now to our epigram, we nd
that it is conventional to a degree one would not normally expect in Calli-
machus. The whole opening phrase right up through the caesura ( P,
/ ) corresponds word for word to an old dedicatory formula,
which we know from inscriptional evidence.14 The structure, too, is con-
ventional. As Friedlnder and Hofeit put it, in the dedicatory epigram it
is not rare to nd a break between the objective formula of dedication in
the hexameter and a prayer directly addressed to the god in the pentame-
ter (p. 66 FH).15 How, one wonders, can we distinguish the poem from
any inscribed dedication that might be found in a public setting? What
sets it apart as Callimachean?
The work of an epigramof votive or sepulchral type, at leastis to
14. Cf. CEG I 413 110 FH: P, / [ / ] /

/ , . Cf. also CEG I 407 125 FH: / / []D
{} / [_] " / " " .
15. The request, , is likewise conventional (cf. CEG I 345: 1 ,
P; CEG I 367, 418), as is the plea (cf. CEG I 275.4).
make a passerby pause and, however briey, connect.16 It seeks to engage

and involve a readers thoughtswhich, after all, have their own preoccu-
pations and agendaand to elicit a response ranging anywhere from pity
or sympathetic witness to approbation or mere acknowledgment. We nd
the process expressed with rare explicitness in a sixth-century b.c. Attic
verse inscription on the marble base of a grave stele (CEG I 28 83 FH
GV 1225), which virtually accosts the distracted passerby.17
,
h[] : / | : :
| : : : ,.
[Man, as you stride along the road with your mind on other subjects,
stop and weep to see the tomb of Thrason.]
The involvement of the reader in Callimachus poem looks very differ-

ent. Everything is indirection. A woman dedicates an to Artemis,
beseeches the goddess to accept it, and asks for protection on her own be-
half. There is no appeal for our attention. The goddess Artemis is the ad-
dressee; the reader is ignored. At rst sight, there is little that prompts us
to participate in this ostensibly closed and straightforward dialogue (inas-
much as we take the trouble to read the poem at all, we appear to remain
eavesdroppersand unacknowledged ones at that). And this may explain
why scholars have for the most part passed this couplet by in silence.
Passing by is here no idle metaphor. Our experience as readers is not
so very different from that of the wayfarer going past an inscription. To
be sure, the mere fact of our encounter with Callimachus poem already
assumes a voluntary act, that is, that we have chosen to take up the book
(or scroll) of poetry. And this signals a certain willingness on our part to
listen to what the text is saying. Nevertheless, who is to say that we will
choose to concentrate on this particular couplet?18 We may well be more
or less receptive to the poems bid for our attention, understated as it is
no one would argue that it thrusts itself upon our consciousness. And the
possibility of neglect is heightened by the very format of the anthology
16. On this topic, see the ne essay by G. B. Walsh (1991).

17. For a similar view of this epigram, cf. Peek 1960: 16; Walsh 1991: 80.
18. Frasers characterization of itwithout elaborationas a charming couplet (1972:
II 329 n. 35) suggests a low-intensity experience of the poem, with light readerly involve-
ment, yet evidently still mildly pleasurable: the reader is charmed. For the self-conscious ac-
commodation of different levels of readerly involvement in Callimachus, see my essay on Im-
personation of Voice, chap. 2 in the present volume. For a Hellenistic model of detached
reading, cf. Nussbaum 1993: esp. 13645.
or collection, which I consider very likely for Callimachus19 and which

invites us to dip in here or there, to pick and choose whatever happens to
catch our eye, rather than read its parts with equal intensity throughout, as
one might a single continuous workthough even then such concentra-
tion is an ideal that scarcely corresponds to the reality of reading. The pic-
ture of the wanderer striding along the road with his mind on other things,
which we saw in the Attic inscription, is thus not a bad image for the
reader of epigram. Only that, in the case of Callimachus couplet, no voice
calls us to stand and pay attention.
Yet the door to participation is left open, maybe more than a crack.
First of all, as we already saw in the case of the Midas epigram, the couplet
tantalizes with the specicity of its deictic pronouns (/ . . . J), which
point to a concrete object and place. If this couplet was ever inscribed,
then their identity was clear to the worshipper (p. 110 no. 114 FH). And
I certainly wouldnt exclude that possibility: it may well be that Calli-
machus composed epigrams now and again on commission. But even then,
I think we must reckon with the likelihood that so powerful an exponent of
book poetry as Callimachus would at the same time have contemplated his
poems place in a book. Set in the scroll of Callimachus epigrams or in an
anthology, the couplet becomesself-consciously, I believedislocated
or, better, unmoored; / and J oat free, a provocation to imagina-
tive play. Where is this place? What was this ?
Further, the private quality of the dialogue between woman and divinity
is belied by the conventional third-person voice. This may or may not be
Phileratis own, and it introduces the possibility of another perspective. Is it
the poets, the stonecutters, or (of greatest consequence) our own, when we
say the words out loud or in our minds?20 Callimachus knew full well how,
19. We cannot say for sure whether Callimachus collected his epigrams into a book. But
given the poets well-known role in editing his Aetia and Iambi (cf. esp. Krevans 1984) and
given the likelihood that his two sepulchral epigrams (21 Pf. and 35 Pf.) were intended to be
read togetheras one would in a collectionI think it probable that he did indeed put to-
gether such a book.
Epigram collections of the early Hellenistic period are known to us from papyri. The
most important is the third-century b.c. Milan Posidippus papyrus (P. Mil. Vogl. VIII 309),
containing ca. six hundred verses (112 poems, only two previously known) from a collection
of this poets epigrams, arranged according to theme. For a thoroughgoing assessment of the
papyrus as epigram book, see Gutzwiller 2005b. For other early Hellenistic epigram collec-
tions, cf. Pack no. 1593 SH 961; the Elephantine Scholia Pack nos. 1924, 1594, 1596.
See, further, the still unpublished P. Vindob. G 40611 of the third century b.c., which contains
the incipits of an extensive, multivolume epigram collection; cf. Harrauer 1981. On the
arrangement of epigram collections generally, see Krevans 2007.
20. In connection with the indeterminate nature of the voice, cf. Roland Barthes (1974:
41) description of the plural text: The more indeterminate the origin of the statement, the
in the act of reading, a reader can be drawn into collaborating with the text:
one might call this the Acontius effect, after that characters manipulation
of his beloved Cydippe when he tosses her an apple inscribed with the
words By Artemis, I will marry Acontius, which she read out loud, thus
unwittingly binding herself (Callim. fr. 67, Dieg. Z 1 with Pf. ad loc.).21 If
we, then, speak these words on Phileratis behalf and lend our voices to her
plea, will we not want to know more about her? Once we decide to construe
the perspective as our own, wont that prompt us to involve ourselves fur-
ther through a desire to learn as much of the context as we can?22
In what circumstances, we might ask, would this woman offer an
to Artemis and plead for her protection? Arguably, she would do so
on the occasion of impending childbirth (for Artemis special role in this re-
gard, cf. Callim. Hymn 3.21ff.). If so, the nal imperative will have the
poignant sense of save from death, keep alive (death in childbirth was a
sad fact of life, as a glance at the funerary inscriptions in Peeks Griechische
Vers-Inschriften reveals). And if this is indeed the situation, is it meaningful
that Phileratis appears alone. No mention of parents. No mention of chil-
dren. No mention of husband (contrast the inscription CEG I 413, cited in
n. 14 in the present essay). Does Phileratis solitude in this poem suggest
that the she dedicates is nothing so costly as a statue (as assumed by
almost everyone),23 that the term denotes some humbler item, as it so often
does in genuine dedications?24 And what of her name? Though the female
form is unique, the masculine is conspicuously Rhodian.25 Does this sug-
more plural the text. In modern texts, the voices are so treated that any reference is possible:
the discourse, or better, the language, speaks: nothing more. By contrast, in the classic text
the majority of utterances are assigned an origin, we can identify their parentage, who is speak-
ing . . . ; however, it may happen that in the classic text, always haunted by the appropriation of
speech, the voice gets lost, as though it had leaked out through a hole in the discourse.
21. In speaking of inscribed epigrams of the more oral culture of archaic and classical
Greece and how they enlist the viewer in their own behalf, Svenbro goes so far as to describe
the reader as dispossessed of his own voice and having to submit to the written word,
adding, in these circumstances, the reader has but one means of resistance: he can refuse to
read (1993: 47); cf., generally, Svenbros chap. 3, The Reader and the Reading Voice, and
chap. 9, The Inner Voice: On the Invention of Silent Reading. On epigrams appropriation
of the readers voice, cf. also Day 1989: esp. 2628; 2007: 2947; Kurke 1993: esp. 14446.
22. Compared with earlier times, the relatively greater prominence of silent, more
purely visual reading in the Hellenistic period offered readers greater latitude in deciding
their degree of involvement with a text. For in their silence, they might initially approach a
text quite casually and noncommittallyjust trying its perspective on for size, while actually
still withholding identication. For silent reading as mainly a postclassical phenomenon, cf.
Svenbro 1993: 16768.
23. Cf. GP ad loc.
24. E.g., ceramics; cf. CEG I 28992, 298, 334.
25. Fraser 1972: vol. 3, 826 n. 216. Cf. CEG II 690 from Rhodes (ca. 360350).
gest a Rhodian setting? Or, considering the rootless life of the Hellenistic
diaspora, may we imagine this solitary woman in a foreign land (in Alexan-
dria, perhaps), cut loose from her family?
I have spun out a pathetic tale, which readers may or may not nd plau-
sible. But that is not paramount. The point is that the poemwith that ex-
pressive brevity that is the marker of the genreinvites such speculative
play and that Ergnzungsspiel constitutes to a signicant degree the aes-
thetic pleasure of reading the poem. It is here, if anywhere, that the specif-
ically Callimachean quality of the piece is to be found. Those readers
who do not indulge in such play (perhaps constructing tales more plausible
than mine) are missing out on the fun.26
Now one could scarcely hope to nd a passage in Callimachus where
such a process of supplementation is referred to explicitly and that would
allow us to say for sure that he made deliberate use of Ergnzungsspiel.27
But there is such a passage, I believeone whose poetic signicance has
not yet been fully appreciated. It is fragment 57.12 Pf. ( SH 264.12).
" 6, / JD

@ / M [], / 6D
[[The reader] can imagine [this] for himself, and thus cut down the
length of the song.
But all that he answered to the questions, I will relate.]
The verses are attributed with fair certainty to Callimachus epinikion for
the Ptolemaic queen Berenike II, the Victoria Berenices. And they probably
belong to the mythical portion of the poem, where Herakles returns to his
humble host Molorchos after having killed the Nemean Lion. The speaker
is evidently the poet himself, who addresses these words to his audience.
The subject of " 6 is thus a reader or listener, as
26. Cf., similarly, Hunter 1992a: esp. 114: Much of what I have to say will be speculative,
butlike many of the best Greek epigramsthese poems are very clearly written as a provo-
cation to speculation. Perhaps no literary genre makes such a direct appeal to the readers pow-
ers of intellectual reconstruction, to the need to interpret, as does that of epigram; the demand
for concision makes narrative silences an almost constitutive part of the genre. In these cir-
cumstances, the refusal to speculate amounts to no less than a refusal to read.
27. Iser (1984: 176) cites a passage from Sternes Tristram Shandy (1956: 79) as early
evidence (eighteenth century!) that Autor und Leser . . . in sich das Spiel der Phantasie
[teilen]. The passage reads . . . no author, who understands the just boundaries of decorum
and good-breeding, would presume to think all: The truest respect which you can pay to the
readers understanding, is to halve this matter amicably, and leave him something to imagine,
in his turn, as well as yourself. Sterne, however, may here be alluding to the ancient view of
Theophrastus, cited in n. 13 above.
Pfeiffer suggested.28 Therese Fuhrer, who recently dealt with this passage
in her study on Callimachean epinician,29 argues persuasively that what
this reader is supposed to ll in for himself is the heros well-known and
stereotypical combat against the monstrous liona tale that the sophisti-
cated poet naturally wanted to avoid treating in extenso.30 Fuhrer demon-
strates that these verses belong to a type of transitional formula that may
be found in Pindar: an Abbruchsformelhere in the form of a praeteritio in
which the reader is invited to supply the omission. Indeed, she suggests
that Callimachus may have been alluding to a specic case in Pythian
4.24748, which likewise involves the omission of an heroic combat (hero
versus monstrous beast), namely, Jasons struggle with the hydra.31 The
passage is worth recalling here.
/ D O

, : D
[It is too far for me to go along the carriageway:

for time is pressing
and I know a shortcut.]
In discussing Callimachus interest in this type of formula, Fuhrer

stresses the similarities between examples in his own verse and in that of
his predecessors. But I would insist on a basic difference, whichto my
mindis given short shrift in Fuhrers account. In all earlier instances
and the passage from Pythian 4 is a good exampleit is exclusively the
poet who undertakes to shorten the poem. In fragment 57, by contrast, it
is the reader, for the poet expressly invites him to imagine the rest for
himself and thus abridge the poem. This invitation to the reader is, so far
as I can see, unparalleled in earlier literature,32 and it remained so until
28. Pf. ad v. 1: " sc. C vel C ipse excogitet quid aliud fecerint.
29. Fuhrer 1992: 7175, 12125.
30. Ibid., 7275.
31. Ibid., 74. For such formulas generally, cf. Braswell 1988; ad vv. 24748; Thummer
1968: 12225; in rhetoric, cf. Krischer 1977.
32. This is the case, even though Fuhrer shows (1992: 123 n. 457) that already Pindar
mit dem Wissen des Publikums rechnet when he interrupts a myth. To be sure, Fuhrer says
that Callimachus, by means of this device [sc., Abbruchsformel], challenges the reader to draw
on his own erudition and knowledge of mythology to understand the poets learned and witty
allusion, but she clearly means challenge here quite generally, not as an explicit invitation
(cf. Fuhrer 1988: quote on 66). In the Augustan period, cf. Ovid Amores 1.5.2325: singula
quid referam? nil non laudabile vidi, / et nudam pressi corpus ad usque meum. / cetera quis nescit?
Asking Who doesnt know the rest? is like saying You, the readers, know the rest, so I
dont need to go into details.
the time of Augustus. Indeed, the first really comparable instances appear
in Lucian.33
The explicit call to supply what is missing here in fragment 57 is con-
rmation of the strongest sort that Callimachus knowingly and deliber-
ately used Ergnzungsspiel in his epigrams. To be sure, the invitation to
play is never so direct as it is in this fragment. Yet we do quite often nd
Ergnzungsspiel enacted in the epigrams; that is, we can observe characters
in the poems themselves engaging in the game.34 This is the case, for in-
stance, in a sepulchral epigram (58 Pf. 50 GP AP 7.277).35
; 6
, b

# 6/ , , + M
M
6 2 D "+ "
\, ,J / , .
[Who are you, shipwrecked man? Leontichos found the corpse

here on the beach and covered it with this tomb,
weeping for his own doomed life. For his way is not
peaceful either. Rather he roams the sea like the shearwater.]
The scene is the seashore (6 . . . 6/ , ). And we nd ourselves

M, v. 2). There is no further specication of place.
at a grave site (M
A certain Leontichos has found a corpse on the beach and heaped a grave
mound over it. We may assume that it was also he who commissioned the
inscription.
Who asks the question with which the poem begins? Since the rest of
the poem retreats into the third person, it is difcult to say. It may be
Leontichos, the author of the epigram, or, nally, we the readers. Initially,
of course, it was Leontichos who asked this question as he happened on
33. Dial. Deor. 19 (11) 2, Selene to Aphrodite about Endymion: ,D "
; Cf. also Aristainetos 1.16.33ff.: / (, C )
, b , "+ ; 2.3.1516: T -
; , 6 6
J. The idea that a reader must imagine more than is present in the text occasionally
appears in ancient works of literary theory. Ps.-Longinus, for instance, says that poetry partakes
of the sublime when a readers thought exceeds what is said (
, 7.3). Cf. also n. 13 above. Schmidt has collected further examples in
Horatianum (1990), particularly in the section Wenig Worteviel Sinn (9094).
34. Cf. the discussion of this phenomenon by Walsh (1991: esp. 98), who, however, is in-
terested mainly in the primary concern with information, and so with the way one acquires
information. Problems look for solution; inference . . . structures feeling.
35. On this poem, see also Bruss 2005: 15659; Meyer 2005: 2068.
the corpse.36 This is clear from his reaction, when he sheds tears for his
own life. In funerary epigram, tears generally ow for the dead. And after
the two nite verbs of verse 2, which take as object, we expect the
participle to do the samean expectation perhaps heightened
by the slight pause for the caesura after 6. It is surpris-
ing, then, that Leontichos grieves for his own life (2 ).37 The anony-
mous corpse, however, presents a blank to all who encounter it (it is thus
an ideal object of Ergnzungsspiel), and Leontichos mourns his own lot be-
cause he sees in that blank a reection of himself. Put somewhat differ-
ently, Leontichos is the answer that he nds to the question, Who are
you?38 In effect, then, he writes the epitaph for himself.
And yet he does not go that far. We saw how, in the dedication of
Phileratis, readers could decide to what extent to involve themselves. And
their willingness to do so depended on how they were inclinedtheir rel-
ative interest or indifference, alertness or inattention. So with Leontichos,
the possible limits of Ergnzungsspiel come into view, inasmuch as his will-
ingness to play may only go so far. Leontichos draws on his own life expe-
rience in order to ll out the meager traces he encounters in the corpse
and so form a coherent picture. Yet Ergnzungsspiel in this instance cuts
very close to the bone. If continued play means in effect constructing ones
own tombstone, then perhaps the stakes are just too high. And in fact,
Leontichos stops before getting to that point. He avoids naming his coun-
try of origin, his father, his family, or other identifying traits; that is, he
36. Following Agar, Gow and Page print # /, thus transforming most of the epigram
into the corpses replya pointless change, since the corpse does not answer the question (as
surely it could). Construed thus, moreover, that question can only be taken as the readers.
But surely the point is that it was originally Leontichos own and that because the identity of
the corpse remains a blank, Leontichos can see in it a reection of himself.
37. Cf. GV 1231 ( GG 170), where the passerby is asked to mourn the death of a six-
year-old boy as though the loss were his own:
,
/ 6

, ,
a 2 6 .
This epigram differs from that by Callimachus inasmuch as a sympathetic response here
causes the passerby to grieve for another; in Callimachus it prompts him to grieve for him-
self. The basic psychological attitude is described with greater explicitness in the passage
from the Iliad (19.3012) where Patroklos is mourned: , /
, / "
/ 2. Cf. Edwards 1991: ad loc.
38. A reader might at rst think that Leontichos is indeed the answer to the question
;, since the name is placed precisely where such answers usually come in dialogue epi-
grams. For such poems, cf., e.g., Leonidas 70 GP AP 7.163; Callim. 34.2 Pf. AP 6.351;
GP ad Antipater 21. For inscriptional examples, cf. GV 1831ff.
withholds (as though superstitiously) details that would truly make the
commemoration serve as his own prospective epitaph and ensure that we,
the readers, will be able to mourn his end.39 Instead, he moves onrestless
like a seabirdbefore having identied himself further. Perhaps we would
like to have known more about himhis compassionate gesture40 may
well have piqued our interest. But it is hard to begrudge his evasion. For
not everyone is always prepared to play the game to its bitter endif at all.
And we, too, will be moving on: even if we choose to linger over this par-
ticular poem, the shore upon which it is setan archetypically liminal
locationis not where we shall stay. Our community and our home, our
occupations, family, and friends are all elsewhere.
Precisely such a familial or community setting is, in my opinion,
evoked in our next examples, two well-known epigrams of Callimachus.
Here we return from Ergnzungsspiel played in the poem to that performed
by the reader. One of the poems is that on the tomb of his father (21 Pf.
29 GP AP 7.525).
K 6 ,
: .
, / D C @
V, C / A . . .41
[Whoever you are who bends your step past my tomb, know that I
am both child and father of Callimachus the Cyrenaean.
You are sure to know them both. The one led his countrys
troops, the other sang songs beyond the reach of envy . . .]
The second epigram is that for the poets own tomb (35 Pf. 30 GP AP
7.415).
+
"

,, " / :M .
[You bend your step past the tomb of Battus son, well skilled
in song, well able to raise a welcome laugh over wine.]
39. Nor does he seem to draw the obvious conclusion found in epitaphs of other
(e.g., Leonidas 60 and 61 GP AP 7.264, 266), i.e., to give up sailing altogether.
40. N.b. the respect for the dead man expressed by the very dignied form of address,
b . The b postpositum is highly poetic; cf. Pf. ad fr. 103.1; Mineur 1984, ad v. 118;
Allen, Halliday, and Sikes 1936, ad Hom. Hymn 3.14. For an instance in which the rst ele-
.
ment is nominative pro vocativo, cf., e.g., Il. 4.189: b
41. This is not the place to discuss the much debated problem of the nal distichon (vv.
56). Cf. Livrea 1992, with bibliography.
It has long been recognized that these two poems go together and supple-
ment each other.42 For the actual tenant of the tomb remains unnamed in
his respective epigram. Only when we compare the two does it emerge
echt alexandrinisch, as Gabathuler puts it43that the rst poem is for
Battus, son of Callimachus of Cyrene, the second for Callimachus, son of
Battus of Cyrene.44
The missing name is clearly an enticement to Ergnzungsspiel. Yet
what, in this instance, are the rules of the game?45 Claude Meillier has ob-
served that it is not unusual for the name of the deceased to be left out in
an inscribed sepulchral epigram. When this is the case, one can generally
nd it inscribed extra metrum above the poem or below it.46 Literary epi-
gram, however, does not use such extra metrum inscriptions: it did not go
so far in adopting the conventions of its inscriptional counterpart. Never-
theless, Meilliers observation that the name is to be found outside the
poem itself may point us in the right direction. We must simply take it a
little bit further.
We know that ancient families often had family grave plots, wherejust
as todaythe tombstones of various family members stood next to each
other. A good example is a grave stele of the rst half of the fourth century
b.c. from the Piraeus (CEG II 512 GV 1386 Clairmont 74). The name
and parentage of the deceased is inscribed extra metrum above an empty
space, where a painting once stood: Telemachos, son of Spoudokrates,
from Phyla ( | | ). Below the vacant space
comes the epigram itself.
42. For such epigram pairs on monuments, see Gutzwiller 1998: 22931; Fantuzzi 2007;
Fantuzzi 2008. For examples in literary epigramincluding in the two poems under discus-
sion heresee Kirstein 2002.
43. Gabathuler 1937: 56.
44. Thus Wilamowitz 1924: I 175 n. 2; Pf. ad Ep. 21.56. Reitzenstein (1893: 139 with n.
2) made the interesting suggestion that we view funerary epigrams that poets write for them-
selves as concluding poems for collections: e.g., the epigram of Nossis (11 GP AP 7.718),
that of Leonidas (93 GP AP 7.715), and the three that Meleager wrote on himself (24 GP
AP 7.41719). We encounter such an epigram in situ in the nal poem of Propertius
Monobiblos. Cf., further, Gabathuler 1937: 4849, 56.
45. Walsh (1991: 94) writes: [T]he riddle is solved by the younger Callimachus fame
everyone must know his patronym. Fame, more or less, is the point of the poem. But if fame
is the point, why are the poems so carefully harmonized, so as to supplement each other? Fer-
guson (1980: 141) aptly cites the remark of Abraham Mendelssohn, son of the philosopher
and father of the composer: I used to be the son of my father, but now I am the father of my
son. How many, one might ask in response to Walsh, would know the name of this son of a
famous father and this father of a famous son?
46. Meillier 1979: 139. I would cite the following examples: CEG I 77, 89; CEG II 477,
512, 520, 524, 528 (?), 53132, 564, 57071, 585, 589 (This is the sister of Smikythosthe
sister herself remains unnamed), 590, 59496, 613, 67071, 67778, 684, 703 (?), 722, 724, 741.
g /
|
( (/ |
. / 6 , , |
" .
[O man that for your ever-remembered excellence won

great praise from all the citizens and are sorely missed
by your children and dear wife.I lie, mother, on the right
of your tomb and am not deprived of your love.]
Interesting here that the deceased explicitly refers to his mothers neigh-
boring tombwithout, however, calling her by name. She appears simply
as , just as the father of Callimachus in epigram 21 was simply called
(v. 2). Fortunately, however, the tombstone of this very woman,
the mother of Telemachos, has been found.47 Its inscription reads as fol-
lows (IG II/III ed. min., vol. 3.2 no. 7695):
M
[Melite, wife of Spoudokrates, from Phyla]
Thus we only learn the mothers name from this second tomb. Here, to be
sure, the names of the deceased are present on their own tombs. This was
not the case in Callimachus: his omission of the same shows that he is
playing with the conventions of real-life sepulchral epigrams.48 But Calli-
machus can reckon with the readers ability to see through his game and
realize that the poems supplement each other. For the reader knows about
such family grave plots and so possesses the information necessary to play
the game. One of the pleasures of Ergnzungsspiel, in this instance, is that
the reader must translate the context of such real-life family plots onto the
very different landscape of the scroll: the Sitz im Leben becomes the Sitz im
Buch.49 And if we do this, if we imagine the Callimachus family plot set on
the papyrus, it follows with virtual certainty that his two epigrams (though
separated in the tradition) were juxtaposed on the scroll. The only uncer-
tainty is which came rst, which second.
The two epigrams do not, of course, refer to each other as overtly as in
the sepulchral poem for Telemachos, where the deceased explicitly says, I
47. Cf. Conze 18931922: no. 803, plate 150. On family grave plots, cf. Humphreys
1980; Garland 1982; Garland 1985: 1067.
48. Khnken 1973: 426.
49. For this literary landscape, cf. Bing 1988: 3940.
lie, mother, on the right of your tomb. Nor are there the visual links that
might obtain between tombs in a genuine family plot (common elements
of style, matching decoration, placement on the plot). Instead, Calli-
machus devises a subtler way of expressing family relationshipnamely,
by having each poems opening words unmistakably echo the others:
K 6 . . . (21.1)
. . . (35.1)
The repeated phrase () is utterly convincing as a

traditional funerary formulaso convincing, in fact, that, until now, no
one has noticed that the expression is not attested before
Callimachus50 and appears again only much later.51 Did Callimachus here
coin his own conventional (i.e., familial) funerary idiom? And is that how
he suggests the relationship, or family resemblance, that exists, on the one
hand, between the poems and, on the other, between the deceased?52 Is the
repeated phrase a signpost to help orient us in the landscape of the book,
as we engage in Ergnzungsspiel?
Before closing, I want to discuss one last epigram of Callimachus and
thus deliberately extend the meaning of Ergnzungsspiel. It is epigram 22
Pf. ( 36 GP AP 7.518).
/ , \
/.
6 =,
" J # , "
, / / ,+ .
50. It is formed on analogy with such phrases of journeying as I (Eur. Phoen.

302; Soph. Philoct. 291; Theocr. 7.21: , Y 1 I;) and
(often in Euripides, e.g., Suppl. 171 with Collard 1975: ad loc. Cf. also
in Homer (Il. 15.269, 22.24). N.b. that inverts the common Homeric
[sc., ] (Il. 6.514, 13.515, etc.). See, further, Meyer 2005: 178 n. 176.
51. Cf., from the second or rst century b.c., GV 1990.5 no. 38.5 Bernand: , 1 /
T V . Cf. AP 8.188 (Gregory of Nazianzus, fourth cen-
tury). Faraone refers to GV 2036.11 (T 6 ) but adds a nonexistent
after and sets the poem in the fourth or third century b.c., rather than a.d. as it
should be (1986: 55 n. 8).
52. The relationship between father and son, expressed by the identical phrase, corre-
sponds to that in Ep. 21 between grandfather and grandson, expressed in their identical name.
As Reitzenstein observed (1908: 8586), the name Callimachus is here used in its etymological
sense, an able warrior. The family resemblance can still be felt across generations in the fact
that the grandfather was a Callimachus in a martial sense, the grandson in a literary sense
(C . . . @ / V, C / A , vv. 34). On isonumia, the reuse of
a family name over several generations, cf. Svenbro 1993: chap. 4.
[Astakides the Cretan, the goatherd, was abducted by a nymph

from the mountain, so now its sacred Astakides.
No longer beneath Diktaean oaks, no longer of Daphnis
will we herdsmen sing, but evermore of Astakides.]
The sophisticated Alexandrian poet presents himself hereno doubt, with

a smilein the guise of a lowly Diktaean herdsman.53 Wilamowitz summed
up the situation as follows: A herdsman has vanished on Mt. Dikte,
6. The other herdsmen tell a tale, as they would today, that a
Nereid got him. But back then a Nereid was no devil, and being trans-
ported to fairyland did not cost one ones eternal bliss; it bestowed it. The
herdsmen will now sing a ballad on the abduction of Astakides, and he will
become a \ ,, as Daphnis was before him.54
Gow and Page found the epigram puzzling and suspected that it was
simply a joke, though if so the point is . . . lost (ad loc.). I believe that
their suspicion was correct. But in order to get the joke, we must look for
additional information beyond the poem itself (as with the two funerary
epigrams treated earlier). Here, however, our search takes us beyond even
the poems immediate vicinity on the scroll. It leads into the broader liter-
ary landscape of Hellenistic bucolic.
Pfeiffer observed (ad loc.) that the repeated "at the start of
verse 3 and after the bucolic diaeresisrecalls anaphora bucolica. This
type of repetition is a fairly common feature in Theocritus. I doubt, how-
ever, that Callimachus merely intended a general stylistic reference.
Rather, he had a certain eidyllion in mind. For anaphora bucolica is the
specic and very distinctive mark of one particular Theocritean song: the
Daphnis song in Theocritus 1 (vv. 64142).55 Here it occurs twenty-two
times in seventy-eight verses, over four times as often as in any other
poem by Theocritus. Moreover, it appears in three of the rst four verses
of the song, thus constituting a kind of metrical signature. Given that the
53. Cf. Hymn Art. 17082, where the poet adopts the guise of a simple farmer and
whereas herethat guise actually has to do with literature, not farming. On this passage,
cf. Bing 1988: 8389.
54. Ein Hirt ist im Diktischen Gebirge verschwunden, 6. Da erzhlen
sich die Hirten, was sie sich auch heute erzhlen wrden, eine Nereide hat ihn geholt. Aber
damals war die Nereide kein Teufel, und die Entrckung ins Feenland kostete nicht die ewige
Seligkeit, sondern verlieh sie. Die Hirten werden nun eine Ballade vom Raube des Astakides
singen, er wird ein A , werden, wie es bisher Daphnis war (Wilamowitz 1906:
176 n. 1). Cf. Conner 1988 for this topic generally, 16566 on our poem.
55. Schmidt (1987: 93 with n. 65) has likewise pointed to the inuence of Theocr. 1 for
the anaphora bucolica of Ep. 22 but does so without seeing the consequences for our un-
derstanding of Callimachus poem.
anaphora of " in our epigram also occurs in connection with Daphnis

(" . . . , vv. 34),56 that the speaker presents himself
as a ( . . . , v. 4), like Thyrsis in Idyll 1 (vv. 7, 15),
and that in both poems the song is performed under tree cover (#
# , Theocr. 1.21),57 it seems likely that Calli-
machus was directly alluding toperhaps gently mockingthat song, as
though to say, Enough already with Daphnis! . . . Enough!58
This tone of gentle mockery ts perfectly with that of the rest of the
poem. The exaltation of the goatherdthe lowest grade of herdsman, ac-
cording to Gow (ad Theocr. 1.86)to heroic status is expressed with a
witty turn: the nymph abducted Astakides, and now . . . The pause after
at the caesura arouses our anticipation: and now what? The position
of immediately following is thus emphatic. Placed thus, it becomes

the comic contrast to the , that Astakides was.59 Further, with the
bucolic scene set in Crete (rather than Sicily, where the Daphnis legend is
usually set) and with the speaker a Cretan herdsman, we would do well to
look out for tricks. Though the speaker claims that herdsmen will sing
evermore of Astakides (/ / ,+ , v. 4), this previously
unattested Astakides does not in fact appear in poetry ever again:
. In the short space of this epigram, however, he is ubiquitous.
The threefold repetition of his name60 (which incidentally means son of a
lobster61), each time at an emphatic position,62 seems exaggerateda case
56. For the anaphora in ", compare Theocr. 1.11617: _ 65

"/ / a, / "/ . Has Callimachus deliberately reversed Theocritus
anaphora from bucolic diaeresis verse start to verse start bucolic diaeresis?
57. The setting of the Daphnis song in Theocr. 1 is right across from a
C
place YZ / C (vv. 2223). This sounds curiously like
Callimachus setting ( singing # ). Did the poet of the epigram pointedly set
his herdsmens song at a site that recalled the unused one just opposite in Theocritus? Such
allusions as those considered here and in the previous note wouldin the narrowness of their
focus, the detail of their referencevirtually preclude detection and appreciation if experi-
enced aurally (e.g., if one merely heard this epigram at a symposium). As stressed in connec-
tion with the Midas epigram in Platos Phaedrus, one needs the textsand time to examine
themto get the most out of Ergnzungsspiel.
58. Callimachus seems to be contrasting the luckless love of Daphnis with Astakides
successful consummation. Herdsmen will hereafter sing only of love fullled.
59. The term is pointed, moreover, because it denotes what belongs to the god
(something dedicated to the god), which human hands may no longer lay claim to. Astakides
now belongs entirely to the nymph and is no longer fair game for human affections.
60. Might Callimachus here be alluding to another poem by Theocritus, namely, the
Hylas? There a heroized victim of nympholepsy is likewise named three times (13.5859), ap-
parently as an aition of the triple cry in the Hylas cult (cf. Gow 1950: ad loc.).
61. Larson (1997) argues that Astakides is not a proper name but a humorously pseudon-
ymous ethnic, based on Bithynian Astakos.
62. Verse start, verse end, and right before the caesura.
of goatherd boosterism run riot. Surely it strikes a humorous note.63 In

short, I believe that Callimachus was here poking fun at Theocritus not just
for the anaphora bucolica of the Daphnis song but for what has been
called the most prominent single characteristic of Theokritos style, his
well-known taste for repetition.64
As already mentioned, this instance of Ergnzungsspiel looks rather dif-
ferent from that in our previous examples. There, the game was sparked
by the fact that in the ever more literary epigrams of the Hellenistic period,
the deixis of the old inscriptional poems lost its real point of reference. But
the Hellenistic poets were able to turn this referential vacuum to advantage
and give it appeal, by inviting their readers to supply the missing references
themselves. Here, on the other hand, the references that must be supplied
are not to some concrete object or place but, rather, to another work of lit-
erature. No doubt this is Ergnzungsspiel in an extended sense. But I would
see in it a different manifestation of what is essentially the selfsame game.
For its motive force is the same: a preference for the kind of text that
Barthes has (perversely) termed writerly, that is, one that puts the reader
to work (if of course he is so inclined), allowing him to be no longer [just]
a consumer, but a producer of the text.65 This preference in fact represents
a typical trait of Hellenistic poetry, for the authors of the age ask their read-
ers to supply a great deal. They are expected to recognize and bring to the
text an understanding not just of literary allusions (as argued earlier) but
those to history, geography, medicine, religion, and so on. And this, too,
can be considered a type of Ergnzungsspiel. The words of Wolfgang Iser
could well describe the particular pleasure of Hellenistic poetry: Reading
becomes fun, he says, only then, when we play a productive role, and that
means when texts give us the chance to use our abilities.66
63. Threefold repetition of a name is entirely atypical of Callimachus. The poet gener-
ally avoids repetition of names by using patronymics or ethnics (cf., e.g., Ep. 2.1, 4; 6.1, 4;
10.1, 3; 27.3, 4). Cf. Lapp 1965: 25: ad nomina propria sive vitanda sive varianda Calli-
machus saepissime ea circumscriptione utitur, cui nomen est antonomasiae.
64. Cf. Dover 1971: xlv: The most prominent single characteristic of Theokritos style
is his repetition or partial repetition of words.
65. Barthes 1974: 4. Barthes opposes to this the readerly text, which forms the enor-
mous mass of our literature (5), wherein the reader is plunged into a kind of idleness (4)
and in which reading is merely the reactive complement of a writing which we endow with
all the glamour of creation and anteriority (10). We call any readerly text a classic text, (4).
66. Das Lesen wird erst dort zum Vergngen, wo unsere Produktivitt ins Spiel
kommt, und das heit, wo Texte eine Chance bieten, unsere Vermgen zu bettigen (Iser
1984: 176).
chapter 6
Text or Performance / Text and Performance

ALAN CAMERONS
CALLIMACHUS AND HIS CRITICS
D iscontinuity, change, innovation: these are the terms most scholars in

this century have stressedone-sidedly, perhapsin characterizing
Hellenistic poetry. Tradition (they argue), though mined, sifted, painstak-
ingly studied and mastered, is deployed not for reproductive ends but in the
service of something different and new.1 Alan Cameron now shifts the ac-
cent among these terms, placing it squarely on tradition: Not only is there
no evidence that third-century artists and writers thought of themselves as
epigones living in a postclassical age. The real break came two centuries
later (2728).2 Hellenistic poetryon his viewwas not so very different
This is an updated version of the essay that appeared in La letteratura hellenistica: Problemi e
prospettive di ricerca, ed. R. Pretagostini. Rome 2000. 2000, Edizioni Quasar di Severino
Tognon srl.
1. See especially Wilamowitz description (1924: I). For him, the neue Dichtung (148) is
Lesepoesie (149), unclassical, learned, rened, and aimed at a gebildeten Leserkreis (151).
Cf. the helpful discussion in Schwinge 1985: esp. 15463. See also Pfeiffers comparable char-
acterization of the poetry of this era (1968: 88): the great old poetical forms . . . belonged
to ages gone for ever . . . Poetry had to be rescued from the dangerous situation in which it lay,
and the writing of poetry had to become a particularly serious work of discipline and wide
knowledge, and . The new writers had to look back to the old masters . . . not to
imitate themthis was regarded as impossible or at least as undesirablebut in order to be
trained by them in their own new poetical technique. In their wake, cf., e.g., Bulloch 1985b;
Zanker 1987; Bing 1988; Goldhill 1991; Hunter 1996.
2. Cameron 1995. I cite Cameron from this book throughout unless otherwise noted.
106
Text or Performance / Text and Performance 107
from what preceded it, even in archaic times. In tendency, then, Camerons
work can be placed alongside G. Hutchinsonsthough to be fair to the lat-
ter, the continuity for which Cameron argues in poetic convention, prac-
tice, and reception is incomparably more radical.3 From this perspective,
Cameron whips up a blizzard of polemic against almost any consensus one
might care to name about Callimachus and Hellenistic poetry. This is not
the place to take on the totality of Camerons diffuse and complicated argu-
ment. I mean, rather, to address just two aspects of it(1) the kind of audi-
ence to which the chief poets of the age primarily addressed themselves and
(2) the essential literariness of their works.
In Callimachus and His Critics, Alan Cameron attacks the widespread
notion that Hellenistic poets were conned to an ivory tower, writing only
for an elite circle of readers, remote from the general public. In so doing,
he stresses that traditional poetic venues such as musical contests, sym-
posia, and religious festivals continued to ourish, indeed proliferated, in
the Hellenistic age and that they were sponsored by the very kings who are
supposed to have encouraged the depolitization and hence marginalization
of song. There are instructive pages cataloging the abundant evidence
that lesser poets continued to perform their work publicly, and many pas-
sages in the major poets [that] imply that they did the same (30). Yet he
overstates his case against the bookish aspect of Hellenistic poetry, by
denying altogether that some of it was indeed intended for a select circle of
readers, who might well be described as denizens of an ivory tower.
The lengths to which Cameron will go to deny the existence of such a
group may be seen in his treatment of what he calls the one well-known
text which has often been cited as if it lent some support (31) to the image
of the ivory tower, that is, the famous fragment from the Silloi of Timon of
Phlius (SH 786).
+ 6 ,M M

6 M.
3. Hutchinson is balanced (to a fault) and far more circumspect: Crude notions of liter-
ary history often leads to ideas of a poetry absolutely different from, or much the same as, the
poetry of the classical era, and an obsession with the relationship. Swollen conceptions of the
part played in this poetry by learning and still more by theories about literature lead in prac-
tice to narrow and dull conceptions of the poems (1988: 1). Of Callimachus, he writes: [T]he
difculty of reading him has often been grossly exaggerated . . . There is demonstrably far
more to these works than the conduct of scholarly quizzes and polemics; we cannot reduce
them to such a compass by conjuring up readers whose concern with erudition drives out all
other interests and responses (67).
According to Cameron, these verses are normally understood as mean-

ing, Many bookish scribblers are fed in populous Egypt, forever squab-
bling in the birdcage of the Muses. The image has always been identi-
ed, says Cameronciting Luciano Canfora for the communis opinioas
rare birds, remote and precious creatures4 kept in a zoo, cut off from real
life (31). He challenges this interpretation on three counts (none of which
are new). First, suggests that the birds were being raised to be
eaten. Timon was thus thinking of a farm, not a zoo. Second, in
verse 2 is likely to mean not scribblers, as some assume, but, rather,
those enclosed by a fence. Cameron distances himself from the mo-
nastic associations of its usual translation cloisterlings. He says that
could suggest the fences of a farm as easily as the cages of a
zoo (32). Finally, he notes that usually refers to an open basket,
not a closed birdcagethough he admits in a footnote that Athenaeus,
who cites the verses, elsewhere has meaning birdcage and may
have understood Timons usage in that way. Still, asserts Cameron, in the
context, it is surely a birdsnest (32).
Surely is one of Camerons favorite words. But despite its promise of
certainty, readers will do well to construe it with caution: In Callimachus
and His Critics, for surely always read possibly. Let us grant Cameron his
birdsnest, however: it certainly is possible. Assuming, then, that the
birds are squabbling in a nest, he argues that the initial verb, ,
surely evokes a nest full of young birds squawking incessantly in their
rivalry for the scraps of food in the parent-birds beak (32). Thus, he
argues, there is no suggestion of caged birds, of unworldly scholars shut
away in a library; the focus of the passage is, rather, on the contentious ri-
valry of the Alexandrian scholars.
There can be no doubt that one of Timons points is the quarrelsome-
ness of the scholars in the Museum of Alexandria. But beyond that, what is
gained by moving these creatures from a zoo to a farm and out of a cage
into a nest? At best, it is like comparing free-range chickens with the coop-
bred variety. Both are captive, the freer ones still conned within fences;
and that is the case even if we admit that in verse 2 does not
convey the seclusion of a cloisterling. Cameron may be right that the
point is not the seclusion of these birds as oddities, but their value as delica-
cies for the table (32). But whether oddities or delicacies, they are
kept in a special place and fed, so as to make an exit for one purpose alone,
4. Canfora 1989: 37.

and that is not for a life as a wandering songbird.5 Similarly the young in
the nest is scarcely an image of worldliness. These birds are unedged,
conned to the nest, unable to nourish themselves, and so dependent on
their parent bird. As such, the verses still lend potent support to the notion
of the Museum of Alexandria as an ivory tower.
Timons satirical picture of contentious scholars kept on a special farm
or within the nurturing connes of a nest is of course an exaggeration.
That is the nature of satire. Those working at the museum were not actu-
ally conned, and they plied their craft in a variety of settings. Yet it is clear
that its poets, though perhaps occasionally creating works for a broader
public, often wrote for an elite group of insiders. (Incidentally, that does
not mean that their audience was all in one place: the ivory tower is not
limited to a single location; its manifestations are scattered throughout the
world, and there is communication between them.)6
A clear instance of this double aspect of the poet may be seen in Philikos
of Corcyra. On the one hand, this gure might play a very public role as
priest of Dionysus and march at the head of the technitai in the procession
of Ptolemy Philadelphus; he mightas one of the tragic Pleiascompose
works for the normally very public genre of drama. Yet he might also adopt
a very different pose in his Hymn to Demeter (SH 677):
, , # (Men of let-
ters, I bring you a gift of Philikos newly written composition). Cameron
plays down the difference by simply excising all reference to writing from
his translation (42): becomes new-fangled rather than
newly (or innovatively) written, and is rendered with the
neutral scholars. But in conjunction with , with which the
poet stresses the materiality of the text, the word demands to
be taken concretely as readers, men of letters. It is hard to imagine a
poem of the archaic or classical period addressed to such a group (where
would it have been found?). Considering the low estimates for literacy in
Hellenistic times,7 an intended audience whose self-denition centered on
this skill would have been a tiny elite indeed.
5. Cf. Cameron 1965.

6. Cameron himself (1995: 204) imagines the elder Nicander writing regional epic for
public performance in competitions at sacred festivals, while at the same time suggesting that
the Theriaca and Alexipharmaca might not have been appropriate to such contexts. They
would not have been any more appropriate at symposia. On how Nicander envisioned the
transmission of the Alexipharmaca, see n. 16 in the present essay.
7. Harris (1989: 11646) surveys literacy in Hellenistic times and estimates that levels were
probably not much higher than the 20 to 30 percent maximum he posits for the classical era.
While Cameron is right to remind us of the public aspect and function

of such poets as Philikos, he simply dismisses important evidence for po-
etry aimed at and experienced by a cultivated readership. The signicance
of acrostics is a good instance (3738). Here we encounter a technique of
argumentation that recurs often throughout Camerons book. I would call
it the Cameron two-step: step 1, assert that a particular scholarly ortho-
doxy is untrue; step 2, claim that even if true, it is unimportant. Thus
Cameron rst of all asserts that there is nothing particularly Hellenistic
about acrostics. Though in Greece we do not nd them before the latter
half of the fourth century, Cameron assures us that they are pre-Hellenic
(the earliest known are Babylonian) (37). Second, even if they do appear
fairly late in Greece, they are not important as a gauge of how the litera-
ture of the age was intended for reading.
Even the striking acrostics in Aratus Phainomena, which include the
important programmatic term (vv. 78387), are denigrated as no
more than purely external embellishments added to long poems for publi-
cation, telling us nothing about earlier performances (38). Well, we have
no knowledge at all of either prepublication performances or later ones,
and if we presume they occured, we do so on faith. But we do have three
epigrams by contemporaries of AratusCallimachus (AP 9.507 27 Pf.
56 GP), Ptolemy Philadelphus (SH 712), and Leonidas of Tarentum (AP
9.25 101 GP)that allude to the acrostic (Bing 1994; Gutzwiller
2007: 102; Prioux 2007a). Thus the earliest reception of the poem for
which we have any evidence at alland it is very early indeed!is a recep-
tion by reading. That reading, moreover, must have been remarkably
attuned to the visual dimension of the text, since these readers noticed
something that went unnoticed in all the voluminous scholia on Aratus. It
may be that the Phainomena was performed publicly as well. But there is no
reason to assume that it was not intended in the rst instance for reading.
Indeed, what evidence we have suggests it.8
The same can be said of the tragedian Chairemon, who embedded his
name acrostically in a series of hexameters, probably from The Centaur, a
8. Cameron (1995: 45) argues that one factor particularly to be borne in mind when we
are considering visual effects, such as acrostics, is that many (especially older) people did
not actually read in person because of the physical strain. Here he invokes the example of
Pliny the Elder, whose slaves both read to him and wrote for him. According to Camerons
chronology, Callimachus wrote the Aetia prologue around 270, i.e., within a few years of the
publication of Aratus Phainomena, and there (in his late forties) he was already lamenting
that old age weighed on him like Sicily on Enceladus (fr. 1.3336). Are we to imagine, then,
that it was Callimachus slave who recognized Aratus acrostic as he recited the Phainomena
to his master?
play known for its mixture of all meters (Aristot. Poet. 1.1447b21,
24.1460a2). Aristotle, Chairemons contemporary, describes him as one of
the (Rhet. 3.12.1413b8), that is, one who wrote to be read.
Cameron and others object that there is epigraphic evidence that his plays
were performed and that we must thus construe only as
better read than acted; understood thus, the word would still allow that
his plays were originally meant for the stage. But Cameron does not men-
tion that the inscription (TrGF I 71 T5) and the one performance to which
it refers (a production of Achilles, the Slayer of Thersites) are from the third
century b.c., that is, well after Chairemons lifetime. The notions that it
must have been performed within his lifetime as well, or that it was origi-
nally intended for performance (and that we must shape our understand-
ing of Aristotle accordingly) are just inferences from that one testimonium
concerning a single play. Why should we not presume instead that a play
rst meant for reading was eventually performed? The second part of
Goethes Faust was certainly never meant to be staged, yet it has often
been performed in this century. As to Chairemon, the evidence of his
acrostic and the characterization by Aristotle showsas in the case of
Aratusthat reading gured in the authors plans for the reception of his
work and that from the earliest moment we can traceand again, it is
very earlythat reception was expressly associated with reading.
It is not that Cameron sees no role at all for reading in the reception of
Hellenistic poetry. Rather, he sees it as distinctly secondary, even for the
scholarly elite and still more so for others. Michael Grants assertion that by
the third century, people were reading much more than they were sitting
and listening (1982: 260) mighton Camerons viewbe true of Calli-
machus and his colleagues at the museum, if we judge by the thousands of
manuscripts they catalogued. But of the educated public at large, he
continues, it is much more nearly false (44). Yet despite the proliferation
of agonstic festivals and despite the ongoing importance of symposia as a
place for reciting verse, on most days of the year in any given locality, if that
educated public at large that Cameron invokes wanted a bit of poetry, it
probably read it.
Cameron himself points to a text that suggests this (though his intent is
actually to show that poets were not just writing for an elite). I am refer-
ring to the third-century b.c. Lille papyrus of Callimachus Victoria Bere-
nices (SH frs. 25469). This, our earliest text of Callimachus, is remarkable
for its interlinear commentary, which breaks into the poem at irregular in-
tervals andthough written scarcely a generation after the poets death
offers the most elementary information on the preceding verses. Here,
says Cameron (56), is a concrete illustration of the sort of aids with

which people of modest cultural attainments might tackle so difcult a
poemEgyptians who needed to be told that Berenice was not really the
daughter of Philadelphus. But while, as Cameron points out, the papyrus
shows the reach of Callimachus poetry far beyond the ivory tower, it also
shows that even in the backwoods of the Fayum, people with a quite
humble level of education would nonetheless read an expressly occasional
poem like the Victoria Berenices. For this papyrus does not lend itself to
performance or even to reading out loud. Its very format, in which com-
mentary interrupts the poem at odd intervals, in midsentence and even
midclause (cf., e.g., lines 18ff. of the papyrus), militates against recitation.
Of course, there were plenty of musical festivals in the Fayum,9 no dearth
of symposia, and it is possible to imagine some connection for our pa-
pyrus. For instance, was its owner preparing to hear the epinician at a fes-
tival? Was he informing himself so as to be able to answer questions upon
reciting it at a symposium? Had he been bafed by the poems oblique
allusiveness when hed heard it performed and so decided to study it after
the fact? Any such scenario, which maintains the possibility that the poem
was primarily encountered in performance, is conceivable (I suppose). But
the fact remains that the Lille papyrus, our earliest text of Callimachus,
proves only that the Victoria Berenices was readand in all likelihood not
read out loud.
Just as Cameron attempts to undercut the idea of a Hellenistic book
culture, so, too, he denies that poetry of earlier ages had a more oral, im-
provisational character. That, according to Cameron, is only to romanti-
cize the otherness of archaic poetry (72). On the contrary, writing (he
says) was as essential a feature of poetic composition in the archaic and
classical periods as it was in the Hellenistic: Even in the archaic period
many symposiasts used written song-books (84).10 The fact that archaic
and classical poets do not refer to themselves as writing their songs
while that technology is so prominent and explicit a feature of the Hel-
lenistic poets self-imageis deemed insignicant: since the earlier poets
clearly did write, it doesnt matter that they do not talk about it. We rec-
ognize here the old Cameron two-step: step 1, earlier poets employed
writing every bit as much as their Hellenistic counterparts; step 2, if they
describe the process of composition only within an oral framework, that is
9. Cf. Koenen 1977, which deals with a festival in the Arsinoite or Herakleopolitan
nome.
10. One would love to know what evidence there is for this assertion!
not important. John Herington might attempt to understand the Aristo-

phanic scenes in which Euripides and Agathon are portrayed composing
tragedies without recourse to writing by suggesting that in the archaic and
classical periods writing was not essential but secondary, occurring only
after the crucial phase of composition was over (1985: 4647). But to
Cameron, that distinction is meaningless: it may be, as Herington claims,
that Euripides did not take to his pen . . . until late in the composition,
but that says nothing about a supposed predominance of oral style or cul-
ture. That is how Gibbon wrote; indeed, says Cameron (87), I write
that way myself. Besides comparing apples and oranges (ancient poetic
composition in a traditional genre and modern scholarly prose), the juxta-
position of Euripides, Gibbon, and Cameron is simply breathtaking. Judg-
ing from the length and complexity of his book, it must have been a hectic
week when Cameron nally put it all on paper. One can only hope that on
the seventh day he rested!
I want to turn nally to epigram, a genre that for many scholars has
embodied pure book-poetry (76) but that for Cameron is a further in-
stance of performance poetry tied to a certain social setting. In his view,
epigram was the poetic form par excellence of the Hellenistic sympo-
sium (100). He thus revives R. Reitzensteins proposal (1893: 87104) that
poets could compose epigrams extempore at symposia, often reacting im-
mediately to other such poems composed and performed by their drinking
companions.
In his chapter The Symposium (71103), Cameron spreads before us
a truly dazzling array of texts testifying to the continuing vitality of the
symposium and to its importance in Hellenistic times as a locus of poetry.
Still, in all the rich catalog of sources describing performance at such gath-
erings, there is not one that explicitly mentions the recitation of an epi-
gram in this period.11 For that reason, we still should be very cautious in
assessing the frequent representation of sympotic situations in epigram.12
As with Callimachus hymns, we must ask to what extent the description of
11. Even the Elephantine papyrus (pp. 19092 Powell), which Cameron hails as one of
the most interesting of all extant symposium texts and the clearest proof that singing [at
the symposium] was not entirely a thing of the past (1995: 74), does not actually constitute
proof for sympotic performance of epigram in the Hellenistic era. Indeed, the nal epigram-
matic poem servesas Powell himself noted (ad loc.)Scoliorum tanquam 6 and
so is evidence of editorial arrangement for the scroll.
12. Cf. Gutzwiller 1998: 11516: The imitation of oral speech, no more guarantees that
a sympotic epigram was recited than the imitation of written speech guarantees that an inscrip-
tional epigram was inscribed. Reitzensteins fundamental mistake was to confuse the represen-
tation of a speech act with the speech act itself. See, further, Bing and Bruss 2007: 1214.
an occasion may be a literary evocation, to what extent it points to the

actual circumstances of performance.
I would not rule out the symposium as one possible setting for poetic
creation and performance. But it is scarcely sufcient to explain the range
and quality of the epigrammatic corpus as we have it or the pointed re-
sponses of one epigram to another. There is a basic unlikelihood that the
poets in question always happened to attend the same parties, and it is hard
to believe that (without prior warning that such and such a theme would
be on the evenings program) they were able to produce extempore even
preliminary versions of many of their nest epigrams.
Reitzenstein (1893: 1034 n. 1) and Cameron both adduce the model of
Catullus 50 as the best commentary (88) on poetic creation and perfor-
mance at symposia. Here each participant writes versiculos now with this
meter, now that / capping each others jokes and toasts (vv. 46, trans. G.
Lee). Concerning these verses, Cameron observes: Typically (of course)
symposiasts sang others poetry. But the poets themselves (whether Hel-
lenistic or archaic) will naturally have jotted down ideas while still uncer-
tain about the form or details of a poem taking shape in their mindsand
also so as to have a record of their improvisations, in the hope of working
them up into something more polished one day (88). But does this really
provide a valid parallel to the epigrams of the Hellenistic poets? What sur-
vives of the occasion Catullus describes issignicantlynot the versiculos
but the brilliantly nished poem (n.b. poema in v. 16), which we have be-
fore us and which the poet says he wrote in the privacy of his home the
night after (vv. 1416). The nished poem is a plea for another meeting,
that I might be with you and talk (ut tecum loquerer simulque ut essem, v.
13). The implied situation is that speaker and addressee are not together
and that the poem is sentlike a letterto communicate his desire. Else-
where in Catullus (poem 14) we nd an anthology of verse sent to the poet
by the same Gaius Licinius Calvus who is the addressee in poem 50, evi-
dently read in private and then sent back with a response.
This situation may provide a useful model for understanding the allu-
sive interplay of Hellenistic epigrams. Cameron is right (if a bit literal-
minded) to insist that there is no ancient form of which it can be said with
less plausibility that it was written for the book. The average book-roll
contains 700 to 1000 lines, whereas some of the nest epigrams consist of
only two to six lines (77). Callimachus, he points out, wrote epigrams
throughout his life. . . . Did he publish new books every few years, or just
one towards the end of his life? (78) But that does not mean that epi-
grams only, or chief, mode of publication was through performance at
symposia, rather than in written form. I consider it likely that letters (or
short scrolls?)13in addition to presumably far rarer epigram books
served as a critical medium by which epigram was disseminated, thus per-
mitting authors to react to and vary the epigrams of their contemporaries
(as we know they loved to do) in works of their own.14 The letters of
Archimedes (in Sicily) to Eratosthenes and Dositheus (in Alexandria) may
serve as a model for the sort of lively cultural exchange that could occur
between far-ung correspondents in the third century b.c. Signicantly, in
one such letter, Archimedes sent Eratosthenes his cattle problem in the
form of an epigram (SH 201).15
In addition to this report about Archimedes poem sent in a letter, we
possess an actual papyrus letter whose exclusive purpose was to convey to
its recipient two epigrams. I am referring to the two anonymous funerary
epigrams (SH 977) sentpossibly in the poets own handto Zenon, agent
of Apollonius, chief nancial administrator of Ptolemy Philadelphus. They
celebrate Zenons hunting dog Tauron, who perished while trying to save
his master from an attacking boar, and both were probably intended for in-
scription on his tomb. If such poems as these were circulating by letter, it is
perfectly plausible that many others were as welland not just epigrams.16
It might be an interesting point of departure for a future study to ask
whether the later genre of epistolary poetry had a basis in actual practice.
13. Cf. Turner 1968: 140.

14. Cf. Ludwig 1968; Tarn 1979.
15. Eratosthenes dedicatory epigram for a mechanical device (mesolabon) solving the
Delian problem of doubling a cube (35, p. 66 Powell) is likewise transmitted in a letter to
King Ptolemy (Eutoc. comm. in Archim. sphaer. cyl. pp. 88.496.27 Heiberg-Stamatis). This
letter has often been viewed as a late forgery, but both letter and epigram have recently been
defended as authentic by Geus (2002: 195205). According to the letter, Eratosthenes dedi-
cated a stele with a representation of his device, an abbreviated version of his proof, a diagram,
and thereafter his epigram inscribed on it (/ " + 6). Eratosthenes says that he
has written these things as well in the letter so that you [sc., King Ptolemy] may have them as
on the dedication (# "
, G (J E 6 M
, 96.13
15). In other words, the letter provides ease of consultation and the possibility of repeated
viewing, which the inscribed monument (and, a fortiori, oral performance) does not.
16. The proem to Nicanders Alexipharmaca, for instance, is very suggestive of how po-
etic texts traveled in written form from place to place. Here the poet, in his native Clarus
(near Colophon), assures his addressee, Protagoras, far off in Cyzicus, that although a great
space separates us, I can easily tell you the remedies for poison drinks (, . . . +
6, / 8 / " /, vv. 35).
That ease of communication for Nicander is most likely produced by his ability to send the
poem in written form to his addressee. The mode of dissemination evoked in the proem for
this particular instance becomes paradigmatic for readers of the Alexipharmaca generally,
wherever they happen to be.
chapter 7
The Un-Read Muse?

