Sei sulla pagina 1di 916







B. H R O U D A







O ' F A H E Y









/ 6 81-'


L i b r a r y of C o n g r e s s Cataloging-in-Publication D a t a
Handbook of Ugaritic studies / [edited] by Wilfred G.E. Watson and
Nicolas Wyatt.
cm. (Handbuch der Orientalistik. Erste Abteilung, Nahe
und der Mittlere Osten, ISSN 0169-9423 ; 39. Bd.)
Includes bibliographical references and index.
ISBN 9004109889 (alk. paper)
1. Ugarit (Extinct city) 2. Cuneiform inscriptions, Ugaritic.
3. Ugaritic philology. I. Watson, Wilfred G.E. II. Wyatt.
Nicolas. III. Series.
D i e D e u t s c h e Bibliothek - C I P - E i n h e i t s a u f n a h m e
Handbuch der Orientalistik.
Leiden ; Boston ; Kln : Brill
Teilw. hrsg. von II. Altenmller . - Teilw. hrsg. von B. Spuler
Teilw. mit Parallelt.: Handbook of oriental studies

Abt. 1. Der Nahe und Mittlere Osten = The Near and Middle East /
hrsg. von H. Altenmller ...
Teilw. hrsg. von B. Spuler
Bd. 39. Handbook of Ugaritic studies. -


H a n d b o o k of U g a r i t i c s t u d i e s / by Wilfred G.E. Watson and

Nicolas Wyatt (eds.). - Leiden ; Boston ; Kln : Brill, 1999
(Handbuch der Orientalistik : Abt. 1, Der Nahe und Mittlere Osten :
ISBN 90-04-10988-9

ISSN 0169-9423
ISBN 90 04 10988 9
Copyright 1999 by Koninklijke Brill NV, Leiden, The Netherlands
All rights reserved. No part of this publication may be reproduced, translated, stored
in a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill
provided that the appropriate fees are paid directly to The Copyright Clearance
Center, 222 Rosewood Drive, Suite 910, Danvers MA 01923, USA.
Fees are subject to change.

This handbook is dedicated to Cyrus H. Gordon


List of Illustrations

Chapter O n e


General Introduction



Chapter Two Ras Shamra, Minet el-Beida and

Ras Ibn Hani: T h e Material Sources


Chapter Three T h e Written Sources

1 T h e Syllabic Akkadian Texts



T h e Alphabetic Ugaritic Tablets




T h e Hurrian and Hittite Texts




Chapter Four T h e Ugaritic Language

1 T h e Decipherment of Ugaritic


T h e Ugaritic Script




Ugaritic G r a m m a r



Ugaritic Lexicography

Ugaritic Words in Syllabic Texts






Chapter Five Ugaritic Stylistics

1 Ugaritic Prose



Ugaritic Poetry


Chapter Six T h e Ugaritic Literary Texts

1 T h e Mythological Texts



T h e Legend of Keret

T h e Story of Aqhat

T h e Rpum Texts







T h e Incantations



Chapter Seven T h e Ugaritic Cultic Texts

1 T h e Rituals



T h e Offering Lists and the God Lists



T h e O m e n Texts



Chapter Eight T h e Correspondence of Ugarit

1 T h e Ugaritic Letters



T h e Akkadian Letters


Chapter Nine

T h e Legal Texts from Ugarit





T h e Akkadian Legal Texts
T h e Ugaritic Legal Texts
T h e Hittite Legal Text

Chapter T e n


T h e Economy of Ugarit




T h e Administrative Texts
Crafts and Industries

Chapter Eleven

T h e Society of Ugarit



Peoples, Cultures and Social Movements



T h e Royal Family, Administration and Commerce

T h e Family and the Collective
Crafts and Professions
T h e Army

Chapter Twelve

T h e Onomastics of Ugarit





Personal Names and Prosopography

Ugaritic Place Names

Chapter Thirteen
An Overview


T h e Religion of Ugarit:


Chapter Fourteen

T h e Iconography of Ugarit



Chapter Fifteen

A Political History of Ugarit




Preliminary Remarks
Ugarit in the Early and Middle Bronze Ages
Ugarit under Egyptian Influence
Ugarit under Hittite Rule
Ugarit in the Age of 'Pax Hethiticd
T h e Weakening Grip of the Hittites
T h e Last Years of Ugarit

Chapter Sixteen T h e Tablets and the Computer

1 T h e Current State of Ugaritic Sudies and



Storage and Analysis of the Texts



List of Contributors
Index of Topics
Index of Personal Names
Index of Divine Names




Ugaritic Words
Syllabic Akkadian
Various Languages
K T U Texts
R I H Texts
RS Texts



Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9


Western Syria in the Late Bronze Age

A list of offerings with the first tablet number
( K T U 1.39 = RS 1.001)
T h e alphabet tablet from Ugarit (14th/13th
cent, BCE)
T h e long cuneiform alphabet from Ugarit and its
relationship to Phoenician and Canaanite
T h e short cuneiform alphabet from Ugarit
T h e Ugaritic script in relation to the Western and
Southern scripts
T h e alphabet tablet from Beth Shemesh
Tablet with the South Arabic Alphabet
Deciphered alphabets of the South-Eastern
h-l-h-m- tradition
T h e spread of cuneiform alphabets in the Eastern
Calcite statue (of El?)
'Baal au foudre' stela
Menacing god
Gold decorated patera
'Qedeshet'-type gold pendant
Ivory bed panel: royal couple caressing
Cylinder-seal impression (Minet el-Beidah)
Decorated rhyton
Module 1
Module 2
Module 3



T h e Handbook of Ugantic Studies is the product of the labours of a

large team of scholars from many countries. Its gestation has been
quite lengthy, with many emergencies, false alarms, high bloodpressure, worrying scans, premature contractions and so forth.
T h e original editor, Johannes de Moor of K a m p e n , began the
organization of the volume, drew up an outline and undertook the
arduous task of contacting contributors from a r o u n d the globe.
However, for personal reasons, he felt compelled to withdraw from
the enterprise at an early stage and the publishers then invited Wilfred
Watson (Newcastle) to take over. This, of course, was felt to be a
great honour, but due to the need for a fellow-worker, Nicolas Wyatt
(Edinburgh) was then invited to act as co-editor. T h e use of e-mail
has enabled the editors to work closely together on all the stages of
the production of the Handbook and to maintain contact with many
of the contributors. It was also helpful for the translation of contributions in German, Italian and Spanish (15 out of the 47 sections)
prepared by Watson, with some revision by Wyatt and the contributors concerned.
Unfortunately, there was a gap of several months before the project was resumed under its new editors and for a variety of reasons
a number of scholars withdrew from the project. Only when it was
reestablished under the direction of the new editorial team did the
entire membership of the Mission de Ras Shamra withdraw. This
meant that new contributors had to be found, some at quite short
notice. Further withdrawals at intervals right to the end of the project have discouraged us, and we have to thank Patricia Radder of
Brill, as well as those contributors who generously stepped into the
breaches left by others, in some cases very late in the day, to enable
us finally to make the volume ready for delivery.
Since this volume has been published in English, we have by and
large standardized ancient names where there are recognized English
equivalents, so that, for instance, ancient 'Karkami' and 'Kargami'
are rendered 'Carchemish'. Similarly, 'Ilu' becomes , 'Ba'lu'
becomes 'Baal', and so on. We have not however imposed total
consistency, so that 'Hatti' and 'Muki', for instance, which have no

standard m o d e m forms, retain their diacritics. For Ugaritic texts

K T U numbers are followed, with cross-reference to RS numbers,
and in some instances where contributors have added PRU, Ug or
R S O numbers these have been retained.
We feel obliged to honour the Kotharat, the goddesses of childbirth, whose gracious intervention has finally achieved a safe parturition. It is frequendy observed, to move to the spheres of influence
of other deifies, that among the most traumatic human experiences
other than death are house-buying and divorce. Perhaps we should
add to this list the editing of large reference books. sbc alpm Iktrt:
seven oxen to the Kotharat!
G.E. W A T S O N
University of Newcastle


August 1998



New College, Edinburgh

Western Syria in the Late Bronze Age







As yet no handbook of Ugaritic studies has been published with the

scope and range of the present volume. However, there have been
forerunners on a smaller scale. T h e first substantial work of this kind
was by DUSSAUD 1937', 1941 2 . In two volumes, it was greatly influenced
by the Hebrew Old Testament, of course, and with some of the misconceptions of the time (particularly in respect of alleged geographical allusions). Another comprehensive survey by DE LANGHE 1945,
described the discovery and history of Tell Ras Shamra, the texts
found there, the writing and language of the alphabetic texts, their
archaeological, geographic and historical background (with a list of
the personal names) and a description of the family, social and political life of Ugarit. Here SCHAEFFER 1939a may also be mentioned.
Later came D R O W E R 1975, K I N E T 1981 and CURTIS 1985. T w o more
recent works which cover some of the same ground as the present
volume and also include maps and rich illustrations are BALDACCI
1996 and CUNCHILLOS 1992b. Some recent encyclopaedia entries
include C A Q U O T 1979b, C O U R T O I S 1979, Y O N 1992a, P A R D E E
BORDREUIL 1992; and brief descriptions of the language are provided
by PARDEE 1997d and W A T S O N 1994b. A survey of work up to 1980
is provided by Y O U N G , G . D . 1981, Cunchillos has produced a useful handbook for students (CUNCHILLOS 1992b) and the proceedings
of further international conferences have been published (UBL 11
Manchester 1992, ALASP 7Mnster 1993, R S O 11Paris 1993,
UBL 12Edinburgh 1994). T h e successive volumes of Ugartica and
PRU, together with other volumes of the R S O series have provided
updated instruments of research.



T h e tablets, our primary source of information, are unfortunately

deteriorating fast. O n a more positive note, photographs of the texts
are in preparation (ZUCKERMAN - ZUCKERMAN 1997) and computer
programs have been applied to scan them (see section 15). A new
edition of the tablets in transcription has been published ( T 7 7 2 ) ' as
well as a concordance of the tablet-numbers ( T E O = R S O 5/1). An
edition of the texts is now available both on the Internet and on
C D - R O M (CUNCHILLOS - V I T A 1993a; CUNCHILLOS 1998b, 1998c,
and ongoing work in the United States with the West Semitic Research
Project) which has the a d v a n t a g e of being u p d a t e d continually.
O t h e r reference works available are concordances (CUNCHILLOS
V I T A 1995a, which largely replaces W H I T A K E R 1972) and word lists
(DIETRICH - L O R E T Z 1996b). O f particular significance is the first
volume of the new dictionary (DLU = DEL O L M O L E T E - SANMARTIN
1996). In addition, three teaching g r a m m a r s (SEGERT 1984; CUNCHILLOS - ZAMORA 1995; SIVAN 1997) and outlines of g r a m m a r (CAZELLES
1979; CUNCHILLOS 1992b; PARDEE 1997d) have been published.
Comprehensive coverage of the Akkadian of Ugarit is available
(HUEHNERGARD 1989; VAN SOLDT 1991a) and the Hurrian of Ugarit
has also been studied ( D I E T R I C H - M A Y E R 1995, etc.). Studies are
available on the town of Ugarit - Ras S h a m r a (SAAD 1978; R S O
1, 3, 6, 8, 10) and on Ugaritic religion (DEL O L M O L E T E 1992a =
1999) a n d ( W Y A T T 1996a), sociology ( V I T A 1995a; A B O U D 1971;
H E L T Z E R 1976, 1982) and other topics ( G R A Y 1965; RSP 1-3) including the hippiatric texts ( C O H E N - SIVAN 1983; PARDEE 1985;
SANMARTIN 1988a; C O H E N 1996).
Ugaritic remains a flourishing discipline. UF is now in its 30th
year and periodicals such as /1/0, AuOr, JNES, JSS, SEL, Semitica,
Syria, WO and
carry articles and reviews on Ugaritic. Unfortunately,
the Newsletter for Ugaritic Studies has been discontinued (last issue
A p r i l - O c t o b e r 1989) in spite of efforts to reactivate it.
Several translations have been available: C L E A R 1976 2 ; C O O G A N
1978; GINSBERG 1969 3 , 129-55; HVIDBERG - HANSEN 1990; DE M O O R
1987; R I N - R I N 1992; GIBSON 1978 (a revision of D R I V E R 1956),
DEL O L M O L E T E 1981a and X E L L A 1982a. A n u m b e r of new trans' For corrections see T R O P P E R 1995b, 1995-6, 1998; P A R D E E 1998.
See T R O P P E R 1997b for a survey a n d evaluation of recent work.

lations have appeared including TO 2, CS i, with the translations of

Ugaritic texts by PARDEE and others (also PARDEE in R S O 4, 12),
as well as PARKER (ed.) 1997 and most recently, W Y A T T 1998C, with
extensive footnotes, and DEL O L M O L E T E 1998b. Monographs on single texts include G R A Y 1964 2 (on Keret), M A R G A L I T 1989a (on the
Aqhat text) and SMITH 1994 (on the Baal Cycle).
Full bibliographies on Ugaritic up to 1988 are provided by D I E T RICH - L O R E T Z et al. 1967-86; D I E T R I C H - L O R E T Z 1996a; C U N CHILLOS 1990 ( T E O 2 = R S O 5/2); while BORDREUIL - PARDEE 1989
(TEO 1 = R S O 5 / 1 ) is particularly useful for determining the locations of tablets in various museums, their condition and what they
Work currently in hand includes a three-volume work on Ugaritic
g r a m m a r ( T R O P P E R ) ; a series of articles on toponyms (VAN SOLDT
1996; 1998); an English translation of DEL O L M O LETE 1992a (1999),
new editions of the ritual texts (PARDEE in press) and the letters
(PARDEE), the second instalment of which is eventually to appear in
an English language edition, a study of religion (WYATT), and further volumes of SMITH 1994 and CS.


With this description of the present state of studies, which shows the
discipline to be in a healthy condition, it is important to point to
further work that is required, although some indications are provided in the various contributions. Topics to be studied more exhaustively include the alphabetic and syllabic personal names, as well as
archaeology; and as yet there is no comprehensive translation of the
so-called administrative texts. T h e texts in Ugaritic have tended to
be the focus of attention, with the result that other areas have suffered
from comparative neglect.


Due to the international nature of this undertaking, which entailed

a large n u m b e r of scholars, and in spite of the advantages of communication by e-mail, the articles in the Handbook cover the material to different depths and there is also inevitably some overlap

between them. Even so, many contributors discuss a range of topics either not previously dealt with, such as iconography and technology, or with more detail than previously available. T h e main
thrust of the book has been to provide surveys of what has been
achieved, a task which often proved difficult either due to the absence
of previous surveys or because of the sheer range of opinions voiced.
It is hoped that a balance has been struck in respect of the amount
of detail provided and coverage is intended to be comprehensive and
representative rather than complete. Finally, the extensive consolidated bibliography will certainly be of use for reference, filling the
gap between 1988 (covered by A O A T 20/6) and 1998. This work
has appeared in the seventieth anniversary of the discovery of Ras
Shamra, a propitious portent, perhaps, of discoveries to come.















Seventy years have elapsed since a chance discovery was made close
to the coast of Syria which was to spark off a series of archaeological investigations which have continued right up to the present. Not
only have the excavations revealed an important commercial centre
the ancient city of Ugaritwhich flourished in the second millennium BCE, thereby shedding light on the history and culture of the
area and of the wider ancient Near Eastern world. They have also
yielded a hitherto unknown language or dialectUgariticwhich
has made an important contribution to the study of the north-west
Semitic languages in addition to giving access to the life and thought
of the people of the city. T h e facts, firsdy that the newly discovered
language was seen to be akin to Hebrew, secondly that the texts,
once deciphered, were found to contain references to deities mentioned in the Hebrew Bible, in particular the god Baal, and thirdly
that the site was geographically rather closer to the land occupied
by the Israelites than the other great centres of ancient Near Eastern
civilization (though the considerable distance has sometimes been
minimised) all doubtless contributed to the early claims that a site
of major significance had been discovered. This had its pluses and
its minuses. It brought the discoveries to earlier prominence and to
a wider audience than might otherwise have been the case. But the
issue of the relevance of the discoveries at Ugarit for the study of
the Hebrew Bible, exacerbated by the tendency to assume that Ugarit
was a Canaanite city, has often been unduly dominant, at the expense
of an appreciation of Ugarit and its texts in their own right. T h e
excavation of other ancient cites in Syria, notably Ebla and Emar,
has helped to redress the balance somewhat and enabled Ugarit to
be seen in its rather more immediate geographical milieu.



T h e chance discovery alluded to above took place in the spring of

1928 some 10 km to the north of Latakia, close to a small bay, the
white rocks at whose entrance had given it the name Minet el-Beida
(formerly known as Leukos Limen, both names meaning 'white harbour'). A local farmer was halted in the task of ploughing his land
when his ploughshare struck a large piece of stone which, on closer
examination, turned out to be one of a number of stone slabs which
formed the roof of a vaulted tomb. It appears that a number of
antiquities had alread been found in the vicinity, so the discovery
was brought to the attention of the Service des Antiquits en Syrie et au
Liban. Its director at the time was Charles Virolleaud, who was subsequendy to play a major role in the decipherment of Ugaritic and
the early publication of the Ugaritic texts. He sent a member of his
staff", Lon Albanse, to visit the site and it was identified as a necropolis. Some pieces of ceramic were found which appeared to be of
Mycenean or Cypriot origin and to date from approximately the
thirteenth century BCE. However the site was not, at that stage,
thought to be particularly interesting.
Fortunately a plan of the tomb and some pottery samples were
sent to the Louvre in Paris for further examination. There they came
to the notice of Ren Dussaud, who was at the time Keeper of the
Department of Oriental Antiquities, and who noted that the tomb
appeared to be reminiscent of Cretan funerary vaults. He suggested
that what had been discovered might be the necropolis of a significant
city. Albanse had already noticed that there was a mound nearby
whose shape suggested that it might be a tell. This hill was known
as Ras Shamra, the name (which means 'fennel head [land]') being
derived from the plants which grew on its surface. So it was decided
that excavations should be carried out on the site, under the auspices of the French Acadmie des Inscriptions et Belles Lettres, under the
direction of Claude F.A. Schaeffer.
It was in the Spring of 1929 that the first team of archaeologists
(accompanied by a detachment of soldiers to safeguard them) arrived
on the site, their equipment having been transported on the backs
of camels because the roads in the vicinity could not be used by
motor vehicles. T h e initial work undertaken involved a survey of the
vicinity, and traces of occupation stretching from the Neolithic period
to the time of the Romans were discovered. It was on April 2nd

that excavations proper began, and within a relatively short period

early suspicions were confirmed when it became clear that a considerable complex had been found. It emerged that the 'necropolis'
comprised two different areas, the seaward of which revealed finds
of pottery and animal bones but no human remains. T h e other area
comprised well-built vaulted tombs containing a variety of artefacts.
Some of the earliest finds within the tombs were of considerable
interest and importance. For example, they began to give an inkling
of the cosmopolitan nature of the site, revealing artefacts which suggested an Egyptian or Cypriot origin. Among these earliest discoveries were a number which have come to have particular prominence
in treatments of the discoveries from Ras Shamra. O n e was a small
statuette (AO 1 1 . 5 9 8 , C A Q U O T - SZNYCER 1 9 8 0 , pi. IXd), just 2 2 cm
in height, of a male figure with one arm raised above his head and
the other reaching forward, and with one leg in front of the other
as though marching or else poised to throw something. T h e figure
wore the accoutrements of a warriorhelmet, armbands and greaves,
and the fact that the helmet (and head) were covered in gold leaf
and that the armbands and greaves were of silver suggested that this
was perhaps a deity. T h e figure was initially identified with Resheph,
a god of plague, on the basis of other known representations of the
god. However, it has subsequently been thought much more likely
that the figure represented Baal, depicted as the storm-god armed
(originally) with club and spear, symbolising the thunder and lightning, his 'weapons'. This latter identification is supported by the striking similarity between the pose affected by the statuette and that of
the figure on the famous 'Baal stele' (RS 4 . 4 2 7 ; C A Q U O T - SZNYCER
1980, pl. X) which was not, of course, known at the time of the
earlier suggested identification (Fig. 11, p. 590).
Another important early find was in fact made in a tomb which
showed every sign of having suffered at the hands of tomb-robbers
who perhaps did not regard a small ivory box-lid (AO 1 1 . 6 0 1 ,
C A Q U O T - SZNYCER 1 9 8 0 , plates IV & V), only 1 3 cm high, as an object of value. Carved on it was a seated female figure with an ornate
hairstyle, naked above the waist but wearing a very full skirt, holding
what appear to be ears of corn in either hand and flanked by animals (probably goats or ibexes) standing on their hind legs. T h e style
was unmistakably Mycenean, and it was suggested that the depiction was of a fertility goddessperhaps the 'Mistress of the Animals'
though the precise identity of the figure was unclear.

After just over a month (on May 9th) attention was turned from the
necropolis to the tell itself, just over a kilometer away, in order to
answer the question whether it did indeed contain the ruins of a city
of which the necropolis was the cemetery. T h e summit of the mound,
which was some 17-20 m above the surrounding terrain, was very uneven
but showed no clear sign of any ancient structures. After a survey
of the tell's surface, the decision was taken to make the first trial
excavation at the point on the mound's surface which was closest to
the sea. T h e decision was influenced by reports that local inhabitants
had come across artefacts, some of gold, in an olive orchard which
lay below that part of the tell. Schaeffer wondered whether the objects might have come from a royal palace. His acumen was well and
speedily rewarded. Almost as soon as digging commenced, the excavators came upon the foundations of a large edifice which seemed
to have suffered destruction by fire. A bronze nail embedded between
the blocks of a pillar, and a bronze dagger which had suffered distortion as a result of the intense heat of the conflagration, pointed
to a date for the edifice in the second millennium BCE. Confirmation
of this dating was provided by the discovery of parts of an Egyptian
statue made of granite and bearing a hieroglyphic inscription whose
style of writing was dated to the New Kingdom period.
As excavation continued, the scale and plan of the building began
to become clearer. So that the extent of the building might be assessed,
another trench was begun some 20 m to the east. More foundations
were revealed, whose depth and direction suggested that they belonged
to the same building, but here the rooms seemed to be smaller and
it was thought likely that they were storerooms. Pieces of ceramic
pointed to a date for the building which coincided with that of the
necropolis and suggested that both had ceased to be used by the
end of the thirteenth or the beginning of the twelfth century BCE.
O n May 14th, less than a week after excavations had begun on
the tell, in the corner of one of the small storerooms, a tablet of baked
clay bearing a cuneiform inscription was discovered. Soon others
began to emerge within a small radius. Some had been rendered
very friable, presumably by the fire of which evidence had already
been found, and great care was needed to prevent them crumbling.
It was necessary for some to be removed still encased in soil and
allowed to dry slowly. This first epigraphic find involved a total of
twenty tablets of varying sizes, and perhaps the most intriguing fea-

ture was that they had revealed a hitherto unknown cuneiform script.
Shortly afterwards, in a newly opened area in another part of the
tell, a deposit of 74 bronze artefacts was discovered under a stone
slab (see SCHAEFFER 1939, pl. X X I I fig. 2 for photograph of hoard
in situ). These objects included various tools and weapons, all apparendy unused, and a small tripod decorated with pomegranate flowers.
It was soon noticed that some of the tools bore incised inscriptions
in the same script as that which had just been found on the clay
tablets ( K T U 6.6, 6.10, 6.7, 6.8 and 6.9 = RS 1.[051]; 1 .[052];
1. [053] ; 1. [054] ; 1.[055] respectively). Charles Virolleaud, who examined the newly found tablets and the inscribed tools, suggested
that the writing on the tools might perhaps provide the clue to the
decipherment of the hitherto unknown script. (The inscription on
the tools in fact turned out to read (or, in the case of K T U 6.7,
include) the words rb khnm, 'chief of the priests', suggesting that the
building where they had been discovered was the home of the chief
priest, and that the bronze items were perhaps a dedicatory offering
made by one of the metal-workers of the city.)


T h e earliest excavations had done more than enough to suggest that

an important city had been discovered, and that a campaign lasting
a number of seasons was justified. At that stage the identity of the
newly discovered city was not known. However, a tablet unearthed
in 1931 was to provide the clue. Schaeffer, in the context of giving
a preliminary report of the 1931 campaign (Schaeffer 1932) made
public the fact that the tablet contained a phrase which was transliterated as nqmd mlk grt ('Niqmaddu, king of Ugarit'), suggesting that
this might be the ancient city of Ugarit whose existence was already
known from e.g. the Tell el-Amarna letters, which suggested that its
location must have been somewhere in the vicinity of Ras Shamra.
(The identification had apparently already been suggested by Albright
(ALBRIGHT 1 9 3 1 - 2 , 1 6 5 n. 9).) Excavations continued until the outbreak of the Second World War, initially both in the region of the
harbour as well as on the tell. Soundings were taken on the tell to
attempt to establish the antiquity of the site and its principal occupation levels (CONTENSON 1 9 9 2 ) . These indicated that occupation

went back as far as the Neolithic period, and suggested five major
phases of occupation:


ca. 1200-1600
ca. 1600-2100
ca. 2100-3500
ca. 3 5 0 0 - 4 0 0 0

(Subsequent excavations have revealed a much more complex stratigraphy, and discerned some twenty occupation levels stretching from
the Early Neolithic period \ca 6 5 0 0 - 6 0 0 0 BCE] down to a R o m a n
occupation in the first and second centuries CE.)
In the pre-war seasons of excavation on the tell, work was carried
out in particular on the acropolis, revealing what came to be identified
as the Temple of Baal, the Temple of Dagan and the House of the
High Priest. This last emerged as much more than simply a dwelling
place for an important cultic official, but as a temple library and
scribal school where texts were written and stored and where new
scribes could learn their art. After the interruption of the war years,
a limited resumption was possible in 1948, but it was not until 1950
that full-scale work could again be undertaken. Excavations were
concentrated first in the area of the Royal Palace. More will be said
about this imposing edifice later, but it is appropriate to note that
this was not just the residence of the royal family and court but also
an administrative headquarters. A number of groups of texts were
discoveredfive which have been labelled as the Eastern Archives,
the Central archives, the Southern Archives, the South-Western Archives and the Western Archives. A further group was found within
what was originally identified as a firing kiln and therefore thought
to comprise the last tablets to be written before the destruction of
the city at the end of the LBA (but see M I L L A R D 1995).
In the Royal Palace area, excavations were extended southwards
to reveal what came to be known as the 'Southern Palace'. Subsequently a building which has been named the 'Northern Palace' was
found in the proximity of the Royal Palace. T o the west of the Royal
Palace was found a residential area which included a number of
buildings which have been named as a result of discoveries made
within them or because of evidence of their ownership. These included
the houses of Raap'abu and of Rap'anu, in the latter of which was
found an archive of texts, and the so-called 'House of the Scholar'

and 'House of Alabasters'. Excavations were also carried out on the

northern side of the tell in the so-called 'Lower City' below the
acropolis, in the 'Southern Acropolis' (where a house which is thought
to have belonged to a diviner was revealed, since it contained clay
models of livers and a number of para-mythological texts) and in
the 'Southern City', which seems to have contained a public square
and a building which housed a library of texts. Between 1978 and
1984, excavations concentrated on the 'City Centre' which seems to
have been primarily a residential area (YON et al. 1 9 8 7 ) . Perhaps the
most significant of the discoveries made in that part of the city was
another temple, originally known as the 'Rhyton Temple' because
of the finding of a n u m b e r of distinctively shaped drinking vessels in its vicinity. As will be noted later, it is possible that this
temple was in fact dedicated to the worship of El, the head of the



In the foregoing brief account of the principal areas of excavation,

a number of references have been made to the discovery of archives
or libraries of texts. It is therefore appropriate to make some general observations about the textual discoveries. T h e cosmopolitan
nature of the city is borne out by the number of different languages
evidenced in the texts. Many tablets were written in Akkadian, a
language which has been described as something of a lingua franca in
the ancient Near East, there were texts in Sumerian, H u m a n , Hittite,
in Egyptian hieroglyphs and in the linear script of Cyprus, as well
as those in the hitherto unknown language which was thought to be
the local language and hence known as Ugaritic. It was clear that the
method of writing the unknown language was that employed throughout Mesopotamia, i.e. a cuneiform script produced by means of the
use of a clay stylus to inscribe clay tablets which were subsequently
baked hard. It was also clear that the basis of the writing of the language was very different since only about thirty different signs were
employed. Thus it seemed that a significant advance had been made
over the earlier cuneiform scripts which were based on ideographic
or syllabic principles and involved much larger numbers of signs.
This is not the place to enter into the argument as to whether Ugaritic
is truly alphabetic (because three of its signs could be described as

'syllabic' in that they convey the consonant aleph followed by the

vowels a, i, and u respectively; see below, 4.1, 4.2). Suffice it to
say that the script is generally regarded as alphabetic, and therefore
as one of the earliest if not the earliest example of an alphabet.
(Another alphabetic script was being developed further south in the
area of the Sinai peninsular for the writing of a Semitic language
or dialect.) H o w this particular script was produced is not clear. It
is possible that was a conscious modification of the more complex
cuneiform scripts already in use, but it may have been developed
from a linear script to enable it to be more easily written on clay
tablets by employing the cuneiform method of writing ( M I L L A R D 1 9 7 9 ;
D I E T R I C H - L O R E T Z 1 9 8 9 ) . It is noteworthy that this newly developed script was used not only for the writing of the local language
but also, at Ugarit, for the writing of Hurrian (LAROCHE 1968a).
Indeed, one of the thirty signs of the 'Ugaritic' alphabet may have
been developed for the writing of Hurrian (SEGERT 1983b).
This newly-discovered language was in fact deciphered remarkably rapidly. T h e first texts were published commendably quickly by
Virolleaud, enabling other scholars to work on them. Notable among
those who did so were E. D h o r m e of the Ecole Biblique in Jerusalem
and H. Bauer of the University of Halle, both of whom were experienced in cracking codes. It is not appropriate to go into detail
about the decipherment here, since it will be dealt with in another
chapter (see C A T H C A R T , below 4.1). Suffice it to say that it was a
mixture of great erudition and inspired guesswork. T h e speed with
which the decipherment was achieved can be illustrated by the fact
that Virolleaud was able to publish his translation of what he called
the 'Epic of Aleyan and Mot' as early as 1931 (VIROLLEAUD 1931a).
It is important to note that, although the language has come to
be known as Ugaritic, examples of this script have been found in
other locations, some relatively close to Ras Shamra (e.g. Ras Ibn
Hani and, a little further afield, at Tell Sukas [ K T U 4.766 = Varia:
T S 4001] and Tell Nebi-Mend (Qadesh on the Orontes) ( K T U 6.71
= T N M 022). Some have been found at much greater distances, in
Lebanon ( K T U 6.2 = Varia: K L 67:428p; 6.67 = K L 77:66; 6.70
= Sar 3102), in Cyprus ( K T U 6.68 = H S T ) and also in Israel.
Tablets bearing the Ugaritic script have been found at T a a n a c h
( K T U 4.767 = T T 433) and at Beth Shemesh ( K T U 5.24 = 8.1 =
AS 33.5.165), and an inscribed dagger was found in the vicinity of
Mount T a b o r ( K T U 6.1 = PAM = IAA 44.318). Although these

examples are limited in number, their existence is significant not

least in the context of the discussion as to whether the beliefs and
practices alluded to in the Ras Shamra texts were limited to the
immediate proximity of Ugarit or whether they reflect a much more
widespread phenomenon. This in turn leads to the further question,
hinted at earlier, as to whether it is correct to describe Ugarit as a
'Canaanite' city, and its religion and culture as 'Canaanite'.
Before leaving the textual discoveries, it is appropriate to mention
the variety of different types of literature contained therein. T h e
number of languages represented among the texts has already been
noted, and one particular type of text which must have been very
important in ancient Ugarit was the 'dictionary' or 'word-list' in
which words in Akkadian might be listed with their Sumerian or
Hurrian equivalents. Such texts are also of great importance for the
modern study of the languages of the ancient Near East. Reference
has also been made to the discovery of archives, notably in the Royal
Palace, and many texts of a diplomatic, legal, administrative or commercial nature were found (see PRU 2-6). Other finds include private
correspondence and even veterinary texts (PARDEE 1985). Various
types of text might be included within the category 'religious' (see
below, 6, 7). Lists of deities and of sacrifices give an indication
of the large number of deities worshipped and perhaps hint at their
relative importance. Some texts can perhaps be described as 'rituals'. Then there are those longer texts which contain myths or legends, describing the exploits of the deities and of other 'heroes' who
may have been regarded as h u m a n beings but who had encounters
with or stood in a close relationship to the gods. It is these myths
and legends, and particularly the stories of the activities of the god
Baal, which have been taken up by students of the Hebrew Bible
in the hope that they may shed light on the relationship not only
between the Israelite deity Yahweh and his arch-enemy Baal, but
also on the connection, if any, between Yahweh and El the head of
the Ugaritic (and 'Canaanite') pantheon.


In the course of the earlier account of the excavations, a number

of major buildings was noted, and it is appropriate to say a little
more about some of them now in the context of an attempt to give

something of an impression of the city as it must have been in the

years prior to its destruction. Ugarit must have been dominated by
its acropolis, on which the most prominent building was the Temple
of Baala fact which is doubtless an indication of the importance
of the deity to the people of the city. T h e identification of the building was made possible by the discovery there or nearby of stelae
naming or depicting him. T w o are particularly noteworthy. O n e was
a dedicatory stele presented by a person named M a m y who seems
to have been the equivalent of the Egyptian ambassador to Ugarit
( R S 1. [089] + 2. [033] + 5.185; Y O N 1991, 328, fig. 8; C A Q U O T SZNYCER 1980 plate XII). T h e other, found a little to the west of
the temple itself, has come to be the most familiar of the depictions
of Baal (RS 4.427). He is shown standing with one leg in front of
the other, wearing a helmet which seems to be decorated with horns
(a symbol of divinity or perhaps fertility), a skirt or loin cloth and
a scabbard. His right arm is raised above his head, holding what
appears to be a club (probably a symbol of the thunder), and his
left arm is stretched in front of him, holding an object which is
pointed like a spear and which is probably a stylized lightning-flash
(Fig. 11, p. 590; see S C H A E F F E R 1934, F E N T O N 1996). T h u s Baal
appears as the divine warrior, armed with the weapons associated
with the god of rain and storm. Another small figure on the stele
may perhaps be the king. (A number of other statuettes, including
that already mentioned, have been identified as representations of
Baal because they show a figure in the same pose and garb.)
T h e temple building itself conformed to a very familiar pattern
(SCHAEFFER 1 9 4 9 , 4 fig. 2 ) . It was surrounded by a walled enclosure
or sacred precinct, within which was discovered what was probably
the base of an altar which would have been approached by two
stone steps. T h e temple proper comprised an outer room which was
approached by a monumental staircase, part of which has been preserved, and an inner room containing a structure of large stone
blocks which may have been an altar or a platform, perhaps where
an image of the god would originally have been located. This pattern of inner sanctum, outer room and sacred precint or courtyard
is reminiscent of other Semitic sanctuaries, notably Solomon's Temple
in Jerusalem as described in the Hebrew Bible.
T h e r e was a n o t h e r temple on the acropolis which had been
identified as the Temple of Dagan because of the discovery outside
its southern faade of two stelae containing dedications to this deity

( K T U 6.13 = RS 6.021; K T U 6.14 = RS 6.028; YON 1991, 334,

fig. 14). T h a t Dagan should have had a temple in such a prominent position on the acropolis is perhaps somewhat surprising since
he plays no active role in the mythological texts so far known.
However, the texts do suggest that he was considered to be Baal's
father, which might account for his temple being located in the same
vicinity. A feature of this temple was the thickness of the walls, a
fact which even gave rise to the speculation that some particularly
esoteric practices were performed therein. However this is far from
certain and there are other possible explanations, e.g. that the walls
supported an upper storey or tower. T h e general plan of the temple was similar to that of Baal. Close to the Temple of Dagan were
found a number of drinking troughs which, it has been suggested,
might have been used for the pouring of libations. But is not absolutely
clear that they actually had any direct association with the temple.
Between the two temples on the acropolis lay the building which
was identified as the high priest's house, thanks to the discovery of
the cache of bronze tools and weapons, some of which, as has already
been noted, bore the dedication 'chief of the priests'. It was built of
dressed stones, with rooms opening off a central courtyard. Within
the building were found three groups of texts, including those longer
texts in the Ugaritic language which record the activities of the gods,
in particular the god Baal. Other texts had the appearance of being
writing exercises. This suggested that the building functioned not
only as a residence for the high priest but that it was also a school
where scribes could learn to write and where texts were copied and
stored. It may also therefore have been a temple library. Whether
some of the mythological texts were actually used in the cult of the
temple, and if so in what way, it is impossible to be certain, but some
of them do contain hints that they may have been read or enacted
in the context of the worship of the temple.
In addition to the two temples on the acropolis, for a long time
the only other building identified as a sanctuary was located in the
vicinity of the Royal Palace (SAAD, 1 9 7 9 , 1 1 5 - 6 ) . However, in the
course of the excavations in the City Centre, another building was
unearthed which was identified as a sanctuary. This is the edifice
which, as noted earlier, came to be known as the 'Rhyton Temple'
because of the discovery of a number of conical drinking vessels in
its vicinity which were thought to be cult items associated with the
sanctuary ( Y O N et al. 1 9 8 7 , 2 1 3 - 4 8 ; Y O N 1 9 9 6 , 4 0 5 - 2 2 ) . T h e building

contained an entrance porch leading to the main central area, comprising a large room containing benches and a platform or altar and
another smaller room. There appear also to have been various annexes
to the main building. But is it possible to say to which deity this
temple was dedicated? T h e discovery of a stone statue of a figure
seated on a throne may provide the clue since it is thought to be a
representation of EL, the head of the pantheon ( R S 8 8 . 7 0 ; Y O N 1 9 9 6 ,
422 fig. 4c). It would be surprising if there were not a temple dedicated to EL in Ugarit. T h e texts do suggest that EL was perhaps
receding somewhat into the background, in favour of the younger,
more active Baal (though it should be remembered that many of
these texts are particularly concerned with Baal and the construction of his 'palace' or temple, and doubdess reflect the beliefs of his
worshippers). Nevertheless, EL is still the head of the pantheon who
presides over the assembly of the gods, and whose permission is
needed for major projects such as the building of Baal's palace/temple). So perhaps the 'Rhyton Temple' is in fact the Temple of EL.
T h e most impressive building in the city, certainly so far as its
size was concerned, was doubtless the Royal Palace. There is reason to believe that it began (perhaps in the 15th century) as a relatively small building comprising a number of rooms arranged around
two courtyards, but that it developed thereafter in a n u m b e r of
phases of construction until, by the 13th century, it was a huge complex containing some ninety rooms, five large courtyards and some
smaller courts and what has been described as the 'garden'. At its
zenith it measured some 120 m by 85 m. (On the stages of development of the Royal Palace, see SCHAEFFER 1962, 9-17.) O n e of
the courtyards contained an ornamental pool surrounded by two
tiers of shaped stones. Elaborate arrangements were made for the
palace's water supply, a covered channel having been constructed to
bring water from a trough which was next to a well some distance
from the palace itself. Servants presumably drew water from the well
to feed the trough and in turn the channel bringing water to the
T h e main entrance to the palace seems to have been from the
west, through a doorway approached by low steps and flanked by
two columns whose bases remain in situ. A feature of the Royal
Palace is the high quality of the stone-work which must have involved
skilled masons. O n some of the interior walls, traces of plaster can
be seen and gaps between some of the courses of stone suggest that

originally wooden beams were employed in the construction. T h e

presence of a number of staircases shows that there was an upper
storey, and it is suggested that the private living quarters of the royal
family may have been on the first floor, and that the ground floor
would have been the location of various public and reception areas
as well as the administrative headquarters, archives and store rooms
(for the texts found in the archives of the Royal Palace, see PRU
2-6). T h e discovery of a kiln in the courtyard where the ornamental pool was situated shows that texts were written in the palace and
not simply stored there, and the presence of writing exercises and
glossaries suggests that here too, as in the House of the High Priest,
scribes were trained in the art of writing.
T h e Royal Palace seems to have been guarded by a tower and fortress located at the western extremity of the tell. Access to the fortress
from outside the city was via what has become known as the 'postern'
gate. In the vicinity of the Royal Palace were a number of other impressive buildings, including what may have been official residences
and the royal stables. In a residential area within this north-western
part of the city were houses whose occupants must have been significant
or wealthy citizens. T h e owners of some of the houses are known.
T h a t of Rap'anu contained a library of texts and had over thirty
rooms. T h a t of Rasap'abu, a tax collector, also contained a library, as
did the house of an unknown person nicknamed 'the Scholar'. T h e
presence of some forty alabaster vessels led to another of these houses
being named the 'House of Alabasters'. Evidence of staircases points
to the fact that many of the houses would have had an upper storey,
where it is probable that much of the living accommodation would
have been located.
Arrangements for sanitation and water supply in the larger houses
was often impressive. Some were built round courtyards which contained a well and perhaps a trough into which the water, once drawn,
would be poured. Used water was conveyed away along channels
and gutters. Beneath the houses (or courtyards) were carefully constructed family tombs with vaulted ceilings reminiscent of a type of
Cretan tomb. A staircase would lead down into the funerary vault,
which was paved and whose walls contained niches or 'windows'. It
seems likely that the bodies were not placed in coffins, but were laid
direcdy on the floor, probably wrapped in shrouds. T h e discovery
of various items of funerary equipment (despite evidence of the activity of tomb robbers who had presumably removed objects which

they considered to be of value) suggests that perhaps it was felt necessary to make some sort of provision for the dead, a fact which, if
correct, would be of relevance for the question whether the people
of Ugarit believed in some form of afterlife. T h e presence of cups
may also indicate the notion of the provision of sustenance for the
dead. Earlier descriptions of these tombs noted the presence of clay
pipes which were thought to be for the purpose of providing liquid
(or perhaps even libations) for the dead. But it is perhaps more likely
that these were less glamorously part of the drainage system (PITARD

In some of the excavated areas further from the palace, for example to the south of the acropolis and in the so-called 'Southern City',
the houses were often rather smaller and closer together, built along
narrow streets. In the latter area, where it seems likely that some of
the city's artisans and craftsmen lived, there is evidence that houses
were built around a public square, close to which was a large building which contained a library of texts.
Preliminary excavation reports have been published in the journal Syria.


This survey has already mentioned craftsmen and metal-workers,

scribes and texts, stonework and carved ivory, elaborate drinking vessels, statuettes and other representations of deities. All such things
suggest that the city of Ugarit was a place of some culture. A feature of the discoveries at Ugarit is the variety of artistic influences
which they reveal, notably from the Aegean world, but also from
further afield, e.g. Egypt. It is not always clear to what extent such
objects are imports (Ugarit was after all a centre of commerce), or
local products influenced by the artistic styles and techniques of other
regions with which the city had contact.
Ugarit seems to have been renowned for its metal-working. A feature of the discoveries has been the number of different types of
weight which have been found. M a n y of these are geometrically
shaped, but others take the form of e.g. a bull (SCHAEFFER 1 9 3 9 , pll.
X X , XXI) or even a h u m a n head (ibid. pl. XXI). A weight in the
shape of a head was found among a set of weights discovered along
with the bronze pans from a pair of scales, and gave rise to the

speculation that this was perhaps a replica of the head of the metal
worker himself, and that he was therby making it absolutely clear
whose weights they were! T h e actual value of the various weights
suggests that both the Mesopotamian and Egyptian weighing systems
were in use in Ugarit.
Two particularly fine examples of metal-work deserve mention.
These were a bowl (sometimes described as a cup) and a rimmed
plate or patera, both made of gold, which were unearthed together
in 1 9 3 2 (Fig. 1 3 , p. 5 9 4 ; SCHAEFFER 1 9 3 9 pll. XVII, XVIII^ ibid.,
1 9 4 9 , 1 - 4 8 ) . T h e bowl, 17 cm in diameter, was embossed with various decorative features, notably three concentric circles of animallike figures, some of which appear to be winged and are probably
mythological creatures. T h e principal scene depicted on the rimmed
plate (19 cm in diameter) is easier to interpret. It is a hunting scene,
showing a figure (often thought to be the king) in a chariot, armed
with a bow and arrows, in pursuit of various animals and followed
by a dog. Noteworthy among other items of gold which have been
found at Ugarit are a number of pendants, some of which were decorated with geometric devices, e.g. stars (see SCHAEFFER 1 9 3 9 , pi.
X X X I I fig. 1). Other pendants depicted a naked female, sometimes
showing the full figure from head to feet (Fig. 14, p. 596), and sometimes showing just the head and torso with particular concentration
on the breasts and pubic region (see SCHAEFFER 1 9 3 9 , pl. X X I X
fig. 1). It is thought likely that these were representations of a goddess, probably associated with fertility.
In addition to evidence of skilled metal work, it is also clear that
carved ivory was used as a decorative feature. Mention has already
been made of the little ivory box-lid, of Mycenaean style, perhaps
depicting a fertility goddess. Various ivory items were found during
excavations in the Royal Palace in 1952, one of which was particularly impressive (SCHAEFFER 1954b; C A Q U O T - SZNYCER 1980, pll.
XXVIII, XXIX). This was a large ivory panel, measuring approximately 1 m by 50 cm, comprising eight smaller plaques, six of which
were carved with scenes and two (at either end) were representations of trees. Because of the fragile state of the panel, it could only
be removed from the ground with some difficulty, but, when the
task was eventually achieved, it became evident that this was an even
more impressive piece than had at first been appreciated. There was,
in fact, another set of plaques underneath the first, making it clear
that this was a double-sided panel, comprising sixteen panels in total,

which had probably decorated an item of furnitureperhaps a couch

or a bed. Some of the pictures are clearer than others, and it is possible that they are to be understood as depicting scenes from the
life of the king. In one he appears to be about to thrust a spear
into an animal, while in another he is about to put out the eyes of
an enemy w h o m he grasps by the hair. But not all are quite so
gruesome since he is also shown with (and perhaps about to embrace)
his wife. O n e other panel deserves special mention. It shows two
smaller figures sucking the breasts of a larger winged figure with a
horned head-dress, presumably a goddess (SCHAEFFER 1954a, pl. VIII;
C A Q U O T - SZNYCER 1980, pi. XXIXb). This depiction calls to mind
an indication in the story of Keret that his hoped-for heir will suck
the breasts of Anat (or perhaps R a h m a y [ W Y A T T 1998C, 209]) and
Before leaving the heading of 'Art and Culture', it is important
that a word is said about music at Ugarit (CAUBET 1996a). A m o n g
the discoveries have been a n u m b e r of objects which are clearly or
probably musical instruments. These include horns made from the
ivory of the elephant and of the hippopotamus, a pair of small bronze
cymbals, and what are probably scrapers and clappers used in musical accompaniments. T h e picture provided by such discoveries is
enhanced by a number of representations of musical activities. Cylinder
seal impressions (AMIET 1992, nos. 265, 273) have been interpreted
as showing dancers and, perhaps, acrobats. A small bronze shows a
kneeling figure playing cymbals or a tambourine, and a rather damaged limestone figure may represent someone playing a double-flute.
(Line drawings of the above are given in C A U B E T 1996a.) Particularly
intriguing was the discovery of a tablet (RS 15.030+ = Ug 5, 463,
487) inscribed with the words of a hymn or prayer in the Hurrian
language and which seemed also to carry a sort of musical notation
indicating chords. T h e precise relationship between the hymn and
the music is not clear, but an attempt has actually been made to
reconstruct this piece of music, and to record it sung to the accompaniment of lyres ( K I L M E R 1974; K I L M E R et al. 1976).
For further discussion on the religious aspect of Ugaritian art, see
below 1 3 . 1 2 ( W Y A T T ) and 1 4 (CORNELIUS).


T o the south of Minet el-Beida, in the vicinity of the first discoveries,

were located the remains of an occupation which dated from the
14th century (and possibly even the late 15th century) BCE and lasted
until the time of the destruction of Ugarit. It has been suggested
that these might be the ruins of a quite separate city, Mahadou, but
it is probably appropriate to regard this as the port area of Ugarit,
to be associated with the city's commercial activities which many
textual discoveries attest, and that it was in this area that those particularly involved in maritime activities would have lived. T h e houses,
some of which suggest that they may have belonged to people of
substance, were built along straight, intersecting streets. Many were
arranged around courtyards, often containing a well, and comprised
several rooms. Underneath one of the rooms would be a vaulted
tomb approached by a staircase (see R S O 1, 3). Other buildings
were stores and warehouses, and one was found to contain some 24
large pottery jars, many in a remarkable state of preservation. In
fact, large numbers of pottery vessels of various types but often suggesting a Cypriot or Mycenean origin and probably used in commercial activities, were found in the area. (In this context it is relevant
to note that, on the cliff north of Minet el-Beida, a misfired Cypriot
vase was discovered, suggesting that there was a pottery workshop
nearby, and that perhaps Cypriot settlers lived in the vicinity of
Ugarit.) O t h e r discoveries included an Egyptian axe and several
Egyptian-style ivory cosmetic boxes, some of which were in the shape
of a duck (SCHAEFFER 1939b pl. X I V fig. 1). These pointed to trading links further south as well as with the Aegean world. Metal-work
finds included various bronze tools and weapons as well as silver
rings and lead ingots. Weights made of stone or haematite, cylinder
seals and, perhaps not surprisingly, stone anchors (some of which
had also been found on the tell; FROST 1969; SCHAEFFER 1978; FROST
1991) featured among the discoveries. Another noteworthy find was
a deposit of murex shells, for which the Mediterranean coast is noted,
used for the making of the purple dye.
T h e port area may also have contained its places of worship, as
is indicated by the presence of altars and other cultic installations,
including what may have been a small sanctuary. O n the port area
see now Y O N 1 9 9 7 .


About 8 km north of Latakia and 4.5 km south-west of Ras Shamra

there is a small cape which juts out into the Mediterranean. There
are some grounds for believing that it may have been an island in
the second millennium BCE (see below). It was already known that
there had been a Roman occupation there because of the remains
of buildings which were observable, and Gabriel Saad had noted
in 1965 that there was what appeared to be a low tell in the middle part of the cape. But it was yet another chance discovery which
led to excavations being carried out in this area. In the course of
earth-moving activities associated with urban developments (including plans to build the Meridien hotel) on the cape, a tomb was discovered in 1973 which gave rise to the possibility that there might
have been a significant occupation in the LBA. Thus, as a matter
of urgency, excavations were undertaken u n d e r the auspices of
the Direction gnrale des antiquits et des muses de Syrie, and a joint FrancoSyrian team was established, in charge of which were A. Bounni and
J . Lagarce ( L A G A R C E 1995).
When work began in earnest in 1975 it took place on the southern side of the low tell, in the face of constant difficulties caused by
the presence of modern earth-moving machinery in the vicinity. It
was undertaken in the hope of clarifying the different phases of occupation. Evidence soon emerged of thick rubble walls which had often
been reduced to foundation-level by the prdations of later builders
in the Iron Age and the Hellenistic period who used its stones. What
was clear was that these were the walls of a building from the LBA.
This became known as the 'Southern Palace'. T h e following year,
a survey was undertaken using geophysical techniques which measured electrical resistivity. This provided useful indications of the
westward extent of the 'Southern Palace'. It was decided to remove
the surface soil by means of mechanical diggers, thereby making it
possible to gain a clearer impression of the plan of the southern part
of the building. It was in the course of this activity that the presence of pottery reminiscent of Mycenaean ware from the beginning
of the Iron Age was first noticed. Excavations in the area of the
'Southern Palace' continued until 1980, particular attention being
paid to the Hellenistic remains.
Studies of the geomorphology of the cape which have been carried out since 1976 have, as has already been indicated, given rise

to the speculation that it may have been an island during the second millennium BCE. Sections of stone paving which showed through
the sand from place to place to the south-east of the 'Southern
Palace' seemed to be the vestiges of a roadway. These pieces of paving seemed to predate the sand-bar on which part of the Hellenistic
town was situated. Further study of the roadway has been impossible since 1976, but, in 1991, radiocarbon tests were carried out on
two samples of natural cement formed between the paving-blocks of
the roadway when they were submerged as a result of an earlier
phase of erosion. These tests yielded dates of 1179-860 and 791-441
and led to the suggestion that the road must have been constructed
prior to 1179/860 and was subsequently submerged. T h e most likely
time for its construction, in view of what is known of the site, would
have been the LBA, the period of building of the 'Southern Palace'
(and other important buildings as will be noted later), and it is possible that the roadway originally led to the eastern entrance to the
'Southern Palace'.
In 1977, a new area close to the tomb which had been discovered in 1973 was opened up. It soon became clear that this was a
site of considerable importance and led to subsequent excavations
being concentrated in this area and on the edifice which has come
to be known as the 'Northern Palace'. Various soundings were undertaken with a view to establishing the extent of the building in the
LBA. It became clear that this was a major building including not
only residential areas but also an administrative centre and that it
housed workshops, e.g. for metal-working. It proved difficult to produce a plan of this 'Northern Palace', partly due to the fact that,
as in the 'Southern Palace', stones had been removed to be re-used
later (but see the plan in L A G A R C E 1 9 9 5 , 1 5 4 ) . Some indication of
its extent was provided by the presence of what appeared to be a
street running along its western periphery and, less certainly, another
to its east. T h e block which lay between these two limits has been
analysed as divisible into two quite distinct sections. T h e first, to the
south-west, was basically rectangular and arranged around a central
courtyard; the second, to the north-west, was a much more confused
conglomeration of rooms. It is thought unlikely that these two sections of the building were entirely separate, and that there must
therefore have been a corridor or passageway linking them. There
is some evidence which makes it possible to suggest the original function of some of the rooms. For example, an impressive room off the

central courtyard, approached by an entrance flanked by two columns,

may have been a throne room (room XII). Apparendy next door to
this 'throne room' was a workshop which seems to have been used
for working in bone (room XX). This juxtaposition of rooms seems
rather strange, and has led to the suggestion that the latter may
originally have been on the first floor but that, with the destruction
of the building, its contents fell through to ground level where they
were subsequently found. T h e presence of staircases suggests that the
building had more than one storey. Caution is therefore necessary
in identifying the original function of the building's rooms. Nevertheless,
it has been thought possible to make some judgements about the
likely use of certain parts of the building in the LBA.
T h e main access to the 'Northern Palace' was probably from the
south and into the rectangular south-western area of the building,
which seems from its stonework, general plan and lay-out around
the courtyard, to have been the most prestigious part of the edifice.
It has been described as the 'reception area'. By contrast, the northwestern part of the building followed a much less geometric plan
and gives the impression of being something of an annexe. This is
not to suggest that this part of the building was unimportant. It was
in this area that rooms housing tablets were located, and perhaps
where administrative activities were undertaken. Access from one
part of the building to the other may have been via a courtyard
and a room with benches which is perhaps to be understood as a
guardroom controlling access to the less public areas of the 'Northern
T h e 'Northern Palace' appears, then, to have been an important
residence, administrative and production centre. But whose residence?
The very nature of the building would suggest the likelihood that it was
a royal residence and one particular tablet ( K T U 2.82 = R I H 78/12)
found on the site may provide the clue since it is addressed ' T o the
queen, my mother'. Although the queen is not named, it is suggested
that she may have been Ahatmilku, the mother of 'Ammittamru II
who reigned in Ugarit in the middle of the thirteenth century BCE.
A seal impression bearing the imprint of this king was found in connection with administrative documents in the 'Northern Palace' in
1982 and 1983, indicating that some at least of the texts discovered
there originated during his reign.
Further excavations sought to establish the relationship between

the 'Northern Palace' and the buildings on cither side of it. T h e r e

are some grounds for the belief that the building which lay immediately to the east may have functioned as a service building for the
'Northern Palace', not least because there did not appear to be sufficient room for a completely separate building between the palace
and the eastern extremity of the city. More importantly, several of its
rooms seem to have served a utilitarian function, including cookery
and the baking of bread. T h e rooms seem to have been well constructed and paved, and one housed a toilet.
T h e building to the south-west (which became known as Building B)
was separated from the 'Northern Palace' by a street, though it is
possible that the buildings may have been joined in an unexcavated
area. T h e r e is some evidence of 'city planning' in this area, though
not to the extent which would have yielded a number of blocks separated by parallel streets. It seems that some of the streets went
round corners, which would have had the effect of reducing the
strength of the wind, and that at least one was a cul de sac. But the
symmetry of this building with the 'Northern Palace' is noteworthy
and it is not impossible that the two buildings were constructed to
the same basic plan. It too comprised a paved area (perhaps a courtyard) off which several rooms opened, one of which contained a
staircase. T h e symmetry even extends to the presence of two wells
in 'opposite' rooms, and kilns in 'opposite' rooms. T h e kiln in building was well-preserved and impressive though its precise function
was unclear. It was located in a room along whose eastern wall ran
a bench covered in white mortar which supported several vessels
and a lamp. T h e r e were pieces of ceramic and bone on the floor.
Another intriguing discovery in this building comprised about ten
ceramic objects which have been described as scoops. It has been
suggested that they may have been used for the distribution of rations,
a possibility which might support the suggestion that this too was a
public building.
H o w long did this LBA occupation last on Ras Ibn Hani? T h e r e
is some evidence of repairs or resurfacing of the floors, particularly
in the 'Northern Palace', which might suggest a relatively lengthy
period of occupation. It is also possible that there there is evidence
for the secondary usage of certain rooms, e.g. in the 'bakery' in the
'service building', and that a funerary vault under the 'Northern
Palace' was not, in its present state, contemporary with the building's

construction. However, there is no clear evidence, e.g. from the pottery found on the site, for a beginning of this occupation earlier than
the 13th century BCE. Pottery may be of more use in suggesting the
possibility that the end of Ras Ibn Hani mirrored the end of the
city of Ugarit itself, in view of the fact that people using a particular type of Mycenaean ware seem to have occupied the site immediately after its destruction. This would lend support to the view that
both Ugarit and Ras Ibn Hani were destroyed in the context of the
advance of the 'Sea Peoples' as they pressed south through the regions
of the Levantine coast.
T h e 'Northern Palace' seems to have suffered a violent destruction by fire, but not before it was abandoned and emptied of essential moveable items by the inhabitants. This fits with what appears
to have been the case in Ugarit. A similar situation seems to have
occurred with the 'Southern Palace', i.e. that it was emptied prior
to being destroyed by fire. However, it is not clear that this was the
case in 'Building B', where evidence of fire seems to be restricted
to a room which was probably used for cooking or baking and which
may therefore have been caused by that activity. It does not seem
to have been the result of a major conflagration involving the whole
building. T h u s caution is needed. Nevertheless, it is possible to suggest that archaeology (to some limited extent supported by the texts
found on the site) points to a foundation of what might be thought
of as this outer suburb of Ugarit not earlier than the late 14th century and more likely in the early 13th century BCE. This expansion
of Ugarit may reflect a period of relative stability and prosperity. Its
destruction was not later that the early 12th century and perhaps
more likely at the end of the 13th century, probably at the hands
of the 'Sea Peoples'.


It seems appropriate, therefore to think of ancient Ugarit as comprising not merely the city on the tell of Ras Shamra, but also, at
its zenith in the LBA, the city proper together with the port area
and the oudying suburb of Ras Ibn Hani. It was an important strategic and commercial centre, standing at the 'crossroads' of major land
and sea routes, and was doubtless quite cosmopolitan. It was a city
of impressive buildings, high culture and literary artistry, which has

bequeathed to later generations a script which may represent one of

the major steps forward in the development of writing systems. It is
perhaps fitting that the n a m e of the city should be best known
because of the language and method of writing to which it has given
its nameUgaritic.








T h e most surprising discovery made by the late C.F.A. Schaeffer

during his first campaign at Tell Ras Shamra was undoubtedly the
group of tablets written in an unknown cuneiform script.' This discovery attracted so much attention that the decipherment of the
script was accomplished in less than a year. 2 Overshadowed by the
tablets in the new (alphabetic) script was the find of a number of
texts written in a script already well-known to the excavators, the
Mcsopotamian (syllabic) cuneiform script. T h e first campaign yielded
only a handful of these texts 3 but during the years to come, and especially after the discovery of the royal palace, many syllabic texts came
to light. Not all of these were written in the languages of Mesopotamia, Sumerian 4 and Akkadian. A n u m b e r of tablets had been drawn
up in Hurrian 3 and in Hittite. 6 However, the number of tablets in
Sumerian and Akkadian (the latter is by far the most important) is
very large indeed and new ones are being discovered regularly. 7
In this chapter of the handbook I shall study the syllabic Akkadian
texts found at Ugarit by looking at their archaeological context and





1932, 9. S e e



V I R O L L E A U D 1929, 3 0 4 - 5 and PL. I . X X V I L X X V I I . For a complete list of the

P A R D E E 1989, 1 6 - 2 3 ; VAN
tablets found during the first campaign, see BORDRF.UIL
S O L D T 1991a, 5 3 2 - 5 .
Sumerian is only attested in schooltexts: lexical, literary and religious texts
copied by apprentice scribes, see below.
VAN S O L D T 1991a, 3 3 9 4 0 .
L A R O C H E , Ugaritica 5 , 7 6 9 - 7 9 .
See, for example, B O R D R E U I L - P A R D E E 1995a; D I E T R I C H
L O R E T Z 1994b; Y O N

1 9 9 5 ; BORDREIJIL



their general contents. A study of the former involves a survey of

the various archives where the texts have been discovered, a study
of the latter will encompass such varied aspects as the genres found
in the texts, their distribution over the archives, the education of the
scribes, and a short description of the characteristics of their Akkadian.
Naturally, in discussing the archives and genres I cannot avoid mentioning the alphabetic texts as well. T h e focus, however, will be on
the syllabic texts.

The archives

In contrast to many other excavators, Schaeffer at least attempted

to keep a record of every individual object which he found during
his excavations. He did so by assigning topographical points (points
topographiques, hence p.t.) which were written on a label attached to
the object and entered in a plan of the excavated area. Moreover,
he kept a notebook in which every object was described and listed
with its p.t. and the depth at which it had been found. A combination of the plan and the elevation of the findspot (deduced from
the depths and an elevation plan of 1928, before the excavations
had begun) would giveat least in principlethe correct findspot.8
First I shall discuss the archives found in the royal palace, then
we shall take a look at the ones found in private houses. In order
to save space I shall refer to the pertinent chapters in VAN S O L D T
1991a, where all the previous literature can be found. New publications are added wherever necessary.

The palace archives

T h e Western Archive 9 was located in rooms 3, 4 and 5 near the

main entrance 10 and contained almost only administrative texts, mainly

There are, however, many problems with the way the excavator kept his record.
During the first nine campaigns new p.t.s. were given for every new pit that was
opened. This led to duplicate numbers with the threat of confusion. Therefore, a
unified system was set up in 1938 which also covered previous seasons (VAN S O L D T
1991a, 673-4; the new p.t.s. are sometimes provided by B O R D R E U I L - P A R D E E
1989, 16-50). However, the multitude of p.t.s. given in the record for a single tablet
from the house of the High Priest (1929 1934) makes any attempt to locate them




See the plans in



1995a, 194 5.

in alphabetic cuneiform. Noteworthy are a few letters and schooltexts and especially two 'work copies' of the treaty with the Hittites,
in which the tribute is stipulated." Similar tablets were found elsewhere in the palace. Dated texts point to the time of 'Ammittamru
II and later (VAN S O L D T 1991a, 57-8), with the notable exception of
the translated 12 treaty from the time of Niqmaddu . Whether the
tablets had been stored on an upper storey cannot be ascertained.
No stairs were found in this part of the building.
T h e Eastern Archive 13 (rooms 54-56) is more diverse in contents
than the Western Archive, although administrative textsmainly in
Ugariticstill form by far the biggest group. Interesting is a small
group of juridical texts, most of which are styled as private contracts. 14 It is only from rooms 54 and 55 that we have royal deeds. 15
T h e most remarkable group of texts from this archive, however, is
formed by the letters. More than fifty letters were found, a fair
number of which can be ranked as international correspondence.
T h e letters were mainly addressed to king Ibirnu and his(?)16 queen
Taryelli. 17 Only a few texts survive from before this king. 18 T h e
tablets were at least partly stored on an upper storey.
The Central Archive 19 consists of three different wings with different


RS 1 1.732 (PRU 3, 181; 4, 47) and RS 1 1.772 (KTU 3.1). For the latter, see
K T U 2 199-200 and VAN S O L D T 1990a, 354-7.
Translations of Akkadian texts into Ugaritic such as K T U 3.1 = RS 11.772
are not really summaries nor are they faithful copies of the original. For K T U 3.1
see K N O P P E R S (1993), who suggests that the text is a covering letter including a tribute list sent by Niqmaddu. However, since the text is in Ugaritic, I tend to regard
it as a simplified copy for the use of the administrators in the Western Palace
archive. According to M I L L A R D 1995, 120, not all letters in Ugaritic which were
sent by foreign powers need to have had Akkadian or Hittite originals. The messenger could have memorized the message which was then written down in Ugarit.
VAN S O L D T 1991a, 6 0 - 7 3 .
RS 15.37 (PRU 3, 35), 15.81 (PRU 3, 37), 15.173 (PRU 3, 40), 15.180 (PRU
3, 36), 15.182 (PRU 3, 35), 17.248 (PRU 4, 236), 17.388 (PRU 6, no. 50), 17.426
(PRU 6, no. 51). Note also 15.128 (KTU 3.3), a legal text in Ugaritic.
RS 15.113 (PRU 3, 168), 15.114 (PRU 3, 112), 15.131 (PRU 3, 133).
VAN S O L D T 1985-6, 71; 1991a, 15-8.
That the diplomatic correspondence in general was kept in the eastern archive
as contended by Courtois 1988 is not true; many international letters were found
in other archives as well. It is possible, however, that the correspondence of Ibirnu
was concentrated in this wing of the palace.
Niqmepa': RS 15.117 (KTU 7.63); 'Ammittamru II: RS 15.114 (PRU 3, 112),
RS 15.131 (PRU 3, 133) and probably RS 17.383 (PRU 4, 221).
VAN S O L D T 1991a, 7 4 - 9 6 .

T h e northern wing is made up of rooms 30 and 31 and is the

most important. Here almost all the royal deeds concerning real
estate were filed and the distribution of the texts suggests that most
of them were kept in dossiers on the upper storey which, after the
destruction, fell on top of the debris in court IV. 20 Genres other than
legal texts as well as texts in Ugaritic are rare in this part of the
T h e eastern wing (room 64) contained mainly economic texts,
some of them in Ugaritic. T h e few legal texts are not concerned
with real estate. More important is a group of letters between king
'Ammittamru II and the king of Carchemish. A few other letters
(mainly in Ugaritic) are addressed to the queen. This queen was
most probably 'Ammittamru's mother Ahatmilku, although Taryelli
cannot be excluded. 21 Part of the tablets had been stored on the
upper storey.
T h e southern wing archive (room 66; tablets were also found in
65 and 67) consisted mainly of administrative texts, almost all in
Ugaritic. Of interest is a small group of legal texts from room 66
dealing with the sale of land to queen Taryelli and witnessed by a
certain Tipit-Ba'lu. 22
T h e Southern Archive 23 was located in a late addition to the
palace, rooms 68 and 69 to the south of court V. In this archive
all tablets were found that regulate the relations of Ugarit with the
foreign powers, the Hittite king and the king of Carchemish. T h e
archive proper was probably kept on an upper storey, which could
be reached through room 69, while room 68 served as a secretariate. Apart from the many treaties and international juridical texts, 24
however, we also find a number of local real estate transfers, as well
as a few administrative texts, some of them in Ugaritic. T h e dates
obtained from the texts cover the entire historical period at Ugarit.
T h e Southwestern Archive 25 (rooms 80 and 81) contained mainly
administrative texts in Ugaritic. O t h e r genres, such as letters and
schooltexts, were written in Ugaritic as well. Of special interest is the



1986, 2 0 0 - 3 ;








Published as Usfltica 5, nos. 159-61. For the seal of Tipit-Ba'lu, see ibid..,
p. 261.
VAN S O L D T 1991a, 97-109.
Published in PRU 4.
VAN S O L D T 1991a, 114-24.

synoptic table of scripts found in room 81, 26 A substantial group of

religious texts was written in Hurrian in syllabic script. 27 Two legal
texts 28 were found as well. All available dates point to the reign of
'Ammittamru II and later. T h e tablets could have been stored on
an upper storey, although one can also think of shelves along one
of the walls.
Miscellaneous tablet finds in the palace. Several groups of texts
have been found outside the archival rooms described above, such
as rooms 73 and 90 (both mainly administrative). T h e most important find, however, is that of a cluster of tablets in court V, formerly
believed to have been put there for baking. 29 As it turned out later,
the traces of oven material proved to be burnt debris, and the cluster of tablets was probably a mixed lot, some of which may even
have belonged to an older, discarded archive (at least according to
M A R G U E R O N 1995b, 66-7; M I L L A R D 1995, 119 speaks o f ' t h e result
of inverting a carefully packed basket'). A number of them, however,
have to be dated shordy before the final destruction of the city, as
shown by the translation into Ugaritic of a letter to king 'Ammurapi 3
( K T U 2.39 = RS 18.38).

The private archives

A number of houses proved to be important findspots of cuneiform

tablets. T h e owners of these private archives may have been scribes
themselves or may have needed the services of scribes. This is not
always clear and sometimes it even proves impossible to ascertain
the very name of the owner.
T h e so-called Southern Palace 30 probably was the house of the
chief administrator (atammu rabu) Yabni-apu (abbreviated Yabninu), 31
who had an administrative archive in rooms 203 and 204. Almost
all tablets are administrative in nature (all but a few in Akkadian)
and from a letter we learn about a scribe who, surprisingly enough,
has a good Akkadian name: Nahi-a1mu. 32 Provided the scribe him-

19.159 (KTU 5.14), see VAN S O L D T 1991a, 122 and 325.

Ugantua 5, 465-96.
RS 18.283 (PRU 6, no. 67) and 19.98 (PRU 6, no. 31). Especially the latter,
a real estate transfer, seems out of place.
VAN S O L D T 1991a, 1 1 0 - 4 .
C O U R T O I S 1990; VAN S O L D T 1991a, 1 4 9 - 5 8 and 1991b, 3 4 0 .
VAN S O L D T 1991a, 1 5 5 - 7 . C O U R T O I S 1990 (only Yabninu).
Spelled Vna-f}-ial-mu by his Ugaritic colleague (RS 19.53 = PRU 6, no.
18:2). F or the name, see (IAD s.v. nafyu le. For Middle Assyrian references, see



self was Mesopotamia!! and not a native of Ugarit in Mesopotamian

disguise, the rigid Mesopotamian orthography observed for texts from
this archive 33 may at least partly be due to him. In view of the many
Assyrianisms, both in sign forms and in grammar (see 3.1.5), we
have to assume that Nahi-a1mu was an Assyrian scribe working in
Ugarit. T h e archive was in use until the destruction of the city.
T h e house of Rasap'abu 3 4 is one of the three buildings containing
an archive discovered in the residential quarter east of the royal palace
(the two following archives were found here as well). T h e owner was
a man called Rasap-'abu, the supervisor of the harbour (Ma'hadu)
and the archive mainly consists of legal texts (partly dealing with
Rasap-'abu's affairs) and administrative texts. T h e legal texts date
from the reign of Niqmaddu II through that of c Ammittamru II.
There are only very few school texts from this house; two of them
are practice letters in Ugaritic. 3 '
T h e house of the lettr36 is named after its contents: literary, religious, and lexical texts. T h e house may originally have belonged
with that of Rasap-'abu, the name of the owner is still unknown.
Apparently, the house served as a school.
T h e house of Rap'nu 3 7 can easily lay claim to the status of the
most important school in Ugarit. Among its hundreds of tablets the
most important category is that of the lexical texts. Strangely enough,
hardly any literary and religious texts have come to light, which
seems to indicate that the scribes did not practise their knowledge
in context very much. 38 T h e presumed name of the owner is known
from three letters in which he appears twice as recipient and once
as sender. 39 Another important group of texts in this archive is formed
by the international letters. A large part of the royal correspondence
was kept here (and not in the palace), and it is here that we find
some of the famous letters dealing with the threat of the Sea Peoples.40
Since the architectural remains and the archive are still unpublished


1970 I, 344;














RS 17.63 (KTU 5.10) and 17.117 (5.11).



1979, 91. For Middle Babylonian,




1991a, 165-81.
VAN S O L D T 1995a, 179.
' Cf. Ugaritica 5, nos. 53-5.
For the correspondence in general, see Ugaritica 5, nos. 20-80; The Alashiya
letters are nos. 22 and 24 (no. 23 comes from the antiques market).


not much can be said about the stratigraphy of the building. T h e

dates obtained from the tablets are generally late: 'Ammittamru II
through c Ammurapi\
T h e Tablet House (Maison-aux-tablettes)^ has recentiy been published by C A L L O T (1994, 53-61). In the centre of a residential area
opened up in 1959 the excavators discovered a big house with a
large number of tablets. Most of these tablets are school texts: lexical and literary texts written by apprentice scribes. T h e few letters
and legal texts do not help to identify the owner of this important
house. T h e majority of the tablets had been stored on an upper
storey. However, a small group of texts was found below floor level.42
In view of the joins which can be made between tablets of both
groups (VAN S O L D T 1991a, 185, 187, 192)43 the two groups cannot
be separated in time (contra C A L L O T 1994, 61). Dates deduced from
the texts point to the time of 'Ammittamru II and later.
T h e Archives on the South Acropolis. In a heavily damaged building on the southern part of the Acropolis two archives were discovered; the first was named the archive of the Hurrian Priest, the
second the Lamatu-archive. 44 This first archive was located in two
rooms (10 and 11) in the northeastern wing of the house, one of
which ( 1 0 ) was interpreted as a cella ( C O U R T O I S 1 9 6 9 ) . T h e archive
contained almost exclusively texts in alphabetic cuneiform. T h e second archive was housed in the southwestern wing of the house and
contained a large number of syllabic cuneiform texts. Since most of
these are lexical and literary texts the wing must have housed a
school. T h a t the teacher(s) in this school must either have been
Babylonian or have been trained by Babylonian scribes can be seen
from the ductus in which they wrote and from their Akkadian. 45 T h e
few letters and legal texts do not allow an identification of the owner.
T h e tablets in the Lamatu-archive had partly eroded down the slope
of the Acropolis, but it is clear from the many joins that they belong
with the archive. 46 Circumstantial evidence points to a relatively late
date for both archives.



1991a, 1 8 2 - 9 3 .









The most striking example is RS 22.403 + 431B + 433A-C, a copy of Lu 1,

pieces of which were found at 0.70, 1.00 and 2.50 m below the surface.
For the former, see VAN S O L D T 1991a, 193-203, for the latter, ibid., 204-11.
VAN S O L D T 1991a, 209, 3 7 3 - 4 , 521.
VAN S O L D T 1991a, 209-11.

T h e Library of the High Priest 47 on the Acropolis was the first

archive to be discovered and, as it happens, it contained the literary texts in Ugaritic which brought Ugarit instant fame. Apart from
these important texts, the building must have housed a school, as
can be seen from a number of lexical texts. 48 As pointed out at the
beginning, the findspots of the tablets cannot always be identified
with certainty. An assessment of the stratigraphy will have to await
new research. Dates obtained from the texts probably point to the
period from Niqmaddu II (the Ugaritic literary texts)49 until the end
of Ugarit's existence (most other texts).
T h e house of Urtenu 5 0 was discovered by accident in 1971 after
a tablet had been found in debris from building activities by the
Syrian army. In 1973 the excavators were allowed to search the
debris but a regular excavation could only be carried out from 1986.
At the end of the 1994 season more than 500 texts had been recovered from this house, 51 thereby making it the biggest archive discovered so far. T h e great majority of the texts is in Akkadian, but
there are a few important texts in Ugaritic as well. O n e of these is
a literary fragment written by Ili-malku,' 2 the scribe who wrote a
number of literary texts in the house of the high priest. T w o groups
stand out among the Akkadian texts, a number of international letters
dealing with important historical events, 53 and a group of lexical texts
written by apprentice scribes. 54 T h e owner of the house, (the scribe?)
Urtenu, is known from a number of letters and an incantation in

1991a, 2 1 2 - 2 0 ; C U N C H I L L O S 1989.
1995a, 194.
For arguments for this date, see VAN S O L D T 1991a, 27-9. However, according
to B O R D R E U I L
M A L B R A N - L A B A T 1995, 447-8, the king mentioned in the colophons
is more likely to be Niqmaddu III than Niqmaddu II.
Y O N 1995; For Urtenu, see already VAN S O L D T 1991a, 221. B O R D R E U I L
M A L B R A N - L A B A T 1995; R S O 7 ; B O R D R E U I L
P A R D E E 1995b, 3 1 2 ; M A L B R A N - L A B A T
1995a, L O M B A R D 1995. For previous literature, see VAN S O L D T 1991a, 221-3.
YON 1995, 439.














In view of a number of syllabic spellings with -a-, I prefer the vocalization Ili-malku
to the more traditional Ili-milku, see provisionally van Soldt 1991a, 21 n. 182 and



3 n.



For example, the battle of Nihriya, R S O 7, no. 46, a letter from NorthBabylonia (no. 47, in line 3 read DIi-tlmi-, 'Shepherd of M a n ' [<Itr-Mer]),
and a letter from the Hittite king concerning the Sikila'ites, who are said to live on
boats (no. 1 2 ) . For a short description of the letters found in 1 9 9 4 , see B O R D R E U I L MALBRAN-LABAT



R S O 7, nos. 48-77.

Ugaritic. 53 His son Ur-Teub was also a prominent resident. 56 Almost

all tablets were found in a layer of ca. 1.50 m above the floor, only
two were lying under the floor. According to Lombard, the latter
probably form an older group, whereas the former belong to the
destruction level and had probably fallen from a higher elevation,
perhaps an upper storey. 57 T h e dates obtained from the texts point
to the time of N i q m e p a ' and later. 58
Miscellaneous tablet finds. For the sake of completeness, a few
more clusters of tablets should be mentioned. First, there is the small
group of texts found in the centre of Ugarit (Centre Ville) mainly consisting of schooltexts 59 and, second, we have a substantial group of mainly
Ugaritic texts from the northern palace at Ras Ibn Hani. 6 0 T h e latter can be dated to the period shortly before Ugarit's destruction.


Text genres

After this survey of the different archives it seems appropriate to

summarize the contents of the tablets and to look at their distribution in place (see VAN S O L D T 1991a, 133-40, 2 2 6 - 7 ) and time. For
reasons of convenience I shall discuss them by genre.
T h e treaties form one of the most important groups of texts. 61
T h e y provide us with a wealth of historical material not only pertaining to Ugarit itself but also to the states with which it was in
contact. Since Ugarit was a vassal of the Hittite king, the treaties
describe the relations between the two states. These important documents, which date from the entire historic period, were kept on file
in a separate wing of the palace, the southern archive. Copies and
summaries to be used in other archives were prepared and have
been found here and there. 62
VAN S O L D T 1991a, 221; B O R D R E U I L
P A R D E E 1995, 31. See also the literature
mentioned in note 50.
R S O 7, no. 25:1. New attestations can be found in B O R D R E U I L
MALBRANI^ABAT 1995, 446 and 449, where we probably have to read Ur-Teub instead of
Ur-Ba'al. For Ur-Teub as the son of Urtenu see B O R D R E U I L
1995, 446. For the name, see G R O N D A H L 1967, 423.
L O M B A R D 1 9 9 5 , 2 3 2 and cf. 2 3 7 . For the texts found in 1 9 9 4 , see Y O N 1 9 9 5 .



1991A, 2 2 2 - 2 2 3 ;









J . and E . L A G A R C E 1995; B O U N N I - S A L I B Y
L A G A R C E 1996. For earlier literature, see VAN S O L D T 1991a, 225.
Almost all treaties have been published in PRU 4. An additional fragment was
published as R S O 7, no. 1.
See my remarks to the texts from the western palace archive. Note also the

T h e international legal texts are records which deal with legal

matters involving persons from outside Ugarit. Most, but not all, of
these texts are verdicts passed by the Hittite king or by his viceroy,
the king of Carchemish. Sometimes a representative of the crown,
either a prince or a high official (often a kartappu), is deemed worthy to supervise a case. 63 Occasionally, we also find settlements
between kings of two different vassal states without Hittite supervision. 64 Ninety-five percent of these texts were found in the palace
and of these, eighty-five percent in the southern archive. 63 This suggests that the international legal texts, just like the treaties, were in
principle filed in the southern palace archive. Since the dates range
from Hattusili III to Ta1mi-Teub of Carchemish, the texts were
apparently kept on file for a long time.
T h e domestic legal texts deal with matters pertaining to the citystate only. They can be found in almost every archive 66 and could
date to any king of Ugarit. This means that these documents were
considered important enough to be kept on file for a long time. A
special case are the texts dealing with land transfers supervised by
the crown: with just a few exceptions all these texts were found in
the northern wing of the central palace archive. Legal texts dealing
with land transfers found elsewhere are rare and they usually belong
to private archives. 67 T h e texts from the palace were kept on the
upper storey and grouped according to dossier. Their dates range
from Niqmaddu II to Niqmaddu III. 68
T h e international letters are state letters exchanged between members of the royal family as well as high officials (such as the skinu)
of Ugarit and foreign rulers or their representatives. In view of the
importance of the contents of many of these letters it is surprising

list K T U 4.610 = RS 19.017 from the southwestern archive, in which the share of
the tribute for the Hittite king (argmn p) is calculated for every town and guild.
Compare, for example RS 17.244 (PRU 4, 231) in which two high Hittite
officials are supervising a settlement in court. In RS 17.314 (PRU 4, 189) prince
Arma-ziti is present.
Compare, for example, RS 17.228 (PRU 4, 141), a settlement between
'Ammitttamru II and augamuwa of Amurru on the famous matter of the former's (ex-)wife.
VAN S O L D T 1991a, 2 2 6 - 3 1 .
No legal texts have been found in the house of the Hurrian Priest or that of
the High Priest.
For example, RS 22.399 + 23.77 from the 'Tablet House'.
O r Ibirnu, if RS 15.139 (PRU 3, 166) still belongs with these texts.

to find them not only in the palace but also in some private archives. 69
Moreover, except for a later copy(!) of a letter from the time of
Niqmaddu II 70 filed in the southern palace archive (and probably
considered a state document), all the letters are dated to the time
of 'Ammittamru II or later. 71 Both this dispersion and the relatively
short period of preservation indicate that these texts were not considered as important as the treaties and the legal texts. Quite a few
letters give the name of the king of Ugarit as the sender, from which
we may conclude that in a n u m b e r of cases a copy of the letter was
kept. For the few international letters in Ugaritic, see the remarks
on the Western Palace archive.
T h e private letters are letters sent from private persons to the
royal family or to a high official (usually the skinu) or between private persons, one of whom can be outside Ugarit. T h e former are
usually found in the palace, the latter in private archives. Some letters of the first group, however, have also been found in the houses
of R a p ' n u and the 'Tablet House', 7 2 a situation familiar from that
of the international letters. This genre is found in almost every single archive; only the house of Rasap-'abu has not produced any private letters. T h e texts can be written in Akkadian or in Ugaritic,
two letters are in Hurrian. 7 3 By their very nature private letters can
seldom be dated. T h e few indications that we have point to a late
date, the reign of 'Ammittamru II and later.
T h e administrative texts are documents usually styled as lists in
which persons, either as individuals or as a group (towns and guilds),
are mentioned who receive or deliver commodities, who pay taxes
or are just listed with their place of residence. T h e majority of these
texts is written in Ugaritic, the rest is in Akkadian. T h e r e is not a
single archive which has not produced at least one administrative
text. T h e r e are, however, a n u m b e r of archives which can be labeled
'administrative' by the sheer quantities of texts of this genre in comparison to other genres. Such archives are the western (75%), the
eastern (64%), and the southwestern (75%) palace archives (76% of
all administrative texts come from the palace) and the southern palace

Especially in the houses of Rap'anu and Urtenu.

RS 17.334 (PRU 4, 54).
VAN S O L D T 1991a, 2 2 8 .
Rap'anu: Ugaritica 5, nos. 20, 44, 48, 49, 52; K T U 2.68 - RS 20.199; 'Tablet
House': RS 22.347 and 22.419.
R S 11.853 (PRU 3, 327) and R S 23.31, see VAN S O L D T 1991a, 364.

(90%). T h a t the last archive is mainly administrative is not surprising: its owner is Yabni-apu (Yabninu), the chief administrator
(.atammu rabu). Dates obtained from a number of texts point to the
last 50 years of Ugarit's existence, a time span which serves as an
indication for their limited importance. 7 4
T h e lexical texts form the primary study material for students of
Akkadian ( 3.1.4). Thus they can be expected in buildings which
served as schools. Since the palace has produced only two lexical
texts and two syllabic practice texts it probably did not house a
school. 75 Different is the situation in the private houses, where large
numbers of these texts have been uncovered. 7 6 Especially the houses
of the Lettr, R a p ' n u , Urtenu, the High Priest, the 'Tablet House'
and the Lamatu-archive were very rich in lexical material. All these
houses must have had a school within their walls. Dates are usually
lacking, so that we cannot be sure if the texts were kept for longer
than one or two generations. T h e lexical texts are closely associated
with the next genre.
T h e literary and religious texts are actually two groups with a
different purpose. Both groups, however, served as advanced study
material for students of Akkadian 77 ( 4) and were normally found
in archives which also show a sizable number of lexical texts. As with
the latter, the palace archives have produced just a handful, whereas the house of the Lettr, the Tablet House and especially the
Lamatu-archive have been rich sources. Also interesting is the number of texts from the City Centre: five out of eleven (the others were
lexical texts). Most conspicuous is the (almost) complete absence of
this genre from the houses of R a p ' n u and the High Priest. For the
possible implications, see 3.1.4. In the house of Urtenu literary
and lexical texts have recently been uncovered. 78 Dates are difficult
to ascertain for this genre, a characteristic shared by the lexical texts.
As for the Ugaritic literary and religious texts, for which the alphabetic
script had perhaps been introduced, 7 9 they are discussed in another

VAN S O L D T 1991a,
Not to be included
the southwestern palace




139, 231.
7 4 7 - 8 ; 1995a, 1 9 4 .
are, of course, the many texts in Hurrian, especially from
archive (Ugaritica 5, 462-96).




