The Carian Language

HANDBOOK OF ORIENTAL STUDIES

SECTION ONE
THE NEAR AND MIDDLE EAST

Ancient Near East

Editor-in-Chief
W. H. van Soldt

Editors
G. Beckman • C. Leitz • B. A. Levine
P. Michalowski • P. Miglus

Middle East
R. S. O’Fahey • C. H. M. Versteegh

VOLUME EIGHTY-SIX
The Carian Language

by

Ignacio J. Adiego

with an appendix by

Koray Konuk

BRILL
LEIDEN • BOSTON
2007
 This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

Adiego Lajara, Ignacio-Javier.

The Carian language / by Ignacio J. Adiego ; with an appendix by Koray Konuk.
p. cm. — (Handbook of Oriental studies. Section 1, The Near and Middle
East ; v. 86).
Includes bibliographical references.
ISBN-13 : 978-90-04-15281-6 (hardback)
ISBN-10 : 90-04-15281-4 (hardback)
1. Carian language. 2. Carian language—Writing. 3. Inscriptions, Carian—Egypt.
4. Inscriptions, Carian—Turkey—Caria. I. Title. II.

P946.A35 2006
491’.998—dc22 2006051655

ISSN 0169-9423
ISBN-10 90 04 15281 4
ISBN-13 978 90 04 15281 6
© Copyright 2007 by Koninklijke Brill NV, Leiden, The Netherlands.
Koninklijke Brill NV incorporates the imprints Brill Hotei Publishers,
IDC Publishers, Martinus Nijhoff Publishers, and VSP.
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 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.
PRINTED IN THE NETHERLANDS
Günter Neumann
In memoriam
 CONTENTS

Foreword ........................................................................................ xi
Acknowledgments for the Use of Illustrations ............................ xiii

Chapter One Introduction .......................................................... 1

Chapter Two The Indirect Sources .......................................... 7

A. The Glosses .......................................................................... 7
1. Glosses und Pseudo-glosses .................................................... 7
2. Interpretation ........................................................................ 10
B. The Proper Names .............................................................. 12
1. The Unity and Continuity of Anatolian Onomastics .............. 12
2. Present Compilations of Carian and Anatolian Names ............ 14
3. Carian Names from Indirect Sources vs. Those from Direct
Sources ................................................................................ 15

Chapter Three The Inscriptions ................................................ 17

A. Introduction .......................................................................... 17
1. The Revised System of Transcription of Carian Letters .......... 18
2. Vocalism .............................................................................. 18
3. Consonantism ...................................................................... 19
B. ‘Para-Carian’ or ‘Caroide’ Inscriptions .............................. 22
1. ‘Para-Carian’ Inscriptions from Caria .................................... 22
2. ‘Para-Carian’ Inscriptions from Other Places .......................... 23
3. The Ostrakon of Hou and the Naukratis Fragment ................ 26
4. Carian Grafitti from Sardis .................................................. 27
5. Carian Inscription from Old Smyrna .................................... 29
C. The Carian Inscriptions from Egypt .................................. 30
1. Sais (E.Sa) ........................................................................ 32
2. Memphis (E.Me) ................................................................ 34
3. Abydos (E.Ab) .................................................................... 79
4. Thebes, Tomb of Montuemhat (E.Th) .................................. 95
5. Luxor Temple (E.Lu) .......................................................... 106
6. Murwàw (E.Mu) ................................................................ 109
7. Silsilis (E.Si) ...................................................................... 110
8. Abu Simbel (E.AS) .............................................................. 115
viii CONTENTS

9. Buhen (E.Bu) .................................................................. 119

10. Gebel Sheik Suleiman (E.SS) ............................................ 123
11. Unknown Origin, Likely from Egypt (E.xx) ........................ 124
D. The Carian Inscriptions from Caria .................................. 128
1. Tralleis (C.Tr) .................................................................. 130
2. Alabanda and Surroundings (C.Al) ...................................... 132
3. Euromos (C.Eu) ................................................................ 132
4. Kindye (C.Kn) .................................................................. 134
5. Hyllarima (C.Hy) ............................................................ 135
6. Mylasa (C.My) ................................................................ 137
7. Sanctuary of Sinuri near Mylasa (C.Si) .............................. 138
8. Kildara (C.Ki) .................................................................. 141
9. Stratonikeia (C.St) ............................................................ 142
10. Halikarnassos (C.Ha) ........................................................ 144
11. Didyma (Ionia, near Milet) (C.Di) .................................... 145
12. Iasos (C.Ia) ...................................................................... 145
13. Keramos (C.Ke) ................................................................ 150
14. Kaunos (C.Ka) .................................................................. 151
15. Krya (C.Kr) ...................................................................... 158
16. Inscriptions of Unknown Origin, Presumably from Caria
(C.xx) .................................................................................. 159
E. The Carian Inscriptions from Greece ................................ 164

Chapter Four The History of the Decipherment .................. 166

A. The ‘Semisyllabic Era’ (1887–1962) .................................. 166
B. The ‘Greek Alphabetic’ Era ................................................ 176
C. The ‘Egyptian Approach’ .................................................... 187
1. The First Attempts ............................................................ 187
2. The Seminal Work of Ray ................................................ 191
3. The Definitive Decipherment (‘Ray-Schürr-Adiego System’) ...... 197

Chapter Five The Carian Alphabet .......................................... 205

A. Alphabetic Varieties ............................................................ 205
1. Alphabetic Varieties of Caria Proper .................................... 206
2. Inscriptions from Continental Greece .................................... 219
3. Egyptian Alphabets ............................................................ 219
4. The Classification of the Alphabets of Caria Proper .............. 223
5. The Relationship between the Alphabet from Egypt and
the Local Alphabets from Caria Proper .............................. 226
6. The Common Origin of the Carian Alphabetic Varieties ........ 228
B. The Origin of the Carian Alphabet .................................. 230
 CONTENTS ix

Chapter Six Phonological Features ............................................ 234

A. The Phonological System .................................................... 234
1. Vowels and Semivowels ........................................................ 234
2. Consonants .......................................................................... 242
3. Letters of Uncertain Value .................................................... 251
4. Letters of Unknown Value .................................................... 253
5. Phonotactics ........................................................................ 254
B. Overview of the Historical Phonology of Carian ............ 256
1. Vocalism ............................................................................ 257
2. Consonants .......................................................................... 259
3. Some Secondary Changes ...................................................... 262

Chapter Seven Analyzing Carian Inscriptions .......................... 264

A. Basic Onomastic Formulae .................................................. 264
1. Inscriptions Consisting of Only an Individual Name .............. 264
2. Inscriptions Consisting of Only a Twofold Onomastic
Formula .............................................................................. 265
B. The Structure of the Stelae from Memphis ...................... 267
1. Threefold Formulae .............................................................. 267
2. Stelae for Women ................................................................ 271
3. Inscriptions with Ted and En ............................................ 273
4. Other More Complex Funerary Inscriptions ............................ 275
5. The Rest of the Inscriptions from the Memphis Corpus .......... 276
6. A First Summary ................................................................ 279
C. Analyzing Brief Inscriptions ................................................ 280
1. Inscriptions on Objects .......................................................... 281
2. Funerary Inscriptions of Caria and Athens ............................ 287
3. The Longest Graffito from Abu Simbel (E.AS 7) .................. 293
D. The Longer Inscriptions ...................................................... 294
1. The Kaunos Bilingual Inscription .......................................... 295
2. The Kaunos Inscription C.Ka 2 ............................................ 301
3. Sinuri’s Longer Inscription (C.Si 2) ...................................... 302
4. The Hyllarima Inscription (C.Hy 1) .................................... 305
5. Other Inscriptions from Caria .............................................. 308
E. Summary and Some Controversial Questions .................. 310

Chapter Eight Morphological Features ...................................... 312

A. Nominal Inflection .............................................................. 312
1. Nominative Singular ............................................................ 312
2. Accusative Singular .............................................................. 313
x CONTENTS

3. Genitive Singular .................................................................. 314

4. The -s Ending and the Problem of Carian Datives ................ 314
5. Other Possible Datives .......................................................... 317
6. Nominative Plural ................................................................ 318
7. Accusative Plural .................................................................. 318
8. Other Possible Case Endings ................................................ 319
B. Pronominal Inflection .......................................................... 319
C. Verbal Inflection .................................................................. 321

Chapter Nine The General Vocabulary and the Proper

Names .......................................................................................... 326
A. General vocabulary .............................................................. 326
B. Proper Names ...................................................................... 328
1. Theophores .......................................................................... 331
2. Some Nominal Stems ............................................................ 333
3. Verbal Stems ...................................................................... 339
4. Adverbs .............................................................................. 339
5. Lallnamen .......................................................................... 340
6. Suffixes ................................................................................ 341

Chapter Ten Carian as an Indo-European Anatolian

Language .................................................................................... 345

Chapter Eleven Carian Glossary ................................................ 348

Appendices
A. Carian Inscriptions in Transcription .................................. 443
B. Carian Glosses ...................................................................... 455
C. Carian Names in Greek Sources ........................................ 456
D. Concordances ........................................................................ 464
E. Coin Legends in Carian (by K. Konuk) ............................ 471

Abbreviations and Bibliography .................................................... 493

Table I: The Carian Alphabet .................................................... 508

Table II: Carian signs in coin legends ........................................ 509
Index .............................................................................................. 511
Plates .............................................................................................. 519
Maps .......................................................................................... 521
Coins .......................................................................................... 523
 FOREWORD

Fifteen years after my doctoral dissertation on Carian (Studia Carica.

Investigaciones sobre la escritura y lengua carias, y su relación con la familia
lingüística anatolia indoeuropea, University of Barcelona, 1990), directed by
Pere J. Quetglas and supervised by Jürgen Untermann, and twelve
years after my book Studia Carica (Barcelona 1993), a revised edition of
my dissertation, I now offer a new work on the Carian script and lan-
guage. This is not an English translation of the above titles. In fact,
very little remains of the structure and content of these former works,
mainly due to three, closely interrelated reasons. Firstly, both my dis-
sertation and the subsequent book were conceived to defend a new
proposal for deciphering Carian, beginning with the fundamental work
of the British Egyptologist John D. Ray, which, in the 1980s, marked
a dramatic breakthrough in the history of research on Carian. My aim
at the time was to further develop the decipherment, after noting that
Ray’s system, although certainly well founded, was by no means totally
satisfactory. Now, following the discovery of the bilingual inscription of
Kaunos in 1996, the system proposed in my book Studia Carica has
been proved correct and is accepted—with slight modifications—as the
standard tool for transcribing the Carian signs. Therefore, a new book
on Carian no longer needs to offer a ‘combative’ justification for a par-
ticular decipherment system, but should rather take the definitive deci-
pherment as a starting point for the analysis of Carian texts. Secondly,
the corpus of Carian inscriptions has been augmented by the appear-
ance of several new texts, among which the Carian-Greek bilingual of
Kaunos mentioned above is undoubtedly the most relevant. In addi-
tion, some other inscriptions that were already known have subsequently
been revised, yielding new possible interpretations. Thirdly, the new
decipherment encouraged many scholars to apply themselves to the
study of Carian, so that although many uncertainties remain and our
knowledge of Carian language continues to be very limited, important
progress has been made in recent years, and this must now be incor-
porated into an updated analysis of the subject.
I wish to repeat here my most profound gratitude to those that
helped me when I began to work on Carian, particularly the afore-
mentioned Pere J. Quetglas and J. Untermann, without whose help
xii FOREWORD

and encouragement I could not have undertaken a doctoral disserta-

tion on a topic that was then beset by so many risks. I am also grate-
ful to Theo van den Hout, who invited me to prepare this book for
Brill; to Brill for accepting this project; to Koray Konuk, for giving me
the opportunity to substantially improve this book with an appendix
on Carian coin legends; and to Wolfgang Blümel, H. Craig Melchert
and Diether Schürr, not only for their kind assistance with some of the
problems that arose during the preparation of this work, but also for
the consistently productive and exciting interchange of ideas about
Carian. Georg Rehrenböck (Kleinasiatische Kommission, Österreichi-
sche Akademie der Wissenchaften) also deserves a special mention for
his extreme kindness and unfailing generosity in answering all of my
queries about copies of Carian inscriptions conserved in Vienna.
I am equally indebted to Peter Cottee for his accurate revision of
my imperfect English. Needless to say, all the possible errors and omis-
sions are of my own doing.
On a more personal level, I must express my gratitude to my wife,
Anna, and to my daughters, Alba-Artemísia and Laura-Neït, for bear-
ing with such patience the long period during which the book was put
together.
This book is dedicated to the memory of Günter Neumann: for a
decipherment to be successful it must not only be correct, but also
credible and convincing, and he showed me how to achieve this with
his open-minded consideration of ideas that questioned the prevailing
communis opinio on Carian, his astuteness in refraining from excessive
speculation, his numerous suggestions of improvements, his discreet but
very effective work in favour of the new decipherment, and his advice
to wait patiently for a proposal marked with the signum veritatis (to quote
his own expression) to finally achieve general acceptance.
ACKNOWLEDGMENTS FOR THE USE OF ILLUSTRATIONS

Drawings of inscriptions come from O. Masson, Carian Inscriptions from

North Saqqara and Buhen (Masson 1978), and are reproduced by kind
permission of the Egypt Exploration Society.
Drawings taken from O. Masson-J. Yoyotte Objets pharaoniques à inscrip-
tion carienne (Masson-Yoyotte 1956) and O. Masson “Remarques sur les
graffites cariens d’Abou Simbel”, Hommages à la mémoire de S.
Sauneron II (Masson 1979) are reproduced with the permission of the
Institut français d’archéologie orientale.
Drawings of graffiti published in The Epigraphic Survey, Reliefs and
Inscriptions at Luxor Temple II, The facade, Portals, Upper Register Scenes,
Columns, Marginalia, and Statuary in the Colonnade Hall (ESS 1998) are
reproduced by courtesy of the Oriental Institute of the University of
Chicago.
Drawings extracted from L. Deroy “Les inscriptions cariennes de
Carie”, L’Antiquité Classique 24 (Deroy 1955) are reproduced with per-
mission of the editors of the journal.
 CHAPTER ONE

INTRODUCTION

It is not clear when Caria and the Carians enter into ancient History.
This is dependent on equating classical Caria with the land of Karkiya/
Karkisa mentioned in Hittite sources. This supposition, eminently suit-
able from a purely linguistic point of view (karkº in Karkisa, Karkiya
is practically identical to the Old Persian word for ‘Carian’, kºka-), is
complicated by the uncertainties regarding the exact location of Karkisa/
Karkiya on the map, a problem intimately bound to the complex ques-
tion of Hittite geography, a topic still subject to controversy despite the
great progress made in recent years.
In any case, no information about the language of the land of
Karkiya/Karkisa can be obtained from Hittite sources, so that even if
the equation could be confirmed, its value for the study of the Carian
language would be very limited. The only relevant (but overly generic)
datum is that Karkiya/Karkisa is a land located in the western region
of Anatolia, an area occupied by Luwian population groups, and thus
consistent with the clear similarities between Carian and Luwian, Lycian
and other Indo-European Anatolian dialects that can now, since the
decipherment of Carian, be clearly traced (see Chapter 10).
Classical Caria, the country situated in western Anatolia between
Lydia and Lycia, must therefore be the starting point of the research
on Carian language. It is during this period that we find both direct
documentation of Carian and a wealth of information about this land
and its inhabitants in indirect, mostly Greek, sources.
Particularly meaningful are the consistent ties that we can establish
between various types of records on Carian and the Carians regard-
ing one of the most remarkable characteristics of Carian language doc-
umentation: the fact that the greatest number of Carian inscriptions
have been found in Egypt, and not in Caria itself.
From Greek sources, we know that Carian and Ionian mercenaries
were employed by the pharaoh Psammetichus I (664–610) for consol-
idating his throne (Herod. I, 151). According to Herodotus, these mer-
cenaries were based in the Delta area, near Bubastis (Herod. II, 154).
It is no coincidence then that the oldest datable Carian document from
2 CHAPTER ONE

Egypt, a bilingual inscription on a statuette of the goddess Neith (E.Sa

2) from the times of Psammetichus I, can be traced to Sais, another
Egyptian city situated on the Delta.
A further connection can be drawn between Carian documents and
historical facts under the reign of the grandson of Psammetichus I,
Psammetichus II (595–589): a well-known Greek graffito from Abu
Simbel attests the participation of foreign mercenaries in his Nubian
campaign (593/592, see below p. 31 for this dating), and this infor-
mation is consistent with the existence of Carian graffiti in Abu Simbel
and in other locations further to the south (Buhen, Gebel el-Sheik el-
Suleiman).
Under Amasis (568–526), Ionian and Carian settlements were moved
from the Delta to the city of Memphis (Herod. II, 154), where a ‘Carian
quarter’ (KarikÒn) and a ‘Greek quarter’ (ÑEllhnikÒn) existed for many
years (Aristagoras of Milet apud Stephan of Byzantium, s. v. ÑEllhnikÒn).
Once again this event can be linked to epigraphical evidence: the most
important sub-corpus of Carian inscriptions is the collection of funer-
ary stelae found in Saqqâra, one of the necropoleis of Memphis. As
for the rest of Carian inscriptions found in Egypt—mostly graffiti—
from Thebes, Abydos, Silsilis, etc.), no connections can be established
with historical facts, and we can only assume that they are the marks
of Carian visitors, similar to Greek graffiti found in these and another
parts of Egypt.
Caria itself does not offer such striking results. The Carian inscrip-
tions found in Caria are far less numerous than those from Egypt,
come from different cities, and appear more heterogeneous, both in
content and in form, thus constituting a very fragmentary and incom-
plete view and lacking a clear connection with historical facts.1 In fact,
the sole inscription that gives any indication of a link to the history of
Caria is the bilingual inscription from the temple of the god Sinuri
(C.Si 2), which can confidently be interpreted as a decree enacted by
the Carian dynasts of the Hekatomnid era, Idrieus and Ada, whose
joint reign is dated in the period 351/350–344/343. But not even this
inscription has any real implications for Carian history: it is simply part
of a wider corpus of regulations of a local syngeneia—mostly in Greek—
produced by the satrapal couple. In the case of another of the most

1
This is not the place for a history of Caria. I refer the reader to Hornblower
(1982).
 INTRODUCTION 3

important Carian inscriptions from Caria itself, namely the bilingual

proxeny decree of Kaunos (C.Ka 5), dated at the end of the IV cen-
tury BC, the two Athenian citizens honoured as proxenoi are not clearly
identifiable with any figures found in classical sources. The new inscrip-
tion from Hyllarima (C.Hy 1b), and perhaps also the inscription recently
found in Mylasa (C.My 1), offer lists of priests that can be linked to
local cults. The two texts from Stratonikeia also contain lists of names,
but in these cases their religious character cannot be confirmed. Nothing
can be ascertained about the exact content of the inscription from
Kildara (C.Ki 1) or the longest Carian inscription from Kaunos (C.Ka
2), but their official character is beyond doubt. This heterogeneous cor-
pus is completed by a number of funerary, and therefore private, inscrip-
tions and by others about whose content and functionality we know
nothing. The absence of a precise dating for most of these texts fur-
ther compounds the difficulties in placing them in a definite context.
If a conclusion must be drawn exclusively from the present corpus
of Carian inscriptions from Caria, we must conclude that written use
of Carian was very limited and confined to certain local communities,
mainly linked to religious cults and to private funerary contexts: in this
latter case there are so few examples that it would be tenuous to make
a comparison with the much richer documents of neighbouring Lycia.
The only notable exception to this extreme scarcity of documentation
is Kaunos, where a small, but very significant corpus of inscriptions
has been established, thanks particularly to the archaeological excava-
tions carried on in the last forty years. But this is an exception only if
considered in relative terms compared to the rest of the documenta-
tion. In absolute terms, Kaunos provides us with a very a modest col-
lection of Carian inscriptions.
It is astonishing that we have at our disposal only a single, reliable
example of the use of Carian as a co-official language (alongside Greek)
in the Hekatomnid period: the aforementioned decree from Sinuri, C.Si
2. This is particularly surprising since it is generally believed that dur-
ing the activity of the Hekatomnid dynasty, a ‘Karianization’ (to use
Hornblower’s word) took place, so that “Mausolus and his family, them-
selves native Karians, encouraged the institutions (such as the koinã),
the cults, and the language of Karia” (Hornblower 1982:352; the empha-
sis is mine). The Carian documentation found to this day in no way
supports this view. It is true that this current interpretation of the writ-
ten use of Carian must be viewed with some caution, since it is pos-
sible that the lack of documentation is in fact a result of our insufficient
4 CHAPTER ONE

archaeological knowledge of the land, and the perishable materials on

which many inscriptions were written. Moreover, both the hypotheses
on the origin of Carian letters envisaged here (p. 231) and the strongly
differentiated local alphabetic variants point clearly to a much wider
and prolonged use of Carian script. However, at least in the case of
the alleged importance of the Carian language as a tool for ‘Karianization’
in the Hekatomnid period, there is compelling counter-evidence: the
total disappearance of Carian letters, and their replacement by single
Greek legends, from the well-documented coinage of the Hekatomnids
from Hekatomnos onward,2 which would seem to disprove the theory
of the co-official status of Carian.
Carian belongs to the Indo–European family of Anatolian languages,
which also comprises Hittite, Cuneiform and Hieroglyphic Luwian,
Palaic, Lycian, Lydian and lesser known Sidetic and Pisidian. More
precisely, it forms part of the so-called ‘Luwic’3 group. This classifica-
tion of Carian has already been proposed by several scholars—notably
V. V. ”evoro“kin—on the basis of indirect documentation affirmed in
Classical sources, mostly Carian proper names that showed strong con-
nections with Anatolian. But this is now clearly confirmed by direct
sources, now that the Carian alphabet has been deciphered, and deci-
sive linguistic evidence has been obtained from Carian texts.
This book is intended to show the new look of Carian following the
decipherment of the Carian writing system. My aim is to offer as com-
plete and updated a view as possible of our present knowledge of
Carian, paying special attention to existing documentation, the Carian
alphabet, its origin and variants, and to the still developing linguistic
interpretation of the materials and its repercussions for the classification
of Carian as an Anatolian language. In Chapter 2, a brief summary
of the indirect sources on Carian is outlined. Chapter 3 is devoted to
the presentation of the entire corpus of Carian inscriptions available to

On Hekatomnids’ coinage, see the decisive work of Koray Konuk (Konuk 1998a).
2

From now on, I shall use the term ‘Luwic’ for this group of dialects, following
3

the suggestion made by Melchert, as a convenient and non-confusing form to refer to

a series of Anatolian dialects that share important differentiating issogloses but whose
exact internal relationship—originating from a common branch, or rather the result
of an areal convergence—is still debated (see Melchert 2003:176). The former use of
‘Luwian’, both for two dialects (Hieroglyphic and Cuneiform) and for the entire group
(embracing also Lycian, Carian, Sidetic and Pisidian), could lead to confusion and must
be therefore abandoned.
 INTRODUCTION 5

me at the time this book was written.4 Only the coin legends are not
discussed there, since they are the subject of a specific appendix, kindly
prepared by Koray Konuk. Chapter 3 does not constitute an epigraphical
edition. The inscriptions are simply intended to be a useful tool with
which to tackle the following chapters and the analysis of Carian writ-
ing and language in general, hence the inclusion of drawings and the
observations about reading problems.
Chapter 4 offers a general history of the decipherment of Carian. It
is based mostly on the corresponding chapter of my Studia Carica (Adiego
1993a), but has been expanded to include a succinct exposition of the
decisive progress of decipherment, accomplished during the final decade
of the 20th Century.
Chapter 5 deals with the Carian alphabet, offering an analysis of
the different local alphabetic variants and some reflections on the pos-
sible origins of this extremely peculiar writing system.
Chapters 6 to 11 focus on linguistic aspects. Chapter 6 offers an
overview of Carian phonology from a synchronic and, insofar as is pos-
sible, diachronic perspective. Chapter 7 introduces the possibility of
analysing a great number of Carian inscriptions, from the briefest and
most transparent, to the more extensive, wherein the difficulties of inter-
pretation are practically insurmountable. Chapter 8 discusses the (scarce)
morphological traits that have so far been identified in Carian. Chapter
9 is of a lexical nature: an inventory of all the Carian common words
to have been identified is put together and analyzed, and the same is
done for the proper names, in this latter case in the context of Anatolian
onomastics. As a means of concluding the study, chapter 10 presents
the evidence that suggests Carian can be classified in the group of
Anatolian ‘Luwic’ dialects. Finally, Chapter 11 provides a glossary of
all the forms shown in Carian inscriptions, inspired by similar works
such as the Lydisches Wörterbuch by Roberto Gusmani, or the more recent
Dictionary of the Lycian Language by H. Craig Melchert. The book is
accompanied by five appendices: an editio minor of the inscriptions in
transcription, the collection of glosses and a list of proper names found
in Greek sources, a table of concordances with other editions of Carian

4
The only unavailable source is a Carian inscription found in Greece and pub-
lished some years ago (see G. Neumann, “Epigraphische Mitteilungen—Kleinasien” in:
Kadmos 39, 2000:190). Despite my efforts, it was impossible to obtain when prepar-
ing this book.
6 CHAPTER ONE

inscriptions, and a special section by Koray Konuk on Carian coins

with Carian legends, as mentioned above. The bibliography that com-
pletes the book is intended to be an exhaustive collection of all publi-
cations on Carian language and inscriptions from Sir Archibald Sayce
to the present, together with other works not specifically devoted to
Carian, but that are cited thoroughout this study.
Despite my best efforts, it is inevitable that a study of a language
that is only partially understood, and surrounded by uncertainties, will
contain a number of provisional results, fragile hypotheses and even
mere speculations, not necessarily shared by other scholars. This will
not come as a surprise to seasoned specialists in the study of Trümmer-
sprachen, but I consider it necessary to advise scholars of related disci-
plines against taking all of the interpretations included in my work to
be demonstrated or indeed definitive. Much more will have to be done,
and much more new material will have to be made available before
we are able to paint a confident picture of this language. In any case,
I have tried at all times to separate the assertions presently considered
plausible by the communis opinio (for example, the overall validity of the
decipherment system Ray–Schürr–Adiego) from those interpretations
that rest on more fragile indices, and which are shared by only a hand-
ful of scholars, or indeed those hypotheses that are my own creation.
I have tried where possible to avoid resorting to overly hypothetical
explanations, and to steer clear of long, confusing paragraphs of a crit-
ical nature, which also explains the limited use I have made of works
by other scholars, when their results seemed to me excessively tenuous
or somewhat premature. Only in the glossary, for the sake of com-
prehensiveness, have I included some interpretations suggested for differ-
ent Carian words that I don’t share or that I find too speculative.
Proposed in this study is a new system of classification for Carian
inscriptions, which is intended to be more rational than previous offerings,
and also some modifications of the transcription system of Carian signs
(the reasons for this are explained in pp. 18–20). I hope that these
innovations will not create any difficulties for the reader. For more
clarity, Appendix C (Concordances) can be used to find the equiva-
lences between the new classification system and the older ones, and
Table I at the end of the book shows the transcription system of Carian
that is used here.
 CHAPTER TWO

THE INDIRECT SOURCES

A. THE GLOSSES

“Sur le chapitre des gloses, on peut passer assez rapidement. Le matériel

est d’une valeur très médiocre” (Masson 1973:190).
Olivier Masson’s categorical statement is absolutely true. There are
very few Carian glosses, and the possibilities of connecting them both
to the direct documentation of the Carian language and to the com-
mon Anatolian lexicon are very limited or simply non-existent. Despite
this, we will examine this modest corpus.

1. Glosses and Pseudo-glosses

The first problem that must be addressed is the distinction between
true Carian glosses and merely fictitious forms. From the time of the
first compilation of scientific interest—Sayce (1887[92]:116–120)—to the
most recent—by Dorsi himself (Dorsi 1979)—the inventory of Carian
glosses was artificially expanded by invalid entries until Dorsi significantly
reduced it. Sayce offered a total of 21 glosses, which Dorsi then short-
ened to six definite ones, and three doubtful, adding three relevant
names of gods. Between Sayce and Dorsi, in Brandenstein (1935a) this
artificial expansion of the number of glosses reached its peak: Brandenstein
lists more than sixty words under the title “Wörter, Glossen, usw.”
I have analysed elsewhere the methodological errors that led to this
situation (Adiego 1992b). Here we need simply remember that a source
of tremendous confusion was the lack of differentiation in Brandenstein
(1935a) between actual glosses and a modern, semantic explanation of
Carian proper names (for the most part very weakly argued). As a
result, some theories were constructed based only on a purely hypo-
thetical meaning attributed to Carian names.
As has been said, Dorsi’s corpus consists only of six definite glosses,
and three doubtful ones. I choose to disregard the three god names,
since no meaning is offered in ancient sources, and they therefore
8 CHAPTER TWO

contribute nothing to our knowledge of the Carian lexicon.1 The six

definite glosses are êla ‘horse’, bãnda ‘victory’, soËa(n)2 ‘tomb’, g°la
‘king’, g¤ssa ‘stone’, and ko›on ‘sheep’. Of these, the first five are linked
by the fact that they are all found in the work of Stephan of Byzantium,
who quotes them when speaking about the Carian place names ÉAlã-
banda, ÑUlloÊala, Souãggela and MonÒgissa (see appendix B for the
relevant passages).
As for the sixth gloss, although the Carian character of the word
concerned is expressly mentioned in the sources, there are some prob-
lems concerning the exact form of the word. The sources are two scho-
lia to the Iliad XIV, 255, one in Eusthatius’ scholia, the other in the
manuscript T of the Scholia Vetera in Iliadem,3 both of which are related
to the name of the island of Kos (K«w). In T, the word is given as
ko›on, while Eustathius’s passage is ambiguous: he states that K«n (acc.)
is spelled KÒon by some, but then immediately adds that the Carians
use this name for sheep (fas‹ d¢ toÁw Kãraw oÏtv kale›n tå prÒbata).4
It has traditionally been interpreted5 that the Carian word was in fact
identical to the name of Kos (hence k«w ‘sheep’), but Dorsi believed
that Eustathius left out the actual Carian word simply through over-
sight, which would be ko›on as it appears in the other scholion.
However, Erbse, in a work published after Dorsi’s article (Erbse 1986:
389–390), is suspicious of the form ko›on of T, offering it inter cruces.
His preference is the kÒon form of Eustathius, interpreting the passage
of Eustathius’ work mentioned above as a true gloss6 and noting more-
over that this latter form appears in another passage by Eustathius

1
The three god names are ÖImbramow = Hermes (St. Byz. s. v. ÖImbrow; but note
the variant reading ÖImbrasow, which Dorsi does not mention, reported by the Scholia
vetera in Theogoniam v. 338, and Eustathius, Commentarii ad Iliadem XIV, 281), Mãsariw
= Dionysus (St. Byz. s. v. Mãstaura), and ÉOsog«a = Zenoposeidon (Strabo XIV,
659, Pausanias VIII; 10, 4).
2
As Dorsi rightly points out, “la n finale di soËan può essere un semplice morfema
di accusativo greco (peraltro non necessario: cfr. sopra êla), ma può anche essere stata
suggerita (a torto o a ragione) dalla scomposizione del toponimo in souan-gela” (Dorsi
1979:29).
3
Edited by Hartmut Erbse: Scholia Graeca in Homeri Iliadem (Scholia uetera), Berlin,
1979–1984. T is the Cod. Brit. Mus. Burney 86, to be dated in 1014 or 1059 A.D.
4
Eusthatius, Commentarii ad Homeri Iliadem 983, 33 (= ad Iliad. XIV, 255). I follow
the Leipzig edition of 1827 (re-edited in Hildesheim-New York 1970).
5
Cf. Sayce (1887’92]:118): K«w (k«n) ‘a sheep’, although he adds the spelling kÒow
in Eusthatius. See also Brandenstein (1935:142).
6
At least this is what I have deduced from the very terse and implicit treatment
of the problem in Erbse (1986).
 THE INDIRECT SOURCES 9

(318,41 = ad Iliad. II, 677).7 Although Erbse does not discount the
possibility that Eusthatius could have replaced ko›on with kÒon in order
to “improve” the etymological explanation in both passages, he sees
kÒon as the genuine form, opting for the simpler solution that ko›on is
a graphical error.
The three glosses that Dorsi lists as dubious are: tãba ‘rock’, tou-
ssÊloi ‘dwarfs, pigmies’, and tumn¤a ‘stick’, all of which are also attested
by Stephan of Byzantium. The inclusion of tumn¤a, even if it is branded
as dubious, seems rather inadequate; it is true that Stephan mentions
this word when referring to the Carian city of TumnhsÒw, but he attrib-
utes it to the language of the inhabitants of Xanthos, a Lycian city,
implying that the word must be Lycian rather than Carian.8
In the case of tãba ‘rock’, the word is cited by Stephan of Byzantium
in connection with a Lydian city called Tãbai. He adds that the word
tãba, which he does not attribute specifically to any language, is trans-
lated in Greek as ‘rock’. Following this, he mentions another Tãbai,
in this case situated in Caria, but it seems to be merely a passing ref-
erence. From his observations then, the gloss had to be interpreted as
Lydian, but Dorsi rightly observes (1979:29) that modern scholarship
coincides in its estimation that no Lydian city of this name existed and
that the two cities are in fact only one, situated in Caria (cf. Zgusta
KON § 1277–1, Blümel KarON:179). It is therefore feasible, although
impossible to demonstrate, that Stephan’s mistaken belief that the place
name belonged to a Lydian country could have led to an error when
attempting to establish the origin of the gloss. The problem becomes
even more intractable given the existence of Tabhnoi (pl.), the name
given to the inhabitants of some part of Lydia, which suggests that a
Lydian Tabai or similar might actually have existed (see Zgusta
KON:593).
As for toussÊloi, the text is ambiguous and obscure. The word
appears under the entry Kãttouza, a Thracian city inhabited by pig-
mies. Reference to Carians is therefore secondary and open to various
interpretations.

7
However, in this case, the word kÒon is not attributed expressly to the Carians.
8
Jãnyioi går tØn =ãbdon tumn¤an l°gousin.
10 CHAPTER TWO

2. Interpretation
If the collection of Carian glosses is very small and their value mediocre,
the attempts to interpret them are equally disappointing. The only
attractive etymological interpretation is that suggested by Carruba,
regarding ko›on ‘sheep’ (Carruba 1965). Carruba proposed connecting
it with Cuneiform Luwian ¢àôa/ì-, Hieroglyphic Luwian ha-wa/i/- <
PIE *h2e/owo- “sheep”. The new form for the gloss defended by Erbse
(1986), kÒon, would support this etymological explanation, because it
would come from an intermediate form *kÒWon, a plausible Carian
result of Proto-Anatolian (henceforth PA) *Hàwo- (the stem was not in
-i- originally in Anatolian, cf. Lyc. xawa-, and see Melchert CLL s. v.
¢àôa/ì-). The treatment of PA laryngeal as velar stop in Carian (as in
Lycian) is now clearly confirmed (see below p. 260).
As for the possibility of finding some of these glosses in the Carian
inscriptions, the results are equally discouraging, although in theory
words meaning ‘tomb’ or ‘king’ are likely to appear. In fact, thanks to
the bilingual inscription of Athens (G. 1), we know that ≤jas is proba-
bly one of the Carian words used for referring to a tomb or a funer-
ary monument. The word appears as ≤as in Euromos (C.Eu 1). The
possibility of connecting these forms with the gloss soËa(n) was con-
ceived by Meier-Brügger (Meier-Brügger 1979:81),9 but we must be
aware that in order to connect all of these forms, a lot of non-trivial
sound changes are needed (*/swa/- > */swa/- > */sja/ > ≤ja- (> ≤a-),
for instance), and in any case, -s in ≤( j)as?, vs. -n or -Ø in soËa(n)
would remain unexplained.
In the case of the word for ‘king’, it is commonly supposed that it
could be very similar in Carian to the corresponding word in Lycian,
xñtawat(i)- /k–dawati/. Adiego (1994a:240) proposed that the form
ºk?dow“ (part of a word esak?dow“, E.AS 7) could be the Carian word
for ‘king’, and this hypothesis was substantiated in Adiego (1995:18–21)
by the Lycian-Greek-Aramaic inscription of Xanthos (Lycia, N 320),
wherein the Carian divinity “King of Kaunos” appears in Aramaic as
KNDWÍ (KNDWS) KBYD”Y. This seems to imply that KNDWÍ-
KNDWS could be the Carian word for ‘king’ (see below Chapter 11
s. v. esak?dow“ for more details).