INSCRIBED EPIGRAM AND ITS
READERS IN ANTIQUITY
I n Frank Capras 1939 political satire Mr. Smith Goes to Washington, there
is a scene that stages the act of reading an inscription and presents it in
particularly vivid paradigmatic detail. The movies idealistic young hero,
Jefferson Smith, memorably played by Jimmy Stewart, has become the
surprise choice of the corrupt party bosses of his native Montana to re-
place the states junior senator, who has suddenly died in ofce; Smith,
they hope, will be an easily manipulated stooge for the big-money inter-
ests of the state. But shortly after his arrival in Washington and presaging
his later troublesome independence, this callow, hayseed senator man-
ages to slip away from his handlers to see the sights of the nations capitol.
Wide-eyed with awe, Smith travels past the White House and the Wash-
ington Monument and nally comes to the Lincoln Memorial. Here he
pauses attentively before the seated statue of Abraham Lincoln, then turns
to see the great inscription of Lincolns Gettysburg Address, which towers
above him, its monumental lettering with dark highlights easily legible to
the movie audience. Beside him, he notices a little boy holding the hand of
This is a revised and updated version of an essay that appeared in Hellenistic Epigrams, ed.
M. A. Harder, R. F. Regtuit and G. C. Wakker, 3966. Hellenistica Groningana 6. 2002. Lou-
vain. 2002 Peeters. Heartfelt thanks to Rip Cohen and Prof. Jongsook Leeexceptional
readers bothfor their help and encouragement at critical points in the genesis of this essay.
116
The Un-Read Muse? 117
Fig. 2. Jimmy Stewart as Jefferson Smith, at the Lincoln Memorial, in Frank

Capras Mr. Smith Goes to Washington. (Courtesy of Columbia Pictures.)
an elderly man. The boy reads the inscription aloud to the frail old man,
who listens with rapt attention, occasionally helping the boy with unfamil-
iar words. They are joined by a black man who doffs his hat and reads
along as well. Smith watches them, as do we, the audience, looking over
their shoulders to Lincolns words. It is hard to imagine more ideally fo-
cused or receptive readers of an inscribed text.
In his autobiography, Capra relates how he took this scene directly
from life, when heamong dozens of other tourists reading the engraved
wordsobserved a little boy reading the Gettsyburg Address, as well as
the Second Inaugural, to a very old man. Never, says Capra, had Lin-
colns impassioned, moral indictment of slavery sounded so eloquent, so
moving, so powerful as when that young boy read it to his grandfather.
That scene must go into our lm, I thought. We must make the lm if
only to hear a boy read Lincoln to his grandpa (Capra 1972: 25960). As
Capra says, there were dozens of tourists reading the speeches carved on
the walls. Countless visitors had done the same before, and countless have
done so since. This helps explain why this moment in the lm is apt to
strike so resonant a chord in an American audience.
What Capra experienced in encountering this inscribed textwhat he
allows us to experience, peering over the shoulder of his readersis a
thing many Americans would nd familiar. In observing this act of reading
the inscription in the movie, we in fact observe ourselves as readers of in-
scriptionsas a nation of inscription readers (though a few of us may be
slightly more jaded than the readers in the lm). For with these texts, the
memorial constitutes itself as a national charter, embodying some of
Americas understanding of itself and shaping its self-denition. In this it
resembles that ur-charter of a nation, likewise carved in stone by the orig-
inal poet for his people to read, to remember, and to live by: the Ten
Commandments.
Descriptions of people encountering inscribed monuments, pausing to
read and reect on them, are common and privileged in narrativeafter
antiquity. They recur steadily, from Dantes Inferno, where readers (to-
gether with the narrator and his guide, Virgil) come face-to-face with the
inscription on the gate of hell and participate in its exegesis (Inferno III vv.
118),1 up through modern times.2 But do they occur in ancient narrative
as well? To judge by epigram, the genre actually or ctitiously inscribed on
ancient monuments, one would certainly expect they would. For from ear-
liest times, these poems insistently envision and dramatize that encounter
between reader and monument. Again and again the texts call out to the
1. Like the narrator, we scan the verses of the famous epigram ending with the words
Abandon all hope ye who enter here (v. 9) and participate in his perplexity as he tells his
guide that for him the text is duro (v. 12)presumably, difcult to understand, tough to
take, as well as cut in stone. We hear the response of his teacher, Virgil, who is described
as persona accorta (v. 13), meaning agile, intelligent, attuned, a reference both to his
general wisdom and, more specically, in this context, to his competence as a reader. Virgils
answer conrms that characterization. He is indeed an attentive reader, echoing yet reshap-
ing the words of the inscription: abandon all distrust; all cowardice should perish here (vv.
1415). The scene thus describes not only the confrontation with the inscribed text but the
process of its receptiona reception all the more problematic as it is adapted within the text
to a distinctly unimplied reader, i.e., one who does not fulll the prerequisite of being dead.
But if unimplied, that reader is nonetheless ideal. For he brings that indispensable (if rare)
quality of attentiveness: he lavishes care on his reading.
Cf. also the fourteenth-century Le livre du cuer damours espris, by Ren d Anjou, where
the heart goes about reading texts inscribed on walls, etc., including its own tombstone. De-
scriptions of people contemplating inscriptions seem to have enjoyed a particular vogue in
fourteenth-century literature.
2. Among many instances, see, especially, Poussins inuential painting Et in Arcadia Ego
(with Panofsky 1955), or Cavafys poem In the Month of Athyr.
readerb /, b
, C, or the likeinviting him to stop
and pay attention, acknowledge some achievement, utter a prayer, deliver
some message. In this sense, they are meant to be interactive texts; they
presuppose audience participation.3
One might reasonably expect, therefore, to nd numerous depictions
outside epigram (in ancient epic, drama, history, oratory, philosophy, the
novel) of people pausing before inscriptions to read and ponder thema
kind of ecphrasis of the inscribed monument comparable to the sort so
common across various genres, where characters stop to contemplate a
work of art (Goldhill 1994). Such ecphrases of inscriptions would prove an
interesting yardstick against which to measure the imagined encounter
within the epigrams themselves.4 But they simply do not existnot, that is,
until late: the earliest appears to be from Nonnus in the fth century a.d.5
Indeed, an examination of the evidence leaves one rather wondering,
did anybody actually read inscribed epigram, and if so, who? Oddly, there
is scant indication that anyone did. And here there is a poignant contradic-
tion between the elaborate devices whereby the text tries to involve the
reader and the surprising lack of reader response. For although much of
epigram is constructed like a trap for a reader, it appears that very few
people in antiquity ever stopped to take the bait. Those that didas I sug-
gest in this essaywere in all ways exceptional persons, at least up through
the early Hellenistic period. An ordinary reader of inscriptionsthe
kind of everyman Frank Capra imagined at the Lincoln Memorialdid,
I think, appear, linked with epigrams development in the Hellenistic era
into a genre composed or collected for the book. In that form, suddenly
accessible to a wider audience, it came to be used even in schools. And with
that institutional setting as its springboard, I suggest that book epigram
3. Cf. Svenbro 1993: 44: Just as he foresees his own absence, the writer foresees the
presence of his writing before the reader. The reading constitutes a meeting between the
reader and the written marks of someone who is absent. The writer foresees that meeting,
plans it carefully. He counts on the reader and the reading aloud that the reader will accom-
plish, for in a culture in which klos has a fundamental part to play, what is written remains
incomplete until such time as it is provided with a voice.
4. In a sense, of course, any artwork is a text, and its viewer is a reader. Why limit
oneself to inscriptions, then, in looking for an independent gauge for their readers response?
Thus, e.g., Goldhill (1999) includes the reading of artworks when he poses the question in
his title Body/Politics: Is There a History of Reading? But given the historical tension
between oral and literate culture in ancient Greece and the fact that the technology of writ-
ing only gradually acquired authority and prestige in the course of the classical age, it is
reasonable to operate with a narrower sense of these terms and focus here on the readers
experience of inscribed epigram in particular.
5. See the close of this chapter for a detailed discussion of this ecphrasis.
was in turn able to generate new interest and a different kind of readership
for the inscribed variety. Prior to that, however, if there were a goddess of
inscribed epigramthe Musa Lapidaria of Courtneys 1995 titleone
would be tempted to call her the Un-Read Muse.
Even early on, of course, there were exceptions. For not all inscriptions
are created equal. In every culture, different sorts of inscribed text elicit
varying responses. Or from the readers perspective, ones willingness to
engage a text will vary according to its kind. These varying degrees of re-
ceptiveness to a text are, of course, operative in all reading, across the spec-
trum of literary genres from lyric poetry to legal contract, not just in epi-
gram. And there are differences within genres as well. The inscriptions at
the Lincoln Memorial are special cases inasmuch as many of those en-
countering them are familiar with the text from having read it elsewhere
(very likelyand signicantlyas a set text in school, a locus for the read-
ing of epigram to which I will return toward the end of this essay): the cul-
ture thus bestows on them an exalted status.6 In antiquity, the Delphic
maxims engraved on Apollos temple already in the sixth century b.c. en-
joyed a comparable status and garnered similar attention.7
At the other end of the cultural hierarchy stands another exception: the
humble grafto. Though commanding none of the prestige of the Delphic
maxims, this sort of inscription sometimes manages to arouse readerly in-
6. Another instance is Emma Lazarus The New Colossus, engraved in the pedestal of
the Statue of Liberty. Lazarus originally wrote the poemwhich contains the famous lines
Give me your tired, your poor, / Your huddled masses yearning to breathe free . . .in
1893 to aid fund-raising efforts for the statues pedestal. Thereafter, the poem languished in
obscurity until Georgina Schuyler rediscovered it and led a campaign to bring it back to pub-
lic notice. As a result, the poem was inscribed on a bronze tablet in 1903 and dedicated inside
the pedestal of the Statue of Liberty, from whence its impact has stretched far beyond Liberty
Island (Moreno 2000: 17375, s.v. The New Colossus). For Lazarus poem as school text,
cf. the Public Broadcasting Services Web site Learning Adventures in Citizenship: Raise a
New Torch, http://www.pbskids.org/bigapplehistory/activities/a_immigration/activity2/
index.html.
7. They functioned as the gods salutation to those entering, instead of , as Plato
has Kritias say at Charm. 164d: "

,
(cf.
also Protag. 343ab). On their location, cf. Paus. X 24. On their impact, cf. Plato Charm.
164d, as well as the report that the Hipparchan herms were set up in competition: G
+ 6 " ,
+
, ; 8
? (ps.-Plato Hipparch. 228e; discussion of the account later in the present essay). See,
further, the comment made in Plutarch De E apud Delphos (385d) in the very presence of these
inscriptions: @ + , + , @
@ / 2
. Indeed, the entire dialogue consists of an on-site discussion of the E inscribed
at the temple.
terest. Because it can be so low, so deliciously personal, unfettered by deco-

rum, exploiting safe anonymity to indulge in virulent, often obscene ad
hominem attack or to extol an object of desire, the grafto pricks our
curiosity and incites us to answer. In ancient times, as now, grafti were
read. We know that because their readers not infrequently talk back,
engraving their retorts close by.8 Precisely that kind of response is vividly
portrayed in Aristophanes Wasps (vv. 9799) when the household slave
Xanthias, describing Philokleons passion for jury duty, says that if he sees
the name of Pyrilampes son written on a door as , he goes and
writes next to it [i.e., beautiful voting funnel] (X :J
/ 6 Y , / ,5
). Such grafto repartee appears even among the very
earliest Greek inscriptions, for example, in seventh-century b.c. (?) Thera,
where we nd on a single boulder a series of erotic grafti in which each
writer tries to cap the previous one(s) in praising a beloved.9
But while readers may have been receptive to certain kinds of inscrip-
tion as a result of cultural conditioning or personal inclination, it appears
they reacted to the great mass of such texts differentlyor, better, indif-
ferently. Writing of epitaph, George Walsh (1991: 94) comments that a
basic precondition [of the genre] is the readers indifference. To be sure,
Walsh considers that most readers would acknowledge the epitaphs
information hastily and move on without thinking about it (95). But I
believe that a more radical indifference was at work and that it extended
beyond epitaph to votive and honoric epigram as well. As Rosalind
Thomas (1989: 35) has noted in connection with public stelae of the late
fth- and fourth-century Athens, It is not clear that Athenians actually
read inscriptions much.10
One reason for readers indifference lies in the nature of the inscribed
8. As Hellenistic authors of book epigram responded to poems of their contemporaries

or predecessors.
9. IG XII.3, 540; cf. also 536. See the discussion of B. Powell (1991: 17180, esp. 17476,
17980). Cf. also the grafti recently found by Merle Langdon near the coastal road to
Sounion, which were inscribed by sixth-century b.c. herdsmen. Later, see the Greek grafto
from the Domus Tiberiana on the Palatine, where the response to earlier grafti is sweeping
and witty: Many have inscribed many things, I alone have written nothing (
6 / 65 _ (). Cf. Castrn and Lilius 1970. A comparably generalized
grafto answer is the elegiac couplet found at several locations in Pompeii in a variety of
hands: admiror te paries non cecidisse ruinis / qui tot scriptorum taedia sustineas (CIL IV 1904, 1906,
2461, 2487; cf. Franklin 1991: 8283). For more grafto repartee at Pompeii, cf., e.g., CIL IV
2175 (hic ego puellas multas / futui), 2176 (Felix / bene futuis); Franklin 1991: 8284, 8889).
10. For the view that inscriptions were, on the contrary, widely read, see, most recently,
the chapter entitled Consultation in Sickinger 1999: 16087.
monument: it is stuck in one place. The stele is a gure for immobility al-
ready in Homer, where it twice appears in that capacity in similes (of a
warrior unable to move at Il. 13.437 and of the horses of Patroklos unwill-
ing to move at 17.43233). Rooted to the spot, it has to wait, relying on
the uncertain prospect of a literate person (not just any viewer) rst of all
seeing and then taking the trouble to read it. To initiate communication is
beyond its competence. Unlike a scroll, moreover, which can accompany a
traveler wherever he goes, the monument has only a eeting chance: the
moment the viewer comes before it.11
Add to this the physical location of the inscription: epigrams of the ar-
chaic period were placed not at eye level on their monuments but low
down. Classical epigrams tended to be placed higher (Clairmont: 7, 46).
But a painting on the tondo of a red-gure cup by the Ancona Painter (g.
3, p. 123, ca. 500480 b.c.), which shows a youth stooping over so as to read
a stele, leaves no doubt that inscriptions were not designed for ease of in-
spection.12 The famous marble relief of the mourning Athena (g. 4, p. 124,
ca. 460) from the Athenian Acropolis, which has sometimesimplaus-
iblybeen interpreted as depicting the goddess reading, gives a fair notion
of how laborious that would be. That impression of difculty is conrmed
for the Hellenistic period in Herodas fourth mimiamb, which describes two
women making an offering at a shrine of Asclepius and seeing the sights.
When one, Kokkale, asks her friend Kynno who was the sculptor and ded-
icator of a particular statue, Kynno replies, Dont you see the letters on the
base? (" CJ / 6 J /;). That is, Kokkale needed
prompting to look below. The inscription was not immediately obvious.13
11. In an important article, Scodel (1992: 71) argues that the wayfarer who cannot read
need not nd a literate helper to read it [sc., the inscribed monument] to him. Rather, because
its metrical form was an aid to memory, she claims, inhabitants near an impressive tomb will
surely have known the inscription by heart. Scodel does not, however, cite any evidence for
such communal familiarity. This is not to deny that a few exceptional epigrams, such as that by
Simonides on the dead at Thermopylai, were remembered orally beyond their monuments.
12. Illustrated in Boardman 1989: no. 79. Svenbro (1993: 19495) argues that the boys
posture communicates that he is ready to be buggered. His position is that which, in
iconography, suggests katapgn. Reading a stele, then, exposes you to indignity.
13. And when, a few lines on, Kokkale in turn exclaims, Kynno, dont you see this statue
of Batale, daughter of Myttes, what a gait it has? ( , " CJ,
, / @ , ;), she not only echoes the words of her com-
panion (" CJ " CJ, ) but also mimics her act of readingor, rather,
since the names she mentions mean, respectively, Fucked-up-her-ass and Cunt, perhaps
she is making them up, as if she was reading. Indeed, one may wonder whether Kokkale is
illiterate and therefore had to ask her friend about the rst statue.
For another scene where a reader does not see an inscription and must have his attention
drawn to it, cf. Glaukos 3 GP AP 9.341 where the nymphs direct Pan to a couplet inscribed
Fig. 3. Youth reading stele. Tondo of Attic red-figure kylix, Ancona Painter,
500480 B.C. Corpus Vasorum Antiquorum 16.2 no. 62 ( Copyright by
Hessische Hausstiftung, Museum Schloss Fasanerie, Eichenzdell/Fulda
[Germany].)
Indeed, on sculptural monuments or votive objects at least, it is safe to say

that inscriptions were in general secondary features of the monument, not
meant to strike viewers at rst sight or excite their attention.
A further obstacle might be weathering, the corrosive effect of which
on a stele is acknowledged by no less an expert on engraved monuments
than Simonides (PMG 581), the poet most renowned for his work in this
for him on a tree by Daphnis. It is unclear whether the nymphs or the god himself then reads
the inscription aloud (cf. GP ad v. 4); if the former, it may cause us to wonder again whether
we are supposed to imagine the god as illiterate.
Fig. 4. Pensive Athena, votive relief from the Acropolis, Athens, ca. 470450
B.C. (Courtesy of Scala/Art Resource, NY.)
medium: What man in his right mind would commend Kleoboulos, in-
habitant of Lindos, who set the strength of a stele against ever-owing
rivers, the sprouting plants of spring, the ame of the sun, the golden
moon, or the wash of the sea? All things are weaker than the gods. Stone is
shattered even by mortal hands. This is the devising of a fool (
, , /
/ , / /
/ ; / 9 6 \D + /
D / 9 ).14
Shifting from physical impediments to those of content, there is the
very sentiment of epigram: it is thoroughly hacknied, its poetic expression
largely formulaic. Even poems engraved at great cost on imposing civic or
private monuments are apt to be numbingly conventional.15 Pervasive ref-
erence to epic diction and values notwithstanding, they share nothing of
epics ability to divert and enchant. Consequently, and despite their pleas,
a reader is likely to ignore them.
Finally, there is the fact that a monument has no standing by which to
stake its claim on the readers attention. A person may appeal to another by
reference to common bondsfor example, Friends, Romans, country-
men, lend me your earsor to the mere fact of shared humanity. But a
monument is inanimate, with no more voice than a dead man (conversely,
for Theognis, a dead man is mute as a stone [O / , vv.
56869]). That deciency is limned in an epigram by Aratus (AP 12.129
1 GP) where the speaker evokes the handsome youth Philokles of Argos,
celebrated in inscriptions from Corinth to Oropos (evidently he de-
served his name, whether it means beloved by fame or in love with
renown).
/ P , G
" D
/
E . / 4 D
/ " 6,
M
" ,, 2 / 6 .
2 " Brunck : P 4 4 Brunck : 4 P
14. An instance of an epigram half worn away on its monument will concern us later in
this essay.
15. This is apparent in the ease with which they may be categorized according to rhetor-
ical type, as in W. Peeks arrangement in his Inhaltsbersicht (1955: xixxxii).
[Argive Philokles is gorgeous in Argos; Corinths

columns declare the same, Megaras tombstones too.
As far as the Baths of Amphiaraos you can read
that hes gorgeous. But all we lack are a few letters.
For it is not stones that attest to this fellow, but Prieneus,
who saw for himself: he is superior to another.]
Inscribed monuments conventionally speak ( , v. 2) and bear wit-

ness (6, v. 5),16 but this epigramby dropping the functional
terms and (v. 2) and frankly calling these stones stones
(, v. 5)insists nally on the irreducible difference between inani-
mate object and sentient being and so compels readers to acknowledge
that such expressions are merely metaphors (" 6,
/ " ,, vv. 56). To be sure, the speaker depicts him-
self as a reader of such inscriptionsa person of the sort mentioned before,
who in the course of traveling reads grafti, prompted perhaps by their
prurient appeal. Maybe we are even meant to imagine that the in-
scriptions inspired him to see for himself what all the fuss over Philokles
was about.17 Nevertheless, the speaker insists on the insufciency of stone
as a witness. For he has something better: though he may lack the en-
graved letters of the inscriptional medium (4 , v.
4), he has the physical presence of his beloved (hence deictic M / in v. 5),
whose surpassing beauty he can attest to viva voce and as an eyewitness.18
16. For a discussion of speaking objects, see, e.g., Svenbro 1999, esp. the section enti-
tled The I and the Voice (4650).
17. Do the inscriptions in Aratus epigram imply that Philokles was once present
for amorous encounters at the places mentioned: temples or stoas in Corinth, cemeteries in
Megara, baths in Oropus?
18. Contrary to earlier critics, I propose that Prieneus is the speaker throughout the
poem, not different from the rst-person plural in v. 4. Moreover, M / in v. 5 may
refer to Philokles, the subject of the rst four lines, rather than to a different and, in that case,
strangely unnamed youth, as most critics would have it (for the deictic pronoun referring to
the same subject, rather than a different one, cf. already the inscription on the Ischia Cup
(CEG I 454): [,] _ / h / [ [] . . . ). The
point, then, is that Philokles is there in the presence of the speaker, who is able to assess his
charms rsthand. On this view, too, (v. 5) acquires added point, whether as a proper
name or an ethnic: for it suggests that this speaker takes Philokles fame beyond the coastal
strip between Corinth and Oropos, extending it at least as far as Priene in Asia Minor. Of
course, the epigram slyly plays with the fact that it says little more about the beloved than
would a inscription and conveys Prieneus testimony to us by means of another written
medium, the papyrus (the statement in v. 5 that it is not stones that attest raises the ques-
tion of just how Prieneus testimony is conveyed and suggests the advantage of papyrus in ex-
tending the poems reception far beyond Priene).
A persons authority thus clearly trumps the prerogative of a stone, and

an inscription must rely on the uncertain prospect that a passerby will
himself summon up the interest to stop and read. But will the reader do
so? Will he make the effort? It is fashionable to think that he will. For ex-
ample, Svenbro (1993: 47) condently speaks of how the ancient reading
of inscriptions takes the form of an exercise of power over the voice of the
reader: the voice has to submit to the written word, becoming its in-
strument of communication and implying a reader dispossessed of his
own voice. Mary Depew (1997: 239, 245), following Joseph W. Day
(1989: 2628; 1994; 2000), speaks of the inscription readers (re-)activa-
tion of the original act of dedicating the monument, claiming that the
whole ritual occasion is brought repeatedly to life in the here-and-now,
whenever anyone reads the text.19 Yet Svenbros and Depews readers
display a willingness to engage the inscribed text, which scarcely exists in
the ancient evidence. Leaving aside modern scholars attempts to imagina-
tively construct the experience of the reader, our sources bespeak, rather, a
pervasive indifference.
Paradoxically, an important aspect of this indifference may be seen al-
ready in Homer.20 This poet does not, of course, present us with inscrip-
tions. But his account of grave monuments is revealing. At Iliad 7.67ff., for
instance, when Hektor proposes that one of the Achaeans meet him in
single combat and that the victor should return the others body to be
buried, his proposal turns into an elaborate vaunt in which he imagines his
adversarys tomb and its reception by future generations.
+ 6 6 ,
= 2 /, 85
6 ;M.
:J 4
q 6 : D
+ ,
@ / K. 90
O 6D / 6 _ / 4.
19. The term (re-)activation occurs in Day 2000. Cf. also Kurke 1993: 146: [in an
athletes victory monument] the passerbys reading aloud and gaze recreate the original an-
nouncement and crowning of the victor . . . perpetually regenerating his kudos. See, further,
Day 2007.
20. Some of the same points about characters indifference to actual tomb monuments in
Homer are covered by Scodel (1992: 66).
[His corpse I will return among the well-benched ships

so that the Achaeans with the owing hair may duly bury him
and pile up a tomb for him beside the broad Hellespont.
And one day a man among those still to come will say
as he sails by upon the wine-dark sea in his ship with many oars,
This is the tomb of a man who died long ago,
one of the bravest, and glorious Hektor killed him.
Thus someone will say, and my fame will never perish.]
The verses near the end of this passage, imagining the speech of the per-
son passing by the tomb (This is the tomb of a man who died long ago, /
one of the bravest), bear a close resemblance to the diction of inscribed
sepulchral epigram,21 and that similarity extends to the concern with the
response of future viewers. Like Hektor, the author of an ancient epigram
conventionally projects the encounter with the monument into the here
and now of a future reader:22 the commemorative text is perpetually pres-
ent, available, and repeatable by potential viewers.
One may doubt, of course, whether a tomb would truly continue to
function as a memorial to the slayer (!) of its occupant, the more so as it
does not offer the guidance of an inscription but is envisaged, rather, as
being seen from ships at sea, a location that would, in any case, not have
permitted close inspection.23 Nevertheless, it is typical of passages such as
these that a character imagines the response of a future passerby, rather
than describing an actual response at a present grave site.24
What would that actual response have looked like? The striking fact is
that when a preexisting tomb appears in Homer, no one ever responds to it
qua tomb. Rather, it serves merely as a landmark or focus of activity,25 such
as the tomb of Aisyetes at Iliad 2.79194, which is used as a lookout, or the
tomb of Ilos, used both as a gathering place (Il. 10.415) and as somewhere
to shoot fromParis leans against the stele on the tomb so as to steady his
aim and be able to hide (Il. 11.37172, 379).26
21. The resemblance has been variously explained. Cf. Lumpp 1963; Raubitschek 1968:
57; Svenbro 1993: 53; Scodel 1992: 59; R. F. Thomas 1998: 2057.
22. As Doris Meyer puts it (2005: 55) with reference to CEG I 162, Auffllig ist . . . die
explizite Ausrichtung des Denkmals auf die vom Erbauer oder Epigrammautor imaginierten
Betrachter und Leser.
23. For other Homeric tombs conceived as visible from afar at sea, cf. those of Elpenor
(Od. 11.7576) and Achilleus and Patroklos (Od. 24.8084). See Pearce 1983: 11015.
24. Cf. also the imagined tomb of Menelaus at Il. 4.16981, which a Trojan will insult
and jump upon, or the tomb of Achilleus and Patroklos in Od. 24.8084.
25. These categories are from Sourvinou-Inwood 1995: 129, 132.
26. Cf. also Il. 11.166, 24.349.
Just how precarious the act of memorialization can be is best repre-

sented in a passage from book 23 (vv. 326ff.) that shares with the speech of
the passerby imagined by Hektor in book 7 that epigrammatic phrase
. Here Nestor instructs Antilochos about how
to keep his chariot close to the turning point.
6 / , " .
I " @ / =/ #+ :
X X D + " =M,
+
2 6 5
6 J C , / 330
A ,
X 6 ,
/ ( /.
[I will show you a very clear marker [], which you cannot fail to
notice.
There is a withered trunk rising about six feet over the ground,
of oak or pine, which the rains have not rotted away,
and two white stones are propped against it, one on either side,
at the joining of the ways, and there is smooth running for horses on
either side.
It is either the grave marker [] of someone who died long ago
or was made to be a turning post in the time of earlier men,
and now swift-footed brilliant Achilleus has set it as the racing goal.]
The turning post, it seems, may have been the tomb of some unspecied
person from long ago, no longer remembered. The tomb may have become
simply part of the landscape. Now it has been marked out ( , v. 358)
by Achilleus as the point around which the charioteers will race, alienated
from what may have been its original function. The uncertainty about what
this marker was, its potential instability of meaning over time, is wonder-
fully conveyed by the semantic slippage between (a landmark) in verse
326 and (a gravemark) in verse 331. It turns out that this marker is not
so very clear, / , as Nestor claimed in verse 326.27
27. Cf. Nagy 1990: 20222, esp. 21517 of his chapter Sema and Noesis: The Heros
Tomb and the Reading of Symbols in Homer and Hesiod. Nagy focuses more on the pos-
sibility that the monument may be a tomb than on Nestors uncertainty about whether or not
it is one.
For another heroic tomb whose identity is mistaken, cf. the steep hill near Troy that men
call Batieia (Il. 2.81115) but that the gods know as the tomb of bounding Myrina. The
This is the pattern, then, in Homer: those who conceive or build sepul-
chral monuments envision future viewers to acknowledge and nd them
meaningful; existing monuments, however, are ignored or misconstrued. If
I am correct in discerning this pattern, we may well ask: does a person dis-
charge his duty to the dead in building a tomb or writing an epitaph when
he conceives that it might be seen; that is, is the crucial gesture the act
itself of imagining the monuments receptionan act of consolation both
for the dead man and the survivorand not its actual reception in the fu-
ture? I would argue that this is largely the case with inscribed monu-
mentsand not just with tombs. The meaning lies in the perspective of
those setting the monument in the present rather than in that of those who
will see it in the future.28
Nestors ignorance about the tomb should not surprise us, then. Ne-
glecting the dead has a long and glorious tradition.29 This perspective casts
human name is explained by Epaphroditos (quoted in Herodian De prosodia 3.1.277.25) as

coming from the path trampled by horses () at a turning post or the bramble (). In
any case, from a human perspective, it is viewed simply as a steep hill. It is only the gods who
recognize it as the tomb of Myrine. For a later heroic tomb whose identity is mistaken,
though still recognized as a tomb, cf. that of the Argonaut Idmon, which later generations
take for that of Agamestor (Ap. Rh. 2.84250).
28. In this, my position resembles that of Rosalind Thomas (1989: 49), who writes of
public inscriptions of the late fth and the fourth centuries b.c.: They . . . had a signicance
quite independent of whether they were read much by most Athenians. For, more than doc-
uments, they were also stone memorials or symbols of the honour, treaty or decision that they
recorded.
29. Archilochus, for instance, writes: No one once hes dead is held in respect among
the citizens, not even if hes very famous. Instead, the living chase the favor of the living; the
dead mans always got it worst (_ , / "+ 5 / D
+
/ , / , M , fr. 133
West). Semonides even suggests that if we were sensible, we would not think about a dead
man for more than a day ( + " [ 6, / : ,
? , fr. 2), for we have a long time to be dead, while we live a few miserable years in
number ( \ 6 , / /
(, fr. 3).
Cf. also Stesichorus (PMG 245), who says, when a man dies, all the goodwill of men per-
ishes ( / ); this is cited in Stobaeus under
the rubric the memory of most men fades quickly after death (@
? ).
Perhaps it was due to attitudes like these that, though a family typically underwent a pe-
riod of mourning for thirty days following the funeral, the care for graves subsequently be-
came part of an annual civic festival honoring the dead (cf. Burkert 1985: 194). How long
would people have recalled the date (or season) of a loved ones death or even their identity?
The span of living memory is, in any case, notoriously short. Historians conventionally as-
sume three to four generations, eighty years at most, before the last living witness to an event
has died, as well as those who would have heard his testimony as adults. Even with the sup-
port of a written epitaph, the meager data provided on a tomb will not sufce to keep alive an
image of the deceased with any detail. Regarding the span of living memory in traditional so-
cieties and the so-called oating gap that comes between it and accounts of mythic origins,
see Assmann 1999: 4852.
a harsh light on the convention in ancient epigram of asking the readers to

convey a message to a distant place.30 For instance, in an epitaph on a
limestone block of the rst half of the third century b.c. in Alexandria (GV
1353 no. 30 Bernand), wayfarers are asked to bridge the distance be-
tween Egypt and the dead womans hometown of Herakleia to report her
demise in childbirth.
;, C, A G,
,D b
A , / /D " 6
A 6.
[Travelers, if one of you should reach her native Herakleia,

say that pangs of birth led off the daughter of Polykrates
to Hades, Agathokleia. For they did not come upon her
easy as her child approached the light.]
Everything in this poem suggests movement. All verbs apart from the im-
perative say are verbs of motion: G, A, A, 6
(even the b are on the move!). It is as though the epitaph tries to spur
the desired movement through the suggestive power of its verbs. For of all
those plural wayfarers (C, v. 1), surely someones way will lead to
Herakleia! But which Herakleia? tienne Bernand comments (ad loc.) that
les villes du nom dHracle sont si nombreuses quil nest gure possible
de savoir quelle est celle dont est originaire la dfunte. Despite the com-
monness of such requests, there is, to my knowledge, no instance in an-
cient literature in which the hoped-for communication is carried out, no
narrative describing how a traveler, on reading such an epigram and arriv-
ing elsewhere, reports on his encounter with the text.31
It is instructive to compare a modern instance inspired by ancient epi-
gram and recounting precisely such a scene. I am referring to Shelleys
30. In addition to the famous b /, . . . , cf. the literary ex-

amples in Asclepiades (31 GP AP 7.500) and Callimachus (Ep. 12 Pf. 43 GP AP
7.521), with Dan Seldens ne reading (1998: 31415).
31. Richard Hunter suggests to me that one possible reading of Callimachus Ep. 2.1
(Pf.), , ;, would be to take it as a report of what someone saw
on a grave stele. And indeed, the phrasing in a rst-century b.c. stele from Rhodes (GV
1625.12) seems (anachronistically) to provide a model from which Callimachus words could
appear to have been quoted: 6 V+ / . (N.b.
that and 6 are in the same position.) But given Callimachus silence
about his informants source and given the unlikelihood that this person would have chanced
to stop before that very tombstone and read it, it seems more plausible to think that the
model for how he came by the news is that provided in the poem itself: word of mouth.
sonnet Ozymandias, whose traveler from an antique land describes a

monument, reports the text engraved on it, and turns out to be as attentive
a reader as those we met in the Lincoln Memorial.
I met a traveler from an antique land / who said two vast and trunkless
legs of stone / stand in the desert. Near them on the sand, / half sunk, a
shattered visage lies, whose frown / and wrinkled lip and sneer of cold
command / tell that its sculptor well those passions read, / which yet
survive, stamped on these lifeless things, / the hand that mocked them,
and the heart that fed. / And on the pedestal these words appear: / My
name is Ozymandias, king of kings: / look on my works ye mighty, and
despair. / Nothing beside remains. Round the decay / of that colossal
wreck, boundless and bare / the lone and level sands stretch far away.
This traveler has not merely stopped to read the text; he has thoroughly
studied the monument and observed its physical setting, his careful and at-
tentive examination clashing starkly with the monuments general neglect
and decay. To the modern sensibility (which we, as present-day scholars,
share in a fortiori), the neglected monument itself deserves a careful read-
ing. Evidently, too, he thinks this experience worth communicating to the
poems speaker and so to us. Instead of telling us what it all means, he de-
scribes the various elements of the ensemble. Yet the terms of his descrip-
tion leave no doubt that he has pondered the whole and constructed a
meaning, which points to the gap between powers insolent aspiration and
its futility in the face of time and human neglect.
The absence of any comparable scene in ancient literature is sobering.
In this light, how should we assess the common pleaas in the epitaph of
Agathokleiathat the reader convey to others what he has seen? Is it just a
topos, more conventional than real? Perhaps we can say so, if we are
careful to add that conventional does not suggest meaningless. For
here again it is pertinent to ask if what matters most from the standpoint of
those setting up the inscription is the act itself of making the requesta
gesture of respect toward the deadand not whether anyone carries it out.
After all, if a person went to the expense of purchasing a stone, composing
or commissioning an epitaph in verse, and paying for its inscription, can
we not assume he could resort to more practicalless painfully contin-
gentmeans to get its message to its hoped-for recipient?
Epigrams orientation toward an imagined encounter with a reader is
already present in inscriptions of the archaic period and is a long-estab-
lished custom by the time of the genres rst great master, Simonides. Sim-
ilarly, public inscriptions not uncommonly state that they are set up so that
anyone who wants may see them.32 Yet it has often been noted that au-
thors through the fth and into the mid-fourth centuries rarely cite in-
scriptions (epigrams or other types) and that when they do, they often cite
from hearsay, not from autopsy.33 There is thus a considerable lag between
the expectations conventionally expressed in epigram concerning reader
reception and our evidence that anyone actually fullled them.34 Granted
that this is largely an argument from silence, yet that silence weighs heav-
ier when we recall how low estimates are for literacy during this time and
32. Cf. R. Thomas 1989: 6061 with n. 151.

33. For a balanced overview for the classical period, cf. Higbie 1999.
R. Thomas (1989: 38) accounts for the dearth of reference by suggesting that a written
document may have an immensely important function even though it was seldom read. Its
function lies in its nonwritten aspect: Public inscriptions were regarded as memorials as well
as written documents. They therefore had a signicance quite independent of whether they
were read much by most Athenians. For, more than documents, they were also stone memo-
rials or symbols of the honour, treaty or decision that they recordedmaterial objects which
were a reminder and symbol of the decision they recorded, as well as documents with written
contents (ibid., 49).
I agree that certain inscribed monuments may have held symbolic importance in their
communities quite apart from whether they were read, but I doubt that most treaties ever at-
tained that status. A telling example of what kind of monument might function as symbol (in
Thomas sense) rather than as a document for reading occurs in Platos Critias (119c120b, not
cited by Thomas). There, the ten kings of Atlantis meet every fourth or fth year at the shrine
of Poseidon, beside a brass column inscribed by the islands earliest kings with the laws of the
land ( # 6 J 4J, 119cd). In Platos
account, there is much activity in the presence of this monument: the kings, bringing a bull
for sacrice, slaughter it atop the column ( " (, 119e); they allow
its blood to drench the letters of the inscription ( , 119e); thereafter, they
cleanse the column completely ( , 120a), pour libations, swear to
give judgment according to the laws on the column and to punish any king who oversteps
them. Sitting all night on the bare earth beside the column, the embers of their sacrice grown
cold, other res in the sanctuary extinguished, they consult and give judgments. While the
column is clearly the focus of the action, it is striking that that action does not include reading
(the kings swear an oath inscribed on the column [119e120b], but nothing suggests that they
read itit is the written record of an oral custom). The monument is the focus because it
embodies the communitys charter, guaranteeing its continued stability. As such, it is the ob-
ject of veneration. On this function of inscriptions and this passage in particular, cf. Steiner
1998: 6870.
34. A telling instance of this lag appears in Herodotus account (8.22.1) of how The-
mistokles sought out places (plural!) where drinking water could be found at Artemesium and
there inscribed a lengthy message in the rocks for the Ionians, urging them to not ght
against their kin, the Athenians; i.e., he chose settings where he imagined they would pause and
have the leisure to read (contrast the Homeric tomb viewers, who see the monument from afar
as they sail past). While the authenticity of the inscribed message has been justly doubted
(Macan 1908, ad loc., cited in West 1985a: 286 n. 34), the evidence that it was read is even
slimmer and more dubious (Hdt. 8.85). In other words, here, too, the scene between reader
and inscription exists more in the imagination of its author than in actual fact.
how gradually interest in the written word took hold (Harris 1989, 1996).
The sad truth behind the conventional expectation may exceptionally be
revealed in an epigram on a stele of the third or second century b.c. from
Rhodes (GV 1248.12). Here, rather than hail a passerby, the young de-
ceased addresses neighboring tombs and stelae, bidding them to weep (!)
and spread the report to all the rest: V ,
/ . Here the tombs themselves comprise a reliably
steadfast circle of mourners, and the injunction to weep is no stranger than
the image of gravestones passing the word from monument to monument
up through the necropolis. Did the audience for other kinds of inscription
resemble this one?
When, in the mid-fourth century, Athenians began to be more docu-
ment-minded, as Rosalind Thomas calls it (1989: 47), a crop of epigrams
starts appearing in the orators.35 Why, asks Thomas, were they not
quoted before? (8687).36 Her answer is that their citation must be seen
against the increasing respect for written documents at this time (86).
Increasing respect is relative, however.37 Even the changes in attitude
toward written documents in the fourth century did not erase the funda-
mental apathy of most readers toward their inscribed heritage. This be-
comes apparent in Socrates account (ps.-Plato Hipparch. 228d229b) of
the herms set up by Hipparchus, brother of the tyrant Hippias, between
520 and 514 b.c. According to Socrates, Hipparchus set up these herms at
the midpoint between the city and every deme and inscribed them with
elegiac couplets so as to educate the country folk. These are imagined as
passing up and down the roads, just like the wayfarers envisioned in epi-
gram, and pausing to read the inscriptions, which will inspire them to
come to the city for further education and betterment (
"
6
M
6 ).
35. Aeschin. 3.184 FGE p. 257, 3.187 FGE p. 420; Lycurg. In Leocr. 109 FGE
p. 230; Demosth. De cor. 18.289 FGE p. 433; ps.-Demosth. 7.39 AP 9.786 FGE p. 374.
36. Thomas continues: Apparently all kinds of material, most of it documentary, must
now be brought forward and quoted in order to prove the virtue of the Athenian ancestors.
There were certainly relevant . . . epigrams which Isocrates, Lysias or Andocides could have
quoted. Yet they did not do so (1989: 87).
37. Simon Goldhill (1999: 93100) has recently drawn attention to Isocrates Pana-
thenaicus as a crucial document in the history of reading. There (24647), one of Isocrates
pupils contrasts those who read lazily ( 8 ) and those who go
through a text with care and are able to comprehend the complexity and nuance of Isocrates
speech.
6 + bD 6 + [229] 6/

; 2 6 C ; @ 6 M

I, 6 + 6

/ ;D

6 ;
. ( +
6 J J CM
6D ( +
, 6 ME
[229]
/ ;D 6.
65 " + 6 = " M [ 6

6M M = .
[The inscriptions are twofold: on the left side [229a] of each Hermes
there is one in which the god says that he stands at the midpoint be-
tween city and township, while on the right side he says:
The memorial of Hipparchus: walk with just intent.
There are many other ne inscriptions from his poems on other gures
of Hermes, and this one in particular, on the Steiria road, in which he
says: [229b]
The memorial of Hipparchus: deceive not a friend.
I therefore should never dare, I am sure, to deceive you, who are my

friend, or disobey the great Hipparchus.]
(translation adapated from A. J. Taylor)
Socrates presents himself here as one who has paused along the road and
carefully read the epigrams he cites (no doubt he ironically suggests he is
like an agroikos come from the country to learn wisdom). An attentive
reader, then, like the one described in Shelley, he applies his reading to
the present circumstance. For as the dialogue is largely about those who
seek prot by unjust means, the admonition to walk with just intent is
apropos. And since those means potentially include even deceiving a
friend, the second maxim is equally apt. As he says in conclusion, he
would never deceive his friend or disobey the great Hipparchus.
Judging by the number of demes, there must have been at least 150 of
these herms scattered throughout Attica38a display of power as grandi-

ose for its scope as the statue of Ozymandias was for its colossal size. For
Hipparchus devised a network of inscriptions that did not have to compete
for the readers attention with the mass of monuments found in a market-
place, a necropolis, or a shrine; they were set, rather, in isolation at the
midway point between the city and every deme, scarcely avoidable by a
traveler, no matter what direction he might take. In other words, Hip-
parchus created a web of discourse from which it would be difcult for
readers to escape. Yet no further source in ancient literature refers to these
many . . . ne inscriptions.39 Were it not for the pseudo-Platonic text
and the extraordinary reader it envisions in Socrateswe would not even
know of these two.40 But that is not surprising. It was laborious to read the
inscription.
Just how laborious, however, is something we do not grasp until we
consider the appearance of the herm. The fragment of a Hipparchan
hermwhich no one would have thought to identify with the Hipparchan
herms were it not for pseudo-Platowas discovered at Koropi in the Attic
Mesogaea (149 FH CEG I 304).
[6] h h |
[ h()D] []
[Half way between Kephale and the city [here stands] glorious Hermes.
This is a monument of Hipparchus . . . ]
The fragment is inscribed from top to bottom along the right-hand mar-
gin of the shaft, which is 1.28 meters long as it survives and of whose
38. Cf. Crome 19356: 306, who rightly stresses the herms function as milestones and
their signicance as an instrument of Peisistratid policy: Aus ihrer Aufstellung drfen wir
schlieen, dass die Peisistratiden, die groen Frderer der Kleinbauern, sich sehr um den
Ausbau der attischen Wege bemht und diese genau vermessen haben mssen. Auch der
Zwlfgtteraltar des jngeren Peisistratos . . . ist ein Monument dieser groartigen Ttigkeit
der Tyrannen, die das Land ordnete und den Verkehr hob . . . Man hat ihn mit Recht als Zen-
tralmeilenstein bezeichnet (3067).
39. We do nd the following notice in Hesychius s.v. ; ;: ;
;, ` K 6 , " 6, 6 E (
. But the closing explanation of why Hipparchus set up the
herms clearly reveals that the information is dependent on Plato.
40. Crome (19356: 307) argues, sicher wurden mit der Veruchung der Tyrannen
auch die meisten Hermen als Denkmale des ermordeten Hipparch zerstrt. But any wide-
spread destruction is contradicted by Socrates statement that [t]here are many other ne
inscriptions [( + . . . . . . 6] from his poems on other
gures of Hermes.
length the inscription takes up slightly more than a meter (Peek 1935). We
may assume from the description in Plato that the pentameter was simi-
larly inscribed down the margin to the left. It is one thing, though, to read
this in Plato, quite another to stand before the stone itself and tilt ones
head now in one direction, now in another, so as to read the couplet.41 It is
anything but user-friendly. Such an inscription cannot be scanned at a
glance while walking. Even with perfect eyesight, one must stop, draw
near, make an effort.42 Perhaps it takes a Socrates to do so.
Other readers of inscriptions are singled out as in some way unusual.
The periegete Polemon of Ilion from the transition from the third to the
second century bore the nickname , stele glutton: he was
probably notorious for snooping about in shrines and cemeteries, gathering
inscriptions. In the second century, Polybius, who puts a premium on eye-
witness testimony, remarks on the well-known and apparently eccentric
claim to fame of the historian Timaios (late fourth through early third cen-
tury): his peculiar attention to written records andas Polybius derisively
sayshis appearance of accuracy in dealing with them ( +
6 . . . 6

6 , , 12.10.4). Polybius mocks him as a
man who discovered inscriptions at the back of buildings and lists of prox-
enoi on temple door jambs ( C 4
6
65 6, 12.11.2: cf.
Stein 1931: 3; Walbank 1972: 82).43 We hear of the Egyptian prince Nane-
ferkaptah in the Demotic Tale of Setne Khamwas (P. Cairo 30646, of Ptole-
maic date), who [had no] occupation on earth but walking on the desert of
Memphis, reading the writings that were in the tombs of the Pharaohs
and on the stelae of the scribes of the House of Life and the writings that
were on [the other monuments, for his zeal] concerning writings was
very great. 44 A temple priest ridicules him for reading inscriptions:
41. Hansen observes (ad loc.) that the sides of the inscription are the reverse of those
mentioned in Plato. Noting that there is more than one way to inscribe a herm, he sensibly
remarks, fortasse non omnes Hermae Hipparchi eandem dispositionem habuerunt.
42. Such an act is made explicit in the third-century b.c. iambic epigram from Alexandria
(no. 63.15 Bernand GV 1620.15) where the reader is asked to drop to one knee so as to
be able to read the inscription carefully (with both eyes, v. 5): C " , Z
/ / 6 / D /
(), b /, 6 / . On this epi-
gram, see, further, Meyer 2005: 106.
43. Walbank (op. cit.) does note, however, that Polybius is elsewhere quite ready to boast
of his own discovery of an inscription left by Hannibal [3.33.178, 56.14] in the Temple of
Hera on the Lacinian Promontory and to use the statistics which it contained.
44. Trans. Lichtheim (1980: 128). The passage is referred to by Thompson (1992).
These, he says, are of no importance to anyone; it is far better to read a

book that [the god] Thoth wrote with his own hand and that this priest
can show him.
Just how exceptional it was to read an inscription and how starkly it de-
parted from the more normal inattention is vividly conveyed in an anecdote
told by Cicero concerning an inscribed sepulchral epigram in iambic trime-
ter (senariolos, little senarii) of the third century b.c. (Tusc. Disp. 5.23).
ex eadem urbe humilem homunculum a pulvere et radio excitabo qui

multis annis post fuit, Archimedum; cuius ego quaestor ignoratum ab
Syracusanis, cum esse omnino negarent, saeptum undique et vestitum
vepribus et dumetis indagavi sepulcrum; tenebam enim quosdam
senariolos, quos in eius monumento esse inscriptos acceperam, qui
declarabant in summo sepulcro sphaeram esse positam cum cylindro.
Ego autem, cum omnia collustrarem oculisest enim ad portas
Agragantinas magna frequentia sepulcrorum, animum adverti
columellam non multum e dumis eminentem, in qua inerat sphaerae
gura et cylindri. Atque ego statim Syracusaniserant autem
principes mecumdixi me illud ipsum arbitrari esse quod quaererem.
Immissi cum falcibus multi purgarunt et aperuerunt locum: quo cum
patefactus esset aditus ad adversam basim accessimus; apparebat
epigramma exesis posterioribus partibus versiculorum dimidiatis fere.
Ita nobilissima Graeciae civitas, quondam vero etiam doctissima, sui
civis unius acutissimi monumentum ignorasset, nisi ab homine
Arpinate didicisset.
[But from Dionysiuss own city of Syracuse I will summon up from the
dustwhere his measuring rod once traced its linesan obscure little
man who lived many years later, Archimedes. When I was questor in
Sicily [in 75 b.c., 137 years after the death of Archimedes] I managed to
track down his grave. The Syracusians knew nothing about it, and in-
deed denied that any such thing existed. But there it was, completely
surrounded and hidden by bushes of brambles and thorns. I remem-
bered having heard of some simple lines of verse which had been in-
scribed on his tomb, referring to a sphere and cylinder modeled in
stone on top of the grave. And so I took a good look round all the nu-
merous tombs that stand beside the Agrigentine Gate. Finally I noted
a little column just visible above the scrub: it was surmounted by a
sphere and a cylinder. I immediately said to the Syracusans, some of
whose leading citizens were with me at the time, that I believed this
was the very object I had been looking for. Men were sent in with
sickles to clear the site, and when a path to the monument had been
opened we walked right up to it. And the verses were still visible,
though approximately the second half of each line had been worn
away. So one of the most famous cities in the Greek world, and in for-
mer days a great centre of learning as well, would have remained in
total ignorance of the tomb of the most brilliant citizen it had ever
produced, had a man from Arpinum not come and pointed it out!]
(trans. M. Grant, Penguin)
It is hard to imagine indifference more perfect than that described here

though one should also consider Callimachus account of the tomb of Si-
monides in Akragas (Aet. fr. 64 Pf.), in which an epigram is likewise ignored
and to which Ciceros narrative forms a Sicilian pendant (Bing 1988: 67
70). The scene of a necropolis so rank with brambles and thorns that
scythes are needed to clear an approach is remarkable, especially consid-
ering that within sepulchral epigram itself, such overgrown tombs are
imagined only for the likes of the misanthrope Timon or Hipponax.45
Thus, as in the case of the envisioned response of the reader, the conven-
tional expectation within the genre does not conform to the reality outside
it. Cicero himself is quite anomalous in wanting to view the tomb. From his
magician-like conjuring of Archimedes out of dust (humilem homunculum a
pulvere . . . excitabo), to his account of how the verses he had known of sud-
denly appear,46 to his exultation that he, this man from Arpinum, can teach
the Syracusans about one of their own (n.b. the culminating urry of
superlativesnobilissima, doctissima, acutissimiwhose sound is picked up in
the closing didicisset), it is clear that Cicero thinks hes something special
and he is! Only the fewest would ever have bothered. While Pierre Laurens
45. For Timon, cf. Zenodotus AP 7.315 3 GP ( / 6 , ,

8 2 / X
, / E 6/ 6 / = 6 :
6 / :, Brittle earth, may you encircle me all round with prickly box thorn
or straggling limbs of twisted bramble, so that not so much as a bird in Spring will lightly
perch on me); cf. also Hegesippus AP 7.320 8 GP. The Timon epigrams are perceptively
discussed by Fantuzzi (2000: 17480; in Fantuzzi and Hunter 2004: 3026). For Hipponax,
cf. Alkaios of Messene AP 7.536 13 GP. See also the imagined tomb of a lena at Propertius
4.5.1 (terra tuum spinis obducat, lena, sepulcrum).
46. One would like to know how Cicero learned of the verses that were inscribed on the
tomb. My hunch is that Archimedes wrote them while still alive and that they became part of
his legend through oral tradition at an early stage. But the fact that Cicero does not quote the
verses is suspicious. The tomb has never been found. Could it be that he has fabricated the
whole scene and thatonce againit represents only how he envisions himself as a reader of
inscriptions?
(1989: 34) may cite this passage to celebrate the intellectual curiosity of les
chasseurs dinscriptions, what it actually conveys is the utter inattention of
the Greeks to their monuments.47
It is against this backround of indifference that we must measure the
claims of Svenbro, Day, and Depew (noted earlier in this chapter) when
they attempt to reconstruct the readers experience of the inscribed text.
Studying these texts today, we should be particularly wary of importing
our readerly bias and scholarly valuesunthinkable that others would not
lavish the same care on these poems that we do!into ancient circum-
stances. For we all run the risk of adopting the perspective of those who set
up the inscriptions and envisioned its encounter with the reader. What is
most striking in Ciceros account is that, far from being dispossessed of
his voice (as Svenbro suggests could happen), he never even quotes us the
inscription. In one of the most detailed accounts of a reader coming face-
to-face with an inscribed epigram, the poem itself is left a blank. Ulti-
mately, the passage reads like a memorial more to Cicero than to Archi-
medes. For even when a person chooses to engage with an inscribed text,
there may be disparate degrees of readerly involvement. Rarely, if ever, I
would argue, will readers so immerse themselves in the text that they iden-
tify with the speaking voice and so (re-)activate the occasion at which the
monument was set.
To be sure, not all sites suffered neglect comparable with that of the
Syracusan necropolis. Prominent shrines, such as Delphi, were rich and
well tended. There were even guides ( or 6) to draw vis-
itors attention to particularly important inscriptions, some of which they
may even have read aloud to them. That is not to say, however, that
tourists showed interest. On the contrary, our most detailed source for
such guided tours, Plutarchs Why the Pythia No Longer Gives Oracles in
Verse, merely reinforces our impression of readerly apathy. For although
the characters in Plutarchs dialogue display a voracious appetite for anti-
quarian detail, they draw the line at inscriptions. Indeed, Plutarch recounts
their pique when the Delphic guides persisted in their standard lecture
although we asked them to cut short their set speeches and most of the
epigrams [or simply inscriptions, including epigrams]: /
+ ? 6-
8 6, (395a). Plutarch has his
47. It is perhaps with a mind toward the precariousness of commemoration through epi-
taphs and monuments that Thucydides has Pericles state the superiority of oral remembrance
of the dead: the celebrated in each of us (Thuc. 2.43.23).
visitors encounter and contemplate numerous statues and dedications, but

not inscriptions.48
I want to suggestno more, the evidence is slimthat attitudes to-
ward inscribed epigram began to change in the course of the Hellenistic
era and that this was due again to the interest of exceptional persons:
poets, above all, who clearly stopped to read what stood on monuments,
absorbed the conventions of the genre, and sensed its untapped potential
as literature, but also scholars who began making collections of inscribed
epigrams. Through them the genre acquired a parallel life in a new
medium, the papyrus.49 There, suddenly, the encounter between reader
and epigram was much easier, far likelier. For epigrams on scrollsunlike
their stationary archetypescould move about, readily disseminated, be-
coming themselves the wayfarers (C, , etc.) that their
readers had by necessity been before. Because of this newfound mobility,
epigram begins to display for the rst time a pervasiveindeed, exuber-
antintertextuality: poets avidly respond to and vary each others book
epigrams in a manner rarely seen between inscribed ones.50 The resultant
interplay has been singled out as typical, in fact, and termed the art of
variation in the Hellenistic epigram (Ludwig 1968; Tarn 1979). Easily
transportable, epigrams even made their way into the schools. As is clear
from the third-century b.c. Livre dcolier, children were assigned to read
them. The epigrams of Callimachus, no less, came to be read in school
(Athen. 15.669c test. 41 Pf.). And thus an encounter of a sort (at least)
48. At lesser shrines, there might be no guide to point things out to the traveler; cf. the
proem to Longus Daphnis and Chloe, where, at a grove of the nymphs in the countryside of
Lesbos, the narrator must make the effort to nd one ( 6 , praef. 3).
On guides generally, cf. Casson 1974: 26467; Habicht 1985: 14546; Jones 2001.
49. On this development, see Ergnzungsspiel in the Epigrams of Callimachus and
Between Literature and the Monuments in the present volume. Cf. also Gutzwiller 1998:
47114.
50. A telling instance of inscriptional intertextuality in the Hellenistic era may be seen in
an epigram from Thermon, IG IX2 51 (rst quarter of the third century b.c.), whose nal two
lines contain a sustained and marked allusion to an inscribed epigram of Simonides, pos-
sibly for the dead at Plataea (AP 7.251.34 IX FGE p. 199). By what mechanism, we may
ask, does one inscribed text refer to another? Had this allusion been found in a text of an ear-
lier era, the answer would most likely be that the poem of Simonides was one of those rare
inscribed epigrams that became part of the oral heritage, inspired by a particularly famous
battle. Only such poems would have had sufciently wide dissemination to motivate refer-
ence in subsequent inscriptions, and only such would have had a reasonable chance of being
recognized by readers. (An author could, of course, have become acquainted with an in-
scribed epigram in situ, but then only few readers, if any, would recognize his allusion.) By
the early third century b.c., however, an inscribed epigram could also have been disseminated
in writing and become known to poets and readers elsewhere through that medium.
between reader and monument was assured.51 Brought into the schools on
papyrus, its modern medium, book epigram may now, in turn, have
heightened the appeal of epigrams on monuments and made more readers
willing to stop and contemplate inscriptions carved in stone.
One senses a nascent taste for inscribed epigram in a scene from the
Life of Aesop (78.1, rst century a.d.).52 Here we read that Aesops master
Xanthus walked out with Aesop to the edge of the city, enjoying his com-
pany, and when he came to the cemetery, he took pleasure in reading the
epitaphs ( 6 1 M ,M , , J
, CY 6 6 6
6). The narrative then describes how Aesops eye is
caught by a particular epitaph andin a paradigmatic dramatization of the
act of readingthe leisurely process of how he entertains its various pos-
sible meanings.53 What is amazing in this scene is to nd a character who
goes somewhere simply for the pleasure of reading inscriptions. Even
more astonishing is that it is not the exceptional and brilliant Aesop who is
described as enjoying them but, rather, his conventional master.54 With a
51. In their editio princeps, O. Guraud and P. Jouguet (1938: XIV) argued that the
scroll was a manual for schoolchildren. More recently, however, it has been seen as a text for
a schoolmaster from which he could draw for classroom use (Cribiore 1996: 269 no. 379 and
pp. 12128). On the use of epigrams in schools, see Wimann 2002; Rossi 2002: 16385.
52. On the date of the Life, see Hopkins 1993: 11 n. 14.
53. At a slightly later date, there is perhaps evidence for a similar scene of a reader
puzzling over epitaphs: in Photius summary of Antonius Diogenes Wonders beyond Thule
(Bibliotheca 166, 111b), Alexander the Great is bafed by riddling inscriptions on grave vaults
and ultimately nds there the cypress tablets on which the novel itself is written.
54. It is from roughly this period that we get a description like that in Ovid (Met.
11.42729) when Alcyone tries to dissuade her husband Ceyx from going to sea by telling
him of the wreckage she has seen on the shore and how she has often read the names on
cenotaphs: Aequora me terrent et ponti tristis imago: / et laceras nuper tabulas in litore
vidi, / et saepe in tumulis sine corpore nomina legi. I owe this reference to Philip Hardie.
Riemer Faber draws my attention to Calpurnius Siculus, Ecl. 1.2032, a detailed descrip-
tion of reading an oracular inscription, which in many ways anticipates the ecphrases of
inscription reading in Nonnus, analyzed at the close of this chapter.
In the latter part of the rst century a.d., Pliny the Younger (Ep. 7.29) describes his reac-
tion upon encountering an inscription on the road to Tibur, less than a mile away from
Rome, which records how the senate honored Pallas, the freedman and nancial secretary of
the emperor Claudius: This inscription more than anything makes me realize what a farce it
is when [honors] can be thrown away on such dirt and lth, and that rascal could presume to
accept and refuse them, all with a show of setting posterity an example of moderation.
Plinys reaction assumes that, however outrageous, the inscription will be read: it is an effec-
tiveand easily manipulatedmedium for setting posterity an example.
Somewhat later, in Xenophon of Ephesus Ephesian Tale (second century a.d.), the cli-
mactic recognition scene hinges on the reading of dedicatory inscriptions. And as in the Life
of Aesop, the readers are no longer the exceptional persons that once would have likely
stopped to read: Leukon and Rhodeslaves of the storys hero and heroinevisit the temple
of Helios in Rhodes, where their masters had once offered up a gold panoply and tablets con-
reader like that, an everyman like Capras, perhaps we have moved a step
closer to our scene at the Lincoln Memorial.
This essay began from questions raised in the face of a striking contradic-
tion: Encounters between readers and monuments are ubiquitously evoked
in inscribed and literary epigram and indeed form one of the genres most
salient features, suggesting that such encounters must have been common
in everyday life; yet apart from in the epigrams themselves, we nd scarcely
any descriptions of readers pausing before inscribed monuments and con-
templating their texts, no ecphrases of inscriptions corresponding to the
numerous examples across a variety of genres that take as their subject
works of art. In probing the roots of this contradiction, we traced a perva-
sive indifference toward inscribed textsdespite their blunt appeals to
passersbyextending until the Hellenistic age. At that time, interest in
such texts seems gradually to grow, due perhaps in some measure to the
sudden popularity of literary epigrams, which even led to their adoption for
use in schools. Even then, however, as mentioned near the start of this
chapter, it was a long time before an ecphrasis of an inscribed monument
would appear. When nally it did materialize, how was it presented? In
closing, let us explore that question by briey considering a passage from
Nonnus Dionysiaka (12.29115).
In this passage, one of the ;, who alone among her sisters still has
no privilege yet proleptically is called nursemaid of ripened fruit to come
(6 + 4, v. 29) and the grape-bearing Season
( K, v. 21), begs Helios to tell her, though grapes do not
yet exist, on which of the Seasons Time will bestow the honor of growing
taining a dedicatory epigram (1.12). Beside these, they make a new offering (5.10), honoring
their masters with an inscribed pillar and listing themselves as donors (5.1112). Shortly
thereafter, their mistress Anthia arrives in Rhodes, goes to the temple, and dedicates a lock of
her hair, together with an inscription, on behalf of her missing husband Habrocomes. Stop-
ping at the temple again later the same day, the two servants are thunderstruck to nd the
new dedication. Upon reading it, they realize that Anthia must be nearby, and their reading
precipitates the reunion of the pair.
I should note a further scene of recognition in Xenophons Ephesian Tale, which involves
reading inscriptions but at the same time describes what may have been common limits of
readerly interest with regard to this kind of text. Not long after the servants set up their in-
scribed pillar, their master Habrocomes comes to the temple and, noticing their dedication
and, above all, its inscription, starts to lament all he has lost. Just then Leukon and Rhode ar-
rive, see him, andas they do not yet recognize himquestion him: Why are you . . .
lamenting beside offerings that have nothing to do with you? What concern of yours are
these? And what do you have in common with the people named here? (5.10). This leads to
a joyous recognition. Interestingly, however, the servants assume that only a person with
some connection to the dedicators would trouble to linger at a dedication and show sympa-
thetic interest.
this fruit ( ,; / , ,

, @ / 65 , vv. 25
27). With raised nger, Helios points her to the separately fastened tablets
of Harmonia ( ; 2, v. 32) beside a wall just op-
posite (M MM, v. 30).55 There she may read her destiny
from pictures and from an inscribed prophetic text. Evidently the poem is
interested in specifying a locus legendi in this way not just for its characters
but for its readers as well, thus ensuring that we can visualize the act of
reading in the passage that follows.
The subsequent portrayal of the motions and attitude of inscription
reading is remarkable, enacted by a reader who, far from being indifferent,
is engrossed in the text, since she has a personal stake in its content. She
dashes over to the inscription ((, v. 41) and, though Helios had told her
to look specically at tablets 3 and 4 (12.3740), she cannot keep her eyes
from straying over the rst two tablets (= , v. 42).56 When
she nally runs over to the third tablet ( @ 6, v.
64), she stands xed to the spot, gaping ( . . . , vv. 65
66) at the letters of glowing color engraved with the craftsmans red ochre
( , J M, v. 67).57 Restlessly, she
passes over the text at a glance ( + , 2
K, v. 90). Striking here is the play on , line of text, in ,
pass over; remarkable, too, is the vivid gesture of readerly impatience con-
veyed by the adjective . At last she arrives at the part that concerns
her (, G, @J . . . , v. 91); there she brings her feet to rest
and reads ( _, / , v. 96),
nding her role conrmed as the special Season of the grape. The ecphras-
tic climax, however, comes when she sees on a neighboring wall (@J
M, v. 104) a four-line oracle of inscribed hexameters (-
55. The term 2 may mean either yoked together or yoked singly (LSJ s.v.).
Given the context in Nonnus, however, where the Season runs from one tablet to another (cf.
esp. v. 64), it appears more likely that each of the are imagined as separately fas-
tened. Considering their position M (v. 30), could it be that the were fas-
tened to that wall? That might be supported by the later description of the tablets of Har-
monia (41.360), where they are described as the glorious oracles of the wall (
). Or were they, rather, secured on a stand or in a frame near the wall? By the
Hellenistic era, the term was used generally of inscribed pillars or tablets (LSJ s.v. II).
For the controversy surrounding the meaning of the term earlier in connection with the in-
scriptions bearing the laws of Solon, see Sickinger 1999: 2627.
56. The detailed description of their contents is clearly focalized through her eyes.
57. may play on the alphabets Phoenician inventor, Cadmus (discussed
shortly in text). Red epigraphic highlighting is stressed also in the ecphrasis at Nonnus
41.33999, specically at v. 363, where the oracle is engraved on the tablet in ruddy ochre
(6 . . . : M).
6 4 , v. 107)an epigram, in other words (vv.