Note that the literary texts are the oldest alphabetic texts uncovered so far, at




In conclusion, one can say that the genres can be divided into
two groups, those that were kept on file for future reference and
those that were discarded not more than two, or at most three generations after they had been drawn up. T h e former group consists
of the treaties and legal texts, both international and domestic. They
were usually kept on file inside one of the palace archives. T h e latter group comprises the letters both international and private, the
administrative texts and the schooltexts (the lexical, literary and religious texts). These genres, especially the international letters, are
much more dispersed and probably had only ephemeral importance.


Scribal education80

As indicated in the previous paragraph, the education of scribes-tobe took place in private houses and not in the palace. T h e syllabic
cuneiform script and the languages for which it was used in Mesopotamia, Sumerian and Akkadian, was a highly complicated tool
which could only be learned during many years of study and practice. T h e study material was organized in a didactic manner, in an
order which ensured a rising level of difficulty. No school books were
available to the students; teachers knew the texts by heart and taught
from memory. 8 1 T h e order of the schooltexts can be deduced from
combinations of texts on single tablets and from catchlines. T h e most
elementary exercise was a list of signs arranged by their phonetic
values (tu - ta - ti, bu - ba - bi, etc.), then a few lists with simple
ideograms followed, of which we have unilingual (Sumerian) as well
as bilingual versions (Sumerian - Akkadian). These lists are the
'Silbenalphabet/-vokabular A' and the Syllabary A with its vocabulary. T h e latter was a complete list of simple signs, which in Ugarit
also had a trilingual and even a quadrilingual version: columns in
Hurrian and Ugaritic were added to the Sumerian and Akkadian.
After this probably came a long list of divine names (the 'Weidner
G o d List'), a list of grammatical forms and a table of measures, the

least if the date of Niqmaddu II can be upheld ( B O R D R E U I L

1995, 447-8 suggest Niqmaddu III). If so, it is possible that the alphabetic script gradually came to be used for other genres, such as letters and administrative texts,
as well.
See in general, VAN S O L D T 1995a.
Only two texts can be shown to have been written outside Ugarit: Ugaritica 5
nos. 119 and 169 = RS 20.121 and 25.421.

exact order is still uncertain. 82 W h e n the student had mastered these

exercises he was ready for the larger compendia, the series Harrahubullu, Lu, Izi and Diri. Harra-hubullu is a series of 15 tablets
which contains the terminology for writing legal and administrative
texts as well as a thematic catalogue of material culture. T h e series
Lu continues with a list of professions, and the last two series, Izi
and Diri, give compound ideograms.
W h e n the student started to work on a new lexical text he would
normally make an excerpt on which he wrote the ideogram, its pronunciation and the Akkadian translation (in a number of cases the
latter two are broken off and one may wonder if that was done for
didactic purposes). After this exceipt the student wrote the whole
text with an Akkadian translation (or even with Hurrian and Ugaritic
translations). T h e last stage was the unilingual version; the student did
not need the translation any more, but he would now memorize the
whole text.
W h e n the student had completed these lists he could practise his
knowledge in context. T o this end, he would write literary and religious texts, also from dictation, which often proved to be too difficult
for the would-be scribes. T h e many mistakes that they made testify
to the problems they had in understanding the texts. Some archives
are almost devoid of literary texts, although they produced plenty of
lexical texts (the houses of R a p ' n u and the High Priest). Perhaps
this means that the level of practising Akkadian words in context
was not taught in these schools.
W e know very little about the teachers. It seems that the profession of scribe was often kept in one family and that sons were probably taught by their fathers. 83 T h e names of the scribes who taught
the first students at Ugarit how to read and write are, unfortunately,
unknown. T h a t they probably came from a Hurrian-speaking area
can be deduced from the strong Hurrian influence on the Akkadian
of the first period (see 3.1.5) as well as the Hurrian translations in
a n u m b e r of vocabularies. At any rate, the study material derived
from Mesopotamia after the Old Babylonian period from a centre
other than Nippur (probably North Babylonia). 8 ' 1







1995a, 172 3.
1995a, 1 8 1 - 2 , 2 1 1 - 2 .
1995a, 1 8 2 .



The Akkadian of Ugant85

In this final paragraph I shall present a brief description of the

Akkadian as we find it in the texts written at Ugarit. T h e corpus
consists of all the genres enumerated in 3.1.4, with one restriction:
the literary and religious texts have to be excluded from a study of
syntax. As we have seen, these texts were written by the scribes as
part of their training. T h e y ultimately derive from Mesopotamia as
study material and therefore cannot be used for a syntactical study. 86
However, they should certainly be used for the study of paleography, orthography and morphology (contra D I E T R I C H 1 9 9 6 , 4 0 ) . 8 7 Rather
than giving a brief survey of all the grammatical features which we
find in the texts, I will limit myself to a description of the various
influences which can be detected during the ca. 150 years that texts
were written in the city. Since the study material came from Babylonia
and was probably early Middle Babylonian it is not surprising to
find that the Akkadian at Ugarit was basically Babylonian, 88 but with
a n u m b e r of influences from other languages: Hurrian, Assyrian and
Hurrian has left its mark in various ways. We find Hurrian translations in lexical texts, a n u m b e r of (mainly religious) Hurrian texts,
including two letters, 89 lots of Hurrian names, 90 and a few Hurrian

1989; VAN S O L D T 1991a.

1991a, 4 7 5 .
I am afraid I cannot agree with D I E T R I C H ' S statement ( 1 9 9 6 , 4 0 ) that 'these
[lexical and literary] texts cannot really be used in the [general] definition of the
"Akkadian of Ugarit"'. These texts were written from dictation by scribes whose
native language was Ugaritic and they show the same orthographical and morphological characteristics as the records of daily life. The scribes built their scribal career
on what they learned at school. They did not adopt a completely different grammar
once they had finished. That the texts ultimately came from a Hurrian-speaking
area (see 4) is irrelevant for the Akkadian of Ugarit. In Bogazky the texts show
Hittite influence, in Emar and Ugarit, West Semitic influence. It is these influences
which have to be detected and which I have tried to describe. In my opinion, the
lexical and literary texts are as important as the other genres and they should be
treated as such, with some reservations. Both Huehnergard and I have carefully
kept the two groups (school texts and records of daily life) apart in our grammars
(see, for example, H U E H N E R G A R D 1989, 9 - 1 0 and VAN S O L D T 1991a, 3 7 2 - 3 , 4 7 3 - 4 ) .
T o mention a few examples, there is only one clear case of Assyrian vowel
harmony (VAN S O L D T 1991a, 3 9 1 - 2 ) , the Middle Babylonian 2-umlaut occurs frequently (ibid. 390-1), and for / w / in anlaut the texts write wa- rather than - (ibid.,
3 8 9 - 9 0 ) . Many more examples can be cited.
See the list in VAN S O L D T 1991a, 339-40. For the alphabetic Hurrian texts,
see K T U 2, 654.
G R O N D A H L 1967, 203-67.




words in legal and administradve texts, such as the conjunction undu,

"when", 9 ' and the verbal form pu-fyu-ka-ru-i = pg.ugar.od.i, "she has
made an exchange". 9 2 T h e strongest influence, however, is found in
the orthography of the Akkadian texts. Documents written during
the reign of king Niqmaddu II and his immediate successors show
a number of features reminiscent of the orthography in the Akkadian
Mittanni letters. First, there is the occasional confusion of stops (such
as DU for / / , PA for / b / , etc.) and, second, there is the use of signs
to mark I el as opposed to H/ (such as the use of ci for / qe/ instead
of or QI). Both p h e n o m e n a can also be found in the Mittanni letters of Turatta. They probably find their origin in the syllabary
used by the Mittannian scribes for their Hurrian texts. 93 These features are very strong during the time of Niqmaddu II, but they start
disappearing during the time of his successors. During the reign of
his grand-son 'Ammittamru II they have practically disappeared.
Thus, forms of the verb leq are always spelled with 'Mittannian' GI
during the reigns of Niqmaddu II, but are slowly being replaced by
'Babylonian' during the reigns of Niqmepa c and c Ammittamru II,
which in turn must give way to the 'Assyrian' QI in the time Ibirnu
and Niqmaddu III. 94 T h e same phenomenon can be observed for
the conjunction undu.95 Probably also of Mittannian origin is the construction X-/w sa Y for the construct state or a simple construction
with sa during the reigns of earlier kings. 96
Assyrian influence is relatively weak (but present) in the older texts
from Ugarit. However, its influence grows gradually during the historic period, specifically in morphology. 97 Thus, we find the frequent
use of the Assyrian pronominal forms st and sit for Babylonian
and , and -sunu for Babylonian -unti. W e also find the Assyrian
genitive in -e/e and the -prefix in the third person of verbs pmae
aleph. T h e latter appears to be on the increase. 98 Assyrian features
appear to be especially prominent in the archives of the Southern
Palace and the Tablet House; compare, for example, the use of



2 0 1 ; VAN S O L D T



1991a, 350 note 209; see S P E I S E R 1955, 164b, who suggests that
the form is either a noun or a verbal form. H U E H N E R G A R D 1989, 93 opts for a noun.
VAN S O L D T 1991a, 375-81; 1995b, 208-9.
VAN S O L D T 1991a, 2 6 3 .
VAN S O L D T 1991a, 4 6 4 .
VAN S O L D T 1991a, 517; cf. H U E H N E R G A R D 1989, 227-9.
VAN S O L D T 1991a, 471.
VAN S O L D T 1991a, 4 3 0 .


Assyrian signs f o r m s " and the verb tadnu in texts from the Southern
Palace, 100 see below. Assyrian sign forms are also attested in some
texts from outside Ugarit. 101
Ugaritic made by far the strongest impact on the Akkadian texts.
However, it is not constant. In the older texts the Ugaritic influence
is mainly noticeable in syntax, but gradually it also becomes stronger
in morphology. It is impossible to list all the phenomena that can
be ascribed to Ugaritic influence. Instead, I shall select a few salient
examples. Most important of these is, again, syntax. From the earliest texts on, the Akkadian documents do not conform to standard
Babylonian grammar, but follow patterns that have their origin in
the language of Ugarit. Thus, instead of the order subject - object indirect object - verb (the verb is normally the last constituent of the
clause), the normal order of constuents in a main clause in Babylonian,
the texts from Ugarit usually have subject - verb - object - indirect object
and, even when they do use the Babylonian word order, the verb
can still be followed by an adverb. 102 In morphology the growing
preference for nouns with case vowels and triptotic declension (also
in lexical and literary texts!) in the construct state is the most obvious feature. 103 O t h e r examples are the lack of a subjunctive in subordinate clauses, the application of the Barth-Ginsberg law in verbal
prefixes, and the occasional use of Ugaritic verbal forms. 104
These three types of influence affected every single archive. However,
there is a certain distribution according to archive. For example,
Ugaritic influence, although noticeable everywhere, is particularly
strong in the domestic legal texts and, since most of these come from
the central palace archive, this type of influence is most obvious
there. Assyrian influence is strong in two archives: the Tablet House
and the Southern Palace. For the latter it is not difficult to find an
H U E H N E R G A R D 1989, 277; most of his examples in the second paragraph come
from the Southern Palace archive.
VAN S O L D T 1991a, 438; H U E H N E R G A R D 1989, 278.
As, for example, in R S O 7, no. 12, see also the photo, Ugaritica 7, pl. XI.
Note that the letter also shows a number of Assyrian grammatical characteristics.
VAN S O L D T 1991a, 4 8 5 .
H U E H N E R G A R D 1981; VAN S O L D T 1991a, 417-26; for the chronological distribution, see ibid., 418.
For the subjunctive, see VAN S O L D T 1991a, 440 and H U E H N E R G A R D 1989, 169.
For the Barth-Ginsberg law, see VAN S O L D T 1991 a, 431. The prefixes ta- and ti- are
not in free variation (thus H U E H N E R G A R D 1989, 159), but show a distribution in time:
ta- in older texts (Niqmaddu II and Niqmepa'), ti- in later texts ('Ammittamru II).
For Ugaritic verbal forms, see VAN S O L D T 1991a, 432, 436-7, 441-2.

explanation: the only scribe attested so far for this archive appears
to have been an Assyrian (see the Southern Palace archive, above).
In the former, we probably have to think of a solution along similar lines. At any rate, it is clear that foreign scribes were working
in Ugarit. T h e same is probably true for the Lamatu-archive. T h e
schooltexts found here show reasonably good Babylonian written with
Babylonian signs in a Babylonian orthography. T h e number of mistakes is much smaller than in archives like that of R a p ' n u and that
of the Tablet House. T h a t the students who wrote these texts (and
the one teacher mentioned in one of them) were natives of Ugarit
is clear from the colophons. 105 T h e name of a teacher from Babylonia
does not appear in the documents at our disposal.
A final word on the political situation during the time the texts
were written at Ugarit. As explained in VAN S O L D T 1991a ( 5 2 2 - 3 ) ,
the development of the various language influences is closely connected with the political situation in the final stage of the Late Bronze
Age, the period from which we have written documents from Ugarit. 106
At the beginning of this period the Mittanni empire was still an
important power, although under heavy attack from the Hittites. T h e
state of Mittanni was a cultural centre which was a much closer
neighbour to Ugarit than Assyria or Babylonia and we may assume
that Mittannian teachers probably worked in Syria. This would at
least explain the strong Hurrian influence that we find during the
earliest Ugaritic kings. After the defeat of Mittanni, Assyria slowly
took its place and Assyrian troops reached the Euphrates during the
first half of the thirteenth century. It is therefore not surprising to
find an increasing Assyrian influence and even an Assyrian scribe at
Ugarit. With Mittanni out of the way, the road to Babylon was open
again and correspondence with this part of Mesopotamia shows that
contacts existed. 107 Finally, Ugaritic influence, already strong at the
beginning of the historic period, becomes stronger as time passes.
T h a t the native language of the scribes at Ugarit interfered more
and more with their Akkadian is to be expected and provides a good
parallel for developments in other cities outside Mesopotamia where
Akkadian was written.
VAN S O L D T 1 9 8 8 , 3 1 6 ; 1 9 9 5 , 2 1 0 (students:. . . -iskur and . . . -la-na) and 2 1 1
(teacher: Gamir-Haddu son of Nu'me-Rasap).
For a survey of the history of Ugarit, see the chapter by I. S I N G E R ( 1 5 ) .
Babylonian letters are R S O 7, nos. 39, 40 and 47. The last one is from the
area of Mari.







During the first season of excavations at Ras Shamra in 1929, a number of clay tablets were found in the ruins of a house on the city's
acropolis. T o the great surprise of the excavators, the tablets were
written in a previously unknown script, rather than in the expected
Akkadian cuneiform. Charles Virolleaud, who was assigned the publication responsibilities, copied these peculiar texts and published the
facsimiles in Syria the following year, to allow scholars to work on
decipherment of the script. Within months Virolleaud, Bauer and
D h o r m e had each independently worked out a substantial percentage of the script (on the history of the decipherment, see 4.1). It
proved to be a cuneiform adaptation of the linear alphabetic script,
the ancestor of all the western alphabets, that had been invented in
the Levant sometime in the first half of the second millennium BCE.
With the decipherment of the script, scholars began to realize the
extraordinary treasure that had been found at Ugarit. T h e texts were
written in the local Semitic language of the town, and a number of
them had a religious content. During the excavations of the next
two seasons, astounding new tablets in the same language and script
emerged from the rubble in and around the house, and it became
clear that Schaeffer had discovered a library of 'Canaanite' religious
and literary texts unparalleled anywhere else in the Levant. For the
first time native texts which allowed a direct view into Canaanite
mythology, legend and cultic practice were available. But it also
became immediately clear that these texts showed astonishing cultural relationships with the Israelite literature of the Hebrew Bible.
In fact, the impact of these texts on biblical studies can hardly be
overestimated, and an argument can be made that the Ugaritic tablets
are the most significant single discovery this century for the study
of Israelite religion and the Hebrew Bible. 2

' I thank Dennis Pardee, who read over this chapter and made a number of
valuable corrections and comments. Any remaining inaccuracies, of course, are my
own responsibility.
Numerous books and articles have dealt with the relationship between the

T h e library recovered during the first few seasons of excavations

was not the only collection of alphabetic cuneiform tablets to be
found at Ugarit. In succeeding years, archive after archive was discovered throughout the ancient city, in the royal palace and related
buildings, in the homes of high government officials, and in those
of priests and scribes. Hardly a season of excavations has occurred
over the past seventy years without the discovery of more documents.
Surprisingly, no other excavation in the Levant has recovered a
substantial number of tablets in alphabetic cuneiform, with the exception of nearby Ras Ibn Hani, 4.5 km southwest of Ugarit, where a
palace belonging to the kings of Ugarit was excavated. Several sites,
including Minet el-Beida, Ugarit's harbour town, Tell Sukas, and
Tell Nebi M e n d in Syria; Kamid el-Loz and Sarepta in Lebanon;
and Tell Taanak, M o u n t T a b o r , and Beth Shemesh in Israel, have
p r o d u c e d individual tablets or inscribed objects with alphabetic
cuneiform inscriptions, thus indicating that the script at least was
known across the entire Levant. 3 Presumably other archives in the
script will eventually be discovered, but for now the tablets from
Ugarit essentially stand in lonely splendour.
As of the 1996 campaign there were approximately 1046 published alphabetic tablets and fragments of identifiable genres known
from Ugarit, as well as 217 unclassified fragments with legible signs,
69 inscribed objects other than tablets, ca. 416 unpublished fragments, most of them tiny, unclassified pieces, and 90 unpublished
tablets and fragments from the 1986-96 excavation of the house of
Urtenu, currently in preparation by the epigraphic team of the Mission de Ras Shamra-0ugat. Ras Ibn Hani has produced 39 published
texts of identifiable genres, one inscribed object, four unclassified
fragments and 93 as yet unpublished pieces. 4
Several editions of the tablets have been published over the years.
Most texts found between 1929 and 1939 received their initial publication in the journal Syra. A comprehensive edition of all the texts

Ugaritic texts and Israelite religion. Most recently see B R O O K E - C U R T I S - H E A L E Y

1994, W Y A T T 1996a and P I T A R D 1998.
All of these inscriptions have been included in KTU2.
The numbers here are calculations based on my analysis of B O R D R E U I L
1989, KTU2, and private communication concerning the recent finds in the house
of Urtenu from D. Pardee, to whom I am very grateful. Because of the complex
nature of the fragmentary state of the tablets, the numbers in this article must be
considered approximate.

from that period, including transcripdons, facsimiles and photographs,

was published in 1963 ( H E R D N E R 1963). T h e finds of the p o s t - W W
II campaigns have been published in the three official series of the
Mission archologique de Ras Shamra-Ougant, Palais Royal d'Ugarit (PRU
2 a n d 5), Ugaritica (5 a n d 7) and Ras Shamra-0ugart ( R S O 7). N e w
editions (transcriptions, facsimiles, photos a n d translations) of m a n y
of these texts, including the paramythological ( P A R D E E 1988b) and
cultic texts, as well as the letters, are in the process of publication
by the Mission's current epigraphic team, P. Bordreuil and D. Pardee,
in t h e R S O series. N e w tablet finds will also a p p e a r in R S O .
Comprehensive collections of the texts (transcriptions) have been produced by G O R D O N (1965), KTU (1976, second edition 1995), and
CUNCHILLOS - V I T A (1993a). And an edition of the texts for the computer, T h e Ugaritic Tablets Digital Edition, is also in progress (see
T h e r e is no comprehensive translation of the entire corpus, but
several translations of the m a j o r religious and literary texts and a few
translations of the letters have appeared. T h e most significant ones
are the following: G O R D O N 1949 and 1977; D R I V E R 1956; AISTLEITNER




1 and





1981a; X E L L A 1981; DE M O O R 1987; PARDEE 1997a;

(ed.) 1997; W Y A T T 1998C. G O R D O N 1949 and TO 2 include
translations of letters, in addition to the religious texts. M a n y of the
economic texts were translated into French in their editio princeps. A
complete listing of all inscriptions f o u n d at Ugarit a n d Ras Ibn H a n i
may be f o u n d in BORDREUIL - PARDEE 1989.


The archives and their locations

O n e of the most extraordinary aspects of the epigraphic finds at

Ugarit is the great n u m b e r of separate archives found in the city.
T h e largest quantity of tablets came from the Royal Palace, but substantial archives have also been discovered in eleven other locations.
Nine m a y be described as the private residences of m e m b e r s of the
Ugaritian u p p e r class. T h e s e include the House of the High Priest
on the Acropolis; the Houses of Rap'anu, Rashapabu and 'the Scholar'
(Maison du Lettr) in the residential quarter to the east of the Royal
Palace; the House of Yabninu (originally called the Southern Palace,
but now identified [ C O U R T O I S 1 9 9 0 , Y O N 1 9 9 8 : 5 7 ] as a private residence) just south of the Royal Palace; the House of the H u r r i a n

Priest (also known as the House of the Priest with the Liver and
Lung Models) in the area called the South Acropolis, which possessed two separate archives known as the Cella of Tablets, and the
Library of Lamashtu Texts; the House of Literary Texts in the area
called the Southern City (la Ville Sud), and the House of Urtenu in
the South Centre area. In addition to the private archives, a group
of administrative texts originated in the area just north of the palace,
called in the reports 'the Northwest Hill (Butte Nord-ouest) of the tell',
but the tablets are not identified as belonging to a specific building.
And finally, an important archive was found in the Northern Palace
at Ras Ibn Hani, an additional residence of the Ugaritian king. In
addition to these, smaller finds of texts occurred at numerous other
locations, both in houses and in plazas and streets. 5
T h e tablets from the Royal Palace were found primarily in five
discrete archives located in various wings of the building. All of these
archives were originally located on the upper floor of the palace and
fell into the ground-level rooms when the palace was destroyed. 6
Each of them contained both alphabetic and Akkadian tablets. Three
proved to possess more alphabetic than syllabic texts. These were
the West Archive (found in Rooms 3-5), which consisted of a number of documents dealing with the administration of the towns and
villages in the kingdom of Ugarit; the East Archive (Rooms 52~56),
another collection of administrative records concerning various goods,
weapons, personnel, etc.; and the Southwest Archive (Room 81),
again primarily administrative, but containing a few religious and
scribal texts as well. In addition, about 75 alphabetic texts (ca. 30

A good sense of the extent of finds throughout the city can be gained by examining the 'Index des Points Topographiques par Quartier et Locus' in B O R D R E U I L P A R D E E 1989, 423-45. Here each inscription is listed according to find spot. In
addition to the archives described above, small finds of texts were made in eleven
other buildings, as well as in numerous courtyards and plazas. In addition, there
are many tablets whose exact find spots are not known. The following description
of the archives is based on the extensive article on the archaeology of Ugarit by
C O U R T O I S (1979, 1155-1285), the new and important book by Y O N (1998a, 998b),
the discussions in VAN S O L D T 1991a, 47-231, and B O R D R E U I L - P A R D E E 1989.
This circumstance appears to be typical of most of the archives found at Ugarit.
Most of the tablets located in the private houses were also stored on the upper
floor and were retrieved in the rubble of the collapse. The chaotic nature of a
house collapse explains why tablets belonging to the same archive may be strewn
through more than one ground-level room and even outside in the street or plaza
abutting the house. See YON 1998a, 59, 78, 84, 101, 106 for discussions of this situation in specific houses.

intact) were found in Court V, in what was originally thought to

have been an oven for baking tablets. Now it appears that they were
simply part of the Southern Archive which fell from the upper floor
when the palace was destroyed ( Y o n 1 9 9 8 : 4 9 - 5 0 ) . These tablets
included about ten letters that appear to date to the very end of
Ugarit's existence, along with standard administrative texts. T h e other
two archives contained primarily Akkadian texts (the Central Archive
preserving numerous juridical texts, and the Southern Archive holding
important international documents), but both also had some alphabetic texts as well. Besides these large archives, numerous tablets and
fragments were found elsewhere in the palace. In fact, inscribed objects
were recovered from some 57 of the rooms and courtyards of the
Royal Palace (cf. B o r d r e u i l - P a r d e e 1 9 8 9 , 4 2 4 - 3 3 ) .
T h e archives from the other governmental locations varied somewhat
in genre. T h e collection found to the north of the Royal Palace (the
Northwest hill of the tell) was a mixture of Akkadian and alphabetic
administrative documents. But the Northern Palace at Ras Ibn Hani
contained tablets of several genres, including some religious texts and
letters, though the majority of the tablets again were administrative.
T h e contents of the archives found in the private houses show
considerable diversity. Collections of tablets were recovered from the
houses of four high government officials, those of Rap'anu, Rashapabu,
Yabninu, and Urtenu. In all four cases, most of the tablets were
written in Akkadian (see 3.1), but a few were in the alphabetic
script. T h e R a p ' a n u archive produced a number of important Akkadian international documents, but it also contained fourteen alphabetic administrative documents and three scribal exercise tablets. T h e
Rashapabu collection primarily consisted of personal documents in
Akkadian and contained only two scribal exercises in the alphabetic
script, along with two administrative tablets, and, most significandy,
the best preserved hippiatric text from Ugarit ( K T U 1.85 = RS
1 7 . 1 2 0 , see below). T h e texts in the House of Yabninu were largely
Akkadian, with a few alphabetic administrative documents. And the
House of Urtenu, in addition to its largely Akkadian material and
alphabetic administrative tablets, has so far yielded some 27 letters
and four religious texts.
T h e other five major house archives primarily contained religious
and literary texts. T h r e e of these, the House of the Scholar (Maison
du Lettr), the House of l iterary Texts, and the library of the Lamashtu
Texts, consisted almost entirely of Akkadian texts of various types,

with very few alphabetic inscriptions. T h e House of the Scholar produced only two alphabetic tablets, a letter and an administrative text,
within its lexical, religious and literary library. T h e House of Literary
Texts contained significant Babylonian literary works, but it also produced the famous alphabetic text in which someone (Anat?) appears
to eat the flesh and drink the blood of Baal ( K T U 1 . 9 6 = RS 2 2 . 2 2 5 ;
but cf. 6 . 5 . 3 ) , as well as a fragmentary hippiatric text, an abecedary
and a few administrative tablets. T h e Library of the Lamashtu Texts
added only two administrative tablets to the alphabetic corpus.
T h e other two locations contained the extensive collections of
Ugaritic literary and religious documents. T h e archive in the House
of the High Priest, the first one discovered at Ugarit, produced the
single most famous collection of texts from the site. Here were found
the major literary texts of Ugarit, alongside a number of cultic tablets,
letters, and administrative texts. T h e 'Cella of Tablets' in the House
of the Hurrian Priest provided a collection of paramythological and
cultic tablets all together in a single room. In addition to the tablets,
some thirty liver models and one lung model were found in the
room, several of which had alphabetic inscriptions naming the participants in various extispicy rituals.'


Genres of texts in alphabetic cuneiform

T h e alphabetic script was used for a variety of purposes by the

scribes of Ugarit, but it was generally limited to texts that dealt with
local matters, religious, economic or administrative. Texts with an
international function were written in Akkadian." Most of the alphabetic tablets are written in the local West Semitic language, Ugaritic,
but the script was also used to write texts in Hurrian.
T h e r e are a n u m b e r of ways in which the alphabetic tablets
may be classified by genre. G o r d o n ( 1 9 6 5 , 2 9 0 - 1 ) divided them into
ten basic categories, 9 while KTU2 places the tablets into six genres
A particularly good discussion of the archaeological aspects of the archive discovered in the Cella of Tablets may be found in P A R D E E 1988b, 4-12 (archaeological description by J.-C. Courtois) and 261 6 (Pardee's own discussion).
There are a few letters in alphabetic Ugaritic that have international content,
but it is most likely that they are Ugaritic translations of Akkadian originals or
BORDREUIL 1 9 9 2 : 6 . 7 1 1 ) .
drafts of letters to be translated into Akkadian ( P A R D E E






2.38 =








2.39 =







They are: (1) literary texts, (2) religious or ritual texts, (3) letters, (4) tribute,

(excluding the categories, 'Not Classified Texts,' and 'Illegible T a b lets.'). 10 P a r d e e - B o r d r e u i l ( 1 9 9 2 ) 7 0 6 - 2 1 subsume all of them
under three large umbrella designations ('Religious, Epistolary and
Administrative'). T h e following description makes use of the genre
divisions chosen by the editors of the Handbook, so that the reader
may easily move from this summary to the more detailed discussions in the succeeding chapters.

Literary texts

As of 1996 approximately 179 alphabetic tablets and fragments that

can be identified as religious a n d / o r literary documents had been
discovered at Ugarit and Ras Ibn H a n i . " O f these, some 148 are
in the Ugaritic language, while twenty-eight are Hurrian texts, and
three are in the Akkadian language, but written in the alphabetic
script. 12 These documents can be divided into two general categories:
(a) Literary texts, including the large poetic narratives. These constitute the only surviving collection of Canaanite mythological and
literary works and are thus the most important source currently
known for the study of Canaanite religion and culture. Their influence

(5) hippiatric texts, (6) administrative, statistical and business documents, (7) tags,
labels or indications of ownership, (8) Hurrian texts, (9) Akkadian texts written alphabetically, and (10) Miscellaneous.
They are (1) Literary and Religious Texts, (2) Letters, (3) Legal Texts, (4)
Economic Texts, (5) Scribal Exercises, (6) Inscriptions on Seals, Labels, Ivories, etc.
(KTU2, ix).
T h e find spots for the religious texts are as follows: fifty-six tablets, including
the major literary texts were found in the House of the High Priest during the first
four seasons of excavation (1929-32). Twenty others were found elsewhere on the
acropolis. T h e Cella of Tablets in the House of the Hurrian Priest, excavated during the twenty-fourth season of excavations (1961), provided sixty religious texts.
Seventeen others were found in various rooms of the Royal Palace ( K T U 1.78 =
RS 12.061, K T U 1.79 = RS 13.006, K T U 1.80 = RS 15.072, K T U 1.81 = RS
15.130, K T U 1.82 = RS 15.134, K T U 1.83 = RS 16.266, K T U 1.84 = RS
17.100[a]+, K T U 1.86 = RS 18.041, K T U 1.87 = RS 18.056, K T U 1.88 =
RS 18.107, K T U 1.89 = RS 18.[508], K T U 1.90 = RS 19.013, K T U 1.91 = RS
19.015, K T U 1.92 = RS 19.039, K T U 1.93 = RS 19.054, K T U 1.94 = RS 19.059,
K T U 1.95 = RS 19.179), and fourteen came from the North Palace at Ras Ibn
Hani (KTU 1.163-176 = RIH 78/14, 77/02B+, 77/04+, 77/08A+, 77/10A, 77/10B+,
78/20, 78/11, 78/16, 7 8 / 0 1 + , 78/04, 7 8 / 0 9 + , 77/18, 78/26). T h e few other
tablets of this genre were found in various houses on the site. This includes three
unpublished texts from the house of Urtenu, RS 88.0237, 92.2016 and 92.2014.
T h e alphabetic Akkadian texts are K T U 1.67 = RS 5.199; K T U 1.69 = RS
5.213; K T U 1.70 = RS 5.156+. O n e tablet, K T U 1.73 = RS 5.303 bis, has seven
lines of Akkadian in the alphabetic script, followed by eleven lines in Ugaritic.

on biblical studies has been enormous, by illuminating the cultural

background of Israelite religion and its early development. In addition, they have played an important role in advancing the study of
Hebrew grammar, philology and poetics, (b) Cultic texts, which have
also served the dual function of clarifying numerous elements of
Ugaritic/Canaanite and Israelite religious practice.
2.3.1 (a) T h e literary texts. This category includes the largest tablets
found at Ugarit, a series of carefully inscribed, multi-columned tablets
containing substantial narrative poems. T h r e e distinct multi-tablet
narratives have been recognized, along with some single-tablet poems
and fragments. T h e major narratives are: (1) the Baal-Anat Cycle, a
series of six tablets which recount three primary stories about Baal,
the Canaanite storm/fertility deity who was the patron deity of Ugarit,
and his rise to power among the gods. T h e first gives an account
of the conflict between Baal and Y a m m , 'Sea,' for dominance in the
council of the gods. T h e second story deals with the building of
Baal's palace, while the third describes his defeat by Mot, 'Death,'
and his subsequent rescue by his sister/wife(?), Anat. (2) The Legend
of Keret, a narrative poem about a king of the land of Hubur, who,
with the help of El, ruler of the Canaanite gods, is able to secure
a wife and produce progeny, only to find himself dealing with devastating illness and a rebellion by his ambitious eldest son. T h e narrative is, unfortunately, incomplete. Substantial portions of three
tablets of the epic are preserved, but at least one additional tablet
must have existed, since the third tablet ends in the middle of a
scene. (3) The Legend of Aqhat, a narrative poem concerning Danel, a
pious man who longs for a son. His desire is granted by El, and
Danel and his wife have the child Aqhat. When Aqhat has grown
to be a young man, he is murdered on the orders of the goddess
Anat after he refuses to give the goddess his special bow. Aqhat's
sister seeks out the murderer to avenge her brother's death, finds
him and is about to kill him when the final preserved tablet ends.
Again at least one tablet is missing at the end of the story, and it
is not certain whether the first preserved tablet is actually the beginning of the tale.
In addition to these texts, there are several other mythological
narratives whose stories are shorter and do not extend beyond a single tablet. For example, The Birth of the Pleasant Gods ( K T U 1.23 =
RS 2.002) and The Marriage ofNikkal and Yarih ( K T U 1.24 = RS 5.194)

provide reladvely straightforward narratives about the gods. Several

other, more fragmentary mythical tales have been preserved, but the
interpretation of their story lines is much more uncertain. In this
category we can place some texts concerning Baal (e.g., K T U 1.10 =
RS 3.362; K T U 1.12 = R S 2. [012]), three fragments dealing with
a group of beings called the Rpum ( K T U 1.20 = RS 3.348; K T U
I.21 = R S 2.[019]; and K T U 1.22 = R S 2. [024]), a small text
describing a conflict between a deity whose name is not preserved
(probably either Anat or Baal) and Y a m / N a h a r , the sea monster of
chaos ( K T U 1.83 = R S 16.266), 13 and a newly discovered mythological fragment (RS 92.2016), from the house of Urtenu in the
southern part of the city. T h e r e are also a number of tablets that
have a clear ritual function in which mythic themes play an important role. P a r d e e (1988b 261-6) has designated this type of text as
'paramythological.' Included in this category are such texts as (1)
K T U 1.108 = RS 24.252, a description of a feast a m o n g the gods
that appears to conclude with the call for a blessing upon the king
of Ugarit and the city itself. (2) K T U 1.100 = R S 24.244, a text
describing the plea of a horse goddess to several deities, seeking a
cure for a snakebite. At the end of the text the god Horon disperses
the venom. (3) K T U 1.114 = R S 24.258, an account of a feast
(marzihu) held by El, during which he gets drunk. This story concludes with a recipe for dealing with a hangover. (4) K T U 1.124 =
RS 24.272, a text that concerns healing the sick, but which describes
in mythic form a consultation with a deity called Ditanu.
Almost all of the major literary tablets (the Baal-Anat Cycle, the
Legends of Keret and Aqhat, T h e Birth of the Pleasant Gods, T h e
Marriage of Nikkal, the Rpum texts, etc.) were found in the archive
of the House of the High Priest on the acropolis. The three multi-tablet
works and a few additional fragments (two of the Rpum texts, and
the mythological fragment found in the house of Urtenu in 1992
(RS 92.2016), were produced by a single scribe named Ilimilku. A
colophon at the end of K T U 1.6 = RS 2. [009] indicates that he
did his work during the reign of a king Niqmaddu of Ugarit. Until
recendy scholars have generally assumed that this king was Niqmaddu
II, who reigned during the mid-fourteenth century. But with the discovery of a fragment of an Ilimilku mythological text (RS 92.2016)
The recognition that the monster being fought in this text is Y a m / N a h a r is
new. See the edition of the text in P I T A R D , 1 9 9 8 . Cf. also W Y A T T 1 9 9 8 C , 3 6 8 - 9 .

in the house of Urtenu, an official who served Niqmaddu III during the latter part of the thirteenth century, it seems more likely that
Ilimilku produced the tablets during the latter's reign. 14
2.3.1 (b) T h e cultic texts. Somewhat over one hundred alphabetic
tablets and fragments may be considered cultic texts. Most of them
are in the Ugaritic language, but twenty-nine are in Hurrian and
several contain a mixture of Ugaritic and Hurrian elements. 15 These
texts are considerably shorter than the literary tablets (the mythological and legendary narratives constitute approximately sixty percent of the preserved lines of the Ugaritic religious texts). These texts
may be divided into three general types:
(1) Ritual texts, which provide information about the performance
of rituals. Some of the literary tablets described above also fit into
this category. These texts are often very difficult to understand, since
they are 'professional texts,' intended for the use of the priests. They
contain obscure technical vocabulary, and do not provide explanations of the concisely described events that make up the rituals.
Examples of such ritual texts include K T U 1.161 = RS 34.126, a
funerary ritual for the deceased King Niqmaddu III; and an expiatory ritual found in several exemplars, but best preserved in K T U
1.40 = RS 1.002. (See 7.1).
(2) Offering and deity lists. This is the most common type of cultic
text found at Ugarit. Most of these simply provide listings of offerings
made to the gods. They usually give the name of the deity, the thing
offered, and sometimes the name of the type of offering (e.g., srp,
'burnt offering,' lmm, 'peace offering'). These lists are important for
identifying the gods actively worshipped at Ugarit, and, to an extent,
their relative positions in relation to one another. (See 7.2).
(3) Omen texts. T h e examination of the internal organs of sacrificed
animals and the study of abnormal animal and h u m a n foetuses were
two ways in which the priests looked for signs of the future. This
type of practice is well known throughout the Near East, and a number of texts from Ugarit reflect the local version of it. Five inscribed

On the date of Ilimilku, see B O R D R E U I L
M A L B R A N - I ^ A B A T 1995, 4 4 7 - 8 . For
the chronology of the kings of Ugarit, see the discussion in VAN S O L D T 1991a, 1 4 6 .
See, for example K T U 1.132 = RS 24.291, a sacrificial list that is made up
of a peculiar mixture of Hurrian and Ugaritic phrases, and K T U 1.148 = RS
24.643, another sacrificial list that has two sections in Ugaritic and one in Hurrian.

liver models and a lung model with an inscription on it illustrate

the practice of exdspicy. 16 In addition some texts listing omens related
to abnormal foetuses have been recovered ( K T U 1.103 + 1.145 =
R S 24.247; and K T U 1.140 = RS 24.302). From Ras ibn Hani
comes a tablet with omens related to astronomical phenomena ( K T U
1.163 = R I H 78/14). (See 7.3).
2.3.2 Correspondence
With the discovery of twenty-four as yet unpublished letters in the
house of Urtenu between 1988 and 1996, there are now over 110
letters in the alphabetic script, including those from Ras Ibn Hani.
Most of these are correspondence between members of the royal
court, including the royal family and the major officials of Ugarit.
While Akkadian was used in international correspondence, the Ugaritic
language and script were the vehicles for local and more personal
communication in the city. T h e r e are a number of letters to and
from the king of Ugarit, rarely named, and there are several letters
to the queen as well. O t h e r letters are from one official to another,
and there are a few personal letters from and to non-royal family
Nearly half of the letters (approximately 49) were found in the
Royal Palace, while 27 come from the house of Urtenu, whose key
position in the Ugaritic government at the end of the 13th and
beginning of the twelfth centuries is abundantly illustrated in them.
T h e other letters in alphabetic script were found in various buildings throughout the site. (See 8).

Legal texts

A very small n u m b e r of alphabetic tablets that deal with specifically

legal matters have been discovered so far at Ugaritbarely a dozen,
published and unpublished. A m o n g these tablets, we find a few contracts, agreements in which persons take on the responsibility to act
as security for loans, receipts for payment of redemption money to
release persons from a type of servitude known as unt, and royal
grants of property. This category overlaps with certain aspects of the


The liver texts are K T U 1.141-144, 155 = RS 24.312, 24.323, 24.326, 24.327
and 24.654. The lung model is K T U 1.127 = RS 24.277.

administrative texts, discussed below ( 9.2), and some scholars might

place a number of additional texts in this category.

Administrative texts

This is by far the most common genre of alphabetic document recovered at Ugarit. Nearly eight hundred have been published, and many
of the as yet unpublished texts fit into this category as well.17 T h e
majority of these tablets arc lists compiled for various purposes. There
are lists of names, professions, towns and villages, as well as records
of land transfers, deliveries of goods, distribution of rations, inventories, and payment of taxes. Most of these tablets are quite small,
possessing fewer than twenty lines. Only a handful of administrative
tablets exceed fifty lines of text. (See 10.1).

The hippiatrc texts

Four exemplars of a text describing treatments for various medical

conditions of horses have been found at Ugarit. W h a t makes this
text interesting is that the prescriptions for treatment of the various
equine ailments make no use of magical or other religious elements,
but rather focus on recipes for the mixing of various plants together,
which are then administered to the sick animal. T h u s they are medical in nature and find their closest parallels in Mesopotamian texts
that deal with human illnesses. Only one of the Ugaritic hippiatric
tablets is nearly complete ( K T U 1.85 = RS 17.120), but the occurrence of three other copies indicates the value placed on the treatments described here.

P A R D E E - B O R D R E U I L ( 1 9 9 2 ) , writing before the discoveries of 1 9 9 4 and 1 9 9 6 ,

estimated that there are approximately 900 administrative texts from Ugarit and
Ras ibn Hani. We can now add about forty from the 1994 finds and five from








W a l t e r



Due to the integration of Ugarit into the cuneiform writing tradition of Middle Babylonian Koine during the 14th century, clay tablets
in palace, priest and private libraries have reached us which show
that the indigenous Ugaritic population of the harbour town at the
close of the 13th century was multilingual: Besides documents in the
local language of Ugaritic and Middle Babylonian Koine were found
some in Hurrian and Hittite, demonstrating the ethnic mix of the
T h e discussion which follows starts from the evidence for Hurrian
and Hittite, with Hurrian plainly having priority since the documents
found so far outnumber those in Hittite. In terms of topic, linguistic,
cultic and historical questions are to the fore.


The Human texts

First we provide a list of the texts grouped according to which of

the various libraries they come f r o m a n d we depend on material
published before 1998. 2

Texts from palace libraries

Within the palace, only texts and fragments in syllabic script were
found. 3 Apart from a letter (RS 11.853) and an Akkadian-Hurrian
wisdom text (RS 15.010), these are exclusively songs which are religious in content with indications of melody and directions for playing, most of them have been preserved only as fragments: RS 14.015,
14.018, 15.030 + 049 + 17.387, 18.282, 19.084, 1 9 . 1 4 2 - 1 5 1 ,
19.153-155, 19.164evidently, to this group of texts also belong the

Isolated Cypriot texts and documents, with A1aia-Cyprus as their theme, provide grounds for the assumption that the close connections of Ugarit with its neighbouring Mediterranean island also entailed a Cypriot component of the population.
Cf. VAN S O L D T 1991a, 339-40.
If the fragment K T U 4.669 + 7.130 = RS 19.174A + 19.174c, the contents
of which cannot be defined, should prove to be Hurrian, it would be the only
Hurrian text in alphabetic script to be found in the palace.

scattered finds RS 9.253 and 9.483A from the northwest region of

the palace (Butte nord-ouest).
Except for RS 1 5 . 0 3 0 + . . RS 1 5 . 0 1 0 from the eastern archive
of the palace and the scattered finds RS 1 4 . 0 1 5 and 1 4 . 0 1 8 , all the
tablets come from the southwestern archive. T h e letter and the scattered finds were published in L a r o c h e 1 9 5 5 , the wisdom text in
N o u g a y r o l - L a r o c h e 1 9 5 5 and the remainder in L a r o c h e

Texts from the priests' libraries

T h e library of the Grand Prtre

During the first five campaigns (1929-33), on the acropolis and in

the maison du Grand Prtre4 between the two main temples of Baal
and Dagn, numerous clay tablets in alphabetic script were found
which throw light on the cultic and religious life of Ugarit. These
are written in Hurrian or a Mischsprache of Hurrian and Ugaritic.
From the house of the Grand Prtre come the following textswhere
possible we have given an indication of genre for each text:


RS 1. [066]
RS 1. [067]
1.34 z= RS 1 .[076]
1.35 = RS 1. [069]
RS 1 .[070]
RS 1 .[071]
RS 1.004
1.44 = RS 1.007

9. K T U 1.51
10. K T U 1.52
11. K T U 1.54

12. K T U 1.59
13. K T U 7.40
14. K T U 7.43



RS 1.027
RS 1.028 + 035
RS 1.034 + 045
RS 1. [ 0 4 9 a ]
RS 1 .[074]
RS 1.031



- uncertain
incense incantation for
- uncertain
incense incantation for
auka 6




Most recent edition: D I E T R I C H

MAYER 1994, 7 4 - 8 1 .
Most recent edition: D I E T R I C H
MAYER 1994, 8 1 - 5 .
' The tablets of numbers 1-13 were found together in room 1.

- list of sacrifices 8

15. K T U 1.60 = RS 2. [006]

16. K T U 1.64 = R S 3.372

In addition, from the area of the acropolis come the two tablets:
17. K T U 1.66 = RS 5.182
18. K T U 1.68 = RS 5.200

- uncertain
- uncertain

Even though no precise information about the findspot is available,

the following tablets must also be assigned to the same area:
list of sacrifices 9

19. K T U 1.26 = RS 1-11.[048]

20. K T U 1.30 = RS 1-11.[046]

Library of the Prtre Hounite

During the 24th campaign in 1961 a group of clay tablets written

in the alphabetic script was dug up in the southern area of the
'acropolis' in room 10 of the house of the Prtre Hounite,w of which
the contents were cultic and religious themes. O f these, eleven are
definitely in pure Hurrian or in a mixture of Hurrian and Ugaritic:
1. K T U 1.110 = R S 24.254
2. K T U 1.111 = RS 24.255
3. K T U 1.116 = RS 24.261
4. K T U 1.120 = RS 24.269 + 297
5. K T U [.125
6. K T U 1.128

RS 24.274
RS 24.278

7. K T U 1.131 = RS 24.285


Most recent edition:

Most recent edition:













1989, 298; cf.



list of sacrifices of the

oracular decision for a
royal sin-offering 12
offering at the Festival
of Atarte 13
fragment of an
death ritual 14
incense incantation for
incense incantation for
Ushara 1 6

1998; 1997b.
S O L D T 1991a, 194: 'Hurrian Priest'.














1995, 1 7 22.
1997b; 1998.
1994, 87 94.



1994, 94


8. K T U 1.132 = RS 24.291
9. K T U 1.135 = RS 24.295
10. K T U 1.148 = RS 24.643

11. K T U 1.149 = RS 24.644


ritual for the palace

goddess Pidray' 7
- palace offering from a
ceremonial ritual 18
palace offering from a
ceremonial ritual for
- unintelligible

Texts from private libraries

In the private libraries east of the palacefor instance in the House

of Rap'nu 1 9 and southeast of the palacefor instance in the Maison
aux tablettes20were found numerous fragments of multilingual lexical lists with a column in Hurrianthese texts have been discussed
above ( 3.1). 2 '
Noteworthy is the as yet unpublished letter RS 23.031 22 from outside Ugarit, which was discovered as a 'scattered find' on the edge
of the Maison aux tablettes. It is written in Akkadian heavily interspersed with Hurrian.

Formal aspects

As the foregoing lists show, the Hurrian texts in alphabetic script

are chiefly lists of sacrifices set within rituals, and incantations. T h e
first thing that deserves to be established is that the symbiosis between
Hurrians and Ugaritians, already evident in the use of the Ugaritic
alphabetic script, under the influence of an Akkadian tradition finds
its written expression both in language, i.e. grammar and lexicon,
and in cult and pantheon. 2 3 This is most evident in the lists of
sacrifices and in an oracular decision ( K T U 1.111), where the text
throughout is a mixture of languages: T h e ritual sections are in
Ugaritic, the lists of gods which are concerned with the sacrifices


Most recent edition:

Most recent edition:







1997b; 1998.
1989, 228; VAN S O L D T 1991a, 165-10.
1989, 282.292; VAN S O L D T 1991a, 182 91.





Photo: B O R D R E U I L P A R D E E 1 9 8 9 , 2 9 7 : Fig. 3 8 a : reverse with seal impression.
On the problem of symbiosis and its effects, e.g. of Akkadian on the Hurrian
M A Y E R 1992, 39-40.
of the Mittanni Letter see D I E T R I C H



and which come from both cultures, are however in Hurrian and
seem to be almost fossilized.
An essentially formal aspect is also the re-use of previously inscribed
tablets for which the term 'palimpsest' can be used, as shown for
example by K T U 1.114, K T U 1.116 and K T U 1.131, to mention
the three most important, where individual wedges and traces of
signs especially on uninscribed sections can only be explained if the
tablet had already been written on, before the new text was applied. 24
All these tablets come from the archive of the Prtre Hounite. T h e
priest responsible had re-used an already inscribed tablet after a new
covering for a text for a particular occasion. T h e reasons for this
practice can only be speculated on: perhaps to save on clay for a
text of lesser importance which was used only for a particular occasion a n d / o r lack of clay suitable for making tablets.


The Hittite texts

In spite of the close political relationship to the Hittite kingdom, especially to the sub-capital Carchemish, in Ugarit only a private Hittite
legal document (RS 17.109) and a polyglot Sumerian literary text
with a translation into Akkadian and Hittite (RS 25.421) have been
found. 2 5


The language of the Hurrian texts from Ugarit

Previous research

As yet there is no adequate appraisal of the Hurrians and of the

Hurrian of Ugarit in current descriptions of Hurrian studies. Detailed studies of g r a m m a r and lexicon dependent on the Mittanni
Letter cannot hide the fact that they are all based on the bad transcription 26 of a defective copy. 27 With these premisses it was without
exception difficult to establish links with Ugarit and also with Bogazky
where, to some extent, there was a very different Hurrian dialect.
In this connection the as yet unpublished textual material from

D I E T R I C H - L O R E T Z 1993 for K T U 1.114; D I E T R I C H
for K T U 1.116; 1995, 23 n. 58: as yet unnoticed.
See K M M E L 1969; N E U 1995b, 126-8.









1994, 96 n. 78,

E m a r must play a role which has still to be determined. As shown

by comparisons between the Akkadian literary texts from E m a r and
Ugarit, there were close scholarly ties between both places which
found expression in their common tradition ( v a n S o l d t 1994). Similarly, on the basis of their proximity in time and space, E m a r could
provide illumination on several points regarding the development of
the Hurrian of Ugarit.
V a n Soldt has shown in detail that the Babylonian of Ugarit came
under Hurrian influence only until the mid-13th century. Afterwards,
it was increasingly open to Assyrian influence ( v a n S o l d t 1991a,
519-23)for us today this drastic change is certainly one of the tangible results of the defeat of Hanigalbat. Remarkably, so far only
very few tablets in Hurrian from the period of Hurrian domination
before the mid-13th century have been found from excavations: Most
of them come from the end of the 13th century and so from the
time of Assyrian influence. At the same time these texts also show
that Hurrian played a role even in the royal house up to the fall of
Ugarit which is not to be underestimated.
T w o opposing traditions have to be distinguished in the Hurrian
texts from Ugarit: O n the one hand there are texts from the palace
in syllabic cuneiform, which in morphology, less so in orthography,
exhibit interesting parallels with the Mittanni Letter. O n the other
hand, there are texts in alphabetic cuneiform, which were principally
found in two priestly libraries and reflect cultic practice. T h e latter,
with their consonantal script, present many problems for the recording of the Hurrian of Ugarit and also give us much to think about.
In any case, the conclusions which S p e i s e r (1941-2) and B u s h (1964)
drew in respect of the consonantal inventory of Hurrian on the basis
of texts in the Ugaritic cuneiform alphabet require correction. T h e
main problem is that the consonantal script itself still conceals barely
comprehensible difficulties even for Semitists. 2 " All the more reason
for Hurrian studies to exercise caution in the adoption of the preliminary results of Semitic research. O n top of that comes the difficulty
that H u r r i a n itself through a long tradition in the West(-Syrian)
cuneiform system has been moulded by Middle Babylonian Koine
and has thus lost many of its characteristic features.

See most recently





Phonology and orthography


T h e Hurrian texts written in the alphabetic script distinguish only

two sibilants: t denotes the voiceless sibilantsuch as unt (unub e)
( K T U 1.128:6.7.9), tdn (tad=en) ( K T U 1.44:3[to], similarly K T U
1.128:8), tritt (tan=at = t=a) ( K T U 1.44:11-2, K T U 1.128:20-21)and d the voiced sibilante.g. endr (en-nadura) K T U 1.128:4.6, jjllyd
[hill-ill-y add) ( K T U 1.44:4-5). This matches the observations made
by T r o p p e r concerning the inventory of consonants in the alphabetic
script of Ugarit. 29 Thus, all attempts to reconstruct the phonemic
inventory of H u r r i a n n o t only the sibilantson the basis of alphabetic spellings are premature and to be rejected.
If a different sibilant grapheme occurs then the word in question
can only be a loanword from Semitic culture. T h u s words like sp
(sup(p) = e) ( K T U 1.54:13), bsl (bask) ( K T U 1.44:5), zg (zagge) ( K T U
1.44:10) or rznn (rz-enni= ne) ( K T U 1.128:12) are equally Semitic
loanwords as e.g. kl (sukkalle) ( K T U 1.44:10, K T U 1.54:14, K T U
1.128:16, K T U 1.131:15) uhr (uhara) ( K T U 1.131:1.11.12) or mr
(maru) ( K T U 1.131:15).
It is noteworthy that the ergative ending
on a participle of
action, which as yet is only attested with the verb at 'to release',
occurs as t ( K T U 1.44:8; K T U 1.54:13), as ( K T U 1.128:17) and
as d, ( K T U 1.131:11). This results therefore in some uncertainty in
the way the ergative ending is written, which as is known from the
morphology of Ugaritic and Koine texts, can be conditioned by the
proximity of t with a partial assimilation completed ad libitum.
T h e graphic representation of sibilants in the root tae 'gift', known
from the Mittanni Letter, is by means of a in tzgd
K T U 1.128:8 and corresponds to the signs containing in the Koine
syllabary. Besides, this also corresponds to the spellings in the Hurrian
incantation from Mari: ta-zi-.30

1994b, 25-49: sibilants and interdentals.

1939, 2, 7.8.9; cf. also the spellings of the
1.128:13) and Kzg (KTU 1.116:6.14).






(KTU Velars
T h e emphatic velar q also indicates that a word in a Hurrian context is a loanword from Semitic: ql (ql) 'with a loud voice' ( K T U
1.44:4). T h e liquids r and / as word-initial
Since there is no initial I or r in Hurrian, it can be assumed that
words beginning this way correspond to Semitic foreign and loanwords. T h u s we derive rznn (rzenni-ne) ( K T U 1.128:12) from Akk.
rsu 'helper' and li (le'e) ( K T U 1.128:4) from Akk. leu 'capable, able'
and consider them as foreign words.
In a borrowed word, initial r can be avoided either by a preposed
vowel (cf. the D N Irappa < Raapu) or by means of a metathesis between r and the following consonant, as in the case of trbn (,tarb = enne)
( K T U 1.54:3), which derives from Akk. rsibtu 'one who commands
respect'. T h e uncertainty in respect of liquids emerges also in the
attribute bsl (basle) ( K T U 1.44:5), occurring in a context which is
parallel to that of srbn (sarb-enne) 'rich in willows' ( K T U 1.131:3.5).
Since loanwords of this kind occur in a H u m a n i z e d form, it has
to be accepted that they belong to an earlier stage of borrowing
than those which still retain an unchanged Semitic phonology. T h e tenuis d
T h e tenuis d is inserted in earlier borrowings where the underlying
Semitic word has the emphatic dental /: pdrrn (padr-ma(n)) K T U
1.44:11 and K T U 1.128:20, which must go back to Akk. patru 'to
Before the stressed vowel u an original t is 'weakened' to d as
shown by the example mrdml (mrdmele) K T U 1.131:2, which is the
H u m a n i z e d form of Akk. mrtma-ilim 'El's daughter'. This phenomenon also occurs in the Mittanni-Letter, where it also applies to
Hurrian words.
Matres lectionis
In line with usage in late Ugaritic texts, the incantations show a certain tendency to plene writings, with y for | i / / | and if for | u / / | :
kby kbny (kubb = i kib-enn = iye), mryt (mite) ( K T U 1.131:2), nw (re'u)
K T U 1.42:38 and ilwny (iln) ( K T U 1.128:17-8). Hiatus
T h e auka-Incantation transmits this divine name in the form twtk
( K T U 1.54:2.11.13) and is thus outside normal tradition, which uses
the M-aleph for the glide: tutk ( K T U 1.42:22; K T U 1.59:1.5; K T U
1.60:2; K T U 1.64:26; K T U 1.116:3.9.13; K T U 1.120:3; K T U 1.135:2;
K T U 1.149:10.11).
T h e word 'heaven' only occurs in the spelling with a w-aleph: hum
(haurunne) ( K T U 1.68:27 and K T U 1.128:2).

Morphology Nouns
T h e following determined or undetermined cases are attested: the
absolutive, ergative, directive, ablative and comitative.
T h e absolutive: Alongside the use of the absolutive as an adverbial it also functions as a vocative. As far as can be ascertained,
the forms of this absolutive used as a vocative are not determined.
T h e r e are too few occurrences for making a reliable assertion as
to whether there really is a connection between indeterminateness and the vocative.
T h e ergative shows no deviation from normal usage.
T h e directive and the ablative. Apart from the use of the directive,
which is well documented, and in Hurrian corresponds to West
Semitic and Ugaritic I 'to, for, towards' it also occurs in the El
Incantation with an ablative function, whereas the ablative occurs
only once: il.dn (ile-dan) 'away from ( K T U 1.128:16), where
it has been retained in a fixed formula. In general the findings
of the Mittanni Letter confirm this, where the ablative seems to
vanish in favour of the directive. 3 '
T h e comitative. T h e determinate plural of the comitative is relatively well documented: ath.ndrm (athe- nadura - ma[n\) ( K T U 1.128:3),
in(.)dr (en-nadura) ( K T U 1.128:4.6), trnd.rm
( K T U 1.128:5). However, it occurs as indeterminate in the UharaIncantation: armdr (ar-urn-adura)
( K T U 1.131:14). Pronouns
T h e enclitic personal pronoun, 2. sg., occurs in hldp (held= a=ppa[n])
'you (sg.) are exalted' ( K T U 1.128:4.5).





150-1. Verbs
Transitive verbs are attested in the 3 pi. imperative of the indicative
( K T U 1.44:4-5) and the cohortative singular: agrl
uwln (agr-ile uw=ilen) shall incense (and) slaughter!' ( K T U 1.128:19).
In the intransitive conjugation the 1. sg.umtn (um-a-ttari)
1.131:12)and the 2. sg hldp (held= a= ppa[n]) ( K T U 1.128:4.5)are attested.


T w o simple intransitive/stative clauses form the end of the incantation in the style of noun clauses: gr a*r*mdr mr kl '(Here is) the
igT-vessel with the (offering that) has been brought. Maru (is the
Uhara) messenger' ( K T U 1.131:13-15). It cannot therefore be excluded that both gr and ski are in the locative-stative: *sagara and
*ukkalla respectively.
A special indication of Semitic influence on the Hurrian of Ugarit
is the compound: within the god-lists, unnamed gods are defined by
apposition. Thereby the apposition is, on the model of the Semitic
construct state, so closely connected that it forms a unit with it of
which only the apposition has the case-ending syntactically required
for both words; cf. such expressions as en tlnd 'for the /z/fe-god(s)'
( K T U 1.110:1; K T U 1.111:7) or ewrn prznd 'for the lord over the
decision' ( K T U 1.110:4) instead of *end tlnd or *ewmd prznd.


Incantations are particularly subject to the usual rules of poetic language with stylistic usages such as parallelism and chiasmus: Simple
parallelism is evident for example in K T U 1.131 in the opening
address (uhr mryt mrdml ttl srbn 'Ushara, Lady of Mari, || daughter
of El from Tuttul rich in willows', lines 1-3), with the chiastic positioning of its members in the following predicate (kby kbny mdm kt*[y\
'who plans her destiny, || who utters wisdom', lines 3-4). In the
second case the inversion of predicate and object in the second member requires the addition of the enclitic -m(a).
A typical poetic characteristic is also the emphasis of an independent imperative by the enclitic -m(a), as shown by the form him
'speak up!' ( K T U 1.131:13). In a non-poetic context this -m(a) irrespective of whether its origin is Hurrian or Ugariticcan also be
attached to nouns when at the head of a list: atf}lm 'a/AZ-offering
( f o r ) . . . ' ( K T U 1.110:1; K T U 1.111:3.8).



Whereas the rituals and sacrificial texts are written in a H u r r o Ugaritic Mischsprache, the incantations, in 'good' Hurrian probably
also reflect authentic Hurrian thought. However, they are directed
to a mixed Hurro-Syrian pantheon: to Kumarbe of Uriga and K u m m a
including the cult centre Tuttul, to Akkadian auka of Niniveh, to
Ushara of Mri and to El of Ugarit. In line with this mixed tradition, which ultimately represents the result of a lengthy symbiosis, it
is not surprising that the texts contain numerous Semitisms and also
countless borrowings from neighbouring Asia Minor. This finds expression not only in the names of the gods, their messengers and
their places of worship, but also in the lexicon. Here we have provided a list of lexemes which as yet have not been explained or have
been explained inadequately.

Semitic words

hannuge (hnng-) 'merciful' ( K T U 1.132:9) < Ug. hnn;

bazzizi= (hzz-) 'wisdom' ( K T U 1.116:5; K T U 1.125:11) < hassu;
iln (ilwny) 'divine' ( K T U 1.128:17-8 ) < *ilnyu;
le'e (H) 'powerful one' ( K T U 1.128:4) < lew,
kelage (lklg-) 'a pot' ( K T U 1.128:11) < (WSem.) qlh (Ug. qlh, Eg. qrh.t);
kalle- (Id-) 'cup, bowl' ( K T U 1.128:11) < kallu;
kad(d)ale= (kdl-) a container ( K T U 1.128:11) < ka/undulu;
mmurte- (mmrt-) 'audience gift' ( K T U 1.128:10) < nmurtu:
mnte (mryt) 'Lady of Mari' ( K T U 1.131:2) < mntu;
mrdmele (mrdml) 'El's daughter' ( K T U 1.131:2) < *mrtma-ilim;
mulugi= (mlg-) approximately 'gift' ( K T U 1.116:31) < mulgu
nadi- (nd-) 'what is laid' ( K T U 1.116:4) (bis) < nadi(u)
ninnaggi= (nng-) 'incense bowl' ( K T U 1.116:31) < *ninnakku <
nignakk/ qqu;
padri= (pdr-) 'release, solution' ( K T U 1.44:11; K T U 1.128:20)
< patru;
qle (ql) '(loud) voice' ( K T U 1.44:4) < (WSem.) *qlu;
re' (riw) 'shepherd' ( K T U 1.42:38) < r';
rzenne (rzn-) 'helper' ( K T U 1.128:12) < rsu;
sikitt-enne(sktn-) 'one who creates living things' ( K T U 1.125:1)
< sikittu;
sup(p) (sp) 'to pray' ( K T U 1.54:13) < suppw,
sarb/ basle (srb-/bsl) 'rich in willows' ( K T U 1.44:5; K T U 1.131:5) <

agare (gr) a container ( K T U 1.131:13) < akr (?);

s'ukkalle (ski) 'messenger of the gods' ( K T U 1.44:10; K T U 1.54:14;
K T U 1.128:16; K T U 1.131:15) < sukkallu;
tarb- enne (trbn) 'one who commands respect' ( K T U 1.54:3) < ribtu;
zogge (zg) 'first-rate, foremost' ( K T U 1.44:10) < sank < sag.

Hurrian and non-Semitic words

ag=uge= (agg-) approximately 'tray' ( K T U 1.128:10);

agruthe (agrth-) 'incense holder' ( K T U 1.125:14);
arum (arm-) approximately 'offering' ( K T U 1.131:14);
at= (at-) 'to release' ( K T U 1.44:6.8; K T U 1.54:10.12; K T U
1.128:16.17; K T U 1.131:10.11);
el= (el) 'to speak' ( K T U 1.128:1) ( Ur. ale);
ep= (ep-) 'to receive' ( K T U 1.44:3; K T U 1.131:5) (cf. Hitt. e/ap(p)-);
liubruthe- (ffbrt['}-) 'receptacle for a sacrifice or holy water' ( K T U
kubb- (kb-) 'to plan' ( K T U 1.131:3.4) (cf. Glossenkeil-W\ti. kup= I);
kiyade= (kyd-) 'sea' ( K T U 1.125:12);
muti= (mt-) 'justice' ( K T U 1.116:5);
nirul (nrl-) 'to prove to be good/merciful' ( K T U 1.125:3);
tagi= (tg-) 'beautiful' ( K T U 1.125:18; K T U 1.132:11);
taz-uge (tzg-) approximately 'offering table' K T U 1.128:8;
tad- (Id-) 'to accept' ( K T U 1.44:3 (bis); K T U 1.116:4; K T U
1.128:8) (cf. Ur. sat=);
tay-enn(tyn-) 'located by water' ( K T U 1.125:3);
tun (tn-) ' h a n d ' ( K T U 1.44:7; K T U 1.54:11; K T U 1.128:17;
K T U 1.131:11);
tiw= (tw-) 'to escort' ( K T U 1.125:4) (cf. Ur. sm=);
ude= (udr) 'result of extispicy' ( K T U 1.116:27; K T U 1.125:17)
(< Sum. uzu).


The pantheon of the cultic texts

Corresponding to their mixed language the rituals and offering lists

also reflect a mixed Hurro-Ugaritic pantheon.
As there is a representative list of Hurro-Ugaritic deities in the
central second section of K T U 1.116, in other sacrificial texts it is
also possible to start from corresponding sequences. T h u s it would
be possible to draw up a more or less 'canonical' list of the HurroUgaritic 'palace pantheon', corresponding to the Akkado-Ugaritic

'canon' 3 2 according to the alphabetic texts K T U 1.47 and K T U

1.118 and the syllabic text R S 20.24. 33
Besides K T U 1.116, the offering lists in K T U 1.26, 1.60, 1.110,
1.111, 1.125, 1.132 and 1.135 are of prime significancefrom these
are derived K T U 1.26 a n d 1.60 f r o m the House of the Grand
Prtre, the others, as well as K T U 1.116 from the House of the Prtre
K T U 1.26, 1.60 and 1.135 are only fragmentary and and each
comprises only the middle section of the tablet; the upper or lower
halves are missing. In preserved passages a pantheon is given which
largely follows the sequence of the 'palace pantheon', with minor
Starting with the structure of these three tablets, which is like
the structure in K T U 1.116, then the information which probably concerns specially celebrated deities has not been preserved.
K T U 1.110 is completely preserved and provides a list of deities
who are to be honoured seven times with an unspecified sacrifice. 34
These deities are the 'great ones', also given in the list of palacegods in K T U 1.116; auka and her female retinue and H a m m u ,
however, are not mentioned. Where they occur, the sequence has
minor discrepancies in the lower section of the list. T h u s Anat,
Simige and Nikkal come before the city god, and Nubadig closes
the list.
Even when subordinate deities as well as deified paraphernalia
and expressions of well-being are lacking, the agreement otherwise
with K T U 1.116 indicates that this is also a sacrifice for the most
important deities of the 'palace pantheon'.
K T U 1.111 describes an oracular decision for a three-day royal
offering for sin. 35 Here only the 'great' gods are listed. T h e only
goddess to appear is Nikkal, as consort of the moon god K u u h /
Yarih who here plays a major role.
K T U 1.125 is directed specially to El as a death ritual. O f the
'great' gods, only Teub, Kuuh, Ea and Attabi are mentioned
beside El, and there are no goddesses. Instead in K T U 1.116 there






1988b, 300-5.












are deified paraphernalia and expressions of well-being and in addition, deities from the realm of the dead ( D i e t r i c h - M a y e r 1997).
K T U 1.132 describes a three-day ritual for the palace goddess 36
Pidray. T h e sacrifices which had to be made during the feast are
principally directed to the palace goddess and her retinue, 37 in
which the daily list of offerings begins with the talli-deity (in tlnd:
lines 4, 18 and 22).
This brief description of the relevant texts with lists of offerings shows
that K T U 1.125 and K T U 1.132 evidently lie outside the frame
which the other texts profess. For the investigation of a H u r r o Ugaritic pantheon they are thus only of limited use.
T h e list provided in the following table shows the differences in
sequence. They seem to be independent of the rank of individual
gods. O n e gets the impression that the god-lists were compiled individually for particular occasions. T h u s a group of repeatedly recurring 'chief deities' can be drawn up which occurs to some extent in
almost all the textsas expected the exceptions are K T U 1.125 and
K T U 1.132. This group regularly begins with the talli-god and
includes the father-god, El, Teub, auka, Kumarbe, Kuuh, Ea,
Atabe, the city-god, H a m m u , Nubadig, Anat, imige, Piaaphe,
Hebat, Daqit, Hudena-Hude11ura, Ihara, Allai and Nikkal as well
as Ninatta-Kulitta and finally Adammait thus includes 23 deities
and to some extent represents the kernel of a 'palace pantheon'. T h e
lists in K T U 1.110 and K T U 1.111 are shorter, with 13 and 11
deities respectively. 38
These findings are in general also confirmed by K T U 1.42, which
is provided in the last column of the following table of sacrificial
lists for comparison.
K T U 1.116 extends this group of gods by a series of numina which
are hard to identify, to which are added deified paraphernalia and
expressions of well-being. In the other texts these are only quite inadequately documented, which may be partly due to the incomplete
condition of the tablets and partly due to an existing but not completely transparent eclecticism, as in the cases of K T U 1.125 and
K T U 1.132.

Against KTU'1 line 3 is to be read b bt mlk.

The only deity who is certainly male is Nubadig (line 10).
No account is taken here of the deities mlk, 'ttr and yrfa, mentioned in a Ugaritic
context and not in a closed list of sacrifices.







in tin
in atn
(/ kmrb)

















in ardn
in h,mn











































(table cont.)





















Of the deities listed, El, Ea, c Anat and Nikkal, for example, originally come from the Syro-Canaanite world of gods, as their names
indicate, whereas Teub, Kuuh, K u m a r b e , imige, auka and
Nubadig are evidently Hurrian. Which particular numina lie hidden
behind ta//z-deities, the father-gods and the city-gods, and where they
come from, to the extent that they do not denote groups of gods
here, must remain open for the time being.
What decisions can be made concerning the insertion of separate
sections in the Syro-Ugaritic and Hurrian panthea? At all events the
lists represent a mixed religious tradition. Therefore, one can hardly
speak of 'canon' of the pantheon, as given in the Akkado-Ugaritic


Historical Aspects

T h e Uhara-incantation K T U 1.131 shows that this text is written

in 'good' Hurrian an observation which agrees essentially with the
incantations of Kumarbi ( K T U 1.44), awuka from Niniveh ( K T U
1.54) and El ( K T U 1.128).39 A reason for this phenomenon may lie
in the long tradition of the texts which, as for example K T U 1.131,
may go back to the Yamhad-pcriod of the 18th 17th centuries. If the
question is again posed concerning the reason for such a long and
good tradition, then the explanation lies in the genre of incantation.






Here also the Hurrian wording has still been conscientiously transmitted even when Hurrian was no longer a commonly understood
language in Ugarit. Incantations, in fact, as in any language, have
to be difficult to pronounce and difficult to understand in order to
be effective. 40
From this observation can be derived a bridge to the mixed language particularly of the sacrificial texts. Starting from the existence
of a Hurrian pantheon and the rituals connected with it, then it can
be shown that it was no longer possible to carry out these sacrifices
and rituals at the time these texts were composed, since the Hurrian
instructions had become unintelligiblenothing is more important
for a ritual than its correct performance. T h u s these rubrics had to
be supplemented or replaced by rubrics in Ugaritic. An immediate
result was that originally non-Hurrian deities were accepted into the
rituals and lists.41 By their nature the incantations remained unaffected,
although they were directed to deities who, like El, belonged to the
Semitic pantheon.
It is also clear from this that at the time when the texts were
written down by the Prtre Hounite and his school, the Hurrian cult
was no longer alive but was only cultivated in certain fields. This is
certainly the result of a long process and can scarcely be traced back
to a specific historical date. T h u s it is in any case feasible that the
seizure of power towards the middle of the 2nd millennium by a
ruling house from the south 42 whose followers we care to name
'Ugaritians' as shorthand for their written tradition, had taken the
first steps towards the elimination of the Hurrian cult from Ugarit.
T h e n their kings, who had erected their palaces to the south of the
'acropolis' ( D i e t r i c h 1 9 9 7 ) , performed their cult also in the temples
on the 'acropolis'; they moved there with their Semitic deities. T h e
dominant role assumed by these deities is particularly clear from the
god-lists, where of the individual gods, El is at the top and others
such as Ea, Attabi, Anat and Nikkal are ranked below him only as
of minor importance and cannot be omitted from any list. Quite
clearly, at the same time the Semitic language of the new masters
took root in the Ugaritic cult, so that ultimately Hurrian was eliminated from the ritual texts. 43



On this cf. VAN D I J K 1982.

Cf. for example, D I A K O N O V 1981, 86-7.
D I E T R I C H - L O R E T Z 1988b, 3 1 0 - 1 .
The deathblow to the Hurrian cult in Ugarit must already have been pre-

In connection with these remarks on the history of the cult and

religion, the question can be posed concerning the Hurrian component of the population of Ugarit. It is hardly likely that the Hurrian
cult was cultivated in a city without its own priests and congregation. How large this congregation was is difficult to say, as we do
not know how widespread the cult was in town and 'acropolis'. O n
the one hand, it might have been restricted to Hurrians in certain
social classes, and on the other there might have been participants
who did not belong to the Hurrian population.
This provides no answer to the question concerning the level of
the Hurrian element of the population in Ugarit. 44 T h e onomasticon alone is no proof for the ethnic origin of the person bearing a
particular namethere are sufficient examples in the ancient Near
East of that. It is also quite likely that in the land around Ugarit
there were villages with an above-average Hurrian component in the
population. 45
T h e publication of the Babylonian texts of the priests' libraries of
E m a r has provided close connections, in terms of literature and the
history of religions, between Ugarit and E m a r at the end of the 13th
century. T h u s it is to be hoped that the Hurrian texts from Emar,
which are as yet unpublished, can throw some light on the questions which with regard to the Hurrians of Ugarit still remain open.
(Translation: W.G.E.


pared by the complete suppression of Hanigalbat by Adad-nrr I (1305 1275)

and Shalmaneser I (1275-1245), which entailed the end of the Hurrian kingdom
in Syria and thus the termination of its language, cf. VAN S O L D T 1991a, 521-2.
Cf. however H A A S 1978, 66.
Just as, for example, there were Kassite villages in the area of Arraphe, cf.






T h e


Kevin J.




C a t h c a r t


T h e decipherment of Ugaritic is associated with the names of three

persons in particular: Hans Bauer, Edouard D h o r m e and Charles
Virolleaud. T h e first lot of tablets, which had been discovered at
Ras Shamra in May 1929, was brought to Paris and given to Virolleaud for cleaning, transcription and decipherment. It was immediately clear to the French Assyriologist that the tablets comprised two
quite distinct kinds of documentssome in a well-known language,
Akkadian, but many others in an unknown language written in an
extremely simplified cuneiform (Virolleaud 1929, 305-6). T h e number of signs was twenty-six or twenty-seven, so the script was almost
certainly an alphabet. Virolleaud thought, quite righdy, that the decipherment of the newly discovered cuneiform writing would be made
easier because the words were generally separated from one another
by a vertical sign. He also noticed that the words were short, consisting of one, two, three or four letters, and he concluded that the
vowels were not written.


Charles Virolleaud

Virolleaud provided a starting-point for the decipherment of the Ras

Shamra alphabet by his acute observation that a short text engraved
on four of five bronze axes or adzes also occurred in the first line
of one of the clay tablets (1929, 306-7). T h e text on the tablet was
preceded by another single-letter word Y^/. Since contemporary
Akkadian letters began with the preposition ana, 'to', it seemed likely
that W / w a s a l s o a preposition. A fifth adze had the same inscrip-

tion but the words were preceded by another word of four letters:
^ 5J3> ^
c n > Virolleaud suggested that this word should mean
'axe', pointing to a tenth-century arrow-head from Sidon: KAI 20
hs cd\ 'the arrow-head of Addo'. T h u s the first word must denote
the object on which it is written and the second word has to be the
name of the owner. Virolleaud believed that in the absence of a
bilingual inscription, success in working out the meaning of the words
on the adzes would lead to the decipherment of other texts. In this
his first paper, which was the text of his communication to the
Acadmie des Inscriptions et Belles-Lettres in Paris on 20 September 1929,
he made no hard suggestions as to the identity of the language. He
excluded 'Mittannian', then suggested that perhaps a search should
be made in Asia Minor for the key to the new writing, and even
wondered whether Cypriot colonists had invented the cuneiform
alphabet. More to the point, although he did not realise it at the
time, he posed the question whether the Ras Shamra alphabet was
older than that of the Phoenicians or was an imitation or adaptation
of it ( V i r o l l e a u d 1929, 310). Some months later, however, Virolleaud
was expressing the view that the search for decipherment should be
directed towards Cyprus and the Aegean world: 'Malgr la difficult
actuelle d'un dchiffrement, il faut probablement orienter la recherche
vers Chypre et le monde gen' ( V i r o l l e a u d 1930, 353).


Ren Dussaud

It is worth mentioning that, at an early stage, Ren Dussaud ( D u s s a u d

1929, 298-9 n. 3) was firmly of the view that the alphabetic writing of Ras Shamra had been invented under the influence of the
Phoenician alphabet; but he did not elaborate on his views. In a
monograph published eight years later ( D u s s a u d 1937, 49 n. 2), he
recalled that Virolleaud in his communications of 20 Sept. 1929 and
14 Feb. 1930, and in his publication of the texts discovered in 1929,
did not yet regard the new cuneiform characters of Ras Shamra as
letters of the 'proto-Phoenician' alphabet. However, Virolleaud's
observations about the inscriptions on the five adzes were very useful indeed, as will be seen below, and his generosity in publishing
the newly discovered texts as soon as possible, so that scholars could
study them, was very laudable. T h e account of his contribution to
the decipherment of Ugaritic must properly be given below after the
contributions of Bauer and D h o r m e have been described.


Hans Bauer

Virolleaud's article containing transcriptions of the Ugaritic texts

reached H a n s Bauer on 22 April, 1930. In less than a week he had
already m a d e enough progress to inform Dussaud in Paris that he
had identified m a n y of the letters ( D u s s a u d 1930a, 130-1; 1930b,
200). According to a report in the Berlin newspaper Die Vossische
Leitung, he claimed to have identified twenty signs with certainty and
four others tentatively ( B a u e r 1930a). In fact Bauer had correctly
identified the signs for b, g, d, h, w, h, y, I, n, \ r and t. H e had also
identified two aleph-sigris but did not of course understand their precise value. Some of his incorrect readings were obstacles to progress:
for example k (for m), m (for s) and a second w (for k). However,
Bauer's basic a p p r o a c h and m e t h o d were excellent. T h e sixteen-page
introduction in his Entzifferung has been described as 'a masterpiece
of decipherment description' by Cyrus G o r d o n ( G o r d o n 1982, 109,
n. 6). Having assumed that the language was Semitic, he worked
with the knowledge that certain consonants were commonly found
as prefixes and suffixes and he quickly identified the sign for / through
Virolleaud's observation that the sign at the beginning of the first
line on the clay tablet was a preposition, the equivalent of Akk.
ana. Bauer received E d o u a r d D h o r m e ' s first p a p e r with his independent results ( D h o r m e 1930a) too late for correcting the errors in
his first m a j o r publication on the decipherment ( B a u e r 1930c), but
in a subsequent publication ( B a u e r 1930d) he gratefully acknowledged the signs for k, m, s, p, q and furnished by D h o r m e . Bauer's
e m e n d e d alphabet, known as the '5 O c t o b e r alphabet', was also
more accurate due to his identification of the word hrsn (earlier read
grzn), 'adze'.


Edouard Dhorme

At the Ecole Biblique in J e r u s a l e m , D h o r m e had also m a d e rapid

progress in the decipherment. H e and Bauer had worked (on opposite sides) as cryptanalysts during the First World W a r . It is clear,
however, that they had a good academic relationship and with their
excellent skills both of them accomplished m u c h in a relatively short
period. Like Bauer, D h o r m e identified the sign \f\f\f at the beginning of the first line of the clay tablet as 'Phoenician /', 'to'. T h e n he
worked out the frequendy occurring name bcl. Unfortunately, although

he had now identified the consonant b, problems arose when he read

bn where he should have read bt, for the letters and t are very
frequent, and he was thrown off course. He was able to rectify matters when he saw Bauer's announcement that he had discovered the
key to the Ras Shamra texts. He paid particular attention to Bauer's
suggestion ( B a u e r 1930a) that the word for 'adze' found on one of
the implements was grzn as in Hebrew (see the Siloam tunnel inscription and the Hebrew Bible). Even though Bauer's reading was erroneous, the signs for r and were read correctly. Since D h o r m e had
already identified the sign for k at the end of the word mlk, he could
now read Irb k-nm, which he quickly recognised as Irb khnm, ' T o the
chief of priests'. T h e title rb khnm was also on the adzes. So he had
now worked out the sign for h. In his alphabet, D h o r m e had correctly identified the signs for b, d, h, w, y, k, l, m, n, s, c, p, q, r, s, t
and two aleph-sigris, although like Bauer he did not have the full
truth about these last signs. Due to the incorrect reading grzn, he
gave the sign for h the value g, and s the value z- He also expressed
the opinion that for some consonants like aleph and 'ayin two different
signs could be used, and he understood this as evidence that the
final simplification of the writing system had not taken place ( D h o r m e
1930a, 575).
This first article by D h o r m e was dated 15 August 1930, but there
was a post scriptum, dated 14 September 1930, in which he commented on Bauer's first results. T h e signs which Dhorme read as m
and s were read by Bauer as k and m respectively. In his second
and third papers, dated 26 November, 1930 and 8 December 1930
respectively ( D h o r m e 1930b, 3 - 4 ; 1931, 32), Dhorme pointed out
that the divergences between his alphabet and Bauer's were greater
than he thought. When Bauer's Entzifferung appeared at the beginning of October it was clear that the two scholars differed not only
on k, m and but also on i. p, and q. Dhorme's first paper had
appeared too late to be considered by Bauer. As we have seen, Bauer
immediately corrected his alphabet in the light of Dhorme's readings, all of which he accepted and gratefully acknowledged.
Virolleaud's method of decipherment also involved the recognition of the preposition /, 'to'. H e looked for words containing / which
were common in West Semitic and soon identified mlk and its plural
mlkm. H e found the name of the god bcl and also the feminine form
b'lt. As a result of this he soon identified bt which could be 'daughter' or 'house'. T h e r e was a three-letter word in the texts with / in

middle position and the same sign before and after it. T h e only
word in the Semitic languages to fit this combination of signs was
the numeral 'three', which is l in Hebrew and tit in Arabic. T h e
early decipherers did not immediately establish the correct value of
Ugar. t. Virolleaud also recognised that the Ras Shamra alphabet
possessed three aleph-signs. H e seemed to have worked out the value
of more than twenty signs, but when the French newspapers reported
his decipherment of a 'mysterious alphabet', as if it were the first
successful attempt, Dhorme, clearly irked by the announcement, commented 'Nos lecteurs savent quoi s'en tenir sur la porte de ces
affirmations' ( D h o r m e 1931, 33).
An examination of J o h a n n e s Friedrich's publications on Ugaritic
( F r i e d r i c h 1933a; 1933b) shows that his grasp of the issues was very
sound in the early days of Ugaritic studies. T h e r e was still much
work to be done, but with an ever-increasing supply of texts, scholars would arrive at a complete decipherment.