9
It is somewhat remarkable that Meier-Brügger’s proposal, which implies an ≤ (a
sort of sibilant) value for z, was formulated some time before this value of z was con-
vincingly established by J. D. Ray in the context of a wider system of decipherment.
 THE INDIRECT SOURCES 11

If we accept these identifications, the problem of g°la is very simi-

lar to that of soËa(n): g°la and a Carian word kdow (“ ) have some
points in common (the velar initial, a possible correspondence -l- :
-d- (<*/nd/),10 but a series of fragile hypotheses about sound changes
would be needed to bring both forms closer.
An alternative approach to these glosses has been to question their
validity by taking into account that some of the place names glossed
by Stephan of Byzantium allow an analysis based on Anatolian ono-
mastics, which is clearly at odds with the etymological explanation of
the Greek source. For example, in the case of ÉAlãbanda, Brandenstein
(1936) observed that the word displayed the typical Anatolian suffix
-anda, frequently used for place names, thus ruling out the segmenta-
tion ÉAlã-banda on which Stephan’s etymological explanation was based,
and consequently the validity of his information.
However, discrediting the information provided by Stephan is hardly
reconcilable with the idea, expressed by Dorsi (1979:34), that these
Carian glosses must all be considered as qualitatively important, since
they may have borrowed from Carian writers of the Alexandrian era.
This would imply that we are dealing with first-hand evidence, going
back to authors on Carian topics who could be still fluent in Carian,
and who are therefore extremely reliable sources.
In my opinion, there is no implicit contradiction in accepting both
the information provided by Stephan of Byzantium and the different
interpretation of the place name from the perspective of Anatolian ono-
mastics: a folk-etymological analysis of ÉAlã-banda as a compound of
two authentic Carian names does not say anything about the actual
origin of this Carian place name. In other words, even if we accept
that Stephan’s etymological explanation of the place name ÉAlã-banda
was probably flawed, this does not necessarily invalidate the informa-
tion that êla means ‘horse’ and bãnda ‘victory’. A good example of
this is to take just to one of the place names involved: Souãggela was
later replaced by Yeãggela, undoubtedly by a process of folk-etymo-
logical hellenisation (yeÒw + êggelow). If someone had glossed this place
name as “coming from yeÒw ‘god’, and êggelow ‘messenger’” and as
meaning ‘messenger of God’ or similar, the information about these

10
In the Carian language of Thebes, the letter & d is not used, and if one accepts
that Carian of Thebes mlane corresponds to Carian of elsewhere mdane, a sound change
d > l could be imaginable.
12 CHAPTER TWO

latter words would be correct, despite the fact that the place name was
not created ex nihilo with this slightly absurd meaning, but rather was
the result of an intricate process of deformation.11 A similar explana-
tion could be acceptable for all the Carian glosses based on place
names.12

B. THE PROPER NAMES

1. The Unity and Continuity of Anatolian Onomastics

Before the decipherment of Carian writing, our only source of reliable
information on the Carian language was (besides the scarce glosses
analysed in the foregoing section) the onomastic corpus, essentially per-
son and place names transmitted by indirect sources, particularly Greek
ones. The most notable aspects of the Carian onomastics observed dur-
ing the history of the research are shown here:

(1) Carian onomastics is to a large extent inseparable from the ono-

mastics of other minorasiatic regions that have also reached us through
indirect sources.
(2) These onomastics of Asia Minor in Greek transmission, which
can be ascribed to a period that stretches from the middle of the first
millennium B.C. to the first centuries A.D., turn out to be a clear con-
tinuation of the Hittite-Luwian onomastics of the second millennium
B.C.
(3) A great part of these onomastics, traceable from the beginning
of the second millennium B.C. to the start of Christian Era, can clearly
be linguistically interpreted as belonging to the Indo-European Anatolian
group. We are dealing therefore with onomastics that are, linguistically
speaking, Anatolian.
(4) In any case, Carian proper names display certain characteristics
that make them distinguishable from other Anatolian onomastics:

11
An intermediate phase Yuagg°l/a/ (ethnic Yuagg[°leuw] is also documented, see
Hornblower (1992:99, n. 160).
12
For some etymological proposals (all rather provisional) formulated about Carian
glosses, see Adiego (1993a:22). For ÉAlãbanda as ‘rich in horses’ (my suggestion, purely
hypothetical), see Adiego (1993a:21).
 THE INDIRECT SOURCES 13

several name types, for instance those in -vllow, -vlliw, are typically
Carian, and this singularity can be attributed to specific phonological
and/or morphological traits of the Carian language.
The unity of the Anatolian onomastics transmitted by Greek sources
was established by Paul Kretschmer in chapter X (“Die kleinasiatische
Sprachen”) of his Einleitung in die Geschichte der griechischen Sprache (Kretschmer
1896). Kretschmer believed that in these onomastics a Pre-Indo-European
substratum common to all Asia Minor (with the exception of Phrygia)
was recognizable, given that many concrete elements (lexemes as well
as suffixes) appeared in different regions of the Anatolian Peninsula.
The hypothesis of a single minorasiatic group sui generis (Kretschmer
1896:292) constituted in those days an innovation, because the earlier
theories of other scholars (reviewed summarily in Kretschmer 1896:289–
292) had tried to establish different linguistic groups and attribute a
different external kinship to each one (with Indo-European, Semitic or
Caucasian languages). It is true that Kretschmer was wrong in classi-
fying this sui generis group as non-Indo-European, but in his defense,
we must bear in mind that the existence of an Indo-European Anatolian
family was in those days difficult to imagine. But in fact, this negation
of the Indo-European character of minorasiatic languages had a posi-
tive effect: it obliged scholars to adopt the combinatory method for
analyzing Lycian (and later Lydian) inscriptions, and to discard more
fragile etymological approaches.
Moreover, Kretschmer’s seminal work already outlined some of the
ideas—either new or systematized by him—that in the course of the 20th
Century have become vital to the research on Anatolian: the identification
of -ss- and -nd- as suffixes and their possible connection with Greek
place names in -ss- and -nd-; the frequent appearance of the so-called
Lallnamen (names whose structure seems to be characteristic of children’s
language: CV, CVCV, VCV, etc., like Dada, Nana, Ada . . .); and the
isolation of lexical items that enter in compounding or derivation, as
pig-, imbr-, tarku- or -muhw, nowadays easily interpretable as Indo-
European Anatolian stems.
The approach begun by Kretschmer reached its peak in the monu-
mental work of Johannes Sundwall devoted to Lycian indigenous names
(Sundwall 1913). Sundwall tried to establish a systematic study of Lycian
onomastics, isolating and grouping the different formative elements in
the proper names. Although it contains mistaken readings and names
that are clearly Greek incorrectly analyzed as indigenous, Sundwall’s
14 CHAPTER TWO

book was a fundamental work in the history of the research on Anatolian

onomastics. It was only to be superseded by the work of Ladislav Zgusta,
who in two books of enormous significance produced an exhaustive
collection of the Minor Asian person and place names of Classical
sources (Zgusta 1964, 1984).
When a connection could be established between these onomastics
and those of Hittite and Luwian of the second millennium, it heralded
a dramatic change of focus in the study of Minor Asian onomastics
from Classical sources. Obviously, this step forward became possible
only when these languages in Cuneiform writing were sufficiently under-
stood. Friedrich (1931) had already drawn attention to the lexeme muwa-
in Hittite and Luwian proper names, a lexeme identical to the element
muwa- present in the indigenous names of the first millennium. The
idea of a linguistic continuity between the second and first millennia
that could be detected thorough the proper names was given further
weight by new evidence in Goetze (1951), but the definitive work on
this subject is the fundamental book of Houwink Ten Cate (1961), who
demonstrated with a great number of examples that the Lycian names
contained lexical elements of Anatolian origin. This theory came as no
surprise, however; parallel to the study of onomastics, Lycian had been
clearly established as a member of the Anatolian family, first in Pedersen
(1945) and later by the work of Laroche, who demonstrated that Lycian
was very similar to Luwian (Laroche 1958, 1960, 1968). More impor-
tantly, given the fact (already established by Kretschmer many years
before, cf. above) that Anatolian onomastics in Classical sources showed
a clear unity that suggested the existence of a linguistic group, the attri-
bution of Lycian to the Indo-European Anatolian family together with
Hittite, Luwian, and Palaic, made it very likely that the languages spo-
ken in Cilicia, Isauria, Caria and in the other Minor Asian regions
where this type of onomastics is well documented also belonged to this
same Anatolian Indo-European group. In the case of Carian, this hypoth-
esis constitutes one of the basic tenets of the proposals formulated by
”evoro“kin since his first works, and seems to be confirmed by the
definitive decipherment of Carian.

2. Present Compilations of Carian and Anatolian Names

Nowadays, we have at our disposal two complete and very up-to-date
repertories of Carian place and person names respectively, both com-
piled by Wolfgang Blümel (Blümel 1992, 1998a). For the other Anatolian
 THE INDIRECT SOURCES 15

proper names from Classical sources, the books of Ladislav Zgusta

(Zgusta KPN and the supplement Zgusta 1970 for the personal names,
Zgusta KON for the place names) are still invaluable, although in the
case of Lycian, there are now two updated repertories of personal names
compiled respectively by Anne-Valerie Schweyer and Nicola Cau, which
appeared after Zgusta’s works (Schweyer 2002:95–128, Cau 2003). It
is to be hoped that these updates of Zgusta’s continue to appear, since
not only do they incorporate new forms, but also offer revised forms
and correct some mistakes, inevitable in works of such dimensions as
Zgusta KPN and KON.
For Anatolian onomastics of the second millennium, Laroche’s reper-
tories remain indispensable: Laroche (LNH) for person names, and
Laroche (TA1) and (TA2) for place names.

3. Carian Names from Indirect Sources vs. Those from Direct Sources
A methodological problem arises when dealing with the Carian names
found in indirect sources. In Adiego (1993a), these were analysed
(although not exhaustively, looking only at the most significant aspects)
dispensing totally with those Carian names directly attested in Carian
inscriptions. The reasoning seemed be well founded at the time; the
principal aim of the book was to defend a new decipherment of Carian
wherein the use of Carian onomastics from indirect sources played a
fundamental role, so it seemed preferable to avoid mixing forms directly
obtained from the decipherment with indirect forms from Greek sources,
which were the key to that decipherment (In Adiego 1993:26, n. 4,
”evoro“kin’s mixing of forms in his book of 1965 was severely criticized).
Nowadays the situation is clearly different; this book is not conceived
as a justification of a concrete proposal of decipherment, but is meant
to offer as complete a picture as possible of our current knowledge of
Carian, now that the new decipherment has been universally accepted.
I therefore believe that all the data available from the indirect attes-
tation of Carian names must be taken into account together with the
Carian onomastics from direct sources. The study of Carian onomas-
tics, both from indirect and direct sources, will consequently be dealt
with in a specific section (Chapter 8). In Appendix C, a list of the
Carian personal and place names, mostly based on the compilations
published by Wolfgang Blümel (Blümel KarPN and KarON) is pro-
vided. The list is merely intended to offer a convenient compendium
of Carian indirect onomastics: for the data corresponding to each name,
16 CHAPTER TWO

the reader should turn to the articles written by Blümel. For extremely
useful, updated accounts of the data obtainable using only the Carian
names from indirect sources, without discussion of the direct documen-
tation, I refer the reader to Neumann (1988) and, above all, Neumann
(1994).
 CHAPTER THREE

THE INSCRIPTIONS

A. INTRODUCTION

The most direct and important sources of Carian language are obvi-
ously the inscriptions in Carian alphabet, although strangely the bulk
of this epigraphic corpus does not come from Caria itself, but from
various other locations in Egypt. The historical reasons for this curi-
ous circumstance have been covered in Chapter 1.
Inscriptions on funerary stelae and other objects, mainly from Memphis
and Sais, and graffiti found in other parts of Egypt are the result of
this long presence of Carian-speakers in Egypt. About 170 inscriptions
have been found in Egypt to date. All these texts are relatively short,
given their typology (onomastic formulae in funerary texts—Carians
were somewhat laconic when writing epitaphs—and brief graffiti).1
The epigraphic material found in Caria itself is far less abundant
(approximately 30 inscriptions), but it includes several texts that are
more extensive than those discovered in Egypt, particularly the fol-
lowing three: a decree from Kaunos whose precise terms are still
unknown (C.Ka 2), the proxeny decree for two Athenian citizens writ-
ten in Carian and Greek, also from Kaunos (C.Ka 5), and a decree
enacted by the Carian satraps Idrieus and Ada, possibly concerning a
syngeneia of the temple of the god Sinuri, near Mylasa (C.Si 2). To these
three inscriptions now must be added the new inscriptions of Mylasa
(C.My 1) and Hyllarima (C.Hy 1), the latter in fact a fragment that
completes the inscription already known.
Besides Egypt and Caria, we know of several other inscriptions found
in the bordering regions of Lydia and Lycia, as well as in Greece. For
convenience, I will classify the texts of Tralleis and of Krya (on the
Gulf of Telmessos) as Carian, since we are dealing in both cases
with areas very close to Caria. It is logical to assume that there was

1
On Carians in Egypt, see Masson (1969), Masson (1977[78]) and now Vittmann
(2003:155–179).
18 CHAPTER THREE

a linguistic continuity beyond the alleged political boundaries of Caria,

and Carian was spoken in some border zones of Lydia and Lycia.2 For
the same reason I classify as Carian the graffito from Didyma, near
Milet, situated on the Carian border.

1. The Revised System of Transcription of Carian Letters

Before dealing with the Carian corpus of inscriptions, it seems appro-
priate to present here the system of transcription of Carian letters that
will be used throughout this book, since certain new conventions are
introduced here. The phonological reasons behind the new transcrip-
tion procedures for several signs will be dealt with in Chapter 7, and
here I will limit myself to giving a brief justification of them.
Although the bilingual of Kaunos has confirmed the overall ‘Ray-
Schürr-Adiego’ system (see Chapter 4), and by extension the deci-
pherment tables in Adiego (1993a), (1994a) remain essentially valid,3 I
believe that this is a good opportunity to introduce slight modifications
to the system. These will allow us to adjust the transcriptions of sev-
eral letters more precisely to their actual phonological value, and to
simplify other conventions in transcribing Carian signs. Moreover, since
the publication of the works mentioned above, hypothetical sound val-
ues have been proposed for some formerly undeciphered letters, which
will also be considered in the revised system presented here.

2. Vocalism
The method of transcribing vowels used until now has been largely
superseded by our improved understanding of the Carian vocalic sys-
tem. In the case of i / j and u / v, a purely diacritical distinction

2
Tralleis was situated north of the River Maiander, which served as the traditional
boundary between Lydia and Caria, but this boundary was undoubtedly permeable to
contact between people, see Hornblower (1982:2). According to Strabo (XIV, 1, 42),
Tralleis was inhabited by Lydians, Carians and Ionians, and the Carian flavour of the
alternation -ll-/-ld- in Tralle›w vs. Tralde›w was already noted by Benveniste (apud
Robert 1945:20, n. 2). As for the Krya inscription, it clearly belongs to the Kaunian
alphabetic variety, which is congruent with the geographical proximity of the two
places.
3
As a exception note only the Kaunian letters T t (vs. “ in the rest of the alpha-
bets) and / “ (a specific Kaunian sign), whose value has been established from the
bilingual (about these letters see here pp. 228–229).
 THE INSCRIPTIONS 19

(“i”/“í”, “u”/“ú”) was employed in order to reflect the alternations

detected between each pair of letters. At present, it seems clear that,
at least originally, the alternation lay in a vowel vs. semivowel oppo-
sition, so that a transcription i / j, u / w would be more adequate. In
the Carian alphabets of Caria itself, j and v were abandoned, and
i and u took up their functions (cf. Kaunian u≤ol≤ vs. Carian of Egypt
w≤ol≤ ).
As for W and w, their character of ‘u-sounds’ led to a conventional
and perhaps flawed transcription “ù” and “w”, respectively. The idea
that these letters represent a /y/ sound, already proposed in Adiego,
has been confirmed, at least in the case of W, thanks to the Kaunos
bilingual (see p. 237). Therefore a transcription of W by means of y
must be preferred. Consequently w, a letter that alternates with W, will
be transcribed as ÿ. It is possible that w represented the semivowel cor-
responding to /y/, id est, /w/, but it is not absolutely certain, and
moreover the transcription w could lead to confusion due to its formal
resemblance to y (cf. the similar problem caused in Lydian transcrip-
tion between n and v).
The letter V/W is only documented in Sinuri, Kildara and Mylasa,
where no traces of W are found. Assuming that a similar process of
eliminating semivowel letters occurred, the simplest solution is to inter-
pret V/W as the Sinuri-Kildara-Mylasa form for W = y, despite of the
(apparently) more formal proximity of V/W to w than to W. For this
reason, V/W will be transcribed by <y>, in the same way that T, the
Kaunian form of t = t, is transcribed simply by <t>.

3. Consonantism
The case of Kaunian T = t and / = “ has already been mentioned;
given that there are no doubts about these values, and that these let-
ters are simply the forms that t and f, respectively, adopted in the
Kaunian alphabet, it is not necessary to resort to more complex tran-
scriptions, such as t2—used by Marek-Frei in their works—or “2.
The transcription of x X merits further consideration. In Ray, and
in my early works, it was transcribed by h. A new transcription was
introduced—suggested by Neumann—in my subsequent works: x, which
has been commonly accepted. While this latter transcription is more
consistent with the tectal and stop value of x X, discernable from the
Carian-Egyptian equivalence ursxli- = 3rskr, a more in-depth analysis
of Carian has allowed us to establish a more precise value for this
20 CHAPTER THREE

letter, which cannot be a velar aspirated stop or a velar fricative, the

kind of sounds that the transcription x would imply. Rather there is
evidence that points to a palatal stop (see Chapter 7), hence the pro-
posal that it should be transcribed as ∞, the grapheme used in Indo-
European Linguistics for PIE voiceless palatal stop.4
The remaining consonants whose values are well known remain
unchanged. Further modifications involve several letters of less certain
value. In the case of 1 = z (instead of z), the reason is to avoid employ-
ing too many Greek letters in transcribing Carian. Insofar as z was not
used at all, I think it reasonable to substitute it for z, although I rec-
ognize that z is a somewhat ambiguous letter and its IPA value is cer-
tainly not the same as 1 presumably had in Carian. However, I believe
that this simplification is desirable, in the same way that there is a ten-
dency to use x instead of x in transcribing Lycian.
As for 0 and %, it is likely that they represent several types of tec-
tal, but the evidence is inconclusive, and an exact value is impossible
to determine. The possible identification ≤u0li- = Souaggela points to
a *ng origin for 0, in parallel to & d < *nd and Ø b < *mb, so that
a corresponding g transcription is adopted, although it is true that the
option of <g> was still available. Regarding %, the only clue we have
is Schürr’s proposal to identify pr%idas with Bragx¤dai. Although he
now seems to have abandoned this hypothesis, I still see it as an attrac-
tive idea that should not be ruled out. The purely conventional—and
far from certain—transcription of <ã> is provisionally adopted here,
although there is no strong supporting evidence.
A rather different problem is posed by O, an exclusively Kaunian
letter. In Adiego (2002) I offered arguments in favour of identifying it
with the far more widespread c letter, absent in Kaunos (see here p.
252). In any case, I recognize that my arguments are not particularly
strong, so I adopt, also cautiously, a transcription t2.
To avoid confusion, I shall not attach a question mark to these rather
uncertain values in transcribing Carian texts, but instead I will indi-
cate such cases in the sign tables.
The following table shows the new system of transcription of Carian
used in this book:

4
A likely alternative was <c>, which would coincide with the letter used in IPA
for this type of sound. However, this is a very ambiguous letter in Indo-European stud-
ies, so a more precise letter like ∞ seems preferable to me.
 THE INSCRIPTIONS 21

Nº (Masson) Letters Transcription

1(+8) a A ~ À (E) a
3 dDG d
4 l l
5(+41) WùVW y
6 rR r
7 s2L l
9 qQ q
10 b5B b
11 mM m
12 o o
14 tT t
15 f F S/ _ “
17 s s
18 H ?
19 uU u
20 ñ ñ
21 xX ∞
22 nN n
24(+2) p p (¯) p
25 zZ ≤
26(+8) IíÎÏyìYI i
27 ee e
28 w ÿ
29+30 kK k
31 & d
32 vÚ w
33(+34) 0 8? g
35(+36) 199 z
37 % ã
38 jT j
39 _ 1 ?
40(+23?) c C / O? t / t2?
42 6 ®
43–44–45 Ø ® 4 B &? b? b
46 ÿ b2?

Note finally two conventions also used throughout this book. Dagger (†)
indicates a transcription of a word or letter by means of a different
system of decipherment. Double-dagger (‡) is used to mark an old read-
ing of a word or letter now discarded. A combination of both signs
(†‡) serves to signal both the use of a superseded deciphering system
and an erroneous reading.
22 CHAPTER THREE

B. ‘PARA-CARIAN’ OR ‘CAROIDE’ INSCRIPTIONS

A serious problem is caused by a heterogeneous body of inscriptions,

whose only common characteristic is the use of writing systems that
have, or seem to have, certain similarities with the Carian alphabet,
but which cannot be definitively classified as Carian inscriptions. A
detailed analysis of this material cannot be provided here, so I merely
intend to establish the problem in precise terms and to enumerate
thecorpus of these inscriptions regularly referred to as “para-Carian”
or “caroide”.
For years, the temptation has existed to attribute any inscription from
Asia Minor written in an unknown or barely recognizable alphabet to
Carian. In a sort of obscurum per obscurius, such materials were classed
as Carian at a time when the Carian alphabet itself remained un-deci-
phered. Today, we have a better understanding of the Carian alpha-
bet (letter values, geographical variants, a complete inventory of signs)
and we can reject the theory that these materials are Carian (canoni-
cal Carian, at least).

1. “Para-Carian” Inscriptions from Caria

The “para-Carian” or “caroide” inscriptions discovered in the Carian
area can be separated geographically in five groups: (1) Khalketor, (2)
Ancin (south of Alabanda), (3) Labraunda, (4) Stratonikeia (a single
inscription), (5) Aphrodisias (also a single text). Groups 2–4 are in fac-
trock graffiti that are practically impossible to read. From the existing
editions of the inscriptions, nothing can be acceptably identified as
Carian.
The cases of Khalketor and Aphrodisias are different; in Khalketor,
at least one of the two inscriptions discovered is clearly readable. But
this simply allows us to state that the alphabet of Khalketor has noth-
ing to do with the Carian writing system. Concerning the language of
texts from Khalketor, all attempts to interpret them using our limited
knowledge of Carian have proved fruitless. In recent years the theory
has been suggested that we are in fact dealing with much more recent
inscriptions, perhaps written in Turkish.
Finally, the Aphrodisias inscription consists of eight letters (the last
one incomplete) and does not pose problems of reading. Some letters
resemble Carian ones, but others are totally alien to the Carian alpha-
bet. As a matter of fact, we cannot discard the possibility that this
inscription actually represents the Carian alphabetic variety of Aphrodisias,
 THE INSCRIPTIONS 23

but in this case, we ought to accept that this variety was extremely
uncharacteristic. Since further Carian epigraphic testimonies from
Aphrodisias are lacking, this text (which is also uninterpretable) will be
excluded from our corpus.

These texts had been included in the successive inventories of Carian inscrip-
tions: Ancin = D 5, Labraunda = D 17 (recent edition: Meier-Brügger 1983),
Estratonicea = 26* (Hanfmann-Masson 1967); Chalketor = D 4 and 27*
(Neumann 1969a; revised in Blümel 1988); for the Aphrodisias text, see
Innocente (1994:107–108 [text nº 7]). Blümel (1998:168) mentions the exis-
tence of several other inscriptions in Chalketor’s alphabet. The ‘Turkish
approach’ has been suggested by Hasan Malay to Blümel, see Blümel (1998:
169).
I have some doubts about the true Carian character of the two identical
inscriptions from Keramos (here C.Ke 1, C.Ke 2), but I retain them, though
in a provisory way, in the corpus insofar as all the letters that appear there
can be analysed as Carian.
The so-called ‘tegola de Iasos’, an inscription consisting of eight signs pub-
lished by Lucia Innocente (Innocente 2002) could be Carian, but, as Innocente
herself observes, the very few letters that can be identified are not unequivo-
cally Carian, and no valuable results can be obtained. Therefore, I exclude
it from the present corpus.
For similar reasons I exclude from the corpus the alleged Carian inscrip-
tion from Labraunda, recently published by Belli and Gusmani (Belli-Gusmani
2001). According to Gusmani, the inscription reads (a) e E 2 e s m (b) M
U.. While part (a) has a Carian flavour, the co-occurrence of e and 2 would
be very strange. Also, a form such as E is rather puzzling. As for (b), as
Gusmani rightly observes (Belli-Gusmani 2001:41), a form M instead of the
typical Carian form s for s is surprising.
As for the alleged Carian fragment from Kaunos 51*, consisting of only
two letters c a (see Frei-Marek 2000:125–126), I consider it as non-Carian: the
letter c is totally absent from the Kaunian alphabetic inventory, and the form
of a is very different from the all the variants attested for this letter in the
Carian alphabets.

2. ‘Para-Carian’ Inscriptions from Other Places

The rest of the ‘para-Carian’ or ‘Caroide’ inscriptions come from other
locations: from the islands of Cos and Rhodes, possibly part of the
Carian linguistic area; from the neighbouring countries Lydia and Lycia;
from Egypt, and also from Persia. Even some texts of unknown origin
have been arbitrarily interpreted as “Carian”. None of these texts can
be included in a strict corpus of Carian inscriptions, but given their
presence in former collections (particularly in Meier-Brügger 1983), I
will include them here, accompanied by a brief note:
24 CHAPTER THREE

• Inscription from the area of Telmessos (Lycia; 36* = Meier-Brügger

1981). It seems to be related to the Chalketor corpus, but its authen-
ticity is not absolutely conclusive.
• Inscriptions of Belevi (Ephesos, Lydia; 24* = Dressler 1966–67).
Three inscriptions on stone fragments. Some letters seem distinctly
Carian, whilst others are clearly alien to the Carian alphabet.
• Inscription of Persia (32* = Pugliese Carratelli 1974[77]). An inscrip-
tion on a bronze bowl, whose only alleged Carian trait is a sign, &,
which recalls an identical Mylasean letter now known thanks to the
new inscription C.My 1, and also, although more remotely, the typ-
ical Carian letter &. However, the presence of letters alien to the
Carian signs inventory (P, K, Z) invalidates their classfication as
Carian.
• Tablets Peiser-Böhl-von Grotthus (23*). A group of three tablets, the
first two edited by Böhl (1932/33), the third by Friedrich (1965), and
carefully analyzed by Meriggi (1966). The exact origin of the tablets
is unknown, as is their dating, but they have traditionally been attrib-
uted to Cappadocia. This means that the labelling of their writing
system as ‘para-Carian’ is based merely on the alleged graphic sim-
ilarities between this system and the Carian alphabet.

Such similarities are indeed clear. It is particularly striking that several

typical—and in some cases exclusive—Carian letters such as p, z, k,
v, j seem to appear. But there are also Greek letters alien to the
Carian alphabet (for instance K), and a number of signs that cannot
be compared either to the Carian or to the Greek alphabet.
The main difficulty in interpreting these three tablets lies in many
cases in the impossibility of distinguishing between which signs are actu-
ally different, and which are mere variants. For example, the third
tablet (Friedrich 1965), contains several types of r pointing in different
directions and with slight modifications of shape, and it is impossible
to know whether we are dealing with the same sign in all instances.
As an indication of these difficulties, it is worth considering that whereas
Friedrich identifies 217 different signs in this tablet, Meriggi reduces
this number to 61, and Nahm (1974) further still to 34 (although in
this latter case, Nahm applies somewhat unconvincing criteria). As a
result, it is hard to ascertain whether the graphic system is alphabetic,
syllabic, of a mixture of both. Furthermore, we are unaware of the lan-
guage in which the tablets are written. All these problems make these
texts undecipherable, and the Carian specialist must limit himself to
 THE INSCRIPTIONS 25

noting the rather surprising graphic affinities between the Peiser-Böhl-

von Grotthus tablets and the Carian alphabet. Only a significant increase
of the documentation available would allow us to move forward in
their study.

The only effort at decipherment of which I am aware is that of Nahm, regard-

ing the third tablet (Nahm 1974). After reducing the inventory of different
signs as far as possible (cf. above), he transcribes with Greek letters those signs
that show a certain similarity to them. His observations on several recurrent
sequences are of some interest, but they are conditioned by the somewhat
questionable attempt to decrease the total number of signs.

• Graffito of Cos (29* = Metzger 1973). A graffito inscribed on an

amphora coming from the island of Cos. In this case, doubts arise
from the almost complete illegibility of the inscription. ”evoro“kin
(apud Metzger 1973:77) and Metzger claim to recognize some Carian
letters such as j, v, or i, but a complete reading of the text is
unfeasible.
• Graffito of Ialysos (Rhodes) (Innocente 1994:101–104 [text nº 1]).
Innocente has defended the inclusion of this four-letter graffito, for-
merly considered to be Lycian, in the Carian corpus. It is true that
all the letters of the graffiti could be Carian, but none of the possi-
ble readings of this brief inscription offers compelling results (← u“tu,
uatu, even utku /→ ut“u, utku)
• “Poetto’s two para-Carian inscriptions”. Innocente (1994:108–109
[texts nº 8 and 9]) includes in her collection of supposed Carian
‘monstrosities’ two texts whose existence has been noted by Massimo
Poetto: a cylindrical seal of unknown origin and extremely old pos-
sible dating (1000 B.C.!), and an object classified among the “Grie-
chische geometrische Gemmen”, also of unidentified provenance.
Each inscription contains only a few signs, and their attribution to
Carian is based only on the possible resemblance of some letters to
those of the Carian alphabet. Consequently, I believe that it is prefer-
able to exclude them from the Carian corpus.
• The “Oxford Para-Carian Inscription”, a tablet from Sardis first pub-
lished by Sayce, and recently re-published with excellent photogra-
phy by Innocente (Innocente 1995). The situation is similar to the
preceding para-Carian texts: whilst some letters essentially resemble
certain Carian signs, others are completely alien to known Carian
repertoires (for example the three-straight sigma), and no linguistic
interpretation is possible.
26 CHAPTER THREE

• The most extreme case of resorting to Carian script as a panacea,

is the attempt of to classify as Carian a number of masons’ marks
found in very different locations. Gosline (1992) proposed a Carian
origin for the masons’ marks discovered in Elephantine (Egypt); in
another paper, the same author proposes the same origin for the
masons’ marks of Pasagardae (Persia) (Gosline 1998). This line of
argument has also been taken up by Franklin (2001) in order to
explain masons’ marks found in North Israel (Samaria and Megiddo),
dated to as far back as the IX century B.C. More recently, Avishur
and Heltzer devoted a brief article to connecting these latter exam-
ples from Israel with the biblical references to the kàrì (Avishur-
Heltzer 2003).

Although it remains possible that Carian masons were responsible for

some of these marks (in the case of North Israel, however, I would
think it unlikely),5 a careful observation of the collections of marks does
not throw up compelling evidence of a connection with Carian: where
the letters that can be easily identified are not specifically Carian (for
example, an inconsequential letter such as a), the links proposed with
distinctively Carian letters (w, j) are extremely tenuous.

3. The Ostrakon of Hou and the Naukratis Fragment

None of the para-Carian or Caroide texts mentioned above comes from
Egypt. In fact, the only examples of para-Carian inscriptions found in
Egypt until now are the so-called Carian Ostrakon of Hou (Diospolis
Parva), and the inscribed fragment of a piece of pottery known as the
Naukratis fragment.
The Ostrakon was discovered by Flinders Petrie in 1899. Sayce (apud
Petrie 1901) considered it to be “Kaunian”, and later re-edited it him-
self as a Carian inscription (Sayce 1905). Masson and Yoyotte excluded
the text from their edition of “pharaonic objects”, but some years later,
Masson published it as a truly Carian text (Masson 1967). This view,
however, seems to be too optimistic: once again, as in the para-Carian

5
Possible mentions of Carians in Biblical sources are, according to Hornblower, “an
inviting but nebulous topic” (Hornblower 1982:16, n. 82). Avishur-Heltzer (2003) does
not add any new evidence for these alleged references, and employs dangereously cir-
cular reasoning: the kàrì of the Bible are Carians precisely because the masons’ marks
show Carian letters (!).
 THE INSCRIPTIONS 27

texts mentioned above, a few resemblances to Carian letters are found

alongside clearly non-Carian letters, and none of the sequences of let-
ters offers the slightest possibility of a connection with known Carian
words and sequences. It is rather surprising that Masson, generally scep-
tical about the acceptance of other documents as Carian (compare for
instance his exclusion of the scarabaeum inscription, E.xx 5, much more
likely to be Carian, Masson 1959b) defended the addition of this text
to the Carian corpus (Masson 1967, a view repeated in Masson 1969:32).
I include here Masson’s drawing of the ostrakon:

Ostrakon of Hou (Masson 1967)

The text inscribed on a piece of pottery (the ‘Naukratis fragment’), also

found by Flinders Petrie, consists of only three letters, whose Carian
character is unverifiable, and must therefore be excluded from a seri-
ously conceived repertory of Carian texts. See Masson-Yoyotte (1956:
14–15) for further remarks.

4. Carian Grafitti from Sardis

The Carian graffiti found in Sardis deserve closer attention. The Carian
character of these texts, edited by Gusmani (1975) is in my opinion
beyond doubt, since they contain some of the most typical Carian let-
ters, such as p, z or %. However, the scarcity of the inscriptions,
their highly fragmentary nature, and the serious difficulties encountered
when analyzing and interpreting the texts (the direction of reading and
the exact identification of the letters remain unclear in most cases) mean
28 CHAPTER THREE

that we must exclude them, at least provisionally, from the current

Carian corpus.
A good indication of the uniqueness of and the difficulties presented
by the graffiti from Sardis can be seen in the longest example: the join-
ing made by Gusmani (1990) of two fragments (C I 1 and C I 5 in
Gusmani 1975):

Sardis C I 1 + C I 5 (Gusmani 1990)

At first sight, many typical Carian signs can be easily recognized: v

w w ÿ, p p, j j, z ≤, i i, etc., so that the attribution to Carian is
incontestable. The first difficulties arise when one tries to establish the
orientation of the reading. For example, the second line offers contra-
dictory evidence: there are two m, oppositely oriented (m M). Even a
reversal of the text cannot be rejected, due to the form of a. Nearly
every possibility of reading direction implies a different interpretation
of the letters (for instance B = b, but u = u). A further complication
is that some letters are not easily identifiable: Z in the third line could
simply be a variant of L l, but it could also be interpreted as a form
of z; the letter 8 in the left part of the second line resembles a let-
ter only documented in the alphabetic variant of Kaunos, a trait that
seems suspicious. Similar problems of identification are found in other
graffiti: a letter E appears at least three times. Are we dealing, then,
with an ancient form of e e (note the alternative form e in Hyllarima
and Mylasa) or rather with a quadrangular variant of i?
But the main problem with this and the other Sardis graffiti is that
none of the different possible readings yields a sequence that can be
compared to the rest of the Carian documentation. The possible con-
nections are far from conclusive, and only can be based on a handful
of signs. For example, in C I 2, ]-?-mzto [, transcribed as ]-?-m≤to [,
a possible ending of a name in genitive -m≤ could be identified,
 THE INSCRIPTIONS 29

but the brevity of the fragment makes this interpretation simply an ad

hoc one.
In conclusion, the presence of the Carian alphabet in Sardis is unde-
niable, and this may have interesting implications for the chronology
and spread of Carian writing, but from a linguistic point of view, the
scarcity of these materials renders them useless.

5. Carian Inscription from Old Smyrna

A similar problem to that of the graffiti from Sardis is posed by an
inscription from Old Smyrna, the photography and drawing of which
was published by Jeffery (1964, nº 23). It contains clearly Carian let-
ters (note &, e and perhaps also p, f), but it is impossible to obtain
a comprehensive and satisfactory reading of the inscription given the
ambiguity of many of the signs, which can be interpreted in very
different ways, and the impossibility of establishing the exact direction
of the writing.6
One of the most significant features of this inscription is its possible
age: the object is dated at the end of the 7th century (so it would be
more or less contemporary with Carian graffiti from Abu Simbel). There
are no reasons to suggest that the inscription should be dated much
later than the object: the alleged palaeographical reasons put forward
by Jeffery are untenable from a Carian perspective (Adiego, 1993a:86).

Old Smyrna ( Jeffery 1964)

6
Schürr (2001c) offers a very hypothetical interpretation of the text: em-?-l / salpde/
. . . ubrod bore“. No clear connection with the Carian corpus can be established for any
of these sequences.
30 CHAPTER THREE

C. THE CARIAN INSCRIPTIONS FROM EGYPT

The epigraphic materials from Egypt (about 170 inscriptions, to which

50 new and still unedited inscriptions found by ”evoro“kin must be
added, cf. below) are the most important part of the direct documen-
tation of Carian.
From a typological point of view, a differentiation can be made
between funerary and votive inscriptions on the one hand, and graffiti
on the other. The first group has been the best source for decipher-
ing Carian, mainly due to the relative lack of reading problems, the
regularity of the graphic system used, and the excellent editions of the
inscriptions that subsequently appeared.
The scenario presented by the graffiti is quite different. In contrast
to the high degree of standardization of the alphabet used in the votive
and funerary inscriptions, the graffiti show a markedly less careful use
of writing, as a result of the spontaneity involved in their execution.
This problem is further complicated by the lack of modern and reli-
able editions of the main body of graffiti. Of the graffiti already known
since the times of Sayce, only the graffiti of Abu Simbel and Buhen
have been revised and edited with a sufficient degree of reliability to
be used without significant reservations. The rest of the documentation
is still pending new collations, which in some cases will become unfea-
sible due to the likely destruction of the graffiti.
The typological classification outlined above can easily be linked to
geographical distribution: the funerary and votive inscriptions come
from Memphis and Sais, whereas the graffiti have been found in var-
ious locations further to the south. The historical reason for this is
clear; the graffiti are present because of the temporary presence of
Carian visitors, particularly during the military campaigns, while the
funerary and votive monuments exist due to the settlement of Carian
people in cities like Memphis or Sais.
When considering the dating of the Egyptian corpus, it is essential
to mention the important work of Kammerzell on the Memphis Corpus.
It had been assumed that this corpus should be dated after the settle-
ment of Carians in Memphis by order of Amasis, i.e. around the mid-
dle of the VI century (Amasis is believed to have ruled Egypt between
570 and 526 B.C.).7 But Kammerzell, after carefully analyzing the typol-

7
Masson (1978:6–7), Martin-Nicholls apud Masson (1978).
 THE INSCRIPTIONS 31

ogy of the funerary stelae, has established that some of these inscrip-
tions can be assigned earlier dates. It is true that not all of the method-
ological procedures used by Kammerzell in order to determine the
chronology of Caromemphite stelae are similarly convincing, but at
least in the case of E.Me 7, wherein the same person is mentioned in
both the Carian and the Egyptian parts, a dating before 570 seems
very probable. A more speculative suggestion is his identification of the
name pikre- of the stela E.Me 3, as corresponding to the Pigres (P¤grhw)
mentioned in Polyaenus’ Stratagemata (7, 3). This figure was referred to
as an adviser of Psammetichus I in the early years of his reign, so this
stela should be dated approximately between 660 and 620 B.C. But it
is very dubious to base, as Kammerzell does, the chronological attri-
bution of E.Me 3 and typologically similar stelae only on this indemon-
strable personal identification.8
Until now, the oldest Carian inscription from Egypt is the base of
a statue of the goddess Isis, which can be dated to the second half of
VII century thanks to the presence of a cartouche displaying the name
of the pharaoh Psammetichus I. This document is therefore chrono-
logically very close to the arrival of Carian and Ionian mercenaries in
Egypt and their subsequent settlement in the Eastern Delta (Masson
1969:35–36, 1977[78]:335). As for the rest of the Carian documenta-
tion from Egypt, a precise dating can be given only to the graffiti from
Abu-Simbel (E.AS); there is no doubt that these graffiti were inscribed
in the course of the great Nubian campaign ordered by Psammetichus
II and conducted by Potasimto, as is particularly evident in a long
Greek graffito, where mention is made of this historical context. This
campaign has commonly been dated in 591 B.C., but Ray (1982:85)
suggests revising this chronology slightly, to situate the event in 593/92
B.C. The Buhen (E.Bu), Gebel Sheik Suleiman (E.SS) and Murwàw
(E.Mu) graffiti are also likely to date from the same period.
The inscription on a bronze lion “de provenance égyptienne” (E.xx.7)
is dated by Masson (1976) at around 500 B.C., given the Achaemenid
artistic influence visible in the figure of the lion.
For the remaining inscriptions (mainly graffiti plus some texts found
inscribed on various objects), there is no certain dating. Only in the
case of the graffiti from Abydos do we have some idea: Masson has

8
See the same criticism in Masson (1995:176).
32 CHAPTER THREE

suggested that they may be contemporary with the oldest Greek graffiti
from the same location, so that they can be dated around the end of
the V century.