11013)from which she learns which god is to be her particular patron.
For while Zeus has given the laurel to Apollo, roses to Aphrodite, olives to
Athena, and corn to Demeter, he has granted the vine to Dionysus (
? M, v. 113). At this, she bounds away in joy ( /
A, v. 115), having nished reading the inscription.
Signicantly, a comparable scene of inscription reading occurs else-
where in the Dionysiaka (41.33999). This one likewise concerns the tablets
of Harmonia.58 But here it is Aphrodite who reads, so as to learnamong
other thingsthe fate of the city Beirut. The scene contains many of the
same elements found in book 12. As Helios there guides the Season to the
prophetic tablets by the wall (M M / 4
6, 12.3031), so here Harmonia herself leads Aphrodite to the or-
acles of the wall (? 6 , 41.360). As the let-
ters in the rst instance were glowing red, engraved with the craftsmans
red ochre ( , J M, 12.67), so
here the oracle contains red letters ( , 41.352), en-
graved in ruddy ochre ( : M, 41.363). In either case,
the motions of reading are the same, with the deitys eye roving from one
tablet to another (/ @ . . . ( K, /
, 12.5255 / @ / . . . , /
6, 41.36870) and from wall to neighboring wall (@J
M, 12.104 @J M, 41.370); in each, a
similar phrase describes how the reader comes upon the passage she needs
(, G, @J . . . , 12.91 , ( , @J . . . ,
41.361). Both scenes culminate when the avid reader nds a particularly
gratifying prophecy (# + / . . .
, 12.1089
( # , 41.388), quoted verbatim, after which she
leaves the scene ( / A, 12.115 , 2 , ( -
, 14.400). In brief, Nonnus, who is elsewhere at pains to evoke every
traditional element of Homeric epic one could think of (from invocations
of the Muse, to catalogs, epithets, similes, etc.), here creates something
new: a novel typical scene, one more in keeping with his ageinscription
reading as typical scene.
That is no shock coming from Nonnus, who was above all things a
learned poet.59 The fth-century poet of Egyptian Panopolis may have
retrojected an aspect of his scholarly concern with texts of all kindseven
58. These are here consistently called rather than , but at 12.69,
are evidently used synonymously with : 6 .
59. Cameron 2000: 186.
with the mechanics of inscription and the practice of writingwhen, in

book 4 of the Dionysiaka, he describes the invention of the alphabet by Cad-
mus, who turned with a lathe the engraved mark of sounding silence
( , v. 263), carving the cuttings
at an angle with back-faring hand ( 4
, v. 268).60 Cadmus concern with writing is presented as the di-
rect upshot of his upbringing in Egypt. For he had spent nine years there
together with his father, Agenor. While there, he learned to write, squeez-
ing out the milk of the holy books, not to be divulged (
, v. 267), and while a boy in the temple full of
stone images, he had come to know the inscriptions, cunningly wrought,
carved deep into the wall ( + / M
M / , vv. 27375). In this
light, the prominence of Harmonias inscribed tablets in the ecphrases
detailed earlier, does not surprise us. After all, she was Cadmus bride, who
sailed with him to Greece even as he brought with him the gift of writing
(vv. 23348). Indeed, the seeds of Harmonias link to the written text may
have been sownin allusively erotic termsat the very moment that Cad-
mus invented his letters, mingling consonant and vowel in a line of
blended harmony ( + / Z 6
vv. 26162).61
60. The phrase 4 evidently refers to the direction of the script from
right to left.
61. I am grateful to Prof. R. Hschele for discussing these passages from Nonnus with
me and for her stimulating suggestions.
chapter 8
Allusion from the Broad,

Well-Trodden Street
THE ODYSSEY IN INSCRIBED
AND LITERARY EPIGRAM
A round 1980, in the hometown of the University of Michigan, Ann

Arbor, a large billboard raised prominently above State Street adver-
tised the merits of Miller Lite. Lite beer was at that time a relatively new
productMiller had introduced it into the American market just a couple
of years earlier, in 1977and perhaps advertising strategies were not yet as
relentlessly geared to a young, single, sports-oriented, pleasure-seeking
audience as they are today. But even then, this particular billboard stood
out as something unusual.
Drawing on a painting of the year 1787 by J. H. W. Tischbein, it por-
trayed the German poet Goethe, age thirty-eight, reclining on some low-
lying blocks from an ancient ruin in the landscape of Campania during his
italienische Reise. He is elegantly clad in a owing cream-colored cape, cov-
ering stylish knickerbockers, white hose, and black shoes; on his head, he
sports a dapper broad-brimmed hat, which he wears at a rakish tilt. Alto-
gether the painting encapsulates the image of the cosmopolitan dandy of
the Enlightenment, drinking deep of the rich cultural elixir of the Italian
countryside. What was novel in the Ann Arbor billboard, however, was
that its artist waggishly altered the poets outstretched arm in the original
147
Fig. 5. J. H. W. Tischbein, Goethe in the Campagna, 178687, Frankfurt

a. M., Stdelsches Kunstinstitut. (Courtesy of Stdel Museum/ARTOTHEK.)
painting, raising it up and placing in its hand a mug of beer. Further, com-
ing from the poets mouth were now the words More Lite!
Billboards are characteristically among the more lowly instruments of
American commerce: set on the busy highway and hawking their wares to
the eeting throng, they tend to keep their message simple and blunt.
Here, however, one really wonders what the marketing department of
Miller Breweries was thinking, since this one seems to expect something
more of its target audience. In the rst place, simply to recognize that this
is Goethe entails a considerable degree of cultural literacy on the part of
prospective viewers. And unless the onlooker can identify this gure, he
will be unable to grasp a further and still more learned transaction. For the
phrase More Lite is an allusion, playing on Goethes dying words,
Mehr Licht!an ambiguous utterance, which has itself provoked a raft
of scholarly exegesis. Thus we nd a most recondite literary allusion em-
bedded in a strikingly humble public medium.
In the eld of Hellenistic poetry, learned allusion has of course been a
Allusion from the Broad, Well-Trodden Street 149
staple of scholarly investigation for more than a century, especially as it is

deployed by such elite authors as Aratus, Apollonius of Rhodes, Calli-
machus, and Theocritus. Here it is viewed as one of the hallmarks of
rened Hellenistic artistry. In this essay, I will try to add a new perspective
to this topic by extending the scope of discussion to include allusion even
in Gebrauchspoesie, or utilitarian versethat is, in the anonymous verse
inscriptions known as epigram that lined the broad, well-trodden streets
and public spaces of Greek cities, even as billboards do in our own time.
Specically, I will probe the popular limits of allusion by examining epi-
graphic references to the Odyssey to see whether they are qualitatively dif-
ferent in inscribed epigram from those found in its literary counterpart. By
so doing, I hope to delineate a spectrum of readership and readerly expec-
tation concerning allusion, from the simplest to the most challenging.
In this regard, the Ann Arbor ad may be helpful. For it raises a number
of questions that can be fruitful for us in dealing with erudite allusion in
Hellenistic poetry. First is that ever thorny question of authorial intent:
What did the author want? Or in the case of our billboard, what purpose
would an allusion to Goethe serve in the advertising campaign of Miller
Breweries? Second, to what extent does context (here the very location of
the billboard in a particular place) determine the character of the allusion
we are likely to nd? That is, does the physical setting of performance and
reception matter in determining just how demanding an allusion will
probably be (how broad the range of sources it draws on, how familiar or
obscure the sources, how elaborately detailed its allusive play)? Finally,
what does the allusion suggest about its implied audience?
The answer to the rst question is simple enough. What Miller in-
tended by alluding to Goethes words and image is fairly certain: it aimed to
sell beer. A more delicate matter is, how did it do so? Though I have no way
of knowing whether the ad achieved its aim, it is reasonable to think that it
sought to carve out a new segment of the market by arousing the interest of
those who did not t the traditional beer-drinking prole. It targeted a cul-
tivated consumer, college educated, perhaps with advanced degrees and
considerable earning powera consumer, moreover, who might be con-
cerned to watch his waistline and so to welcome a less-lling brew. Even if
such a customer might not recognize the allusion to Goethe, he would still
recognize the afuence and renement of the gure demanding More
Lite! and thus associate lite beer with such a drinker.
Let us move on to our second question: To what extent does geo-
graphic setting determine the character of an allusion? Or differently,
does the real-estate maxim also hold true for allusion: is it all location,
location, location? Our billboard suggests that this is indeed a crucial fac-
tor. For can we imagine this ad placed anywhere but in a university town?
Could the brewery have hoped to nd a comprehending audience in an-
other setting? In that regard, it was only shrewd marketing to locate the ad
just over a block from the central campus, within easy walking distance,
and close to bookstores and cafs. Pedestrian accessibility was certainly
essential in view of the high proportion of students without cars. Thus,
had it been placed even on the fringes of Ann Arbor (say, by the nearby in-
terstate) the billboard would have lost much of its impact.1
Finally, as to the billboards implied audience, I already surmised that
those who would appreciate the image and its allusion would likely belong
or aspire to that cultivated, afuent, and upwardly mobile group that here
largely coincides with the universitys forty thousand students, together
with its faculty and staff. Yet I want to point to an undergraduate subgroup
for whom a modicum of literary knowledge was obligatory. I mean the stu-
dents of the Honors College, Michigans crme de la crme, who at this
time were required to enroll in a full years course in the Great Books,
which included Goethe in the second semester. Was the billboard con-
ceived with these very students in mind (indeed, does its presence suggest
that one of them had found a job with Miller Breweries)? For to them, as
to the upper classes of ancient Greece and Rome, their literary education
served as a badge of elite status. Such students might well recognize the
allusion to Goethe and thus have the pleasure of nding their hard-won
culture afrmed out in the real world (if only a block from campus). For
a reader able to detect an allusion may feel the warm glow of belonging to
a privileged clubthe cognoscenti, who share a common cultural lan-
guageand the intensity of that feeling may increase in proportion to the
difculty of the allusion.
Let us now turn to ancient verse inscription and its use of literary allu-
sion. To begin with, it is worth emphasizing that allusion is not particularly
common in inscribed epigram of any era. That may at rst seem surprising.
After all, its metrical formpredominantly hexameter and elegiac cou-
pletmakes it a close relative of those grander, culturally more estimable
1. A further reason that Millers marketing department may have chosen Ann Arbor was
that it was founded by Germans and retained some measure of German heritage. When I was
a graduate student, there were still several German restaurants in town, numerous German
names attached to leading businesses, and many antiquarian copies of (yes, you guessed it)
Goethe available in the used bookstores. Among other things, then, the advertising agency
may have taken that heritage into account (just as it might Hispanic or African American her-
itage in other parts of the country), in the belief that Goethe might still be a recognizable
quantity and carry some weight with this population.
genres of epic and elegy, from which it undoubtedly draws a good measure
of its style. Thus one would expect inscribed epigram to be generically pre-
disposed to allude to specic subjects or incidents from those kindred forms
of verse and to do so in language pointedly recalling that used in those con-
texts. But this is not the case.
The inuence of epic on inscribed epigram is indeed pervasive right
from the start, but as Friedlnder and Hofeit demonstrated quite clearly
in their collection of Greek verse inscriptions of the archaic era,2 it nor-
mally manifests itself in nonspecic epicizing; that is, the diction does not
attempt to evoke the details of particular epic scenes. Rather, it gives the
inscriptions broadly heroic coloring, nothing more. In the words of
Friedlnder and Hofeit, the thorough epic style . . . shaped the concep-
tion of life (1948: 30), giving to the solemn moments of their lives a
Homeric cast (9); in this way, men and action receive something of the
dignity and strength of the epos (13). Epigrams inscribed in such public
settings as cemeteries, shrines, or civic gathering places during the archaic
era aimed not to suggest comparisons with particular poetic models but to
evoke a generic heroic tone,3 and that is in keeping with the often formu-
laic, thoroughly conventional character of sentiment expressed in in-
scribed epigram.
Yet if we look for learned allusions in inscribed epigram of the archaic
era, it is instructive to see where it occurs. One of our earliest verse in-
scriptions, that on the humble clay kotyle known as the Ischia Cup, from a
cremation burial of the late eighth century b.c., contains what has been
called Europes rst literary allusion (CEG I 454).4
: ( [ ] : _[]: . |
h / [ : []: " |
h h : [] : /.
1 ( [ ] Heubeck : [,] G. Buchner and C. F. Russo, Rend. Linc., 8th ser., 10 (1955):
226 n. 2
[Nestor had a drinking cup [or I am Nestors drinking cup] good to

drink from, but whoever drinks from this one will at once be seized by
desire for fair-crowned Aphrodite.]
2. Friedlnder and Hofeit 1948.

3. Irwin (2005: 6381) stresses the desire to heroicise the dead (69) as a function of
aristocratic self-representation in archaic funerary epigram. This constitutes generic,
rather than specic, allusion. Cf. Pucci 1994: 2427; Irwin 2005: 19596.
4. B. Powell 1991: 167. The supplement in v. 1 is based on Heubeck 1979: 11016.
As is generally agreed, the versesconsisting of a somewhat awed iambic

line followed by two dactylic hexameterspointedly and humorously refer
to the cup of the hero Nestor described at Iliad 11.63136.5 As the Iliadic
passage notes that Nestor brought this cup from home (T : V / C
, v. 632), it may well be the same cup that appears in his household
and is offered to Athena at Odyssey 3.5153. It could thus be viewed as this
heros standard epic prop.
A reader needs more than casual acquaintance with epic for this allu-
sion to work. It does not sufce merely to know that the hero Nestor had
a cup. The reader must recall that this cup was elaborately wrought, stud-
ded with golden nails; that it had four handles, and on either side of each
perched two golden doves, feeding, and there were double bases beneath
it (Il. 11.63335); he must recall, further, that it was hugeanother man
with great effort could lift it full from the table, / but Nestor, aged as he
was, lifted without strain (vv. 63536). The reader must thus have the sort
of mastery of the tradition that comes from repeated, attentive exposure.6
For the playful point of this allusion depends on his being able to set that
epic description against the physical fact of the humble clay kotyledeco-
rated with modest geometric motifs and measuring a diminutive ten by f-
teen centimeterswhich he holds in his hands. Indeed, it helps if he
knows that despite the wealth of incident in the books following the cups
description in Iliad 11, the hero is still guzzling from it at the start of book
14the strength of the ancient heros elbow is surprising, as Hains-
worth wryly observes7for this protracted bout of drinking wittily moti-
5. For a good discussion of the allusion in Nestors cup and its humor, cf. Fantuzzi and
Hunter 2004: 28687.
6. At this early date, such exposure is likely to have been oral and part of a still-uid oral
tradition. That is suggested by the fact that the Ischia Cup shows no verbal correspondence to
the description of Nestors cup in Homer, referring, rather, to the content of the epic scene.
Precise verbal allusion, by contrast, is more likely to arise in a context where xed texts may be
weighed against each other for comparison. That, as we shall see, is the case already a century
or so later in Alcaeus adaptation of Hes. WD 58288 in fr. 347 Lobel-Page (discussed in n. 12
in the present essay). There, verbal correspondence between alluding text and model stretches
over fully seven verses. The striking span over which Alcaeus recurs to his model in relentless,
specic detail may suggest that the poet was not striving to be subtle with his allusion: he
wanted, rather, to ensure recognition of the antecedent passage by an audience relying on its
memory in the eeting circumstances of performance. Such an audience must in turn be dis-
tinguished from a different, more bookish one able to consult and juxtapose written texts so as
to detect the most recondite and intricate allusions. This latter sort of audience, as well as the
kind of poet who writes for it, belong to a different age, as we shall see near the end of this
essay when discussing Callimachus allusion to the Odyssey in Ep. 43 Pf. For a nuanced discus-
sion of the problem of allusion in a largely oral culture, see Irwin 2005: 2226, 100101, 114
15, 11819, 16264, 19496). For allusion in literary epigram, see Sens 2007.
7. Hainsworth 1993: 293 ad vv. 63637.
vates the epigrams characterization of the vessel as _[] (v. 1).8 The
tiny and humble kotyle, by contrast, works at once (", v. 2)9 andde-
spite its very different size, material, and embellishmenthas a power of
its own, that of Aphrodite, which makes its drinker long for sex. Rather
than fortify the hero, then, in the context of martial epic, it motivates the
lover at a symposium.
Authorial intent may be ultimately unknowable, particularly across so
vast a chasm of time. Yet it is important to make informed, plausible con-
jecture about it within the historical context. In this instance, I suggest that
the poets motive in deploying this allusion was arguably to exploit the
knowledge of the elite group so that it may dene itself. It does this
through the opposition of its own sympotic values, with their focus on
merriment and eros, to the values of epic, which favor grand heroic action.
That opposition appears as well in contrasting the poetic genres of epic, on
the one hand, and, on the other, those shorter forms at home in the sym-
posium, such as skolia, lyric, or even an epigram incised on a cup.10 Indeed,
the cup brilliantly solves a perennial problem of the epigrammatic genre,
namely, of alluding to so expansive a narrative form as epic within the brief
connes of epigram, by humorously guring the difference in genre
through the contrast between the huge heroic cup and the unassuming
kotyle. It makes the difference part of its witty point.11
Signicantly, this pointedly allusive text was not set out on a grave stele
along the public highway or dedicated in a religious shrine with its throng-
ing worshipers, nor did it belong on a civic monument in the busy market-
place. Rather, the modest clay vessel points to the context of the sympo-
sium, where the small and privileged circle of like-minded hetairoi gather
8. In the compound _, " signals either the abundance of drink in the cup or (hu-
morously with reference to Nestors strength and prodigious capacity for drink) the ease of
drinking. Cf. LSJ s.v. VI.
9. The epigram is inscribed upside down and is thus legible when the cup rests on its
rim. In other words, it communicates its message to the drinker either after he has emptied
and laid down the cup (as a commentary on the effect of the drink he has just nished) or pre-
liminary to drinking (in anticipation of its results).
10. The mix of iambic and hexameter lines may serve as a generic marker, signaling
mock-epic, as in the Margites, which consisted of hexameters irregularly interspersed with
iambics. Cf. Aristot. Poet. 4. 1148b24; Hephaest. Isag. 4 (p. 59.21 Consbruch); Hephaest. De
poem. 3.4.
11. Cf. B. Powell 1991: 16566: The clay skyphos bears as much resemblance to the
elaborate gold masterwork of Homers Nestor as its owner bears to the great Trojan ghter
except of course that both are heavy drinkers! Homer himself parodied epic exaggeration in
his description of the cup (Anbody else could scarce have lifted it . . .), speaking of the cup as
he did the mighty stones upon the windy plain of Troy which not two men could bear, such
as men are today.
to enjoy the pleasures of wine, verse, and eros. Indeed, the inscription
evokes that setting and those very elements of sympotic activity, since
through the medium of poetryit pictures what happens whenever some-
one drinks from the cup: he suddenly feels in the mood. The paradig-
matic scene that the text envisions applies to anyone at the partyand thus
the inscription denes its target audience, as well as those likely to appre-
ciate its allusion, as coextensive with the sympotic group. That delineation
of an implied audience is reinforced inasmuch as the cup is a functional ob-
ject designed to be of use at the symposium for any member of the exclu-
sive fellowship. This contrasts with the Homeric cup, which only the hero
can handle with ease, and even more so with inscriptions on such public
monuments as tombstones or dedications, which promiscuously address
anyone at all, regardless of origins, who will take the trouble to read them.
However surprising it is to nd a detailed literary allusion in an inscription
(particularly at so early a date), it is less surprising for having an incompa-
rably smaller, more privileged and poetically knowledgeable target audi-
ence than that for an inscription on a public monument.12 That that audi-
ence was indeed privileged is demonstrated by the sepulchral context in
which the cup was found. For it comes from a family plot containing no-
tably opulent burials.13 These are distinguished by a conspicuous quantity
12. Another instance within the context of the symposium is Alcaeus reworking of Hes-
iod WD 58288 in fr. 347 L-P (also mentioned in n. 6 in the present essay). Concerning this
fragment, Page (1955: 306) comments, Nowhere else in Greek poetry, except in deliberate
parodies, is so extensive and close a copy of one poet by another to be found. As Rsler (1980:
25664) plausibly argues, the point of the allusion lies in how Alcaeus applies Hesiods gener-
alized advice about what to do in the heat of summer to the circumstances of his sympotic
hetairoi. Hesiod, the didactic poet, describes an indenite, recurrent situation concerning
which he gives guidancein the heat of summer, when cicadas sing, artichokes ower, women
are lascivious, men weak, etc., then one may (potential optative, vv. 588589) sit in the shade of
a rock and drink a sober mix of three parts water to one of wine (v. 596). For Alcaeus, those
same conditions (reiterated in terms remarkably close to Hesiods) justify an urgent, immedi-
ate appeal in the here and now of the sympotic group: soak your lungs with wine, for the star
is coming around ( :, , ). Drink is impera-
tive in Alcaeus world, for unlike in Hesiods, everything is thirsty ( + /, v. 2),
not just Alcaeus companions. According to Rsler (1980: 263), Alkaios konnte (dies eben
beweist die Tatsache, da er ein solches Gedicht fate) davon ausgehen, da die 2 die
Anspielung zu verstehen und seinen Text zu dem von ihm evozierten Text Hesiods in
Beziehung zu setzen vermochten, da sie den berraschenden Umstand, trotz der ihnen ver-
trauten Thematik diesmal nichts Eigenes zu hren zu bekommen, sogleich bemerkten und
die Nachbildung als dichterische Leistung spezischen Ranges zu wrdigen wuten. For
the differing demands put on the audience with regard to recognizing allusion in Alcaeus vis-
-vis the Ischia Cup, cf. the reections in n. 6 in the present essay; the intensied demands
placed on readers of Hellenistic literary epigram are discussed later in this essay.
13. Ridgway 1992: 11516.
of precious metal and rich personal ornaments that denote membership

of an upper class (Ridgeway 1992: 115). The burial of the ten-year-old
male from which we have Nestors kotyle further suggests the family inter-
est in symposia, as it contains no less than four kraters (ibid., 116). As in
the case of the Ann Arbor billboard, then, a learned allusion is at home in
certain settings more than in others. It requires an environment where it is
apt to nd an appreciative audience.
I will return later in this essay to the idea of the symposium as an arena
congenial to allusive play. For now, it is worth noting that we may envision
other inscriptional settings (in addition to the symposium) where allusion
would be apt to nd discerning readers. An epigram inscribed in a palace
precinct can aspire to a different readership from one set in the agora. A
poem addressed to a king may display a higher level of sophistication and
polish from one aimed at a more general audience. Thus, in an epigram on
the base of a bronze satyr statue in Pergamon dedicated by the statesman
and admiral Dionysodorus, son of Deinokrates (MS 06/02/05), the im-
plicit addressee is the Pergamene king Attalus I.14 As Parsons notes in
characterizing the poems renement and literary self-consciousness, it
would be difcult to nd a more Alexandrian epigram, though it is lo-
cated in Pergamon.15 For in it, the satyr of the statue, speaking in artful
phalaecians and Doric dialect, says that he is Pratinaean in theme ( +
, v. 5). He is thus not simply the familiar follower of
Dionysus, ubiquitous in art and myth, but a specic instance thereof from
literature, the prototype of a satyr in a particular poetic genre. For in say-
ing that he is Pratinaean in theme, the satyr alludes to the inventor of
satyr play, Pratinas of Phlius. He thus comments on his own literary pedi-
gree, doing so, moreover, in the dialect native to his place of poetic ori-
gina clear sign that his audience (or at least the king to whom he was
dedicated) had an interest in and knowledge of literary history.16
14. The poems ostensible addressee is Dionysus. As the epigram is, however, dedicated
both to the god and to Attalus and closes with a plea for both the gods and the kings favor, it
is clear that it aims for the ear of the king.
Unfortunately, the statues original location is unknown since, according to H. Mller in
his editio princeps (1989: 500), the base was incorporated into a wall already in Hellenistic
times. For this inscription, cf., further, Lebek 1990; Kassel 1990; Kerkhecker 1991; Parsons
1993: esp. 1314; Lehnus 1996. A further instance of a prominent Attalid monument pre-
senting a particularly artful epigram is that on Delos, celebrating that dynastys founder
Philetaerus, for his victory over the Gauls (most likely in 275/4 b.c.). On this poem, IG XI 4,
1105 Durrbach, Choix no. 31, see Bing and Bruss 2007: 811.
15. Difcilmente si pu trovare un epigramma pi alessandrino (Parsons 1993: 14).
16. One is tempted to take here in its scholarly sense of lemma, a heading or
rubric of a comment in the scholia.
Not all inscriptional allusions are equally learned. Do those that are less
so belong to a different setting? Inscriptions may, of course, reect the
tastes and literary education of those who commission them (or of the poets
they hire), regardless of their setting.17 But as with the Ischia Cup, poems
inscribed in busy public settings often envision their target audience, fore-
seeing a reader who is a kind of everyman, an ordinary wayfarer, and
addressing him as , , C, or the like.18 Given the
overwhelmingly public character of most inscribed monuments, it is rea-
sonable to assume their content was geared toward an implied audience
approximating that generic passerby imagined in so many poems.
What sort of poetic allusion, then, if any, may we expect in such in-
17. Thus CEG II 597, brought to my attention by Kathryn Gutzwiller, is a sepulchral

poem of ca. 33020 b.c. from Rhamnous, which incorporates a subtle Odyssean allusion. It
commemorates Hieron, who went to Hades last of ve brothers, leaving his spirit far be-
hind, subject to shining old age ( # , v. 4). The words at
the start of this verse quote Teiresias prophecy that death will come to Odysseus when the
hero is worn out by shining old age ( Y a M , Od. 11.136 23.283).
Apart from our epigram, the combination Y a M occurs only in these two passages
in all Greek poetry. Hierons prosperous old age is thus evidently assimilated in a meaningful
and gratifying way to that of Odysseus. Perhaps the reference to Hierons family, from which
his brothers departed before him (j (, v. 2), suggests a potential audience suf-
ciently educated and interested to appreciate the allusion. On family as a potential audience,
see my discussion of GG 424 later in the present essay.
Another instance worth mention is CEG I 40, an Attic grave stele from ca. 530520,
which relates how a certain Peisianax erected a sema for Damasistratos: 6
[] (v. 2). The words may recall Il. 16.43157, where Hera dissuades Zeus from saving
Sarpedon, urging her husband instead to abandon him to death, to be buried by family with
tomb and stele: 6 (v. 457 v. 675). To judge from Merkelbach and
Staubers Sachregister: Homer: Zitate, Variationen, Anspielungen (MS vol. 5 VII p. 335),
this Homeric phrase is quoted more often than any other in inscribed epigram. We may
doubt, however, that these quotesmostly in marginally competent poemsactually evoke
the Iliadic scene. Indeed, even in Homer, the phrase is used of other funerary customs as well,
such as lamenting the dead (Il. 23.9), cutting ones hair and weeping (Od. 4.19698), washing
and laying out the body (Od. 24.18990), and closing the eyes of the corpse (Od. 24.296). The
phrase became proverbial in sepulchral contexts (Stobaeus cites it twice, in his sections
and ). Can we then be sure that even the early CEG I 40 truly aims to
compare the deceased specically with Sarpedon and evoke the gods discussioni.e., that it
is pointedly allusive in any meaningful sense? Or was the Homeric phrase recalled here,
rather, in a nonspecic fashion, as generally applicable in honoring the dead?
It is important to point out that the index of Homeric Zitate, Variationen, Anspielungen
in MS vol. 5 consists overwhelmingly of nonspecic epicizing, not intended to activate in the
reader the recollection of specic Homeric scenes.
18. An interesting exception is the sepulchral epigram from Alexandria for Aline (GV
1312 no. 34 Bernand). Despite its inauspicious setting along the road where (as it says)
cowherds and shepherds drive their animals, the poem self-consciously resists the preponder-
ance of potential passersby so as to speak only to a very special reader, one reared in the
Muses labors. On this poem, see Bing 1998b: 13335.
scriptions? Those found on public monuments of both the classical and

Hellenistic eras characteristically appear less specic, more limited in min-
ing the implications of an antecedent passage, and, on the whole, more
generic than that on the Ischia Cup or in literary epigram.19 In a second-
century b.c. marble epitaph from Cos, for instance (GV 1729 GG 207),
of which I cite just the rst four verses, a cherished slave is compared with
Homers famous swineherd Eumaios.
[] + ; [ ] [] V
" ( 6 D
+ , / ,

P/, . . . .
[Homers pen once cried aloud the faithful nature
of Eumaios in golden papyrus columns,
but, you, Inachos, even when youre in Hades, the uent stone,
through its ever-remembering script, will sing your prudent
counsel. . . .]
Here it may be better to speak of reference rather than allusion, since

there is nothing veiled about this eulogy: the epigram points explicitly to
Homer and to his character Eumaios. In keeping with that straightforward
mention is the fact that no direct knowledge of the Odyssey is actually re-
quired to appreciate the reference.20 The reader must only be familiar with
19. Allusion is rare even on prominently placed public monuments that would be easily
accessible for repeated viewing. The epigram on the tyrant slayers monument in the Athenian
agora is a good case in point (CEG I 430 150 FH FGE pp. 18689). It conforms entirely
to the pattern of nonspecic epicizing demonstrated by Friedlnder and Hofeit with re-
gard to verse inscriptions of the archaic era (discussed earlier in the present essay). The state-
ment of the rst verse, that truly a great light dawned for the Athenians (V / /
/) when Aristogeiton and Harmodius killed Hipparchus, is strikingly Homeric. The
phrase / appears at Il. 15.669 (in a clearly unrelated context), and the metaphor of a
heros bravery being a light to his people is likewise Iliadic. But although it is twice used of
Patroklos (Il. 11.797, 16.39) and though one might be tempted to see a parallel between the
pairs Harmodius and Aristogeiton, on the one hand, and, on the other, Achilleus and Patrok-
los, it is also used of Teucer (Il. 8.282) and Aias (Il. 6.6), and the metaphor becomes common
thereafter (cf. LSJ s.v. II). Thus, while we can certainly speak of Homeric coloring, the
epigram contains no pointed allusion to a particular Homeric passage or scene.
20. A comparably overt Iliadic reference appears already in the third epigram quoted by
Aeschines (Against Ctesiphon 18385) and Plutarch (Cimon 7.38.2 Simonides 40 FGE pp.
25559) as inscribed on the herms celebrating Athens victory over the Persian remnant at
Eion in 475 b.c. The poem cites as a paradigm for this action the Athenian leader at Troy,
Menestheus, whom Homer once called the best marshaler of battle among the well-armed
Danaans of those who came along (@ / K ( /
( = , vv. 34; cf. Il. 2.55254).
the cast of leading characters and possess summary information about

their part in the poem.21 Even if we dig a bit further for possible allusion,
what we nd is in much the same vein, if marginally subtler. For instance,
in the context of the Odyssean comparison in which the servant Inachos is
identied with Eumaios while his master is implicitly identied with
Odysseus, the phrase (v. 3) suggests a possible contrast
with Odysseus. For Inachos has only one , and that is a
very prudent one. Odysseus standing epithet in Homer, , occurs
with such frequency (eighty-ve times in the Odysssey alone, twenty-ve
times in the Iliad) and would be so broadly familiar that, once again, we
may wonder how much direct knowledge of the Odyssey is required to ap-
preciate the allusion.22
To make its points, then, the epigram demands of its audience only a
rudimentary level of literary schoolinga level one might already attain in
the primary phase of instruction. For poetry lay at the core of Greek edu-
cation from the moment a student had mastered his letters (Plato Protag.
325e, 339a), and even a minimal cultural package, as Raffaella Cribiore
has called it, contained some Homer.23 As is well known, papyri of
Homer are ten times as numerous as those of the next most popular au-
thor, Euripides, with copies of the Iliad three times as frequent as those of
the Odyssey. Over half the Iliad papyri cover the rst six books. Similarly, in
21. The same is true of inscribed epitaphs that, e.g., compare a deceased wife with Pene-
lope. Cf. discussion later in the present essay and, generally, GV p. 517ff. It is worth noting
that, for whatever reason (and it would be interesting to speculate), such summary reference
to paradigmatic Homeric characters does not appear in epigram before the fourth century
(e.g., in CEG II 888.3850, esp. 4445). Cf. also the epitaph of the poet Philikos of Corcyra
(SH 980 GLP 106). Since Corcyra was equated with Scheria, the island of the Phaeacians,
Philikos is said (vv. 59) to have happily seen the festive old age of Alkinoos, a man who
knew how to live, born of Alkinoos line . . . from Demodokos . . . (translation adapted
from Page: " + ,
" / / , 6D /
/ 6 5 6 G / ] . Thus Demodokos is seen as either Philikos
ancestor or his model. On this poem, cf. Fantuzzi 2007a.
22. A more sophisticated allusion may be lurking behind , / in v. 3. The
phrase and its position is Homeric (at Od. 11.211, Odysseus complains that he cannot em-
brace his mother in Hades: so that even in Hades with our arms embracing / we can both
take the satisfaction of dismal mourning? vv. 21112). It also occurs at Theocr. 16.30, where
the topic is likewise posthumous fame through song. Given the reference to Eumaios in
Theocr. 16.5455, there may be direct inuence. Conference participants in Giessen won-
dered, further, whether and (vv. 34) might cleverly recall the Iliads
opening words, .
23. Cf. Cribiore 2001: 17879: Some Homer, a bit of Euripides, and some gnomic quo-
tations from Isocrates formed the cultural package of students at the primary level. Cf. also
ibid., 194: Starting from Homer is mandatory, since he was considered in antiquity the poet
par excellence . . . and the teacher who inspired reverence from the early years of study.
school exercises preserved on papyrus, the rst two books of the Iliad are
the most popular.24 Though the Odyssey was not nearly as prevalent as a
school text, it is interesting that the allusions to that poem that I have
found in inscribed epigrams tend to favor the beginning of the epic.
That is the case already in the classical era. The epigram CEG II 755
provides a good baseline. Dated by Hansen to between 400 and 350, it was
inscribed on a column found in a well by the Asclepieion in Athens and
dedicated to Asclepius. The names of the dedicators and the god appear
extra metrum. The remainder consists of the following lone hexameter:
[] 5 | [,]5 .
Concerning this verse, Hansen laconically remarks, Ecce Ulixes Athe-

niensis (1989: II ad v. 2). And he is doubtless right. Suffered and saw a
lot: that, in a nutshell, is the hero of the Odysseys opening lines ( /
C / 6 M , 1.4; / : , 1.3).
When one adds the fact that the subject of either poem ultimately survived
with the aid of a god, the equivalence is striking.25 Admittedly, this nut-
shell is a tiny one, even by the standards of epigram. Yet though it is re-
duced here to its barest minimum, the Odyssean prologue was clearly
meant to be heard. And it can be, because the prologue was such a cultural
commonplace that the mere use of with aorist forms of and
C sufce to evoke it.
A comparable, though more elaborate, instance appears in a votive epi-
gram of the Hellenistic era, dedicated to Pan and transmitted as a grafto
on a piece of broken sandstone from Upper Egypt, probably during the
reign (221204) of Ptolemy IV Philopator and Arsinoe III (Householder
and Prakken 1945 FGE Anon. 150, pp. 46265 no. 164 Bernand).
" M 6[]| , T
[6]| ,
, []| / - - - -
[+ (?) 6 ]| / [M ],

" | 2[] 6 , 5
| 6 []
,[ | ] A "
[|] [] [ 6][] .
24. Cribiore 2001: 194.

25. Contrast the epigram GV 726, discussed in n. 40 in the present essay.
N[ , | ], / [] \ [] |

[ ][] [6]/ ,, [ | ] , | 10
[]" [] [+] , b [] [] |, [] []
[ ?- - - | ] / .
" [- - - - - - - - - - - - - - - - | ] | . . . . .
[This is for Pan, skillful in hunting, attentive to prayers, who saved

me from the land of the Trogodytes, when I had suffered much
during two
labors, as I returned from the sacred land that produces myrrh, and
from the Koloboi.
You saved us as we wandered on the Red Sea,
sending a fair wind to our ships as they were spun about on the sea,
piping with shrill blasts from your reeds,
until at last you brought us yourself into the harbor of Ptolemais,
piloting us with your hands, most tenacious in the hunt.
Now, dear Pan, save the city of Alexander, which formerly he rst
built in Egypt, most famous of cities,
and I will proclaim your power, dear Pan, when I come safely
to Ptolemy . . . . . . . . . . . . . . . . . and Arsinoe.
Skilled in the hunt. . . .]
The grafto probably comes from the shrine of Pan at El Kanas near Re-
deseh, a watering stop on the caravan route between Edfu (Apollinopolis)
in Upper Egypt and the Red Sea. The shrine was originally built by the
pharaoh Seti I to the god Amon Re, whom the Greeks identied with Pan
of the Prosperous Journey, or Pan Euodos.
The invocation of Pan as _ (v. 1: cf. [] [ 6] []-
, v. 8; ", v. 13) suggests that the god helped the dedicator in
connection with a hunting expedition, one consisting of multiple ships
(, v. 5). The geographical references of verses 2 and 3 situate the ex-
pedition along the southern coast of the Red Sea and beyond, as does the
dedicators description of how the god guided his ships to safety at the
southern Red Sea port of Ptolemais (v. 7). The latter reference also sug-
gests that the object of the expedition was elephants, as Ptolemais is ex-
plicitly attested as being founded as a staging ground for elephant hunts
(the multiple ships of v. 5 accord well with such game).26 Judging by the
Y
26. It was called Ptolemais Epitheras (6 ) or, as Strabo describes it, J
6 (16.4; cf. Pliny NH 6.171). Mention of Ptolemais also provides a terminus post
quem for the poem, as the city was founded (according to the hieroglyphic Pithom Stele of
noninscriptional handwriting, the poem is a draft that the dedicator made

on his way back north (to Alexandria?) following the perilous expedition,
perhaps with the intention of having it later formally inscribed on costlier
material.
The poem paints this expedition as a voyage into the beyond, an exotic
exploit among peoples whose very names carry a whiff of danger and
adventure. The term (v. 2), though here apparently referring
to actual tribes inhabiting the region southeast of Egypt, along the south-
ern half of the Red Sea, is often used of primitive peoples on the mar-
gins of the known world. In manuscripts, the name was frequently altered
to , an etymologizing explanation of the fact that they are
sometimes described as living in holes () or caves. Already in
Herodotus (4.183) their speech resembles the squeaking of batscave
dwellers par excellencerather than any human language. Similarly
creepy are the Koloboi, whose name means the mutilated ones, most
likely a reference to their practice of male genital mutilation (Strabo
16.4.5, 13).
Within this fantastical landscape, the storm-tossed dedicator appears as
a latter-day Odysseusthough in a narrowly restricted sense: he is the epic
hero strictly of the Odyssey prologue. By describing him as
(v. 2), the epigram clearly recalls verse 4 of the Odyssey, / C /
6 M .27 Similarly, the phrase in
verse 4 of the epigram echoes T / verses 12 of the
Odyssey.28 By recalling a passage that Greeks of even minimal education
would have known and could easily recognize, here at a shrine of Pan on
the very fringes of Greek society, our epigram offers its readers a chance to
feel their Greekness, to identify with the best-known hero of their culture.
I doubt that our dedicator was the only Greek voyager in these parts to
264) by Eumedes between 270 and 264 during the reign of Ptolemy II Philadelphus. Philadel-
phus was keen on acquiring elephants for military use, and we know of other Red Sea ports
founded during his reign for this purpose. The date of Ptolemais and the handwriting of the
grafto make it just possible that the Ptolemy and Arsinoe to whom the narrator says he will
proclaim the power of Pan are Philadelphus and Arsinoe II. It is more likely, though, that they
are Ptolemy IV Philopator and Arsinoe III (221204) and that the narrators plea to the god to
now save Alexandria (v. 9) refers to unrest among both Greeks and native Egyptians, which
increased markedly during their reign.
27. Indeed, is an Odyssean phrase (cf. Od. 7.224, 13.131), always applied
to Odysseus.
28. Cf. also the standing phrase 6 (e.g., Od. 3.106). Further aeld,
we may nd a more learned allusion in v. 7 of the epigram, A ". The phrase occurs
once in Homer (Od. 13.323), likewise with a mortal addressing a god (Odysseus to Athena)
who led him to safety: A " .
whom this comparison would have occurred.29 The narrators plea to the
god to now save Alexandria (v. 9) may refer to contemporary unrest
among native Egyptians, which increased markedly during the reign of
Ptolemy IV Philopator and which threatened the city.30 At such a time, a
reminder of ones Greek heritage may have been especially welcome.
Our next poem comes from an early third-century b.c. white marble
stele from Pherae in Thessaly (GG 424 SEG 23 no. 434k).
; | (, . |
, , C , . |
/ : 6 | |
1 [].
[Am I passing the tomb of Lampis? Tell me, stone.
Yes, stranger, she who was devoted to her children and her parents.
But go in safety and fervently beg the gods
that you may come to your shared house with such a wife.]
This sepulchral poem for Lampis is framed as a dialogue between passerby

and stele, a form that became popular and was subject to great elaboration
in the Hellenistic period, particularly in literary epigram.31 What immedi-
ately catches our attention in this example is the striking exhortation to the
stone in the opening line: (, .32 Can one fail to be charmed by this
unexpected variation on the epic (, , with its clear recollection
of the rst line of the Odyssey (cf. also Il. 2.761)?33 It is not simply the
29. It is worth noting that Odysseus was available as a model of suffering and exotic
journies as early as Theognis, though while our epigram evokes the hero through the famil-
iar phrases of the Odyssean prologue, Theognis makes the reference explicit (112324):
D / /, / @ / / / A 6.
30. The aorist imperativestrengthened by suggests impending danger to which
the deity is asked to respond immediately. Cf. Od. 4.765 (Penelope praying to Athena on be-
half of Telemachus): , ; Eur. IT 1082ff.: b
/,
\ / " / ( 6 , / ;
Aristoph. Vesp. 393: [sc., addressed to the Athenian hero Lycus] 6
. If the aorist imperative refers to present danger, we may be justied in
looking (as did Householder and Prakken, in their editio princeps) for historically attested
threats to Alexandria as one factor in dating the epigram.
31. The earliest examples are by Anyte (A. Plan. 231 19 GP) and Phalaecus (AP 13.5
2 GP). The only earlier inscribed instance that survives intact is in a dedicatory inscription
on a statue base in Halicarnassus from ca. 475 (CEG I 429). Cf. Kassel 1983; Rasch 1910.
32. CEG I 429 likewise begins with a remarkable exhortation: " ,
/ [] | (cunningly wrought voice of the stone, say who set up this offering).
33. Perhaps there is some play here with the folk etymology relating meaning
rock with meaning the living, breathing people with their ability to speaka play that
goes back to Il. 24.611: 1 + (cf. Hes. fr. 234; Pindar Ol. 9.4146).
wording, however, that echoes Homer. The part here played by the stone
is functionally equivalent to that of the goddess of song. The written text
displaces the living voice of the Muse as repository of kleos. For in the
sepulchral context, the stele is an agent of memoryand so a memorial,
every bit as much as is the Muse, Mnemosynes daughter. Simi-
larly, the situation implicit at the start of the epic is also present in our epi-
gramthe colloquy with an authoritative source: as the Homeric poem
may be viewed as one long response to the poets initial request for knowl-
edge, so the tale of Lampis dutiful existence may be seen as the stones
response to a comparable opening request.34 In alluding to the Odysseys
introductory appeal, then, the epigram implicitly suggests a genealogy of
its own form; it traces the origins of dialogue epigram to the conversa-
tion between poet and Muse at the start of epic song.
I believe that the clear allusion of (, to (, xes
the basic level to which the poem is pitched: toward readers of modest
educational attainment, who might be familiar with the best-known parts
of the Odyssey, such as its prologue and famous opening line, yet not much
more. The allusion is obvious enough, moreover, to be grasped by those
readers quickly. For sepulchral monuments have only a eeting chance to
make their case, set as they normally are along busy streets and appealing
to the wayfarer hurrying in or out of town on business. For an allusion to
have an impact under such circumstances, it cannot demand too much.
Yet the poem may contain other, more recondite allusions. In view of
the clear Odyssean color of the rst line, perhaps we may see in Lampis a
subtly Penelopean shading. It is worth noting here thatas with Eumaios
beforePenelope is certainly a model of virtue in inscribed epigram.
Thus, in a second-century b.c. poem from Kleonai in the Peloponnese
(GV 1735.14), we hear, Homer praised above all the much-admired
child of Ikarios, Penelope, in his writing tablets, but no one is powerful
enough to sing your virtue and renown, [Nomonia], with a clear voice
(/ + K[] / A []/ 6 (
D / / # _ 6[ ] / ,[]
[].). In a rst-century b.c. epitaph from
Y
Didyma (MS 01/19/43 v. 4), we learn of Gorgo, who in Miletus was a
Penelope among the Ionians ( [] 6 /).35 In
that light, our epigrams emphasis on marriage as a parternship of equals in
34. For ( used in inscribed epigram to question a monument, cf. MS 20/06/01

(sixth century a.d.), 01/12/02 v. 1. Cf. also AP 7.679.2.
35. Cf. AP 15.8.6; GV 727, 1736, 1737.
verse 4 may be signicant. Normally, a groom is described as taking a bride

to his house.36 Here, they go there together (reading adverbially) or
else to their shared house (reading adjectivally with ).37 In
either case, the emphatic placement of 1 or at the start of each half
of the pentameter suggests that a real partnership is envisioned, like that
enjoyed by Odysseus and Penelope.38 Further, despite its reference to chil-
dren and parents, the epigram tactfully omits all mention of Lampis hus-
band39notwithstanding the fact that the stone addresses itself to a male
passerby, exhorting him to pray that he share a house with such a wife. Does
the idea of the missing husband evoke the absent Odysseus? Finally, the ex-
hortation in verse 3 to pray earnestly to the gods ( 6
) recalls an Odyssean formula in the same metrical position, and he
prayed to all the gods ( 6 , Od. 14.423; cf. 21.203,
20.238), always occurring in connection with a prayer for Odysseus to
come home to his home ( . . . @ ; cf. in
v. 4 of Lampis epitaph). Implicitly likened to Odysseus, then, the reader is
urged to seek a bride who is similar to Penelopean Lampis.
These latter allusions are, of course, not remotely as clear as (,
. Yet if they are at all plausible, they suggest that the inscription could
also speak to other readers with diverse levels of literary knowledge. Who
would they be?40 It is important here to stress that one cannot confront
36. Headlam (Headlam and Knox 1922: ad Herodas 5.70) cites Il. 9.288; Hes. WD 695;
Hes. Th. 410; etc.
37. Another possibility is that refers to Hades, the house shared by all. Cf.
CEG II 593.4: . . . / ( . To be sure, the desire of
husband and wife to be together in death is not uncommon (cf. Lattimore 1962: 24750), and
Euripides Admetus even begs his dying wife to take him along with her to Hades (Alc. 382).
But a prospective wish of traveling to Hades together with ones wife (i.e., of dying at the
same time), expressed (as here) on behalf of a passerby, is neither realistic nor a commonly
voiced ideal. On balance, then, refers more likely to a house than to Hades.
38. The ideal of such harmony in marriage (C) is articulated by Odysseus
himself in his wish for Nausikaa (Od. 6.18085).
39. Was Lampis a widow? If that was the case, one might expect some reference to her
pious regard for her husband or for his tomb, to match that mentioned with respect to par-
ents and children.
40. This question comes up similarly and with particular urgency in a late epitaph (GV
627, second to third century a.d.) for a doctor from Thasos named Antiochus (the following
discussion owes much to the stimulating comments of Regina Hschele).
/ C , T 6 ,
T
: ( ,
a 1 6,
J , .
The deceased doctor presents himself in rst-person speech in terms of the Odyssey prologue.
His medium, appropriately enough, is dactylic hexameter. Like the epic hero of Od. 1.3, he
this question in the same way for an inscribed monument as for a text on a
scroll; their modes of reception are too different. A scroll is portable; its
reader may keep it with him, study it at his leisure repeatedly and under
different circumstances, compare its text with those on other scrolls. In-
scribed monuments, by contrast, are xed in a single place, their connec-
tion to the reader precarious, dependent as it is on his presence and good-
will. Thus, when the epitaph of Lampis urges its reader, in verse 3 to go
in safety (/ : ), it does not reckon with return visits for
further contemplation. To my knowledge, no ancient inscription envisions
such a second chance: the etiquette requires Farewell, not Au revoir.41
Who, then, would be likely to grasp those more recondite allusions?
Notwithstanding the fact that the epigram is explicitly addressed to an
anonymous wayfarer (the of v. 2), the most likely answer seems to be
the members of Lampis family. These were plausibly offered a papyrus
copy of the poem for approval before it was set in stone.42 From that
says, I have seen the cities of many men and known their mind. Thus far the Odyssean
color could not be more obviousas with (, in Lampis epitaph.
The allusion becomes marginally more complex, though still readily comprehensible, in
v. 3, where the doctor may be claiming to have outdone his epic model inasmuch as he has
saved many people. Odysseus had been unable to do so (/ "/ u 2 6,
, Od. 1.6). Indeed, his nickname , preserver (v. 1), conrms the doctors
claim 1 6 (v. 3); Homer, by comparison, plays on the derivation of Odysseus
name from 4 (Od. 1.6062, etc.) and highlights his heros destructive aspect:
( (Od. 1.2).
Though glorifying its subject through the comparison with Odysseus in vv. 23, the epi-
grams relation to the Odyssey acquires greater complexity in v. 4, when it suddenly tilts to-
ward self-mockery. This physician, it turns out, is no Odysseus after all. Not for him the
happy nostos, for he failed to make it home, having been poisoned while still on his travels
(J , v. 4). Did the doctor mistake his own ? Was he a victim of foul
play? The mystery is left unsolved. Here, too, however, there is a specically Odyssean point.
At Od. 1.261, Athena (as Mentes) says she rst encountered Odysseus when he was on a jour-
ney gathering deadly poison to smear on his arrows ( ). The
real Odysseus is thus a master of , able to withstand their effects (as in the Circe
episode). Poor Antiochus was simply not as cunning as Odysseus.
In view of the epigrams apparently self-deprecating wit, we are left wondering, who
would have written this, and for whose delectationthe doctor himself as he lay dying, a poet
commissioned for the task (perhaps imagining that family members wouldnt get the refer-
ence), a family member (one of those mentioned in the prose inscription under the epigram,
i.e., this Odyssean doctors loyal parents, son, and wifedescribed as , despite his
being frequently away)? Humor is not especially common in inscribed epitaphs, so one may
well wonder at whom this allusion is aimed.
41. For the dynamic of the readers interaction with the inscribed text, cf. Ergnzungs-
spiel in the Epigrams of Callimachus and The Un-Read Muse? in the present volume.
42. Such a scenario seems likely in the case of a papyrus letter containing a pair of sepul-
chral epigrams to commemorate Tauron, the hunting dog of Zenon, agent of Apollonius,
chief nancial administrator of Ptolemy Philadelphus (SH 977).
source, repeated reading would be as easy as with any scroll. But even at
the grave site, it is family members who might be expected to make regu-
lar visits and so have the chance to scrutinize the poem most thoroughly.43
In any case, whether dealing with that baseline audience to whom the
more obvious allusion is pitched or with those apt to detect the subtler ex-
amples, the readers of the inscribed epigram look rather different and op-
erate under starkly different conditions from those we generally imagine
as consumers of literary texts and interpreters of their allusions.
I want to illustrate that difference through an example of allusion in a
literary epigram by Callimachus (43 Pf. 13 GP AP 12.134) and so
bring us back to the symposium as a favored setting for allusive play.
K ( C 6D E
(,;) ,

?/ (, + 8
b 6 D
e D " 8
,, / : 5 (.
[We didnt notice that the stranger has a wound. What a painful
sigh he heaved through his breast (did you see?)
when he drank his third cup, and the roses, shedding their petals,
fell from the mans garlands all onto the ground.
Hes burned, and bad! Im not just guessing,
by the gods. A thief knows the tracks of a thief.]
The poem is a self-conscious variation on one by Asclepiades (18 GP AP

12.135), where a speaker illustrates the opening apophthegm that wine is
the test of love (, ( (, v. 1), by recalling a symposium in
which frequent toasts ( . . . , v. 2) revealed the symp-
toms of love in Nikagoras, despite his denials: he cried, he hung his head,
he looked downcast, his wreath would not stay on (6, 6,
+ (, " ( , vv. 34). Our epigram shapes this
unelaborated recollection into a drama that unfolds before our eyes. The
speaker becomes a character at the symposium, an intelligent, compassion-
ate observer whowith a condential ,; (v. 2)invites us to imagine
ourselves his drinking companions, pointing out to us a guests painful sigh
43. I suggested this explanation for the subtle Odyssean allusion in CEG II 597, dis-
cussed in n. 17 in the present essay. For continuing visits to family grave sites, cf. Garland
1985: 10420; cf. also Venit 2002: 18789.
already at the third toast (as opposed to . . . in Ascle-

piades v. 2)44 and the rose petals from his wreath fallen to the ground
telltale signs of previously undetected erotic suffering (6, v. 1).
With a witty closing apophthegm to match the one with which Asclepiades
began,45 the speaker acknowledges that he speaks from experience (
/ : 5 (, v. 6).
The concluding proverb is attested in Aristotle: ( + ,
(Eth. Eud. 1235a9). To this, Callimachus signicantly
adds tracks, :, a reference to the signs of erotic afiction detected by
the speaker, but alsoas Marco Fantuzzi astutely observesa metapoetic
comment: Callimachus knows well the : of his model Asclepiades. But
further, as poets of erotic verse, both authors know well and draw on the
tradition of erotic literature that includes such physical signs. From this
perspective, as Fantuzzi suggests, the phrase " 8 in verse 5
seems to be used pointedly: the speakers conjecture is not musically out of
step with the tradition.46
Inasmuch as the reader is addressed in the parenthetical ,; of verse
2 as a fellow observer, he is invited to join in the process of discovering the
: embedded in the scene. And indeed, there are further tracks in this
poem, which to my knowledge no reader before now has noted. For our
poet has modeled his dramatic situation on two Odyssean passages, Odys-
sey 8.9395 and 8.53235. There, at the feast of the Phaeacians, following
the rst and third song of Demodokos, Alkinoos observes Odysseus
whom he, too, refers to as C in each passage (8.101, 541), in the
same metrical position as in our poemsuffering with sighs in the midst
of libations (cf. 8.8889): There, shedding tears, he went unnoticed by all
the others [6, as in our poem], / but Alkinoos alone understood
what he did and noticed, / since he was sitting next to him and heard him
groaning heavily ((/ + 6 , /
/ , 6/ V/ 6 / \ / " , 1 +
note how Callimachus elaborates 1 -
into a painful sigh he brought up out of his chest). By recalling
44. Thus Giangrande (1968: 12022).

45. Thus Ludwig (1968: 313).
46. Fantuzzi and Hunter 2004: 33840 with n. 181. The word 8 may denote any
regularly recurring motion (e.g., in music) and, by extension, a recognizable pattern in
human behavior (cf. Archilochus 128.7 West, Anacreon PMG 416.2), artifacts, or natural
phenomena (cf. LSJ s.v.). Hence, " 8 may mean that the speakers guess is not
out of line (i.e., it ts the pattern). Gow and Page (GP ad loc.) suggest the translation My
, in step, rhythmically.
conjecture is not out of step, comparing such phrases as 6 8M
Alkinoos, Callimachus characterizes his speaker as an ideally knowing and

sympathetic participant in the symposium, for the Phaeacian king came to
be viewed as the paradigmatic sympotic host (as we see, e.g., in the sepul-
chral epigram for the third century b.c. Corcyrean poet Philikos, who
saw the festive old age of Alkinoos, a man who knew how to live (" +
,5 " / / , 6, SH
980.56).47 The allusion also underlines the very different circumstances
between epic and epigram: while Alkinoos recognizes the marks of heroic
suffering, our speaker detects those of erotic distress.
Aristotleit is interesting to notecites this same Odyssean model in
his Poetics (16.1455a) to illustrate a type of recognition () that
he terms recognition through memory ( ). There,
it is Odysseus who, spurred by the text of Demodokos songs, tearfully re-
members his own exploits and so precipitates his recognition ( ? 6
/ M, 6,
@ ). Was this Aristotelian passage lurking in Calli-
machus mind when he incorporated the reference to Odyssey 8 into his epi-
gram, and could he have seen in it a commentary on allusion? I suspect he
did. For the challenge of allusion in the reception of a poem might likewise
be termed recognition through memory: for readers, as for Odysseus, a
text may trigger the memory of an earlierreaderlyexperience and so
bring about a recognition.
But to an extent that goes beyond what Aristotle suggests about the
case of Odysseus, the reference here is oblique, condensed, and deliber-
ately veiled. Indeed, allusive artistry is gured as a larcenous cat-and-
mouse game between poet and reader in which the former conceals the
lootthat is his reference to poetic precursorsby cloaking it in a differ-
ent context, while challenging us to play detective and nd him out:
/ : 5 (.48 Metapoetically speaking, Callimachus here made
47. Cf. also the late second-century b.c. inscribed epigram on Delos by Antisthenes of
Paphos (ID 1553, with forthcoming discussion by Garulli) honoring Simalos of Salamis in
Cyprus as a second Alkinoos: / [] , / ,
[] (vv. 12); Athen. 1.9a: And Alkinoos, whose choice inclined to
a luxurious life; he feasted the Phaeacians, who lived most luxuriously, and entertained the
stranger Odysseus (/ + C J? 1 2
/ ).
It is worth noting as well that Alkinoos and his queen, Arete, had currency in third-cen-
tury b.c. poetry as epic models of the Ptolemaic royal couple; cf. the comments on Theocr. Id.
17.3839 in Hunter 2003b: 12829. See also Fantuzzi 2006.
48. It is probably not coincidental that verbs of concealment abound in Callimachus
model passage in Odyssey 8: Odysseus there draws a mantle over his head (
25 . . . / :, vv. 8485), hides his face ( + , v. 85; cf.
off like a bandit, his theft from Homer undetected for so long
6, he might have saidwhile we the readers appear momentarily
clueless, unable to read the signs, even in a poem that is all about noticing
(literary) tracks.49
In what kind of setting will readers have engaged in such a learned pas-
time? Denitely not as they glimpsed an inscribed monument in a busy
street or bustling shrine. In such surroundings, they would not have been
equal to the games demands, for their encounter with the text would be
eeting. People rarely went out with the aim of reading inscribed poems.
As one such poem puts it, the everyman addressee strides along the road
with his mind on other things ( h[] : / | : :
, CEG I 28). If he succumbs to the epigrams appeal and pauses
to read, the stop is envisioned as incidental and brief. As we have seen,
poems destined for public settings adapt their allusions to these difcult cir-
cumstances of reception and to a readership of modest education, by mak-
ing them unmistakable.
That is a far cry from the audience Callimachus requires for his cun-
ning game of detection. It is no coincidence that he evokes an idealized
image of a sympotic gatheringsuch as that proposed for the Ischia Cup
as the setting for his game. Yet for the reader to discover the layers of allu-
sion (not just Asclepiades but the Odyssey, too), he must bring to the task an
ingenuity and erudition beyond that assumed by the humble cup; he must
have time; he must also have the patience to persist with repeated reread-
ing and reection and, yes, a thievishness to match that of the author. Even
then, one can only say Thank goodness there is no statute of limitations on
discovering allusions. Here, once more, the medium of communication
and context of reception are crucial. The sine qua non of such allusive play
is, to my mind, the epigrams setting on a tablet or scroll. For there it can
be carried around with ease, is available wherever and whenever its reader
wishes, and may be set beside other texts, commentaries, glossaries, and so
on. Against the backdrop of reading allusion in inscription, this sort of
reception, with its literary detective work, turns out to be an elite pastime
indeed.
, v. 92), wipes away the traces of his tears (/ 4, v. 88),
etc. For a helpful discussion of allusion as theft, cf. Hinds 1998: 2225. As far as I know, this
gure for allusion has yet to be investigated in Greek literature.
49. Here (with 6 in v. 1 ,; in v. 2 and ( in v. 6) and elsewhere in his
epigrams, Callimachus is interested in the mental process that leads a speaker from igno-
rance to comprehension. (Cf., e.g., 30 Pf. 12 GP AP 12.71 (" (, v. 2; (, v. 5);
41 Pf. 4 GP AP 12.73 (" ,/, v. 1; ,/ @, v. 6); 15 Pf. 40 GP AP 7.522 ( /
6; v. 1 _ / [ 6, / , . . . , vv. 12). Cf., generally, Walsh 1990: 121.
By way of a coda, I want to return briey to the symposium as a site con-

ducive to allusive play. This setting and the poetry proper to it has brack-
eted my discussion of Odyssean allusion on inscribed monuments, provid-
ingfrom within an elite domain where learned poetic reference was more
commona foil for the less-challenging verse set out in public spaces,
along the broad, well-trodden street. Our trajectory, however, from Ischia
Cup to Callimachean epigram has also encompassed divergent eras, sepa-
rated by almost half a millennium, from early Greece to the third century
b.c. Between them, the symposium had accrued new meaning.
A dening feature of early sympotic verse had been its occasionality.
The poem on the Ischia Cup bespoke its occasion with particular clarity,
as it was inscribed on an object specially made for use at an actual sympo-
sium. Though enduring into the Hellenistic era as a vibrant social institu-
tion, the symposium had by then also become literature. As evoked in
Callimachus, there is no necessary link to an actual occasion: the setting
of his epigram is neither so specic as to be xed to a particular historic
moment (as, e.g., in the poetry of Alcaeus) nor so generic as to apply to
any sympotic gathering at which it might be performed (as with the
Theognidea). The idealized party in Callimachus is located, rather, in the
space of the scroll and set to unfold in the act of reading. That, as we saw,
was where one might best engage in that game of detection that the
poems cunning speaker invited us to play. And there the symposium be-
comes something different. It becomes, I would argue, a gure for the
diachronic community of readersa community whose members count
as their close companions the poet-readers of all eras, as well as their cre-
ations. This literary fellowship, in other words, cuts across time, so that
even now, if we bring to it the requisite skills, we can join the privileged
circle. Even in the present, those belonging to this hetaireia can, if they so
wish, cross the threshold at any time and nd the festivities in progress
simply by opening the book. The hetairoi are always there: their party is
never-ending.50
A well-known epigram of Asclepiades (16 GP AP 12.50) may provide
a glimpse into that readerly party.
50. Cf. the Greek view of the afterlife as an eternal symposium, at least for the blessed
(Plato Resp. 363cd; Pindar fr. 129, the Tomb of the Diver at Paestum). Socrates looks for-
ward to meeting Orpheus, Musaeus, Homer, and Hesiod in the underworld (Apol. 41a): X "
/ M ;M ; M 6 M / [ # ;
65 + 6 , / 6
. 6 ( "M
[
: ? " . . . Cf. Cic. Sen. 83; Ael. VH 13.20.
/, /. ; ;
" + 6,
"/ 6 M ,1
P. 6 J ;
D . 5
, ;
V
D " ( " ,
, / .
[Drink, Asclepiades. Why those tears? Whats wrong?

Youre not the only one harsh Kypris dragged off.
Did bitter Eros sharpen his arrows and his bows
against you alone? Why sit in the ash if youre alive?
Lets drink the pure potion of Bacchus. The dawn is nger thin.
Shall we wait to see the lulling lamp again?
We drink. There is no love. And it wont be long now,
poor guy, until we rest through the vast night.51]
It has long been recognized that Asclepiades draws heavily here on earlier
sympotic song. The poem signals as much to the reader by insisting
twicethat Asclepiades is not the only one to have been injured by
Aphrodite and Eros (vv. 24); there have been others before. Readers
should thus be on the lookout for (literary) precedents. They hear this,
moreover, from a speaker who evidently knows the (poetic) subject of
which he speaks. Who is he? This is a question to which I will return.
Asclepiades chief model is clearly Alcaeus. The specics of his debt
have been analyzed most recently by Richard Hunter (2008), who focuses
primarily on Alcaeus fragment 346 (Voigt). The archaic model, as Hunter
plausibly suggests, is especially evident at the start of each of the epigrams
two quatrains, with their typically Alcaean calls to drink. But rather than
follow Hunters view that fragment 346 does double duty as the source for
both these echoesthe rst a somewhat muted one (D , fr.
346.1 /, . . . , v. 1), the second conspicuous and far-reaching
(D / 4; , fr. 346.1
D D / V ,
, vv. 56)I would argue, rather, that the Hellenistic poet con-
ated two Alcaean models, with verse 1 actually alluding to fragment 38.1
51. Translation adapted from Bing and Cohen 1991: 13334.

(Voigt) ( [ . . . . . . . ] / / ( . . . [] /,

/. ; ; v. 1).52 Seen thus, the formal
correspondence is very close indeed: Both poets start from (1) an impera-
tive call to drink, directed toward (2) a singular addressee, mentioned (3)
by name in the vocative, and followed most likely by (4) a question intro-
duced by .53 And just as Asclepiades justies his appeal by reference to
the inevitability of death (vv. 4, 78), so does Alcaeus in fragment 38 (vv. 2
10)something that is not the case, so far as we can tell, in fragment 346.
As Hunter notes, the epigram shifts in verse 5 from the private to the
public, from the anti-social self-absorption of the injured lover to the sol-
idarity of the drinking group, marked by the switch to the rst-person plu-
ral (2008). It is here, in this second, more prominent echo (vv. 56 fr.
346.1), that the voice of Alcaeus is most clearly to be heard. Indeed, the
similarities are so greatwith the plural exhortation to drink (
), the question of whether to wait for the lamps ( /
4; V . . . , ), the reference to the ngers
breadth of daylight ( )54that we seem
to hear Alcaeus very words.55 Bearing that in mind, the speakers call for
drink in response to the specically amorous plight of Asclepiades may
be viewed as likewise true to Alcaeus words: the archaic poet justied his
demand for wine by calling it : ,
/ (/ (for the son of Zeus and Semele gave
wine to men to make them forget their worry, fr. 346.34). Our speaker
52. The link to Alcaeus fr. 38 is highlighted most recently also by Guichard (2004: 262
63). Cf. also Acosta-Hughes and Barbantani 2007: 45455.
53. It is, of course, not certain that the letters [ represent the interrogative pronoun
. This interpretation, which I adopt from the text of Page (1955: 300), is, however, very
plausible.
54. Striking here is not just the metaphorical use of but the ellipse of the verb.
In addition, Guichard (2004: 262) suggests that Asclepiades signals these echoes through
dialectal coloring: Llama la atencin el que todos los dorismos [, v. 5 , v. 6,
, v. 8] se encuentren en partes del poema que hacen variaciones del texto de Alceo (cf.
also 26667).
55. Of course, those words are also signicantly altered throughout, generally with an
eye toward intensifying the archaic model. For instance, Alcaeus demand to combine two
parts wine to one of water (( ( , fr. 346.34)already a strong mix!
becomes in Asclepiades a call to drink wine straight ( , v. 5). Alcaeus urges his
companions to start drinking already before the lamps have been lit and while there is still a
trace of daylight ( / 4; , fr. 346.1); in Asclepiades, the drink-
ing has already stretched through the night, dawn has been sighted, and the speaker wonders
if they should just keep drinking until the lamps are lit again ( . / V
, ; vv. 56). Despite extending their drinking bout right through
the next day, " , / , / .
merely instantiates and claries the previously unnamed ; their na-

ture is erotic.56
This leads us back to what Hunter (2008) terms the vexed question of
Who speaks? Scholars have construed the unidentied speaker in vari-
ous ways. Some consider that Asclepiades is here talking to himself; oth-
ers that he soliloquizes in the rst quatrain, while addressing his hetairoi in
the second. A different approach is to take the speaker to be one of those
drink mates, doing his best to console the lovelorn Asclepiades.57 I
would like to propose a further possibility, one that develops with greater
specicity the idea that the speaker is a fellow reveler at the symposium.
Given how distinctly we heard Alcaeus words in this poem and how
pervasive his presence and inspiration is throughout, could not the speaker
be the archaic poet himself ? All poets, most especially those of the Hel-
lenistic era, acknowledge in one way or another that their predecessors
speak to them. Asclepiades, I suggest, has dramatized the encounter.58 He
does so, moreover, in a typically involving way. Naming no name beside his
own, he relies on his readers knowledge of literary history to detect the
source of the speakers words.
The setting in which one may hear Alcaeus speak to his Hellenistic coun-
terpart is, of course, no ordinary symposium. It is, rather, that perpetual
party that stands for the diachronic fellowship of those immersed in books.59
56. Hunter rightly stresses that Asclepiades reading [of ] may of course
have been a creative misreading, as we are unable to know whether Alcaeus were po-
litical or personal in nature (2008).
57. For the rst possibility, cf., most recently, Guichard 2004: 26263: La invitacin a
beber es un estilema de la poesa simposiaca; la novedad de Asclepades consiste en que se
hace la invitacin a s mismo. The second possibility is raised by Hunter (2008). The third
and most common has been argued recently by Gutzwiller (1998: 14849).
58. The device is comparable to those whereby poets seek to negotiate the gulf between
present and pastfor instance, Homers reincarnation as Ennius or Hipponax return from
Hades to address the scholars of Alexandria (cf. Bing 1988: 6272). Similar strategies crop up
in the Renaissance, e.g., the dialogue staged by Petrarch in his Secretum between himself and
St. Augustine or Petrarchs letters to ancient authors in the Familiarium Rerum Libri. On the
latter, see the insightful analysis of Hinds (2004).
Note that Hipponax statement in Callimachus that he does not have much time before
going (back) to Hades ( "/ " / D / ]

/, vv. 3335) is comparable to Alcaeus closing remark in the epigram that there is
not much time any longer before we will sleep the sleep of death ( "
, / , / , vv. 78).
59. A far later, but remarkably comparable, instance appears in Machiavellis letter to
Francesco Vettori, 10 December 1513, which dramatizes how the Renaissance present may
commune with the classical past:
When evening comes, I return home and enter my study; on the threshold I take off my
workday clothes, covered with mud and dirt, and put on the garments of court and
Here, for however long they attend the festivities, the normal linearity of
time dissolves, and three distinct moments converge: the archaic past of the
lyric poet, the Hellenistic present of the epigrammatist, and the future of
Asclepiades readers. To the extent that we recognize the voice and feel our-
selves addressed as fellow revelers in the exhortation of drink (v. 5) and in the
poems rst-person plurals (vv. 68), we may join the group. Those who do
are among the passionate readers for whom an ancient poets words are as
full of life as a contemporarys. It is to such an audience and in that readerly
environmentfar from the broad, well-trodden streetthat the allusive
artistry of elite Hellenistic poetry is aimed.
palace. Fitted out appropriately, I step inside the venerable courts of the ancients, where,
solicitously received by them, I nourish myself on that food that alone is mine and for
which I was born; where I am unashamed to converse with them and to question them
about the motives for their actions, and they, out of their human kindness, answer me.
And for four hours at a time I feel no boredom, I forget all my troubles, I do not dread
poverty, and I am not terried by death. I absorb myself into them completely.
(Translation from Atkinson and Sices 1996: 26265)
Note here particularly how, with its evocation of food and table talk in the company of the
ancients, this passage resembles quite closely that transtemporal symposium where (accord-
ing to my interpretation) Alcaeus can address Asclepiades. I am grateful to my colleague
Garth Tissol for bringing this passage to my attention.
chapter 9
Reimagining Posidippus

I n 1992, an Italian bank, the Cassa di Risparmio delle Provincie Lom-

barde (CARIPLO), made a generous donation to the Universit degli
Studi di Milano, allowing it to acquire a mummy pectoral that contained a
hidden treasure. The pectoral was made of cartonnage, a kind of . . . pa-
pier mch built up of sheets of used papyrus.1 In this instance, the car-
tonnage consisted of some documentary papyri and a substantial scroll,
which turned out to hold remains of 112 epigrams, some six hundred
verses divided into sixteen columns. The papyrologists Guido Bastianini
and Claudio Gallazzi made the nd public in a preliminary report in 1993.
Here they established that two of the poems were previously known and
attributed already in antiquity to the mid-third-century b.c. epigrammatist
Posidippus of Pella.2 Moreover, as the scroll contained no further indica-
tions of authorship and as diction, meter, style, and range of historical ref-
erence in the remaining epigrams were largely in accord, they concluded
that Posidippus was in all likelihood the author of the entire collection.3
1. Turner 1968: 24.