Ugarit, Home of the Oldest Alphabets

In 1929 and 1930, C.F. Schaeffer discovered clay tablets from the
14th and 13th centuries bce on the tell known as Ras Shamra
ancient Ugaritin Northern Syria, close to the modern harbour
town of Latakia. These tablets did not belong within Mesopotamian
cuneiform tradition but were instead written from left to right, in
a script which comprised only 30 different signs. T h e decipherers
Bauer and Dhorme immediately surmised that in spite of the unusual
direction of the writing, they contained a hitherto unknown Semitic
alphabetic script (Fig. 1). Comparison with other Semitic languages
indigenous to Northwest Syria and Palestine soon confirmed this hypothesis. Thereafter, most of the signs could very quickly be identified

Fig. 1 A list of offerings with the first tablet number (KTU 1.39 =
RS 1 . 0 0 1 ; Photo: U G A R I T - F O R S C H U N G Archive)

Fig. 2

The alphabet tablet from Ugarit (14th/13th cent, bce)

(Photo: D i e t r i c h ) (KTU 5.6 = RS 12.068)

and the language they represented could be described as an idiom

which in terms of content seemed to be comparable to Canaanite
texts, but from a phonological perspective, however, was more like
Immediately after the decipherment, the question arose concerning the classification of the language of the texts from Ugarit, i.e.
Ugaritic, within the family of Northwest Semitic. T h e problem of
the dextrograde direction of writing could be explained as a relic of
the cuneiform tradition from which the Ugaritians already had the
idea of converting their texts onto clay tablets. This debate is still
open, as shown by discussions in articles, books and lectures during
specialist conferences.
Independently of the classification of the vocabulary with its own
peculiar phonology, within a closer or more distandy related dialect,
the discovery of school tablets in 1949 created a new dust storm as
they were written in letters in alphabetic sequence (Fig. 2). T h u s we
were dealing, completely unexpectedly, with the oldest documentation of the alphabet as a literary product: in the schools of Ugarit,
an alphabet was taught and learned, within a framework exactly as
known from the texts of the Hebrew Bible and from Middle Phoenician
several centuries later.
T o all intents the Ugaritic alphabet looks like, so to speak, the
Phoenician alphabet with 22 consonants, extended by the insertion
of 5 additional letters plus 3 at the end. Consequentiy, the Phoenician

alphabet formed the forerunner of the longer Ugaritic alphabet. T h e

events can be outlined as follows (Fig. 3).
T h e establishment of a longer alphabet comprising 30 letters could,
however, also be understood as the original and earlier one, from
which a shorter alphabet developed through the loss of 8 letters, i.e.
the later Phoenician alphabetadvocated by Albright and his students up to the present as the 'reduction theory'. However, that
would contradict what happens when an alphabet is adopted and
where necessary extended by adding further letters, according to the
principle 'an alphabet comes after a language'. This became clear
when a group of tablets was published, also from Ugarit and contemporary with the others, which displayed the following 'Phoenician'
characteristics (Fig. 4): a script going from right to left and a short
22-letter alphabet. This proved that in Ugarit two alphabets coexisted: a shorter alphabet with the characteristics of later Phoenician,
and a longer one, expressly for writing down Ugaritic. T h e result of this
was that our monograph on the alphabet was called 'The Cuneiform
Alphabets' (Die Keilalphabete, D i e t r i c h - L o r e t z 1988a).1
With regard to the nature of the letters in the texts written in
alphabetic cuneiform, the following outline can be given: T h e letters, which were written from left to right and determine the script,


m .







^ - ^ r r r - r

Fig. 3



FT K * * - *


^ c



The long cuneiform alphabet from Ugarit and its relationship to

Phoenician and Canaanite ( D i e t r i c h - L o r e t z 1988a, 128)

See also




KTU 1.77
KTU 4.31







KTU 4.710

KTU 1.77
KTU 4.31
KTU 4.710


m ?
m ?



The short cuneiform alphabet from Ugarit

1988a, 270)



represent vowelless consonants, opening or closing syllables, with an

implicit vowel of unspecified nature and lengthexcept, of course,
for the aleph-letters. More precise details on the nature of a consonant and the vowel most probably assigned to it can be m a d e
by means of the transmission of Ugaritic words in syllabic spelling
1987b) or by comparison with cognate languages,
where the vowels are known.
Now, however, we are faced with the task of allocating a homeland and if possible also the nearest cognate to both alphabets: in
view of the clear line of development of the expansion of a previous alphabet, theoretically speaking it was no longer necessary to
look for the closest cognate to the shorter alphabet tradition, as
attested directly and indirectly in Ugarit. T h e y were there and had
to be sought within the framework of the old and possibly protoCanaanite alphabets, the history of which was already established.
T h e cognates of the longer alphabet tradition attested in Ugarit were,
however, much more difficult to determine. T o be able to tackle the
related questions it was necessary to discover what the cuneiform
signs made in clay might possibly have looked like as strokes, i.e. in
linear form (Fig. 5). It was clear that they could have been transposed from linear shapes to cuneiform signs and no longer needed

to be discussed. Thus it was necessary to discover which principles

the scribe and artist might have followed. We have considered these
principles and have attempted to connect the reconstructed linear
signs with similar signs in the other alphabets of Semitic languages.
In this way we showed that the linear form of the Ugaridc alphabet
exhibits the direction both of the Phoenician alphabet and of the
old South Arabian alphabetic tradition. Accordingly, we started to
look for epigraphic and historical supporting evidence.
As a result of our endeavours we received quite unexpected proof
from St Petersburg: L u n d i n (1987) suspected that the small clay tablet
dug up by E. Grant in 1933 in Beth Shemesh, in Southern Palestine
dating to the 14th/13th century bce (Fig. 6) contained the beginning of the alphabet in the South Arabian tradition, of which the
earliest witness so far dates to the mid first millennium bce, i.e. about
700 to 800 years later. After closer inspection of the photographs
from the excavation, and of other photographs, we came to the conclusion that on this small clay tablet there was not only the beginning of the South Arabian alphabet (h-l-h-m) but the complete alphabet,
written from left to right round the tablet. T w o things were clear,
therefore. First, that the old South Arabian alphabet with the corresponding sequence of letters was already in existence at this time.
Second, it was also clear that following the general trend in the midsecond millennium bce, of transposing both script and language to
clay in line with the Babylonian model, it was in cuneiform script.
T h e r e was no need to emphasize that the signs used were not very
different from those used in Ugarit: the forms underlying both attempts
at transposition lay in their linear shape, as Fig. 5 shows, and they
are very similar ( D i e t r i c h - L o r e t z 1988a, 277-96).
In respect of the geographical distance of Beth Shemesh in Southern
Palestine from Ugarit, doubts about a connection between the alphabet traditions are justified; the rejection of such a connection would have
been premature, however. For the Syrian-French excavations in 1988
have brought to light in south east Ugarit the archive of a trader
and dignitary called U r t e n u , in which a 'palimpsest'-tablet was
inscribed with the South Arabian alphabet of the /W-A-m-tradition
(Fig. 7; B o r d r e u i l - P a r d e e 1995a). With this tablet the proof has
finally come that the connection of the Ugaritic Long Alphabet to
the South Arabian alphabet, which also has a tradition going back
to the 14th-13th century bce, was completely justified.
A comparison of the sequence of letters in the old Arabic alphabet


Western alphabets
proto- Ane. Phoen.
Can. Can. Ah. Byblos






cuneiform linear
I 1

1 <- - r


i p
t1 "


F i t


fct-J ^






Il **





; J



X +



+ y





-o -

_ H-X 4

t f * -M



7 V ?

e m r H*

W f

-> 1


A 3

n ?





d n

- ri




- > O

\ X

. . . Ill- -w
V - n>rt



-a cr) r r


OSA Iih.

4 X




/ <






^J " F


Southern alphabets







- -t

? I C


Fig. 5

The Ugaritic script in relation to the Western and Southern scripts

( D i e t r i c h - L o r e t z 1988a, 102)

Fig. 6

Fig. 7

The Alphabet tablet from Beth Shemesh

( D I E T R I C H - L O R E T Z 1988a, 285)

Tablet with the South Arabic Alphabet





Beth eme
{J. Ryckmans

H L H M Q. w
H L H M Q. w
H L H M Q. w

Beth eme



' '

{J. Ryckmans


Fig. 8




N U [] F






Deciphered Alphabets of the South-Eastern A-/-^-m-Tradition

with the traditions of Beth Shemesh and Ugarit gives the following
results (Fig. 8)in the third line the South Semitic alphabet is given
in brackets for convenience, as established by R y c k m a n s ( 1 9 8 5 ) from
texts dating to the first millennium bce.
This comparison shows clearly that the sequence of letters of the
cuneiform alphabets attested in Beth Shemesh and Ugarit as well as
the South Arabic-South Semitic alphabet reconstructed by Ryckmans
show very few divergences. T h e fact that the signs, as shown by
hand-copies and photographs, vary slightly as well, leads to the conclusion that the traditions, certainly not least in view of their geographical separation, must to some extent have developed independently.
These divergences are basically so slight that their common origin
never completely vanished from view.
This report on the discovery of the two cuneiform alphabets as transmitted in Ugarit as well as their summary comparison should show that
the commercial centre Ugarit in the third quarter of the second millennium was a turning point in the early history of the alphabet in
two ways:
1. With the earliest recorded alphabet so f a r w h e t h e r in the '-b-g
sequence of the north west tradition or the h-l-h-m sequence of
the south eastern traditionwe are quite unexpectedlyor rather,
as was to be expectedat a point in its development which already
has the appearance of the conclusion to a forerunner. Just when
we thought we were close to the origin of the evolution of the
alphabet and finally were able to propose a date for the 'discovery
of the alphabet', we have to envisage a world with two alphabets (Fig. 9).

2. T h e origin of the cuneiform alphabet from Ugarit also reflects a

history of two alphabets: a Levantine tradition with 22 letters was
shall we say overlaid by an Arabic tradition so that, with some
additions, a 30-letter alphabet was derived. T h u s we find ourselves in the midst of a historical process on which it is worthwhile to reflect a little: the inhabitants of a city who towards the
middle of the second millennium had migrated from south eastern Palestine to the northern Levant, had developed a commercial ruling dynasty in Ugarit which in the history of literature
and the art of the scribe had thus created a lasting monument,
so that on arrival its native tongue could be inserted into an existing scribal school and writing tradition and thus the alphabet
already discovered could be expanded. T h e authentic Ugaritic
alphabet which arose in this way is ultimately to be perceived as
a witness of inter-cultural activity with branches as far afield as
Cyprus (Hala Sultan Teke).
With the examination of the cuneiform alphabets transmitted in
Ugarit we find that in terms of the history of the alphabet we are
not, as many would argue, at its inception. Instead, we are once
again at the conclusion of separate developments which nevertheless
in terms of their significance represent an important step in universal history. Although we find ourselves in the mid-second millennium
bce, we have still not yet reached a point at which we can speculate on the origin of the alphabet.



.ArstaiTai Hamm
W Bars
f is
Ris amrj/Uqarit

Hata Sultan Tekke


Neti Mwd/CLadeS.


"TK3md ei-Lz/Kumidi


Mountains of


Heni J


Ugaritic cuneiform

Proto-Canaanlte Inscriptions

Ancient Canaanite Inscriptions

Ancient Phoenician Inscriptions

(up to the end of the 8th cent. BCE)

Ancient Aramaic Inscriptions

(up to the end of the 8th cent. BCE)

Ancient Hebrew Inscriptions

(up to the end of the 8th cent. BCE)





drawing: G. Neuber

Fig. 9

The spread of cuneiform alphabets in the Eastern Mediterranean

( D I E T R I C H - L O R E T Z 1988a, 359: Map)







The classification of Ugaritic

Ugaritic, the local language of the city state of Ugaritic, is one of the
Semitic languages. T h e classification of Ugaritic within Semitic is still
a matter of dispute.
T h e Semitic languages can be broadly divided into an East Semitic
branch (Akkadian) and a West Semitic branch. West Semitic can be
further divided into a South Semitic and a central Semitic branch
(Canaanite, Aramaic and North Arabic). Of the central Semitic languages, Canaanite and Aramaic can be traced back to a common
former stage which is called 'Northwest Semitic'. T h e r e is hardly any
doubt nowadays that Ugaritic is more closely related to Canaanite
and Aramaic than to North Arabic. 1 T h u s Ugaritic is shown to be
a Northwest Semitic language. Since Ugaritic is closer to the later
Canaanite languages than to Aramaic due to more important linguistic isoglosses, either it belongs to Canaanite or it is the (only) representative of a separate language branch of Northwest Semitic closely
related to Canaanite.3.1.2

Current research in Ugaritic grammar T h e four most important overall summaries of Ugaritic

g r a m m a r published so far are:
1. C.H. G o r d o n , Ugaritic Grammar (Rome 1940) with two further editions: Ugaritic Manual (Rome 1955) und Ugaritic Textbook (Rome 1965).
2 . S . S e g e r t , A Basic Grammar of the Ugaritic Language (Berkeley, Los
Angeles, London 1984).

' The opinion maintained in the early period of research that Ugaritic is particularly close to North Arabic can no longer be held today. The two most important arguments for this theory, the extensive phoneme inventory of Ugaritic and
Ugaritic-Arabic isoglosses have been substantially modified by the results of recent
For most recent discussion see especially ISAKSSON 1989 and T R O P P E R 1994b.

D. S i v a n , Ugaritic Grammar (Encyclopaedia Miqra'it 9 ; Jerusalem

1993) [in Hebrew],
4. D. S i v a n , A Grammar of the Ugaritic Language ( H d O 1/28 Leiden,
New York, Cologne 1997).

By now, Gordon's summaries have been largely superseded. T h e

later grammars listed merely present a concise grammatical outline
of Ugaritic. As yet there is no complete, modern reference grammar
of Ugaritic. In his Habilitation submission presented in 1997 at the
Freie Universitt Berlin with the dtie Untersuchungen zur ugaritischen Grammatik. Schrift- Laut- und Formenlehre3 the author has laid the foundation
stone for such a grammar. In spite of research spanning several decades, a large number of problems of Ugaritic g r a m m a r remain unsolved, due to the
following factors:
(a) Ugaritic is a dead language attested only in writing.
(b) Closely cognate contemporary languages are known only sketchily.
(c) T h e corpus of Ugaritic texts is recorded in a writing system without vowels, except for a few words in syllabic spellings.
(d) T h e range of grammatically important Ugaritic texts is restricted.
(e) M a n y texts are of uncertain interpretation.
(f ) T h e corpus of Ugaritic texts contains a set of very different genres, each with its own grammatical features.

Aim of this grammatical outline

Here follows a brief grammatical outline of spelling, phonology, morphology and morpho-syntax of the Ugaritic language. It contains
only the central topics of grammar. Facts on which scholars are
widely in agreement are described without reference to secondary
literature. However, areas where there are problems are provided
with a short commentary. In this way the outline of grammar will
describe the present state of research on Ugaritic grammar.

Research on Ugaritic Grammar. Script, Phonology and Morphology.



The principles of consonantal orthography

T h e basic principle of the Ugaritic alphabetic script is that each consonantal phoneme of the language is equivalent to one sign in the
script. By the introduction of the three additional signs of the Ugaritic
alphabet, < i > , < u > and < s > , the principle is violated in two ways:
the phoneme / V is represented in Ugariticdepending on the following vowelby < a > , < i > or < u > (cf. 3.2); the phoneme / s / is
represented by <s> or < s > . 4
Lengthened (doubled) consonants were not differentiated from single consonants. They can be determined only by comparative philology or on the basis of syllabic spellings.

The aleph signs

a. T h e most remarkable feature of the Ugaritic alphabet is that

it has three different signs for writing the glottal stop / V , i.e. an
a-aleph = < a > , an z'-aleph = < i > and a w-aleph = < u > .
b. It is generally agreed that the three Ugaritic aleph signs are used
in syllable-initial position: < a > represents the syllables a/ and /'/,
< i > represents / V , / , and probably also / W (shcwa) and
lasdy < u > represents I'ul, / and /'0/.
c. T h e way in which syllable-final, i.e. vowelless aleph is written is
a matter of dispute. T h e r e are quite different theories. T h e extreme
views are as follows: 1.) Any syllable-closing aleph is represented
by <i>. 2.) T h e choice of sign for syllable-closing aleph depends on
the quality of the preceding vowel. 3.) Syllable-closing aleph is no
longer expressed in Ugaritic: / V becomes / / , / f / becomes I I I and
/ u } / becomes / / . T h e aleph signs < a > , < i > and < u > act as vowel
letters (matres lectionis) for the resulting contracted vowel: / / , ll and

For the phonetic value of this grapheme, which gained entry into the Ugaritic
alphabet only at a later date, see S E G E R T 1 9 8 3 (<s> = [su]) and recently T R O P P E R
1995. Thus the grapheme <s> has the value ['s]. It was inserted after the Ugaritic
phoneme / s / written with <s>, which originally (also) had the phonetic value ['s],
had been deaffricated to [s] in certain phonetic contexts.
For the proponents of these theories see V E R R E E T 1983a, 2 2 3 - 6 .

d. T h e theory of the position of aleph defended here is more complex than the proposals mentioned above. 6 O n the one hand it envisages the possibility that the syllable-closing glottal stop in Ugaritic
was not always strongly articulated (= quiescent aleph). O n the other
hand it follows that in Ugaritic / a V after the loss of a syllable-closing
glottal stopmost probably irrespective of stressbecame either / /
or / 0 / (cf. Heb. ns't < *naa>t as distinct from Heb. ro < *ra> ).
O n this basis the following 'rules' can be formulated: the (articulated) syllable-closing glottal stop is written with an < i > , irrespective
of the preceding vowel. If the glottal stop is not articulated, i.e. the
aleph is quiescent, then the following applies: < a > stands for / / <
*a}; < i > stands for III < *i'; < u > stands either for // < *u} or
for loi < V .
e. T h e graphic notation of a syllable-closing glottal stop is non-homogeneous in the corpus of Ugaritic texts, as some alephs evidendy
represent a strong aleph, others a quiescent aleph. T h e former are
phonemic spellings, the latter phonetic spellings: e.g. yihd lya'fyud-l
'he takes/took' ( K T U 1.6 = RS 2. [009]+5.155 1, etc.) againstyahd
lyhud-I <*ya}(}ud- ( K T U 4 . 4 4 = RS 9.453:28) or yuhd(m) lyffVd-l
<*ya'hud- ( K T U 1.4 = RS 2.[008]+3.341+3.347 iv 16; K T U 1.22
= R S 2.[024] ii 17*; K T U 1.103+ = R S 24.247+: 17) and tuhd/tfj. Vdl<*ta'hud- ( K T U 1.2 = RS 3.367 i 40).7 Within a word there
are more spellings with strong aleph, but at the end of words there
are more with quiescent aleph.

Vowel notation

In principle, in alphabetic spelling vowels are ignored, apart from

the aleph signs which in syllable-initial position express the vowel
following an / V (as well) ( 3.2.2b). Exceptions are rare. T h e sporadic use of < y > as a vowel letter (mater lectionis) for / / is worth
emphasizing, especially at word-close, e.g. ily ugrt /}il 3UgaritV/ 'the
gods of Ugarit' ( K T U 2.26 = R S 16.264:4-5). 8

On this see T R O P P E R 1990b.

Cf. Heb. yhz (18x; alongside 3 X ye"hz).
On vowel letters in Ugaritic see BI.AU - L O E W E N S T A M M 1970 and
L O R E T Z 1973c.




The consonantal phoneme system

According to the Ugaritic alphabet, which comprises twenty-seven

'ordinary signs' and three 'supplementary signs', there were at least
27 consonantal phonemes in Ugaritic. This relatively wide range of
consonants can be subdivided into obstruents (plosives and fricatives), nasals ( / m / , / / ) , liquids ( / / / , /r/) and semi-vowels (/w/,
Depending on how they were articulated (unvoiced, emphatic, voiced)
most of the obstruents can be arranged in rows of three columns,
as follows:
(alveolar) sibilants
(palato-alveolar) sibilants


/ y





Comments on uncertain consonantal phonemes

a. T h e Semitic phoneme / d ! is written consistently with < d > only

in texts K T U 1.12 = RS 2.[018] and K T U 1.24 = R S 5.194. In
the other texts it is written predominantly with the sign < d > a p a r t
from specific phonetic environments 1 0 which indicates a conditioned
coalescence of / d / and / d / in Ugaritic.
b. Etymological / / usually occurs in Ugaritic spelling as <z>. However, there are a few words in which < g > stands for etymological
/ / . " Incontrovertible examples are Ingr, 'to guard', Vgm\ 'to be
thirsty' and gr, 'mountain'.

On this topic see FRONZAROLI 1955b.

Often in the vicinity of Irl (cf. UT 5.3), /m/ and /n/.
See FRONZAROLI 1955b, 33-5, / 5.7 and D I E T R I C H - L O R E T Z , 1967, 312-4.
RSSLER 1961 has a different view.

c. Ugaritic has four sibilants, i.e. / s / , / s / , / z/ and / / . T h e first

three phonemes stand out by their distinctive feature of affrication
and should have been realised as ['s - 's5 - d z]. In contrast, // is
a fricative palato-alveolar sibilant, i.e. [].
d. T h e lateral phonemes /s2/ (= / / ) and / / / attested in other
Semitic languages are regularly represented in Ugaritic spelling by
the graphemes <> and <s>. Rarelyfor instance in K T U 1.12 =
RS 2.[012]etymological / / / appears in Ugaritic as < ? > .


Equivalence table of selected Semitic consonants12
















s 3 /s




? g*
d, d

s, z*




The vocalic phoneme system

Proto-Semitic has a) three short vowels, / / , //, / / , b) three long

vowels, / / , / / , / / , and two dipththongs, / a y / and /aw/. Ugaritic
has preserved the primary vowels / a / , /i/ and /u/ as both long
and short. T h e PS diphthongs *ay and *aw, however, have been
contracted to monophthongs: *ay > //\ *aw > // (see

Sound changes

A large number of sound changes are documented in Ugaritic spelling

following phonological rules. T h e most fundamental are set out here.


OSA Old South Arabic; Arab. = Classical Arabic; Aram. = Aramaic; Eth. =
Ethiopie; Akk. = Akkadian; PS = Proto-Semitic.

Consonantal sound changes

(a) Sound change *w > lyl at word-initial, e.g. \ybl < *wbl, 'to
carry, bring'. 13
(b) Voicingdevoicing, 14 especially with labials, e.g. ybcl /yib'alu/ 'he
makes' ( K T U 1.17 = RS 2. [004] vi 24; <p'l) and tpky /tapkiyu/
'you (sg.) weep' ( K T U 1.107 = RS 24.251+: 11; <bky).
(c) Regressive assimilation: *dt > / t t / , e.g. aht /'ah(h)att-/ < *'ahadt'one' ( K T U 1.48 = RS 1.019:13 etc.); *nC > /CC, e.g. ap = syllabic spelling ap-pu /'appu/ < * 'anpu 'nose' ( K T U 1.2 = R S
3.367 i 13 etc.); *IC > /CC/, only in qh 'to take' (see

Vocalic sound changes 1 ' 1

(a) Vowel harmony: *qattv\l > /qv\ttv\l/, e.g. ibr /'ibbr-/ < *'abbr
'bull' ( K T U 1.10 = R S 3 . 3 6 2 + iii 35 etc.); *>ViC.v2C(C) >
^V2C.V2C(C) ( V = short vowel), e.g. urbt /'urubbat-/ < *'arubbat'opening, hatch' ( K T U 1.4 = R S 2. [008]+ 61, etc.), irt irit-/
< *>arit- 'wish' ( K T U 1.104 = RS 24.248:1, etc.).
(b) Vowel syncope: pretonic: (Cv)CvCvC.v > (Cv)CvCC.v (V = short
vowel), e.g. rit /ra'sat-/ < *ra'at- 'heads' ( K T U 1.2 = RS 3.367
i 23.24, etc.); post-tonic: C.vCvCvCv > C.vCCvCv or C.vCCvCvCv >
CvCCvCCv (V = short vowel).

Sound changes in diphthongs and triphthongs 16

(a) Contraction of diphthongs: *aw > /<?/; *ay > /<?/; *iy > //\



(b) Preservation of some types of triphthong ( / u w a / , /iyv/,

as opposed to the contraction of other types of triphthong (*awu> //\ *aw> //*ayu > //; *ay > / f / ; *awa> //\
*aya > //; *uwu > //).


However, word-initial /w/ is preserved (a) in the conjunction w, 'and', (b)

before the vowel lui (D-stem infinitives of the I-y < I-w roots, e.g. wld lwullad-1
'to bear (a child)' [KTU 1.14 = RS 2.[003]+ iii "48 etc.] and wpt-m lwuppat-1 'to
insult' [KTU 1.4 = RS 2.[008]+ vi 13]).
On this topic see G A R R 1 9 8 6 and V O I G T 1 9 9 1 .
For syllabic spellings see H U E H N E R G A R D 1987b, 268-83.
For syllabic spellings see H U E H N E R G A R D 1987b, 288-92.

Sound changes within syllables

(a) Prothesis (to avoid a word-initial consonant cluster): e.g. usbc/(')usbac-/

< *sibac- 'finger (pi.)' ( K T U 1.2 = R S 3.367 iv 14, etc.).
(b) Specific pausal forms: words at the end of a sentence occasionally have a phonetically altered form (e.g. reduction of the ending or special lengthening of the stressed syllable).


Morphology and morphosyntax

The pronoun
T h e personal pronoun

a. Nominative forms: ank syll. a-na-ku anku/ (longer form)

or an /3an/ (shorter form); at = syll. at-ta /}atta/ < *}anta; at /3atR/ < *'anti; hw = syll. -wa /huwa/ < *hu}a; hy /hiya/ < ^hi'a; atm /'attum/ < *'antum; 2.c.du. atm
attum/ < *'antum; 3.c.du. hm /hum/.17
b. Oblique forms (gen./acc.): hwt /huwati/ ( K T U 1.3 = R S
2. [014] vi 20 etc.); hyt /hiyati/ ( K T U 1.3 = R S 2. [014]
iii 10 etc.); 3.m.p1. hmt /humti/ ( K T U 1.19 = R S 3.322+ iii 9
etc.); 3.c.du. hmt /humti/ ( K T U 1.17 = R S 2. [004] 20.30; K T U
1.19 = RS 3.322+ iii 44).

Pronominal suffixes possessive suffix -0 or -y / - / , -y /-ya/;R object suffix

- /-n/; -k /-ka/; -k /-k/; -h = syll. - /-h/; -h /-ha/; -n /-na/(?); -km /-kurnu/; 2.f.p1. -kn
/-kun(r)a/; -hm /-hum/;
3.f.p1. -hn /-hun[n)a/; I.e.du. -ny
/-nay/; 2.c.du. -km /-kum/; 3.c.du. -hm /-hum/.

Demonstrative pronouns

- 'this' (adjectival): hnd (many); hndn ( K T U 2.71 = RS 29.095:10).

- 'this' (nominal): hndt ( K T U 1.19 = R S 3.322+ iv 62; K T U 2.38 =
RS 18.031:12; K T U 2.45 = R S 18.140:7).

17,, 3 m./ and I.e.du. forms are not attested.

After a short H-/ vowel and after various long vowels it is / -ya/, otherwise it
is / I / .

- 'that' (adjectival): hnhmt ( K T U 3.3 = RS 15.128:8; perhaps also

K T U 4.659 = RS 19.166:6).
- 'that' (nominal): hnk ( K T U 2.33 = R S 16.402:23); hnkt ( K T U 2.41
= RS 18.147:13; K T U 2.21 = RS 15.174:10).

T h e determinative pronoun (relative pronoun)

T h e forms of the Ugaritic determinative pronoun, which also functions as a relative pronoun, are: d /d/ (Nom.), / d / (gen.),
/d/ (acc.) and d = /d/ (only K T U 1.24 = R S 5.194:45 [gen.]); dt = /dtu/, /dti/, /data/; c(?).p1. dt /dtV/. T h e r e is also an
indeclinable variant d = /da/(?).19

Interrogative pronouns

- 'who?': my /mya/(?) (several occurrences); mn ( K T U 1.3 = RS

2.[014] iii 37; K T U 1.3 iv 4; perhaps K T U 1.5 = RS 2. [022] +
iv 23.
- 'what?': mh /mah(a)/ (several occurrences); mhy ( K T U 2.14 = RS
[Varia 4]:9 only); mat ( K T U 1.14 = R S 2. [003]+ i 38 only); mn
(uncertain occurrences: K T U 1.5 = RS 2. [022] + iv 23; K T U 2.45 =
R S 18.140:25; K T U 2.72 = RS 34.124:22; in K T U 1.16 = R S
3.325+ ii 19.20 it means 'how many?').

Indefinite pronouns

- 'anyone': mnk ( K T U 3.2 = R S 15.111:12 [mnk mnkm]); mnkrn ( K T U

2.19 = RS 15.125:12; K T U 3.2 = RS 15.111:13); mnmn ( K T U
1.123 = R S 24.271:22 [mr mnmn]);20 mnn ( K T U 5.9 = RS 16.265
i 2).
- 'anything': mhk ( K T U 2.38 = RS 18.031:26); mhkm ( K T U 2.30 =
RS 16.379:22; K T U 2.71 = RS 29.095:14 [<h>mhkm]); mnm (many
- 'whatever' (adjectival): ay ( K T U 1.23 = RS 2.002:6; K T U 1.24 =
R S 5.194:44).

It occurs only as a determinative pronoun before a noun clause or as a relative pronoun before a nominal relative clause.
Cf. Akk. mammon < *man-man 'anybody' as well as the expression mar marnmana(ma), 'anybody's son' (CAD M / l , 200-1).


The noun
Noun formation

Several different patterns are used for the formation of nouns (nouns
and adjectives) in Semitic: monosyllabic forms; polysyllabic forms;
forms with lengthened components; forms with prefixes, infixes and
suffixes. Most patterns can be assigned to specific classes of meaning. T h e r e are severe limitations on identifying nominal patterns in
Ugaritic as the alphabetic spelling often permits no conclusions regarding formation. 21 T h e Ugaritic noun forms attested in syllabic spelling
are rich in information. 2 2


T h e r e are two grammatical genders: masculine (masc.) and feminine

(fem.). Masc. nouns are basically unmarked, whereas as a rule fem.
nouns have a special ending.
By far the commonest feminine morpheme is -(a)t. T h e choice of
the m o r p h e m e variant -at instead of -t is largely dependent on syllable structure. T h e -(amending also denotes nomina unitatis (singular
nouns), e.g. mnht '(single) gift' ( K T U 4.709 = R S [Varia 13]:6) in
relation with the generic name mnh 'gift(s)' ( K T U 1.2 = RS 3.367
i 38 etc.). Besides the feminine ending -t = /-{a)/ probably a rare
feminine ending y /-ayV/ is also attested: ncmy 'the (exceedingly)
lovely' ( K T U 1.5 = RS 2. [022]+ vi 6.28; K T U 1.17 = RS 2. [004]
ii 42). T h e r e are also grammatically feminine nouns without a feminine ending, e.g. um 'mother'.


T h e r e are three numbers in Ugaritic: singular (sg.), dual (du.) and

plural (pl.). T h e sg. is unmarked. T h e du. and pi. are denoted by
special morphemes.
In Ugaritic the du. is productive and is marked by the m o r p h e m e
- 0 = / / which always coalesces with the case ending: nominative
-a, oblique (gen./acc.) -e. In the absolute state the ending is lengthened by mimation: nom. -m /-mi/, obi. -m /-ma/ (alternatively: /-mi/). As a rule, the dual ending is added onto the singular



On this topic see

1 9 8 4 4 3 ; SIVAN




1987b, 302 17.

form. It comes after the gender m o r p h e m e -t /-(a)t/ of (marked)

fem. nouns.
T h e pi. is marked by a m o r p h e m e which causes vowel lengthening. With fem. nouns the plural marker comes before the gender
marker and the case endings: nominative /-tu/, oblique /-ti/; in
the masc. noun it merges with the case endings: nominative / - / ,
oblique /-/ (absolute state: /-ma/, /-ma/). T h e nominal base of
the pi. is mosdy the same as the sg. In certain nominal patterns and
certain weak root classes, the plural basis differs from the base of
the singular. It should be emphasized that <qVtl>-forms generally
have a bisyllabic base <qVtal> in the plural (e.g. pi. ram/1 /ra'as-/
of sg. ris / ra's-/).


a. Ugaritic has three main cases: nominative (nom.), genitive (gen.)

and accusative (acc.). They are marked by vocalic morphemes which
follow the gender marker in fem. nouns.
In the (masc. a n d fem.) sg. the three main cases are mostly
differentiated by three different vowel endings: nom. -u, gen. -i, acc.
-a (triptotic endings). Nouns of certain patterns (including certain personal names) have only two different case-endings in the sg.: nom.
-u, gen./acc. = oblique (obi.) -a (diptotic). In the du. and masc. pi.
the inflection is exclusively diptotic: d u . n o m . -,;
nom. -; obi. -f.23
b. Besides the three main cases, Ugaritic has two further cases with
primary adverbial function, i.e. the terminative and the locative. Both
are comparatively little used.
T h e terminative functions as an independent adverbial case, primarily for denoting direction. It is marked by the --ending which,
in connection with the so-called 'he locale (locative h) of Hebrew
g r a m m a r is probably to be vocalised as /-ah/. T h e terminative ending is probably added on to the uninflected noun stem: arsh =
/',arsah/ 'towards the earth' ( K T U 1.14 = RS 2. [003]+ i 29). It only
occurs for certain in the abs. state.
T h e locative is marked by the ending / - / , e.g. sbu p /sabVu/

Occasionally the oblique ending seems (already) to have assumed the function
of the nominative ending, e.g. ily ugrt (= /ill UgaritV/) tgrk tlmk 'may the gods of
Ugaritic guard you (and) grant you well-being' (KTU 2.16 = RS 15.008:4-6).

'at sunset' ( K T U 1.41 = RS 1.003+:47.53). Examples are difficult

to identify as the locative ending is only evident from spellings in
forms of III-' roots. T h e r e seem to be several different functions of
the locative. It denotes place (locative), time, the ablative, the instrumental, measure and quantity, final nuances (with infinitives or verbal nouns) and the paronomastic infinitive (e.g. bt krt bu tbu, 'she did
enter Kit's house' [ K T U 1.16 = RS 3.325+ vi 3]).
c. T h e form of the noun in the imperative and in direct address
the vocadveis expressed by various syntagmata: (a) by an unintroduced noun, (b) by the noun introduced by the particle /, (c) by
a noun introduced by the particle y or (d) by a noun with a 1st
pers. noun suffix. There is no explicit information on the case-endings
of the vocative in Ugaritic in spite of a few occurrences of III-' radical nouns. It is uncertain whether a noun in all the constructions
just mentioned (a-d) has the same morphological form. It is also
uncertain whether the vocative has the same form as one of the
three main cases. T h e r e are indications that in the singular the vocative can be expressed without any case-endings and that the accusative
case serves as a vocative.
d. In the corpus of Ugaritic texts there are occasionally nouns
without any inflection. 24 This could be a relic of what is known as
the 'absolute case', 25 comparable with the 'absolute state' of Akkadian g r a m m a r (cf. GAG 62 c-j).


T h e noun has two states which depend on the syntactic position of

a noun: 1. the absolute state (abs. st.), 2. the construct state (cstr. st.).
T h e abs. st. is unmarked in the singular and so is the same as the
cstr. st. In the dual and plural it is sometimes marked by a final
-m, known as nominal mimation. T h e cstr. st. is unmarked for all
numbers and differs from the abs. st. in the dual and plural by the
lack of mimation. Both states in Ugaritic have (the same) case-endings.


f o r syllabic spellings see H U E H N E R G A R D 1987b, 300-1.

On the absolute case in Hamito-Semitic see SASSE 1984.


Ugaritic has no morphological marker for determination or indtermination. T h e r e is neither a definite article nor a specific determined
case, and mimation on nouns has neither a determinative nor an
indeterminative function. 26

Cardinal numbers

a. T h e cardinal numbers 1 - 1 0 are as follows:


ahd /'ah(h)ad-/;


in /tin/ (nom.), /tin/ (obi.); tt /titt/ (nom.), /titt/ (obi.)

tit / talf, tilt /taltat-/
arb'/'arba'-/; arb't /'arba'at-/
hm /hami-/\ hmt
It / M - / ; ttt
b' /sab'-/; b' t /sab'at-/
tmn / tamn/ < *tamniy-; tmnt /tamnt-/
ts" /tie-/\ t' t /lis'at-/
'r /'a(a)r-/ < *Caar-; 'rt /Cas\a)rat-/

T h e uninflected (masc.) forms of the cardinal numerals 3 - 1 0 can be

coupled with nouns of either gender. In the Baal Cycle, the Aqhat
Epic and a few other poetic texts, however, fem. numerals are generally used with masculine countables (syntax with 'polarity of gender'). In prose, fem. numerals are used exclusively with the ellipsis
of tql 'shekel' and ym 'day'.
b. T h e numerals from 11-19 are made up of the units 1 - 9 and the
expression for 'ten' (Cr / Crt / esrh). T h e sequence is mosdy 'unit
ten', e.g. (a) hm Cr, (b) hm Crh and (c) hmt Crt '15'. Type (a) is only
used together with masc. nouns. Type (b) occurs with fem. and masc.
nouns. Type (c) is used only with the ellipsis of tql 'shekel' or ym
'day'here as an ordinal numeral. Numerals 12-19 can also be constructed in the reverse sequence ('tenunit'). In these cases the unit
is always followed by the word kbd which can be rendered 'plus',
e.g. csr arbc kbd '14'.
c. T h e numeral 20 is formed from the dual or plural form of Cr
' 10', the tens from 30 to 90 from the plural forms of the numerals
3 to 9: Crm, tltm, hmsm, ttm, sb'm, tmnym, ts'm. T h e cardinal numerals



1 9 8 4 , 5 2 . 6 ,



2 1 - 9 9 comprise two or at most three words: the ten, the unit and
usually a word linking the ten and the unit, e.g. tt I ttm '66' ( K T U 1.4
= R S 2. [008]+ vii 9) or tmnym tmn kbd '88' ( K T U 4.179 = RS
d. '100' is mit /mi't-/,
'200' mitm (dual of mit). T h e hundreds from
300 are formed by connecting a unit and mat /ma'at-/ (pi. of mit),
e.g. tit mat.
e. '1,000' is alp /calp-/, '2,000' alpm (dual of alp). T h e thousands from
3,000 are formed from a unit and alpm (pi. of alp), e.g. hms alpm
'5,000' ( K T U 4.181 = RS 15.106:2).
f. T h e word for '10,000' (or 'myriad') is rbt or rbbt.

The verb


T h e inflected verb differentiates gender, number, person, aspect/tense

(imperfective or perfective; antecedent, contemporaneous, subsequent),
mood (indicative or volitive [imperative, jussive]), diathesis (active, reflexive, passive) and aspect (e.g. factitive, causative). Gender, number
and person are differentiated by various prefixes a n d / o r suffixes. Various
verb stems differentiate diathesis and aspect (see; aspect/tense
and mood are differentiated by a) subtypes of the prefix conjugation, b) the imperative and c) the suffix conjugation.
Alongside the genuine (finite) verbal forms two nominal (infinitive)
categories occur in connection with the verb system, i.e. participles
and infinitives. T h e y are morphologically and semantically directly
related to verbal categories. Besides gender and number their inflection also differentiates diathesis and aspect.

Morphological classes of the basic stem

T h e imperative

a. T h e imperative (impv.) is the mood of c o m m a n d in the 2nd pers.

Morphologically, it is identical with the short form of the prefix conjugation without the prefix and phonemically monosyllabic, i.e. <qtVl>
(V = / a / , / i / or / / ; the same thematic vowel as in the prefix conjugation). As a word-initial consonant cluster is not tolerated, the
impv. becomes bisyllabic, usually by insertion of an auxiliary vowel

after the first radical ( anaptyxis), generally / 2 / , m o r e rarely / / :

e.g. isp /}isfn/ < *'Vsup{ 'collect!' ( K T U 1.107 = R S 24.251+:33
etc.); uhd /"hud/ < *>hud 'seize!' ( K T U 1.82 = R S 15.134:6).
b. T h e inflected endings of the impv. are the same as the endings
of the 2nd person of the short form of the prefix conjugation: qHVl, qHVl, q'tVl ( not attested); c.du. q'tVl.
c. Besides the uninflected form of the line with
H e b r e w t h e r e is probably a lengthened ('emphatic') form qUVla,
marked by the suffixed m o r p h e m e /-a/,'21 e.g. sa /sa'a/ < *sa'a (^Ins3)
'raise/lift up!' ( K T U 1.5 = R S 2.[022]+ 13; K T U 1.14 = R S
2. [003]+ ii 22). T h e prefix conjugation
a. 'Prefix conjugation' (PC) is the generic term for various different
morphological subtypes which have differing verbal meanings. Inflection
is by means of prefixes and suffixes. In morphological terms and
functions the following subtypes of the PC can be distinguished (cf.


short form
extended short form
long form


a) perfective aspect, preterite

b) 'jussive' mood
jussive/cohortative mood
imperfective aspect, present


PC s p
PC s j
PC s e

b. T h e prefix c o n s o n a n t s of the P C are: y-\ a n d

2.m./ t-;
3.m.p1. t-/*y-;26 3.f.p1. and 2.m./f.p1. t-\
3.f.du. a n d 2.c.du. /-.
-; 3.m.du. y-/1-;


It is the same morpheme as occurs in the PCfe (cohortative mood); see
Normally a /-prefix (see D O B R U S I N 1981). There was also a variant with a
jy-prefixprobably attested only in two cases (KTU 1.4 = RS 2-1008]+ 17: yblk
'they should bring' || tblh, K T U 1.4 = RS 2.[008]+ 38.40:yblnn 'they brought').
This variant is no (longer) productive and only attested in grammatical parallelism
with the 'normal' /-prefix form.
Forms with the y- and /-prefix both occur (with almost the same frequency).

c. T h e personal suffixes of the PC are the endings of the short form

of the prefix conjugation (PC S ): 3.m./,, -0; /-/; 3./ / - / ; 3./ /-a/(?); 30 3./2.du. /-/.
d. T h e PC in the basic stem of the underlying 'strong' roots has the
following structure: <CV\qtV2l> (paradigm root Vqtl; C = any prefix
consonant; V | = prefix vowel; V 2 = thematic vowel). T h e thematic
vowel (TV) is / a / , / z / or / / , the prefix vowel (PV) either / a /
(before T V /u/ or til) or lit (before T V /a/). T h e following PCpatterns occur: <Caqtul>, <Caqtil> und <Ciqtal>.3] T h e choice of
T V is essentially dependent a) on the semantic class of the root (roots
with a fundamentally stative meaning usually have the T V / a / , roots
with a fientic basic meaning have either t u l or t i t as a T V , b) and
on the phonetic quality of the third and second root consonant (roots
with gutturals in second or third position often have / a t as TV).
e. T h e paradigm of the PC S is therefore (paradigm root ^lqtl, T V / u l ) :





taqtul- (as masc.)

- ( ? )

f. T h e forms of the PC S without endings have a morphological variant with the suffixed m o r p h e m e I-at instead of - 0 (= PC s e). 32 It is
only attested in connection with jussive forms and so can be termed
a lengthened or 'emphatic' jussive. T h e lengthened jussive is attested
in an unequivocal spelling a large n u m b e r of times only in the 1st
p. sg. In analogy with Hebrew this form can be called 'cohortative'.
As the lengthened jussive is not attested for every person and is not
in functional opposition to the ordinary jussive (= PC s j), it is not an




3 1 7 - 9 has a different view and postulates a P C L tqltn = /taq-

tulna/ and a PC Iqtl = /laqtul/ for the 3.f.p1. However, a PC-form of the 3.f.p1.
without the -n ending does not exist.
For these vocalic sequences see already B A R T H 1894, 4-5. For the validity of
'Barth's Law' in Ugaritic see especially V E R R E E T 1983b.
It is therefore the same morpheme as probably also occurs in Ugaritic on the
lengthened impv. (; see

autonomous grammatical category. In other words, unlike Arabic,

Ugaritic does not have a specific 'yaqtula-mood'.
g. T h e long form of the prefix conjugation (PC 1 ) differs from the
PC S by an additional suffixed morpheme. Forms without an ending
in the PC S have the ending / - u / in the PC1'; forms with a vocalic
ending in the PC'S (except for the 3./2. have the additional
ending -n = /-na/ or /-ni/ in the PC 1 :

1 .c.




taqtul-na (?)
*taqtul-naa (?)

h. In earlier research the question was hotly debated whether in

Ugaritic there was also a long form of the prefix conjugation with
the pattern <CaqattVl>, comparable to Akkadian iparrVs, Ethiopie
y^qathl or similar formations in modern South Arabic languages. 33
F e n t o n (1970) and M a r c u s (1975, 75-104, esp. 97ff.) demonstrated
independently, however, that this category does not exist in Ugaritic.
T h e Ugaritic texts published over the last twenty years provide
absolute proof of this. T h e suffix conjugation
a. T h e Suffix conjugation (SC) is not a homogeneous category in
Ugaritic, as is also the case in other West Semitic languages. T h e r e
is a fundamental difference between SC-forms with stative meaning
and those with a fientic (perfective, mostly preterite) meaning. T h e
former can be called 'statives', the latter 'perfects' (abbreviations:
'SCs' and 'SCp'). T h e subtypes mentioned also differ from each other
morphologically by different thematic vowels (see c).
b. T h e paradigm of the suffix conjugation is as follows:








qat VI-
qatVl-/ (?)
*qat Vl-tun(n)a
*qat Vl-na/


qatVl-tum (also masc.)


c. T h e thematic vowels of the SC are / a / , / i / and / u / . / a / is reserved exclusively for the fientic-perfective subtype of the SC (SCp),
/ u / for the stative subtype of the SC (SCs). I i i occurs in both subtypes. In the fientic subtype, however, it is limited to roots with I I /
III guttural, where the thematic vowel of the PC is / a / . T h e system
of thematic vowels in the SC and their equivalents in the PC can
be summarized as follows: SC qati/ ula - PC Ciqtal (stative); SC qatala PC Caqtu/il (fientic); SC qatila - PC Ciqtal (fientic II/III-guttural). Finite Verb Forms with Energie Ending
a. Besides the inflectional endings, finite verb formsespecially in
poetry and in object suffixesoften exhibit a so-called energic ending,36 spelled either -n or -nn. At least two perhaps even three different
alloforms of the energic m o r p h e m e can be distinguished (energic
types I / I I / [ I I I ] ) . As the energic endings may occur, basically, on all
finite verb forms, whether they are indicative or volitive, the energic is not a mood in the strict sense.
b. By far the the most commonly attested allomorph of the energic
is -n /-()nnV/ (= energic type I). It may stand alone and or before
the 3rd pers. sg. pronominal suffixes. In combination with 3rd pers.
sg. suffixes, the ending is -nh, to be vocalised as I-anna-hI or / -annaha/ respectively.
c. T h e r e is also an energic allomorph: -nn = l-ninl(?) (= energic
type II). It occurs exclusively in combination with 3rd pers. sg.
suffixes. T h e initial consonant, I hi, of the pronominal suffix is thus

Alternatively /i/- vowel, i.e. / qatVl-tim(V)/. Similarly for 2.f.p1. (/qatVl-tin(n)/)
and 2.c.du. (/qatVl-tim/).
Alternatively: /-til (as in Canaanite). In favour of /-t/ however is that the
independent personal pronoun in Ugaritic also ended in / - / (/'anku/).
On this topic see especially V E R R E E T 1 9 8 8 , 7 9 - 9 8 and K R E B E R N I K 1 9 9 3 .

always assimilated to the second /n/ of the energic ending: -nn

/-nVnn/ < *-nin-hu or /-nVnna/ < *-nin-ha respectively.
d. Perhaps Ugaritic also had a third allomorph of the energic, i.e.
-n = /-an/ (= energic type III). T h e orthographical proof for this
allomorph has not yet been produced. Aspect and tense
Verbal aspects and (relative) tenses are mainly differentiated by the
subtypes of the PC and the fientic variant of the SC. T h e functions
of these categories can be set out in the following table (paradigm
root ^lqtl,


yaqtul (PCsp)
*yaqtul(a) (PC s j/e)



qatala (SCp)
*qatala (SCp)
qatala (SCp)

yaqtulu (PC L )
yaqtulu (PC L )
yaqtulu (PC L )

All the fields on the right = imperfective column of the table are
filled by the long form of the prefix conjugation (PC 1 ). T h e left =
perfective column includes the short form of the prefix conjugation
(PC S ) and the perfective-fientic suffix conjugation (SCp). In the field
'perfective anteriority', the P C s p and the SCp have practically the
same function. T h e field 'perfective-contemporaneous' is empty because facts which occur simultaneously are essentially imperfective.
Only a special function of the SCp, i.e. the function of the so-called
'performative perfect', 37 can be placed in this field. T h e field 'perfective-posteriority'with reference to indicative statementsis only
covered by the SCp. T h e function of the variants of the PC S also
in that slot is exclusively volitive (jussive).
T h e table shows clearly that the PC1 is always imperfective and
the PC S is always perfective. As the PC L is used for simultaneous
situations, this category is conventionally labelled the 'present'. As,
on the other hand, the indicative P C s p generally expresses previous
events, this category is conventionally called the 'preterite'. These
labels, which suggest an opposition of tense between PC 1 ' and PC S ,
are not in fact correct, as the PC1- can also denote previous events,

E.g. I rgmt Ik hereby surely tell you . . .' (KTU 1.2 = RS 3.367 iv 7).

provided that they are imperfective. An imperfective presentation is

demanded especially for situations which are marked by the features
of plurality or repetition. 38 T h e P C s p occurs for certain only in narrative verse and is used there as the usual narrative form for single
and instantaneous actions of the past. 39 In other genres perfective
actions of the past are obviously always expressed by the SCp.
However, the SCp also occurs in narrative poetry, sometimes as a
free variant of PC s p. 4 0 Moods
In the Ugaritic verbal system the 'indicative' (= declarative mode) and
'volitive' (= wish and c o m m a n d mode) moods are differentiated. T h e
categories PC s p, PC1- and SC (SCp and SCs) are used for indicative
statements. 41 T h e following have volitive functions: a) the imperative,
b) the PC S ] (jussive) and the PC s e (cohortative) as well asrelatively
rarelyc) both subtypes of the suffix conjugation, i.e. SCp and SCs.
T h e r e is no specific use of mood in subordinate clauses. Volitive
moods, i.e. P C s j and PC s e, occur only in subordinate clauses with
volitive (final) meaning. In Ugaritic there is no specific 'subordinating mood', comparable to the Akkadian 'subjunctive', which only
occurs in dependent clauses. 42 Participles
T h e pattern for the formation of the active participle of the basic
stem (G-ptc. act.) is <qtil>. For the passive participle of the basic
stem (G-ptc. pass.) probably the pattern <qatl> is generally used,
as in Canaanite. 4 3
E.g. p'rih I tmgyn hdm / rih I ymgy apsh 'His feet did not reach the footstool,
his head did not reach as far as its (upper) end' (KTU 1.6 = RS 2.[009]+ i 59-61
[general-continuing event, expressed by the PCL]).
See for instance 1u ilm rathm 'the gods lifted up their heads' (KTU 1.2 = RS
3.367+ i 29). The morpho-syntactic autonomy of the category PC s p is, however,
questioned by some scholars; for discussion see M . S . S M I T H 1994, 39-41. According
to Smith, PO s p and PC 1 although in morphological contrastare free variants.
On the parallelism between PC s p and SCp see M.S. S M I T H 1994, 49-51 and
1995, 797-9. On other functions of the Ugaritic S C see M . S . S M I T H 1994, 45-57
and 1995.
For the use of these categories see
For a different view cf. V E R R E E T 1 9 8 8 , esp. 8 - 1 0 .
Over the years scholars have repeatedly stated that in Ugaritic there could
also or only be other patterns for the passive participle of the basic stem. The discussion centres on the patterns <qatl> (G-Ptc. pass, in Aramaic) and <maqtl> (G-Ptc.
pass, in Arabic). So far, however, no proofs have been provided. Infinitives/Verbal nouns

As in Hebrew, an absolute infinitive and a construct infinitive may
be differentiated. T h e former corresponds syntactically to a noun in
the absolute state, the latter to a noun in the construct state or a
noun after a preposition.
T h e infinitive (inf.) of the basic stem generally has the pattern
<qatl>. Besides this there is in Ugaritic a series of differently constructed verbal nouns of the basic stem. T w o patterns are noteworthy. O n e is <qitl>, which occurs several times in syllabic spellings, e.g.
ni-ib-r /nigru/ 'guard' (Ug 5, 137 = RS 20.123+ i 5').44 T h e other is
<ti/al(a)t>, which occurs only in I - w / y roots and ^1hlk, e.g. sat /si'at-/
'going out, expression' ( K T U 1.4 = RS 2. [008]+ vii 30.32); d't /dacat-/
'knowlege' ( K T U 1.2 = RS 3.367+ i 16.32); Ikt /likt-/ 'going' ( K T U
1.10 = RS 3.362+ ii 28, 29). Whereas the pattern <qatl> is used
for both the absolute and the construct infinitives, other patterns can
only denote the construct infinitive. T h e system of verbal stems: basic and derived stems
a. T h e Semitic languages use a large number of different patterns
to express aspect and diathesis, called 'verbal stems'. T h e simple
basic stem of Semitic is morphologically unmarked. T h e 'derived'
verbal stems are, instead, indicated by specific morphological markers.
b. Ugaritic has the following ten verbal stems, which can be arranged
as follows:




basic stem
passive basic stem
reflexive basic stem
passive intensive

basic lexical function

passive of G45
reflexive of G etc.46
intensive, factitive etc.
passive of D

For further examples see H U E H N E R G A R D 1987b, 305-6.

On occurrences of the Ugaritic Gp-stem see M A R C U S 1971.
E.g. reciprocal, durative and iterative. At times no clear difference in meaning between Gt and G can be established. On the Gt and tD stems in Ugaritic




In stative verbs the function is chiefly factitive/causative (e.g. 'be good': 'to
make [someone/somethingl good'). In intransitive-fientic verbs the D-stem chiefly
provides transitivity. In transitive-fientic verbs it strengthens or pluralizes the verbal

(table cont.)





reflexive intensive
or causative
or passive

reflexive of D etc.
passive of
reflexive of etc.48
reflexive, passive49 etc.50


Four of the ten patterns can be called 'cardinal stems': G, D, and

N. G is unmarked (= Heb. qal); D is marked by gemination of the
middle radical ( Heb. piel); is marked by the prefix /- (= Heb.
hiphil in function); 51 is marked by the prefix n- (= Heb. niphal).
All the 'cardinal stems'except for have both a passive and
a reflexive variant. T h e reflexive forms have an additional element
t\ in the Gt it comes after the first radical, in tD (= Heb. hithpael)
before the first radical and in t directly after the causative marker
-. T h e passive forms, i.e. Gp, D p (= Heb. puai) and p ( Heb.
hophal in function) are differentiated from the corresponding active
forms only by different vowels. As this characteristic is usually not
expressed in the alphabetic spelling, in most cases examples of the
passive stems can be determined only from syntax or context. Some
authors 52 doubt 5 3 their existence in Ugaritic.
c. T h e paradigm of the verb stems (forms are vocalized; finite forms
are always; ptc. and inf. u n i n f e c t e d ; n.o. = no [certain]
meaning (plurality of agents or objects; repetition of an action). The D-stem is also
used for denominative meanings.
On the , p and t in Ugaritic see T R O P P E R 1990a, 2 1 - 1 1 1 .
Chiefly or only in prose texts.
E.g. reciprocal, ingressive and inchoative.
As non-sibilant causative morphemes occur in other Northwest Semitic languages
(e.g. the causative marker h- in the Heb. hiphil), repeated attempts have been made
by sholars over the years to prove these types of causative also occur in Ugaritic.
The attempts in question have not been convincing, however (see T R O P P E R 1990a,
The most uncompromising opponent of the existence of the passive stem in
Ugaritic is V E R R E E T 1 9 8 5 , 3 2 4 - 3 0 . It should be noted, however, that all the central Semitic languages have passive stems.
There are no other verbal stems in Ugaritic. On the so-called 'lengthened stems'
(L) see under and, on the so-called 'reduplicated stems' see
under In Ugaritic there are no stems corresponding to IX, XI or XII-XV
of Arabic.

PC s j


yiqtati/ al54
yuqattal ?
y Vtaqtil



q1 tu/il

quti/ ala
qutti/ala ?
('i)tqatti/ ala57
uqta/ ila



naqtVl ?

qtil (act.)
qatl (pass.)

muqtatil ?
muqattal ?











Morphological peculiarities of the 'weak' verbs

a. Five I - ' verbs have irregular G-PC-forms of the type yuC2C3

/yCi VC3/ instead of or as well as yiC2C3 - /yaCC2 VC:/ (cf. 3.2.2e):
^bd 'to perish', ^hb 'to love', ^hd 'to seize', kl 'to eat', yl'sp 'to
gather'. 6 1
b. I-h verbs usually have strong forms. Exceptions are verbs with / / /
as the second radical, i.e. ^lhlk 'to go' and ^hlm 'to strike, hit'. Both
verbs have G-PC-forms without /h/, e.g. ylk- /yalik-/ o r y l m - /yalum-/.
Whereas the remaining forms from VA/m are strong (e.g. G-impv.
him- /hum-/), ^Jhlk is weak in other ways, i.e. it produces forms without /h/: G-Impv. Ik- /lik-/; G-verbal noun Ikt /likt-/; Gt-PC ytlk
/yitalik/. -PC-forms from ^lhlk are instead strong: ashlk aahlik/
( K T U 1.3 = R S 2.[Ol4]+ 2, etc.).62

In forms with endings there was probably syncope of the corresponding vowel:
/yiqlatl/ < *yiqtatVl (cf. The same applies to other forms of the paradigm with similar syllabic structure.
Occurrences: Imthsh 'her fighting' (KTU 1.3 = RS 2. [014]+ ii 19); trntbs 'fighting'
(KTU 1.3 = RS 2.[014]+ ii 29); tljtsb 'quarrel' (KTU 1.3 ii 20.30).
/y/t/nuqattil/, but /'aqattil/ < *'uqattil (vowel harmony; see
Alternatively: /taqatti/ala/. The only certain example: w Ikms /wa-t(a)kamm Vsa/
'he fell to his knees' (KTU 1.12 = RS 2. [012] ii 54). Another possible example:
tmz' (KTU 1.19 = RS 3.322+ i 36.46).
Possibly /tuqattil/, cf. the uncertain syllabic spelling: tu-a-p-[ku(?)] /tuhappiku/
(Ug 5 137 = RS 20.123+ ii 23).
/y/t/nuaqtil/\ however /'aaqtil/ < *'uaqtil (vowel harmony).
< *yinqatil (chiefly undifferentiated by spelling).
The same verbs also have similar forms in Hebrew (verbs with 'weak aleph'
as the first radical). On the topic see esp. T R O P P E R 1990b, 367.
On the morphology of I-h verbs see T R O P P E R 1990d.

c. In I-n verbs and Mqh 'to take', the first radicalwhen vowellessis
assimilated to the following consonant, e.g.ygr /yaggur-/ < *yangur- (Vngr
'to guard' G-PC) or yqh /yiqqah/ < *yilqah (Mqh G-PC). 6 3 In most
I-n verbs the G-impv is formed without the first root, e.g. la/sa'a/
< *a3a (^ln' 'to raise', lengthened impv. [ K T U 1.4 = R S
2.[008]+ viii 5]). u /a'/ (VraT, m. pl. [ K T U 1.2 = R S 3.367 i 27
etc.]), sk /saf/ (<nsk 'to pour', f. sg. [ K T U 1.3 = R S 2. [014]+ iii
16 etc.]) as well as qh/qah/ ilqh, m. sg. [ K T U 1.4 = R S 2. [008] +
ii 32, etc.]). T h e verb ^lngr 'to protect' constitutes an exception: impv.
m. sg. ngr /mgur/ 'look out!' ( K T U 1.4 viii 14 [alternatively: N-impv.
/naggVr/ < *nangVra\).
d. III- J verbs arebesides \\\-w/y verbsof central importance for
understanding the Ugaritic verbal system as in principle in such verbs
the spelling allows verbal aspects and moods to be differentiated
clearly. PC1 forms occurs with the spellingyC\C 2 u (; for forms
of the PC S , however, the spellingyC\C 2 i ( is expected. In fact,
though, the situation is more complex as in Ugaritic word-final aleph
was no longer reliably pronounced. 6 4 O n this basis only verbs with
the PC thematic vowel H / permit an unequivocal differentiation of
the underlying classes: e.g. P Q ysu /yasi'u/ 'he goes out' i^yf) versus
PC S ysi /'yasiV (or /yasV/ < *yas) 'he should go out / he went out'.
In PC S forms with the thematic vowels / / and / a / this differentiation
is not given with certainty: spellings such as ybu (yn') and yu (yn1)
can be understood as PC 1, (/yabu'u/ or /yia?u/), but possibly they
could also be PC S forms (/'yab/ < *yabu', /yis/ < *yina').65
e. Verbs with /w/ or /y/ as the first, second or third radical present several forms which are irregular with respect to the forms of
the paradigm of the 'strong verb', as the semi-vowels /w/ or / y / ,
depending on position in the syllable, can occur either as consonants
or as vowels.
f. T h e paradigm of I - w / y verbs, which in Ugaritic includes ^lytn 'to
give', is characterised by the occurrence of forms both with and
without a (consonantal) first radical. If the word begins with a semi-

However, there is no assimilation of / / / in the N-stem: SC nlqht /nalqahat/
(KTU 4.659 = RS 19.166:1).
See 3.2.2d (quiescent aleph).
O n the grapheme < u > for / 0 / < *aJ see 3.2.2d. There are no attested forms
of the spelling ysi. On the topic see T R O P P E R 1990c.

vowel it is normally retained, as long as there no aphaeresis of the

whole initial syllable occurs as in impv. G. As word-initial *w in
Ugaritic generally becomes /y/ (see, except before a / u / vowel (inf. D), it is not possible to distinguish I-w roots from I-y
roots by spelling in most cases. In non-initial position the semi-vowel
(Iwl) is retained only after a / / - v o w e l (D-PC); in the other cases
either it causes a lengthening of the preceding vowel (e.g. in forms
of the stem) or it disappears unreplaced (e.g. G - P C forms). T h e
function of the inf. cstr. is generally filled by verbal nouns of the
pattern <ti/al(a)t> (see
Significant forms of the paradigm for l-w verbs i^yrd < *wrd 'to
descend'): 66














g. In forms of II-w/y verbs, the second (weak) radical never occurs

as a c o n s o n a n t but always causes a lengthening of the original
preceding or following vowel, G"PC' /yaqmu/ < *yaqwumu,
G - S C /qma/ < *qawama (yqurni).68 T h e thematic vowel of the , is usually / / in II-; verbs (e.g. /yaqmu/), in II-y verbs usually H/ (e.g. lyatuI).
Instead of the ('normal') intensive stems (D, D p and tD), I I - w l y
verbs form stems which are marked by lengthening the vowel between
the first and second radical and reduplication of the third radical (=
Heb. polel, polal, hithpolel). T h e y are conditioned variants of the
'normal' intensive stems (D, D p , tD) and have the same functions.
H e r e they are denoted by the symbols D*, Dp* and tD*. 69


The paradigm for I-y verbs is still blank at many points due to the lack of
significant forms.
' However, /'a(md1.
Conventionally vocalized as /yaqmu/ or /qma/.
These stems are conventionally called 'lengthened stems' and denoted by 'L',
'Lp' and 'tL'. The term 'lengthened stem' has been taken from Arabic but is

Significant forms of the paradigms for II-; and II-y verbs i^qwm
'to rise' and ^yt 'to place'): 70


yuq/ mamu
yitq/ mimu




qum/ qm






inf. abs.


h. T h e paradigm of I I I - w / y verbs is marked by the occurrence of

forms both with and without a (consonantal) third radical. In originally syllable-closing position (also when final) the weak third radical is always vocalic (contraction of diphthongs *iy > / / , *ay > //,
*uw > //, *aw > /0/ [see]). In intervocalic position it is
partly preserved as a consonant (e.g. ybky = /yabkiyu/ 'he weeps'
[G-PC 1]), partly the relevant triphthong is contracted. Which
triphthongs in particular remain and which are contracted is still not
clearly explained 74 in spite of extensive research. 70 It seems that occasionally paradigmatically identical forms occur both with and without contraction.
Forms III-; and III-jy verbs are usually orthographically identical.
Some indications of different paradigms of these two classes are provided however by forms such as atwt atawat/ (y'tw) 'she came'
( K T U 1.4 = R S 2.[008]+ iv 32) versus mgyt /magayat/ ^mgy) 'she
arrived' ( K T U 1.4 = R S 2. [008]+ ii 23, etc.).

misleading. Ugaritic 'L'-stems correspond to the III and VI verbal stems of Arabic
neither in form nor in function.
The II? forms in the derived stems are the same as II-w forms.
Shortening of the long vowel between the first and third radical when the syllable in question is closed, e.g. PC * /yaqum/ and PC L/S 3.f.p1. /laqumn/, analogically, impv. /qum/, PC S II-j /tait/, impv. U-y /it/.
Before personal endings beginning with a consonant, either Iqam-I < *qmor /q/m/ (cf. Heb. bnt [byn 'to understand', G-SC]). The same applies
analogously to the SC-forms of the derived stems.
Uncontracted forms of the type *qyim or *q'im are unattested.









The rules concerning triphthongs in respect of noun forms (cf. cannot be transferred to verbs without being modified.

O n the basis of comparative linguistics, one can conclude that in

P C as well as in SC different thematic vowels existed (III-w: / u / and
/a/; III^: /i/ and /a/). In III-y the commonest type was definitely
*yaqtiy- (PC) versus *qataya (SC), in III-w *yaqtuw- (PC) versus *qatawa
Significant forms of the lll-y paradigm, basic stem (Ibny 'to build'):
yabmyu (PC1); yabnf6 (PCS).
impv. b'n (; niy or bvn (; bVniy or bVn (


banaya or (rarely) ban (; banayat or bant (; banta

(; banay (

bniyu/a ( nom./acc.); bniyi ( gen.); bnt- (



i. T h e paradigm of weak geminate verbs (II-gem.) still presents m a n y

uncertainties. In the basic stem they are both 'strongly' constructed
forms, i.e. forms with reduplicated 2nd = 3rd radical, and 'weak'
forms, i.e. forms with doubled 2nd = 3rd radical. Doubled consonants, when final, are reduced to a single consonant. T h e distribution of strong and weak forms is not as in classical Arabic. However,
there are parallels with the H e b r e w paradigm for II-gem.
Significant forms of the II-gem. paradigm, basic stem iysbb 'to go

with endings: yasubb- (e.g. PC1 yasubbu).

without endings: yasbub or yasub < *yasubb.


sub < *subb (; subb (

sabba (; sabbata (;80 sabbanVy (I.e.du),

Forms of the derived stemsas far as can be ascertainedare strong

throughout. In agreement with H e b r e w (plel = p c e c ) it is possible
that instead of or alongside 'usual' D-stems (D, Dp, tD) the Ugaritic
II-gem. verbs form so-called 'lengthened stems' (D*, Dp*, tD*), 8i e.g.
Analogously III-w: PC s j tdu /tad'/ < *tad'uw 'she flew' (KTU 1.16 =
RS 3.325+ vi 6.7).
Analogously III-w: du /ds'/ ( (KTU 1.19 = RS 3.322+ iii 28); di /di/
< *ds'iyi ( (KTU 1.16 = RS 3.325+ 48); du /d$'/ < *ds'uy ( (KTU
1.19 = RS 3.322+ iii 14).
Analogously III-;: dit /d't-/ (KTU 1.108 = RS 24.252:8).
II -gem. verbs which are also I- provide an exception: all the forms have a
reduplicated 2nd = 3rd radical, e.g. tlt/latut/ < *lanlut 'they (3.f.du.) trembled'
(KTU 1.3 = RS 2. [014]+ iii 33, etc.).
Cf. Heb. sabbot < *sabbt.
For the term 'lengthened stem' and the sigla see g.

D*-PC *yusabib- instead of D-PC *yusabbib- (vowel lengthening instead

of consonant lengthening). T h e orthography is ambiguous.
j . Only a few Ugaritic verbs have four radicals. T h e three most
important formations are Crr/l-C3-C4,82
b) Ci-C 2 -C r C 2 83 u n d c) Cr
C2-C3-C3.84 T h e forms of types (b) and (c) are conventionally understood as 'reduplicated stems' (R-stems) of roots with three or two
radicals. 85


a. Adverbs of place:
tm, tmt, tmn, tmny. 'there'; 7, cln 'above'; I pnm 'before'; b'dn 'behind';
atr '(directly) after'; pnm 'within/inside'.
b. Adverbs of time:
ht, htm, cnt 'now'; idk, ap(.)hn, apnk, b km 'then; thereupon; a!}r
'after(wards), later', atr '(directly) after'; ahrm 'in succession'(P); clm
'on the following/next (day)' (alternatively: 'further').
c. Modal adverbs:
k, kd, kmt 'thus, in this way'; Ibdm 'alone'.
d. Interrogative and indefinite adverbs:
iy, i, 'where?'; an 'whither?'; ik, ikm, iky, 'how? why?'; Im 'what for?


T h e Ugaritic prepositions mostly denote an adverbial position but

in connection with certain verbscan also be used directionally.
T h e y can then fundamentally express both directions, terminative
and ablative. 87
82 <grd (KTU
1.14 = RS 2.[003+] i 11.23: SC grd); ^lprsh (KTU 1.2 = RS
3.367 iv 22.25: SC [probably N-stem] yprsh).
83 <gtgl (KTU 1.13 = RS 1.006:33: SC glgl)\ -Jgrgr (KTU 1.23 = RS 2.002:66:
PC tgrgr); Ihmhm (KTU 1.17 = RS 2.[004] i 40.41: SC hmhmt); ylkrkr (KTU 1.4 =
RS 2.[008]+ iv 29: PC ykrkr, <qiqi)\ (KTU 1.114 = RS 24.258:5: PC yqtql).
84 ylshrr (KTU 1.3 = RS 2. [014]+ 17; KTU 1.6 = RS 2. [009]+ ii 24, etc.:
SC shut).
See t / T 9.41-2.
On this topic see especially A A R T U N 1974, 1978.
On this topic see P A R D E E 1975, 1976.

a. monoconsonantal prepositions:
b (also: by), syll. bi-i /b/ 'in, at, on, with, from'; / (also: ly), syll.
le-e // (or / l e / ) 'towards, for, against; from; away from; to';
k /ka/ with the (lengthened) variant kmm 'as, like'.
b. prepositions formed from bi- or triconsonantal roots:
yd /yada/ 'next to, together (with)'; 'm /'imma/ '(together) with;
towards'; bn /bna/ 'between'; cd /cad/('?) 'until'; 7 /Cal// 'upon;
towards; down from; away from'; tht /tahta/ 'under, beneath'; qdm
/qudma/(?) 'before; in front of'; b'd /ba'da/ 'behind'; atr /'at(a)ra/
'in the wake of; (directly/immediately) after'; tk /tk/(?) '(right) in
the midst of; into'; qrb /qarba/ 'in the interior; into'.
c. Composite prepositions: (preposition b or / + noun):
bd /bdi/ < *bi yadi ' i n / f r o m the hand of; from'; b tk /bi tki/ 'in
the midst of'; b qrb /bi qarbi/ 'in the midst of'. I p /li p/ 'according to, in the m a n n e r of'; I pn /li pan/ 'to the front of; before;
before (temporal); away from (spatial)'; I p'n /li paen/ 'at (both)
feet of; (low) before'; I ir /li in/ 'on top of; on; onto (movement); from o f f / o n (movement)'; I bl /li bal/ 'without'.
d. Prepositions can be lengthened by the enclitic particles -m or -90
with no essential change in meaning. T h e forms b-m, k-m, l-m, cm-m,
l-n und cm-n are attested. T h e y are especially favoured in poetry.


a. coordinating:
- w /wa/ 'and; but': copulative conjunction.
p /pa/ 'and t h e n / t h e r e u p o n / c o n s e q u e n d y ' : copulative conjunction;
it marks a temporal or logical sequence. 90
- ap (extended variant: apn) 'thus, just as; even'. 91
- u /'0/ < *}aw 'or': disjunctive conjunction.
b. subordinating:
- ahr 'after': to introduce a temporal clause.
- id 'when; as soon as': to introduce a temporal clause.
d 'while; as long as; until': to introduce a temporal clause.

See d.
On Ugaritic p see W A T S O N 1990e, 1994e.
Also functions as an asseverative particle.

- hm /him/ with the phonetic variant im /'im/ 'if, in case': to introduce a conditional clause.
k /k/ (variant spelling: ky) 1. 'because': to introduce a causal
clause; 2. 'that': to introduce an object clause; 3. 'as, when i f : to
introduce a temporal or conditional clause.
- km /kma/ 'as; as soon as'.
- him !; as; as soon as'.


- Presentation particles ('lo!'): hn (extended variants: hnn\ hnny); hi

(extended variants: him, hin, hlk); mk.92
- vocative particles: y /y/\ l (cf.
i /'/ 'truly!' (only in oaths); an 'oh!' (exclamation).

Asseverative particles

- k, al, dm, I, mc: 'truly!; certainly!'.

Optative particles

- I /l/ (alternatively: /la/):

sive (PC s j/e).
- ahl: 'alas!; if only!'.

proclitic optative particle before a jus-


- / / l a / : for negating words and verbal clauses. 93

- al /'al/: for negating volitive verbal clauses (only before P C s j / e ) .
bl /bal/ (extended variant: bit): for negating words and noun clauses;
in verbal clauses only in an interrogative sense.

Existential particles

- it /'it/ < *'itay: 'there is/are'.

- in /'na/ < *'ayna (extended variants: inm; inn): 'there is/are not'.


Alternatively: 'then, after that' (adverb of time).




Enclitic particles

T h e most important enclitic particles in Ugaritic are -m,94 -n, -y, -k

and -t. T h e enclitics -m and -n, which are by far the most frequent,
serve generally to emphasise certain constituents of a sentence. T h e
enclitic -y obviously acts as a marker of direct speech; 95 -k and -t
occur chiefly in connection with pronouns and adverbs. 96
(Translation: W.G.E.



O n which see W A T S O N 1992c, 1 9 9 4 F .

See T R O P P E R 1994a.
Note the special abbreviations: C = any consonant; c. = common gender; obi. =
oblique case; PC = prefix conjugation; PC S = short form of the prefix conjugation;
PC s e = lengthened short form of the prefix conjugation (cohortative); PC s j = short
form of the prefix conjugation with jussive function; PC s p = short form of the prefix
conjugation with perfective-preterite function; PC1' = long form of the prefix conjugation; PV = vowel of prefix; SC = suffix conjugation; SCp = fientic sub-type
of the suffix conjugation; SCs = stative sub-type of the suffix conjugation; syll. =
syllabic; T V = thematic vowel; V - any vowel.

U g a r i t i c

W i l f r e d


L e x i c o g r a p h y

G . E .

W a t s o n

Previous work

Not unexpectedly, the meanings of Ugaritic words have been a matter for discussion and analysis right from the earliest days of Ugaritic
studies and it is due to the difficulties inherent in these texts that
many words remain unexplained to this day. Several surveys of previous studies are available 1 and there is no need to repeat all this
material here. As yet, the only complete dictionary is Aistleitner's
Wrterbuch der ugartischen Sprache (1963, etc.). T h e only other comparable 'dictionary' is the glossary which formed part of Gordon's series
of handbooks to Ugaritic: with each new edition, the glossary was
revised as new texts were discovered and different solutions were
proposed. 2 Like the Wrterbuch, Gordon's glossary listed all the words
found in the Ugaritic texts discovered at the time, including personal and place names. Partial glossaries are to be found in Segert's
g r a m m a r ( S e g e r t 1984, 175-205) and in the various translations of
the Ugaritic texts now available. 3 Specialised glossaries are included
in studies on the following: the hippiatric texts, 4 the ritual texts. 5 Also
important are the studies of prepositions 6 and the particles 7 and in
spite of its age, d e M o o r 1971 remains a mine of lexical information.
Of particular interest are the study of terms connected with textiles used in the Ugaritic texts ( R i b i c h i n i - X e l l a 1 9 8 5 ) and the
discussion of terms for sheep and goats ( d e l O l m o L e t e 1 9 9 3 ,
1 8 3 - 9 7 ) and of terms for sacrifice ( d e l O l m o
L e t e
1 9 9 5 ) . Useful,
too, are Pardee's listing of lexical items with bibliography ( P a r d e e
1987) and the studies of syllabic spellings. 9 T h e personal names of






1 9 7 3 ; DEL O L M O L E T E 1965








1988; WATSON













































see review by



Ugarit are also a source of lexical items even though their meaning
may not have been noticed overtly either by those who gave them
or by those who bore them. T h e classic collection by G r o n d a h l
( G r o n d a h l 1967) is a useful if somewhat dated reference work in
this respect. Some recent studies have provided additional material. 10
T o p o n y m s also contribute lexical items but their origins are more
difficult to ascertain."
Several series of articles on Ugaritic lexical problems have been
written by various authors, some of which are to be continued, 1 2 as
well as sets of studies on Ugaritic semantics. 13 T h e r e are many notes
and articles on individual words or groups of words which cannot
be listed here. 14 It is very helpful when a study is devoted to words
belonging to a particular semantic field: sociology ( R a i n e y 1963),
fabrics and dyes ( v a n S o l d t 1990), sacrifice ( d e l O l m o L e t e 1995),
crafts ( S a n m a r t i n 1995) and the army ( V i t a 1995a). For various reasons, some words receive more attention than others, for example,
words which occur in the mythological texts. 15 A reverse glossary
(English-Ugaritic) is provided in UT, 530-7. 1 6
In recent years actual dictionaries are starting to be published.
O n e is the Diccionario de la lengua ugartica (DLU) by del O l m o Lete
and Sanmartin, a two-volume work of which the first volume has
appeared and the second is at an advanced stage of preparation.
Another is Cohen's Comprehensive Dictionary of the Ugaritic Language, as
yet unpublished. 1 7 T h e third such lexicon (UHw), ] R which was initiated in Mnster at Ugarit-Forschungen several years ago, will soon
be ready for publication but is available in the form of a word-list
( D i e t r i c h - L o r e t z 1996b).
1990a, 1990b, 1993, 1995b, 1996a.





1 9 8 7 ; cf.


and D I E T R I C H - L O R E T Z SANMARTIN; bibliography in SEL

5 1988, 2-12. Many of their studies are concerned with differentiating homonyms.







1 9 9 1 ; BERGER

1978, 1979, 1980, 1988;




1984, WATSON





1 9 7 0 ; MARGAI.IT

1978b, 1980. See also


1 9 8 2 ; SANMARTIN



1965, 1979,

AJJAN - V I T A L E 1 9 7 6 ,




1 9 8 4 ; SANMARTIN


See, for example, W A T S O N 1996c.

E.g. lmm, which denotes a type of sacrifice, has 25 entries in P A R D E E 1987, 410.
A semantic glossary is provided in C A R T U ,
As mentioned in SIVAN 1997, xix.
18 Ugatisches Handwrterbuch.
Supplemented by the extremely helpful 'glossary' in D I E T R I C H
L O R E T Z 1996a,
543926. Card indices of lexical material are also held in research institutes (see,
for example, R I B I C H I N I - X E L L A 1985, 1 1 ) .


The texts

T h e texts under discussion are, of course, those in Ugaridc found

at Ras Shamra, Ras Ibn Hani and elsewhere (conveniently collected
in KTU'2), supplemented by more recent discoveries. T h e Akkadian
texts are relevant chiefly for the lexical material they provide, either
directly (as in the lexical texts) or indirectly (see K h n e 1974, 1975).
O f particular interest is the treaty 20 ( K T U 3.1 = R S 11.722+) of
which large sections in Ugaritic correspond to its Akkadian exemplars (RS 1 1.732, 17.227, 17.382; K n o p p e r s 1993 with previous bibliography). It can also be noted that some Ugaritic letters may in
fact be translations from Akkadian, Egyptian and Hittite.



Aside from the large n u m b e r of words which are known from comm o n Semitic (um, 'mother', klb, 'dog', etc.)21 it is difficult to determine the m e a n i n g of m a n y lexical items in Ugaritic for several
reasons. For one thing, the corpus is small and the range of significant
contexts is accordingly quite limited. Also, vowels (aside from the
use of the three aleph signs, 'a, 'i, 'u) are not indicated, and it is therefore not always easy to distinguish homographs. While prose and
verse texts share much of the vocabulary (e.g. thm, 'message'), certain
words are found only in non-literary contexts (e.g. gzl 'spinner' [ K T U
4.358 = RS 18.048:9]) whereas others occur only in verse (e.g klat
'both' [ K T U 1.1 = RS 3.361 iv 10 etc.]; phi 'stallion' [ K T U 1.4 =
R S 2.[008]+ iv 5.9.15, etc.]; rt 'dirt' [ K T U 1.16 = R S 3.325+
29, etc.]). 22 M a n y words are difficult or obscure, or have uncertain
etymologies, e.g. ilqsm, bnn, b^r, gml, gpr, ddym, dnt II, dqr, dnt, kb, kbm,
kdr, kmlt, knh, kpsln, krln, cbk, cprt, crgz, ctrb, gb, gbt, gprt, dmt,23 tkt, etc.,
and the meanings of others (aktn, agzr, aqhr, askrr, idm, idrp, idt, udbr,
unk, Cd, etc.) cannot as yet be determined. Some words occur in broken or difficult contexts, e.g. hkm, hnn, hsm, htn, kdt, gbz, gbt, gdm (see
DLU for details). With the discovery of new texts, previously unknown
words continue to be added to the lexicon but the meanings of these


O r letter accompanying a treaty ( K N O P P F . R S 1 9 9 3 ) .

Even here there can be false assumptions, as SANMARTIN ( 1 9 9 6 ) has shown in
respect of ahl which means 'town' (Akkadian alum) rather than 'tent' (as in Hebrew).
For the vocabulary of the Ugaritic letters see C U N C H I L L O S 8.1.5.


But see WYATT


can also be uncertain, e.g. udn, 'to give ear', ghr, 'to sound forth(?)',
zb, 'to foam(?)', qnn, perhaps 'to stand up', in RS 92.2014 ( P a r d e e
1997a, 327-8)



In view of the vast literature on Ugaritic lexicography and the often

conflicting or at least divergent conclusions reached by scholars, there
have been several attempts to set out solid methodological principles
for the determination of meanings. 24 These are discussed here briefly.
O n c e the conect reading of the text has been established, the context is
of crucial importance. In fact, all agree that context is the most
important single element for ascertaining what a word may mean.
Syllabic spellings must also be taken into account, 25 and finally, comparative philology can be used. For this approach to be valid a set of
rules must be applied: context is more significant than etymology;
without context, etymology can only uphold a hypothetical proposal;
phonological rules should only be flouted with supporting evidence;
words in another language may not necessarily have the same meaning in Ugaritic; homographs and homonyms should be assumed only
as a last resort; syntagmata and idioms as well as words need to be
compared; a distinction must be made between the (archaic) poetic
texts and the language of the letters, rituals and administrative texts;
in the poetic texts it is important to determine stichometry and parallelism; in general, the rules of g r a m m a r and syntax should be
applied. Finally, non-linguistic evidence should not be neglected.
Some illustration of these principles is provided below.


Use of cognate languages

Comparison with other Semitic languages can provide a significant

contribution to determining the meanings of words, but a degree of
caution is required. Healey has surveyed the contributions available from Hebrew, Phoenician, Arabic, Akkadian, South Arabian 2 6
and Ethiopie ( L e s l a u 1 9 6 8 ) , particularly Aramaic and Syriac 1 9 8 8 ) . 2 7




See S I V A N 1984a; H U E H N E R G A R D 1987b.

See especially R E N D S B U R G 1 9 8 7 .



1 6 9 ; DE M O O R





His conclusion, though, is that context is 'the ultimate arbiter'. 2 8

Arabic has been much used (and misused) as a resource for determining the meaning of Ugaritic words. This approach has been
examined in detail by Renfroe 2 9 who has shown that there are many
genuine Arabic-Ugaritic isoglosses but an equal if not greater number of spurious ones. In many cases we may simply have to say that
there is insufficient evidence for any firm conclusions. It always has
to be remembered that the meaning of a word in a cognate language cannot simply be transferred to Ugaritic and at times is no
more than a guide. T h e same applies to the contribution from Eblaite


4.6 Methodolog))
T h e first task necessary before resolving the meaning of a Ugaritic
word is to survey all previous attempts, which is often very timeconsuming, with no guarantee of complete coverage. T h e scholar
must then establish the correct reading on the tablet, determine the
context, perhaps use etymology based on established language laws,
refer to a wide range of Semitic languages, if necessary, use other
languages (including Egyptian, Hittite, Hurrian, and even Sanskrit
and Sumerian) and avoid the multiplication of homonyms and homographs. These rules, however, are an over-simplification. In practice,
several other factors need to be taken into account, as the following examples show.


Selected examples

Some examples can help to illustrate the above. Evidence from cognate (Semitic) languages can come from Phoenician ( d e l O l m o L e t e
1 9 8 6 ) , Hebrew, Akkadian, Aramaic and Arabic as well as from such
languages as Ethiopie 30 and even Syriac. For example, the verb nsr,
parallel to bky 'to weep' ( K T U 1.16 = R S 3.325+ vi 4-5), can be
explained from Syriac n'sar/nasar, 'to sigh, groan, m u r m u r , howl,
shriek, lament'. 3 1 Choice of the correct cognate is important; for










1985, 1986a, 1986b, 1989, 1992.

See D I E T R I C H - L O R E T Z 199Id on 'r.
H E A L E Y 1 9 7 6 ; SANMARTIN 1 9 7 8 , 4 5 1 . However, cf.

WYATT 1998C, 237,



instance in the expression b bz'zm ( K T U 1 . 8 0 = RS 1 5 . 0 7 2 : 4 ) which

could mean 'from the booty of goats', in view of Heb. baz, 'plunder,
spoil' (HALOT, 117). However, in the contexta list of sacrifices
it is more likely to mean 'a young animal from the udder of the
goats', i.e. an unweaned kid, where Ugaritic bz is an isogloss of
Arabic buzz, Jewish-Aramaic bzz3, etc. all denoting 'teat' ( S a n m a r t i n
1 9 7 9 , 7 2 3 - 4 ) . Extra-linguistic evidence can also help determine meanings, for example, glp may denote murex used as a body-dye, since
this type of shellfish was common near Ras Shamra ( d e M o o r 1 9 6 8 ) .
Correct syntactic analysis is important for determining the meanings of words as shown by Husser ( H u s s e r 1995) in respect of atr
in I'pr dmr atrh K T U 1.17 = RS 2.[004] i 2 7 - 8 (and par.). This
expression has been understood in various ways but because of the
parallelism with the previous line (lars mssu qtrh, 'who makes his spirit
come out towards the earth'), which refers to correct burial, the
preposition I also means 'to(wards)'. It is probably to be translated
'who protects his step towards the dust', and therefore atr cannot
mean 'place', 'shrine' or even 'sanctuary'. T h e meaning of a word
can depend on several factors including the structure of a text and
recognition of the correct meaning of another word in the same passage. For instance, in K T U 4.392 = RS 18.130, prs means neither
'steering pole (of a chariot)' nor 'horse' but more probably a type
of 'ration': Ihms mrkbt hm csrh prs bt mrkbt, 'For the five chariots of
the five divisions: ten />r-rations from the chariot-house', since hm
rh does not mean 'fifteen' but 'five divisions (of the army)' followed
by the numeral 'ten' ( V i t a 1996).32 It is also important to compare
not just isolated words but syntagmata in Ugaritic with those in another
(Semitic) language. Del O l m o Lete has provided a list of syntagmata
common to Ugaritic and Phoenician. 33 O f interest, too, is the term
hrs which occurs in the economic texts in connection with chariots
(e.g. K T U 4.145 = RS 15.034:8-9) as a syntagm in the form w.hrs
and means 'precisely, exactly', a usage borrowed from Akkadian. 34
T h e personal names provide a wealth of vocabulary, with many
items not otherwise attested, e.g. rgln ( K T U 4.619 = RS 19.047:7),


T h e text remains difficult because the term ant (line 2) is not yet understood.
1986b, 46~7 = 1996a, 32-3. For comparison with a syntagm
from Aramaic cf. W A T S O N 1992d.
DEL O L M O L E T E 1979; cf. V I T A 1995a, 57.


which is formed from the word rgl, 'leg'. T h e same applies to placenames such as bir, 'well' ( K T U 1.91 = RS 19.015:29, etc.). For both
types of names syllabic spellings can be of use in determining meanings.