1. Sais (E.Sa)
The sub-corpus of Sais is currently constituted by two bronze votive
objects, included in Masson-Yoyotte (1956). Both texts are bilingual and
have proved fundamental to the deciphering of Carian.

E.Sa 1 (= MY L)

E.Sa 1 (Masson-Yoyotte 1956)

“arkbiom: zidks mdane: ÿn-[?]

mo | den: tumn
(Egyptian text: Jtm ntr ‘3 dj ‘n¢ snb ”3rkbym)

Inscription on a reliquary for mummified reptiles. The only problem-

atic letter is the last letter of the first line. ”evoro“kin’s attempt to com-
plete ÿn[s]mo (see ”evoro“kin 1965:119–120) is only based on the
comparison with ÿnsmsos in E.AS 3 (cf. also E.Mu 1). Masson-Yoyotte
(1956) mention two extremely divergent proposals made by Wild and
Raphäel after examining the original (K k vs. 1 z, respectively!), and
leave the letter without interpretation. The photography in Masson-
Yoyotte is in this case unusable. The Carian inscription is accompa-
nied by an Egyptian formula, “Atum the great god may give life and
health to ”3rkbym”. The non-Egyptian, and presumably Carian, char-
acter of the name ”3rkbym was already noted by Sayce (1905:124), see
Masson-Yoyotte (1956:52). Also correct are the observations in Masson-
 THE INSCRIPTIONS 33

Yoyotte (1956:52) regarding the vocalisation o of ºym, based on its

spelling using the Egyptian word for ‘sea’ ( ym = *yòm cf. Coptic eiom),
and about the possibility that the name would be *Sarkebivmow in a
Greek transcription (the form is not yet documented, but a form Kebivmow
has appeared in the meantime). Surprisingly, though, none of this evi-
dence encouraged Masson and Yoyotte to attempt to find this name
in the Carian text, and the name was not identified there until Kowalski
(1975).

Photograph and drawing: Masson-Yoyotte (1956:53, pl. VI).

E.Sa 2 (= MY M)

E.Sa 2

pdnejt qÿri≤ ∞i

Inscription found on the base of a statuette of the Goddess Neith. For

this text in scriptio continua, I adopt the division suggested by Meier-
Brügger (1979a:81–82), contra Masson (1978) and Meier-Brügger (1979b).
In the Egyptian texts that accompany the Carian inscriptions, a P3-dj-
Njt son of K3rr is mentioned.9 These two names clearly correspond to
the pdnejt son of qÿri- in the Carian part. The use of the biliteral sign
k3 for a syllable /ku/ or sim. (= Carian qÿº) has been correctly identified
by Vittmann, see Vittmann (1996).
The presence of a cartouche with the name of Psammetichus I allows
us to date the object to the times of this Saite pharaoh (663–609 B.C.),
which would mean that this is the oldest datable inscription of the
Carian corpus from Egypt.

Photograph and drawing: Masson-Yoyotte (1956:63, pl. VIII).

9
Other names belonging to the genealogy of the dead are also mentioned. About
these forms and the possible genealogical tree of P3-dj-Njt, see Masson-Yoyotte (1956:61).
34 CHAPTER THREE

2. Memphis (E.Me)
The inscriptions published in Masson (1978) and those of Memphite
origin included in Masson-Yoyotte (1956) are grouped together under
this label. This corpus has been crucial for the decipherment of Carian.
The excellent quality and preservation of a great number of stelae, the
fact that they appear on monuments, which implies a very standard-
ized use of writing, the geographical and chronological consistency of
the corpus, and the fact that it includes some bilingual texts (E.Me 5,
E.Me 7, E.Me 8, E.Me 9, E.Me 15) make this sub-corpus the most
important direct documentation of Carian.
To this sub-corpus, we must add the so-called stela of Abusir (Masson
1978:91, Kammerzell 1993:138–139), although we need also to con-
sider that its reading is very difficult. Very recently, Diether Schürr has
tried to improve the reading of this inscription (Schürr 2003), and his
efforts will be taken into account here. Finally, I also include the frag-
ment 180* from Kammerzell (1993) (here E.Me 66), although it is a
largely unusable document.
Generally, I adopt the readings given in Masson-Yoyotte (1956) and
Masson (1978). The differences, concerning certain details of reading
and, above all, the order in which some texts must be read, will be
duly indicated.

E.Me 1 (= MY A)

E.Me 1 (Masson-Yoyotte 1956)

←
ttbazi[≤] | p. iub[a]Ωi≤ | aor[≤]
 THE INSCRIPTIONS 35

Inscription on a false-door stela. Reading according to Schürr (1996),

the main difference being the interpretation of the first letter of the
second name: Masson’s reading was l l, which must be discarded.

Photograph: Masson (1953: pl. XII). Drawings: Masson-Yoyotte (1956:2–3).

E.Me 2 (= MY B)

E.Me 2 (Masson-Yoyotte 1956)

←
uksmu | lkor≤ | mrsi≤

False-door stela. Unproblematic reading

Photograph: Masson (1953: pl. XIII). Drawings: Masson-Yoyotte (1956:3, 5).

36 CHAPTER THREE

E.Me 3 (= MY D)

E.Me 3

←
pikre≤ ue “arwljat≤ msnord≤

The fifth letter of the second line is not in fact k k, but an inverted
form of l l, given the clear onomastic identification (“ar)wljat = Uliatow.
For the penultimate letter of the second line, I adopt the reading d d
instead of Masson i i, according to Schürr (2001b:103), who follows
on from a new direct reading of ”evoro“kin: cf. also E.Me 48. Close
observation of the photograph in Masson-Yoyotte (1956) supports the
new reading.

Photograph: Masson-Yoyotte (1956: pl. IX). Drawing: Masson-Yoyotte (1956:10),

Kammerzell (1993:146).
 THE INSCRIPTIONS 37

E.Me 4 (= MY E)

E.Me 4 (Masson-Yoyotte 1956)

←
terÿez≤ | upe | nuol∞. [—]sarmrol∞yt

The reading of the text after the second division mark is very doubt-
ful, and the photography in Masson-Yoyotte is not a great help. The
ending in t -t is unexpected. Could it simply be a z?

Photograph: Masson-Yoyotte (1956: pl. I). Drawings: Masson-Yoyotte (1956:19).

38 CHAPTER THREE

E.Me 5 (= MY F)

E.Me 5 (Masson-Yoyotte 1956)

←
psm“kwneit≤ | ue | naria≤ | ≤ugliq | sarl?

The inscription, published in Masson-Yoyotte (1956), was re-edited in

Masson (1978). Despite this new edition, many doubts remain about
the reading of the second, third, and final word. Particularly trouble-
some is the last letter: I adopt Schürr’s reading, l l (the photography
points to l rather than to À).
The first word must not be divided in two, contra Masson.
Bilingual stela. In the Egyptian part, the name of the dead figure is
mentioned: Psmtk-‘wj-Njt, son of W3˙-jb-r‘-[. . . .]. The first name cor-
responds clearly to psm“kwneit-, but there is no connection between the
Carian and Egytian patronyms.

Photograph: Masson-Yoyotte (1956: pl. II). Drawings: Masson-Yoyotte (1956:21, 25),

Masson (1978:92).
 THE INSCRIPTIONS 39

E.Me 6 (= MY G)

E.Me 6 (Masson-Yoyotte 1956)

←
triqo: parma≤≤ ∞i
klorul ∞i

There are no problems of reading.

The stela also contains an inscription in Egyptian, but the Egyptian
names (P3-dj-st, and his mother T3-dj(t)-wsir) do not correspond to the
Carian ones. Therefore, either the stela seems to have been reused, or
we must accept a double denomination—Egyptian and Carian—at least
in the case of the deceased (the second name could be the father’s
name in the Carian text).

Photograph: Masson-Yoyotte (1956, pl. III). Drawing: Masson-Yoyotte (1956:28, 30).

40 CHAPTER THREE

E.Me 7 (= MY H)

E.Me 7

←
tamou tanai≤ qarsio[-?]

Curved stela with images and hieroghlyphical inscriptions.

The reading adopted here is taken from Schürr (2001b:118). The
controversial point is the last letter: in Masson-Yoyotte (1956), it was
read as o o, but in later works, Masson preferred the reading z ≤ (see
for instance the index of words in Masson 1978). The photography
points clearly to o, and the most likely explanation is that the word
is incomplete, since the stone is broken just after this letter. The break
in the text is not particularly big, so only one letter, or at the most,
two, have been lost. Given the typical structure of Carian onomastic
formulae, an integration qarsio[≤] would be a good solution, but I pre-
fer to leave the question open.
The stela provides an Egyptian inscription that also mentions the
dead man T3j-Óp-jm=w son of T3[. . .]. The correspondence to the
Carian text is evident: tamou, son of tanai.

Photograph: Masson-Yoyotte (1956: pl. IVa). Drawing: Masson-Yoyotte (1956:33).

E.Me 8 (= MY K)

E.Me 8a

E.Me 8b
 THE INSCRIPTIONS 41

→
a. paraeym: armon ∞i
b. para!eym: sb polo

Inscription on a bronze Apis. The fourth letter b, E, has always been

a source of difficulty, due to the clear alternation with a (paraeWm
/ parEeWm). It has even been considered an independent sign (Masson
nº 10), and the alternation a / E has received varying explanations.
In my opinion, the simplest solution is to interpret E as an a, its
strange form perhaps being the result of a spelling error and subse-
quent correction.10
The Apis also contains an Egyptian inscription: Ó3py dj ‘n¢ Prjm
p3w˙m “Apis may give life to Prjm the dragoman”. The non-Egyptian
name, Prjm, is logically the transcription of the Carian name Paraeym
mentioned twice in the Carian section. For the problems posed by the
Egyptian word p3w˙m, apparently translated in Carian as armon, see
Chapter 11, s. v. armon.

Photographs: Masson-Yoyotte (1956: Pl. Va, Pl. VIa). Drawing: Masson-Yoyotte (1956:
43, 48).

E.Me 9 (= M 1)

E.Me 9 (Masson 1978)

10
Stephen Durnford (pers. comm. to C. Melchert) has suggested an excellent and
very likely explanation for E as a spelling error: the scribe mistakenly wrote e right
after r, omitting a. He then immediately noticed his error and repaired it by simply
turning e into a by adding the horizontal bar (E).
42 CHAPTER THREE

←
arli“≤: upe: arlio-
[m≤] ∞i: yjas[i≤]

The reading proposed here is an attempt to improve Masson’s edition.

I read the penultimate letter of the first line as i i, instead of Masson’s
j j, according to the photography. As for the integration of the sec-
ond line, both arlio[m≤] and yjas[ i≤] are forms documented in other
Saqqâra inscriptions. The space for integration seems to be adequate,
and from a syntactic point of view, one expects two genitives in -≤
(arli“-≤ . . . arliom-≤ . . . yjasi-≤ ).
The inscription is bilingual. The Egyptian part contains the names
Jr“(3) son of Jrym3, two non-Egyptian names that are undoubtedly the
adaptation of arli“- and arliom-, respectively.

Photographs: Masson (1978, pl. I, 1; II, 1). Drawings: Masson (1978:20, pl. XXXI, 1).

E.Me 10 (= M 2)

E.Me 10 (Masson 1978)

←
[—]q. årm≤: q[—]≤ ∞i: p∂uüi≤ mno≤
[mw]don≤ ∞[i —]w≤ord≤ ∞i

New reading following the work of Schürr (2002:169, n. 9 [but note

ibid. the error u≤ord for w≤ord !]).

Photograph: Masson (1978, pl. I, 2). Drawings: Masson (1978:21, pl. XXXI, 2).
 THE INSCRIPTIONS 43

E.Me 11 (= M 3)

E.Me 11 (Masson 1978)

→
(a) wår[—]t[——]i[—]≤ | mdaÿn
(b) [—15—]a[–]i≤ | mdaÿn

The stela represents a man and a woman. According to Masson, the

drawing included in his edition and reproduced here is too optimistic
with regard to the reading of some signs, so it is preferable to make
a very prudent reading. This would mean that only the respective final
words of each line, in both cases mdaÿn, and the existence of a pre-
ceding form in -≤—also in both lines—are definite. For the icono-
graphical importance of this stela, see Martin-Nicholls apud Masson
(1978:61–70).

Photograph: Masson (1978, pl. II, 2, I II). Drawings: Masson (1978:22, pl. XXXII).
44 CHAPTER THREE

E.Me 12 (= M 4)

E.Me 12 (Masson 1978)

→
pjabrm | w≤ol≤ | mwdon≤ ∞i
kbjom≤ | m[no≤]

In the stela, the prothesis of a woman is represented, which indicates a

female character for the name pjabrm.
The possible integration of the last word was suggested by Kammerzell
(1993:213).

Photograph: Masson (1978, pl. IV, 1). Drawings: Masson (1978:23, pl. XXXIII, 1).
 THE INSCRIPTIONS 45

E.Me 13 (= M 5)

E.Me 13 (Masson 1978)

→
“dtat≤ | upa | w | wet≤ ∞i | mwdon≤ ∞i

As in E.M 13, here the prothesis of a woman is represented.

Photograph: Masson (1978, pl. V, 1). Drawings: Masson (1978:24, pl. XXXIII, 2).
46 CHAPTER THREE

E.Me 14 (= M 6)

E.Me 14 (Masson 1978)

→
irow | pikarm≤ | mwdon!≤

The surprising form of the last word, mvdoUZ mwdou≤ [sic], instead
of the usual mvdoNz mwdon≤, was convincingly argued by Masson as
a simple case of an incomplete sign U u for N n.

Photograph: Masson (1978, pl. V, 2). Drawings: Masson (1978:24, pl. XXXIV, 2).
 THE INSCRIPTIONS 47

E.Me 15 (= M 7)

E.Me 15 (Masson 1978)

←
arli“≤
urs∞le≤
kidbsi≤
(Egyptian text: Jr“(3) s3 n 3rskr s3 J‘˙(?)[. . .)

Bilingual inscription. The Egyptian part was interpreted by Martin and

Nicholls (apud Masson 1978:86) as Jr“(3) s3 Nrskr s3 J‘˙(?)[. . . “’Jresh(a)
son of Nerseker son of Ja˙(?)-. . .”. While the correspondence of each
first name poses no problems ( Jr“(3) = arli“, as in E.Me 9), the diver-
gences between Nrskr ~ urs∞le- have been debated at length. The cor-
rect solution was formulated by Kammerzell (1993:12), who discarded
Martin-Nicholls’s analysis and argued persuasively in favour of an inter-
pretation, Jr“(3) s3 n 3rskr s3 J‘˙- ‘‘Jr“(3) son of 3rskr son of J‘˙(?)-. . .”.
This solution is far more suitable for the second name in the Carian
part, and is now the commonly accepted interpretation.

Photograph: Masson (1978, pl. VI). Drawings: Masson (1978:25, pl. XXXV, 1).
48 CHAPTER THREE

E.Me 16 (= M 8)

E.Me 16 (Masson 1978)

←
irow | p. ikra≤ ∞i
semw≤ | mno≤
mwdon≤ ∞i

No reading problems, with the exception of the initial letter of the second
word, for which Masson’s intepretation, p, seems to be the best solution.

Photograph: Masson (1978, pl. VII, 1). Drawings: Masson (1978:26, pl. XXXV, 2).

E.Me 17 (= M 9)

E.Me 17 (Masson 1978)

 THE INSCRIPTIONS 49

←
“arnai≤
upe | quq≤
bem≤ ∞i md-
aÿn

No reading problems.

Photograph: Masson (1978, pl. VII, 2). Drawings: Masson (1978:27, pl. XXXV, 3).

E.Me 18 (= M 10)

E.Me 18 (Masson 1978)

←
(a) ta“ubt≤
kuari≤b-
ar | ≤en
niqau≤
ptnupi

(b) idmuon≤
∞i | mdayn
∞i
50 CHAPTER THREE

The two parts have been inscribed by different hands. Whilst neither
section poses problems of reading, the overall structure of the text
remains obscure.

Photograph: Masson (1978, pl. VIII, 1). Drawings: Masson (1978:28, pl. XXXV, 4).

E.Me 19 (= M 11)

E.Me 19

←
pnu≤ol
zmu≤ ∞i

Curved stela (‘stèle cintrée’). No problems of reading.

Photograph: Masson (1978, pl. VIII, 2). Drawing: Masson (1978:29).

 THE INSCRIPTIONS 51

E.Me 20 (= M 12)

E.Me 20

←
uqsi | “rwli≤ ∞i | mwdon≤ ∞i

False-door stela. Unproblematic reading.

Photograph: Masson (1978, pl. IX, 1). Drawing: Masson (1978:29).

E.Me 21 (= M 13)

E.Me 21 (Masson 1978)

52 CHAPTER THREE

←
punw≤ol≤: somne≤
qÿblsi≤ ∞i

False-door stela. No reading problems.

Photograph: Masson (1978, pl. IX, 2). Drawings: Masson (1978:30, pl- XXXVI, 1).

E.Me 22 (= M 14)

E.Me 22

←
artay≤: upe: [. . .

False-door stela. In the damaged part there is sufficient space for a

complete onomastic formula.

Photograph: Masson (1978, pl. X, 1). Drawing: Masson (1978:30).

E.Me 23 (= M 15)

E.Me 23
 THE INSCRIPTIONS 53

→
ap[—]ws
a[rb]ikarm≤ ∞i

False-door stela. The integration of the second word was proposed by

Kammerzell (1993:214), who also claimed to have identified a former
inscription under the current one. Nevertheless, his reading of this pre-
vious text does not lead to any connection with the rest of the Carian
materials: (a)?p[. . .]ws # --]b[-]aubm[-].

Photograph: Masson (1978, pl. X, 2). Drawing: Masson (1978:31).

E.Me 24 (= M 16)

E.Me 24

→
tdu≤ol
kbos | “amsqi[. . .?

False-door stela. The crack in the stone makes it impossible to estab-

lish whether the last word is complete.

Photograph: Masson (1978, pl. XI, 1). Drawing: Masson (1978:31).

54 CHAPTER THREE

E.Me 25 (= M 17)

E.Me 25

←
“ayriq | parpeym≤ ∞i
yiasi

False-door stela. The order of reading adopted here is different to that

used by Masson (who began from yiasi ), and is the same as in Kammerzell
(1993:214). Contra Masson and Kammerzell, I believe that there is no
text after yiasi. It is true that the stone is damaged, but the existing
part shows evidence enough to assume that yiasi is a complete word
and that the text finishes here.

Photograph: Masson (1978, pl. XI, 2). Drawing: Masson (1978:32).

 THE INSCRIPTIONS 55

E.Me 26 (= M 18)

E.Me 26

[. . .]u≤ | upe sa | triel≤ | mrsi≤

False-door stela. Kammerzell completes the first word as [arm]u≤, but

he does not propose any argument for this integration.11 In any case,
the space preceding u≤ is small, so it is likely that only two or three
letters have disappeared.

Photograph: Masson (1978, pl. XII, 1). Drawing: Masson (1978:32).

11
The inscription appears marked with an asterisk (*) in the transliteration appen-
dix (Kammerzell 1993:214), which means that he controlled the text, but if he was
able to recognize any letter before ]u≤, one would expect it to be marked as a doubt-
ful reading, rather than using [ ].
56 CHAPTER THREE

E.Me 27 (= M 19)

E.Me 27

←
irow≤: psHÿm[-]≤
pttu≤: mno≤

False-door stela.

Photograph: Masson (1978, pl. XII, 2). Drawing: Masson (1978:33).

E.Me 28 (= M 20)

E.Me 28 (Masson 1978)

 THE INSCRIPTIONS 57

←
sanuq≤ | ue | pntmun≤ ∞i
mwdon≤ ∞i

False-door stela. Unproblematic.

Photograph: Masson (1978, pl. XIII, 1). Drawings: Masson (1978:33, pl. XXXVI 2).

E.Me 29 (= M 21)

E.Me 29

←
s[—]et≤ | [ue] | ynemori≤ | mwdon≤

False-door stela.
The second word is definitely ue (already suggested by Masson 1978).

Photograph: Masson (1978, pl. XIII, 2). Drawing: Masson (1978:34).

58 CHAPTER THREE

E.Me 30 (= M 22)

E.Me 30

←
“aru≤ol
pleq≤ ∞i: ≤ugli≤

False-door stela. Unproblematic.

Photograph: Masson (1978, pl. XIV, 1). Drawing: Masson (1978:34).

E.Me 31 (= M 23)

E.Me 31
 THE INSCRIPTIONS 59

←
wnuti≤ | kwar≤ mHm≤ ∞i | mwdon≤ [∞]i

False-door stela. The integration of the last word, already proposed by

Masson, is suitable. For the second sequence, I adopt the segmenta-
tion kwar≤ mHm≤ suggested by Schürr (apud Vittmann 2001:48, n. 40).

Photograph: Masson (1978, pl. XIV, 2). Drawing: Masson (1978:35).

E.Me 32 (= M 24)

E.Me 32

←
iturow≤ | kbjom≤ | ∞i en | mw[d]on≤ ∞i

False-door stela. Unproblematic. The integration of mw[d ]on≤ is clear.

Photograph: Masson (1978, pl. XV, 1). Drawing: Masson (1978:35).

60 CHAPTER THREE

E.Me 33 (= M 25)

E.Me 33

←
(a) idmns | myre≤ ∞i | mdayn ∞i
(b) idmns | myre≤ ∞i

False-door stela. As Masson (1978) observes, both (a) and (b)—in fact
a partial copy of (a)—seem to have been written by the same hand.

Photograph: Masson (1978, pl. XV, 2). Drawing: Masson (1978:36).

E.Me 34 (= M 26)

E.Me 34 (Masson 1978)

 THE INSCRIPTIONS 61

←
me®≤ | somne≤ | t®∞ata[r]≤

False-door stela. The integration of the last word, proposed by Masson,

is based on the occurence of the same word in complete form in E.Me
41.

Photograph: Masson (1978, pl. XVI, 1). Drawings: Masson (1978:36, pl. XXXVI, 3).

E.Me 35 (= M 27)

E.Me 35

←
ntokris | dw≤ol≤ | mwdon≤ ∞i

False-door stela. The current solution to the uncertainty regarding the

fourth letter of the first word (a form of l l or an inverted K k?, see
Masson 1978:37) is to favour the last option, which facilitates a very
good onomastic identification (ntokris = Nitokris, a well known Egyptian
female name, see Chapter 11, s. v.).

Photograph: Masson (1978, pl. XVI, 2). Drawing: Masson (1978:37).

62 CHAPTER THREE

E.Me 36 (= M 28)

E.Me 36

← wksmu≤ | wpe | lkor≤ ∞j

→ qarpsi≤

False-door stela. This order of reading was already adopted in Adiego

1993a—see also Kammerzell 1993—and differs from Masson’s (he read
qarpsis as the first word).

Photograph: Masson (1978, pl. XVII, 1). Drawing: Masson (1978:37).

E.Me 37 (= M 29)

E.Me 37
 THE INSCRIPTIONS 63

←
qlali≤ | [. . .]
tkrabi≤

False-door stela. It is not possible to calculate the extent of the letters

lost after qlali≤ (see Masson 1978:38).

Photograph: Masson (1978, pl. XVII, 2). Drawing: Masson (1978:38).

E.Me 38 (= M 30)

E.Me 38

←
“ÿin≤ | upe | arie?≤ ∞i ted

False-door stela. This reading was made by ”evoro“kin, who has con-
trolled the original. It differs from Masson’s interpretation in identifying
e (with doubts) in arie ?≤ and, particularly, in the interpretation of the
antepenultimate letter, a clear t t, not o o as indicated by Masson.

Photograph: Masson (1978, pl. XVIII, 1). Drawing: Masson (1978:39).

64 CHAPTER THREE

E.Me 39 (= M 31)

E.Me 39

←
[. . .]s? | ar∞ila≤
mno≤

False-door stela. Different reading order (cf. also Kammerzell 1993:215):

Masson’s reading began at mno≤.

Photograph: Masson (1978, pl. XVIII, 2). Drawing: Masson (1978:39).

E.Me 40 (= M 32)

E.Me 40
 THE INSCRIPTIONS 65

→
plqo | pikrm≤ ∞i | mwdon≤ ∞i

False-door stela. The sign l of the second word is actually a k inverted

(= /k/) (already noted in Adiego 1993a).

Photograph: Masson (1978, pl. XIX, 1). Drawing: Masson (1978:40).

E.Me 41 (= M 33)

E.Me 41

→
|? or≤ | wpe | qdar®ou≤ | t®∞atar≤

False-door stela. The initial vertical small stroke and the brevity of the
first name, or≤, perplexed Masson, who thought that, ‘pour des raisons
obscures, le lapicide n’ait jamais gravé le début du premier mot.’ In
fact, it seems to me most likely that the stoke is an accidental and
intrusive mark. In any case, contrary to Masson, I see no problem in
accepting the existence of a name or-≤ in Carian.

Photograph: Masson (1978, pl. XIX, 2). Drawing: Masson (1978:40).

66 CHAPTER THREE

E.Me 42 (= M 34)

E.Me 42 (Masson 1978)

←
(1) arjom≤: ue: mwsat≤: ∞i: mwdon≤: ∞i
(2) tbridbd≤: ∞i

False-door stela. Very clear reading (‘belle stèle intacte’, Masson 1978:41).

Photograph: Masson (1978, pl. XX, 1). Drawings: Masson (1978:41, pl. XXXVI, 4).

E.Me 43 (= M 35)

E.Me 43
 THE INSCRIPTIONS 67

←
(a) lÿ∞si≤ | upe | “rquq≤ ∞i | ksolb≤
(b) arliom≤ | mno≤ ∞i

False-door stela. Different reading order (already in Adiego 1993a,

Kammerzell 1993): Masson read firstly (b), then (a). The reading of the
initial letter of the third word of (a) is also different: f “ instead of
Masson’s a “.

Photograph: Masson (1978, pl. XX, 2). Drawing: Masson (1978:41).

E.Me 44 (= M 36)

E.Me 44 (Masson 1978)

←
(a) apmen “rquq≤ kojol ∞i
(b) mwton≤ ∞i

False-door stela. Reading order different from Masson’s (already in

Adiego 1993a, Kammerzell 1993): f “ instead of a a at the beginning
of the second word of (a).

Photograph: Masson (1978, pl. XX,I 1). Drawings: Masson (1978:42, pl. XXXVII, 1).
68 CHAPTER THREE

E.Me 45 (= M 37)

E.Me 45

→
[q?]lalis
[?]iam≤ ∞i
alos ∞arnos

False-door stela. The integration of the initial letter of the first word
is already in Kammerzell (1993), from the parallel form qlali≤ in E.Me
37.

Photograph: Masson (1978, pl. XXI, 2). Drawing: Masson (1978:43).

E.Me 46 (= M 38)

E.Me 46
 THE INSCRIPTIONS 69

←
(a) ÿasd≤ | yi≤{∞}biks≤ ∞i
(b) mwdon≤ ∞i

False-door stela. Different reading order (Adiego 1993a, Kammerzell

1993). Masson’s reading began with (b).
Masson (1978) reads a sign X ∞ between ≤ and b in the second word
of (a). A careful observation of the photography shows that this alleged
sign appears placed between ≤ and b without keeping distance with
them, unlike the other letters of the inscription. As Schürr (1996b:151)
aleady pointed out, this ∞ seems to be ‘eine Illusion’, and I have there-
fore decided to discard it.
As for the final sequence of this word, I maintain that the reading
generally accepted until now, yi≤biks∞i≤ (see Masson’s drawing in the
bottom part of the illustration), must be replaced by a more natural
ÿi≤biks≤ ∞i (cf. ÿ≤biks in C.xx 2): in fact, the three later letters can be
read beginning at ≤.

Photographs: Masson (1978, pl. XXII, 1, 2). Drawing: Masson (1978:43).

E.Me 47 (= M 39)

E.Me 47 (Masson 1978)

←
tqtes | paraibrel≤ ∞i | mn[o-?]

With this inscription, a series of texts on different objects and frag-

ments begins (E.Me 47–64). In this case, it is a limestone plaque whose
use is unclear (Masson 1978:43).

Photograph: Masson (1978, pl. XXIII, 1). Drawings: Masson (1978:43, pl. XXXVII, 2).
70 CHAPTER THREE

E.Me 48 (= M 40)

E.Me 48 (Masson 1978)

←
[—] j[-]≤
[-]owt≤
∞i: msn-
ord≤

Note our reading of d d instead of i i in the final word, cf. above,

E.Me 3.

Photograph: Masson (1978, pl. XXIII, 2). Drawings: Masson (1978:44, pl. XXXVII, 3).

E.Me 49 (= M 41)

E.Me 49 (Masson 1978)

 THE INSCRIPTIONS 71

→ loubaw | si-
← ral | pnld≤wl

Possibly an oblong stela that has been reused.

A very strange inscription, without clear connections with the rest of
the Memphis corpus. Moreover, the reading is very difficult and there-
fore not absolutely certain.

Photograph: Masson (1978, pl. XXIII, 3). Drawing: Masson (1978:45).

E.Me 50 (= M 42)

E.Me 50 (Masson 1978)

←
(a) “enurt
(b) p∞simt≤ ∞i

Also likely to be a reused oblong stela. No problems of reading.

Drawings: Masson (1978:46, pl. XXXVI, 4).

E.Me 51 (= M 43)

E.Me 51 (Masson 1978)

72 CHAPTER THREE

←
arli“≤ | psikro≤ [?
ue

Upper fragment of a curved stela (‘stèle cintrée’). The reading offered

here is new: in my opinion, the last sign of the second word is clearly
z /≤/, and not N /n/ (see plate xxiv in Masson 1978). As for ue,
which appears just below arli“≤, it is possible that it must be read imme-
diately after this latter word (arli““´ue | psikro≤ ), which would represent
a more logical structure (ue, upe, etc. always appear in a second position).

Photograph: Masson (1978, pl. XXIV, 1). Drawings: Masson (1978:46, XXXVII, 5).

E.Me 52 (= M 44)

E.Me 52

←
[. . .] ardybyr≤ | md[. . .]

Upper fragment of a false-door stela.

Photograph: Masson (1978, pl. XXIV, 2). Drawing: Masson (1978:46).

 THE INSCRIPTIONS 73

E.Me 53 (= M 45)

E.Me 53

←
[. . .]q≤si≤

Fragment of a false-door stela.

Photograph: Masson (1978, pl. XXIV, 3). Drawing: Masson (1978:47).

E.Me 54 (= M 45a)

E.Me 54

←
[. . .] mrsj[. . .]

Fragment of a false-door stela.

Photograph: Masson (1978, pl. XXV, 1). Drawing: Masson (1978:47).

74 CHAPTER THREE

E.Me 55 (= M 46)

E.Me 55

←
[. . .] psma[≤/“k . . .]

The integration proposed here is based on the identification of the

beginning of the typical Egyptian name, Psammetichus. Given that this
name can appear in Carian spelled either with “ or ≤, I choose not to
discard either possibility.

Photograph: Masson (1978, pl. XXV, 2). Drawing: Masson (1978:47).

E.Me 56 (= M 47)

E.Me 56

←
[. . .] “ark[bi/jom . . .?]

Fragment of a false-door stela.

I follow the suggestion of Meier-Brügger (1979b) and Ray (1982b:
189) of reading f “ instead of i i (Masson 1978) at the beginning.
 THE INSCRIPTIONS 75

The integration is based on the form “arkbiom, which appears in other

inscriptions.

Photograph: Masson (1978, pl. XXV, 3). Drawing: Masson (1978:48).

E.Me 57 (= M 47a)

E.Me 57

←
[. . .]i≤ ∞i

Fragment of a false-door stela. Clear ending in genitive + particle ∞i.

Photograph: Masson (1978, pl. XXVI, 1). Drawing: Masson (1978:48).

E.Me 58 (= M 47b)

E.Me 58

←
[. . .]s≤ ∞i

Undetermined fragment. A typical ending in genitive + particle ∞i, as

in E.Me 57.

Photograph: Masson (1978, pl. XXVI, 2). Drawing: Masson (1978:48).

76 CHAPTER THREE

E.Me 59 (= M 48)

E.Me 59

←
[. . .]utr[. . .]

Fragment of a false-door stela.

Photograph: Masson (1978, pl. XXVI, 3). Drawing: Masson (1978:48).

E.Me 60 (= M 48a)

E.Me 60

→ (?)
[. . .]∞≤

Fragment of a false-door stela.

Photograph: Masson (1978, pl. XXVI, 1). Drawing: Masson (1978:48).

 THE INSCRIPTIONS 77

E.Me 61 (= M 48b)

E.Me 61 (Masson 1978)

[. . .]i

Undetermined fragment.

Drawing: Masson (1978:48).

E.Me 62 (= M 48c)

E.Me 62 (Masson 1978)

[. . .]≤[. . .]

Undetermined fragment.

Drawing: Masson (1978:49).

E.Me 63 (= M 48d)

E.Me 63

←
(a) idyes≤
(b) m [?
78 CHAPTER THREE

Fragment of a false-door stela. Although Masson indicates the presence

of a M m, he does not transcribe it in his edition of the text. I follow
Kammerzell (1993) in including the letter in the transcription.

Photograph: Masson (1978, pl. XXVI, 3). Drawing: Masson (1978:49).

E.Me 64 (= M 49)

E.Me 64 (Masson 1978)

←
(a) [. . . u?]p. e : pd[
(b) [. . .]mi [. . .]

Fragment of a curved stela. I concur with Kammerzell (1993:217) in

the interpretation of the first word.

Drawing: Masson (1978:49).

E.Me 65 (= Stela of Abusir)

E.Me 65 (Schürr 2003)

←
u[. . .]m | punm[-]≤ | mudo[n]≤

See Schürr (2003) for a new attempt at reading this extremely difficult
inscription. I adopt his reading (with the integration of [n] in mudo[-]≤)
and reproduce his drawing (Schürr 2003:94). Former references: Masson
(1978:91), who was the first to confirm the Carian character of the
 THE INSCRIPTIONS 79

inscription, already suggested by L. H. Jeffery: Kammerzell (1993:

138–139).

Photograph: Masson (1978, pl. XXX). Drawings: Masson (1978:91), Kammerzell

(1993:138), Schürr (2003:94).

E.Me 66 (= Kammerzell *180)

—].[. . u][. . .]p[-]n[—

Concerning this inscription on a fragment of stela, conserved in the

British Museum, see Kammerzell (1993:144, n. 101). To my knowl-
edge, neither the photography nor the drawing have yet been pub-
lished, so I limit myself to follow Kammerzell’s reading.

3. Abydos (E.Ab)
The Carian inscriptions from Abydos were found and copied for the
first time by Archibald H. Sayce, who published them in his most
important work on Carian (Sayce 1887[1892]). He discovered them in
the temples of Ramses II and, more significantly, Seti I.
Regrettably, there is not a definitive edition of the Abydos graffiti.
Jean Yoyotte seemed to have revised these graffiti and even to have
found some others in 1955–1956, but his work has never been pub-
lished. In his index of Carian words, Masson (1978) pointed out that
he had adopted, “insofar as it was possible”, Yoyotte’s revision, and
added the most important unpublished graffiti found by Yoyotte. Meier-
Brügger (1979b) followed Masson in his collection of Carian inscrip-
tions in transcription. ”evoro“kin also (1965) used Yoyotte’s new readings.
My attempts to obtain more information on this sub-corpus from Jean
Yoyotte were unsuccessful.
As a result, our present knowledge of the Carian graffiti of Abydos
is unsatisfactory. Neither Masson (1978) nor Meier-Brügger (1979b) offer
any additional information on the readings adopted for each graffiti,
so that in cases where Masson’s readings are not coincident with those
made by Sayce (Sayce 1887[92] followed by Friedrich 1932), it is impos-
sible to ascertain whether this is due to the revisions of Yoyotte, or it
is simply a new interpretation of Sayce’s copies based on mere divina-
tio. However, thanks to ”evoro“kin (1965) and to some notices scattered
throughout the works of Masson, it is possible to establish in a great
number of cases that the text is in fact the result of Yoyotte’s collation.
80 CHAPTER THREE

As for the unpublished graffiti, Masson and Meier-Brügger take into

account, as already established, only the most important examples: they
offer only 8 from a total of at least 34. Moreover, given that both
authors limit themselves to transcribing the texts, nothing can be said
about the graphic traits of these inscriptions. The only exception is
E.Ab 43 (= Ab. 34 Y)—in fact an already known graffito—published
in Murray (1904: 10 and pl. XII) and included in ”evoro“kin (1965)
(= 24 ”), though absent in the editions of Sayce and Friedrich.
Confusion with the Abydos graffiti is even augmented by the fact
that Bork (1930) re-ordered Sayce’s list of inscriptions, putting together
under the same reference number those inscriptions that he believed
had the same content. This decision was not only disputable from an
epigraphical point of view, but was in fact revealed as altogether erro-
neous, since in some cases Bork grouped together texts which were not
strictly identical. This explains the posterior use of redundant reference
numbers like “2a”, “2b” in works by Friedrich or Masson. In the new
number system introduced here, I prefer to assign each inscription a
different number, even in the case of inscriptions that seem to have
the same content.
Given the lack of an up-to-date edition, I adopt in general terms
Masson’s readings, but in some cases I also include new attempts at
improving these readings, mainly from Schürr (in various papers) and
Vittmann (2001). Since the Abydos graffiti pose such problems, in the
following sections, I will comment briefly upon each inscription and
will explain the reading adopted here. For want of a better solution, I
shall limit myself to reproducing Sayce’s drawings,12 although in some
cases the present readings may not coincide.

E.Ab 1 (= Ab. 1 F = Sayce 1887[92] 1)

E.Ab 1 (Friedrich 1932)

12
For technical reasons, the source of my illustrations will be Friedrich (1932), where
Sayce’s drawings are reproduced.
 THE INSCRIPTIONS 81

←
pisiri

Revised by Yoyotte (Masson 1974:131). There are no reading problems.

E.Ab 2 (= Ab. 2a F = Sayce 1887[92] 2)

E.Ab 2 (Friedrich 1932)

←
panejt iarja≤

Expressly mentioned as revised in Masson-Yoyotte (1956:13, 63). Masson’s

reading is adopted, but with different segmentation (the first name is
clearly an Egyptian one whose final element is -nejt).

E.Ab 3 (= Ab. 2b F = Sayce 1887[92] 3)

E.Ab 3 (Friedrich 1932)

→
ptn“e | ibarsi≤

Revised by Yoyotte, according to Masson-Yoyotte (1956:63). For the

second name, I follow Schürr’s proposal of reading ibarsi≤, not ‡irarsi≤
(Masson).