2. The two were the poem on the snakestone, cited by Tzetzes (Chil. 7.660 20 GP
II 39III 7 BG 15 AB), and the poem on a statue of Alexander the Great by the sculp-
tor Lysippus, preserved in the Planudean anthology (APlan 119 18 GP X 3033 BG
65 AB).
3. That is now the communis opinio, as one can see from the unanimity of the contributions
in Gutzwiller 2005b. It is worth recalling, however, that some scholars have questioned the at-
tribution in varying degrees: Lloyd-Jones (2003) believes that the Milan papyrus may have
been part of a multiauthored collection (possibly the so-called Soros); Ferrari (2004) considers
the epigrams of the papyrus to be mostly by Posidippus but thinks it included variations
177
This epigrammatists work had until then been known only through a
fragmentary elegy of twenty-eight verses on his old age, which has been
characterized as the sphrags, or seal, poem of Posidippus (SH 705 118
AB), and some twenty-ve epigrams, mainly erotic and sympotic. There
had been earlier indications that there was also a political side to his poetry,
for several epigrams celebrated Ptolemaic monuments and cult. Signi-
cantly, these appeared only in poems preserved outside the Greek Anthol-
ogy, from the Firmin-Didot papyrus,4 or in Athenaeus.5
Bastianini and Gallazzi now indicated that there were important new
poems celebrating Ptolemaic royalty. Indeed, their sketch of the scrolls
contents promised something altogether different from the mainly erotic/
sympotic subject matter that had previously shaped our view of this poet.
They described how the epigrams were organized on the scroll into at least
nine sections according to topic, each with its own heading, the titles point-
ing to nothing resembling the erotic/sympotic subject matter known be-
fore. Some categories were familiar from those of the Greek Anthology,
some bizarrely different. In sequence, the headings announced poems on
Stones ([] ), Bird Signs (,), Dedications (-
), Epitaphs [6], Statuary ( ), Equestrian
Matters ( ), Shipwrecks (), Miraculous Cures (,-
), and a mysterious topic called Tropoi ().
On the basis of the documentary papyri preserved in the cartonnage, as
well as to judge by the hand, Bastianini and Gallazzi dated the scroll to the
late third or early second century b.c. Since we know that Posidippus was
active from about the 280s into the 240s, this meant that the papyrus was
on individual poems by other authors; Schrder (2004) subjects the new poems to a harshly
critical stylistic analysis and nds them unworthy eines erstrangigen hellenistischen Epi-
grammatikers (31), einer allgemein anerkannten Koryphe ihres Fachs (32), eines hoch-
hellenistischen Meisterdichters (42). To my mind, howeer, Schrder simply assumes the
high quality of the previously known poems of the old Posidippus, rather than demonstrat-
ing it. One may well ask, should Posidippus be considered in the rst rank of Hellenistic
poetson a level, say, with Theocritus or Callimachusrather than as a highly successful
second-stringer? In my view, it might be closer to the mark to say that Posidippus resembles
Salieri to Callimachus Mozartthough Posidippus at his best is probably better than
Mozarts contemporary.
4. Weil 1879. The epigrams (1112 GP 11516 AB) concern the dedication of the
statue of Zeus Soter atop the Lighthouse of Alexandria and the establishment of the shrine
of Arsinoe-Aphrodite Zephyritis at Cape Zephyrium by the Ptolemaic admiral Kallikrates of
Samos. On these poems, see my essay on Posidippus and the Admiral, ch. 12 in the pres-
ent volume.
5. Athen. 7.318d 13 GP 119 AB, a further poem on the shrine of Arsinoe-Aphrodite
Zephyritis; cf. also the reference at 10.415a 143 AB to an epigram on Aglais, who blew the
trumpet for the Ptolemaiia in the rst great procession of Ptolemy Philadelphus.
Reimagining Posidippus 179
produced not long after the poets own lifetime. This raised intriguing
questions: Was the order of the poems artful or academic? If artful, might
it derive ultimately from Posidippus himself; that is, was the poet simulta-
neously an editor, like his contemporary Callimachus? Ancient evidence
already suggested that Posidippus engaged in such activity (see the discus-
sion later in this essay on the celebrated case of Berisos). Did the Milan
papyrus constitute a concrete instance? In any case, the wealth of new
material evident even in this preliminary sketch justied its title, Il poeta
ritrovato.
The papyrologists, however, did not publish their ndings in a scholarly
journal. These appeared, rather, in the banks in-house publication, a glossy
magazine with the reassuringly steadfast title Ca de Sass (House of Stone).6
This report included a tantalizing photograph of the papyrus, just too small
to read (even with magniciation), and reproduced four of the poems in
full. That same year, Bastianini and Gallazzi brought out another publi-
cation, Posidippo Epigrammi: this slender, boxed volume intended as a gift
for the banks investors, beautifully produced on heavy paper and with ex-
quisite typesetting, contained a text and translation of twenty-ve epigrams
from the scroll, including the four already published.7 These publications
caused an enormous stir among scholars of Hellenistic poetry, but also
much frustration. First of all, they were hard to come by. Samizdat copies,
or copies of copies, circulated, some hopelessly faded from repeated copy-
ing. This led to ludicrous, if touching, scenes of scholars huddled together
at desks, bright lights trained on the washed-out letters of a much-thumbed
duplicate, excitedly trying to decipher the letters of a photocopy.8 Far more
seriously, however, the textseven when easily legibleoffered no schol-
arly apparatus or diacritical signs to indicate lacunae, uncertain readings, or
even whether they were extracts from longer epigrams. Thus, from a schol-
arly standpoint, they were virtually unusable.
That is how matters stood for about eight years. Then, in the autumn
6. Bastianini and Gallazzi 1993a: 3439.

7. Bastianini and Gallazzi 1993b; cf. the closing words of the preface: Questo prezioso
libretto contiene una parte signicativa di quegli epigrammi, stampati con la cura e la per-
fezione che un piccolo tesoro ritrovato richiede e che la tecnica rafnata del Polilo ancora
consente.
8. As for me, it occurred to me that the Cassa di Risparmio delle Provincie Lombarde
might have a branch ofce in New York. A glance in the phone book conrmed that it did.
Thus, during a visit to the city, I walked into the ofce and asked if they happened to have a
spare copy of each of the preliminary publications. The friendly receptionist asked why I
wanted them, and when I told her of my interest in Hellenistic poetry, she said that there
were plenty in the storage room and asked how many I wanted.
of 2001, the editio princeps nally appeared, a lavishly produced edition in

two volumes, with introduction, diplomatic and critical texts on facing
pages, and extensive commentary thereafter. The commentary, in particu-
lar, is a work of deep scholarship, which will doubtless provide the basis for
all work on the papyrus for the foreseeable future. The editors, Bastianini
and Gallazzi, acknowledge the far-reaching contribution of Colin Austin,
who shed light on innumerable doubtful readings and proposed many per-
suasive reconstructions. Within a year of the appearance of the rst edi-
tion, Austin and Bastianini produced an attractive editio minor, containing
not just the new epigrams but all known works of Posidippus, including
some that are doubtfully attributed. In this immensely useful book, with
facing-page translations in Italian and English as well as a comprehensive
index verborum and nominum, Posidippus suddenly emergedif we take
the uncertain attributions for genuineas the single best-preserved poet
of Hellenistic epigram.
Faced with the sudden quadrupling of his oeuvre, it makes sense to step
back for a moment and reevaluate our view of the poet or, better, reimagine
him. Who was this Posidippus? This much is certain: his is no household
name, even among classicists. No Latin poet ever declared his ambition to
be the Roman Posidippus. No Latin poet ever even mentioned Posidippus,
though a good case can be made that some read him and were inuenced by
what they read (Hutchinson 2002; R. F. Thomas 2004; Barchiesi 2005).9
What, then, do we know? My family is from Pella, he states in his
sphragis, the elegy on old age ( , SH 705.17 118.17
AB). That link with Pella, the royal seat of Macedon, which the poet
proudly declares in this elegy, is conrmed by an inscription from Ther-
mon of the year 263/2 (IG IX 12i, 17A.24 test. 3 AB), listing those to
whom the Aetolians granted proxeny. Among many others, they bestowed
this honor on Posidippus, the poet of epigram, from Pella ( [] -
6 ). Strong identication with his
Macedonian heritage, such as Posidippus displays here in his seal poem,
also emerges in the new poems of the Milan papyrus as something he
shares with his foremost patrons, the Ptolemies.10 As to Posidippus fur-
9. For his inuence on Greek poets of the imperial age, see Magnelli 2005; Gronewald
2004.
10. Marco Fantuzzi (2005: 25152) has productively explored the remarkable empha-
sis on the Ptolemies Macedonian ethnicity, especially in the : the Ptolemies deliber-
ately presented themselves not as kings of Egypt, but as Macedonians in the context of
pan-Hellenic games from which non-Greeks were excluded. With this ethnic designation the
Ptolemies may also have strengthened their persistent claim to be the real and legitimate suc-
cessors of Alexander the Great.
ther designation in the inscription as 6, I will return to

that shortly.
Pella has added a further piece to the biographical puzzle in recent
years, with a nd suggesting that family tradition may have predisposed the
poet toward initiation in the mysteries. That same seal poem of Posidippus
ends with the following wish: But in old age may I travel the mystic path to
Rhadamanthys, longed for by my people and all the commuity, on my feet
without a stick, sure of speech among the crowd, and leaving to my children
my house and my wealth (" 65 / , 6 ;-
/ 6, / 6
4 / @ / = 6, SH
705.2428 118.2428 AB). Excavations at a cemetery in the vicinity of
Pella suggest that a recent ancestor either paved this mystic path for
Posidippus or at least pointed it out to him. For a leaf-shaped lamella made
of gold, discovered in a tomb there of the late fourth century b.c., contained
the inscription D
(Attention Persephone: Posidippus the initiate is pious). As Matthew
Dickie has argued, this is not the epigrammatist Posidippus, but a citizen
of Pella from the generation of the grandfather of the poet and probably
related: If there was one member of the family who was an initiate in the
mysteries, it becomes more likely that the epigrammatist was also an ini-
tiate.11
In this light, it is striking that the rst three sepulchral epigrams in the
new papyrus may commemorate initiates in the mysteries (4244 AB), the
last one, signicantly, from Pella (44 AB). It is for a girl called Niko.
6 [ ] []
[ (] [] " [
, , 6[ ] [
5 [ ] A 6 4.
[The [youngest] of twelve children, a [lovely]
maiden, [was wept by] Pella and her ecstatic friends
three times, alas, [when Fate] brought down the servant of Dionysus
Niko, from the [Bacchic] mountains.]
11. Dickie 1998. As Dickie says (74), The name Posidippus is common enough, but the
likelihood is that Pella will not have had many men called Posidippus who were also persons
of substance, as the inhabitant of the grave evidently was, and who were unrelated to each
other. For the original publication of the lamella, cf. Lilibaki-Akamati 1989: 95101. Cf.
also Dickie 1995; Dickie 2005; Rossi 1996.
The maiden Niko evidently died while participating in the Dionysiac

mysteries. Destiny (Moira, the supplement in v. 3, looks plausible) had to
bring her back from the mountains (A 6 4, v. 4), sinceas
] . . . 4 here indicates (BG ad loc.)that was where the
[
frenzied Dionysiac ritual took place. Youthful though she was (vv. 12), she
probably died of exhaustion during the rites (Dignas 2004: 183 rightly
compares the tale from Plutarchs Mulierum Virtutes 13 underlining the
potential dangers from fatigue for participants in maenadic rites). The
threefold cry of ritual lament, , , at the start of verse 3, with which her
civic and religious community (and by implication, the epigrams readers)
bewail her, is balanced against the epigrams closing phrase, 6 4. This
is a pointed reversal of the ritual cry , = , = in the Bacchic rites,
the well-known refrain of the maenads in Euripides Bacchae (116, 165).12
Thus, as the epitaph suggests, as the body is brought back from the moun-
tain, the place of burial supplants the place of ecstasy, and the cry of grief
supplants the frenzied Bacchic call.
Besides his Pellaean origin, his familys link to local mystery cult, and
the poets interest in initiates of the mysteries from this vicinity, another
crucial fact about Posidippus emerges in the inscription at Thermon.
Otto Weinreich, the scholar who rst drew attention to the importance of
this text, shrewdly noted that the reference to Posidippus suggests that he
was being honored specically in his capacity as 6, that
is, as poet of epigrams to be inscribed on monuments and honoring no-
table men of the Aetolian League.13 That 6 refers to
inscribed verse is likely, since at this early date, the term 6 still
carried its literal meaning, inscription.
Even more decisively in favor of that words epigraphic connotation is
its very use here in an inscription. / is proper toindeed, vir-
tually a terminus technicus ininscriptions. Consequently, this verb and its
derivatives prompt medium-specic expectations in readers of an inscrip-
tion. That is especially so at a time when the metaphorical meaning of
6 (i.e., poems that had the form, but not the context, of physical
epigrams) was still new. The earliest probable use of 6 in its
12. Thus also Dignas (2004: 18283), who points to the epigram from Miletus (MS I
01/20/21) commemorating a priestess of Dionysus, who led her celebrants to the mountain
(# , v. 3). Cf. also Theocr. 26.2: 6 = "
, = V
6.
13. Weinreich 1918: 439: Die Vermutung liegt nahe, dass er solche [Epigramme] fr
den tolischen Bund gedichtet hatetwa Grab- oder Weihepigramme fr hervorragende
Mnner des Bundes, Steinepigrame fr historische Persnlichkeiten dieser Zeit.
broader, metaphorical sense of a short occasional poem is the title -

6 (P. Petrie II 49a SH 961) from around 250 b.c.
(Puelma 1996). The use of 6 in the inscription at Ther-
mon (263/2 b.c.) is clearly earlier. What is more, it occurs at a provincial
shrineno Panhellenic sanctuarywhere one would expect traditional ter-
minology, rather than innovative usage. Coming so early, in Aetolia, and
particularly in an inscription, 6 would scarcely have meant
anything other than a poet of epigrams to be inscribed on monuments.14
Why insist on this point? The inscriptional connotation of the term
6 at Thermon deserves emphasis, I think, because it deci-
sively shapes our image of Posidippus poetic enterprise, suggesting that
he actively wrote epigrams for inscription, even as he also wrote them for
the scroll. If this is right, he is a poet who worked on two different tracks,
that is, in two different media, though in a single genrethe rst such
poet that we know of with certainty.15 As I shall show later in this essay and
in subsequent essays in this book, he was keenly aware of the aesthetic
potential of each form. In either case, he understood that the medium was
key to the message.
The bestowal of proxeny here indicates that an epigrammatist might
achieve considerable status from such commissionseven though verse
inscriptions were traditionally anonymous, their poets named only in ex-
ceptional cases. The presence of a Posidippus, together with a certain
Asclepiades (possibly the contemporary epigrammatist and associate of
our poet), on a proxeny list at Delphi in the mid-270s, (Fouilles de Delphes
III 3 no. 192 test. 2 AB), suggests that Posidippus may already have
won considerable recognition for his inscribed poems a decade before the
14. That medium-specic connotation may be claried by reference to a modern example.

Consider the phrase iron curtain. Its original use was theatrical, designating the safety curtain
that falls between stage and audience to prevent resa standard element of theater equipment
since the eighteenth century (the OED cites the earliest use of the phrase in 1794). If a book on
stagecraft referred to the iron curtain within, say, ten years of 1946, when Winston Churchill
rst used that term to describe the cold war divide between East and West, it would in all like-
lihood refer to the traditional re-retardant barrier (still common in British theaters). Set in a
theatrical handbook, the conventional theater-specic sense of iron curtain is a priori more
likely to be operative than Churchills metaphorical usage, which at that time was still quite
new. In the meantime, of course, the metaphor has completely displaced the original use of the
term, so that most people are now unlikely even to recognize its theatrical roots.
15. An earlier poet, such as Simonides, certainly wrote epigram for inscription, but it is
doubtful that he wrote book epigram as well or that he collected his own inscribed epigrams
into a book. On the question of when a so-called sylloge Simonidea came into existence, cf.
Page 1981: 12223; Sider 2007. A later instance is Antipater of Sidon, from whom we have,
in addition to his many literary epigrams, an inscribed example from Delos (Inscr. de Dlos
2549 Roussel-Launey 42 GP; cf. Peek 1957).
honor at Thermon. The most sought-after poets of this sort might also
grow rich, as the story of Hieron II and the poet Archimelus shows: the
Sicilian tyrant paid this poet fteen hundred bushels of wheat (one thou-
sand medimnoi ) for an epigram celebrating his royal transport ship and
even shipped the wheat to the poet in the Piraeus at his own expense
(Athen. 5.209b SH 202).16 Posidippus himself, in his elegy on old age,
contemplates leaving to his children his house and his fortune (
= 6, 118.28 AB SH 705.28). As the inscrip-
tions from Thermon and Delphi suggest, a considerable part of that for-
tune likely came from inscriptional commissions.
The poems of the Milan papyrus make it abundantly clear that, like
Archimelus, Posidippus cultivated close ties to those with money and
power, particularly the Ptolemies and their courtiers. The new epigrams
are striking for their Ptolemaic orientation: the rst section, on stones,
closes with a prayer to Poseidon to preserve all of Ptolemys domains, the
islands, land, and shores (20.56 AB). Four of six dedicatory epigrams cel-
ebrate Arsinoe II Philadelphus in various divine aspects. Ptolemy Philadel-
phus commissions one of the statues in the andriantopoiika (63 AB). Five of
eighteen poems of the hippika celebrate equestrian victories of Ptolemaic
queens. The earliest (87 AB) marks an Olympian victory with four-horse
chariot by Berenike I (wife of Ptolemy I) and probably dates to the 280s, so
constituting our earliest datable poem by Posidippus (Cameron 1995:
24344). Others come from the years 248 and 247 and are for Berenike II,
soon to be the wife of Ptolemy III Euergetes (78, 79, 82 AB);17 these are
the poets latest datable poems. By any measure, these poems evidence an
amazing run of royal patronage over more than three decades. The section
also celebrates a Pythian victory of the Ptolemaic admiral Kallikrates of
Samos. Finally, the iamatika begin with a poem honoring a high Ptolemaic
ofcial, Medeios of Olynthos, son of Lampon. These poems come in
addition to several known before, in praise of such courtiers as Sostratus
of Cnidus, who dedicated the statue of Zeus Soter atop the Lighthouse of
Alexandria (115 AB 11 GP P. Firmin-Didot), and the establishment of
the temple of Arsinoe-Aphrodite Zephyritis by Kallikrates of Samos (116
AB 12 GP P. Firmin-Didot; 119 AB 13 GP Athen. 7.318d).
16. C / ; / 6 ,
6 , W , ,
,
6.
17. Thompson (2005: 27479) suggests, however, that the Berenike of these poems was
the true daughter of Ptolemy Philadelphus and Arsinoe I. This would date 7879 and 82 AB
prior to 252 b.c., since this Berenike left Egypt in that year to marry Antiochus II.
While we do not know whether any of these poems or others among the
new ones were written for inscription, it would be surprising if none was.
That is especially so given how shrewdly the Ptolemies set monuments at
conspicuous sites throughout the Greek world (Hintzen-Bohlen 1992) and
given the implication of the term 6 in connection with
Posidippus at Thermon.
What we can say more certainly regarding a link to inscription is that
the new poems clearly show that Posidippus was conscious of epigraphic
genres and exploited them in his poetry, in ways that his contemporaries
do not. He goes beyond the common use of epigraphic conventions from
sepulchral or dedicatory inscriptions. This is evident, for instance, in the
section on cures, the iamatika.18 These poems clearly draw on the tradition
of prose inscriptions set up by temple authorities on the basis of stories
culled from individual votives at shrines of Asclepius and recounting the
greatest miracles performed by the god. Set on the scroll, the sequence of
seven epigrams appears to simulate the experience of a pilgrim wandering
through the sanctuary and reading inscriptions.19
It is one of the peculiarities of Austin and Bastianinis editio minor that it
omits entirely and without explanation any poems from the epigraphic tra-
dition that might potentially lay claim to Posidippan authorshipthis,
though it includes any number of dubious poems surviving on papyrus,
attributed solely on the basis of suitable style and content, which it cau-
tiously marks with an asterisk. Given what we know about the different
elds in which Posidippus plied his trade, the omission of inscribed epi-
grams is an odd editorial choice.20
If only for the sake of experiment, so as to imagine the sort of poem Po-
sidippus might compose for a monument, let us consider one such epi-
gram, nely inscribed on a statue base from the sanctuary of Apollo at
Thermon, the meeting place of the Aetolian League. The epigram con-
tains, as we shall see, elements tailored specically toward viewers reading
the poem in this particular epigraphic setting. It commemorates the dedi-
18. For a fuller discussion, cf. Posidippus Iamatika in the present volume. Cf. also
Zanetto 2002; Di Nino 2005.
19. Lelli argues for a similar reliance on una tradizione epigraca duso as a particu-
larly characteristic trait in other sections of the Milan papyrus, especially the oionoskopika
(2004: 8993, here 91).
20. Fernndez-Galiano, by contrast, in his edition and commentary of Posidippus (1987:
2830), made sure to include several inscribed poems previously suggested as being Posidip-
pan by Peek (1953: 438) and other epigraphers. Fernndez-Galiano is properly cautious
about the attributions, justifying their inclusion because of la posibilidad, si bien no la prob-
abilidad that they may be autnticamente posidipeos (1987: 30).
cation by a certain Drakon of a bronze statue of his son, Skorpion, killed

in an ambush near Teithron while on a cavalry mission to aid the Pho-
cians. The incident has been plausibly linked to the Phocian liberation of
nearby Elateia from Antigonid dominance in 285 b.c., a date that accords
well with the lettering of the inscription. The epigram runs as follows (IG
IX2 51 31 Fernndez-Galiano):
6 1 G
[] #
, ,
.
/ ,
+ G 5
/ ,
@ 6 6/ , ,
D u " .
[In a golden grove, as you came with cavalry to the rescue
of the Phocians, beneath the brow of Teithrons hill,
without warning, a stealthy enemy ambush killed you, you who
contrived things worthy of your fatherland,
worthy of your forefathers, the sons of Oeneus.
As a memorial, your father, Drakon, dedicated this statue of your form
in bronze beside the tripods of Apollo,
which will lead you, though dead, up into the light,
Skorpion, since the excellence of the noble does not perish.]
Any number of factors conspire to make plausible the attribution of

this poem to Posidippus. First, it comes from Thermon, the site where he
was granted proxeny in his function as 6, maker of in-
scribed epigrams. Second, its date ts perfectly with Posidippus career.
Third, the statues dedicator, Drakon, is known to us: we nd him named
in the very same inscription in which Posidippus was honored, appearing
as sponsor of a proxeny (though not of the poets; the proxeny list is twenty
years later than the epigram for Drakon); he comes up, further, as Drakon
the hipparch ( ) in another inscription of the
period (IG IX2 8, 13)a position that ts nicely with his sons role in lead-
ing the cavalry (1 GM, v. 1) in our poem (Weinreich 1918). Setting,
date, and cast of characters thus overlap in a most suggestive manner.
Further, this is a good poem. It has a vividness and formal ambition
well suited to Posidippus. Its style is generally consistent with that of his
other epigrams. The meter shows nothing that would strike one as anom-
alous or un-Posidippan.21 Similarly, we nd certain commonalities in dic-
tion. In verse 1, for instance, the separation of the adjective from
its noun by means of the preposition 6 in the phrase 6
resembles that in 33.4 AB: M . . . 6 M (of the bedroom of
Athena in the house of Zeus). In verse 2, the position in the pentameter of
the verb , describing a murderous ambush and combined with an
indication of locale where the murder took place, is similar to that in 29.4
AB: 6 ,. Finally, compare the phrase
G / in verse 5 with Posidippus 18.3 GP ( 65.3 AB): [sc.
] T / / (.
Formal aspects, too, are shaped with care. The epigram falls neatly into
two parts, according to setting. The rst four verses recall Skorpions
death in the golden grove beneath the brow of Teithrons hills, whereas
the last four shift the scene to the memorial at Apollos shrine at Thermon
+ . . . / . The latter four verses, it
should be noted, do not evoke the site with the geographic specicity of
the poems rst half. They dont need to: that would have been immedi-
ately apparent from the statues physical location; the only further require-
ment was to draw viewers attention to the bronze efgys prestigious
placement close to the tripods of the god.
In the poems rst one and a half lines, we focus exclusively on the
noble Skorpion, the addressee, riding to the rescue oblivious of the danger
that lurks out of sight, especially in so glorious a setting as a golden
grove of Apollo22 the grand opening phrase 6 distracts
us, the readers, as well. The verb shatters the idyll in verse 2. But
the poet meaningfully puts off its subject, rst underlining the element of
stealth through the addition of qualiers, (concealed) and
(of which there had been no report). The careful deferral of
until near the end of verse 3 further mirrors the stealth of the ambush and
gives the added connotation unspeakable: the heinousness of the
21. Cf. Fernndez-Galiano 1987: 5863; Fantuzzi 2002. The breach of Meyers Second
Law (according to which an iambic word rarely stands before masculine caesura) with in
v. 5 is not uncommon in Posidippus (cf., e.g., 115.9 AB 11.9 GP; 132.1 AB 15.1 GP). In
v. 7, the apparent breach of Naekes Law, avoiding word end after a spondaic fourth foot, is
mitigated by the fact that the preposition is considered a single unit with its following noun.
22. Pausanias (10.33.12) describes a grove that may be plausibly identied with that in
our poem. It lay between Teithron and Drymaea and was dedicated to Apollo by the Teithro-
nians: From Teithronion it is twenty stades to Drymaea. At the Cephisus, where this road
joins the straight one from Amphikleia, the Teithronians have a grove and altars of Apollo. A
temple has also been built there, but no cult statue.
deed stands in stark contrast to the radiant glory of the grove. Skorpions
nobility is highlighted again by the emphatic and stylized repetition of the
things he accomplished, worthy of his fatherland and worthy of his mythic
forebears ( , / , ), an instance of the bucolic
anaphora especially favored by Theocritus in his pastoral poetry.23 The
participle in verse 4 certainly points to what Skorpion had con-
trived, but more poignantly, it may also refer to what he had intended.
These verses do not, however, concern Skorpion alone: their emphasis on
deeds worthy of the fatherland and of mythic forefathers serves implicitly
to exhort those viewers most likely to see the statue at this regional shrine.
Fellow Aetolians, who share that same fatherland, might count themselves
descendants of Oineus and might well intend great deeds on behalf of their
country. The epigrams meaning, then, is carefully adapted to its physical
milieu and keyed to a local audience.
Finally, it is notable that the wording and sentiment of the last couplet
deliberately echo the end of an epigram attributed to Simonides (AP
7.251.34 pp. 199200 IX FGE) that, according to Page (1981: 198),
was widely known and referred to events that were particularly fa-
mous. Bergk plausibly identied the epigram as one of those mentioned
by Pausanias (9.2.4; cf. Page 1981: 19798) for the dead at Plataea: "+
, 6 / / / /
6 /. (though they died, they are not dead, since Arete, by giving
them glory, leads them up, from above, out of the house of Hades). The
allusion is cued by the presence of the verb and the prominent role
of the nal word of the poem from Thermon. In our poem, the
memorial will lead up () into the light (, ) one of the dead
(6 );24 in Simonides, Arete leads up () out of Hades (6
/) those who died (). In each case, excellence ( ) al-
lows them, in some sense, not to perish (" "+
). The epigram of Simonides was sufciently famous as
23. On bucolic anaphora, see Ergnzungsspiel in the Epigrams of Callimachus in the

present volume for my comments regarding Callimachus epigram on Astakides the goatherd
in relation to Theocr. Id. 1.
24. The pronoun @ probably refers to the bronze statue ( / , vv. 56). One
might, however, consider whether Drakon could be the antecedent, in which case he will be
leading him back from the dead by setting such memorials as this one and spreading his fame.
This would be quite a claim for an Aetolian hipparch. But it could involve a play on the names
of father and son, recalling traditions about the snake () being untouched by death
(cf. how it sloughs off its skin according to Ibycus [PMG 342] and Sophocles [TrGF IV F
362]) and its links with healing (cf. Asclepius). Both snake and scorpion are who
can suddenly emerge from the earth and into the light.
to have been copied in a second-century b.c. epitaph from Knossos as well

(GV 1513, whose rst couplet reads: "+ 5 =/ e,
/ / 6 /). In our epigram, the
gural embodiment of Skorpions excellence and fame, his efgy (together
with the epigram itself ), replaces Simonides / and the Knossian
as the agent that ensures his presence in the upper world.
The evocation of this Simonidean model in the Thermon epigram, im-
plicitly comparing Skorpion to those who fell in a great battle of the past,
suggests considerable sophistication on the part of the poet. While Po-
sidippus is not the most allusive of Alexandrian authors, he was certainly
capable of such an allusion as this (see generally my discussion of Posidip-
pus 7 AB in the chapter of this volume on The Politics and Poetics of
Geography, with its n. 17). Its presence here also hints at a level of learn-
ing on the part of his patron, Drakon, sufcient to appreciate his sons
assimilation to heroic models of the past. The ne quality of the inscrip-
tion and costly material of the statue, coupled with Drakons high position
as commander of cavalry, suggest that he had enough ambition to hire
such a prominent poet as Posidippus. It will have been for such poems as
this oneof substantial quality, commemorating important local events,
and written for people of consequencethat Posidippus received the
honor of proxeny from the Aetolians.
I have thus far tried to form a picture of one aspect of Posidippus epi-
grammatic work, namely, that in the medium of inscription. The Ther-
mon decree makes him the earliest poet (and one of the only ones) of
whom we can certainly say that he wrote directly for monuments, and that
coheres with the unusual awareness of the inscribed medium he displays
elsewhere (e.g., in the iamatika). There may well be inscribed epigrams,
such as that for Skorpion and his father, which we can plausibly ascribe to
him. Given what we know, we should, in any case, not exclude inscrip-
tional poems a limine, as Austin and Bastianini do, thus disregarding an en-
tire important part of Posidippus attested activity as though it had never
existed. Rather, we should keep our eyes peeled, stay continually on the
lookout for, and carefully examine inscriptional poems that we could po-
tentially attribute to our poet (if only as a check on the easier inclination to
stick with what has come to us via manuscripts and papyri), remaining fully
conscious of the uncertainties that attend such an undertaking.
There is, however, another prominent aspect to this epigrammatist, for
he is equally interested in the impact of the scroll. The formative impor-
tance of papyrus for the poet is memorably expressed in one of the old
epigrams of Posidippus (6.3 GP 137 AB), where the speaker describes
his soul (vv. 34) as having previously labored in the papyrus scrolls (? +
6 ). This toil among the books allows him not just
to withstand Pothos but to reap other harvests while reproaching the griev-
ous god. Though not as assertive in his erudition as Callimachus, the doctus
poeta par excellence, Posidippus nonetheless shared many of the learnedly
bookish interests of his contemporaries.
An epigram quoted by Athenaeus likewise stresses the power of the pa-
pyrus. It is an epitaph for Doricha, the notorious hetaira of Naukratis in
Egypt and mistress of Sapphos brother Charaxus (Athen. 13.596c 17
GP 122 AB). In this poem, which I treat in detail in chapter 13 The
Politics and Poetics of Geography in the present volume, we hear that
the bright resounding papyrus columns of Sapphos / dear song abide and
will yet abide (M + ( / Mb
, vv. 56) and that they will preserve the name
of Doricha for as long as a ship sails out from the Nile across the salt sea
((/ [ :J 6/ Z , v. 8), that is, through the ceaseless
export of papyrus rolls going out into the world from Egypt. Posidippus,
then, was keenly interested in books as a vehicle of literary dissemination.
Through the medium of the scroll, even marginal gures of literary biog-
raphy, such as Doricha, could survive in public consciousness.
Ancient testimony further suggests that Posidippus was directly en-
gaged in shaping his own works into collections, that is, he acted as poet
and editor at once (just like Callimachus). This appears in the celebrated
case of Berisos, where we see Posidippus embroiled in a scholarly dispute
on Homeric interpretation.25 At Iliad 11.101, we hear that Agamemnon
went on to slay Isos and Antiphos (" C 8/ / P
6). Doubtless with sly humor, Posidippus interpreted the phrase
8/ / as referring to an otherwise unknown bastard son of Priam,
Berisos, evidently reading the line as " C P
6 (but he slew Berisos and Antiphos). According to the (A)
scholium on this line, Posidippus mentioned this Berisos in an epigram
from a collection called The Heap, or Soros.26 That same scholium ( SH
701 144 AB) tells us: Zenodotus [the editor of Homer and rst librarian
of the Library of Alexandria] read the line without the rho (i.e., as /),
and Aristarchus says that this Berisos is not now to be found in the [col-
lected] Epigrams of Posidippus, but that he did nd it in the so-called Soros.
25. On Berisos, see Nagy 2004: 6164.

26. Reitzenstein (1893: 96102) and Cameron (1993: 36976) believe that the Soros was
a collection by multiple authors. Gutzwiller (2005b: 7 n. 19) rightly cautions, however, that
the only poet known to have been included in this collection was Posidippus.
He says it is likely that Posidippus deleted it on being criticized (Z

( /. 6 C / 6
6 , / 6 M M M
#. _ 6 " ). Thus Arist-
archus at least supposed that Posidippus took some ack for his reading and
that he thus expunged the characterwho was, in any case, a bastard
() and spurious (Il. 11.1023)from a subsequent collection. Besides
imputing to Posidippus an awareness of scholarly disputes, this passage sug-
gests that Posidippus published his epigrams in at least two collections dur-
ing his lifetime and thatlike Callimachushe was involved in shaping
and editing the different editions. That, in turn, has implications for our
assessment of the organization of the Milan papyrus. For it squares well
with the artful arrangement of the poems that scholars have noted within
individual sections of the papyrus,27 and even of the sections vis--vis each
other.28
If the iamatika showed how Posidippus exploited his familiarity with
epigraphic models to take epigram in new directions, other sections of the
Milan papyrus demonstrate that he mined his deep knowledge of the lit-
erary tradition to the same purpose. The most surprising and innovative
section in the papyrusand one that forces us to reimagine Posidippus
conception of epigram quite radicallyis the oionoskopika.29 These are
didactic poems in epigrammatic form, mainly on bird omens. They read
like Aratus transformed into elegiac couplets. In fact, however, we can
probably trace the source for these epigrams to an earlier epic, the lost
Hesiodic Ornithomanteia, which interestingly was athetized by Apollonius
of Rhodes (Paus. 9.31.4ff.). If these are not untrodden paths in the Cal-
limachean manner, I dont know what is. That is so despite the fact that
Posidippus is mentioned together with Asclepiades in the Florentine
Scholia to the Aetia 1.1 as one of the Telchines, that is, as one of Calli-
machus literary adversaries.
Indeed, apart from his admiration for Antimachus Lyde (AP 12.168
9 GP 140 AB), which Callimachus abhorred (fr. 398), Posidippus liter-
ary values seem quite compatible with those of his distinguished contem-
porary. That comes out in one of the new epigrams, commemorating a
27. See The Politics and Poetics of Geography in the Milan Posidippus and Posidip-
pus Iamatika in the present volume. Cf., further, Fantuzzi 2004a; Baumbach and Trampe-
dach 2004; Sens 2005. For doubts as to whether the arrangement reects the hand of a poet
or an editor, see Krevans 2005.
28. See especially Gutzwiller 2004, 2005a.
29. On this section, see, above all, Baumbach and Trampedach 2004; Sider 2005.
statue commissioned from the sculptor Hekataios by none other than

Ptolemy Philadelphus and representing Philitas of Cos, one of Calli-
machus chief models (X 1625 BG 63 AB). The poem stresses how,
with utmost skill and precision, the sculptor portrayed Philitas as an eld-
erly gure, an anxious perfectionist, with no trace about him of the heroic.
Alex Sens has shown persuasively how the literary and artistic ideals
embodied here in Philitas, in his sculptural image, and in the epigram itself
reect the rened Hellenistic aesthetic that especially prized works of
painstaking techne, exquisite craftsmanship, and polish and that abandoned
the grand heroism of earlier traditions for art on a more human scale (Sens
2005; Prioux 2007: 1974, who however also differentiates Posidippus and
Callimachus aesthetics, cf. 77130; see also my chapter 1, The Unruly
Tongue in the present volume). What is there here that Callimachus
would not admire?
From the various sources about Posidippus activity, then, what emerges
is an outline of a gure whose activities as epigrammatistunlike that of
any other Hellenistic poet we can tracebridged the competing media of
inscription and scroll. In the one, his work was set in stone, xed in a single
place, its meaning dependent partly on its physical location in the landscape
of a given shrine, cemetery, or civic space and partly on the importance his
patrons attached to that particular place. In the other, his epigrams func-
tioned in the very different terrain of the book, which has different expres-
sive potential and elicits from the reader a different response.
The two media in which Posidippus found fame as an epigrammatist
the monument and the scrollare nicely merged in a passage from his seal
poem (SH 705 118 AB). Here the poet hopes to be honored with a
statue of himself in his native Pella; that is, he yearns to see himself com-
memorated in a monument, like those for which he himself had written
epigrams.30 He even imagines the particulars of the setting: his efgy is to
stand in the crowded marketplace ( , , v. 18).
30. Indeed, Posidippus hopes that Apollo will grant him a hero cult, like that which he
had recently proclaimed for the iambic poet Archilochus on Paros. There, a certain Mnesiepes
had sought the gods sanction for the founding of an Archilocheion, an event to which
Posidippus seems to refer specically in his seal poem. For there, in bidding his readers to
give to the Parian nightingale a mournful / ood, shedding empty tears from the eyes and
groaning in lamentation (/ 6 + 6.[ /

[ / , SH 705.1820 118.1820 AB)that is, specifying pre-
cisely the sort of honor we would expect in hero cultPosidippus wishes that Apollo will give
an oracle about him, as he did concerning Archilochus, prescribing honors for him, like those
for Archilochus (SH 705.1117 118.1117 AB). On the cult of Archilochus on Paros, see
D. Clay 2004; Bing 1993a.
The adjective , people-bearing, here has special meaning for a

poet who wrote for inscription, for he could only hope for his poems to be
read on busy thoroughfares, at crowded shrines, or, as here, in the market-
place. No isolated spot would sufce. Posidippus wants his statue set in the
busiest, most central spot of his hometownin Pellaean terms, Times
Square and 42nd Street. That bespeaks the shrewd appraisal of one who
knows from direct, considered experience just where a monument can
achieve its maximum effect.
But for the ambitious epigrammatist, this medium has obvious limita-
tions. Like the monuments on which he worked, his would be xed in a
single place, at Pella. Further, inscribed verse traditionally remained anony-
mousonly in rare exceptions does a poem on stone reveal its author. De-
spite the recognition Posidippus achieved as an epigrammatist for monu-
ments, his ambition here is clearly for more than what was associated with
local commissions. Rather, seeking the fame that an epigrammatist might
nd solely through a books virtually unlimited potential for dissemination,
he calls on the Muses to join him, writing down the song in the golden
columns of my tablets ( 6 , v. 6), and
hopes that the Macedonians may do me honor, both those on the islands
and the dwellers near the coast of all Asia (= []
G / 6 [ / G / / V, vv. 1415). In keeping
with this grandiose conception, he imagines the statue of himself bearing a
most signicant object: he will stand there unwinding a book (
2, v. 17).31 Thus we nd that in reimagining Posidippus, we must
ultimately see him as he imagined himself, that is, in a form that embodies
both aspects of his work: that with regard to monuments and that involving
the book. For as the elegy on old age, his seal poem, clearly shows, he wants
it both ways. For himself he wants both the marble and the scroll.
31. Gutzwiller (2005a: 31719) suggests that the seal poem might have been the nal
poem of the Milan papyrus and that the book Posidippus holds may embody that very collec-
tion. Cf also Gauly 2005. R. Hschele notes (in her 2007 dissertation) how the Ver-
steinerung of that collection in the statue wittily reverses epigrams development from
inscribed monument to scroll: Der Gesang Poseidipps und der Musen wird als schriftlich
xiert vorgestellt, das (damit identische?) Buch erscheint, zumindest in der Phantasie des
Dichters, in Stein verewigt.
c h a p t e r 10
Between Literature
and the Monuments

Sostratus and the Lighthouse:

Zeus Soter in Posidippus 11 GP ( 115 AB)
I n the summer of 1995, a team of underwater archaeologists, led by Jean-

Yves Empereur (head of the Centre de recherches Alexandrines and
director of research at the CNRS), made a spectacular nd working in the
waters beside Fort Qait Bey. There, just east of the fort and beyond the
citys eastern harbor, they discoveredtogether with a variety of sculp-
tural remainsa number of massive architectural blocks that, because of
their enormous size and location right next to where scholars generally
place the ancient Pharos, seem very likely to have belonged to the great
Lighthouse of Alexandria itself. For the rst time, it seems that we may be
able to recover something of this monuments physical aspect from its
material remains.1
With this renewed focus on the Pharos, it is time to take another look
at an important literary source for this building: the epigram of Posidippus
This is an updated version of an article that appeared in Genre in Hellenistic Poetry, ed. M. A.
Harder, R. F. Regtuit, and G. C. Wakker, 2143. Hellenistica Groningana 3. 1998. 1998
Egbert Forsten Groningen.
1. For the excavations, see Empereur 1995a, 1995b, 1996a, 1996b, 1999, 2000), Em-
pereur and Grimal (1997). Exciting underwater explorations of the royal quarter in the east-
ern harbor of Alexandria have been undertaken as well, cf. Goddio, etc. (1998, 2006).
194
Between Literature and the Monuments 195
(P. Firmin-Didot 11 GP 115 AB).2 For the sake of convenience, I print

the text as constituted by Austin and Bastianini.
; , , b
,
( D
" 6 ,M _ / 6
6.
" = ,
5
@/ /
A, + 6
= 6 ,
6/ " "/ [ Z
, , Z C J . 10
[Savior among Greeks, this watchman of Pharos, was set up, Lord
Proteus,
by Sostratus the Cnidian, son of Dexiphanes.
Since in Egypt there are no lookout points on mountains as on the
islands
and the breakwater lies low for anchorage,
for that reason, sheer and steep, this tower appears to cleave
the air across countless leagues
by day, and all night long quickly the mariner on the wave
will see the great re blazing from its peak,
and though he may run to the Bulls Horn itself, he would not miss
Zeus Soter, O Proteus, in sailing hither.]
This poem carries great evidentiary weight in discussions of the light-

house. For it is our earliest evidence altogether, probably contempo-
rary with the buildings completion. Its author was a prominent poet of
the age, acclaimed in many parts of the Greek world, including Egypt,
which he evidently visited during the first half of the third century b.c.,
2. The papyrus was found among the papers of two Macedonian brothers connected
with the Serapaeum of Memphis in the mid-second century b.c., Ptolemaeus and Apollo-
nius; it contains a variety of literary passages, a very personal selection, as it seems (cf.
Thompson 1987; 1988: 25961). Cameron (1993: 7 with n. 25) suggests that Posidippus
Macedonian origin may have made him an appealing choice for the brothers. According to
Thompson, our poem in particularas well as the other epigram of Posidippus preserved in
the papyrus (12 GP 116 AB)is in the rude hand of the elder brother, Ptolemaeus.
This brothers possible connections with Alexandria and with the Pharos are discussed later
in the present essay.
writing numerous poems in honor of Ptolemaic monuments, occasions,

and cult.3
The text of the epigram as a whole has been studied and improved by
many scholars.4 But I want to focus mainly on the rst and last couplet, par-
ticularly on the meaning of in verse 1 (and its echo in verse 10),
which I believe has been misunderstood; it may have been so even in antiq-
uity. The interpretation I propose has important consequences. It affects
how we assess Sostratus of Cnidus, for it provides a new answer to the ques-
tion of what precisely he did in connection with the Pharos and, conse-
quently, what Posidippus commemorated in his epigram.
My argument builds on two important points made by P. M. Fraser and
buttressed by F. Chamoux (and indirectly by J. Scherer). The rst concerns
Sostratus of Cnidus and the nature of his involvement with the lighthouse,
which is unclear. Many modern scholars view him as its architect.5 In doing
so, they follow Pliny and Lucian.6 However, as Fraser puts it in a penetrat-
ing analysis of the evidence generally cited in support, the claim [that
Sostratus was architect] lacks any real substance.7 For in fact, no one ear-
lier than Pliny speaks of Sostratus in this capacity, and his information
probably rests on a misunderstanding of a word such as ( in verse 2,
3. In addition to the previously known epigrams on the Pharos, the cult of Arsinoe-
Aphrodite at Cape Zephyrion (12 GP 116 AB; 13 GP 119 AB), and the epithalamium of
Arsinoe (P. Petrie II 49a SH 961 114 AB, plausibly by Posidippus), see the numerous new
epigrams of the Milan papyrus, on which, most recently, cf. Ambhl 2007.
4. For bibliography on our poem, cf. Chamoux 1975: 214 n. 2; Fraser 1972: I 1720, 568
with nn. 129 and 132. More recently, see Austin 2002.
5. Thus e.g., Bernand (1966: 1034; 1995, 5053) and Praux (1978: I 222).
6. The relevant passages are as follows:
magnicatur et alia turris a rege facta in insula Pharo portum optinente, Alexandreae,
quam constitisse DCCC talentis tradunt, magno animo, ne quid omittamus, Ptolemaei
regis, quo in ea permiserat Sostrati Cnidii architecti structura ipsa nomen inscribi.
(Pliny NH xxxvi 83)
6 , 6D , 6 J
CY
M , ( Z, E / "
6 1 , ,
, O , " , : 6 , ID ,
" ( ( +
" # = (, 6 +
M 6 6 _ , ,, @
6, 4 + M ,
( , #+
. a "/ 6 6 "+ # 4
, [ 2 J C J "
2, / , ? .
(Lucian Hist. conscrib. 62)
7. Fraser 1972: I 19; for his discussion as a whole, see 1920 with nn. 104 and 111.
which can refer either to the act of building or to dedication.8 Scholars

have noted, moreover, how odd it would be for the buildings architect,
rather than the dedicant, to be named in its dedication9something Pliny
and Lucian implicitly recognize, since their stories seek to explain the
presence of his name in the inscription (in one case attributing it to the
kings great magnanimity, in the other to Sostratus trickery). Strabo, our
earliest source after Posidippus, visited Alexandria for an extended period
starting in 24 b.c. and knew the Pharos rsthand. In contrast to Pliny, he
makes no mention of Sostratus as architect of the lighthouse, saying,
rather, that he dedicated it: / (sc., ) . . .

, O ? 6 (791.6). That is cer-
tainly more plausible. But while we must be grateful to Fraser for casting
rightful doubt on Sostratus as architect, I would argue on the basis of
Posidippus epigram that (contrary to Strabo) we should not even consider
Sostratus the Pharos dedicant.10
The second point concerns the statue that crowned the Pharos. Ancient
depictions (in coins, mosaics, etc.) clearly and repeatedly show a large male
gure atop the lighthouse, and on the basis of the pictorial tradition, schol-
ars have eagerly speculated about that gures identity. But the evidence is
inconclusive.11 Again it was Fraser who rst publicly insisted that Po-
sidippus nal couplet identies the statue unequivocally. There, in describ-
ing the sailor trying to make it to harbor, the poet says, he would not miss,
in sailing hither, O Proteus, his target, Zeus Soter ("/ [ Z /
, , Z C J , vv. 910, trans. Fraser). Earlier
critics, perhaps blinded by the confusing images of the lighthouse and by
Lucians report that it was dedicated , muddled their trans-
8. Testimony sometimes taken as evidence of Sostratus architectural activity elsewhere

is equally late and does not stand up to scrutiny, as Fraser clearly shows.
9. Fraser (1972: I 20 n. 116) aptly cites Wilamowitz (1924: I 154 n. 2): Immer noch
begegnet man der Torheit, Sostratus wre Architekt gewesen, als ob er sich dann als Bauherr
htte nennen knnen.
10. Moreover, I questionpace Wilamowitz (1924: I 154 n. 2)whether Sostratus
would have had the immense fortune required to build such a monument as the Pharos (eight
hundred talents according to Pliny): nicht gerade preiswert, as G. Weber trenchantly com-
ments (1993: 333 n. 2). Hlbl voices similar doubts (1994: 66). But implausible as it might at
rst seem that the king would allow a private citizeneven a prominent courtier and friend
of kings, as Strabo calls Sostratus (791.6)to take responsibility for a structure that would
so utterly dominate his citys skyline, Weber does well to remind us of the very conspicuous
dedications of Ptolemaic (1993: 33334).
11. The evidence is reviewed by Chamoux (1975: 219220) and Fraser (1972: I 18 with
n. 103).
lations.12 But Fraser (1972: I 19) recognized that the words "/ [ Z /
. . . Z must refer to the statue: this is clearly what is implied
by the poets phrase, not fail to hit Zeus Soter; . . . conversely if he was not
referring to some specic representation of Zeus Soter, his reference to him
is meaningless. Around the time Fraser completed his manuscript, J.
Scherer apparently proposed the same thing per litteras to F. Chamoux.13
And Chamoux himself added the appealing suggestion that the statues
presence atop the Pharos is reected in a propemptic fragment of Calli-
machus, invoking Zeus as watchman of the harbor (fr. 400 Pf.): ; , `
6 1 / 9, Z
(Ship, that snatched away the only sweet light of my life, I en-
treat you by Zeus the watchman of the harbor). As a longtime resident of
Alexandria, Callimachus would naturally have been thinking of that Zeus
who, from atop the Pharos, was watchman of the harbor in his adoptive
city: thus contends Chamoux.14
While scholars since Fraser and Chamoux have generally accepted
12. Especially egregious in this regard was Page (1962: 447), who simply ignored the ex-
plicit naming of Zeus in the last verse, translating . . . Z as the God of Safety,
with the explanatory note The lighthouse was inscribed .
It is well to note that no one before Lucian records this dedication ; Strabo
says only , which accords well with the single savior god
atop the monument, commemorated in Posidippus poem. I agree with Frasers suggestion
that the original dedication, , was erroneously altered in quotation or paraphrase to
(1972: I 19). By contrast, Bernand (1996) arguesagainfor the authenticity
of the plural dedication, with Poseidon as the second god. The weakness in his argument is
that it starts from late evidence, rather than recognizing the unequivocal identication in our
earliest source, the poem by Posidippus.
13. Chamoux 1975: 21819. Commenting on the phrase "/ [ Z / . . .
Z, Chamoux writes: On entend dordinaire que laide de ce dieu ne lui fera pas dfaut.
Mais traduire ainsi, crivait J. Scherer, cest rendre dune facon trs approximative " [
Z. ;, cest manquer le but. Ce vers doit trouver son explication dans le mon-
ument lui-mme . . . Le Phare portait son sommet une statue de divinit . . . Posidon serait
bien sa place ici, mais Zeus, pre des Dioscoures, ne serait pas non plus dplac. Et, ceci
accept, les deux derniers vers de lpigramme sont dune limpidit parfaite: le Zeus dont il
sagit nest pas celui qui habite quelque part dans lOlympe, mais trs prcisement le Zeus qui
surmonte le Phare; il est parce que, si on le manque (Z), si on ne vogue pas
droit sur lui, on fait invitablement naufrage . . . Il faut donc entendre la phrase, comme le
grec y invite, dune facon concrte: celui qui navigue dans ces parages ne saurait manquer
( tomberait droit sur) Zeus Sauveur. Posedippos dit en style potique exactement la mme
chose que Strabon dans son honnte prose: O/ " , / (letter
of 3 April 1967).
Scherer and (implicitly) Chamoux equate the function Posidippus gives the statue with
that which Strabo attributes to the lighthouse. I believe that this discrepancy may be due to
Strabo erroneously recalling the dedicatory inscription of the statue as applying to the build-
ing as a whole: hinc illae lacrimae.
14. Chamoux 1975: 22122.
. . . Z in verse 10 as pointedly referring to the statue on the

Pharos, so identifying it as Zeus Soter,15 they have not taken the next logi-
cal step; that is, they have not gone back to verse 1 to revisit the poems be-
ginning in the light of their new understanding. Consequently, ;
, , . is typically taken to refer to the lighthouse in
toto, as in the translations of Bernand (1995: 50), Cette sauvegarde des
Grecs . . . , cest Sstratos qui l a rig,16 or of Adam (1995: 27), Pour le
salut des Grecs, cette tour qui veille sur Pharos . . . fut leve par Sostra-
tus.17 All such renderings have in common that they treat as mean-
ing something different in verse 1 from in verse 10. Underlying their will-
ingness to accept this difference is, I believe, the critics hopeperhaps
unconsciousof reconciling Posidippus words with the evidence of later
sources (Strabo, Pliny, Lucian) concerning the role of Sostratus of Cnidus.
In translating so, however, they must ignore the plain sense of the words
not to speak of weakening the very carefully constructed frame of the
poem, in which the invocation of Proteus and mention of a savior in the
last line echo precisely these elements in the rst.
A crucial step toward reinterpreting the poem consists in taking
(v. 1) in its more usual sense, that is, as a strictly personal noun,
meaning savior.18 In that sense, it is the epithet of many gods (as well as
heroes and kings). But one god alone could be describedin the majestic
words of the exordiumas the ; tout court. That god can
be none other than Zeus, an identification that accords well with his pres-
ence in the poems last line ( . . . Z, v. 10) as the sailors goal
when heading into port in Alexandria. ; , then, is possessive geni-
tive, so that Zeus is among Greeks the [one known as] savior. By con-
trast, if referred to the lighthouse, ; would have to be
objective genitive (as in the translations cited earlier). But that is implau-
sible, since the lighthouse helps all, not just Greeks. And Alexandrias
port certainly drew merchants from all corners of the Mediterranean and
beyond.19
To be sure, if the epigram apostrophized Greeks, the possessive genitive
15. See, e.g., Bowman 1986: 206; Green 1990: 158; Bernand 1995: 53.
16. This rendering is almost the same as Chamoux (1975: 215).
17. In the same vein, cf. Dombart 1967: 81: Proteus! Meergott! Fr Griechen zur Ret-
tung erbaute des Pharos / Warte Dexiphanes Sohn, Sostratus, Cnidusgeboren. This is cited
approvingly by Hesberg (1981: 69).
18. Cf. LSJ s.v. That it could, however, be used of a building is demonstrated by Eur.
Med. 360: X .
19. Cf. Frasers discussion, e.g., of trade with Carthage in the third century b.c. (1972: I
14884, esp. 15253).
; might seem superuous: why tell Greeks that Zeus is the savior
god par excellence among Greeks? But it is perfectly apt when we recall
that the poems addressee is Egyptian: the sea god Proteus (b ,
v. 1), known to Homer as Egyptian Proteus, 1 , (Od.
4.385).20 According to pseudo-Callisthenes (1.32.13), Proteus was in fact
worshiped as hero by the local Egyptians before the founding of Alexan-
dria; Alexander the Great refurbished his shrine. And authors as early as
Herodotus (2.112f.) and Euripides (Helen 45) view him as an Egyptian
pharaoh. The savior god of the Greeks and the local Egyptian deity thus
frame the rst verse.
The identication of with Zeus also perfectly suits the following
appositive phrase , watchman of Pharos. For starting al-
ready in Homer, Zeus is regularly portrayed keeping watch from heights,
such as Mt. Ida (e.g., Il. 8.397). Such a conception is thoroughly part of his
image.21 Moreover, looking back from the end of the poem,
acquires an added sense: as the god is the target that guides the sailor into
port at verses 910 ("/ [ Z / . . . Z C J ), so
he becomes the mark on which one xes the eye at verse 1 (LSJ s.v.
II).
If, then, the ; of verse 1 should be identied as Zeus,
the sentence as a whole (Sostratus set up the savior of the Greeks, the
watchman of Pharos, ; , . . . /
() must surely refer to that same statue of Zeus Soter mentioned in
verse 10. The expression . . . ( is entirely appropriate to the
dedication of a statue, for we nd comparable usage in inscriptional evi-
dence, as, for instance, when the city of Phocis dedicates statues of the
Dioscuri to Poseidon in the fourth century b.c. (CEG II.807.23, on a
marble statue base): ? " / / ?
(according to its vow, the city dedicated to the god these demi-
god saviors). Note that the inscription does not explicitly call the statues
statues: the city dedicated saviors, just as Sostratus set up the savior. In
short, the opening couplet of the epigram tells us quite straightforwardly
20. Similarly, when addressing the Persian king Cyrus, Croesus speaks of Apollo as
; (Hdt. 1.87.3). It is worth noting that there is also a tradition that denies Pro-
teus Egyptian origin and attempts to recoup him for the Greeks. Thus at the start of the
Victoria Berenices (SH 254.5), Callimachus calls him [, prophet of Pallene,
in the Chalcidice.
21. It is borne out also by the fragment of Callimachus (400 Pf.) adduced by Chamoux
and cited earlier in this essay, in which Zeus is invoked as : Z
.

that Sostratus of Cnidus set up the statue of Zeus Soter atop the Pharos
the same one referred to at the end of the poem.22 It is this statues dedica-
tion, not that of the lighthouse altogether, that Posidippus commemorates
with his epigram.
How does this t with what follows? After all, the middle section of the
poem (vv. 38) clearly deals not with the statue but with the towering
lighthouse as a whole. The answer lies in how we interpret in line 3:
" 6 ,M _ . . . It should be taken as anticipatory,
establishing why a great tower is needed for the statue of Zeus to stand on
(cf. Denniston 1954: 6872, esp. IV 23). The train of thought is as fol-
lows: Sostratus installed a lookout (sc., watchman, , v. 1); since
(, v. 3) Egypt has no lookout places (, v. 3), therefore ( ,
v. 5) this tower serves its purpose. In other words, because there are no
mountainous heights for Zeus the watchman to watch from, this tower
( @/, v. 6) serves as a substitute.
Nothing in the text, then, suggests that Sostratus dedicated the light-
house itself (much less was its architect). Posidippus assigns him a more
limited, if more plausible, role: he gave the statue with which the monu-
ment was crownedan impressive enough donation when one considers
its prominence in artistic representations.23
Still one wonders how Strabo and others after him came to associate
Sostratus with the building as a whole? Any answer must of course remain
speculative. But Chamoux (1975: 221) made an interesting observation
that may point us toward an answer. Looking at Strabos statement con-
cerning Sostratus dedication ( / [sc., ] . . .
, O ? 6 , 791.6) and the
inscription itself as reported by Lucian (
#+ ), he suggested that both passages
may depend on Posidippus, that is, that when Strabo speaks of the 6-
, he is referring to our poem: Rien nempche de croire que la
formule de Strabon et ultrieurement celle de Lucien rsument tout sim-
plement lpigramme de Poseidippos, qui aurait servi de ddicace effec-
tive au monument . . . Le thme du salut, qui inspire le pome du premier
vers, ; , au dernier, Z, a pu suggrer Lu-
cien la formulation resserre quil nous a transmise et qui correspond la
22. One may note that Posidippus thus places Zeus at the beginning and at the end of his
poem, precisely as suggested by his contemporaries Theocritus (17.1: 6
6 ) and Aratus (Phaen. 14: M

a ).
23. Cf. especially the Begram Vase and the mosaic from Qasr-el-Libya near Cyrene; cf.
Fraser 1972: I 18 n. 103.
paraphrase de Strabon.24 I believe Chamoux was correct in proposing that

these later authors depend on our poem. I think, however, that Strabos
dependence extends well beyond what he suggested.
For on reading Strabo, one is struck by a range of similarities with Po-
sidippus epigram: in the function ascribed to the monument, the reasons
given for why it is needed, as well as the whole train of thought, which is
remarkably similar. Here is what Strabo says (791.6):
/ [sc., ]
, , O ? 6 .
_ 2 , 6
+ , ( #
, O/ "
, .
[Sostratus the Cnidian, friend of kings, dedicated this tower for the
safety of sailors, as the inscription says. Since it is without shelter and
the shore is at on either side, and there are reefs and shallows, some
high and conspicuous marker was needed for those sailing in from the
sea, so that they could locate the entrance to the harbor.]
In the following comparison, I always start by quoting Posidippus. Both

texts begin by naming Sostratus of Cnidus and reporting his dedication
((, v. 2 ), which is linked with safety (, v. 1
). Thereafter we learn that the need for the dedication
arises from the atness of the coast around the harbor (" 6 ,M
_ / 6 / 6, vv. 3
4 _ 2 ), whose
entrance is perilous (6/ " 6 +
, v. 9). For this reason, a marker is necessary (
. . . / . . . @/, vv. 56 ( ). Both texts stress the
height of the marker (" = , , v. 5 # )
and its radiance ( / ., v. 6 ). Fi-
nally, both texts close by returning to the theme of safety: by means of this
dedication, sailors may avoid the perils at the mouth of the harbor and
24. Chamoux justly casts doubt on the authenticity of the inscription in Lucian because
of the very questionable story in which it is embedded (for the entire text, cf. n. 6 in the pres-
ent essay): lanecdote quil ajoute ce tmoignage est si fantaisiste quelle le rend fort sus-
pect: comme tant dautres logoi qui encombrent les compilations rdiges lpoque impri-
ale, lhistoriette relative Sostratus a pu tre forge pour piquer la curiosit du public, et le
texte de linscription reconstitu pour la circonstance (1975: 220).
reach safe haven ("/ [ Z / . . . Z C J , vv. 9