Lexical tablets

Of considerable importance are the polyglot vocabularies which have

been found in Ugarit. These list the equivalents of words in four
languages (Sumerian, Akkadian, Hurrian and Ugaritic) and in five
of the eight tablets discovered so far, the Ugaritic column has been
preserved (details in H u e h n e r g a r d 1987b, 21-3). According to
Huehnergard, about 114 Ugaritic words have been vocalized in syllabic cuneiform spellings. For example:









'song' 36

T h e sources for the vocalization of Ugaritic are (1) the three aleph
signs, (2) syllabic spellings of Ugaritic words and to a lesser extent
(3) comparative Semitics. T h e aleph signs give some indication of the
associated vowels (or the absence of a vowel; see T h e syllabically written Ugaritic words and names are particularly helpful.
T h e list of such items in v a n S o l d t (1991a, 301-8) has 156 entries.
H u e h n e r g a r d provides a glossary with approximately 280 entries
1987b, 103-94). Similarly, S i v a n 1984, 185-295,
although his sources are not confined to texts found at Ras Shamra
(see 4.5). 3 ' Reference to other Semitic languages can only provide
an indicadon of possible spellings and has to be used with caution.


Non-Semitic words in Ugantic

T h e city of Ras Shamra was a melting pot of several nationalities

speaking different languages and both court and administration dealt
with documents in several languages. In ritual, particularly, sections
of text were written in Hurrian and Hurrian words occur liberally

As H U E H N E R G A R D 1987b, 97 notes, za-am-ma-r stands for zamru (the double

-mm- is incorrect) and all the forms are nouns rather than infinitives.
VAN S O L D T 1991a, 7 4 7 - 5 3 : 'Appendix C : The lexical texts at Ugarit'.
See VAN S O L D T 1989d for review.

in the Ugaritic texts. It is not suprising, then, that m a n y words in

the Ugaritic lexicon are in fact b o r r o w e d f r o m H u r r i a n , occasionally f r o m Hittite a n d m o r e rarely f r o m Egyptian or f r o m other nonSemitic languages. O v e r the years m o r e a n d m o r e such words have
been identified. 3 8 T h u s , although most Ugaritic lexical items have a
Semitic etymology, several are (or m a y be) of non-Semitic origin.
S o m e of these words are listed here u n d e r the following headings:
(1) H u r r i a n words, (2) Hittite words, (3) Egyptian words, (4) Sumerian
words, (5) I n d o - E u r o p e a n words, (6) words from other languages.
4.9.1 H u r r i a n words include alhn, 'steward' ( K T U 4.392 = R S
18.130:4; cf. K T U 4 . 1 0 2 = R S 11.857:25; K T U 4 . 3 3 7 = R S
18.024:11) b o r r o w e d f r o m H u r r o - U r a r t i a n allae-hhi-nn, 'housekeeper',
all, '(festive) g a r m e n t ' ( K T U 1.12 = R S 2.[012] ii 47, etc.), H u r r i a n
allu (Neu 1996, 314, . 22); itnn, 'gift' ( K T U 1.100 = R S 24.244:74),
Hurrian uatnannuf grbz, 'helmet' ( K T U 4.363 = R S 18.055:2), Hurrian
gurpisi; hdm, 'footstool' ( K T U 1.3 = R S 2.[014] + ii 22, etc.), H u r r i a n
atm ( W a t s o n 1996b); hbrt, 'vessel, container' ( K T U 1.4 = R S 2. [008] +
ii 9), H u r r i a n hubrushi; hptr, 'pot, c a u l d r o n ' (K TU 1.4 ii 8), H u r r i a n
huppataru; hrd, 'warrior', Hurrian huradi/e ( S t i e g l i t z 1981); kht, 'throne'
( K T U 1.2 = R S 3.367 i 23, K T U 1.4 vi 51 etc.), 40 probably H u r r o U r a r t i a n ; kkrdn, ' c h e f ' ( K T U 4.126 = R S 14.084:27); kmn, '(a surface measure)' ( K T U 1.3 = R S 2. [014]+ iv 38 etc.), H u r r i a n kumnw,
llh ( K T U 4.363 = R S 18.055:5) denotes part of trappings or harness, H u r r i a n lulahhi, (DLU, 245); gr, 'total', H u r r i a n heyar; pg[n)dr,
'a type of fabric' ( K T U 4.270 = R S 17.111:10), H u r r i a n pahandam-f
tbl, 'smith' ( K T U 4.790 = R S 86.2235:15), Hurrian tabid- ( D i e t r i c h L o r e t z 1990); tgpt, 'fe1t(?)' ( K T U 4.183 = R S 15.116 ii 10, etc.),
H u r r i a n tahape ( W a t s o n 1995c, 540); tkt, 'chariot' ( K T U 1.4 = R S
2. [008]+ 7, etc.), H u r r i a n uktu ( L o r e t z 1996). Note that some
words are Semitic with H u r r i a n endings, e.g. hdgl, 'arrowsmith' ( K T U
4.138 = R S 15.016:2, etc.) which is a H u r r i a n form of U g . hz,
'arrow' with the Hurrian -(hii)li ending ( S a n m a r t i n 1995, 179). Others


See the list provided by DE M O O R 1973, 98. Not all are correct, of course.
For additional material the indices of Ugarit-Forschungen and other periodicals may
be consulted. For a survey see W A T S O N 1995c, 1996c. See also P A R D E E 1996.
Borrowed through Middle Assyrian utnannu: cf. VON S O D E N 1988.
However, cf. DEL O L M O L E T E
SANMARTIN 1 9 9 5 .








are Semitic words in H u r r i a n guise, e.g. kid, 'bow' ( K T U 4.277 =

R S 17.141:1) is a form of qatu, 'bow' ( D i e t r i c h - L o r e t z 1978b).
4.9.2 Hittite words: ans, 'small of the back' ( K T U 1.3 = R S 2. [014] +
iii 3 5 ) , Hittite anaa ( d e M o o r 1 9 8 0 ) ; uiyn, '(an official)' ( K T U 6 . 2 9
= R S 1 7 . 3 6 4 : 3 ) , Hittite ur(/i]yanni; dgt, 'incense' ( K T U 1 . 1 9 =
3 . 3 2 2 + iv 2 3 , etc.), Hittite tuhhui-/tuhf}uwai-;
htt, silver' ( K T U 1 . 1 4
= R S 2 . [ 0 0 3 ] + iv 1; K T U
1 . 1 4 ii 1 7 ) , Hittite (or Haitian) (fottu()-;
hndlt, '(coloured wool)' ( K T U 4 . 1 8 2 = R S 1 5 . 1 1 5 : 1 7 ) , Hittite siGhandala;
hsn, 'domestic' ( K T U 4 . 1 3 7 = R S 1 5 . 0 1 5 + : 1 . 1 0 ) , Hittite baann-;
mtyn, '(garment)' ( K T U 4 . 1 4 6 = R S 1 5 . 0 3 5 : 5 ) , Hittite
garment)' ( R i b i c h i n i - X e l l a 1 9 8 5 , 5 2 ) ; spsg, 'glass' ( K T U 1 . 1 7 =
R S 2. [004] vi 3 6 - 7 etc.), Hittite zapzagi-, which denotes precious
stones or a mineral ( N e u 1 9 9 5 ) ; tpnr, 'chief scribe' ( K T U 3 . 1 = R S
1 1 . 7 7 2 + : 3 2 ; K T U 4 . 4 4 = R S 9 . 4 5 3 : 2 8 ) , Hittite tuppanuri, etc.
4.9.3 Egyptian words: 4 2 br, ' b o a t , (war)ship' ( K T U 4.81 = R S
11.779:2-3, etc.); ht, ' b r e a d ' ( K T U 1.41 = R S 1.003+:22), Egyptian
ht3 ( W a t s o n 1995a, 223-4); htt, 'silver' (see above), Egyptian hd; kw,
'drinking vessel' ( K T U 4.691 = R S 20.010:6), Egyptian kb; krk,
' p i c k a x e ' ( K T U 4 . 3 9 0 = R S 18.119:8 etc.), E g y p t i a n grg 'pick'
( S a n m a r t i n 1987b, 151); ktp, '(weapon)' ( K T U 1.6 = R S 2. [009] +
2); rr, p r a y ' (enclitic of entreaty; K T U 1.4 = R S 2. [008]+ i
20, etc.), tkt, 'ship' ( K T U 4.81 = R S 11.779; K T U 4.366 = R S
18.074); etc.
4.9.4 Sumerian words: ad, 'father' ( K T U 1.23 = R S 2.002:32 etc.);
ilg, 'stone' ( K T U 4.751 = R S 29.096:11); ksu, 'seat, throne', ( K T U
1.3 vi 15 etc.), krs/su, '(a type of forage or fodder)' ( K T U 4.225 =
R S 16.198 [a]+:16); plk, 'spindle' ( K T U 1.4 ii 3.4), etc.
4.9.5 I n d o - E u r o p e a n / I n d o - A r y a n words: agn, 'cauldron' ( K T U 1.23
= R S 2 . 0 0 2 : 1 5 . 3 1 . 3 6 ) m a y be cognate with Sanskrit agni 'fire'; mryn
probably Indo-Aryan, e.g. Sanskrit marya, 'hero'; sm, 'king', ( K T U
1 . 2 2 = R S 2 . [ 0 2 4 ] i 1 8 ) ; ssw/ssw, 'horse' ( K T U 1 . 7 1 = R S 5 . 3 0 0 : 7
etc.), Sanskrit asva ( D i e t r i c h - L o r e t z 1 9 8 3 ) ; tnn, 'archer' ( K T U
4 . 3 5 = R S 8 . 1 8 3 + ii 1 1 etc.) a n d p e r h a p s others such as smrgt




now needs updating.

'emerald', 43 Sanskrit marakata and Greek smaragdos (also found as maragdos), both meaning 'emerald' (WYATT 1 9 9 8 C , 9 1 , n. 9 0 ) .
4.9.6 Words from other languages: adr, 'door (?)' ( K T U 4.195 =
R S 15.184:5), explained by anduru of uncertain origin; 44 irp, 'vase,
container' ( K T U 4.123 = RS 13.014:20), is perhaps Hurro-Hittite,
unless to be explained by Egyptian irp, 'wine' and therefore, possibly, 'wine-container'; utiyn ( K T U 3.1 = R S 11.772+:30 has the syllabic spelling u-r[i-ia]-ni ( P R U 3 203 = R S 16.257+ iv 21) and may
derive from Hurrian, Hittite or some other language (cf. DLU, 62).
Generally speaking, in the case of some loanwords it is difficult
to know whether they have been loaned directly, or indirectly through
another language such as Akkadian, or even whether they are in
fact Kulturwrter or Wanderwrter. Some words may even have been
borrowed back from the language which initially borrowed them,
e.g. kht, 'throne', from Hurrian keshi, itself a loan from Semitic ksu
(DEL OLMO LETE - SANMARTIN 1995) and the same may apply to
mgn, 'gift' and mryn, 'warrior' ( O ' C O N N O R 1989). T h e r e were also
inner-Semitic borrowings, 45 and a distinction must be made between
cognates and actual loans, such as nmrt from Akk. namurratu, 'splendour' (PARDEE 1988b, 115).46



H o m o n y m s can be distinguished by context, comparative philology

and occasionally from syllabic spellings. Simple examples of homonyms
are bt 'house' and bt ' d a u g h t e r ' , both nouns; from comparative
Semitics and (where attested) syllabic spellings, it is possible to determine that the first word corresponds to / b t u / and the second to
/ b i t t u / (SIVAN 1984, 210.212). In the case of weak verbs it is also
difficult to determine the correct form of the root (e.g. does gl derive
from gll, gly or gyl?). Since the Ugaritic corpus is so small, it is quite
possible that a 'word' which occurs only a few times may have as


K T U 1.4 = RS 2.[008]+ i 32, which is read mrfrt in KTU2.

Listed as = daltu, 'door' in C A D A / 2 , 117; AHw, 51a ( C E C C H I N I 1984, 47).
Including loans from Ugaritic to Akkadian, e.g. Ug. mit, 'oar', which was borrowed by Ugaritic Akkadian ( V I T A 1995b).
In K T U 1.108 = RS 24.252:21.24; nouns with preformative n- are Akkadian,
not Ugaritic, as P A R D E E notes (ibid.).

many different meanings. 47 Examples include b'r I, 'to burn', b'r II,
'to a b a n d o n ' (only in the D stem); 48 gl I, 'shout of joy', gl II, 'cup'
and gl III, '(type of field)'; ptt I, 'linen' and ptt II, '(make-up) case'
( K T U 4.247 = R S 16.399:22; SANMARTIN 1987a, 54, n. 7).
A clear example of the importance of distinguishing homonyms
(and incidentally of correct word division) is provided by
yrk tcl bgr
She climbed the mountain by the flank,
mslmt bgr tliyt by the incline, the immense mountain.
wfl bkm ban She climbed bkm, Araru,
bm an wbspn Araru, Sapnu,
bn'm bgr tliyt the fair, the immense mountain
(KTU 1.10 = RS 3.362+ iii 27-31)
Although the sequence bkm looks like the particle bkm, 'thereupon',
this is impossible here as such particles are never postpositive: they
always come first in the clause (RENFROE 1 9 9 2 , 58). Instead, here
km means 'hill, m o u n d ' (as proposed by AARTUN 1 9 6 8 , 2 9 1 ) and it
is preceded by the preposition b (as part of the syntagm cly + b, 'to
climb'). 49 Hence the third line should be translated 'She climbed the
m o u n d , Araru'. 5 0


Ghost words

Non-existent words are due to scribal error, false readings, incorrect

analysis or incorrect word division. Examples of words written incorrecdy are any ( K T U 1.16 = R S 3.325+ i 7-8) which is to be read
liny, a place-name and tdrs ( K T U 1.45 = RS 1.008+:5), to be read
tdrq, 'tread'. T h e word ski, 'vizier' in K T U 3.1 = R S 1 1.772+:38
(KNOPPERS 1993) may have to be read skn (so KTU ). A classic example is ulp, taken by some scholars to mean 'noble, chief or the like',
though it is really to be understood as u, 'and' + Ip, 'like' 51 or as
u + / + p, 'whether from the mouth o f (see 13.6.2). In some cases,
the word division is uncertain, e.g. the sequence grbtil ( K T U 1.19 =


See especially the studies by

















However, cf. DLU, 107.

Where tp is the preposition I + p, 'mouth'; cf.
2 9 1 - 2 , with bibliography.




R S 3.322+ iii 47) could be analyzed as gr bt il 'resident in the house

of Ilu', as grbt il, 'leprosy of Ilu' 52 or even as grb til 'may you seek
asylum as a leper'; 53 kgmn may = k + gmn, 'like a funeral offering(?)'
or kgmn = Hurrian 'three-year old'. 54 O n the other hand, in K T U
1.96 = RS 22.225:1, most scholars corrected cnn to cnt[ 'Anat', thus
eliminating a previously unnoticed word which may mean 'evil eye'. 55


Future research

Although the core vocabulary of the Ugaritic texts is now understood to a large extent, there still remain many lexical items which
either need to be determined or require further clarification. For example, in the Keret epic, msb'thn bslh ttpl, ' T h e seventh of them fell
by (the) sW ( K T U 1.14 = R S 2. [003]+ i 20-1), it is uncertain
whether the deaths described refer to his wives or to his children or
indeed to the way the last victim died. T h e word slh could mean 'a
throwing weapon', 'a sword', 'war', 'lightning', the god 'Salhu', 'parapet' or a disease which affects babies (Babylonian ulhu). If the last
meaning applies, then this death must have affected Kirta's children,
which in turn is significant for the meaning of the epic (WATSON
However, the task of determining the meaning and or etymology
of individual words is not simply a luxury for students of Ugaritic.
Scholars in other branches of Semitic studies or in other disciplines
frequently need to consult reference works on Ugaritic for their own
purposes. Examples include the compilers of DNWSI and HALOT or
of encyclopaedias of various kinds or of comparative studies (e.g.
H O C H 1994). It is important, therefore, to establish as accurately as
possible what Ugaritic words mean. O u r main difficulties in understanding correctly many a difficult passage are principally that there
is no similar passage in Ugaritic or that the context is uncertain. In
other words, the Ugaritic corpus is simply too small owing to lack
of texts. Future discoveries and continuing research are our only
hopes in this exercise.






1986, correcting amd, the first word of the line, to tmd. See


1989a, 47~8.













1998c, 375 n. 1. See






E m b e d d e d within the syllabic cuneiform texts written by scribes at

Ugarit are over three hundred Ugaritic lexical items. These Ugaritic
forms appear in all genres of Akkadian texts.
In one group of syllabic cuneiform texts, those of the polyglot Syllabary A Vocabulary (Sa Voc.), the Ugaritic words were intentionally recorded by the scribes. T h e S a Voc. was a Mesopotamian lexical
series in which columns of individual cuneiform signs, in a fixed
order, were equated with one or more Akkadian words in a second
column (LANDSBERGER - H A L L O C K 1 9 5 5 ) . This lexical series was imported to scribal centres in the west, including Hattua, Emar, and
Ugarit. T h e Ugarit exemplars of the S a Voc. are unusual in that
they do not have merely the two columns of the exemplars found
elsewhere; instead, they add either one additional column giving lexical equivalents in Hurrian 1 or, more often, two additional columns
with equivalents in both Hurrian and Ugaritic. Six exemplars of this
quadrilingual type are known (VAN S O L D T 1 9 9 0 , 7 2 8 - 3 0 ) , on which
more than one hundred Ugaritic words are wholly or partly preserved. Thanks to the presence of Akkadian equivalents (and, when
those are broken away, the fixed order of the cuneiform signs), the
meanings of the Ugaritic words in the S a Voc. exemplars can be
established with more precision and certainty than is the case with
the Ugaritic lexical items attested in other text genres. Nearly all
parts of speech are attested a m o n g these forms as the examples on
p. 135 illustrate.
Apart from the S a Voc. exemplars, Ugaritic words appear in Akkadian texts either (a) as parts of the names of local geographical features or plots of land or, much more often, (b) by chance, essentially
lapsus calami in which the scribe either forgot the appropriate Akkadian
word and substituted a Ugaritic form or thought that the (Ugaritic)
form he was writing was proper Akkadian. T h e former group, which
by their nature are substantives and adjectives, occur in legal and

R S 21.062 (Ug 5 no. 135); also R S 94.2939, discussed by M. S A L V I N I and

B. A N D R - S A L V I N I at the 45th Rencontre Assyriologique Internationale, Cambridge,
Mass., July 6 1998.

in LH LO
S rc en
?. S CO
-50 S?
? o f cb
( Il C
II ; ~
sa " + :3 + :S + :s
+ :=
a h Il> 13 O
TtCMco CMco Tt- co CO
2 - CO^ - - - - ^ CO co CO
. ~' O
o oM o oCM o
o o

CM co CM C
C/} C/3 C/3 </3 c/3 (
/3 C/3 C/3 C C/3 C

ci ci Ci 0 Ci 0













- S
















! J





















economic texts; many of them correspond to designations attested

in alphabetic texts, such that alphabetic gt X corresponds to syllabic
A. (me5/1,i ' a) * or ( ) AN.ZA.GR X (see H U E H N E R G A R D 1987a, 11
n. 51), as in
gt gwl =
(/guw(w)1i/ 'circuit');
gt dpm = A.A-,ia d-ip-ra-ni-ma (/diprnrma/ 'junipers');
gt m'br = "AN.ZA.GR ma-ba-n (/ma'bari/ 'ford');
gt 'mq = [AN.ZA].GR: am-qa (/'amqa/ 'stronghold?');
gt gl = A.* 3 : hu-li (/gli/ 'low ground?').
T h e largest n u m b e r of Ugaritic words in syllabic texts, however, are
those that appear, seemingly at random, for an expected Akkadian
form. About a fourth of the Akkadian texts contain one or more
such Ugaritic words. T h e y are found in all genres, although they
are, understandably, relatively u n c o m m o n in texts that are copies of
Mesopotamian originals, i.e. lexical texts (other than the S a Voc.)
and literary texts; note, however, the following examples:
lexical EME.UR.GI 7 = la-a-nu UR.GI 7 me 'hound's-tongue' (a plant
name), with Ugaritic /1anu/ for Akkadian lin(u) (RS 22.034 +
349 = MSL 10 107ff. A, 110);
literary lip-hu-d-ma 'may they fear' RS 17.155 = Ug 5, no. 17a r. 7',
in which the root is Northwest-Semitic p-h/h-d 'to fear' (VON S O D E N
1969) but the form, as is usually the case in the few examples found
in literary texts, has been made to conform to an Akkadian paradigm (here, precative).
Ugaritic vocabulary is much c o m m o n e r in the many legal and economic texts written in syllabic cuneiform. Some economic texts contain several Ugaritic words, or even a preponderance of them (e.g.
RS 19.071= PRU 6, no. 114). An extreme instance is R S 17.240 =
PRU 6, no. 136, a list of men of different professions who are owed
a shekel (of silver), in which it is likely that every syllabically-written
word is Ugaritic; this text may be compared with the very similar
alphabetic text K T U 4.99 = R S 11.845, in which most of the same
terms occur, albeit in the plural and not in the same order:
RS 17.240 = PRU 6, no. 136
r 1 GN UG[U I
1 GN MIN 1[
1 GN MIN 1[
1 MIN '"[.TAM]



1 MIN '"SAN [GA]

KTU 4.99 = RS 11.845

pslm (line 17)

khnm (line 9)



1 MIN ^karbi-s[]



mf}sm (line 15)

kbm (line 7)
(cf. ngr krm in KTU 4.609 = RS

1 MIN '> 5 (QA)-/a-%]

1 M I N xHa-si-[ru]

ia-q- [u]

ysrm (line 11) or


ia-si- [hu] y shm

(line 19)
yqm (line 6)

1 MIN '"UGULA ma-[i

1 M I N ^mur- [


1 MIN lna-s[i-ku]

mru ibr^n* (line 12) and mru skn

(line 13)
nsk ksp (line 14)

As the examples cited thus far suggest, most of the Ugaritic forms
that occur in Akkadian texts (except for the S a Voc. quadrilinguals)
are nouns. A few finite verbs are also found, however, such as the
following suffix-conjugation forms:

3ms ta-ba-'a /taba'a/ 'he departed' RS 19.032 = PRU 6, no. 77:1;

3mp sa-ma-t /samat/ 'they devolved' RS 16.147 = PRU 3, 90b: 13;
3ms al/a-li-ma /a11ima/ 'it delivered' RS 20.012 = Ug 5, no. 96

?N 3mp na-ap-ta-ru /naptar/ 'they exchanged?' RS 15.123 + 16.152

= PRU 3, 89a:5.
T h e syllabically-written Ugaritic words are usually not identified as
such by any graphic device; they simply occur within an otherwise
Akkadian context, as in
ul-ma-tu GALme sa S'SM '4 large ship's hammers' (/hu1mtu/)'
RS 19.112 = PRU 6, no. 141:4;
'AN.ZA.GR TN (kVR<me) (la) al/a-li-ma 'the manor of TN has
(not) delivered (/sallima/) slaves' RS 20.012 Ug 5 no. 96, passim,
ma-a-ra sa TN 'and the tithe (/ma'ara/, acc.) of TN' RS 16.244
= PRU 3, 93b:7.


In many instances, however, the Ugaritic words are preceded by a

special sign that is written with two small angled wedges. This sign,
usually termed a 'gloss mark' Glossenkr), has several functions in
the syllabic texts from Ugarit (see H u e h n e r g a r d 1987a, 204-8), but
its most c o m m o n use is to mark the word that follows it as nonAkkadian (i.e. in all but a few examples, as Ugaritic). T h e gloss mark
is indicated by a colon in transliteration:
Emcs-/w KAme- sa : ma-a'-a-ri-a 'the grain and beer of its (sc. a
TN) tithe (/ma'sari/)' RS 16.153 = PRU 3, 146-7:10-1 (compare
the last example cited above);

i-na A. : ad-ma-ni 'in "redland (/'admni/) field'" RS 15.145 = PRU

3, 122-3:8, 12;
u -tu4 PN a-na 'MJLUGAL'-ft : sa-ma-ta 'and PN's field devolved
(/samata/) upon the queen' RS 15.086.15-16 = PRU 3, 51-2.
Most of the syllabically-written Ugaritic words are also attested in
alphabetic texts. O v e r one-fifth of the forms, however, are thus far
unknown in alphabetic form. In the case of some presumably comm o n words, such as the first example cited below, the absence of
an alphabetic attestation m a y be due to the poetic nature of m u c h
of the Ugaritic corpus.
ri-i\g]-lu /riglu/ 'foot' RS 20.123+ = Ug 5, no. 137 i 10';
b[i]-f}i-ru UTUu-g[a-ri-it] 'the elite troops (/bihir/) of Ug[arit]' RS
17.432 = PRU 6, no. 71:5';
[k]a-ma-'a-\lu) /kama'tu/ 'truffles' RS 19.035B + = PRU 6, no. 159:3';
ti-ib-nu /tibnu/ 'straw' RS 20.149 = Ug 5, no. 130 iii 17'.


Several Ugaritic consonantal p h o n e m e s do not occur in Akkadian.

T h e s e were generally represented in syllabic writings by signs whose
consonantal c o m p o n e n t approximated that of the Ugaritic sound:
/ 0 / appears only rarely, as in i-zi-ir-\tu^\ /'iirtu/ 'help' RS 20.149
= Ug 5, no. 130 iii 7';
/ 0 / is written with -signs: mu-a-bu /mo6abu/ 'seat' RS 20.123+ =
Ug 5, no. 137 iii 32"; ^u-uq-du(-)ma /9uqdu/ 'almond' RS 19.035B +
= PRU 6, no. 159:4'; ma--I}a-tu-ma /ma0hatma/ '(cloths)' RS
19.028 = PRU 6, no. 126:1;
/ z / probably appears in: zu-ur-PI /zurwu/ '(aromatic) resin' EA 48:8;
/ h / and / g / are written with -signs: ha-ra-^u} /harrau/ 'artisan' RS
20. 189a + ( L a r o c h e 1979b, 479) 7; fr-qu / h q u / 'lap' R S
20.123+ = Ug 5, no. 137) i 9'; ^ha-ma-ru- / g a m a r u - h u / 'his
apprentice' RS 19.042 = PRU 6, no. 79, 11; /}u-ul-ma-tu4 /gu1matu/
'darkness' RS 20.123+ = Ug 5, no. 137: i 15';
/ V , / h / , and / V are sometimes written with the Akkadian '-sign, as
in m-1 a-tu /ni J tu/ '(implements)' RS 19.135 = PRU 6, no. 142:2;
&meisa-'a-tu /s'tu/ '(wooden) bowls' RS 19.064 = PRU 6, no.
163: r. 4'; ma-a'-sa-H /ma'sari/ 'tithe' RS 16.153 = PRU 3, 146-7:11;
sometimes indicated by 'broken writings', as in ma-a-a-li /ma 5 a1i/
orac1e(?)' RS 15.092 = PRU 3, 54ff:25; tu-a-p-[ku] /tuhappiku/
'to be upset' RS 20.123+ = Ug 5, no. 137: ii 23'; si-il-a /sil'a/
'c1iff(?)' RS 16.249 = PRU 3, 96ff.:5; and sometimes, especially
word-initially, not represented, as in a-na-ku /'anku/ RS 20.149 =
Ug 5, no. 130: iii 12'; -[P]I-[/]u 4 /huwtu/ 'word' RS 20. 189A +
( L a r o c h e 1979b, 479) 12; ab-du / ' a b d u / 'slave' R S 20.123+ = Ug
5, no. 137: iii 4.

T h e greatest linguistic benefit of the syllabically-written words is the

evidence they provide for the vocalization of Ugaritic. T h e y show,
for example, that the patterns of some Ugaritic words differed from
those of their Northwest Semitic and Arabic cognates, as in
da-ab-hu / d a b h u / 'sacrifice' RS 20.123+ = Ug 5, no. 137: iii 6, versus
Aramaic, Hebrew and Arabic *dibh\
: ma-ad-da-t /maddatu/ 'measurement' RS 17.022 + 087 = Ug 5,
no. 5:9, versus Hebrew midd;
a-du-r /'aduru/ 'mighty' RS 20.123+ = Ug 5, no. 137: ii 34', versus
Hebrew 'addr.
A n u m b e r of phonological processes are also exposed by the vocalized syllabic forms. A m o n g these are
vowel assimilation around gutturals: tu--ru /tuhru/ < *tahru 'pure'
RS 20.123+ = Ug 5, no. 137: ii 1; [ u ' r u d u ' m e J W ^ ' - [ j ] ^ W m e
/mihrsma/ < *mahsma '(implements)' RS 19.135 = PRU 6, no.
142:4; ta-a-ma-tu4 /tahmatu/ < *tihmatu 'sea' RS 20.123+ = Ug
5, no. 137: iii 34";
raising of a and before w and y: [h]u-V\-tu^ /huwwatu/ < *hawwatu
'land' RS 20.123+ = Ug 5, no. 137: ii 10'; h-V\-ma /hiyyma/ <
*hayyma 'life' RS 20.426c + 20 = Ug 5, no. 131:6';
optional syncope of short vowels in open syllables: na-ba-ki-ma and :
na-ab-ki-ma /nab(a)kma/ 'springs' RS 16.150 = PRU 3, 47a: 16 and
RS 16.263 = PRU 3, 49b:5;' [b]a-ma-ru-m[a] and 1meif}a-am-rumamei/gam(a)rma/ 'apprentices' RS 15.042 + 110 = PRU 3, 196:
i 1 and RS 25.428:6 (see PRU 6, 150 n. 3); -W-[<] 4 /antu/ <
*anatu 'year' RS 20. 189a + b ( L a r o c h e 1979b, 479) 11 (see v a n
S o l d t 1990b).
T h e Ugaritic vocabulary attested in Akkadian texts has been studied
in detail in B o y d 1975, S i v a n 1984a (see the reviews of H u e h n e r g a r d
1987b, v a n S o l d t 1989d), H u e h n e r g a r d 1987a (see the important
review of v a n S o l d t 1990b), and v a n S o l d t 1991a. Several studies of individual lexical items have also appeared, including, recendy,
1988; S a n m a r t i n 1987b, 1992; v a n S o l d t 1989a; V i t a
1995b, 1996a; W e g n e r 1995; W i l h e l m 1992; X e l l a 1990.








It is rather difficult to give a precise definition of Ugaritic prose

texts. In general, they include all those texts that are assumed not
to be poetic texts, or at any rate do not reveal clear marks or criteria of poetry as found in the major Ugaritic myths and legends.
However, the distinction between poetry and prose is rather clear,
where poetic sections of myths and legends are interrupted by prose
sentences containing ritual prescriptions, instructions for performance
and recitation, or colophons ( K T U 1.4 = R S 2. [008]+ le.edge, K T U
1.6 = R S 2.[009]+ vi 5 4 - 8 ; K T U 1.16 = R S 3.325+ le.edge, K T U
1.17 = R S 2. [004] le.edge). For instance K T U 1.4 4 2 - 3 : wtb
Imspr. . ktlakn glmm 'and repeat the recitation that the lads were sent'
and K T U 1.19 = R S 3.322+ vi. le.edge whndt.ytb.Imspr 'and this (passage) should be recited once more' (referring to the legend from iv
23 onwards). W e shall see that hndt is a typical prose word. A similar line of instruction is included in the myth of K T U 1.23 = R S
2.002.56 ytbn yspr 1hm 1slmm* wyr pf}r klat, ' O n e shall repeat the
recitation five times before the images and the congregation together
shall sing. . .'. K T U 1.23 is a good example of how prose ritual prescriptions are interspersed in a poetic text, particularly in the opening sections ( K T U 1.23.12, 14-5, 18-22), but in the expiation ritual
K T U 1.40 = RS 1.002+.35 the instruction w.tb.lmpsr 'and start to
recite again . . .' appears in a prose discourse.
T h e majority of documents published in KTUX and 2 are prose texts.
T h e largest group, the economic or administrative texts, usually contain lists of persons and cities, or villages, often introduced by a label
or heading identifying the nature and purpose of the list and some-

times also preceded by the general marker spr. O f t e n only these

headings permit some grammatical and syntactic analysis. Together
with the letters, they may help to assess the criteria and character
of the Ugaritic vernacular used during the years of Ugarit's final
flourishing, basically the last fifty years following the reign of Ammittamru III, though some older documents survived ( K T U 3.1, 3.4,
7.65 = R S 11.772+, 16.191+ and 16.402[B]) mentioning such kings
as N i q m a d u II and his son Niqmepa. It has been assumed that these
latest texts reflect the m o r e developed language of everyday use
(SEGERT 1984, 13.1).

T h e purpose of this chapter is to review the different types of

prose style and syntax found in distinctive prose genres such as letters, contracts, or rituals. Distinct use of a given verb form may
occur in different types of discourse. Prose discourse is a constellation of functionally used verbal or noun clauses pertaining to a given
type of prose. Discourse types may, for instance, be narrative, precative, persuasive, prescripdve or performative and each function implies
the use of certain modes of verbs and noun clauses. This review
starts from the assumption that each type of prose is, in effect, such
a cluster of functionally and semantically used verbal or noun clause
types ( L O N G A C R E 1 9 9 2 , 1 7 7 - 8 ) . It implies that reports are basically
narradve, ritual and medical instructions prescripdve, and prose incantantions and letters persuasive. T h e borderlines between the different
types of discourse are not always well defined; performative elements
may also occur in rituals and incantations. Letters may contain narrative parts in so far as they function as reports. Such a functional
approach related to context and genre for the study of verb and
syntax in Ugaritic prose is more appropriate than the generic aspectual a n d temporal distinction made, for instance, by S E G E R T 1984
(particularly 64.2, but see R A I N E Y 1987, 397; T R O P P E R 1993a,
389ff.). W e cannot deal extensively here with the function of perfect
and imperfect in poetry in comparison with its function in prose,
but there is more overlap between prose and poetry than Segert
suggests. H e states that the perfect a n d imperfect acquired temporal character in the late Ugaritic vernacular (about 1200 BCE;
S E G E R T 1984, 64.21). However the perfect is used in poetry as a
narrative m o d e describing a completed action in the past, whereas
it still may a p p e a r in its constative and performative function in late
Ugaritic prose depending on the context, for instance in contracts
and rituals. Segert's assumption may in general apply to Ugaritic

correspondence and administradve texts, but the modes of use m a y

be different for other types of discourse.


Classification of the prose texts

Administrative texts

KTU2 lists 792 texts as economic or administrative. T h e y are by far

the largest group of prose texts. Not all of t h e m are administrative
texts (e.g. K T U 4 . 6 6 9 + = R S 1 9 . 1 7 4 A B is possibly H u r r i a n [ D I J K S T R A
1 9 9 4 , 1 2 5 - 6 ] and 4 . 6 5 9 = R S 1 9 . 1 6 6 a sales contract for a female
slave) a n d m a n y fragments are chips a n d bits that m a y in time be
joined to other documents (e.g. K T U 4.412 + 545 + 518 + 512 =
R S 1 8 . 2 5 1 + 1 8 . [ 4 7 1 ] + 1 8 . [ 4 3 5 ] + 1 8 . [ 4 2 6 ] ) . However, some texts
listed as religious texts, letters or juridical documents would better
be assessed as administrative documents (e.g. K T U 1 . 9 1 = R S 1 9 . 0 1 5 ,
2 . 2 7 = R S 1 6 . 3 7 8 A , 2 . 6 9 = R S 2 4 . 6 6 0 C a n d spr mnh bd mnny
K T U 3 . 1 0 = R I H 8 4 / 3 3 , c o m p a r e also K T U 4 . 9 1 = R S 1 1 . 7 9 5 ) .


T h e second largest corpus of Ugaritic prose texts that are susceptible of basic linguistic analysis are letters ( K T U 2.1-83). T h e r e is
some doubt as to the epistolary nature of some of the texts ( K T U
2.2 = R S 3.334, 2.5 = R S 1.020, 2.7 = R S 1.026+, 2.19 = R S
15.125 [manumission of a royal slave], 2.27 = R S 16.378A, 2.31 =
R S 16.394, 2.60 = R S 18.[528], 2.62 = R S 19.022 and 2.69 = R S
24.660G). Quite a few letters are purely formal epistles, or contain
only short messages apart from the usual airs and graces ( K T U 2.4 =
R S 1.018, 2.10 = R S 4.475, 2 . 1 1 - 7 = R S 8.315, 9.479a, 11.872,
[Varia 4], 15.007, 15.008, 15.098, 2.24 = R S 16.137 [b\+, 2.26 =
R S 16.264, 2 . 3 0 - 3 1 = R S 16.379, 16.394, 2.40 = R S 18.040, 2 . 6 3 64 = R S 19.029, 19.102, 2.68 = R S 20.199 a n d 2.71 = R S 29.095),
or they are too broken for coherent translation ( K T U 2.1 = R S
3.427, 2.3 = R S 1.013+, 2.6 = R S 1.021, 2 . 8 - 9 = R S 1.032, 2.[026],
2.18 = R S 15.107, 2.20 = R S 15.158, 2.22 = R S 15.191 [a], 2.25 =
R S 16.196, 2.35 = R S 17.327, 2 . 4 8 - 5 9 = R S 18.285[a], 18.286[ab],
18.287, 18.[312, 364, 380, 386, 387, 400, 443, 482, 500], 2 . 6 5 - 6 7
= R S 19.158B, 19.181 AB, 2 . 7 7 - 8 0 = R I H 7 7 / 0 1 , 77/21A, 7 7 / 2 5 ,
7 8 / 2 1 and 2.83 = R I H 78/25). Only a few offer larger portions of
prose to give an impression of the 13th century West Semitic per-

suasive m o d e of discourse used in diplomatic and business letters







































a n d 2 . 8 1 = R I H 7 8 / 0 3 + 7 8 / 3 0 ) . Twenty or so more letters

were found in 1 9 9 4 in the house of U r t e n u ( M a l b r a n - L a b a t 1 9 9 6 ;
D i e t r i c h - L o r e t z 1 9 9 7 ) , but are not yet available for analysis.


Ritual texts

Quite a large group are about fifty Ugaritic rituals and five lists of
gods. With this group should also be mentioned about 26 completely
or partially Hurrian ritual texts. T h e Ugaritic rituals include a series
of monthly rituals as a kind of service book through the cultic year.
T h e y contain prescriptions for daily sacrifices, seasonal festivals a n d
prayers. T h e H u r r i a n texts include sacrificial lists, sacrificial agr hid.
hymns a n d perhaps incantations:

Monthly rituals through the year

yrh ryn K T U 1.411| 1.87 = R S 1.003+, 18.056 and partial duplicates 1.39 = R S 1.001a.2-10|| 1.41 = R S 1.003+.11-9, 1.126 = R S 1.41.44-9;
yrh sm[et] K T U 1.87.54ff.|| 1.46+ = R S 1.009+ [ D i j k s t r a 1984, 6 9 f f ]
and partial duplicates 1.109 = R S 24.2531| 1.46+. 10-32, 1.130 =
R S 24.284111.46+. 11-21, 1.58? = R S 1.047, 1.134? = R S 24.294;
yrh n[ql] K T U 1.138 = R S 24.298;
yrh ib'lt K T U 1.119 = R S 24.266;
yrh hyr K T U 1.105 = R S 24.249, 1.112 = 24.256, 1.132 = R S
24.291 (partially Hurrian, continuation of 1.112? = R S 24.256), 1.148
= R S 24.643 rev?;
yrh gn? K T U 1.106 = R S 24.250+, partial duplicates 1.134 = R S
24.294 obv., 1.171 = R I H 7 8 / 1 6 .

Related texts with daily rituals a n d lists of sacrifices

K T U 1.48 = R S 1.019, 1.49 = R S 1.022, 1.50 = R S 1.023, 1.53 =

R S 1.033, 1.57 = R S 1.046, 1.58 = R S 1.047, 1.76 = R S 6.215,
1.81 = R S 15.130, 1.91 = R S 19.015, 1.104 + 7 . 1 3 3 = R S
24.248+24.305 ( D i j k s t r a 1998, 280-2), 1.110 = R S 24.254 (Hurrian

with Ugaritic gloss bW pamt), 1.111 = R S 24.255 (obv. Hurrian);

1.134 = R S 24.294, 1.136 + 1.137 = R S 24.296ab, 1.146 = R S
24.253, 1.156 = 24.656, 1.159 + 1.160 = R S 28.059AB, 1.162 = R S
[Varia 20], 1.165 = R I H 7 7 / 0 4 + 7 7 / 1 1 , 1.170 = R I H 7 8 / 1 1 ,
1.171 = R I H 7 8 / 1 6 , 1.173 = R I H 7 8 / 0 4 , 7.46 = R S 1.042, 7.177
= R S 24.653B;
id yph/ydbh
mlk: K T U 1.41 = R S 1 . 0 0 3 + . 5 0 - 5 , 1.90 = R S
19.01311 1.168 = R I H 7 7 / 1 0 b + 7 7 / 2 2 , 1.115 = R S 24.260, 1.164 =
R I H 7 7 / 0 2 B + , 1.139? = R S 24.300.



km t'rb GM,) bt mlk: K T U 1.43 = R S 1.005, 1.148 = R S 2 4 . 6 4 3 . 1 8 22, 1.139? = R S 24.300.

Occasional sacrificial festivals

K T U 1.91 = R S 19.015 obv. a catalogue of royal festivals;

spr dbh K T U 1.161 = R S 34.126;
dbh il bldn K T U 1.162 = 1.91.6 (= R S [Varia 20], R S 19.015.6);
dbh spn K T U 1.148 = R S 2 4 . 6 4 3 . 1 - 1 2 , 1.91.3;
dbh cttrt qrat.bgrn (partially H u r r i a n ) K T U 1.116 = R S 24.261;
a lung model with ritual a n d sacrificial instruction K T U 1.127 =
R S 24.277.

G o d lists

K T U 1.47 = R S 1.017, 1.74(?) = R S 6.138, 1.102 = R S 24.246,

1.113 = R S 24.257 (deified kings), 1.118 = R S 24.264+.

Expiation rituals

K T U 1.4011 1.84111.121 + 1.122 + 1.153 + 1.154 + 7.162? = R S

1.002 II 17.100[A]+ || 24.270a[b] + 24.650b + 24.652G+ + 24.652b?.

H u r r i a n rituals

K T U 1.26 = R S 1 - 1 1 . [ 0 4 8 ] , 1.30 = R S 1 - 1 1 . [046], 1.32 = R S

1.[066], 1.33 = R S 1.[067], 1.34 = R S 1.[076], 1.35 + 1.36 +
1.37 = R S 1. [069 + 070 + 071], 1.42 = R S 1.004, 1.44 = R S
1.007, 1.51 + 52 = R S 1.027 + 1.028+, 1.54 = R S 1.034+, 1.59 =
R S 1.[049a], 1.60 = R S 2.[006], 1.64 = R S 3.372, 1.66 = R S 5.182,
1.68 = R S 5.200, 1.110 = R S 24.254, 1.111 = R S 24.255 (rev.

Ugaritic), 1.116 = RS 24.261, 1.120 = R S 24.269+, 1.125 = R S

24.274, 1.128 = R S 24.278, 1.131 = RS 24.285, 1.132 = R S 24.291,
1.135 = R S 24.295, 1.148 = R S 24.643.13-7, 1.149 + 150 = R S
24.644 + 24.644[a], 4.669+ = R S 19.174A.
Administrative texts quite often also contain information about
rituals, in particular when they deal with the distribution and allocation of wine, food and other commodities for the cult, e.g. hmyn.bdbh
mlkt bmdr' 'five (kd) of wine for the sacrifice of the Q u e e n in the
sown land' ( K T U 4.149 = R S 15.039.14 6, see further K T U 1.91 =
R S 19.015, 4.168 = R S 15.082, 4.182 = R S 15.115, 4.213 =
R S 16.127.24, 4.219 = R S 16.179.2-3). Most ritual texts stem from
the High Priest's house and the house of the Hurrian Priest (PH
rooms 10-11) and just a few from Ras Ibn Hani.

Religious texts in literary prose

T h e r e is quite a large group of smaller texts and fragments which

may be classified as literary religious texts, if not myths. Some of
them were clearly composed as poetry (e.g. K T U 1.10 = RS 3.362+,
1.12 = R S 2.[012], 1.92 = RS 19.039+, 1.96 = R S 22.225 and
1.100 = RS 24.244), but some may be fragments of myths and incantations in prose or a kind of poetic prose ( K T U 1.9 = R S 5.229,
1.24 = RS 5.194, 1.25 = R S 5.259, 1.45 = RS 1.008+, 1.65 = R S
4.474, 1.82 = RS 15.134, 1.83 = R S 16.266, 1.96 = RS 22.225,
1.107 = R S 24.251+, 1.151 (?) = R S 24.647 and 1.169 = R I H
78/20). T h e r e are also occasionally prayers, one in poetic form ( K T U
1.119 = R S 24.266.26-36), but also in prose ( K T U 1.65 = R S 4.474
and 1.123? = R S 24.271), a blessing or dedication ( K T U 1.77 =
R S 6.411), a small collection of fables ( K T U 1.93 = R S 19.054)
and, perhaps, a wisdom text ( K T U 2.21| 2.5? = R S 3.334, 1.020).
With this group, we may also mention the few Akkadian .n/?M-prayers
in Ugaritic alphabetic script ( K T U 1.67 (+) 1.69 = R S 5.199 +
5.213, 1.70 = R S 5.156+, 1.73 = R S 5.303fc, 7.50 = R S 5.157,
7.52 = RS 5.196 and 7.55 = R S 5.218).

Other miscellaneous prose texts

Minor groups of prose texts are the juridical texts, medical prescriptions and omens. T h e juridical texts are mainly found in K T U
Section 3, but see also K T U 2.19 = R S 15.125 (manumission of a
slave) and K T U 4.659 = R S 19.166 (sales record of a female slave?)

a n d the custom licences (or h a r b o u r dues?) a n d caravan licences


4.172 =


15.093, 4.266


17.074, 4.336



and 4 . 3 8 8 = R S 1 8 . 1 1 3 ) . As sealed documents, the licences can be

considered a kind of juridical contract. T h e distinction between legal
documents and administrative records is not always clear. K T U 3.7 =
R S 1 8 . 1 1 8 and 3 . 1 0 = R I H 8 4 / 3 3 are rather administrative lists,
though they deal with legal charges such as ^ - o b l i g a t i o n s and debts,
as do K T U 4 . 3 3 8 = R S 1 8 . 0 2 5 and 4 . 3 4 7 = R S 1 8 . 0 3 5 + . Medical
prescriptions are found in K T U 1 . 1 7 5 = R I H 7 7 / 1 8 . T h e y are
sometimes included in other texts K T U 1 . 1 1 4 = R S 2 4 . 2 5 8 . 2 9 - 3 2
(an incantation for medical treatment of delirium) a n d 1.124 = R S
2 4 . 2 7 2 (an oracular report), a n d also perhaps the fragment K T U
1 . 8 8 = R S 1 8 . 1 0 7 . T h e s e prose texts and sections are related to the
hippiatric medical text of which four copies have been discovered

1.71 =



1.72 =



1.85 =




T h e large corpus of ancient N e a r

Eastern o m e n literature was also represented in Ugarit by a d r e a m
book (spr hlmm K T U 1 . 8 6 = R S 1 8 . 0 4 1 ) , a collection of astronomical omens ( K T U 1 . 1 6 3 = R I H 7 8 / 1 4 ) and birth omens of the summa
izbu type ( K T U 1.103+ = R S 2 4 . 2 4 7 + a n d 1.140 = R S 24.302:
D i e t r i c h - L o r e t z 1990a); also omens inscribed on fields of lung
and liver models ( K T U 1 . 1 2 7 = R S 2 4 . 2 7 7 , 1.141-4 = R S 24.312,
2 4 . 3 2 3 , 2 4 . 3 2 6 , 2 4 . 3 2 7 and 1 . 1 5 5 = R S 2 4 . 6 5 4 ) and a report of an
astronomical o m e n ( K T U 1.78 = R S 12.061). T h e r e is, perhaps,
also a protocol of n e c r o m a n c y with some ritual prescriptions ( K T U
1.124 = R S 2 4 . 2 7 2 : D i e t r i c h - L o r e t z 1990a).
1.97 =


23.484; C o h e n



Administrative prose

By far the largest group of prose texts are the administrative texts,
which include census lists of persons, guilds and cities, p a y m e n t rolls,
receipts and records of received or distributed commodities. T h e y
are a main source for private n a m e s a n d also a lexicographic goldmine, though m a n y words are still poorly understood. For the structure of the language they are less informative, since their syntactical
structure and style is often very simple. M a n y texts only have a simple label as heading mentioning a guild (hrtm 'ploughmen', K T U
4.65 = R S 11.602, 4.122 = R S 13.012; tnnm a kind of soldier, K T U
4.66 = R S 11.656; mrynm 'knights', K T U 4.623 = R S 19.049[b];
nqdm, 'sheep breeders', K T U 4.681 = R S 19.180; mdrglm 'guards ? ',
K T U 4.751 = R S 29.096; khnm, 'priests', K T U 4.761 = R S 34.123)

or a village/city/gentilic. T h e s e guild markers themselves are listed

as such too ( K T U 4.29 = R S 3.320, 4.38 = R S 8.272, 4.47 = R S
10.043, 4.68 = R S 11.716.60ff., 4.99 = R S 11.845, etc.); likewise
geographical markers in topographical lists (K TU 4.63 = R S 10.052,
4.232 = R S 16.355, etc.). Both serve as headings in texts which
contain persons grouped by trade, profession or provenance ( K T U
4.35 = R S 8 . 1 8 3 + , 4.69 = R S 11.715+, 4.71 (+) 72 = R S 11.721,
11.722, 4.103 = R S 11.858, 4.183 = R S 15.116, 4 . 4 1 2 + = R S
18.251 a n d 4.633 = R S 19.086A).
T h e different p a r t s of these simple syntactical structures are:
(1) heading (with or without introductory spr); (2) lists of persons,
towns, etc. (together with n u m b e r , commodity, etc.); (3) s u m m a r y or
total (with or without tgmr). T h e s e sections are often extended by
descriptive, or restrictive remarks in relative clauses. T h e style is usually concise in the extreme, leaving out self-evident terms a n d phrases
(e.g. tql, kbd, dd, tgmr, etc.). F r o m such texts, only a few prose sections
can be gleaned, in particular f r o m texts such as K T U 4.145 = R S
15.034, which is a small report revealing the p o o r condition of the
king's chariotry:
(1) tmn.mrkbt.dt. (2) ' (3) yd.apnthn (4) yd.hzhn (5) yd trhn/
(6) (7) inn. utpt/ (8) w.tlt.smdm.w.hrs (9)
Eight chariots, which entered the royal palace with their wheels, their
ax1es(?), their bearings(?), but two chariots have no quiver; and of three
two-horse carriages(?), the wheels are in the hands of the chief smith,
who took (them) out for repair.
It is a good example of the descriptive style found in administrative
texts (also K T U 4.136 = R S 15.013). T h e nature of the d o c u m e n t s
is often indicated by the w o r d spr, while the subject m a t t e r m a y be
persons, trades a n d professions, or commodities, tribute, rations a n d
fields u n d e r these headings extended with different types of relative
sentences: spr np d.crb bt.mlk w.b, 'List of people w h o entered
the royal palace, but who were not put into the list. . . ' ( K T U 4.338 =
R S 18.025.1-3). Usually, clusters with construct nouns do not exceed
three nouns or names, such as spr argmn p ( K T U 4.610 = R S 19.017);
spr ksp mnny ( K T U 4.791 = R I H 8 4 / 0 4 ) ; but cf. spr hpr.bns.mlk ( K T U
4.609 = R S 19.016), spr hr's qst iptl ( K T U 4.215 = R S 16.130).
D o c u m e n t s often have no h e a d i n g and start in mdias res. C o m modities listed and other entries are occasionally extended by descriptive relative n o u n clauses or participles:, '. . . a robe

that has a string(?) of carbuncles on it' ( K T U 4.132 = R S 15.004.4);

w.lp. d sgr.b/, '. . . a garment that has a fibula' ( K T U 4.166 = R S
15.078); tit mrkbt spyt.bhrs [.] f ir[.] smdm.trm.d[.l.s]py/w.trm.
spym/tit mrkbt d.l.spy, 'Three chariots covered with gold, ten pairs of
tr which are [not co]vered and a doub1e(?) set of tr covered, three
chariots that are not c o v e r e d . . . ' ( K T U 4.167 = R S 15.079.1-7),
but also verbal clauses: yn[hnm] (2) b.dbh.mlk, 'Wine that is
delivered into the hands of the pr[iests] for the sacrifice of the king'
( K T U 1.91 = RS 19.015.1-2; T R O P P E R 1991b, 355); qmh. d.kly.ksh.illdrm
bd.zlb[n], 'Flour that was completely spent according to the order(?)
of Illdrm into the hand of %b[n] . . . ' ( K T U 4.362 = RS 18.052.1-2),
ksp.d.slmyrmn.', 'Silver that Yrmn paid for (the mortgage on?) the
house . . . ' ( K T U 4.755 = R S 31.080; also K T U 4.95 = R S 11.836+,
4.166 = R S 15.078, 4.213 = RS 16.127, 4.290 = R S 17.297 and
4.348 = R S 18.036) and . . . prs qmh d nlm, '. . . a /w-measure of flour
that has been paid for' ( K T U 4.328 = R S 18.008.1). In relative
noun clauses the expression for existence it is often added ( K T U
4.235 = RS 16.369, 4.422 = R S 18.293, 4.617 = RS 19.044, 4.752 =
RS 29.097 [ S E G E R T 1982, 55.7]), and the usual negation in such
relative noun clauses is in(n) ( K T U 4.53 = RS 10.090, 4.180 = R S
15.105, 4.214 = R S 16.128 and 4.379 = R S 18.098).
Ugaritic scribes seem to have used two conventions to sum up the
totals of their administrative documents, either by writing the totals
(U.NGIN = napharu/gabbu) in cuneiform Sumero-Babylonian shorth a n d ( K T U 4.48 = RS 10.045, 4.63 = R S 10.052, 4.68 = R S
11.716, 4.69 = RS 11.715+, 4.71, 4.72 = RS 11.721, 11.722, 4.90 =
R S 11.797, 4.93 = R S 11.776+, 4.100 = R S 11.850, 4.102 = R S
11.857, 4.165 = R S 15.076, 4.219 = RS 16.179, 4.232 = R S 16.355,
4.299 = R S 17.345, 4.308 = RS 17.386, 4.340 = R S 18.027, 4.435 =
R S 18.[306], 4.610 = R S 19.017, 4.704 = R S 21.002, 4.745 = R S
25.417, 4.754 = R S 31.043, and 4.784 = R S [Varia 38],2; VAN
S O L D T , 1995, 485-6) or in Ugaritic tgmr, and sometimes even both
ways (e.g. sb'.mat ttm kbd/7 me-at 60 H I . M E , K T U 4.340 = RS
18. [027]). As with the headings, the pattern of such totals is not consistent. Usually the tgmr of the commodity received or distributed,
or the group or city is mentioned first (e.g. K T U 1.91 = R S 19.015,
4.67 = R S 11.714, 4.156 = R S 15.053, 4.269 = R S 17.106 and
K T U 4.151 = R S 15.044, 4.179 = RS 15.103, 4.777 = R I H 8 3 / 0 7 +
respectively), but it may also follow the total a m o u n t ( K T U 4.230
= R S 16.341, 4.764 = R S 34.176 and 4.137 = R S 15.015+, 4.141

= R S 15.022+ and 4.173 = R S 15.094 respectively), but the word

tgmr is also often left out ( K T U 4.164 = R S 15.075, 4.344 = RS
18.030, 4.427 = RS 18.299, 4.163 = R S 15.073.15ff, and 4.174 =
RS 15.095 respectively). T h e style of these texts is basically descriptive and strongly paratactic. Complex syntactic structures with subordinate clauses are almost absent.


Literary prose of incantations, stones and reports

Everybody knows, or rather thinks he knows the difference between

prose and poetry ( W a t s o n 1984d [1995] 44). T h e problem is to
establish sound criteria. We cannot deal here with this question in
extenso (see P a r d e e 1993a). T h e criteria often used to distinguish prose
from poetry in Hebrew literature, namely the absence or rarity of
prose elements such as the relative marker }"ser (less often the relative
pronoun), the definite article, the object marker and the narrative
waw, are not very helpful for Ugaritic prose. T h e existence of a narrative waw discourse in Ugaritic is still disputed and indeed, if it is
not to be found in the context of the incantation K T U 1.100 = RS
24.244.67-8: mgy.hrn.l bth w (68)ystqilh^rh, ' H o r o n reached his house,
and he entered his court', it occurs, perhaps, in a few prose texts
such as, for instance, the report about a necromantic inquiry ( K T U
1.124 = R S 24.272): ky mgy.adn (2) ilm rbm.'m dtn (3) wysal.mtpt.yld (4)
w/ny.nn.dtn (5) t'ny. . . uymg (11) mlakk.'m dtn (12) Iqh mtpt (13) uy'ny.nn
dtn . . .' W h e n the Lord of the Great Gods came to Ditanu and asked
for the boy's (oracular) decision, Ditanu answered him: "You will
a n s w e r . . . and your messenger to Ditanu arrived after he received the
(oracular) decision." T h e n Ditanu answered:. . .' As long as no vocalized narrative texts are available, the question will remain undecided.
O n the other hand, the use of parallelism or parallelizing style in
Ugaritic texts is not confined to poetry. As in the Hebrew Bible, it
is also found in many prose texts. T h e greater use of relative pronouns and particularly, a set of demonstratives (hnd-hnk/hndt-hnkt/hnhmt,
including the definite article hn-, R a i n e y 1971, 160; C u n c h i l l o s
1983b) and interrogative pronouns (mn(m)-mnk(m); mh-mhkm, etc.) in
prose texts such as economic texts, letters, contracts, etc. ( W a t s o n
1984d [1995], 62) is a fairly clear criterion, but not the only one.
Several texts show a mixed style of prose and poetry. It is hard to
say whether these literary texts are prose containing poetic elements,
or a kind of poetry in which the rules of poetic parallelism are weakly

used. T h e r e can be no doubt that m a j o r works from Ugarit such as

the Ba'al Cycle, the legends of Aqhat and Keret, the astral myth of
Shahar and Shalim were composed as oral poetry, in which inserted
prose elements clearly stand out in their context. Some minor myths
and incantations were also composed in the concise prosody of the
major works ( K T U 1 . 1 0 = R S 3 . 3 6 2 + , 1 . 1 2 = R S 2 . [ 0 1 2 ] , 1 . 8 3 =
R S 1 6 . 2 6 6 and 1 . 9 2 = R S 1 9 . 0 3 9 + ) , but a few are in prose, or
have prose sections alternating with mythical passages in poetry. For
instance, K T U 1 . 1 0 7 = R S 2 4 . 2 5 1 + clearly opens with a mythical
poem ( K T U 1 . 1 0 7 . 1 - 1 4 ) , but the poison-expelling charms on the
reverse are in prose ( K T U 1 . 1 0 7 . 3 2 - 4 5 ) . T h e incantation text K T U
1 . 8 2 = R S 1 5 . 1 3 4 , so far as it is readable and understandable, is
also couched in prose. Its sections are marked by quick changes of
subject and by subordinate clauses in persuasive style, but no clear
parallelism can be traced:
\y\mhs.bcl [xxjy.tnn.wygl.wynsk.'d [x]
[x]xy.lars[.id\y.alt.I ly
[xx] tgwln.ntk
[xx]wptk.l mt.brtk
[wa]tm.prtl.l rih.hmt.tmt.
Let Ba'al smite the breed(?) of Tunan and reveal and pour out the . . .
(2). . . on the earth. Then I shall not feel the curse, then the curse
will not for me be (3) harmfu1(?). The archer Reshef (is) between the
two of you. He will shoot at his kidneys and his heart. Let your
mouth . . . in the lowland, let your mouth resound in the woods, when
you grind(?) your teeth (5) [against him]. And your lips will surely
rejoice, if you keep until death your covenant. (6) [I shall
] myself,
I shall ring out with the Sun's voice: the life of the serpents I shall
take away, Ba'al, and nail down the iron pole on his head
(KTU 1.82.1-7)
Likewise the minor myths of Yarikh and Nikkal ( K T U 1.24 - R S
5 . 1 9 4 ) , Astarte the Huntress ( K T U 1 . 9 2 = R S 1 9 . 0 3 9 ) a n d El's
drunkeness ( K T U 1 . 1 1 4 = R S 2 4 . 2 5 8 ) show a mixed style of prose
and poetry. Of the last text the medical prescription is entirely in
prose ( K T U 1 . 1 1 4 . 2 9 - 3 1 ) . T h e first text has a narrative framework
in prose, whereas some of the speeches are couched in verse. This
may also be true of the fable K T U 1 . 9 3 = R S 1 9 . 0 5 4 ( D I J K S T R A


(1) arh td.rgm.b gr (2) b py.t'lgt.b 1n[y] (3) gr.tyb.b npy.rg[m] (4) hzvt.b'l.itm['.y
gr'] (5) M ly.y p. i[k(?)] (6) hkr(.b]ry[ } . . .

The Cow let out a cry against the Mountain: 'In my mouth is stammering, on [my] tongue is agitation, in my throat rolls "thun[der]".
Heed the word of Ba'al [O Mountain!] Listen to me, you braggart!
Why this distress of my [bo]dy . . . ? . . . '
K T U 1.92 and 1.114 contain prosodie phrases and epic formulae
borrowed from the major myths ( D i j k s t r a 1 9 9 4 , 1 1 6 ) , but the narratives are basically prose compositions despite this poetic flavour.
For K T U 1.92, one has to assume that many verses have been shortened to monocola, if it was a piece of narrative poetry. In K T U
1 . 1 1 4 , the sequences of perfect (with inversion of the subject) and
imperfect forms suggest the transformation of poetically structured
verses into prose sentences (so also K T U 1.96 = R S 22.225.Iff.).
il dbh . . . sh . . . tlhmn.ilm.wtstn . . . After El slaughtered . . . called . . .
the gods ate and drank . . .
After he rebuked El his father,
b il.abh.gcr. ytb il.. .
El sat down . . .
El sat down at his
il.ytb.b mrzhh yt. .
marzeah, he drank . . .
After El wanted to go home,
il.hlk I bth . . .y'msn.nn . . .
they carried him . . .
A similar mixed style is also present in the ritual K T U 1.161 - R S
34.126, the incantation K T U 1.169 = R I H 7 8 / 2 0 and the prayers
K T U 1.65 = R S 4.474, 1.108 = R S 24.252 and perhaps 1.123 =
R S 24.271. W e observe in these texts a certain repetitive style, even
chains of adverbial clauses and comparisons: k qtr.urbtm.k btn. cmdm
(3) kyim.zrh.k Ibim.skh, 'like smoke from a chimney; like a snake from
a pillar; like a mountain-goat to the hill-top; like a lion to a lair'
( K T U 1.169.3-4); il h il add (10) bcd spn bcd[ (11) ugrt (12) b mrh il
(13) b nit il (14) bsmd il (15) b dtn il, etc., El, hurry! El, stand
up, on behalf of Saphon, on behalf of Ugarit, with the javelin of
El, with the spade(?) of El, with the span of El, with the threshings1edge(?) of El, etc.' ( K T U 1.65.9-15); b mrmt (8) b miyt.bzlm.b qd,
'. . . on the heights, in the lowland, in darkness and in the sanctuary' ( K T U 1.169.7-8); b cz (22) [rpi.] mlk.'lm. b dmrh.bI (23) [anh].bhtkh.b
nmrth . . ., 'in the safety [of the Healer], the eternal king, in his protection, in his strength, in his dominion, in his benevolent power . . .'

( K T U 1.108.21-3; also 1.108.4-5). C o m p a r e also the chain of adverbial p h r a s e s a n d //^-greetings K T U 1 . 1 6 1 . 2 2 - 6 (with tht), 3 1 - 4
and the chain of epithets in K T U 1.100.1: um phi bt.' . ., ' T h e m o t h e r of the he-ass, the she-ass, daughter of
the spring, daughter of the stone, daughter of heaven and flood . . .'.
Similarly, the chains of epithets and participial predicates in the liturgical prayer K T U 1.108.Iff. T h e y are all instances of poetic prose
with repetition and even occasional parallelism within a prose context. Despite these poetic elements, such passages as K T U 1.65.9ff.
and K T U 1.108.Iff. form one extended prose sentence, bursting the
bounds of an originally poetic structure. Even the small fragment
K T U 1.83 = R S 16.266, p e r h a p s part of an incantation (DE M O O R
1987, 181-2), shows this mixed style:
. . . [ts]un. b ars (4) mhnm. trp ym (5) 1nm.tlhk (6) mm. ttrp (7) ym dnbtm.
(8) tnr. I btn (9) ist. trks (10) I miym Ibnm* (11) pi. tbtn. yymm* (12) hmlt.

ht. ynh*[r] (13) itph. mk* [ ] (14) thmr. [ ] . . . (text newly collated by



. . . [She we]nt out into the land of Mahanayim to vanquish(?) Yam

with the forked tongue that licks the sky. She vanquished Yam with
the forked tail, Tunan she muzzled. She bound him onto the heights
of Lebanon (saying): You swill no longer humiliate me, Yam. Mankind
may be scared, River (but) you will not see [me] collapse. You may
foam [. . .]
Related to the persuasive prose style of incantation is the prose of
letters and an occasional wisdom text such as K T U 2.2 = R S 3.334
(II 2.5? = R S 1.020), perhaps dictated as a scribal exercise:

[/] rHSyy[al]
\s\gr.l lmt.l[m]
b*th.p lmt.p* lm

(5) b*t.lbn.trgm*[?]
(6) / stmt.I lm.b[tk]
(7) by.nt.mlit.t[mla]

[w]h*t.msgr.bnk[ ]

(12) [wh]n.thmM[ ]

First of all, my friend, one should a[sk] for

[p]eace. A man should keep the bolt(?)
secure. (If) it is not safe, will its house be
at peace? When it is safe, then the house
will have
peace. To a man you should say:
'It is not safe, so [your] house will not be
at peace!'
Believe me, after a full year has surely
children you asked for will come to you.
Sons and daughters will Ba'al give you.
My son, spoil (it) and . . .
you will have [spoi]1ed the bolt(?) of your
[And 1o]ok, the word of Ba'al. . .


The prose of letters

Letters form the most interesting group in which to study the syntax of Ugaritic prose in the latter days of its floruit. W e shall not
deal here with the well known formulae of sender and addressee
(thm X rgm l Y), the airs and graccs (yslm Ik, ilm tgrk tlmk, with its
variants), the prostration formula (lpen PN[bCd bCd/tnid] mrhqtm qlt/qlny),
enquiries about health (al lm formula), requests for an answer (wrgm
[t]ttb l/cm-PN), to reply or to pay attention to the message (wbcly. . .ydc
[rgmh]), or even the closing remark to reassure somebody (wap mhkm
b Ibk al tst, etc.). These have all been thoroughly and properly studied
in the past ( A h l 1 9 7 3 ; K r i s t e n s e n 1 9 7 7 ; P a r d e e 1 9 8 4 ; P a r d e e
1987; C u n c h i l l o s 1983a, 1989a). Here attention will be
paid to the larger passages of prose only in so far as they are preserved and help to give a glimpse of late colloquial Ugaritic.
Characteristic of this Ugaritic 12th-century written vernacular is
the increased use of plene writing with -y in prepositions by, ly, ky,
iky, construct state: ily ugrt and verbal forms tmgyy, etc., but also an
increase in enclitic - j as a marker of direct speech, in particular in
letters ( T r o p p e r 1994d, 474-5). T h e difference between plene written -y and enclitic -y is not always easy to detect ( T r o p p e r 1994d,
480-1). O t h e r changes in vocalization and phonology include the
quiescent aleph, sbcd < bC,id, yr < yar, etc., though also incidentally
found in poetry, cbdnn < acbdnn; cdbk < a'dbk ( K T U 1.6 = RS 2. [009] +
ii 21; 1.18 = R S 3.340 iv 22);ytmr < yitmr ( K T U 1.3 = RS 2. [014] +
i 22), the shift of ' < h, for instance in im < hm ( K T U 2.15 = RS
15.007.8; 2.72 = RS 34.124.9, 10, 17; 3.9 = RS [Varia 14].6; T r o p p e r
1989b, 421-3); vowel harmony ulp (*ullupi) < alp (*allupi); ihy and
uhy < ahy, ibr (Hbbiru) < abr (*abbiru). See further below 8.1.
T h e g r a m m a r and syntax of this late Ugaritic prose are enriched
by the use of the article and demonstrative element hn, rarely independent in hn Ws, K T U 1.40 = R S 1.002 (perhaps also 1.114 =
RS 24.258.28 and in assimilated form in K T U 2.70 = RS 29.093.15-6,
w.hwt (16) hbt, 'and I repaired the house'), but frequently as part of
a set of demonstrative pronouns: masc. hndhnk; fem. hndthnkt pi.
hnhmt. Morphological developments may include the loss of diptotic
plural and the occurrence of imperatives with prothetic aleph i: ibky,
ihn ( K T U 1.161 = RS 34.126; T r o p p e r 1993a, 391-2); igr ( K T U
2.33 = RS 16.402.1); or aleph a: add ( K T U 1.65 = RS 4.474.9).
Furthermore, we may note the wider use of the absolute infinitive

with separate personal p r o n o u n to continue a finite verb or imperative, wtbc ank ( K T U 2 . 1 7 = R S 1 5 . 0 9 8 . 6 ) ; w.ttb.ank ( K T U 2 . 3 8 =
R S 1 8 . 0 3 1 . 2 3 ) ; wrgm hw/ank ( K T U 2 . 4 2 = R S 1 8 . 1 1 3A. 1 9 , 2 5 ) ; hbt
hw (6) hrd hw (7) qrt, '. . . it eliminated the guard a n d pillaged
the c i t y . . . ' ( K T U 2 . 6 1 = R S 1 9 . 0 1 1 . 5 - 7 ) ; w.ybl.hw ( K T U 2 . 7 2 =
R S 3 4 . 1 2 4 . 2 7 a n d passim), though this is incidentally also attested in
poetry ( S E G E R T 1 9 8 4 , 6 4 . 4 2 ) .
T h e variation in the use of the conjunction p(m)- is significant,
a n d greater than in poetry (DE M O O R 1969, 2 0 1 - 2 ; K T U 2.2 = R S
3.334.4, 2.3 = R S 1.013+.19, 2.10 = R S 4.475.12, 2.14 = R S [Varia
4].12, 2.15 = R S 15.007.7, 2.23 = R S 16.078+.17, 2.26 = R S
16.264.7, 2.33 = R S 16.402.28, 2.70 = R S 29.093.27, 2.71 = R S
29.095.11 [pm], 2.72 = R S 34.124.1 1, 22, 42, 2.73 = R S 17.434.14;
W A T S O N 1990e, 1994e) a n d the occurrence of pi and pn 'lest, you
may not' ( K T U 1.83 = R S 16.266.11, 1.114 = R S 24.258.12). Also
new, specific verbs such as dhl, 'to be afraid' ( K T U 2.16 = R S
15.008.12, 2.31 = R S 16.394.21); hbt, 'to knock down, eliminate'
( K T U 2.4 = R S 1.018.19, 2.47 = R S 18.148.16, 2.61 = R S 19.011.5),
or verb forms like tn cm/l 'put something at the disposal of P N '
( K T U 2.36+ = R S 17.435+.6, 13, 2.45 = R S 18.140.19, 2.50 = R S
18.287.16, perhaps also 2.32 = R S 16.401.7, 10, 2.39 = R S 18.038.35,
2.79 = R I H 77/25.3), presumably a -stem ofy/ntn;
-stem 'hr 'to
withhold, keep back (things)' ( K T U 2.42 = R S 18.113A.11, 2.79 =
R I H 7 7 / 2 5 . 4 ) a n d the Gt-stem sal 'to make a request, e n q u i r e '
(2.17 = R S 15.098.15, 2.42 = R S 18.113A.23, 2.70 = R S 29.093.12,
2.71 = R S 29.095.10).
Further, we may note the continued use of bl in compounds like
blym, 'never' ( K T U 2.45 = R S 18.140.23; 4.272 = R S 17.118.7),
bl bns, 'nobody' ( K T U 2.45 = R S 18.140.25), || Ibl ks, 'without a knife, or cup ( K T U 1.96 = R S 22.225.4-5) a n d bl sml ( K T U
1.169 = R I H 78/20.7), though also used in poetry: bl spr/hg ( K T U
1.14 = R S 2. [003]+ ii 37-8); blmt, 'immortality' ( K T U 1.17 = R S
2. [004] vi 27, etc.), the increased use of the perfect or participle Nstem ntkp ( K T U 2.10 = R S 4.475.14); nUi (2.34 = R S 17.139.13);
nskh (2.38 = R S 18.031.15); nmkr ( K T U 2.48 = R S 18.285[A].5); nplt
(2.82 = R I H 7 8 / 1 2 . 4 , 11); also in administrative and legal texts: nkly
( K T U 4.213 = R S 16.127.24, 4.230 = R S 16.341.15; 4.280 = R S
17.236.6, etc.); night ( K T U 4.659 = R S 19.166.1) a n d nlm ( K T U
4.328 = R S 18.008), but also a new preposition like ml(y), 'opposite
(me)' ( K T U 2.50 = R S 18.287.12; 2.75 = R S 34.148.11). T h e r e is no

clear evidence that a special subjunctive m o o d was maintained in

prose texts ( T r o p p e r 1991b, 3 5 3 - 5 , pace V e r r e e t 1988). T h e deictic
or a n a p h o r i c use of -n, in p a r t i c u l a r in the apodosis of o m e n s
( D i e t r i c h - L o r e t z 1990a, 104; T r o p p e r 1994c, 466-7), but also
elsewhere ( K T U 1.124 = R S 24.272.14, 2.37 = R S 17.438.10, 2.39
= R S 18.038.21, 2.42 = R S 18.113a.6, 10, 26) may suggest influence
of the H u r r i a n 'article' -ni (but see T r o p p e r 1993b, 468).
It is inherent in the nature of letters that we find narrative parts
(reports using the perfect, e.g. K T U 2.38 = R S 18.031) and prescriptive sentences (instructions); but the f u n d a m e n t a l convention is
that of an oral message exchanged between parties. This means that
the m o d e of discourse in letters is usually persuasive, a m o d e of discourse couched in a kind of virtual verbal exchange between sender
and addressee, but from the temporal perspective of the writer. A
m o d e of discourse that varies with the field (diplomacy, international
commerce, royal bureaucracy) and tenor (grade of social relationships between parties, see also P a r d e e - B o r d r e u i l 1992, 711).
Letters are essentially a verbatim account of verbal exchanges between
parties w h o argue their case. As part of the discourse the words of
the other party are quite often quoted or referred to ( D i j k s t r a 1987).
This style of persuasion is marked by statements accentuated with
particles or adverbs such as ap/ p/ hn, emphatic use of the separate
personal pronouns, interrogative particles and pronouns, such as ik,
'how', Im, 'why', my/mn(m), 'who(ever)', mndc, 'whoever knows, perhaps', rhetorical questions and other turns of speech introduced by
ht, 'now', hm . . . p/w/zero,
' i f . . . then', etc. If we are aware of the
sequence of inferences and thrust for persuasion m a d e coherent by
a set of characteristic phrases and other cohesive devices in gramm a r and style, the letters are essentially a genre of persuasive discourse, even if stories are told, oaths are sworn and instructions are
given within such a persuasive m o d e of discourse. Unfortunately only
a few of the m o r e elaborate letters are complete or have a passage
that allows for coherent translation and rhetorical analysis. Any translation given below admits that other translations are possible in some
instances, but I a m concerned here only with the mode of discourse
and the general thrust of a given passage. An example of such a
dialogical discourse of persuasion including an oath is K T U 2.10 =
R S 4.475.5-15,
. . . trgds (6) w.l.klby (7) rrt.hti (8) (9) hm.inmm (10) nhtu.w.lak
'my.wyd (12) ilm.pkmtm (13) 'z.mid (14) hrn.nlkp (15) m'nk


. . . from Trgds and Klby I have heard that we suffered a heavy defeat.
However, if we did not suffer a defeat, send me a message and, as
for the hand of the gods, it will indeed be as strong as Death (against
you) if your answer be negative(?).
Letters often include reports. A good example is the letter from the
king of Tyre to the king of Ugarit about the shipwreck of a Ugaritian
fleet sent to Egypt, but there are also short reports such as K T U
2.17 = RS 15.098, 2.30 = RS 16.379, 2.33 = RS 16.402.4ff., 2.40 =
RS 18.040, 2.61 = RS 19.011 and 2.75 = RS 34.148. We often
find here as a typical report marker the formula 'the king my lord
may know it/his word!' ( K T U 2.17, 2.33, 2.40, 2.75; perhaps also
2.35 = RS 17.327), or advice not to fear or to worry too much
about the reported developments ( K T U 2.16 = RS 15.008, 2.30,
2.38 = RS 18.031 and 2.71 = RS 29.095).
(10) anykn.dt (11) likt.msrm (12) (12) (14) gm.adr (15)
nkh.wb.tmtt (17) lqh.kl.dr'\ 18) bdnhm.w.ank (19) (20)
(21) w.Ub.anUhm (22) (25) by.'/cy.'ryt (26) w.ahy.mhk (27)
This fleet of yours, which you sent to Egypt, was shipwrecked at Tyre.
It was hit by a heavy storm. And the master of shipwrecks took all
the cargo from their holds. However, I in turn took all their cargo,
all the livestock for their provision from the hand of the master of
shipwrecks and returned it to them. And your second fleet is in for
repair(?) at Acre, but my brother should worry about nothing.
(KTU 2.38.10-27).
In particular, the mode and tenor of the discourse in international
diplomatic letters becomes very persuasive, if not suggestive. In many
instances we are here perhaps dealing with drafts and translations
of official letters, of which the originals were sent and received by
the royal chancelleries in official Akkadian. They negotiate about
tribute, settle border disputes and deal with conflicting interests and
loyalties. Consider the following anthology: bt. mlk. thmk. hin [y] (6) \lj\rs.a rgmny [ .\'m. p. tn ,\ank\ (7) [w]at.m[h]r.k[x][mt] (8) \ht\.Iqdm.udh.mgt.wmlkn.[] (9) [m]hrt[.]nib.'mnkm.I.qrb.[xx]
(10) [xj.i[x][xY.] (11) [w.]ma[k]tk/my.l.likt
[x][xx] (13) | xxx] Jtnt.Cmy.'m.pJtn[t]
When you sent your message to the royal palace (saying), 'Herewith
I put the gold of my tribute at the disposal of the Sun', [as for] you,
the equivalent of the . . . as was settled in the trea[ty], you should present it now! Its payment(?) is due. And the king [said?], 'Tomorrow
we shall return to you to deliver the . . ., but you have not come to
me [. . . nor] did you send your embassy to me. [Now was this] as

arranged with me? So, the gold [of your tribute that was] put here
at my disposal, I shall put (it) at the disposal of the Sun.
(KTU 2.36+ = RS 17.435+.5-13)
Li[th]dn.p (6) ad[nk.'] bdk.ukJkn (7) k.'[bdm.]sglth.hw (8) w.b[nh].uk.ngr (9)
rg[mh.l]adny.l.yfysr (10) w.[ap.y]d'.l.ydet
(11) ht[.hm].l.p.belk (12) 'b[dm.] (13) ht[.hm].p.b'lk (14) yd'm.l.yd't
(15) 'myJps.b'lk (16) nt.ntm.lm.<.>l.tlk
Did not the Sun his father and his servant make an agreement(?) either
that he would record that he and his sons would be servants of his
own property, or that he would keep his promise (saying): 'My father
will not lack anything' and [also]: acknowledge you fully'.
Now, if to the Sun your lord, you are servant of his own property,
so then, if you recognize the Sun your lord fully, why did not you
come to the Sun your lord for one, for two years?
(KTU 2.39 = RS 18.038.5-16)
Likewise K T U 2.23 = RS 16.078+. T h e tenor of such letters is often
haughty, if not aloof on the part of the Great King, his queen and
his officials. T h e Great King speaks about himself in the third person, the greetings are curt and there is no love lost, whereas the
attitude of the vassal king is submissive and the airs and graces are
elaborate, if not exhaustive. More than half the letter from Ammittamru
to the Egyptian Pharaoh is filled with the repeated string of royal
tides:^wt msrm . . . , ' . . . the
Sun, the great king, the king of Egypt, the benevolent king, the righteous king, the king of kings, lord of the whole country of Egypt . . .'
( K T U 2.81 = R I H 7 8 / 0 3 + ; see somewhat less tediously K T U 2.23
= R S 16.078+, 2.76 = RS 34.356.1-2, 9 - 1 0 [a draft?]). All along,
the writer repeatedly praises his overlord as benevolent king, trying
to negotiate a lower tribute: [mtn.] (25) ['bdk.b]
nt.qdm.alpm.mznh (26) [ht. cbdk.] yir.snp.ln.dym.hw, '. . . Another matter,
my benevolent lord. Look, the silver which [your servant] has
paid for many years, two thousand (shekels) is its weight. (Now your
servant] asks, will two thirds be sufficient for us? . . .'. Another flower
of such submissive speech is K T U 2.23 = RS 16.078+. 15-24:[k.cbdk.]d (16) ar[.hym.lp] (17) mlk.r[b.b'l}y.p.l. (18)[h. ]/ (19)
/./w.//[/.] spn.b'ly (20) w. urk.ym. b'ly (21) (22) il.msrm. dt.tgm
(23) np.p [.]mlk.rbMy
. . . And I am [your servant] who begs [for life to] the Sun, the great
king, my lord. Then do I not pray for the life of his soul before Ba'al
Saphon my lord, and length of days for my lord before Amun and

before the gods of Egypt who protect the soul of the Sun, the Great
King, my lord?
In the exchange of messages between the king a n d officials, we sometimes find such elaborate phrases in addition to the usual formulae
of submission, ankn.rgmt.l.b'l.spn. (7) Ups.'lm.l.'ttrt (8) l.' (9)
nmiy.mlk.'lm, ' . . . I pray to Ba'al Saphon, to the eternal Sun, to Astarte,
to Anat, to all the gods of A1aia for the splendour of an everlasting kingship!' ( K T U 2.42 = R S 18.1 13a.4-9).
N o t only are developments reported, but problems are also discussed a n d instructions given in the same persuasive sort of style.
T h e U g a n d a n king reports violations of his territory by Egyptian caravans, a n d the Hittite q u e e n , probably in consultation with the Hittite
deputy-king of C a r c h e m i s h , instructs him to direct t h e m past Q a d e s h
through the valley of the O r o n t e s ( K T U 2.36+ = R S 17.435+ .16ffi;
D I J K S T R A 1989, 142-4). An interesting instance is the letter f r o m
G e n e r a l Iwri-tarruma ( K T U 2.33 = R S 16.402), reporting an attack
by the kings of Mugihe (Alalakh) a n d , p e r h a p s , Nuhai against
N i q m a d d u II. After some explanations a b o u t the course of strategy
taken, he c o m e s with a m a z i n g e l o q u e n c e to the subject of reinforcements:
w.mlk.b'ly (23) ImJkn.hnk (24) l.'bdh.alpm.sswm (25) rgmt.Hy.lh.lm (26)<b>'ly (27) (28) (29) riry.lh.l pn.ib
(30) (31) b' (32) alpm.wm. hnd (33) w.mlk.b'
(34) bnny.'mn. (35) mlakty.hnd (36) ylak.'my (37) w( (38) [a]lpm[.]umi
And the king my lord, why did he assign such a thing to his servant?
Two thousand horses, you said, would come soon! Why has the king,
my lord, not provided them yet? Look, the enemies are pressing me
hard, but I cannot put my womenfolk and children just in front of
the enemy! Now, if the king my lord orders it, they will arrive here,
those two thousand horses. And the king my lord may also send to
me mediators(?) with this my embassy. And let them come up soon
hither, the two thousand horses [and] let him not [. . .] and withdraw!
(KTU 2.33 = RS 16.402, 22-39)
Fragments of such eloquent pieces of prose, in which someone is
pleading his case, are also found, for instance, in K T U 2.41 = R S
18.075, 2.42 = R S 18.113A, 2.45 = R S 18.140, etc., but unfortunately they are too broken for their lines of reasoning to be followed
in detail. In K T U 2.70 = R S 29.093, we find a complaint a n d a
request. Obviously one of the senders of this letter is a w o m a n (as
also K T U 2.11 = R S 8.315):'yn ( 12) ystal.'m.amtk (13) lak.lh.wkhdnn (14) ( 15) Iqht.w.hwt

(16) hbt.wlm.tb (17) bn.'yn (18) w.lqh.tqlm (19)
tn.'bdk (21) tmt.'mnk (22) k I On.akl.Uun (23) w.k tal (24) bt.'bdk (25) w.kymgy
(26) ebdk.l ihn (27) 'mk.p.l.ysb' I (28) hpn.l b'ly (29) 'bdk

Here, the son of came to request (silver) from your handmaid. It

was sent to him, but he hid it. I enlisted a contractor and I repaired
the house. Why has the son of returned to take the two shekels
of silver from the hand of your handmaid?
And two of your servants are there with you in order that you may
give food to them. Now if your servants ask for accommodation and
if they arrive in peace to you, will not then a handful satisfy them?
T o my lord belongs everything that your servants own.
(KTU 2.70 = RS 29.093.11-29)
Letters evidently often react to messages received about information,
allegations of disloyalty, requests for help or neglect to pay outstanding debts a n d tardiness in fulfilling obligations. T h e r e is nothing new
u n d e r the sun! M a n y of the letters refer to messages received and
even quote from them (examples in D i j k s t r a 1987a, 3 8 - 9 to which
K T U 2.36+ = R S 17.435+ passim, m a y be added).
lm.tlikn.hpt.hndn (11) p.mrrt.mlk (12)* (13) (14)
msm't.mlk (15) tnm (16) 'my.wttbm.lby (17) (18) ky.tdbr.umy
(19) (20) (21) msqt.yt_bt (22) (23) likt.ank.lht (24)
bt.mlk.amr (25)ybnn.hlk (26) 'm mlk.amr (26) (28) hrs.w. mrdt<t>.l
(29) mlk.amr.w.lqh.hw (30) mn.b.qrnh (31) w.ysq.hw.l.ri (32) bt.mlk.amr (33)[at.brt]

Why did they release these two, as if they were not subordinate to
the king? Either being the son of Qln, or the Son of Alyy, or a subordinate of the king, these two should have come to me together and
should have broken my heart. As for the letter about the daughter of
the king of Amurruwhen my mother speaks on behalf of the city:
'If now the city will not go on to live in anxiety, who then must I
send with the letter about the daughter of the king of Amurru?' Let
Yabninu go to the king of Amurru and let him bring a hundred
(shekels) of gold and the repudiated woman(?) to the king of Amurru
and let him take oil in his horn and pour it on the head of the daughter of the king of Amurru. Whatever sin [she] committed [she will be
free?] . . .
(KTU 2.72 = RS 34.124.10-33)
It is characteristic of this m o d e of prose to construct complicated
sentences in an elaborate rhetorical, often conditionally phrased style.
For instance, the passage quoted from K T U 2.39 = R S 18.038.1 1 - 6
is probably one long sentence. Sentences introduced with conditional

hm, temporal k(y)- (preceded or followed by the conjunction p(m)-,

but often also simply the copula w-, K T U 2.10 = R S 4 . 4 7 5 . 9 - 1 0 ,
2.31 = R S 16.394.16ff., etc. or without connection, K T U 2.33 =
R S 16.402.30-1) are numerous, as are statements a n d conditional
sentences emphasized with hn, ht and ap (e.g. ht hm . . . K TU 2.10.8-9,
2.33.30, 2.39.1 I f f ; wap ht... 2.3 = R S 1.013+.20; wap.ank.. . 2 . 1 1 =
R S 8.315.13, 2.33.15, 2.41 = R S 18.075.19), and rhetorical a n d real
questions with or without ik(y), mh(y), e.g. w.k tal bt cbdk w k ymgy
bdk l lm cmk pi ysbc Ihpn, 'And if your servants ask for a c c o m m o d a tion and if they reach you safely, would not a handful be sufficient?'
( K T U 2.70 = R S 29.093.23-8, also 2.23 = R S 16.078+. 17ff. and
perhaps, 2.39.5ff).
Quite often in the prose of letters and elsewhere (e.g the narrative K T U 1.114 = R S 24.258, the ritual K T U 1.112 = R S 24.256.6-7)
the preferred word order V S is changed to SV: wum tsmf} mad. . .,
'and my mother should rejoice g r e a d y . . . ' ( K T U 2.16 = R S 15.008.10); . . ., 'And look, the enemies are pressing m e hard . . .'
( K T U 2.33 = R S 16.402.27), perhaps for emphasis (other examples
T R O P P E R 1994c, 467-70). A related interesting p h e n o m e n o n in this
m o d e of prose is casus pendens or nominative absolute ( W A L T K E O ' C O N N O R 1990, 4.7), e.g. in the oath-sentence: wyd ilm p kmtm cz
mid hm ntkp m'nk, 'as for the h a n d of the gods, it will indeed be as
strong as Death, if your answer is negative(?)' ( K T U 2.10 = R S
4.475.11-5); note the incongruity of yd (f.) and cz (m.)5 but this seems
to be the idea (pace P A R D E E 1987; W A T S O N 1990, 8 1 - 2 , 1994, 495);
[w].b.ym.k.ybt.mlk (15) [t]ydr.w.ap.ank (16) [i]hd.lgr. amn, '[And] as for
today, if the king stays in [T]ydr, I for my part took hold of M o u n t
A m a n u s . . . ' ( K T U 2.33 = R S 16.402+. 15-6; for the geographical
n a m e ty[n\dr, see D I E T R I C H - L O R E T Z 1994, 65-7); wmlk bcly Im skn
hnk rbdh 'and as for the king my lord, why did he assign such a
thing to his servant?' ( K T U 2.33.22-4); hn.mrt d tt asu b Idtk, 'Look,
the patrimony which was (legally) settled, I shall release after you
have given birth' ( K T U 2.34 = R S 17.139.32-3); cmy p b'Ik snt sntm
Im Itlk, 'to the king your lord, why did you not come for one or
two years? ( K T U 2.39 = R S 18.038.15-6); wlht akl Iy likt cm p bclk
ky akl bhwtk. inn. pn tubd, 'As for the tablet about food, when you
sent to the Sun the message that there is no food in your country,
the Sun was indeed d i s h e a r t e n e d . . . ' ( K T U 2.39.17ff). T h e r e are
m a n y other examples, for instance, in administrative texts: mlb.trmnm
(6) (7) (8) ytn.lhm, '. . . As for the clothing of the

trrnnm-gods when it is old, then in the house of the king clothing

should be given to them' ( K T U 4.168 = RS 15.082.5-9).