E.Ab 4 (= Ab. 3b F = Sayce 1887[92] 24)

E.Ab 4 (Friedrich 1932)

82 CHAPTER THREE

←
“amow ltari≤

Revised by Yoyotte (cf. Masson-Yoyotte 1956:39). Masson’s reading is

adopted, but with different segmentation (Masson’s reading would indi-
cate ‡“amowl tari≤.

E.Ab 5 (= Ab. 3c F = Sayce 1887[92] 25)

E.Ab 5 (Friedrich 1932)

←
“amow ltari[≤]

Revised by Yoyotte (cf. Masson-Yoyotte 1956:39).

Masson groups Ab. 3b F and Ab. 3c F under a single epigraph, “Ab
3bc F”. This procedure, debatable even if the content were exactly the
same, is inappropriate here, as the second word appears complete in
the case of E.Ab 4 = Ab 3b F, but incomplete in E.Ab 5 = Ab. 3c
F (ltari[≤]). On segmentation, see above E.Ab 4.

E.Ab 6 (= Ab. 4 F = Sayce 1887[92] 5)

E.Ab 6 (Friedrich 1932)

←
“aru≤ol | ÿrsbe | pdubi≤

Revised by Yoyotte (cf. Masson 1978:34).

 THE INSCRIPTIONS 83

E.Ab 7 (= Ab 5a F = Sayce 1887[92] 6)

E.Ab 7 (Friedrich 1932)

←
plat | pals≤

E.Ab 8 (= Ab 5b F = Sayce 1887[92] 9)

E.Ab 8 (Friedrich 1932)

←
plat pals≤

E.Ab 9 (= Ab 5c F = Sayce 1887[92] 10)

E.Ab 9 (Friedrich 1932)

←
plat pals≤

E.Ab 7, 8 and 9 do not pose reading problems. The three inscriptions

are grouped together by Masson.
84 CHAPTER THREE

E.Ab 10 = (Ab. 6 F = Sayce 1887[92] 7)

E.Ab 10 (Friedrich 1932)

←
piubez
qurbo≤

Revised by Yoyotte (Masson-Yoyotte 1956:9), this inscription formed

by two personal names still raises serious reading problems. Masson’s
reading (1978) distances the first name from the very similar example
found in E.Ab 15. Schürr’s reading seems preferable (Schürr 2000:172,
n. 7), because it allows us to connect this name with the second name
in E.Me 1 and with the name mentioned in E.Ab 15 (although we
cannot establish an absolutely certain link). Regarding the second name,
although in Masson-Yoyotte (1956:9) the reading ‡≤urbo≤ is preferred
(thus also ”evoro“kin 1965), in Masson (1978), Sayce’s reading qurbo≤
is restored.

E.Ab 11 (= excluded by Masson [Ab 7 F = Sayce 1887[92] 8])

E.Ab 11 (Friedrich 1932)

←
≤?
[. . .]it

Since the letters are apparently Carian, I reintroduce this very frag-
mentary inscription to the corpus, even though it is of no use to us.
 THE INSCRIPTIONS 85

E.Ab 12 (= Ab. 8a F = Sayce 1887[92] 11)

E.Ab 12 (Friedrich 1932)

←
untri uantrpo

E.Ab 13 (= Ab. 8b F = Sayce 1887[92] 20)

E.Ab 13 (Friedrich 1932)

←
untri | uantrpu≤

E.Ab 14 = Ab. 9 F (= Sayce 1887[92] 12)

E.Ab 14 (Friedrich 1932)

←
abrq∞[. . .?

Regarding E.Ab 12, 13 and 14, I am unsure as to whether Masson’s

readings are the result of a revision of the texts. Lacking a better read-
ing, I adopt Masson’s.

E.Ab 15 = Ab. 10 F (= Sayce 1887[92] 13)

E.Ab 15 (Friedrich 1932)

86 CHAPTER THREE

←
pdubez or≤

Revised by Yoyotte (Masson-Yoyotte 1956:13). The reading in Masson

(1978) is not the same as in Masson-Yoyotte (1956:13): ‡pdubtzor≤ vs.
‡pdubtzt r≤ (with r≤ as a part of a different graffito!) respectively. I adopt
the reading proposed by Schürr and Vittmann (see Vittmann 2001:42).

E.Ab 16 = Ab. 11 F (= Sayce 1887[92] 14)

E.Ab 16 (Friedrich 1932)

←
nprosn≤

Revised? In any case, it does not pose reading problems.

E.Ab 17 = Ab. 12 F (= Sayce 1887[92] 15)

E.Ab 17 (Friedrich 1932)

←
pa[-]in[-]t≤

Revised? I follow Masson (1978)

E.Ab 18 = Ab. 13a F (= Sayce 1887[92] 16)

E.Ab 18 (Friedrich 1932)

←
tamosi | inut≤
 THE INSCRIPTIONS 87

E.Ab 19 = Ab. 13b F (= Sayce 1887[92] 17)

E.Ab 19 (Friedrich 1932)

←
tamosi utnu≤

E.Ab 20 = Ab. 14 F (= Sayce 1887[92] 18)

E.Ab 20 (Friedrich 1932)

←
ninut | tamosi≤

E.Ab 21 = Excluded by Masson [Ab. 15 F (= Sayce 1887[92] 19)]

E.Ab 21 (Friedrich 1932)

←
to[-]a[—] l
tamosi u?tnu≤?

These four graffiti seem to coincide in that they contain the same per-
sonal name. Rather surprisingly, in Masson’s list, this name is read as
trmosi- in E.Ab 18, E.Ab 19, and as tamosi- in E.Ab 20, while E.Ab 21
is excluded. These readings were adopted in Adiego (1993a). It is quite
possible—although I cannot confirm it—that this discrepancy has arisen
because only E.Ab 20 was revised by Yoyotte.13

13
Vittmann (2001:43) mentions Yoyotte’s collation of E.Ab 20.
88 CHAPTER THREE

Vittmann (2001:43) has correctly noted this incongruity, and has

offered a convincing reading of tamosi- everywhere, a view that is sup-
ported here.14 More problematic, and still unresolved, are the respec-
tive readings of the names that accompany tamosi- in the four inscriptions.
They look very similar, but it is not easy to imagine a complete iden-
tity for all the forms. In the case of E.Ab 20, I continue to work with
”evoro“kin’s suggestion of reading ninut instead of ‡∞inut. For the other
names found in the remaining graffiti, Masson’s readings are adopted.
I include Ab 15 F, excluded by Masson for reasons that remain unknown
for me.

E.Ab 22 = Excluded by Masson [Ab. 16 F (= Sayce 1887[92] 21])

E.Ab 22 (Friedrich 1932)

←
[-]untlau[-]|

This brief graffito is very difficult to read, but it seems to show true
Carian letters. For this reason, I include it in the list (as ”evoro“kin
did: ”evoro“kin (1965), 34 ”).

E.Ab 23 = Ab. 17 F (= Sayce 1887[92] 22)

E.Ab 23 (Friedrich 1932)

←
be≤ol

14
A theory already envisaged by Ray, see Ray (1994:205).
 THE INSCRIPTIONS 89

E.Ab 24 = Ab. 18 F (= Sayce 1887[92] 23)

E.Ab 24 (Friedrich 1932)

←
[. . .] arli“

Revised by Yoyotte (cf. Masson 1978:51). This revision identifies the

well-known Carian name arli“ (contrary to Sayce’s reading, ‡araii ).

E.Ab 25 = Ab. 19 F (= Sayce 1887[92] 26)

E.Ab 25 (Friedrich 1932)

←
ttubazi kattÿri≤

Revised by Yoyotte (Masson-Yoyotte 1956:4, 33). According to Masson-

Yoyotte (1956:33), the sign for t here adopts the form C, as in E.Me
7. The segmentation adopted here differs from that of Masson and is
based on Schürr’s works.

E.Ab 26 = Ab. 20 F (= Sayce 1887[92] 27)

E.Ab 26 (Friedrich 1932)

←
[. . .]pri | ptnuq?i?

Masson (1978:28) points out that this graffito was not rediscovered by
Yoyotte in 1956. Note that in Adiego (1993a) the end of the (incom-
plete) first word was transcribed incorrectly as †]pre, instead of ]pri (a
90 CHAPTER THREE

mistake noted by Vittmann 2001:44). Here I adopt one of the two

interesting corrections proposed by Vittmann for the second word
(Vittmann 2001:44).15

E.Ab 27 = Ab. 21 F (= Sayce 1887[92] 28)

E.Ab 27 (Friedrich 1932)

←
yysmt≤oHa[

Divergences between Sayce’s and Masson’s readings lead me to think

that the graffito was revised by Yoyotte, but I cannot confirm this. I
follow Masson’s reading. Note that the drawing is based on Sayce/
Friedrich editions.

E.Ab 28 = Ab. 22 F (= Sayce 1887[92] 29)

E.Ab 28 (Friedrich 1932)

←
Hosurz | srton[-]t[. . .?]
(or: → . . . +t[-]nota/rs | za/rusoH/l? Schürr)

In Masson-Yoyotte (1956:68) this graffito is referred to as “not redis-

covered”. Indeed, the reading adopted by Masson seems to be based
on Sayce’s drawing. I maintain Masson’s reading, but I also consider
as a plausible alternative the suggestions made in Schürr (1996a:65)
that the text could be read in the opposite direction and that the let-
ter H could in fact be L = l (a possibility to be ruled out if Masson’s
reading is upheld, since it is very unlikely that Carian L l would appear
at the beginning of a word).

15
The other possibility suggested by Vittmann is ptnuti. Both readings allow us to
identify good correspondences with Egyptian personal names (see Chapter 11, s. v. ptnuq?i ).
 THE INSCRIPTIONS 91

E.Ab 29 = Ab. 24 F (= Sayce 1887[92] 31)

E.Ab 29 (Friedrich 1932)

←
[. . .]r[--]tnit

Graffito consisting of only five signs. The readings given by Sayce and
Masson coincide.

E.Ab 30 = Ab. 25 F (= Sayce 1874 5+6 = Sayce 1887[92] 32)

E.Ab 30 (Friedrich 1932)

←
bid≤lemsa: “a[ru]≤ol: “aÿdiq≤
[. . .]allia: bsis

I am unaware of whether it has been revised. In any case, Sayce’s

reading is generally followed by Masson, who only improves the read-
ing of the second word in order to obtain a well-known Carian name
(fa[ru]zoL = “ar[u≤]ol).

E.Ab 31 = Ab. 26a F (= Sayce 1887[92] 33)

E.Ab 31 (Friedrich 1932)

←
∞aye
92 CHAPTER THREE

E.Ab 32 = Ab. 26b F (= Sayce 1887[92] 34)

E.Ab 32 (Friedrich 1932)

←
∞arr≤

E.Ab 33 = Ab. 27 F (= Sayce 1887[92] 35)

E.Ab 33 (Friedrich 1932)

←
“arpt≤ | p[-]lu≤
For E.Ab 31–33 I follow Masson’s readings.

E.Ab 34 = Excluded by Masson [= Ab. 28 F (= Sayce 1887[92] 36)

E.Ab 34 (Friedrich 1932)

←
dbkrm [-]kb?[

Despite the difficulties of interpretation, Schürr has been able to iden-

tify a good Carian sequence, dbkrmº (Schürr 1996b:154, n. 8; see Chapter
11, s. v.). For this reason I have decided to reinstate this graffito to the
Abydos sub-corpus.
 THE INSCRIPTIONS 93

E.Ab 35 = Ab. 29 F (= Sayce 1874 4 = 1887[92] 37)

E.Ab 35 (Friedrich 1932)

←
u≤ol | mi∞≤≤ kdu.usi≤

For this graffito, I adopt the reading proposed by Schürr (2001b:108,

111), apparently based on the collations of Yoyotte and ”evoro“kin.
E.Ab 36—E. Ab 43: “inédits de Yoyotte”, unpublished new graffiti
found by Yoyotte in 1955–1956. I include only those transcribed by
Masson (1978). As mentioned above, E.Ab 43 = Ab. 34 Y is not strictly
an unpublished graffito: a drawing of it appeared in Murray (1904),
and it is also included in ”evoro“kin (1965) (= 24 ”).

E.Ab 36 = Ab. 8 Y
[-]ars, ∞[-]urb≤

E.Ab 37 = Ab. 9 Y
“arur≤

E.Ab 38 = Ab. 15 Y
piew

E.Ab 39 = Ab. 26 Y
uarila[-]os≤

E.Ab 40 = Ab. 27 Y
ialli | q∞blio≤

E.Ab 41 = Ab. 28 Y
ttbazi kt?tri≤
I follow Schürr’s reading (see Schürr 1996a:60).

E.Ab 42 = Ab. 29 Y
“aru≤[. . ?
94 CHAPTER THREE

E.Ab 43 = Ab 34 Y = Murray (1904)

E.Ab 43 (Murray 1904)

?-ras
Although Murray’s drawing points clearly to an initial k k (by exten-
sion, kras), Masson’s reading leaves the sign unread.

Graffiti excluded from our collection:

Ab. 3a F (= Sayce 1887[92] 4), also excluded by Masson.16 It seems
to begin with “am[, which explains how Bork could link it to E.Ab 4
and E.Ab 5.
Ab. 23 F (= Sayce 1887[92] 30), also excluded by Masson and by
”evoro“kin. Apparently it has not been revised since Sayce (Masson-
Yoyotte 1956:46: ‘non retrouvé par J. Yoyotte; Masson 1974:131 n. 43:
“non revisé”). The graffito could possibly be Greek.
Ab. 30 F (= Sayce 1887[92] 38), also excluded by Masson. It is an
almost illegible graffito, which cannot even be definitely identified as
Carian.
The drawings made by Sayce are reproduced below:

Ab 3a F (Friedrich 1932)

Ab 23 F (Friedrich 1932)

16
”evoro“kin (1965:313–314) unifies Ab 3a, b, c under a sole entry (20 ”).
 THE INSCRIPTIONS 95

Ab 30 F (Friedrich 1932)

4. Thebes, Tomb of Montuemhat (E.Th)

The case of the corpus of Carian inscriptions found in Thebes (tomb
of Montuemhat) is even worse than that of Abydos: the corpus has
never been published correctly. Leclant, the first to find fifteen graffiti
engraved on the walls of the open court of the tomb, never produced
an edition of them, instead publishing only the photographs of two
graffiti (E.Th 13 and E.Th 14, Leclant (1951, tab. LXIV). ”evoro“kin
included in his book on Carian (”evoro“kin 1965) a merely provisional
edition, with drawings of these graffiti. This edition forms the basis of
Masson’s transcriptions (in Masson 1978, reproduced in Meier-Brügger
1979b), although Masson aknowledges that these were improved thanks
to the photographs and notes of Leclant that he had at his disposal.
On the other hand, some years later, ”evoro“kin himself revised Leclant’s
graffiti and discovered some new examples, although these have not
been published either, and the new graffiti are circulated only in pri-
vate copies. In the Carian conference held in Rome in 1993, Masson
announced a new enterprise in order to adequately publish the Theban
corpus (cf. Masson 1994a), but I do not know if this enterprise is in
fact underway.
More recently, Diether Schürr, who was able to gain access to the
entire dossier of drawings, has drawn up a provisional corpus of Theban
graffiti in transcription, which he has kindly made known to me. This
corpus consists of four parts: (A) inscriptions found in the entrance of
the temple. I already knew of the drawings of these texts thanks to a
private copy made by ”evoro“kin and circulated among scholars; (B)
inscriptions found at the end of the corridor; (C) graffiti located in the
vestibule, the drawings of which I have never seen, and from which I
knew only certain texts, occasionally cited by ”evoro“kin in recent arti-
cles; and (D) graffiti from the open court, in fact already mentioned
in Leclant’s corpus (with some additions).
I choose to adopt Masson’s readings for (D) (Leclant’s graffiti), intro-
ducing some corrections suggested by ”evoro“kin in published or unpub-
lished works, and I include the drawings published in ”evoro“kin (1965).
For (A), I provide the transcription, following Schürr and controlling
”evoro“kin’s drawings, but I refrain from reproducing the drawings,
96 CHAPTER THREE

since they are as yet unpublished; my aim is merely to contribute to

their use in linguistic research, not to actually produce an epigraphi-
cal edition, which must be made by ”evoro“kin himself. For (B) and
(C), I limit myself to reproducing Schürr’s transcriptions, adapting them
to the decipherment system followed in this study. According to stan-
dard practice within the Carian corpus, I begin my numeration of the
inscriptions with the first to be published (Schürr D), which are then
followed by the remaining inscriptions (A, B, C).

E.Th 1 = 47 ” (= D 1)

E.Th 1 (”evoro“kin 1965)

←
uarbe

E.Th 2 = 48 ” (= D 3)

E.Th 2 (”evoro“kin 1965)

←
dtÿbr | kbokt≤
k≤atÿbr

E.Th 3 = 49 ” (= D 2)

E.Th 3 (”evoro“kin 1965)

pla?t
 THE INSCRIPTIONS 97

E.Th 4 = 50 ” (= D 5)

E.Th 4 (”evoro“kin 1965)

→ dokmmpint seqqejewsk | mqtjq

← ÿpdnmwd

The second line was usually read in the opposite direction. The new
reading was suggested by Schürr. I am somewhat sceptical about the
last sign, now read & d, a letter rarely found in Thebes (in fact, it only
appears in one other graffito, still unpublished and whose drawing I
have not seen; E. Th 14, see below). Former readings pointed rather
to B b (cf. ”evoro“kin’s drawing here reproduced). The reading of the
letters preceding qt jq at the end of the first line probably comes from
a new collation of the graffito, the old drawing showing a lacuna and
some illegible signs.

E.Th 5 = 51 ” (= D 6)

E.Th 5 (”evoro“kin 1965)

←
dÿbr | t®∞atr≤
The second word, which posed serious problems in ”evoro“kin (1965),
was re-read in a new collation by ”evoro“kin as t®∞atr≤, undoubtedly
using the model of the form t®∞atar≤ that appears twice in Saqqâra
(”evoro“kin, “Corrections to Existing Copies”, ms.). Schürr, however,
does not rule out a reading of l l instead of ®.
98 CHAPTER THREE

E.Th 6 = 52 ” (= D 7)

E.Th 6 (”evoro“kin 1965)

→
bebnd

The last letter, d, does not appear in the old drawing by ”evoro“kin
and was thus absent in older collections of Carian inscriptions (for
instance, Adiego 1993a, where the reading bebn was given).

E.Th 7 = 53 ” (= D 10)

E.Th 7 (”evoro“kin 1965)

→
wljat

The doubts about the second and fourth letters, reflected in the draw-
ing, can be dismissed due to the clear onomastic connection of the
word (→ wljat).

E.Th 8 = 54 ” (= D 11)

E.Th 8 (”evoro“kin 1965)

→
qutbe
 THE INSCRIPTIONS 99

E.Th 9 = 55 ” (= D12)

E.Th 9 (”evoro“kin 1965)

→
kudtubr
The initial letter is now read as k k.

E.Th 10 = 56 ” (= D 8)

E.Th 10 (”evoro“kin 1965)

→
a?q≤baq ewm ≤emot qtblo owdown[. . .]mwarudk≤o mlane

The most difficult letter to determine is the first one: read k by ”evoro“kin
(1965) and in the Masson/Meier-Brügger collections. In Adiego (1993a),
a reading w was introduced, based on ”evoro“kin’s observations. Schürr
now reads it as a?. For the rest of the letters, the present reading differs
very little from that offered in Adiego (1993a). Only a letter b imme-
diately after owdown is not read by Schürr.

E.Th 11 = 57 ” + 58 ” (= D 12)

(”evoro“kin 1965, 57 ”)
100 CHAPTER THREE

E.Th 11 (”evoro“kin 1965, 58 ”)

→
psma≤[k] [?
| nm[
mplat | o[

Schürr’s dossier puts these two lines together, formerly interpreted as

different graffiti. The clearest element is the presence of the name
psma≤k.

E.Th 12 = 59 ” (= D 13)

E.Th 12 (”evoro“kin 1965)

→
?-˚bjqmq ewmlane qeb≤t | u[. . .]ü≤q | qwsal | mqabaewleqo“oski. oms

This reading coincides almost exactly with that presented in Adiego

(1993a), where some corrections from the work of ”evoro“kin had
already been introduced. The sole difference introduced by Schürr’s
reading involves the final sequence kioms, formerly read as ‡lkoms.
 THE INSCRIPTIONS 101

E.Th 13 = 60 ” (= D 14)

E.Th 13 (”evoro“kin 1965)

→ dbiks | kbjoms | wdwn | sb a≤b≤t

← ewm

The reading corresponds precisely to that offered in Adiego (1993a).

Note that ”evoro“kin’s drawing shows sbb? instead of sba in the fourth
word. This latter, corrected reading was introduced by ”evoro“kin him-
self after revising the inscription (“Corrections to Existing Copies”, ms.),
and was already reflected in Adiego (1993a).

E.Th 14 (= D 4)
]q[. . .]btdeo

This graffito from the open court was absent from the corpus of pub-
lished inscriptions. The presence of d is surprising, as in the new read-
ing of E.Th 4 (see above). Schürr observes that a vertical p p appears
under the graffito.

E.Th 15 (= D 15)
(Very uncertain reading)

Also a new graffito. Schürr mentions two very different alternative read-
ings, suggested by Kayser and ”evoro“kin respectively (see ”evoro“kin
1994:145).17

E.Th 16 (= A 1)
∂saml-?-?-o (vacat) dy “a

Three unclear signs found under the end of A 1.

17
The readings are, respectively, [1] u?∞a pn-?-ek . . . i / plsiwbms and [2] ub∞l pn-?-
ekleai/pksiwrm≤.
102 CHAPTER THREE

E.Th 17 (= A 2)
ku

E.Th 18 (= A 3)
t n

Very uncertain reading.

E.Th 19 (= A 4)
dbikrm

k presents an inverted form l, apparently l, but the identification is

based on the good onomastic correspondence of the reading dbikrm (see
Chapter 11, s. v.).

E.Th 20 (= A 5)
orbá ˚ r i“

E.Th 21 (= A 6)
mmn∞al

An inverse reading la∞mmn is also possible.

E.Th 22 (= A 7)
mwk | te

E.Th 23 (= A 8)
bebi

E.Th 24 (= A 9)
kow[?-?]

The drawing made by ”evoro“kin available to me points to koweq.

E.Th 25 (= A 10)
ktmno

Inverted k (l). The interpretation as k, not l, is supported by the

good onomastic connection of ktmno (see Chapter 11, s. v.).

E.Th 26 (= A 11)
brsi yri≤
 THE INSCRIPTIONS 103

E.Th 27 (= A 12)
pnw≤ol | mlqi≤

E.Th 28 (= A 13)
bejeym | teboot
K
bebi. nt ken

The strange ‘diamond-sign’ also appears in E.Si 3, preceding precisely

the same word bebint, but it is not present in the other examples of this
word (E.Th 30, E.AS 7), which makes it difficult to consider it an actual
letter. The exact function therefore remains unexplained (see pp. 253–254).

E.Th 29 (= A 14)
]ke

Some uncertain signs follow.

E.Th 30 (= A 15)
bebint | psrkrte | mumn“tnse-?
»ßwk˚n

E.Th 31 (= A 16)
(Very uncertain reading)

“Impossible to describe” (Schürr). Allegedly a sign with the form s

appears twice, and a sign % appears once, both of which are alien to
the standard inventory of Carian letters (for this latter sign see also
E.Th 34 below).

E.Th 32 (= A 17)
tqlow

A reading in the opposite direction is also possible (wolqt)

E.Th 33 (= A 18)
∞lbiks≤

The second letter could also be a k (”evoro“kin’s reading).

E.Th 34 (= A 19)
sl∞maewm | urt | kwri≤ | prna∞non | dm-?-n | maãtnor | qanor | uro
104 CHAPTER THREE

The uncertain letter in the fifth word seems to be similar to %, but

”evoro“kin’s drawing in fact displays a sort of % form. Schürr ques-
tions whether it could merely be o o.

E.Th 35 (= A 20)
lÿ∞se | “i“≤ | mlan[-?]

”evoro“kin’s drawing and reading point to a final letter e e, somewhat

damaged. Schürr omits it in his reading.

E.Th 36 (= A 21)
\ or≤
The first sign, apparently a \, could be the rest of a letter.

E.Th 37 (= A 22)
ktmn
Perhaps an incomplete form of ktmno, cf. E.Th 25 above.

E.Th 38 (= A 23)
]bewmsmnwdiq tebwnqmw

For wdiq (and the suggested segmentation involved), see below E.Th 46.

E.Th 39 (= A 24)
krws | ko“m≤

E.Th 40 (= B 1)
pnu≤ol

From this graffito on, I shall simply reproduce Schürr’s readings, as I

have not seen the drawings of this part of the corpus. I only comment
on the forms that offer some parallel with known words.

E.Th 41 (= B 2)
tmonks

E.Th 42 (= B 3)
rdudmm»≤

E.Th 43 (= B 4)
p
 THE INSCRIPTIONS 105

E.Th 44 (= B 5)
dquq | ewmlane | tebot | gkem≤

The first word shows a good onomastic identification, see Chapter 11,
s. v. The second word contains the typically Theban sequence (m)lane.
The third word is the same as in E.Th 28.

E.Th 45 (= B 6)
krwß

Cf. the first word of E.Th 39.

E.Th 46 (= B 7)
prpwri∞ kblow≤

Apparently a sequence of two names. The first seems to belong to the

family of yriq/ydiq names, but spelled with w, not y/ÿ, and with ∞ (as
in Euromos idyri∞≤). For the first spelling, cf. perhaps the sequence ºwdiq
in E.Th 38, although I do not think we should rule out that v w
may in fact be a variant of W y, so that the name could be read *prpyri∞.

E.Th 47 (= B 8)
w.dbo≤kn ewál.å»e ˚[

E.Th 48 (= B 9)
brsi

A well-known Carian name, see Chapter 11, s. v. and cf. above E.Th 26.

E.Th 49 (= B 10)
bal ewlane | “rb˚[-]sal|

E.Th 50 (= C 1)
pn-?

E.Th 51 (= C 2)
p

E.Th 52 (= C 3)
plqodse | ewm-?-?-?-? | rqemw | k-?[
106 CHAPTER THREE

In plqodse, perhaps the name plqo should be identified, see Chapter 11,
s. v.

E.Th 53 (= C 4)
dr“≤iem

5. Luxor Temple (E.Lu)

Some Carian graffiti were recently identified in the Luxor Temple.18
They are very brief, damaged texts, and only one of the seven I con-
sider to be genuinely Carian19 suggests a clear connection with the rest
of the Carian Corpus.

E.Lu 1 (= G 19)

E.Lu 1 (ESS 1998)

→
ds-?

E.Lu 2 (= G 21)

E.Lu 2 (ESS 1998)

18
They have been published in: ESS (1998). I am very grateful to Richard Jasnow
and to Theo van den Hout for the information provided about this sub corpus.
19
The graffito G 12 is excluded here: it could also be Carian, but none of its signs
can be clearly identified.
 THE INSCRIPTIONS 107

→
rsy
suso
“?rquq [. . .?

The clearest graffito of the collection. The last word is—if the reading
of the first letter here proposed is accepted—the well-known Carian
name “rquq.

E.Lu 3 (= G 22)

E.Lu 3 (ESS 1998)

Very uncertain reading: o or t followed by m and another m or rather

u. The order of reading is unclear.

E.Lu 4 (= G 23)

E.Lu 4 (ESS 1998)

→
?-?-[-]ms[-]ry-?-?

Very uncertain letters at the beginning and the end of the graffito. The
letter immediately after y could be j j, and the last letter a D d.
108 CHAPTER THREE

E.Lu 5 (= G 24)

E.Lu 5 (ESS 1998)

←
b?s?ui∞am | oã?

Van den Hout (pers. comm.) suggests reading the apparent interpunc-
tion trace .as L l. The resulting sequence mlo would have a good par-
allel in C.Ka 9 [. . .]ois?ur?mlo. The presence of % h, not a typical letter
in the Carian alphabet of Egypt, is not certain: the letter could be also
r r. But note that % also appears in the nearby corpus of Thebes.

E.Lu 6 (= G 25)

E.Lu 6 (ESS 1998)

→
| urq

The reading is far from certain. The last sign could also be ≤.
 THE INSCRIPTIONS 109

E.Lu 7 (= G 26)

E.Lu 7 (ESS 1998)

tksr or rather tasr?

Neither of the two possible readings offers a good connection with other
Carian sequences.
Summing up, although these graffiti are practically unusable (with
the exception of E.Lu 2), the presence of Carian graffiti among the
graffiti of the Luxor temple is indisputable, as demonstrated by the
presence of genuine Carian letters such as I or f.

6. Murwàw (E.Mu)
In his monumental work on the inscriptions of Lower Nubia, Zbyn^k
¥ába published, in collaboration with Fritz Hintze, a Carian graffito
found in the region of Murwàw, about two kilometres north of the
temple of Dendùr (¥ába 1974[79] nº 196 & fig. 323).
This inscription does not appear either in Masson’s indexes (Masson
1978) or in the transcribed corpus of Meier-Brügger (Meier-Brügger
1979b). Our only source is the unsatisfactory edition of ¥ába, which
includes a fairly poor quality photograph, an inaccurate and unreliable
facsimile (reproduced here), a transliteration to the systems of Bork-
Friedrich and ”evoro“kin, and some rather confusing notes. Only the
reading of two words known elsewhere is certain.
110 CHAPTER THREE

E.Mu 1 (= ¥ába 1971, 196)

E.Mu 1 (¥ába 1971)

→
p?owk | wljat≤ | ÿnsmsos
saaw?on sa?awon

Despite the drawing, the readings wljat≤ and ÿnsmsos are practically
assured because they are known elsewhere: wljat≤ (not ‡wljot≤ !) = E.Th
7 wljat, and ÿnsmsos = E.AS 3 ÿnsmsos. The remaining words pose very
serious reading problems. As well as powk, an alternative reading sowk
is possible, and even the last letter of this word is far from certain: it
could also be t t, which gives sowt/powt as further alternative readings.
The alleged letter / in the second line—unexpected in an Egyptian
inscription—seems in fact to be ~ a. The two words of the second
line are generally taken as almost identical (already in ¥ába, see draw-
ing), but they could be different, coinciding only in their final part
(-on). In this sense, the letter under s in the second word might be not
an a but in fact a b b, so that a reading sb-awon would be possible. In
conclusion, only s—on is completely certain for both words of the sec-
ond line. Therefore, the reading adopted here is merely provisional.

7. Silsilis (E.Si)
There is not a modern edition of the Silsilis graffiti, it should be added
at this point that in all probability producing this new edition will never
be possible: according to Masson (1969:32), a great number of these
graffiti must have been destroyed.
The corpus consists of 11 generally very short and impenetrable
graffiti. Only a pair of words that correspond to the rest of the Carian
documentation can be clearly interpreted: psma≤k≤, the name Psammetichus,
well documented in the Carian of Egypt (E.Si 2, E.Si 7), and be?bint
(E.Si 4), also attested in Thebes. I adopt the readings of Masson (1978),
with the exception of E.Si 2 and E.Si 10. In the first case, I follow
 THE INSCRIPTIONS 111

Schürr’s proposal of reading the text as boustrophedon, and modifying the

interpretation of some signs from the drawings of Sayce and Legrain
(Schürr 2000:172). In the case of E.Si 11, Schürr has argued con-
vincingly for a new reading, based on a photograph published by Hans
A. Winckler in 1939 (see below).
I provisionally reinstate the inscription Si 61 F (= Sayce 1906, n. 5),
excluded from Masson’s corpus. This is also present in ”evoro“kin (1965)
(69 ”), see also Schürr (1996b:149).20

E.Si 1 = Si 39 F

E.Si 1 (Friedrich 1932)

←
∞iqud | marariso[-. . .]

E.Si 2 = Si 53 F

Sayce I

Sayce II

20
For technical reasons, the source of my illustrations will be Friedrich (1932), where
Sayce’s drawings are reproduced.
112 CHAPTER THREE

Legrain
E.Si 2 (Friedrich 1932)

→ [—]e∞ld | wa | psma≤k≤ |
← uejresi | qan | kolt | kowrn[. . . ?

I offer the three existing copies of this graffito (two by Sayce, the third
by Legrain 1905) and adopt Schürr’s proposed reading (Schürr 2000:172).
In any case, as noted above, the only clearly recognizable form is the
name psma≤k≤ in the first line.

E.Si 3 = Si 54 F

E.Si 3 (Friedrich 1932)

←
irasa | n[-]eakrnanb

E.Si 4 = Si 55 F

Sayce

Legrain
E.Si 4 (Friedrich 1932)
 THE INSCRIPTIONS 113

→
[. . .]t bebint | sqlumidun | sqla

For the diamond-sign preceding the same word, see above E.Th 28.
The reading bebint (instead of ‡bÿbint in Adiego 1993a) now seems prefer-
able, given the clearer evidence of the Theban inscriptions.

E.Si 5 = Si 56 F

E.Si 5 (Friedrich 1932)

→
betkrqit[—. . .]

E.Si 6 = Si 57 F

E.Si 6 (Friedrich 1932)

→
bÿta“ | sursiabk | dr[-. . .]
qku

E.Si 7 = Si 58 F

E.Si 7 (Friedrich 1932)

→
psma≤k
114 CHAPTER THREE

E.Si 8 = Si 59 F

E.Si 8 (Friedrich 1932)

→
bij≤≤pe (. . . ?)

E.Si 9 = Si 60 F

E.Si 9 (Friedrich 1932)

→
[. . .]rbn“a[—. . .

E.Si 10 = Si 61 F

E.Si 10 (Friedrich 1932)

←
∞?mpi

E.Si 11 = Si 62 F

E.Si 11 (Friedrich 1932, modified)

 THE INSCRIPTIONS 115

←
dmo“bqs

Reading taken from Schürr. The photograph of the inscription pub-

lished by Winckler in 1939 (which I was made aware of by Schürr)
quite clearly substantiates the reading proposed by Schürr. It also serves
to definitively rule out the alleged presence of a letter b in the Carian
alphabet of Silsilis.21

8. Abu Simbel (E.AS)

Fortunately the current situation of the Abu Simbel corpus is very
different to those mentioned above; we have at our disposal a recent
and very careful edition of these graffiti, thanks to the efforts of Olivier
Masson (Masson 1979), who was also responsible for the publication,
some years ago and in collaboration with André Bernand, of the cor-
pus of Greek graffiti from the same location (Bernand-Masson 1957).
The Carian graffiti from Abu Simbel are strongly tied to the history
of research on Carian: it was Richard Lepsius who first edited and cor-
rectly identified as Carian these inscriptions found on the legs of the
two colossi of Abu Simbel (Lepsius 1844, Abt. 6, Bl. 98 [Kar. 1–3],
and Bl. 99 [Kar. 4–7]). This theory was confirmed by Sayce (1874,
1887[92]) when texts in a similar alphabet were found in Caria and
bordering areas. However, from an epigraphical point of view, Sayce
could not contribute to an improved edition of the text, and he had
to be satisfied with Lepsius’ copies (see Masson 1979:35–36 for details).
Masson’s edition is based on the drawings and photographs made
by André Bernand and Abd el Latif Ahmed Aly in 1956. The draw-
ings had already been published provisionally (Bernand-Aly 1959?), and
reproduced in ”evoro“kin (1965). The corpus presented here introduces
some corrections and additions made by Diether Schürr.

21
See Schürr (1996b) where, contrastingly, the presence of a letter / is argued.
Later, Schürr argued in favour of a f, which seems more likely. This latter reading
is the one that I adopt here. For the alleged initial letter “my” in the inscriptions, see
the convincing arguments in Schürr (1996b) for rejecting it.
116 CHAPTER THREE

E.AS 1 = AS 1

E.AS 1 (Masson 1979)

←
par≤olou
[. . .]oe

Graffito discovered by Bernand.

E.AS 2 = AS 2

E.AS 2 (Masson 1979)

←
“abd?aikal
Also a new graffito found by Bernand, absent from Lepsius’ corpus.
Schürr’s reading is followed here.

E.AS 3 = AS 3 = Lepsius Kar 4

E.AS 3 (Masson 1979)

←
pisma“k | “arnw≤ | ÿnsmsos
 THE INSCRIPTIONS 117

E.AS 4 = AS 4 + Lepsius Kar 7

E.AS 4 (Masson 1979 + Lepsius 1844)

→
a∞akowr | emsglpn | b[. . .] pisma[“/≤k . . .]

The first part of the graffiti (which appears in the drawing) was rec-
ognized by Bernand, while Lepsius was only able to identify some illeg-
ible characters (Bernand apud Masson 1979:39, n. 39). But as Schürr
notes ( per litteras), Lepsius Kar. 7 (left out by Masson because it was
not re-found by Bernand) is actually the continuation of E.AS 4, and
the beginning of the Egyptian name (in its Carian adaptation) pisma“k/
pisma≤k is easily readable in Lepsius’ copy.
The direction of the reading offered for AS 4 in earlier works (for
instance Adiego 1993a) was incorrect.

E.AS 5 = AS 5 = Lepsius Kar 6

E.AS 5 (Masson 1979)

←
pnyri≤ru | iÿkr≤ | “a[--]i≤b?wn

Schürr suggests that the final letters may be part of another graffito.
118 CHAPTER THREE

E.AS 6 = AS 6 = Lepsius Kar 3

E.AS 6 (Masson 1979)

←
platt
slaÿ≤ ∞i

E.AS 7 = AS 7 = Lepsius Kar 1

E.AS 7 (Masson 1979)

←
naz ∞i∞ | bÿ“ | esak?dow“ | mÿqudem | pisma≤k | bebint | mo | ne
| psÿ“[|?] ai[-]iqom

Schürr’s reading (Schürr 2001b:108, and per litteras) is adopted here.

E.AS 8 = AS 8

E.AS 8 (Masson 1979)

 THE INSCRIPTIONS 119

←
nid≤kusas | meÿqak | sn≤ | ≤t≤ | ≤uni≤ | k“mmsm[. . .]
[. . .]r≤wk[-]“[

Graffiti discovered by Bernand. Masson’s reading is followed, despite

the significant problems posed by the final part.

E.AS 9 = Kar 2

E.AS 9 (Lepsius 1844)

→
ÿsm [?

This graffito is omitted without any reference in Masson (1979). It had

already been identified as a part of a Greek graffito in Bernand-Masson
(1957:38; the graffito is Bernand-Masson nº 32), but Schürr ( per litteras)
maintains that it is in fact Carian. In fact, no connection seems to exist
between the clear drawing wsm of Lepsius and the Greek graffiti as
reproduced in Bernand-Masson loc. cit.