10 O/ " , ). In brief, we may say of
these texts as a whole what J. Scherer observed with respect to their con-
clusions: Poseidippos dit en style potique exactement la mme chose que
Strabon dans son honnte prose.25
Our comparison certainly suggests that Strabo knew Posidippus poem.
But should we follow Chamoux (1975: 221) in thinking that the epigram
was actually inscribed on the Pharos?26 After all, Strabo might have been
working from a transcription. Here one of the recent nds from the exca-
vations by the harbor may shed a tantalizing, if perhaps evanescent, ray of
light. For the divers recovered a fragmentary white marble block contain-
ing part of a Greek inscription with inlaid bronze letters (.45 m in height),
such as is attested for the Pharos eastern face in a tenth-century Arabic
source (Fraser 1972: I 18 n. 104).27 The letters are legible. Could
they come from the phrase X at the start of Posidippus
verse 5?28 To put it mildly, this is extremely uncertain. But one can imagine
Strabo either glimpsing the poem up on the lighthouse as he sailed in or
out of the harbor and construing it erroneously as referring to the building
itself on which (in our imaginary scenario) it would have been inscribed, or
mistakenly recalling the poem from such an occasion. Maybe, too, he was
inuenced by lore about the Pharos, which at that late datesome 250
years after the factmay already have colored its history. If this scenario is
at all plausible, it might explain how Sostratus was transformed into the
sponsor of the lighthouse as a whole when, originally, as a careful reading
of Posidippus demonstrates with fair certainty, he had been responsible
only for the statue of Zeus Soter at its peak.
Inscribed Poems and Their Fictions
In entertaining the possibility that Posidippus poem was originally in-

scribed, we come up against an awkward, nagging question. For like an itch
that lingers, now forgotten, now acute, yet always just beyond our reach,
25. See the letter cited in Chamoux 1975: 21819 (quoted in n.13 of this chapter).
26. So, too, Fraser (1972: I 568).
27. Together with the other recent underwater nds, the block is now exhibited in the
archaeological park at Kom ed Dik.
28. I must say that the letters are followed by what might be the upper left-hand traces
of a . I think, however, that this reading is scarcely the certainty that Empereur seems to
suggest in his preliminary report (1995a: 757). On the preceding line, just to the right of that
possible , is the lower part of what could be an or similarly rounded letter.
the possibility of inscription bedevils scholars who deal with funerary or

dedicatory epigram preserved in literary sources (the Greek Anthology,
citations in ancient authors, oras herepapyri). Wilamowitz summed it
up under the rubric AUFSCHRIFT ODER NICHT.29 For as is well
known, epigramthough continuing to be used in inscriptionsgradually
outgrew its chiseled origins, acquiring a parallel life as a 6,
where it might be composed strictly as literature.30 The result is a herme-
neutic crux deriving from the fact that, in spite of its development away
from inscription, epigram retained the generic conventions of its incised
counterpart; that is, besides keeping its traditional meters, it continued to
use deictic markers to refer to (now ctional) monuments or votive objects
and their physical settings, and it continued to name the donors and de-
ceased, their families, afliations, cities of origin (again, no longer real).
Consequently, it is often impossible to tell whether epigrams that come to
us via literature were once actually inscribed or are quasi-inscriptional,
merely adopting the pose. Aufschrift? . . . oder nicht? On paper they can look
exactly alike.31
Rather than try to resolve this conundrum, I intend to dwell, in what
follows, on one of its causes, epigrams migration to papyrus, and on a con-
sequence, the genres striking lability with regard to medium. Only then
will I turn to a basic difference between inscribed and quasi-inscriptional
poems, not, however, a formal one, such as various scholars have tried to
establish, but, rather, one of reader response.32
29. This title stands blazoned in capitals, like an ancient inscription, to lead off the dis-
cussion of epigram in Wilamowitz 1924: I 119.
30. I trace this development in Ergnzungsspiel in the Epigrams of Callimachus in the
present volume.
31. The terms used by Richard Thomas (1998) to make this distinction, functional
versus literary epigram, may be misleading since they imply that inscribed epigrams may
not be literature or literary and that literary epigram is without function. The terms in-
scribed and quasi-inscriptional are less problematic, at least for dealing with funerary and
dedicatory epigram. I owe the latter term to Hutchinson (1988: 2021).
32. Most recently, Khnken (1993: 12021 and passim) has tried to establish two tests
for distinguishing inscribed from quasi-inscriptional poems of votive or sepulchral type.
First, quasi-inscriptional poems include descriptions of physical context that would be self-
evident and unnecessary if a reader was facing a genuine monument (berinformation),
while, conversely, poems originally inscribed but known only through literary sources may be
identied through lack of contextual information (Unvollstndigkeit der Angaben). Sec-
ond, there is often an element of play with dedicatory and sepulchral convention, which
serves no function other than an artistic one.
While potentially useful, these tests do not have as broad an application as Khnken
assumes: lack of information need not indicate that a poem was originally inscribed. On the
contrary, as I have argued in detail in my chapter on Ergnzungsspiel, chap. 5 in the present
To begin, then, the problem voiced with epigrammatic pith in Wilam-

owitz alternatives arises not just because of a formal resemblance between
inscribed and quasi-inscriptional poems but because of their shared locus
on the scroll. So long as the two kinds of poem stand apart, respectively
xed upon stone or papyrus, there will be no confusion. Only as boundaries
break down and poems can migrate from one medium into another are we
faced with uncertainty. Such uncertainty is, I would argue, peculiarly Hel-
lenistic. It was at this time that, like their quasi-inscriptional kin, inscribed
epigrams began to appear in books. Plucked from the rich semantic eld of
their physical context (stripped even of such particulars of place as ancient
authors cite in quoting inscriptions),33 they were planted instead on the
arid, inexpressive plane of the papyrus, side by side with works that merely
simulate inscription. Upon this common ground, the formal similarity of
inscribed and quasi-inscriptional poems becomes a problem.
The epigram of Posidippus is a case in point. It is, however, difcult to
appraise, since our papyrus sourcethe selection of literary texts made by
Ptolemaeus and Apollonius, Macedonian brothers connected with the
Memphite Serapaeum in the mid-second century b.c. (cf. Thompson
1987: 1067)contains contextual clues that could point in either direc-
tion: Aufschrift oder nicht.
Speaking for inscription is the plausible thesis that its scribe, Ptole-
maeus, picked this poem for his selection to commemorate a personal
experience (Thompson 1987, 1988). The basis for this proposal comes
from another papyrus of the same archive (Urkunden der Ptolemerzeit
78.2839), likewise written by Ptolemaeus, in which he recounts a dream
set in Alexandria.
Me 6 /- | Y , 6 . , |
, " A " |
" | ,[] [[.]]
= | / . . .
volume, such Unvollstndigkeit der Angaben may be precisely one of those dedicatory and
sepulchral conventions (cf. Khnkens test 2) playfully exploited for artistic purposea delib-
erate spur to the readers imagination. On the other hand, numerous inscribed epigrams con-
tain superabundant information that cannot be explained with special reasons (cf., e.g., 781
Kaibel, concerning a heroon of Antigonus Gonatas). Incidentally, we nd berinformation
used as a criterion to ascertain the literariness of an epigram already in Hesberg 1981: 100105.
33. What Rosalind Thomas says of modern collections of epigram, applies as well to
their incipient counterparts in the third century b.c.: the inscription is usually isolated . . .
from the surrounding material, thus the total impact of the memorial is lost (1992: 63).
[I thought I was in Alexandria, on top of a great tower. I had a beautiful

face and did not want to show my face to anyone because of its beauty.
And an old woman sat down next to me, and [there was] a crowd of
people to the north and east of me . . .]
The vividness of this dream caused U. Wilcken (1927: 362 ad 29), the edi-
tor of the papyrus, to suggest that Ptolemaeus could have visited Alexan-
dria in his youth. And D. J. Thompson (1988: 261 n. 288) cautiously pro-
posed that the great tower was perhaps indeed the Pharos. Is it going
too far to observe that the dreamers location 6 is curiously like that
of the statue of Zeus Soter on the Pharos? (Does the dreamer identify with
the statue? That beautiful face he shrinks from showing is entirely apt
for a divine visage.) Epigram and dream may thus be linked more closely
than at rst appears: both may have as their subject that statue of Zeus
Soter on top of the Pharos. If Ptolemaeus recalled the epigram of Posidip-
pus as a memento of a personal encounter with the monumentsome-
thing he may have dreamed about long afterwardthen it is as likely for
him as for Strabo that he saw the poem in situ, inscribed on the lighthouse
itself. Unlike Strabo, however, he would seem to have understood that it
was about the dedication of that statue.
Yet the matter is more complex: W. Peek (1953: 432) observed that the
heading in the papyrus, 6, and the word be-
tween our epigram and the next suggest that the poems were copied from
an anthology.34 Both the label and the designation are typical of col-
lections, an editorial practice adopted here almost mechanically. The pos-
sibility that the poem was copied, not simply remembered, raises difcult
questions: Did Ptolemaeus have to rely on an intermediary source as his
memory receded with time? Did he, alternatively, never see the epigram
inscribed at all and know it only from literature (thus perhaps supplement-
ing a distant sighting of the Pharos)? Or might it never have been meant
for inscription at all?35
34. Thompson (1987: 113) notes that such labeling is reminiscent of a school-book
copy and, consequently, that the literary texts of the archive may represent those typically
taught at school. In fact, we know that epigrams on Ptolemaic monuments were included in
third-century b.c. schoolbooks, namely SH 978, 979; cf. Guraud and Jouguet 1938: 2026.
But Thompson concludes that the selection is more plausibly explained as a personal one:
The inhabitants of Ptolemaic Memphis clearly had available a wide range of literature which
included poetry and scientic and other learned works. This is not perhaps surprising in the
second city of Egypt, and the wealth of Alexandrias libraries was only two or three days dis-
tant by boat (1987: 109).
35. The latter possibility is certainly not to be ruled out. In that case, Strabos reference
to the Pharos 6 would not depend on Posidippus, though his description of the
The uncertainty underlying these questions points up not just how

hard it is to distinguish inscribed epigram from quasi-inscriptional. It re-
minds us, too, how easily the genre glides between media, from stone to
scroll or vice versageneric mobility quite literally construed (Thomas
1998: 205). Starting in the Hellenistic age, this genre will not stay put.
Epigrams easy, two-way passage between media is a characteristic hall-
mark of the age. We observe it, for instance, in scholarly projects early in
the Hellenistic period where, for the rst time, epigrams on stone were
systematically sought out, transcribed, and collected in booksfor in-
stance, in Philochorus / / or Polemon of Ilions
6 (cf. Ergnzungsspiel in the Epigrams of
Callimachus, chap. 5 in the present volume).36 Knowledge of such collec-
tions, where epigrams appeared at a remove from their inscribed setting,
doubtless encouraged poets to mimic and adapt the form for purely liter-
ary aims or else to realize that there might be (literary) life for their epi-
grams after inscription.
Inversely, poems destined for inscription might rst be written and
transmitted on papyrus. This was probably the case with two anonymous
funerary epigrams (SH 977) sent by letterperhaps in the hand of the
poet himselfto Zenon, agent of Apollonius, chief nancial administrator
of Ptolemy Philadelphus, to commemorate Zenons hunting dog Tauron,
who was fatally gored while protecting his master from an attacking boar.37
building, its setting, why it was needed, etc., would still arguably have been inuenced by
knowing the (quasi-inscriptional) epigram.
36. One gets a sense of how such projects shaped contemporary imaginings from a work
like Euhemerus Sacred Record, the ; / , whose author claims to have visited a far-
off island south of Arabia and seen there a large golden stele; its inscribed contentsa record
of the deeds of Ouranos and Zeus, revealed as illustrious men, not gods at allevidently
formed the heart of the work.
37. I consider it likely that letters (or short scrolls; cf. Turner 1968: 140)in addition to
presumably far rarer epigram booksserved as a critical medium by which much quasi-
inscriptional epigram was disseminated to poets (and other literati) in Egypt and abroad,
thus permitting authors to react to and vary the epigrams of their contemporaries (as we
know they loved to do) in works of their own (Ludwig 1968; Tarn 1979). The letters of
Archimedes (in Sicily) to Eratosthenes and Dositheus (in Alexandria) may serve as a model
for the sort of lively cultural exchange that could occur between far-ung correspondents in
the third century b.c. Signicantly, in one such letter, Archimedes sent Eratosthenes his
cattle problem in the form of an epigram (SH 201).
Such a scenario is on the whole more plausible than that proposed by Reitzenstein
(1893: 87104), that poets would compose epigrams extempore at symposia, often reacting
immediately to other such poems composed and performed by their drinking companions
a thesis now revived by Cameron (1995: 71103, esp. 7984) and Puelma (1996: 12627
with n. 15; cf. 130: modisch gewordenen Improvisationsdichtung im ktiven und erweit-
erten Epigrammstil).
The epigramsone elegiac, the other iambicare plausibly explained as

doublets, both of which would have been inscribed on the tomb (a well-
known practice in genuine funerary epigram). Though the poems deixis
evokes the physical presence of the grave site (@/ , v. 1;
/, v. 14; 6

/ . . . , v. 24), the medium of the letter
suggests distance, that is, that the epitaphs were sent to Zenon in the
Fayum from elsewhere (Parsons and Lloyd-Jones, SH 977 [ad loc.] plaus-
ibly weigh Alexandria). In that case, how tenuous, how conventional the
link between epigram and its physical object turns out to be: there is no
need to imagine the epigrammatist ever having visited the site or encoun-
tered Tauronexcept in Zenons description upon commissioning the
poems. Even inscribed epigrams, then, may be conceived and composed
remote from their setting and monument, and the physical context that
plays a signicant role in the experience of epigram in situ would here be
initially lled in purely in the imaginations of poet and patron. It is a thin
line between this situation and that of an epigram composed for the scroll
from the start.
On the other hand, even strictly quasi-inscriptional poems may sec-
ondarily become inscribed.38 That is the case with an epigram of Leonidas
of Tarentum about three brothers who dedicate their hunting nets to Pan
(AP 6.13 46 GP). The poem appears in a room of the Casa degli Epi-
grammi at Pompei, painted as a caption for a picture showing the dedica-
tory scene described in the poem (Dilthey 1876; Leach 1988: 21922,
37576).39 The poem is thus ostensibly reconnected with an object, the
I would not rule out the symposium as one possible setting for poetic creation and per-
formance. But it is scarcely sufcient to explain the range and quality of the epigrammatic
corpus as we have it. Besides the basic unlikelihood that the poets in question always hap-
pened to attend the same parties, it is hard to believe that (without prior warning that such
and such a theme would be on the evenings program) they were able to produce extempore
many of their nest epigrams. The model of Catullus 50, which Reitzenstein adduces (1893:
1034 n. 1: cf. also Cameron 1995: 88; Puelma 1996: 127 n. 15), in which each participant
writes versiculos . . . reddens mutua per iocum atque vinum (v. 6), is not a valid parallel. Cf. my
discussion in Text or Performance, chap. 6 in the present volume.
38. The urge to impart tangible reality to a literary ction is, of course, known from ear-
liest times. The epigram on the Ischia Cup of the eighth century b.c. (CEG I 454),
[,] _[] (assuming the second word is correctly restored), humorously por-
trays this tiny clay vessel (10 15 cm) as the huge, gold-decorated chalice of Nestor de-
scribed in the Iliad (11.63136), which only that great hero can lift with ease. Of course, in
that instance a ctional object is made real and given an epigram. In the case I am about to
discuss, a ctional epigram is inscribed on a real object. For more on the Ischia Cup, see my
discussion in Allusion from the Broad, Well-Trodden Street, chap. 8 in the present volume.
39. The room contains ve pictures in all, each coupled with epigrams (some from the
Greek Anthology). Of these, three seem to contain dedicatory scenes.
painting, which envisions a functional context for the dedication by giving

concrete physical detail to the dedicants, the image of Pan, and their set-
ting in a landscape. Still, this is clearly not an epigram set on its monu-
mentthough it certainly recalls that generic possibility. Rather, the
painting simply illustrates the text.40 It is thus crucially different from, for
example, the painting and epigram described by Herodotus (4.88) as dedi-
cated by Mandrocles in the temple of Hera on Samos to commemorate his
bridging of the Bosporus for the Persian king Darius.41 There, the paint-
ing is itself the dedication, appropriately set in a sacred shrine. At Pompei,
it decorates a rooma reading room, most likelyin a private house,
solely to signal its owners literary taste and cultural aspiration.42
A secondarily inscribed poem that more obviously attempts to mimic
genuine inscribed epigram is the fragmentary metrical inscription on a
Parian marble plaque displayed in the Archaeological Museum of Venice
and dated by Guarducci, on the basis of the lettering, to the second cen-
tury b.c. (1942: 2930 with photo).43 That dating, however, was attacked
and tartly corrected by W. Peek (1975: 26; see also Prioux 2002), who
noted that the verses had long been recognized as Theocritus epigram 17
(Gow), on a statue of Anacreon, and that the inscriptions authenticity had
already been questioned in the nineteenth century. Indeed, according to
Peek, the plaque is too thin ever to have carried a statue; one need only
glance at the photo to see that it cannot be ancient; the poem was copied
onto stone in the Renaissance and is merely a jeu desprit (though given
that it was rst described by the sixteenth-century artist and forger Pirro
Ligorio, I do not know why it could not be a deliberate counterfeit. Per-
haps that would be too obvious?). I have argued elsewhere that this same
epigram (as transmitted in the manuscripts) clearly signals to the reader its
ctionality qua inscription both through its style and its content (Bing
1988). It is interesting, that, despite such signals, there remains an innate
40. In this it resembles the mythological panels of the domus Musae in Assisi, which il-
lustrate epigrams incised in plaster outside the frames (Guarducci 1979; Leach 1988: 37576).
41. On this epigram, see Ergnzungsspiel in the Epigrams of Callimachus in the pres-
ent volume.
42. Cf. Leach 1988: 220: the room is a chamber tted out for the display of objets dart.
Such gallery rooms are often mentioned in descriptions of the furnishings of rich villas from
the Republican period onward. Many were doubtless of imposing proportions, but this small
room may be aptly explained by Ciceros mention, in a letter to M. Fabius Gallus (Ad Fam.
7.23.13), of a little reading room he wanted to decorate with tabellae. In fact, both the loca-
tion of the Casa degli Epigrammi room in a quiet corner at the back of the peristyle, and the
nature of its subject matter are appropriate for a reading room.
43. Cf. also the description in the museum guide (Tamaro 1953: 29).
potentiala kind of genetic memory of inscription within the genrethat

permits, perhaps invites, even clearly quasi-inscriptional specimens to be
transferred spielerisch auf einen Stein (thus Peek).
Given, then, that the boundaries between stone and scroll are quite
permeable and that migration across them is easy, one might ask, does it
really matter whether epigrams are inscribed or quasi-inscriptional? Many
scholars play down the difference: To a certain point the distinction is
invalid, says Richard Thomas, for instance (1998: 207).44 Beneath this
widespread view is an all-too-exclusive focus on strictly formal aspects of
the genre, on the basis of which, to be sure, it is hard to distinguish the
two.45 But I would insist that the difference mattersindeed, is crucialin
one important respect: how we experience the poem, or (in literary critical
terms) from the standpoint of reader response. Epigram for the scroll
involves another horizon of expectation, elicits a different response, and
has different expressive potential from epigram composed for the stone. In
either case, context shapes our reading, whether book context or physical.
But due to this genres peculiar history, the notion of context is intricate.
To readers of sepulchral, votive, and honoric epigram, their daily experi-
ence of such poetry in cemeteries, temple precincts, marketplaces, and so
on, links each instanceas it does no other genresto one permanent
manifestation in a single place.46 There it is set quite literally in stoneor
wood, clay, metal, and so on. When, however, they encounter epigram on
the scroll, its formal resemblance to the inscribed variety prompts readers
to experience the poems context as at least partly a lack of context. The
lack elicits a response, which is to use imagination to ll out the picture. In
instances when an epigram migrates from one medium into another, the
audiences memory also plays a role: what it knows of contextual history
will condition its response.
44. Cf. also Lausberg 1982: 97.

45. The presumption of formal identity underlies even Peeks Griechische Vers-Inschriften
(italics mine). Where it comes as a shock to the uninitiated to nd a wide assortment of poems
from the Greek Anthology. Though perhaps not as literal-minded as the inscriber of Theocri-
tus epigram on Anacreon (already treated), Peek does something comparable, namely, placing
quasi-inscriptional poems in an epigraphic context. For Peek sets inscribed and literarily trans-
mitted epigrams side by side according to type, in consecutive numeration and without typo-
graphic distinction. Above all, that authorial anonymity that is so striking a feature of most
inscribed epigram (especially funerary epigram) is imposed on the literary examples as though
to simulate the style of inscription. It is only upon looking at the adnotatio critica that one nds
that a given poem comes from the Greek Anthology and belongs to a particular author (cf.
Peek 1955: XVI).
46. By contrast, occasional poetrys existence between performances is not xed in a
unique location, and reperformance may occur in a variety of settings.
An example from the Greek Anthology may illustrate these points (AP
6.138).
+ / , / 6
( 6 /, .
[Previously Kalliteles dedicated me, but this his

descendants set up. Grant them favor in return.]
This epigram belongs to a series ascribed to Anacreon (AP 6.13445) and

added to the dedicatory section of Meleagers Garland, most likely from a
scholarly collection, since the rst nine are arranged alphabetically (Page
1981: 12324). As transmitted, the poems frame of reference is strikingly
unclear. To whom does the speaking voice (, v. 1) belong? The verb
() points to a dedicatory object, but what was ita statue, a col-
umn, a stele? Such questions multiply with in the next clause: was
this another sort of object altogether or similar, and where are we to
imagine it in relation to the speaking voice? Next, is there anything we
can say about Kallitelesapart from the fact that his ( consider-
ately mention his earlier dedication?47 And what deity is being asked to
give favor in return for ? Engraved on its object, set in a certain
place, such references would be unambiguous. Here on the page, they
lose their denition.
Such referential open-endedness encourages readers to interpret
names, actions, and objects as typical rather than historically specic (cf.
Gutzwiller 1998). I have argued elsewhere in this volume that in their
quasi-inscriptional verse, Hellenistic poets recognized and eagerly ex-
ploited this open-endedness as a spur to imaginative playErgnzungs-
spiel. I argued that in doing so, they took advantage both of the genres
traditional brevity and of the spare referential terrain of the scroll, which,
if not quite semantically neutral, still leaves far more to the imagination
than would the physical context of inscription.
The lemma to our couplet provides a modest instance of imaginative
supplementation encouraged by this setting. It consists of only three
words "
Cthe last of which demonstrates that its author
gave some thought as to which divinity might be addressed and looked for
help to whatever sources were at hand. Naturally enough, he based his
suggestion on the poems context. Since the previous poem was addressed
47. The name is quite common throughout the Greek world; cf. Fraser and Matthews
1987: I and II.
to Apollo, its lemma describing it as / . . . ,48 he sug-

gests that Apollo was similarly (C) the addressee hereas though the
dedications stood side by side in a precinct, rather than a text, and as
though we readers were touring a shrine, not perusing a book.49 A poems
place on the scroll, then, clearly contributes to the production of meaning,
stimulating readerly play. Epigrams in books possess this potential, a po-
tential not lost on their authors, who, starting in the third century b.c., also
sometimes edited and ordered their work for collections.50
The previous part of the lemma " , by the same (i.e.,
Anacreon)also reveals how placement on the scroll helps generate
meaning. Authors of inscribed epigram traditionally remain anonymous,
their identity subordinated to that of the person dedicating or honored by
the monument.51 By contrast, epigrams in books tend to be tied to specic
48. AP 6.137: , /, , / , "

/ #. Perhaps the similarity of requests ( ) also
suggested to the author of the lemma that the dedications belonged to the same god.
49. It appears that the poems in the section of the Milan Posidippus papyrus entitled
iamatika were meant, within the connes of a book, to create the impression for the reader of
touring a sanctuary of Asclepius. See Posidippus Iamatika in the present volume.
50. Puelma (1996: 12627) contends that the complete absence of the term epigram,
and of programmatic statements about the genre, among its authors until the rst century
a.d. suggest that poets did not consider it vollwertige Kunstdichtung . . . , die in durchkom-
ponierter Buchform der ffentlichkeit . . . prsentiert werden sollten. Rather, it was eine
zwang- und anspruchslose polymetrische Alltagspoesie, die Stoff einer mehr privatem Zweck
dienenden Auswahlkollektion poetischer Parerga bot als fr eine eigentliche kuntsgeme
Verffentlichung. Puelma nds conrmation in titles of early epigram collections, such as
or , and in their principles of organization: mechanischen Prinzipien
einer Inventarisierung (wie thematische Gliederung, alphabetische Reihenfolge, Autoren-
gliederung). But given that the values expressed in such titles lie at the heart of the poetic
program of the Hellenistic avant-garde, they can hardly suggest works unworthy of deliber-
ate and artful publication, norgiven what we know about how Hellenistic poets edited and
organized their workscan one assert that thematic (or even alphabetic) order is incon-
sistent with authorial intent (after all, Philitas, the author of , also wrote P
). Finally, the very earliest use of the term epigram in its broader meaning of short
occasional poem (rather than inscription) contradicts the notion that such poetry was artisti-
cally too humble to publish in a collection. For that rst instance occurs in the title
6 in a third-century b.c. papyrus (P. Petrie II 49a SH 961) with epigrams by
Posidippus and others and opening with a long dedicatory elegy that appears to be an epi-
thalamion for Queen Arsinoe: evidently an epigram collection might be an appropriate gift
for a royal wedding. This accords with the potentially high status of those who wrote epi-
gram: e.g., Posidippus was granted proxeny by Thermon specically in his capacity as
6, and Archimelus was paid one thousand medimnoi of wheat for celebrating
the ship of Hieron II with an epigram (Athen. 5.209ce). Why would such poets consider
their epigrams unworthy of collection and artful publication?
51. Cf. Gutzwiller 1998: 4748. The rare instances in which the poet is named begin in
the fourth century. Cf. Hansens list in the note ad CEG II 888 ii; cf., further, Page 1981: 121
n. 2. All are dedicatory until the second century b.c.
poets. Thus, inasmuch as our couplet is set in a group ascribed to

Anacreon, readers are disposed to relate it to other poems " , and
they expect to ndindeed, will imaginecharacteristics they associate
with this particular poet, whether of style, content, poetic persona, or the
like. It is no surprise, then, that various critics have supposed our couplet
to embody that peculiarly Anacreontic virtue of elegant simplicity: its
context on the scroll suggests it. More on this in a moment.
In this instance, however, the play of imagination has been drastically
curtailed by the discovery of the couplet inscribed in a single line on a
marble herm near Daphni in Attica (IG I.381 107 FH Anacreon IX
FGE p. 140 CEG I 313).
+ h D [ + 6]
( ( [ h ].
[Previously Kalliteles made the dedication, but this one his

descendants set up. Grant them favor in return.]
We see at once how the focus narrows. refers to the herm, and the
closing imperative is thus addressed to Hermes, not Apollo. ,
which follows in the Greek Anthology, is absent from the in-
scription, so the speaking voice is impersonal. This initially seems to open
the nature of Kalliteles dedication ( h ) to a broader
range of interpretation. But in fact, again, it limits it. For while implies
the concurrent presence of two dedications, h s lack of object in
the inscription suggests that Kalliteles once dedicated a herm, which his
descendants have replaced; the earlier herm is no longer there.52 Epi-
graphic considerations push the date toward the second quarter of the fth
centuryso we can denitely rule out Anacreon as author.53 And once we
know the place in which the inscription stood, we can try to identify this
particular Kalliteles by connecting him with comparably named individu-
als from the same region.54 We will do this because, as Gutzwiller points
52. Cf., e.g., FH ad loc., followed by FGE ad loc. K.-H. Stanzel suggests to me that the
initial verbs lack of object gives concrete expression to the absence of that earlier herm.
While it is possible that the earlier dedication was not a herm (cf. Hansens comments ad
CEG I 313), that is not likely. For the lack of a denite object, immediately followed by ,
this one, strongly suggests that the object of h was essentially the same. Indeed, on
rst reading, one could initially construe as the object of that verbuntil readjusting
the sense to take account of the subsequent verse.
53. Cf. Trypanis 1951: 33; FGE ad loc.
54. Cf., e.g., Beckby 1957: ad loc.: Bekannt ist ein Bildhauer Kalliteles von Aigina, der
um 460 blhte; er kommt also nicht in Frage, vielleicht ein Vorfahr.
out (1998:8), readers of inscribed epigram conventionally assume that the

people named in it are real.
In all, then, a modest distich becomesin aesthetic termsmore mod-
est still, reduced from a work of literature that could pass for Anacreon
to a rather undistinguished piece of Gebrauchspoesie,55 with no pretension
beyond its immediate pragmatic purpose. While Friedlnder and Hofeit
suggest that it was the epigrams elegant simplicity that must have ap-
pealed to the collector,56 there is, in fact, nothing in the diction to justify
the characterization elegantno individual word that is especially po-
etic or prosaic. Indeed, h s lack of object makes one stumble (per-
haps the reason for in AP ). Simplicity, on the other hand, is there in
spades.57 Far more, it is the fact of its inclusion in a collection and there-
after in the Greek Anthologycoupled with the forgetting of its origins
that rst lends the couplet dignity and artistic worth, opening the way for
a different experience of the poem.
To sum up, a readers response to the genre is conditioned by context,
though context of various kinds. First, for cases that stay xed in their re-
spective media, we make the distinction between seeing epigrams in-
scribed and seeing them on the scroll. For inscriptions, a reading com-
prises not just the text but its physical context: the kind of object on which
it is engraved, its geographical setting, sociohistorical circumstances, and
so on. On paper, by contrast, an epigrams surroundings are pared to a
minimum. The reader construes such textual landmarks as titles and
headings, identies authors (like Anacreon in the preceding example),
or interprets a poems setting in relation to others on the page or within
the collection. Yet the physical context of inscription remains an essential
reference point even in quasi-inscriptional poems, only now the details
are left to the readers imagining. A readers urge to engage his imagina-
tion in this way is strong, virtually a reex, or habit of mind, based on
deeply engrained expectations from daily experience of the genre in in-
scription. Put somewhat differently, the reader supplies those elements
that he is used to seeing in life and that he was used to seeing even before
55. On the distinctions between literary and subliterary epigram, cf. the very important
comments of Lausberg (1982: 9697 with n. 4).
56. Friedlnder and Hofeit 1948: 104, echoed by Pfohl (1968: 159: Darin [sc., in der
AP] scheint unser Distichon wegen seiner elegant simplicity einen Platz verdient zu haben)
and Lausberg (1982: 435).
57. In other words, the answer to Pfohls groe Frage . . . , warum die Inschrift ber ihr
Monument hinaus bekannt wurde may be far more banal than the one he offered: man
denkt an eingngige Formulierungen mit ansprechendem Inhalt, an bedeutende Denkmler
(1968: 157).
he had learned how to read (the sight of inscribed monuments would,

after all, have been familiar almost from the moment he grew conscious
of the outside world). Thus, in concert with the genres traditional conci-
sion, the extent of the readers role in constructing meaning exceeds what
is found in other genres.
Turning now to epigrams that migrate between media, the couplet of
Anacreon clearly shows that contextual history also makes a difference. If
we know it was inscribed, and where, our reading changes.58 For we bring
contextual data from one medium to bear on the other. Thus the genre con-
sists not just of its formal aspects (the references to objects, settings, donors,
etc., mentioned earlier). It is shaped as well by its function or Sitz im
Leben, which alters our experience of the poem.59 To be sure, if unaware of
its double status, we may still read an epigram as quasi-inscriptional.60 Yet
58. It is interesting, however, that the couplets context in the Greek Anthology still ex-
erts residual inuence on interpretation, as when Friedlnder, Hofeit, and others assert the
couplets elegant simplicity.
59. On this important topic, cf. Jauss 1982: 103. Jauss comments on medieval genres
and their liberation from function and movement toward the autonomy of literature in
the Renaissance apply also to the development of epigram in the Hellenistic era: the distinc-
tion between . . . functional constraint and literariness has meaning in the Middle Ages only
when it is understood as the process of gradual literarization of genres that originally are tied
to cultic, religious and social functions. For ancient genres, cf. Kppels discussion (1992:
1721) of Der Sitz im Leben als gattungskonstituierende Komponente.
60. Most readers of the Greek Anthology are probably oblivious to our couplets origin
(since few, apart from specialists, will trouble to read the critical apparatus or notes): so long
as they do not know of or recall its inscription, they retain the possibility of reading the poem
as quasi-inscriptional. The less notable the monument, the more likely that is (readers are less
likely to forget the fact of an epigrams inscription when it belongs, e.g., to a famous building
like the Pharos). Inscribed epigram, such as Anacreons, may thus acquire a second life
entirely independent of its rst, in which it functions to quite different effect.
Further, I suspect that a poem might be conceived for the scroll as well as the stone,
whether simultaneously or as an afterthought on the part of its author. Such ambidextrous epi-
grams are particularly likely in the Hellenistic period, when poets began shaping their work
into collections. A good candidate among epigrams is Callimachus 33 Pf. ( 21 GP), which
if ever inscribedwas probably on a very humble object that few of its readers on the scroll
would have seen or known of. I discuss this poem in detail in Ergnzungsspiel in the Epigrams
of Callimachus in the present volume. (The obvious nonepigrammatic model is Callimachus
Aetia, where such evidently occasional, once-independent poems as the Coma Berenices enjoy a
parallel life in the larger work to quite different effect.) But we must distinguish such a case,
where an author contemplates two distinct media, aware that his readers experience in each
will be different, from that of Anacreons couplet, meantand meant onlyfor a single
stone at a certain site. In the latter case, though the readers experience may also change with
the medium, its literariness is not essential but extrinsic. One might compare the status of a
found object in art, arrogated to new ends by someone unconnected with its prior function.
Yet as in the case of Anacreons couplet, so long as we know its earlier function (i.e., so long
as it is recognizably found), its origin shapes our response to its altered state.
we can never read the couplet by Anacreon in the same way again once we
recall its setting on the herm. And, after all, that simply raties the memori-
alizing function of the monument: it wants us to remember it.
Returning nally to the epigram of Posidippus, I would stress that while
knowing of this (or any other) poems inscribed origin will inuence our
reading, it need not always have that narrowing effect on the sense that we
found with Anacreons couplet. We cannot count on a neat distinction
between indeterminacy of meaning on the scroll and xity in stone. On the
contrary, knowledge of the inscribed context may produce ambiguity. Ear-
lier, I proposed two hypothetical scenarios in which Posidippus epigram
might have been seen inscribed on the Pharos. In the rst, I suggested that
Strabo could have misconstrued Sostratus role in connection with the
lighthouse if the epigram was inscribed on the tower rather than right be-
neath the statue of Zeus, where it would have been hard to see. In the sec-
ond, I raised the possibility that the scribe Ptolemaeus may have identied
in his dream with the statue of Zeus Soter atop the Pharos, a place he may
actually have visited, and that he transcribed the poem because of that visit
and, in particular, his interest in the statue. On this view, he would have
understood correctly that the subject of the poem was the dedication of the
statue. Perhaps neither imagined scenario is true. But if they are at all plau-
sible, they reveal the interpretive pitfalls that may affect the intelligibility of
an inscribed monument. An inscriptions physical context seems so stable
and univocal. But at times, there is no clear unambiguous marker, like a
statue of Zeus Soter, to guide us to a sure conclusion.
c h a p t e r 11
Posidippus Iamatika

T he Milan papyrus confronts its modern readers with many surprises,

among themdue to its singular subject matterthe short section
entitled ,. To help us get our bearings in the terrain of this ex-
traordinary new text, I want, in this essay, to pose some rudimentary
questions, such as the following: (1) What are the epigrams about? (2) Do
they function as an ensemble? (3) Where should we seek their generic
antecedents or models? and nally, (4) What kind of reader response do
they elicit?
To begin with the most basic, what are these epigrams about? The
, comprise only seven poems, all four verses long except the rst,
which is eight,1 and all concerning cures, ,. These cures are sudden
and miraculous; they appear as testimony to the benecent power of the
healer god Asclepius.2 With the exception of iamatika 1 (95 AB), the
This is an updated version of an article that appeared in Labored in Papyrus Leaves: Perspectives
on an Epigram Collection Attributed to Posidippus (P. Mil. Vogl. VIII 309). ed. B. Acosta-Hughes,
E. Kosmetatou and M. Baumbach, 27691. Cambridge, MA 2004. 2004 Center for Hel-
lenic Studies. Trustees for Harvard University. For their penetratingand therapeuticcri-
tique of earlier drafts, I thank Profs. D. Bright, J. Lee, and C. Perkell. If this essay nonethe-
less remains uncured of all its defects, that is due entirely to the authors pathological
stubbornness.
1. In other words, it totals thirty-two verses, as does the following section, . The
only section with fewer verses is that immediately preceding, the with twenty-six
(8994 AB). Benjamin Acosta-Hughes points out (in conversation) that these short sections
grouped together resemble Callimachus Iambi 2 and 3.
2. For suddenness as a characteristic of miraculous cures, see Weinreich (1909: 19798).
217
poems consistently invoke Asclepius3 or have to do with his cult.4 Indeed,

even the doctor of poem 1, who discovered a cure for the bite of the
Libyan asp, owes his success to having received all the panacea of Ascle-
pius sons (95.56 AB).5
Proceeding to our second question, and in light of so unied a subject
matter, we ask whether the poems comprise an artfully arranged set, that
is, a deliberately planned sequence with marked beginning and end. Let us
assume, as a working hypothesis, that they do. Given the section heading
,, this is an inherently plausible hypothesis. The title, like that of
almost all the other sections, calls attention to the process of editing and
classication, in that it consists of the substantivized neuter plural with
the denominative sufx -, indicating a collection of things classed to-
gether, in this case things having to do with cures, ,.6 Whether the
editorial hand belonged to the poet himself or to a somewhat later com-
piler, we need not say. In any case, the hypothesis allows us to ask
potentially useful questions about the ways in which the poems may work
together to form a meaningful whole. What, for instance, can we say
about the topics covered in the epigrams? Do they signify as a group? I
believe they do. For the god is portrayed in the poems as curing an im-
pressive range of illnesses. Taking the epigrams in order, we ndin ad-
dition to the snakebite of the rst poemcases of paralysis, epilepsy (the
sacred disease), an infected wound made by a metal weapon, deafness, and
blindness. It is striking that there is no overlap in these illnesses, that they
are each quite different in kind. I think we may thus plausibly conclude
that they were intended to be a representative assortment, to stand collec-
tively for the entire spectrum of disease.
That exemplarity squares well with another consideration, namely, the
identity of the doctor who is the subject of the initial poem, Medeios of
Olynthos, son of Lampon. It is highly probable that this doctor was the
man of identical name and patronymic known to have held the prestigious
3. 96.1 AB: /; 97.1 AB: /; 98.3 AB: ; 101.1 AB: / .
4. 99.3 AB: / " / ; 100.1 AB: \ a , (incubation).
5. His dedication is strictly speaking to Apollo. It is a statue portraying a wasted remnant
of a manthe sort of patient he used to save through his discovery. Aristotle (History of Ani-
mals 8.607a) mentions a remedy for this snakes bitethe so-called septic drugwithout spec-
ifying its inventor: 6 # = [sc., ], .
Other authors insist that the bite of the asp is incurable and fatal: cf. Ap. Rh. 4.1508ff.; Ael. NA
1.54, 9.15. The identity of this doctor, Medeios of Olynthos, is discussed shortly in the text.
6. Cf. Khner and Blass 18902: II 294 no. 334.5. This is by far the earliest instance of
the adjective ,. The same appears to be the case for (cf. Diodorus 7.1.13),
(otherwise rst in Polybius 27.18.23), and .
Posidippus Iamatika 219
position of eponymous priest of Alexander and the Theoi Adelphoi at Alexan-

dria in the year 259/8 b.c. and attested in the following year as overseeing
the proceeds of the royal tax for medical services, ,.7 He was, in
other words, a VIP, a player in the upper reaches of Philadelphus court.
The section thus opens in a distinctly Ptolemaic key. But more, considering
the rst poems markedly greater length (double that of any other in the
section) and emphatic position, we may reasonably wonder whether the
Ptolemaic note carries over into the remaining six poems and whether that
representative assortment reects the medical interests of this courtier and
was perhaps even compiled in his honor.8
Further signs of unied design appear when we ask if the epigrams col-
lectively evoke a particular social context. Again, we can answer in the afr-
mative; they clearly do. The context is that of sanctuaries of Asclepius, such
as Epidaurus, the most famous, where pilgrims ocked from throughout the
Greek world to seek the gods aid in overcoming illness. In the shrine, they
offered sacrice and prayers andmost strikinglyunderwent incubation in
the hope that the god would appear to them in their sleep, either to cure out-
right or show them the road to health. Taken together, the epigrams echo
the rhythms of life at precisely such a shrine: patients journey to the god;9
they sacrice and pray;10 they experience the gods power in the night11 and
7. I have presented the case for identication fully in Bing 2002a. Bastianini and Gal-
lazzi (BG ad loc.) overlook the possibility.
8. Cf. my discussion of the who is the subject of the sections nal poem
(101.1 AB), at the end of the present essay. One may wonder why Posidippus is silent con-
cerning Medeios political appointments. Though there can be no certainty, I can imagine a
couple of scenarios: perhaps the poem comes from early in Medeios life and reects his ac-
tivities prior to his remarkable political rise, or perhaps it comes from late in his life, when he
was thinking about how he wanted to be remembered. The imperfect tense of the verb refer-
ring to his therapeutic activity, 6, [the ones] he used to save (95.3 AB), would suit a
time when he looked back on his medical career as something in the past. It brings to mind
the epitaph of Aeschylus (Vita 25), which makes no mention of his poetry at all, saying of his
life only that the famous grove of Marathon could tell of his prowess and the thick-haired
Mede learned it well.
9. 96.12 AB: Antichares came to you, Asclepius, with two canes, / dragging his step
along the path ( + + /, /, 1 / V /
: 6).
10. 96.3 AB: [+ ] ; 99.3 AB: / " / ; 101.3 AB: ,
/ #.
11. 97.4 AB: , e , divinity, you came and wiped away [the
disease] in a single night; 98.34 AB: , / _[ , ] , E 6/ 4 /
1 , [ 6] , when painless [he beheld you gracious], Paean. So after
the dream / being cured [he escaped] his great toil; 100.1 AB: ?/ ( Z \
a ,, When Zenon had to sleep that gentle sleep. On Posidippus use of incubatory
formulae, cf. Di Nino 2005: 6063.
make thank offerings;12 nally, they go back home again when theyre
done.13
Still other meanings emerge from these poems if we view them as an en-
semble. I would suggest that the section offers readers the impression, as
they turn from poem to poem, of strolling through a shrine of Asclepius. It
allows them to play the part of imaginary pilgrim orin a more detached
modeuninvolved observer. In fact, as we shall presently learn, the ofcial
testimonials collected and inscribed by the authorities at Epidaurus them-
selves provide the model for such a reading of the iamatika: they repeatedly
envision visitors to the shrine strolling about and perusing inscriptions.14
In this guise, Posidippus reader encounters diverse monumentsa cross-
section typical of what a pilgrim might actually have seen at a healing
shrineand pauses to read about them, as the pilgrim might have done
through inscriptions. There is the statue in 95 AB; the votive phiale of
Coan Soses (97 AB); numerous testimonies recounting the wondrous cures
of the god, such as those on the countless pinakes that lled the shrine (96,
98100 AB); and, nally, a worshipers prayer (101 AB).15
The evocation of such a setting in these poems leads to our third ques-
tion: to what genre do Posidippus , belong? What is their model?
There are notably few epigrams about cures in the Greek Anthology, and
the handful we possess are mostly late in date.16 Among inscribed epi-
grams, there is a comparable dearth.17 Those that we have, most datable in
the fourth century b.c. and after, are strikingly reticent about anything
miraculous in the cures. Take, for example, the hexameter epigram by the
orator Aeschines that survives both in the Greek Anthology and as an in-
scription at Epidaurus and happens to be our earliest acrostic (CEG 776).
[A, /] / | [/
].
+ , +
6 (, 5 _ / ,
12. 97.12 AB: , , /, /

.
13. 99.3 AB: :/ . See Di Nino 2005: 6366.
14. The most explicit is IG IV2, 121.22, examined later in this essay, but cf. also IG IV2,
1 no. 121.33.
15. Though framed as a maxim, the epigram in fact functions as a prayer, though indi-
rect, for good fortune and health. Cf. my discussion later in this essay.
16. Cf. AP 6.203 (Philip of Thessalonica), 6.330 (Aeschines), 9.46 (Antipater of Thessa-
lonica), 9.298 (Antiphilos), 9. 511 (Anonymus).
17. Cf. CEG II 776 AP 6.330 (Aeschines acrostic inscription from the rst half of the
fourth century cited in the following text), 808, 818; Kaibel 8035; IG IV2, 1 no. 125 Edel-
stein and Edelstein 1998: T 431. Except for the last, these references are from Rossis discus-
sion of Theocr. Ep. 8 (2001: 197).
, 6, /, ,
I ( 6, 6 .
[Aeschines, son of Atrometus, from Athens dedicated [this] to Asclepius.

Despairing of mortal skill and putting all hope
in the divine, I left Athens of the fair youths
and, coming to your grove, Asclepius, was cured
in three months of a sore Id had on my head for a year.]
Here there is neither a spectacular illness (just a headsore) nor sudden

miraculous cure. It takes three months for the sore to heal.18 A comparable
reticence appears in other fourth-century inscriptions.19 None of them
sound like what we nd in Posidippus, where the sudden, miraculous cure,
which may include paradox, is the rule.
Interestingly, what the marvelous tales in Posidippus Iamatika most
closely recall, by contrast, are the prose inscriptions set up by temple au-
thorities in the second half of the fourth century b.c. at the sanctuary of
Epidaurus (IG IV2, 1 nos. 12124).20 Four large, carefully incised stelae
survive,21 containing accounts of some sixty-six miraculous cures.22 They
18. Zanetto (2002: 74) strangely prints Stadtmllers in place of , perhaps so

as to make the cure appear more miraculous.
19. CEG II 808: [] / , / , | h h
/ (
/ (Androcritus, the son of Hagillus of Aegina, dedicated me to Asclepius for his
healing skill); 818: / ( , /[], / | #
/ / (In exchange for good works, Asclepius, Antiphilos dedicated these; / they are
gifts for himself and his children).
20. Zanetto (2002), in an article unknown to me at the time I wrote this essay, reached
similar conclusions about the Epidaurian prose inscriptions as Posidippus model. For treat-
ments of these inscriptions, see, in addition to IG, Herzog 1931; Edelstein and Edelstein
1998: no. 423; LiDonnici 1995. For further inscriptions (not, however, including the afore-
mentioned stelae) from Epidaurus and elsewhere, cf. Girone 1998.
21. Pausanias evidently saw these stelae (2.27.3), since his description is a close match:
In my day there are six left of the stone tablets standing in the enclosure, though there were
more in antiquity. The names of men and women healed by Asclepius are engraved on them,
with the diseases and how they were healed; the inscriptions are in Doric. Cf. Tzifopoulos
(1991: 1920).
22. Though similar inscriptions also occur in the second century b.c. at the sanctuary of
Lebena on the southern coast of Crete, they are not found elsewhere. Lynn LiDonnici has
observed (1995: 42): The preserved nds from the three known major mainland Asklepieia,
Epidaurus, Corinth and Athens, present the appearance of a . . . regional style or preference
for certain types of votives over others. Epidauros is best known for narrative inscriptions;
Corinth lacks inscriptions but is rich in terra-cotta body-part votives, while Athens and
Piraeus have many stone votive reliefs, without any text. Each of these types is poorly repre-
sented from the other sites. This may reect the taste of the respective districts and the avail-
ability in each area of craftsmen and materials.
are entitled not , but [/] /

/ . You could just as well call them Asclepius greatest hits: they
were evidently culled from the great mass of votive tablets ( pinakes) that
lled the shrine and that were periodically cleared and buried by temple
personnel so as to make room for more.23
We catch a glimpse of the transition from private votive to that ofcial,
collective text, as well as the addition of a notably miraculous element in
the process, in the very rst narrative on the stele, for here the private and
relatively modest votive inscription is quoted within the text.24
[K]5 / ( 6. a / 61 A |
[] 6 6 D E + | []
6 6 " 6
6, (, T " | []1
" 6 9 |
[].
+ 6 6D " |
[] , , / ( E 6 6 |
5 , ( 6 ( #.
[Cleo was with child for ve years. After she had been pregnant for ve
years she came as a suppliant to the god and slept in the Abaton. As
soon as she left it and got outside the temple precincts she bore a son
who, immediately after birth, washed himself at the fountain and
walked about with his mother. In return for this favor she inscribed on
her offering: Admirable is not the greatness of the tablet, but the di-
vinity, in that Cleo carried the burden in her stomach for ve years,
until she slept in the Temple and he made her sound.]
(IG IV2, 1 no. 121.39).25
The only miraculous element in the quoted tablet is the ve-year preg-
nancy. The framing narrative embellishes this nucleus with Kleos urgent
23. LiDonnici (1995: 66) suggests that collection may have occurred every few years as
the sanctuary became overloaded with votives and that there may have been several epi-
sodes of collection and arrangement of tales onto successively larger and probably less nu-
merous stelai.
24. The included votive seems to have been metrical (two hexameters and a pentameter):
" [] , ,
/ ( E 6 6 5 , (
.
6 ( #
The inscription does its best, however, to mask the meter and assimilate it to the surrounding
prose by breaking lines in midverse and midword ( | [], | ) and using scriptio
plena (6 6, ( # ).
25. Trans. Edelstein.
departure from the temple and immediate birth. It has her deliver, more-
over, a child of incredible maturity, who washes himself in the fountain as
soon as he is born and is then able to walk aroundand presumably home,
as wellwith his mother. The modest votive is thus transformed from a
personal commemoration into a wonder tale celebrating the gods miracu-
lous benevolencean aretalogy, in other words.
We see here how the texts selected for preservation were exceptional
and wondrous in character (or capable of being made so), able to fulll an
ongoing aretalogical function that set them apart from workaday com-
memorations of less spectacular illnesses. Four of Posidippus epigrams
(96, 98100 AB) share this aretalogical character, apparently not conform-
ing to any other epigrammatic type (though cf. the discussion of 100 AB
later in the present essay).
I would suggest that inscriptions such as those at Epidaurus provided
Posidippus with a model for these poems, particularly with regard to his
content. In general, the maladies enumerated in those poems concerned
with Asclepius nd close parallels in the Epidaurian inscriptions, as do the
character of the cures.26 Recent scholars have explored how the early Hel-
lenistic poets display in their epigrams a keen awareness of their genres in-
scriptional roots and often playas epigrammatists of later generations do
noton contemporary epigraphic topics and style.27 Posidippus was well
positioned to indulge in such play, for in addition to being a master of the
literary tradition6 , as he describes his soul in a
previously known epigram (AP 12.98.3 6.3 GP AB 137.3)he also
knew his way around monuments, winning acclaim at particular shrines as
26. The lame man who approaches the god on two canes in 96 AB resembles a paralytic
in the inscription ,5 | [, ] # 6
(IG IV2, 1 no.
123.123ff.). The epilepsy that plagued Coan Soses in 97 AB is likewise represented at Epi-
daurus (IG IV2, 1 no. 123.115). Numerous cases in the inscription record cures from weapons
lodged and festering in the body. Posidippus Archytas, who had kept the deadly bronze for
six years / in his thigh . . . a festering wound (98 AB), recalls the Epidaurian example of
Euhippos, who had had for six years the point of a spear in his jaw (IG IV2, 1 no. 121.95
Edelstein and Edelstein 1998: T 423, 12; cf. Zanetto 2002: 7578; cf. also IG IV2, 1 no.
122.55 and 64 Edelstein and Edelstein 1998: T 423, 30, 32). Interestingly, there dont ap-
pear to be any cases of deafness among the Epidaurian iamata to compare with that of Asklas
the Cretan in 99 AB. But cf. the late Epidaurian inscription of Cuttius the Gaul (IG IV2, 1,
no. 440). Finally, blindness is a common ailment in the sanctuary of Asclepius (cf. IG IV2, 1
no. 121.33, 72, 120, 125; no. 122.7, 64; no. 123.129; cf. also AP 9.298). In iamatika 6 (100
AB), Posidippus gives it a paradoxical twist, however, by having the aged Zenons restored
sight last for only two days before he dies.
27. See particularly Rossi (2001); Fantuzzi and Hunter (2004: 291338), in the section
Funerary and Dedicatory Epigrams: Epigraphic Conventions and Epigrammatic Varia-
tions of the chapter The Epigram.
a poet of inscribed verse. An inscription at Thermon from 263/2 records

that the Aetolian League granted him proxeny in his capacity as
6 (IG IX 12; 17A.24 test. 3 AB), and he also seems to
appear in a proxeny list from the 270s at Delphi (Fouilles de Delphes III 3
no. 192 test. 2 AB). As to the iamatika, we need not think exclusively of
Epidaurus. I wonder whether Posidippus work in Egypt exposed him to
the cult of Imouthes/Asclepius at Memphis, where, as Dorothy J. Thomp-
son notes, the prayers and expectations of the Egyptian stelae with tales
of miracles the god performs are indeed similar in tone and content to
contemporary Greek inscriptions from the shrine of the god Asklepios at
Epidaurus (1988: 210). Poems by Posidippus certainly made it to Mem-
phis, as we know from the Firmin-Didot papyrus, so why not the poet
himself? In the ,, Posidippus used his familiarity with such settings
to draw on an epigraphic model that scholars had not previously contem-
plated. Yet, in principle, it is a typically Hellenistic move: the poet here
translates the subject matter of a prose genre into poetic form and shifts it
from its inscriptional medium onto the scroll.28
28. In his Plutus, Aristophanes re-created the workings of such a shrine in comic verse
and to hilarious effect. Comedy, however, does not evoke the inscribed tradition, while epi-
gram insists on it through its generic history and retention of epigraphic conventions.
Interestingly, among the Epidaurian inscriptions, we may nd an epigraphic counterpart
to the transferral of prose narrative into versein this instance, inscribed verse. On the rst
of the Epidaurian stelae (IG IV2, 1 no. 121.107 Edelstein and Edelstein 1998: T 423, 15),
there is a third-person account of the miraculous transformation of Hermodikos of Lamp-
sakos from helpless paralytic into mighty muscleman, capable of superhuman feats.
; . 6 | ,
6 6 6 , | , C D C +
| | A.
[Hermodikos of Lampsakos was paralyzed in body. This one, when he slept in the Temple,
the god healed and ordered him upon coming out to bring to the Temple as large a stone as
he could. The man brought the stone which now lies before the Abaton.] (trans. Edelstein)
As the commentary in IG notes, this stone has been found: repertus est lapis proxime a templi
latere orientali. Pondus computandum fecit Bl. C. 334 kg. Herzog (1931: 102) estimated the
weight at 375 kilograms, that is, 1,045 pounds.
An inscription postdating this narrative, apparently from the third century b.c.litterae
saec. III a. Chr. elegantes, as stated in the IGtranslates it into rst-person poetry (IG IV2, 1
no. 125 Edelstein and Edelstein 1998: T 431 Girone 1998: II.3, pp. 5357).
;E[ ]
[]/, /, | [

] | , [ C
] , | |
= D | , 6 |
| a
|
( g | |
D 1 , | ,
| , .
This brings us to our nal question, regarding reader response. How

should a reader evaluate the miraculous cures in Posidippus epigrams, sit-
uated as they are at a remove from both their generic model (the prose in-
scription, already remote from its source in private votives) and its physical
setting in a shrine?29 Do the poems endorse the cures or subvert them? In
inquiring thus, we implicitly ask as well how Posidippus, an erudite and so-
phisticated reader, evaluated the cures he encountered on stone at such
shrines as Epidaurus and how his literary reworking of such material re-
ects his response, as mediated for his readership. Indeed, it is well to bear
in mind the multiple layers of interpretive mediation in play here: in expe-
riencing the iamatika, an audience is reading Posidippus readings of the
Epidaurian authorities readings of the personal votives. In any case, to
contemplate the iamatika on the scroll is to have an altogether different ex-
perience from that of the stricken pilgrim encountering engraved iamata at
a shrine, and the distance between these experiences may incline the poets
readers toward a more detached perhaps even skeptical stancean inclina-
tion no doubt even stronger for a modern scholar/reader, with the added
distance of time and the attendant change in mentality.30
Leaving the epigrams aside for a moment, it is useful to recall that
people in ancient times could be quite skeptical of what went on in healing
sanctuaries. While the archaeological record leaves no doubt that such a
sanctuary as Epidaurus enjoyed enormous popular esteem, particularly
[Hermodikos of Lampsakos
As an example of your power, Asclepius, I have put up this
stone which I had lifted up, clear for all to see,
a manifestation of your art. For before I came under the care of your hands
and those of your children, I was stricken by a wretched illness,
an abscess in my chest, my hands paralyzed. But you, Paean,
by ordering me to lift up this rock made me live free from disease.]
(translation after Edelstein)
It may be, of course, that the prose version was taken from an original verse inscription (dam-
aged or worn over time) and hence reinscribed. Could it be, however, that the verse inscrip-
tion is the secondary phenomenon and that such literary epigrams as Posidippus prompted,
in their turn, demand for poetic versions? It is worth noting that the epigram heightens the
miraculous element by specifying paralysis of the hands and an abscess in the chestpre-
cisely those parts of the body with which Hermodikos presumably hefted the huge boulder
while the prose inscription leaves the ailment as the more general paralysis of the body,
.
29. For epigrams shift from monument to scroll and its impact on reader response, cf.
Ergnzungsspiel in the Epigrams of Callimachus, Between Literature and the Monu-
ments, and The Un-Read Muse? in the present volume. See also Gutzwiller 1998: 47
114; Fantuzzi and Hunter 2004: 291338.
30. Cf. the comments of Di Nino (2005: 7576), which tend in a similar direction.
from the mid-fourth century onward,31 nevertheless the cult did not elicit
universal trust. There is the story that tells how Diogenes the Cynic once
saw a woman prostrate herself before the god: Wishing to free her [and
those like her] of their superstition, . . . he dedicated to Asclepius a erce
rufan who, whenever people prostrated themselves, would run up to them
and beat them up ( " . . . M
/M , T 1 6 6 -
, Diog. Laert. 6.3738). Now one might consider the skepticism of a
Diogenes an extreme case, but we nd similar sentiments voiced among the
inscribed Epidaurian , themselves.
One text in particular, IG IV2, 1 no. 121.22, placed near the start of the
rst stele and so perhaps intended programmatically, presents starkly con-
icting assessments of the gods miraculous cures. Perhaps it can also help
us set interpretive parameters and can provide a baseline for evaluating
possible reader response to Posidippus iamatika.
/ 1 ( | 2
D + 1 6
|
, # 6 | | []. 6 + =
,D 6 # | [] "
, 6 | [] 6 6
6 # 1 | D E / ,
/ I 6 | D 6 +
6, 6 , | | , (
6 6 | , " / "
. @ ( | "[] " 6 ,
( , , P | = []. Z +
# 6.
[A man whose ngers, with the exception of one, were paralyzed, came
as a suppliant to the god. While looking at the tablets in the Temple he
expressed incredulity regarding the cures and scoffed at the inscrip-
tions. But in his sleep he saw a vision. It seemed to him that, as he was
playing at dice below the temple and was about to cast the dice, the
god appeared, sprang upon his hand, and stretched out his [the pa-
tients] ngers. When the god had stepped aside it seemed to him [the
patient] that he [the patient] bent his hand and stretched out all his
ngers one by one. When he had straightened them all, the god asked
31. Cf. Tomlinson 1983: 25; LiDonnici 1995: 1011.

him if he would still be incredulous of the inscriptions on the tablets in

the Temple. He answered that he would not. Since, then, formerly
you were incredulous of the cures, though they were not incredible,
for the future, he said, your name shall be Incredulous. When day
dawned he walked out sound.32]
I believe that we have here one potential road map for interpreting the
iamatika of Posidippus. For this is a text about how to read accounts of
miraculous cures. In it, we are presented with two models of reader re-
sponse, twin poles marking the endpoints along an axis of belief, the one
skeptical, the other favoring credence. Of course, the narrative strongly
endorses the latter: readers should believe wholeheartedly in the powers of
the god. That conclusion should not surprise, given the documents setting
at the gods chief shrine. On the other hand, it scarcely compels Posidip-
pus readers, who operate under quite different circumstances, to discard
whatever skepticism they may have had. Posidippus does not stage as a
guide for his audience any comparable encounter with divinity (nor even
provide them an authoritative voice or point of view), and so their experi-
ence must a fortiori remain open to both poles of interpretation set out on
the Epidaurian stele and to all gradations in between.
Consider the case of Soses of Cos in 97 AB.
, , /,
,
# 1 6 {} / 9 {} ,
, e .
[In payment to you for curing his sickness, Asclepius, Coan

Soses dedicates a silver libation bowl,
he whose six-year illness, together with the sacred disease,
divinity, you came and wiped away in a single night.]
Nothing in this epigram militates against our considering it a stock expres-

sion of popular piety. To be sure, a silver phiale is a particularly handsome
gift, but , ( :), thank offerings for cure, are attested in nu-
merous inscriptions at Epidaurus (cf. LSJ s.v. :). Similarly epilepsy,
the sacred disease, is well known in the Epidaurian stelae (IG IV2, 1 no.
123.115). Six years is a conventional duration for an illness (IG IV2, 1 no.
32. Trans. Edelstein.

121.95; Hippocr. Epid. 5.46; 98 AB). And the traditions of the gods healing
hand (which here wipes away the disease) and of his nocturnal appear-
ance to the pilgrim are likewise quite common, as Weinreich has shown in
detail.33
But Soses turns up again in a starkly different light after just ve
poems, in epigram 2 of the section entitled (103 AB).34
"/[{}] 6 _
, _ [ ]_ D
/ {} ? [ : ] , , / 65
/ [, C, ],
[You didnt even ask, for customs sake, what land Im from;
no, nor who I am, nor descended from whom. You just walk by.
Come on, [look at] me [lying] peaceably. Im the son
of Alkaios, Soses of Cos, [alive once, same] as you.]
Though he had appeared cured of both epilepsy and his unspecied six-
year illness in the third iamatikon (97 AB), poor Soses is envisioned here as
having suffered a grave setbackthe gravest: he is dead. Might the title
refer among other things to such sharp turns of fortune as we see
between these two poems (LSJ s.v. I)? In any case, it is hard, retrospec-
tively, not to nd humor in the earlier poem. Fordespite his pious grat-
itude beforeSoses has become a cranky old corpse. That silver phiale,
precious as it was, did not ensure happiness. Now Soses makes no allusion
to prior blessings, no reference to divinity at all. He just berates the
passerby for his breach of decorum in not inquiring about his identity.
The fourfold repetition of negatives in metrically emphatic positions
including verse start, bucolic diaeresis, and following the caesura of the
pentameteris a humorously over-the-top way of having Soses express
his indignation. One thing is certain: Nothing about the way he lies in
his tomb is peaceable (? [ . . . ] ). On the contrary he
makes darn sure that, willy-nilly, the passerby will hear his full prove-
nence, patronymic included (which had been omitted from the rst poem),
especially as he had so rudely failed to ask about it. In light of this unex-
33. Weinreich (1909: 145, 7679).