Performative and prescriptive prose

In this group I would include the medical texts, omens, rituals and
contracts, though a distinction from other prose texts is not always
easy. For instance, the protocol of necromancy through the medium
Dtn, one of the royal deified ancestors, contains as a report an amalgam of narrative discourse, ritual and medical prescriptions ( D i e t r i c h
L o r e t z 1990a, 212, 216). T h e Ugaritic liturgical prayer K T U
1.108 = RS 24.252 is an interesting amalgam of performative ritual language and descriptive hymnic prose with occasional parallelism. Though the style of contracts is basically performative and
very formal, lym hnd RN mlk ugrt ytn bt/d. . ., ' O n this day, R N the
King of Ugarit gave the house, the field . . ( K T U 3.2 = RS 15.111;
3.5 = RS 16.382); lym hnd iwrkl pdy . . ., ' O n this day, Iwrikalli
redeemed P N . . . ' ( K T U 3.4 = RS 16.191+), we occasionally find
instances of persuasive style complete with metaphor: . . . l.yihd stqlm
(2) b (3) d brt.kmt. (4) br stqlm (5) b unt cd clm, 'No one shall
take Stqlm in corve-service. As the Sun who is free, so Stqlm is
free from corve-service for e v e r . . .' ( K T U 2.19 = RS 15.125).
T h e largest category in this section comprises rituals. They are
written with an exasperating concision ( P a r d e e - B o r d r e u i l 1 9 9 2 ,
709). Many of the ritual texts look like administrative texts: a list of
gods and the sacrifices administered to them. Some of them are even
simple onomastica of gods ( K T U 1 . 4 7 || 1 . 1 1 8 ; Akkadian RS 2 0 . 2 4 ,
N o u g a y r o l , 1 9 6 8 , 4 2 6 4 ) , perhaps a kind of canonical list of gods.
Excerpts from this list and others, are extended in the rituals by
sacrifices administered to them. This may happen by simple juxtaposition of name and sacrifice, e.g. b'l i, atrt s tkmn wnm 'nt , rp
/, etc. ( K T U 1 . 4 1 = RS 1 . 0 0 3 + . 15-6, with parallels), but also with
a dative I (wtn sm lbclt bhtm, csrm lins ihn, K T U 1 . 4 1 . 5 , see also 1 . 8 1 =
RS 1 5 . 1 3 0 ; in H u m a n texts dative -d/-da, plural -tt/-asta). These
lists can be preceded by date formulae of months and days: byrh ris
yn. bymhdt. . . btltt esrt, 'In the month Risyn (First Wine) on the day
of the new moon . . . on the 13th day . . .' Occasionally, and often
interspersed between this lists of sacrifices, we find references to
processions and other cultic rites to be performed. A basic question
remains as to whether these rituals were meant to be 'prescriptive'

or 'descriptive' ( L e v i n e 1963; P a r d e e - B o r d r e u i l 1992, 709). M y

own preference is to see the references to ritual activities as habitual. This is consonant with the frequent use of imperfects indicating an incomplete action, if not an action to be performed, i.e. a
jussive mode, or a prescriptive imperfect ( P a r d e e - B o r d r e u i l 1992).
This prescriptive nature of the rituals is also consonant with the fact
that some rituals have complete or partial duplicates. T h e ritual prescriptions usually mention or imply the king and the priest as officiants:
btltt 'srt. yrtfis. mlk.brr, 'on the 13th day the king shall wash himself
clean' ( K T U 1.41 = R S 1.003+.3 a n d passim), wynt qrt/db I cnt walp
w l il wb urbt y tk gdlt ilhm, ' a n d he shall p r e p a r e a city-pigeon before
Anat, a cow a n d a sheep for El a n d in the chimney he shall p o u r
out (the blood o f ) the cow for the ancestor-gods(?)' ( K T U 1.41.10
a n d passim), wtlhm. att. . . kl lylhm bh . . . , ' . . . a w o m a n m a y eat (from
i t ) . . . nobody should eat f r o m i t . . . ( K T U 1.115 = R S 24.260.8,
10).;. . .yt.rpu.mlk.Clm. wyt (2) [/] gtr.wyqr. . wtst.'nt.gtr. . ., '. . . let the
Healer, the eternal king, drink a n d let [the god] G a r u - a n d - Y a q a r u
d r i n k . . . a n d let A n a t of G a r u d r i n k . . .' ( K T U 1.108 = R S
24.252.1-2, 6). This prescriptive prose style is particularly clear in
the ritual where the king a n d the officiant priests celebrate together:
id[.yd]bh.mlk (51) ar[b]'.arb'.mtbt.azmr.bhJJr[p]
(53) yr[gm.]mlk.sbu.p.w.hl.mlk
w.l[b][y pn]h.t[t]tbn (55) b.b[t] y [u.l.]smm. yd[h]
Then the king shall sacrifice to PRGL SRQN on the roof, on which
there are four by four dwellings of foliage: a sheep as a burnt offering
and a cow and a sheep as a peace-offering. Seven times the king shall
say whole-heartedly (the prayer): 'Host of the Sun and army of Maliku'.
Clothe him with covers(?) and wipe his face, and let him return to
the temple; and when he is present, he shall raise his hands to heaven.
(KTU 1.41 = RS 1.003+.50-5)
. . . [ s]l[m.u]hiy ylb (23) mlk.ylk.lqli.ilm (24) atr.ilm.ylk.p'nm (25) mlk.p'nm.yl[k.]
(26) b' pamt.lklhm
. . . [the] sta[tue of U]hari one shall dress. The king shall proceed to
accompany the gods, people shall go behind the gods barefoot. The
king shall also go barefoot, seven times for both(?) of them.
(KTU 1.43 = RS 1.005.22-7)
A n o t h e r instance of performative style in ritual, but in the second
person plural, is the sacrificial ritual of the ancestor cult ( K T U 1.161 =
R S 34.126). T h e perfective forms have been u n d e r s t o o d as prescriptive narrative (as also in the H e b r e w Bible Lev. 8 - 9 ; d e M o o r

1976, 335; H E A L E Y 1978b, 85), but

performative or precative perfects:

would suggest that they are

spr dbh (2) qritm.rp.a[rs] (3) qbitm.qbs.d[dn] (4) qra.ulkn.rp[a] . . . tqdm

sr (31) lm.lm 'mr[pi] (32) (33) lm.bth.lm.ugrt (34) lm.lgrh
Book of the Sacrifice of the 'Shades': Y o u shall n o w invoke the Healers
of the E a r t h , you shall n o w s u m m o n the Assembly of D D N : 'Invited
be U L K N , the H e a l e r . . .' Y o u shall a p p r o a c h the festive assembly by
saying the peace: Peace to A m m u r a p i ; p e a c e to his sons; p e a c e to
Tryl\ p e a c e to his house; p e a c e to Ugarit; p e a c e to its gates.

Performative imperatives and jussive imperfects 2nd person sing, and

plur. also occur in rituals: b tC Crh (2) trbd.'rs []pd-{?>)-iy.b t.mlk.. . (25)
I pn ll.trr (26) V/. . ., ' O n the 19th day, you shall prepare the nuptial bed of Pidriya in the suite of the king. . . before the night, you
shall shake up the b e d . . . ' ( K T U 1.132 = RS 24.291.1-3, 25-6).
Note also the poetic prose incipit of the prayer in K T U 1.119 =
RS 24.266: k gr 'z.tgrkm.qrd (27) hmytkm.''l.tsun, 'When a strong
one attacks your gates, a warrior your walls, you shall raise your
eyes to Ba'al (saying:)...' (R.P. G O R D O N 1991, 161-3). Such performative style is characteristic of Ugaritic rituals, just as for ancient
Near Eastern rituals in general. A last good example is the expiatory ritual K T U 1.40 = RS 1.002 found in several mutilated copies:
A n d present a y o u n g he-ass to obtain the re1ease(?) of the Ugaritians
a n d the expiation of the sojourners within the walls of Ugarit, the
expiation of Tman, the expiation of crmt, the expiation of Ugarit a n d
the expiation of N i q m a d d u . W h e t h e r y o u r faithfulness d e p a r t e d f r o m
the Q a t i a n clans, the D a d m i a n clans, the H u m a n clans, the Hattic
clans, the A1aian clans, the clans of Gbr, the clans w h o r o b b e d you,
the clans of y o u r faithfu1(?), the clans of Qrzblwhether
y o u r faithfulness d e p a r t e d either because of y o u r a n g e r or because of y o u r i m p a tience, or because of the quarrels you h a d , w h e t h e r your faithfulness
d e p a r t e d for sacrifices a n d oblation, o u r sacrifice we should sacrifice.
T h i s is the oblation we oblate, this the offering we offer. Let it rise
to the F a t h e r of the gods, let it rise to the family of the gods, to
T u k m a n a n d u n a m , this he-ass.
( K T U 1.40.26-34)

T h e omen texts show a fixed pattern of prose sentences and syntax,

like their Mesopotamian counterparts. Obviously, this type of literature derived from Babylonian tradition. Fragments of a dreambook
( K T U 1.86 RS 18.041) and a collection of astronomical omens
( K T U 1.163 = R I H 78/14) have been preserved. T h e birth omens
are represented by two main groups: the summa izbu 'If an abortion'

( K T U 1.103+ = RS 24.247+) and the umma sinnitu (.. .) ulid 'If a

woman gives birth to a . . .' omens ( K T U 1.140 = RS 24.302). T h e
former is the better preserved collection and shows the well-known
protasis-apodosis k tld X . . . Y(-r) structure. Characteristic of omens
is the subject (+) verb order in the apodosis ( T r o p p e r 1 9 9 4 c ,
O m e n s of small livestock. [If] a ewe(?) bears a stone, then the m a j o r ity in the land will fall victim; (if) a snake follows after it, the y o u n g
of its cattle will be weak, (if) also a . . .; famine will be in the land,
(if) it has no . . ., the country wall be destroyed; a n d (if) [its belly] is
open, a famine will be in the land . . .

Certain omens were checked by a second opinion of the haruspex

( K T U 1.78 = RS 12.061), but inspection of the omina also drew
forth ritual activity to eliminate the effects of bad omens. T h e lung
model K T U 1.127 = RS 24.277, for instance, reveals an interesting instance of relationship between omen interpretation and ritual.
It may have been an instruction model, but this is far from certain,
since parallels are still lacking ( M e y e r in D i e t r i c h - L o r e t z 1990a,
270-1). Nevertheless certain parts arc marked by 'borders' containing small texts with ritual instructions, seemingly derived from omina
observations. T h e most interesting instance is the instruction of a
scape-goat rite to eliminate the danger of a city taken or a plague
L o r e t z 1990a, 32-38, 270-1):
hm qrt mt y'l bn (31) bt bn bus yqh

z (32) wyhdy mrhqm

If a city is besieged (and) if plague attacks a m a n , the citizen's household will take a goat a n d banish it to the remotest parts.

(KTU 1.127.30-2)
However not many other examples of such characteristic prose can
be gleaned from the rituals. T h e style is often extremely concise and
many phrases are still poorly understood, for instance Crb.p. whl.mlk
( K T U 1.87 = RS 18.056.56-7; 1.46+ = RS 1.009.9-10, 44, 1.112 =
RS 24.256.9, 1.119 = RS 24.266.4, 2 3 - 4 , 1.126 = R S 24.276.23?,
1.132 = RS 24.291.27-8), variant sbu p (whl ym crb p) whl mlk
( K T U 1.41 = RS 1.003+.47-8, 5 3 - 4 , 1.112.14-5), also ttb rgm (bgn)
whl mlk ( K T U 1.106 = RS 24.250+.23-4, 33). T h e context suggests
a kind of morning or evening prayer spoken by the king.









Almost from the very beginning of Ugaritic studies, account was

taken of the verse component, 1 but as yet there has been no fullscale description of Ugaritic poetry nor has there been an exhaustive examination of the principles involved, although several partial
surveys are available. 2 T h e best and fullest account so far is P a r k e r
(1989, 7-98) 3 while a very detailed analysis of parallelism including
phonological features in only one short passage ( K T U 1.3 = RS
2.[014]+ i 2~25) has also been completed ( P a r d e e 1988c, 1-67). In
view of this state of affairs, the presentation here is not systematic;
instead, it sets out a number of related topics under several headings. First, though, some account is required of the nature of the
material under consideration and the problems it entails.

The texts

T h e corpus of Ugaritic poetry is fairly easy to define: it includes all

the mythological and epic texts, which are mostly in narrative verse,
and excludes letters, legal and economic texts as well as most of the
ritual texts as non-poetic. 4 Some texts, however, are borderline, e.g.
K T U 1.119 = RS 24.266:28-36); K T U 1.161 = RS 34.126 ( P a r d e e
1993a) and some verse texts contain prose elements ( C r o s s 1974)
e.g. K T U 1.113 = R S 24.247:1-11 = verse; 12ff. = prose list. Some



1 9 3 6 (19412); GASTER






as well as C O P P E N S 1 9 4 6 , 1 9 4 4 , Y O U N G 1 9 4 8 , 1 9 4 9 , etc., though inevitably

the relationship to Hebrew poetry was to the fore.
T / T 1 3 , K O S M A L A 1 9 6 6 , 1 7 2 - 6 ; G R A Y 1 9 6 5 passim, S E G E R T 1 9 7 9 ; 1 9 8 3 ; 1 9 8 4 ,

DEI. O I . M O LF.TF. 1 9 8 1 ,






For a critique cf. DEL O I . M O L E T E 1990, esp. 190-4.

'Features known as parallelistic may also occur in prose texts. Some Ugaritic
epistolary formulae exhibit parallelism, and even some of the expressions in the
body of the letters are arranged in parallelistic structures. This criterion alone cannot determine whether the text is poetry or prose. The other distinguishing feature
of a poetic text is its division into prosodie units of approximately the same length'



speech introductions are extra-colonic but others comprise an integral

part of the verse (see T h u s , the corpus includes the
Baal Cycle (KTU 1.1-6 = RS 3.361, 3.367, 3.346, 2. [014]+, 2. [008]+,
2.[022]+, 2.[009]+); the 'Stories' of Keret and Aqhat ( K T U 1.14-16
= R S 2.[003]+, 3.343+, 3.325+; K T U 1.17-9 = RS 2. [004], 3.340,
3.322+); the Rapi'uma texts ( K T U 1.20-1.22 = RS 3.348, 2.[019],
2.[024]); a wedding poem ( K T U 1.24 = R S 5.194); incantations
( K T U 1.82 = R S 15.134; K T U 1.96 = R S 22.225; K T U 1.100 =
RS 24.244; K T U 1.107 = R S 24.251+); a prayer ( K T U 1.119 =
RS 24.266:28-36; see d e l O l m o L e t e 1987, W a t s o n 1996); a mythic
marriage ritual ( K T U 1.23 = RS 2.002; cf. W a t s o n 1994a); a funeral
ritual ( K T U 1.161 = RS 34.126; see P a r d e e 1993a) and various
fragments. 5 A broad distinction can therefore be made between longer
texts (mosdy narrative) and relatively short texts (the remainder), with
K T U 1.100 = RS 24.244 occupying a mid-posidon. O n the whole
problem see 5.1.4.

Problems in studying Ugaritic verse

Apart from the poor condition of some tablets which

readings uncertain as well as leaving large gaps in the
and the fact that the corpus is relatively small, certain
culties combine to make the analysis of Ugaritic verse

makes many
poetic texts,
specific diffiproblematic.

Principally, for most of the tablets the stichometry is uncertain or

at least not made clear.6 Exceptions include K T U 1.10 = RS 3.362+
(and to some extent K T U 1.23 = R S 2.002), where the verse-line
corresponds to tablet line (cf. d e l O l m o L e t e 1991a, 463 and
n. 3, W a n s b r o u g h 1983 and W a t s o n 1982). Occasionally, stretches
of text are written with correct stichometry (e.g K T U 1.15 = RS
3.343+ iii 1-23, with 23 consecutive verse-lines which match the
lines on the tablet) and some of the worst sections are in Aqht.
At times, the vocabulary poses difficulties (e.g. gmn in K T U 1.6 =
RS 2. [009]+ i 19ff.). Generally, this is not particularly an obstacle
to determining poetic structure. However, when whole passages
which are repeated are not entirely understood, (e.g. K T U 1.3 =
RS 2. [014]+ iii 14-7 and par.) problems do arise.




1.8 = R S 3.364; K T U 1.83 = R S 16.266see P I T A R D 1998, DEL

1996; K T U 1.92 = R S 19.039+see DE M O O R 1985, D I J K S T R A 1994,
1989b; K T U 1.93 = R S 19.054see D I J K S T R A 1986.









T h e lack of vocalization and in general the absence of a tradition

of pronunciation (although syllabic spellings are of some help where
available) mean that we do not know how this poetry was recited
or sung and there is no indication of metre (see below), especially
as it is not known for certain where the stress lay.
T h e r e are several scribal mistakes; in addition, very often lines
appear to be omitted (as is apparent from comparison of nearparallel passages) 7 but it is not always clear when this was intentional (see 5.7 on expansion/contraction) and when not (see d e
M o o r 1978a, 130-1).
Almost all Ugaritic narrative is in verse, with no strictly comparable material in prose. Exceptions are the letters and to some
extent the ritual texts, but these have their own special styles and
to some extent Akkadian influence is evident in the letters, some
of which were translations (see, e.g. M a r q u e z R o w e 1 9 9 2 ) . It is,
for instance, difficult to describe 'normal' syntax and then compare it with the syntax of poetry, due to the lack of material ( S i v a n
1 9 9 7 , 2 1 0 ) (however see 5 . 2 . 3 . 3 and 5 . 2 . 3 . 5 below on verb
forms and ellipsis).


O n the positive side, some assistance is provided by the tablets and

several factors make study of Ugaritic verse easier. Quite often, as
has been noted, whole passages of verse are repeated. Sometimes
the correct stichometry is used and as has been said, the corpus is
reasonably well-defined. Enough of the texts is understood for a
degree of certainty in describing their poetic aspect and information
from similar traditions (Akkadian, Hebrew, Phoenician) can be of
some help. Also, text editions with translations now tend to include
comments on poetic structure. 8 M.S. Smith, who provides vocalization, semantic parallelism, word-count, syllable count, comments:
Some remarks bearing on various sorts of parallelismsyntactic,
morphological and sonant. . .often follow the presentation of cola,
although syntactic parallelism is not treated according to any specific
system. Rather, it has been my interest to indicate how these sorts of

See, e.g.





1991, P A R D E E 1988a, S M I T H 1994a, W Y A T T 1998c, etc.; see also

1980b (with the critique by DEL O I . M O L E T E 1983) and M A R G A L I T 1989a.

parallelism may bind and contrast cola, especially in the absence of

apparent semantic parallelism (M.S. S m i t h 1994a, xxxiv).9


It is now generally accepted that Ugaritic verse is not metrical. Young

discussed the possible existence of metre in Ugaritic verse and concluded: 'If there is any metric [i.e. metrical] system in Ugaritic, it
should show itself in some regular manifestation observable in the
texts themselves without our trying to fit any system into t h e m '
(YOUNG 1950, 124). T w o factors led him to this conclusion: (1) 'there
is no consistency in the sequence of similar stich combinations within
a poem or within sections of a poem, much less a consistency of an
accent-per-word pattern for the successive stichs themselves' (YOUNG
1950, 124) (2) 'if an accentual meter existed at Ugarit, it might be
seen in consistency between parallel passages within single poems, if
meter were important to the composer, singer, or reciter of the
poetry. N o such consistency is found' (YOUNG 1950, 128). He added:
poetry in which the outstanding feature is parallelism of thought;
a poetry written in a language in which the majority of words are
of one, two, or three syllables, and in a language in which almost
any clause can be couched in from two to four words, is a poetry
which naturally lends itself to the creation of lines of uniform metric length' (YOUNG 1950, 132). His conclusions have been accepted
by Pardee who states that since there is no 'regular, predictable or
at least observable recurrence' of rhythmic units in Ugaritic poetry
it has no metre (PARDEE 1981, 116) and PARKER (1989, 9-10) provides additional arguments. 1 0 It has been suggested by de M o o r that
Ugaritic verse is written in free rhythm to match its musical accompaniment. Such music was probably led by a soloist and would fit the
short stichoi of ancient West Semitic poetry. T h e fact that consecutive lines had a degree of regularity can be explained by the poets'
'pursuit of symmetry' (DE M O O R 1978a, 132). Ultimately, 'parallelism
was the primary structural principle of Ugaritic poetry and . . . length

T h e aspect of literary translations of these texts cannot be discussed here; cf.












1990 and



endorsed by Z E V I T 1 9 8 3 (but cf. P A R D E E 1 9 8 1 - 2 ,

argues that the word-divider had a metrical function, but

1995, 215;








of line was only prescriptive in the general principle of approximation' ( P a r d e e 1 9 8 1 , 1 2 6 ) .



In general, the language of Ugaritic verse is archaic ( P a r d e e 1 9 8 1 - 2 ,

267) and to some extent it also has its own vocabulary (see above). In
respect of verb forms, the rules followed are unlike those for prose.
T h e use of the qtl (or qatala) form in Ugaritic verse is distinctive:
'While *qatala constitutes the characteristic form for past narration
in the prose texts, its poetic usage is more restricted; the prefix forms
[yqtl, etc.] arc the norm for poetic narration' ( M . S . S m i t h 1 9 9 5 ,
7 8 9 , following F e n t o n 1 9 7 3 ) . " Besides being used for the stative,
reporting the past, continuing other perfects, the pluperfect, subordination and the performative, there arc three usages specific to
Ugaritic verse. These are 'contrast with prefix forms', 'report of action
c o m m a n d e d in the imperative' and 'delimitation of a section' (M.S.
S m i t h 1 9 9 5 , 7 9 0 , with further references). Different or identical verb
forms can occur in parallel lines (see 5 . 2 . 3 . 3 ) . Little research has
been carried out on syntax, but for the Story 12 of Keret, it has been
shown 'that word and sentence order remains completely unaffected
by the type of verbal form present' ( W i l s o n 1 9 8 2 , 3 1 ) . 1 3


Parallelism and the verse-line


T h e basic component of Ugaritic verse is the verse-line which can

be divided into two (parallel) half-lines or provided with a parallel
line to form a bicolon. It is generally accepted, then, that parallelism
is a fundamental component of Ugaritic verse, and it differs from
prose precisely because parallelism is so prevalent." T h e r e are several

However, cf.

According to G I B S O N 1975, Keret is a myth and Ai/liai a folktale. On literary
forms in Ugaritic cf. DEL O I . M O L E T E 1984b.
W I L S O N ( 1 9 8 2 , 3 1 ) concludes: 'Such interchangeabilitv of verb forms with no
appreciable effect on word order or sentence structure may well indicate a "frozen"
state for some poetic passages, in which word order is fixed and immovable. Such
passages could be inserted at any point of a narrative with only the necessary change
of verbal form in the new context'. See also SIVAN 1 9 9 7 , 2 1 0 - 4 .
'There is little disagreement that the most obvious and pervasive convention
of the Ugaritic poems is parallelism' ( P A R K E R 1989, 7, cf. 10).

types of parallelism, depending on meaning (i.e. semantic parallelism

which can be synonymousincluding numerical parallelism, antithetic or contrasting, alternating), syntax (grammatical; nominal and
verbal; chiastic) and the lines (or parts of lines) comprising parallelism can have various degrees of separation (standard or near, internal, and distant) and can be grouped into bicola, tricola, etc. These
types may or may not overlap.

Semantic parallelism

T h e r e are various sub-types of semantic parallelism.

Synonymous parallelism

This is the standard form of parallelism, 15 where line A and line

say virtually the same thing:
(A) They ply with gifts Lady Athirat of the Sea,
tmgnn rbt atrt ym
tg zyn qnyt ilm
(B) they implore the Progenitrix of the gods
(KTU 1.4 = RS 2. [008]+ iii 25-6)

Numerical parallelism

Since numbers have no synonyms, synonymous parallelism involving

numbers takes the form of || + 1 (where is an integer), as in
hm tn dbhm ska b'l
lit rkb crpt
dbh bit wdbh dnt
wdbh tdmm amht
(KTU 1.4 iii 17-21)

For two sacrifices Baal hates,

three, the Cloud-rider:
a sacrifice of shame and a sacrifice of prostitution
and a sacrifice of handmaidens' lechery

T h e r e are several other examples. 16

Antithetic parallelism

Although relatively rare, contrasting or antithetic parallelism does

occur, 17 e.g.

P A R K E R 1974, P A R D E E 1988b, S E G E R T 1984, 109.

See the bibliography in W A T S O N 1991b, 241, n. 2 and 242, n. 3, esp.
1973 and 1981 and L E E 1973.
W A T S O N 1986 = 1994b, 468-77.



t pt lars pt 1mm
set a lip to the earth, a(nother) lip to the sky
(KTU 1.23 = RS 2.002:61-2)18
and, in spite of the missing text:
[A1]oud did Anat laugh,
[g]/n tshq cnt
wblb tqny [. . .]
but in her heart she was hatching [a plot?]
(KTU 1.17 = RS 2. [004] vi 41-2)

Alternating parallelism

More problematic is the existence of what appear to be non-parallel bicola (SEGERT 1983, 300). Such non-parallel lines can be explained
in several ways. For example, a line such as hm it Ihm wtn wnlhm, 'If
there is food, then give so that we may eat' appears to be prose,
but in view of its counterpart, the lines are evidently in alternating
parallelism (indicated by A / A ' and B/B'):
hm [it 1 ]hm
(A) If there is food,
wtn wnlhm
(B) then give so that we may eat
hm it [yn]
(A') If there is wine,
\yi\tn. wnt
(B') then give, so that we may drink
(KTU 1.23 = RS 2.002:70-2). 19
Elsewhere the overriding pattern of parallel couplets tended to generate non-parallel couplets.

Grammatical and syntactical parallelism

A complete match in grammatical terms is evident in

lhm qy ilm
Give food, drink to the gods,
sad kbd hmt
wait on, honour them
(KTU 1.17 = RS 2. [004] 19-20).21
T h e verbs in the parallel
giving rise to types such
etc. 22 O n qtl in Ugaritic
Note also that a verbal


lines of a couplet can be in various 'tenses',

qtl || qtl, qtl || yqtl, yqtl || qtl and yqtl || yqtl
verse see above.
clause can be in parallel with a noun clause:




adduces K T U 1 . 1 9 = R S 3 . 3 2 2 + iv 4 6 5 0 .
V = 'verb'; = 'object'.
Parallel to K T U 1 . 1 7 = R S 2. [ 0 0 4 ] 2 9 - 3 0 , perhaps with hendiadys here.






1993, 2 0 4




1989 = 1994c, 240-9.








al trgn ybtltm
Do not delude me Virgin,
dm Igzr srgk hhm
for your delusion to a hero is sheer rubbish.
(KTU 1.17 = RS 2. [004]+ vi 34-5)
Perhaps chiastic parallelism belongs here, as in
sb ksp Irqm
turned had the silver into sheets
hrs nsb llbnt
the gold had turned into bricks
(KTU 1.4 = RS 2. [008]+ vi 34-5)
where the components of line 1 are switched around in line 2. 23 O n
chiasmus see W e l c h ( 1 9 7 4 ) . 2 4

Parallelism based on degree of separation

Internal or half-line parallelism occurs when only one line is involved,

for example:
rgm cs 1ht abn
Word of tree and whisper of stone
(KTU i . 3 = RS 2.[014]+ iii 22-3 and par.).
Here, the two halves of the line are in parallel. 2 ' 1 Standard or near
parallelism is none other than the couplet formed by adjacent lines
(see above). In distant parallelism, which serves to bind together longer
sections of text, there is a gap between the lines involved ( P a r d e e
1988c, 193-201, esp. 199-200). In general, the standard verse pattern comprises two sequential lines, as if parallelism were the accepted
norm, although it is not always present. Quite often couplets can be
formed from two (formulaic) monocola ( P a r k e r 1989, 23).

Ellipsis and ballast variant

Ellipsis (or gapping) is the absence of one or more elements (e.g. a

verb) from a line which would be expected but is or are understood
to be present. 26 For example, the verb tbl (+ suffix) is taken to be
present in the second line of
tblk grm mid ksp
May the mountains bring you plenty of silver,
tblk gb'm mhmd hrs
the hills bring you choicest gold27
(KTU 1.4 (= RS "2. [008]+) 31-3)



Note also 'gender-matched parallelism' (on which cf. W A T S O N 1981a).

See W A T S O N 1984b, 1985, 1988b = 1994b, 104 44. K O R P E L - DE M O O R 1998, 11.
See S I V A N 1997, 215-6.
The 'missing' element is indicated by (DE M O O R 1993, 200).


1 9 8 3 C , DE M O O R



A ballast variant is the use of a longer expression (usually in the seco n d line) for its corresponding a n d evidently shorter equivalent (usually in the first line) a n d is related to ellipsis. ' T h e principle of ellipsis
in poetry is the converse of (and goes h a n d in h a n d with) the principle of ballast variants.' (UT 13.105). For example:
wykn bnh bbt
Sired be a son for him in the house,
wykn 1rs bqrb hklh Sired be a scion within his palace
(KTU 1.17 = RS 2. [004] i 25-6 || 42-3)
H e r e , the ballast variant of bbt in the first colon is bqrb hklh in the
second colon, which compensates for the the ellipsis of wykn in the
second colon Similary, zbl b'l ars is the longer equivalent of aliyn bcl
(matching the ellipsis of widc) in:
wid' khy aliyn b'l
kit zbl b'l ars

Then I shall know that Mightiest Baal lives,

I shall know that the Prince, Lord of the Earth,
(KTU 1.6 iii 8-9; also K T U 1.17 i 36-7; K T U 1.18 iv 17-8).
Ellipsis in the first colon is rare, occurring almost exclusively in 'staircase parallelism' (see
ht ibk b'lm tmhs
Now, your enemy, Baal, do strike,
ht ibk tmhs
Now, your enemy strike etc.
(KTU 1.2 = RS 3.367 iv 8-9)
As A l t e r (1985, 23-6). has explained, such ellipsis frees space in
the second line for some elaboration of a parallel item in the first
line a n d also isolates the e x p a n d e d topic for attention. 2 8




Lines of verse can occur in relative isolation (monocola) or clustered

in sets varying f r o m two (couplets) to seven or more. T h e s e combinations of cola 29 are described here.


See also G R E E N S T E I N
single word in K T U 1 . 2 =

2. [ 0 2 4 ]
1.3 =



2. [ 0 0 3 ] +



iv 4 2 - 3 ;




1.16 =








and ellipsis of an expression in

2. [ 0 1 4 ] +




1.3 =




discuss ellipsis of a
iv 5 ; K T U












Although the standard strophic form is the couplet, single lines or

monocola occur very frequently. T h e y appear as introductory monocola drawing attention to speaker and in this form are prose, e.g.
rr Imtt hry 'Listen, Lady Hurriya' ( K T U 1.16 = R S 3.325+ vi
16-7). W h e n expanded to a bicolon or tricolon (see section on expansion/contraction), they are verse. Several are speech-openers.
It is sometimes difficult to determine whether a line of verse forms
part of a longer unit (bicolon, tricolon, etc.) or is entirely separate, e.g.

d Ihm sty ilm

While they ate the gods drank.
(KTU 1.4 = RS 2. [008]+ vi 55)
Is this line a monocolon or an introduction to the next two lines?
T h e function of the monocolon is to open and close sections of verse.
T h e y can also mark a climax, as in
sgrthn. abkrn
The youngest of them I shall make the firstborn
(KTU 1.15 = RS 3.343+ iii 16).
Some monocola exhibit inner parallelism 30 and so conform to the
prevailing feature of Ugaritic verse. An example is the standard formula
balp Id rbt kmn

by the t h o u s a n d iddu, the m y r i a d kumnu

(KTU 1.3 = RS 2. [014]+ iv 38 and many times elsewhere).


The tricolon

T r u e tricola, where all three lines are in parallelism (AAA"), are

rare, e.g.
I ys' alt tbtk
Surely he will
I yhpk ksa mlkk
Surely he will
I ytbr ht mtptk
Surely he will
( . 6 = RS 2. [009]+ vi 27-9

remove the prop of your seat,

upset the throne of your kingship,
smash the sceptre of your rule.
(and par.).

Some comprise a monocolon followed by a bicolon (ABB'):

She prepared a lamb in flour
tcdb imr bphd
hps ktr whss
for the appetite of Kothar-and-Hasis
1brlt hyn dhrydm
for the craving of Hayin, skilful with both hands.
(KTU 1.17 = RS 2. [004] 22-5)




1984b, 1985a, 1988b = 1994b, 104-44.

or, have the reverse formation (AA'B):

ydd wqlsn
He stood and insulted me,
yqm wywptn
He got up and spat on me,
btk phr bn ilm
in the gathering of El's sons
(KTU 1.4 = RS 2. [008]+ iii 12-4).
A special type of tricolon, called 'staircase parallelism', where the
initial line is interrupted by a vocative and then repeated in complete form, occurs several times, 31 e.g.
ir hym laqht gzr
Ask for life, Hero Aqhat,
irs hym watnk
Ask for life I and I shall give it to you,
blmt waslhk
for non-death, and I shall grant (it) to you.
(KTU 1.17 = RS 2. [004] vi 26-8)
O n e function of tricola is to slow the pace of the narrative which
then requires more attention ( P a r k e r 1 9 8 9 , 1 6 ) .

The quatrain or tetracolon

T h e r e are various forms of the four-line strophe, including

AAA" A' "
Her father set the beams of the scales,
her mother, the trays of the scales.
Her brothers settled the pointer,
her sisters, the stones of the scales.
(KTU 1.24 = RS 5.194:33-7)
adnh yt msb mznm
umh kp mznm
ihh yt'r mrrm
ahth labn mznm

tmh ht atrt wbnh
Let Athirat and her sons now rejoice,
ilt wsbrt aryh
the goddess and the throng of her kin,
kmt aliyn b'l
for dead is Mighty Baal,
khlq zbl b'l ars
expired has the Prince, the earthlord
(KTU 1.6 = RS 2. [009]+ i 39-43)
Some are formed when one couplet is inserted into another, as in
dnilysb palth
Daniel went round his blighted land,
bsql yph bpalt
he saw a stalk in the blighted land,
bsql yph byglm
he saw a stalk in the weeds,
bsql yhbq wynsq
he hugged and kissed the stalk.
(KTU 1.19 = RS 3.322+ ii 12-5).32











'Staircase parallelism' is more correctly analysed as apostrophe plus ep-

analepsis ( W A N S B R O U G H 1 9 8 2 ) .
As identified by D I E T R I C H


1973a (but cf. 1973b) and




The pentacolon

An example is:
bh p'nm ttt
On her, her feet quake,
behind her, her loins burst,
b dn ksl ttbr
'In pnh td'
above, her face perspires,
tgs pnt kslh
the joints of her loins quiver,
ans dt irh
the muscles of her back.
(KTU 1.4 = RS 2. [008]+ ii 16-20; || KTU 1.3 = RS 2. [014]+ iii
32ff. KTU 1.19 = RS 3.322+ ii 44ff.)
Again, much like the quatrain (see above) pentacola arise when a
tricolon is inserted into a bicolon, as in
Who sings and plays
dyr vuydmr
bknr wtlb
with lyre and flute,
btp wmsltm
with drum and cymbals,
bmrqdm dsn
with ivory castanets
bhbr ktr tbm
in the company of sweet(-voiced) Kothar.
(KTU 1.108 = RS 24.252:3-5)33

Longer sequences

Hexacola are relatively rare in Ugaritic (cf. L o r e t z 1 9 8 9 ) . T h e following comprises an introductory line and a five-line speech:
tm ydr krt t' itt
There Keret the votary vowed a gift:34
atrt srm wlilt sdynm 'Oh, Athirat of Tyre, and goddess of Sidon,
hm hiy bty iqh
if I take Hurriy to my house,
asr'b glmt hgry
and bring the damsel into my dwelling,
tnh kspm atn
twice her mass in silver shall I give,
wtltth hrsm
and three times her mass in gold!'
( K T U . 14 = RS 3.343+ iv 36-43)
O t h e r hexacola may be K T U 1.3 = RS 2.[014]+ iii 3 - 8 a ; K T U
1.3 iv 4 8 - 5 3 () 3 9 - 4 4 || K T U 1.4 = R S 2. [008]+ i 12-8; iv

1981a, 34 and DE M O O R 1978a, 137, n. 41. Other examples in W A T S O N 1997a,

30-5 and 1997b.
For another example cf. D I E T R I C H - L O R E T Z 1 9 8 2 .
The meaning of iitt (= i + lit?) is disputed (cf. DLU, 1 and 60); here I follow
W Y A T T 1998C, 2 0 0 - 1 ; see esp. nn. 115 and 117. Another possible translation is
'The gift of 'Atiratu of Tyre, the goddess of Sidon (is this):' ( P A R D E E 1997a, 336).
For yet another translation cf. M A R G A L I T 1997.

50-7); K T U 1.12 = R S 2. [012] ii 5 8 - 6 1 ; K T U 1.17 = R S 2.[004]

vi 43-5. 3 3 O t h e r sets are heptacola, K T U 1.6 = R S 2. [009]+ 11 - 9
and its near-parallel K T U 1.6 ii 31 -5,36 the eight-line sequence K T U
1.5 = R S 2.[022]+ i 14-22 and the nine-colon set K T U 1.107 = R S
24.251+ 3 8 - 4 4 .


Strophe and


A stanza is 'a fixed . . . or variable . . . grouping of lines that is organized into thematic, metrical, rhetorical, musical, or narrative sections' ( M e y e r s - S i m m s 1989, 288), though whether this definition
would be recognised by the poets of Ugarit is a moot point. T h e
only p o e m with an a p p a r e n t sequence of strophes or stanzas 37 is
K T U 1.100 = R S 24.244 which is divided into 14 sections by ruled
lines. O f these, sections 2 - 1 1 have the same n u m b e r of verse-lines
(i.e. 10) due principally to almost wholesale repetition. T h e first section ( K T U 1.100:17) has 14 lines because the initial couplet has
been expanded to a quatrain (see W a t s o n 1997a, 3 5 ) . 3 I ! T h e last two
sections differ completely from all the others. T h u s , although these
sections are actually marked off on the tablet, as P a r d e e (1978, 104)
comments: 'this is unquestionably o w i n g . . . to the extra-poetic structure of the text and the comparable length of the sections is owing
to the repetition within this structure' and 'any attempt to find strophes in Ugaritic poetry as a prosodie or poetic e l e m e n t . . . is d o o m e d
to failure'. 3 9 Analysis is limited to identifying shorter or longer sets
of verse-lines (couplets, tricola, quatrains, etc., as above), with no
regularity or predictability. 40 Even so, this remains a useful exercise,
and Ugaritic poetry can be segmented into sections based on content,
sometimes marked off by features such as certain particles (apnk, etc.)









See previous note. Cf. VVYATT 1998c, 1 3 5 n. 8 3 and 1 4 1 n. 108.

The terms 'strophe' and 'stanza' as denoting lines of verse grouped into sets
are used almost interchangeably (cf. C U D D O N 1 9 9 2 , 9 1 5 - 6 , 9 2 1 ) , although stanza is
more correct. On strophe, M E Y E R S - SIMMS 1 9 8 9 , 2 9 1 note: 'In contemporary usage,
the term usually refers to any stanzaic unit containing irregular lines'.
Each 'stanza' has 10 lines (= 1 + 9); the first has 14 (4 + 1 + 2 + 7).
See also P A R D E E 1993a, 208, n. 2 (ruled lines do not mark off'strophes').
A very interesting attempt at dividing the Story of Keret into (three) chapters,
each further subdivided into 6 cantos of 5 canticles or strophes, has been made by
S P R O N K 1 9 8 8 , although the incomplete form of the original text precludes cast iron
conclusions. See also LICHTENSTEIN 1 9 6 6 . On a smaller scale cf. D I E T R I C H - L O R E T Z
1 9 7 8 , HUSSER




a tricolon, peculiar syntax ( d e M o o r 1993, 197-200), speech introductions and the like, though these generally reinforce what has already
been indicated by the meaning of the passage concerned (see, for
example, the headings in d e l O l m o L e t e 1991a, 158-235, etc.).41



Repetition takes on various forms. Repetition of sound takes the form of

end-rhyme, alliteration, assonance and wordplay and is discussed
below ( 5.2.10). Repetition of words can occur at the beginning of
a line (anaphoric), at line end (cataphoric), as immediate repetition of
words (epizeuxis), and in the form of identical word pairs 42 (or repeated
over several cola, as keywords). Sometimes whole lines (or sets of lines)
are simply repeated, which in terms of strophe and stanza, results in
envelope figure and refrains. Lasdy, complete passages recur, sometimes
unchanged, sometimes in altered form (see 5.2.7; P a r k e r 1989,
26-52). A selection of these types of repetition is considered here. 43

Repetition of words

For example, repetition of one word at line-initial, as in

idk al ttn pnm
Then you shall set off
towards Mount Trgzz,
'm gr trgzz
m gr trmg
towards Mount Thrmg,
towards the two hillocks at the edge of the earth
'm tlm gsr ars
(KTU 1.4 = RS 2. [008]+ viii 1-4)
T h e function here is apparently to depict a long journey. Repetition
of a word consecutively occurs rarely and most examples come from
a single text ( K T U 1.23 = R S 2.002). O n e function may be to
d e m a n d attention or convey urgency:
J n k r n S r p
Hey! Watchman! Watchman! Open!
(KTU 1.23 = RS 2.002:69-70)
Similarly, y mt mt, O h , husband, husband!' ( K T U 1.23:40 and 46);
y ad ad and ad ad, 'Father, father!'; urn urn, 'Mother, mother!'. M o r e


For an attempt along these lines cf. S A U R E N - K E S T E M O N T 1 9 7 1 , although their

scheme was much too rigid.
See especially B O R N E M A N N 1 9 7 0 .
On repetition in Ugaritic see Z U R R O 1 9 8 7 and H E N S - P I A Z Z A 1 9 9 2 .

striking is the repetition of six consecutive lines beginning tld pgt. . .

'She shall bear a girl. . ( K T U 1.15 = RS 3.343+ iii 7-12). 44

Repetition of lines

ytlk llbnn ufsh

They went to Lebanon and its trees,
1ryn mhmd arzh
to Siryon (and) its choice cedars;
hn llbnn w'sh
Yes, Lebanon and its trees,
1yn mhmd arzh
to Siryon (and) its choice cedars.
(KTU 1.4 = RS 2. [008]+ vi 18-21)
Again, the purpose of the repetition, perhaps, is to depict a long

Repetition of passages

Ugaritic narrative poems are constructed using passages which are

repeated, sometimes verbatim, sometimes with slight variations. 43 Sets
that recur several times include the list setting out the six duties of
the model son ( K T U 1.17 = R S 2. [004] i 2 6 - 3 3 and par.; H u s s e r
1995); the 'gift-list' in the Keret Story ( K T U 1.14 = RS 2.[003] +
iii 2 2 - 5 and par.; H e n s - P i a z z a 1992) and the 'peace-offering ritual'
( K T U 1.3 = R S 2.[014]+ iii 14-7). Typically, a passage comprising
a ' c o m m a n d ' is then repeated for the 'performance' or fulfilment of
that c o m m a n d , e.g. K T U 1.14 ii 6iii 19 and K T U 1.14 iii 52-iv
31, or else an invitation ( K T U 1.22 = R S 2. [024] iii 2~4) and its
acceptance (lines 5 - 8 ; see d e l O l m o L e t e 1981a, 58-60) It is significant, perhaps, that segments of verse (couplets, etc.) could occur in
different sequences, for example, in K T U 1.14 4 0 - 2 (restored) ||
vi 10-2, the couplets

wng mlk Ibty

rhq krt Ihzry
al tsr udm rbt
wudm trrt
udm ytnt il
wusn ab adm


And depart, king, from my house,

go far, Keret, from my dwelling!
Do not besiege Greater Udum,
or Lesser Udum;
Udum is a present of El
and a gift of the father of Man

come in the sequences ABC ( K T U 1.14 iii 27-32), BCA ( K T U 1.14

40-5) and ACB ( K T U 1.14 vi 10-5) and in addition, the word

According to P A R D E E 1997a, 338, n. 51 this may be a list of the daughters'

names, now lost.
Cf. DEL O L M O L E T E 1991a, 58-62, P A R K E R 1989, 26-52.

pair ng II rhq is inverted the last two texts. Unless due to dictation
error, this indicates that verse was composed in formulaic passages,
perhaps orally.


Expansion and contraction

In Ugaritic the poets were free to expand single lines to bicola and
in turn form tricola from bicola. T h e process could also be reversed,
with longer strophes becoming shorter. It is certainly the case that
a line can be expanded to a couplet, as in:
rr rr latiyn b'l
Listen, please, Mightiest Baal!
(KTU 1.4 = RS 2. [008]+ vi 4)
which becomes
m' laliyn b' I
Listen, Mightiest Baal,
bn Irkb 'rpt
understand, Cloud-rider!
(KTU 1.4 59-60)46
A single line can also be extended to a tricolon. For instance, Pughatu's
core epithet is tkmt mym '(she who) shoulders water', and it can occur
alone ( K T U 1.19 = R S 3.322+ iv 28); it can also be expanded by
two further epithets: hspt Is'r tl, 'scooping dew from the barley' and
yd't hlk kbkbm, 'knowing the course of the stars' (cf. M a r g a l i t 1989a,
364-5). T h e prose formula wrgm Ikrt t' thm pbl mlk,47 'And say to Krt
the Noble, "Message of King P b l " ' ( K T U 1.14 = RS 2.[003]+
32-3), which derives from everyday usage, is expanded by applying
parallelism to each half to form two bicola:
wrgm Ibn ilm mt
And say to divine Mot,
tny lydd il gzr
Repeat to El's beloved, the hero
thm aliyn b'l
"Message of Mightiest Baal,
hwt aliy qrdm
The word of the Mightiest warrior".
(KTU 1.4 = RS 2. [014]+ viii 29-35)48
However, it is sometimes difficult to determine whether a poet has
intentionally added or omitted a line (or lines) or whether these are
accepted variants. For example, in


See L O E W E N S T A M M 1980a = 1992, 230-9,

Restored from similar formulae.
L O E W E N S T A M M 1980b, 256-61.


1978b, 1980.

sh hrn bbhtk
Call a caravan into your house,
dbt bqrb hklk
merchandise into the midst of your palace.
tblk grm mid ksp
The rocks will bring you much silver,
gb'm mhmd hrs
the hills attractive gold.
yblk udr ilqsm
let the quarries bring you choice gems.
(KTU 1.4 = RS 2. [008]+ 15-6 and KTU 1.4 38-40)
T h e last line is present in the two parallel passages but has been
omitted in K T U 1.4 = RS 2.[008]+ 31-3. Is this a mistake or a
deliberate contraction? Since it is of little significance, it was probably left out unintentionally. However, in the two club-naming passages, it is only when the c o m m a n d line
yprsh ym wyql lars
May Yam crumple and fall to earth
(KTU 1.2 = RS 3.367 iv 22-3)
is given to the weapon that it is effective and achieves the intended
result (line 26); this line does not appear either in the previous n a m ing or in the unsuccessful previous attempt, which indicates its omission there to be intentional. Each case has to be judged carefully
because the copyists 49 did occasionally leave out lines by mistake but
generally speaking the poets could expand or contract as they saw fit.


Word pairs

It has been noticed that there seems to be a large set of word pairs
which recur in Ugaritic verse and that many have equivalents in
other verse traditions of the ancient N e a r East. 50 T h e problem then
arises: Did the poets have a sort of 'dictionary of word pairs' on
which they drew to compose their verse, which was probably originally oral, 51 or were these pairs the side-effects of their use of parallelism and of (oral) formulaic language?'' 2 As is evident from the
following, the matter is unresolved.
In such word pairs, the - w o r d ' is usually c o m m o n e r than the
'B-word', e.g. klb || inr, 'dog' || 'puppy' ( K T U 1.16 = RS 3.325+
i 12; K T U 1.114 = RS 24.258:13) and any particular -word may








Cf. AVISHUR 1984, D A H O O D 1972, 1975, 1981, DEL O I . M O L E T E 1984a, W A I T E R S

1976. However, see the cautionary remarks of VAN DER L U G T
DE M O O R 1974.
See P A R D E E 1988a, 160, D I E T R I C H - L O R E T Z 1980b.
Another possibility is that they arose through word association (BERLIN 1 9 8 3 ) ,
but see below.

be paired with several different B-words (e.g. ib, 'enemy' || srt, 'adversary' or qm 'one who rises against' or nu, 'hater'). 5 3 Often a word
pair is related to a theme, e.g. ars || d, 'earth' || 'field', and is connected with fertility as in
n'm lars mtr b'l
Pleasant to the earth is Baal's rain,
wild mtr 'ly
and to the field the rain of the Most High.
(KTU 1.16 = RS 3.325+ iii 5-6, 7-8)
as also in K T U 1.3 = R S 2.[014]+ iii 16-7 and par., K T U
R S 2. [022]+ 18-9; K T U 1.6 = R S 2. [009]+ ii 16-7, 19-20.
word pairs are bound to a formula or formulaic expression
fewer that are non-formulaic and m a n y of these are repeated
as Aitken has shown for the Aqhat Narrative. H e comments:

1.5 =

This calls into question the notion of the word pair as a compositional
device, functioning independently of the formulas and formulaic expressions or of a thematic or/and formulaic context within the tradition.
While there are indications that a narrator could 'learn' a word pair
as a word pair in one context and use it quite independendy in another,
this was the exception rather than the rule. In the overwhelming majority of examples, the repository of word pairs is the formulas and themes
of the narrative tradition, and their appearance in the narrative is a
spontaneous reflex of the formulary and thematic habits of that tradition, and not of the production and manipulation of word pairs.
Neither spontaneous word association, nor the 'learning' and subsequent deployment of 'generally useful' word pairs has played a significant
part in the generation of recurrent word pairs within the narrative
( A i t k e n 1989b, 38).
Very rarely, word pairs are reversed, generally to denote some form
of reversal of events. C o m p a r e
al tt urbt bbhtm
You shall not install a window in the mansion,
hin bqrb hklm
an aperture within the palace.
(KTU 1.4 = RS 2.[008]+ 64-5 || 61-2; vi 5-6, 8-9)
with the reversal of urbt || hin, which matches the sense, in
ypth hin bbhtm
Let an aperture be opened in my mansion,
urbt bqrb hklm
a window within the palace.
(KTU 1.4 vii 17-9 II 25-7).55


Cf. AVISHUR 1984, 344-9 and S E G E R T 1984, 108.

See also K U G E L 1981, 27-40.
See W A T S O N 1981b = 1994b, 262-6.

W o r d pairs, then, were an integral part of the poet's composing technique and the very traditional character of versification in Ugaritic
resulted in most pairs remaining 'fixed'.


Formulae and formulaic patterns


T h e Ugaritic poems were composed using traditional formulae, sometimes with modifications or complete transformations. 5 6 A c o m m o n
type of formula is the one which introduces speech and it can take
many forms. Typical examples are
wy'n Itpn il dpid
Answer did Lutpan, kindly god
(KTU 1.1 = RS 3.361 iv 13)
where the slot Itpn il dpid could be filled by the appropriate name
or epithet, and
yu gh wysh

He raised his voice and exclaimed

which can be altered to suit gender and n u m b e r where necessary.

T h e r e is a whole range of such formulaic introductions. 5 . It was also
accepted convention that such introductions could occasionally be
omitted, either because they were implicit or for dramatic effect.'' 8

Formulaic patterns

T h e formulaic patterns to portray the passing of time are of two

types. T h e first has as its core two expressions, one denoting time
and the other an activity, as in the single line:
hn ym ysq yn
Behold, for a day they pour wine
(KTU 1.22 = RS 2.[024] i 17)
and either or both expressions can be extended to cover more time
(up to a sequence of seven days) or further activity (cf. K T U 1.16 =
R S 3.325+ vi 21-4; K T U 1.17 = R S 2. [004] ii 30-40). In the second pattern, the time expression denotes a period of days, months
or years and activity is continuous, e.g.




1 9 8 1 , also M E I E R 1 9 8 6 .
1983a = 1994b, 414- 24; 1992a. Some of them had elaborate preambles;
cf. W A T S O N 1994d.
W A T S O N 1990 = 1994b, 425-30.

ym ymm yt'qn
A day, two days elapse,
lymm lyrhm
from days to months,
rhm 'nt tngth
Maid Anat (still) looks for him.
(KTU 1.6 = RS 5.180+ ii 26-7).
O n c e again these patterns can be extended. They are used for themes
such as making a journey ( K T U 1.14 iv 31-5), preparing a banquet
( K T U 1.22 = R S 2.[024] i 21-6) or to depict a ruling monarch
( K T U 1.6 5 - 1 0 ; K T U 1.16 vi 21-4). T h e y also have other functions within the wider framework of the narratives. 59


Sound patterns

In spite of the lack of vocalization, some idea of the patterns of

sound exploited by the poets can be gained from the texts and a
few examples are provided here.


Particular words and forms were often chosen for reasons of alliteration although this feature should not be exaggerated. 6 0 First comes
the simple word-initial type, as in
ap ank ahwy aqht gzr
I too shall revive Hero Aqhat
(KTU 1.17 = RS 2. [004] vi 32-3).
Quite often consecutive lines of verse began with the same letter
which may indicate they were intended to be read as well as performed. An example, with line-initial i- twice, is
in b'l bbhth
Baal is not in his mansion,
il hd bqrb hklh
the god Hadd, in his palace.
(KTU 1.10 = RS 3.362+ ii 4-5)
1980 = 1994c, 431-4). In addition, multiple alliteration
seems to have played a part in longer sections of verse as has been
shown by P a r d e e ( 1 9 8 8 c ) .


A I T K E N 1987. See also LOEWENSTAMM 1965 = 1980, 192-209, P A R K E R 1989,46-52
('Repetition with a numerical framework'), T R O P P E R 1995 and YVHITAKER 1969. On
theme cf. A I T K E N 1990, 1991a, H I L L E R S 1973, L L O Y D 1990 and on motif, W A T S O N 1984a.
M A R G A L I T 1975, 310-3, 1979, 1980a.



Undoubtedly, plays on words formed part of Ugaritic verse but

because our knowledge of the language is limited, many puns escape
us.1'1 A few examples can be given, however.
ytt nhm mhrk
I hereby give you a snake as your bridal gift
bn bin itnnk
a serpent's son as your present.
(KTU 1.100 = RS 24.244:75-6)
Here, the use of the rare word itnn is apposite as it evokes tnn, the
mythical serpent monster, and the wordplay between itnn ('present')
and nh || bn btn, both denoting 'serpent' is transparent.
Another example is
' ym
Yam was strong, he did not sink.
(KTU 1.2 = RS 3.367 iv 17)
T h e play between Ym (probably pronounced yammu) and the verb
form ymk (yamukku) is self-evident. Another verb such as yql, 'he fell',
would have been less effective. 62


Figurative language

Ugaritic poetry is relatively rich in figurative language and includes

similes, metaphors and metaphorical expressions, personification and
apostrophe 6 5 as well as imagery. Occasionally it is difficult to know
where to draw the line between mythological language (as in 'the rain
of the Cloud-Rider', which refers to Baal) and extravagant expression (e.g. 'rain with which the stars anointed her'). Here, examples
are provided under appropriate headings, though it is not always
easy to make clear-cut distinctions. 64


In similes, the particle k (or km), 'like', is used, as in

tit kyn udm't
She drinks tears like wine
(KTU 1.6 = RS 2. [009]+ i 10)


For a complete survey see W A T S O N 1999.

'Janus parallelism' is another form of wordplay; cf.
See, e.g. W A T S O N 1984c = 1994b, 460-4.
See, in general, K O R P E L 1 9 9 0 .



though it can be omitted through ellipsis. Similes rarely come singly,

as they are mostly in sets of two, for example:
klbs km 1p dm ahh

He was clothed like a mantle in his brothers'

km all dm aryh
like a cape, in his kinsmen's blood.
(KTU 1.12 = RS 2.[012] ii 46-7) 65
or in sets of three:
thth kkdrt
Beneath her like balls were hea[ds],
lh kirbym kp
above her like locusts were palms,
kqsm grmn kp mhr
like grasshoppers in a swarm, warrior palms.
(KTU 1.3 = RS 2. [014]+ ii 9-11)

and once, a set of four: K T U 1.169 = R I H 7 8 / 2 0 : 3 - 4 . Similes can

be drawn out at length (as extended similes), indicative, perhaps of an
oral, improvising style, e.g.
km tdd cnt sd
Just as Anat hurries to the chase,
(and) sets the birds of the skies soaring away,
ttr cpt smm
tbl} alpm ap sin
(so) they butchered oxen as well as sheep,
ql trm wmri ilm
felled bulls and the fattest of rams, etc.
(KTU 1.22 = RS 2. [024] i 10-4).66
Only one cumulative simile has been identified:
klb arh I'glh
Like a cow's heart for its calf,
klb tat limrh
like a ewe's heart for its lamb,
so is Anat's heart after Baal.
km lb cnt atr b'l
(KTU 1.6 = RS 2.[009] ii 6-9)


M e t a p h o r is little used in Ugaritic verse. For example, fertility is

expressed metaphorically as
mm mn tmtm
Let the heavens rain oil,
nhlm tlk nbtm
the wadis run with honey.
(KTU 1.6 = RS 2. [009]+ iii 6-7;
M a n y metaphorical expressions are used such as dm csm 'blood of trees',
for grape juice and tl mm 'dew of heaven' for rain. In some metaphors
the mythological meaning may be muted, e.g. wytn qlh b'rpt, 'and may
he (Baal) give his voice in the clouds' ( K T U 1.4 = RS 2. [008]+ 8).

For another example cf. W A T S O N 1991a.

See also I R W I N 1983, but cf. P A R D E E 1988a, 127-9.

However, 'the lack of comparable contexts in Ugaritic prose makes

the recognition and evaluation of these and other poetic figures
difficult and uncertain' ( S e g e r t 1 9 7 9 , 7 3 3 ) .


Besides expressions such qr cnk 'the well of your eyes' ( K T U 1.16 =

R S 3.325+ i 26) to denote tears, weapons are said to fly off and
strike like birds of prey ( K T U 1.2 = RS 3.367 iv 13-7), a tree
stands for descendants ( K T U 1.19 = R S 3.322+ iii 53-4; cf. P a r d e e
1997a, 354, n. 121), Mot's domain is described as being a town (qrt)
called 'Miry' {hmiy), in a land called 'Filth or M u d ' {fih: K T U 1.4
= R S 2. [008]+ viii 12-4 || K T U 1.5 = R S 2. [022]+ ii 15-6), a
mountain weeps (KTU 1.16 = RS 3.325+ i 7) and so on. Occasionally
the imagery is obscure due to our lack of knowledge (e.g. K T U
1.16 = R S 3.325+ vi 5 7 - 8 ; K T U 1.17 = R S 2.[004] vi 36-7).
Hyperbole also occurs ( W a t s o n 1979 = 1994b, 452-60). T h e use of
abstract for concrete is extremely rare, perhaps only srt, 'adversary'
(see above) and t'dt 'legation' 67 ( K T U 1.2 = R S 3.367 i 22; cf. G r a y
1965, 22, n. 6).68


An example

In order to illustrate some of the poetic structures, rhetorical techniques and expressive language described above, a passage is set out
here in tabular form, with analysis, comments and discussion.

A king is chosen ( K T U 1.6 = R S 2.[009]+ i 43-65)



line translation

introd. to sp.

"gm ysh il
Irbt alrt ym


Aloud shouted El
to Lady A.Y.


rr ''Irbt alrt ym


"Listen, Lady A.Y.


Give (me) one of your mlk

sons so I can make
him king!"



In 46ahd bbnk amlkn 04


As aptly translated by M . S . S M I T H 1994a, 2 6 6 .

For an example of imagery cf. W A T S O N 1992b.

keyword ||m

(table cont.)


line translation

introd. to sp.


wt'n rbt atrt ym


And Lady A.Y. replied

keyword ||m


bl nmlk yd' yIhn


'Shall we not make a

person of intellect

introd. to sp.

uy'n Itpn il dpid


And El, merciful god,



dq anm lyrz





'One feeble of strength

cannot run
(or) like B. release the
like the son of


And Lady A.Y. replied


'Shall we not make

Awesome Athtar king?
Let Awesome Athtar
be king!'


'm b'l ly'db mrh


'm bn dgn ktmsrn 10

introd. to sp.



bit nmlk 'ttr '



ymlk 'ttr 'rz



w'n rbt atrt ym







apnk 'ttr '

j7 bsrrt spn

14 Then Awesome Athtar

15 went up to the heights
of Sapnu,
ylb Ikht aliyn 59b'l 16 he sat on the throne of
Mighty Baal;


p'nh Itmgyn a)hdm

his feet did not reach

the footstool,
his head did not reach
its extremity.




rilh lymgy




introd. to sp.

uy'n 'ttr '


And Awesome Athtar





shall not be king in

the heights of Sapnu'


lamlk bsrrt spn

yrd 'ttr 'rz

yrd "'Mt aliyn b'l


'uymlk bars il klh

Awesome Athtar came

22 came down from the
throne of Mighty Baal,
23 and was king in all the
vast earth.







T h e following selective remarks are set out in the sequence of topics adopted above, and there is a brief overall evaluation (on the
whole passage see now X e l l a 1996a).
T h e passage is a combination of speech and narrative, linked by the
formulaic introductions to speech which are all monocola (except for
0 1 - 0 2 which is a non-parallel bicolon).
In 13 of the 23 lines, the verse-line corresponds to the line on the
tablet (i.e. 05-16, 0 9 - 1 5 and 20-23). Lines 14-18 could be analysed
in several other ways (e.g. monocolon + two bicola) or one could
argue that they form a pentacolon. fiy O n the other hand, although
amlkn (04) could be a single-word separate line, it would seem to
belong to tn ahd bahk amlkn, like the corresponding expression tn ahd
bahk ispa (see below). Similarly, some scholars argue that ktmsm (10)
comprises a separate line, 70 although this view has not been accepted
here. Others argue that 0 8 - 1 0 comprises a couplet.' 1
Difficulties are caused by the lack of a clear translation, especially
of 0 8 - 1 0 (survey: d e l O l m o L e t e 1984, 77). T h e expression bars il
klh has been translated in various ways. 72 T h e epithet ydc ylhn, perhaps 'he knows, he understands', may be a hendiadys and occurs
only here. Also unique is aps, 'edge' and the form nmlk is found only
here (twice);73 on bit, see below. T h e verb form amlk is used elsewhere only in K T U 1.16 = R S 3.325+ vi 37, 53, also in the context of a failed would-be usurper to the throne. T h e formula tn ahd
bahk, 'Give (me) one of your brothers (so that I may . . .)' occurs later
in this text ( K T U 1.6 = R S 2. [009]+ 19-20, where Mot is speaking to Baal), but nowhere else.


According to M A R G A L I T (1980b, 143), lines 01 13 are all 'monostichs'.

See e.g. G O O D 1994, P A R D E E 1997a, 269, n. 246.
K O R P F . I . - DE M O O R 1 9 8 6 , 1 8 0 = 1 9 8 8 , 12. For a different approach see


1965, 441


E.g. 'la terre dont il est matre', by X E L L A 1 9 9 6 , 3 8 7 , discussion 3 9 0 - 1 .

T h e first occurrence (06) could be an N-stem, but this verb form is rare and
here it is causative, probably D (SIVAN 1 9 9 7 , 1 1 6 . 1 3 5 ) .

T h e best examples are 12-13 and 17-18. In many 'couplets' it is
absent (eg. 01-02); 'staircase parallelism' is present in lines 21-23.
Antithetic, distant parallelism is effected by lines 16 and 22 (_ytb Ikht
aliyn b'lyrd Ikht aliyn b'l) and lines 20 and 23 (lamlk bsrrt spn ymlk
bars klh).
Strophes and stanzas
T h e strophic sequence of monocola, bicola and tricola is unpredictable, 74 and inasmuch as there is any demarcation into stanzas or
sections, these seem to be 0 1 - 0 6 , 0 7 - 1 3 , 14-18 and 19-23. T h e
only clear division is signalled by the combination of apnk and a tricolon (14-17).
T h e title rbt atrt ym, 'Lady Athirat Yam', occurs 4 times in 23 lines,
ttr crz, 'Awesome Athtar', occurs 5 times, but mlk occurs 6 times. In
view of the content, which primarily concerns kingship, the keyword
here, then, is mlk, 'to rule, be king'. Crucially, it is missing from the
first narrative section (14-18), where Athtar attempts to sit on Baal's
throne and fails miserably (mlk - 0). However, in the second narrative
section, where he actually rules over the earth, it does occur, only
once but to effect. It is interesting that when bl is repeated it takes on
a form with enclitic -t which is found only here ( A a r t u n 1974, 27).75
Sound patterns
Consecutive lines beginning with the same letter are 0 9 - 1 0 ( f ), 15-16
(y) and 2 1 - 2 2 (y again). Alliteration also occurs in the n a m e + epithet sequences, notably 'ttr 'rz, and perhaps in the obscure yd' ylhn;
see also srrt spn. T h e r e is probably wordplay between 'rz and yrzWord pairs
These include p'n || ris (found only here) and the repeated pairs / ||
I, mlk mlk, mgy || mgy and 'm || 'm.
Figurative language
Apart from the comparisons in the difficult tricolon where the wouldbe king is compared to Baal, there is virtually no figurative language,
but Athtar does perform a symbolic act (descent from the throne).
' 4 It is possible that lines 2 0 - 2 3 form a split couplet with inserted bicolon; cf.
1997a, 3 1 .
For a similar usage in respect of enclitic -m see W A T S O N 1992c, 238-9.




T h e passage demonstrates the difficulties caused by uncertain stichometry and obscure language, especially in 06 and 08-10. Also,
it is not always clear how lines were grouped together to form couplets or higher units. However, overall the structure is quite evident
and there are no textual problems to complicate matters. T h e mix
of speech and narrative is fairly typical and the use of repetition and
stock formulae is offset by the presence of rare vocabulary and grammatical forms. T h e interpretation of the passage is quite clear (no
suitable successor to Baal has been found) 76 and is reinforced in
particular by the keyword (mlk), while of course the passage forms
part of a larger whole (the Baal Cycle).


The character of Ugaritic poetry

M u c h of the above is concerned with techniques and rhetorical

devices, 77 but this does not mean that the aesthetic dimension is
absent. Even though we do not know who the poets of Ugarit were
nor for w h o m they composed their verse, it is evident that they were
masters of the language and well able to manipulate it in a variety
of ways, attracting and holding the attention of their listeners or
readers. 7 " T o do justice to such aspects would require detailed study
of each composition for which there is no space here. 7u Enough has
been provided, one hopes, to whet the appetite for closer reading.
As in some other ancient N e a r Eastern verse traditions, perhaps
the most salient feature of Ugaritic poetry is its unpredictability, a feature which runs right across the board from prosodie structure to
complete compositions. This means in effect that, with a few exceptions, length of line, whether two lines will be parallel or not and
if so, the type of parallelism adopted, whether a speech will or will
not have an introduction and whether the introduction will comprise one line or several, sequences of bicola, tricola and so on, how


For a nuanccd approach cf. W Y A T T 1 9 9 8 C , 1 3 2 , n. 7 5 .

Though the survey has not been exhaustive, e.g. rhetorical questions have not
been considered (cf. H E L D 1969).
See, for instance, W A T S O N 1988a = 1994b, 434-45 on delaying devices in
Ugaritic verse.
'' For examples of close analysis see H E T T E M A 1989 90, M A R G A L I T 1989a, T S E V A T




many lines there are in a 'strophe' or 'stanza', etc., are all completely variable.
Generally speaking, studies of Ugaritic poetry pay considerable
attention to similarities with other verse traditions, particularly ancient
Hebrew (e.g. A v i s h u r 1 9 9 4 , P a r k e r 1 9 8 9 ) . While this is useful, especially when features from other traditions can throw light on Ugaritic,
it is also of interest to determine in what respects Ugaritic differs from
such verse traditions. T h e question to be asked is: what is unique or
special to Ugaritic verse? According to S e g e r t ( 1 9 7 9 , 7 3 1 ) , 'The most
prominent feature of Ugaritic poetry is its parallelistic structure. It
can be said that no other literature of the ancient N e a r East, Semitic
or non-Semitic, exhibits such consistent application of this structure'.
O t h e r features which could be mentioned include the use of verse
for narrative (which though rare or virtually unknown in Hebrew or
Phoenician, is common in Mesopotamian tradition) and a general lack
of hymns (though this could be due to chance). In addition, there
is a tendency to alter repeated (parallel) passages slightly. Special
verse patterns such as 'staircase parallelism', and its combination with
anadiplosis (notably in K T U 1.10 = RS 3.362+), the use of word
pairs in fixed sequences, with variation and inversion rare, the use
of chiasmus to show two or more individuals acting as one (e.g.
K T U 1.17 = R S 2. [004] 10-1) and the split couplet 80 all seem to
be peculiar to Ugaritic. Also, unlike Hebrew and Akkadian verse,
there is no clustering of similes or of lines with inner parallelism. 81
However, descriptions of actions as preludes to speech and lengthy
introductions to speech are c o m m o n e r in Ugaritic than in Hebrew
verse, whereas abrupt or unmarked speech is rare. 82 'If there is a
specific contribution made by Ugaritic to a poetic tradition . . . [it]
seems to lie in injecting originality into a well-worn, stereotyped body
of versification . . . T h e single copies of the Ugaritic tablets that have
been found are, perhaps, the work of a single school which re-worked
stock and static verse and made it sparkle with new life'. 83










= 1994b, 157.
cf. W A T S O N 1990 = 1994b, 425-30.
= 1994b, 68.


T h e







The Baal cycle

The tablets

T h e Baal mythological cycle is the largest text from ancicnt Ugarit,

taking up six tablets ( K T U 1.1 6: 1.1 = RS 3.361, 1.2 iii = R S
3.346, 1.2 iiv = R S 3.367, 1.3 = RS 2.(014]+, 1.4 = RS 2.[008]+,
1.5 = RS 2. [022]+, 1.6 = RS 2.[009]+) and comprising in its surviving portions around fifty per cent of the contents. Each tablet is
written on both sides and usually has six columns, although 1.2 has
only four and 1.4 has eight. K T U 1.6 has the heading 'of Baal' and,
although the other tablets have their tops missing and it remains a
conjectural point, this was probably written on them all and is the
title of the whole composition. K T U 1.4 has a note on the edge,
' T h e scribe is Ilimilku, the master, Niqmad being king of Ugarit',
while K T U 1.6 has at the close a full colophon, giving the information that Ilimilku was not only the scribe but a student or assistant of a high religious officer of King Niqmad, probably the second
of that name. So the composition was officially approved. At a banquet scene in K T U 1.3 i there is talk of chanting and singing and
of a minstrel (n'm). This suggests a possible Sitz im I^ben in the royal
palace, though equally possible is some festival in the temple of Baal,
in the library of which the tablets were inscribed. No doubt the king
himself was often present. For a translation and full bibliography
consult W y a t t , 1998c, 33 146.

The contents

T h e plot or story-line centres around a contest (under the overall

supervision of El) for the kingship over gods and men, and divides
itself usefully into three parts, cach consisting roughly of two tablets.

Baal and Y a m - N a h a r ( K T U 1.1-2)

T h e first part leads up to Baal's defeat of his rival Prince Yam (sea),
also called J u d g e N a h a r (river), with the help of two maces constructed for him by the divine craftsman, Kothar-and-Hasis, a story
told in the last column of 1.2 (iv). T h e evidence of what happened
prior to the battle is, however, not at all so clear, since 1.1 is very
imperfectly preserved. But near the beginning (1.1 ii) a message is
sent from the supreme god, El, to Anat, Baal's sister, calling on her
to perform what seems to be some kind of ritual, which involved
setting an offering of war in the earth (perhaps the burying of
weapons) and, following this, offerings of love and peace. This ritual may originally have had to do with ceremonies for the cessation
of hostilities; but it was not, as far as we can tell, performed by
Anat, so it is more likely that the passage uses ritual language to
express a wish on El's part that the notorious goddess of war and
love should a b a n d o n her more savage ways and, in particular, not
employ them in her brother's support. It is an important indication
of the way El desires things to work out. Thereafter in 1.1 iii Kotharand-Hasis is summoned to El's distant abode, perhaps as an ally of
Baal, to be given a similar warning; for clearly Y a m - N a h a r is at this
point being favoured by El, since in 1.1 iv he accords him royal status in a kind of ceremony of coronation.
By 1.2 i, however, Y a m - N a h a r is worried; for he sends an embassy
to the divine assembly, complaining that Baal has been reviling him
and demanding his surrender. El appears to sympathize but Baal,
who is present, objects strongly and sends an angry reply back to
Yam-Nahar. In 1.2 iii, a large fragment (perhaps out of place in its
present position), El instructs Kothar-and-Hasis to build a palace for
Y a m - N a h a r , and the claims of a minor rival, Athtar, are dismissed.
W h e n , after a sizeable gap, the text resumes, the battle between Baal
and Y a m - N a h a r is joined, Baal with the encouragement of Kotharand-Hasis wins and, though it is not according to his plans, the
supreme god has presumably to accept that Baal is now king.

T h e Palace of Baal ( K T U 1.3-4)

These tablets concern the building of a palace for Baal, from which
he may exercise his newly achieved kingly power. After a victory
banquet (1.3 i), the goddess Anat resumes her warlike ways and

mercilessly slaughters the inhabitants of two u n n a m e d towns, thereafter repeating the process with a n u m b e r of soldiers and guests in
her own mansion (1.3 ii). At the beginning of the next column (1.3
iii) she sings of her affection for her brother, but Baal, perhaps like
El before him, perturbed by her behaviour, sends messengers to her
requesting her to perform the same 'ritual' for peace as El had previously asked for, but also tells her to visit him and help him search
for the secret of the lightning. She has to be reassured that YamN a h a r and his cohorts, w h o m she had worsted in the past, had been
finally dealt with and were no longer a threat to Baal (1.3 iv); but
she decides to call on him to find out for herself what is really worrying him. It transpires that what Baal has set his mind upon is to
have a palace like other gods, and she herself goes to El's abode to
press Baal's suit, which she does in too threatening a m a n n e r and
apparendy has it turned down (1.3 v).
A new plan has to be concocted, the working out of which takes
up the last column of 1.3, all of 1.4 i-iv and half of 1.4 v. This
involves the preparation of gifts for El's consort, Athirat, and the
enlisting of her intercession with the supreme deity. T h o u g h we know
from elsewhere that she is no friend of Baal's, she persuades him to
let Baal have his palace. Anat takes the good news to Baal, and he
immediately summons Kothar-and-Hasis to build a palace for him
on his sacred mountain Saphon (1.4 27ff.). At the end of of colu m n the divine craftsman suggests that it should have windows in
it, but at the beginning of column vi Baal refuses to entertain the
idea lest, it seems, his old enemy Y a m - N a h a r may gain entrance
and again wreak havoc on earth. The house is soon finished and a
celebratory feast held (1.4 vi). Column vii tells how Baal then marches
through the surrounding territory, annexing a large n u m b e r of cities
and towns and thereby forming an empire for himself. Returning
home flushed with success, he puts away his former fears and resolves
that after all he will have windows in his palace. H e thunders out
of them; the earth reels and people far and near are terrified, his
enemies cling to the rocks in dismay, and he openly taunts them:
would anyone, prince or commoner, now dare to resist his royal
power? Column 1.4 viii nicely leads into the third main division of
the cycle as Baal sends messengers to the underworld abode of Mot
inviting him to a feast to acknowledge his sovereignty.

Baal and Mot ( K T U


Mot's embassy back to Baal dismisses this invitation contemptuously

(1.5 i). Rather Baal, just as he once transfixed Yam-Nahar, will soon
be descending to Mot's subterranean domains where he will be swallowed down and torn limb from limb by him. Baal, now in dread
of Mot, sends an abject reply (1.5 ii). In 1.5 (after two very poorly
preserved columns) Mot's messenger is advising Baal to take his wind
and cloud and his other attendants down to the underworld and
assume the condition of the strengthless dead. Baal obeys, but on
his way he has connexion with a heifer, who is delivered of a boy
w h o m Baal clothes in his own robe. 1.5 vi relates how the substitute's dead body was found at the edges of the earth and El, on
hearing the news in his mountainous home, descends from his throne
and wallows in sackcloth and ashes, wondering what will now become
of Baal's followers.
In K T U 1.6 i Anat and the sun-goddess Shapsh together bury
the surrogate corpse, and Anat goes on to the abode of El to inform
Athirat and her brood that they can now rejoice because Baal is
dead. A short interlude follows in which Athirat proposes Athtar for
the vacant kingship, but on proving unfit to take Baal's place he
resigns. In K T U 1.6 ii Anat, seemingly now aware of what has happened, goes to the underworld to search for the real Baal. She confronts Mot and summarily slays him, and then (1.6 iii) returns to El
to inform him that Mot is no more. She invites El to have a dream,
in which he sees the heavens raining oil and the valleys running
with honey; this shows that Baal still lives. In 1.6 iv Anat is invited
by El to speak to Shapsh, and they make plans for Baal's return.
K T U 1.6 and vi tell of Baal's restoration to full vigour in the
world above, and also of Mot's recovery after seven years. T h e y
argue threateningly with each other, at least some of the time on
Mt Saphon, and finally fall to fighting. T h e y butt and gore like wild
animals, and both fall exhausted to the ground. At this juncture the
goddess Shapsh arrives to warn Mot that fighting with Baal is useless, for El is now on Baal's side. Mot, at last afraid, picks himself
up and declares that Baal is rightfully king.