9. Buhen (E.Bu)
Also in the case of the graffiti from Buhen, we can refer to an excel-
lent, recent edition, made by Masson and published together with the
Saqqâra corpus (Masson 1978).
In 1895 Sayce published seven inscriptions copied in the Southern
Temple of Buhen (then mentioned as “Temple of Thothmes III at
Wadi Halfa (Sayce 1895). He had serious doubts about the Carian
nature of two of these (numbers 6 and 7). They have not been redis-
covered since and must remain outside our collection.
For his edition, Masson was able to re-read three of the five clearly
Carian graffiti (E.Bu 2, 3, 6), and he also added a new and important
graffito found during the British campaign 1962–1963 (E.Bu 1). For
the remaining Carian graffiti (4 and 5), Masson was obliged to work
only from Sayce’s copies.
120 CHAPTER THREE

Despite the quality of his edition, Masson was unable to find a sat-
isfactory solution to the problem posed by the same word that appears
repeatedly in different inscriptions from Buhen. Following the deci-
pherment of Carian, we now know that this word is the name Psam-
metichus in Carian, and that the many divergent readings adopted by
Masson must be brought together. This correction affects E.Bu 1 ( psma≤k≤,
not ‡psma≤u≤ ), E.Bu 4 ( psma≤k, not ‡psma≤m) and E.Bu 5 (identical cor-
rection). Other discrepancies with Masson’s edition affect the word ar®i“
(not ‡arli“ !) in E.Bu 1 and E.Bu 2 (”evoro“kin’s correction, see ”evoro“kin
1984[86]:199), and ibrsi≤ (not ‡iyrsi≤ !) in E.Bu 4 (Schürr’s correction:
Schürr 1991–1993). These corrections were already set out in Adiego
(1993a).

E.Bu 1 = M 50

E.Bu 1 (Masson 1978)

←
[—]msal | ar-
[® ]i“ | psma≤-
k≤ | urm≤ | an-
kbu“ | trel
kbou≤
 THE INSCRIPTIONS 121

E.Bu 2 = M 51

E.Bu 2 (Masson 1978)

←
euml?bna-
sal | ar®i“
pdtom≤
urom≤ | an-
kbu“

E.Bu 3 = M 52

E.Bu 3 (Masson 1978)

←
[-]tmai≤[—]
122 CHAPTER THREE

E.Bu 4 (= M 53)

Sayce (1895)

Masson’s copy of Sayce’s drawing

E.Bu 4

←
psma≤k
ibrsi≤

This constitutes a good example of a biased interpretation of a text

known only from a single copy; the inscription was not found again
when Masson was preparing his edition, so he had to use Sayce’s draw-
ing. Masson interpreted the second letter of the second line as a W,
modifying Sayce’s copy in order to approximate the traces of the let-
ter to a more standard form. However, as Schürr (1991–1993:167) has
convincingly demonstrated, this letter is actually Ø b, and not W y,
since we are dealing with the typical Carian name ib(a)rsi, brsi, attested
by a number of examples. The other problematic letter is the last one
of the first line, for which an interpretation as k k is now beyond
doubt, as the whole word is the Carian form of the Egyptian name
Psammetichus.

E.Bu 5 (= M 54)

E.Bu 5 (Masson 1978)

←
psma≤k

As in E.Bu 4, the last sign is now interpreted as k, in order to obtain

the well-known Egyptian name psma≤k.
 THE INSCRIPTIONS 123

E.Bu 6 (= M 55)

E.Bu 6 (Masson 1978)

←
eypsal
puor≤ | aor≤
ursea∞k ∞i

This inscription is extremely difficult to read, so some uncertainties

remain. In any case, the reading aor≤ for the third word should be pre-
ferred to ‡a≤r≤, as found in previous works, given the good parallel in
E.Me 1 and the sound onomastic identification (see Chapter 11, s. v.
aor≤ ).

10. Gebel Sheik Suleiman (E.SS)

E.SS 1 (= 72 F)

This graffito, which constitutes the southernmost document of Carian

to be found in Africa, is known only from two different copies: Sayce
(Sayce 1910 = Friedrich 1932) and A. J. Arkell (reproduced in ”evoro“kin
1965) respectively. As we are informed by Masson (1978: 35, n. 1, and
98), this inscription, which never was photographed, now lies beneath
the waters of the Nile. The correction introduced in Adiego (1993a) is
maintained here: the second word of the second line is read “ÿin≤, not
‡“ÿ“n≤, because it is undoubtedly the same word that appears in E.Me
38.
124 CHAPTER THREE

Friedrich (1932)

”evoro“kin (1965)

←
n≤n[-]s“|“aru≤ol
pneit|“ÿin≤
parÿd∞≤

11. Unknown Origin, Likely from Egypt (E.xx)

Under this title I include seven inscriptions whose Egyptian origin seems
clear, but for which no exact location can be determined.

E.xx 1 (= MY C)

E.xx 1 (Masson-Yoyotte 1956)

 THE INSCRIPTIONS 125

←
qorb | isor≤ ∞i | ≤ugli≤

This stela is known as the “stèle de Grenoble”. The attribution of an

origin from Bubastis was discarded in Masson-Yoyotte (1956:7). In any
case, the alphabet is clearly of the same kind as that used in Memphis
and Sais. Unproblematic reading.

Photograph: Masson-Yoyotte (1956: pl. IV). Drawings: Masson-Yoyotte (1956:8–9).

E.xx 2 (= MY I)

E.xx 2

←
wliat

This so-called “ichneumon of Berlin”, is actually a reliquary for a

mummified shrew with a Carian name followed by an Egyptian votive
formula ‘X (the god incarnated by the animal) may give life’.

Photograph: Masson-Yoyotte (1956: pl. 4). Drawing: Masson-Yoyotte (1956:37). My

drawing of the shrew is based on the photograph.

E.xx 3 (= MY a)

E.xx 3
126 CHAPTER THREE

→
ionel≤
Inscription on the mount of a ring.

Photograph: Masson-Yoyotte (1956: pl. VIII). Drawing: Masson-Yoyotte (1956:11). My

drawing of the object is based on the photograph.

E.xx 4 (= MY b)

E.xx 4

→
pduba

Word engraved on a vase.

Photograph: Masson-Yoyotte (1956: pl. VIII). Drawing: Masson-Yoyotte (1956:12). My

drawing of the object is based on the photograph.

E.xx 5 (= MY c)

E.xx 5 (Friedrich 1932)

 THE INSCRIPTIONS 127

←
ow∞meb≤t

This is the “scarabaeum with a Carian inscription”, first edited by

Sayce (1887[92]). It was briefly referred to in Masson-Yoyotte (1956)
(MY c), but neither photography nor drawing or transcription was
attempted at the time, given that neither editor knew where the object
was. The object was rediscovered a short time after the publication of
Masson-Yoyotte (1956), and Masson dealt with it in an article (Masson
1959b). In the absence of more information, Masson considered it to
be “pseudo-Carian”. In Masson (1978) it was omitted.
Contrastingly, ”evoro“kin included it in his book on Carian (”evoro“kin
1965: 127; 6 ”) and offered a very plausible reading (ow∞meb≤t). The
final sequence ºb≤t is a particularly attractive suggestion, because it also
appears in two graffiti from Thebes (E.Th 12 and E.Th 13). I have
therefore included alongside the other Carian texts.

Photograph: Masson (1959: pl. 2). Drawing: Masson (1959: pl. 3).

E.xx 6 (= 4 ”)

E.xx 6

←
“arnajs | sb taqbos

Inscription on the so-called “Leningrad Isis”, published by ”evoro“kin

(”evoro“kin 1964b; ”evoro“kin 1965). As in the case of E.xx 2, a Carian
128 CHAPTER THREE

onomastic formula is completed by an Egyptian votive phrase, here Js.t

dj ‘n˙ ‘Isis may give life’.
I adopt here a new suggested reading formulated by Diether Schürr.

Photographs: ”evoro“kin (1964b:pl. I–IV). Drawing: ”evoro“kin (1964b:58–59). My

drawing of the inscription is based on the photograph.

E.xx 7 (= Lion)

E.xx 7 (Masson 1976)

←
ntros: prãidas
or“a
nu mdane: uksi wrm≤

Inscription on a lion of Egyptian origin, edited by Masson (Masson

1976). Some years ago, I suggested to Masson the possibility of read-
ing ntros ntros, instead of noros ‡noros (sic in Masson 1976), a
suggestion based both on the photograph and the comparable sequence
ntro ntro in C.xx 1. Yoyotte’s revision of the inscription made this
correction possible, which was already included in Adiego (1993a) and
is now widely accepted.

Photograph: Masson (1976: pl. 1). Drawing: Masson (1976 pl. 2).

D. THE CARIAN INSCRIPTIONS FROM CARIA

The first and last specific edition of all the Carian inscriptions from
Caria available at the time was produced by Louis Robert and Louis
Deroy, in two complementary articles: Robert (1950), which includes
 THE INSCRIPTIONS 129

photographs and epigraphical information, and Deroy (1955), a col-

lection of drawings based on the photographs published by Robert.
The numbering of the inscriptions introduced by Robert and adopted
by Deroy became the standard (hence the use of “D” or, less frequently,
“R.-D.” followed by a number to refer to the Carian inscriptions from
Caria). This ordering did not follow a coherent set of criteria.22
As for the inscriptions published after the Robert-Deroy edition, in
Meier-Brügger (1983) a catalogue was established (although the inscrip-
tions were not edited) and the numbering system of Robert-Deroy was
continued, but an asterisk mark was added to the new inscriptions
instead of the “D” and “R.-D” abbreviations.23 The numbering crite-
rion employed was the order of publication of the new inscriptions. A
source of further confusion was the fact that both Robert-Deroy’s and
Meier-Brügger’s catalogues included Para-Carian inscriptions, here omit-
ted from the true Carian corpus. The inscriptions from locations other
than Caria were also listed, so this catalogue actually comprises all the
non-Egyptian-originated epigraphical documents of Carian.
In this study, a new classification system, parallel to that introduced
for Carian inscriptions from Egypt, will be introduced. The inscriptions
are classified according to the locations in which they were found.
We are clearly in a better position to analyse the corpus from Caria
than the corpus of Egyptian texts: with the sole exception of C.Tr 2,
at least photographic evidence is available for all the inscriptions.
Moreover, some of the texts edited by Robert and Deroy have been
revised thanks to the efforts of Blümel, Gusmani, Frei and Marek,
whereas the texts found after Deroy’s compilation have generally been
edited satisfactorily.

22
1–3 are already published. 4–5 are inscriptions neglected by scholars prior to
Robert (in fact, “Para-Carian” inscriptions from Chalketor and Ancin, south of Alabanda).
6–13 are inscriptions found by Robert himself. 14 is a text discovered by G. E. Bean,
15 is a Carian (Kaunian) inscription from Lycia already known and previously pub-
lished (as 1–3), and 16, offered in addendum, is the great inscription of Kaunos, also
discovered by Bean. In Deroy’s article, the confusion continues: D 17 are Para-Carian
graffiti from Labraunda, D 18 Carian coins, and D 19, the Greek-Carian bilingual
inscription from Athens.
23
For the continuation of the numbering after Meier-Brügger (1983), see Meier-
Brügger (1994:113), Frei-Marek (1997:6, n. 10), Frei-Marek (2000:85).
130 CHAPTER THREE

1. Tralleis (C.Tr)

C.Tr 1 (= D 1)

Deroy (1955) Kubitschek’s copy conserved in Vienna

C.Tr 1

←
sdi amt[
pau≤
art{ }mon

This inscription has not been re-found since it was seen and copied at
the end of the 19th Century, but Deroy’s drawing, based both on a
photograph of the cast and a copy conserved in the notes of Kubitschek
in the Austrian Academy of Vienna (which I reproduce here)24 is quite
reliable. Moreover, the inscription does not pose any particular prob-
lems of reading. The third letter of the first line is definitely Y i.
I adopt the reading suggested by Schürr (cf. Schürr 2001b:109,
n. 12), but I am unsure about the segmentation of the first line.

Photograph of a cast and drawing in Deroy (1955:307 and pl. I).

24
I am grateful to Dr. Georg Rehrenböck (Austrian Academy of Sciences, Vienna),
for your extreme kindness in sending me a reproduction of Kubitschek’s drawing and
notes concerning this inscription.
 THE INSCRIPTIONS 131

C.Tr 2 (= D 2)

Pappakonstatinou’s drawing Kubitschek’s copy conserved in Vienna

(Meier-Brügger 1978)

C.Tr 2

→
an sidi a-
rtmi pau≤
parãaq?

This inscription, which has since disappeared, is known only through

drawings.
I retain the reading adopted in Adiego (1993a), based on the image
published by Meier-Brügger (1978:81), a direct reproduction of the
drawing edited by Pappakonstantinou in 1895. The copy conserved in
Vienna among Kubitschek’s travel notes (see above C.Tr 1), brought
to my knowledge by Dr. Georg Rehrenböck, is concurrent with Pappa-
konstantinou’s drawing, the only surprising difference being the omis-
sion (accidental?) of r in the third line ( paãaq? instead of parãaq), see
the illustration.
The only problem of reading is the last letter of the inscription:
Schürr (2001b:109, n. 12) argues that it is ≤ (“das letzte Zeichen wohl
≤ ”), but Pappakonstantinou’s and Kubitschek’s drawings points rather
to a Q q.

Drawings: Deroy (1955:308), Meier-Brügger (1978:81).

132 CHAPTER THREE

2. Alabanda and Surroundings (C.Al)

C.Al 1 (= D 13)

(Deroy 1955)

→
sdi a[-]mob[

Found by Robert in Eski Çine, south of Alabanda. To my knowledge,

the inscription has not been seen again. I adopt Schürr’s reading (Schürr
2001b:109, n. 12).

Photograph: (1950: pl. II 1, VIII 2, XXI 2). Drawings: Robert (1950:17), Deroy (1955:319).

3. Euromos (C.Eu)

C.Eu 1 (= D 3)

Drawing conserved in Vienna

→
≤as: ktais idyri∞≤: mn[os?]

This inscription was copied, photographed and edited for the first and
last time, by E. Hula and E. Szanto in 1894. As Meier-Brügger has
demonstrated (Meier-Brügger 1978:78–79), the drawing made by the
first editors is considerably better than that published in Sayce (1905)
and Deroy (1955). I reproduce here a copy of the original drawing by
Hula and Szanto conserved in Vienna, kindly sent to me by Georg
Rehrenböck.
 THE INSCRIPTIONS 133

The main difficulty is posed by the alleged sequence WW yy. Some

years ago, ”evoro“kin (in his unpublished document “Corrections to
Existing Copies”) proposed to correct it on the basis of the direct study
of the cast: according to him, the sequence was actually a Wr yr. This
correction is eminently suitable from a linguistic point of view, since a
sequence WW yy is hardly acceptable; moreover, a reading Wr yr yields
a good onomastic identification (see Chapter 11, s. v. idyri∞≤ ). However,
considered from the point of view of the local alphabet of Euromos,
we would expect a form R—rather than a form r—for r, as can be
seen in the other inscription from Euromos (C.Eu 2).
Thanks to the kindness of Dr. Rehrenböck, I have been able to con-
trol directly the two (very similar) casts of the inscription conserved in
Vienna. Both Dr. Rehrenböck and I agree that, at first sight, an W can
be recognised in both casts. Without totally ruling out the possibility
of an original mistake made by the engraver, one can suppose that
either the intermediate trait or the lower trait is intrusive. It is true
that the intermediate trait seems weaker, favouring thus a reading R
(which would be more consistent with the testimony of C.Eu 2), but it
is frankly impossible to decide definitively between a reading r, or a
reading R.
The last word, incomplete, must be without any doubt the Carian
word for ‘son’, mno-. The difficulty lies in establishing whether it was
inflected. In my proposal of integration, I tentatively suggest the com-
pletion mn[os?], based on the assumed concordance of this word with
the name ktais, very possibly a name with an -s ending.

Photograph: Robert (1950:Pl. VI 4). Drawings: Deroy (1955:309), Meier-Brügger

(1978:80).

C.Eu 2 (= D 8)

C.Eu 2 (Blümel 1988)

134 CHAPTER THREE

→
omob ∞i: temazi
≤dun: ≤o≤niabkol
armon qyrbmudolo
manon

This inscription was re-edited almost simultaneously, and with divergent

readings, by Blümel (1988) and Gusmani (1990). Wolfgang Blümel was
kind enough to allow me to see a cast of the inscription, and I can
confirm the accuracy of his edition. I reproduce Blümel’s drawing here.
The text does not pose reading problems. The most remarkable inno-
vation of the current transcription is the interpretation of the letter e
not as a variant (Kaunian-like type) of L l, but rather as a local form
of e e, as in Hyllarima and probably also in Mylasa (see below p. 225).
Photographs: Robert (1950), pl. IV. Drawings: Deroy (1955:316),
Blümel (1988:262), Gusmani (1990:49).

4. Kindye (C.Kn)

C.Kn 1 (= D 6)

C.Kn 1 (Deroy 1955)

→
pareÿs

Extremely fragmentary inscription found by Laumonier, which has been

not recovered.

Drawing: Deroy (1955:314). Photograph: Laumonier (1933:35).

 THE INSCRIPTIONS 135

5. Hyllarima (C.Hy)

C.Hy 1 (= D 7 + new fragment)

C.Hy 1 (col a: Adiego-Debord-Varinlio<lu 2005; col. b: Laumonier 1934)

→
(a) “asqariod dymda
muot armotrqdosq

brsi ari“≤ brsi≤

mane: u≤ol≤
rtim u≤ol≤ pur?i≤

u≤bzol tñu≤ brsi≤

pau mane≤ ybr-
s≤

(b) kdu≤opizipususot
mol“ msot ylarmit

(Greek text—only the oldest inscriptions that appear also in the illustration):
(a) flere›ew ye«n pãntvn:
ÑErm¤aw Fan°v ÑErm¤adow
(b) flereÁw ye«n pãntvn:
ÑUssvllow ÉArrissiow

The stone containing the Col. b, was discovered by Laumonier in 1933.

In 2004, a new bilingual inscription was found in Hyllarima, which
136 CHAPTER THREE

turned out to be a fragment of the same stone, the two pieces fitting
together. This sensational discovery provides us with almost the com-
plete marble stela (only the lower part is missing), and the result is a
complex mixture of inscriptions from different periods, which makes
the interpretation of the Carian text difficult, both internally and in
connection with the Greek texts.
The first problem is to establish whether the two first Carian lines
of each column must be read as two complete lines (the first line of
column (a) being followed immediately by the first line of column (b),
and likewise for the two second lines), or if one must begin reading
from the first two lines of (a) and then the first two lines of (b), or
whether column (a) and column (b) in fact represent two independent
inscriptions.
Also problematic, and related to this discussion, is the connection
between Carian and Greek texts. It is clear that the majority of Greek
texts seem to have been engraved long after the Carian lines,25 but the
first four lines of column (b) could be contemporary with the Carian
texts. Whether this contemporaneity implies a connection between the
two texts or not is impossible to decide.
The problem is compounded somewhat by the fact that in the Carian
text of (a), the last three lines, clearly separated from the preceding
ones, show marked differences in the shape of the letters, and thus
seem to have been written long after the preceding lines. A similar sit-
uation can be observed in the four Greek lines mentioned, the last two
showing divergent traits to the first. This opens the way for a number
of different hypotheses about the order in which the sections of the
Carian inscription and the sections of the oldest Greek inscription were
engraved.
Independent from the problems of interpretation, which will be dealt
with in pp. 305–308, the inscription of Hyllarima is an exceptional
document: it is the best-preserved long Carian inscription, the text being
complete and with only one difficulty of reading: the antepenultimate
letter of the fifth line of column (a), apparently a R r.

25
The Greek texts are the following: in col. (a), immediately after the Carian inscrip-
tion, a list of Apollo’s priests at the time of the joint rulership of Antiochos and his
son, dated in 263–262. In col. (b), after the Carian text, a list of priests of all the gods
followed by a sale of the priesthood of all the gods, and a land renting document. In
the lateral side of col. (a), another sale of priesthood of the same date as that of col.
(b), the name of the divinities implied not being readable. In the lateral side of col.
(b), other land renting documents.
 THE INSCRIPTIONS 137

Col. b, photographs: Laumonier (1934:347), Robert (1950:pl. V), drawings:

Laumonier (1934:346), Deroy (1955:315). Edition of the new frragment in
Adiego-Debord-Varinlio<lu (2005).

6. Mylasa (C.My)

C.My 1

C.My 1

→
idrayridsemdbq mol“ ty∞[
tsial tusol≤: moi m[-]sao[
banol paruos≤: p?au paryri∞≤
qzali obrbi≤: tsial obrbi≤
banol yrqso≤: paryri∞ psoir≤
[-]bdo pnu≤o≤: myze trdy≤
“arkbiom qzali≤: ≤umo kbdmu≤
skdubrotoz≤: pau ∞toi≤
[-]qo idyri∞≤: ksbo idu≤ol≤
[-]obiokli≤: ∞toi yrqso≤

Inscription found in 2004 in Kırca<iz, near Mylasa. As with other

Carian inscriptions, this text poses serious reading problems regarding
the letters z, Q and o, which in several cases can barely be distin-
guished. It also seems that the engraver was not overly experienced in
138 CHAPTER THREE

his task—as the editors of the inscription have pointed out (Blümel-
Kızıl 2004: 138)—and that some difficult readings could in fact be mere
errors.
The reading adopted here follows in general terms that given in
Adiego (2005), but with the inclusion of improvements suggested more
recently by Blümel (Blümel 2005). I am aware of some reading prob-
lems that still exist. In any case, this inscription is a very important
document both for Carian onomastics and for our knowledge of the
Carian alphabetic variety of Mylasa.

Photograph: Blümel-Kızıl (2004). The drawing presented here has been made on the
photograph.

7. Sancutary of Sinuri near Mylasa (C.Si)

C.Si 1 (= D 9)

C.Si 1 (Deroy 1955)

→
adymd“: yri∞ñ: t[-]rsi: [. . .?]
tbe≤
(vacat)
yri∞ñ: binq: sñaidlo
 THE INSCRIPTIONS 139

Found by Pierre Devambez in 1935, it has been not recovered since.

The drawing reproduced here is from Deroy (1955). Several reading
problems therefore remain unresolved. I adopt the reading that can be
deduced from Deroy’s drawing.

Photographs: Robert (1950: pl. 1). Drawing: Deroy (1955:316).

C.Si 2 (= D 10)

C.Si 2 (Deroy 1955)

→
(a) [—]ryin ∞tmño≤: sb ada ∞tmño≤
eri: pisñoi mda: pñmnn≤ñ: pda-
∞m≤uñ ∞i “aoyr∞ri mt∞elã
ñmailo mda lrHñ: stspñ vacat
sm“s[—5—] sb añmsñsi mda
sm[—7—]a∞e[
∞[—8—]tuñdñ[
ñe-?-[

(b) pim[. . .]
Ha?[. . .]
140 CHAPTER THREE

(Greek text:)
Robert (1945), n. 75
[ ÉIdri°vw ÑEkatÒmnv]
[ ka‹ ÖA]daw ÑEkatÒmnv ka[
[ ?S]uennitvn fler°iow [
[ ]now Ponmoonnou [
[ét°l]eian pãntvn ep[

[ÉId]rieÁw ÑEkatÒ[mnv ka‹ ÖAda]

ÑEkatÒmnv ¶[dvkan?
ét°leia[n
sunge[n
tis o[
oug.[

Robert (1945), n. 74 (+ C.Si 2).

] ßkastow
]areihw
?k]atÉ §niautÚn
?§]p‹ Nhsaiou
]w
to›]w Suenn¤oiw

Bilingual inscription found by Louis Robert. The Carian part was edited
by Robert (1950) and Deroy (1955). The Greek part can be found in
Robert (1945). Regrettably, neither Blümel’s efforts to locate the inscrip-
tion years later nor the steps given by Schürr in order to obtain Robert’s
cast have been successful.
In any case, the reading of the three first lines is quite certain,
with the important correction made by Schürr, consisting in reading d
d instead of F, in the name ada. Things are not so clear from the fourth
line onwards: H in lrHñ could be U (therefore lruñ), and the t t in stspñ could
be o o (sospñ), etc. Also problematic is the reading of the entire fifth line.
Schürr’s suggestion of linking this fragment of stone with another
from Sinuri that contains two decrees in Greek by Idrieus and Ada
(Robert 1945, nº 75) seems to me a very attractive theory, and it is
adopted here (see Schürr 1992:136–138; cf. Adiego 2000:134–135 for
details; I reproduce the text Greek edited in this latter article).26

26
This edition was revised by W. Blümel.
 THE INSCRIPTIONS 141

This inscription can be now dated more accurately: the mention of

the joint ruling of Idrieus and Ada situates it between 351/350 and
344/343 (cf. Hornblower 1982:45).

Photographs: Robert (1950:pl. II and III). Drawing: Deroy (1955:317).

8. Kildara (C.Ki)

C.Ki 1 (= D 11)

C.Ki 1 (Blümel-Adiego 1993)

→
[. . . . . . . (.)]zolba∞a[..(.)] kil[
[. . .]uda[. . .] trqdimr qrds tazomd[
kilarad[-]ybzsdmHnmkda[-]aHuq[
iasoum

Greek text:
¶doje KildareËsin, §kklhs¤hw genom°nhw: ÑUss[vllvi?]
Samvou eÈerg°thi genom°nvi Kildar°vn ét°[leian]
doËnai ka‹ proedr¤an ka‹ §sagvgØn ka‹ §jagv[gØn]
ka‹ §n efirÆnhi ésule¤ ka‹ ésponde‹ ka‹ aÈt«i
ka‹ §kgÒnoiw: ka‹ Kildar°aw e‰nai ín y°[lvsin?]

Discovered by Louis Robert in 1934, this inscription is followed by a

Greek decree honouring Uss[vllow], son of Samvow. The inscription
was seen again by Blümel, and a new edition of the Carian text was
subsequently published in Kadmos (Blümel-Adiego 1993).27 The present

27
The Greek inscription has also been published by W. Blümel in Die Inschriften von
Mylasa. II, Inschriften aus der Umgebung der Stadt (= I. K. 35), Bonn 1988.
142 CHAPTER THREE

drawing and reading synthesize the readings of Deroy and Blümel-

Adiego, both conserving the letters that were impossible to identify in
the revision, and incorporating the new readings of some letters.
I still see the exact reading of the last word as problematic; the read-
ing m of the last letter implies the presence of a M oriented to the
left, contrary to all the other instances of the letter in the inscription
(m), and also the rightward direction of the writing. The reading of the
remaining letters is also uncertain. I believe therefore that a reading
*ßAmous “amous, instead of ÎAsouM iasoum, cannot be altogether
dismissed. Such a reading would be compelling, insofar as it would
allow us to identify in the Carian text the name of the father of the
individual honoured in the Greek section, and consequently to estab-
lish a connection between the texts. But for want of sufficient evidence,
I maintain the old reading, iasoum.

Photographs: Robert (1950:pl. VI 2, VIII 1, IX 2, X), Blümel-Adiego (1993:pl. 1).

Drawings: Deroy (1955:318), Blümel-Adiego (1993:89).

9. Stratonikeia (C.St)

C.St 1 (= D 12)

C.St 1

→
]sel“ a[—]a[———]om≤
]som[n?]e brsi≤ ula[——]ol
]latmne≤ ≤ysñal[
 THE INSCRIPTIONS 143

] ari“ maqly≤[
]sel“ piks[
]sel“ p[

This inscription, found by Robert in 1946, has not been revised since
then. The present reading and modified drawing are based on my own
interpretation of the photograph published by Robert (1950). The most
remarkable innovation appears in the second line. I propose that a pos-
sible personal name som[n]e (see Chapter 11, s. v. somne/somne≤ ) can be
recognized, followed by the well-known name brsi≤, in genitive: the let-
ter R r seems to be certain (and not W y, as formerly read), and the
sign preceding it (b) seems to be the local form of $ 4 b, although
neither Robert’s photograph nor Deroy’s drawing provides a clear image
of what the exact form of the letter was. Also new is the reading Q
q, not o o, of the seventh sign of the fourth line.
For the first line, I adopt the reading of Schürr (2001b:106).

Photograph: Robert (1950:pl. VI). Drawing: Deroy (1955:319). The drawing offered
here has been made on the photograph.

C.St 2 (= 36*)

C.St 2 (}ahin 1980 modiWed)

→
u≤ol≤ uodrou u[
mute≤ ymezus[
∞diye≤ uodryia[
uliade pidaru[
144 CHAPTER THREE

mañ“qaraH≤rl-?-[
dar“qemorms[
Hda“qedormñs[

Inscription edited by }ahin (1980).

I adopt Schürr’s suggestion that l. 7, sign 5 is Q q, not o o. In fact,
some rounded signs of the inscriptions are difficult to distinguish cor-
rectly (they could be o, z, or Q).

Photograph: }ahin (1980:pl. V 1). Drawing: }ahin (1980:206). The drawing presented
here is a modified version of }ahin’s, in order to introduce the new readings of some
letters.

10. Halikarnassos (C.Ha)

C.Ha 1 (= 33*)

C.Ha 1

←
smdÿbrs | psnlo | ml orkn tÿn | snn

Inscription on a bronze phiale (6th B.C.) published by Jucker and Meier-

Brügger ( Jucker-Meier-Brügger 1978). The attribution to Halikarnassos
is based on the information given by the dealer to Jucker (“Karien,
Bereich von Bodrum”, Jucker-Meier-Brügger 1978:104; cf. also 109:
“Der Händler, der als Fundort die Gegend von Halikarnass (. . .)
nannte . .”). It is true that a phiale is an easily movable object, so it
could have been made elsewhere. For this reason, the possibility that
the alphabetic variant used here may represent that of Halikarnassos
must be viewed with caution.
The text does not pose reading problems.

Photograph: Jucker-Meier-Brügger (1978:pl. 2). Drawing: Jucker-Meier-Brügger (1978:109).

My drawing is based on the photograph.
 THE INSCRIPTIONS 145

11. Didyma (Ionia, near Milet) (C.Di)

C.Di 1 (= 21*)

C.Di 1

←
]ub“ÿ

Inscription on a broken fragment, datable to the 6th century B.C.

Although this graffito is located not in Caria, but in Ionia, I include
it in the Carian section of the corpus given the clear proximity of
Didyma (and Milet) to the Carian country.
This is a very difficult text: not even the direction of reading is clear.
It is ascribed to the Carian corpus based on the clear presence of the
letter w ÿ. The overall reading is far from certain (see remarks in Adiego
1993a:80). The reading proposed here (the same as in Adiego 1993a)
is merely hypothetical. Steinherr’s reading, adapted to the present deci-
pherment system, would be ‡ul“t, whilst Innocente (1994:106) prefers
to read ‡∞a“ÿ, in reverse direction

Photograph: Naumann-Tuchelt (1963/64: pl. 25). Drawing: Adiego (1993a:324). See

also Tuchelt (1970), which includes a reading and some notes made by Steinherr. My
new drawing is based on the photograph.

12. Iasos (C.Ia)

Italian excavations in the Carian city of Iasos have brought several
Carian inscriptions to light. Their value, with the exception of C.Ia 3,
is very limited: the texts are extremely short and fragmentary, and no
parallels with the rest of the Carian corpus can be traced.
146 CHAPTER THREE

C.Ia 1 (= 20* a)

(Pugliese Carratelli 1985[86])

→
]la
limtaoa | [
om

Inscription on the neck of a vase. I have not seen any photography,

and my only sources of information are the drawings and comments
in Levi-Pugliese Carratelli (1961–62 [1963]:632) and Pugliese Carratelli
(1985[86]:151).

Drawings: Levi-Pugliese Carratelli (1961–62 [1963]:632), Pugliese Carratelli (1985[86]:

151).

C.Ia 2 (= 20* b)

(Pugliese Carratelli 1985[86])

←
]ue∞l | ∞ob[

Although a reverse reading is possible, given that all the letters present
in the inscription are symmetrical, the theory is practically dismissed,
insofar as it would situate L l as the initial letter of a word, which is
highly improbable.

Photograph: Levi-Pugliese Carratelli (1961–62 [1963]:632); Drawings: Levi-Pugliese

Carratelli (1961–62 [1963]:632), Pugliese Carratelli (1985[86]: 151).
 THE INSCRIPTIONS 147

C.Ia 3 (= 38* a)

C.Ia 3

→
?] are“ | “anne mlne | siyklo≤ | “ann | trqule | ∞lmud [?

Inscription on a cratera. Published for the first time by Pugliese Carratelli

(1985[86]:150–151). Gusmani (1988:145–149) offered a new reading
that challenged Pugliese Carratelli regarding the identification of the
two ~ in the second and fourth sequences: Gusmani proposed that
they be read in both cases as /. This new reading would subsequently
be discredited, because this letter is typical only in the Kaunian alpha-
bet, where its existence is inseparable from the fact that in this alpha-
bet there is no t t, and t is represented by T, so that a diacritically
differentiated / (and variants) is used to represented “, instead of
the f F of other alphabets. In the present inscription from Iasos, both
t and f do appear, which leaves the need for a letter such as /
unexplained.
Careful observation of the photograph provided by Pugliese Carratelli
shows that, at least in its second appearance, the letter in question
148 CHAPTER THREE

clearly has a form of ~ very similar to the example in the first word
(the only difference being the somewhat rounded upper trace). It can
therefore extremely unlikely to be a letter comparable to / and vari-
ants. Therefore, Pugliese Carratelli’s reading is preferable.

Photograph: Pugliese Carratelli (1985[86]:pl. I). Drawings: Pugliese Carratelli (1985[86]:

150), Gusmani (1988:146). My drawing has been made directly on the photography
published by Pugliese Carratelli.

C.Ia 4 (= 38* b)

(Pugliese Carratelli 1985[86])

→
n[. . .]
pr[. . .]
is[. . .]

Very fragmentary inscription published by Pugliese Carratelli (1985[86]).

The few remaining letters are easily readable.

Photograph and drawing: Pugliese Carratelli (1985[86]:152).

C.Ia 5 (= 47*)

C.Ia 5 (Berti-Innocente 1998)

→
baqgk[. . .]
 THE INSCRIPTIONS 149

C.Ia 6 (= 48*)

C.Ia 6 (Berti-Innocente 1998)

→
[. . .]b?e≤

C.Ia 7 (= Berti-Innocente 2005)

C.Ia 7 (Berti-Innocente 2005)

←
?]y?n“

C.Ia 5–7 are three very recently discovered graffiti on vases, edited by
Fede Berti and Lucia Innocente (Berti-Innocente 1998 for C.Ia 5, 6;
Berti-Innocente 2005 for C.Ia 7). Note the surprising letter v in C.Ia
7, apparently a variant of Mylasa W, Sinuri-Kildara V (cf. Berti-Innocente
2005:21).
150 CHAPTER THREE

13. Keramos (C.Ke)

C.Ke 1 (= *39a)

C.Ke 1 (Varinlio[lu 1986)

→
uso-
t

C.Ke 2 (= *39b)

→
uso-
t

This pair of inscriptions with the same text, first published in Varinlio[lu
(1986), were briefly the focus of attention when Ray ingeniously sug-
gested that the sole word they contained could be identified with the
well known Carian name Ussollos, u≤ol (Ray 1988). This identification
was based on an assumption that has since turned out to be false: that
the letter c could be equivalent here and in Hyllarima to the letter L
l. It is now clear that c does not represent l in Hyllarima, where
there is a particular letter for this sound (L), and the supposedly equiv-
alent sequence ºusoc in Hyllarima has nothing to do with the Carian
name mentioned (even the segmentation is far from certain!). As already
mentioned (p. 23), I even have a number of doubts concerning the
true Carian nature of these documents, which contain an initial sign
| that makes no sense as a letter in Carian. The disposition of the let-
ters is also a little strange, and their inclusion in the Carian corpus
must be accepted not without certain reservations.

Drawing: Varinlio[lu (1986).

 THE INSCRIPTIONS 151

14. Kaunos (C.Ka)

C.Ka 1 (= D 14)

C.Ka 1 (Deroy 1955)

→
sñis: sdisa-
s: psu≤ol≤
mal≤: mno≤

This funerary inscription, discovered by G. E. Bean and published by

Robert, does not pose any reading problems.

Photographs: Robert (1950: pl. VI 3 and IX, 1); Masson (1973[75]: pl. I, 1) Drawing:
Deroy (1955:320).

C.Ka 2 (= D 16)

C.Ka 2
152 CHAPTER THREE

→
[ui?]omlã qrds grdso[-]i[
[-]r sb a∞mnnartnyr obsmns[
[-]∞arlanoã sb z“ariosã i∞[
[-]nudrma ∞yrpai sarni“ sb u[
[-]aH punot2 otr“ bi sb a∞tmsk[m
[-]d bi 1aitk ouor gdb“laã1_i[-]
[-] sarni“ sb 1orsol“ sb uHbit
[-]bi qrdsol“ ait 1mali H∞it
[-]intnor ∞yrapai≤ umot2 oba
[-]diurt obsmsmñ1ñ ouor mt1_yr
[—]abrun∞ur[-]“yn“Hynn sb vacat
[——————]tbsms _1mali [
[——————]maH sb an[
[.............................]ba vacat

This inscription, the longest Carian text known to date, has recently
been revised by Marek and Frei, and this new reading is followed here.
The segmentations are purely hypothetical, mostly based on the iden-
tification of some clear words (for example the conjunction sb) and
endings (-“, -≤ ).

Photographs: Robert (1950: pl. XXIX, XXX, XXVIII, 2); Masson (1973[75]: pl. II,
2); Frei-Marek (2000:86). Drawings: Steinherr (1950[51]:331), Deroy (1955:321), Masson
(1973[75]:125). For an evaluation of these drawings, see Meier-Brügger (1978). I have
made the drawing on the basis of the excellent photograph published by Frei-Marek.

C.Ka 3 (= 28*)

C.Ka 3 (Roos 1972 modified)

 THE INSCRIPTIONS 153

→
“oru≤
ann ibrs≤

This funerary inscription, first published in Roos (1972), has been revised
by Schürr (see Schürr 1996c, Adiego 1996), and the new reading has
highlighted some important corrections: the first letter is not—as Roos’
drawing indicated—an unexpected form of a letter pi, but in fact a
variant of the typical Kaunian letter /. This new reading is now
confirmed by the photograph published by Frei-Marek (2000). Schürr’s
revision also allows us to identify the well-known Carian stem ibrs-.

Photograph: Frei-Marek (2000:126). Regrettably, this photograph shows only part of

the inscription. Drawings: Roos (1972:93, and pl. 40), Schürr (1996c:158). My draw-
ing is based on that of Roos, but implements the new readings made by Schürr and
the information available from Marek’s photography.

C.Ka 4 (= 30*)

C.Ka 4

→
[. . .]u≤ou≤ ibrsdr[-]
[. . .]a yomln r1_i
[. . .]dar1_ idym“

This fragment, published by Masson (1973[75]), has been re-read and

re-published in Frei-Marek (2000). The new revision has allowed some
154 CHAPTER THREE

corrections of former readings and, perhaps more importantly, has lead

to the rejection of Masson’s suggestion (supported by Peter Hermann)
that this was a further fragment of the long stela C.Ka 2: the thick-
ness of C.Ka 4 is more than 3 cm greater than that of C.Ka 2 (Frei-
Marek 2000:99).