34. As Bastianini and Gallazzi note, the reconstruction of the second couplet is quite
speculative (cf. BG ad XV 3031 and following). I prefer their alternative supplement, C,
rather than C, ],
], , for v. 31, since it better suits the feisty speaking voice
of the rst couplet.
pected reversal in fortune, the rst poemand with it, Sosesappear

tinged with comic irony.35
An epigram that permits a comparably double reading is the sixth
iamatikon (100 AB), about the elderly blind man Zenon.
?/ ( Z \ a ,,
6/ , 6 ,
4 # / V +
[ ] 1 ,/{} /.
[When Zenon had to sleep that gentle sleep,
in blindness for the twenty-fth summer,
at age eighty he was cured. But glimpsing
the sun only twice, he beheld oppressive Hades.]
A pious reading (which can certainly be justied given the epigrams loca-
tion in the iamatika) might construe this poem as conventionally aretalog-
ical. When Zenon, elderly though he was and blind for a quarter century
already, sought divine help for his afiction through incubation (When
[he] had to sleep that gentle sleep, v. 1), the god gladdened his nal days
by miraculously restoring his sight. If, according to the proverb, one
should count no man happy until he dies, then surely (that pious reading
suggests) Zenon may be accounted such.
But the poem also allows a darker construction: on a purely formal level
it could just as well be an epitaph. One would not have been surprised to
nd it in book 7 of the Greek Anthology.36 Indeed, I believe one could as-
sign it to a well-known, presumably epideictic sepulchral type, the paradox
in death. Unique among the poems of the iamatika, this one contains no
mention of the healing god. That, of course, could be supplied from the
context within the section. Thus it was obvious to read the rst verse as re-
ferring to incubation. In itself, however, the phrase a , in the rst
verse could just as well refer to death (as at GV 455, 1874.7; Anthologia
Planudea 375: for death as a sleep, cf. LSJ s.v. a); that is, when Zenon
was going to die, his sight was suddenly restoreda bitter blessing, as it
turned out. For with its pointed nal words, ,/{} /, the epigram
35. Indeed, we may even nd an explicit link to the earlier poem if, instead of Bastianini
and Gallazzis in v. 1, we read , i.e., on account of [my unspecied]
sickness. The passerby is squeamish about stopping at the tomb due to that disease, which
proved fatal: the god, it appears, had not denitively cured Soses.
36. I owe this observation to Richard Thomas.
manifestly plays on the etymology of Hades as invisible, , the place

where nothing is seen (cf., e.g., Plato Crat. 403a).37 Rather than enjoy the
sightless sleep of death (a ,), Zenon now can see. But what does
he behold? Nothing, forevera grimly ironic, paradoxical demise.38
A further epigram that seems to invite double reading is that on Asklas
of Crete (99 AB).
C 65 / [ , ] /
, /
"1 / " / :/
8 / .
[Asklas the Cretan, deaf and unable to hear either

the [crash] of the surf or clatter of winds,
suddenly because of his vows for Asclepius went home
a man about to hear conversations even through brick walls.]
How we interpret this poem depends on a linguistic nicetya nuance of

aspectto which a hasty reader might turn a deaf ear. Until the nal line,
a pious interpretation seems perfectly appropriate. Asklas deafness is ab-
solute; it cuts him off from even the loudest sounds of nature. Upon visit-
ing the shrine, he makes vows to Asclepius, then heads back home. The
start of verse 3, "1, leaves us primed for a sudden, miraculous cure in
the manner of the Epidaurian inscriptions, and the beginning of verse 4
appears to conrm that expectation ( 8 /). But the
future participle , pointedly placed as the poems last word,
creates a space for irony. For it suggests that Asklas left the shrine as yet
unaware of what he had acquired there, a man about to hear conversations
even through brick walls. That ignorance sets the pilgrim in a comic
light. The verb of cognition moreover functions as a cue, inviting the
reader, retrospectively, to hear humor in other elements of the poem.
Asklas gained not simply the ability to hear but (as he presumably discov-
ered not long after) the superhuman capacity to overhear conversations
even through brick walls. No doubt is miraculous; but it is
also very funny, suggesting a range of domestic or public contexts in which
his newfound talent might be used. There is comic potential, too, in his
37. Indeed, in an active sense, the adjective can also mean blind (cf. LSJ s.v.).
38. For a comparable paradox in death, cf., e.g., Dioskorides 33 GP AP 7.76, where
a mariner abandons seafaring for farming, only to be overtaken after death by the ooding
Nile, which consigns him to the watery grave of a shipwrecked man ( , v. 6).
ethnicity: Cretans famously prized reticence, so the prospect of indiscrim-

inately hearing everyones conversation might seem more torment than
blessing. Bastianini and Gallazzi may have been right when they com-
mented that the poet describes this cure con una sfumatura di sorriso
(BG ad loc.).
Of course, it is important to recall that humor is not necessarily at odds
with a pious reading. Greek religion embraces the comic in ways that
startle modern sensibilities schooled in the Judeo-Christian tradition. The
tales in the Epidaurian inscriptions, too, are at times distinctlyand no
doubt deliberatelyfunny.39 Thus humor alone should not sufce to make
Posidippus readers incredulousunless that incredulity has the semantic
range it has in the narrative of the skeptical visitor in the Epidaurian stele
(previously cited), where the term could signify distinct things under dif-
ferent circumstances and where the name Incredulous in fact bespoke
credence.
I want to close with a look at the last of the iamatika (101 AB), an epi-
gram once again susceptible to double reading but that perhaps illuminates
a different point along the interpretive axis previously traced.
, ,
= [ ], /
/ 4
, / #D D ,
V # .
[The noblest man, Asclepius, asks for moderate wealth

great is your power to bestow it when you wish
and he asks for health: remedies both. For these appear to be
a towering citadel for human conduct.]
Coming after the particular instances of divine cures in the previous

poems, this epigram appears to conrm their value by generalizing the im-
portance of health.40 With its idealized subjectthe indenite
[ and its impersonal, metaphor-rich summation ( D
39. Weinrich (1909: 8990) makes the following comment about several of the Epidau-
rian miracles, e.g., a cure for baldness (IG IV2, 1 no. 121.122) and a cure for lice, in which the
god sweeps away the vermin with a broom (IG IV2, 1 no. 122.45): Bei diesen Wundern mge
man bedenken, dass die Aretalogie nicht nur erbauen, sondern auch unterhalten will. Des-
halb wird Humor und Komik nicht verschmht.
40. BG (ad loc.) cite numerous instances of the view that health (often coupled with
wealth) is the most important of goods, such as PMG 890, the paean of Ariphron PMG 813,
Simonides PMG 604, and Pindar Ol. 5.23.
, / V # , vv. 34), the poem functions

as a gnomic conclusion drawn from the aforegoing tales.
We have seen, however, that the outcome of divine therapy is not al-
ways expected or, indeed, happy. For Soses of Cos, to be healed of his
epilepsy was doubtless a blessing (97 AB), but his epitaph a mere twenty
versesve poemslater in the (103 AB) exposes its ultimate futil-
ity. Miraculous cures have their limits, for as that poems fragmentary nal
words seem to suggest, Soses is only mortal (same] as you, C, ],
). That is one condition the god cannot cure.41 Similarly, for elderly
Zenon (100 AB), it appeared that the gift of sight arrived with unforeseen
consequences. Take care, the poem seems to suggest, what you beg from
the gods: they may grant it. The same could apply to the deaf man Asklas
of Crete (99 AB), who leaves the shrine with more than he bargained for,
the potentially disagreeable ability to hear even through brick walls.
At issue here may be less skepticism or belief than how one thinks about
wondrous cures. One potential response to the iamatika might be to sup-
pose that they do not so much tempt one to disbelieve in the possibility of
miracles as make one question their efcacy in creating human happiness,
in fullling ones desires: for miracles sometimes prove to be either incon-
venient or useless for humans locked in the condition of mortality. As such,
the cures represented in these poems become yet one more example of a far
broader theme, to wit, the problematic nature of divine-human interaction.
In this light, the nal epigram invites a different, less staunchly afrma-
tive reading. The poem emphasizes moderation. The key adjective in this
respect, , in verse 1, is accentuated through its placement following
the bucolic diaeresis, at the other end of the line from its noun. Exemplify-
ing a human standard of measure, = . . . contrasts in an essen-
tial way with the divinitys expansive . . . of verse 2. Similarly,
note the ontological opposition inherent in the juxtaposition
[ ], /
(v. 1). In light of this emphasis, one may plausibly take
not just as modifying = in verse 1 but as extending to #
in verse 3.42 This possibility is all the more appealing given how closely the
poem binds together = and #, classing them into the single cate-
gory and subsuming them into the unitary image .43 On this
interpretation, the noblest person requests not only moderate wealth but
41. Asclepius notoriously tried to raise a mortal from the dead but was struck by Zeus
lightning for his transgression; cf. Pindar Pyth. 3.5558.
42. For two-termination, cf. LSJ s.v. Or could this be simply an anacoluthon?
43. In this way, the epigram actualizes the traditional concept of , which BG
trace in their commentary (ad vv. 1921).
moderate health. That is, he does not rely on the prospect of a divine
cure, whicheven if he were so fortunate as to receive onemight lead to
unforeseen and untoward consequences. Rather, he prays for what human
methods can achieve, a more modest general tness of body and mind that
is a foundation for proper conduct. Of course, the only humanly wrought
cure in the iamatika was that of Medeios, son of Lampon, in the rst poem
(discussed earlier in the present essay). It now appears that he was indeed an
, one of the foremost of the Ptolemaic (Bing 2002a).
Perhaps it is he that is meant here and implicitly exhorted to pray.44
I use the word pray advisedly here, for although this epigram is
framed as a maxim, the invocation of Asclepius in verse 1 and parentheti-
cal address to the god in verse 2 suggest that the poem is in fact an indirect
prayer. Perhaps that indirection is meant to characterize the tact and mod-
eration not just of the noble doctor Medeios, son of Lampon, but of the
speaking voice itself. It tells us what an should do but leaves
unspoken the implication that such a pronouncement itself bespeaks an
here obliquely requesting moderate wealth and health on
its own behalf. Discreet yet authoritative, this anonymous voice may plau-
sibly be identied as the poets. Ending the section, then, with a tradi-
tional form of poetic closure, a prayer, the poet pleads for something
other than a miracle and more moderatecounsel that may be compatible
with the Stoic orientation that some scholars have found in Posidippus
(Gutzwiller 1998: 15762). At the close of the iamatika, its readers must
decide if that plea retrospectively colors their response to the wonders
they encountered before.
44. If that is correct, then he, given the wondrous nature of his cure for snakebite, may
have been less ready to pray for only moderate health, i.e., to take as modifying
# as well as =for that reading subtly undercuts his achievement. Nothing com-
pelled him to read it thus, however: # can just as well stand alone.
c h a p t e r 12
Posidippus and the Admiral

KALLIKRATES OF SAMOS IN THE EPIGRAMS OF THE
MILAN POSIDIPPUS PAPYRUS (P. MIL. VOGL. VIII 309)
T he new epigrams of Posidippus, published as P. Mil. Vogl. VIII 309,

cast a sudden dazzling light on an array of important topics in Hel-
lenistic studies, ranging from Ptolemaic patronage of the arts to the early
form of the poetry book. Not least among the scrolls attractions are previ-
ously unknown poems about Kallikrates of Samos and his famous founda-
tion, the shrine of Arsinoe-Aphrodite Zephyritis at Cape Zephyrion near
Alexandria. These provide fresh insight into the interests of this promi-
nent Ptolemaic courtier and so oblige us to consider anew some aspects of
his career and objectives. That is what I propose to do in this essay. In light
of both new evidence and old, Kallikrates will emerge as a gure who pro-
moted a consistent agenda in his actions on behalf of his sovereigns,
Ptolemy II and Arsinoe Philadelphus. An exponent of that intercultural
poetics most recently described by Susan Stephens (2003), he sought to
mediate between old Hellas and the sometimes strange new world of
Ptolemaic Egypt, bridging the gap between the two, whether by bringing
Greek tradition to bear on his Egyptian milieu or by spreading abroad his
rulers novel cultural policies.
To start, let us review what is known of his life: Kallikrates of Samos,
This is an updated version of Posidippus and the Admiral: Kallikrates of Samos in the Milan
Epigrams, pp. 24366 2003 GRBS.
234
Posidippus and the Admiral 235
the son of Boiskos, was a man of power and inuence.1 Supreme com-
mander of the Ptolemaic navy, or nauarch, for some twenty years from the
270s into the 250s b.c.,2 he belonged to the inner circle of the court and
was described by Philadelphus himself as among his philoi (Welles 1934,
no. 14.9). His achievements and faithful devotion to the crown were such
that Ptolemy chose him to be the rst eponymous priest of the dynastic
cult of Alexander and the Theoi Adelphoi in 272/1 (P. Hibeh II 199.ii.12), a
signal honor. At Olympia, Kallikrates made a lavish dedication to Zeus
Olympios in honor of his king and queen, setting up statues of each atop a
pair of ionic columns ten meters high (discussed later in the present essay).
A new detail furnished by the Milan Posidippus is the information that
Kallikrates was active also at another Panhellenic shrine, at Delphi, where
(as we shall see) his colts won the chariot race. In consequence, Kallikrates
made a grand statuary dedication to the Theoi Adelphoi (74 AB). Further
dedications by him are recorded in Samos (a statue of one Tinnis, daugh-
ter of Dionysodoros, apparently to Hera, IG XII.6, 446) and possibly at
Kourion in Cyprus (a stele to Apollo, Mitford 1971: 11718 no. 58).
In Egypt, he dedicated a sanctuary of Isis and Anubis on behalf of
Ptolemy and Arsinoe at Canopus.3 Most famously, at some point shortly
before or after Arsinoes death, he founded the shrine of Arsinoe-Aphrodite
Zephyritis at Cape Zephyrion between Alexandria and Canopus.4 This
shrine is unique in being the only edice of the third century b.c. com-
memorated in multiple epigrams by various leading Hellenistic poets.5 As
we will see, the Milan Posidippus adds at least one more poem to the
rosterI believe, however, that it in fact adds several. Both old and new
epigrams show clearly that through this coastal shrine, Kallikrates sought
1. In the following, I rely on the compilation of sources in Mooren 1975: 5860 no. 010.
2. Cf. Hauben 1970: 69. Kallikrates appears as during Arsinoes lifetime (i.e.,
before 268) in a Samian dedication (OGIS I 29) and in his own foundation, on behalf of
Ptolemy and Arsinoe, of a temple of Isis and Anubis at Canopus (SB I 429). In about 257 b.c.,
he was probably still nauarch, as he had his agent Zoilos write to Apollonios, Philadelphus
chief nancial ofcer, to collect a tax () for the upkeep of the navy (P. Mich. I 100).
3. SB I 429. The attention to Egyptian gods may also be reected in the offering of a
rhyton in the form of Bes at the shrine of Aphrodite-Zephyritis, described in an epigram by
Hedylus (4 GP), discussed later in the present essay.
4. I am persuaded by the dating of Arsinoes death to 268 proposed by Grzybek (1990:
10312), Hazzard (1987; 2000: 3), etc., rather than the previously held 270. See the discus-
sion of their ndings in Hauben 1992: esp. 160ff.
5. The previously known epigrams are Posidippus 116 AB 12 GP P. Firmin-Didot,
and 119 AB 13 GP Athen. 7.318d; Callim. Ep. 14 GP 5 Pf. Athen. 7.318b; Hedylus
4 GP Athen. 11.497d. Among the new poems of the Milan papyrus, cf. 39 AB and possibly
36 and 37, all of which I discuss later in the present essay. Cf. Hesberg 1981: 63.
to identify the queen with the maritime Aphrodite, who held particular sig-
nicance for sailors. As Louis Robert stressed, we see a link between the
character of the cult and the duties of the nauarch. Through this founda-
tion, the queen became patron of the eet and of the Ptolemaic maritime
empire.6
Finally, Kallikrates was himself the object of numerous honors. At
Olous on Crete, he was acclaimed as benefactor and granted proxeny (I.
Cret. I xxii 4.A.3538). He was honored with statues by the nesiotai in
Delos (Durrbach, Choix 25.12) and at Palai-Paphos and Kourion in
Cyprus (Mitford 1961: 9 no. 18), as well as together with Ptolemy and Ar-
sinoeand on a strikingly equal footingin his native Samos (OGIS I
29.34, with II p. 539).
In all, our evidence depicts a man whobefitting the mobility im-
plicit in the office of nauarchglides easily between the old Greek world
and the new and is at home in both. On the one hand, we find him linked
with venerable Hellenic shrines, making offerings to traditional Greek
deities, such as Zeus at Olympia, Hera at Samos, Apollo on Cyprus; yet
at the same time, he serves in the very untraditionally Greek role of
priest of his divinized, yet living, sovereigns in their dynastic cult. Again,
on the one hand, Kallikrates aimed his newly founded cult of Arsinoe-
Aphrodite Zephyritis at a mostly Greek constituency, for according to
the testimony of an epigram by Posidippus (116 AB 12 GP), it is Greek
women and men who are its beneficiariesthe poem addresses them:
/A/ 6 Z / / ; Z
, / G / Z 6 , but come to that
Aphrodite who will be called Zephyritis, chaste daughters of the Greeks
and men who work the sea; yet at the same time he founded a cult of the
Egyptian deities Isis and Anubis at Canopus, whose priest is recorded as
being the Egyptian Pasis.7
The important point is not that Kallikrates simply moved between
these worlds but that he actively sought to bind them together, establish-
ing links from one to the other, integrating his adoptive homeland into the
cultural fabric of old Greece. I believe that the new epigrams help illus-
trate this point. The rst I will examine commemorates Kallikrates vic-
tory at the Pythian Games in the chariot race for colts (74 AB).
6. Robert 1966: 2012.

7. SB I 429. Following Fraser (1967: 40), Hauben (1970: 40 with n. 7) believes that Pasis
usurped the proprietary rights to this temple at a later time.
6 ? @/

, 1 6
V , /{}, 6 D
8 + , E 5
? , D
\ + [/ ]
6[ ] " 8 6[,
? / D / 6[
[] [ 10
[] D 6 [ +
[] A/ [,
/ / []{} ,5 6 []
9[ ?] E / (.
[At Delphi this lly, when vying with the four-horse chariots,
nimbly ran a dead heat against a Thessalian rig
and won by a nose. Great then was the uproar of the drivers,
Phoebus, in the presence of the amphictyonic judges.
But they at once threw their rods onto the ground, so that by lot
the charioteers should carry off the crown of victory.
But she, the right-hand trace-horse, bent down in the pureness
of her heart and lifted up the rod herself,
a wondrous female among the males. The mingled throng
then clamored with a single voice
for the judges to proclaim the great crown hers. And amid noisy
applause
the Samian man Kallikrates carried off the laurel.
But here he set as vivid image of those former contests
a bronze chariot and charioteer for the Theoi Adelphoi.]
With a vividness no doubt meant to match the ,5 6 dedicated

by Kallikrates (vv. 1314), the poem describes the high drama of the nish
line at Delphi following this race. In bold strokes, it evokes the tumult of
the crowdthe 1 . . . 6 (v. 3), the shout of the mingled
throng (vv. 910), its noisy applause (6 [, v. 11). In between,
there is the powerful gesture of the judges in inging down their rods,
signaling (as was the norm in case of a dead heat) that the victory was to
be decided by lot and so considered sacred (hiera), that is, left up to the
god.8 Yet for all its dramatic immediacy, the epigram leaves no doubt that
it is at a remove from the action. As Bastianini and Gallazzi recognized
(BG ad XII 7), the opening words, 6 , show clearly that the ded-
ication was not at Delphi, since a reference such as this would be super-
uous in that setting. Further, the poem implies a temporal disjunction
between the dedication of the statue and the victory in those contests
back then ( / [], v. 13). To be sure, may in part be
intended to speak to imagined viewers of future generations,9 but that
does not prevent it from referring in the rst instance to events already at
a distance at the time of dedication. All this raises the question of how to
interpret the demonstrative E / of verse 14. To what location does it
refer? Where, if not at Delphi, did Kallikrates set up his elaborate bronze
statuary group?
To commemorate an agonistic triumph, victors commonly dedicated
statues in their native cities.10 We may wonder, therefore, whether the des-
ignation of Kallikrates as [ in verse 12 hints at a location in
Samos. That, however, is unlikely, sinceas Hauben points out in another
contextit would be peculiar that a Samian would mention his ethnic in
his own native city.11 Not Delphi, not Samos: where then?
While we cannot be certain, I believe that the evidence points toward
Egypt. That is the inference I draw from the striking fact that the hono-
rand of the dedication is not Kallikrates (though it doubtless does him
honor, too) or even the god of Delphithough he is the poems addressee,
with /{} (v. 4)but, rather, the Theoi Adelphoi. What have they to do
with this victory?
An answer may lie in how that victory is presented. I already noted how
the epigram stressed the thronging multitudes, their noise, and the judges
casting away their rods (8 . . . , v. 5). Juxtaposed to
these in meaningful contrast is the silent solitary gure of the female colt,
who, answering the judges gesture, bent down to draw to herselfand so
claimthe rod ( [/ . . . / . . . 8 6[, vv. 78).12
Signicantly, the colts deed derives ] / 6[ ] (vv. 78). It
8. Cf. Moretti 1953: 201.

9. Ebert 1972: ad no. 38.4. Cf. ibid., ad no. 26.1/2: Die historisch auswertende
Interpretation von H.T. Wade-Gery ( JHS 53, 1933, 71ff.), der vielfach aus schliet,
dass jeweils zwischen dem im Epigramm angezeigten Ereignis und der Abfassung des Epi-
gramms geraume Zeit verossen sei, erscheint mir sehr fragwrdig.
10. Cf. Ebert 1972: 1112.
11. Hauben 1970: 37 n. 4, referring to OGIS I 29.
12. Both senses, draw to oneself and claim for oneself, are clearly in play here. Cf.
LSJ s.v. 6 III 2 and 5.
is because she acts out of the purity of her heart that she is able to resolve
the matter of the lots: for lots, by their very nature, involve divinity,
nowhere more so than at Delphi, where we know there were lot oracles.13
To gain the favor of the Delphic god (the poems addressee), the competi-
tor in the race must be a match for him in purity, a quality here evoked in
Apollos byname (v. 4).14 Where human judgment fails, the gods
will is decisive. Here it is expressed through the action of this untainted
creature, which constitutes a sign whose validity is at once acknowledged
by all.15
Up until this point, throughout this drama of decision, Kallikrates has
remained discretely in the background, the text reserving any mention of
him until near the end (v. 12).16 Indeed, the poet has been at pains to take
the outcome out of the hands of Kallikrates and place it instead in the
realm of the divine. Now, when he enters, it is as a modest [ (v.
12), without titles or patronymic. He appears, rather, in a role analogous to
that of his colt (or rather, in retrospect, it appears that the colt was a stand-
in for Kallikrates). He is likewise a solitary silent gure in the crowd (6
[, v. 11). And as the colt drew up to herself the rod (8
6[, v. 8), so he bore aloft the laurel ( A /, v. 12). The
13. Cf. Burkert 1985: 116.

14. Certainly the Hellenistic poets heard that meaning in the name. Cf. Cuypers com-
ment on the use of the verb in connection with the seer Phineus at Ap. Rh. 2.302:
. . . / J , having cleansed (the old mans skin) all aroundbut also
Phoebused, appropriate treatment for a seer . . . There may be some truth lurking behind
our scholiasts " , D , @
C /, (1997: ad loc.). For this verb in a similar sense elsewhere
in Hellenistic poetry, cf. Callim Hymn 5.11; Theocr. 17.134; Lyc. 6, 1166.
15. Bingen (2002) argues, differently, that in the phrase 8 +
(v. 5) does not mean that the judges threw their rods onto the ground at
once, in no time ( ), as the original editors and Austin suppose, but that it means
too few: Trop peu de ceux-ci [sc., the judges] jetrent terre leur bton pour que ce ft par
tirage au sort que les cochers emportent la couronne de la victoire (186). Though Bingens
interpretation of is syntactically and lexically possible (cf. 4 too few at
Thuc. 1.50.5 and Hdt. 6.109), its meaning in the context is implausible. For on his view, the
dramatic core of the poem embodied in the extraordinary act of the colt, which is taken as a
sign by all present, is in fact superuous: a sufcient number of judges was sure enough of the
outcome that no decision by lot was actually necessary. Once the tumult died downwe infer
from Bingens argumentthe ordinary human decision-making process would have arrived
at the right conclusion. This, however, is to make the sign, together with the poets implica-
tion that the god was responsible for it, quite pointless.
Bingens point may in any case be moot if we accept Jankos plausible correction (2005) of
to , i.e., judges at the games.
16. Ebert (1972: 79 ad 19.2), notes the in hellenistischer Zeit aufkommenden Sitte, ger-
ade in lngeren Gedichten, wohl um eine gewisse Spannung zu erzeugen, die Namensnen-
nung hinauszuzgern.
actions of colt and master are clearly analogous and imply that the gods
favor extended to Kallikrates. Perhaps in making his dedication to the
Theoi Adelphoi, his patron deities and real-life benefactors, Kallikrates
sought to acknowledge that, in a larger sense that reached beyond the
present victory, he owed his success to higher powers. The message is,
Not mine, the glory, but theirs.17
Bastianini and Gallazzi argued that the terminus post quem of our poem
must be the year 270. They noted that the dynastic cult of the Theoi
Adelphoi was founded in 272/1 (i.e., in the fourteenth year of Philadelphus
reign as reckoned from his coregency with Soter in 285/4; cf. P. Hibeh II
199) and that the rst Pythian Games thereafter fell in the year 270.
Those, they argued, must be the rst for which our poem could have been
composed.18 Given, however, that by characterizing the games as /
[] (v. 13) the epigram may point to a certain lag between
Kallikrates victory and his dedication, I would propose another possible
date. The Pythian Games of 274 fell in August/September of that year (cf.
BG ad XII 6), while the year in which the cult of the Theoi Adelphoi rst ap-
pears in operation began either in March or May of 272.19 No doubt the
decision to create the cult and to name Kallikrates its rst ofcial went
back to the previous year. Thus Kallikrates eponymous priesthood com-
menced less than two years after those games, when their memory might
still be fresh. In this light, I would suggest that the admirals striking deci-
sion to commemorate his victory with a dedication to the Theoi Adelphoi
may have been related to these circumstances: At no other time would he
have had so strong a motive or have made so potent an impression as dur-
ing that year when he was rst eponymous priest of this newly established
cult. Plausibly, then, we may ascribe Kallikrates offering for the Theoi
Adelphoi to the year 272/1.
And where did he make it? What setting should we contemplate for the
poems nal words, E / (? Fraser stressed that unlike casual ruler-
worship [the Ptolemaic dynastic cult] was centered on the capital of the
kingdom, and cult-centres elsewhere were subordinate to this central ad-
ministration.20 Alexandria thus seems an obvious choice. But wherever
17. Fantuzzi and Hunter (2004: 39293) aptly compare the epigram from Cyrene (SEG
17.817) in which Eupolemus dedicates Victory, to King Magas as his rightful (392).
In Posidippus epigram Callicrates may well have wanted to make a similar point: victory
both depended upon and belonged to the Ptolemaic house. (393).
18. BG ad XII 6: 270 dovrebbe dunque essere il terminus a quo per il fatto narrato e per
la composizione dell epigramma stesso.
19. It ran either from 15 March 272 to 4 March 271 or from 17 May 272 to 5 May 271.
Cf. Hauben 1992: 161.
20. Fraser 1972: I 214.
Kallikrates made his offering, clearly its goal was to link the new Ptole-
maic cult to the luster of Apollo and the prestige of a victory at his Pan-
hellenic shrine at Delphi. Not that this goal was unique or even novel:
Greek city-states had long recognized the benets that befell them when
their citizens brought victory crowns back home from the games. But the
hunger for prestige and the legitimation offered by such victories was
particularly urgent among the ambitious Ptolemies, with their recently
established kingdom and new-built capital. In his Victory of Sosibios (fr. 384
Pf. with II pp. 121, 125), Callimachus dramatizes this situation so as to
highlight the somewhat embarrassing disparity between the kingdoms
grand presumptions and the little it at rst had to show for them: the Nile
itself declares that until Sosibios, no one had brought a trophy back to
the city from these sepulchral festivals [i.e., the Panhellenic Games] and,
great though I am, I, whose sources no mortal man knows, in this one
thing alone was more insignicant than those streams that the white
ankles of women cross without difculty and that children pass over on
foot without wetting their knees ("] 6[] A/
/ ] / ] , T "/ @ ,
C / , 2 M
/ ( / [] W
/ [ ] M (, vv. 29
34). In a series of epigrams celebrating equestrian victories, the Milan pa-
pyrus makes clear that the Ptolemies pursuit of such prestige was far more
determined than we had hitherto imagined.21 With his conspicuous vic-
tory at Pytho, Kallikrates doubtless helped his sovereigns attain their goal.
The same holds true of his dedication to Zeus Olympios in honor of
Ptolemy and Arsinoe at Olympia. It is worth recalling this remarkable
monument here, for it helps us visualize on just how grand a scale and with
what audacious and meaningful design Kallikrates sought to glorify his
sibling lords.22 Here, too, he made sure to present them in terms evoking
divinity, specically as a couple, like Zeus and Hera. At the same time, it is
striking again, by contrast, how modestly he presented himselfjust as in
21. Cf. particularly Fantuzzi 2004a, 2005.

22. The dedications appear to have no connection with specic victories, as the inscrip-
tions make no mention thereof. However, the new epigrams reveal that both Ptolemy and
Arsinoe won such victories. The 126th and 127th Olympiads (276 and 272 b.c.) are the only
festivals overlapping Ptolemys marriage to Arsinoe II, which occurred sometime after 279
(cf. Hauben 1970: 35). It was evidently at one of these that Arsinoe won the triple Olympic
victory celebrated by Posidippus in 78.78 AB. It was possibly at the same festival that
Philadelphus won the Olympic chariot victory mentioned in the same poem (78.57 AB),
though there can be no certainty about this. Another epigram of Posidippus (88 AB) deals
with a chariot victory of Ptolemy II, which he won while coregent with Ptolemy I (i.e., before
283, prior to his marriage to Arsinoe).
the epigram for his Delphic victory.23 The dedication consists of a monu-
mental pedestal 20 m. long by 4 m. wide by 1.12 m. high in the middle of
which was an exedra 2.35 m. long by 1.68 m. wide, articulated with a bench.
Bracketing the pedestal on either end, large stylobate blocks supported tall
(8.93 m.) Ionic columns,24 each crowned by a bronze statue resting on a
statue base atop the capital. As the texts symmetrically inscribed in four
lines on the plinths at the base of each column reveal, one statue was of
King Ptolemy Philadelphus, the other of his queen Arsinoe (OGIS I 2627).
Hoepfner (1971: 4549) persuasively described how the columns with
their statues dominate the eastern end of the broad plaza in front of and
facing the great temples of Zeus and Hera. Indeed, as his site plan indi-
cates, they are carefully aligned so as to stand in relationshipat precisely
the same angleto the end columns of the temples of Hera and Zeus.
Thus Ptolemy and Arsinoe as king and queen, but also as brother and sis-
ter, become the visual counterparts to the divine couple, the brother-sister
sovereigns Zeus and Hera. Theocritus made a similar analogy in his Idyll
17 (vv. 13134).25 Further, the pair of columns serve quite literally to exalt
the royal couple above the plane of the viewer, strongly suggesting their
divinity. Philadelphus in fact established the cult of the Theoi Adelphoi
within his own and his sisters lifetime, making them gods on earth. The
divinization of the brother-sister pair was a momentous step that Ptole-
maic apologists sought to make intelligible and to legitimize through
Greek precedent.26 The monument does just that. Finally, Hoepfner sug-
gests that Kallikrates dedication injects a subtly Egyptianizing element
into this most Greek milieu. On his view (1971: 4748), the highly unusual
double-columned monument deliberately evokes pharaonic tradition, re-
calling the paired obelisks with which Egyptian rulers marked important
occasions. It was at about this time, he notes, that Ptolemy rst brought an
obelisk to Alexandria to be set in honor of his sister-wife before her shrine,
the Arsinoeion (cf. Pliny NH 36.68, 34.148). Certainly the inclusion of
such an Egyptianizing component is consistent with Kallikrates attempts
to mediate between the new world and the old.
But even as the monument exalts its Ptolemaic sovereigns to the skies,
drawing the gaze of the viewer heavenward, it also communicates the rela-
23. For the dimensions and reconstruction, I rely on Hintzen-Bohlen (1992: 7778), fol-
lowing Hoepfner (1971).
24. The stylobate blocks are of white marble like the columns and thus visually form a
unity with these, distinct from the pedestal, which is built of limestone. Cf. Hoepfner 1971: 15.
25. On this analogy in Theocritus and elsewhere, see Hunter 2003b: 19293 ad 13132.
26. On this topic, cf., most recently, Huss 2001: 3079, 325.
tive humility of its sponsor, Kallikrates of Samosjust as in the epigram

celebrating his dedication to the Theoi Adelphoi on the occasion of his vic-
tory at Pytho. For the viewer encounters Kallikrates name at eye level
(Hoepfner 1971: 46), inscribed in lines of text each a mere three centime-
ters high and cut just two to three millimeters deep into the stone (ibid.,
15)a strikingly unobtrusive, even self-effacing declaration of authorship,
when set against the imposing size of the monument as a whole.
I turn now to some of the new epigrams concerning the shrine of Arsi-
noe-Aphrodite Zephyritis. Their appearance in the Milan papyrus adds new
force to the old observation that Meleager was not interested in collecting
poems about Ptolemaic monuments or occasions. For none of the poems on
this shrine (including those by Callimachus and Hedylus) are transmitted in
the Palatine or Planudean anthologies, nor is the epigram by Posidippus
about Sostratus of Cnidus and the Pharos.27 I start with the new poem that
is closest to the old Posidippan examples about this shrine (39 AB).
9 q
, " /,
] 6 , ! C y
,
, D / _ + 5
D
G , 9
" # 6.
[Whether you plan to cross the sea by ship or make fast the stern-cable
from the shore, say hail to Arsinoe of Fair Sailing,
and call the goddess-queen from her temple, which Boiskos son,
the nauarch, Samian Kallikrates dedicated
especially for you, sailor. In pursuit of fair sailing
often another man has called upon this goddess.
Therefore whether youre on dry land or casting off for sea,
you will nd her attentive to your prayers.]
We recall that in one of the previously known epigrams by Posidippus

about this cult (12 GP 116 AB), the shrine itself had taken voice to call
27. On the latter poem, see my treatment in Between Literature and the Monuments
in the present volume. Similarly absent is Posidippus scoptic epigram (143 AB SH 702, re-
ferred to in Athen. 10.415 ab and Ael. VH 1.26) about Aglais, daughter of Megakles, who
blew the trumpet in the rst great Ptolemaiia.
upon two sorts of audience: But come to that Aphrodite who shall be
called Zephyritis, you chaste daughters of the Greeks, and men who work
the sea (vv. 79). The new poem addresses only the second of these groups.
It tells the sailor that Kallikrates dedicated the temple especially for him
(, , v. 5) and that he should greet its goddess, Arsi-
noe, as Euploia (She of Fair Sailing). With this cult title, the poem makes
plain what was previously implicit in the second of the earlier known Po-
sidippan poems on this shrine (13 GP 119 AB), where we heard that
Arsinoe-Aphrodite would grant fair sailing, "y (v. 5). The new
poem thus conrms Louis Roberts suspicion that Aphrodite was wor-
shiped at Zephyrion as Euploia, a title with important associations.
Through it, Kallikrates forged a link between Arsinoe and such celebrated
old-world exemplars of Aphrodite Euploia as the Knidian Aphrodite of
Praxiteles28 and Konons temple in the Piraeus (Paus. 1.1.3). Indeed,
Robert plausibly argued that the many humble private dedications to Arsi-
noe Philadelphus, which cluster particularly in the far-ung port cities of
the Ptolemaic empire and also the numerous ports renamed for Arsinoe,
reect Kallikrates activities as ardent propagandist for this cult,29 pro-
moting its spread far beyond the connes of Egypt. In other words, the
links Kallikrates forged went in both directions: from old Hellas to Egypt
and from Egypt back to old Hellas.
A nal observation about this poem: the assimilation to Aphrodite is
here so complete that, unlike in the previously known poems, Aphrodite is
not even mentioned. Arsinoe is invoked as theos entirely on her own
(] 6 , v. 3;
, 6). Scholars
have argued, on the basis of Posidippus description of Arsinoe as -
in epigram 12.5 GP ( 116.5 AB), that Kallikrates founded the cult
at Zephyrion while Arsinoe was still the reigning queen, since that designa-
tion appears in inscriptions that presuppose Arsinoe alive.30 Be that as it
may, the new poem addresses her as (v. 3), a comparable term, but
one customarily applied to divinities. By invoking her in this way and by
calling her theos without mention of Aphrodite, the new poem may suggest
that we are dealing with the sovereign as she was divinized after her death.31
28. Posidippus himself appears linked to both the Knidian and Zephyrian cults of Eu-
ploia, since he is said to have written about Praxiteles statue in a work entitled Peri Knidou
(147 AB SH 706).
29. Robert 1966: 2012, 208.
30. For the pros and cons, cf. Hauben 1970: 4445.
31. Probably in 268 (cf. n. 4 in the present essay). The rst Kanephoros of Arsinoe
Philadelphus is recorded in 268/7, i.e., following her demise, and thus speaks for the establish-
ment of a separate cult for the divinized queen shortly after her death. Cf. Hauben 1992: 161.
That also seems to be the case in the next poem I want to deal with
(36 AB).
/,

/ ,
E , , / = 4 1

A 4 D
u 6, , 6 , , 5
, 6 ( .
? + ,
;5 [.
[Arsinoe, for you this gift is dedicated, to be swept by the wind

across your garments folds, a linen cloth from Naukratis,
with which in a dream, dear one, you wished to wipe your dulcet sweat
when youd ceased from your busy labors.
Thus you appeared, O dear to your brother, in your hand the pointed
spear,
queen, and on your arm a hollow shield.
At your request, this white strip of cloth
the maiden Hegeso, of Macedonian heritage, gave in offering.]
What has this poem to do with Kallikrates or the cult he established at

Cape Zephyrion? The poem records a dedication for the divinized Arsinoe
Philadelphus, with no apparent reference to Aphrodite, her cult title, or to
seafaring, such as we see in the other poems. Yet there may be grounds for
making a connection. First of all, the mention of Naukratis as the source of
the . . . (v. 2) evokes an Egyptian milieu. That squares with
Hegesos proud insistence on her Macedonian roots: in a largely alien envi-
ronment, Greeks might appeal to their Greekness with particular empha-
sis.32 The affectionate bond evoked when Hegeso addresses Arsinoe as
(v. 3) suggests a relationship between the goddess and her maiden-
worshiper, dened by more than just her dream. Moreover, the goddess
title Phil-adelphos, indicating she is both dear to and loving her
brother, necessarily links her function to love and marriage. We note that in
one of the previously known Posidippus epigrams (13 GP 119 AB), it is
32. See the similar insistence on the part of Apollonios, one of the Macedonian brothers
connected with the Memphite Serapaeum in the mid-second century b.c., as described by
Thompson (1988: 261).
expressly as Arsinoe Philadelphus (i.e., Arsinoe Who Loves Her Brother)

that the queen is equated with Kypris (vv. 12). Most suggestive, however,
is the specication that the offering is to be swept by the wind / across
your garments folds ( , v. 1). Does this detail
serve any purpose in the poem beyond the merely ornamental? So far as I
can see, of the numerous literary instances we have of women dedicating
textilesfrom the Trojan women of Iliad 6 offering the peplos to Athena
in her temple (6.9092, 27173, 28895), to the bride presenting her veil
to Hera in an epigram by Archilochus (AP 6.133 FGE p. 14748), to
Nossis and her mother dedicating a ne linen garment to Hera Lakinia at
her shrine near Croton (3 GP AP 6.265)none of them pictures the
dedication billowing in the wind.33 I would suggest that
hints at Arsinoes identity as Aphrodite Zephyritis.34 Indeed, the celebrity
of this cult was such that perhaps the mere mention of wind in conjunc-
tion with Arsinoe could trigger the association.35 If this is so, then Hegeso,
the Macedonian maiden, represents that other constituency to which Po-
sidippus addressed himself in the earlier-known poem I cited before (12.9
GP 116 AB). She is one of those chaste daughters of the Greeks
(; Z . . . ) whotogether with the men who work
the seawere urged to come to the temple of Arsinoe-Aphrodite. And
while sailors came to the shrine to pray for fair sailing, the young women
came in the hope of a smooth voyage on the sea of love and marriage.36 Is
that what Hegeso desires?
On the face of it, the idea does not look promising: Arsinoe appeared to
the maiden equipped with spear and shield and sweating from her labors.
Bastianini and Gallazzi note a dream recounted in Plutarch's Life of Lucul-
lus 10 in which Athena sweats due to her efforts on behalf of Cyzicus in the
Mithridatic War, and they wonder whether we should assume a compara-
ble martial context here in connection with that same goddess. Specically,
33. Cf. Rouse 1902: 27477.

34. Note that the windswept garment accords with the iconography of Aphrodite gener-
ally; cf. LIMC II 93650, 985.
35. Recall the that her temple is said to occupy in Posidippus 12.3 GP
116.3 AB.
36. This point is convincingly demonstrated by Gutzwiller (1992b: esp. 198202) with
reference to Callimachus epigram on the offering of the nautilus shell at the shrine of
Aphrodite Zephyritis. Thus, in the closing words of Posidippus 12 GP ( 116 AB),
" (v. 10), the qualication refers to a haven not just from every wave
in a literal sense but also from every erotic wave. Cf. Theogn. 45960 (of the ):
+ / 6 / V
( ; 127374: 6 +
/ 6 6.
they weigh the Chremonidean War, which pitted Ptolemy Philadelphus

against Antigonus Gonatas. It may be that Arsinoes weapons, toil, and
sweat suggest something on these lines. But it is worth recalling that
Aphrodite may be portrayed as armed quite apart from any overt military
function, at various well-known cult sites, particularly on Kythera and in
Laconia.37 As early as Saphhos appeal to the goddess to ght alongside
her, ( (1.28 Lobel-Page), it is possible to speak of Aphrodites
erotic activities in martial terms. But even if we accept a possible reference
to a contemporary war, it would not preclude an erotic subtext. For why is
Hegeso portrayed here as ? Could it be that the maiden was
thinking about an armed Arsinoe-Aphrodite, even dreaming of her, because
she cared about someone involved in a war, a prospective husband perhaps?
One last poem from the new epigrams, though poorly preserved, may
refer to the shrine of Arsinoe-Aphrodite Zephyritis. If so, then the epigram
suggests again how this cult could serve as a conduit through which the cul-
tural heritage of Greece might enter the new terrain of Egypt (37 AB).
/ , [] # [ ]
[] A/ /[
[] 6 / [
[] [
[] , [ 5
[] [
/, [b ] , A [ /],
[,] [.
3 " BG : I/ " [] ex. gr. Austin : ], ], 6] vel sim. BG :
[ / a Lapini : [ / Bettarini : / @[ Puelma 4 [] ex. gr.
Austin : [ ] Lapini : [] Bettarini 5 [ ] , ex. gr. Austin :
[ ] Luppe : [ ] Bettarini 6 [/ (] [ ex. gr.
Austin : [/ (] Bettarini : [ (] Bing : [ Gronewald :
[ Puelma 7 [, /] Austin : [,/ ] Luppe
[Arsinoe, Arions dolphin brought you this lyre

that once resounded at the touch [of a singer]
. . . from the wave. But when [
that one . . . crossed the foaming sea
37. Nilsson (GGR1 I 490 n. 7) cites the temple of Aphrodite Areia (Paus. 3.17.5; cf. Plut.
Mor. p. 239 A; IG V 1.602; Aphrodite 6 (early third century a.d.). There is an armed
xoanon of Aphrodite Ourania on Kythera (Paus. 3.23.1). Cf. Magnelli 1999 on Alexander
Aetolus fr. 9 ( 8 Powell p. 127); cf. also Flemberg 1991: 2942. For the armed Aphrodite in
art, cf. LIMC II 24345.
many things . . . and various with [

his voice . . .
But this offering, O Philadelphus, which Arion played
please accept it, a dedication of . . . your temple custodian.]
This poem records the dedication to Arsinoe Philadelphus, by her temple

keeper (), of a lyre brought ashore by Arions dolphin ( . . .
/[), a reference to the late seventh-century musician of Met-
hymna.38 It seems plausible to infer from the text that this seaborne offer-
ing was made inperhaps even found nearthe temple of Arsinoe-
Aphrodite Zephyritis, the celebrated foundation of the Ptolemaic admiral
Kallikrates of Samos, which (as Posidippus put it in the epigram from the
Didot papyrus, 12.23 GP 116.23 AB) occupied the windy headland
among the encircling waves (6 M ( / . . .
, vv. 23).39 If this is correct, then the poem represents a
striking example of how an object, the lyre, may be made to embody the
cultural/historical heritage and become (quite literally) the vehicle by
38. Contrary to my view in the original version of this article, I now hesitantly follow
BG, who suggest that the masculine articles of vv. 78 refer to a synonym of the lyre lost in
the lacuna at the end of v. 7though it must be admitted that no plausible candidate has been
found until now. Alternately, Austin (in BG) proposes that the articles refer to a word, such as
,, and that the epigram thus represents itself as a song of Arion. More speculatively,
Bettarini (2003) goes beyond Austin, proposing that the missing word is a. This would
refer to a separate hymn, perhaps by Arion, dedicated with the lyre and possibly inscribed
in the temple as well (cf. his n. 81). That hymn would be comparable to or even a version of
that at PMG 939, which Aelian (NA 12.45) says Arion composed to thank Poseidon for his
rescue by the dolphin. In the context of Ptolemaic Egypt, the maritime deity now honored is
Arsinoe-Aphrodite Zephyritis rather than Poseidon. This latter point is also eloquently ar-
gued by Fantuzzi (2004b: 3133); cf. also Puelma 2006: 7071.
Different reconstructions of the closing lines and other parts of the poem have been sug-
gested by Lapini (2003: esp. 3942) and Luppe (2003). Lapini thinks the object dedicated was
a statue of the dolphin-riding poet Arion with his lyre, like that at Cape Taenarum commem-
orating Arions landing there (Hdt. 1.24). Thus he suggests the supplement A [,1
/], with the verb used in the sense of forged. This view has now been taken up and
eloquently argued by Puelma (2006) in a detailed discussion of the whole poem. It does not
explain, however, why the lyre alone ( [] , v. 1) is highlighted with deixis as the ded-
ication at the start of the poem (pace Puelma 2006: 70 n. 27). Luppe makes the more plaus-
ible proposal that the dedication was a sculpture of the lyre, hence [,(), while the end of
the lacuna ] conceals the artists name. On this view, however, there remains a worrying
disjunction between the lyre itself, which Arions dolphin brought to you, Arsinoe, and its
presumed sculptural representation dedicated in the temple: or does Luppe mean that the
dolphin brought the statue ashore (inscribed with the artists name)?
39. The identication of the site is also made by Bastianini and Gallazzi (BG in their in-
troduction to VI 1825), though their characterization of the shrine as the tempio di Arsi-
noe a Canopo is misleading.
which that heritage is transmitted to a new place. For the epigram clearly
alludes to the legend of Arion as told by Herodotus (1.24). In that account,
upon being threatened with robbery and death by the crew of the ship on
which he was sailing, Arionthe best singer in the worldleaped into
the sea in full citharodic regalia. He was saved, however, through the mirac-
ulous intervention of a dolphinthe most musical of creatureswho
caught him up and carried him safely ashore at Cape Taenarum. There, a
statue of a man riding a dolphin was dedicated in a temple of Poseidon to
commemorate the singers deliverance. Clearly Posidippus poem evokes
not just the story of the rescue but also the subsequent dedicationboth of
them made at a coastal shrine on a rugged Cape. By describing how this
lyretogether with the tradition it evokescame to Egypt, the poet links
the third-century b.c. shrine of Arsinoe to one of the great gures of ar-
chaic poetry from the seventh century b.c. and, with him, to the rich tradi-
tion of Lesbic lyric, including Terpander, Sappho, and Alcaeus.
But a further, less obvious model may be oating just below the surface
here as well. For the story of the lyres wondrous appearance on Egyptian
shores may be intended to recall and provide a modern counterpart to a
well-known Lesbian tale linked with Methymna, likewise about a won-
drous poetic windfallthe story of Orpheus lyre, which, after the leg-
endary singer had been torn apart by the Thracian maenads, oated across
the sea together with his severed head until running aground, as Ovid put
it, on Methymnaean Lesbos shore (et Methymnaeae potiuntur litore Lesbi,
Met. 11.55).40 This tale is memorably recounted in an elegy of Phanocles,
plausibly of early Hellenistic date (fr. 1, pp. 1067 Powell).
T / + M , " / "
, 9 J 8 C

\M , G/ 6 J
9, 8.
M 6 D
/ J 15
6 ,
V / u
/ , / Z, (
/ 6 ,
6 + M , !
( a. 20
/ 1
(, / 6 .

. Cf.
40. Lucian (Adv. indoct. 11) retells this story at length, calling it a
Burkert 1983: 202 with nn. 30 and 33.
[They cut off his head with a sword of bronze and threw it at once
in the sea along with the Thracian lyre,
binding them strongly with a nail, so they would both be carried
on the sea together, soaked by the billowing surf.
And the foaming sea drove them to sacred Lesbos.
And the clear echo of the lyre spread across the sea
and over the islands and sea-beaten shores, where men
interred the clear-sounding Orphic head
and set in the tomb the bright-ringing lyre, which used to persuade
even mute stones and the hateful water of Phorkos.
From that time forth, songs and lovely kithara music
have occupied the island, the most musical of them all.]
In effect, this tradition about the lyre of Orpheus constitutes an aition of the
poetry of Lesbos, lending the authority of one of poetrys founding fathers
to that islands status as a great repository and source of song. Transmission
and preservation of the lyre here function virtually as a charter. Inasmuch as
this early Hellenistic text invests the geographic transfer of this instrument
with such poetological signicance, I think it plausible to take the tale in
Posidippus epigram in a similar waythat is, to see it as emblematic of the
Ptolemies claim to be the true inheritors and guardians of the literary
legacy of Hellas, in particular here the great tradition of Lesbian song. The
Lesbian lyre has been passed on; today its home is Egypt.41
At the same time, our epigram on the dolphin and the lyre may point to
important political ties in the mid-third century b.c. between Egypt and
Arions native Methymna on Lesbos. That city came to serve as an impor-
tant strategic base for Ptolemaic interests in the northern Aegean, and its
third-century b.c. coinage included the image of the citharodic dolphin
rider as its emblem.42 The extent of Egypts inuence here is plain in epi-
41. In this sense, the poem serves a function comparable to that of the epigram about the
seal ring of Polycrates (9 AB), with its image of the lyre. For by setting that lyre-bearing gem
in the context of his , with its pointedly Ptolemaic orientation, Posidippus similarly
linked his monarchs with a grand gure of the lyric heritage (in that case, Anacreon), as well
as with Polycrates as an important archaic model of artistic patronage. Gronewald (2004) has
plausibly suggested that our poem on Arions lyre served as the model for an epigram by
Philip of Thessalonica (AP 9.88 Garland of Philip 40); cf. also Puelma 2006. If that is so, it
is striking that Philip leaves no trace of Posidippus concern with the transmission and appro-
priation of the cultural heritage within a contemporary political context.
42. Cf. Wroth 1964: 179 no. 16 (note also the lyre and dolphin on no. 14), 180 no. 27,
181 no. 35. Cf. also Buchholz 1975: plate 12. The lyre alone appears already on Methymnean
coinage of the late fth and fourth centuries, but this is an emblem it shares with the coins of
Methymnas rival Mytilene.
graphic sources, which attest to, for instance, a priest of the divinized
Ptolemy there between 267 and 260 (IG XII Suppl. 115), worship of Arsi-
noe Philadelphus (CIG II Add. no. 2168; IG XII 2, 513), regular celebra-
tions of the Ptolemaia on the model of those in Alexandria (IG XII Suppl.
115; cf. also at Eresos IG XII 2, 527 and Suppl. p. 33), and perhaps a
month named Ptolemaion (IG XII Suppl. 115).43 Hence, a poem com-
memorating Egypts acquisition of the lyre through the miraculous agency
of Arions dolphin would certainly have conveyed a potent political mes-
sage at just the time when Methymna had become a vital part of the
Ptolemies maritime empire and when the nauarch Kallikrates had recently
founded his cult of Arsinoe Zephyritis to mark the deied queens special
patronage of the Ptolemaic navy. Indeed, if E. Puglias recent suggestion44
that the dedicator of the lyre, the unnamed temple keeper () of
verse 8, was none other than Kallikrates himself, then this gift would have
been all the more meaningful for coming from the very man responsible
for those close ties between Methymna and Egypt.
We see, then, how the shrine established by Kallikrates became a focal
point into which might ow and from which might spread the broad polit-
ical/cultural interests of the Ptolemaic courtinterests that were far-ung
both geographically and chronologically. In this respect, the poem about
the lyres migration takes its place beside comparable epigrams concerning
objects dedicated in this shrine, such as one by Callimachus (14 GP:
65, Z 5 Pf. Athen. 7.318) that charts the passage to
the land of Egypt of the roving nautilus shell and its itinerant dedicator
a certain Selenaia from Aeolian Smyrna (v. 12) who found the shell on the
beach at Iulis on Keos, a way station perhaps on her way from Smyrna to
Alexandria.45 The poem conforms to what Selden has described with re-
spect to Callimachean dedicatory epigram generally; that is, it locates the
offering at the site of multiple displacements, leaving traces of . . . alter-
ity that make the dedication signicant.46 The lyre, just like the shell and
43. See, generally, Brun 1991; Labarre 1996: 5456. Labarre doubts Bruns dating of the
priesthood for the divinized Ptolemy, placing it ratheras originally proposed by Habicht
(1970: 109) and Bagnall (1976: 162)in the reign of Philopator. The month name Ptolemaion
is plausibly conjectured in IG. Cf. Buchholz 1975: 230; Trmpy 1997: 247.
44. In Puelma 2006: 71 with a n. 32.
45. It has been observed that in the mid-third century b.c., the nearby Kean town of
Koresia became an important Ptolemaic port, which (signicantly) was renamed Arsinoe fol-
lowing the queens death (cf. Robert 1960). It is probably also important that Keos has liter-
ary signicance as the birthplace of Simonides and Bacchylides.
46. Selden 1998: 309, with his illuminating discussion of the Selenaia epigram (30913).
In a forthcoming article New Inscriptions on Old ShellsCallimachus Nautilus Epigram
its dedicator, nds its way to the shrine from elsewhere and carries some-
thing of its origins into its new environment.
The same may be said about another offering in the temple of Arsinoe
to whom the West Wind is dear (, 1)that is, Arsinoe-
Zephyritisdescribed in an epigram of Hedylus (4 GP Athen. 11.497D).
Contrary to what we found in the previous poems, however, its alterity
bears the exotic stamp of Egyptian cultureand here we recall Kallikrates
interest in fostering Egyptian gods as well as Greek ones, manifested in his
founding of a cult of Anubis and Isis on behalf of Ptolemy and Arsinoe at
Canopus (mentioned earlier in the present essay). The offering is a rhyton
conceived and dedicated by the Greek engineer Ktesibios in the shape of an
Egyptian god, the dancer Bes (4 ,, v. 3). When
wine ows through its mouthpiece, the god seems to trumpet a shrill note
(T 1 V / , 8 ,, vv. 45), which is
like the ancestral melody that Lord Nile produced from his sacred waters
( C . . . / # 6 #, vv. 89). In
short, the poem represents an Egyptian melody played by an Egyptian god
at a shrine in Egypt but mediated by an epigram in Greek commemorating
a Greeks dedication to a deied Greek queen.
Even in those poems, then, in which Kallikrates is not actually named,
his shrine remains true to that project that, as we have seen, the admiral
seems to have set himself, namely, to mediate between old world and new.
Situated on the windswept frontier between those worlds, simultaneously
a point of convergence and a clearinghouse, it continued to serve as a con-
duit through which political/cultural traditions of Greece could enter into
Egypt and from which the Ptolemies could broadcast their own peculiar
contributions to that legacy.
(5 pf. 14 G.-P.) and the Aesthetics of Inscribed Seashells, Chad Schroeder makes a fasci-
nating case that this epigram also bears traces of Egyptian tradition, in which shells were dec-
orated and inscribed for dedication. For more explicit reference to Egyptian culture in a ded-
ication at the temple of Arsinoe-Zephyritis, see the next example in this essay, Hedylus 4 GP.
c h a p t e r 13
The Politics and Poetics of Geography

in the Milan Posidippus Section One,
on Stones 120 AB

A t the start of his sixth Olympian ode, Pindarcomparing the con-

struction of his song to that of a conspicuous palace whose portal is
raised on golden columnsmemorably states that when a work of poetry
is begun we must make the entrance far-shining [, v. 4]. The new
Posidippus papyrus was arranged in at least nine sections, each headed by
a title. Of these nine, the rst and longest with its 126 verses, was evidently
on Stones. Its fragmentary title may with some plausibility be restored as
[] in light of the consistent subject matter of its twenty-one epi-
grams.1 That it was the rst section of the roll seems likely. For although
stichometric annotations appear in the margins at the end of each section
to record its length, only in that on stones is there a further stichometric
note at the bottom of the rst column, strongly suggesting that it held a
special place as the rst of the roll.2
This essay is adapted from that originally published in K. Gutzwiller ed., The New Posidippus:
A Hellenistic Poetry Book pp. 11940, 2005 Oxford University Press.
1. Diodorus Siculus cites as the title of a work by Orpheus (7.1.1), perhaps the
same mentioned by the Suda (s.v. Orpheus) as being on eighty gemstones and their engrav-
ing. And indeed, such a work is extantthe Orphic : cf. Halleux and Schamp 1985; cf.,
generally, Plantzos 1999: 10.
2. Cf., generally, BG p. 13. To be sure, the start of the papyrus was repaired with a new pro-
tokollon, leaving the chance that there were segments preceding what is now the rst column
253
Further, as an opening to a poetic work, the section on stones provides

precisely that splendid introduction Pindar recommended. Indeed, that
special brilliance embodied in Pindars insistently recurs in the
terms Posidippus uses to describe the luminosity of his gemstones: "
(8.6 AB), " (3.1), (16.5); such nouns as (6.6) and
(7.6); the verbs (11.1), (6.3), and (7.6); such
adjectives as (16.1); and above all, terms evoking celestial phenom-
ena, such as (5.1), (4.3), V (16.6), V
(14.1), , (14.6), and (13.4).
An eye-catching start. But to what end? In what follows, I argue that the
Lithika draw our attention, right from the rst, to fundamental themes that
gure prominently elsewhere in the scroll, illuminating important aspects
of both politics and poetics and frequently tying them together. In particu-
lar, I hope to show how the stones exemplify, in their geographical distribu-
tion and social construction, both the territorial and cultural/artistic aims of
the Ptolemies and of their poet, Posidippus.
To begin with poetics, many of the poems of this section linger on the
artists exquisite workmanship, and it is tempting to read them program-
matically: as art contemplating art, they invite a self-reexive interpreta-
tion likewise apt for the beginning of a work.3 David Schur suggests that
such an interpretation may extend even to the title, for inasmuch as epi-
gram is by its origin and history inescapably linked with stone, the possi-
bility that this opening section was called things having to do
with stonessuggests a high degree of generic self-awareness.4 Indeed,
in this sense, could be taken as a title for the overall collection.
In any case, the concern with artistry seems to reect a number of stan-
dard Alexandrian preoccupations. It is linked here, for instance, to
diminutive works in a minor artistic genre, Kleinkunst as the Germans call it.
Though I shrink to mention it, the term slender, delicate, re-
nedappears in the very rst poem (1.4 AB). This, of course, was a cru-
cial watchword of Hellenistic poetics: notably in the programmatic opening
poem of Callimachus Aetia, with its slender Muse (fr. 1.24; cf. AP 9.507
Ep. 27 Pf. 56 GP), and in Aratus celebrated acrostic at Phainomena 783
(selis) of writing. Nonetheless it is likelier that the section on stones formed the original open-
ing. Damage such as that repaired by the new protokollon is characteristic of the outermost
(i.e., the opening) part of the scroll, where the papyrus is most exposed to mishap (cf. Turner
1968: 5). Johnson (2005) is more skeptical that this formed the beginning of the roll.
3. One thinks, for instance, of Theocritus 1 or the prologue to Callimachus Aetia.
4. Cf. Schur 2004.
The Politics and Poetics of Geography in the Milan Posidippus 255
87; thereafter also in Augustan poetry.5 A section leading off with gems pro-
vides an ideal platform to highlight this rened aesthetic, for again and again
Posidippus calls attention to the delicacy of the engravers art. Delicacy such
as this is, moreover, a result of and . In one of the two previ-
ously known epigrams, the speaker claims that the engraved chariot on a
snakestone must have been carved with the superhuman vision of a Lynceus
(/ # 6 15.4 AB AP 16.119.4 20.4
GP), nding it a marvel that such toil () did not hurt (cognate
6) the artists eyes (/ , C
/ " 6 , 15.78). Yet other epi-
6
grams in the section suggest that tiny works like thesethough marginal
are prized by connoisseurs, particularly royalty. A sadly fragmentary poem (9
AB) evokes the famous seal ring of Polycrates, which in Herodotus (3.41)
was the Samian tyrants most treasured possession. The epigram introduces
a signicant innovation vis--vis Herodotus in that it has Polycrates take as
his emblem on the seal the lyre of a singer-man, strumming his song at
your feet ( / ] [ [] ,
9.12 AB). This most likely refers to Anacreon, whom Herodotus describes
as present in Polycrates banquet hall (3.121). Thus Posidippus brings us
back to the contemplation of poetry. Indeed, inasmuch as this tiny seal
([, 9.1 AB) is embodied in the comparably brief compass of his epi-
gram, he suggests that the one art form can stand for the other. In so doing,
he indirectly presents us with a sphragis of his own, bearing the stamp and
conveying to us the impression of an important poetic forerunner.
But more than possibly delineating programmatic parameters, the sec-
tion on Stones explores and maps out a political landscape reecting cer-
tain aspirations of sovereignty that set the tone for the whole work (i.e., it
outlines the world we are dealing with).7 In this regard, we do well to re-
call the long section in Theocritus Idyll, the Encomiium to Ptolemy, which
recounts the lands Ptolemy Philadelphus counted as his own in addition to
Egypt (vv. 8693).
5. For the term and its impact in Roman poetry, cf., e.g., Hopkinson 1988: 90, 98101.
We will see that Posidippus uses metapoetic water imagery in ways similar to Callimachus (cf.
n. 17 in the present essay). Further, the similarities between the seal poem on his old age
(SH 705 118 AB) and the prologue to Callimachus Aetia have long been noted.
6. For the emphasis on techne in Hellenistic poetics, cf. Callim. Aet. 1.1718. For toil
as part of the poets self-image, cf. the paignion of Philitas (10, p. 92 Powell 12 Sbardella,
with commentary). Posidippus (AP 12.98 6 GP 137 AB) describes his as 6
(v. 3). Cf. generally Prioux 2007b.
7. Susan Stephens thinks in similar terms (2004: 17071).
/ 86
/ , D
,
,
, 6 90
6, + ,
M
[And he cuts off for himself part of Phoenicia and of Arabia

and of Syria and Libya and the black Ethiopians.
And he holds sway over all the Pamphylians and Cilician spearmen,
and over the Lycians and war-loving Carians
and over the Cycladic isles, for his ships are the best
that sail the sea: indeed all the sea and all the earth,
and the thundering rivers are ruled by Ptolemy.]
As Gow comments (ad v. 92), the hyperbole in Theocritus is remarkable.