Since the early days of Ugaritic research the Baal cycle has nearly
always been interpreted as a ritual and seasonal text, either enacted

dramatically at Ugarit's New Year festival in the autumn or, more

broadly, as reflecting ritual events that took place at certain points
of the Syrian agricultural year (see e.g. H v i d b e r g 1962 [1938],
G a s t e r 1950, d e M o o r 1971, 1972). Such views, in various shapes,
are still influential in ancient Near Eastern, including Old Testament,
studies, but over the years they have been increasingly and quite
severely criticised ( F o n t e n r o s e 1966, G r a b b e 1976, M . S . S m i t h 1986,
W y a t t 1996a, 1998a). Few would argue today that where we have
a myth, a ritual basis must be assumed or that the proper ordering
of the seasons and their role in natural, or even h u m a n fertility were
the only or even the chief interests of Ugaritic mythology.
A good example of an agrarian meaning being read into a passage is the description of Anat's destruction of Mot in K T U 1.6 ii
3Iff., which speaks of her threshing him with a 'blade', winnowing
him with a sieve, burning him with fire, grinding him with millstones, and then throwing his remains into the open fields for the
birds to eat. Commentators have seen in this a mythological counterpart of a ritual ceremony held each year at the time of the grain
harvest. H o w Mot's discomfiture at this time is to be equated with
his normal role later in bringing about the summer dryness, or with
his ongoing role in swallowing h u m a n bodies, is not said. M u c h
more likely is the view that the whole scene is metaphorical ( L o e w e n s t a m m 1972); cf. the not dissimilar language used of the destruction
of the golden calf in Exodus 32:20. In other words, Anat destroys
Mot thoroughly; there is nothing more to it than that.
O f late it has been argued (e.g. by W y a t t 1996a, 1998a) that
some of the themes in the Baal cycle, notably the Chaoskampf, are
much more archetypal than merely agricultural, and that they may
really be referring to issues of kingship and the exercise of power in
the ancient world. T h e way the gods behave is a mirror image of
the way rulers in this world behave or should behave. Approaches
along these lines could be much more fruitful.
T h e part played in the story of the cycle by the supreme god El
is particularly revealing here (see further G i b s o n 1 9 8 4 ) . T h e three
great deities (and one minor candidate), whose exploits fill the cycle,
are battling essentially for supremacy over the earth. T h e kingship
to which they aspire is, as El's viceregent, to control the earth, YamN a h a r through the waters which surround it and are the source of
its streams and rivers, Baal through his rains and thunder and the
air which people breath, Mot through the droughts and dryness he

can cause. Baal is the victor in the contest, becoming the 'prince,
lord of earth' ( K T U 1.3 i 3 - 4 ; 1.5 vi 10; 1.6 iii 9; 1.6 iv 29), holding at bay the unruly waters from outside and bringing the dry season to an end by his rains. But there is much more than these
naturalistic roles to the three of them. Y a m - N a h a r also engenders
moral chaos or evil in the lives of mankind, Mot eventually finishes
everyone off, while Baal by his control of the atmosphere can be
regarded as the life-force in the world of Ugarit. H e did not only
secure year by year the ordered succession of the seasons, but every
day of every year he had to wage a constant battle against evil and
death in the lives of humanity, so that the one did not cause too
m a n y depredations or the other gain too m a n y victims too soon.
T h a t was really why Baal became the favourite god of the people
of Ugarit.
But he became this in the last resort u n d e r the supervision, indeed
by the connivance of El reigning from his distant abode beyond the
earth. From the standpoint of people on earth, looking around them
in fear and trembling, the encounters between the negative and positive forces in their environment were tense and awesome affairs,
and Baal's victory was always in the balance and never certain. But
the people of Ugarit could also take comfort from their belief that
beyond the squabbling powers that impinged so insistendy on their
everyday lives stood a remoter but by no means disinterested figure,
the god El, who had fathered the gods, the nice and nasty both,
and had created the universe, contrary powers and all, who must
therefore have planned it that way and built both good and evil into
its very fabric, and who could for that very reason be trusted to
uphold its equilibrium. For all his mistakes, his choosing the wrong
side, his blustering, his pleading, his changes of mind, his putting
up with impertinence from his underlings, hisif you willlack of
power, he was the one ultimately in charge, whom the other gods
had to visit for approval whenever they had an enterprise planned,
and whose was the final decision, which, however reluctandy, they
had to accept. And perhaps most remarkably of all, he masterminded
the balance he sought, not by calling on openly superior force but
by relying upon an engaging mixture of diplomacy and conciliation,
sharpness and persuasion. H e arranged it so that usually good and
life triumphed, but even evil and death were his 'darling' and 'beloved'
children ( K T U 1.1 iv 20; 1.4 viii 23~24) and had, as it were, their
rights too. This is the view of reality espoused by the people of

Ugarit, their explanation of the divine ways with the universe and
with h u m a n beings, their estimate of power and the m a n n e r it is
exercised; it is, for all the fancifulness with which ancient peoples in
their myths expressed themselves, mature and not lacking either faith
or irony. It may have involved naturalistic reasoning, but it involved
a great deal more besides.


Other mythological texts

T h e r e are in KTU quite a few other mythological texts or at any

rate partly mythological texts, that is, passages embedded in legends,
hymns or rituals. T h e larger of these, like Keret ( K T U 1.14-16 =
R S 2.[003]+, 3.343+ and 3.325+) and Aqhat ( K T U 1.17-19 = R S
2.[004], 3.340 and 3.322+) which some may prefer to call legends
or sagas, or the Rpum texts ( K T U 1.20-22 = RS 3.348, 2.[019] and
2.[024]), are given separate treatment later in this chapter ( 6 . 2 - 4
below). M a n y of the rest are mere fragments, e.g. extracts from the
Baal cycle used probably for scribal practice ( K T U 1.7 = RS 5.180+,
K T U 1.8 = R S 3.364, and K T U 1.133 = R S 24.293). I concentrate
here on the most intriguing and (relatively) well preserved texts. They
are all commented on, with up-to-date bibliography, in W y a t t , 1998c.

Baal and Anat ( K T U 1.10 = R S 3.362+)

This tablet has three columns of text on one side of the tablet only.
O f the first column little readable text survives, but at the beginning
of the second Anat calls on Baal (also called Hadd) in his palace.
O n being told that he is out hunting in the Shamak marsh, she follows him there, and is warmly welcomed. She sees a cow giving
birth and is apparently seized with passion, as is Baal who mounts
her before returning to his sacred mountain. As a consequence she
gives birth to a bull and, on her taking the news to Baal, he rejoices.
Some scholars attach the tablet to the Baal cycle, but it is more
likely that it belongs to a series recounting his dalliances with his
sister (e.g. K T U 1.11 = R S 3.319 and K T U 1.13 = RS 1.006). It
is not obviously thogonie, ritualistic or seasonal.

The Devourers ( K T U 1.12 = R S 2.[012])

This difficult text survives in two columns. T h e first tells of the conception of monstrous creatures by the handmaids of the deities Yarih

(the moon-god) and Athirat (the wife of El), who complain to El that
they are being caused distress by carrying them. T h e head of the
gods, doubtless their begetter, is amused by this, and instructs the
handmaids to go into the desert to bear their offspring. In his naming of them they are likened to bulls and steers. Baal is present, and
he expresses a great interest in them, perhaps for hunting purposes.
In the second column, after a long gap, the offspring, called the
'devourers', set upon and destroy Baal, who falls into a swamp, after
which the earth suffers a drought for seven or eight years. H e is
eventually found by his brothers and and restored. At the end of
the text a few lines tell the king to perform a water ritual, presumably to guard against the disaster caused by the 'devourers'. This
disaster cannot be a seasonal disaster but is a long-lasting one. T h e
real point does not seem to be about Baal but about El fathering
such dangerous creatures.

Hymn to Anat ( K T U 1.13 = RS 1.006)

A hymnic text, interesting mythologically for its portrayal of the complex character of Anat, called the 'virgin' in the Baal cycle, and in
the final lines here described as voracious to bring forth, although
her w o m b had not known conception. T h e r e is no need to connect
the hymn with any specific ritual, e.g. an incantation against infertility.

The Gracious Gods ( K T U 1.23 = R S 2.002)

A quite substantial text, written on both sides of a single column

tablet and nearly complete. It is clearly a cultic tablet, most of the
obverse consisting of little hymns, blessings on the king and queen
and the ministerial personnel, instructions about repeating certain
lines, about niches for the gods, about incense offerings, and so on.
T h e r e are also citations from a few mythological texts, in some cases
merely a heading, but in two cases rather fuller; in 11. 8 - 1 1 , there
is a short excerpt about Mot-and-Shar ('death and the prince', a byname of the god of death), and in 11. 3 0 - 7 6 a longer story about
El's seduction of two w o m e n (perhaps the goddesses Athirat and
Anat), who give birth to Shahar and Shalim and then to the gracious gods as a whole. Mot-and-Shar holds in his hands the sceptres of bereavement and widowhood, and is felled by the 'pruners
of the vine'. His removal from the scene makes it possible for El,

the progenitor of the gods, to father offspring on the two women.

It is interesting that the first deities to be born are S h a h a r and
Shalim, whose names mean 'dawn' and 'dusk'; we may compare the
beginning of Genesis where the division of day and night is the first
act of creation. But it is the behaviour of the gracious gods after
birth that is worth remarking on. T h e y open their mouths greedily
to swallow the birds of the air and the fish of the sea, and are sent
off by El with their mothers into the desert where for seven or eight
years they hunt for food. T h e y eventually come upon someone callcd
the 'watchman of the sown land' who invites them in to continue
their eating and drinking.
Is the meaning of this myth that the gods are not satisfied with
the natural provision of the open country, but require in addition
the offerings of the cultivated land which h u m a n beings bring them?
It is unlikely that such a profound observation, in effect that though
mankind are clearly dependent on the gods, they in their turn are
dependent (or at least partially so) on mankind, would be confined
to one particular ceremony, agricultural, fertility-angled or otherwise.
Doubdess this myth found expression on numerous liturgical occasions
at Ugarit. It is a not untransparent theogony or explanation of how
and why the gods came into existence.

Nikkal and the Kotharat ( K T U 1.24 = RS 5.194)

T h e mythological portion of this text, written like the last one on

two sides of a single tablet, relates how a deity called Hirhib, king
of summer, who behaves like a typical eastern marriage-broker, arranged the betrothal of a lunar goddess Nikkal to the moon god
Yarih. It is probably, like the second narrative in the previous text,
an extract from a fuller thogonie myth. This tale is preceded and
followed by hymns of praise and invocation to Nikkal, Hirhib and
the Kotharat, who are the sages-femmes of the Ugaritic pantheon. In
the first hymn the Kotharat are summoned to oversee the birth of
a son to the two moon deities. T h e last lines of the second hymn
with their allusions to incantations to the Kotharat, betray the purpose of the whole poem, which is to secure for a h u m a n girl Prbht
the same blessing and protection in her forthcoming marriage as had
been enjoyed by the goddess Nikkal in hers. Probably, with the necessary change of the girl's name, the text was regularly recited at
ceremonies of engagement and courtship.


Shapsh and the Mare ( K T U 1.100 = RS 24.244)

This is a long and excellently preserved but difficult text containing

in the opinion of most commentators a charm against snake-bite.
T h e daughter of the sun-goddess Shapsh (or perhaps simply a mare,
as her n a m e may be translated) calls on her to carry a message to
El, Baal and various other deities in order to obtain help from them
in curing the malady. Only when the god H o r o n (apparently a
chthonic deity) is a p p r o a c h e d is a positive response forthcoming.
According to others, however, the text is chiefly a mythical narrative,
not a charm and the serpent mentioned represents some cosmic disaster which is removed by H o r o n .

El's Banquet ( K T U 1.114 = R S 24.258)

This is description of a banquet to which El invites the other gods

and at which he falls outrageously drunk. T h e last lines on the reverse
contain an incantation for the cure of a disease or perhaps, as has
been suggested, a hangover.
T h e texts assembled here, some of recent discovery, give us a glimpse
of the diversity of Ugaritic mythology but, apart from the Baal cycle
and other larger texts like Keret and Aqhat, their extent is not very
great. W e have some way to go before a comprehensive account is
possible. Perhaps the only things we can say is that myths are not
always, if much at all, connected with ritual, and especially that Near
Eastern, including Ugarit, mythology is not always, if much at all,
obsessed with matters of seasonal agriculture or fertility.








The history of the text: discovery, publication, editions

2.1.1 T h e poem of Keret is one of the three major literary works

which gifted Canaanite poets of the Late Bronze Age (ca. 1500-1200
bce) bequeathed serendipitously to 20th century civilization. Excavated
at modern Ras Shamra on the northern Mediterranean coast by a
French archaeological team in the early thirties (1930-31), the poem
was published in three u n e q u a l instalments by the Assyriologist
V i r o l l e a u d , beginning with a monographic study in 1936 and concluding several years later with a series of articles in the periodical
Syna (vols. 22-23) published during the war. Accordingly, only the
monograph was widely known and accessible before 1945, although
a great deal of interpretation, much of it fanciful, was current in the
2.1.2 After the war, the text was republished or re-edited several
times. G o r d o n reproduced Virolleaud's text in the successive editions
of his Ugaritic G r a m m a r . A new and critical edition was published
by H e r d n e r in 1 9 6 3 . This much acclaimed two-volume work consists of all the alphabetic texts, literary and other discovered in the
thirties, together with photographic plates and autograph facsimiles.
Generally abbreviated as CT[C)A, this edition established the by-now
standard numeration of the Keret text as 1 4 - 1 5 - 1 6 , corresponding to
Virolleaud's IK-IIIK-IIK respectively.
2.1.3 CT[C)A was followed by a new edition published in 1976 by
a team of Ugaritic specialists affiliated with the University of Mnster
(Westphalia) headed by Oswald Loretz. Entitled Die Keilalphabetischen
Texte aus Ugart (KTU), it contains (in transliteration only) all the alphabetic texts discovered up to 1970. In this edition the Keret poem appears as K T U 1 . 1 4 - 1 5 - 1 6 = RS 2.[003]+ 3.343+ 3.325+, the initial
cipher indicative of its classification as a literary-poetic text. A revised
edition appeared in 1995. It too has the transliterated text only.
2.1.4 T h e text of the poem, labelled krt by the ancient scribe after
the royal hero of the story, is distributed over three rectangular clay

tablets of similar size (21x17 c m ; 15x17 cm; 2 3 x 1 7 . 5 cm). Each

tablet comprised originally six double-ruled columns, three on each
side, a n d altogether contained a p p r o x i m a t e l y a t h o u s a n d lines of
compactly written text. O f the three, only K T U 1.14 = R S 2. [003] +
is relatively well-preserved, a n d with the aid of the m a n y duplicate
passages in this portion of the p o e m can be restored nearly to its
pristine state. Most Ugaritic specialists hold that the extant tablets
never comprised the entire p o e m a n d assume that one or m o r e
tablets have been lost, especially at the conclusion ( K T U 1.16 = R S
3 . 3 2 5 + vi); this despite the colophon at the end of K T U 1.16 = R S
3 . 3 2 5 + vi n a m i n g the scribe Ilumilki w h o committed the p o e m to
writing (spr).
2.1.5 Since its discovery, Keret has been translated m a n y times a n d
into several languages, either as part of an anthology of Ugaritic literary texts or m o r e broadly of ancient N e a r Eastern texts.
T h e m a j o r translations, a n d the most widely cited, are the
ones by (a) G i n s b e r g , first in his A S O R m o n o g r a p h ( 1 9 4 6 ) a n d subsequently as part of P r i t c h a r d ' s anthology (AJVET); (b) G . R . D r i v e r
( 1 9 5 6 ) later substantially revised by G i b s o n ( 1 9 7 8 ) ; (c) H e r d n e r , in
a work jointly p r o d u c e d with C a q u o t a n d S z n y c e r (TOi) in 1 9 7 4 ,
a n d (d) d e l O l m o L e t e in 1 9 8 1 , where the Spanish translation is
a c c o m p a n i e d by extensive analytical discussion a n d a unique synoptic
presentation of alternative translations. 1


The history of (misinterpretation

2.2.1 T h e history of the p o e m ' s interpretation during the past sixty

years can be divided into three phases. Initially it was the subject
of extravagant claims of historicity a n d 'biblicization'. 2 T h e hero was
thought to have been a Phoenician king whose legions, including
m e m b e r s of the Israelite tribes of Asher a n d Zebulon, waged w a r
in the N e g e b region of Palestine a n d in E d o m . Progeny or devotees
of the biblical T e r a h , the father of A b r a h a m , were also thought to
be involved. However, rapid progress in Ugaritic philology p u t a

Also noteworthy are G O R D O N 1977, 34-59; DE M O O R 1987, G R E E N S T E I N 1997,

9 - 4 8 ; L O R E T Z 1997 and P A R D E E 1997a, 333-43. See also W Y A T T 1998C, 175-243.
Uncannily reminiscent of the Eblaite euphoria some 40 years later; plus a
change. . . .

quick and merciful end to this euphoric era and to the fata M o r g a n a
of biblical persons and tribal entities. O f the alleged geographical
allusions, only the shrine of Asherah in the environs of Tyre and
Sidon would survive the debacle; and even this determination was
destined not to go uncontested (Cf. A s t o u r 1973, 29-39). Still, there is wide if not consensual agreement today that
two fundamental insights of this early era in Ugaritic studies retain
their validity, viz., (a) Late Bronze Age Keret, like Aqhat, reflects a literary genre qualitatively different from the mythological tales of Baal,
Anat, and the members of the Ugaritic pantheon generallythis
notwithstanding the prominent roles of divine actors in both poems;
(b) the main works of Ugaritic literatureBaal-Mot, Keret and Aqhat
are 'classics' of Late Bronze Age Canaanite civilization and culture
and as such were known in Iron Age C a n a a n generally and in ancient
Israel specifically.
2.2.2 A second and similarly fleeting phase in the interpretation of
Keret was introduced by the Scandinavian, secondarily British, Myth
and Ritual School, representedat its most extremeby the publications of Engnell and Mowinckel. T h e former considered the poem
'a ritual for the Ugaritic sukkot festival' ( E n g n e l l 1967, 149) T h e
wedding party for Keret and his bride described in K T U 1.15 =
RS 3.343+ is characterized as 'originally the [ ] of the god
and goddess, celebrated annually and co-experienced by the participants as they watched the cult-drama and also when indulging in
sacral prostitution' ( E n g n e l l 1967, 148). For Mowinckel, on the other
hand, the poem exemplified myth attenuated as legend; behind the
portrait of the hero as a Phoenician king stands the figure of Adonis:
'[in Keret] the god is . . . strongly anthropomorphized; the original
god has become the dynastic founder, the mythic first ancestor of
the royal family . . . the poem is no longer a real myth, but a mythic
hero-legend.' 3 If the shortcomings and misconceptions of the French historical school were the result of inadequate philology, those of the
myth-ritualists were the product of faulty methodology. By means of




1941, 142-3, as translated by



1967, 148. See further

careful selection and tendentious interpretation of certain model texts,

'evidence' is created proving the existence of 'an organic [ANEastern]
culture . . . whose special feature is the domination throughout by the
divine kingship idea' ( E n g n e l l 1967, 2). This 'pattern' is then applied
to other texts assumed a priori to reflect this 'pattern'. T h e explanatory value of this theory is commensurate with its (non-)falsifiability.
Basic to this approach, which breathed its last in Ugaritic studies
with G r a y ' s monograph in the mid-fifties, 4 is the axiom that any
ancient N e a r Eastern literary text, be it myth or legend, is necessarily 'functional', and almost invariably so in the cultic sphere where
the ancients are presumed to have spent all their leisure time. In
Gray's words, 'the t e x t . . . was not an aesthetic exercise'presumably the author's understanding of literary creativity'but served a
practical purpose in the community where it was current to achieve
some desired end or to conserve . . . all the social conventions and
the social order'. 5 It is also typical of this approach that its advocates do not
feel constrained to demonstrate precisely how this 'conservation of
values' is actually implemented in the poem or how an audience
might infer such a conclusion. It never occurs either to Engnell or
to Gray to query whether the story might not be understood by at
least some readers or listeners in a quite different, even opposed,
manner, e.g., as underlining the perilousness of a social order predicated on the health of an individual, mortal king supported in turn
by a bunch of rather inept gods; or that the author of Keret, far from
preaching the doctrine of divine kingship, might in fact be condemning it by means of a lethal dose of parody. In the final analysis, the failure of the Myth and Ritual school lies in its denial of the
literary ontology of the text. 6
revised and attenuated in G R A Y 1 9 6 5 .
4 - 5 . In all fairness it should be noted that he concedes that the
poem 'was not deliberately [so] designed' (ibid., 5). But nowhere are we informed
what this 'original' design may have been, since it surely was not a mere 'aesthetic
exercise'. This concession, however, contradicts the Myth-Ritual postulate of a cultic 'Sitz-im-Leben' for all texts with divine characters and in fact heralds the school's
Contrast the astute observation of DE L A N G H E 1 9 5 8 , 1 3 1 , citing B A U M G A R T N E R
( 1 9 4 1 , 8 9 - 9 1 ) that 'aesthetic interests stand side by side with religious interests' in
the Ugaritic literary texts. However, very few specialists in Ugaritic have taken this
admonition to heart, either before or after. This is true even of B E R N H A R D T ( 1 9 5 6 ) ,
who is at pains to criticizevery successfullythe Myth-Ritualists but whose own






O n e cannot take leave of the 'pre-historic' era of Ugaritic

studies without taking note of the important study by the Danish
Semitist P e d e r s e n . Published in 1 9 4 1 with only V i r o l l e a u d ' s monograph at hand ( 1 9 3 6 ) , Die Krt Legende is probably the only study of
this era whose influence abides to the present. This influence has
unfortunately perpetuated a basic misconception in the interpretation of the poem.

to his credit, rejects outright the Myth-Ritual

interpretation of M o w i n c k e l E n g n e l l ' s study ( 1 9 6 7 ) had yet to app e a r w h e n he states categorically that Keret 'ist nicht der Ausdruck
kultischer Vorgnge und ist kein Mythus' ( P e d e r s e n 1 9 4 1 , 6 4 ) . Methodologically he stands close to the French School. He assumes that
we are dealing with a historiographie work whose historical kernel
is heavily overlaid with legend'wie der israelitischen Passahlegende'
( P e d e r s e n 1 9 4 1 , 6 4 ) . T h e hero 'ist ein Urknig, Grnder einer Dynastie'
( P e d e r s e n 1 9 4 1 , 6 5 ) i n reality, not just in the plot; and the principal theme is 'die Sicherung der Dynastie durch N a c h k o m m e n schaft' ( P e d e r s e n 1 9 4 1 , 6 4 ) . In other words, Keret is at heart a work
of propaganda commissioned by a royal house and executed by a
poet with the soul of a priest.

T h e theme of divinely sanctioned dynastic kingship, moreover, is deemed to reflect a society 'deren Knigsgeschlecht schon
als eingewurzelt betrachtet werden kann', analogous to the Israelite
society which spawned the Davidic royal ideology: 'hier wie dort
handelt es sich u m Legenden welchen den festen Bestand der herrschenden Dynastie besttigen und begrnden' ( P e d e r s e n 1 9 4 1 , 1 0 4 ) .
T h e author of Keret is so to speak a 'kept w o m a n ' of the political
authorities. His hand is free to write but his soul is in bondage. O n e
does not normally take the work of such writers seriously, whatever
their technical virtuosity.
This view of the Keret poem as a work of royal propag a n d a b y implication if not explicitly, by the ruling house of the

position is merely a re-statement of PEDERSEN'S: 'Richtiger wird man jedoch von

einer Besttigung der Erwhlung der Keret-Dynastie sprechen' {ibid., 119). This is
the raison d'tre of a political manifesto, not a work of art.

kingdom of Ugarit whose dynasty Keret is presumed to have founded

is very widespread in contemporary Ugaritic scholarship; and it is
hugely mistaken. 7 T h e curious omission, on these assumptions, of any reference to K e r e t o r his son and heir Yassibas king(s) of Ugarit
has long been noted. It is reinforced by a similar omission in the
so-called 'Ugaritic king list' ( K T U 1.113 = R S 24.257). However,
the real shortcoming of this view is that it fails to distinguish the
substance of the plot from the authorial intention, the creation from
the creator. T h e statements placed in the mouths of the characters
are naively taken as the author's own point of view. 8 In fact, a close, methodologically unbiased scrutiny of how
the author of Keret depicts his characters must surely lead to the conclusion that far from endorsing sacral dynastic kingship the poet actually ridicules it. T h e opening scene, for example, portrays the king
as a hapless soul who has gone through seven wivesthe first of
w h o m simply 'walked out' (tbc) on him!and who can think of no
better expedient than, like a baby, to cry himself to sleep (only the
soothing lullaby is missing. . .). Subsequently he will conscript all the
men of his kingdomincluding the disabled and the newly-wed
for a 'historic' military campaign to the Bashan for the grand purpose o f . . . obtaining a wife! It is inconceivable that this entire scenario
should have evoked from a contemporary audience anything but
gales of laughter. 9

Cf. e.g., B E R N H A R D T ( 1 9 5 6 , 1 2 0 ) : 'der text [steht] in enger Verbindung mit der

durch Keret begrndeten Herrscherdynastie . . . und [hat] als Tendenzdichtung die
Aufgabe . . . die besondere gttliche Erwhlung gerade dieser Dynastie ganz augenfllig darzustellen.' Cf. also above, . 6.
Thus, even if it be true 'dass wir in Keret einen typischen Vertreter des altorientalischen Sakralknigtums vor uns haben' ( B E R N H A R D T 1 9 5 6 , 1 1 6 ) , it does not
follow that this is an ideology which the author either espouses or wishes to propagate. Can one legitimately infer from the detailed description of the Persian monarchy in Esther that the author is desirous of propagating an ideology of oriental
despotism? Indeed, but for the strong nationalistic motives attributed to its (supposedly) Jewish author, the book of Esther might well have been understood as
political satire.
T h e fact that the latter stratagem is concocted by Keret's divine patronwith
Baal nowhere in sight!does not make it more 'respectable'; it merely adds to the
scope of the ridicule. El in Late Bronze Age Ugarit is a museum piece and a soulbrother of Shakespeare's Falstaff. But the most telling refutation of the dynastic interpretation comes from the final scene of the poem (a scene unknown to
Pedersen at the time of his essay) in its portrayal of Yassib, the king's
eldest son and divinely-ordained (as well as politically confirmed)
heir. It is difficult to imagine a less favourable comment on dynastic kingship or a more incongruous endorsement of a royal line supposedly founded by Keret. In short, there is altogether too much comedy
ody in Keret for it ever to have served as propaganda for
but the joy of living. For the author of Keret, not even the
sacred, much less the political institution of kingship. His
devotion are given unconditionally only to his art." 1

and paranything
gods are
love and

2.2.4 G i n s b e r g ' s short monograph published in 1946, inaugurated
a new era in the poem's interpretationor rather, explication. O n e
of his severest critics, Gaster, hailed it as 'a marked and revolutionary advance in our understanding [of the text]' ( G a s t e r 1947,
385). Ginsberg's was the first study to have addressed the material
in its (extant) entirety: K T U 1 . 1 4 - 1 5 - 1 6 = RS 2. [003]+ 3.343+
3.325+, are fully at his disposal and will be so henceforth for the
scholarly world to study and analyse. It is Ginsberg's contribution to
have been the first to establish the narrative coherence of the text.
However, Gastera dues-paying m e m b e r of the Myth and Ritual
schooltook Ginsberg to task for 'his obvious lack of acquaintance
with c o m m o n facts and methods of comparative religion, anthropology, and folklore . . . This leads . . . to an egregious disregard for
the cultural context and background of the narrative.' ( G a s t e r 1947,
286-7). But Ginsberg consciously and deliberately eschewed 'metaphysical' interpretation; he was a devout positivist. His strength lay

In a recent interview to a Montreal newspaper on the occasion of his 85th

birthday (The Gazette, Mar. 7/97), Irving LaytonCanada's (unofficial) poet-laureate
offered the following assessment of his life in the service of his art: 'Poetry never
let me down. My worry is, have I ever let poetry down? I should like to think that
I've never dishonoured poetry or turned my back on it. . . . A world without poetry
would be just intolerable. Unbearable.' O n e cannot mistake the (unintentional) piety
of this inveterate God-baiter and iconoclast. T h e ancient Canaanite bard would
surely have given this credo his unqualified assent.

in his philological dexterity at the level of grammatical analysis. H e

was primarily interested in words and how they combine to form
grammatical structures. T h e ideational content is secondary and the
literary craftsmanship incidental. Nevertheless, Ginsberg took a definite
stand on several 'metaphysical' issues (without however making them
a part of his discussion or interpretation). H e considered it 'probable' that the story 'contains a certain core of history'; he also deemed
it 'probable' that text K T U 1.14 = R S 2. [003]+ was preceded
'by one or more lost tablets'. Probability becomes certainty on the
question of the poem's alleged non-conclusion in K T U 1.16 = RS
3.325+." For all his skill in explicating the text, Ginsberg either misconstrued or overlooked several key elements in the plot, beginning
with the mistaken notion 12 of Keret as the victim of the catastrophic
loss of countless children deemed to have perished in bunches: a
third, a fourth, a fifth, etc. 13 H e is completely unaware that the real
reason for convening the nobility of Bt-Hbr ( K T U 1.15 = RS 3.343+
iv-vi) is to confirm Yassib as Keret's successor. In fact, Ginsberg's
translation of K T U 1.15 = R S 3.343+ v-vi lacks the thread of narrative coherence. This in turn leads him to wonder whether K T U
1.16 = R S 3.325+ i is the direct continuation of K T U 1.15 = R S
3.343+. T h e r e are also some incongruities in Ginsberg's rendering of K T U 1.16 = R S 3.325+. T h e phrase pnh. tgr. ysu ( K T U 1.16
= R S 3.325+ i 52-3), referring to the king's daughter 'Octavia' as
she emerges from the gateway of her (convent) residence, is under" The assumption of a story with large gaps before, within, and after the extant
tablets is very useful for the philologist operating exclusively with the microscope
of comparative grammar. If we possess only a half of the original work, then we
obviously are severely handicapped, if not outright precluded, from interpreting it
macroscopically on the metaphysical level. It is also something to fall back on when
one's interpretation of a given section, at the philological level, seems literarily vapid
or even inconsistent with that of a preceding or following section, whether in terms
of characterization or plot. It is not without irony that one notes the common
ground shared by Ginsberg and the myth-ritualist: neither takes seriously the poem
of Keret as literature or its author as artist. Both approach the text as so much grist
for the grinding.
Corrected early on by Cassuto but ignored by Ginsberg, even in his later work.
Here loo one senses how the lack of esteem for the pagan as poet enables the
Western scholar to attribute to him such a literary inanity, not to say moral obtuseness, in so quantifying human beings. Surely the biblical prejudice of the idolatrous
Canaanite and his 'debased' culture is here subverting the scholarly enterprise.

stood to mean 'Its sheen (i.e., of brother Ilhu's lance \mrh\) lights up
the gateway'. H e makes no effort to translate K T U 1.16 = RS
3.325+ ii 2 4 - 3 4 , although the text is quite well preserved; and he
passes over in silence the sudden a p p e a r a n c e of Octavia in her
father's bed-chamber in lines 5 0 - 1 . In col. iii, 8 - 9 , the language
tnnth . . . tltth does not evoke in Ginsberg's mind the association with
Keret's vow in K T U 1.14 = R S 2.[003]+, and his understanding
of the a C tiqat episode (v 28-vi 14) is both faulty and incomplete;
especially curious in his failure to render yqrs, while citing the analogy with Gilg. I ii 34 and its reference to tta iqtaris. Finally, the
curse which concludes K T U 1.16 R S 3.325+ (vi 54-8) is deemed
by Ginsberg to be 'unintelligible', in which case one wonders at the
certitude which informs his opinion as to the non-conclusion of the
poem at this point. T h e foregoing critique, be it noted, is based not on Ginsberg's
early translation ( G i n s b e r g 1946) but rather on his contribution to
P r i t c h a r d ' s anthology, first published in 1950 and subsequently
(unrevised!) in 1955 and 1969. T h e authority of Ginsberg's name
he was widely considered to be the 'doyen of Ugaritic studies' in the
fifties and sixtiesand the popularity of Pritchard's anthology, which
soon became a standard reference work for biblical and ancient Near
Eastern studies, go a long way towards explaining the rather limited
progress made subsequently in the elucidation of the poem at the
most basic level of narrative explication. T h e unspoken if not also
unconscious assumption is that short of a windfall discovery of additional copies, Ginsberg's translations of the major Ugaritic poetic
texts ( K T U 1.1-6 = R S 3.361, 3.367, 3.346, 2.[014]+, 2.[008]+,
2.[022]+, 2.[009]+; K T U 1.14-16 = R S 2. [003]+, 3.343+, 3.325+;
K T U 1.17-19 = RS 2.[004], 3.340, 3.322+) have defined the limits of what scholars can ever hope to know of them.
2.2.5 A new phase in the study of Keret is introduced by M e r r i l l ' s
short essay ( 1 9 6 8 ) , marking the first serious attempt to deal with the
poem as a literary uvre and providing the inspiration for an important essay by P a r k e r ( 1 9 7 7 ) nearly a decade later. ' T h e hypothesis of this paper', writes Merrill, 'is that the
poem . . . points to the 'house of Keret' as the basic issue. Every part

of the narrative finds its focus and delineation in this motif.' ( M e r r i l l

1968, 7). T h e story, it is supposed, 'begins with the ruined and
impoverished house of Keret. T h e king stands alone, without heir,
wife, or progeny.' ( M e r r i l l 1968, 9). By the end of K T U 1.15 =
R S 3.343+, 'the narrative of the king who has lost his 'house' and
regains it appears to be complete in itself. It has a beginning, a middle, and an end . . .'; and with mild surprise, 'yet the story continues' ( M e r r i l l 1968, 9-10). T h e 'fact' that the story continues beyond
its 'logical' conclusion leads Merrill to the conclusiontaken up
and elaborated subsequently by Parkerthat the unfulfilled vow to
Asherah, and her ensuing wrath, 'become the basis for the addition
of the other 'narratives' which are woven around the central concern for the 'house of Keret' and find their sub-themes in the three
areas of fertility, salubrity, and sovereignty.' Keret, on this hypothesis, is a composite work, although Merrill stops short of assuming
multiple authorship. It is the merit of Merrill's essay to have dealt with the story
in its own terms and with a vocabulary drawn from the field of literary criticism rather than comparative religion or Semitic linguistics. Implicit at least is the assumption of an author who has something
interesting, perhaps even important to say, and who commands the
necessary tools of the trade which he employs with the skill and
imagination worthy of an artist. This approach also implies an audience who can appreciate such a work, not as a cultic libretto or a
catechism of theological-political indoctrination, but as an artistic
endeavour, which, like good wine, is to be savoured and enjoyed. But for all the freshness and originality of its approach,
Merrill's essay, like Parker's subsequently, goes astray in its effort to
determine what the author is trying to say, as well as the specific
techniques which he has chosen for this purpose. T h e hypothesis of
a composite work, and a fortiori of multiple authorship, is symptomatic of a basic misconception, or rather, misperception. T h e view of an ancient work of Semitic literature as composite comes easily to scholars trained primarily in Old Testament
exegesis, as their partiality to myth-ritualism and cultic solutions generally tends to reflect their roles as (practising) theologians in the
Judaeo-Christian tradition. But it is nonetheless a view quite unfounded

here in Keret and in Ugaritic literature generally. 14 T h e r e is no evidence for a 'history' of any of the m a j o r Ugaritic poems, although
such is not to be precluded a limine.^ T h e vow-to-Asherah episode, it must be insisted, is absolutely central to the plot of the story for the simple reason that it
alone supplies the story with its dramatic quality. Without the vowepisode the story is a tale not worth the telling, much less the price
of admission to its performance. T h e absence of a corresponding
instruction in the dream-episode does not prove the vow to be secondary: if someone were intent on tampering with the original by
'grafting' on the vow episode, he would have had little difficulty
making the necessary emendation in K T U 1.14 = R S 2.[003]+. 1 6 T h e omission, on the other hand, speaks volumes for the
authorial intention. T h e r e is nothing more characteristic of the (male)
dramatis personae in Keret than their personal shortcomings and imperfectionsincluding most definitely the head of the pantheon who
(like Y H W H in the Garden-of-Eden story) fails to anticipate his clientservant's initiative. 1 ' However, there is a second and more basic problem in
Merrill's theory, viz., his initial assumption that the well-being of

It would be inappropriate in this connection to cite in rebuttal the complicated history of the Gilgamesh epic for obvious reasons related to the chronological spans of the respective works.
One should also not wish to deny the existence of 'parallel traditions' in Ugaritic literature, notably the stories dealing with the construction of Baal's palace
(KTU 1.3 II 1.4). However it has yet to be demonstrated (though often assumed)
that 1.3 and 1.4 belong to a single literary work or that they constitute a consecutive narrative.
T o be noted in this connection are the ill-preserved conversations of the
Udumite king, first with his wife Na'amat (KTU 1.14 = RS 2.[003]+ 14-23)
and subsequently with his messengers, commissioned to scale Mt Inbb and offer
sacrifice to the gods {ibid., 24-9; cf. Margalit, 224 31), both of which are unforeseen in Keret's dream. Since no authorial design can be discerned in their omission from the dream, and since nothing in the sequel would seem to presuppose
these conversations, the theoretical possibility of a 'second hand' can be entertained
here. However, as presently constituted the scene has the positive effect of 'humanizing the enemy', a sentiment very close to the (original) author's heart, as is evident from the emotional departure-scene which follows shortly at the beginning of
K T U 1.15 = RS 3.343+.
It should not be overlooked that once Keret awakes, El 'disappears' from the
story. He will return as a guest at the wedding reception, but he cannot be supposed to have monitored his client's actions in the interim.

KereCs dynasty stands at the centre of the poet's concern and creation.
It is simply not true that the king is portrayed at the beginning of
the poem as impoverished note how easily Keret dismisses El's offer
of 'silver and gold' in the dream, and the king of U d m ' s bribe subsequently during the siege. H e lacks progeny, but not for having
been bereaved; like Dan'el, he lacks a male heir for not having sired
one! N o w just as the birth of a son in Aqhat does not signal the
completion of the story but more nearly its commencement, so too
does the birth of Keret's offspring provide the impetus for moving
the story to its climax. T h e truly important developments in the
story come after the birth: in the case of Aqhat, the lad's treacherous m u r d e r by the goddess Anat and her Sutean mercenary, followed by the homicidal act of blood-redemption by the hero's sister.
In the case of Keret, the 'meat' of the story is the king's illness and
the behaviour of his offspring in response. T w o of them, without
aspirations to the throne, are devoted, loving, and obedient. T h e
third, predesdned by birth as heir-apparent, is the spoiled-brat antithesis. T h e attempted putsch by Yassib and the thunderous curse called
down on his head by his enraged father ( K T U 1 . 1 6 = R S 3 . 3 2 5 +
vi) bring the story full circle as it drives home the principal message: Keret is miserable at the beginning of the story for want of a
son and heir; he is equally miserable at its conclusion precisely
because of his son and heir. If the curse were not so funnyYassib
examining his teeth in the cup of his h a n d t h e ending would indeed
be sad. This is the essence of the poem as tragi-comedy, mixing the
tears of laughter with those of pain. T h e fate of the 'house of Keret' is thus of no particular
interest either to the poet or his audience. T h e real 'star' of Keret is
neither the king nor the gods but the invisible Moira who like the
poet delights in irony and makes the h u m a n life-experience at once
fascinating and unpredictablethe very qualities required of a good
For P a r k e r ( 1 9 7 7 , 1 6 7 ) , the poem of Keret is a conflation
of three originally independent stories executed by different poets at
different times and with variable degrees of editorial skill. ' O u r conclusion . . . is that the first section [= A] of Keret originally stood on

its own, and the material dealing with Keret's sickness [= B] was
attached to it by the insertion of the promise [= vow] passage into
the journey to U d m . . 18 U p to this point Parker is echoing Merrill.
But he goes further in positing multiple authorship and in his understanding of the Yassib episode ( K T U 1.16 = R S 3.325+ vi) as an
'originally independent story . . . tacked [iic] onto section B' ( P a r k e r
1977, 169). T h e alleged 'third story' (C) is of course quite incomplete,
and is assumed to have continued on (a) no longer extant tablet(s).
It follows accordingly that 'we are scarcely in a position to speak of
the theme or function of the whole work' ( P a r k e r 1 9 7 7 , 1 7 4 ) . Here
too (cf. critique of Ginsberg above) the assumption of incompleteness serves as a safeguard against criticism based on literary considerations; the 'answers' to difficult questions can be conveniently
assumed to lie in the unattested and empirically unverifiable 'hereafter'. E.g. it would be most surprising if a literary work which
had evolved in this 'tacky' way could be shown to have a unifying
theme or structure. Yet according to Parker, the combination of story
A and the 'neatly grafted' story yields, remarkably, a unified theme
described by him as 'the vulnerability and helplessness of the king
on the one hand, but also the benevolent power and wisdom of El
on the other." 9 We have discussed earlier some of the weaknesses in Merrill's
argument for the secondary nature of the vow episode. In his monograph P a r k e r ( 1 9 8 9 ) tries to meet one of these objections, but in
so doing actually reinforces it. Parker acknowledges that (a) the reason for suspecting the
vow is its absence from the list of detailed instructions in the dreamtheophany of K T U 1.14 = RS 2. [003]+ which the king subsequently


'The poets [TTD] who thus extended the poem . .

(PARKER 1 9 7 7 , 167).
1977, 174. In P A R K E R 1989 he claims to have discovered significant
structural differences between A and B. But at best these differences do not necessitate a distinction of authors. They are certainly consistent with the stylistic versatility and literary virtuosity of a single writer.


carries out to the letter and which include, incidentally, a sacrifice

to the god Baal, presumably to enlist his support for the venture;
(b) a good 'grafter' would accordingly have encountered little difficulty in making the necessary adjustment. Therefore (c) the omission was motivated ideologically, viz., by reverence for El's reputation.
T h e grafter did not want to make El responsible for the subsequent
debacle. Setting aside the conjectural and suspiciously ad hoc nature
of this latter supposition, it is surely clear that it effectively undermines
the case for multiple authorship. T h e same pious concern for El's
reputation could as easily have motivated the original author of the
poem! In other words, the vow-episode loses through this 'explanadon'
its entire value as an empirical indication of multiple authorship. 2 0 T h e case for viewing the Yassib episode as secondary
Parker's own contribution to the hypothesis of a composite work
is devoid of even the prima facie evidence supporting the secondariness
of the vow episode. O n e suspects that the very idea owes its birth
to the widely held view of the poem as lacking, in its extant form,
a conclusion, and hence needs have been continued elsewhere. It
seems improbable that Yassib would make his one and only appearance at the end of the story. T h e missing conclusion, it is supposed,
will have described how Yassib was punished for his insolence by
forfeiting his claim to the throne in favour of his younger sister
Octavia, the favourite of El and the gods (.sgrthn. abkm. etc.). Were such a denouement actually attested, it might well
be taken to support a theory of compositeness and multiple authorship, for it would totally contradict, in substance and spirit, much
of what has transpired in the poem up to this point. T h e fear is however unfounded. In point of fact, the role
of Yassib is much more firmly rooted in the story than is readily
apparent from his single appearance in K T U 1.16 = RS 3.325+ vi.
Parker's hypothesis appears to antedate the realization that the baro-

Here too the influence of O.T. scholarship is readily apparent. Bible scholars
commonly assume 'pious glosses' in the text originating with 'pious Jews' of the
post-exilic era.

niai council in K T U 1.15 = RS 3.343+ v-vi was only ostensibly

convened to 'weep for Keret', and tbat its agenda was secretly political: to confirm Yassib as successor in the event of Keret's demise
a decision ultimately taken, albeit after stormy debate. 2 ' With this
'certificate' in hand, Yassib has no reason to challenge his father
during the latter's illness (as Parker suggests he ought to have done
if his role were original); he need but bide his time until the king's
imminent demise. It is only when his expectations are frustrated by
Keret's miraculous recovery that he makes a pathetic attempt to unseat
him. Yassib, for whose confirmation so m u c h energy had been
expended but whose true character the author has skillfully concealed up to this point in the story, is now revealed at the conclusion for the 'wimp' that he is! T h e r e are two points to be emphasized in connection with
Parker's hypothesis: (a) that story the king's illnessis securely
tied to the figure of Yassib and his succession and can never have
existed independently thereof; (b) that the case for the Yassib episode
as an independent story C hangs entirely on the assumption that the
poem is not concluded at the end of K T U 1.16 = R S 3.325+ vi. T h e final point to be m a d e against Parker's case is the catastrophic consequences which the deletion of the Yassib episode has
on the literary structure and message of the poem. If Keret were a typical 'happy ending' narrative, the story
(Parker's story in particular) ought to have ended with the king's
recovery (much as Merrill's 'original story' ought to have ended with
the wedding reception in K T U 1.15 = RS 3.343+!). For if at the
beginning ( K T U 1.14 = R S 2.[003]+) the king is alone and in tears,
and then subsequently, facing death, he is tearfully embracingpossibly for the last timehis beloved 'blossom' Octavia ( K T U 1.16 =
R S 3.325+ ii 50ff.), he is surely smiling from ear to ear, surrounded
by his faithful wife and adoring children, at the feast described (laconically) in K T U 1.16 = R S 3.325+ vi 15-21; and his resumption of


The language of the council's decision can be presumed identical with the resolution presented by its president, the rk-il (KTU 1.15 = RS 3.343+ 18-21): 'rb.
p. lymg19 krt/sbia. p'10 b'lny/uymlkn [ j ] ^ ( . ) 'In 'When Keret arrives at the western
horizon (i.e., dies)|Our lord, at the setting s u n | T h e n will [Ya]ssib rule over us'.
Cf. M A R G A L I T 1982, 425; 1995, 252-2.

work as king of Bt-Hbr (ibid., 11. 22-4) ought to have been greeted
with much fanfare and public celebration. T h e 'addition' of the episode of filial infidelitya sin punishable by death in the Bible and which the very name 'Aqhat' (lit.,
'the-obedient-one') attests to as heinous in ancient Canaanite societyturns this would-be happy ending on its head at the same time
as it brings the story full-circle to tragi-comic conclusion. In K T U
1.14 = R S 2. [003]+ Keret is miserable for want of a son and heir;
at the end of K T U 1.16 = R S 3.325+ vi he is miserable for having obtained a son and heir. Nothing more can or need be said.
2.2.7 But a word should be said, in conclusion of our critique,
on vestiges of the historical approach to the poem still current.

Parker writes:

I would see the origin of the poem in a story about a king who undertook a campaign against another king to claim the king's daughter as
his bride. Negotiations between the king resulted in the ceding of the
woman in question, and hence in the marriage of the two and the birth
of children. . . . It is this much that forms the most solid basis for those
who claim that die poem reflects historical events . . .' ( P A R K E R 1 9 8 9 , 39) This statement, I submit, is as true (or false) of Keret as it
is (mutatis mutandis) of Hamlet, Julius Caesar, or Antony and Cleopatra,
none of which can be considered 'historical' works reflecting historical events. T h e y are works of the imagination, pure fiction, in which
historically attested personal a n d geographical names, scraps of history, social and religious customs are expertly utilized as trappings
for the plot and its characters by craftsmen minutely knowledgeable
in historical arcana andall importandymasters of disingenuity in
the service of artistic integrity. But even if the poet be inspired by a 'real event'which
in the case of Keret one is entitled to doubtthis determination is
no more consequential for understanding the poem and its author,
than is the Danish chronicle which inspired Shakespeare's Hamlet.
At most, such knowledge can produce some learned footnotes to the
text, enhancing its appreciation by cognoscenti but irrelevant and
boring for poet and audience alike.


The story in outline, the message in detail

2.3.1 Although there is no hard evidence to indicate that the poem

of Keret was ever the subject of dramatic presentation in a theatre
or like setting, it is useful, and certainly not misleading to summarize its contents as if it were. T h e material is most amenable. T h e 'prologue' in the opening lines of K T U 1.14 = RS
2. [003]+, now largely defective, introduced the hero, Keret, as king
of Bt-Hbr, situated '[by the se]a'. T h e king is a man of valour
([gbr. hyl]) and a devotee of El (glm. il) who is his 'patron' (ab); but
he is wretched for want of wife and children to fill his 'naked' (crwt)
palace. Seven dmes was Keret wed, but each marriage ended abrupdy,
for the most part tragically with the death of the spouse; in one
case, in childbirth. T h e absence of a (male) heir apparent causes his
seven brothers to cast greedy eyes on his throne.
Comment: (1) 'Bt-Hbr by-the-sea' ([gblt. y]rr) is a pseudonymic riddle to be solved by the audience in the course of the p o e m / p l a y .
(2) By presenting the king as a devotee of Elin contrast to the Baalworshipping poet and his audiencethe author conveys the message
that the story is about a historical figure of long-ago, the era of the
ancestors. 22 Keret is thus a 'patriarchal narrative'. T h e members of
the hero's clan (lim || umt) are to be found roaming the steppelands
between the (Phoenician) coast and the Euphrates ([ f ]</. nhr). (3)
Except for the royal backdrop, the scene is uncannily reminiscent of


For the author of Keret, this 'patriarchal era' began with Ditanu (Ug. dtn, var.
ddri) cited obliquely in K T U 1.15 = RS 3.343+ iii 2 - 3 || 13-5) thought to have
lived in the early MBA (ca. 2100 BCE; cf. K I T C H E N 1977, 131-42; H F . L T Z E R 1981,
1 10) and developing into an eponymous ancestor. It needs be emphasized however that Keret's Ditanu-ancestry does not make him a direct ancestor of the Ugaritic
kings Niqmaddu and Ammittamru ( K T U 1.161 = RS 34.126). Like Abraham,
Ditanu is 'the father of many [Amorite] nations'. This much however can be said:
the city-state kingdoms which speckled the Phoenician and north-Syrian coastline
in the early 2nd millennium BCE were all ruled by classes of Amorite stock. The
dynastic houses of Byblos and nearby Ugarit in particular could accordingly have
been related by ties of blood a n d / o r marriage, and both of them to clans residing
in Bashan (cf. K T U 1.108 = RS 24.252: 23-4). T h e phrase qbs. dtn, roughly 'union
of Ditanu' (KTU 1.15 = RS 3.343+ iii 2 - 3 || 13-5), like its parallel rpi. ars (ibid.,
K T U 1.108) denotes the transnational aspect of this consanguinity, the word ars
'land' contrasting with socio-political terms like qrt 'city' and mlk 'kingdom'. (This
usage of ars [Heb. 'eres] survives in the Hebrew Bible in the phrase 'am-h-'res,
denoting an institution of landed gentry who can 'make or break' a king. Cf. provisionally M A R G A L I T 1995, 255-6.)

the El-worshipping, Harran-based patriarchs of the Bible, one of

whom complains bitterly that he is without (legitimate) heir and who
subsequently sends his trusted servant to obtain a wife for his son
in the 'old country'.

Act 1, Scene 1 T h e curtain rises on the king about to retire for the night
to his sleeping c h a m b e r which he enters shedding tears of self-pity.
Curled up in bed in a foetal position, he falls asleep. His patron
deity now appears in his dream, having heard the heart-rending sobs
of his valiant servant. Wise but not omniscient, El inquires as to the
cause of the king's distress. Ever the jester, he speculates whether
Keret, dissatisfied with the modest extent of his kingdom, has designs
on his own; or perhaps, he wonders aloud, the king is short of money
to cover his regal expenses. In reply, the king assures his patron that
he wants for nothing material, and that his only wish, and the
panacea to his pain, is to sire a family, sons in particular. El is sympathetic; and the remainder of the dream (and
scene) is devoted to divine monologue wherein the deity issues a
series of detailed instructions as part of an elaborate plan for the
hero to realize his ambition. At the centre of the plan is the full
mobilization of the kingdom for a military expedition to the (Bashanite) kingdom of Udum(u), to be followed by a siege and ultimatum
to its king: surrender your eldest daughter, the fair Hry, to be Keret's
wife (or face the consequences).

Scene 2 T h e King awakens with a start, but with total recall of the
dream which he immediately begins to implement to the letter. He
attends first to his personwashing (for cleanliness) and rouging (for
war) then to the gods to w h o m he offers sacrifice, and then to the
business of war.
Comment: (1) El's oneiric thcophany is part and parcel of n o m a d i c /
Amurritic religion centering on El and his consort Asherah. El resides
in the subterranean fresh-water deep which feeds the palm-trees of
the oasis where the n o m a d pitches camp. W h e n the latter retires for
the night and sets his sleepy head down to rest, he is lulled to sleep

by the gurgling stream nearby. It is both logical and natural that

El, residing close by, should pay him an occasional visit in his dream,
especially when the nomad is troubled. (2) Again, the correspondence
with the biblical tradition-complex (in its and versions particularly)
is uncanny. T h e El(-Shaddai)-worshipping patriarchs receive their
divine visitations in nocturnal dreams. T h e same is true of the Elworshipping Aramean clairvoyant Balaam of the D A P T . Y H W H , by
contrast, never appears in a dream to his servant Moses, 23 and there
are no dream-theophanies in the Hexateuch outside Genesis and
N u m . 2 2 - 4 . (3) T h e characterization of both El and his protg is
parodical. T h e valiant warrior of the prologue is totally deconstructed
by the pathetic king crying himself to sleep like a baby. Crying is
womanish (2 Sam. 1:24; Lam. 1:2, etc.) and a sign of weakness in
men except in well-defined special circumstances. It never occurs to
the biblical author to depict A b r a h a m as weeping in Gen. 15; nor
does the son-less Dan'el weep when petitioning for a son (1.17). Both
are in contrast with the similarly situated H a n n a h (1 Sam. 1:10).
T h e proverbially wise El (by dint of age and experience) is parodied
by means of the patently ridiculous plan which he concocts: the total
mobilization of the kingdom, including the sick, the blind, and the
newly-wed normally exempt from military draft, and a strenuous and
very expensive seven-days march to the hinterland region of Bashan
for no better reason or exigency than to obtain (yet) a(nother) wife
for the king. Were this not enough, the poet will subsequently inform
us that Pbl, the king of Udum(u), is himself a devotee of El, and
his kingdom a 'gift' (un) from this same deity ( K T U 1.14 = R S
2. [003]+ vi 12-4). All El need have done was to send an oneiric
message to Pbl and the fair princess would have been on her way
to Bt-Hbr. 2 4 This contrasting of exaggerated means utilized for trivial ends is of course a staple of comedy and burlesque (cf. the M a r x

Cf. Num. 12:6 8. A later tradition, no longer familiar with the religio-historical presuppositions of the patriarchal faith, attributed this fact to the uniqueness
of Moses' prophetic status.
H e could also have spared U d u m the pains of siege, and its monarch the
pangs of uncertainty, by revealing himself in a dream to Pbl and thereby confirm
Keret's ultimatum as indeed inspired and supported by divine degree. But then El
would be seen to be truly wise and compassionate rather than the comical dotard
intended by the poet.

brothers), a contrast further accentuated here by the disparity between the normally peacable and compassionate El (Itpn. dpid) advising,
and devizing, a strategem of war. 25 (4) T h e use of parody at this
early stage in the story must be understood as setting the tone for
all that ensues. It is the dramaturgic equivalent of Shylock's 'poundof-flesh' bond contracted (ostensibly) 'in a merry sport', and to the
over-reaction of foolish king Ahasuerus (Est. 1) to the queen's refusal
of a royal summons (itself a parody of a king 'ruling from India to
Ethiopia'). It serves notice that the poem of Keret is a species of 'mock
epic', perhaps the oldest of its kind in recorded history.

Scene 3 T h e a r m y of Bt-Hbr marches in battle array to U d m

(= U d u m u in the land of G a < s h u - > r u [EA 256]). T h e march is
broken up into two more or less equal segments: 3 days from BtH b r to Tyre, where Keret pays an unscheduled (or at the least unanticipated in the dream) visit to the shrine of Asherah, El's wife, where
he takes a vow (cf. Gen. 28) that if his mission be successful (one
senses clearly the insecurity of this valiant warrior) he will pay to
Asherah's shrine 'twice [his bride's] weight in silver, thrice in gold'. T h r e e days later, on the seventh day of the campaign,
Keret's army arrives at U d u m and camps outside its walls after having cleared the countryside. T h e r e follow the futile efforts of Pbl,
king of U d u m , and his queen N a ' a m a t to relieve the siege, first
by offering Keret a bribe of silver, gold, three horses and chariot
(with attendant squire), and, simultaneously, sending messengers to
offer sacrifice atop nearby Mt. Inbb, the mythological abode of the
(war-)goddess Anat. T o no avail; Keret is a d a m a n t (and the gods,
by implication, unresponsive): only the surrender of beautiful H r y
in the description of whose (as yet unseen) beauty (he has only El's
word for it) the king waxes poeticwill suffice to remove the siege
(cf. mutatis mutandis 2 Sam. 20:14-22). T h e scene concludes with Hry
taking tearful leave of her family and friends as she sets out for
Keret's c a m p and her new life as queen of Bt-Hbr.


O n e may note the uncanny if fortuitous resemblance of El's plan with that of
Portia's 'virtuous father' (The Merchant of Venice), mocking the 'holy men [who]
at their death have good inspirations'. El's plan is similarly 'inspired'.

Comment: (1) O n the identifications of U d m and Mt. Inbb respectively, cf. M a r g a l i t 1995, 2 2 5 - 4 3 . (2) Although formally a married
couple, El and Asherah do not live together (cf. K T U 1.4 = R S
2. [008]+ iv).26 (3) T h e fact that the king's initiative, for all its good
intentions, eventually lands him in hot water couldif Keret were a
'serious' piece of literaturebe taken as implying the futility of human
endeavour and the advisability of resignation to divine will. But if,
as I maintain, Keret is tragi-comical, then the crisis precipitated by
the king's ill-fated initiative (the result, be it recalled of his absentmindedness) can and should be seen as contributing to his portrayal
as a pathetic figure, a 'Schlemiel' or 'Sad-Sack' who can do no right,
a master bungler. Keret, like Dan'el, is a 'talker', not a 'doer'. (4)
This characterization of the king is underscored by Pbl and Na'amat's
reluctance to become Keret's in-laws. After all, such an attitude is
not self-evident given Keret's credentials. A king of the backwater
kingdom of U d u m would normally have given his eye-teeth for a
liaison with the royal house of Bt-Hbr, alias Byblos. However,
Keret's reputation as a matrimonial 'jinx' has preceded him to Udum.

Act II Scene 1. T h e reception celebrating the marriage of Keret

and Hry is attended (i.a.) by the gods, including El and Baal. Asherah
is conspicuous by her absence. During dinner, Baal prompts El to
toast the newly-wed couple. El is glad to oblige: raising his wineglass, his blessing consists of a promise that Keret's wife will bear
him multiple offspring (cf. Gen. 15:5, etc.): seven || eight boys and a
like n u m b e r of girls. T h e eldest of the boys, to be named Yassib, will
be Keret's heir (poetically, he will be nursed by goddesses); the youngest
of the girls, 'Octavia', will be El's favourite (bkr, literally, 'first').
Comment: (1) T h e senior gods arrive at the party in pairs; the
'assembly', consisting of the minor (younger) and anonymous gods,


' The separation of El and Asherah on the mythopoeic level is surely a reflection
of the transformation of their originally pastoral-nomadic cult following the sedentarization of their worshippers. El is put out to pasture in the Upper Jordan Valley,
his domain extending from the foot of Mt Hermon near Dan as far as the Sea of
Galilee. But his consort starts up a new career among Tyrians as rbl. alrt. ym 'Lady
Asherah-of-the-Sea' where she is presumably worshipped as the patroness of fishermen (cf. her attendants qdl. wamrr described (KTU 1.3 = RS 2.[014]+ vi 10-1;
K T U 1.4 = RS 2. [008]+ iv 2-4) as dgy. rbl. alrt. y m 'fishermen of Lady Asherahof-the-Sea'.

arrive in threes. T h e absence of Asherah is accentuated by pairing

El with Baal. Anat (here labelled 'Rhmf [Heb. rehem]) is accompanied by the similarly bellicose Reshef. Kathir-and-Hasis, sporting a
binomial n a m e , escorts himself (our poet is a 'kibitzer'). (2) T h e
choice of R h m y as an alias for Anat is motivated by two considerations: (a) the alliteration with Rip; (b) the synonymity with ( c )Anat,
both referring to the female genitalia. 27 Scene 2. T h e scene shifts to the (unspecified) domicile of
Asherah. Seven years, and several birth-days, have elapsed and Keret's
pledge is still unpaid. With the king now in default, Asherah takes
her own vow: to make Keret pay . . . with his life!
Comment: 'Heaven has no rage . . . nor hell no fury, like a woman
scorned'. Cf. also Ps. 50:14, Eccl. 5:3. Scene 3. A party in Keret's h o m e m o r e precisely, his atr
for the nobility of Bt-Hbr, its 'Bulls' and 'Stags' in the poet's saucy
language. In preparation, the king instructs his queen to 'dress-up
like a maiden' (km[.n]crt) by hiding her bulges, doing up her hair,
and manicuring her fingernails. T h e ostensible purpose of the party
is to 'weep' (ritually) for the ailing Keret. But the secret agenda is
political, viz., to confirm the juvenile Yassib as heir and successor
to the throne of Bt-Hbr upon the king's supposedly imminent demise.
O n c e this political purpose is made known to the guests at the party,
a furious debate erupts, accompanied by shouting and clenched fists,
in the course of which the 'president' (irk. il, lit., 'chief member')
stands up to speak and pledges the support of the assembly for the
young prince. T h e ailing king replies. Invoking the private parts of
the president's wife, he blesses him for his support. H e then informs
the council that he expects to die within the month, blaming his
misfortune on Athirat's abiding hatred for his kingdom. His personal
fault he passes over in silence. However, the king's remarks, far from
stilling debate, add fuel to its fire, in the course of which both the
king and his queen are forced to intervene to restore order. T h e
king accuses his opponents of 'drinking his blood', while the queen
reprimands her guests for their indecorum as well as for their insin-


Cf. D E E M 1978; M A R G A L I T 1995, 241 2. T h e basic meaning of 'nh is 'open

up', normally of speech. Its use with sexual activity (cf. Ex. 32:18b) reflects a perceived symmetry between oral and vaginal anatomy (cf. Prov. 30:20).

uation that the king might be feigning illness in order to obtain an

endorsement of the crown-prince as successor to the throne. Hry
assures the noblemen that Keret's illness is, unfortunately, neither
dream nor fantasy; and he has the body sores and fever to prove
it! T h e conclusion of the scene is lost, but a political victory for the
royal family is a necessary inference. 28 From this moment on, Yassib
is heir-apparent in fact as well as in theory, and his enthronement
evidently a matter of days.
Comment (1) T h e location of the banquet in a tent (hmt) set up
in the family atr or burial-ground (cf. K T U 1.17 = RS 2.[004] i-ii)
points to a kispum or mrzfi, i.e., a feast associated with the cult of
the ancestral dead (dbh. ilm). This would furnish a convenient pretext for convening the nobles and a suitable occasion for 'beweeping' the sick king. It also is consistent with the all-male guest list as
well as the king's instructions to his wife to dress appropriately (cf.
K T U 1.4 = R S 2. [008]+ iii 10-22). This banquet is definitely a
'stag affair'.
(2) T h e scene implies the existence of a political group whose
authorization the king requires to transfer power to his son and thus
establish a dynasty. This council of barons is a fcudalistic body whose
duties, and prerogatives, are to 'advise and consent'. Neither servile
nor rubberstamping, it can make or break a king. A residue of tribal confederation, this group of grandees may be seen as the sociopolitical equivalent and 2nd-millennium precursor of the biblical
'am-h'res. (3) It is further implied in this scene that while the legitimacy of the dynastic principle is acknowledged, its cultural roots
are shallow indeed. T h e tribal tradition of charismatic leadership,
understood in terms of military prowess, is still very much alive, and
it goes far towards explaining the fierce opposition to the blankcheque endorsement of the king's son, still very young and untried
in battle and leadership. (4) Keret is in all likelihood (portrayed as)


T h e spirit and circumstances of Keret's speech are strongly reminiscent of the

speech delivered by Hattusi1i before the panku-assembly: 'Behold I have fallen
sick. . . . Behold, Mursilis is now my son . . .' It may not be too venturesome to
suggest accordingly that the missing portion of Keret's speech may have been formulated in a vein similar to the continuation of Hattusilis' address: 'In the hour
when a call to arms goes forth . . . you . . . must be [at hand to help my son]. When
three years have elapsed he shall go on a campaign. . . . If you take him (while still
a child) with you on a campaign, bring [him| back [safely].' (Translation apud



the first m e m b e r of his family to have occupied a throne, which he

may well have seized by overthrowing an incumbent ruler. His rise
to p o w e r a n d this is probably the extent of the story's historicity
(which in any case is only presupposed by the narrative)would
have been a model Idrimi who, with a b a n d of ruffians and outlaws,
conquered Alalakh and set himself up as king. Like Keret's El-religion
and his affiliation with a clan dispersed in the Syrian steppeland,
the present scene reflects the political ethos of an earlier epoch, viz.,
the formative stages of Amorite settlement in Phoenicia and N. Syria
and the struggle to establish the legitimacy of dynastic kingship in
a society barely weaned from non-hereditary charismatic leadership
and tribal organization.

Act III Scene 1. As the scene opens, preparations are underway

for Keret's funeral (although the king is still quite alive). T h e sound
of caterwauling w o m e n t h e poet prefers the comparison with howling dogs and coyotesfills the royal mansion. Overcome emotionally
by these depressing sounds and by the realization which they spur
of his father's imminent demise, the loving and devoted son Ilhu
approaches the king's bedside. With tears rolling down his pubescent
cheeks, he queries his father in disbelief (in the process giving expression to the current ideology of divine kingship in Canaan): 'Is Keret,
the divine offspring of El and Athirat, not immortal?! D o gods die?!' T h e compassionate Keret responds with words of comfort to his distraught son; and by way of occupational therapy counsels him to set out on a mission to sister Octavia, residing elsewhere,
and to bring her home. T o spare her sensitive feelings, the pretext
is to be an invitation to a family feast rather than a funeral. (But
since Keret's funeral will doubtless be followed by a lavish wake
cf. K T U 1.6 = R S 2. [009]+ i 18-31the lie is truly lily-white.) Ilhu
obediently complies and takes his leave.
Comment: T h e text at this point is in disarray, resulting in the dism e m b e r m e n t and dislocation of Ilhu's speech. T h e awareness of this
disturbance by a subsequent copyist led to its rewriting; but the corrected version unfortunately found its way into the second column
of K T U 1.16 = R S 3.325+, causing yet another disturbance. (2) W e
are not informed here either as to the reason for Octavia's residence
away from home or its location. If this information was not forth-

coming in a no longer extant part of the preceding text (e.g., at the

end of K T U 1.15 = RS 3.343+ iii), the poet left it to the audience
to fathom the answers from the continuation. Scene 2. After climbing a mountain and praying there (to
Baal!) for a safe journey, Ilhu sets out. Arriving at Octavia's convent-residence (hmh<m>), he squats on a nearby hillock, with the
gateway-entrance in view, to await his sister. As she emerges from
the gate, on her way to fetch water, Octavia espies her brother. In
her excitement, she drops (more likely, hurls) the encumbering vessel
in her hand and makes a dash to embrace Ilhu, her head now nestling tenderly in his shoulder. But it does not take long for Octavia
to collect h e r s e l f a n d to sense that brother Ilhu has not just droppedby for a chat. Her female intuition tells her that something is awry,
possibly relating to her father. T o her query, 'is father ill?' Ilhu
replies that, of course, all is well with the king, and that he has
come to invite her to a family party. Sensing her incredulity, Ilhu
unfolds his cover-story with lavish improvisation: it will be a sumptuous affair, attended by the gods and the who's-who of Bt-Hbr;
musical entertainment will be provided by nubile lasses singing songs
'to set one on fire'. He himself, Ilhu continues in his prevaricating
best, was asked by the king to go out and trap birds as delicacies
for mother Hry and brother Yassib; 'and since father knows that
my hunting would take me near where you live, he thought it might
be nice to extend you an invitation. So here I am!'. Octavia's response leaves no doubt as to her incredulity.
First she asks her brother to pour a cup of wine from his portable
jug. After removing the plug, Ilhu obliges. She now turns to her
brother and, in a tone mixing hurt pride with barely concealed anxiety, she asks: 'Why do you make a fool of me? H o w long has father
been ill?' Faced with such precocity, Ilhu breaks down and proceeds to tell his sister the sorry truth. Upon hearing this, the heartbroken Octavia cries out and shudders (V hi); she commences a
funereal song-and-dance around her brother. She then repairs with
him to the parental home. Upon arrival, Octavia enters, silently and abashcdly, her
father's bed-chamber. Approaching his bed, she kisses him affectionately

on his feverish forehead. She is his little 'blossom' (ib). She leaves
shortly thereafter, heeding her father's request to climb a mountain
and pray there for his recovery. 29
Comment: (1) T h o u g h not expressly stated, it is a reasonable inference from this portion of the narrative that Octavia has become a
nadtu-priestess, or nun, residing in a cloister and in the service of
the sun-goddess, Shapsh. A m o n g other things, this hypothesis will
explain (a) why Octavia is residing away from her parental home;
(b) why Ilhu does not even consider entering the gateway to notify
his sister of his arrival. As a female retreat, it is presumably off-limits
to men. (2) Octavia's funereal song-and-dance, encircling her brother,
is described in language similar to that used by Ilhu to describe the
wailing-women in Keret's house. This may help to explain the intrusion here of extraneous material originating as a (corrected) version
of Ilhu's plaintive speech to his father. (3) It is typical of Ugaritic
epic literature to portray w o m e n as superior in intellect a n d / o r
courage to men; 30 and the present encounter of brother and sister
is certainly no exception. El compares unfavourably with Athirat,
Dan'el with his daughter Pughat, and Ilhu with Octavia. 31
2.3.7 Scene 3. T h e text of this scene is very fragmentary, and
its contents consequently are obscure. T h e king's illness, like Aqhat's
murder, has resulted in drought, and the stocks of grain, wine, and
oil are depleted. A set of obscure ritual acts, intended presumably
to induce rainfall, is followed by a delegation of farmers to the king,
presumably (since the continuation is lost) to apprise him of the situation and to ask for help.


A considerable part of the text summarized above is missing, and the summation at certain points presupposes the correctness of the restorations. Cf. the discussion in M A R G A L I T 1 9 9 5 , 2 6 4 - 8 9 for this and other matters relating to this passage.
It is also not uncommon in O.T. literature: Adam is clearly inferior in intellect to wife Eve (which is why the 'wily' snake takes her on first). The same holds
true for Isaac and Rebekkah, Barak and Deborah, Sisera and Jael, Haman and
Esther, etc. A notable exception is David and Michal. The latter is possibly the
most 'trag(ed)ic' figure in the entire Bible: bright, beautiful, and courageousand
an habitual 'loser'.
Anat is only seemingly an exception: for while nominally female, she acts and
dresses like a (violent) man, and is therefore the villain of Aqhat. She contrasts both
with her virtuous brother Baal and the heroine Pughat; and the poet does not stop
short of ridiculing the penis-envy of this self-hating goddess by depriving her of the
coveted bow once acquired (KTU 1 . 1 9 = RS 3 . 3 2 2 + i).

Comment: W e have here another expression of the 'ideology of

divine kingship': the illness of the king induces a paralysis of Nature.
However, there is no more reason here than in the previous instance
to assume that this ideological stance reflects the authorial point-ofview. Like the portrait of the sick hero, and (shortly) the inept gods,
this ideology is also subject to satirization, as if what is true of the
great Baal ( K T U 1.5 = RS 2.[022]+ ii 5-7) is true of the pathetic
king of Bt-Hbr. Scene 4. T h e desperate situation created by the king's illness sets the stage for a curious development: the artisan god, Kathirwa-Hasis, whose wisdom (say the gods) is second only to El's, is
approached by a delegation of the divine assembly'El's sons' ((dt.
bn. il)and asked to take an urgent message to a hitherto and otherwise unknown character named lis and his (characteristically unnamed) wife, bearing the title ngr(t) of the H o u s e / T e m p l e of El, (var.
Baal). In the message promptly delivered by the hobbler Kathir-waHasisdescribed by the poet as running with the grace of an ass
lis is instructed to go up to the tower and to shout at the top of
his lungs'like a waterfall || like a bull' to the inhabitants of the
city. T h e sequel is lost, and with it presumably the statement of purpose, viz., a call to prayer and supplication on behalf of the dying
king and the drought-imperiled kingdom.
Comment T h e present scene, if correctly interpreted, brings the
satirical tone of the poem strongly to the fore; indeed, the satire
comes close to becoming farce. El is in deep trouble: his plan for
his protg has miscarried, and the protg himself and his famished
kingdom are teetering on the brink of disaster. El's distress signal
(which one may presume to have been lost in the lacuna at the
beginning of K T U 1.16 = RS 3.325+ iv) has been picked up by
the assembly of his sons which now goes into (pathetically ineffectual)
action. In their infinite divine wisdom they turn to the wise but crippled Kathir to deliver an urgent message to the temple crier (ngr)
or mu'addin32and for good measure, to his wifeto summon the

The semantic correspondence of Ug. ngr and the Arabic mu'addin (< 'dn (II)
'cause-to-hear, announce') is very close indeed. T h e Ugaritic term is cognate with
Akk. nagaru 'Ausrufer, Herold' (., 711). The translation 'herald' in the present
context is, however, misleading in that it implies a palace functionary charged with
making public pronouncements. Ils is rather a temple functionary; and since his j o b
is to summon the faithful to prayer, he is necessarily (a) mortal, and (b) a BronzeAge Canaanite precursor of the Islamic mu'addin.

faithful to prayer and supplication, without which the gods can do

nothing to save Keret and his kingdom. From here it is surely but
a small step to the conclusion that the 'sons of EP (notably excluding heroic and virtuous Baal) can do nothing because they are good
for nothing. Scene 5. T h e race is on to save Keret's life, for which purpose El has convened the divine assembly in emergency session. El
arrives accompanied by wife Athirat. H e opens the session with a
plea to his consort to spare Keret's life. Asherah replies that the
king's life 'is in the hands of his wife Hry': she (now that the king
is incapacitated) has to pay Athirat 'twice her weight in silver, thrice
in gold' (with the interest waived) if Keret is to be cured. Even El cannot apparently raise such a sum, much less
Keret, whose kingdom is down at the heels. T h e 'father of m a n '
turns accordingly to his sons for help; but they remain deafeningly
silent, either because they are cowed by the presence of their mother
or they are simply at a loss for ideas. El is now left with no choice
but to deal personally with the problem. His solution: to create a female exorcist named 'tqt (< etq
'(cause to) remove'), the details of which are obscured by the tablet's
poor state of preservation. But they are not beyond recovery. El creates his creature, in primordial fashion, from (red) clay (cf. J o b 33:6),
inserts snake-poison into her vagina (to thwart reproduction?), christens her over a cup of sanctified wine, and then brings her to life
by 'pouring' into her mouth '[the soul] of a god and the blood of
a [human]'. Fortified by El's blessing and directed by his detailed
instructions, Sctqt flies off to Bt-Hbr.
Comment: (1) T h e divine assembly, as we learn from K T U 1.15 =
R S 3.343+ ii, is m a d e up of the minor deities consisting of the
sonsnot including daughtersof El. It corresponds, one may assume,
to the 'council of princes' reflected in 1 Kgs. 12:8ff. T h e presence
of Athirat is thus a breach of protocol warranted by the emergency
as well as by the divine mother's direct responsibility for creating it.
(2) Athirat's offer at first sight seems to hold the key to a satisfactory resolution of the crisis. However, it must be remembered that
after fifteen births and nursing periods, the once streamlined Hry is
now bulging at both waist and bosom (cf. above, [= K T U

1.15 = R S 3.343+ iv 10-3]). (3) T h e echoes of cosmogony in this

scene are part of the parody. Like the exaggerated mobilization in
K T U 1.14 = RS 2.[003]+, the creation of S'tqt is a case of mockheroic 'overkill'. W h a t S'tqt does for Keret could, and would, have
been done by any run-of-the-mill exorcist in Bt-Hbr. El, says the
poet, may be very wise (who else knows how to create life?); but he
is sorely lacking in c o m m o n sense. Scene 6 describes how S'tqt saved Keret's life. U p o n arrival,
she immediately gets down to business. She fastens a string to the
patient's navel and places a wreath of medicinal leaves on his feverish head, while repeatedly wiping his brow of sweat. She then opens
his mouth and forces him to eatwe are not told w h a t a n d presto!
the king is well. A rejuvenated Keret promptly orders wife H r y to
slaughter a fading lamb for dinner to celebrate his recovery.
Comment T h e poet's feminist bias finds expression here once again.
Just as the w o m a n Athirat foils the male El's plan, so the female
S'tqt saves the day (while foiling the foiler). But woman's superior
intelligence, courage, and enterprise are, alas, not enough to alter
the existential 'female condition': S'tqt, like Athirat, is subordinate
and subject to El. T h e 'tree' of Athirat cannot exist without El's
fertilizing water, and S'tqt will die, without fruit, once her mission
is accomplished, as will Pughat, whose heroism is recounted in a
tale n a m e d for her b r o t h e r (in marked contrast to the 'book of
Esther'). Scene 7. T h e crown-prince and heir-apparent, whose birth,
heralded by the gods, fulfilled his father's most fervent wish, makes
his firstand lastappearance in this scene, as if to say: if you meet
him once, it is enough for a lifetime. Obviously displeased and disappointed by his father's dramatic recoveryYassib, be it recalled,
is still an adolescent!he decides to act. Encouraged by 'the fiend
at his elbow', he enters the throne-room where his father is seated
and orders him to step down, on grounds of incompetence in the
discharge of his royal duties. Keret's reaction is a mixture of rage
and anguish; and as the curtain falls there resounds a curse, at once
fearsome and funny, called down by the king on the head of his
perfidious son.
Comment: (1) Keret's recovery obviated a tragic end to the story;
but the finale is its tragi-comic equivalent. For upon hearing his son's

words, the king might well have wished he were dead. (2) Yassib is
'instructed' (V wsr) by his jinn (Ug. ggr)he is a school-boy listening to the wrong teacher. (3) T h e charges laid by Yassib against his
father are instructive for the insight they provide to the Canaanite
view of kingship. T h e most important task of the king is not to lead
in battle (which is precisely what Keret does in K T U 1.14 = RS
2.[003]+) but to administer justice fairly and compassionately (something he is never seen to do). 33 T h e Canaanite king is first and foremost a 'judge', in contrast to his Amorite counterpart whose claim
is based on personal charisma as a warrior proven in batde (gibbr
hayyil). In this sociological sense, Aqhat is older than Keret, in that the
former describes its (male) heroes, young and old, in terms derived
from the military lexicon. Keret is a mlk,H Dan'el a gzr. (4) In addition to ferocity and hilarity, the curse also contains the most important clues, suitably and cleverly embroidered into the finale, to the
identity of pseudonymic Bt-Hbr, lit., 'House-of-Union'. 3 5 T h e king
calls on 'Astarte-name-of-Baal'i.e., Baalatand Yassib's dislodged
teeth are to fall out 'altogether', for which the poet chooses the rare
(b)gbl (Palmyrene-Aramaic gbl 'community', M H e b . gbl 'to mixtogether (as porridge)', which plays on the original form of 'Byblos',
i.e. G B L / G u b l a (Heb. Gba).


The moral of the story (in sum)

2.4.1 T h e moral of the story is clear, a proud and praiseworthy

testimonial to the venerable Stoic tradition c o m m a n d i n g the allegiance of the wise throughout the ages:

Contrast the description of the (non-royal!) judge Dan'el! For all their impudence, Yassib's words thus contain a germ of truth. This motif of 'truth from the
mouth of babes' is especially prominent in Aqhat (cf. M A R G A L I T 1989, passim).
Noteworthy too is Absalom, like Yassib motivated by a desire to depose his father,
who sets himself up as a judge in the gateway, intercepting his father's 'clients', in
order to establish his credentials for kingship.
As well as (Heb. s'a) a title which he (presumably) shares with the other
members of the Bt-Hbr nobility.
Cf. Akk. /)ibru(m), a Canaanite loanword denoting 'clan' (OB) and 'in gathering (of fruit)' (LB)cf. AHw, 344. In the 11th cent. Egyptian Wen-Amon story,
f}-b-r denotes a joint commercial venture (AJVET 27, n. 17). ^l(fbr and Igbl are thus
fully synonymous terms. G r a n d the plans of gods and man,

But when the day is done
Bones broadly scattered dry in the sun,
For ironic Moira the fray hath won.
And nought remains for Apollo's progeny,
But to sing her praise
In comic agony.
2.4.2 'Life', not 'kingship', stands at the centre of Keret as it does in
Aqhat and Baal-Mot. But whereas Baal-Mot focuses on the uncanny
dialectic of Life and Death on the (awesome) cosmic plane, Aqhat
and Keret focus on the (absurd) human-life condition, the former on
its tragic aspect, the latter on the tragi-comic ('if it weren't so funny,
it would be sad . . .'). If Aqhat is a C a n a a n i t e Hamlet, Keret is a
Canaanite Merchant of Venice. Like his great English counterpart,
the Canaanite bard is a master at manipulating emotion; but to mistake him for a 'politician' (or a 'preacher'), and his art for propaganda
(or a sermon), is at once an insult and a betrayal. O u r poet- indeed any poet (of integrity)writes (or sings)
for an audience which is 'free', not 'captive'. He is by nature the foe
of tyranny, be it of the body or of the mind. He is anathema equally
in Plato's Republic and in Augustine's 'City of God'. He kneels (only)
in the Temple of Moira, at the feet of Apollo. 36

What J. H U I Z I N G A (Homo loudens) has said of 'play' is equally true of writing
poetry: 'all play is a voluntary activity. Play to order is no longer play; it could at
best be a forcible imitation of it. By this quality of freedom alone, play marks itself
off from the course of the natural process. It is something added thereto and spread
out over it like a flowering, an o r n a m e n t , a g a r m e n t . ' (Beacon ed., 1955, 7).
Elsewhere (ibid., 132) he rightly observes that poetry as such is a form of play.









W y a t t


Tablets R S 2.[004], 3.340 and 3.322+349+366, discovered in the

'High Priest's House' on the acropolis at Ras Shamra-Ugarit in 1930
and 1931, 1 were quickly established as constituting the same literary
work.- T h e mention of Danel by name in another third-season find
from the same location, R S 3.348 (IV D = 1 R p = U T 121 = C T A
20 = K T U 1.20), led to the initial incorporation of this tablet in
the series, but its successor Rpum tablets were never thus regarded,
and for practical purposes it too was eventually discarded from the
sequence. 3
N o authorship is mentioned on any of the tablets. However, the
lower edge below K T U 1.17 vi reads [ ]prln, (KTU2 prln) and is generally restored on the basis of K T U 1.6 = RS 2. [009]+ vi 5 4 - 5 as
[spr.ilmlk.bny.lmd.atn.]prln, thus restoring the n a m e of Ilimilku, the
scribe to w h o m K T U 1.1-6 = R S 3.361, 3.367, 3.346, 2.[014]+,
2. [008]+, 2. [022]+, 2. [009]+ and K T U 1.14-16 = R S 2. [003]+,
3.343+, 3.325+ are attributed (with colopha at 1.4 viii lower edge,
1.6 vi 5 4 - 8 and 1.16 vi 59 lower edge). RS 92.20 1 6 4 (as yet unpublished) also apparently bears the name of Ilimilku. In the case
of the published tablets, the script is similar in all the tablets at-

See B O R D R E U I L - P A R D E E ( 1 9 8 9 ,
systems for the texts are as follows:

26, 30-32).

The most widely used numbering






2. [004]



3 D


2 Aqht
3 Aqht
1 Aqht

C T A 17
C T A 18
C T A 19



K T U 1.17
K T U 1.18
K T U 1.19

T h e tablets are located as follows: K T U 1.17, 19 in the Louvre (AO 17. 324 and
A O 17.323 respectively), K T U 1.18 in the British Museum (AO 17.325 = BM
Published by V I R O L L E A U D 1936a (Editio Princeps).
Typical expressions are 'clearly at least one further tablet must have followed':
G I B S O N 1 9 7 5 , 6 6 . 'At least four tablets': DE M O O R 1 9 8 7 , 2 2 4 . P A R K E R 1 9 9 2 , 9 9 ,
1 3 4 - 5 , still evidently had a fourth tablet in mind, but refrained from identifying it
with K T U 1 . 2 0 = RS 3 . 3 4 8 . See also P A R K E R 1 9 9 7 , 4 9 .
Provisionally K T U 9.432. See C A U O T 1992, B O R D R E U I L 1995a, 2.

tributed to Ilimilku, so that the identification is reasonable. 5

As will be clear from the synopsis below, considerable portions of
Aqhat are missing. K T U 1.17 is a tablet originally of six columns,
of which two are entirely missing, while the beginning and end of
the four surviving columns are missing (an estimated twelve or so
lines in all on the recto, and on the verso, with a shearing break
down the upper surface, perhaps twenty or so), with the wedgeshaped breaks on columns i and vi resulting in even further loss.
K T U 1.18 is a tablet of four columns originally, of which two are
entirely missing. Again, the beginning a n d end of the surviving
columns are missing (some twenty-five lines or so), and wedge-shaped
breaks further reduce the surviving text, with no complete lines in
col. i, and only nine complete in col. iv. K T U 1.19 is the best preserved of the three; the only substantial losses here result from friable edges along the three sections into which the tablet has broken,
with the additional loss of a central section in col. i, the bottom
right hand corner of col. ii, the top corner of col. iii, and a small
vertical section in the lower part of col. iv. T h e surface is also eroded
at the beginning of col. i and in the upper central section of col.
iv. At a rough estimate, from these three tablets slightly over half
the lines are missing, say approximately 430 from an original 840
(fourteen columns of ca 60 lines each), or fifty-one percent. While
these figures are not set in stone (or clay), any adequate interpretation of the text must remain provisional, having to take into account
the fact that it can never tell more than half the story.


Synopsis of the story

Substantial portions of the text are missing, as we have noted. T h e

following narrative sequence can however be understood.

Danel has no son and so performs devotions to obtain divine

assistance. For six days he sacrifices to the gods, apparently
spending every night in the temple. O n the seventh day Baal
intercedes for him, asking El to provide a son who will perform all the filial duties necessary for a father to be blessed.
El blesses Danel and promises . . .