Photographs: Masson (1973[75]: pl. I, 2 and II, 1); Frei-Marek (2000:98). Drawings:
Masson (1973[75]:125), Frei-Marek (2000:97). My drawing is based on the photograph
published by Frei and Marek.

C.Ka 5 (= 44*)

C.Ka 5 (Deroy 1955)

 THE INSCRIPTIONS 155

→
kbidn uiomln i[—]
inis drual nik[—]
lan lysiklas[-?]
otonosn sb lys[ikl]
an lysikratas[-?]
otonosn sarni[“]
mdot2 un sb undo[—]
tl“ kbdyn“ sb b2o[—]
ol“ otr“ sb a∞t[ms]—
kmt absims sb [—]
yt2 oru sb a∞t[—]
bu∞y[——]i[——]i
[—]≤ un moa[-]lboror
[—]Hl∞sasot2 ort
tab sb ort[-] sb Hor-
ouo bi mslmnlia
purmoruos mnos
aitusi

Greek text:

¶doje Kaun[¤]oiw §p‹ dhmio[u]-

rgoË ÑIpposy°nouw: Nikok-
l°a Lusikl°ouw ÉAyhna›o[n]
ka‹ Lusikl°a Lusikrãt[ouw]
[ÉA]yhna›on proj°nouw e[ ‰nai k-]
a‹ eÈerg°taw Kaun¤v[n auto-]
Êw ka‹ §kgÒnouw ka‹ [——]
n auto›w §[. . .

This is the now well-known bilingual inscription honouring two Athenians.

The reading I adopt here is that published in Frei-Marek (1998), which
improves the editio princeps (Frei-Marek 1997) and adds a new frag-
ment of the stela. Concerning the dating, Frei-Marek (1997) proposes
connecting the proxeny decree contained in the stela with events that
happened in 322 B.C. However, Descat (1998) has argued in favour
of a slightly more recent dating: 314 B.C. From a linguistic point of
view this discrepancy is not particularly relevant. In both cases, the use
of Carian writing at the very end of 4th century, in a post-Hekatomnid
era, seems meaningful enough.
156 CHAPTER THREE

The segmentation of the words is based in part on my own analy-

sis of the text, and must be taken as provisional.

Photographs: Frei-Marek (1997), (1998). The drawing here is mine.

C.Ka 6 (= 45*; Schmaltz 1998)

C.Ka 6

→
or

Photograph: Schmaltz (1998: Abb. 4). My drawing is based on the photograph.

C.Ka 7 (= 46*; Schmaltz 1998)

]no≤?(or better: ]noñ?)

C.Ka 6 and 7 are two very brief graffiti found on vases and published
by Schmaltz (1998). No photograph is available for the second of these
in Schmaltz, so we must make do with the reading he offers, leaving
a doubt about the last letter (z or ñ? Schmaltz 1998:209).

C.Ka 8 (= 49* Frei-Marek 2000:116–119)

C.Ka 8

→
potko≤l≤? aba?d?
ya
 THE INSCRIPTIONS 157

Inscription on two blocks from to the wall of the so-called Demeter’s

temple, published for the first time in Frei-Marek (2000:116–119). The
reading offered here differs slightly from that of the first editors: instead
of potko≤lo, I prefer to read—with reservations—potko≤l≤? with a final Z
≤, not o o: the differentiating trace is quite clear, and I do not believe,
contra Frei-Marek, that the different form of the vertical trace regard-
ing the other letter ≤ can is an obstacle to this reading. Moreover, I
give (with some doubts) the reading of the last two signs of the first
line. Both are clearly visible in the photograph, and even the two edi-
tors suggest these readings in their comment on the single signs of the
inscriptions, although they decide against including them in their edi-
tion of the text.

Photograph: Frei-Marek (2000:116). My drawing is based directly on that photograph.

C.Ka 9 (= 50* Frei Marek 2000:120–125)

C.Ka 9

→
[. . .]ois?ur?mlo

Inscription on a broken fragment, published in Frei-Marek (2000:120–125).

This text poses serious problems of reading, since if the we adopt the
interpretation of signs proposed by Frei-Marek, it is not easy to estab-
lish the direction in which the letters must be read: Frei-Marek sug-
gests four possibilities. I believe however that Schürr (2001a) offers the
correct solution to the problem in reading i, not Z, for sign 2, and
s, not m, for sign 3. If the s reading is not absolutely certain (the
apparent trace that would allow to read it as s could be an intrusive
mark), in the case of i, Schürr’s proposal must be accepted, and this
latter reading would mean that the orientation of the text no longer
158 CHAPTER THREE

poses a problem: the fragment must be oriented as it appears in my

drawing, and it must be read from left to right (the photograph in
Frei-Marek 2000 must therefore be inverted; the photography is cor-
rectly reproduced in Schürr 2001a:63). Some doubts remain, however,
about the sign read as r r by Schürr.

Photograph: Frei-Marek (200:120). My drawing has been obtained directly from this
photograph.

15. Krya (C.Kr)

C.Kr 1 (= D 15)

C-Kr 1

→
qot2omu sdisa-
s? n≤ “odubr≤ or rather: mn≤ “odubr≤?
sb mno≤ knor
noril?amsor rather: norimams?

The single inscription from Krya (the modern Ta{yaka) in the gulf of
Telmessos, whose alphabet is clearly similar to the Kaunian variety, is
not free from reading problems: there are several drawings of the inscrip-
tion, yet none of these coincides exactly each the others. Sayce
(1887[1892]: pl. III) reproduced two divergent drawings, by von Hammer-
Purgstall and by Forbes and Hoskyns. A further drawing was published
by Cecil Smith in 1888. Kalinka’s edition in the Tituli Linguae Lyciae
offered a new drawing and reading, made in collaboration with Heberdey
(Kalinka 1901:93). Deroy also included a new drawing based on two
photographs sent by G. E. Bean (Deroy 1955). Finally, Gusmani (1990)
 THE INSCRIPTIONS 159

produced a revision of the inscription, following the autopsy made by

Roos. Gusmani’s reading was used in Adiego (1993a), although the
drawing included there was a reproduction of Deroy’s copy (Gusmani
did not publish any drawing in his paper).
From these various readings and the photograph published by Deroy,
and if we disregard several old reading errors that have since been
superseded, it is clear that controversy still exists regarding two letters:
the first letter of the second line, and the fifth letter of the last line.
In the first case, Kalinka’s and Deroy’s readings, which both propose
a letter M m, have been challenged by Gusmani’s suggestion that this
letter could in fact be s s. As for the second problematic letter,
Gusmani, following Roos, proposes M m instead of Deroy’s l l (Hammer-
Purgstall, Smith, and Kalinka’s readings B can be ruled out, as this
kind of letter is alien to the Kaunian alphabetic variety, to which this
inscription belongs).
The drawing I offer here is based directly on the photograph of the
calque published by Deroy (1950). From this photograph it is impos-
sible to tell if the first letter of the second line is M, as it would seem
to be, or actually a broken s, as Gusmani suggests. This latter read-
ing has the advantage that it creates the sequence sdis as, parallel to
sdis as in C.Ka 1 (see p. 291 for an interpretation of this sequence).
For the other problematic letter, the drawing points to l, but the
horizontal line could be an intrusive one. The most prudent solution
is to offer a reading that leaves open the alternative readings for both
letters.

Photograph: Deroy (1955:pl. II). Drawing: Deroy (1955:320).

16. Inscriptions of Unknown Origin, Presumably from Caria (C.xx)

I group together under this title diverse inscriptions whose common
characteristics are their unknown origins and the fact that the object
on which they are incised points to a Greek-Carian background and
not, as in the case of the “pharaonic objects”, to an Egyptian one.
160 CHAPTER THREE

C.xx 1 (= 34*)

C.xx 1

←
“rquq | qtblem≤ | ÿbt | snn | orkn | ntro | pjdl

Text on a bronze phiale published by Gusmani (1978). Some doubts

have arisen about the letter À, which Gusmani interpreted as a, but
which is in fact l (as shown by the clear onomastic identification, qtblem≤
= Kutbelhmiw). Also concerning the last word, a reading pjdl must be
preferred over a: from Gusmani’s photographs, it is beyond doubt that
the final letter is the same as the third letter of qtblem≤.

Photographs: Gusmani (1978: Pl. I, II). Drawing: Gusmani (1978:69). My two draw-
ings showing the part of the phiale with the inscription have been made on the photo-
graphs of Gusmani.
 THE INSCRIPTIONS 161

C.xx 2 (=35*)

C.xx 2

→
ÿ≤biks not: alosd ∞arnosd: jzpe mdane

This inscription, on a bronze dinos, was published by Gusmani together

with the previous text (Gusmani 1978). There are no reading prob-
lems. The segmentation of the first sequence as ÿ≤biks not is based on
my own analysis of the text (see below p. 284).

Photographs: Gusmani (1978: pl. III and IV). Drawing: Gusmani (1978:71). My draw-
ing has been made by combining the two photographs of Gusmani (1978).
162 CHAPTER THREE

C.xx 3 (= 40*)

C.xx 3

←
akymyduÿeryly[vacat]d

An inscription on a cult object conserved in Geneva, edited by Meier-

Brügger (1994:112–113; photograph and transcription). I adopt Schürr’s
proposed reading (in Schürr 2001c).

Photograph: Meier-Brügger (1994:112). Drawing: Schürr 2001c:118). My drawing is

based on the photograph published by Meier-Brügger.

Meier-Brügger (1994:113) suggests that C.xx1, C.xx 2, C.xx 3 and also

C.Ha 1, all of which have surfaced in the European antiques trade in
recent years, come from the same Carian location, perhaps a sanctu-
ary of the god *Natr- (or rather ntro- = Apollo according to the Lycian
trilingual, see Chapter 11, s. v. ntro for further details), from which they
may have been stolen.

C.xx 4 (= 41*)

C.xx 4 (Zalhaas-Neumann 1994)

 THE INSCRIPTIONS 163

→
kdu≤ol“

C. xx 5 (= 41*)

C.xx 5 (Zalhaas-Neumann 1994)

→
kdu≤ol“

Regularized form of the letters of both inscriptions:

(Zalhaas-Neumann 1994)

C.xx 4 and C.xx 5 are two identical inscriptions on two bracelets pub-
lished by Zalhaas-Neumann (1994). Although the origin of both objects,
conserved in Munich’s Prähistorische Staatssamlung, is unknown, their
inscriptions are not only clearly Carian, but even display manifest con-
nections with the Kaunian alphabetic variant: note the form Z of z
≤ and particularly the presence of the letter / “, characteristic in this
alphabet (see already Zalhaas-Neumann 1994:166). Note however the
use of l l instead of l (L, Kaunos 2) in ºu≤ol (vs. u≤ol in other places).
There are no reading problems.

Photographs: Zalhaas-Neumann (1994: between 166–167). Drawings: Zalhaas-Neumann

(1994:161–161, 164).
164 CHAPTER THREE

E. THE CARIAN INSCRIPTIONS FROM GREECE

G 1 (= D 16, Athens)

G 1

→
seÇma ma tÒde: Tur[
KarÚw toÇ SkÊl[akow]
≤jas: san tur
[ÉA]ristokleÇw §p[o¤e-]

This is the well-known Athenian bilingual, dated ca. 525/520 B.C. I

maintain the reading adopted in Adiego (1993a), confirmed by my own
revision of the inscription.

Photograph: Masson (1977:between 90–91). Drawing: Deroy (1955:335). My drawing

has been made on my own photographs.

G 2 (= 42*, Thessaloniki)

G 2 (Tzanavari-Christidis (1995)
 THE INSCRIPTIONS 165

→
qlali≤ | k?[

This inscription was found in Thessaloniki and edited by Tzanavari

and Christidis (Tzanavari-Christidis 1995). It appears on a fragment of
a skyphos. The vase is dated by Tzanavari to the third quarter of the
5th century B.C. The Carian character of the inscription is indisputable.

Photograph: Tzanavari-Christidis (1995:pl. I). Drawing: Tzanavari-Christidis (1995:14).

 CHAPTER FOUR

THE HISTORY OF THE DECIPHERMENT

The history of the research to decipher the Carian Alphabet can be

divided for convenience into three periods:1

1. The ‘semisyllabic era’

2. The ‘alphabetic Greek era’
3. The ‘Egyptian approach’ and the definitive decipherment of Carian

It must be said that only the third period constitutes the true history
of the Carian decipherment: until the pioneering work of Zauzich (1972),
who was the first to correctly identify the solution—although his approach
was unsuccessful—practically all the previous efforts had produced no
results. This does not mean, however, that all the work preceding
Zauzich is worthless: for example, all the studies devoted to clarifying
the inventory of Carian letters—and in this sense, the roles played by
Olivier Masson or Vitali ”evoro“kin should not be dismissed—must be
considered important contributions to the decipherment of Carian
Indeed, some intuitions have proved to be right, as for instance Steinherr’s
interpretation of the bilingual coin of Erbbina (see below), although it
is true that cases such as this are the exception.

A. THE ‘SEMISYLLABIC ERA’ (1887–1962)

The ‘semisyllabic era’ was dominated by the figures of Sayce and Bork,
and lasted until the 1960s. As a conventional date for its end, I use
1962, the year ”evoro“kin’s influential article appeared in RHA (”evoro“kin

1
This periodization is different from that adopted in Adiego (1993a). There, it was
more important to separate the previous unsucessful or incomplete efforts (including
Zauzich and Ray) from the principal goal of the book: to present a complete deci-
pherment of the Carian alphabet.
 THE HISTORY OF THE DECIPHERMENT 167

1962). It is true that the foundations of the alleged Carian semisyl-

labism had already been questioned at the very beginning of the 1950s,
when Theodor BOSSERT observed that the then recently discovered long
inscription of Kaunos pointed to a purely alphabetic system, but it is
also true that the semisyllabic transcription was still used (indeed in a
conventional way) by Masson (1959b).
To Archibald H. SAYCE we owe, among many other things, the first
paper to offer an overall table of sound values for Carian letters (Sayce
1887[92]).2
In order to decipher Carian, Sayce started with the mixed origin of
the Carian writing system: he stated that, on the one hand, a great
number of letters were of Greek-Phoenician origin, whereas some signs
came from what he called an “Old Asianic Syllabary”, traces of which
would also remain in the Lycian alphabet and whose last vestige would
be the Cypriot syllabary:
It is clear at first sight that the main part of the letters is derived from
the Phoenicio-Greek Alphabet, but that, as in the case of the Lykian
alphabet, certain other characters have been added to express sounds
which were unrepresented in the Greek. Now Dr. Deecke, Dr. Isaac
Taylor and myself have pointed out that these additional characters
have in the case of Lykian been taken from the old Asianic syllabary,
a local form of which continued to be used in Cyprus down to a late
date. A probability therefore arises that the additional characters in the
Karian inscriptions also come from the same source (Sayce 1887[92]:
128).

2
Sayce had devoted a previous article to Carian (Sayce 1874), but Sayce (1887[92])
is the first (and last) complete study of the Carian alphabet by the British scholar.
168 CHAPTER FOUR

The following table reproduces that of Sayce (1887[92]):

Sayce (1887[92])
 THE HISTORY OF THE DECIPHERMENT 169

Therefore, whereas a r W m, for instance, come from a Greek model,

the letter & has for Sayce the same origin and value as the equivalent
sign in the Cypriot Syllabary. A dental value is assigned to 1 (dh), only
on the basis of its similar value in Lycian, assuming that both letters
have the same origin.
In one case, Sayce was forced to accept that a sign coming from
that alleged syllabary was used in Carian for a sound existing in Greek:
the sign p, which he transcribed as mi, me, m on the basis of the value
mi of a similar sign in the Cypriot syllabary. This anomaly is justified
by the concurrence of the letter s with s.
Sayce opens a series of interpretations of letters that will be contin-
ued by many other scholars: the differentiation between m and M and
the vocalic value of t are also present in ”evoro“kin’s decipherment.
The vocalic character of e and w and the dental character of 1, also
defended by ”evoro“kin, are nowadays confirmed, following the definitive
decipherment of Carian. Yet this does not mean that the arguments
then used were as compelling as the present ones, and indeed it is clear
that Sayce’s decipherment, like all theories prior to the Egyptian approach,
was wrong.
It is worth noting that Sayce’s system was not exactly semisyllabic.
For Sayce, the syllabic component of Carian writing is in fact of a his-
torical character: the Carian alphabet comes in part from a syllabary,
but it is not actually a syllabary or semisyllabary as such. Our label
“semisyllabic era” must therefore be interpreted in a broad sense. The
semisyllabic character of Carian writing was developed by Bork.
As for the Carian language, Sayce was commendably prudent.
Moreover, he was able to establish two basic and correct principles of
interpreting the texts: that the Carian inscriptions of Egypt must con-
tain essentially person names, and that -z ending was of genitival char-
acter. However, Sundwall (1911) was very much to the point in criticizing
the failure of Sayce’s decipherment to offer forms comparable to the
Carian onomastics known from the Greek sources. The possible equiv-
alences are scarce and—we are now able to say—ill founded: †Ü-z-â-
kho-e = Osogva (E.xx 3, now reads in the opposite direction: ionel),
†‡L-e-le-kh-ä = Lelegians (mythical inhabitants of Caria) (E.AS 7: bebint).
The transcription system of Sayce (1887[92]) is also used in his subse-
quent contributions to Carian, although there are doubts, modifications
(the clearest ones regarding vocalism, in Sayce 1906) and some incongruities.
170 CHAPTER FOUR

As for the connection of Carian to other languages, although Sayce

assumes provisionally that Carian belongs to the “Aryan family” (Sayce
1887[92]:117), this hypothesis is not based on data from his decipher-
ment, but merely on a combination of his own impression that Lydian
was an Indo-European Language, and the notices given by Herodot of
a kinship between Lydians and Carians.
Sayce’s decipherment was only used until the appearance of Bork’s
articles (Bork 1930, 1931). Yet despite the failure of his decipherment,
and the dilettantism of many of his proposals, Sayce deserves recognition
as the founder of Carian studies: he localized and edited all the texts
then available and established the basis for a scientific study of Carian.
Although they cannot be considered decipherments per se, since they
are merely contributions to the line of research begun by Sayce, some
other works must be mentioned: KRETSCHMER (1896:377–384), where
an o-value for z was proposed, and where the typical final sequence
-xi, read †hä (now ∞i ) was compared with Lycian genitival ending -he ;
TORP (1903:43–50), wherein Sayce’s system was adopted in an attempt
to find a sigmatic nominative in Carian; and SUNDWALL (1911), who
formulated several interesting ideas, for instance, that e was different
from k, and w from K, and that k and K were mere variants of one
sign. He also tried to analyze the Carian inscriptions in the wider back-
ground of Minor Asian onomastics.
In Torp (1903:44) a value i for t is proposed for the first time, an idea
later repeated by ”evoro“kin.
The next decipherment was proposed by F. BORK, who in two papers
(Bork 1930, Bork 1931) laid the foundations of the semisyllabic theory
(the so-called Bork-Friedrich system).
In Bork (1930) the proposed decipherment differed from Sayce’s in
two basic points:

a) A clear-cut distinction between alphabetic signs and syllabic signs.

Moreover, the latter have the same value as the (allegedly) corre-
sponding signs of the Cypriot syllabary. Needless to say, in most
cases, the similarities between Cypriot signs and Carian were extremely
questionable, both formally and phonetically.
b) A fixation of the sound values for the Carian signs more precise
than that of Sayce, based on the assumption that the sound inven-
tory of Carian must be not very different from that of Lycian, given
their common origin (according to Bork) and their geographical
proximity.
 THE HISTORY OF THE DECIPHERMENT 171

This latter point could have offered interesting results—at least regard-
ing the criterion of geographic proximity, since it is well known that
nearby languages tend to share phonetic traits—, but Bork had a very
sui generis vision of Lycian phonetics, far removed from the communis
opinio. For both languages Bork assumed the existence of sounds like
/pf/ or /k’h’/ (this latter an affricate palatal), and in Carian he estab-
lished five series of consonants (labials, dentals, palatals, gutturals and
velars), with three articulation modes for each one ( fortes, affricates and
spirants). Bork’s transcriptions therefore appear complex, with an array
of aspirates mixed with syllabic signs. This is his decipherment system,
as appears in Bork (1930):

Bork (1930)
172 CHAPTER FOUR

This is, v. gr. his transcription of E.Sa 1:

R-a-v-ro-l.h’(e).on | No-qh-ro-s.n R(e)-a-kh’.ja | va-kh’-?-no | re.ja-kh’

| h(e)-u.vu.kh’

[“arkbiom zidks mdane yn[-?]mo den tumn]

Bork (1931) focused on the linguistic analysis of Carian. The gram-

matical elements that Bork believed to have recognized are not wor-
thy of consideration, as his analyses are totally arbitrary. Similarly, the
meanings he attributes to the words are capricious. Finally, the typical
Carian onomastics known from Greek sources do not appear anywhere
in his study. To quote a few examples, in the aforementioned inscrip-
tion E.Sa 1, Bork claimed to recognize two Carian names hardly rec-
oncilable with the indirect sources, Ravrol and Noqhros. Other “Carian
personal names” were forms like Kh’asja, Ravuvoc, or Rejajape, equally
aberrant.
Bork (1931) offers the first attempt to situate linguistically the Carian
language based on a particular decipherment. According to Bork, Carian
would belong to an “Old Caucasian” family that would also include
Sumerian, Elamite, “Mitani” (nowadays known as Hurrian), Lycian and
“Alasian” (the language of the Cypriot Minoic inscriptions). At present,
none of these languages is believed to have any connection with the
others: Sumerian is an isolated language, Elamite has been linked—
very tentatively, however—to the Dravidian family, Hurrian forms an
independent group together with Urartian, Lycian is an Anatolian Indo-
European dialect, and “Alasian” is yet to be deciphered. Needless to
say, all these speculations, based on an invalid decipherment and a
nonexistent linguistic family, have been totally superseded.
The relative success of an overly unscientific decipherment must be
attributed to the authority of J. FRIEDRICH and W. BRANDENSTEIN.
Friedrich used it in his important work Kleinasiatische Sprachdenkmäler
(Friedrich 1932) for transcribing the Carian inscriptions. However,
Friedrich himself recognized that Carian could still not be seen as com-
prehensively deciphered (Friedrich 1932:91) and uses a simplified tran-
scription that substitutes <b> for <pf>, <k’> for <k’h’>, etc., which
in fact means destroying Bork’s complex speculations about the sound
system of Carian:
THE HISTORY OF THE DECIPHERMENT 173

Friedrich (1932)
174 CHAPTER FOUR

The so-called Bork-Friedrich system was adopted in a purely instrumen-

tal way by Masson in his editions of Carian texts until 1976.
Brandenstein also adopted Bork’s system—although with some modi-
fications—in his article “Karische Sprache” for the Pauly-Wissowa’s
Realencyclopädie (Brandenstein 1935a): he not only took on the simpli-
fications of Friedrich (1932), but also recovered two values proposed
by Sayce: t = ä (Bork he) and i = e (Bork h’e, Friedrich he). His pro-
posal of attributing a value p to x X is new, and based on the alleged
labial value of X in Lydian (at present, this Lydian letter is in fact com-
monly interpreted as representing a labiovelar [transcribed by q]).3 This
proposal, put forward by Brandenstein in a former paper (Brandenstein
1934b) was unsuccessful:

Brandenstein (1934b)

3
See now Gérard (2005:56–57).
 THE HISTORY OF THE DECIPHERMENT 175

Brandenstein made some efforts to find Carian personal names in the

inscriptions: in E.Ab 16, read as †k-mi-v-o-s-k-vo (now nprosn≤ ), he believed
he had identified the final element of the Carian names Se-skvw, Sa-
skvw. In other cases, he included Bork’s interpretations. In general,
Brandenstein’s contributions are scarce and not particularly convincing.
After Bork’s decipherment and until the end of 50s, I know of only
one attempt to decipher Carian letters: that of A. MENTZ (Mentz 1940).
Its only interest lies in being the first work to employ the Carian-
Egyptian bilingual inscriptions in order to establish sound values for
the Carian signs, and in having ruled out the alleged semisyllabic char-
acter of Carian writing.4 Otherwise, Mentz’s decipherment is an exer-
cise in dilettantism, and has received harsh, but merited, criticism: “ganz
dilettantisch in Lesung und Deutung” (Friedrich, quoted by Masson
1973:207); “tentative ambitieuse (. . .) qui n’a pas convaincu” (Masson
ibid.); “It is difficult to imagine that there can exist works on Carian
language more prone to fantasizing than Bork’s ‘researches’. Mentz
showed that it was indeed possible” (”evoro“kin 1965:51; my translation).

Mentz (1940)

4
“Die entscheidende Erkenntnis ist, daß das karische Alphabet keine Mischung von
176 CHAPTER FOUR

Mentz’s use of bilingual inscriptions is very surprising, as he tries to

find Egyptian verbs and nouns in the Carian parts. The resulting Carian
is a somewhat bizarre, mixed language. In his translation, the Carian
short graffiti turn out to be of a pleading tone: “Râ, gib”, “Râ Amon,
gib Leben”, “Hathor, gib”. Such demonstrations of Carian devotion to
Egyptian deities may indeed be praiseworthy, but when they are based
only on an analysis like E.Si 11 = †‡pn–obpya “Apis Ptah”, they are
highly suspect.

B. THE ‘GREEK ALPHABETIC’ ERA

In 1949, George E. Bean found an inscription of Kaunos (C.Ka 2)

that remains to this day the longest Carian text conserved. Almost
immediately, Th. BOSSERT (apud Steinherr 1950–51:332) rightly observed
that this long inscription, where less than 30 different letters were used,
pointed clearly to a pure alphabetic system: “Nach dem Zeichenbestand
der Kaunos-Inschrift zu urteilen kann von einer Mischung von Alphabet-
und Silbenscrhrift nicht die Rede sein. Es handelt sich um eine reine
Buchstabenschrift (. . .)”. It is true, as we have seen above, that Mentz
was in fact the first to discard the semisyllabic approach and to defend
the purely alphabetic nature of the Carian writing system, but his only
argument was his decipherment itself, which we have already established
as being even worse than Bork’s (if this is possible). In any case, it was
Bossert’s well-reasoned contribution that put an end to the alleged semi-
syllabism of Carian. In general, the proposals of decipherment of this
new period concur in their assumption that the Carian writing system
is alphabetic, and in attributing Greek values to letters of Greek shape.5
Between Mentz’s and Stoltenberg’s (see below) decipherments, it is
also fair to mention the contributions of F. STEINHERR (Steinherr 1950–51,
1955). Steinherr did not offer a concrete decipherment, rather he lim-
ited himself to formulating remarks on certain Carian signs. His obser-
vations are generally very thought-provoking and, in some cases,
surprisingly accurate. Among them, the following four must be highlighted:

Laut- und Silbenschrift ist, wie man seit Sayce allgemein annahm, sondern ein rein
Lautschrift.” (Mentz 1940:279)
5
The sole exception is Stoltenberg’s decipherment, where syllabic values are assigned
to a few letters in a rather capricious way. But even he refuses any connection to
Cypriot syllabary, see below.
 THE HISTORY OF THE DECIPHERMENT 177

1. Steinherr (1950–51:336) proposed a liquid value for the sign z resort-

ing, among other arguments, to the comparison of the “genitival”
endings in -z with the oblique case in -l of Lydian. This proposal
anticipated an idea developed later by ”evoro“kin, but which has
since been superseded.
2. Steinherr suggested that j had to have a value near to i, because
both alternate in some inscriptions. Consequently, he proposed a
transcription ê / e, respectively. This observation, for many years
ignored (a nasal-like value for j was for a long time the preferred
theory), has since proved to be correct, now that the Carian alpha-
bet has been definitively deciphered (i = i, j = j ).
3. Steinherr pointed out that the sign f F (and variants) could hardly
have an r value, because it often appears in an initial position,
whereas Anatolian languages show a clear reluctance to an initial r,
a well-founded argument reiterated by Ray (1987).
4. For this latter sign, Steinherr suggested a t value. We now know
that he was wrong, but only partially: the sign has in general an “
value, but the bilingual inscription of Kaunos has revealed that in
the Kaunian alphabet the letter T was actually used for t. Moreover,
in at least one case his analysis was right: he proposed that the let-
ters i F in the bilingual coin of Erbbina, dynast of Telmessos, read
t e, could correspond to the Lycian legend teleb(ehi) erbbina ‘Telmessos
/ Erbbina’ in other coins (Steinherr 1955:184–192). Similar rea-
soning is nowadays used to explain the Carian legend i F, by assum-
ing that the alphabet used is of Kaunian type (a very plausible
assumption from a geographical point of view), see Adiego (1998b:
58–60), Meier-Brügger (1998:45).

Steinherr’s work can hardly be considered a decipherment, but it does

deserve a great deal of respect: before the “Egyptian approach”, this
was the only moment in the history of decipherment that a correct
method of analyzing the Carian alphabet had been envisaged, even if
it was in a very limited way.
The fifties ended with another failed attempt at decipherment, by
H. L. STOLTENBERG (Stoltenberg 1958a, b, 1959).
Stoltenberg started with the automatic attribution of Greek values to
the allegedly Greek letters: a = a, o = o, m = n, p = m, and so on.
As for the apparently non-Greek signs, he rejected any comparison with
the Cypriot syllabary and proposed as an alternative to resort to the
Minor Asian onomastics of Greek sources, and the comparison with
178 CHAPTER FOUR

‘Termilian’ (= Lycian). This latter hypothesis would in principle be

acceptable, were it not that that Stoltenberg, like Bork, had a very
peculiar vision of Lycian: he believed that Lycian belonged to a group
that he called ‘Laric’ together with Etruscan, ‘Tyrsenian’ (= Lemnian,
an Etruscan dialect of the Lemnos island) and the language of the
“Cretan pelasgians”. Moreover, this alleged ‘Laric’ group would be con-
nected with Uralic languages. This classification of Lycian is clearly
wrong, and most of the meanings for Lycian words that Stoltenberg
proposed are also completely incorrect (see Neumann 1969b:364).
If this approach does not promise good results, any hope of success
vanishes when the values attributed to the ‘non-Greek’ letters are taken
into account: let us examine the following table that contains his deci-
pherment system (Stoltenberg 1958a):

Stoltenberg (1958a)
 THE HISTORY OF THE DECIPHERMENT 179

By attributing values such as ija, nda, and öm to some letters, it is

easy to obtain words of a falsely Anatolian flavour. For instance,
‡1wmuip ← ‡piumz, now read piubez (E.Ab 10), transcribed as †‡m-
e-u-n-uwa-nda can be easily compared with the typical Anatolian names
in -anda, -wanda. However, even resorting to these tricks Stoltenberg
was unable to obtain convincing results.
It is worth remembering that during the fifties the Carian corpus
underwent a considerable improvement. It was during this period that
the complete corpus of Carian inscriptions of Caria to date was pub-
lished by Robert (1950) and Deroy (1955), Masson and Yoyotte col-
lected the inscriptions on pharaonic objects (Masson-Yoyotte 1956), and
a provisory edition of graffiti of Abu-Simbel appeared (Bernand-Aly
1959?). Moreover, the important Greek-Carian bilingual stela of Athens
was discovered, as well as graffiti in Thebes.
At the beginning of the sixties appeared the first works of Vitali
”EVORO“KIN, one of the most prominent Carian scholars, and the author
of the first book devoted exclusively to the decipherment of Carian
(”evoro“kin 1965).
”evoro“kin’s method offers some innovative elements. In order to
separate vowels from consonants, he resorts to typology: given that the
languages combine vowels and consonants in different ways, which im-
plies different patterns of sequences such as CVC. VCV, etc., ”evoro“kin
tries to establish the patterns of Carian on the basis of personal and
place names from Greek sources. Remarkably, the results are not far
removed from the distribution of vowels and consonants in Sayce’s
decipherment. The explanation for these rather surprising results is that
Sayce himself took into account the distributive properties of the signs,
albeit in a rudimentary way.
After establishing a first separation between vowels and consonants,
”evoro“kin moves on to the statistical comparison between Carian signs
and the phonemes present in the names of Greek transmission. However,
in some cases this criterion seems to be used merely to confirm a value
obtained by other means. These other methods are the formal anal-
ogy with Greek letters and the identification of onomastic elements in
Carian words.
The following tables reflect the system of ”evoro“kin as it appears
in ”evoro“kin (1965) and thirty years later, in ”evoro“kin 1994, respectively:
180 CHAPTER FOUR

”evoro“kin (1965)
THE HISTORY OF THE DECIPHERMENT 181

”evoro“kin 1994
182 CHAPTER FOUR

It is clear that ”evoro“kin follows the traditional equivalence between

Carian signs and Greek letters in a great number of cases. Regarding
the “non-Greek” signs, one also comes across values already outlined
by other scholars: t = i, suggested by Torp 1903; z = l (1965) / L
(1994), for which a liquid value was defended by Steinherr. A less exact
precedent is the value ù for e (Sayce ô ) or the dental character of c,
already suggested by Bork. However, other proposals are totally new:
j = ñ (based undoubtedly on the Lycian N ñ ); & = t (1965) / D
(1994), a dental sound, based primarily on the sequences m& that
”evoro“kin identified with the typically Minor Asian consonantal group
o
nd o); k K = p (1965), b (1994), based on statistical criteria and on the
comparison with names from indirect sources.
Careful examination of both tables shows that the changes intro-
duced by ”evoroskin during three decades of research on Carian were
of no great importance. Perhaps the most significant modification is v
= U, resulting from the observation of Memphis corpus, where it alter-
nates with u = u (formerly, ”evoro“kin took v as a sort of e). In the
remaining cases, the modifications are aimed at adjusting the tran-
scription to etymological explanations. This is the case, for instance, of
x X, transcribed as q (formerly h) in order to better reflect his ety-
mology of xi †qe [now ∞i ] < PIE *kwe, or of k K, previously tran-
scribed as p and more recently as b.
One of the most characteristic values that ”evoro“kin defended was
z = L (formerly transcribed as l). This was a not totally new idea (it
was previously considered by Steinherr 1950–51, see above), but in
”evoro“kin’s hands it took on greater importance. Using this value, he
believed he could recognize the Carian name Lujhw in zu0Li-, read
as †Luxze- [now ≤ugli-], comparing the genitival ending -z = †-L with
the genitival adjective ending -li- in Lydian, and he claimed to find
typical Carian names in -vllow/-vldow in some sequences uz = †uL
[now u≤ ]. Obviously, these proposals have since been proved erroneous,
as the sign has been definitively identified as representing a sibilant
sound.
However, the main problem with ”evoro“kin’s decipherment is that
the typical Carian names attested by indirect sources do not appear
anywhere in Carian inscriptions if we adopt his system of decipher-
ment. In this sense, a case such as Lujhw = †Luxze- is almost unique.
In most cases, the only means of establishing connections between
”evoro“kin’s readings and the stock of Anatolian names in Greek (and
Hittite) sources is to resort to partial comparisons of constituents, ety-
 THE HISTORY OF THE DECIPHERMENT 183

mologically reconstructed forms, complex sound changes, and so on.

Indeed, such techniques can be used to explain any form as Anatolian
and/or Indo-European. However, almost none of the typical Carian
names (for instance Ussvllow and related forms) are obtained with this
decipherment. Masson rightly pointed out this deficiency:
Contrairement à bien d’autres tentatives, celle de ”evoro“kin paraît raison-
nable: les discussions philologiques sont judicieuses et les comparaisons
linguistiques sont intéressantes, sinon toujours convaincantes. Peut-on alors
parler d’un véritablement déchiffrement? Il ne le paraît pas, car l’ensemble
des lectures ne donne pas une impression d’évidence, et les noms d’hommes
cariens (. . .) ne se retrouvent pas de manière tangible.
This inability to find recognizable Carian names could explain why, in
his works of the 80s and 90s, ”evoro“kin focused his efforts on the
search for nouns and verbs in Carian texts, paying far less attention
to the amount of onomastic formulae provided by the recently pub-
lished corpus of Memphis inscriptions. ”evoro“kin seemed to feel more
comfortable translating long and complex texts in a rather speculative
way than analyzing the abundant personal names in the laconic funer-
ary inscriptions from Saqqâra.
Nowadays, in the knowledge that ”evoro“kin’s decipherment was a
failed attempt, continuing critical analysis seems unnecessary, and it is
preferable to instead note some important merits of his work. The first,
although apparently negative, does in fact have a positive reading:
despite his methodological weakness, due above all to the abuse of
etymological and analytical speculations, ”evoro“kin’s study remains
the only serious effort to interpret the Carian alphabet as a ‘normal’
Anatolian alphabet, similar to Phrygian, Lydian and Lycian writing sys-
tems, wherein letters with Greek shape have their Greek or like-Greek
sound value, and other letters are added to reflect peculiar sounds not
found in Greek. After ”evoro“kin’s failure, this method became imprac-
ticable, and a different approach became necessary.
The other merits of his work are decidedly more positive: ”evoro“kin
contributed decisively to the demonstration that Carian writing is purely
alphabetic. He showed that the seemingly large number of letters was
due to the existence of alphabetic varieties, which he was able to sep-
arate correctly. Moreover, ”evoro“kin has always defended the suppo-
sition that Carian was an Anatolian Indo-European language, ruling
out the attempts of the dilettanti to connect it with other languages, and
emphasizing the need to include the other Anatolian languages in an
analysis of Carian forms.
184 CHAPTER FOUR

Reviews of ”evoro“kin 1965: Pisani 1967, Gusmani 1967 (laudatory,

although he believes that it is precipitate to consider the decipherment
as done), Zgusta 1968 (totally favourable: “We can then, conclude that . . . the
basic step is done and the Carian inscriptions are really deciphered”
(p. 154)).
In contrast to the predominantly serious approach adopted by ”evoro“kin
the so-called decipherment of R. SHAFER (Shafer 1965) can be taken
as an example of how a language must not be deciphered. He com-
bines a lack of rigor (different values assigned to the same sign, absence
of a table to display sound values, senseless hypotheses about the mean-
ing of the inscriptions . . .) and an arrogant dismissal of other scholars,
which explains the harsh but just opinion formulated by Masson: “ten-
tative vaine et très prétentieuse d’un renouvellement complet des lec-
tures et des interprétations” (Masson 1973:211).
As is often the case in other attempts to decipher Carian, the begin-
ning of Shafer’s work seems very promising, despite the somewhat inap-
propriate tone. He rightly criticizes Bork-Friedrich’s decipherment system,
the outcome of which is a language he derisively terms “Super-Hawaiian”
because of the profusion of vowels and semivowels, phonologically very
distant from the Carian names in Greek sources.
But the interest of his paper ends here. When trying to establish
sound values for Carian letters, the solutions he resorts to are routinely
bizarre: for instance, he attributes a strange value br to p, takes the
clear interpunction sign | as a letter i, and imagines surprising evolu-
tions of the letters, like L > R > r. As for the linguistic interpretation,
in the Egyptian-Carian bilinguals, he claims to identify stereotyped for-
mulae referring to pharaohs in the Carian part. In many cases, he even
reads the texts in the opposite direction to that commonly accepted.
One of his “interpretations” surpasses Mentz’s litanies:
†(q)-lolak qirmdun b(a)bu kimrda (E.Si 4)
According to Shafer, this means: “In the language of the Leleges qirm-
dun is spoken kimrda” or “The Leleges qirmdun pronounce kimrda”. It is
not easy to imagine the reason for which a Carian mercenary would
feel the need to write a sentence such as this, which instead seems to
belong to a sort of Carian Appendix Probi !
This period of the history of Carian decipherment ends with another
somewhat superficial effort, that of Ju. V. OTKUP“‘IKOV (Otkup“‘ikov
1966; cf also Otkup“‘ikov 1968, on the origin of the Carian alphabet).
His proposal is shown below:
THE HISTORY OF THE DECIPHERMENT 185

Otkup“‘ikov (1966)
186 CHAPTER FOUR

These are the more notable elements of his decipherment:

1. Otkup“‘ikov assumes that some signs have two forms, one frequent
(‘asto), and other more infrequent or rare (redko). For instance, B
would be the infrequent form corresponding to the frequent a (!).
2. By this logic, Otkup“‘ikov converts the Carian alphabet into a pure
Greek alphabet, wherein the apparently non-Greek letters are sub-
sumed beneath the Greek ones. The values attributed are in gen-
eral also the Greek ones, note for example j = ks.
3. In fact, for Otkup“‘ikov, the Carian language is merely a Greek dialect.