But Callimachus is no less hyperbolic when he speaks in his fourth hymn
(vv. 16670) of that same sovereign, beneath whose crown shall come
not unwilling to be ruled by a Macedonianboth continents and the lands
that are set in the sea, as far as where the ends of the earth are and the
source from whence the swift horses of Helios carry him (ME # /
G " / j
, / @ C b G /
/ ).8
The rst words of the poem in Posidippus section on Stones are
/ ;, the Indian river Hydaspes (1.1 AB)the source, it
appears, of the gem that formed the subject of the poem. This river was
essentially the furthest limit of Alexanders conquests in the East. The sec-
tion ends with a poem about a massive boulder in Euboea, in the tradi-
tional Greek motherland, and closes with a prayer on behalf of Ptolemy:
But now, Geraestian lord [Poseidon], along with the islands of Ptolemy
preserve his land unshaken and also his shores ( , / ,
/ : ,, 20.5
6). This closing prayer, suitably open-ended in its denition of Ptolemaic
boundaries, retrospectively colors our understanding of the geographic
parameters within which the section unfolds, setting it all in a Ptolemaic
8. Cf. Hunter 2003b: ad 8692, on both these passages as reecting conventions of the
oriental rhetoric of kingship (p. 168) and, in particular, of Alexander as a model.
perspective:9 the range of lands available to the Ptolemies as a source of

wealth corresponds to no less than the empire won by Alexander. In addi-
tion to India, which was famously rich in gems10 (cf. also 2.4, 8.5), the stones
in these epigrams derive notably from Persia (4.5, 5.2, 11.2, 13.3), Arabia
(7.1, 16.1), Lydia (8.1), Mysia (17.1), Samos (9.1), and the depths of the sea
(11.1, 12.1). As Theocritus says in Idyll 17, In wealth Ptolemy could out-
weigh all other kings, so much comes each day to his sumptuous house from
everywhere (PM + D / 6/
I 6 ( , / , 17.9597).11
The evocation of Alexanders empire in the geographic distribution of
stones in the nds conrmation in a little-noted passage from
Theoprastus Characters. There, in describing the alazon, the idle
boaster, Theophrastus tells how this character is apt to claim that he was
on close terms with Alexander the Great himself on his Eastern campaigns
(23.13).
C + 5 , . . . +
6
J CM
, E / / 6, E "M
9. Lest we miss how the reader retroactively acquires a Ptolemaic orientation, the rst
two poems of the following sectionthe oionoskopikareinforce the shift by turning our at-
tention toward Egypt. The section is launched with an epigram about launching a ship (21
AB), beginning and ending with the words (for similarly artful equivalence
between beginnings and ends of a journey and of a poem, cf., Wray 2000, 24145. It is fol-
lowed by one in which the speaker presents himself as a traveler heading toward Egypt: For
us who are about to seek out the Egyptian sea, may the Thracian crane be our guide along
the forestays (? / , /
? , 22.34 AB. The word suggests that the speaker is far from Egypt
and that the , is his goal (cf. LSJ s.v. I 2). In all likelihood, his jour-
ney is from Europe (Macedonia?) to Egypt, since, as Kannicht observes in his commentary
on Euripides Helena (1969: ad 147894), the crane is known above all as an intercontinental
traveler, and only his migratory ight in autumn from north to south appears in literature:
der notorische Kranichzug [ist] immer nur der des Herbstes, weil eben nur der Herbstzug
beobachtet werden kann; denn nur im Herbst lassen sich die Kraniche sozusagen Zeit, fallen
hier und da in die Lnder, die auf ihrer Route liegen, ein, rasten, und iegen
weiter. Deshalb ist in den literarischen Zeugnissen fast ausnahmslos der Herbstliche Nord-
Sd Zug gemeint . . . Der Rckug im Frhjahr bleibt dagegen in der Regel unbemerkt, weil
die Kraniche dann ohne Verzug und in sehr groer Hhe direkt ihre nrdlichen Brutgegen-
den aniegen (p. 388). For the transitional function of the rst two poems of the oionosko-
pika, cf., further, Petrain 2003: 381; Stephens 2005.
10. Plantzos 1999: 106.
11. Nenna describes Ptolemaic interest in gems from the Arabian desert as follows:
Une des premires dcisions des Ptolmees fut de rorganiser lexploitation des carrires du
dsert arabique. Cette industrie minire, fonde sur la recherche des meraudes, topazes ou
de pierres semi-prcieuses . . . tait soigneusement surveille par les collecteurs et les con-
voyeurs de pierreries appoints par le souverain (1998: 156).
,, @ 6D

6 J /Y, @ ,
6 J "J, D
, " 6
[The alazon is the sort who . . . on a journey is apt to put one over on a
travel companion by relating how he campaigned with Alexander, and
how Alexander felt about him, and how many gem-studded goblets he
brought back as booty, and arguing that the craftsmen in Asia are bet-
ter than those in Europe (he says all this even though hes never been
out of town).] (translation after Rusten 2002)
Wonderful how Theophrastus here slyly suggests that the journey and
companionship of the road prompt the alazon to recall his imagined
journeys and friendship with Alexanderonly to divulge in the end that
the present journey is strictly parochial: the boaster has never in his life
strayed beyond the narrow limits of the polis. Most interesting in this pas-
sage for our purposes, however, is what this provincial alazon associates
with those campaigns: gem-studded goblets and superb craftsmen. The
passage is revealing in a number of ways. First, it suggests that the topic of
rare gems and their artistic use was easily linked in the public mind with
Alexander. Second, it shows how the Macedonians conquests were viewed
as having made accessible the great mineral wealth of the East. That is
what Alexanders comrades in arms would have brought back from their
campaigns as booty (@ . . . 6; cf. LSJ s.v. II 2). Finally, it re-
veals that people from all strata of societynot just the ruling classes
were aware of these precious objects, knew of their exotic provenance, and
had some appreciation for the artistry involved in turning them into jew-
els: traveling companions might idly chat about them to pass the time on a
journey, and debate the relative merits of Asian and European craftsmen.
That broad awareness, together with the association of gemstones with
Alexander, would make a politically charged topic and an appealing
vehicle for promoting Ptolemaic aspirations.
Those aspirations were not restricted to the Ptolemies, however; they
could extend as well to those in their employ. It is striking that the territorial
aims of the Ptolemies on view in these poems match those of Posidippus
himself. For the poet sets similar parameters to his own fame in his seal
poem, the elegy in which he considers his legacy from the perspective of
old age (SH 705 118 AB). There, he asks Apollo to give an oracle so that
the Macedonians may do me honor, both those on the islands and the
dwellers near the coast of all Asia (= [] G / 6
[ / G / / V, vv. 1415). In particular, the

phrase all Asia (/ ) invites readers to think of this continent in
the most expansive terms, as, for example, in Herodotus (4.3641; cf. 2.16),
where Asia extends from the Hellespont all the way to India.12 In this way,
the lines attest to the radical expansion of Macedonian power in the after-
math of Alexanders conquests, notwithstanding its humble beginnings,13
for here again we are dealing with territory that may be considered coexten-
sive with Alexanders dominions. Now the Ptolemies were eager to present
themselves as Macedonians, as the Milan papyrus repeatedly demon-
strates (78.32f., 82.3, 87.2, 88.4 AB; Paus. 6.3.1, 10.7.3; Callim. Hymn
4.166f., cited above),14 and so was Posidippus, one of the principal poets
who served them: My family is from Pella, he declares in the very next
line ( , SH 705.17 118.17 AB), thus identifying him-
self with his public and also justifying his expectation of honor from them.
We may thus see Posidippus hopes for recognition from the Macedonians
as being bounded only by the limits of Ptolemaic territorial ambition. Here,
then, political and poetic aspiration coincide, and the geographic breadth of
opportunity available to the poet expands with that of his patrons.15
Within these geographic parameters, the details of the journey from
physical source to cultural application are a persistent theme of the :
the passage from earth, sea, or shore to the artisan who shapes the natural
object and, through his representational skills, makes it signify; how, fur-
ther, it may be bound in a frame as a jewel and, nally, take its place in a so-
cial setting, on a persons nger, neck, breast, or furniture. To illustrate, let
me cite just one poem, 7 AB (with the admittedly uncertain, but plausible,
supplements of Bastianini and Gallazzi).
12. Contrast the more limited sense noted by Dodds (1960) for / at Eur.
Ba. 17, where it is used in the restricted sense of western Asia Minor, as the context shows.
Posidippus reference to the dwellers near the coast of all Asia accords with the idea in
Herodotus (4.44) that Asia is largely dened and bounded by its coasts.
13. This contrast is stressed by Polybius (1.2.2): The Macedonians ruled Europe from
the regions along the Adriatic Sea to the Danube River, which seems an altogether small part
of the above-mentioned land. Later, by overthrowing the Persian dynasty, they added sover-
eignty over Asia ( + " V / I
6 P , T 1 [ D +
/ , ).
14. Cf. Fantuzzi 2005: 25052.
15. It is worth considering, as Kathryn Gutzwiller suggests, whether this poem may have
been placed as the closing piece in the missing nal part of the Milan papyrus (2005a: 317
19). Certainly it would there have stood in appropriate balance to the geopoetic concerns of
the opening section in the . For the relationship of Posidippus seal to the biographical
literature of the time, see Bing 1993a.
6 /A / 4[ ] ,
, 9 b / [6 ]
[ :, T] []
(D [ @ ]
[ , E] 6
.
[Rolling the yellow debris from the Arabian mountains,
the storm-swollen river carried swiftly to the sea
this honey-colored stone, which the hand of Cronius
carved. Bound fast with gold for sweet
Niconoe it blazes as a necklace chain, so that on her breast
its honeyed radiance gleams together with her fair skin.]
The poem starts with a detailed echo of a Homeric simile in which Hector,
in his assault on the Achaean ships, is likened to a great stone pried loose
from its rock by a winter storm, and carried irresistibly down into the
plain, but then coming, powerless, to a halt (Il. 13.13743).
. . . 4 u ,
@ eJ
8 M =M ( D

a / M , / #/ " 140

aD T / (, G
,, / _ 6 D
u K + . . .
[. . . like a great rolling stone from a rock face

that a river swollen with winter rain has wrenched from its socket
and with immense washing broken the hold of the unwilling rock face;
the springing boulder ies on, and the forest thunders beneath it;
and the stone runs unwavering on a strong course, till it reaches
the at land, then rolls no longer for all its onrush;
so Hector for a while threatened lightly as far as the sea . . .]
Posidippus varies the source of stone from the Iliads rocky height
(, v. 138) to Arabian mountains; for Homers (v. 142),
he substitutes the more modern form (7.1); he apparently reverses
the Homeric (v. 138) with . . .
(7.2); and he transposes Homers mention of the sea as stopping point
( , v. 143) from the framing narrative about Hector into his
account of the stone (, 9, 7.2). If we follow the reconstruction of the

editors, the yellow debris reveals a honey-colored stone, which the skillful
hand of an artist like Croniusmentioned by Pliny as being second in
fame after Alexanders personal gem engraver, Pyrgoteles (NH 37.4.8)
transforms into a jewel. Shaped thus, and translated to another kind of hill
(6 , 7.5),16 it can match the radiance of human beauty in the lovely
Niconoe. While discarding the overt form of the Homeric simile, the poet
recalls its function by subtly comparing and equating (, 7.6)
the stones honeyed color ( [ :, 7.3;
, 7.6) to the fairness of the womans skin ( , 7.6)and per-
haps also to her sweetness, if we accept the thrust of the supplement
] (7.4). But unlike Hector, whose onslaughtlike the boulders
nally grinds to an impotent halt, Niconoes power is undimmed, for the
meaning of her name implies that the possessor of such a beautiful gem has
been and will be triumphant in her plans, whatever those may be. In that
light, the costly journey from exotic source through the artists transform-
ing hand and into the possession of a Greek owner proves worthwhile in-
deeda potent gure for the exploitation and mastery of the earth in an
implicitly Ptolemaic context.
At the same time, from a poetological standpoint, the Homeric simile
of the massive stone thundering down the mountain is an apt image of the
epic hero. The development of that simile, then, by Posidippuswhich
focuses on the ne gem extracted from whats swept down by the torrent
and how it is artfully cut into a jewelmay stand as an allegory of the Hel-
lenistic poets reception and transformation of Homeric raw material to
new ends.17 Such a poetological subtext will hardly surprise in a section
dealing so insistently with sources.
Without a doubt, it is in the that the politics and poetics of
geography (as I call it in the present essays title) intersect most promi-
nently. Yet this section represents only a particularly rich lode in a vein
that ramies throughout the scroll. One could, for instance, productively
16. For this double meaning of , cf. Pindar Pyth. 4.8; Callim. Hymn 4.48 with
Mineurs note (1984).
17. In its programmatic Bildersprache, it resembles Callimachus Hymn to Apollo (vv. 108
12), with its opposed images of the masses of waste and refuse that the great stream of the
Assyrian river sweeps along and the trickling drops from the pure and unpolluted spring.
For Posidippus, however, the debris-carrying torrent is itself the ultimate source of the gem.
For another Iliadic simile about surging water (11.49297) and its impact as a stylistic
metaphor in Hellenistic and Roman poetry, as well as in rhetoric, cf. the illuminating discus-
sion of Hunter (2003a: 21923). Concerning Homeric allusion elsewhere in the Milan pa-
pyrus, cf. Petrain 2003; Hunter 2004; and diNino 2009 (forthcoming).
mine poems of the or of the for the convergence

of geopolitics and geopoetics. In my essay Posidippus and the Admiral
in the present volume, I treat in detail an exemplary instance, 37 AB, from
the section on dedications, the , which suggests the ongoing
importance of this theme throughout Posidippus work. Let us set aside
our gemstones for a moment, then, to explore it briey. As in the poem on
the seal ring of Polycrates, this one concerns a lyre, and as in many of the
, it charts an objects route from exotic source into a new Ptolemaic
context. The poem commemorates the dedication to Arsinoe Philadel-
phus, by her temple keeper (), of a lyre brought ashore, most
likely near the temple of Arsinoe-Aphrodite Zephyritis, by Arions dol-
phin ( . . . /[). The lyre evokes not only the great seventh-
century musician of Methymna but, with him, the whole rich tradition of
Archaic Lesbic lyric (including Terpander, Sappho, and Alcaeus), linking
it with the third-century shrine of Arsinoe. Its transfer to Egypt exempli-
es the Ptolemies claim to be the true inheritors and guardians of the lit-
erary legacy of Hellas, in particular (here) the great tradition of Lesbic
song. At the same time, Egypts acquisition of this lyre may have a politi-
cal dimension, since the Ptolemies made Arions native Methymna an im-
portant strategic base for their interests in the northern Aegean in the
mid-third century; the city became a vital part of their maritime empire.18
As in the , then, we see how a precious object embodying Egypts
geopolitical/geopoetic ambitions enters the eager, inclusive embrace of
Ptolemaic power.
The ip side of this ambitionthat is, to make Egypt also a source dis-
seminating that poetic heritage to the worldis apparent elsewhere in
Posidippus, namely, in his epitaph for Doricha, a hetaira of Naucratis in
Egypt, famous as the mistress of Sapphos brother Charaxus (Athen.
13.596c 17 GP 122 AB).19 In commenting on the enduring power of
Sapphos verse, Posidippus here signicantly links it with its written
medium: The bright resounding papyrus columns of Sapphos / dear song
abide and will yet abide (M + ( /
18. I lay out the details of these ties in Posidippus and the Admiral in the present
volume.
19. Cf. already Hdt. 2.135, where the author calls her by her nickname Rhodopis (cf.
Page 1955: 49 n. 1). However, Lidov (2002) proposes that Rhodopis and Doricha only came
to be identied with one another in the Hellenistic biographical tradition, for which Posidip-
pus is our earliest source. Doricha also appears in Sapphos poetry (frs. 7, 15 L-P), though not
immediately connected with Charaxus (frs. 5 and, generally, 202 L-P). On Posidippus recep-
tion of Sappho and recontextualization in Hellenistic Egypt, see Acosta-Hughes and Barban-
tani (2007: 439, 447, 452).
Mb , vv. 56). The importance of writing

as a means of preservation is further stressed in the following couplet,
which addresses Doricha: Most blessed is your name, which Naucratis
will preserve E / as long as a ship sails out from the Nile across the salt
sea. (_ , T E , / (/ [ :J
6/ Z , vv. 78). In the sepulchral context, we would
normally expect E (v. 7) to be used in its conventional deictic sense and
mean here. But Posidippus pointedly thwarts epitaphic expectation. For
here refers back to the ability of the papyrus scroll to bestow perma-
E
nence on its subject ( . . . ( /
, vv. 56) and therefore means that it is thus, in this way20that
is, through the medium of the scrollthat Naucratis will preserve the het-
airas name. It will do so, moreover, as long as a ship sails out from the
Nile across the salt sea (v. 8). P. A. Rosenmeyer elucidated the particular
point of this conclusion, acutely observing that the ship was probably
laden with papyri.21 Naukratis was indeed ideally located in the Delta to
play a signicant part in the papyrus trade.22 Its very name (, v.
7), meaning the city whose power is in ships, bespeaks its mercantile
strength and age-old standing as a base of maritime trade between Egypt
and far-ung points in the Mediterranean.23 That connotation of the citys
name is deftly activated by Posidippus in the nal line ((/ [ :J
6/ Z , v. 8), inasmuch as the envisioned there implies
that Naucratite shipswith their precious cargo of scrollswill sail down
the Nile and out to sea forever.24 Here, then, Posidippus, famous in the
Greek world as a poet of epigrams for monuments,25 shows himself equally
aware of the memorializing power of papyrus. In particular, he is mindful
of Egypts special role not just in collecting and preserving but in dissemi-
nating the precious poetic heritage of Greece.
Precious objects could be of various sorts, however, even amongst the
20. On E , cf. Gow and Page (GP ad 17.7), who note that the translation thus seems
more likely than here.
21. Rosenmeyer 1997: 132.
22. E. Marion Smith (1926: 35).
23. See, most recently, Mller 2000; Hckmann and Kreikenbom 2001.
24. The verses may be viewed as an updatefrom the perspective of book-conscious
Hellenistic Egyptof what is implied already about the easy dissemination of written verse
in such an image as Pindars at Nem. 5.23: / 6 C ( / M, / ,
/ / , / (on every merchant ship, on every boat, sweet song, go forth
from Aegina proclaiming the news).
25. Cf. the honors bestowed on him as 6 at Thermon, which are dis-
cussed in greater detail in the chapter on Reimagining Posidippus in the present volume.
. Returning now to our stones, we nd there Ptolemaic interests of

another, though related, kind, reected in the epigram that follows the ini-
tial sixteen-poem seriesthe gemstone sequence, as Gutzwiller has
called it (2004: 88). This epigram has to do with a stone whose natural
properties, rather than any artful craftsmanship applied to it secondarily,
make it a marvel worthy of contemplation (17 AB).26
C P
D
+ I 8

, / 6 ,
6D T 6 2 "
, 5
, .
[Look upon this stone, such a one as Mysian Mt. Olympus grew,
wondrous for its double power.
With one side it easily pulls the iron set before it,
like a magnet-stone; with the other it drives it far away
with opposing effect. Even that is the marvel, how from one itself
it imitates two stones with regard to movements.]
This poem suggests that the Ptolemies were interested not simply in
claiming the wealth of the world but also in gathering together its wonders
(note , v. 2; , v. 5). But what is the nature of this marvel?
Bastianini and Gallazzi were puzzled by the idea that one stone imitates
two: It would not be clear, they say, what the two are that are
imitated by the stone described: one could be the , which at-
tracts iron; the other, however, which repels it, would not be described in
any way. Above all, nally, the phrase would add nothing of substance with
regard to what was previously said, while the connective T neces-
sarily implies that the topic is something new.27 Without minimizing the
difculty of the nal couplet, I think that the commentators fail to account
here for the long-standing ignorance in ancient sources about magnetic
polaritythat is, that every magnet has two poles and that when two mag-
26. On this poem, cf. Luppe 2001.

27. BG ad III 19: non si capirebbe quali siano le due imitate dalla pietra de-
scritta: una potrebbe essere il che attira il ferro, l altra per, che lo respinge,
non sarebbe indicata in nessun modo; sopratutto, inne, la frase non aggiungerebbe nulla, in
sostanza, respetto a ci che detto prima, mentre il nesso T implica necessaria-
mente che si dica qualcosa di nuovo.
netized objects of the same pole meet, they repel, while opposite poles at-
tract. Surprisingly, the earliest known textprior to the discovery of our
epigramto describe a stone that can simultaneously attract and repel
comes from the fth century a.d.28 In rare instances, authors note that
magnets can occasionally repel, but as A. Radl remarks in his study Der
Magnetstein in der Antike, that is an exception: such a faculty is more
usually attributed then to another stone altogether.29
Thus the standard view in ancient sources is that magnets attract. That
is why our epigram compares its subject to a magnet when seeking to illus-
trate that ability alone: With one side it easily pulls the iron set before it, /
like a magnet-stone ( + I 8 /
, vv. 34). Those that repel, by contrast, are normally considered
different stoneshere the second of the . . . that our stone
imitates. This epigram is thus by far the earliest evidence for a single stone
that incorporates both powers. Inasmuch as this ies squarely in the face of
prevailing wisdom, it would indeed be perceived as a marvel (, v. 5)
that this stone from Mysian Olympos, being only one, imitates two stones
with regard to movements.30 The unexpectedness and novelty of the phe-
nomenon justies our taking the relative phrase T as retrospec-
tive, looking back to the preceding description of the stones paradoxical
doubleness and amplifying it (even this is a marvel) rather than introduc-
ing some new wonder.31
The poem reects the Hellenistic interest in paradoxography, which
was given a decisive impetus and new scope through the conquests of
28. Cf. Marcellus Empiricus De medicamentis liber 1.63; thereafter Johannes Philoponus
commentary on Aristotles Physics 403.24. Radl (1988: 7), discounting the testimony of Mar-
cellus, concludes: Lediglich einmal [sc., in Philoponus] wird als ganz besondere Kuriositt
bemerkt, da ein und derselbe Stein sowohl anzieht als auch abstt. Cf. also H. Rommel,
RE XIV/1 477: Neben der Anziehung beobachtete man auch die Abstoung . . . ohne da
man sich aber ber die Polaritt ganz klar wurde.
29. ber eine abstoende Wirkung wird ebenfalls berichtet; aber nur als Ausnahme,
die dann eher einem anderen Stein zugeschrieben wird (Radl 1988: 7). Cf. Lucretius
6.10423; t quoque ut a lapide hoc ferri natura recedat / interdum. Similarly, Plutarch (De Iside
et Osiride 62, 376b, citing Manetho [FGrHist III C, n. 609 F 21]), speaks of a stone that at dif-
ferent times ( . . . ) attracts and repels. Pliny (NH 20.2) sees the stone
that repels as a different stone.
30. The phrase 6 2 " recalls philosophical discussions concerning d ". Cf.
Plato Parm. 137b3 and esp. Aristot. Met. 1001b5, which pointedly asks from whence is there
to be another one besides the one itself? (6 d ( " I;). That
seems precisely what is miraculous about the one stone that embodies the function of two.
31. The editors take T in the latter sense as meaning and this, too, is a won-
der. Granted that that is its sense at 8.7 AB, we need not assume that the phrase will always
be used in the same way.
Alexander the Great. One may see how the inux of carved gems into the
Ptolemaic realm, as described in the , is paralleled by the collecting
of wonders initiated by Callimachus in his , 9
.32 Or in verse paradoxography of the same pe-
riod, there are the poems of the Egyptian Archelaus (SH 12529), who,
according to Antigonus of Carystus, explained paradoxes in epigrams to
Ptolemy.33 In other words, what ows into the Ptolemies domain is not
merely material wealth (=)as Theocritus 17.9597 (quoted earlier)
might suggestbut comprises as well the wealth of cultural/scientic in-
formation, including strange fact and fancy, that accompanied territorial
expansion (or its ambitions).34
The Ptolemaic interests as surveyed in the section on Stones are not
simply geographic but cut across time. We already mentioned the epi-
gram concerning the ring of Polycrates. There are other antique gems
described in this section as well, particularly associated with Persian roy-
alty (4, 8 AB), through which Posidippus suggests the Ptolemaic appro-
priationfollowing the ever-present model of Alexanderof the artistic
inheritance of the Persian Empire as well. Perhaps we may find traces
of the Persian legacyagain via Alexanderin yet another poem where
it has not previously been suspected. I mean in the poorly preserved 18
AB, which, with its opening invitation to recline (, v. 1) and
its apparent references to a young wine steward (,] 1 , v.
3) as well as to an amphora (] , v. 4), seems to point to a sympotic
context.
]/ 6/ (/, 6
, [] D
] 65 [] [
,] 1 [] [
8] I [ ] D
V]D + / [ ] , ? [ 5
] + [] []
32. Callimachus is seen as the rst fully edged exponent of the genre paradoxography,
exerting important inuence on his contemporary Antigonus of Carystus, who wrote
; . Cf. Susemihl 1891: 463ff.
33. The reference is to either Ptolemy II Philadelphus or Ptolemy III Euergetes (De
mirab. 19.34 PGR p. 42 Giannini); cf. Fraser 1972: I 77880. Cf. also the verse and prose
paradoxes of Callimachus student Philostephanus of Cyrene.
34. Our poem is of course only an explicit example of a scholarly/paradoxographical in-
terest that pervades the whole section, as has been shown by Martyn Smith in his examination
of Posidippus use of technical treatises (2004). For paradoxography in the Milan papyrus
generally, cf. Krevans 2005: 8992, 96.
] [ 6] [

] + 6/ d [], / [
1 / [] Austin 3 [] [ Austin 4 [ Austin 5 /

[ Fantuzzi : [ Austin : [/ Z Austin 6 [ 1] [] Austin 7
[ / Austin
[Come to me, you nine men, and lie down [together?]

. . . for I . . . three . . . of stone
with a young wine-pourer . . .
easily . . .will hold a six-choes amphora
Look! Here the width is that of ve [men] where [the joint is held
together (?);
here it is of three spans; [here it is much] fatter (?)
with four corners, [but greater (?)] in length
on the one hand six, on the other . . .]
Bastianini and Gallazzi (BG ad III 23) propose that this poems subject and
speaker may be a large and costly stone krater, which declares . . . its own
capacity (more than a six-choes amphora, i.e., more than twenty liters),
which it holds easily. If, say the editors, we are truly dealing with a stone
krater, then we might be able to understand why the epigram would be in-
cluded in this section (BG ad III 2223; cf. also introduction to III 2027).
But something is not right with this picture. A six-choes amphora is not
particularly large. The standard capacity of a storage amphora is twelve
choes ( one ; cf. OCD3 s.v. measures). A container somewhat
more than half that size, even if this were a deliberate understatement,
does not mesh with the description of an object whose width alone sufces
to hold ve men (/ [ ] , v. 5)35 and that appears to have
four corners (, v. 7).
I think it more likely that we are dealing with one of two possibilities.
The rst is a kline, or couch,36 possibly big and sturdy enough to hold the
nine men (not the usual one or two to a couch) who are invited in the
opening line to come to me . . . and lie down. In this connection, M.
35. Much less with an object ve feet thick ([] ), the supplement pro-
posed in BG (ad loc.), discussed what follows.
36. In a similar direction, cf. Luppe 2002. Luppe suggests that because the epigram be-
longs to the , the speaker must be a boulder with three benchlike outcroppings capable
of holding three men each. However, this suggestion, with its bold supplements, lacks plausi-
bility inasmuch as it is not supported by a shred of archaeological evidenceare there ex-
amples of sympotic furniture built thus into a natural outdoor setting?or by any suggestion
as to where such a structure might have been. Cf. further Prioux (2007b: 12729).
Fantuzzi notes that we should reject the editors supplement []

(v. 5), as it ignores the central caesuraan exceedingly rare occur-
rence in the history of epigram, and one that we should certainly not cre-
ate through conjecture.37 He suggests instead / [ ] , a
width of ve men, a characterization thatin addition to observing the
caesuraaptly describes a couch large enough along its length and
breadth to carry nine men. The second possibility is a massive table per-
haps set up in the conventional manner in front of the couch where the
men are reclining.38 Couch or table, I agree with the editors that [ at
the end of line 2 suggests that stone or gems gured in the construction.
Now where would such a grand couch or table be found? In a setting, I
suggest, that evokes a heroic context. A cue that prompts us in this direc-
tion may lie in the appeal to nine men, 6 . This is, rst of all,
39
a Homeric combination. More importantly, it recalls a passage from the
Iliad (7.161) where, in response to Nestors scolding challenge, nine great
heroes spring up to volunteer to ght Hector: / 6 .
Signicantly, this line is quoted verbatim in a sympotic context in Hip-
polochus of Macedons description of the sumptuous wedding feast of
Caranus the Macedonian (Athen. 4.129f ). There, after one of the feasters
empties, at a single draught, a , a capacious six-pint cup of
barely diluted Thasian wine ( : 4 6
a 6), the host responds by offering to give the cup as a Ho-
meric-style to all who can match this feat. At these words, contin-
ues the narrator, citing Homer, all nine sprang up and seized a cup, each
striving to outstrip the others (6/ / 6
Z ). In other words, the con-
vivial agon is made parallel to the heroic agon. I suspect that, in view of the
Macedonian authorship and subject of this description, the Homeric
phrase may have had some currency in Macedonian circles, including that
of the Ptolemies. With its nine men, our epigram similarly evokes the
Homeric model and thereby lends a touch of epic grandeur to its imagined
symposium. At the same time, of course, it humorously undercuts that
grandeur by inverting the Iliadic (the nine sprang up) with
(nine men, lie down, v. 1).
37. The same goes for Luppes proposal (ibid.), [+] .

38. In that case, we might consider whether in v. 2 might refer to the standard
three-legged sympotic table and whether might be part of a plausible supplement in
the gap following .
39. It appears just once in Homer (Il. 16.785), of Patroklos nal murderous assault,
/ 6 (.
Larger than life opulence may be evoked in another way as well. De-
spite its tattered condition, the epigram strikes us at once by its profusion
of numbers and measurements (6 , v. 1; , v. 2; I, v. 4;
/ [ ], v. 5; , v. 6; , v. 7; d, v. 8). This fea-
ture immediately calls to mind the periegetic prose of the Hellenistic pe-
riod, particularly elaborate accounts of gigantic royal tents, ships, and pro-
cessionsgaudy narrative showpieces obsessed with quantity and size40
and meant to convey an impression of endless wealth and power.
Oswyn Murray has emphasized the importance in the development of
the Hellenistic royal symposium of the Persian model, combined with
Macedonian customs of royal feasting. He stresses the central importance
of tryph and feasting in the life of the [Persian] king, recognized as early
as Herodotus, and notes how Alexanders banquets were certainly com-
pared to those of the Persian king in this regard.41 Alexanders lavish
feasting pavilion (Athen. 12.538b539a) was in turn the clear model for the
great banquet tent of Ptolemy Philadelphus described by Callixenus of
Rhodes (Athen. 5.196a197c). That tent, with its 130 gold couches
(5.197a), each with a pair of three-legged gold tables set before it, seems
deliberately to echoand outdoAlexanders furnishings, which included
only a hundred couches of silver: the sole couch of gold belonged to
Alexander himself. Most signicantly for our epigram, in Ptolemys tent
there was a couch (), in full sight of the sympotic assembly, that
could hold a great pile of goblets, cups, and utensilsall of gold, studded
with jewels (` V , ),
and weighing ten thousand silver talents (5.197c). The text does not say of
what material this kline was made, but to bear such a loadfully three
hundred tonsit seems a fair bet that it must have been of stone.42 If the
object described in our epigram is a kline, it was doubtless more modest
than this one. But I think that in exploring the context of royal symposia
40. A verse counterpart might be the description of the statue of Olympian Zeus in Cal-
limachus Iamb. 6. Cf. I. Petrovic 2006; Prioux 2007b: 12729.
41. O. Murray 1996: 1819. It is worth notingespecially in light of Posidippus inter-
est in the ring of Polycrates, mentioned earlierthat a Greek model for the display of deluxe
furniture may also be at work here. A tantalizingly brief notice in Herodotus (3.123) tells us
that Polycrates secretary, Maeandrius, following his masters death, dedicated in the Heraion
of Samos all the furniture from the tyrants banquet halla sight worth seeing, Herodotus
says: 6
6 6
K.
42. Cf. Cleopatras for Antony, 6 ME
6 (Athen. 4.147f ). In this case, the use of stone inlay ap-
pears to have extended to the furniture.
Hellenistic . . . gigantosymposia, as Bergquist has called them (1990: 53)

we are looking in the right place.
A comparable royal context would be suitable if we consider the object
in our poem a table, for we have a detailed source describing Ptolemy
Philadelphus special interest in the manufacture and exquisite craftsman-
ship of a gem-encrusted table made of gold. I am referring to the sumptu-
ous table said to have been given by the king to the high priest of
Jerusalem, Eleazar, and minutely described in an extended ecphrasis in the
Letter of Aristeas (5172). To be sure, its dimensions were smaller than what
our epigram describes: , + a
? (57), that is, roughly three feet long by two feet high. Ptolemy had
hoped to make it colossal in size ( + " C 1
# , 52), but naturally he
bowed to biblical authority in this regard. What the table lacked in size,
however, it more than made up for in its art and precious material. These
are lovingly set out in the elaborate ecphrasis already mentioned, which
like our epigramrepeatedly stresses the dimensions of each detail. I pro-
pose, then, that we may have here in verse a counterpart to the sensational
prose descriptionsthe Hellenistic answer to the tabloids of todayof
spectacular objects on display at royal symposia, quite possibly Ptolemaic
symposia.
To conclude, it is worth noting that the Ptolemaic orientation revealed
in this section on Stones recurs with far greater emphasis in that on dedi-
cations (); that on victories in equestrian contests (),
where numerous epigrams celebrate the institutions and achievements of
the rst three generations of Ptolemaic kings and queens; and that on
cures (,), which seems to take its cue from the interests of a high
Ptolemaic ofcial.43 Posidippus was evidently active in quite disparate
parts of the Greek world and served a variety of masters. We know, from
an inscription in Thermon, that in 263/2 he received proxeny from the
Aetolian League in his function as writer of epigrams (6,
IG IX 12; 17 A test. 3 AB); and he appears to come up in a proxeny list
from the 270s at Delphi as well (Fouilles de Delphes III 3 no. 192 test. 2
AB). Further, he seems to have maintained close ties to his native Pella in
Macedonia (cf. his poetic sphragis, SH 705 118 AB). The remarkable
emphasis on Ptolemaic themes in the Milan papyrusto the exclusion of
almost any other regional powers political interestsstrongly suggests,
43. Cf. Bing 2002a; 2003a; and the chapter on Posidippus Iamatika in the present
volume.
then, that the epigrams in this collection were selected with a Ptolemaic
audience in mind.44 An apparent exception, like epigram 83 AB, commem-
orating a victory of one of the Thessalian Scopadae in a horse race at
Olympia, in fact just adds luster to the Ptolemaic victories so prominently
described by evoking the precedent of Simonidean epinician (cf., similarly,
Theocr. 16.34ff.). Bearing in mind the striking number of epigrams that
focus on womenthe rst seven poems of the ,45 almost all those
from the section on epitaphs, the prominence of Ptolemaic queens among
the and 46we may even contemplate a collection
shaped to the interests of a Ptolemaic queen or to one in her service.
44. This squares with the Ptolemaic emphasis found in the previously known epigrams of
Posidippus, where such notables as the nauarch Kallikrates of Samos or Sostratus of Cnidus,
the dedicator of the statue of Zeus Soter atop the Pharos, are prominent. Cf. Between Liter-
ature and the Monuments and Posidippus and the Admiral in the present volume.
45. Cf. BG p. 25: Nel sottogruppo concernente le pietre incise (I 2III 7), prima trovi-
amo riuniti insieme tutti gli epigrammi che hanno per tema gemme intagliate offerte in dono
a donne (I 235).
46. Note especially how the last two poems of the insistently situate the achieve-
ment of these queens in a female context and according to feminine criteria. Cf. 87.34 AB:
/ 6 [] (sc., the horses of Berenike); 88.56
AB: / @ / 9,
. For the prominence of women in
the poems of the Milan papyrus generally, cf. Bernsdorff 2002: esp. 3841; Hutchinson 2002:
esp. 2.
Bibliography

Acosta-Hughes, B., E. Kosmetatou, and M. Baumbach, eds. 2004. Labored in Papy-

rus Leaves: Perspectives on an Epigram Collection Attributed to Posidippus (P. Mil.
Vogl. VIII 309). Cambridge, MA.
Acosta-Hughes, B., and S. Barbantani. 2007. Inscribing Lyric. In Bing and Bruss
2007: 42957.
Adam, J. P. 1995. Le Phare d Alexandrie. Les Dossiers dArcheologie 201: 2631.
Albert, W. 1988. Das mimetische Gedicht in der Antike: Geschichte und Typologie von
den Anfngen bis in die augusteische Zeit. Frankfurt.
Alfonsi, L. 1943. La poesia amorosa di Filita. Aegyptus 23: 16068.
Allen, T. W., W. R. Halliday, and E. E. Sikes. 1936. The Homeric Hymns. Oxford.
Ambhl, A. 2007. Tell, All Ye Singers, My Fame: Kings, Queens, and Nobility in
Epigram. In Bing and Bruss 2007: 27594.
Angi, F. 2002. Filita di Cos in bronzo (Ermesianatte, fr. 7, 7578 Powell-P. Mil.
Vogl. 309, col. X, ll. 1625). Archiv fr Papyrusforschung 48: 1724.
Asper, M. 1997. Onomata allotria: Zur Genese, Struktur und Funktion poetologischer
Metaphern bei Kallimachos. Stuttgart.
Assmann, J. 1999. Das kulturelle Gedchtnis: Schrift, Erinnerung und politische Iden-
titt in frhen Hochkulturen. Munich.
Atkinson, J. B., and D. Sices. 1996. Machiavelli and His Friends: Their Personal Cor-
respondence. DeKalb, IL.
Austin, C. 2002. Posidippus and the Mysteries . . . of the Text. In Bastianini and
Casanova 2002: 719.
Austin, C., and G. Bastianini. 2002. Posidippi Pellaei quae supersunt omnia. Milan.
Bagnall, R. S. 1971. The Ptolemaic Trierarchs. Chronique dgypte 46: 35662.
Bagnall, R. S. 1976. The Administration of the Ptolemaic Possessions outside Egypt. Leiden.
Bagnall, R. S. 1988. Greeks and Egyptians: Ethnicity, Status and Culture. In
Cleopatras Egypt. Age of the Ptolemies. Catalogue of an Exhibition at the Brook-
lyn Museum. Mainz.
273
274 Bibliography
Barber, E. A. 1936. The Lock of Berenice: Callimachus and Catullus. In Greek

Poetry and Life: Essays Presented to Gilbert Murray on His 70th Birthday, Jan.
2, 1936. 34363. Oxford.
Barchiesi, A. 2005. The Search for the Perfect Book: A PS to the New Posidippus.
In Gutzwiller 2005b: 32042. Oxford.
Barthes, R. 1974. S/Z. New York.
Bassi, K. 1989. The Poetics of Exclusion in Callimachus Hymn to Apollo. TAPA
119: 21931.
Bastianini, G., and A. Casanova, eds. 2002. Il papiro di Posidippo un anno dopo: Atti
del Convegno Internazionale di Studi Firenze 13-14 giugno 2002. Florence.
Bastianini, G., and C. Gallazzi. 1993a. Il poeta ritrovato. Ca de Sass 121: 2839.
(Cassa di Risparmio delle Provincie Lombarde, S.p.A).
Bastianini, G., and C. Gallazzi. 1993b. Posidippo Epigrammi. Milan.
Bastianini, G., and C. Gallazzi. 2001. Papiri dell Universit degli Studi di Milano.
Vol. VIII, Posidippo di Pella: Epigrammi (P. Mil. Vogl. VIII 309). Milan.
Baudy, G. 1991. Das alexandrinische Erntefest: Ein Rekonstruktionsversuch. Mit-
teilungen fr Anthropologie- und Religionsgeschichte 6: 5110.
Baumbach, M., A. Petrovic, I. Petrovic. 2009. Archaic and Classical Greek Epi-
gram. Cambridge.
Baumbach, M., and K. Trampedach. 2004. Winged Words: Poetry and Divina-
tion in Posidippus Oinoskopika. In Acosta-Hughes, Kosmetatou, and Baum-
bach 2004: 12360.
Beckby, H. 1957. Anthologia Graeca. Vol. I. Munich.
Bergquist, B. 1990. Sympotic Space: A Functional Aspect of Greek Dining-
Rooms. In O. Murray 1990: 3765.
Bernand, A. 1966. Alexandrie la Grande. Paris.
Bernand, A. 1995. Alexandrie des Ptolmes. Paris.
Bernand, A. 1996. Les veilleurs du Phare. ZPE 113: 8590.
Bernand, E. 1969. Inscriptions mtriques de lgypte grco-romaine. Paris.
Bernsdorff, H. 2002. Anmerkungen zum neuen Poseidipp (P. Mil. Vogl. VIII 309).
Gttinger Forum fr Altertumswissenschaft 5: 1145.
Berthiaume, G. 1982. Les rles du Mgeiros: Ltude sur la boucherie, la cuisine et la sac-
rice dans la Grce ancienne. Leiden.
Bettarini, L. 2003. Posidippo 37 A.-B. ( P. Mil. Vogl. VIII 309, col. VI, rr. 1825).
Sem. Rom. 6.4364.
Bing, P. 1986. The Alder and the Poet. Philetas 10 (p. 92 Powell). RhM 129:
22226.
Bing, P. 1988. The Well-Read Muse: Present and Past in Callimachus and the Hellenis-
tic Poets. Gttingen.
Bing, P. 1993a. The Bios-Tradition and Poets Lives in Hellenistic Poetry. In
Nomodeiktes: Greek Studies in Honor of M. Ostwald, ed. R. Rosen and J. Farrell,
61931. Ann Arbor, MI.
Bing, P. 1993b. Impersonation of Voice in Callimachus Hymn to Apollo. TAPA 123:
18198.
Bing, P. 1994. Aratus and His Audiences. In Mega nepios: Il destinatario nellepos
didascalico; The Addressee in Didactic Epic, ed. A. Schiesaro, P. Mitsis, and
J. Strauss Clay. Special issue MD 31: 99109.
Bibliography 275
Bing, P. 1995a. Callimachus and the Hymn to Demeter. Syllecta Classica 6:2942.
Bing, P. 1995b. Ergnzungsspiel in the Epigrams of Callimachus. A&A 41: 11531.
Bing, P. 1997. Reconstructing Berenikes Lock. In Collecting Fragments, ed. G. W.
Most, 7894. Aporemata 1. Gttingen.
Bing, P. 1998a. Between Literature and the Monuments. In Genre in Hellenistic
Poetry, ed. M. A. Harder, R. F. Regtuit, and G. C. Wakker, 2143. Hellenistica
Groningana 3. Groningen.
Bing, P. 1998b. La culture littraire dAlexandrie au IIIe sicle avant J.-C. in La
Gloire dAlexandrie. 13335. Muse du Petit Palais, Paris.
Bing, P. 2000. Text or Performance/Text and Performance: Alan Camerons Calli-
machus and His Critics. In La letteratura Ellenistica: Problemi e prospettive di
ricerca, ed. R. Pretagostini, 13948. Rome.
Bing, P. 2002a. Medeios of Olynthos, son of Lampon, and the Iamatika of Posidip-
pus. ZPE 140:297300.
Bing, P. 2002b. The Un-Read Muse? Inscribed Epigram and Its Readers in Antiq-
uity. In Hellenistic Epigrams, ed. M. A. Harder, R. F. Regtuit, and G. C. Wakker,
3966. Hellenistica Groningana 6. Louvain.
Bing, P. 2003a. Posidippus and the Admiral: Kallikrates of Samos in the Epigrams
of the Milan Posidippus Papyrus (P. Mil. Vogl. VIII 309). GRBS 43: 24366.
Bing, P. 2003b. The Unruly Tongue: Philitas of Cos as Scholar and Poet. CP 98:
33048.
Bing, P. 2004. Posidippus Iamatika. In Acosta-Hughes, Kosmetatou, and Baum-
bach 2004: 27691.
Bing, P. 2005. The Politics and Poetics of Geography in the Milan Posidippus,
Section One: On Stones (AB 120). In Gutzwiller 2005b: 11940.
Bing, P., and J. S. Bruss, eds. 2007. Brills Companion to Hellenistic Epigram.
Leiden.
Bing, P., and R. Cohen. 1991. Games of Venus: An Anthology of Greek and Roman
Erotic Verse from Sappho to Ovid. New York and London.
Bing, P., and V. Uhrmeister. 1994. The Unity of Callimachus Hymn to Artemis.
JHS 114: 1934.
Bingen, J. 2002. La victoire pythique de Callicrats de Samos (Posidippe, P. Mil
Vogl. VIII 309, XI.33XII.7). Chronique dgypte 77: 18590.
Boardman, J. 1989. Athenian Red Figure Vases: The Classical Period; A Handbook.
London.
Bonanno, M. G. 1995. Poetae ut Homeri Interpretes (Teocrito, Apollonio). Aevum
Antiquum 8:6585.
Bowie, E. L. 1985. Theocritus Seventh Idyll, Philetas, and Longus. CQ 35: 6791.
Bowman, A. K. 1986. Egypt after the Pharaohs 322 BCAD 642: From Alexander to
the Arab Conquest. London.
Braswell, B. K. 1988. A Commentary on the Fourth Pythian Ode of Pindar. Berlin.
Bremer, J. M. 1990. Pindars Paraxodical 6 and a Recent Controversy about the
Performance of His Epinicia. In The Poets I in Archaic Greek Lyric, ed. S. R.
Slings, 4158. Amsterdam.
Brun, P. 1991. Les Lagides Lesbos: Essai de chronologie. ZPE 85: 99113.
Bruss, J. S. 2005. Hidden Presences: Monuments, Gravesites, and Corpses in Greek Fu-
nerary Epigram. Hellenistica Groningana 10. Louvain.
276 Bibliography
Buchholz, H. G. 1975. Methymna: Archologische Beitrge zur Topographie und Ge-

schichte von Nordlesbos. Mainz.
Bulloch, A. W. 1977. Callimachus Erysichthon, Homer, and Apollonius. AJPh 98:
97123.
Bulloch, A. W. 1985a. Callimachus: The Fifth Hymn. Cambridge.
Bulloch, A. W. 1985b. Hellenistic Poetry. In The Cambridge History of Classical
Literature, eds. P. E. Easterling and B. M. W. Knox. 541621. Cambridge.
Burkert, W. 1979. Kynaithos, Polycrates, and the Homeric Hymn to Apollo. In Ark-
touros: Hellenic Studies presented to Bernard M. W. Knox on the Occasion of His
65th Birthday, eds. G. W. Bowersock, W. Burkert, and M. C. J. Putnam, 5362.
New York.
Burkert, W. 1983. Homo Necans. Berkeley.
Burkert, W. 1985. Greek Religion. Cambridge, MA.
Burnett, A. P. 1989. Performing Pindars Odes. CP 84: 28393.
Cahen, . 1929. Callimaque et son oeuvre potique. Paris.
Cahen, . 19493. Callimaque. Paris.
Cairns, F. 1979. Tibullus: A Hellenistic Poet at Rome. Cambridge.
Calame, C. 1993. Legendary Narration and Poetic Procedure in Callimachus
Hymn to Apollo. In Callimachus, ed. M. A. Harder, R. F. Regtuit, and G. C.
Wakker, 3755. Hellenistica Groningana 1. Groningen.
Caldern Dorda, E. 1990. Ateneo y la de Filetas. Emerita 58: 12529.
Cameron, A. 1965. Wandering Poets: A Literary Movement in Byzantine Egypt.
Historia 14: 470509.
Cameron, A. 1993. The Greek Anthology from Meleager to Planudes. Oxford.
Cameron, A. 1995. Callimachus and His Critics. Princeton.
Cameron, A. 2000. The Poet, the Bishop, and the Harlot. GRBS 41: 17588.
Campbell, D. A. 1991. Greek Lyric. Vol. III. Cambridge, MA.
Canfora, L. 1989. The Vanished Library. Berkeley.
Capra, F. 1972. The Name above the Title: An Autobiography. London.
Carey, C. 1989. The Performance of the Victory Ode. AJPh 110: 54565.
Carey, C. 1991. The Victory Ode in Performance: the case for the chorus. CP 86:
192200.
Casson, L. 1974. Travel in the Ancient World. Toronto.
Castrn, P., and H. Lilius. 1970. Grafti del Palatino II: Domus Tiberiana. Acta Insti-
tuti Finlandiae 4.
Chamoux, F. 1975. Lpigramme de Poseidippos sur le Phare d Alexandrie. In Le
Monde Grec: Pense, littrature, histoire, documents. Hommages Claire Praux,
ed. J. Bingen G. Cambier, and G. Nachtergael, 21422. Brussels.
Clairmont, C. W. 1970. Gravestone and Epigram: Greek Memorials from Archaic and
Classical Period. Mainz.
Clausen, W. 1970. Catullus and Callimachus. HSCP 74: 8594.
Clay, D. 2004. Archilochos Heros: The Cult of Poets in the Greek Polis. Cambridge,
MA.
Clay, J. S. 1989. The Politics of Olympus: Form and Meaning in the Major Homeric
Hymns. Princeton.
Collard, C. 1975. Euripides. Supplices. Groningen.
Bibliography 277
Conner, W. R. 1988. Seized by the Nymphs: Nympholepsy and Symbolic Expres-

sion in Classical Greece. CA 7: 15589.
Conze, A. 18931922. Die attischen Grabreliefs. Vols. IIV.
Courtney, E. 1995. Musa Lapidaria: A selection of Latin Verse Inscriptions. Atlanta.
Cribiore, R. 1996. Writing, Teachers, and Students in Graeco-Roman Egypt. Atlanta.
Cribiore. R. 2001. Gymnastics of the Mind: Greek Education in Hellenistic and Roman
Egypt. Princeton.
Crome, J. F. 19356. ; ;. AM 6061: 300313.
Cuypers, M. P. 1997. Apollonius Rhodius: Argonautica 2.1310; A Commentary. PhD
diss., Leiden.
Day, J. W. 1989. Rituals in Stone: Early Greek Grave Epigrams and Monuments.
JHS 109: 1628.
Day, J. W. 1994. Interactive Offerings: Early Greek Dedicatory Epigrams and Rit-
ual. HSCP 96: 3774.
Day, J. W. 2000. Epigram and Reader: Generic Force as (Re-)Activation of Ritual.
In Matrices of Genre: Authors, Canons, and Society, ed. M. Depew and D. Obbink,
3757. Cambridge, MA.
Day, J. W. 2007. Poems on Stone: The Inscribed Antecedents of Hellenistic Epi-
gram. In Bing and Bruss 2007: 2947.
Deichgrber, K. 1952. Polemon 9, RE XXI, cols. 12881320.
Della Corte, F. 1951. Due Studi Catulliani. Genoa.
Denniston, J. D. 1954. The Greek Particles. Oxford.
Depew, M. 1992. , : Genre, Occasion, and Imitation in Calli-
machus, frr. 191 and 203 Pf. TAPA 122: 31330.
Depew, M. 1993. Mimesis and Aetiology in Callimachus Hymns. In Callimachus,
Hellenistica Groningana 1, ed. M. A. Harder, R. F. Regtuit, and G. C. Wakker,
5777. Groningen.
Depew, M. 1997. Reading Greek Prayers. CA 16: 22958.
Depew, M. 2004. Gender, Power, and Poetics in Callimachus Book of Hymns. In
Callimachus II, Helenistica Groningana 7, ed. M. A. Harder, R. F. Regtuit, and
G. C. Wakker, 11737. Louvain.
Dettori, E. 1999. Filita, frg. 10 Kuch. e la poesia di corteggiamento. Scholia 2: 512.
Dettori, E. 2000a. Filita Grammatico: Testimonianze e frammenti. Rome.
Dettori, E. 2000b. La lologia di Filita di Cos (con qualche osservazione sulla
lologia del III sec. a. C.). In La letteratura Ellenistica: Problemi e prospettive di
ricerca, ed. R. Pretagostini, 18398. Rome.
Deubner, L. 1921. Ein Stilprinzip hellenistischer Dichtkunst. NKA 24 (47/8):
36178.
Dickie, M. 1995. The Dionysiac Mysteries in Pella. ZPE 109: 8186.
Dickie, M. 1998. Poets as Initiates in the Mysteries: Euphorion, Philicus, and Po-
sidippus. A&A 44: 4977.
Dickie, M. 2005. The Eschatology of the Epitaphs in the New Posidippus Papyrus.
In Greek and Roman Poetry, Greek and Roman Historiography, ed. Francis Cairns,
1952. Papers of the Langford Latin Seminar 12. ARCA 44. Cambridge.
Dignas, B. 2004. Posidippus and the Mysteries: Epitymbia Read by the Ancient
Historian. In Acosta-Hughes, Kosmetatou, and Baumbach 2004: 17786.
278 Bibliography
Dilthey, C. 1876. Dipinti pompeiani accompagnati depigrammi greci. AdI 48:

294314.
Di Nino, Margherita Maria. 2005. Posidippo e la letteratura incubatoria. In Po-
sidippo e gli altri: Il poeta, il genere, il contesto culturale e letterario, ed. M. Di Marco,
B. M. Palumbo, and E. Lelli, 4776. Appunti Romani di Filologia 6. Pisa.
Di Nino, Margherita Maria. 2009. Lost at Sea: Pythermus, an anti-Odysseus. AJPh
130.1.
Dodds, E. R. 1960. Eurpides: Bacchae. Oxford.
Dombart, T. 1967. Die sieben Weltwunder des Altertums. Munich.
Dover, K. J. 1964. The Poetry of Archilochos. In Archiloque, Fondation Hardt 10.
183222. Vandoeuvres.
Dover, K. J. 1971. Theocritus: Select Poems. London.
Dyck, A. R. 1987. The Glossographoi. HSCP 91: 11960.
Ebert, J. 1972. Griechische Epigramme auf Sieger an gymnischen und hippischen Ago-
nen. Berlin.
Edelstein, E., and L. Edelstein 1998. Asclepius: A Collection and Interpretation of the
Testimonies. Baltimore.
Edwards, M. E. 1991. The Iliad: A Commentary. Cambridge.
Eichgrn, E. 1961. Kallimachos und Apollonios Rhodios. Berlin.
Ellis, R. 1878. Catulli Veronensis Liber. Oxford.
Empereur, J.-Y. 1995a. Alexandrie (gypte). BCH 119: 74360.
Empereur, J.-Y. 1995b. Le site du Phare dAlexandrie. Archologia 311: 3033.
Empereur, J.-Y. 1996a. Alexandria: The Underwater Site near Qaitbay Fort. Egypt-
ian Archaeology 8:710.
Empereur, J.-Y. 1996b. The Discovery of the Pharos of Alexandria. Minerva 7:56.
Empereur, J.-Y. 1999. Alexandrie (gypte). BCH 123: 54568.
Empereur, J.-Y. 2000. Alexandrie (gypte). BCH 124: 595619.
Empereur, J.-Y., and N. Grimal. 1997. Les Fouilles sous-marines du phare
dAlexandrie. CRAI 3: 693713.
Erbse, H. 1955. Zum Apollonhymnos des Kallimachos. Hermes 83: 41128.
Reprinted in Skiadas 1975: 276300.
Falivene, M. R. 1990. La mimesi in Callimaco: Inni II, IV, V e VI. QUCC, n.s., 36:
10528.
Fantuzzi, M. 1993. Preistoria di un genere letterario: A proposito degli Inni V e VI
di Callimaco. In Tradizione e innovazione nella cultura greca da omero all et el-
lenistica: Scritti in onore di Bruno Gentili, ed. R. Pretagostini, 92746. Rome.
Fantuzzi, M. 2000. Convenzioni epigrache e mode epigrammatiche: Lesempio
delle tombe senza nome. In La letteratura ellenistica. Problemi e prospettive di
ricerca. Seminari Romani di Cultura Greca. ed. R. Pretagostini. Quaderni 1:
16382.
Fantuzzi, M. 2002. La tecnica versicatoria del P. Mil. Vogl. VIII 309. In Bastianini
and Casanova 2002: 7997.
Fantuzzi, M. 2004a. The Structure of the Hippika in P. Mil. Vogl. VIII 309. In
Acosta-Hughes, Kosmetatou, and Baumbach 2004: 21224.
Fantuzzi, M. 2004b. Sugli epp. 37 e 74 Austin-Bastianini del P. Mil Vogl. VIII 309.
ZPE 146: 3135.
Fantuzzi, M. 2005. Posidippus at Court: On the Contribution of the Hippika of P.
Bibliography 279
Mil. Vogl. VIII 309 to the Ideology of Ptolemaic Kingship. In Gutzwiller

2005b: 24968.
Fantuzzi, M. 2007a. Mescolare il ludicro al serio: la poetica del corcirese Filica e
ledonismo dei Feaci (SH 980). In Lepigramma greca. Probleme e perspettive. ed.
G. Lozza, S. Martinelli Tempesta. 5368. Milan.
Fantuzzi, M. 2007b. Epigram and the Theater. In Bing and Bruss 2007. 47795.
Fantuzzi, M. 2008. Typologies of Variation on a Theme in Archaic and Classical
Metrical Inscriptions. In M. Baumbach, A. Petrovic, I. Petrovic. 2009.
Cambridge.
Fantuzzi, M., and R. Hunter. 2004. Tradition and Innovation in Hellenistic Poetry.
Cambridge.
Faraone, C. 1986. Callimachus Epigram 29.56 (Gow-Page). ZPE 63: 5356.
Ferguson, J. 1980. Callimachus. Boston.
Fernndez-Galiano, E. 1987. Posidipo de Pela. Madrid.
Ferrari, F. 2004. Posidippus, the Milan Papyrus, and Some Hellenistic Antholo-
gies. http://www.chs.harvard.edu/_/File/_/ferrari.pdf.
Flemberg, J. 1991. Venus Armata: Studien zur bewaffneten Aphrodite in der griechisch-
rmischen Kunst. Stockholm and Gteborg.
Foley, H. P. 1978. Reverse Similes and Sex Roles in the Odyssey. Arethusa 11: 726.
Foley, H. P. 1994. The Homeric Hymn to Demeter. Princeton.
Fordyce, C. J. 1961. Catullus: A Commentary. Oxford.
Fowler, D. 1990. Deviant focalisation in Virgils Aeneid. PCPS n.s. 36: 4263.
Fowler, D. 1997. On the Shoulders of Giants: Intertextuality and Classical Studies.
In Memory, Allusion, Intertextuality, ed. S. Hinds and D. Fowler. Special issue,
MD 39: 1334 idem, Roman Constructions: Readings in Postmodern Latin.
Oxford 2000: 11537.
Frnkel, H. 1915. De Simia Rhodio. Gttingen.
Franklin, J. L. 1991. Literacy and the Parietal Inscriptions of Pompeii. In Literacy
in the Roman World. ed M. Beard. Ann Arbor. JRA Suppl. 3: 7798.
Fraser, P. M. 1967. Current Problems Concerning the History of the Cult of Sara-
pis. Opusc. Ath. 7: 2345.
Fraser, P. M. 1972. Ptolemaic Alexandria. 3 vols. Oxford.
Fraser, P. M., and E. Matthews. 1987. A Lexicon of Greek Personal Names. Vols. III.
Oxford.
Friedlnder, P., and H. B. Hofeit. 1948. Epigrammata: Greek Inscriptions in Verse
from the Beginnings to the Persian Wars. Berkeley.
Fuhrer, T. 1988. A Pindaric Feature in the Poems of Callimachus. AJPh 109: 5368.
Fuhrer, T. 1992. Die Auseinandersetzung mit den Chorlyrikern in den Epinikien des
Kallimachos. Basel.
Gabathuler, M. 1937. Hellenistische Epigramme auf Dichter. St. Gallen.
Garland, R. 1982. A First Catalogue of Attic Peribolos Tombs. ABSA 77: 12576.
Garland, R. 1985. The Greek Way of Death. London.
Garulli, V. (Forthcoming.) Antisthenes of Paphos. In Hellenism in Cyprus from
Stasinos to Demonax, ed. I. Taifacos.
Garvie, A F. 1986. Aeschylus: Choephoroi. Oxford.
Gauly, B. 2005. Poseidipp und das Gedichtbuch. berlegungen zur Sphragis und
zum Mailnder Papyrus. ZPE 151: 3347.
280 Bibliography
Geffcken, J. 1918. . Glotta 9: 97109.