For recent discussion on the programme of Ilimilku see W Y A T T 1997, 1998a

and below in this chapter. I also enlarge on his significance at 13.4.2 below.









a son. T h e son is probably born in the missing section

between the columns, and news is brought to Danel, who
rejoices and holds a feast in his palace for the goddesses of
Kothar arrives, bringing a composite bow as a gift. He is
feasted/' and the bow is given to A q h a t . . .
A feast is apparently taking place. Anat asks Aqhat to give
her the bow. H e tells her to take the raw materials to
Kothar: he will make her one. She persists in her demand,
and offers him immortality. Aqhat tells her to stop lying,
and says that a bow is in any case a man's weapon. She
departs to El in a rage, accusing Aqhat of impiety.
Anat threatens El that she will be violent if she does not get
her way, and he gives her a free hand. She approaches Aqhat,
seemingly mollified, inviting him to go hunting with her.
After a successful hunt, Anat summons Yatipan, instructing
him to assume the form of a falcon, and pounce on Aqhat,
killing him. H e does so . . .
and the bow falls into the river and is shattered. Anat mingles regret for the bow with a savage assault on Aqhat's
corpse, which she tears in pieces. Back in his capital, unaware
of what has happened, Danel sits to dispense justice; Pughat
sees the withering of the plants, and sensing a disaster, rips
Danel's cloak. H e utters a curse.
Danel, still evidently not appreciating the situation, wishes
that his son would harvest the now shrivelling grain. T w o
messengers arrive and tell of Anat's complicity. D a n e l . . .
then curses the falcons. As each falls from the sky in turn
he examines its entrails. Those falcons found to be empty
are healed. Finally he finds Aqhat's remains in Sumul's
stomach, and buries him. H e then goes round the country,

This is the folk-tale idiom for worship with sacrifices, but uses the figure of
face to face, person to person, communication between deity and devotee.


cursing all the villages in the vicinity of the murder. Aqhat

is mourned for seven years, Danel concluding the rites with
a sacrifice. Pughat then dons a r m o u r beneath her woman's
clothes, and sets off to find Yatipan. Already half drunk,
and thinking that she is Anat, he demands wine, and while
she plies him with it boasts of his exploit. . J

As can be seen from this synopsis, there are tantalizing gaps in the
narrative. Particularly u n c l e a r a n d consequently open to variations
in reconstructive guessworkis the sequence of events in K TU 1.18.
T h e synopsis above represents this author's reading of the story.
Again, the last column of K T U 1.19 breaks off at the most inopportune moment. It is commonly supposed that Pughat went on to
kill Yatipan, which would provide a relatively satisfying dnouement,
but would leave Anat, the true villain of the piece, unscathed. O n
the other hand, as pointed out below, 8 this is to apply our moral
expectations to a divine power, and may misread the author's intention. W h e t h e r or not K T U 1.20-22 = R S 3.348, 2.[019], 2.[024]
have any close relationship with Aqhat must remain an open question. But while they are narrative in form (and they appear to be
three versions of substantially the same narrative, though Pitard opines
below 6.4that tablets K T U 1.21 and 1.22 = RS 2. [019], 2. [024]
may be two parts of one whole), we cannot assume that the mention of Danel proves a link, since a n u m b e r of stories may have
been attached to the same figure.

T h e following translations have been published: V I R O L L E A U D 1936a, G A S T E R

1936, 1937, 1938. 1950, 257-313, 1961, 316-76, C . H . G O R D O N 1949, 84 103,
1977, 9-29, FRONZAROLI 1955a, D R I V E R 1956, 48-67, J I R K U 1962, 115-36, AISTLEITNER
1964, 65-82, G I N S B E R G 1969, 149-55, C A U O T - S Z N Y C E R 1974, 401 58, C L E A R
1976, 5 0 - 6 9 . X E L L A 1976, 1982, 193-216, G I B S O N 1978, 103-22, C O O G A N 1978,
27-47, M A R G A I . I T 1989a, DEL O L M O I J T E 1981a, 327-401, DE M O O R 1987, 224-66,
A I T K E N 1990, BALDACCI 1996, 333-65, P A R D E E 1997a, 343-56, PARKER 1997, 49-80,
W Y A T T 1998c, 245-312.
Other studies include CASSUTO 1938, BARTON 1940, STOCKS 1943, GINSBERG 1945a,
1945b, O B E R M A N N 1946, H E R D N E R 1949b, G R A Y 1957, 73-91 = 1965, 106-26,
EISSFELDT 1966, K O C H 1967, K A P E L R U D 1969, 70-82, H I L L E R S 1973, DIJKSTRA DE M O O R








1 9 7 5 , WATSON



1979, M A R G A I . I T 1981a, 1983a, 1983b, 1984a, 1984b, 1989a, DEL

O L M O I J T E 1984a, 115-42, CAHJOT 1985, 1987, 1990, PARKER 1987, 1989, 99-144,
C O O P E R 1988, DE M O O R 1988a, A I T K E N 1989a, 1989b, 1990, HUSSER 1995, 1996.
For a fuller bibliography down to 1988 see M A R G A L I T 1989a, 503-6.


History of interpretation*

published the editio pnceps of the Aqhat tablets.

H e accepted Ilimilku's authorship of the present narrative, whatever
their antecedents, 1 0 and dated the tablets to the mid-fourteenth century ( 1 9 3 6 , 8 2 ) . O n genre he was imprecise, referring to 'legends'
( 1 9 3 6 , 8 3 ) and to 'mythological texts' ( 1 9 3 6 , 8 5 ) , or even both together
( 1 9 3 6 , 1 0 9 ) without demur. By ordering them as he did (n. 1) he
inevitably missed the logical progression which was subsequently
recognized. M a n y of his explanations of the vocabulary have had to
be revised, but his study is not to be underestimated as important
pioneering work, however m u c h may now be of primarily historical
interest. An interesting instance is his discussion ( 1 9 3 6 , 8 7 - 9 6 ) of the
names of the characters. H e concluded (p. 96) that Aqhat is 'a member of the family of the god of death; he is, in effect, one of the
gods who dies . . .'; he resumed this argument later (p. 110), claiming that Mot 'personifies the ripe ear of wheat', with the result that
Aqhat is also supposed to be 'the harvest-genius'. T h e r e was an unfortunate tendency to draw conclusions of this kind in early Ugaritic
scholarship, which it took decades to escape. Everything was allegorical! T h e result was the wholesale application of various permutations of the seasonal interpretation (a variant on the myth-and-ritual
theme) to all the larger compositions, with a consequent delay in
the recognition of more balanced assessments. Virolleaud also understood the terms qst ('bow') and qs't ('arrows') to mean 'chalice' and
'vases' respectively," which rather destroyed the symbolic centrepiece
of the whole story.


in a brief note on the text, placed the tablets in

what is now regarded as the correct order, and recognized Aqhat's
h u m a n nature. B a r t o n 1 9 4 0 drew attention to the apparent links


There is insufficient space to offer a complete survey here, and much of the
earlier discussion is in any case now outmoded. I shall therefore merely oudine one
or two salient features of early discussion, and concentrate on later work. For literature before their respective publication dates see also the surveys in CAQUOT SZNYCER - HERDNER 1974, 401-15, DEL OLMO LETE 1981a, 327-401, and MARGALIT
1989a, 3-92.
He writes of them being 'redacted in the fourteenth century' and of a considerable lapse of time between their original formation and reduction to writing
by Ilimilku (VIROLLEAUD 1936a, 83). This dating is now in course of modification.
See PARDEE 1997c, 376 n. 2, and below, 13, nn. 284, 289, 311.
VIROLLEAUD 1936a, 117, 203-5.

of the narrative with the Galilee region, 12 while G i n s b e r g 1945a,

1945b, recognized that Danel was a king, and in his detailed treatment of a n u m b e r of key passages broadly set Aqhat studies in their
present mode.
Gaster developed his views through a n u m b e r of articles and two
editions of Thespis ( G a s t e r 1936, 1937, 1938, 1950, 257-313, 1961
[1966 printing], 316-76). We may take his final account as his considered view. H e treated Aqhat as myth, and classified it as 'the disappearing god type'. It is a purely literary work as it stands, but with
its roots in ritual drama: 'it was, au fond, nothing but an artistic transformation of the time-honored seasonal drama'. 1 3 After offering a
synopsis of the narrative, in which the reader may feel uncomfortably that he is being led more by rhetoric than by hard facts, he
launched into his interpretation (1961 [1966], 320-7).
'If our basic approach is correct,' he averred, 'this story will go back
to a primitive seasonal myth relating how a mortal huntsman challenged the supremacy of the goddess of the chase and how his subsequent execution for this impiety caused infertility upon earth.'
H e went on to invoke T a m m u z , Osiris, Adonis 'and the like'. We
can see the patterning process at work. Only the prior assumption
of some seasonal theory (probably also misrepresenting fundamental
elements in these traditions too) could justify a parallel treatment.
More substantial, however, was his invocation of the Orion myth.
This is indeed a widespread tale (in my view possibly quite independent of T a m m u z and company), of the hunter who confronts
and insults a goddess or is in some way brought to her attention. 14
Unfortunately, Orion is the subject of a large n u m b e r of myths, and
not one of them corresponds very closely to the plot of Aqhat. It
requires a synoptic approach to discern any extensive parallels between
what are essentially variations on a theme. Furthermore, Orion is
inseparable from the constellation of the same name, while Aqhat

O n e of the grounds for considering that the rulers of Ugarit were originally
from the Hauran-Galilee region lies in the familiarity of the tradition with the
toponyms of the region. See discussion, with further references, in M A R G A L I T 1989a,



1961 (1966), 316.

See also G R U P P E 1 9 0 6 , i 6 9 - 7 0 (cited Gaster),

i 1 5 1 - 4 ( 4 1 ) , and A S T O U R 1 9 6 7 , 1 6 3 - 7 5 (discussed
below). Graves' explanation of Aqhat ( 1 9 6 0 , i 1 5 3 - 4 ) , apart from calling it a Hittite (!)
myth, interprets it astronomically.





FONTENROSE 1 9 8 1 , G R A V E S 1 9 6 0 ,

has no obvious links with the stars; 15 and while it would be nice
to find them, we should beware of assuming them on the basis of
'parallels' some centuries younger. But Gaster's work on this aspect
deserves more extensive re-evaluation as our knowledge of the stellar dimension to U g a n d a n religion, now known only fragmentarily,
develops further. But even establishing a stellar basis does nothing
for the explication of a text from which any such putative elements
are now clearly missing. 16
D r i v e r (1956, 8) gave only a slight treatment of the significance
of the story. H e stated that the theme of Aqhat 'is a righteous king's
need of a son', but a couple of paragraphs later wrote that 'the main
theme of the myth is clearly the death and resurrection of Aqhat',
thus introducing a new perception of what concerned the narrator,
before concluding that in view of the damaged condition of the material 'no satisfactory interpretation of the myth is possible'! This final
assessment is certainly the most cautious. But it should be noted that
Driver raised two interesting issues: the problem of whether Danel
was a king, and the death and resurrection motif.

(1967, 163-75) referred to Gaster's treatment of Aqhat in

relation to Orion, but, without discounting it and noting its Mesopotamian antecedents, argued that a much closer figure for fruitful
comparison is the Greek Actaeon. H e noted that Actaeon's mother
Autono was daughter of C a d m u s and H a r m o n i a , thus evincing a
Semitic pedigree, since C a d m u s and the whole Boeotian tradition
reflect West Semitic influence. H e argued (p. 165) that the two
names, Aqhat and Actaeon, are related, and that not only are both
torn asunder, one by 'eagles' (rather falcons), 17 the other by dogs
(falcons and dogs are animals in the service of hunters), but (p. 167)
that in both stories there is a seasonal element, the fifty hounds of
Actaeon representing the cycle of the year, while Aqhat's death provokes a severe drought. Like Gaster, he went on to suggest a stellar element behind the Orion parallels (p. 168), and noted that the
latter's name, too, is susceptible of a Semitic etymology (y'Sr, 'r) and
may even appear in the form aryn as an Ugaritic personal name. 1 8

G A S T E R ' S (1961 [1966], 322) linking of the bow with the constellation of Canis
major is certainly intriguing!
"' T h e only hint at a richer background is Pughat's epithetal yd't hlk kbkbm, 'who
know(s) the courses of the stars', K T U 1.19 = RS 3.322+ ii 2 - 3 etc.


WYATT 1998C, 2 8 4 a n d n .





1967, 27, 220,


'Hurrian?': DLU,


He further argued for a link between Sumul {ml), 'mother of the

eagles (falcons)' and Greek Semele. 19
C a q u o t - S z n y c e r ( 1 9 7 4 , 4 0 9 ) drew attention to the incomplete
state of the text and advised caution in interpretation. Critical of
mythological and seasonal interpretations, they insisted (p. 413) that
the story was neither history, not historical epic, nor myth, and was
not the subject of seasonal, national or exceptional recital. T h e y
found a definition of its genre elusive, but compared it with Gilgamesh, Adapa and Etana, ragarding it as an Ugaritic 'classic' (conveniently vague!). They did not use the term, but to judge from their
treatment (p. 414) appear to have assessed it as wisdom literature. 20
G i b s o n ( 1 9 7 5 ) set out to clarify thinking on the nature of myth
and other genres. Noting that one of K i r k ' s ( 1 9 7 1 , 2 6 8 ) features of
myth was the fantastic dimension, he noted such features in Aqhat
(and Keret), but added that 'a speculative or perhaps better, an ideological bias' (p. 62) should be present for a narrative to qualify as
myth. But he denied any link between the present narrative and Ugaritian royal ideology. T h e scenes such as the confrontation between
Anat and Aqhat, which he considered to have an ideological dimension, he regarded (p. 67) as 'secondary, supplying for all their vigour
only the backcloth against which Daniel's piety is put to the test'.
H e further opined that the bow too was a secondary feature, and
that in a putative fourth tablet Aqhat was finally restored to life.

dealt briefly with Aqhat in the broader context of ancient Near Eastern tales. Her treatment was too cursory
to contribute much to the discussion, but she served the useful purpose of highlighting the conventional folklore motifs to be found in
the story.- 1
In his edition of the texts, d e l O l m o L e t e 1 9 8 1 offered an extensive analysis of Aqhat. H e classified it as 'epic', along with Keret, and
ran through the scenes, analyzing the literary sub-type of each episode,
drawing on parallels in biblical and other ancient Near Eastern literatures. In reverting to a general statement of the text's 'sense and
function' (pp. 3 5 4 - 6 4 ) , he reiterated its epic nature, judging it however





read the


sml at





but this is now dis-

M A R G A L I T 1989a, 58, characteristically summed up their exposition as an
account of 'a pastor or a priest. . . [who] teaches platitudes'!
She identifies them according to T H O M P S O N 1 9 5 5 - 8 .

more 'mythical' than Keret, since the deities are more involved as
dramatis persona, and not merely invoked in conventional religious
terms. ' T h e gods avenge themselves' in response to Aqhat's insolence, he stated (p. 355), discerning a general theological argument
here, and the supreme god 'has to yield to the caprice of an inferior deity'. 22 This theological quality makes it difficult to estimate a
historical basis for the story, as though that were desirable. Del O l m o
Lete attempted to give a serious theological account, but his assessment of 'the caprice of the gods, their amoral conduct' (p. 356)
seems to me to misconstrue the significance of mythological 23 action.
H e later (p. 358) drew attention to another theological point, the
contrast between Danel who is obedient to the gods and receives a
positive response, and Aqhat who confronts them and is accordingly
rebuffed. But while this is true in terms of narrative device, and of
psychology and pastoral theology, it perhaps disguises the real problem on a purely metaphysical level, which is that the different deities
encountered in the story are quite differently motivated in their relationships with humans. T h a t is, the deities as reifications of certain
metaphysical principles are credited with their own motivation, which
operates independently of immediate h u m a n motivation. Anat is after
all, as goddess of war and hunting, by nature vicious, pitiless and
scheming. T h a t is the role she is constructed to play. We are left
feeling that no a m o u n t of diplomacy on Aqhat's part would have
saved him. Certainly no a m o u n t of piety on Danel's part does him
any good.
Margalit has written a n u m b e r of studies on Aqhat, culminating
in his large-scale commentary ( M a r g a l i t 1981a, 1983b, 1984c, 1989),
the first study on a single narrative from Ugarit on this scale. 24 This
is extremely thorough, but is a very difficult volume to work with,
in view of its division into separate blocks dealing with the same

For my slightly different assessment of the general theological principles at
work see below,
I am writing here of the mental disposition rather than the literary genre,
though the two naturally overlap. As though grappling with this issue, del Olmo
Lete (p. 356 and n. 90) writes that Aqhat is 'nearer "myth" than the "epic of Kirta"',
and (n.) '[Aqhat's] intermediate position between saga and myth is recognized'.
'Caprice' is also a term appearing in de Moor's assessment (below). For further
observations on the adequate assessment of myth see 13.4 below.
The studies on the Baal cycle by DE M O O R (1971) and VAN Z I J L (1972a) are
not formal commentaries in the same sense, and in any case the Baal material is
altogether more heterogeneous than Aqhat or Keret.

materials from different perspectives and no adequate cross-referencing aids. Trying to achieve this during use is a taxing occupation. It begins (pp. 3-92) with a useful survey of previous work,
though this is perhaps excessively negative, not to say waspish, in its
assessment of others' efforts. 25 T h e r e follows a prosodie analysis (pp.
93-105), separated by nearly four hundred pages from the appendix on the principles of Ugaritic prosody (pp. 495-502). It is fair to
say that Margalit has ploughed a lonely furrow on this topic, for
few have expressed support for his approach, or are as optimistic as
he that he has solved the considerable problems the topic raises.
Following the initial position-statement on prosody, he offered a
textual analysis (pp. 107-14), followed by a text layout (unvocalized,
pp. 117-41); this is followed in turn by a translation (pp. 143-66),
then by textual and epigraphic notes (pp. 167-246), and finally by
a literary commentary, prefaced by short units of the unvocalized
text (again!) and punctuated by excursi on various topics (pp. 2 4 7 469), before an exegetical overview, a brief statement on Ugaritic
literature and the Hebrew Bible and the final appendix. T w o theoretical positions dominate the work, the non-royal nature of Danel
(on which see further below), and the so-called 'Kinneret hypothesis', according to which the narrative is at home in the Galilee region.
He even considered the Kinneret to be an actor in the drama (p. 411:
'the personified Kinnereth, "unwilling" to disclose the identity of the
assailant, must be punished as "accessory after the fact"'). This seems
a trifle excessive.
Parker has written two studies ( P a r k e r 1987, 26 1989, 99-144) on
Aqhat. In the former, he deplored the atomistic nature of previous
philological approaches to the poem, and the patternistic bias of religious approaches. T h e time had come for a literary approach. While
caution must be urged in view of the fragmentary nature of the text,
a useful approach, on essentially form-critical terms, was the establishment of the types of traditional material employed. 27 'Hypothetically


The treatment of K A P E L R U D 1969 encapsulates this rather well: 'Author (sic)

states at the outset (p. 70): "The Aqht text is still an enigma, and so far no satisfactory solution of its problems has been found". Had he limited himself to this
statement, the net balance of author's contribution would have been more positive
than it is in fact.' Some put-down!
Published in the M.H. Pope Festschrift ( M A R K S - G O O D 1987). Given as an
SBL paper in 1980, and discussed briefly in M A R G A L I T 1989a, 7 1 6 .
Cf. discussion of del Olmo Lete above.

any consistent thrust uniting those peculiarities may be treated as

the theme of the whole' (p. 71). H e isolated five main sections, and
treated each in turn. These are as follows, in his treatment.
A), the birth of Aqhat, deals with the familiar theme of the childless hero who appeals to the god for help; the god responds, and
the child is born. T h e Egyptian tale of the Doomed Pnce and the
Hurrian Appu story, the story of H a n n a h and Samuel in I Samuel,
and the Ugaritic Keret story, are cited as comparable examples of
the type.
B), the bow of Aqhat, describes the making and delivery of the
bow. T h e account of Kothar's visit has analogues in Genesis 18:116
and 19:1-16. Sharing certain traits is 2 Kgs 4:8-17, and a modified
version of the form appears in 1 Kgs 17:916.
C), the death of Aqhat, describes Anat's coveting of the bow, her
overtures to the hero, and confrontation with El when rebuffed, culminating in her being given a free hand in accomplishing Aqhat's
death with Yatipan's help. Comparison is made with the hero with
Ishtar in Gilgamesh, both episodes deriving from an older Vorlage, and
also with Anat's dealing with El in K T U 1.3 = R S 2.[014]+. 2 8
D), the consequences of Aqhat's death. Parker noted that the narrative movement almost comes to a standstill in this section, apart
from describing a n u m b e r of ritual activities which accompany the
inevitable environmental consequences of Aqhat's murder. Awareness of Aqhat's death (as the cause of drought) dawns only slowly,
and then Danel's curses are directed towards the birds who have
devoured his son, 29 and to the cities held responsible for unresolved
homicides in their neighbourhoods. N o similarly extended parallels
from ancient near eastern literature are cited.
E), Pughat's mission of vengeance. Comparisons scholars have
made with the stories of J u d i t h and Jael and Sisera are noted. While
the latter connection is discounted, extensive similarities with the former are discussed.

P A R K K R 1 9 8 7 , 7 7 , notes that while the language in Baal and Aqhat is remarkably similar, the theme of the goddess' insubordination before the high god is far
less apposite in the former. He adduces a closer relationship between Aqhat and
P A R K E R 1 9 8 7 , 7 9 , appears to hold the father of the raptors responsible. In fact
it is their mother, Sumul, who is so described, in K T U 1 . 1 9 iii 2 8 - 3 9 . The birds
are identified as vultures, p. 78. For the present author they are rather falcons.

Parker concluded that Aqhat would have originally ended with an

account of the fulfilment of Pughat's vengeance, and a return of fertility to the land, 30 but with no reference to Aqhat's restoration to
life. H e ended with an assessment in which the mythological emphasis drawn by previous scholarship was played down, while the social
dimension was highlighted as the main theme of the author's intention. H e made some interesting observations on gender roles, contrasting Anat's 'innate and blatant masculinity' with Pughat's 'assumed
and concealed masculinity' (p. 82). his conclusion raises a n u m b e r
of questions (expressed rhetorically) rather than providing answers
for them. H e saw the possibility of a critique of the values of the
contemporary monarchy and administration.
Useful as this analysis of the structures and congeners of the Aqhat
story are, it does not actually tell us much about the moral or ideological dimensions which may lie behind it (that is, the author's
intention). If his final questions had been answered, we might have
some clear idea where Parker stood. Margalit, though too harsh in
his judgment, 3 1 is perhaps justified in complaining that this tells us
more about comparative literature than about Aqhat. It certainly
shows the relatively sterile nature of analysis which gives no account
of why an author works in this or that way, beyond the fact that it
all boils down to 'traditional themes'.
Parker returned to the topic in a further study ( P a r k e r 1989).
Here he set out the broad characteristics of Ugaritic narrative verse,
as it was evidenced in particular in the Keret and Aqhat stories. H e
then turned to Aqhat itself (pp. 99-144), and outlined much the same
discussion as above. His conclusion was extended to a demonstration of how, while drawing on common mythic and legendary themes,
the author(s) ('composers') have, 'by adopting, transforming and
combining several different traditional narratives, produced a larger
work of striking unity' (p. 142). Again he emphasized the familial
values promoted in the story, as distinct from conventional mythological themes, raised the question of authorial motive, and now suggested that as a piece of 'classical' literature in Ugarit, Aqhat may

Does he mean an element of the 'fertility cult' here? He does not say. At
most, what can be said is that fertility represents divine blessing, while sterility is
the outcome of a curse, and in broad terms fertility also has to do with royal power
and its effective implementation. The loss of a prince is a threat to a kingdom.
However, P A R K E R notes ( 1 9 8 7 , 8 3 ) the lack of overt emphasis on royal issues.
M A R G A L I T 1989a, 72.

have afforded its readers and hearers the opportunity of seeing themselves mirrored in the world of the story, 'a satisfying portrayal of
life in an idealized past era, a life with its own tragedies, but also
with its own orderly and beautiful institutions that in the end prevailed' (p. 143).
In his translation of the texts, d e M o o r 1 9 8 7 , 2 2 4 - 6 6 , made a
n u m b e r of comments on the literary features of the story. In keeping
with his broader assessment of Ugaritian theology, 32 he saw Aqhat as
dealing 'with life and death, and with the fate of m a n who all too
often appears to be the victim of divine caprice'. Read in the light
of his earlier and later treatments of Ugaritic theology, this is not a
perspective to be taken seriously, since he appears to have envisaged
a culture incapable of the moral insight to question its own bankrupt theology. T h e authors are thus as benighted as their literary
characters. At best Ilimilku reflects a disillusioned and pessimistic
oudook supposedly typical of the Late Bronze Mediterranean world. 33
Into this scenario d e M o o r (1990, 97 = 1997, 99) wove an argument developed some years earlier ( d e M o o r 1988a), discerning in
Aqhat a further outworking of the seasonal pattern he had previously
argued to be the foundation of the Baal Cycle ( d e M o o r 1971). In
the 1988 article he expressed the principle thus: 'Ilimilku . . . deliberately wove a seasonal pattern into the Legend of Aqhatu out of
his conviction that life on earth revolves according to a circular pattern that had been laid down in the pristine age of myth' (p. 61).
H e then proceeded to fix episodes in the narrative in sequence
through the calendrical year in the same m a n n e r as had been done
for the Baal cycle. T h e substantial objections raised by some scholars to the seasonal interpretation were dismissed as of no consequence ( d e M o o r 1988a, 75 n. 6).
(1990) 34 offered a very thorough analysis of the narrative
from a folk-literature perspective, drawing on the work of Propp,
Dundes and Dolezel. H e saw the narrative structure in terms of a
series of different thematic levels, and of alternating patterns; 'lacks'

D E M O O R 1986b, 1990, 42-100 (= 1997, 41-102). For my views on this issue
see 13.3 below.
See DE M O O R 1990, 99 (= 1997, 101). In my view the observations made here
result from a mistranslation of K T U 1.19 ii 34-36. For my translation see W Y A T T
1998c, 301.
This is the published form of an Edinburgh PhD dissertation from 1978. The
latest entries in the bibliography are from 1984.

being 'liquidated' (not the most apposite term in view of Aqhat's liquidation!) as desires were met or situations reversed (e.g. a son for
the hero, a bow for the hero, the bow for Anat, and so on) or a
status quo maintained. A m o n g the oppositions a set of equivalences
(called 'the synonymous sequence') is also developed, and periodic repetition (e.g. searching the falcons' gizzards for the remains of Aqhat
maintain tension and development to a climax. Aitken was able to
achieve this, quite legitimately, in spite of the considerable gaps in
the narrative, and showed the tight construction of the surviving text,
and, as he put it (p. 206), 'of the skill and artistry of its narrator,
the Ugaritic teller of tales'.


Some recurrent and unresolved issues in Aqhat

A n u m b e r of individual episodes and themes in the story have been

the subject of particular discussion.

The Incubation theory

Like O b e r m a n n (1946), Gaster interpreted the temple episode ( K T U

1.17 i) as an incubation scene, 35 as did Gray, 3 6 del O l m o Lete 37 and
Parker. 3 8 This view has however been persuasively challenged by
M a r g a l i t
1989a, 2 6 0 - 6 , and by H u s s e r 1992, 2 9 - 6 2 , 1996, 9 3 - 5 ,
who marshall substantial arguments against the incubation interpretation. Margalit, citing H a m i l t o n ' s (1906) study, observes that
none of the conditions required is fulfilled: chthonian gods are not
involved, no illness is involved, there is no reason to think that
Danel's sleeping is part of the ritual, there is no direct theophany,
and no cultic personnel are involved. Furthermore, no other ancient
Near Eastern candidate fulfils the conditions either, and we are left
with a late hellenistic institution with no obvious points of contact.
Husser's original discussion was complex and extended, taking Oberm a n n on at every juncture. His later paper summarized his main
findings, broadly in accord with Margalit. H e noted that it was gaps
in the text, filled out in academic imagination, which appeared to
justify the incubation interpretation. Furthermore, it was not to Danel


GASTER 1 9 6 1 , 3 1 6 .
GRAY 1 9 6 9 , 2 9 6 .
D E L OLMO LETE 1981a, 3 3 2 - 3 ,
PARKER 1 9 8 7 , 7 2 ; 1 9 9 2 , 1 0 0 .

1984a, 119 2 0 , 1984b.

that Baal drew near (in a theophany), but to El, to whom he speaks
about Danel in the third person. If 'incubation' is to be used to
describe the scene, it requires a considerable extension of the classical meaning of the term.

The occasion of Aqhat's birth

W h e n was Aqhat born? T h e conventional interpretation of K T U

1.17 ii has been that it is part of the build-up to the account of the
birth of Aqhat, which must have been narrated in the gap following. This is explicitly stated, for instance, by M a r g a l i t 1989a, 147,
and in some other discussions appears to be assumed, though not
spelt out (e.g. P a r k e r 1987, 73). But a convincing case has been
m a d e by Caquot - Sznycer, and developed further by Husser, that
the birth must have taken place in the gap between the end of
col. i and the beginning of col. ii.39 T h e later presence of the Kotharat
is to be understood, no doubt, as for that purpose, but they evidently delay for some days after, perhaps to confirm a safe birth
and the healthy state of the child. O n this alternative interpretation,
the arrival of the Kotharat and the counting of days and months in
K T U 1.17 = R S 2. [004] ii refer not to conception and pregnancy,
but to the immediate post-natal period and the child's infancy. 40

Aqhat's encounter with Anat

T h e encounter between Anat and Aqhat in K T U 1.18 i (in particular 1. 24) has also given rise to much discussion. T h e key part of
the text reads at.ah.wan.x[ ]. It has been frequently restored, to read
at.ah.wan.a[htk].41 Was there a sexual encounter between Aqhat and
Anat? This was asserted by A l b r i g h t (1944, 33-4), but rejected by
G i n s b e r g (1945b, 19). T h e best argument against the c o m m o n view,
i.e. that there was, is that of D r e s s l e r 1979, followed by X e l l a
1984b. I have accepted this view. 42
















Note the pluperfect sense of 'rb bbth ktrt, 'the skilful goddesses had entered his
house', in K T U 1 . 1 7 ii 2 6 : H U S S E R 1 9 9 6 , 9 1 - 2 , W Y A T T 1 9 9 8 C , 2 6 4 and n. 6 1 .
Thus KTU1.
Instead of 'You are my brother and I am your sis[ter]', the passage is to be
read 'Come, brother, and I shall [ ]'. The putative reading a[htk\ is impossible. See

1998C, 2 7 9





Danel's social status

There has been some debate as to whether Danel is a king. G i n s b e r g

(1945a, 4 n. 6) noted that Danel sits at both threshing-floor and city
gate to dispense justice, citing 1 Kgs 22:10, 2 Chr. 18:9 as parallels, and also noting the widespread ancient Near Eastern use of the
royal theme of caring for widows and orphans. G i b s o n (1975, 66)
remarked that 'Danel is only once called a king', as though apologizing for the author's loose use of language! He preferred to see in
Danel a patriarch of the J o b a n or Abrahamic kind. T h e formal royal
view has not gone unchallenged, as noted above. M a r g a l i t (1989a,
2 5 3 - 4 , 278, 292-3, 309, 361 2, 410, 424-7) in particular has been
most persistent in arguing that Danel is not to be seen as a king.
His arguments (1989a) were as follows, though the case was m a d e
more by assertion than by demonstration. A premonarchical society
was envisaged by Margalit (p. 309), in which 'notables' dispensed
justice. Danel's 'political status is that of unus (doubtfully primus) inter
pares. H e is one of the "city elders" . . . ' (p. 361). Finally, he did not
recognize the sense 'king' of mlk in K T U 1.19 = RS 3.322+ iii 46.
He construed it (1989a, 163, 410) as 'your down-course' (Ar. may I).
In view of Gibson's comment, it is worth noting that with substantial portions of the text missing, not too much should be made of
the apparent hapax appearance of mlk here. Margalit's case can hardly
be said to be very strong, and much rhetoric and repetition can
hardly substitute for reasoned discussion.
T h e counter-arguments are as follows. Firstly, the title mt rpi, however it is taken, has a royal significance. In my translation (1998c,
250 n. 5) I have taken it in the sense ' m a n (i.e. ruler) of R a p h a ' .
T h e alternative sense is to take it as a promise of Danel's later (postmortem) incorporation into the rpum (deified dead kings of high rank).
T o this cf. the element in the blessing of Keret which foresees his
inclusion a m o n g the rpum, K T U 1.15 = R S 3.343+ iii 2 - 4 , 13-5).
Secondly the blessing formula restored at K T U 1.17 i 3 4 - 6 , on
the strength of K T U 1.15 = R S 3.343+ ii 16-20, is to be construed
as explicitly royal in its object. As Jackson and Dressier have noted,
there is a close affinity between the scene described and the Ugaritian royal seals. 43 Within this formula, the form cbd reappears (pace
Margalit), with reference to Danel, at K T U 1.17 i 36 (previously




1975. See also WYATT 1997, 787-9.

misread as .bdh, with preceding word-divider), and this too is an

explicitly royal title.44
Thirdly, Danel also performs an action reminiscent of El himself
at K T U 1.17 i 10-11, in placing his feet on his footstool. This is
to be understood as an accoutrement of royal rank (cf. Ps. 110:1).
Fourthly, the description in K T U 1.17 i of Danel 'enrobed' (uzr)
for the performance of his devotions may be an allusion to the ritual garment in which Ugaritian kings are shown robed in various
representations. 4 5
Fifthly, as already observed by Ginsberg (above), Danel's sitting
at the city-gate to judge, at K T U 1.17 4 - 8 , 19 i 19-25, and particularly the reference to widows and oiphans as the beneficiaries of his
dispensation of justice (a clich for royal justice throughout ancient
Near Eastern literature; contrast Keret at K T U 1.16 vi 3 3 - 4 , 45-50!)
is the fulfilment of a specifically royal duty. Margalit's attempt to
democratize this is without substance.
Sixthly, the use of hkl at K T U 1.17 i 26, 43, ii 25, is most reasonably to be construed as denoting a royal palace. T o accept it as
less is to require that the term is used as a sustained hyperbole. T h e
former alternadve is altogether the more natural. While it is of course
the pair-word to bt, it is used elsewhere in the Ugaritic corpus only
of kings' and gods' 'houses', that is, palaces in the first instance and
temples in the second.
Finally, the form mlk occurring at K T U 1.19 iii 46 is most reasonably to be taken to m e a n 'king', with reference to Danel himself, as the subject of the verb in the sentence. T h e syntax raises no
problems, and an appeal to Arabic cognates is unnecessary.
These pieces of evidence represent a prima facie case for Danel's
kingship, and it can only be the prior assumption that he is not a
king which can challenge their cumulative force. O n the other hand,
the anti-monarchical case requires the demolition of the whole argument, point by point. T h e royal status of Danel is not necessarily
an essential element in the broad interpretation of the story, for the
heroes of tales are often kings or the sons of kings, and it merely
adds an aristocratic gloss to the narrative; but it also lends some
support to the view expressed below that the story has been given

A king is 'bd DN, 'servant (or "gardener"WYATT 1990b) of DN'. Thus with
Judahite kings 'bdyhuoh, 'servant of Yahweh', or 'gardener of Yahweh'.







an ideological twist by Ilimilku. Quite apart from its interpretative

significance, the fact remains that Danel's kingship appears to be
taken for granted.

The bow

Dressier and Hillers have both drawn attention to the significance

of the bow in the story of Aqhat. It is evidently of some considerable importance, since the desire for it motivates Anat to murder
the hero. In the broadest sense it is a symbol of power. But what
else is it?
H i l l e r s (1973) argued that it was in effect a phallic symbol. Its
theft by Anat would constitute an act of emasculation. He read into
this a n u m b e r of modern psycho-literary insights.
D r e s s l e r ( 1 9 7 5 ) countered that much of Hillers' argument centred on damaged text (and its restoration), hardly a sound basis for
far-reaching hermeneutical claims. T h e bow was indeed a masculine
symbol, but not a phallic one. T h a t is, it pointed to manly virtues,
but without explicidy sexual innuendo. He drew attention to H o f f n e r ' s
( 1 9 6 6 ) discussion of gender-symbols in defence of his own restrained
view, and rebutted each instance of an ancient literary allusion to
bows, arrows and quivers in which Hillers had discerned a sexual

What becomes of dead men?

T h e retort offered by Aqhat to Anat's offer of eternal life in exchange

for the bow has long puzzled scholars. M a r g a l i t ( 1 9 8 9 , 3 0 7 - 1 0 ) ,
while having nothing to say on the specific matter of post-mortem existence, presents Aqhat's observations in K T U 1 . 1 7 vi 3 6 - 3 8 as an
allusion to the Neolithic liming of skulls attested at Jericho. T h e idea
that a LBA text would be preserving ritual details of a very specific
kind (and long discontinued, to judge from mortuary evidence throughout the intervening period) seems most unlikely to me, and may be
discounted. The further idea that LBA people might discover such
ancient skulls (lying in shallow deposits) and contemplate their meaning (p. 309), while a charming idea, is scarcely a sufficient basis for
postulating a systematic anthropology of death. T h e more general issue
of the nature of beliefs regarding the afterlife has been dealt with
at some length by S p r o n k ( 1 9 8 6 ) , who takes a maximalist line, much
in the tradition of D a h o o d ( 1 9 7 0 ) , who cited the present passage in

his argument for a positive view of the afterlife, and even attributed
a firm belief in a beatific vision to the psalmists.
T h e following translation of K T U 1 . 1 7 vi 3 6 ~ 3 8 4 6 represents my
rather less fulsome view of the matter. T h e key word in the problem of translation has been hrs occurring in 1. 37. I have explained
it as representing perhaps a misspelling of the more c o m m o n hrs,
'gold'. D i j k s t r a - d e M o o r ( 1 9 7 5 , 1 9 0 ) offered the same explanation, considering hrs however as a parallel form, d e M o o r ( 1 9 8 7 ,
239) later abandoned this view in favour of 'potash', but I consider
that his initial insight was to be preferred.
Man, (at his) end, what will he receive?
What will he receive, a man (as his destiny?
Silver will be poured on his head,
gold on top of his skull,
[and] the death of all I shall die,
and I shall surely die.
As I read this, Anat does indeed voice the possibility of a life after
death, probably based on the beliefs held regarding dead kings, all
of w h o m are 'divine' in some sense (cf. K T U 1.113 = RS 24.257.
13-26, where each R N is prefixed by il). Some dead kings (though
to judge from the evidence, legendary rather than historical figures,
since no members of the king-list apparently qualify) are even given
the accolade rpum (cf. 6.4), which I understand to denote dead and
deified kings, comparable perhaps to the heroes of the Greek cult.
However, while this idea is here broached, and Ilimilku may indeed
be floating the idea through the agency of Anat's speech, he also
offers through the medium of Aqhat's wholly negative answer a sound
critique of such unrealistic views.
Cf. W Y A T T 1998C, 274 and n. 115. I have modified the third colon here. T h e
colon begins spsg, translated in 1998c as 'a precious substance ? ', and variously translated as 'glaze' (|| 'quicklime': G I B S O N 1978, 109), 'enamel' (|| 'whitewash': DEL
O L M O L E T E 1981a, 378), 'glaze' (|| 'potash': DE M O O R 1987, 239), 'a coating' (|| 'limeplaster': M A R G A I . I T 1989, 151). 1 noted (1998c, 274 n. 115) that this is reminiscent
of Prov. 26:23, where the expression kesep siggim, was read by G I N S B E R G (1945b,
p. 2In. 55) as k-spsg\ym\, 'like glaze' on the strength of the Ugaritic. But this argument may be regarded as circular, in which one unknown is used to interpret
another. Perhaps the original kesep is to be retained in Proverbs, and restored to
the Ug. text as <k>spsg || hrs/*/j'rs. T h e -sg(ym) remains unexplained in both cases,
but that is another issue. But Ugaritic is relatively stereotypical in its use of regular word-pairs, and ksp || for s is fairly commonplace, occurring many times in the
repertoire. Furthermore, the Heb. term is pi. in form, while the Ug. is sg., thus
reducing the appearance of a seemingly identical terminology.


New angles on Aqhat

T h e present author has attempted 4 7 to inteipret the work of Ilimilku

overall, as motivated in all probability by the concerns of his office
as priest and sacrificer of the king. Such a substantial amount of
material ( K T U 1.1-6, 1.14-6, 1.17-9, perhaps 1.10, and now 9.432
= R S 92.2016, still unpublished as this goes to press) by one named
person, however we apportion his editorial and compositional input,
is quite remarkable in the Late Bronze, and we may at least ask
whether he did not have some broad ideological concern to express.
Working on the basis of the view, now under challenge, that the
N i q m a d d u of the colopha is N i q m a d d u II, it is proposed that we
have in this substantial opus a legitimization of the new king, whose
reign may have begun in inauspicious circumstances (death of a prior
claimant? usurpation?), whose claim required every kind of support
available in a hearts and minds campaign. T h e place of Aqhat in
this hypothetical p r o g r a m m e was almost circumstantial, since the
weighty arguments were already spelt out in the Baal myths and in
Keret. But it would tend to reinforce some features of the Keret story:
thus the restitution of the blessing episode to the text would tend to
confirm the author's view that Danel is indeed a king. This in turn
would highlight the significance of the magnificent bow, fit for a
god, and of course of its function as a royal weapon (the king as
hunter is an analogue of the king as warrior), and thus attach a
peculiar importance to Aqhat, the hapless recipient of this wonder.
While not in the same form as the search for Baal and the allusion
to it in searching for help for Keret, Danel's search for Aqhat's
remains is perhaps intended to invoke the same theological associations, as though Baal's death is a type of a king's, here a future
king, with an aspiration to some happy hereafter. This of course is
what Anat offers the prince, but which he sharply rejects. Are we
to see in the raising of such questions a grappling with issues which
tradition has indeed sanctified by habit, but whose answers are no
longer regarded as good enough? Ilimilku might thus be seen as one
of O'Shaunessy's poets, the movers and shakers of the world.
In attempting to identify the poet's motivation (as well as to quantify his personal input into a representation of essentially traditional

W Y A T T 1997, 1998a. See also notes to the text in 1998c, 3 4 - 6 , 1 7 6 - 8 , 2 4 6 - 8

(written before 1998a), where I initially explored some of the ideas in question.

material), I am of course entering into a mode as subjective as that

which I have implicitly or explicitly criticized above in other scholars' work. But there seems to me to be a world of differencehowever difficult it may be to achieve itbetween attempting to foist
on an ancient author one's contemporary prejudices and discovering what were indeed his own concerns in the matter. In the matter of the 'caprice' of the gods, mentioned by two authors cited, I
think we have an example of the all too c o m m o n tendency to allow
m o d e r n theological values, already c o m m o n l y imported into the
Hebrew Bible, and declared to be resident there, to affect the agenda.
Not content with an invented 'biblical world view' which is more
often that of the post-reformation period, there seems to be an
attempt to make Ilimilku himself an early reformer! Yet I see no
tension at all in broad matters of metaphysics between Ilimilku and
his world. So far as the internal and traditional theology of Aqhat is
concerned, and which there is no reason to believe is under attack
by Ilimilku, the kind of perspective outlined below ( 13.4.2) is to
be discerned: polytheistic theory at the same time offers a coherent
overall structure at the macrocosmic level for the maintenance of
meaning and value in the world, and in the interaction of its particular deities accommodates the microcosmic realities of individual
problems, individual decisions and their consequences, and the tensions which are bound to exist between the real and the ideal world.
Anat's behaviour is predictable, and in no way a reflection of inadequate or immoral theology. As the embodiment of precisely the
dysfunctional aspect of the world, represented by all forms of killing
(hunting and war), she is a terrible power to encounter. Aqhat's brisk
rebuttal of her overtures, while c o m m a n d i n g our respect, inevitably
brings on his own head the whole weight of the traditional sanction
on those who blaspheme (sc. question the divinely ordained order of
things). T h e r e is a degree of tragedy and of awful inevitability, as
the sequence of automatic cursing and automatic revenge is set in
train, but this does no more than express in graphic terms the principle of accountability.
O n the matter of genre, there has been much discussion on that
of the Keret and Aqhat stories. Are they myths, legends or sagas?
W h e n myth is defined as 'stories about the gods' (e.g. by Gunkel
and Eissfeldt), then the presence of h u m a n characters in the story
compromises any attempt to categorize the story as myth. Since no
such inhibitions seem to affect discussion of mythology in any arena

other than the biblical one (where also it is arguably an inappropriate basis for classification, since it really belongs to a polemic
rather than a detached analysis), it may be regarded as irrelevant to
serious discussion of the Ugaritic texts. More to the point is perhaps
the matter of authorial intention. If myth be defined, as it is likely
to be in social scientific terms, as stories bearing an ideological (which
may include a religious) or paradigmatic message to their public,
then the issue of the nature of the characters, divine, h u m a n , or even
animal, is secondary. These elements are important. Fables, folk and
fairy tales, though they contain characters drawn from myth, have
lost the absolute imperative of the ideological norm, which demands
obedience to a conventional set of values, duties and taboos, and
effectively authorizes sanctions against non-conformists. But an element of freedom in the development of a tradition exists. T h u s the
myth (e.g. the Chaoskampf, which is integral to the maintenance of
royal ideology) may slowly evolve into other forms, such as 'St George
and the dragon', which retains vestiges of ideology, and the 'Celtic
dragon myth', which does not. T h u s genre is not absolute, as a given
narrative may be developing from one genre to another.
We noted above del O l m o Lete's and Parker's remarks on the
matter of genre. These highlighted, to my mind, the inadequacy of
the 'genre' approach, in so far as it seeks to establish literary types,
each with its own distinctive mental disposition, attitude to history,
to religion, and so on. These divisions are part of our need to structure our thoughts on such issues, and all too frequently bring a sledgeh a m m e r to crack a nut. Absolute categorizations, of the 'saga',
'legend', 'myth', 'epic' kind, are always subjective, if only because
no one agrees on definitions, and merely cut off various avenues of
retreat from the absolute judgments which each genre is felt to entail.
I prefer like many scholars to speak more neutrally of the 'story' of
Aqhat (and of Keret), noting different tendencies on various issues. This
avoids hostages to fortune, and still leaves narrators free to indulge
in a degree of eclecticism, perhaps drawing on different, or even
hybrid genres, and us free to estimate the broad mental stance of
the tradition, unencumbered by the demands imposed by arbitrary
T h e most obvious starting-point for an assessment should be the
global one of the LBA Weltanschauung of Ugaritian culture, as established through broad studies of its cosmology, theology, mythology
and ritual forms. I am entirely happy to call this 'mythological', in

the sense in which a modern religious response in devotion, scripturereading and cult remains mythological, since it operates on a different
level from the purely empirical. 'Mythological' is the more useful in
that while it relates to myth, it may also denote quasi-mythic features in other genres, such as deities featuring as characters, the suspension of empirical laws for narrative effect, and so on. 48
Aqhat is to be seen as a story, built up as Parker showed around
a n u m b e r of motifs, and as Aitken showed around a n u m b e r of
themes, motifs, formulae and word-pairs. T h e stages of its literary
prehistory are no longer recoverable, partly on account of the considerable skills of the tradents, partly because no one motivation
seems evident in its construction; but in the hands of Ilimilku there
is a case to answer that the poet pressed it into service in the interests of royal propaganda. T o this extent it has become an ideological text. And in so far as Ilimilku has brought an ideological element
into traditional material, he has blurred the distinction between genres, and produced composite works.


Some observations on style

A n u m b e r of commentators have remarked on the 'patriarchal' characterization of Danel, undoubtedly with an eye to similarities in the
presentation of the patriarchs of the Genesis narratives. T h e r e too
a domestic, almost bucolic gloss is given to narrative themes which
address the most urgent needs of h u m a n societies, their very physical survival, expressed most typically in the yearning of a m a n for
a son, who will support him in his declining years and perform his
obsequies. T h e r e is a surprising tautness to the text (well illustrated
in Aitken's treatment), with no word too m a n y and an action that
proceeds deliberately, its pace tailored to the various levels of meaning requiring weaving into the whole.
T h e fourfold repetition of the duties of the pious son, for instance,
is no mechanical overkill by a poetaster, but a skilful development
of one of the main themes of the story. While it is unprovable, it
is worth suggesting that this is one of Ilimilku's own insertions into
the traditional Vorlage, since it is so germane to his own concerns, if
my analysis is correct. In K T U 1.17 = R S 2. [004] i 2 6 - 3 3 the for-


Much biblical historiography is 'mythological' in this sense.

mula is voiccd by Baal, interceding on Danel's behalf with El. In

1.17 i 42~7 (and the missing conclusion) it is repeated by El in his
response to Baal. These two narrations therefore take place in heaven,
in the divine abode. As 1.17 ii becomes legible in 11. 1 - 8 the formula is already being repeated in a message by an unnamed deity
to Danel, confirming the promise of the birth that has now taken
place, and finally, the fourth instance, in 1 . 1 7 ii 1 6 - 2 3 , Danel himself repeats the formula as guarantee of divine blessing. Not only is
this a classic instance of the 'semantic rectangle' ( J A M E S O N 1 9 7 2 ,
1 6 3 - 7 ) in use as a literary tool, but it shows a clear progress from
an idea deliberated a m o n g the gods and only finally, via a revelation, appropriated by the h u m a n recipient. T h e theological weight
this confers on the text goes far beyond the merely folkloric.
A similar technique is used in the episode of Danel searching
the gizzards of the raptors. In the three successive scenes in K T U
1.19 = RS 3.322+ iii (edge and) l - 1 4 a , 14b-28a and 2 8 b - 4 0 a we
see Danel progressively home in on the place where his son's remains
lie. Firstly a rather unfocusscd look a m o n g falcon-gizzards in general; then an examination of Hargab's gizzard, and finally the discovery in Sumul's gizzard of fragments of the hero. This time there
is only a threefold repetition; but this too is masterly, for with the
semantic rectangle incomplete it leaves the reader with a sense of
tension, of a search not really brought to a final conclusion. This is
further heightened by the consequent threefold treatment of the cursing of localities held responsible. Again, the sense of tension. The
reason for this is surely that the real culprit is still at large. W e may
see the problem resolved on one level with the probable vengeance
wrought on Yatipan by Pughat, appropriately disguised as Anat, as
the final column breaks off. O n another level Anat is of course to
blame, and yet as a goddess she must remain above formal, or at
any rate articulated suspicion. It is on account of no weakness, and
certainly of no theological bankruptcy, that the goddess appears to
emerge unscathed. While her cruelty is not perhaps to be compared
in too facile a m a n n e r with Yahweh-El Shaddai's inscrutability in
J o b , we are faced with the same fruitless search for the fulfilment
of h u m a n expectations in the face of the divine nature. 4 9


For further comments on Anat's theological significance see below, 13.4.

In both contexts too, in Genesis and Aqhat, we discern ideological concerns either overtly expressed, or simmering beneath the surface. T h e patriarchs are royal and priesdy ancestors, and Danel is
a king, whose fecundity determines the future of his kingdom. T h e
domestic flavour is deceptive: in their final form both traditions are
pregnant with ideological power.








Introduction: the tablets

T h e so-called "Rpurr or 'Rephanr texts, K T U 1.20 = RS 3.348 (CTA

20, U T 121), K T U 1.21 = R S 2.[019] (CTA 21, U T 122) and K T U
1.22 = R S 2. [024] (CTA 22, U T 124, 123), are three small fragments of at least two large, multi-columned tablets.' T h e y preserve
in very broken form a portion of a narrative concerning a group of
beings called rpum.'1 While extremely problematic and ambiguous,
these texts have played an important role in the discussion of Ugaritic
and Canaanite concepts of death and the afterlife, as well as in the
reconstruction of the Ugaritic political and social order.
T w o of the fragments ( K T U 1.21 = R S 2. [019] and 1.22 = R S
2. [024]) were found during the second season of excavations (1930),
either inside the House of the High Priest on the acropolis or in the
rubble directly outside the southern wall of the house. 3 T h e third

K T U 1.20 = R S 3.348 was first published in V I R O L L E A U D 1936a, 228-30, in

his editio pnnceps of the Aqhat Epic. T h e other two, K T U 1.21 + RS 2. [019] and
K T U 1.22 = RS 2. [024], along with a rdition of K T U 1.20, first appeared in
V I R O L L E A U D 1941a. Besides the transcriptions in the standard collections (CTA and
KTU, KTU2 = CAT), a recent edition of the texts, with extensive photographs, may
be found in P I T A R D 1992. Major translations include V I R O L L E A U D 1941a, 1-30;
D R I V E R 1956, 66-71; A I S T L E I T N E R 1964, 83-86; C A Q U O T etat. 1974, 461-80; G O R D O N
1977, 29-31; C O O G A N 1978, 48-51; L ' H E U R E U X 1979, 129-59; DEL O L M O L E T E
1981a, 405-24; S P R O N K 1986, 163 77; DE M O O R 1987, 266-73; D I J K S T R A 1988,
35-51; L E W I S 1996b, 128-31; L E W I S 1997, 196-205; W Y A T T 1998C, 316-23.
T h e meaning of the word is usually related to the root rp', which means, 'to
heal'. T h e correct pronunciation of rpum in Ugarit remains controversial. T h e two
most popular proposals are (1) rpi'ma, vocalizing it as a participle, which would
mean, 'healers', and (2) rapi'ma, a stative form that can be rendered, 'the healthy,
hale ones'. See the discussion of the possibilities in L ' H E U R E U X 1 9 7 9 , 2 1 5 - 2 1 ; P A R D E E





Unfortunately the inventory lists for the first two seasons were lost, probably
during World War II (cf. B O R D R E U I L
P A R D E E 1 9 8 9 , 1 3 ) . It is known that these
tablets were registered with topographic points that were marked on ground-plans
that still exist. T h e tablets of the second season were numbered with points 2 1 0 - 6 4 .
Unfortunately, excavations that year took place all along the western side of the
house, in the street along its southern boundary, in the room to the west of the
southern entry, and in a room to the east of the larger interior courtyard. So these
topographic points are scattered throughout the house (see the plan in B O R D R E U I L P A R D E E 1 9 8 9 , 2 5 ) . Because the third fragment and most of the other literary texts
were found in the entry room or just outside the doorway, it seems probable that

was discovered the following year in or near the house's southern

doorway, in the same area where the Keret and Aqhat epics emerged,
along with some of the Baal tablets. Of the three texts only K T U
1.22 = R S 2. [024] preserves some complete lines. K T U 1.20 = R S
3.348 contains parts of two columns on one face. T h e left column
is m a d e up of the right halves of eleven lines of text, while the right
column preserves somewhat over half of twelve lines. K T U 1.21 =
RS 2.[019] preserves part of a single column on one face (the right
two-thirds of thirteen lines), but only the last five letters of one line
on the other side. T h e largest of the three fragments, K T U 1.22 =
RS 2. [024] preserves a left column of twenty-eight lines, twenty-two
of which are complete, and a badly broken right column of twentysix lines, with only about one-third of each line preserved. O n the
other side of the latter tablet are two identifiable letters, plus fragments of two others, each the first letter of a line. We thus have an
aggregate of 95 attested lines, only 22 of which are complete, while
an additional 24 are somewhat more than half-preserved.
K T U 1.21 = RS 2. [019] and 1.22 = R S 2. [024] may come from
a single tablet. T h e scripts of both appear to have been written by
the same scribe, probably Ilimilku, who also produced the other
major narrative texts in the archive. O n the other hand, K T U 1.20 =
RS 3 . 3 4 8 seems to belong to the work of a different scribe (cf. P i t a r d
1992, 75, n. 8), and thus presumably to a separate tablet. Since the
literary tablets found in this archive range in size from about 200
to about 500 lines, we may conclude that the 95 lines attested on
the rpum texts are probably no more than a quarter and perhaps
only one-tenth of the lines that would have existed on the two tablets
when intact.
In the entire collection of narrative poetic texts from Ugarit few
are as obscure and difficult to interpret as the rpum fragments. Rarely
have the ambiguities of a vowelless script and the piecemeal preservation of so many damaged lines conspired so powerfully to frustrate attempts at drawing decisive conclusions about a text. T h e
broad context within which the action depicted in these fragments
occurs is quite obscure, a n d only a few things about the events
described in the preserved lines can be discerned with certainty. It
is clear that the characters called the rpum (also designated with the
parallel terms ilnym, 'godly ones', and ilm, 'gods') are invited to a
banquet. At least one invitation seems to be given by the king of
the gods, Ilu ( K T U 1.21 = R S 2. [019] ii 8), but virtually identical

invitations are offered five other times in the three fragments, each
in a broken context that does not allow us to confirm whether these
come from the same deity. Each invitation seems to be followed by
a description of the journey to the banquet.' Both K T U 1.22 = R S
2.[024] and K T U 1.20 = R S 3.348 seem to describe the arrival of
the group at a threshing floor, where the banquet is apparently given,
and K T U 1.22 = R S 2. [024] i 10-25 describes the feast as lasting
for a week. O n the seventh day, it appears that the god Baal arrives.
But his function in the story, and indeed the purpose of the gathering of the rpum for the banquet remains unclear, and the text breaks
off at this point. Beyond these few elements of the story, little certainty about the plot of the narrative and its meaning is possible.
Not only is the larger context of the fragments lost, but the broken nature of the narrative makes it impossible to ascertain even
some of the basic elements of the preserved story line. For example, not only is it uncertain whether all six invitations are offered
by the same character, but it is also unclear whether the invitation
is addressed to the same set of rpum, or whether different groups of
rpum are being invited to the feast. Further, the fragments contain
almost no information about the characteristics and identity of the
rpum. In fact, about all we can say with certainty about them is that
they travel by chariot and that they eat heartily at the feast. O n e
passage in the more complete column of K T U 1.22 = RS 2.[024]
(col. i, lines 4-10), seems to describe a company of rpum who arrive
at the banquet, but the text, though completely preserved, is frus-

K T U 1.21 = RS 2.[019] and K T U 1.22 = RS 2. [024] were discovered either in

the street or in the room west of the entry.
Four of the five versions of these lines in K T U 1.21 = RS 2. [019] and K T U
1.22 = RS 2.[024] appear to read (with the possibility of some variation) atrh. rpum.
lldd I atrh.ltdd.ilnym, while the last version of these lines, which directly precedes the
description of their arrival at the banquet, drops the I before the one preserved Idd.
This is also the case in K T U 1.20 = RS 3.348, where the lines appear in col. ii 1.
The ambiguity of the I has encouraged several interpretations of the lines. Some
take it as an emphatic particle, and thus render these lines with the understanding
that the rpum are travelling in each case; thus, for example, 'After him the shades
verily proceed' ( G O R D O N 1977, 30-31). Others propose that the I is a precative particle and that the lines should be translated with jussive force, either as part of the
invitation (e.g., S P R O N K 1986, 169 72: 'May the rpum flutter to the holy place'), or
as an expostulation of the poem's narrator (e.g., L E W I S 1996b, 129: 'To his shrine,
shades, hasten'). DIJKSTRA 1988, 4 1 - 4 3 and others have argued that the / is best
understood as a negative particle. They suggest that the rpum refuse to hasten until
they are finally convinced by the fifth invitation (thus, 'After him the shades did
not move').

tratingly ambiguous. O n e finds critical words here that may be interpreted either as proper names of the rpum, or as verbs describing
actions taken by some of the story's characters. 5 In other cases, one
cannot be sure whether certain construct nouns are to be construed
as singular or plural. T h u s the mhr b'l, mhr ent, and rpu b'l, in lines
8 - 9 , may be titles of individual characters (i.e. 'the warrior of Baal',
'the warrior of Anat', and 'the rpu of Baal'), 6 or they may designate
large numbers of persons (i.e. 'the warriors of Baal', 'the warriors of
Anat', and 'the rpum of Baal'). 7 T h e ambiguity here makes it impossible to use this section to help define the nature of the rpum.
O t h e r problems arise. H o w is one to understand the relationship
between the three fragments? Since K T U 1.22 = R S 2. [024] and
K T U 1.20 = R S 3.348 both describe the arrival of the rpum at the
threshing floor where a banquet is served, is it best to assume that
the two fragments are separate versions of the same scene, fortuitously overlapping, 8 or are they describing two different banquets?
T h e latter interpretation is possible since K T U 1.20 = R S 3.348
depicts the banquet as being hosted by a h u m a n , Danel, a character
also known from the Aqhat epic, while the banquet of K T U 1.22 =
R S 2.[024] may be hosted by Ilu. Since the former text almost certainly comes from a different tablet, it may actually be part of a
completely different narrative that just happened to have a similar
convocation of the rpum as a story element.
In addition, what is the relationship between K T U 1.22 = R S
2.[024] and K T U 1.21 = R S 2.[019]? 9 D o they belong to a single

Beyond this problem of interpretation, the translations just quoted also point up
another major ambiguity in these lines. T h e word alrh may be analyzed as a preposition with a suffix, 'after him', or it can be understood as a noun, atr, 'place, shrine'
with a possessive pronoun, 'to his shrine', or simply a directional marker, 'to the
For example, is tmq in line 8 a verb (as translated by D R I V E R 1956, 69; L ' H E U R E U X
1979, 152-53; C A Q U O T 1974, 474-75; S P R O N K 1986, 171; DE M O O R 1987, 272; L E W I S
1997, 203) or the proper name of a character (as translated by AISTLEITNER 1964, 85;
DEL O L M O L E T E 1981a, 423; DIJKSTRA 1988, 47; W Y A T T 1998c, 321 n. 38)? The same
question arises concerning the word yhpn in line 9. Some commentators take it as
a proper name (AISTLEITNER 1964, 85; DEL O L M O L E T E 1981a, 423; DIJKSTRA 1988,
47; W Y A T T 1998c, 321); others as a verb ( G O R D O N 1966, 141; D R I V E R 1956, 69;
L ' H E U R E U X 1979, 152-53; SPRONK 1986, 171; DE M O O R 1987, 272; L E W I S 1997, 203).
Rendered thus by D R I V E R 1956, 69; C A Q U O T et al. 1974, 474-75; DEL O L M O
L E T E 1981a, 423; DIJKSTRA 1988, 47; W Y A T T 1998C, 321.
Translated thus or similarly in G O R D O N 1 9 6 6 , 1 4 1 ; A I S T L E I T N E R 1 9 6 4 , 8 5 ;


1 9 7 9 , 1 5 2 - 5 3 ; SPRONK 1 9 8 6 , 1 7 1 ; DE M O O R


1987b, 267; DIJKSTRA 1988, 35-39.

1987b for a discussion of this issue.

1 9 8 7 , 2 7 2 ; LEWIS 1 9 9 7 ,


tablet, or may they belong to two separate ones, once again overlapping in the part of the story recounted? T h e latter possibility
cannot be ruled out, since the vast majority of the lines in K T U
1.21 = RS 2. [019] i are repeated in K T U 1.22 = RS 2. [024] ii. In
particular, the former preserves two invitations to the rpum, followed
by two descriptions of them making the journey toward the shrine
or palace, while the latter has a threefold appearance of the same
basic lines. Would such an extraordinary amount of repetition in
two small fragments of a single tablet be plausible, or is it better to
identify them as duplicates? Again, there is no definitive argument
for either interpretation. T h e a m o u n t of repetition is indeed surprising in such a small n u m b e r of preserved lines. O n the other
hand, multiple reiterations are well attested in the Ugaritic poems, 1 0
and it is possible to develop scenarios of the story that would allow
for so m a n y recurrences of the invitation (e.g., that different groups
of rpum are being invited to the feast).
And finally, is there a relationship between these fragments and
the Aqhat epic? T h e appearance of Danel in K T U 1.20 = RS 3.348
has led a n u m b e r of scholars to argue that the rpum texts are the
remains of a fourth tablet of the Aqhat narrative." T h e y interpret
the banquet as a preliminary to the restoration of Aqhat to life or
some similar dnouement which would have brought the story to its
conclusion. T h e problem with this proposal is that, apart from the
presence of Danel in this text, there is nothing in the rpum texts that
suggests that the action described in these fragments relates in any
way to such a proposed scenario. Others have pointed out that Danel
may have been the subject of more than one narrative at Ugarit,
and that his presence here does not require that the fragments be
related to the other known narrative in which he plays a role. 12 At
this point it seems best not to insist on a relationship between them
and the Aqhat story.


The Identity of the R p u m

Undoubtedly the biggest hindrance to gaining a proper understanding of these texts is the uncertainty about the nature of the rpum,
See for example the four-fold repetition of the duties of an ideal son in the
Aqhat epic (KTU 1.17 = RS 2. [004] i 25-33; 42-7; ii 1-9; 14-23).
" For example, S P R O N K 1986, 160-1; C A Q U O T et al. 1974, 463; DE M O O R 1976,
332; M A R G A L I T 1989a, 464-5; G R A Y 1965, 126-9.
E.g. P I T A R D 1992b, 73; D I J K S T R A 1988, 36; and L E W I S 1996b, 119.

the primary characters in the narrative. In spite of their appearance

not only in these texts, but also in a few others, 13 scholars have been
unable to reach a solid consensus on their identity. T h e primary
proposals include the following:
(1) They are ghosts of the dead, most likely the spirits of deceased
kings, and perhaps of the nobility. Several scholars argue that these
spirits were thought of as deified.
(2) T h e y are a group of deities who join Ilu in special gatherings
and who are called upon to protect the king and his city.
(3) T h e y are living m e m b e r s of the aristocracy, perhaps an elite
group of chariot warriors, or perhaps a group of priests involved in
rituals of fertility. In this context the term may be viewed as a tribal
name, probably related to another tribal designation, Ditanu, which
appears in parallel with rpum in the Keret epic and in K T U 1.161 =
R S 34.126.
(4) A n u m b e r of scholars argue that the term may in fact be used
to designate more than one of these three groups.

The r p u m as spirits of the dead

T h e most commonly-accepted proposal is that the rpum are spirits of

the dead. 14 There are several good reasons to support this identification.
(1) T h e cognate of rpum in biblical Hebrew, r'p'm, has as its primary meaning, 'spirit of the dead, ghost'. T h e same meaning attaches
to the word in Phoenician.
(2) K T U 1.161 = R S 34.126, in which the rpu ars, 'the rpum of the
earth', and the rpim qdmym, 'the ancient rpum\ appear in the context
of a funerary ritual, suggests that they have a relationship to the
dead. In this context they are summoned to take part in the funeral
of N i q m a d d u III of Ugarit (late 13th century) and perhaps to bless
the new king, 'Ammurapi. In the ritual, the summons of the 'ancient
rpum' is followed immediately by the invocation of the spirits of two
identifiable, deceased kings of Ugarit (not explicitly referred to as
rpum, however). Scholars have argued that the 'rpum of the earth'

They are also mentioned in K T U 1.161 = RS 34.126, a funerary ritual text;

in the Keret epic (KTU 1.15 = RS 3.343+ iii 3 and 14); at the conclusion of the
Baal cycle (KTU 1.6 = RS 2. [009]+ vi 46), and perhaps in K T U 1.108 = RS
24.252.23-4 and K T U 1.82 = RS 15.134.32, both of which are damaged.
See for example, C A Q U O T 1960; D I E T R I C H - L O R E T Z - S A N M A R T I N 1976c; P O P E
1977; S P R O N K 1986, 161-96; F O R D 1992; W Y A T T 1998C, 3 1 5 n. 1.

and the 'ancient rpun are royal ancestors of 'Ammurapi from the
distant past. Since the word, ars, 'earth', was sometimes used to designate the netherworld, one can render rpu ars, 'the rpum of the
(3) A passage at the end of the Baal epic K T U 1.6 = R S 2. [009]+ vi
45-9) contains two bicola in which the four words rpim, ilnym, ilm,
and mtm may be understood as synonymous with one another: p \ p thtk.ilnym | 'dk.ilm \ hn.mtm.'dk. This may be translated
as 'Shapshu (the sun goddess), you rule over the rpum, | Shapshu,
you rule over the godlike ones. | T h e gods are your company, |
See, the dead are your company'. It should be noted, however, that
these interpretations of K T U 1.161 = RS 34.126 and 1.6 = RS
2. [009]+ are by no means certain (see below). T h e y cannot be considered decisive for favouring this identification of the rpum.

The r p u m as deities

M u c h of the same evidence can be used to argue that the rpum are
better understood as deities, rather than spirits of the dead. 1 5 Some
scholars would identify them specifically as minor netherworld deities, closely associated with Baal. Others have proposed that the term
may designate any deity, major or minor, who is called upon by Ilu
to perform a special function. Arguments for identifying the rpum as
deities include the fact that the term rpum is several times paralleled
by the term ilnym, 'godlike or godly ones', quite plausibly a divine
title, and perhaps by the term ilm, 'gods' as well."' Arguments against
identifying them as the dead (as described above) and for seeing
them as deities instead also include the following:
(1) Later meanings of words do not always parallel earlier meanings
of the cognates in other languages. T h u s the Hebrew and Phoenician

See S C H M I D T 1 9 9 4 , 8 3 - 9 2 and L ' H E U R E U X 1 9 7 9 , 1 1 6 - 9 for a history of this
view. L ' H E U R E U X also argues that the rpum of our texts are deities, not spirits of
the dead, though in other passages the word may designate such spirits. See ibid.
2 0 5 - 6 , 2 1 5 - 3 0 . The idea that the rpum are deities should be distinguished from the
view of those who argue that the dead are deified at Ugarit and thus can be
identified as both spirits of the dead and gods.
The occurrences are: K T U 1 . 2 0 = RS 3 . 3 4 8 i 1 - 2 ; K T U 1 . 2 1 = RS 2 . [ 0 1 9 ]
i 3 - 4 , 11-2; K T U 1.22 = RS 2. [024] ii 5 - 6 ; all relatively certain, but all in broken contexts, and in K T U 1 . 6 = RS 2 . [ 0 0 9 ] + vi 4 5 - 9 . Parallels with the word dm,
'gods', probably occur in K T U 1 . 2 0 = RS 3 . 3 4 8 ii 1 2 and 8 - 9 , both, however,
in very broken contexts.

cognates are far from decisive. T h e context of a word like rpum

within the local literature is more significant for its interpretation.
However, nothing in our three rpum texts specifically points to a
ghostly identity for the rpum. In fact, there is no hint in any of the
fragments that the regular habitation of the rpum is the netherworld.
T h e only clear action tells of the rpum preparing their chariots, travelling to a threshing floor and having a feast, none of which particularly hint that the beings are spirits of the dead. Near Eastern
evidence for spirits of the dead travelling from the netherworld in
horse-drawn chariots is nil, while such transportation for deities,
whether from the netherworld or elsewhere, seems quite reasonable.
T h e lack of netherworld indications in the rpum texts might suggest
that, while K T U 1.161 = R S 34.126 indicates a connection between
the rpum and funerary activities (see below), their sphere of activity
may not necessarily be restricted to that realm.
(2) Although the rpu ars and rpim qdmym of K T U 1.161 = R S 34.126
are involved in a funerary context, this does not mean that they are
necessarily spirits of the dead. T h e y may also be identified plausibly as deities. T h e names of the beings specifically identified in K T U
1.161.4-7 as rpum are not attested as royal names of Ugarit (or elsewhere), but in fact resemble divine names more than h u m a n ones
(see particularly the composite name, sdn-w-rdn, (lines 6, 23) and tr
'limn (lines 7, 23-4). W h e n the two recognizable, deceased kings of
Ugarit are named (lines 11-2, 25-6), each is designated as mlk, 'king',
rather than rpu. T h u s those who are obviously spirits of the dead
appear to be given a different designation in the text. In sum, the
evidence from this tablet does not compel one to identify the rpum
as ghosts.
(3) T h e reference to the rpum in the Baal Epic ( K T U 1.6 = RS
2. [009] vi 45-9), which can be interpreted as equating the rpum with
the dead, is in fact ambiguous, because the word, mtm, often translated 'the dead' in this passage, could actually be a h o m o n y m well
attested at Ugarit that means 'humans'. In this case, the passage
may be translated, 'Shapshu, you rule over the rpum, | Shapshu, you
rule over the godlike ones. | Gods are your company, | See, humans
are your company'. 1 7 If this is the correct rendering, there would
be no direct parallel connection between the first bicolon and the
second, as assumed by those who suggest that all four words refer





to the dead. Rather, the two extol the importance of Shapshu in

her relations with various elements of the world orderthe rpum/ilnym,
minor gods, perhaps related to the netherworld; then ilm/mtm, the
divine and h u m a n spheres as a whole.
Although none of the three points made here disproves the identification
of the rpum with the dead, they each emphasize the fact that none
of the arguments for such identification is fully compelling. O n the
other hand, while the texts are compatible with the view that the
rpum are deities, none of them require that meaning to make sense.

The r p u m as living persons

Some scholars have argued that at least certain of the Ugaritic references to the rpum are best understood as referring to living persons. 18 In this interpretation, they are usually identified as an elite
group of chariot warriors who had strong connections with the king.
Those who support this idea argue primarily from the passage in
the Keret epic, K T U 1.15 = R S 3.343+ iii 2 - 4 and 13-5 in which
Keret is praised by Ilu himself: 'Greatly exalted is Keret in the midst
of the rpum of the earth (or land), | in the gathering of the assembly of Ditanu'. Supporters argue that 'rpum of the earth' in this passage can hardly refer to the spirits of the dead, since it would be
inappropriate in the context of the exaltation of Keret to proclaim
his preeminence a m o n g the dead. 1 9 T h e 'assembly of Ditanu' (qbs
dtn), the phrase that is parallel to rpu ars in the passage, can plausibly be identified as a designation for the leaders of the Ditanu clan.
This suggests an earthly, political and social context for the rpum.
T h e military imagery of our three rpum texts (especially the use of
chariotry and the appearance of what seems to be a military host
in K T U 1.22 = R S 2. [024] ii 4 - 1 0 ) fits reasonably into this reading as well. Most scholars who identify the rpum as living humans

See L ' H E U R E U X 1 9 7 9 , 1 2 0 7 for a history of this type of proposal. More
recently see S C H M I D T 1 9 9 1 , 7 1 - 1 2 1 for a detailed defence of the argument, including a number of new elements. G R A Y 1 9 4 9 argued that the rpum were elite leaders, perhaps heads of clans, or perhaps priests, who accompanied the king in rituals
designed to insure fertility of crops. This view has generally not been followed.
Those who believe that the rpum are spirits of the dead or deities dispute this,
arguing that since the rpum play an important role in the preservation and support
of the king, Keret's exaltation in the midst the ancestors is a reasonable image to
use in the story. See P O P E 1 9 7 7 , 1 6 6 - 7 ; F O R D 1 9 9 2 , 7 3 - 6 .

also assume that even after these chariot warriors died, they continued to be called rpum, so that there were both living and dead
rpum. This would explain the origin of the usage of the term for the
dead. Note is often made of the fact that the Bible refers to an
ancicnt tribe of giants, located in northern C a n a a n and Transjordan
before the establishment of Israel, as rephaim (Gen. 14:5, 15:20; Deut.
2:11, 20; 3:11, 13; Jos. 12:4, 13:12; 17:15). K n o w n for their military prowess, these rephaim are thought to be a dim memory of the
elite warrior class of that name in the Late Bronze Age. So again
we find some plausible arguments, but again none of the texts used
to support the interpretation provide irrefutable evidence. Plausible,
alternative interpretations of each of these passages have been made.
T h u s in the final analysis, no decisive conclusions about the identity of the rpum can yet be drawn. It is quite possible, as several
scholars have argued, that the word had more than one meaning in
the Ugaritic texts and that different contexts require different meanings. O n the other hand, other scholars have made plausible cases
for seeing all the occurrences of the word as referring to a single
group of beings, either spirits, gods, or humans. Only further discoveries of texts relating to the rpum are likely to improve the present situation.
From the preceding discussion, it is clear that these texts are exceedingly ambiguous and that great caution should be used in drawing upon them to reconstruct aspects of Ugaritic or Syro-Palestinian
culture. In many cases such caution has not been employed, so that
the rpum texts have been used extensively as the basis for elaborate
descriptions of Ugaritic concepts concerning afterlife by scholars who
identify the rpum as spirits of the dead. 20 A similar situation has also
occurred in some of the reconstructions of the Ugaritic military and


For example, DE M O O R 1 9 7 6 , 3 2 9 - 3 3 and S P R O N K ( 1 9 8 6 , 1 5 5 - 6 , 1 7 0 - 4 ) , using

K T U 1 . 2 2 = R S 2 . [ 0 2 4 ] ii 5 - 7 and K T U 1 . 2 1 = R S 2 . [ 0 1 9 ] i 5 - 6 as their primary sources, reconstruct an event at the Ugaritic New Year Festival in which the
deceased nobility, the rpum, are brought back to life, along with Baal, their patron.
But such a reconstruction relies for its foundation upon the identification of the
rpum as spirits, which remains uncertain, alongside problematic readings of both key
passages. Their interpretation of the first passage is based on questionable renderings of two problematic verbs, the first of which (his ytb, which occurs twice here)
is based on a misreading of the text (the verb is y'bs, instead), and the second, qym,
more plausibly rendered as a noun than a verb (see VAN DER T O O R N 1 9 9 1 , 5 2 ) . In
the second passage, they must reconstruct both of the critical verbs needed for their

its relation to the royal house by those who view the rpum as living
elite warriors. 2 ' It is important not to place too much interpretational weight on ambiguous and problematic texts such as these.
Before they can be used as sources for dealing with the wider issues
of Canaanite religion and society, a clearer understanding of the
texts themselves is necessary.


See, for example,










T h e incantation can be defined as 'rhythmic or formulaic words of

power to accomplish a desired goal by binding spiritual powers'.' A
n u m b e r of Ugaritic texts written in alphabetic script unearthed in
Ras S h a m r a and Ras Ibn H a n i fit, completely or in part, under this
heading. In none of these texts, however, do we find a word specifically
denoting the incantation, like Akkadian siptu.2 This word is attested
at the end of a syllabic Akkadian incantation found in Ras S h a m r a
'against fire' (RS 17.155). 3 We do find the more general indication
mnt? This has an equivalent in Akkadian mintu. In Akkadian incantations it is used next to siptu. It seems to refer in the first place to
the act of reciting and repeating the incantation, because it is derived
from the verb man,b 'to count' (cf. Hebrew mnh). In K T U 1.24 =
RS 5.194:46-7, mnt denotes the enumeration of goddesses listed in
the following lines. It is paralleled by spr, 'list'. In K T U 1.82 = R S
15.134:20 and K T U 1.100 = R S 24.244 it is used in texts that as
a whole can be labelled as incantations. In other 'pure' incantations,
like K T U 1.96 = R S 22.225 and K T U 1.169 = R I H 7 8 / 2 0 , this
or another general term for the genre is missing.
It appears to be impossible to distinguish the 'Ugaritic Incantations'
sharply. 6 Also, the calling of divine beings by their names at the end
of K T U 1.24 = R S 5.194 may have had some kind of magical purpose. T h e same can be said of other parts of some of the mythical
or ritual texts. Add to this, firstly, that is often difficult to distinguish




For the much larger corpus of Akkadian incantations see the surveys by F R B E R
1981, 1984 and 1987.
Cf. A R N A U D 1995a. Within the context of the incantations' 'fire' can be regarded
as a reference to demons; cf. K T U 1.2 = RS 3.367 i 3, where the demoniacal
helpers of Yam are described as 'one, two fires'.
See on mnt P A R D E E 1 9 8 8 , 2 0 6 - 8 . According to DE M O O R 1 9 8 7 , 2 4 8 the masculine mn is attested in K T U 1 . 1 9 = RS 3 . 3 2 2 + i l l .
Cf. the expression man iptu, 'to recite an incantation', in Akkadian, cf. CAD
, III, 89.
Cf. J E F F E R S 1 9 9 6 , 1 8 , facing the same problem with regard to the more general theme of magic and divination.

a prayer from an incantation, in the second place, that as a rule

magical and related texts are difficult to interpret, and that, finally,
many of the tablets concerned are damaged, then it comes as no
surprise that in the editions of and commentaries on Ugaritic texts
we find different proposals for classification:
According to Avishur the only texts which 'can clearly be classified
as incantations' are K T U 1.100 = R S 24.244; K T U 1.107 = R S
24.251+; and K T U 1.169 = R I H 78/20. 7
Xella lists under 'preghieri ed incantesimi': K T U 1.65 = RS 4.474;
K T U 1.123 = R S 24.271; K T U 1.100; and K T U 1.107. 8
DE M O O R comes to five incantations as 'more or less independent
prayers without ritual prescriptions', next to incantations 'embedded in rituals': K T U 1.82 = R S 15.134; 1.83 = RS 16.266; K T U
1.169; K T U 1.93 = RS 19.054; and K T U 1.108 = RS 24.252. 9
Dietrich - Loretz come to eight alphabetic texts that in their view
are representative of this 'Gattung', leaving out two thematically
related but heavily damaged texts. T h e y subdivide these eight texts
into four categories:
(1) 'Evokationen kniglicher Ahnen': K T U 1.124 = RS 24.272
and K T U 1.161 = R S 34.126;
(2) 'Beschwrungen gegen Krankheit, Unfruchtbarkeit, D m o n e n ,
Folgen von Trunkenheit und Totengeistern': K T U 1.13 = R S
1.006; K T U 1.82; K T U 1.114; K T U 1.169;
(3) 'Beschwrung gegen Schlangengift': K T U 1.100;
(4) 'Beschwrung gegen die schdliche Naturkrfte': K T U 1.23 =
R S 2.002.'
C a q u o t lists K T U 1.82; K T U 1.114 = RS 24.258; K T U 1.100;
K T U 1.107; and K T U 1.169 under the heading 'tablettes mythicomagiques'."





1981, 207-50.

DE MOOR 1987,


1988b, 328-7. In KTU1 the following texts are also marked

as possible incantations: 1.20-22 = RS 2. [024], 2. [019], 2. [024]; 1.65 = RS 4.474;
1.86 = RS 18.041; 1.96 = RS 22.225; 1.107 = RS 24.251+; and 1.123 = RS
24.271. Apparently, these belong to the 'damaged' or 'related' texts referred to by
Dietrich - Loretz.
" C A Q U O T 1989, 51-123. It is remarkable that he pays no attention to these
texts in his survey in C A Q U O T 1979b.



In his description of Ugaritic religious practices in daily life del

O l m o Lete mentions as 'conjuras': K T U 1.100; K T U 1.107; K T U
1.82; K T U 1.96; and K T U 1.169; 12 and as 'recelas mgicas': K T U
1.124 = R S 24.272:13-5 and K T U 1.114:29-31. 1 3
A recently published survey of documents from the biblical world
contains as examples of Ugaritic incantations: K T U 1.100; K T U
1.169; K T U 1.114; and R S 92.2014.' 4
Apparently the old Ugaritic texts on these matters cannot be clearly
classified. It is better in this situation not to put too much weight
on our modern definitions and choose a more general approach to
the Ugaritic texts about h u m a n efforts to have an influence upon
the supernatural, from raising one's hands in prayer to binding hostile spiritual powers by magic. This means that the boundaries between
'literary' and 'cultic', and those separating 'myth', 'incantation', 'ritual', and 'god lists' are not always as clear as editors of a handbook
might want them to be.


Speaking to the gods in hymns and prayer

T h e genre of prayer appears to be rare in the texts of ancient

Ugarit. 15 O n e should not, however, conclude from this that the people
of Ugarit did not have deep religious feelings or that they were reluctant to address their gods direcdy. T h e lack of separate hymns or
prayers is simply due to the fact that praising the gods or seeking
their favours is usually set in a larger context. Recitation of the great
myths can be seen as a means of expressing respect for the gods
and their glorious deeds. For instance, telling each other about Baal's
victory over Yam and Mot with the words of the myth of Baal
expresses one's confidence in the power of the supreme god over
chaos and death. T h e hymn addressed to the sun-goddess Shapash
at the end of the myth ( K T U 1.6 = R S 2.[009]+ vi 45-53) is put
in the mouth of Anat, but it is also a way in which the people of
Ugarit thankfully praise the sun-goddess for watching over the boundaries between night and day, the world of the living and the world
of the dead. Hymnic elements can also be found in the second part



1992a, 241-60 = 1999, 359-87.

1992a, 261 = 1999, 388.
H A L L O 1997, 295-8; 301-5; and 327-8.
Cf. W A T S O N 1984a, 360 and M I L L E R 1988.



of the much debated text K T U 1.108 = RS 24.252, , e which is discussed below together with texts related to necromancy. Hymns also
seem to have formed a standard element in Hurrian prayers accompanying incense offerings ( K T U 1.44 = RS 1.007; K T U 1.51 = R S
1.027; K T U 1.54 = RS 1.034+; K T U 1.128 = RS 24.278; K T U
1.131 = RS 24.285). 17
In the legend of Aqhat we hear of his father Daniel praying (Ug.
sly) for rain ( K T U 1.19 = RS 3.322+ i 38 46). He calls on the
name of Baal, as 'rider of the clouds', and on his 'delightful voice',
that is of the thunder heralding coming showers. In a subsequent
scene, Daniel beseeches (Ug. sly) the gods that the small stalks in
the dry land may shoot up ( K T U 1.19 ii 15-25).
It is more c o m m o n for prayer to be part of ritual actions, as we
can see in the legend of Keret. T h e c o m m a n d to raise the hands
(in prayer) is paralleled by a reference to a sacrifice to El ( K T U
1.14 = R S 2.[003]+ ii 22-3). W e can also find this combination in
the ritual text K T U 1.41 = RS 1.003+, with prescriptions about the
annual celebration of the grape harvest in the month 'First of the
Wine'. T h e text ends with the same call for prayer as in the legend
of Keret. In K T U 1.87 = R S 18.056, a copy of K T U 1.41, these
last lines containing the reference to the king's prayer are missing.
T h e action described in K T U 1.41:50-5 is situated in another place:
not in the temple, but on its roof; and it refers to a sacrifice offered
to an unknown deity (prgl sqrn). This is probably a deity of Hurrian
origin. So the expansion of the text can be explained as due to later
Hurrian influence upon an older Ugaritic ritual. T h e king is said (or
prescribed) to offer a recitation (yrgm mlk), but we hear nothing of
its contents. Perhaps building on the assumption of Hurrian influence
one should think here of something like the Hurrian incense prayers
mentioned above. These texts all follow a similar pattern: after the
heading we read the names of the gods to w h o m the prayer is
addressed, together with a short hymn. T h e gods are asked to come
and receive the offerings and then to do something on behalf of the
suppliant. T h e texts end with mention of the messenger and in some

A V I S H U R 1994, 297-8 even speaks of the entire text as a 'hymn in honor of
EF, because of the striking similarity with Hebrew psalms and its vocabulary being
reminiscent of that found in hymns.