None of these three assumptions are convincingly argued. The group-

ings are largely unjustified and absurd (for example k = u), and
Otkup“‘ikov arbitrarily assigns different values to a single letter: for
instance, q is both a frequent letter with value th and a redko letter
equivalent to o, and similarly in the case of B (=l / a).
As for the supposed Greek character of Carian, the dialect result-
ing from this decipherment proves to be rather curious: ue uh would
mean ‘daughter’ and would be connected with the Greek uflÒw ‘son’(!)
(Otkup“‘ikov 1966:22); paraeWm, read SaWaieien would be an opta-
tive (Otkup“‘ikov 1966:24) and so on. It is not difficult to agree with
Masson’s judgement: “on aura peine à conclure autrement que par un
scepticisme total devant la langue qui nous est proposée . . .” (Masson
1973:193).
Reviews of Otkup“‘ikov (1966): O. Masson (1967), Heubeck (1967–68),
Jordan (1968), all adverse.
Before concluding this section, it is worth mentioning the contributions
made to Carian studies by the Italian scholars Piero MERIGGI and
Roberto GUSMANI. Their presence here can be justified by the fact that,
in their approach to the language, both scholars broadly adopted the
‘Greek alphabetic’ interpretation of the Carian writing system.
In the 60s, Meriggi published two papers on Carian (Meriggi 1966,
1967). The first, though devoted to the study of Para-Carian tablets,
contains very useful inventories of signs. The second consists of a series
of remarks regarding several inscriptions from Caria itself and certain
Carian letters. Although Meriggi did not offer a decipherment as such,
he assumed implicitly that the “Greek” letters had Greek sound val-
ues. In the 70s, Meriggi dedicated two of his last works to Carian:
Meriggi (1978) adds little new material to Meriggi (1967), aside from
 THE HISTORY OF THE DECIPHERMENT 187

his argument in favour of the Luwian character of Carian based in a

process of elimination. More interesting is his review of Masson (1978
(Meriggi 1980), in which he leaves aside any discussion of the letter
values and aims instead to analyze the structure of the inscriptions of
Saqqâra, with some thought provoking hypotheses, such as the mean-
ing ‘son’ for mno-≤ (already suggested in Meriggi 1967:223), or the pos-
sibility that mwdon-≤ could be an ethnic (‘Caromemphite’).
Gusmani’s first approach to the study of Carian—other than his
review of ”evoro“kin (1965), Gusmani 1967—was his edition of the
Carian graffiti from Sardis (Gusmani 1975), wherein he adopts the con-
ventions of ”evoro“kin’s system (except in a few cases, such as j = †≤,
not †ñ). In any case, the very limited linguistic usefulness of these graffiti
does not allow Gusmani to enter into discussions about the sound val-
ues of the signs. Similarly, no linguistic discussion is present in Gusmani
(1978), which contains two Carian inscriptions: Gusmani adopts Masson’s
system of transcription, with the sole difference of & = †≥ (= Masson
“31”).
In fact, the very first work by Gusmani devoted specifically to Carian
decipherment itself is Gusmani (1979). Here he focuses on the letters
2, t, i, e, K and j. The most valuable contribution is the observa-
tion that t alternates with d, which invalidates ”evoro“kin’s theory of
a vocalic value for t. Less successful is his attempt to defend a simi-
lar alternation between e and o, leading to a transcription of e by
means of †ò. Other suggestions are equally unconvincing, with the sole
exception of the alternation j / i. Most of the ideas Gusmani puts
forward in this article are no longer recognised, having been super-
seded by the definitive decipherment.

C. THE ‘EGYPTIAN APPROACH’

1. The First Attempts

The Egyptologist K.-Th. ZAUZICH deserves to be considered as the pio-
neer of the so-called ‘Egyptian approach’: his use of Egyptian-Carian
bilinguals as the starting point to decipher the Carian alphabet is an
approach that has proved successful.
From a historical perspective, it is interesting that for one hundred
years (from Sayce’s first work to Zauzich’s monography), with the sole
188 CHAPTER FOUR

and frustrating exception of Mentz, nobody considered the possibility

of using the bilingual inscriptions to decipher Carian, despite this being
the most customary way of focusing a decipherment.
The flaw in Zauzich’s attempt is its inability to establish a clear con-
nection between Greek and Carian parts of the bilingual inscriptions.
Practically the only correct value that he discovered was p = p, which
appears at the beginning of three names in both the Carian and Greek
part of three bilinguals (E.Sa 2. E.Me 5 and E.Me 8).6 Other than this
equivalence, both the methods and the results are very disappointing:
he corrects Carian texts whose readings are not in doubt, offers uncon-
vincing equivalences, like Prjm = †patalem [now paraeym], assigns the
same value to different letters, as b e = l, z v = s, and, finally, con-
cludes that Carian is a Greek dialect, rendering his study as defective
as Otkup“‘ikov’s (see above). Consequently his decipherment received
fierce criticism from serious scholars like Masson and Heubeck.
Reviews of Zauzich (1972), both very critical: Masson (1973[74]), Heubeck
(1974:97 col. 1): “Die Entzifferung Z[auzich]s hat uns . . . keinen Schritt
weiter gebracht”.
The way opened by Zauzich (1972) was followed by Thomas KOWALSKI,
with more promising results (Kowalski 1975). The most striking ele-
ment of Kowalski’s only paper devoted to Carian is that he sets most
of the values later established by J. D. Ray, but he does so in an
unusual way, often starting from weak or even erroneous suppositions.
Moreover, his article gives the impression of being a condensed ver-
sion of a more extensive study on Carian that was never published:
nowhere does it attempt to justify, for example, the grouping together
of clearly different signs or the separation of mere variants.

6
Zauzich also believed that a Carian name beginning with p could be connected
with an Egyptian name with an initial p in E.Me 6, an incorrect assumption, because
this is a case of a re-used stela, and indicates no correspondence between Egyptian
and Carian epigraphs. It is also true that Zauzich was right in assigning a sibilant
value to z, but this attribution was based on faulty interpretations.
THE HISTORY OF THE DECIPHERMENT 189

Kowalski (1975)
190 CHAPTER FOUR

I consider the main deficiencies of Kowalski’s proposal to be the

following:

1. Kowalski tried to obtain and correlate results from very dissimilar

inscriptions, treating Egyptian-Carian and Greek-Carian inscriptions
in the same way, even though connections are far more difficult to
establish within the latter type, increasing the risk of falling into haz-
ardous speculations. He also considered the inscription E.Me 6 truly
bilingual, when it is in fact a reused stela and indicates no connec-
tion between Carian and Egyptian names.
2. His transcription and interpretation of the frequent sequence xi as
† yrº ‘son’, clearly reminiscent of Zauzich (1972),7 was not only absurd
per se, but also seriously impeded the overall decipherment, because
it implies an rº value for i, a clearly vocalic letter.
3. It has already been mentioned that Kowalski’s article lacks a detailed
explanation of the treatment of certain Carian letters as mere vari-
ants, and vice versa. The above table demonstrates how inconsis-
tent these groupings and separations are: Kowalski separates N (=
t’ ) from ø (= h), L (= ź) from 2 (= ˙), and yY (= c) from i (=
rº ). Conversely, he puts together v and ç (= t). The most surpris-
ing case is z: two different values are offered: ¬, and Æ. In the first
case it is grouped with 9 and 6, and in the second, with ñ. All
these assumptions have turned out to be false.
4. Kowalski does not apply the values obtained from the bilingual
inscriptions to the rest of Carian texts, which would be the only
way to check the validity of the decipherment, particularly regard-
ing those signs whose values are dependent on the weak interpre-
tation of a single inscription.

To sum up, Kowalski’s decipherment—though it far exceeds Zauzich’s

attempt when Egyptian-Carian bilinguals are used, offering correct val-
ues for some important signs (p = p, k = k, r = r, m = m, b = b,
f = “ )—is fatally flawed by excessive speculation and its chaotic treat-
ment of signs and variants that inexplicably disregards the illuminating

7
= Zauzich ‡ir. However, Kowalski does not make reference to Zauzich’s inter-
pretation. Zauzich aimed to connect this form with Greek uflÒw, an assumption unten-
able from the point of view of both Greek dialectology and the structure of Carian
inscriptions.
 THE HISTORY OF THE DECIPHERMENT 191

work of ”evoro“kin and Masson. As a result, its usefulness is extremely

limited, working only for those words where the correctly deciphered
letters mentioned above appear almost exclusively, for instance in
farKbiom = †“arkbrºom [now “arkbiom], a not entirely correct tran-
scription—note rº—based on the Egyptian form ”3rkbym.
The ‘Egyptian approach’ took a brief step back with Jean FAUCOUNAU
(Faucounau 1980). In this article, the Egyptian-Carian bilinguals are
considered, but the overall attempt is rendered worthless by the author’s
customary dilettantism. Faucounau bases his work on two assumptions:
that Carian local alphabets are very dissimilar to each other, and that
they may reflect different dialects or even languages.
Both assumptions are untenable, at least in the form in which
Faucounau develops them: it is true that there are local alphabetic vari-
eties, with some peculiar forms in each case for certain signs, but
Faucounau proposes absurdly that letters such as f or t may have
different values in diverse inscriptions from Egypt, where the general
unity of the Carian alphabet is undeniable. As for the existence of
different languages, not only is there no evidence for this assertion, but
in fact the presence of similar words and endings in distant Carian
sub-corpora points in the opposite direction. The presumed existence
of different alphabetic systems and languages combined with a markedly
unscientific approach result in an continuously ad hoc assignation of
sound values to the letters and to an inconsistent decipherment. Faucounau
also attempts to identify proper names, but in many cases he resorts
to alleged “Carian” names transcribed in Greek (without any reference)
that do not appear in either Zgusta’s or other similar collections.

2. The Seminal Work of Ray

The ‘Egyptian approach’, as the definitive method for deciphering Carian,
begins in earnest with the fundamental work of British Egyptologist
John D. RAY. It must be pointed out that for his decipherment Ray
starts from scratch: it is clear that when he begins to work on Carian,
he is still unfamiliar with Kowalski’s article. Whilst his decipherment
of Carian cannot be considered complete, the definitive decipherment
could not have been produced—at first independently, and later jointly
by Diether Shürr and myself—without the seminal work of Ray, which
acted as its starting point.
Ray’s first work on Carian is his recension of Masson (1978), received
by the editor of the Journal of Egyptian Archaeology ( JEA) in 1980,
192 CHAPTER FOUR

but published three years later (Ray 1983). This brief article is impor-
tant for two reasons: here Ray already establishes a criterion for the
use of bilingual inscriptions, consisting in leaving aside those inscrip-
tions where the individual has an Egyptian name in the Egyptian part
and deciphering only those where the Egyptian part contains a non-
Egyptian, presumably Carian, name. The other contribution is his
hypothesis, now confirmed, that the name Psammetichus can be identified
in the Carian sequence of signs kfa«sip ←/ pismazk →,
which would suggest totally innovative values for allegedly ‘Greek’ signs
as p, m or f.
It is interesting to note that the first assertion is now known to be
partially false: Egyptian names from the Egyptian part of the bilingual
inscriptions do appear in the Carian part (E.Sa 2, E.Me 5). However,
in somewhat paradoxical fashion, this limitation imposed on the inven-
tory of bilingual inscriptions proved instrumental in triggering the first
dramatic steps towards a definitive decipherment. Apart from the two
inscriptions mentioned above, Ray also left out E.Me 6, in this case
with good reason: there is no correspondence between the names in
the Carian and Egyptian parts. So we can see that, unlike Kowalski,
Ray did not obtain incorrect sound values for his decipherment by
using an inappropriate bilingual inscription. Kowalski’s error was, as
we have seen, to include this inscription in his decipherment.
Ray (1981) offered the first complete proposal of decipherment, based
on the inscriptions of Saqqâra. His method consisted in bringing the
sound values obtained from the bilingual texts to the remaining inscrip-
tions, in order to find Carian proper names comparable to those pre-
served in Greek sources. This onomastic comparison as a form of
confirming and developing the decipherment was not new, as the his-
tory of Carian studies shows. The singularity of Ray’s approach to the
onomastics of indirect sources lies rather in the cautiousness with which
he operates: the onomastic identifications are not so numerous, but
some important similarities begin to appear. Another prudent decision
was the initial limitation to the Saqqâra corpus: almost all the bilin-
gual inscriptions used were from Saqqâra, where a very unitary and
standardized alphabet was used. This left little room for confusion
between letters and variants, a risk that the former followers of the
‘Egyptian approach’ did not take care to avoid. It must be added that
Ray was also much more rigorous than his predecessors in the use of
the Carian alphabet: unlike some, he did not distort the Carian sig-
nary in order to obtain a particular equivalence between forms.
 THE HISTORY OF THE DECIPHERMENT 193

In Ray (1982b), undoubtedly his most ambitious and important work,

numerous inscriptions are transcribed and briefly commented, not only
those from Saqqâra, but also the graffiti from Buhen, Abu-Simbel,
Silsilis and Abydos, as well as other inscriptions. Ray (1982a) is an arti-
cle focused rather on historical aspects and theoretical problems of deci-
pherment. In Ray (1987) he responds to criticisms formulated by Gusmani
(1986), by comparing the results of his own decipherment with those
gained from the traditional approach.

Ray (1982b)

The proposed sound values vary slightly from one article to another,
but they can be characterized in the following way:

1) A significant number of signs had been transcribed in a similar man-

ner by Kowalski. The coincidence is logical: these are the letters that
appear in the bilingual inscriptions used by both authors. However, in
contrast to Kowalski, Ray does not offer arbitrary groupings or sepa-
rations of signs.
Ray’s use of bilingual inscriptions is shown in the table of p. 194.
There were admittedly some problems of interpretation: the value of
l, that in Egyptian appears adapted by means of r, and which was
 194

Sound values established by John D. Ray on the basis of Carian-Egyptian bilinguals

Inscription Carian Egyptian Ray’s Current Resulting sound values

forms version transcription interpretation f =“ r = r k = k m = m p = p l = d s = s x = h c = ‘
(Ray 1982b)

E.Sa 1 farKbiom ”3rkbym “-a-r-k-b-e-o-m “arkbiom + + + +

E.Me 5 psmfk . . . Psm∆k . . . p-s-m-“-k-. . . psm“k . . . + + + +
E.Me 7 Camou T3-n-jm.w ‘-a-m-o-u tamou + +
E.Me 7 Canaiz T3[ ‘-a-k‘-a-e-≤ tanai≤ +
E.Me 8 paraeWm Prjm p-a-r-a-j-é-m paraeym + + +
E.Me 9, 15 arlifz Jr“3, Jwr“3 a-r-d-e-“-≤ arli“≤ + + +
E.Me 9 arlio [ Jry[m]3 a-r-d-ê[sic]-o-
[m-≤] arlio[m≤] + + +
E.Me 15 ursxlez Nrskr [sic; u-r-s-h-d-j-≤ urs∞le≤ + + + +
now: 3rskr]
CHAPTER FOUR
 THE HISTORY OF THE DECIPHERMENT 195

troublesome for Ray; the only partial analysis of the name in E.Me 5;
the problematic reading Nrskr (now read 3rskr); the imprecise analysis
of vocalism, hampered by the almost purely consonantal adaptation in
Egyptian. Yet most of the values proposed by Ray have in fact turned
out to be right. Note that a great number of the sound values were
established on the basis of more than one correspondence: m = m results
from its presence at the end of two names (”3rkbym, Prjm), and in the
middle of a third one (Ô3n-j.mw), f = “ is established thanks to E.Sa
1, E.M 5, E.Me 9, and E.Me 15, etc.
2) Other values are the result of graphical alternations in the inscrip-
tions of Saqqâra already noted by Masson (1978) and Meier-Brügger
(1979a): i / j, u / v, etc.
3) Curiously, two of the most important sound values proposed by
Ray were not obtained from the bilingual inscriptions: L = l (Ray
used ld to transcribe it) and z = ≤.
The first equivalence- totally new- is based on C.Ki 1(Kildara) †k-
e-ld-a[ (now kila[ ) interpreted as the Carian form of the place name
Kildara (following Kowalski’s suggestion, although he interpreted L as
†z), and the endings in -oL = †o-ld (now -ol), that Ray correctly com-
pared with the typical Carian names in -vllow from Greek sources.
The second equivalence (z = ≤ ) was already suggested in Meier-Brügger
(1979a)—as Ray pointed out—but based on weak argumentation, and
without exploiting the consequences. Ray’s basis for this equivalence
was initially only the interpretation of the name vzoL- as †ü-≤-o-ld-
(now w≤ol-), corresponding thus to the Carian name Ussvllow (Ray
1981:161), although it was subsequently noted in Ray (1982b:189) that
more convincing evidence could be found: the clear alternation between
z and f = ≤ in the Egyptian name pisma≤k / pisma≤k.
4) For the remaining signs, whose values he could not ascertain from
the bilinguals used, Ray generally kept to the traditional transcription
used by other scholars (”evoro“kin, Meriggi, Gusmani, Masson), id est,
that is based on the similarities with Greek alphabet: d = g, t = q, n
= k, k’ [= kt ?], q = t, even l = d, etc. This is also valid for vocalic
signs: given the consonantal character of the majority of Egyptian tran-
scriptions, Carian-Egyptian bilinguals were useless for establishing vocalic
values, and Ray chose to keep the traditional value e for i, despite
the evidence of onomastic identifications, that pointed clearly to i.
In Ray (1988), several inscriptions from Caria itself are analysed.
The most relevant contribution is the comparison of the final sequence
196 CHAPTER FOUR

in C.Hy 1, read †é-d-a-r-m-e-ld (now ylarmit) with the place name where
the inscription was found, ÑUllãrima.
Two further articles by Ray previous to the definitive decipherment should
be cited here: Ray (1990a) is an ambitious attempt to systematize the
results of his decipherment, but there are few new ideas. The most strik-
ing contribution is the discussion devoted to the sign n, in which Ray
mentions, and rejects, the proposal of a nasal value suggested to him pri-
vately by Schürr.
Ray (1990b) offers an interpretation of the initial line of C.Si 2, which
owes much to Schürr’s brilliant analysis, also privately communicated (see
below). However, in trying to adapt this to his own decipherment system,
Ray seriously distorts Schürr’s ideas and the final result is unconvincing.
The role played by John D. Ray in the decipherment of Carian is cru-
cial: for the first time, true Carian names emerge from the inscriptions,
although in some cases Ray’s transcription does not correspond exactly
to the Greek adaptation: †u-≤-o-ld = Ussvllow, †“-a-r-u-≤-o-ld =
Sarussvllow, †a-r-d-e-“ = Arlissiw. Even more importantly, a great
number of the values involved in these forms were established from
the independent evidence of Egyptian bilinguals and the alternation of
letters.
Yet despite these merits, Ray was unable to offer a complete deci-
pherment system. The values proposed for very important consonan-
tal signs were either wrong (d t q n), or inaccurate (l & X), while
others remained undeciphered. As for vowels, Ray’s transcriptions were
also imprecise, when not mistaken (i = †e, j = †ê, e = †j, W = †é),
to the extent that the Egyptian bilinguals were rendered useless given
the almost purely consonantal notation of Carian names. Only in the
case of a o u v and to some extent w, did his transcriptions seem
to more accurately reflect the actual value of the letters. The sound
system that arose from his proposal was also inadequate, the most
important gap being the absence of a letter for n, the basic nasal con-
sonant in all the sound systems of the world. As for the onomastic
identifications, although the spectacular comparisons cited above pointed
clearly to the accuracy of the Egyptian approach, they were counter-
balanced by the great number of personal names in the Carian inscrip-
tions left without suitable interpretation. Forms such as †ü-d-e-a-q,
†‡e-r-t-u-t, †‡t-q-b-a-j-m-, †“-a-ju-g-e-t-, †‡“-a-r-ü-k-ê-a-q-, †‡m-s-k’-o-r-e-,
(now wliat, “ !rquq, qtbl!em-, “aÿdiq-, “arwl !jat-, msnord-) and many others,
which are now easy to interpret from the point of view of Carian ono-
mastics by using the definitive system, remained obscure in Ray’s system.
 THE HISTORY OF THE DECIPHERMENT 197

The first effort to improve Ray’s dechipherment was made by

Faucounau (Faucounau 1984). But the typical dilettantism of this scholar
converts his work into a practically worthless attempt. The only inter-
esting suggestion was l = l (vs. Ray †d ) later confirmed more definitively.
Much more serious is the work of Frank KAMMERZELL, who devoted
his doctoral dissertation of 1990 to Carian (published with a Nachbemerkung
in 1993: Kammerzell 1993). However, his decipherment differs too lit-
tle from Ray’s: any differences—not in transcription, but in interpre-
tation—are found only in the vocalic system. The consonantal values
for letters are coincident with those already proposed by J. D. Ray.

3. The Definitive Decipherment (‘Ray-Schürr-Adiego System’)

In this history of the decipherment of the Carian alphabet, I must
speak of my own contributions to the subject. In my first work, pro-
duced in 1989 and published one year later (Adiego 1990), I concen-
trated on only two letters, l and n. For the first, new evidence was
presented in favour of an l value, following the suggestion made by
Faucounau. As for n, I tried to demonstrate that this sign was actu-
ally the Carian letter for n, a sound surprisingly absent in Ray’s sys-
tem (see above). In this later case, the evidence was exclusively based
on onomastic identifications that seemed convincing (for instance p-n-
u≤ol = pnu≤ol = Ponussvllow). Shortly after writing this brief article,
I decided to devote my doctoral dissertation to Carian, with the aim
of offering a complete analysis of all Carian materials available (Adiego
1990b). For the decipherment, the most important contribution was to
use as a further clue two Carian-Egyptian bilingual inscriptions disre-
garded by Ray because they contained true Egyptian names in the
Egyptian part. A more attentive study of these two inscriptions (E.Sa
2, E.Me 5) showed that the Carian parts did indeed reflect the same
Egyptian names present in the Egyptian ones. This lead to new val-
ues for very important letters: n, whose value n was now confirmed,
t, which reflected a t sound (transcribed t in Adiego 1990a, t already
in Adiego 1993a), and d, a d sound (d in Adiego 1990a, d already in
Adiego 1993a):
198 CHAPTER FOUR

E.Me 5 E.Sa 2

Carian form psmfkvneitz pdnejt

Egyptian form Psm“k-‘wj-Njt P3-dj-Njt
Ray p-s-m-“-k-ü-k‘-j-e-q-≤ p-g-k‘-j-e-q
Adiego (1993a) p-s-m-“-k-ú-n-e-i-t-≤ p-d-n-e-í-t
[now: psm“kwneit≤ ] [now: pdnejt]

The t value of t allowed me to analyse the Athens bilingual inscrip-

tion in a both straightforward and convincing way: read ≤ías : san tur[
(now transcribed ≤jas : san tur[ ), the third last letters could be connected
to the Carian name present in the Greek part, Tur[.
Another important improvement concerned the vocalic letters: i =
i (not †e), j = í (given its alternation with the former, not †ê [now j ]),
e = e (not † j ) and sound values close to /u/ for the letters w = w
(Ray † ju, but on a weak basis; now [ÿ]), W = ù (Ray †é [now y]). All
these values are now universally accepted.
As for other letters, a labial character was proposed for $ (m [now
b]), based on the good correspondence ( para)i$rel = Imbarhldow, and
a liquid value for 6.
A brief report on this new decipherment system was published in
Kadmos (Adiego 1992a).
An error of decipherment, regarding q, was rectified shortly after
the completion of the doctoral dissertation, in which a more traditional
value t was envisaged (which explains the use of t for transcribing t).
However, a more detailed analysis of E.Sa 2 brought me to the con-
clusion that the second name in the Egyptian part, K3rr, was also
identifiable in the Carian part, qwri-, so that a tectal value for q
was proposed. This value was also reinforced by convincing onomas-
tic identifications, such as quq = Gugow, or qtblem = Kotbelhmow, Kutbelhmiw.
The new value was presented in a lecture held in Cologne in 1992
and incorporated into my book Studia Carica (Adiego 1993a), wherein
a definitive decipherment of Carian was proposed.
THE HISTORY OF THE DECIPHERMENT 199

Adiego (1993a)
200 CHAPTER FOUR

As for the linguistic results obtainable by means of this decipher-

ment, a very prudent method was adopted: both in Adiego (1990a) and
in Adiego (1993a) the attention was focused exclusively on the most
evident grammatical information recognizable in the Carian texts: the
existence of an opposition Nominative -Ø / Genitive (or ‘Possessive’)
-≤. The clear connection with Hittite, Luwian, Lycian, Sidetic, and
Pisidian sigmatic genitives and possessives, and the Lycian, Sidetic, and
Pisidan zero nominatives was seen as a clue for the consideration Carian
as an Indo-European Anatolian language, more closely related to these
later dialects. This assertion was fully consistent with the indices observ-
able in Carian onomastics, as ”evoro“kin and others had already showed.
During these decisive years for the elaboration of my own deci-
pherment system (1989–1992), I was unaware that another scholar,
Diether SCHÜRR, had already obtained to the same results for the sound
values of t, d, q, n, i and j. As early as 1982, also inspired by the
works of Ray, Schürr had observed the significance of the bilinguals
E.Sa 2 and E.Me 6 for deciphering Carian, which explains the coin-
cidence of the values proposed for these letters. However, Schürr merely
communicated his theory privately to Ray, who rejected it, and after
this disappointing answer he became silent until the appearance of my
first works. This explains why his first contribution appeared after Adiego
(1992a) (Schürr 1992). The positive side was that the two studies had
reached very similar conclusions independently of each other, which
served to reinforce the validity of the decipherment.
However, Schürr’s decipherment was not totally coincident with my
own: the main difference lay in the values attributed to vocalic signs.
While he agreed with i = i, and j = í [now j ], he was very scepti-
cal about e = e, or about the u-like values of w and W. In any case,
Schürr has since come to accept these values. He has been also scep-
tical about other aspects of my decipherment, such as my interpreta-
tion of the Athenes bilingual, or the identification of the name ÑUllãrima
in C.Hy 1 ( ylarmit)—an idea formulated by Ray (see above) and eas-
ily integrated into my decipherment system.
Despite these discrepancies, Schürr’s proposals opened up new inter-
pretations of some aspects I had disregarded. The most relevant was
undoubtedly his brilliant explanation of the Sinuri bilingual inscription
(C.Si 2), in which he was able to recognize the Carian version of the
formula ‘Idrieus (son) of Hekatomnos and Ada (daughter) of Hekatomnos’
that appears in Greek, in some decrees enacted by this pair of Carian
satraps. Schürr’s explanation of the letters & and $, which he con-
 THE HISTORY OF THE DECIPHERMENT 201

vincingly interpreted as representing the outcome of Proto-Anatolian

(PA) *nd and *mb respectively, is also relevant.
In 1993, an international congress on Carian decipherment took
place in Rome. In this congress, I presented a long list of onomastic
identifications that could be obtained with the new system (the “Ray-
Schürr-Adiego”).8 The new decipherment clearly won the support (ex-
pressed either in public or privately) of many relevant scholars, like
Günter Neumann and Heiner Eichner, but both ”evoro“kin and Gusmani
rejected it: while the former continued to argue in favour of his own
system, Gusmani declared his scepticism for a decipherment that implied
extremely surprising sound values for letters of Greek shape (the so-
called ‘metakharakterismós’).
I must stress here the role played by Günter NEUMANN in backing
the new decipherment. Although his public position was very cautious
(for instance at the Rome congress), he encouraged it with meaningful
gestures, for instance writing the preface for Adiego (1993a) or offering
support and suggesting ideas in private correspondence. In Neumann
(1993) he expressed his conviction that the ‘Egyptian approach’ of Ray
and myself was a step in the right direction.9 In the same paper, he
suggested linking Carian sb, very likely to be a coordinative conjunc-
tion, to Milyan sebe ‘and’. It is also worth mentioning the clear sup-
port for the decipherment manifested by Heiner EICHNER in the Rome
Symposium (see Eichner 1994).
Even more decisive in gaining wide acceptance of the new deci-
pherment was H. Craig MELCHERT’s contribution. In an important arti-
cle published in Kadmos (Melchert 1993), Melchert not only accepted
the new system, but also employed it to interpret Carian texts, open-
ing new and thought-provoking ways of deciphering the language. Of
particular note was his proposal of connecting Carian ÿbt (in C.xx 1)
to Lycian ubete ‘he offered’). As in the case of Neumann’s sb = Milyan
sebe, a word of common lexicon was recognized, and the generic
connection of Carian with Lycian and in general Anatolian was thus

8
Different ordering of names has been given by scholars to refer to the system of
decipherment that has become standard (alphabetical: Adiego-Ray-Schürr, strictly chrono-
logical regarding the published works: Ray-Adiego-Schürr). The ordering adopted here
responds to the authentic chronological succession of the research, even if Schürr’s first
contribution appeared after my first works.
9
‘. . . aber nun endlich scheint doch ein erfolgreicher und nahezu vollständiger
Abschluß der Entzifferung in Reichweite zu kommen’ (Neumann 1993:296).
202 CHAPTER FOUR

reinforced. Also important was the analysis of snn and orkn as accusatives
in -n (< PA *-n < PIE *-m), an analysis equally relevant for the linguistic
position of Carian, and later confirmed by the Kaunos bilingual.
Another scholar to take the Ray-Schürr-Adiego system as a starting
point for research on Carian was Michael JANDA, who in his contri-
bution to Rome Symposium offered a series of new ideas, among them
the fascinating hypothesis (developed here) regarding the presence of
ethnic names in the onomastic formulae of Saqqâra ( JANDA 1994).
Finally, Ivo HAJNAL was another of the scholars quick to take up the
new decipherment, with excellent results: he devoted a paper to the
Carian particle ∞i, analysing it as originally a relative pronoun from
PIE *k wi- (Hajnal 1997a). Moreover, in an independent way, Hajnal
and Schürr (1996[98]) identified two further Carian words: ted ‘father’
and en ‘mother’, clearly related to the corresponding Lycian and Lydian
nouns.
It is not easy to briefly sum up the amount of new ideas and hypothe-
ses regarding Carian—some confirmed, others ruled out- that arose in
the first half of the 90s: apart from those quoted above, other partic-
ularly sound theories were: the sound value z (< *st) proposed in Schürr
(1996a) for the letter 1; the identification of the name of the Anatolian
Storm God under the forms trqude, trqd (Blümel-Adiego 1993) and of
the name Hekatomnos in the unedited graffiti from Thebes (Adiego
1995); the reconstruction of the Carian name of Kaunos as *kbid-
(Adiego 1995:21), definitively confirmed a year later by the discovery
of the Kaunos bilingual; the attribution to Kaunos of the coins with
the Carian legend kb, interpreted as the beginning of the place name,
discovered by KONUK and also confirmed by the bilingual (Konuk
1998);10 the satem-treatment of PIE palatals in Carian, as in Luwian
and Lycian (Melchert 1993); the identification of the Carian family of
names in (i)br- = Gr. Imb(a)r- made by Schürr (1991–1993), etc.
During the excavations of summer 1996 in Kaunos, a new bilingual
inscription in Greek and Carian was found. The behaviour of the two
scholars responsible for the editing of the inscription, Peter FREI and
Christian MAREK, was exemplary: they edited the text as quickly as
possible, but also extremely accurately (Frei-Marek 1997), and suggested

10
Although this paper appeared after the publication of the Kaunos bilingual, it is
clear that Konuk’s hypothesis was formulated previously (see Konuk 1998:218, n. 14,
and 223 n. 51).
 THE HISTORY OF THE DECIPHERMENT 203

holding an international conference on the subject in the same year as

their article appeared, to which the principal scholars researching Carian
were invited (1997, in Feusisberg, Switzerland).
The new inscription was the outside evidence that was needed to
remove any doubt about the validity of the Ray-Schürr-Adiego system.
The presence of the place name Kaunos, and of the onomastic formulae
of two Athenian citizens in both parts of the inscription led to values
such as k = k, b = b, d = d, i = i, l = l, W = y, r = r, n = n, s
= s, a = a fully coincident with the values established in Adiego (1993a).
The only anomaly, T = t (not = “ according to the Egyptian approach)
is merely apparent, and easy to explain: there is no t (= t ) in the
Kaunian alphabet, so that in its writing system, a letter in the form of
T represents the value t, whereas for “ a clearly diacritical form of the
common Carian form for “ (/) was used in order to distinguish it from
T. This particularity affects only the Kaunian alphabet: in the other
alphabetic varieties, the value “ for f continues to be valid.
Shortly before the Feusisberg meeting, during the summer campaign
of 1997, a new fragment of the inscription was found, which enabled
the beginning of the Greek part to be completed, showing that Frei-
Marek’s integration of the lost endings of the line, based on the Carian(!)
part, was correct. Therefore, the complete edition of the inscription
can be found now in Frei-Marek 1998. The following table shows the
Greek-Carian correspondences obtainable from the bilingual inscription
of Kaunos:

Greek Carian Transcription Lycian evidence

Kaun[¤]oiw, k5idn, kbidn, Cf. Lyc. Xbide- ‘Kaunos’,

Kaun¤v[n k5dWn/ kbdyn“ xbidẽñne/i- ‘Kaunian’
Ipposy°nouw i[—]ini i[—]ini
Nikokl°a ni[—]lan ni[—]lan
Lusikl°ouw lWsiklas[ lysiklas[
Lusikl°a lWs[—]an lys[—]an
Lusikrãt[ouw] lWsikraTas [ lysikratas[
ÉAyhna›o[n] oTonosn otonosn
[ÉA]yhna›on oTonosn otonosn

Unfortunately, the bilingual inscription of Kaunos, so useful for confirming

the decipherment of the writing, has proved to be less useful in deci-
phering the Carian language, given the incompleteness of the Greek
part and the possible non-literal correspondence between the two texts:
204 CHAPTER FOUR

perhaps the most notable new discoveries are: (1) the word otr-“, whose
probable meaning is ‘self ’ and is possibly etymologically connected to
the Lycian atra/etli ‘self ’, (2) the ethnic suffix -yn- in kbdyn-“ ‘Kaunians’,
clearly comparable to Luw. -wanni-, Mil. -wñni-, Lyc. -ñni- (with the
same function in all three languages), (3) the accusative plural ending
-“, related to Lyc. and Mil. -s < *-ns. To these discoveries we must
add the confirmation of -n as an accusative singular ending, and of sb
as a coordinative conjunction. A plausible interpretation of the remain-
ing words and formants has yet to be made.
 CHAPTER FIVE

THE CARIAN ALPHABET

A. ALPHABETIC VARIETIES

One of the areas of Carian studies in which our knowledge has dra-
matically improved in recent years is undoubtedly the writing system.
Thanks to the definitive decipherment and the appearance of new
inscriptions, we can now create an overview of the different alphabetic
varieties that, although still limited and with very important gaps, was
unthinkable some years ago. If we compare the table of the Carian
alphabet established by Masson in the 1970s with the results that will
be uncovered here, it will be easy to acknowledge the progress made
in this field. Masson was able to offer a complete inventory only for
the alphabetic varieties of Egypt and Kaunos; we now have an almost
complete inventory for the alphabets used in two other two Carian
cities (Hyllarima and Mylasa). Moreover, the combination of new infor-
mation obtained both from the decipherment and from these new
inscriptions gives us a much better understanding of the relationship
between the different alphabetic varieties. Certain signs that Masson
had deemed independent are now interpreted as variants, in some cases
within an alphabetic variety, in others belonging to different alphabetic
inventories but functionally equivalent. Finally, the total of 44 different
letters in the table drawn up by Masson, the result of combining the
alphabets of Egypt and Caria could give the false impression that the
Carian alphabet consisted of a surprisingly high number of different
letters. In fact, as we will see, the number of different letters found in
each alphabetic variety barely exceeds 30 (31 in Memphis is currently
the highest number of letters documented for a concrete alphabet).
Many questions remain unresolved, however. The new inscriptions
from Caria complicate to a certain extent former classifications of the
different alphabets of Caria proper, based on a geographical criterion
that now seems to be too rigid. Our current knowledge also makes it
difficult to establish the exact nature of the relationship between the
Carian alphabet of Egypt, a very homogeneous sign inventory, and the
alphabets of Caria proper, which are much more varied between
206 CHAPTER FIVE

themselves, and are also different from the Egyptian alphabet. Finally,
the most intriguing question of all is yet to receive a wholly satisfac-
tory answer: what is the origin of the Carian writing system?
In the following pages, we will see firstly the known inventory for
each Carian alphabetic variety, together with a succinct commentary
on the main characteristics in each case. In the subsequent sections of
this chapter, the open questions mentioned above will be briefly dealt
with.