Gentili, B. 1968. Epigramma ed Elegia. In L pigramme grecque, 3981. Fondation
Hardt 14. Vandoeuvres.
Gentili, B. 1988. Poetry and Its Public in Ancient Greece: From Homer to the Fifth Cen-
tury. Baltimore.
Geus, K. 2002. Eratosthenes von Kyrene: Studien zur hellenistischen Kultur- und Wis-
senschaftsgeschichte. Munich.
Giangrande, G. 1968. Sympotic Literature and Epigram. In Lpigramme grecque,
93177. Fondation Hardt 14. Vandoeuvres.
Giangrande, G. 1980. Scripta Minora Alexandrina. Vols. IIV. Amsterdam.
Giannini, A. 1966. Paradoxographorum Graecorum Reliquiae. Milan.
Girone, M. 1998. : Guarigioni miracolose di Asclepio in testi epigraci. Bari.
Goddio, F., A. Bernand, E. Bernand, I. Darwish, Z. Kiss, and J. Yoyotte. 1998.
Alexandria: The Submerged Royal Quarters. London.
Goddio, F., and M. Clauss (ed.). 2006. Egypts Sunken Treasures. Catalogue of an ex-
hibition in Berlin 5/139/4/06. Munich.
Goldhill, S. 1991. The Poets Voice: Essays on Poetics and Greek Literature. Cambridge.
Goldhill, S. 1994. The Naive and Knowing Eye: Ecphrasis and the Culture of
Viewing in the Hellenistic World. In Art and Text in Ancient Greek Culture, ed.
S. D. Goldhill and R. G. Osborne, 197223. Cambridge.
Goldhill, S. 1999. Body/Politics: Is There a History of Reading. In Contextualizing
Classics: Ideology, Performance, Dialogue; Essays in Honor of John J. Peradotto, ed.
T. M. Falkner, N. Felson, and D. Konstan, 89120. Lanham.
Gow, A. S. F. 1950. Theocritus. 2 vols. Cambridge.
Gow, A. S. F., and A. F. Schoeld. 1953. Nicander, the Poems and Poetical Fragments.
Cambridge.
Grafton, A. 1983. Joseph Scaliger: A Study in the History of Classical Scholarship. Vol.
I. Oxford.
Grant, M. 1982. From Alexander to Cleopatra: The Hellenistic World. New York.
Green, P. 1990. Alexander to Actium: The Historical Evolution of the Hellenistic Age.
Berkeley.
Gronewald, M. 2004. Bemerkungen zu Poseidippos. ZPE 149: 4953.
Grzybek, E. 1990. Du calendrier Macdonien au calendrier Ptolmaique: Problmes de
chronologie hellnistique. Basel.
Guarducci, M. 1942. Le iscrizioni greche di Venezia. Rivista del Ist. d Arch. e. Sto-
ria dell Arte 9: 753.
Guarducci, M. 1979. Domus Musae: Epigra greche e latine in un antica casa di
Assisi. MemLinc., 8th ser., 23: 26997.
Guraud, O., and P. Jouguet. 1938. Un livre dcolier du IIIe sicle avant J.C.
Cairo.
Guichard, L. A. 2004. Asclepades de Samos: Epigramas y fragmentos. Estudio introduc-
torio, revisin del texto, traduccin y comentario. Bern.
Gutzwiller, K. 1992a. Callimachus Lock of Berenice: Fantasy, Romance, and Propa-
ganda. AJPh 113.3: 35985.
Gutzwiller, K. 1992b. The Nautilus, the Halcyon, and Selenaia: Callimachuss Epi-
gram 5 Pf. 14 G.-P. CA 11: 194209.
Gutzwiller, K. 1998. Poetic Garlands: Hellenistic Epigrams in Context. Berkeley.
Bibliography 281
Gutzwiller, K. 2004. A New Hellenistic Poetry Book: P. Mil. Vogl. VIII 309. In
Acosta-Hughes, Kosmetatou, and Baumbach 2004: 8493.
Gutzwiller, K. 2005a. The Literariness of the Milan Papyrus, or What Difference
a Book? In Gutzwiller 2005b: 287319.
Gutzwiller, K., ed. 2005b. The New Posidippus: A Hellenistic Poetry Book. Oxford.
Gutzwiller, K. 2007. A Guide to Hellenistic Literature. Oxford.
Habicht, C. 1970. Gottmenschentum und die griechischen Stdte. Munich.
Habicht, C. 1985. Pausanias Guide to Ancient Greece. Berkeley.
Hainsworth, B. 1993. The Iliad: A Commentary. Vol. III, Books 912. Cambridge.
Halleux, R., and J. Schamp. 1985. Les lapidaires grecs. Paris.
Halperin, M. 1990. One Hundred Years of Homosexuality. New York.
Hansen, P. A. 198389. Carmina Epigraphica Graeca. 2 vols. Berlin.
Harder, M. A. 1992. Insubstantial Voices: Some Observations on the Hymns of
Callimachus. CQ 42: 38494.
Hardie, A. 2003. The Statue(s) of Philitas (P. Mil. Vogl. VIII 309 Col. X 1625 and
Hermesianax fr. 7.7578 P.). ZPE 143: 2736.
Harrauer, H. 1981. Epigrammincipit auf einem Papyrus aus dem 3. Jh. v. Chr. P.
Vindob. G 40611. Ein Vorbericht. In Proceedings of the XVI International Con-
gress of Papyrology, ed. R. S. Bagnall, et al. 4953. Chico.
Harris, W. V. 1989. Ancient Literacy. Cambridge, MA.
Harris, W. V. 1996. Writing and Literacy in the Archaic Greek City. In Energeia:
Studies on Ancient History and Epigraphy presented to H. W. Pleket, ed. J. H. M.
Strube, R. A. Tybout, and H. S. Versnel, 5777. Amsterdam.
Haslam, M. 1993. Callimachus Hymns. In Callimachus, ed. M. A. Harder, R. F.
Regtuit, and G. C. Wakker, 11125. Hellenistica Groningana 1. Groningen.
Hauben, H. 1970. Callicrates of Samos: A Contribution to the Study of the Ptolemaic
Admiralty. Studia Hellenistica 18. Louvain.
Hauben, H. 1992. La chronologie macdonienne et ptolmaique mise l preuve.
Chronique dgypte 67: 14371.
Hazzard, R. A. 1987. The Regnal Years of Ptolemy II Philadelphos. Phoenix 41:
14058.
Hazzard, R. A. 2000. Imagination of a Monarchy: Studies in Ptolemaic Propaganda.
Toronto.
Headlam, W., and A. D. Knox. 1922. Herodas: The Mimes and Fragments. Cambridge.
Heath, M. 1988. Receiving the : The Context and Performance of Epini-
cian. AJPh 109: 18095.
Henrichs, A. 1993. Gods in Action: The Poetics of Divine Performance in the
Hymns of Callimachus. In Callimachus, ed. M. A. Harder, R. F. Regtuit, G. C.
Herington, J. 1985. Poetry into Drama: Early Tragedy and the Greek Poetic Tradition.
Berkeley.
Herter, H. 1929. Kallimachos and Homer: Ein Beitrag zur Interpretation des
Hymnos auf Artemis. In Xenia Bonnensia: Festschrift zum 75jhrigen Bestehen des
Philologischen Vereins und Bonner Kreises, 50105. Bonn. Reprinted in Kleine
Schriften, 371416. Munich, 1975.
Herzog, R. 1931. Die Wunderheilungen von Epidauros: Ein Beitrag zur Geschichte der
Medizin und der Religion. Philologus Suppl. 22.3. Leipzig.
282 Bibliography
Hesberg, H. von. 1981. Bemerkungen zu Architekturepigrammen des 3. Jahrhun-

derts v. Chr. JdI 96: 55119.
Heubeck, A. 1979. Schrift. in Archaeologia Homerica. Vol. III, part 10. Gttingen.
Heyworth, S. J. 2004. Looking into the River: Literary History and Interpretation
in Callimachus, Hymns 5 and 6. In Callimachus II, ed. M. A. Harder, R F. Reg-
tuit, and G. C. Wakker, 13959. Louvain.
Higbie, C. 1999. Craterus and the Use of Inscriptions in Ancient Scholarship.
TAPA 129: 4383.
Hinds, S. 1987. The Metamorphosis of Persephone: Ovid and the Self-Conscious Muse.
Cambridge.
Hinds, S. 1998. Allusion and Intertext: Dynamics of Appropriation in Roman Poetry.
Cambridge.
Hinds, S. 2004. Petrarch, Cicero, Virgil: Virtual Community in Familiares 24,4.
MD 52: 15775.
Hintzen-Bohlen, B. 1992. Herrscherreprsentation im Hellenismus. Cologne.
Hckmann, U., and D. Kreikenbom, eds. 2001. Naukratis: Die Beziehungen zu Ost-
griechenland, gypten und Zypern in archaischer Zeit. Mhnesee.
Hlbl, G. 1994. Geschichte des Ptolemerreiches Darmstadt.
Hoepfner, W. 1971. Zwei Ptolemaierbauten: Das Ptolemaierweihgeschenk in Olympia
und ein Bauvorhaben in Alexandria. MDAI(A) Beiheft 1. Berlin.
Hollis, A. S. 1990. Callimachus: Hecale. Oxford.
Hollis, A. S. 1992. The Nuptial Rite in Catullus 66 and Callimachus Poetry for
Berenice. ZPE 91: 2128.
Hopkins, K. 1993. Novel Evidence for Roman Slavery. Past and Present 138: 327.
Hopkinson, N. 1984. Callimachus: Hymn to Demeter. Cambridge.
Hopkinson, N. 1988. A Hellenistic Anthology. Cambridge.
Horrocks, G. 1997. Greek: A History of the Language and Its Speakers. London.
Hschele, R. 2007. Die bltenlesende Muse: Poetik und Textualitt antiker Epigramm-
sammlungen. Diss. Munich.
Householder, F. W., and D. W. Prakken. 1945. A Ptolemaic Grafto in New York.
TAPA 76: 10816.
Humphreys, S. C. 1980. Family Tombs and Tomb Cult in Ancient Athens: Tradi-
tion or Traditionalism. JHS 100: 96126.
Hunter, R. 1983. A Study of Daphnis and Chloe. Cambridge.
Hunter, R. 1992a. Callimachus and Heraclitus. MD 28: 11323.
Hunter, R. 1992b. Writing the God: Form and Meaning in Callimachus, Hymn to
Athena. MD 29: 934.
Hunter, R. 1996. Theocritus and the Archaeology of Greek Poetry. Cambridge.
Hunter, R. 2003a. Reecting on Writing and Culture: Theocritus and the Style of
Cultural Change. In Written Texts and the Rise of Literate Culture in Ancient
Greece, ed. H. Yunis, 21334. Cambridge.
Hunter, R. 2003b. Theocritus: Encomium of Ptolemy Philadelphus. Berkeley.
Hunter, R. 2004. Notes on the Lithika of Posidippus. In Acosta-Hughes, Kosmeta-
tou, and Baumbach 2004: 94104.
Hunter, R. 2008. Language and Interpretation in Greek Epigram. In Archaic and
Classical Greek Epigram, ed. M. Baumbach, A. Petrovic, and I. Petrovic. Cam-
bridge (Forthcoming).
Bibliography 283
Hunter, R., and T. Fuhrer. 2002. Imaginary Gods? Poetic Theology in the Hymns
of Callimachus. In Callimaque, ed. F. Montanari and L. Lehnus, 14387. Fon-
dation Hardt 48. Vandoeuvres.
Hurst, A. 1994. Contrepoints de Callimaque. MH 51: 15063.
Hu, W. 2001. gypten in Hellenistischer Zeit, 33230 v. Chr. Munich.
Hutchinson, G. O. 1985. Aeschylus: Septem contra Thebas. Oxford.
Hutchinson, G. O. 1988. Hellenistic Poetry. Oxford.
Hutchinson, G. O. 2002. The New Posidippus and Latin Poetry. ZPE 138: 110.
Ingarden, R. 1973. The Cognition of the Literary Work of Art. Evanston.
Irwin, E. 2005. Solon and Early Greek Poetry. Cambridge.
Iser, W. 1984. Der Akt des Lesens. Munich.
Jan, F. von. 1893. De Callimacho Homeri interprete. PhD diss., Strasbourg.
Janko, R. 1981. The Structure of the Homeric Hymns: A Study in Genre. Hermes
109: 924.
Janko, R. 2005. On First Looking into the New Posidippus (Ep. 64, 74 and 87
Austin-Bastianini). In : Mlanges offerts Andr Hurst, ed. A.
Kolde, A. Lukinovich, and A.-L. Rey, 12532. Geneva.
Jauss, H. R. 1982. Theory of Genres and Medieval Literature. In Toward an Aes-
thetic of Reception, 76109. Minneapolis, MN.
Jocelyn, H. D. 1980. A Greek Indecency and Its Students: Z. PCPS,
n.s. 26: 1266.
Johnson, W. 2005. The Posidippus Papyrus: Bookroll and Reader. In Gutzwiller
2005b: 7080.
Jones, C. 2001. Pausanias and His Guides. In Pausanias: Travel and Memory in
Roman Greece, ed. S. E. Alcock, J. F. Cherry, and J. Elsner, 3339. Oxford.
Kaibel, G. 1878. Epigrammata Graeca ex Lapidibus Conlecta. Berlin.
Kannicht, R. 1969. Euripides: Helena. Heidelberg.
Kppel, L. 1992. Paian: Studien zur Geschichte einer Gattung. Berlin.
Kassel, R. 1983. Dialoge mit Statuen. ZPE 51: 112.
Kassel, R. 1990. Die Phalaeceen des neuen hellenistischen Weihepigramms aus
Pergamon. ZPE 84: 299300.
Kerkhecker, A. 1991. Zum neuen hellenistischen Weihepigramm aus Pergamon.
ZPE 86: 2734.
Kirstein, R. 2002. Companion Pieces in the Hellenistic Epigram (Call. 21 and 35
Pf.; Theoc. 7 and 15 Gow; Mart. 2.91 and 2.92; Ammianos AP 11.230 and
11.231). In Hellenistic Epigrams, ed. M. A. Harder, R. F. Regtuit, and G. C.
Wakker, 11335. Hellenistica Groningana 6. Louvain.
Knox, B. M. W. 1968. Silent Reading in Antiquity. GRBS 9.4: 42135.
Knox, P. E. 1993. Philetas and Roman Poetry. Papers on the Leeds International Latin
Seminar 7: 6183.
Koenen, L. 1977. Eine agonistische Inschrift aus gypten und frhptolemische Knigs-
feste. Meisenheim.
Koenen, L. 1993. The Ptolemaic King as a Religious Figure. In Images and Ideolo-
gies: Self-Denition in the Hellenistic World, ed. A. W. Bulloch, E. S. Gruen, A. A.
Long, and A. Stewart, 25115. Berkeley.
Khnken, A. 1973. Schlupointe und Selbstdistanz bei Kallimachos. Hermes 101:
42541.
284 Bibliography
Khnken, A. 1993. Gattungstypik in kallimacheischen Weiheepigrammen. In Reli-

gio Graeca-Romana. Festschrift fr Walter Ptscher, ed. J. Dalfen, G. Petersmann
and F. F. Schwarz. Grazer Beitrge Suppl. 5: 11930.
Koster, S. 1983. Kallimachos als Apollonpriester. In Tessera: Sechs Beitrge zur Poe-
sie und poetischen Theorie der Antike, 921. Erlangen.
Krasser, H. 1995. Horazische Denkguren: Theophilie und Theophanie als Medium der
Poetischen Selbstdarstellung des Odendichters. Gttingen.
Krevans, N. 1984. Callimachus Aetia: The Poet as Editor. PhD diss., Princeton.
Krevans, N. 2005. The Editors Toolbox: Strategies for Selection and Presentation
in the Milan Epigram Papyrus. In Gutzwiller 2005b: 8196.
Krevans, N. 2007. The Arrangement of Epigrams into Collections. In Bing and
Bruss 2007: 13146.
Krischer, T. 1977. Die enkomiastische Topik im Epitaphios des Perikles (zu Thuk.
II 35f.). Mnemosyne 30: 12234.
Kroll, W. 1968. C. Valerius Catullus, herausgegeben und erklrt. Stuttgart.
Kuchenmller, G. 1928. Philetae Coi Reliquiae. Bern.
Khner, R. 189092. Ausfhrliche Grammatik der griechischen Sprache. Erster Teil
(Elementar und Formenlehre). 2 vols. 3 Au. besorgt von F. Blass. Hanover.
Kuiper, K. 189698. Studia Callimachea. Leiden.
Kumpf, M. 1984. Four Indices of the Homeric Hapax Legomena. Hildesheim.
Kurke, L. 1993. The Economy of Kudos. In Cultural Poetics in Archaic Greece: Cult,
Performance, Politics, ed. C. Dougherty and L. Kurke, 13163. Cambridge.
Kyriakou, P. 1995. Homeric Hapax Legomena in the Argonautica of Apollonius Rhodius.
A Literary Study. Stuttgart.
Labarre, G. 1996. Les cits de Lesbos aux poques hellnistique et imprial. Lyon.
Lapini, W. 2003. Note posidipee. ZPE 143: 3952.
Lapp, F. 1965. De Callimachi Cyrenaei Tropis et Figuris. Bonn.
Laronde, A. 1987. Cyrne et la Libye Hellnistique. Paris.
Larson, J. 1997. Astacides the Goatherd. CP 92: 13137.
Latacz, J. 1985. Das Plappermulchen aus dem Katalog. In Catalepton: Festschrift
fr Bernhard Wyss zum 80. Geburtstag. ed. C. Schublin, 7795. Basel.
Lattimore, R. 1962. Themes in Greek and Latin Epitaphs. Urbana.
Laurens, P. 1989. Labeille dans lambre: Clbration de lpigramme de lpoque alexan-
drine la n de la Renaissance. Paris.
Lausberg, M. 1982. Das Einzeldistichon. Munich.
Leach, E. W. 1988. The Rhetoric of Space: Literary and Artistic Representations of
Landscape in Republican and Augustan Rome. Princeton.
Lebek, W. D. 1990. Neue Phalaikeen aus Pergamon. ZPE 82: 29798.
Lee, G. 1991. The Poems of Catullus. Oxford.
Lefkowitz, M. 1985. Pindars Pythian V. In Pindare, 3267. Fondation Hardt 31.
Geneva.
Lefkowitz, M. 1988. Who Sang Pindars Victory Odes. AJPh 109: 111.
Lefkowitz, M., and M. Heath. 1991. Epinician Performance. CP 86: 17391.
Lehnus, L. 1996. On the Metrical Inscription Found at Pergamum (SEG 39.1334).
CQ 46.1: 29597.
Lelli, E. 2004. Posidippo e Callimacho. In Posidippo e gli altri: Il poeta, il genere, il
Bibliography 285
contesto culturale e letterario, ed. M. Di Marco, B. M. Palumbo, and E. Lelli, 77

132. Appunti Romani di Filologia 6. Pisa.
Lichtheim, M. 1980. Ancient Egyptian Literature. Vol. III. Berkeley.
LiDonnici, L. 1995. The Epidaurian Miracle Inscriptions: Text, Translation, and Com-
mentary. Atlanta.
Lidov, J. 2002. Sappho, Herodotus, and the Hetaira. CP 97: 20337.
Lilibaki-Akamati, M. 1989. / . /. P. .
. 3: 91101.
Livrea, E. 1973. Apollonii Rhodii Argonauticon Liber IV. Florence.
Livrea, E. 1992. L Epitao Callimacheo per Batto. Hermes 120: 29198.
Lloyd-Jones, H. 2003. All by Posidippus? In Des Gants Dionysos: Mlanges de
mythologie et de posie grecques offerts Francis Vian, ed. D. Accorinti and P. Chu-
vin, 27780. Alessandria.
Lobel, E., ed. 1952. The Oxyrhynchus Papyri. Vol. XX. London.
Longega, G. 1968. Arsinoe II. Rome.
Ludwig, W. 1968. Die Kunst der Variation im hellenistischen Liebesepigramm. In
Lpigramme grecque, 299334. Fondation Hardt 14. Vandoeuvres.
Lumpp, H.-M. 1963. Die Arniadas-Inschrift aus Korkyra: Homerisches im Epi-
grammEpigrammatisches in Homer. Forschungen und Fortschritte 37.7: 21215.
Luppe, W. 2001. Weitere berlegungen zu Poseidipps -Epigramm Kol. III
14ff. Archiv fr Papyrusforschung 47: 25051.
Luppe, W. 2002. Ein gastlicher Stein: Poseidipp, Epigramm Kol. III 2027 (P. Mil.
Vogl. VIII 309). MH 59: 14244.
Luppe, W. 2003. Ein Weih-Epigramm Poseidipps auf Arsinoe. Archiv fr Papyrus-
forschung 49: 2124.
Macan, R. W. 1908. Herodotus: The Seventh, Eighth, and Ninth Books. London.
Magnelli, E. 1999. Alexandri Aetoli Testimonia et Fragmenta: Introduzione, edizione
critica, traduzione e commento. Florence.
Magnelli, E. 2005. Fortuna del nuovo Posidippo nella poesia imperiale. In Posidippo
e gli altri: Il poeta, il genere, il contesto culturale e letterario, ed. M. Di Marco, B. M.
Palumbo, and E. Lelli, 21727. Appunti Romani di Filologia 6. Pisa.
Malten, L. 1911. Kyrene. Berlin.
Mnnlein-Robert, I. 2007. Stimme, Schrift und Bild: Zum Verhltnis der Knste in der
hellenistischen Dichtung. Heidelberg.
Marinone, N. 1984. Berenice da Callimaco a Catullo. Rome.
Markwald, G. 1986. Die Homerischen Epigramme. Meisenheim.
Martino, F. de. 1981. Un canto bello ma insolito in Esiodo e nellinno omerico
ad Apollo. A&R 26: 4043.
Masson, O. 1976. Le nom de Battos, fondateur de Cyrne, et un groupe de mots
grecs apparents. Glotta 54: 8498.
McLennan, G. R. 1977. Callimachus: Hymn to Zeus. Rome.
Meillier, C. 1979. Callimaque et son temps. Lille.
Meyer, D. 2005. Inszeniertes Lesevergngen: Das inschriftliche Epigramm und seine
Rezeption bei Kallimachos. Stuttgart.
Miller, A. M. 1986. From Delos to Delphi: A Literary Study of the Homeric Hymn to
Apollo. Leiden.
286 Bibliography
Mineur, W. H. 1984. Callimachus: Hymn to Delos. Leiden.

Mitford, T. B. 1961. The Hellenistic Inscriptions of Old Paphos. Ann. Brit. School
Athens. 56: 141.
Mitford, T. B. 1971. The Inscriptions of Kourion. Philadelphia.
Mller, A. 2000. Naukratis: Trade in Archaic Greece. Oxford.
Mooren, L. 1975. The Aulic Titulature in Ptolemaic Egypt: Introduction and Prosopog-
raphy. Brussels.
Moreno, B. 2000. The Statue of Liberty Encyclopedia. New York.
Moretti, L. 1953. Iscrizioni agonistiche grechi. Rome.
Morgan, T. 1998. Literate Education in the Hellenistic and Roman Worlds. Cambridge.
Mller, C. W. 1987. Erysichthon: Der Mythos als narrative Metapher im Demeterhym-
nos des Kallimachos. Wiesbaden.
Mller, C. W. 1990. Philetas oder Philitas? In Beitrge zur hellenistischen Literatur
und ihrer Rezeption in Rom., ed. P. Steinmetz 2737. Palingenesia 28. Stuttgart.
Mller, H. 1989. Ein neues hellenistisches Weihepigramm aus Pergamon. Chiron
19: 499553.
Murray, J. 2004. The Metamorphosis of Erysichthon: Callimachus, Apollonius,
and Ovid. In Callimachus II, Hellenistica Groningana 7, ed. M. A. Harder, R. F.
Regtuit, and G. C. Wakker, 20741. Louvain.
Murray, O., ed. 1990. Sympotica. Oxford.
Murray, O. 1996. Hellenistic Royal Symposia. In Aspects of Hellenistic Kingship, ed.
P. Bilde, T. Engberg-Pedersen, L. Hannestad, and J. Zahle, 1527. Studies in
Hellenistic Civilization 7. Arhus.
Naeke, A. F. 1845. Opuscula Philologica. Volumen II. Callimachi Hecale. Bonn.
Nagy, G. 1990. Greek Mythology and Poetics. Ithaca.
Nagy, G. 2005. Homeric Echoes in Posidippus. In Acosta-Hughes, Kosmetatou,
and Baumbach 2004: 5764. Cambridge, MA.
Nenna, M.-D. 1998. Gemmes et pierres dans le mobilier Alexandrin. In La Gloire
dAlexandrie, 15661. Paris.
Nisbet, R. G. M., and Margaret Hubbard. 1970. A Commentary on Horace: Odes
Book 1. Oxford.
Norsa, M., and G. Vitelli. 1934. di poemi di Callimaco in un papiro di Teb-
tynis. Florence.
Nussbaum, M. C. 1993. Poetry and the Passions: Two Stoic Views. In Passions and
Perceptions, ed. J. Brunschwig and M. C. Nussbaum, 97149. Cambridge.
Page, D. L. 1955. Sappho and Alcaeus. Oxford.
Page, D. L. 1962. Greek Literary Papyri. Cambridge, MA.
Page, D. L., ed. 1981. Further Greek Epigrams. Cambridge.
Panofsky, E. 1955. Meaning in the Visual Arts. Garden City, NY.
Pantos, P. A. 1987. Brnice II Dmter. BCH 111: 34352.
Parsons, P. 1977. Callimachus: Victoria Berenices. ZPE 25: 150.
Parsons, P. 1993. Poesia Ellenistica: Testi e Contesti. Aevum Antiquum 5: 919.
Pearce, T. E. V. 1983. The Tomb by the Sea: The History of a Motif. Latomus 42:
11015.
Peek, W. 1935. Eine Herme des Hipparch. Hermes 70: 46163.
Peek, W. 1953. Poseidippos von Pella. RE 22.1: 42846.
Peek, W. 1955. Griechische Vers-Inschriften. Vol. I, Grab-Epigramme. Berlin.
Bibliography 287
Peek, W. 1957. Antipater von Sidon und Antisthenes von Paphos. Philologus 101:
10112.
Peek, W. 1960. Griechische Grabgedichte. Darmstadt.
Peek, W. 1975. Steinkopie eines Theokritepigramms aus der Renaissance. ZPE 18:
26.
Petrain, D. 2003. Homer, Theocritus, and the Milan Posidippus (P. Mil. Vogl. VIII
309 Col. III 2841). CJ 98: 35988.
Petrovic, A. 2007a. Inscribed Epigram in Pre-Hellenistic Literary Sources. In Bing
and Bruss 2007: 4968.
Petrovic, A. 2007b. Kommentar zu den Simonideischen Versinschriften. Leiden.
Petrovic, I. 2006. Delusions of Grandeur: Homer, Zeus, and the Telchines in Cal-
limachus Reply (Aitia Fr. 1) and Iambus 6. A&A 52: 1641.
Petrovic, I. 2007. Von den Toren des Hades zu den Hallen des Olymp: Artemiskult bei
Theokrit und Kallimachos. Leiden.
Pfeiffer, R. 1932. . Philologus 87: 179228. Reprinted
in Skiadas 1975: 100152.
Pfeiffer, R. 1968. History of Classical Scholarship. Oxford.
Pfohl, G. 1968. Inschriften in der Anthologia Graeca. Euphrosyne 2: 15560.
Plantzos, D. 1999. Hellenistic Engraved Gems. Oxford.
Powell, B. 1991. Homer and the Origin of the Greek Alphabet. Cambridge.
Powell, J. U. 1925. Collectanea Alexandrina. Oxford.
Praux, C. 1978. Le Monde Hellnistique. 2 vols. Paris.
Preger, T. 1891. Inscriptiones Graecae Metricae ex Scriptoribus praeter Anthologiam
Collectae. Leipzig.
Prioux, . 2002. Quatre pigrammes de lAnthologie dans une collection pi-
graphique de Venise. RPh 76: 10811.
Prioux, . 2007a. Deux jeux de mots sur le nom dAratos: Note sur Virgile, B. III,
42 et Aratos, Phaen. 2. RPh 79: 30917.
Prioux, E. 2007b. Regards Alexandrins: Histoire et thorie des arts dans lpigramme
hellnistique. Louvain.
Pucci, P. 1994. Gods Intervention and Epiphany in Sophocles. AJPh 115:1546.
Puelma, M. 1996. /epigramma: Aspekte einer Wortgeschichte. MH
53: 12339.
Puelma, M. 2006. Arions Delphin und die Nachtigall: Kommentar zu Poseidippos
ep. 37 A.-B. ( P. Mil. Vogl. VIII 309, Kol. VI 1825). ZPE 156: 6074.
Putnam, M. C. J. 1960. Catullus 66.7588. CP 55: 22328.
Radl, A. 1988. Der Magnetstein in der Antike: Quellen und Zusammenhnge. Stuttgart.
Rasch, G. 1910. De Anthologiae Graecae epigrammatis quae colloquii formam habent.
PhD diss., Mnster.
Raubitschek, A E. 1968. Das Denkmal-Epigramm. In Lpigramme grecque, 326.
Fondation Hardt 14. Geneva.
Reid, J. 1885. M. Tulli Ciceronis Academica. London.
Reitzenstein, R. 1893. Epigramm und Skolion. Gieen.
Reitzenstein, R. 1908. Horaz und die hellenistische Lyrik. NJA 21: 81102.
Rengakos, A. 1992. Homerische Wrter bei Kallimachos. ZPE 94: 2147.
Rengakos, A. 1993. Der Homertext und die hellenistischen Dichter. Hermes Einzel-
schriften 64. Stuttgart.
288 Bibliography
Rengakos, A. 1994a. Apollonios Rhodios und die antike Homererklrung. Zetemata 92.
Munich.
Rengakos, A. 1994b. Lykophron als Homererklrer. ZPE 102: 11130.
Richardson, N. J. 1974. The Homeric Hymn to Demeter. Oxford.
Ridgway, D. 1992. The First Western Greeks. Cambridge.
Riese, A. 1881. Geographica. RhM 36: 20614.
Robert, L. 1960. Sur un dcret des Korsiens au muse de Smyrne. Hellenica 11
12: 13276.
Robert, L. 1966. Sur un dcret dIlion et sur un papyrus concernant des cultes
royaux. In Essays in Honor of C. Bradford Welles, 175211. American Studies in
Papyrology 1. New Haven.
Rosenmeyer, P. A. 1997. Her Masters Voice: Sapphos Dialogue with Homer. In
Memory, Allusion, Intertextuality, ed. S. Hinds and D. Fowler. Special issue, MD
39: 12349.
Rsler, W. 1980. Dichter und Gruppe: Eine Untersuchung zu den Bedingungen und zur
historischen Funktion frher griechischer Lyrik am Beispiel Alkaios. Munich.
Rossi, L. 1996. Il Testamento di Posidippo e le laminette auree di Pella. ZPE 112:
5965.
Rossi, L. 2001. The Epigrams Ascribed to Theocritus: A Method of Approach. Hellenis-
tica Groningana 5. Louvain.
Rossi, L. 2002. Composition and Reception in AP 9.1583: Aphegheseis, Epideix-
eis and Progymnasmata. In Hellenistic Epigrams, ed. M. A. Harder, R. F. Reg-
tuit, G. C. Wakker, 15174. Hellenistica Groningana 6. Louvain.
Rouse, W. H. D. 1902. Greek Votive Offerings. Cambridge.
Rusten, J. 2002. Theophrastus: Characters. Loeb Classical Library. Cambridge, MA.
Sbardella, L. 2000. Filita: Testimonianze e frammenti poetici. Introduzione, edizione e
commento. Rome.
Scaliger, 1864. Poemata Omnia. Berlin.
Schmidt, E. A. 1987. Bukolische Leidenschaft. Frankfurt.
Schmidt, E. A. 1990. Horatianum. WS 103: 5798.
Schneider, O. 1873. Callimachea. Vol. II. Leipzig.
Schneidewin, F. C. 1838. Delectus poetarum elegiacorum Graecorum. Gttingen.
Schrder, S. 2004. Skeptische berlegungen zum Mailnder Epigrammpapyrus (P.
Mil. Vogl. VIII 309). ZPE 148: 2973.
Schroeder, C. (forthcoming). New Inscriptions on Old ShellsCallimachus
Nautilus Epigram (5 Pf. 14 G.-P.) and the Aesthetics of Inscribed Seashells.
Schur, D. 2004. A Garland of Stones: Hellenistic Lithika as Reections on Poetic
Transformation. In B. Acosta-Hughes, M. Baumbach and E. Kosmetatou 2004:
11822.
Schwinge, E.-R. 1985. Wilamowitz und das Verstndnis der hellenistischen
Poesie. In Wilamowitz nach 50 Jahren, ed. W. M. Calder et al., 15177.
Darmstadt.
Scodel, R. 1992. Inscription, Absence, and Memory: Epic and Early Epitaph. SIFC
10: 5776.
Scodel, R. 2003. A Note on Posidippus 63 AB (P. Mil. Vogl. VIII 309 X 1625).
ZPE 142: 44.
Seaford, R. 1984. Euripides: Cyclops. Oxford.
Seeck, G. A. 1975. Dichterische Technik in Theokrits Thalysien und die Theo-
Bibliography 289
rie der Hirtendichtung. In , Hans Diller zum 70. Geburtstag, ed. K. I.

Vourveris and A. Skiadas. 195209. Athens.
Segal, C. 1984. Underreading and Intertextuality: Sappho, Simaetha, and Odys-
seus in Theocritus Second Idyll. Arethusa 17: 2019.
Selden, D. 1998. Alibis. CA 17: 290412.
Sens, A. 2002. The New Posidippus, Asclepiades, and Hecataeus Philitas-Statue.
http://www.apaclassics.org/Publications/Posidippus/SensPosidippus.pdf.
Sens, A. 2005. The Art of Poetry and the Poetry of Art: The Unity and Poetics of
Posidippus Statue-Poems. In Gutzwiller 2005b: 20625. Oxford.
Sens, A. 2007. One Thing Leads (Back) to Another: Allusion and the Invention of
Tradition in Hellenistic Epigrams. In Bing and Bruss 2007: 37390.
Sickinger, J. P. 1999. Public Records and Archives in Classical Athens. Chapel Hill.
Sider, D. 2005. Posidippus on Weather Signs and the Tradition of Didactic Poetry.
In Gutzwiller 2005b: 16482.
Sider, D. 2007. Sylloge Simonidea. In Bing and Bruss 2007: 11330.
Skiadas, A. D., ed. 1975. Kallimachos. Wege der Forschung 296. Darmstadt.
Skinner, Marilyn B. 1993. Woman and Language in Archaic Greece, or Why Is
Sappho a Woman? In Feminist Theory and the Classics, ed. N. S. Rabinowitz and
A. Richlin, 12544. New York.
Smith, E. M. 1926. Naukratis: A Chapter in the History of the Hellenization of Egypt.
Vienna.
Smith, M. 2004. Elusive Stones: Reading Posidippus Lithika through Technical
Writing on Stones. In Acosta-Hughes, Kosmetatou, and Baumbach 2004:
10517.
Sourvinou-Inwood, C. 1995. Reading Greek Death: To the End of the Classical Period.
Oxford.
Spanoudakis, K. 2002. Philitas of Cos. Leiden.
Stein, A. 1931. Rmische Inschriften in der antiken Literatur. Prague.
Steiner, D. T. 1994. The Tyrants Writ: Myths and Images of Writing in Ancient
Greece. Princeton.
Stephens, S. 2003. Seeing Double: Intercultural Poetics in Ptolemaic Alexandria.
Berkeley.
Stephens, S. 2004. For you, Arsinoe . . . In Acosta-Hughes, Kosmetatou, and Baum-
bach 2004: 16176.
Stephens, S. 2005. Battle of the Books. In Gutzwiller 2005b: 22948.
Sterne. 1956. Tristram Shandy. Vol. II. Everymans Library. London.
Strauss, W. A. 1995. Disjecta membra: Essays in Fragmentation; Samuel Beckett
and Jasper Johns. Dalhousie French Studies 31: 1127.
Susemihl, F. 1891. Geschichte der griechischen Literatur in der Alexanderzeit. Vol. I.
Leipzig.
Svenbro, J. 1976. La parole et le marbre: Aux origines de la potique grecque. Lund.
Svenbro, J. 1993. Phrasikleia: An Anthropology of Reading in Ancient Greece. Ithaca.
Svenbro, J. 1999. Archaic and Classical Greece: The Invention of Silent Reading.
In A History of Reading in the West, ed. G. Cavallo and R. Chartier, 3763.
Amherst.
Svensson, A. 1937. Der Gebrauch des bestimmten Artikels in der nachklassischen
griechischen Epik. Lund.
Tamaro, B. F. 1953. Il Museo Archeologico del Palazzo Reale di Venezia. Rome.
290 Bibliography
Tarn, S. L. 1979. The Art of Variation in the Hellenistic Epigram. Leiden.

Thalmann, G. 1984. Conventions of Form and Thought in Early Greek Epic Poetry.
Baltimore.
Thomas, R. 1989. Oral Tradition and Written Record in Classical Athens. Cambridge.
Thomas, R. 1992. Literacy and Orality in Ancient Greece. Cambridge.
Thomas, R. F. 1998. Melodious Tears: Sepulchral Epigram and Generic Mobility.
In Genre in Hellenistic Poetry, ed. M. A. Harder, R. F. Regtuit, and G. C.
Thomas, R. F. 2004. Drownded in the Tide: The Nauagika and some Prob-
lems in Augustan Poetry. In Acosta-Hughes, Kosmetatou, and Baumbach
2004: 25975.
Thompson, D. J. 1987. Ptolemaios and the Lighthouse. Greek Culture in the
Memphite Serapaeum. PCPS 213: 10521.
Thompson, D. J. 1988. Memphis under the Ptolemies. Princeton.
Thompson, D. J. 1992. Language and Literacy in Early Hellenistic Egypt. In Eth-
nicity in Hellenistic Egypt, ed. P. Bilde, T. Engberg-Pedersen, L. Hannestad, and
J. Zahle, 3952. Studies in Hellenistic Civilization 3. rhus.
Thompson, D. J. 2005. Posidippus, Poet of the Ptolemies. In Gutzwiller 2005b:
26983. Oxford.
Thummer, E. 1968. Pindar: Die isthmischen Gedichte. Vol. I. Heidelberg.
Tomlinson, R. A. 1983. Epidauros. Austin.
Tondriau, J. L. 1948. Princesses ptolmaiques compares ou identies des
desses (IIIeIer sicles avant J.-C.). BSAA 37: 1233.
Trmpy, C. 1997. Untersuchungen zu den altgriechischen Monatsnamen und Monatsfol-
gen. Heidelberg.
Trypanis, C. A. 1951. The Epigrams of Anacreon on Hermae. CQ 45: 3134.
Turner, E. G. 1968. Greek Papyri. Oxford.
Tzifopoulos, J. 1991. Pausanias as Stelokopas: An Epigraphical Commentary of Pau-
sanias Eliakon A and B. PhD Diss., Ohio State Univ.
Venit, M. S. 2002. Monumental Tombs of Ancient Alexandria: The Theater of the Dead.
Cambridge.
Vitelli, G., ed. 1929. Papyri della Societ Italiana. Vol. IX. Florence.
Walbank, F. W. 1972. Polybius. Berkeley.
Walsh, G. B. 1990. Surprised by Self: Audible Thought in Hellenistic Poetry. CP
85: 121.
Walsh, G. B. 1991. Callimachean Passages: The Rhetoric of Epitaph in Epigram.
Arethusa 24: 77105.
Warden, J. 2006. Catullus in the Grove of Callimachus. In Studies in Latin Litera-
ture and Roman History 13, ed. C. Deroux. 97154. Brussels.
Weber, G. 1993. Dichtung und hsche Gesellschaft: Die Rezeption von Zeitgeschichte
am Hof der ersten drei Ptolemer. Stuttgart.
Weber, L. 1917. Steinepigramm und Buchepigramm. Hermes 52: 53657.
Weil, H. 1879. Un papyrus indit de la bibliothque de M. Ambroise Firmin-Didot.
Paris.
Weinreich, O. 1909. Antike Heilungswunder. Gieen. Reprint, Berlin, 1969.
Weinreich, O. 1918. Die Heimat des Epigrammatikers Poseidippos. Hermes 53:
43439.
Bibliography 291
Welles, C. Bradford. 1934. Royal Correspondence in the Hellenistic Period: A Study in

Greek Epigraphy. London.
West, S. 1985a. Herodotus Epigraphical Interests. CQ 35: 278305.
West, S. 1985b. Venus Observed? A Note on Callimachus fr. 110. CQ 35: 6166.
West, S. 1988. A Commentary on Homers Odyssey. Vol. I, Introduction and Books
iviii. Ed. A. Heubeck, S. West, and J. B. Hainsworth. Oxford.
White, D. 1984. The Extramural Sanctuary of Demeter and Persephone at Cyrene:
Final Reports. Vol. I. Background and Introduction to the Excavations. Philadelphia.
Wifstrand, A. 1933. Von Kallimachos zu Nonnos: Metrisch-stilistische Untersuchungen.
Lund.
Wilamowitz-Moellendorff, U. von. 1879. Die Galliamben des Kallimachos und
Catullus. Hermes 14: 194201.
Wilamowitz-Moellendorff, U. von. 1906. Die Textgeschichte der griechischen Buko-
liker. Berlin.
Wilamowitz-Moellendorff, U. von. 1924. Hellenistische Dichtung in der Zeit des
Kallimachos. Vols. III. Berlin.
Wilamowitz-Moellendorff, U. von. 1925. Demeterfest. In Reden und Vortrge, I
22958. Berlin.
Wilcken, U. 1927. Urkunden der Ptolemerzeit. Vol. I. Berlin.
Williams, F. 1978. Callimachus: Hymn to Apollo. Oxford.
Winkler, J. J. 1985. Auctor & Actor: A Narratological Reading of Apuleiuss The
Golden Ass. Berkeley.
Wimann, J. 2002. Hellenistic Epigrams as School-Texts in Classical Antiquity. In
Hellenistic Epigrams, ed. M. A. Harder, R. F. Regtuit, G. C. Wakker, 21530.
Hellenistica Groningana 6. Louvain.
Wormell, D. E. W. 1966. Catullus as Translator. In The Classical Tradition: Literary
and Historical Studies in Honor of Harry Caplan, ed. L. Wallach, 187201.
Ithaca.
Wray, D. 2000. Apollonius Masterplot: Narrative Strategy in Argonautica 1. In
Apollonius Rhodius, ed. M. A. Harder, R. F. Regtuit, and G. C. Wakker, 23965.
Hellenistica Groningana 4. Louvain.
Wroth, W. 1964. Catalogue of the Greek Coins in the British Museum: Troas, Aeolis,
and Lesbos. Bologna.
Youtie, H. C. 1963. The Papyrologist: Articer of Fact. GRBS 4: 1932. Reprinted
in Scriptiunculae, I 923. Amsterdam, 1973.
Zanetto, G. 2002. Posidippo e i miracoli di Asclepio. In Un poeta ritrovato. Posidippo
di Pella. Giornata di studio, Milano, 23 novembre 2001. 7378. Milan.
Zanker, G. 1987. Realism in Alexandrian Poetry: A Literature and Its Audience.
London.
Zanker, G. 1989. Current Trends in the Study of Hellenic Myth in Early Third-
Century Alexandrian Poetry: The Case of Theocritus. A&A 35: 83103.
Zeitlin, F. I. 1990. Playing the Other: Theater, Theatricality, and the Feminine in
Greek Drama. In Nothing to Do with Dionysos? ed. J. J. Winkler and F. I. Zeitlin,
6396. Princeton.
Index of Ancient Passages Cited

(Passages treated in detail are marked in bold print.)
Aelian Antigonus of Carystus

fr. 44: 62 Mir. 8: 2627
Varia Historia 9.14: 14 Antimachus
Aeschines fr. 22.2: 20n.25
Against Ctesiphon Antisthenes of Paphos
18385: 157n.20 ID 1533: 168n.47
184: 87n.6 Anyte
187: 87n.6 19 GP: 162n.31
AP 6.330 CEG 776: 22021 Ap. Rhod.
Aeschylus 2.302: 239n.14
Cho. 1018: 31 2.375: 65
Vita 25: 219n.8 2.84250: 130n.27
Alcaeus 2.1271: 20n.25
fr. 346 Voigt: 17173 Aratus
fr. 347 LP: 152n.6, 154n.12 AP 12.129 1 GP: 12526
Alexander of Aphrodisias Phaen.
Comm. in Arist. Graeca 1.818.48: 14: 201n.22
22 78387: 110, 25455
Alkaios of Messene Archelaus
13 GP AP 7.536: 139n.45 SH 12529: 266
Anacreon Archilochus
AP 6.137: 212n.48 fr. 128.7 W: 167n.46
AP 6.138: 21114 fr. 133 W: 130n.29
PMG 416.2: 167n.46 AP 6.133 Archilochus: 246
Anthologia Planudea Archimedes
375 Anon.: 229 SH 201: 115, 207n.37
293
294 Index of Ancient Passages Cited
Archimelus 7.318d: 178n.5

SH 202: 184, 212n.50 9.401 de: 1415
Arion 9.495e: 1819
PMG 939: 248n.38 10.415a: 178n.5
Ariphron 10.442e: 87
PMG 813: 231n.40 11.483a: 2021
Aristophanes 12.538b539a: 269
Wasps 9799: 121 14.645d: 1920
Wasps 393: 162n.30 15.669c: 141
Aristotle 15.677c: 1819n.19
de Ath. rep. 7: 87n.6
Eth. Eud. 1235a9: 167 Callimachus
Eth. Nic. 1.8.14. 1099a27: 87n.6 Epigrams
History of Animals 8.607a: 218n.5 13 GP 43 Pf.: 152n.6, 16669
Metaphysics 14 GP 5 Pf.: 25152
1001b5: 265n.30 21 GP 33 Pf.: 9195; 215n.60
1091a5: 22 22.2 GP 34.2 Pf.: 98n.38
Poetics 29 GP 21 Pf.: 99102
1.1447b21: 111 30 GP 35 Pf.: 99102
4.1448b24: 153n.10 34.1 GP 2.1 Pf.: 131n.31
16.1455a: 168 36 GP 22 Pf.: 1025
24.1460a2: 111 43 GP 12 Pf.: 131n.30
Rhet. 50 GP 58 Pf.: 9799
1.7.1365a: 87n.6 56 GP 27 Pf.: 110; 254
1.9.1367b: 87n.6 Fragments (Pf.)
3.12.1413b8: 111 1.910: 11
pseudo-Aristotle 1.3235: 15n.11
mirab. auscult. 57.12 (Victoria Berenices): 9597
58 (59): 87n.6 64: 139
131: 8586, 87n.6 67: 94
133: 87n.6 75.89: 26
Asclepiades 110 (Berenikes Lock): chap. 4
16 GP AP 12.50: 17074 passim
18 GP AP 12.135: 16667 112.2: 80
31 GP AP 7.500: 131n.30 112.8: 80
43 GP APl. 120: 14n.6 186.27: 17n.16
Athenaeus 384 (Victory of Sosibios): 241
Deipnosophistae 400 Pf.: 198, 200n.21
1.9a: 168n.47 Fragments (SH )
2.83: 27 25469 (Victoria Berenices): 11112
2.137.7: 48 254.5: 200n.20
3.2: 48 Hymns
4.80: 27 1.92ff: 55n.24
4.129f: 268 2: chap. 2 passim, 63n.46
4.147f: 269n.42 2.106ff: 55n.24
5.196a97c: 269 2.10812: 63n.46, 261n.17
5.209ce: 184, 212n.50 2.11012: 63
3.256: 66 Demetrius
3.268: 55n.24 De eloc. 222: 90n.13
4.16670: 256, 259 Demosthenes
5.11: 239n.14 De corona
6: chap. 3: passim 18.289: 87n.6
pseudo-Callisthenes pseudo-Demosthenes
32.13: 200 7.39: 87n.6
Calpurnius Siculus Diog. Laert.
Ecl.1.2032: 142n.54 6.3738: 226
Catullus Dionysius
14: 114 1 GP AP 7.78: 13
50: 114, 208n.37 Dionysius of Halicarnassus
62.1: 59n.34 De comp. 25 (p.208 Reiske): 12
65.34: 68 Dioskorides
65.16: 68 AP 7.76 33 GP: 230n.38
66 (Coma Berenices): chap. 4:
passim Epicharmus
CEG I frs. 15961: 27
28: 92; 169 Eratosthenes
40: 156n.17 35 Powell: 115n.15
275.4: 91n.15 Euripides
304: 13637 Alc. 382: 164n.37
313: 21314 Bacchae
345: 91n.15 17: 259n.12
407: 91n.14 116: 182
413: 91n.14 165: 182
429: 162n.32 Helen
430: 157n.19 45: 200
454 (Ischia Cup): 126n.18, 15155, 147894: 257n.9
208n.38 IT 1082ff: 162n.30
CEG II Medea 18789: 56
512: 100102 Phoen. 302: 102n.50
593.4: 164n.37
597: 156n.17; 166n.43
755: 159 FGE
807.23: 200 Anon. 102: 87n.6
808: 221n.19 Anon. 150: 15962
818: 221n.19 FH
888.3850: 158n.21 107: 21314
Cicero 114: 93
Tusc. Disp. 5.23: 13840
CIL GG
IV 1904: 121n.9 424: 16266
IV 1906: 121n.9 Glaukos
IV 2175: 121n.9 3 GP AP 9.341: 12223n.13
IV 2461: 121n.9 Gregory of Nazianzus
IV 2487: 121n.9 AP 8.188: 102n.51
GV Hesiod
455: 229 Op. 109: 31
627: 16465n.40 Theog. 291: 31
1231: 98n.37 Hesychius
1248.12: 134 3417, I. 120 Latte: 17
1312: 156n.18 Hippocr. Epid.
1353: 131 5.46: 228
1513: 189 Homer
1620.15: 137n.42 Iliad
1625.12: 131n.31 1.250: 31
1729: 15759 1.449: 31
1735.14: 163 1.461: 31
1874.7: 229 1.465: 31
1990.5: 102n.51 1.45868: 32
2036.11: 102n.51 2.285: 31
2.42131: 32
Hedylus 2.761: 162
4 GP: 235n.3, 235n.5, 252 2.79194: 128
Hegesippus 2.81115: 12930n.27
8 GP AP 7.320: 139n.45 4.16981: 128n.24
Hephaest. 4.189: 99n.40
Isag. 4 3.4: 153n.10 6.6: 157n.19
Hermesianax 6.9092: 246
fr. 7, Powell: 2325 6.27173: 246
fr. 7.7578, Powell: 2425 6.28895: 246
fr. 7.7778, Powell: 12 6.514: 102n.50
Herodotus 7.6791: 12729
1.24: 248n. 38, 249 7.161: 268
1.73: 31 8.282: 157n.19
1.87.3: 200n.20 8.397: 200
2.16: 259 8.470: 31
2.112f: 200 9.206: 1920
2.135: 262n.19 10.375: 48
3.41: 255 10.415: 128
3.121: 255 11.1013: 19091
3.123: 269n.41 11.166: 128n.26
4.3641: 259 11.37172: 128
4.44: 259n.12 11.379: 128
4.88: 87n.6, 8788, 209 11.49297: 261n.17
5.59: 87n.6 11.63136: 15255
5.60: 87n.6 11.797: 157n.19
5.77.2: 87n.6 13.13743: 26061
7.228: 87 n.6 13.437: 122
8.22.1: 133n.34 13.515: 102n.50
8.85: 133n.34 14.1: 152
pseudo-Herodotus 15.269: 102n.50
Life of Homer 5.24: 89n.10 15.669: 157n.19
16.39: 157n.19 19.2058: 57

16.43157: 156n.17 20.49: 31
16.642: 1819 20.238: 164
16.785: 268n.39 21.203: 164
17.13336: 57n.28 22.12: 31
17.43233: 122 24.8084: 128n.23; 128n.24
18.524: 31 24.18990: 156n.17
18.55257: 1618 24.296: 156n.17
18.55859: 30n.52 Homeric Hymn to Apollo: chap. 2:
19.3012: 98n.37 passim
20.495: 31 Homeric Hymn to Demeter: chap. 3:
22.24: 102n.50 passim
23.9: 156n.17 229: 41n.26
23.326ff: 129
24.349: 128n.26 Ibycus
24.611: 162n.33 PMG 342: 188n.24
Odyssey I. Cret.
1.1: 16263 I xxii 4.A.3538: 236
1.2: 165n.40 IG
1.34: 159, 16465n.40 I 381: 21314
1.6: 165n.40 II/III ed. min., vol. 3.2 no. 7695: 101
1.6062: 165n.40 IV2 1, no. 121.19: 22223
1.261: 165n.40 IV2 1, no. 121.22: 22627
3.5153: 152 IV2 1, no. 121.33: 223n.26
3.106: 161n.28 IV2 1, no. 121.72: 223n.26
3.45463: 32 IV2 1, no. 121.95: 223n.26
4.19698: 156n.17 IV2 1, no. 121.107: 22425n.28
4.27879: 48 IV2 1, no. 121.120: 223n.26
4.385: 200 IV2 1, no. 121.122: 231n.39
4.621: 31 IV2 1, no. 121.125: 223n.26
4.693: 31 IV2 1, no. 122.7: 223n.26
4.765: 162n.30 IV2 1, no. 122.45 231n.39
6.18085: 164n.38 IV2 1, no. 122.55: 223n.26
7.224: 161n.27 IV2 1, no. 122.64: 223n.26
8.38: 26n.44 IV2 1, no. 123.115: 223n.26, 227
8.8485: 16869n.48 IV2 1, no. 123.123: 223n.26
8.92: 16869n.48 IV2 1, no. 123.129: 223n.26
8.9395: 16768 IV2 1, no. 125: 22425n.28
8.53235: 16768 IV2 1, no. 440: 223n.26
9.45: 31 IX 12 i, 17A.24: 180, 18283, 224,
11.7576: 128n.23 270
11.211: 158n.22 IX2 8, 13: 186
11.289: 31 IX2 51: 141n.50, 18589
13.131: 161n.27 XI 4, 1105: 155n.14
13.323: 161n.28 XII 2, 513: 251
14.423: 164 XII 2, 527: 251
18.300: 18n.19 XII.3, 540: 121n.9
IG (continued) Ovid
XII.6, 446: 235 Amores I 5.2325: 96n.32
XII Suppl. 115: 251 Ars am. 3.32948: 12
Isocrates Met. 11.42729: 142n.54
Panathenaicus 24647: 134n.37 Pont. 3.1.578: 24
Rem. am. 75960: 12
Leonidas Tr. 1.6.13: 24
46 GP AP 6.13: 2089
60 GP AP 7.264: 99n.39 Pausanius
61 GP AP 7.266: 99n.39 6.3.1: 259
70 GP AP 7.163: 98n.38 9.31.4ff: 191
93 GP AP 7.715: 100n.44 10.7.3: 259
101 GP AP 9.25: 110 10.33.12: 187n.22
Letter of Aristeas Phalaecus
5172: 270 2 GP AP 13.5: 162n.31
Life of Aesop Phanocles
78.1: 14243 fr. 1 Powell: 24950
Lucian Philikos of Corcyra
de historia conscribenda 62: 196n.6 SH 677: 109
Lucretius SH 980: 158n.21
6.104243: 265n.29 Philitas of Cos
Lycophron Grammatical Fragments
Alex.6: 239n.14 33 Kuchenmller/Spanoudakis
Alex.1104: 20n.25 5 Dettori: 1819
Alex.1166: 239n.14 37 Kuchenmller/Spanoudakis
Lycurgus 9 Dettori: 1920
In Leocr. 109: 87n.6 38 Kuchenmller/Spanoudakis
10 Dettori: 2021
41 Kuchenmller/Spanoudakis
Meleager 13 Dettori: 1819n.19
2 GP AP 7.417: 100n.44 46 Kuchenmller/Spanoudakis
3 GP AP 7.418: 100n.44 18 Dettori: 1718
4 GP AP 7.419: 100n.44 Poetic Fragments
10 Powell 12 Sbardella 25
Nicander Spanoudakis: 12
Alexipharmaca 35: 115n.16 16 Powell 18 Sbardella 20
Theriaca Spanoudakis: 2627
75658: 23 Pindar
15: 66 Isthmian Odes
Nonnus 3.67: 66
Dionysiaka Nemean Odes
4.23375: 14546 5.23: 263n.24
12.29115: 14345 Olympian Odes
41.33999: 145 5.23: 231n.40
Nossis 6: 253
3 GP AP 6.265: 246 Paeans
11 GP AP 7.718: 100n.44 7b: 34n.4
Pythian Odes 3.1 AB: 254

3.5558: 232n.41 4 AB: 266
4.24748: 96 4.5 AB: 257
5: 42n.29 5.1 AB: 254
Threnoi 5.2 AB: 257
129: 170n.50 6 AB: 254
Plato 7 AB: 257, 25961
Apol. 41a: 170n.50 8 AB: 254, 266
Charmides 9 AB: 250n.41, 255
164d: 120n.7 11.12 AB: 257
165a: 87n.6 13.3 AB: 257
Crat. 403a: 230 14.1 AB: 254
Critias 119c20b: 133n.33 14.6 AB: 254
Hipparchus 228d29b: 87n.6, 120n.7, 15 AB: 255
13437 16.1 AB: 254
Minos 316e: 17 16.6 AB: 254
Parm. 137 b3: 265n.30 17 AB: 26466
Phaedrus 18 AB: 26670
259bd: 15n.11 20.56 AB: 184, 25657
262de: 2 29.4 AB: 187
263e64a: 2 33.4 AB: 187
264c: 87n.6, 8890 36 AB: 24547
Protagoras 343ab: 120n.7 37 AB: 24751, 262
Resp. 39 AB: 24344
345c: 31 42 AB: 181
363cd: 170n.50 43 AB: 181
Pliny 44 AB: 18182
NH 63 AB: 11n.1, 1214, 184, 192
xxvi 83: 196n.6 65.3 AB: 187
6.171: 160n.26 74 AB: 235, 23643
20.2: 265n.29 78 AB: 184, 241n.22, 259
34.148: 242 79 AB: 184
36.68: 242 82 AB: 184, 259
37.4.8: 261 83 AB: 271
Plutarch 87 AB: 184, 259, 271n.46
Cimon 7.38.2: 157n.20 88 AB: 259, 271n.46
de E apud Delphos 385d: 120n.7 95 AB: 21718, 220, 233
de Iside et Osiride 62, 376b: 96 AB: 218n.3, 219n.9, 219n.10,
265n.29 220, 223, 223n.26
Pythia 395a: 141 97 AB: 218n.3, 219n.11, 220n.12,
Polybius 223n.26, 22729, 232
1.2.2: 259n.13 98 AB: 218n.3, 219n.11, 220,
12.10.4: 137 223n.26, 228
12.11.2: 137 99 AB: 218n.4, 219n.10, 220,
Posidippus 223n.26, 23031, 232
1.1 AB: 256 100 AB: 218n.4, 219n.11, 220,
1.4 AB: 254 223n.26, 22930, 232
Posidippus (continued ) SH
101 AB: 218n.3, 219n.8, 219n.10, 712 Ptolemy Philadelphus: 110
220, 23133 977 Epigrams for Tauron: 115,
103 AB: 22829, 232 165n.42, 2078
113 AB SH 978: 206n.34 979 Epigram for Ptolemy IV:
114 AB SH 961: 183, 196n.3 206n.34
115 AB 11 GP: 184, 187n.21, Simonides
chap. 10: passim AP 7.251.34 Simonides: 141n.50,
116 AB 12 GP: 184, 195n.2, 18889
196n.3, 236, 24344, 246n.35, PMG 581: 89 12325
248 PMG 604: 231n.40
118 AB SH 705: 178, 180, 181, Sophocles
184, 19293, 25859, 270 Philoct. 291: 102n.50
119 AB 13 GP: 178n.5, 184, TrGF IV F 362: 188n.24
196n.3, 244, 24546 Statius
122 AB 17 GP: 190, 26263 Silv. 1.2.252: 12
132.1 AB AP 7.267 15 GP: Stesichorus
187n.21 PMG 245: 130n.29
137 AB AP 12.98 6 GP: Strabo
18990, 223, 255n.6 14.2.19, 657c: 11
140 AB AP 12.168 9 GP: 191 16.4: 160n.26
143 AB SH 702: 178n.5, 243n.27 16.4.5: 161
144 AB SH 701: 19091 16.4.13: 161
147 AB SH 706: 244n.28 791.6: 197, 2013
P.Oxy. Strato Phoinikides
1793: 7680 1 PCG: 15, 2832
2258: 67n.7, 71n.17, 8081
Propertius Theocritus
3.1.16: 12 Epigram
3.5152: 12 17: 20910
9.4344: 12 Idyll
PSI 1092 1: 1035
67n.7, 7072 2.10610: 57
Ptolemaeus 7: 17n.17
UPZ 78.2839: 2056 7.21: 102n.50
10.4: 27
Sappho 16.30: 158n.22
1.28 L-P: 247 16.5455: 158n.22
5 L-P: 262n.19 17.1: 201n.22
7 L-P: 262n.19 17.8693: 25557
15 L-P: 262n.19 17.9597: 257, 266
104 L-P: 59n.34 17.134: 239n.14
202 L-P: 262n.19 26.2: 182n.12
SEG Theognis
17.817: 240n.17 45960: 246n.36
Semonides 11234: 162n.29
fr. 2: 130n.29 127374: 246n.36
Theophrastus Timon of Phlius

Characters 23.13: 25758 SH 786: 1079
Hist. Pl. 6.4.10: 27
Theopompus
Xenophon of Ephesus
FGrHist II 115 fr.285: 87n.6
Ephesian Tale
Thucydides
1.12: 14243n.54
1.132: 87n.6
5.1012: 14243n.54
2.43.23: 140n.47
3.102: 34n.4
6.54: 87n.6 Zenodotus
6.59: 87n.6 3 GP AP 7.315: 139n.45
Subject Index

acrostics, 11011, 22021 Attalids: Attalus I, 155; Philetaerus,

Aesop, 14243 155n.14
Agathon, 113
agon, musical, 107, 111 Baehrens, E., 70
Alexander the Great, 7, 200, 219, 235, Barthes, Roland, 40n.21, 105, 105n.65
25658, 261, 266, 269 Battis, 2325
allusion: in Archaic poetry, 15155, Battus, 44, 99100
152n.6, 155n.12; in Hellenistic Berenike I. See Ptolemies
poetry, 5155, 5657, 5960, 1035, Berenike II. See Ptolemies
110, 141n.50, chap. 8, passim, Bittis. See Battis
18889, 26061; in inscriptions, bucolic anaphora, 1035, 188
chap. 8, passim, 141n.50, 18889
Ancona painter, 12223 Callimachus, 36, 11, chaps. 26
Anjou, Ren d, 118n.1 passim, 16669, 179, 19092, 251,
Antigonus Gonatas, 247 254, 266; and Asclepiades, 16667,
Antigonus of Carystus, 26, 266 191; Epigrams, chap. 5 passim, 166
Apollo, chap. 2 passim 69, 25152; and Euripides, 5657;
Arion, 24749, 251, 262 Hymn to Apollo, chap. 2 passim; Hymn
Aristarchus, 11, 16, 17, 190 to Artemis, 34n.3; Hymn to Delos,
Arsinoe II. See Ptolemies 34n.3, 256; Hymn to Demeter, chap. 3
Arsinoe III. See Ptolemies passim; Hymns, as collection, 64;
Arsinoe-Aphrodite, shrine, of. See Hymns, as related to Homeric Hymns,
Ptolemies 3435, 41, 4748, 4955; Lock of
Artemis, 9195 Berenike, chap. 4 passim; and Pindar,
Asclepiades, 183, 191; and Alcaeus, 96; poetic program, 4546; 6364,
17173; and Callimachus, 16669 110, 254, 261n.17; and Theocritus,
Asclepius, chap. 11 passim 57, 1035
302
Subject Index 303
Cameron, Alan, 5, 15, chap. 6 passim Hippolochus of Macedon, 268

Capra, Frank, 11618, 119, 143 Homer, 1521, 23, 2526, 2832,
Cavafy, Constantine, 118n.2 12730, 15155, 15759, 19091
Chairemon, 11011 Horace, 24
Chalybes, 6566, 6870
Charaxus, brother of Sappho, 190, Ingarden, R., 86n.2
26263 inscriptions. See epigram, inscribed
Cronius, gem-engraver, 26061 Ischia Cup, 15155, 156, 157, 169,
Cyrene, 4344, 62, 99102, 240n.17 170, 208n.38
Iser, Wolfgang, 105
Dante, 118
Daphnis, 1035 Kallikrates of Samos, 184, chap. 12
Delphic maxims, 120 passim, 271n.44
Diogenes, Cynic, 226 Koine Greek, 26
Dionysodorus, son of Deinokrates, 155 Konon, astronomer, 7274
Doric dialect, 53, 155
Doricha, 190, 26263 Lazarus, Emma, 120n.6
Dositheus, astronomer, 115, 207n.37 letters, 11415, 165n.42, 2078
Drakon, father of Skorpion, 18689 library, Alexandrian, 3, 2728
dreams, 2056, 219, 22627, 24547 Ligorio, Pirro, 209
Lincoln memorial, 11619, 120, 132,
elephant hunting, 160 143
elite, literary, 2728, 37, 4546, 6364,
1079, 111, 150, 15355, 16674 Magas, king of Cyrene, 240n.17
Ellis, R., 74n.24 Medeios of Olynthos, 184, 21819,
epigram, chap. 5 passim, 11315; 233
chaps. 713 passim; books, 9293, Methymna, 24851, 262
1012, 21016, chap. 11 passim; Midas Epigram, 8890
Casa degli Epigrammi, 2089; Muret, Marc-Antoine, 6869
humor in, 16465n.40, 22831;
inscribed, 8593, 100102, chaps. Naeke, A. F., 78
78 passim, 18289, 19293, chaps. Nestors Cup. See Ischia Cup
1011 passim; readers of, chap. 5
passim, chaps. 78 passim, chaps. Orpheus, 24950
1011 passim
Pan, 15961
fragments, poetic, chap. 4 passim papyrus scroll, 18993, 20416,
26263
Gebrauchspoesie, 149, 214 paradoxography, 2627, 26366
Gibbon, 113 Pella, 18082, 19293, 259, 270
Goethe, 111, 14750 Penelope, 16364
grafti, 12021, 126 performance, of poetry, 23, 34, 42,
6061n.38, chap. 6 passim, 152n.6,
Harmodius and Aristogeiton, 157n.19 154n.12
Hekataios, sculptor, 1214, 19192 Pergamon, 155
Hieron II, 184, 212n.50 Pfeiffer, Rudolf, 21, 68, 7072, 7481
Hipparchus, 13437, 157n.19 Pharos, lighthouse, 194206
304 Subject Index
Philikos of Corcyra, 10910 reading, 17, 3536, 38, 4142, 44,

Philitas, chap. 1 passim, 63, 192 4546, chap. 5 passim, chap. 7 passim,
Philochoros, 87, 207 chap. 8 passim, chap. 10 passim, chap.
Plato, 12, 8890 11 passim
Polemon of Ilion, 87, 137, 207 Reitzenstein, R., 11314
Poliziano, Angelo, 21, 6568 Riese, A., 70
Polycrates of Samos, 255, 262, 266,
269n.41 Sappho, 2324, 190, 247, 249, 26263
Posidippus, 1214, chaps. 913 passim; Scaliger, Joseph, 6970
and Berisos, 19091; and Schneider, Otto, 7374
Callimachus, 19092, 254, 261n.17; scholars, Alexandrian, chap. 1 passim,
and epigram books, 18991, 14142, 19091. See also elite
21819, chap. 13 passim; and Homer, Shelley, Percy, 13132, 135
19091, 26061, 268; and inscribed similes, 2021, 5657, 122, 26061
epigram, 18289, 194206, 216, Simonides: Sylloge Simonidea, 8687,
chap. 11 passim; poetic program, 183n.15
19192, 25455, 261; and the Soros, Skorpion, son of Drakon, 18689
19091 Socrates, 8890, 13437
Poussin, N., 118n.2 Sostratus of Cnidus, 184, 195203,
Pratinas of Phlius, 155 216, 243, 271n.44
Praxiteles, 244 Stewart, Jimmy, 116, 117 (g. 2)
Ptolemaeus, scribe, 2056, 216 Svenbro, J., 3, 36n.7, 41n.25, 127, 140
Ptolemais, port on Red Sea, 15960 symposium, 107, 11314, 15355,
Ptolemies: Arsinoe II, 6061n.26, 184, 16674
23436, 24152, 262; Arsinoe III,
159, 16061n.26; Arsinoe- Terpander, 249, 262
Aphrodite, shrine, of, 23436, 243 Theoi Adelphoi. See Ptolemies
52, 262; Berenike I, 184; Berenike II, Thermon, 180, 18284, 18589, 224,
chap. 4 passim, 95, 112, 184; and 270
Egyptian culture, 47, chap. 12 Tischbein, J. H. W., 14748 (with
passim; as literary patrons, 1112, 47, g. 5)
1079, 180, 18485, 195203, translation, chap. 4 passim
21819, chap. 12 passim, chap. 13
passim; Ptolemy I Soter, 11, 184, Virgil, 118
240; Ptolemy II Philadelphus, 11, Voice: feminized, 5559; rst person,
12, 27, 109, 112, 115, 16061n.26, 3347; indeterminacy of, 3842;
165n.42, 184, 192, 207, 219, 23435, readers identication with, 4142,
240, 24142, 247, 251, 252, 25557, 5964, 9394, 188
26970; Ptolemy III Euergetes, 80,
184; Ptolemy IV Philopator, 159, Youtie, Herbert, 67, 79
16061n.26, 162; Theoi Adelphoi,
219, 235, 23738, 24043 Zenodotus, 11, 190
Pyrgoteles, gem-engraver, 261 Zenon, 115, 165n.42, 2078

Peter Bing, The Scroll and The Marble

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

Peter Bing, The Scroll and The Marble

Caricato da

Copyright:

Formati disponibili

Contents

pa rt 1 . Reader and Voice in Callimachus and Hellenistic Poetry

pa rt 2 . Epigram and Its Audiences

pa rt 3 . Inscription and Bookroll in Posidippus

Fig. 1. R. Pfeiffer, Callimachus, vol. 1. Oxford: 1949, 11415 7677

AB C. Austin and G. Bastianini. Posidippi Pellaei quae supersunt omnia.

GP A. S. F. Gow and D. L. Page. The Greek Anthology: Hellenistic

D ionysius of Halicarnassus recounts an anecdote that he describes as

than that of inscribed epigram. That becomes especially apparent in the

The Unruly Tongue

P hilitas of Cos stands as a gray eminence at the start of Alexandrian

the newly published epigrams of Posidippus, a near contemporary (col.

  []  [:]   {} ; []

As in Hermesianax image of the scholar frail with all the glosses,

task seems appropriate to the creation of an image of the meticulous

And in Athenaeus we learn of the tragicomic denouement when one of the

As Alan Cameron has recently noted, thin-jokes were a staple of ancient

9. Cameron 1995: 490, 493 with n. 25.

13. Strictly a dis legomenon: it comes up twice in as many lines.

[Amalla. Sheaves, bundles of grain . . . a bunch; one hundred sheaves,

attempt at explaining Homer through Homer (K 6 ;

I should emphasize that that would be quite contrary to the practice of

Pseudo-Alexander comments on this as follows (Comm. in Arist. Graeca

[Simonides, in the stories that he entitles Ataktoi, imitates and re-

David Campbell translates Ataktoi Logoi as Miscellaneous Stories.31 But

31. Campbell 1991: 507, on Simonides 653.

A still stronger piece of evidence for taking Ataktoi Glossai as unruly or

Inasmuch as the collocation of glossa and atakta appears nowhere else in

32. See the discussion of this poem in Bing 1993a.

34. Kuchenmller 1928: 27 n. 5.

properties of the kaktos, a thistleor artichoke-like plantthat Theo-

No doubt the kaktos pricked Philitas interest in recherch words. It was

Appendix: What Does Stratos Phoinikides

We are fortunate to have a contemporary reference to Philitas Ataktoi

" V 6 " "+  . 15

51. For the translation of this verb, cf. Jocelyn 1980.

The word  (articulate people, v. 6), though not appearing as a sub-

action puzzling to a minimally cultured Greek, but in a conventional Ho-

that permitted the unmediated involvement of the poets persona apart

[How Apollos laurel branch is shaking

We are plunged medias in res, into an atmosphere of breathless anticipa-

dications are remarkably vague. Nevertheless, the imperatives in lines 2, 6,

[Apollo doesnt appear to all, but only the noble [@ 6];

performance of the paean, Apollos song.13 But at this point, we notice a

17. It had already been anticipated in v. 8 ( +   6  6)

with the description referring to a cult statue, as argued by Williams (1978:

integrated into the narrative (as indeed it would be in a paean)clearly

Moving on in our examination of the hymn, we must now contend with

[It was Phoebus, too, who told Battus of my fertile city,

Great is the stream of the Assyrian river, but much

privileging the one or the other reading.36 That poetics of exclusion as

Appendix: The Song of the Delian Maidens

I have translated verses 16264 of the Homeric Hymn to Apollo as follows:

imagines the Deliades song as an imitation of the unintelligible tongues

Callimachus and the

T he Hellenistic era saw revived interest in the Homeric hymns, and

case. It is telling that Hopkinson (in his commentary on the hymn),

I. Eludingand Alluding toa Literary Paradigm

The most immediately striking echoes of the Homeric Hymn to Demeter

Thereafter, as in the Homeric hymn, Demeter impersonates a nurtur-

Erysichthon.16 Scholars have often debated whether Callimachus was the

While a critical consensus has developed that #, , and

poem, only to be dismissed as a threat to Demophoon by the goddess

[] [:] {} ; []

" V 6 " "+ . 15

The word (articulate people, v. 6), though not appearing as a sub-

[Apollo doesnt appear to all, but only the noble [@ 6];

17. It had already been anticipated in v. 8 ( + 6 6)

While a critical consensus has developed that #, , and

( : The wind, Boreas.

and its alleged

How, then, are we to imagine the who lies on the

14. Cf. CEG I 413 110 FH: P, / [ / ] /

" 6, / JD

The scene is the seashore (6 . . . 6/ , ). And we nd ourselves

The repeated phrase () is utterly convincing as a

50. It is formed on analogy with such phrases of journeying as I (Eur. Phoen.

anaphora of " in our epigram also occurs in connection with Daphnis

of immediately following is thus emphatic. Placed thus, it becomes

56. For the anaphora in ", compare Theocr. 1.11617: _ 65

2 " Brunck : P 4 4 Brunck : 4 P