Cf. DE M O O R 1987, 252, n. 190 and

interpretation of 'ahl see DEL O L M O L E T E


1984b, 140-1; for a different

1996, 16.


cases with promises of new offerings and a final doxology. T h e words

spoken by the king according to K T U 1.41:53, on the roof of the
temple, could have been something like this calling up the gods and
asking their favours. Because the tablet is damaged here, it is not
clear whether this invocation is accompanied by the king wiping his
face (mh pnh)i9 or by clapping his hands (mh ydh).20 T h e reference to
prayer in line 55, back in the temple, could be related to the closing hymn in the Hurrian incense prayers.
In the older secondary literature K T U 1.65 = R S 4.47 4 21 has
been interpreted as a prayer to El and the assembly of the gods. 22
More recently commentators of the text appear to be reluctant to
classify it. Xella points to the resemblance of the first lines (naming
El, the sons of El, the family of the sons of El, the assembly of the
sons of El, and tkmn-w-snm) with the repeated address of the ritual
text K T U 1.40 = R S 1.003+. H e assumes as a working hypothesis
that K T U 1.65 is some sort of prayer. 2 3 Dietrich - Loretz take this
text as a scribal exercise. 24 In his elaborate study of this text Avishur
also concludes that the old view (of H.L. Ginsberg) that this text is
a prayer, seems to be closest to the truth. Comparison with the
Q p m r a n W a r Scroll (chapters 4 - 6 and 9, about names with as a
second element to be written on banners and weapons) leads him
to classify it as a list of war banners. 2 5 Because these banners are
'battle cries intended to arouse the deity to assist the warriors', this
text resembles a prayer. Although m u c h remains uncertain, one
should not rule out the possibility that this is indeed the text of a
prayer, related to sacrifices as mentioned in K T U 1.40, calling up
the gods (lines 1-5), appealing to the consideration of the supreme


C f . DE M O O R




165, a n d





Cf. L E V I N E
DE T A R R A G O N - R O B E R T S O N 1 9 9 7 , 2 9 9 , 3 0 1 . On the clapping of
hands serving 'to intensify the accompanying words and perhaps even to effectuate
the action' see Fox 1995.
Cf., for instance, B E R N H A R D T 1975, 239-40: 'Bittgesang an El und die Versammlung der Gtter'. See on this text A V I S H U R 1994, 308-9, who also lists and discusses
previous research.
Cf., for instance, B E R N H A R D T 1975, 239-40: 'Bittgesang an El und die Versammlung der Gtter'.
X E L L A 1 9 8 1 , 2 1 3 : 'almeno come ipotesi di lavoro, una sorta di "preghiera"'.
D I E T R I C H - L O R E T Z 1 9 8 1 , 6 4 - 7 ; in their translation of this text in 1 9 8 8 they
call it 'Opfer fr El und seine Reprsentanten'; note also the problems with classifying this text in ATI/ 2 , 91: 'scribal exercise?, invocation of II?, prayer?, incantation?'.





gods by referring to their noble character (6-9) and naming places

and divine attributes in and with which these words have to be
recited (lines 1 Off.). Del O l m o Lete sees it as a 'cultic invocation of
the divine panoply and to its apparent presence in the sanctuary'. 2 6
In his opinion this primarily concerns Baal's weapons, celebrated in
A more generally accepted example of a prayer in alphabetic
Ugaritic is the end of the ritual text K T U 1.119 = R S 24.266. 27
This text starts as a common prosaic ritual prescription about the
right time, place, and sort of sacrifice to the right god. In line 26
there is a transition to a direct address to Baal by referring to the
problem of a strong foe attacking the gates of the city. T h e style
changes here from prose to poetry. 28 T h e prayer (sit, line 34; cf. the
verb in K T U 1.19 = RS 3.322+ i 39) is introduced by the command: 'raise your eyes to Baal' (line 27). T h e request to drive away
the enemy is accompanied by a n u m b e r of vows and sacrifices by
the suppliant, in exchange for Baal's help. T h e text ends with the
statement, repeating the words at the beginning, that Baal will hear
the prayer.
K T U 1.123 = R S 24.271 29 is probably best described as a benediction,'" because of the repeated lm in the opening lines, followed
by a n u m b e r of divine names. T h e text seems to end in a similar
way, the last word being again slm. Lines 14ff. mention righteousness and mercy. This is reminiscent of K T U 1.65 and can be interpreted in the same way as expressing the hope for and confidence
in a positive attitude of the gods towards the one saying these words.
However, any interpretation of this text in its present severely damaged state must remain uncertain.






1992C, 255; cf. also



1992a, 228-9 = 1999,


See on this text X E L L A 1981, 25-34; W A T S O N 1984d, 360-2; DE M O O R 1987,
171-4; M I L L E R 1988; DEL O I . M O L E T E 1989; DEL O L M O L E T E 1992a, 197-205 =
1999, 292-306; P A R D E E 1993, 213-7; AVISHUR 1994, 253-6; W A T S O N 1996b.
O n this phenomenon see especially P A R D E E 1993a.
Cf. X E L L A 1 9 8 1 , 2 1 6 - 3 , with references to the older literature.


1 9 7 0 3 1 2 ; RAINEY




1404 and

suggest that it is a scribal exercise. KTU1,

possible genre: 'prayer, liturgy?, scribal exercise'.





gives as the


Binding hostile supernatural forces by incantations

T h e texts discussed under this heading are the ones that best fit the
narrow definition of an incantation given at the beginning. It concerns
independent texts with words of power used against evil forces from
the realm of gods and demons. T h e interpretation of these texts is
very difficult, not only because of the state of conservation of most
tablets, but also because of the genre of the texts, with unknown
vocabulary and often without a clear structure or line of thought.
T h e best example of an Ugaritic incantation is K T U 1.169 =
R I H 78/20. 3 1 Although there is much difference of opinion a m o n g
the interpreters about many details, it is generally accepted that we
are dealing here with a spell to drive off evil powers causing sickness, with the help of Baal, H o r o n and Ashera. It is not clear which
disease is meant here, 32 nor which power is causing it; according to
some it is indicated by dbbm in lines 1 and 9, although it is translated in different ways: 'flying demons' (de Moor), 'tormenters' (Pardee),
or 'accusers' (Fleming). Others (Dietrich - Loretz, Caquot) relate it
to Akkadian dabbu, 'word', and interpret it as a reference to the
words spoken to expel the (unnamed) demon. This difference of opinion returns in the interpretation of kspm (line 9) as 'sorcerers' indicating the black magic of demons, but according to others the magic
with which one can expel the forces of evil.33
T h e r e is more consensus about the verbs used in connection with
the expelling magic : ydy (line 1), 'to drive off', and gr/(line 9), 'chase
away'. Both are used in the legend of Keret in the repeated question 'who a m o n g the gods is able to cast out {ydy) the disease, to
expel (grs) the illness?' ( K T U 1.16 = R S 3.325+ vi 10-28). In line
10, in close connection with ksp and grs, we find the root hbr. This
is reminiscent of the use of hbr in the H e b r e w Bible and Akkadian
abru, 'to bind', in Mesopotamian incantations. Avishur points to
Deut. 18:10-11 and to Isa. 47:9 with the word pair hbrym || kpym,
'enchantments 11 spells', and to a similar pair in the Akkadian Maql-



A V I S H U R 1981; DE M O O R , 255-7; DE M O O R 1987, 183-6; D I E T R I C H 1988, 333-6; C A Q U O T 1989, 53-60; FLEMING 1991; DEL O L M O L E T E 1992a,
2 5 9 - 6 0 = 1999, 385-6; P A R D E E 1993, 211-3; FLEMING 1997; W Y A T T 1998C, 442~9.
D E M O O R 1980b, 257, and 1987, 184, n. 13, assumes that the patient suffered
from a 'cataleptic or epileptic seizure', Pardee thinks of 'male sexual disfunction'.
According to J E F F E R S 1 9 9 6 , 6 7 - 8 , one can leave open both possibilities.

incantation. 3 4 So here hbr would denote the negative influence of evil

spells. Avishur translates: 'Horon will expel the binders and the Youth
soothsayers', relating the last word (dctm) to Hebrew yd'ny. Dijkstra
interprets these terms in a similar way, but he assumes a positive
meaning: ' H o r o n be the enchanter, and the Young M a n the one
who provides knowledge. 35 In the hymn at the end of the Baal myth
this word pair hbr || dct ( K T U 1.6 = R S 2. [009]+ vi 49-50) would
have the same meaning. Most commentators, however, prefer the
more c o m m o n meaning of hbr, 'friend', and d't, 'intimate'. 3 6
In the text we hear of the one who has to recite the incantation:
'the /^-priest' (line 3). This title is also used in the colophon of the
Baal myth ( K T U 1.6 vi 57) and seems to refer to a high-ranking
teacher. 37 In K T U 1.40 = RS 1.002:32 we find the related verb parallel to dbh, 'to sacrifice'. According to some interpreters this officiant
used a staff as a magic device, 38 but the meaning of the word ht
denoting it (line 5) is, again, disputed. 39 This person executing the
incantation by word and probably also by gestures and other ritual
activities can be compared to the Mesopotamian incantation priest
called ipu. It is interesting to note that this exorcist is often mentioned in the colophon of the incantation texts as a scholar who
wrote and checked the tablet. 40
O t h e r correspondences with Mesopotamian incantations are the
use of similar metaphors, especially the spirits being said to leave
'like smoke' 41 and the naming of gods acting on behalf of the oppressed
against the evil spirits. In some Mesopotamian rituals the incantation priest even says that it is not he himself who speaks, but that
it is an incantation of Ea 42 or Ninkilil, 'lord of the incantation. 4 3 This

1981a, 22-3.






C f . DEL O L M O L E T E -








1988b, 335 translate 'Genossen || Komplizen'. According to J E F F E R S 1 9 9 6 , 33 both

suggested meanings of hbr are related: 'comrades can be linked together by sworn
words, oaths and the like'.

C f . VAN S O L D T


1988 and




Cf. F L E M I N G 1991, 148-50.

Cf. D E L O L M O L E T E - S A N M A R T I N 1996, 202-3.
See the texts mentioned in CAD A, II, 434, s.v. ipu a.
See, for instance, Maql V: 166-169 (translation by F R B E R 1987, 265). Cf.
A V I S H U R 1981, 18; F L E M I N G 1991, 146; and W A T S O N 1994b, 405-6.
Cf. CAD , II, 431-2, s.v. ipu a.2; CAD III, 90, s.v. siptu e.2'.
Cf., for instance, the recently discovered incantation against Lamatu, discussed



can be compared to K T U 1.169 = R I H 7 8 / 2 0 beginning with the

statement that it is the breath of Baal which drives out the evil spirits. Unfortunately, the text is broken here. Next to Baal a special
function seems to be reserved for H o r o n (lines 9 - 1 0 ) and Ashera
(line 16).
A n u m b e r of these basic elements of K T U 1.169 are also found in
another clear example of an incantation in alphabetic Ugaritic script:
K T U 1.82 = R S 15.134. 44 This seems to be a collection of six
different incantations, to be recited on different occasions, but also
sharing c o m m o n elements (such as the reference to the snake in lines
6 and 35). T h e fourth part is explicitly introduced as an incantation
with the technical term mnt (see p. 270) in its first line (= line 20).
Like K T U 1.169, this text is difficult to interpret, but it gives us
more information about the gods invoked to help and especially
about the demons to fight. T h e benign gods are Baal (lines 1 and 6),
his consort Anat (line 11 and twice in line 39), and the sun-goddess
Shapash (line 6). T h e evil forces they have to destroy are:
T u n n a n (line 1), known from the myth of Baal ( K T U 1.3 = R S
2. [014]+ iii 40) as a monstrous helper (dragon) of Yam, the god
of the sea, one of Baal's prime opponents. According to the myth
T u n n a n is slain by Anat. H e also seems to have been mentioned
in the small fragment K T U 1.83 = R S 16.266, next to Yam,
'bound (by c Anak?) on the heights of Lebanon'. 4 5 This reference
to T u n n a n , however, is uncertain, not only because of the poor
state of conservation of the tablet, but also because it is based on
a correction of the text in line 8, reading tnn instead of t'an.*6
Reshep (line 3), the god of pestilence, who is mentioned next to the
'lads of Y a m ' in the legend of Keret as the god who caused the
death of one of the king's wives ( K T U 1.14 = R S 2.[003]+ i 19).
Mot (line 5), the god of death. Next to Yam he is the other powerful opponent of Baal ( K T U 1.5 = R S 2.[022]+).














1988, 336-9; C A Q U O T 1989, 61-70; DEL O L M O L E T E 1992a,

251-5 = 1999, 373-9.
Cf. P I T A R D 1 9 9 8 ; DE M O O R 1 9 8 7 , 1 8 1 - 2 assumes that this text was part of an
C A Q U O T 1 9 8 9 , 2 8 - 3 0 states that tnn is the key term of the text, but he ignores
the fact that for this interpretation the text has to be changed.


Serpents (lines 6 and 35), who are the prime object of another Ugaritic incantation ( K T U 1.100 = R S 24.244 and K T U 1.107 = R S
24.251+; see below).
Creatures of H o r o n (line 13). This reading is uncertain, but the
name of Horon, who is the lord of the demons, returns in lines
27 and 41. Horon is mentioned in the legend of Keret as a threatening power in a curse: 'may H o r o n break your head' ( K T U 1.16 =
3.325+ vi 56-7; this phrase can also be restored in the broken
text K T U 1.2 = R S 3.367 i 7-8). In the Ugaritic incantations
H o r o n plays an ambivalent role: on the one hand he is a fearful
threat, on the other h a n d he can be called u p o n to take the
demonic threat away (cf. K T U 1.100 and K T U 1.169).
In the second part of the text the evil forces are indicated more
'poetically' as 'creatures of agitation' (lines 18 and 41), 'creatures
of insanity' (line 18), 'sons of disease (or: terror)' (line 23), '1egions(P)'
(line 26; cf. Mark 5:9), 'flies (or: accusers)' (line 26), 'those of the
flood(?)' (line 27), 'stupor(?)' (line 28), 'the fugitive' (line 38; cf.
K T U 1.5 = R S 2. [022]+ i 1, where it is used as epithet of a seamonster related to Yam).
It is not clear what is precisely the nature of the distress caused by
these evil forces. In the first lines there seems to be reference to
problems of a girl with her menstruation, that is with her fertility.
T h e names of the demons in the second part of the text point in
more general terms to disease and insanity. W h a t is clear is that
these evil forces have to be driven out (grs, lines 12 and 40; see also
K T U 1.169:9) or have to be b o u n d (rky, lines 10 and 38). Both verbs
are c o m m o n terms in this genre.
For the place of this and other incantations within the religion of
Ugarit it is important to note the close relation with the myth of
Baal. T h e batde described there of Baal and Anat, supported by Shapash, against Yam, Mot and their helpers does not appear to be something from a distant past. It has its repercussions on daily h u m a n life.
T h e victory over the forces of evil has to be gained time and again.
As was remarked above, Horon takes a central place in K T U 1.100 =
R S 24.244. 47 Fortunately, this text is well-preserved. It is in itself not


See on this text especially the elaborate study of

references to previous studies. Cf. also DE M O O R 1 9 8 7 ,




1 4 6 - 5 6 ; DIETRICH -



a pure incantation, but can be classified as a ritual in mythological

form. It contains, however, a n u m b e r of brief incantations indicated
as mnt. T h e text tells of a mare, 'the mother of the stallion'. She is
probably a goddess acting on behalf of her worshippers. W e are told
that she seeks support from the great deities of the Ugaritic pantheon against venomous serpents. T h e sun-goddess Shapash is indicated as her mother and acts as an intermediary. Each request ends
with the same incantation in order to charm (lf}; cf. the use of
Akkadian lahu, 'to whisper', together with siptu, 'incantation', 4 8 and
the use of Hebrew 1h, specifically related to the charming of snakes
i n j e r . 8:17; Ps. 58:5-6; Q o h . 10:11; and Sir. 12:13), expel (ydy) and
bind (ytq) the snake and its poison. Apparently, this incantation was
repeated eleven times, each time on the basis of another authority.
T h e twelfth, Horon, responds in a different way. With a magical
rite, using a m o n g other things a tamarisk and 'the tree of death',
he succeeds in letting the poison 'become weak and flow away'. This
climax of the text is, as is appropriate in magic texts, described with
a n u m b e r of puns. 49 T h e text ends with a dialogue between a groom
and a bride; apparently these are H o r o n and the 'mother of the
stallion'. T h e y speak about marriage with the serpents (a phallic symbol?) as bride-price.
A clearly related text is K T U 1.107 = R S 24.251+. 5 0 Here the
snake is called 'devourer', a c o m m o n designation of demons (lines
10 and 20).51 In the more elaborate mythological part of the text
Shapash plays a more active role. 32
T h e r e can be no doubt about H o r o n being viewed here in a positive way, be it that he is clearly not the first choice. T h e eleven
incantations preceding the final invocation seem to be meant to show
that no other choice was left than to go to Horon's 'fortress', probably an indication of his hardly accessible residence in the nether-

1988, 345-0; L E V I N E - DE T A R R A G O N 1988; C A Q U O T 1989, 79-94; P A R D E E 1997a

and W Y A T T 1998C, 378-87.
Cf. CAD , III, 90, s.v. siptu A.e.3'.
Cf. G R E A V E S 1994 and P A R D E E 1997a, 298, n. 26. See on this phenomenon
in Mesopotamian incantations F R B E R 1986.
Cf. X E L L A 1981, 241-0; P A R D E E 1988b, 227-56; and C A Q U O T 1989, 95-100.
Cf. DE M O O R 1 9 8 1 - 2 , I 1 0 . He refers, among other things, to the clear representation on a Phoenician amulet from Arslan Tash (seventh century BCE) of a
demoniacal man-eater.
O n this part of the text and a number of resemblances with the story of the
Garden of Eden according to the Hebrew Bible, see DE M O O R 1988b.

world. O n e
demons by
It is also
and ritual.

can compare this to Jesus being accused of casting out

Beelzebul, the prince of demons (Mark 3.22).
important to note the combination of myth, incantation,
This appears to be a c o m m o n feature of the Ugaritic

O n e of the new Ugaritic texts discovered in 1992 is the incantation

R S 92.2014, 5 3 which is in many respects similar to K T U 1.100,
K T U 1.107, and K T U 1.169. As in K T U 1.169 the offending evil
force remains unnamed: dy lyd\ 'the one not known'. H e is presented as a foaming snake and as a scorpion and is exorcised in a
magic rite with 'bits of sacred wood'. This is reminiscent of the
action taken by Horon against snakebite according to K T U 1.100
and K T U 1.107. In this way he prevents the serpent from coming
up (ely) and the scorpion from standing up (a new Ugaritic root qnr).
T h e second part of the text (lines 8-13) is an incantation against
dbbm and kpm. These words are also used in K T U 1.169. In R S
92.2014 they are mentioned next to rs\ 'the evil m a n ' and bn nm,
'son of m a n ' , which seems to be an indication of all possible men.
T h e incantation is directed against any evil word spoken: 'may they
pour it to the earth'.
T h e incantation is dedicated to Urtenu, the holder of the archive
to which this tablet belongs: 'for his body (gb), for his members (tmntf.
These two words also occur in K T U 1.169:5-6. T h e incantation has
a function in securing the physical well-being of Urtenu.
O n e final text to be mentioned within this framework of independent incantations is K T U 1.96 = RS 22.225. This is usually interpreted as a short mythological text about Anat literally or, what
seems to be more likely, figuratively devouring Baal. 54 T h e reference
to Anat was found in the first word of the tablet. New collations,
however, show that the first letters are cnn, not W.55 In the first edition of K T U it was suggested that cnn is a scribal error for cnt, but
in the second edition this 'rectification' was left out. A n u m b e r of
scholars now suggest that cnn is related to the repeated reference to
It is briefly described by B O R D R E U I L - P A R D E E 1995, 2 8 and 3 1 ; a first translation was offered by P A R D E E 1997a.
Cf. DE M O O R 1 9 8 7 , 1 0 9 - 1 0 ; and A S T O U R 1 9 8 8 , with a survey of previous
See now L E W I S 1996a, with excellent photographs and drawings.

n, 'eye', in lines 5 - 1 3 and explain the text as an incantation against

the evil eye. 56 This is a well-known object of incantations in Mesopotamia. 5 7 A clear example is also found in a later Phoenician incantation against 'the coming of the big eye' and with m a n y other
descripdons of the evil eye, just as in the second part of K T U 1.96. 58
Even more interesting within this comparison is that on the tablet
of the Phoenician incantation we see a drawing of a demon devouring the one he attacks. In the heading the d e m o n is called mzh,
'sucker', namely of blood. This has a counterpart in K T U 1.96:4-5
which states that the demoniac power eats the flesh and drinks the
blood of his victim.


Conjuring up the spirits of the dead

T h e demons to be driven out with the help of incantations are associated with death and the netherworld. As we have seen above, some
of them are represented as helpers or satellites of Yam and Mot (cf.
K T U 1.82 = R S 15.134:1, 5, 27, 38). It is very likely that as in
Mesopotamia the people of Ugarit feared the influence of malign
spirits of the dead. 5 9 From Mesopotamia we know many incantations with the object of expelling them. In Ugarit we hear more of
their positive counterpart: invoking the dead to ask their advice and
help. 60 This was also an act of veneration. By offering their sacrifices
and honouring them by calling their names, they hoped to prevent
hostilities from the dead towards the living.
T h e interpretation of the texts concerned is a matter of much dispute. According to some scholars there is not enough evidence to
speak of a cult of the dead. In their view there was probably no
more than a funerary cult intended to offer a good burial for the
deceased, helping them on their way to the netherworld; which is
to be clearly distinguished from any belief in supernatural power of

Cf. DEL O L M O L E T E 1992a, 255-9 = 1999, 379-84, and 1992b; his interpretation is accepted by W A T S O N 1992b, 367, n. 5; W A T S O N 1994b, 237; L E W I S 1996a,
118 and W Y A T T 1998C, 375-7. T h e suffixed -n is usually explained as a deictic element, comparable to Hurrian -ne\ cf. D I E T R I C H - L O R E T Z 1990a, 104.



O n this seventh century incantation from Arslan Tash see







1983 and


1986, 145-206 and


1993, 287-30.



the dead. 61 This is not the place to enter that discussion. M a n y of

these texts are already discussed elsewhere in this handbook. Within
the present context the survey can be confined to the elements related
to the incantation texts.
K T U 1.161 = R S 34.126 62 is a ritual text associated with the burial of a king of Ugarit. It reports the invocation of all possible ancestors, with many names of former kings, but also with more general
indications such as 'rephaim of the earth' and 'rephaim of old'.
Apparently, one was anxious not to forget any of the important
deceased ancestors. This can be compared to a similar p h e n o m e n o n
in the 'Genealogy of the H a m m u r a p i Dynasty' 6 3 listing all spirits of
the dead that are considered important to the well-being of the living king: the royal dead, the heroes and also the spirits who might
become hostile, namely 'any dynasty which is not recorded on this
tablet, and any soldier who fell while on his lord's service, princes,
princesses, all humanity, from the east to the west, who have no one
to care for them or to call their names'. 6 4 T h e idea behind this was
that the dead who remained u n n a m e d and uncared for could become
a threat to the living. So especially the unknown spirits of the dead
(cf. dy lydc, 'the unknown one' in R S 92.2014) were feared. Also the
fact that in K T U 1.161 of the great gods only the sun-goddess
Shapash is mentioned, is reminiscent of the incantations. She appears to be the most important intermediary between the living and
the dead.
T h e Rephaim texts ( K T U 1.20-22 = R S 3.348, 2. [019], 2.[024]),
only partly preserved, seem to describe a similar invocation and
actual gathering of the spirits of the dead. T h e state of the tablets
hardly allows any conclusion on their function. 6 5 T h e relation with
the legend of Aqhat suggests that the rephaim may have been called
up by the father of Aqhat on the occasion of the burial of his son.
T h e fact that the rephaim come together on the threshing floor may
indicate that they could be of help in restoring the fertility which
was lost at the death of Aqhat.


Cf. S C H M I D T 1994, 47 1 3 1 and P A R D E E 1996a.

See on this text especially B O R D R E U I L
P A R D E E 1 9 9 1 , 1 5 1 - 6 3 , and also the
recent study of T S U M U R A 1 9 9 3 . A survey of recent research can be found with







Cf. on this part of the text also L A M B E R T 1 9 6 8 .

Note the call for a 'minimalist' approach by P I T A R D 1992a and




K T U 1.108 = R S 24.252 can be regarded as an example of the

belief in ancient Ugarit that the spirits of the dead could be invoked
to help and bless the living. 66 Baal seems to be presented here as
the first of the rephaim. Together with Baal and other gods these
spirits of the dead enjoy a banquet presented to them in order to
propitiate them.
In K T U 1.124 = R S 24.272 we may find another way in which
the spirits of the dead could support the living. 67 T h r o u g h a mediator they give precise advice on how to cure a sick child. If this
interpretation is correct, it would offer a good illustration of the spirits of the dead acting according to the probable meaning of their
name: rp'um, 'healers'.


Elements of incantations in other texts

We have already come across the phenomenon of incantations embedded in other texts. Some of these also deserve our attention.
K T U 1.13 = R S 1.006 is interpreted by Dietrich - Loretz as an
incantation against infertility: 68 a hymn to Anat is followed by a
prayer for fertility, which is answered by a mythological fragment
about Anat and Baal solving a similar problem. In particular, the
urgent call for supernatural assistance (lines 23-29) can be regarded
as an incantation. Note also the special role in this context of 'messengers from heaven' (ml'ak smm, lines 21-22). T h e y belong to the
class of divine beings who, standing in between humans and the
great gods, often play a prominent role in incantations, either as
helpers or as offenders.
A combination of myth and ritual can also be found in K T U 1.23
= RS 2.002. Although there is no consensus about the interpretation of this text, there can be no doubt about the relation to the
question of fertility. 69 T h e text is associated by Dietrich - Loretz
with the incantations because of its beginning: }iqr'a, invoke (the


See for a survey of previous research on this 'Zankapfel der Ugaritologen'

1993, 293-5; cf. also P A R D E E 1988b, 75-118, and A V I S H U R 1994, 277-307.
See on this text D I E T R I C H - L O R E T Z 1990a, 205-40; L O R E T Z 1993, 289-93;
and, for a different interpretation, P A R D E E 1988b, 179-92.
D I E T R I C H - L O R E T Z 1988b, 339-42 with a reconstruction of the ritual; cf. also
DE M O O R 1980a. For a different interpretation see DEL O L M O L E T E 1981b and
M A R G A L I T 1995, 231-8.








gracious gods)'. In their opinion it is an incantation against malign

forces of nature. 7 0 Also the part of the text dealing with mt w sr,
'death and evil' (lines 8-11) resembles the incantations; in particular the reference to the binding (smd, line 10) of the demon-like god
points in this direction.
K T U 1.114 = RS 24.258 is labelled by Dietrich - Loretz as an
incantation for the medical treatment of drunkenness. 7 1 T h e text
describes El drinking himself nearly to death and two goddesses
finding a remedy to cure his sickness. T h e remedy seems to be
described in the last lines as a recipe for humans with the same
problem. 7 2 This and similar texts (cf. K T U 1.23; K T U 1.100; K T U
1.107) can teach us something about the use of myths in the daily
life of ancient Ugarit. Apparently, it was believed that reciting the
right story on the right occasion, combining it with the right prayer
and cultic acts, had magical power. Interpreted in this way, K T U
1.114 is related to 'pure' incantations. T o this can be added that in
lines 19-20 we hear of a demon-like figure threatening El in his
drunkenness (lines 19-20). This hby is described as 'the one with two
horns and a tail'. T h e name itself can be translated as 'crawler' and
seems to refer to a scorpion." In Ugaritic incantations this is a common indication of the evil force to be expelled.
In the myth about the moon-god Yarikh obtaining his bride Nikkal
( K T U 1.24 = R S 5.194) we find some elements related to the incantations in the second section of the text (lines 40-50), which is separated from the rest of the text by a horizontal line. It concerns a
hymn to goddesses called the Kathirat, daughters of the new moon
Hilal. T h e y are described as birds settling down on the trees. T h e
singer calls them by their names, stating that 'their list' (mnthn] cf.
the use of mnt in K T U 1.82:20 a n d K T U 1.100) is on his lips.
According to this text the Kathirat can be regarded as lower goddesses who are especially related to marital affairs. Their status is
comparable to that of demons and (deified) spirits of the dead, who
just like the Kathirat, are often compared to birds. 74 Calling their

1988b, 3 5 0 - 7 .
1988b, 342-5; cf. also P A R D E E 1988b, 13-74;
1997a; and C A T H C A R T 1996; W Y A T T 1998c, 4 0 4 - 1 3 .
Cf. W A T S O N 1990a.















names at the end of the myth about a divine wedding probably

functioned as a way to invoke their blessings on the occasion of a
h u m a n wedding.
For the sake of completeness another two texts deserve some attention. In the second edition of K T U the genre of tablet 1.86 = R S
18.041 is indicated as 'myth?, ritual?, incantation?'. 7 5 Recently, it has
been demonstrated that it is likely that we are dealing here with a
hippiatric text about breeding. 7 6
K T U 1.93 = RS 19.054 is listed by de M o o r a m o n g the incantations, interpreting it as a prayer to Anat for help in reciting his
incantation properly, that is, without stammering. 7 7 If this interpretation of the short and damaged text is correct, 78 then it would illustrate the importance of incantations in the religious life of the people
of Ugarit.




C f . DEL O L M O L E T E -





1987, 186-7.
For other interpretations see




CAQUOT 1989, 3 7 - 9

and the literature listed there.





T h e

M e r l o



X e l l a

The problem of the ritual documents

T h e Ugaritic texts which can be classed as ritual texts or have a ritual background did not at first attract the attention of scholars to
the same extent as the mythological texts, on which an enormous
bibliography has emerged. T h e very formal characteristics of these
documents have contributed to discouraging any approach to them.
T h e y are mostly schematic texts, written in a concise and technical
language which proceeds by allusions, using an accurate and precise
liturgical vocabulary which cannot be studied with the help of, for
example, the parallelism characteristic of poetry. In addition, there
is the frequent use of stereotyped formulae, the understanding of
which depends on actually identifying the rites to which they allude.
Lastly, the focus of the syntax is extremely difficult because 'prescriptive' rubrics and 'descriptive' sections alternate with no obvious
criteria and are accompanied by long lists of gods followed (often
asyndetically) by the victims or offerings intended for them.
A good stimulus to the study of ritual texts, which have been neglected
for so m a n y years (apart perhaps from a couple of specific contributions) 1 came from the article by Levine 2 on the possibility of identifying prescriptive and descriptive rituals, even if this distinction now
seems applicable only to a limited extent. This is either because
1 9 5 5 . Cf. also U R I E 1 9 5 9 .
1963. The distinction he proposes between 'prescriptive' rituals and
'descriptive' rituals is only acceptable in broad outline since it is clear that even a
description is significantly prescriptive in nature (libretto for ceremonies; cf. the use
of the imperfect/future). It is, thus, a spurious problem. Cf. also L E V I N E 1965; 1974,
8ff. and 1983.




several texts do not, strictly speaking, belong to either of the two

categories or else because the descriptive texts are also standard and
are also often devised and written down as m e m o r a n d a .
However, the situation has improved markedly over the last twenty
years, which have seen the first monographs on these documents as
well as a series of minor studies devoted to analysing the lexicon,
structure and function of the various texts, their implications for the
history of religions, the divinities involved in the rites, the typology
of the rites, etc. If we limit ourselves here only to wide-ranging works,
first of all comes the publication in 1979 of a long and excellent
comprehensive t r e a t m e n t of the ritual texts in the Supplment au
Dictionnaire de la Bible written by Caquot. 3 T h e n , at the beginning of
the eighties, there appeared the first monographs on the topic by de
Tarragon 4 (which is more discursive) and by Xella 5 (which is more
systematic). Besides a large n u m b e r of other minor studies (which
appeared chiefly in Ugant-Forschungen), a new wide-ranging contribution appeared in the section written by de T a r r a g o n in the book
edited by Caquot, de T a r r a g o n himself and Cunchillos (TO 2) which
was published in 1989 (although written a few years earlier) even
though it did not provide any remarkable new interpretations (in a
few rare instances even marking a regression). With del O l m o Lete's
monograph, 6 substantial progress has been achieved in spite of the
way the author has chosen to present the material (in practice it is
a comprehensive study of the religion of Ugarit). In this survey can
also be noted the translations m a d e by Dietrich and Loretz of a certain n u m b e r of ritual texts in the collection TUAT and elsewhere 7
as well as the inclusion of several cultic documents in the anthology
of de Moor 8 and in the study by Pardee of texts which he curiously
called 'para-mythological'. 9


Classifications of form and content

T h e ritual texts supply direct and extremely valuable information

about religious practices (chiefly but not exclusively royal and pub3

















lie) of ancient Ugarit, even if we reiterate that the material is difficult,

to be approached with caution, using clear methodological principles. Apart from attempts to consider these documents as a specific
'literary form', the main fixed points of departure for analysis are as
follows. In terms of the history of religions, they have an undoubtedly cultic character, while in terms of linguistics, they belong to a
form of linguistic expression that can be defined as 'chancellery language', 10 which places them on the same level as the economic and
administrative texts and the hippiatric texts.
As for their being documents with a religious purpose, the ritual
texts have to be studied against the background of all that we know
about the beliefs and cults of Ugarit during the Late Bronze Age,
with the proviso of also using, to the greatest extent possible, the
archaeological data from Ras S h a m r a and Ras Ibn Hani in order
to provide a better framework as to how the rites functioned, and who
were their participants and the recipients of the various ceremonies.
O f course, local mythology also has to be a constant and fundamental reference point for understanding the rites and their underlying ideology. T h e objection (which is often raised) that the 'literary'
texts reflect a stage of Ugaritic religion which is different and older
(than the ritual texts) in fact has very little foundation, for it is based
only on the formal opposition between poetic language and the nonpoetic or non-literary language" in which the ritual texts are cast.
In terms of the history of religions it is true that in some cases there
is a certain divergence, for example between the rank and personality of the divine figures described in the myths and those venerated in the cult. However, the history of religions teaches us that
the mythic dimension has its own laws, times and coordinates, such
that the actions or features of a god in the mythological narratives
need not necessary have an exact equivalent in the ritual universe
regulated by man in terms of his own needs (even such banal and
practical needs as the eating and [re-]distribution of meat by means
of ritualizing immolations and celebrating religious feasts). T o cite
only one striking case, it is precisely in the 'later' ritual texts that
the god Elincorrectly considered by some to be in decline with
the passage of timeoccupies a position of undoubted pre-eminence
over all the other deities. 12











If, instead, we consider the content of the texts rather than their
outward form, mythology and culdc documentation comprise two
complementary and fairly organic aspects of Ugaritic religion which,
as in all the cultures of the ancient world, is expressed differently at
different levels. Lastly, as regards the matter of the relationships
between myth and ritual, which are inextricably connected, it is
sufficient to r e m e m b e r the existence of texts such as K T U 1.23 =
R S 2.002, correctly defined as 'cultic myth'; 1 3 or, the fact that recitations, prayers, exorcisms and incantations are religious acts which
can all be projected into the mythic dimension (while the recitation
of a myth is itself a rite!).
Besides these considerations, another important fact to keep in
mind in the study of the Ugaritic rituals is the comparative dimension within the religious traditions of Syria-Palestine and Northern
Mesopotamia. With the increase in our knowledge of the religion
and sacrificial system of centres such as Ebla, 14 Emar 1 5 as well as
Mari (with the requisite changes), 16 it is necessary to get away from
the usual and repeated references to the Bible in order to reconstruct the religious tradition peculiar to Syria, the consistency and
essential continuity of which are perceptible.
It should also be r e m e m b e r e d that additional and often direct
information on the cult and on various rites, both sacrificial and
non-sacrificial, can be gained from other types of text, i.e. the economic and administrative texts, 17 the letters and the mythological
texts. T h e most famous example in the last category comprises the
rites carried out by Keret ( K T U 1.14 = R S 2. [003]+ ii 50ff. and
parallels) which, though to some extent awaiting proof, are undeniably connected with cultic practice.
However, it is understandable how, faced with a mass of documents sharing a definite (and more or less direct) connection with
the cult, but without uniform formal characteristics, specialists have
attempted to sort this material, proposing classifications and subclassifications of various kinds.
In a monograph written in 1 9 8 1 ( X e l l a 1 9 8 1 ) one of the authors
set out a subdivision of the texts as follows: 1) monthly liturgies and


D E L O L M O L E T E 1992a, 15 = 1999, 15.

Cf. in general F R O N Z A R O L I 1993 and P O M P O N I O - X E L L A 1997.
Cf. especially F L E M I N G 1 9 9 2 and, for example, D I E T R I C H 1 9 9 0 .
An excellent synthesis in D U R A N D 1 9 9 5 .
S A N M A R T I N 1998. On these aspects see below.

lists of offerings; 2) divination texts and oracles; 3) prayers; 4) incantations; 5) atonement sacrifices; 6) liturgies for kings; 7) votive texts;
8) lists of gods. 18 This proposed classification has largely been followed by G . del O l m o Lete ( d e l O l m o L e t e 1992a = 1999) with
a series of additional subdivisions which refine the original grid still
f u r t h e r , even for example distinguishing prescriptive rituals into
sacrificial or non-sacrificial, pure or mixed, single or multiple, etc. 19
T h e table proposed by the Spanish scholar is undoubtedly a good
theoretical grid for classification, provided that it is not taken rigidly
as a formal reference point. In this Handbook, which has a more general approach, we shall retain the distinction between prescriptive
and recited rituals. Whether the sacrificial action does or does not
involve blood is a further classification which, in some sense, cuts
across the others and can connote a n u m b e r of other ceremonies.
Likewise, whether the ritual actions are more or less complex, the
length of time they take and where they are performed, are factors
which do not alter the basic typology.
Ultimately it has to be said that our as yet imperfect knowledge
of the sacrificial vocabulary of Ugarit and our continuing ignorance
of the deep structures of Syrian religion in the Bronze Age should
put us on guard against claims of elaborate classification which are
too detailed and too systematic.


Rite and sacrificial material: terminology and


As indicated above, to describe or record various actions of the cult,

the Ugaritic ritual texts use a series of technical terms which sometimes have a more generic meaning and sometimes are very technical. Even if we are still very far from having resolved all the
problems of interpretation, the meanings of some terms have now
been determined with enough certainty. 20 T h e word dbh is a prime

X E L L A 1 9 8 1 ; in fact there is also a section on Hurro-Ugaritic rituals, due only

to the bilingualism marking these documents, recently studied in a systematic way
particularly by M. Dietrich, W. Mayer and M. Dijkstra in various publications
(chiefly UF and A O A T / A L A S P ) .
A different approach has been adopted by D I E T R I C H - L O R E T Z 1988b, 300ff.,
which however is also reductive because it is determined by the anthological nature
of the series (TUAT) in which it appears.
X E L L A 1989; D E L O L M O L E T E 1992a, 19-22 = 1999, 20 1; 1995. Nevertheless
there are still quite a few disputed or even completely enigmatic cases, for example,

example of a term to denote the sacrificial ceremony which is both

specific and generic at the same time.
O f as yet unknown etymology, 21 dbh (generally attested as a noun
although it also occurs in verbal form) is a key term in the Ugaritic
sacrificial lexicon. It has been suggested that a whole series of terms
(for example,ytn, kbd, nkt, np, nrr, na, ntk, ql, qrb, lm, t (?), t'y) should
be considered essentially as its synonyms, 22 but each of t h e m a p a r t
from proven cases of functional synonymityis of course distinct and
must be investigated separately as a specific element of a highly technical lexicon. In fact, dbh is a polysemantic term, as noted already
by Aistleitner, 23 who correctly distinguished between (a) the meaning
'schlachten/Schlachtopfer' and (b) the meaning 'Gastmahl essen/Mahl'.
T h e Ugaritic term is not semantically related to Hebrew zbh (verb
and noun) but instead to Akk. zbu, 'food offering', which in turn is
synonymous with naptnu (originally an accounting term used for food
rations, used in the cult but also in non-cultic texts). T h e so-called
polyglot vocabulary of Ug 5 (137 iii 6)24 gives the equivalents EZEN =
i-si-nu = e-\l]i = da-ab-hu, from which can be deduced the clear
meaning 'feast', 'soire'. This meaning is abundantly confirmed by
the use of the term in the ritual texts. It is a sacrificial meal, i.e. a
sacred banquet, as has been proved by several terms which are parallel or actually synonymous to itdgt, msd, trm, esrt. However, the
meaning of the word sometimes seems to be more generic (precisely
'feast' in general), 23 since dbh sometimes includes not only foodstuffs
(for example, cf. oil and honey in K T U 1.41 = RS 1.003+:20-l)
but also objects, clothes or metals. T h u s it is an offering in a generic
sense even if, as seems to be the case, the gift par excellence to the
gods is still food (cf. the implicit equivalence dbh kispum in K T U
1.142 = R S 24.323). 26
An excellent example of a grandiose and complex dbh is provided
by K T U 1.148 = R S 24.643, 27 the dbh of Sapunu, which seems to

cf. terms such as iyn, m'rb (in spite of its apparently clear etymology), sin, trmt and
others as well.



1992a, 2 0 - 1 = 1999, 21-3, although he notes that the synonymy refers only to the basic meaning of 'to offer'.





Cf. H U E H N E R G A R D 1987b, 117; VAN S O L D T 1991a, 303. Cf. 4.5.

Cf. for example the opening line of K T U 1.161 = RS 34.126: spr. dbh. zlm.
Cf. DLU, 128a for the references.
D E L O L M O L E T E 1992a, 88-91 = 1999, 129-33. (with previous bibliography).





include a whole series of different rites within the ceremony. In this

case, however, it seems that the iZmm-sacrifice does not form part of
the general typology of the dbh, but it cannot be excluded completely
that in the first section of this text the offerings are to be considered as performed as a /r/?-sacrifice (which was not mentioned because
it was implicit). A text such as K T U 1.170 = R I H 7 8 / 1 1 shows
that the dbh included sacrifices such as the srp, as is also indicated
by the semantic field of the Hurrian term athl, sometimes used as
equivalent to dbh and sometimes as equivalent to srp.2*
If, out of several possible cases, we consider the so-called atonement
ritual K T U 1.40 + 84 + 121 + 122 + 154 = RS 1.002,29 significandy,
the terms dbh and tc denote the inner cultic sphere in which the
Ugaritic homo religiosus has committed sin.
Above we have referred to attempts at formal classification of the
ritual texts. Without going into over-precise subdivisions, we note
that the ritual actions can be simply classed into the following categories: 30 ritual blood sacrifices, bloodless sacrificial offerings, processions, enthronements-investitures, cultic meals. T o these can be
added texts or parts of texts comprising recitation (including prayers).
T h e first category is thus represented by ritual actions involving blood.
A m o n g these, the designation attested most often in the rituals of
Ugarit is provided by the brace of terms srp and lmm, which almost
always occur together in close succession. These are two specific
terms for sacrifice, the first of which is evidently connected with the
action of 'burning' the victim (as the root *srp shows) as is the case
in the other Semitic languages (although this does not mean that it
was always a holocaust); the second term, lmm, is possibly to be
translated 'communion sacrifice' or 'peace sacrifice', based on etymology and also on comparison with H e b r e w elamm.U Explicit
confirmation that (with the requisite modifications) in Ugarit also it
is a communal meal of the offerers, comes from K T U 1.115 = R S
24.260:9-10: l il bt lmm kl l y Ihm bh, 'a ram to the god of the
(royal) house as a //mm-offering; all eat it'. 32






1981, 251-76; DEL O L M O L E T E , 1992a, 9 9 - 1 0 9 = 1999, 1 4 4 - 6 0 ; DE

M O O R - SANDERS 1991; W Y A T T 1998C, 342-7.
D E T A R R A G O N 1980, 55-129; DEL O L M O L E T E 1992a, 28-32 = 1999, 3 4 - 4 0 ;
cf. also DEL O L M O L E T E 1995, passim.
Current views and bibliography in DEL O L M O L E T E 1992a, 3 0 - 1 and n. 81 =








1 0 8 ; DEL O L M O L E T E




= 1999, 257-64.

Also part of the other actions of the blood ritual is the act of
'immolation' (nkt) or 'slaughter' (qll) of the victim. Nevertheless, the
ritual texts provide evidence of further types of sacrificial offerings
of which the meaning is not always precise, as in the cases of npt
(related perhaps to Hebrew tenp),33 tzg (which may denote either
the victim or the rite)34 and also t'y, mentioned already, a polysemantic term which also seems to denote an offering connected with
the royal cult. 35
Besides rites involving blood, the Ugaritic texts also mention bloodless ritual actions. A m o n g the ceremonies which do not involve the
sprinkling of blood can be mentioned here the fairly widespread
practice of performing 'libations' (the verb *ntk or the derived noun
mtk), especially of water and wine, documented in several texts (cf.
for example, K T U 1.41 = RS 1.003+:12; K T U 1.107 = RS 24.251:46;
K T U 1.119 = R S 24.266:25), although they provide no information
about the details of the ceremonies. T h e bloodless offerings also
include, of course, offerings of vegetable food, and of various types
of object, metal or cloth, which are amply documented in the ritual texts as well as in the economic and administrative texts. 36
A further category of ritual actions comprises the processions, identifiable by a series of terms (for example, yrd, hlk, Iqh, cly, crb, etc.)
and correcdy included by G. del O l m o Lete among the 'non-sacrificial
liturgies'. 37 T h e processions mentioned in the Ugaritic rituals have
in fact as a central point a series of ceremonial actions in honour
of divine statues, 38 in which, as usual, the role carried out by the
king and his family is completely in the foreground. In this connection, the ritual K T U 1.43 = R S 1.005 can be mentioned which
is focussed completely on the procession of divine statues. 39 It begins
with k t'rb* 'ttrt. hr. gb/bt mlk 'When 'Attart of Hurri 4 0 enters the gb
of the royal palace', and has the important conclusion (lines 24-26):
mlk. ylk. Iqh. ilm || atr. ilm. ylk. penm./mlk. p*c*nm. yl[k]/sbc pamt. Iklhm,
'the king will go to take (the statues o f ) the gods || behind (the stat33




1981, 39-40.
D E L O L M O L E T E 1988a; cf. also F R E I L I C H 1992.
D E L O I . M O L E T E - SANMARTIN 1998; cf. in general
1999, passim.
D E L O L M O L E T E 1992a, 96 = 1999, 1 4 0 - 1 .
D E T A R R A G O N 1 9 8 0 , 9 8 - 1 1 2 ; in general D I E T R I C H
D E L O L M O L E T E 1992a, 189-94 = 1999, 282-91.


C f . BONNET -






1992a =

ues o f ) the gods he will go on f o o t / t h e king will go on foot/seven

times with all of them'.
Further references to ceremonies of transporting statues are also
given in K T U 1.91 = R S 19.015, a key text (classed as administrative) which lists concisely various rites which can be identified in
more detail in other documents. 4 1 Line 10: k t'rb. cttrt. d. bt m[lk]
'When 'Attart of the steppe enters the royal palace' (cf. K T U 1.148 =
R S 24.643:18-22) and in line 11: k t'rbn. rpm. bt. mlk, 'When the
Raaps enter the royal palace' (cf. K T U 1.106 = R S 24.250+, as
well as K T U 1.43 = RS 1.005:9ff.). K T U 1.112 = R S 24.256: 42
6 - 9 , where images of the deities (genuine gods and divinized ancestors) are moved around in various cultic installations as part of the
dynastic cult, is not exactly of the same type.
Another category of ritual actions is that of investiture-enthronement.4:i
Here we are referring, for example, to actions described by verbs
such as ytb 'to sit (down)' or lbs 'to dress/be dressed' (for the semantic field of clothes or ritual dressing cf. especially azr and nps, as well
as other terms) but which, in a religious context, refer to specific
ceremonies the details of which are unknown to us. A clear case of
royal investiture can be identified, for example, in K T U 1.41 = R S
1.003+:53-5, sbu. p whl mlk/w l*[b]n. spm. w mh[pn]h*
b*\t\ w km. ity[u. L] mmyd[h] 'the sun sets and the king is desacralized and, robed splendidly and with his face cleansed, they shall
enthrone him (lit. they shall make him return) in the palace and
once there, he shall raise his hands towards the sky'. 44 Although it
mentions the royal throne, the expression 'and at night the throne
is prepared' alludes to a ritual action which is not completely clear
in K T U 1.106 = RS 24.250+:26~8 (w III tcr[k] ksu).
Important moments in the liturgies, apparently without involving
blood, can be recognized in the invocations, recitations and prayers
(cf. for example dn (?), nsa + yd, pth. + yd (?), sh, qra, r), in acts of
divination and oracles (phy (?), rgm, r, ttb + rgm, etc.) and in expiation rites (cf. K T U 1.40 + 84 + 121 + 122 + 154 = RS 1.002,
cited already and marked by its peculiar structure and lexicon).

1979; DEL O L M O L E T E 1992a, 173-7 = 1999, 257-64.

Rites to be celebrated in the month of tjyr (likewise K T U 1 . 1 0 5 = RS
On the passage quoted cf. X E L L A 1 9 8 8 .
D E L O I . M O L E T E 1992a, 96 = 1999, 1 4 1 .
D E L O L M O L E T E 1992a, 83-5 = 1999, 122-5.




Finally, a m o n g the ritual actions, communal meals can be mentioned (cf. what has been said above concerning dbh and lmrr) which
frequendy acquire religious significance, and include the consumption of drinks (specified by terms such as kly, Ihm, nsl (?), (db, Cr, sty),
either as a convivial occasion which joins together men, the gods
and the spirits of the ancestors, or else as a ritualized and regulated
occasion where food (especially meat) is eaten.
As for the sacrificial material, 45 in rites involving blood, the offerings
of animals were substantially similar to those known in the other
religious traditions of the ancient N e a r East. T h e animals offered
most frequently are bovids (the ox, alp, the bull, tr, the cow, gdlt,
lit. 'female head of cattle'), ovines, denoted genetically as sin (the
ram, i, the sheep, dqt, lit. 'head of small catde' or tat, she-goat, cz),
birds, called generically csr, with the dove, ytnt or the turtle-dove, tr
specified. However other kinds of animals also occur such as donkeys, V and also fish, dg. Besides complete animals, the various parts
of the victim were offered, limbs and entrails (ap, nps, lb, kbd, mtnt
are the easiest to identify) as studied in detail by G. del O l m o Lete. 46
Animal offerings are certainly not the only ones attested in the
Ugaritic rituals and several times gifts were dedicated to the gods
either of vegetables such as wine (yn), oil (mr), e m m e r (ksm), flour
(qmh) honey (nbt), or else of precious metals such as silver (ksp) and
gold (}}rs) or even objects in c o m m o n use including articles of clothing and crockery.


The role of the king

By now it is well known that the king had a role of particular importance within the Ugaritic cult and was by far the principal officiant. 47
T h e texts of Ugarit, in fact, show not only how the palace has control over the personnel appointed to the cult but how the king himself is often the main celebrant within a liturgy and how frequently
the ceremonies take place in locations and internal buildings actually within the royal palace. Moreover, the importance of the dynas-

Cf., for example,
1999, 40-2.

1980, 31-54;


1992a, 32~3 =

L E T E 1989b.
1984; Y O N 1985; D E L O L M O L E T E 1992a, 115-95 = 1999, 166-291;
L E T E 1993b; A B O U D 1994, 123-92.







tic cult together with the veneration of the divinized royal ancestors
has emerged as one of the most typical aspects which characterize
the religion of Ugarit in a peculiar symbiosis between the living a n d
the d e a d for the c o m m o n good. 4 8
T h e high n u m b e r of rituals f o u n d in Ugarit testify to a series of
liturgies where the ritual actions clearly refer to the king, expressed
frequently by the use of fixed 'ritual formulae' which indicate his
state of (ritual) purity a n d refer to special m o m e n t s in the day. 49
T h e best known and best attested form of ritual action where the
king is seen as the protagonist is the one concerning his ritual purification. This must certainly have been achieved by means of special
ablutions, expressed by the f o r m u l a yrths mlk bn 'the king washes
< a n d is> purified'. This formula often introduces a series of ceremonies in which the king takes part, at the close of which there
occurs, connected with nightfall or sunset (sba/u p, crb p) the further formula whl mlk 'and the king is desacralized'. T h u s the purification
of the king seems to be a prerequisite for him to be able to p e r f o r m ,
assist at or take part in the ritual (cf., for example, K T U 1.119 =
R S 2 4 . 2 6 6 : 4 - 6 ; K T U 1.112 = R S 2 4 . 2 5 6 : 1 0 - 5 , etc.).
A n o t h e r series of 'formulae' which are quite similar to each other
even if not exactly identical, allude to the king's role in oracle practice
(mlk brr rgm y ttb/rgm y ttb mlk bn, ttb rgm whl mlk) probably acting as
mediator for the replies a n d always in conditions of ritual purity. 0 "
T h e r e are m a n y other examples of the sovereign being involved
in the cult. A m o n g the most i m p o r t a n t a r e K T U 1.119 = R S
2 4 . 2 6 6 : 1 3 - 4 in which 'the king sacrifices in the temple of , or
K T U 1.164 = R I H 7 7 / 2 b + ' W h e n the king sacrifices in the hmn
(i.e. the palace chapel)'; the king takes part in processions, as in
K T U 1.43 = R S 1.005:23-5, cited already; a cantor has to sing in
front of the king, w h o has his h a n d s spread out: K T U 1.106 = R S
24.250+: 15-7; the king's throne is p r e p a r e d at night, probably to
allow him to be seated d u r i n g the p e r f o r m a n c e of nocturnal rites:
K T U 1.106 = R S 2 4 . 2 5 0 + : 2 7 - 8 , a n d so on.


Restated by D E L O L M O L E T E 1992a, 130 4 = 1999, 192-8 (in respect of K T U

1.161 = RS 34.126) and passim, where five whole chapters (3- 6) are devoted to the
religion and cult of the king.
" X E L L A 1984c.
D E L O L M O L E T E 1992a, 22 = 1999, 24.

T h e ceremony described at the end of K T U 1.41 = R S 1.003:505 is particularly interesting; here the cultic role of the king is very
obvious: the rite, which is still difficult to identify exactly, is performed in the month of riyn and takes place on the terrace of the
royal palace (it is less likely that it was a temple); it exhibits remarkable similarities with the H e b r e w ritual of the New Year which
was also celebrated on the day of the full moon in the first month
of the year, at the season of the wine harvest, with the construction
of huts. 52

Also worthy of mention is the possible sacred marriage ceremony

mentioned in K T U 1.132 = R S 24.291, studied again recently (with
new proposals) by Dietrich and Mayer, 5 3 in which it seems that the
union of the king with the goddess Pidrayu was celebrated. She was
the daughter of Baal and evidently must have embodied the ideal
wife at the highest level. 54
Besides the rites in which the king took part, the locations where
the liturgy took place were often included within the palace complex and the chapels attached to it.
O n e of the places in which ritual actions involving the king's presence were often performed is the hmn, undoubtedly a chapel reserved
for the royal and dynastic cult, a raised and covered structure, a
sort of earthly projection of the heavenly dwelling, which forms part
of the royal palace and was used chiefly for the royal dynastic cult. 55
T h e r e are other places in which the rituals were performed and
probably must have belonged to the king's residence such as the
'terrace' (gg) mentioned above or the 'garden' (gn). This second term,
which is also a name for a m o n t h in the Ugaritic calendar when an
impressive n u m b e r of ceremonies are performed, is already attested
in the texts from Ebla' 6 and denotes a cultic area inside the royal
palace which was a sort of cemetery, the location for funerary rites
intended for the ancestors and the chthonian deities. 37


Lines 50-55 of this tablet are separated from lines 1-49. Only the latter have
a duplicate in K T U 1.87 = RS 18.056.
D E L O L M O L E T E 1992a, 83-5 = 1999, 122-5.
D I E T R I C H - M A Y E R 1996a; cf. the new proposal to read bb*t mlk instead of bit
mlk in line 3.
D E L O L M O L E T E 1992a, 1 4 3 = 1999, 212.
D E L O L M O L E T E 1984d; X E L L A 1991, I69ff. and passim.
X E L L A 1995c.
Cf. for example S P R O N K 1986; L E W I S 1989; D E L O L M O L E T E 1992a, 149-56 =
1999, 219-32. For a more reserved critical attitude cf. VAN DER T O O R N 1991.

T h e reference to gn brings us to the ceremonial role of the king

in his capacity as principal celebrant, a role exercised even more
within the funerary cult in honour of the dead and deified kings of
the dynasty. This fundamental aspect of Ugaritic religion is attested
in a whole series of indications and texts (three, in particular K T U
1.108 = RS 24.252, K T U 1.113 = RS 24.257 and K T U 1.161 =
RS 34.126 which, although their general meaning and many details
are still disputed, throw considerable light on the phenomenon which
seems to be peculiar to the Syrian region).
T h e first of these documents ( K T U 1.108 = RS 24.252) 58 seems
in fact to attribute to the dead king the title of rpu mlk clm 'Rapiu,
eternal king' (lines 1 and 21-22). Unfortunately, the break and the
theoretical possibility of attributing the title rpu to a god (in this case
Baal) rather than to the king makes the interpretation of the whole
text very uncertain. K T U 1.113 = RS 24.257 (see 7.2) is a clear
testimony that the dead kings were considered as divine. Lastly, K T U
1.161 = RS 34.126 59 is introduced as is known as the 'liturgy of
sacrifice of the shades [= protective spirits]' (spr dbh zlm), i.e. a 'libretto'
of the funeral celebration which was performed to accompany the
descent of the dead king into the underworld. T h e liturgy begins
with the invocation of the Rapiuma, the ancestors belonging to the
dynasty, after which we are present at a lament with a ritual meal
followed by the c o m m a n d given by the goddess apa to the dead
king (here probably Niqmaddu III, line 26) to descend into the underworld. T h e text then continues with the sacrifice of a bird (probably as an offering for the dead king) performed seven times and it
ends with a propitiatory blessing of well-being towards the king, the
ruling house and the whole city of Ugarit. It is precisely this final
blessing which shows us the purpose of the cult of the ancestors
which the Ugaritic kings carried out: it expresses in fact the concern for obtaining prosperity, protection and oracular responses 1 '"
from the Rapiuma, fittingly honoured for that purpose. 61
Finally, it can be noted that the role of the royal house in Ugaritic
liturgy does not seem to have been strictly confined to the king.
1988b, 75-118; ( C A Q U O T - ) DE T A R R A G O N
(CUNCHILLOS) 1989, 1 1 1 - 8 ;
1992a, 126-30 = 1999, 184-92, W Y A T T 1998C, 395-8.
D E L O L M O L E T E 1992a, 130-4 = 1999, 192-8, with previous bibliography.
See also W Y A T T 1998C, 4 3 0 - 4 1 .
T R O P P E R 1989a; L O R E T Z 1993.
T h e 'minimalist' position recently adopted by P A R D E E 1996a, 281 does not
seem completely justified.



Already K T U 1.112 = RS 24.256 studied above shows that the

whole royal family took part in ritual actions. In addition, K T U
1.170 = R I H 7 8 / 1 1 , even in a broken context, seems to show the
existence of a sacrifice carried out by the queen, which is confirmed
by the administrative text K T U 4.149 = R S 15.039:14-6 which
records quantities of wine for the sacrifice by the queen on sown


Other cult personnel

Besides the foreground role performed by the king as officiant, we

also know of other personnel appointed to the cult even if, surprisingly enough, neither khnm nor rb khnm ever feature in the ritual
texts, apart from a the doubtful reference to a 'throng of priests',
tltt khn[m], in the second broken incantation against serpents' venom,
K T U 1.107 = R S 24.251 + : 18. Otherwise note can be taken of a
'cantor', sr, in K T U 1.106 = R S 24.250+: 15, a qdl K T U 1.112 =
R S 24.246:21, a category which can perhaps be defined as 'purifiers',
mhllm in K T U 1.119 = RS 24.266:23, whereas a certain exorcist
termed mihi is the h u m a n protagonist in the great incantation against
serpents, K T U 1.100 = R S 24.244.
All this personnel belongs to the category of dependents of the
royal house (bnl mlkf3 and therefore, strictly speaking, a class of
priests independent of the king is not attested in Ugarit. 64
T h e rare mention of personnel with religious functions in the cultic texts is partly completed by references which can be derived from
the administrative and economic texts, 65 although there is no mention
of their cultic duties. In this category of documents we find several
references to 'priests', khnm (in the Akkadian texts: l u . m e s s a n g a ) ; individually or collectively, in 'community', dr khnm ( K T U 4.357 = R S
18.046:24), with their 'chief' rb khnm (in the Akkadian texts: l . u g u l a
s a n g a ) , a title which occurs also on the famous hatchets
and in
the colophon to the 6th tablet of the Baal cycle mentioned above.


Already noted in X E L L A 1 9 8 1 , 1 4 9 .
Cf. for example H E L T Z E R 1982, 131-9; LIPINSKI 1988.
According to the colophon of the tablet K T U 1.6 = RS 2.[009]+ vi 55-7,
even the high priest Attanu-prln was a dependent of king Niqmaddu, albeit at the
highest level. T h e title nqd can be related to the cult only hypothetically.




See most recently






T h e r e are also the 'consecrated ones', qdm, for w h o m a divinatory function has also been proposed, hypothetically, 67 and then the
'singers', rrn, personnel connected in various ways with music in the
cult and perhaps also outside the cult and lastly the 'water carriers
of the sanctuary', ib mqdt, a function which has parallels in the
Hebrew world, in Hcllenistic-Roman Syria and elsewhere.


The times of the celebrations

As is almost the general rule in the ancient Near East, the calendar 6 8 of Ugaritic liturgical celebrations is based on the lunar cycle
and therefore the days in which the cult takes on greater importance are evidently at the beginning or middle of the month, corresponding to the new moon (ym hdt, i.e. the first day of the month)
or the full moon (ym mlat, i.e. the fifteenth day of the month). O n
these days great sacrifices are carried out, accompanied by the usual
rite of purification of the king, as shown for example by K T U 1.46 =
RS 1.009, K T U 1.109 = R S 24.253 and K T U 1.41 = R S 1.003+.
However, there are also other liturgies for other days of the month,
often in 'weekly' cycles (i.e. a quarter of a month), but also on several other days of the month. And then some rites took place in
the evening or at night, as can be deduced from the indications I II
and Ipn II.
With regard to the annual cycle of the cult, however, unfortunately we are not yet completely certain of the exact sequence of
the months of the Ugaritic calendar since the local names are considerably different from the standard names of Mesopotamian tradition. And comparison with what we know of the months of various
Eblaite calendars from over a millennium earlier does not help much
either. Some scholars have tried to reconstruct a particular seasonal
liturgical sequence on the basis of the mythological texts, 69 but this
attempt has not been accepted by scholars. As far as the strictly ritual texts is concerncd, we know of liturgical texts related to certain

S A N M A R T I N 1 9 9 8 , 1 8 0 - 1 . On the root cf. X E L L A 1 9 8 2 .

1980, 17-30; D E L O I . M O L E T E 1992a, 2 2 - 4 = 1999, 2 4 7 ;
C O H E N 1993, 3 7 7 8 3 ; A R N A U D 1993b. For chronological indications to the cult
which can be obtained from the economic and administrative texts cf. DEL O L M O
L E T E - S A N M A R T I N 1998, 186-8.
Especially DE M O O R 1971. Cf. also DE M O O R 1972 and 1988a.






specific m o n t h s such as the m o n t h s of ib'lt ( K T U 1.119 = R S

24.266:1-17), nql ( K T U 1.138 = R S 24.298) and hyr, the second
m o n t h of spring corresponding to A p r i l / M a y ( K T U 1.105 = R S
24.249 and K T U 1.112 = R S 24.256). In addition we are acquainted
with the ritual of the m o n t h of ris yn ( K T U 1.41 = R S 1 . 0 0 3 / K T U
1.87 = R S 18.056). This last m o n t h ('first of the wine') fell in the
time of the wine harvest and corresponded to the month S e p t e m b e r /
October; it was probably the beginning of the cultic year in Ugarit.
Its ritual, with an extremely long list of sacrifices, is marked by an
initial and highly symbolic offering of a bunch of grapes, undoubtedly to be understood as firstfruits, to the god El.


The places for the celebrations

T h e Ugaritic liturgy certainly did not only take place in the temples. In fact the ritual texts mention several sacred places, some of
which have been confirmed from archaeology whereas others still
remain unidentified. 70 For reasons of completeness it is necessary to
note, however, that certain 'cult places', identified on the basis of
archaeological evidence, actually have no equivalents in the specific
terms of the ritual texts. 71
T h e temple of Baal, identified with one of the two great temples
on the acropolis, takes on a foreground role in the ritual texts ( K T U
1.119 = R S 24.266; K T U 1.105 = R S 24.249:19; K T U 1.104 =
R S 24.248:13; K T U 1.109 = RS 24.253:11), a n d also in the mythological texts (especially K T U 1.4 = R S 2. [008]+; but see also K T U
1.17 = R S 2. [004] i 31). It is built on a platform which is approached
by means of a monumental stairway; it comprises a vestibule which
comes before the naos (or inner cella) and, in the southwest corner,
is preceded by a structure which can be explained as the holy of
holies. In front of the entrance, within the court surrounding the
sanctuary to the south, there is a structure measuring 2 x 2 metres,
usually explained as an altar (it is thought that it may be the altar


The first of these is the so-called sanctuary of the Hurrian gods close to the
royal palace, the nature and cultic function of which seem to be certain; the second is the so-called sanctuary of the rhyta, located in the residential area of the city;
cf. Y O N 1 9 9 6 .
On cult places in Ugarit cf. generally Y O N 1984; D E L O L M O L E T E 1992a,
2 4 - 8 = 1999, 27-34; DE T A R R A G O N 1996; DEL O L M O L E T E - S A N M A R T I N 1998,

mentioned in K T U 1.41 = RS 1.003+:41 mdbh b'l). It is interesting

to note that the raised structures mentioned as the 'tower of Baal
of Ugarit' ( K T U 1.119 = RS 24.266:12) and the 'terrace' (gg in
K T U 1.41 = RS 1.003+:50), where sacrifices were offered, are probably confirmed from archaeology by the enormous width of the foundation walls of the temple of Baal and in the purported existence
of a stairway on the eastern side of the temple.
Archaeology has brought to light a second great temple on the
acropolis. Due to the finding of the two inscribed stelae K T U 6.13 =
RS 6.021 and K T U 6.14 = R S 6.028 in its vicinity, it has been
considered as dedicated to the god Dagan. However, unfortunately
the Ugaritic ritual texts never mention a temple of Dagan, 7 2 so that
it seems more reasonable to attribute this temple to the god El in
agreement with the evidence from the ritual texts K T U 1.87 = RS
18.056:42 and K T U 1.119 = R S 24.266:13-4 and with the mythological texts K T U 1.17 = RS 2. [004]+ i 3 1 - 2 and K T U 1.5 = R S
2. [022]+ iv 21. This temple also, like the other one dedicated to
Baal, is built on a platform, has massive foundations which suggest
the existence of a very high upper storey and is subdivided into a
vestibule and a cella.
Besides these temples which have been identified archaeologically,
the ritual texts also mention the temple of Ilat ( K T U 1.41 = R S
1.003+:24 mdbh bt ilt 'altar of the temple of Ilat') and the 'temple
of the lady of the high t e m p l e s ' (ibid. 37 [b]t b'lt bt[m rmm]).
Unfortunately, nothing worthwhile can be said about these two temples given that the context provides nothing useful regarding their
description. From the texts in prose we know of the existence of a
temple dedicated to El (cf. above on the hypothesis regarding Dagan),
a temple to 'Attart ( K T U 4.216 = R S 16.165:2), a temple of Raapgn (ibid, line 3) and other sacred structures dedicated to Iiis ( K T U
4.781 = R I H 8 3 / 2 8 + 31 + 8 4 / 1 5 + 26:2) and to the unknown
deity dml (the letter K T U 2.26 = RS 16.264:6).
Besides the temple structures, a whole series of names of other
specific places of cult are known. T h e most important of these is
the hmn, which has already been mentioned, i.e. the palace chapel
dedicated chiefly to the dynastic cult (and comprising perhaps qd,
ly and kbm, i.e. a small sacred area, steps and other unidentified
Cf. N I E H R
ing in KTU1.


and on





cf. also the new read-

structures). But there are also other places about which nothing certain is known, such as the 'garden' (gn) already mentioned, cited for
example in K T U 1.106 = R S 24.250+:22~3, or else the gb ('sacrificial
pit'?, 'cistern' ? or 'platform' ?, cf. K T U 1.43 = R S 1.005:1-2), the
urbt, the cgml, the gb and others as well (such as the 'tower' mgdl, the
'staircase' (?) m'lt, etc.).
In spite of the lacunae and the uncertainties, the combined use
of textual, archaeological and comparative data makes the ritual system of Ugarit certainly the best known within west Semitic religions
with the prospects of further knowledge in connection with excavations which continue uninterruptedly.

T h e






O l m o



G o d




T h e 'list' as a literary form is well known in the scribal practice of

the ancient N e a r East 1 and is well represented in its archives. T h e s e
provide us with canonical series which comprise a sort of universal
encyclopaedic vocabulary arranged by theme, together with other
lists of similar origin.
These lists, which come from administrative archives, record different
products a n d items that ' c o m e into' or 'leave' the public warehouses,
evidently for the purposes of accounting. In principle, their fate is
irrelevant as far as accounting goes, which m e a n s that there is no
need for a special category for recording products intended to be consumed in the cult as sacrificial victims. In fact, in the archives of
Ugarit, such account records occur mixed together with others which
have all kinds of destinations ( d e l O l m o L e t e - S a n m a r t i n 1998).
However, we are able to identify t h e m and given their syntactic
structure, consider them to be cultic texts. This structure is set out
as the attribution of an offering to a divine recipient following various grammatical models ( d e l O i . m o L e t e 1992a, 1 3 20 = 1999,
1121 ); such an attribution implies a cultic action even though its
model does not need to be m a d e explicit. As a result, in their simplest form these texts a p p e a r as a series of two juxtaposed lists, one
of offerings a n d the other of gods. T h e y can be analysed separately
or in relation to each other, in both cases providing some information on the development of cultic practice in this community. It is
clear that in such a case the 'list of gods' determines the origin and
a r r a n g e m e n t of the ritual, either implying the existence of cultic p a n theons or using other p a n t h e o n s which originated outside the cult.
In terms of administration, t h o u g h , the i m p o r t a n t element is the
record of offerings.
However, this concise form of cultic list is not the only one; there
are others which supply information a b o u t the m o m e n t , place or


1 9 7 7 , 244FF.; L A M B E R T 1 9 5 7 - 7 1 , 4 7 3 - 9 ; C A V I G N E A U X 1 9 8 0



type of cultic act, even providing descriptions of it, which suggests

that at least some of these texts are rather more than mere account
'records' ( d e T a r r a g o n 1995a, 104).
Even so, in principle the references to the recipient or sacred
m o m e n t of offering, even to the rite, seem to have the purpose of
merely certifying or justifying delivery, issue or expense in the accounts,
just like their civil counterparts concerning the moment or method
of a business transaction. Thus, they do not seem to be accounts
which strictly tabulate items but rather they all seem to be tinged
with contextual references. This is possibly connected with the formation and function of the scribe, Quite often the development of
these facts takes on an informative value which transcends mere
accounts and becomes a more or less schematic description, in space
and time, of a complex cultic action. Given the importance of the
cult in ancient cultures, it is legitimate therefore to suppose that these
texts were produced as meta-economic texts, with their own sacral
meaning, and have been analysed as such. 2
Keeping this in mind, it is often quite difficult to determine when
the text is merely a list for accounts (list) and when it is descriptiveprescriptive (ritual). It all depends on what is implied by the information as a whole. Here we shall discuss basically texts which seem to
be pure lists or largely lists, i.e. in which the other information can
be considered (merely) circumstantial. Texts which lie outside this
frame of reference are left to be analysed as a whole as rituals, i.e.
examples of the liturgy of Ugarit, and from them we shall extract
here only the series of offerings and divine names they provide.
O n the other hand, these elements (offering and recipient) which
are supplied together in the lists mentioned, could be recorded quite
separately. This happens especially in the pantheons or n a m e lists
of gods, whereas the possible lists of products or offerings can easily be camouflaged a m o n g texts which we consider to be strictly
administrative if there is no indication of their use or purpose (del





This has been the approach adopted by us (DEL

and earlier by DE T A R R A G O N 1980 and X E L L A 1981.



1992a = 1999),


God Lists

We shall begin with the lists of gods, given the importance this literary form acquired in ancient near Eastern religion: from the beginning, the principal entities to be listed were the gods.
These lists could have functioned merely as templates with which
to fill the records of delivery or lists of offerings, but they also
undoubtedly had a meaning transcending their practical use. This is
suggested by the n u m b e r of copies that were made and their translation into Akkadian, as we shall see below in connection with List
A of the gods. This is a process of systematization which combines
profession of faith in the divine person with the invocation of his
name, elements basic to all ancient Near Eastern religions.
Apart from the 'List , which is arranged in the Mesopotamian
style and so is foreign, there were at least two indigenous lists, originating in a different period and a different ideology, both translated
into Akkadian and occurring together in the cult in the lengthy text
K T U 1.148 = RS 24.643 ( P a r d e e 1997b, 67-71). T h e palace shared
in this religious process of systematic theology in respect of the divine
universe providing its own 'pantheon', also for cultic use. Besides
these and other name-lists or exempt lists we shall provide next those
to be found in ritual contexts, sacrificial and non-sacrificial.

Lists of divine names

T h e List Ann

T h e Ugaritic fragments of this classic list of Mesopotamian gods copy

their model faithfully and are a witness to how venerable and widespread they were throughout the ancient N e a r East, no doubt together with the school tradition of the literature of lists. However,
as such, they provide nothing new or important about the religion
of Ugarit. For that topic we refer to the bibliography. 3 More important as part of the same scholarly tradition is the information provided by the polyglot vocabularies since they allow us to trace the
equivalences made in Ugarit between the Sumerian-Akkadian and

On the Mesopotamian list Anu in general cf. the fundamental article by W E I D N E R

1929 and the bibliography provided later by B O R G E R 1975, I I I 64-5; also L I T R E
1958. For this list in Ugarit cf. N O U G A Y R O L 1968a, 2 1 0 - 3 0 , 246 9, 324. Another
new fragment has been published by A R N A U D 1982a.

Hurrian pantheons and their own; however little of the text has been
preserved ( N o u g a y r o l 1968a, 246-9). List A ( K T U 1.118 = RS 24.268+, K T U 1.47 = RS
1.017, K T U 1.148 = RS 24.643:1-9, R S 20.24)
T h e basic text of the principal or canonical list (A) is K T U 1.118 =
R S 24.268+, whereas K T U 1.47 = RS 1.017 which has the addition of il spn at its beginning, is very damaged, and can be reconstructed from the other text. O n the other hand, the good condition
of the Akkadian text RS 20.24, a version of the previous text, is of
inestimable value for determining what the Ugaritic names/epithets
mean and their relation to the Mesopotamian pantheon ( d e l O l m o
L e t e 1986a, 293-9; H e a l e y 1985, 115-23; 1988a, 103-12). In turn,
K T U 1.148 = R S 24.643:1-9 confirms the fixed character of this
canonical sequence in the cultic domain: its first nine lines give us
a simple listing of the gods of List A, followed by the victim allotted them in the festival of Sapunu (cf. below). This first section is
repeated in lines 10-12 as a sort of summary, in accordance with
the two sets of offerings required by the rp wlmm ritual.


adad be-el hurn ha-zi
adad II
adad III
adad IV
adad V
adad V I
adad V I I



l}uran ha-zi
huranumr u a-mu-tu[m]



b'l spn
ars wmm
grm w\'mqt]



il spn
[il spn]
b'l spn
b'l spn]
[ars] wmm ars wmm
[grm w'mqt] grm w'mqt

HlmT* til-la-ad Aadad
pu-lfur ilnimes







il t'dr b'l
phr ilm

['] ttrt
il t'dr b'l
phr ilm

il 't'dr b'l
phr ilm

Leaving out the addition of the reinterpretative title (il spn), the structure of this list provides the following elements:
a) It begins with a 'triad' which to some extent defines the personality of the s u p r e m e god u n d e r three cultically distinct epithets:
ilib, il, dgn (see also K T U 1.123 = R S 24.271:1-3: ab wilm. . . il. . . il
r dgn . . .). Very probably they suppose a development (il as 'father'
of the divine family) of his personality a n d an attempt at combining p a n t h e o n s , as a reflection of myth, portraying Baelu as 'son of
Dagnu' a n d considering Ilu to be his 'father' (DEL O L M O L E T E 1992a,
56 n. 77 [1999, 74 n. 78] with bibliography).
b) A similar process is assumed in the case of Ba'lu a n d (H)ad(a)du, 4
H e is the great C a n a a n i t e god of the second millennium, the protector of Ugarit (b'l ugrt), defined in principle by the attribute of his
residence spn, specified by the Akkadian version as '(H)Adad, lord
of M o u n t H a z . In his sevenfold epithet are concentrated all the
m a n y (local) e p i p h a n i e s (b'l ugrt, hlb . . .)} In K T U 1.118 = R S
24.264+, a line separates this group f r o m the following heading.
c) T h e descriptive series invoking Ba'lu is followed by a g r o u p of
seven deities, h e a d e d by the dual divinity 'Earth a n d H e a v e n ' , a p p a r endy 'chthonian-astral' or cosmic in nature, in chiastic relation (stellar/

Akkadian makes a clear distinction between the proper noun Adad (IM) and
the common noun be-el, 'lord', which are combined in Ug. b'l (cf. line 4: ''adad be-el
t}uran fra-zi = b'l spn). The Ug. epithet hd of myth is not used here nor is add of
ritual (cf. K T U 1.65 = RS 4.474:9), in spite of the Akkadian translation. But it is
possible that d IM was not read as AAdad in Ugarit; the Akkadian version is a Ugaritic
interpretation (interpretatio) for 'foreigners' (DIETRICH - LORETZ 1981, 67-8).
In this context K T U 4.15 = RS 9.469 must be considered, which is possibly
a list of local or family epithets (stelae?) of Ba'lu.On divine 'heptads' in Mesopotamia
cf. the bibliography given by BORGER, above, n. 2.

terrestrial deities) with this polar pair which heads the group (ktrt,6
yrh, 'ttr || spn, ktr, pdry).
d) T h e next group, again of seven deities, is also headed by the dual
name/divinity 'Mountains and Valleys'. It is combined with six goddesses (1 + 6 ) , apparently arranged hierarchically and related to the
male gods of the previous groups. This separation of the sexes confirms
the absence of 'pairing' in the Ugaritic pantheon.
e) T h e last group is different in each of the various texts. T h e
most complete list is provided by K T U 1.118 = R S 24.264+ and
the Akkadian version, whereas the others omit one or other element.
Here too the series is introduced by a composite divine name, 'the
helper gods of Ba'lu'. T o it corresponds another composite epithet,
'the assembly of the gods', equivalent to 'the family of Ilu'J At all
events, this last group is somewhat of an appendix and possibly contains later additions. In it is developed the process of 'divinizing'
objects (utf}t, km) and persons (mlkm).8
T h e final result is a pantheon of 33 divine invocations, excluding
the title added in K T U 1.118 = R S 24.264:1 (and K T U 1.148 =
R S 24.643:1). T h e hierarchical principle governing the composition
of this 'god list' is not clear; they do not seem to be arranged according to 'personal' importance. Even so, it is possible to say that the
'canonical list' (A) does in fact include the group of principal gods
of Ugarit, exacdy as they appear in myth and the official cult ( K T U
1.148 = RS 24.643:1-9]. It represents a mythologizing expansion
which tends to make organic distinctions between the gods and at
the same time to assimilate other foreign pantheons within its own
religious framework, from the multiple cultural influence which affected
Ugarit: Amorite, Hurrian, Hittite, Sumerian and Akkadian.

This is an overall term for a group of 'seven' deities, as we know from K T U
1.24 = RS 5.194:47-50 (DEL O I . M O LETE 1991, 74-5).
Cf. K T U 1.123 = RS 24.271:32 which refers to drm ilm, 'the two divine families' (?); K T U 1.40 = RS 1.002:25 and par.; and K T U 1.65 = RS 4.474:2-3, with
its invocation of dr/mpl}rt bn it.
In the cult a small reduction of the standard list as it has come down to us
in its final form, is evident: from a pantheon of 33/34 epithets it has become a
group of 28/29, a number which is 'cultically' determined in K T U 1.148 = RS
24.643:1-9, in all likelihood, by the clays of the month, as will be apparent in the
fifth section (lines 23-45). One b'lm as well as 'ttr, uhtt, mlkm and lm are omitted
and the sequence of uf)iy and 'ttrt is inverted.

List ( K T U 1.148 = R S 24.643:23-45; R S 26.142)

K T U 1.148 = R S 24.643:23-45, instead, provides us with a new

list of the il hyr, 'gods of the (month) Hiyaru. W e have no separate
Ugaritic version of this text, but it is evidently a fixed and canonical list as is shown by the coincidence of the cultic use with the
independent Akkadian version (RS 26.142). As both texts are d a m aged, we lack a complete list of gods. 9 W e now set out both texts
arranged in order (the Akkadian text follows the recent reconstruction by A R N A U D 1994, 107-9) a n d their Ugaritic equivalents:
RS 26.142
1. dingir [a-bi\
Ki [ clidim]
Nin.mah x? [
5. Da-gan
U Hal-bi
U d hur.sag Ha-zi
'Ikur' tur
10. [ d hur.]sag Ha-zi
[ d ]-a
A-ta-bv. dBe-el]'ul-ba)-ad-da

K T U 1.148
23. ilib
ars wmm
25. il
b'l hlb
b'l spn
30. ktr
gr witm
rp idrp

15. ' gi.sag.kul' da-ad-me-na

] mr
40. il m[
[...] w thmt
[...] xmr
[z7] sk[r]

Earth and Heavens
Nin.mah [ ]
Ba'lu of Halba.
Ba'lu of Mount Hazi.
Ba'lu the second (?)
[] (?).
[Moun]t Hazi.
Sagru and Itmu

Rapu of idrp
The Damsel

The god of [...]
and Tihmtu
Gods of the bolt
Gods of Dadmena

As yet unpublished texts may perhaps help complete and identify them (BORP A R D E E 1995, 31). P A R D E E (1992, 167) suggests that K T U 1.148 = RS
24.643 is an incomplete tablet which does not reproduce the full god list RS 26.142.


DINGIR.ME la-ab-a-na
20. d U

Ma-[lik. mes]
25. d D[I[n]n

45. [....]

Gods of Labana
Incense burner (?)

In our view this is yet another pantheon used in the cult. As in the
case of K T U 1.148 = RS 24.643:1-9, it is very probable that this
is also a modification of an earlier pantheon, reduced to 28 deities,
in line with the monthly nature of the ritual. 10 Everything indicates
that K T U 1.148 = RS 24.643:23ff. is to be considered a ritual connected with the 'funerary' cult of the month of Hiyaru, which is older
than the 'festival of Sapunu' (lines 1-9). List C ( K T U 1.102 = R S 24.246; K T U 1.139 =
R S 1.001:13-9)
W e have to consider the list of gods provided by K T U 1.102 = R S
24.246:1-14 along the same lines (exempt list and cultic use). It is
a list (C) used in the cult in K T U 1.139 = R S 1.001:13-9, i.e. this
is a standard list not a casual one. Thus, the group of 14/16 deities
mentioned there presupposes a selection which in n a m e and number largely agrees with the one provided by the group of texts which
we call 'dynastic' (cf. below).
1. il bt
5. ktr
10. r'sp
'nt f}bly

T h e 'month' would only be a pattern, without necessarily implying, on the

other hand, a daily celebration in honour of each one of the gods. Also, comparison with RS 26.142:16 suggests correcting ktrt to atrt (nin.mah) in K T U 1.148 =
RS 24.643:25, so written probably because of the sequence mm w ars ktrt in list A
(but cf. atrt in line 31; Pardee reads [-]-rt).

p pgr
iltm hnqlm
yrh kty

This 'list of gods' contains a group of deities exclusive to the quoted

texts (trmn, dqt, trt and the epithets cnt hbly, p pgr, iltm hnqtm; yrh(m)
kty also occur in K T U 1.123 = R S 24.271:7), and also includes the
major deities of Ugarit [b'l, ym, yrh, ktr, cnt, p).

Lists of names of divine kings

List D ( K T U 1.102 = RS 24.246:15-28)

T h e dynastic nature mentioned above explains why the list of K T U

1.102 = RS 24.246:1-14 is continued on the reverse of the tablet
by another (D) of theophoric names, to which in K T U 1.39 = RS
1.001:19" correspond the references to glmt, bclt bhtm, ins ilm. This
suggests that these theophoric names are also divine names, actually
of the kings of Ugarit, of their ins ilm, 'divine peoples', of which
K T U 1.106 = R S 24.250:3-5 already provides proof (three: ydbil,
yaril, cmtr) as recipients of offerings and the colophon K T U 1.6 =
R S 2.[009] vi 58 assigns one (yrgb.bcl) to king Niqmaddu of Ugarit. 12
This interpretation, however, has incurred some opposition, although
alternative suggestions are not convincing. 13
Consequently, we consider K T U 1.102 = RS 24.246 to be one list
or single pantheon which belongs to the dynasty with its gods and
divinized dead, which as such (ins ilm, gtrm, rpum) are only listed globally in the c o m m o n standard patheon under the official title of mlkm.
These divine names are:

Understood in this way, K T U 1.39 = RS 1.001 is a 'generic' royal ritual
which closes by invoking the ins ilm, specified in K T U 1.102 = RS 24.246:15-28,
and as such could be added to it together with the sacrificial element thus supposed, e.g. in K T U 1.106 = RS 24.250+:2-5 (gdlt).
On K T U 1.102 = R S 24.246 in general cf. V I R O L L E A U D 1968, 594; DE M O O R
1970b, 326-7 (see 1990, 241); H E R D N E R 1978, 3-7; D I E T R I C H - L O R E T Z
1975b, 545-6; C A Q U O T 1979b, 1404; S T A M M 1979, 753-8; X E L L A 1981, 328-31;
D E L O L M O L E T E 1986a, 282-5; 1987:43-6; 1992a, 117