1. Alphabetic Varieties of Caria Proper

Alphabet of Tralleis We do not yet possess sufficient Trallean materials

to reconstruct the alphabet used in this Carian-Lydian city: C.Tr 1
offers 12 different letters, C.Tr 2, 12 or perhaps 13.1 Given the brevity
of the texts and the coincidence of some words, the combination of
both inscriptions produces an inventory of only 13–14 different letters.
This means that in all probability less than the half of the complete
inventory of letters is attested. However, despite the very limited num-
ber of letters, some interesting traits can be observed. Perhaps most
striking are the unique characteristics of C.Tr 1: it shows a right-to-
left orientation, totally atypical for Carian inscriptions from Caria proper.
Moreover, the letters for a and r seem to be A and A respectively.
This is a clear contrast with the corresponding forms in C.Tr 2: a
and R. The form of a A is the same as that found in the Sinuri-Kildara
variety, and shows the particularity of a dextrorse orientation in a sin-
istrorse context. As for A, at a glance it seems closer to Egyptian and
Kaunian r (r) than to the more extended shape of the letter in the
alphabets of Caria proper (R). In any case, it is very likely that the
definitive form of both letters is the result of internal tendencies to sym-
metry (A | A): such tendencies to structural uniformities are well known
in the history of the writing system. Regarding the rest of the letters
from C.Tr 1, particularly remarkable is the form adopted by the sign
for i, a right-to-left version (Y) of the i found in Hyllarima (y). Also
in this case, C.Tr 1 diverges—though to a lesser degree—from C.Tr
2, where the corresponding letter is Y. Both y and Y point to an iden-
tical subtype of Carian i characteristic of the Northern zone (see above),

1
The real sign that appears at the very end of C.Tr 2 is much disputed: it could
be a Q (as assumed here) or simply a z. See p. 131 and here below.
 THE CARIAN ALPHABET 207

but it is rather surprising that the form found in C.Tr 1 is in fact

closer to that of the Hyllarimean alphabet than to its equivalent in the
other Trallean inscription.
Compared to C.Tr 1, C.Tr 2 appears more as we would expect: a
for a, R for r. It is worth noting the presence of %, a letter absent
from the alphabet of Memphis. But the most uncertain point of this
inscription is the last letter, as mentioned in p. 131: the drawings are
ambiguous, pointing to either Z or q. The first of these forms would
suggest a z ≤ of Kaunian type, but the second drawing, and the fact
that in the same inscription ≤ appears as z, make it more plausible in
my opinion to interpret the letter as a q, comparable to the form that
this sign adopts in the most local alphabets of Caria proper (Q).

Alphabet of Hyllarima The new fragment of the sole remaining inscrip-

tion of Hyllarima, added to the part of the text already known, has
allowed us to create an almost complete inventory of the letters of the
alphabetic variant used in that city. They are shown below:

Alphabet of Hyllarima
a a U u
D d ñ ñ
l l N n
W y p p
R r z ≤
L l y i
Q q e e
B b K k
m m & d
o o 9 z
t t c t
S “ B b
s s

The most remarkable absence is the letter for ∞ (x X in other alpha-

bets). This absence is most probably a matter of chance. It is also pos-
sible that other letters, less common but nonetheless found in other
alphabets of Caria proper (such as %, H) could have belonged to the
inventory of Hyllarima. In any case, these 25 letters constitute an inven-
tory large enough to allow us to consider our knowledge of Hyllarima’s
alphabet to be fairly comprehensive.
208 CHAPTER FIVE

While most of the letters used in Hyllarima are clearly comparable

to the other Carian alphabets, the new fragment sees the appearance
of three somewhat surprising new forms of letters: L l, e e, and B b.
In the case of L and B, it seems clear that we are dealing with graph-
ical innovations produced within Hyllarima’s alphabet. Both letters have
undoubtedly been created from Greek letters: L is a diacritised form
of B that only can be explained by a Greek lambda (B is b in Hyllarima).2
B is simply a Greek beta, used for the Carian sound represented in other
alphabets by Ø 4 <b> (< *mb). e, however, could be an old variant
of e e assuming a common origin from Greek archaic eta E, and con-
sidering its probable presence in the Carian alphabets of Euromos and
Mylasa (see below).
The reasons for a recent introduction of L and B are not clear. In
the first case, it is worth mentioning the fact that some alphabetic vari-
ants (Thebes, Mylasa) lack a specific letter for l, this sound being cov-
ered by the sign for the ‘basic’ lateral liquid l, l. Perhaps this was
also the original situation in Hyllarima, so a specific letter may have
been created when the local alphabet already existed. Regarding B, the
same explanation could be envisaged, although the alternative suggested
in Adiego (2005) must not be discounted: that the original form of Ø
4 <b> was &, i.e. the form attested in Mylasa, and that the similar-
ity of & to & could mean discarding the first letter in favour of a new
sign directly borrowed from the Greek alphabet, B.3

Alphabet of Euromos Our knowledge of the alphabet used in the Carian

city of Euromos is more limited than that of Hyllarima. The two brief
inscriptions that remain give us an inventory of 20 letters:

Alphabet of Euromos
a a s s
d d U u
l l x ∞
W y nN n
R (/ r?)4 r z ≤
Q q íI i
(continued on next page)
2
However, we must not rule out an alternative origin from Carian l l.
3
For a detailed discussion on these new letters, see Adiego-Debord-Varinlio[lu (2005).
4
On the doubts about the presence of the variant r for r in C.Eu 1, see p. 133.
 THE CARIAN ALPHABET 209

table (cont.)
Alphabet of Euromos

B b e e
m m kK k
o o & d
t t 9 z

The absence of the signs for two very important and widespread Carian
sounds ( p and “ ) is without doubt a coincidence. It is also likely that
the alphabet of Euromos had specific letters for ñ, t and b, as in
Hyllarima. It is particularly unfortunate that we are unable to ascer-
tain whether there was a specific letter for l, or if this sound was
assumed by l l, as is the case in Thebes and Mylasa.5 The letter e,
formerly interpreted as a sign for l similar to the Kaunos form of the
letter (rotated 2 against L in other alphabets), must now be considered
a sign for e: Kaunian 2 is a peculiarity that is hard to separate from
the similarly rotated form Z of z ≤, a letter that appears in Euromos
in the more widespread form z. e as e in Euromos is more consistent
with the use of the sign with this value in the nearby cities of Mylasa
and Hyllarima.6

Alphabet of Mylasa The newly published single inscription of Mylasa,

of significant length, offers us a very wide inventory of letters, which
undoubtedly marks a clear improvement in our knowledge of the local
alphabetic varieties of Caria.

Alphabet of Mylasa
a a U u
D d x ∞
l l N n
W y p p
R r z ≤
Q q I i
B b e e
(continued on next page)

5
The two sole examples of l (in C.Eu 2) appear in contexts without the parallels
that would allow us to decide if they represent l or l.
6
Schürr (pers. comm.) has independently come to the same conclusion.
210 CHAPTER FIVE

table (cont.)
Alphabet of Mylasa

m m k k
o o & d
t t 9 z
F “ & b (?)
s s

The presently known inventory consists of 23 letters. None of the most

widespread Carian signs are missing here (on l, see immediately below).
If we compare this inscription with that of Hyllarima, for example, the
only notable absences are c and ñ, neither of which is a particularly
common Carian letter.
The most notable traits of the alphabet of Mylasa are the following:

• There is not a specific sign for l, the sound involved being repre-
sented by means of l l: note idu≤ol≤ besides dw≤ol≤ (E.Me 35).
• Given the use of l for l in idu≤ol≤, the letter e cannot represent this
sound here, and it is more likely that we are dealing rather with a
letter for e, as in Hyllarima and perhaps also Euromos. Unfortunately,
the only example of this letter in E.My 1 appears in an unclear con-
text.
• For y, a letter similar to that used in Sinuri and Kildara is employed:
W (Sinuri-Kildara V).
• Also very close to Sinuri and Kildara alphabetic variety is the form
adopted by i in Mylasa: I (cf. Sinuri-Kildara Î).
• It is interesting to note the angular form of some letters: besides the
sign for i (I), the same trait is observed in 9 z (9, 1 in other
alphabets), and in & if we accept the interpretation of this latter
sign proposed here (see the following point).
• The most troublesome sign in this alphabet is &. I have proposed
to interpret it as the local form of the letter for b, which in other
alphabets is Ø, 4 (Adiego 2005).

From these observations, the alphabet of Mylasa turns out to be of a

very idiosyncratic nature; while it shows some clear affinities with the
Sinuri-Kildara variety, as could be expected ( W = y, I = i ), the use
of e for e (against Sinuri e) or the lack of a specific sign for l (which
 THE CARIAN ALPHABET 211

does exist in Sinuri-Kildara) prevent us from speaking of a common

local alphabet for the region of Mylasa, Sinuri and Kildara.

Alphabet of Stratonikeia The two inscriptions from Stratonikeia give us

an inventory of 25 letters:

Alphabet of Stratonikeia
a a U u
d d ñ ñ
l l X ∞
W y N n
R r p p
L l z ≤
Q q Ï i
m m e e
o o K k
t t & d
F “ 9 z
s s b (?) b
H ?

It is somewhat strange, but perhaps merely a matter of chance, that

there is no letter for the widespread sound b. Leaving aside this one
uncertainty, the inventory of Stratonikeia, with a number of different
letters identical to that Hyllarima, seems almost complete. When com-
pared with Hyllarima, the only letter that might be missing is c,7 since
the other letters that are not present here are rather idiosyncratic for
some local alphabets or for the alphabet attested in Egypt.
I offer the new suggestion of recognizing in C.St 1 the Stratonikeian
letter for b, a hypothesis based on the identification of the name brsi
in the second line (see p. 143). The problem arises from the fact that
it is not easy to establish the exact form of the letter: the photograph
in Robert (1950) is not absolutely clear, and the drawing published by
Deroy (1955) could not be totally faithful. Given that the question can-
not be satisfactorily resolved, I use conventionally a form based on
Deroy’s drawing (b). In any case, this letter seems to be formally related
to the Egyptian (Ø) and Kaunian (4) signs for b.

7
About the possibility that in other alphabets of Caria proper the undeciphered let-
ter H could in fact represent c t, see p. 253.
212 CHAPTER FIVE

Apart from this new letter, the alphabet of Stratonikeia does not
offer other remarkable traits, with the exception of the form of i, Ï,
until now exclusively documented in this alphabet, and which shows
certain similarities to corresponding signs in Mylasa and Sinuri-Kildara.
Also comparable with Mylasa and Sinuri- Kildara is the form of F “.
Regarding the rest of the letters, we find the typical forms Q, 9 of
Caria proper (against q, 1 respectively, in Egypt and Kaunos). In any
case, it is worth noting the form of ∞, X, against x in most of the
Carian alphabets from Caria proper (including Kaunos).8

Alphabet of Sinuri-Kildara The combination of the letters documented in

the inscriptions of Sinuri and Kildara gives a total of 26 different signs:

Alphabet of Sinuri-Kildara
Sinuri Kildara
A A a
d d d
l l
V V y
R R r
L L l
q (??) Q q
Bb B b
m mM m
o o o
t t t
TF “
s s s
H (?) ?
U u u
ñ ñ
x x ∞
N n n
p p
z ≤
Î Î i
e e
k k
& & d
9 z
% ã?

8
The form X only appears in the inscription C.Ia 3 (but note also x in Iasos, C.Ia
2) and in the inscription C.xx 2 (of unknown origin).
 THE CARIAN ALPHABET 213

An attentive examination of the inventories of letters used in each loca-

tion reveals the close affinity between the alphabets used in Sinuri and
Kildara, which allows us to speak of a single alphabetic variety. Note
particularly the forms of a (A), y (V), i (Î). The new inscription from
Mylasa has served to reinforce this theory, because several letters, which
formerly seemed unimportant given their presence in other varieties,
have become more representative insofar as they are absent, or different,
in the inventory of Mylasa. This is the case of e e (against Mylasa e)
and L (absent in Mylasa).
What is alleged to be the most remarkable difference between Sinuri
and Kildara—the use of q in Sinuri against Q in Kildara—perhaps
does not exist: the reading of q in C.Si 1 is far from assured (see
p. 138). Note that other divergences occur within the alphabet of one
or both places: b and B in Sinuri, m and M in Kildara (but about the
doubts regarding the reading of this latter sign, see p. 142).
We can suppose that this inventory of Sinuri-Kildara is practically
complete. If compared with the inventories of Hyllarima, Mylasa, and
Stratonikeia, only the letters for b and t are missing.9 Conversely, it is
interesting to note the presence of % (as in Tralleis and Kaunos).

Alphabet of Kaunos The alphabetic variety of Kaunos is the best known

of all the varieties attested in Caria proper, thanks for the most part
to the long inscription C.Ka 2, and the bilingual text C.Ka 5. The
inventory of signs, which we can consider complete,10 consists of 29
letters.11

Alphabet of Kaunos
a a ñ ñ
d d x ∞
l l N n
W y p p
r r Z ≤
2 l I i
q q k k
(continued on next page)

9
This latter sound, however, could be represented by H, see the similar observa-
tion on Stratonikeia above.
10
Obviously, it is not impossible that some letter has still not been documented,
but in this case its functional value would be very low.
11
I assume that ÿ is the form of C.Ka 5 corresponding to 4 in C.Ka 4.
214 CHAPTER FIVE

table (cont.)
Alphabet of Kaunos

5 b & d
M m 8 g? (if = 0)
o o 1 z
T t % ã? _
/ “ 1 ?
s s O t (if = c)
H ? 4, ÿ b
U u

The main characteristics of this alphabet, some of which have already

been mentioned, are the following:

• Leaving aside the particular cases of Iasos and Halikarnassos (see

below), this is the alphabet that shows the most resemblances to the
Carian alphabet of Egypt: as the table seen above makes clear, in
this alphabet the forms for r, q and z are identical to those found
in Egyptian inscriptions (r, q, 1 against R, Q, 9 in other alphabets
of Caria proper, respectively).
• However, despite these formal analogies with the Egyptian alpha-
betic model, from a functional point of view, Kaunos behaves as a
typically Carian local alphabet: unlike the Egyptian model, there are
no specific letters for semivowels (j j, v w, perhaps also w ÿ), and
it also lacks 6 ®. Conversely, it does contain letters such as ñ ñ or
% ã?, both of which are absent in the Memphis Alphabet.
• As for the specific traits that characterise Kaunos, the following must
be mentioned:
1. A letter for e is missing. The reasons for this absence seem to be
rather of phonological nature (the Kaunian vowel system would
lack a sound /e/).
2. Some letters have only been documented in Kaunos: 8, O, 1_.
It is very likely that some, if not all of these letters have the same
value as other letters with a different shape in the rest of the
alphabets: 8 seems to be the Kaunian form of 0, and for O I
have proposed an identification with c t (Adiego 2002; see here
p. 253). But even if these correspondences were correct, these
Kaunian letters would remain unique.
 THE CARIAN ALPHABET 215

3. The letters for l and ≤ show a rotation (2, Z) regarding the rest
of alphabetic varieties (L, z).12

Other Less Documented Alphabets (Kindye, Eski Çine, Keramos) Little can be
said about the alphabet varieties of other Carian locations due to the
scarcity of documents.
In Kindye, the most remarkable feature of the single, six-letter inscrip-
tion is the sign w, unknown in the other alphabets of Caria proper.
Merely hypothetically, we could consider the possibility that this is the
form adopted by V y in this alphabetic variety. Note that in Kindye
the ‘Egyptian-like’ letter w exists side by side with R, the typical form
for r in Caria proper (against Egypt and Kaunos r).
In the case of the Eski Çine (near Alabanda) inscription (C.Al 1),
note the shape of the sign for i, apparently ì,13 which recalls the typ-
ical “northern” forms (Tralleis Y, Hyllarima y).

The Problem of the Iasos and Halikarnassos Alphabets The alphabets stud-
ied to date share the characteristic that they are attested in inscriptions
for which an origin from the location where they are found is the sim-
plest and most logical explanation. In the case of Hyllarima, Kaunos,
or Sinuri-Kildara, for example, there is no doubt that each alphabetic
inventory reflects the local writing used in each place; in Kaunos,
Kildara and Hyllarima, evidence is given by the mention of the place
name (kilara, kbid º ‘Kaunos’) or the ethnic name ( ylarmit ‘Hyllarimean(s)’,
cf. also kbdyn“ ‘Kaunians’). In Sinuri, the Carian text of C.Si 2 is insep-
arable from the preceding Greek text, and the two together fit well in
the context of inscriptions concerning the syngeneia in charge of the
sanctuary of Sinuri. The funerary inscriptions of Tralleis and Euromos,
and the long inscription of Mylasa also seem to be closely related to
the places where they were found, and the characteristics of the respec-
tive alphabetic varieties are consistent with the geographical situation.
The case presented by the inscriptions from Halikarnassos and Iasos
is very different.

12
It is true that z can occasionally appear as Z in some Egyptian inscriptions (see
for instance E.Me 14), but this occurs as a result of the disposition of the text on the
stela. The particularity of the Kaunos alphabet is that this rotated form is the canon-
ical form of the letter.
13
At least this is the form that can be hypothesized both from the photograph in
Robert and the drawing in Robert and Deroy, see p. 132.
216 CHAPTER FIVE

For Halikarnassos, there is only one inscription, found on an object

that could be moved very easily, but even its origin in Halikarnassos
cannot be assured. Here is the inventory of letters documented in C.Ha
1 (note that the inscription is written right to left):

Alphabet of C.Ha 1
# r
L l
ª b
M m
o o
t t
s s
N n
p p
w ÿ
K k
& d

The possibility that this alphabet could be that used in Halikarnassos

has important repercussions for the discussion of the origin of the
different Carian local scripts, because although the inventory of letters
that can be obtained is very limited, it is sufficient to draw some clear
parallels with the Carian alphabet attested in Egypt: w ÿ, # r, or the
right to left orientation, a very unusual practice in the alphabets of
Caria proper but well known in Egypt. This would make the region
of Halikarnassos the most likely origin of the Carian alphabet from
Egypt.
The situation is similar, but more complex, in Iasos. Here, six of the
seven inscriptions documented until now appear on fragments of vases,
so they may be of non-Iasian origin. The sole inscription found on a
stone fragment, C.Ia 4, consists of only five letters, the only remark-
able trait being the form R (‘not-Egyptian’) of r (against Egypt-Kaunos
r). Of the remaining inscriptions, the longest is C.Ia 3, where we find,
rather surprisingly, r and R side by side. Also of note is the ‘Egyptian’
or ‘Kaunian’ form for q, q (against Q in the rest of Caria, including
another inscription from Iasos, C.Ia 5). Another ‘Egyptian’ trait could
be recognized in the brief fragment C.Ia 6, where it would possible to
identify the letter Ø b, not attested in this form in the rest of the alpha-
bets of Caria proper. However, this reading is not certain (see p. 149).
 THE CARIAN ALPHABET 217

Similarly puzzling is the letter v in C.Ia 7 (= Mylasa W, Sinuri-Kildara

V?).
The following table synthesizes the letters documented in the corpus
of Iasian inscriptions:

C.Ia 1 C.Ia 2 C.Ia 3 C.Ia 4 C.Ia 5 C.Ia 6 C.Ia 7

~ ~ a a
d
l l l
W v(?) y
rR R r
L L l
q Q q
B B b
m m m
o o o o
t t t
ßF D “
s s s
U u u
x X ∞
ë ë ë n
p p
z ≤
i i I z i
e e e e
k k k
& d
0 g?
Ø? b

Supposing that all the inscriptions found in Iasos reflect the local writ-
ing, we obtain a total of 24/25 different letters, a number very simi-
lar to the almost complete inventories of signs documented in other
Carian cities. Note the presence of 0, a typical ‘Egyptian’ letter, and
the absence of H, c, ñ, %, and of a sign for z (1, 9, 9 in other alpha-
bets). Such absences could be due to chance. None of the specifically
“Egyptian” letters j, v, w appears (but note v!).

The Alphabets of the Inscriptions of Unknown (but Presumably Carian) Origin

It is particularly regrettable that we are ignorant of the origins of the
218 CHAPTER FIVE

two inscriptions on vases, C.xx 1 and C.xx 2, since they show the clos-
est affinities with the Carian alphabet of Egypt: in both inscriptions the
letter j j is found; moreover, in C.xx.2 the letter w appears, also typ-
ically Egyptian; signs for r, q, i, z are of the Egyptian (and Kaunian)
type: r, q, I, 1. In fact, both inscriptions could fit well in the Egyptian
corpus, because all the letters have a similar form to the Carian alpha-
bet of Egypt, and there is no example of a letter alien to that writing
variety. Here is the inventory of letters from both inscriptions:

C.xx 1 C.xx 2
a a
£ d
À l l
# + r
q q
ª b
m m m
o o o
t t t
D “
s s s
u u
X ∞
N nN n
p p p
z ÷ ≤
i i
e e e
w w ÿ
k k k
& d
1 z
j j j

C.xx 3 also shows a notable “Egyptian” trait: the coexistence of w and

W in the same alphabetic inventory. As far as we are aware, in the
alphabets of Caria proper there is only one letter for the sound /y/
and its semivocalic form, /w/ (W, V, perhaps also w in Kindye, but the
evidence is lacking, see above). In any case, unfortunately the inscrip-
tion C.xx 3 is hard to analyse, and the reading of some letters remain
uncertain. This, together with the brevity of the text, prevents us using
it as a valuable source on the Carian alphabet.
 THE CARIAN ALPHABET 219

We must recall the possibility, suggested by Meier-Brügger (1994:113)

that Cxx 1, C.xx 2 and C.xx 3, together with C.Ha 1, could come
from the same place (a sanctuary of the god ntro-), see above p. 162.
The affinities between the alphabets of all these inscriptions are con-
sistent with this hypothesis, and a common origin from Halikarnassos
is an attractive possibility.
Finally, C.xx 4 and C.xx 5, in fact two identical inscriptions con-
taining only a single word, clearly belong to the Kaunian alphabet, as
was observed previously by the editors (Zalhaas-Neumann 1994) and
has been referred to here (see above p. 163). The close connection to
the Kaunian alphabet can be ascertained on the basis of the use of
the letter Z (= Kaunos Z against z in the rest of the alphabets) and,
above all, the letter / for “: this letter only appears in the Kaunian
alphabet, where its presence is closely related to the use of T for t (see
above). But contrary to Kaunos, these inscriptions show the lack of a
specific letter for l, instead of which the letter l (l) is used (as in
Mylasa and Thebes).

2. Inscriptions from Continental Greece

In the case of G 1, G 2 only G 1, the Athens bilingual inscription,
shows some significant traits regarding the Carian alphabet: we note
the presence of the “Egyptian” letter j j, clearly used for the semi-
vowel /j/ (≤jas). However, this “Egyptian” trait is counterbalanced by
the use of R (not ‘Egyptian’ r) for r.

3. Egyptian Alphabets
The main characteristic of the Carian alphabets used in Egypt is their
homogeneity, which contrasts with the pronounced local differences that
can be seen in the alphabetic varieties found in Caria proper. This
homogeneity has been demonstrated by the Saqqâra alphabet, which
offers us a point of reference for analysing the other Egyptian alpha-
bets. It is true that the different sub-corpora of graffiti can at initially
suggest the contrary, because the forms of the letters show a high degree
of variety, but these differences are in reality very superficial, and can
be attributed to the less accurate writing in the graffiti. If we look at
the more pronounced differences for some letters in the varieties of
Caria proper (for example q / Q for q, r / R for r, i / y / Y /
Ï . . . for i ), we can observe that in all the documentation found in
220 CHAPTER FIVE

Egypt only one form for each letter is found (q, r, i), which implies
a common alphabetic model for the Carian of Egypt. Only in the case
of Theban graffiti can we envisage a sufficient number of remarkable
traits to speak of a specific variety, and even in this case the differences
are more functional (the absence of some letters whose function is
fulfilled by another sign) than formal.

Alphabet of Memphis As already said, the corpus of the alphabet of

Memphis (Saqqâra) shows a complete and clear inventory of letters:

Alphabet of Memphis
1 a a 17 nN n
2 dD d 18 p p
3 l l 19 zZ ≤
4 Wù y 20 I i
5 r r 21 kK k
6 L l 22 & d
7 q q 23 0 g
8 bB b 24 1 z
9 mM m 25 c t
10 o o 26 Ø b
11 t t 27 e e
12 fg “ 28 w ÿ
13 s s 29 vÚ w
14 H ? 30 jT_ j
15 uU u 31 6 ®
16 xX ∞

The following traits must be taken into account when making a com-
parison with the best-known alphabets of Caria proper:

• The Memphis Alphabet, along with the other alphabets from Egypt,
has specific letters for the semivowels /j/ /w/ and perhaps also /w/:
j, v and w or W. In this latter case, it is not possible to clearly
establish which of the two letters was originally used for the semi-
vowel, and which for the vowel. In the Carian alphabets of Caria
itself, there is neither j nor v, and in the case of w/W only one of
the signs is used (V/ W, very probably an evolution of w in Sinuri-
Kildara and Mylasa, W in the rest).
• Also specific to Memphis, and to some other Egyptian alphabets, is
the use of the letter 6 for a special liquid sound (® ). No comparable
 THE CARIAN ALPHABET 221

letter has been identified in any alphabet of Caria proper.14 Moreover,

it is interesting to note that if the identification ar®i“ = Arrissiw is
correct, this would mean that, at least in Hyllarima, where the name
Arrissiw can be definitively identified under the form ari“, there is
not any particular strategy for spelling a possible sound cluster -r®-
(Greek rr). A similar situation seems to exist in Stratonikeia, where
the same name also appears (ari“ ).
• Typical shapes of certain letters separate the Egyptian alphabet from
most of the alphabets of Caria proper, as has been noted repeatedly
in the preceding pages: q q, r r, I i, 1 z (against Q, R, and diverse
variants of i, respectively).
• None of the Carian alphabets of Caria proper display a simultane-
ous use of H and c (the only exception would be Kaunos, and only
if the very hypothetical identification O = c is accepted). Obviously,
this does not necessarily mean that both letters must be related, but
the fact certainly merits consideration.
• Two letters documented in the alphabets of Caria proper are absent
from the inventory of Memphis: % and ñ. The first of these letters
does appear in the inscription (of apparently Egyptian origin) E.xx
7 and probably also in Thebes (E.Th 34) and in the Luxor temple
(E.Lu 5). In Caria, it is documented in Tralleis, Sinuri and Kaunos.
More meaningful is the absence of ñ, until now not documented at
all in Egypt, but very widespread in Caria proper (Hyllarima, Sinuri,
Stratonikeia, Kaunos).

The Remaining Carian Inscriptions of Egypt Leaving aside the particular

case of Thebes and E.xx 7 (see immediately below), the alphabet used
in all the inscriptions from Egypt is consistent with the inventory of
letters that makes up the alphabet of Memphis. Even the less commonly
attested letters in Memphis are present to some extent in the different
sub-corpora. This is the case, for instance, of 6 ®, present in Buhen
(and also in Thebes), or Ø b, documented in Abydos and Buhen (and
Thebes). Allegedly ‘abnormal’ letters, such as B in Silsilis (E.Si 11), have
turned out to be the result of an incorrect reading (it must be read g
“ ). Perhaps the most significant variants found in the other Egyptian
inscriptions are B for b (against Memphis b (), and G for d (against

14
However, it is methodologically advisable not to rule out the possibility that in
Kaunos the sign 1, still undeciphered, corresponds to 6 in Memphis.
222 CHAPTER FIVE

Memphis d D). Regarding the remaining letters, the differences are not
so marked, and in fact seem to correspond to the more informal char-
acter of the graffiti or to varying degrees of skill of the engravers.

The Case of Thebes The alphabet of Thebes shows some singularities

when compared to the alphabetic model of Memphis. Unfortunately,
the importance of these differences is hard to estimate without a com-
plete epigraphical edition of the entire Theban corpus. I can only base
my observations on the partial knowledge gained from the drawings of
the inscriptions, thus the conclusions that will be formulated here must
be taken as merely provisional.
One singularity of the alphabet of Thebes is the absence of a letter
for l, a sound represented simply by l l, a feature also apparent in
the alphabet of Mylasa and in the two objects C.xx 4 and C.xx 5.
This particularity is best attested by the clear example pnw≤ol, against
pnu≤ol, punw≤ol≤ in Memphis.
Another trait already observed in this alphabet, namely the absence
of a letter for d (against & in the rest of the alphabets), and the use
of l (and occasionally t)15 in its place, is nowadays less clear: in the cor-
pus of Theban inscriptions provisionally elaborated by Diether Schürr
and followed here, the letter & does appear (in E.Th 4 and E.Th 14).
Rather surprisingly, both inscriptions are not new, but were in fact
already published in ”evoro“kin (1965), where the letters in question
were read in a different way. Due to a lack of further information, the
question remains open. In any case, it must be stressed that the prob-
lem posed by the presence or absence of & in Thebes transcends the
topic of the alphabetic varieties and affects the linguistic interpretation
of Theban inscriptions, because the identification of the Theban word
mlane with the form mdane depends on the hypothetical lack of & in
Thebes and its substitution by l l. Also, from a phonological point of
view, this trait will be inseparable from the absence of a letter for l:
in Thebes, l l would represent three different but related sounds (l,
l , d ) or the convergence of these three sounds in a single one.
In the unpublished inscriptions of Thebes, other rather surprising
traits appear, but they must be considered with caution until a satis-
factory edition of the whole corpus is edited. According to the drawings

15
This latter use would be visible in the name k≤atÿbr if its equivalence with Janduberiw
(attested in Lycia) were accepted (see Chapter 11, s. v.).
 THE CARIAN ALPHABET 223

of ”evoro“kin, at least an example of % ã is attested (in E.Th 34),16

and two new letters, until now unknown in Carian, are present: %
and s. Both of these letters seem to appear in E.Th 31. The first also
seems to be present in E.Th 34. However, in the case of E.Th 31,
Schürr decides against giving a reading of the whole inscription, and
in E.Th 34, he proposes an alternative reading o (see above p. 103–104).
Therefore the existence of both letters seems to be highly doubtful. As
for %, it must be pointed out that this letter can be easily mistaken
for a more standard letter x. Summing up, the alphabet of Thebes
cannot be used in any discussion of the Carian alphabet without the
definitive edition of all the Theban graffiti.
E.xx 7, the inscription of a bronze lion, shows (together with a typ-
ical Egyptian trait, the presence of v w) the use of %, a ‘non-Egyptian’
letter.

4. The Classification of the Alphabets of Caria Proper

We have seen that in contrast to the homogeneity of the alphabet used
in Egypt, a high degree of variation is found between the different local
scripts of Caria proper.
Former studies of the alphabets used in Caria proper have attempted
to classify them in groups on the basis of the marked differences in
the shape of some letters (see ”evoro“kin 1965, followed in Adiego
1993a). These groups coincide roughly with the geographical situation
of the local varieties, so that a Northern group (Tralleis, Hyllarima,
Eski Çine [South of Alabanda]), a NW group (Euromos, Kindye), a
Central group (Stratonikeia), a Western group (Sinuri-Kildara), and a
SW group (Kaunos and the inscription of Krya) have been established.
I now think that this classification should be revised, not least because
our knowledge of the alphabets of Kindye and Eski Çine is so limited
that we cannot possibly use them in a serious discussion of alphabetic
variants. Even in the case of Tralleis and Euromos, we cannot work
with inventories that are sufficiently extensive to give an idea of all the
characteristics of these alphabets.
Moreover, the new inscription of Mylasa presents a strong argument
against such groupings; despite the clear affinities between the alpha-
bet of Mylasa and the Sinuri-Kildara variety, there are also significant

16
Note also its possible (but doubtful) presence in Luxor (E.Lu 5).
224 CHAPTER FIVE

differences, some of which point rather to a connection with some

“northern” and “north-western” alphabets (Hyllarima, Euromos).
In fact, if we leave aside Kindye from the north-western group and
Eski-Çine and Tralleis from the northern group, given the scarcity of
the materials, and if we also claim that Mylasa did not have the same
alphabet as the nearby sanctuary of Sinuri, we come to a surprising
conclusion: with the sole exceptions of Sinuri and Kildara (two places
with an identical alphabet), and Krya in Lycia (where a Kaunian alpha-
bet was used), each of the other places where the Carian alphabet is
attested seem to have had their own alphabet.
For this reason, I believe that a more suitable way of analysing the
connections between the different local alphabets of Carian is to focus
on the ‘isographs’, i.e. the identical or similar shapes of several letters
that are shared by diverse alphabetic varieties.
The following table offers a comparative overview of the most rep-
resentative letters in the main alphabetic varieties of Caria. I leave out
the inscriptions of Kindye and Eski Çine for the reasons stated above,
and the inscriptions of Iasos and Halikarnassos, given the particular
problems mentioned in the preceding pages. The cities are ordered
from North to South and from West to East:

Tralleis Hyllarima Euromos Mylasa Stratonikeia Sinuri- Kaunos

Kildara
a a (A) a a a a A a
y ? W W W W V W
r R (/A!) R R (/r?) R R R r
l ? L ? - L L 2
q Q (?) Q Q Q Q Q q
b ? B B B ? bB 5
t t t t t t t T
“ ? S ? F F F /
≤ z z z z z z Z
i Y y I I Ï Î I
e ? e e e e e -
z ? 9 9 9 9 9 1
b ? B b (?) & (?) ? ? 4

The first, and most important, contrast can be established between the
Kaunos alphabet and the rest of the alphabetic varieties. As we have
commented previously, some (but not all!) of the particularities of the
 THE CARIAN ALPHABET 225

Kaunian alphabet connect it to the Egyptian alphabet: q q, r r, 1 z.

Other singular traits are specific to Kaunos: the rotated form of ≤ and
l, the use of T for t (and the subsequent creation of a specific letter
for “ ) and the form of b. To these we must also add the absence of
a letter for e and the presence of letters not found in the rest of the
varieties (8, 1).
As for the remaining alphabets from Caria, there is clear evidence
of internal affinities that correspond to some extent to the geographi-
cal situation of the cities. But, as we have already seen, with the excep-
tion of Sinuri and Kildara, whose alphabets show sufficient similarities
to speak of a single alphabetic variety, we find different isographs, not
always coincident, rather than real groupings of alphabets.
In this sense, the different forms of i are particularly striking; whereas
the “northern” Alphabets of Tralleis and Hyllarima show forms that
are clearly interrelated and differentiated from the rest (y, Y), the forms
in the “central” alphabets of the Mylasa-Stratonikeia-Sinuri/Kildara tri-
angle point to a common original form.
The coinciding form for y in Mylasa and Sinuri-Kildara is also impor-
tant; in this case, Stratonikeia does not share this characteristic, so we
can limit the extent of V/W to a more restricted area.
The distribution of the forms for e offers a somewhat different per-
spective. If one accepts my interpretation of Euromos and Mylasa e
as an e-sign, we can establish an area that includes these two alpha-
bets and that of Hyllarima, in contrast with Sinuri (a letter for e is not
documented in Kildara, probably a matter of chance) and Stratonikeia.
This distribution of e-signs can be connected with the signs for l.
As in the case of e, Sinuri-Kildara and Stratonikeia coincide in the
form of the letter, L., while in the “e-zone” the situation is totally
different: in Mylasa there is no specific sign for l, l l carrying out its
function, and in Hyllarima a special letter L is used, probably a new
sign introduced more recently on the basis of Greek L or Carian l.
Unfortunately, in the case of Euromos, the only two inscriptions to
have been found do not help us to establish whether this connection
between the signs for e and l also existed there.
The explanation for these ‘isographs’ is not clear: could they be the
result of a common origin for a group of alphabets, or rather the con-
sequence of a diffusion? Our knowledge of Carian writing is still too
incomplete to give a clear answer. The case of Sinuri-Kildara and
Mylasa is particularly puzzling; the corresponding use of a special form
for y and the form of i, both features that point to a clear common
226 CHAPTER FIVE

origin, is counterbalanced by the treatment in Mylasa of e and l. Giving

priority to certain signs to the detriment of others in order to explain
the relationship between both alphabetic variants, would be arbitrary.

5. The Relationship between the Alphabet from Egypt and the Local Alphabets
from Caria Proper
An initial conclusion can be drawn from a comparison of the alpha-
bet from Egypt with the local alphabets from Caria proper: none of
the alphabets of Caria proper identified until now can be considered
as the source of the alphabetic variety attested in Egypt. In theory, we
can formulate two extremely divergent hypotheses in order to explain
this circumstance: (1) the Carian alphabet of Egypt is a sort of
‘Uralphabet’, prior to a differentiation of the original Carian alphabet
into different local varieties, or; (2) the Carian alphabet of Egypt is
merely another local variety, with the particularity that we are yet to
find any Carian inscriptions in the Carian area it comes from, and we
are thus unable to establish where this place was.
Certain characteristics of the Egyptian alphabet, which could be inter-
preted as archaic, could be used to support the first hypothesis, in par-
ticular the existence of pairs of letters for vowels and semivowels (i/j,
u/v, W/w) against the reduction to a single series of vocalic signs in
the alphabets of Caria proper. Moreover, in the latter case, the split
observed within the local writings, where some use W for /y/ while
others resort to a letter that seems to come from w (V/W), could reinforce
this hypothesis of a more archaic system in Egypt. The chronology is
also favourable to this theory (the inscriptions from Egypt belong to
an earlier period) as is the homogeneity of the alphabet used in Egypt
set against the clear fragmentation within local varieties in Caria proper.
However, this hypothesis encounters some difficulties: not all the vari-
ants of the letters found in the local alphabets of Caria proper can be
explained by the letters used in Egypt. The most notable case is the
letter for i; ‘northern’ variants (y, Y) can hardly come directly from a
proto-form identical to the Egyptian shape of the letter (i), but instead
seem to indicate a common form that would share the angular form
of y, Y, and the horizontal trait visible in i (and also in other vari-
ants of the letter: Î, Ï, I). The absence in Egypt of certain letters
found throughout the scripts of Caria proper (% and particularly ñ) is
further evidence that the Egyptian alphabet cannot be considered simply
as the Carian ‘Uralphabet’.
 THE CARIAN ALPHABET 227

Therefore, I believe that the second hypothesis must be preferred,

although it must be qualified taking into account the possible archaic
traits mentioned above. We can base the theory that the Carian alpha-
bet used in Egypt could represent a local variety on the following
evidence:

1. There is a very significant gap in our epigraphical documentation

of Carian in Caria proper; leaving aside the only possible excep-
tion, which I shall address in the following point, there are no
inscriptions from most of the Carian coast. Only in Iasos and Kaunos
have we obtained epigraphical documentation. The rest of the known
local writings come from the inner part of Caria (Tralleis, Hyllarima,
Euromos, Mylasa, Stratonikeia, Sinuri, Kildara). It is therefore pos-
sible that the Carian mercenaries who arrived in Egypt by sea came
from coastal cities where no epigraphical documentation is available.
2. We have already commented on the ‘Egyptian’ features recognizable
in a Carian inscription that could possibly come from Halikarnassos.
It is true that the Halikarnassean origin of this inscription is not
absolutely certain, and that it appears on an object that could easily
be moved, thus there is no guarantee that it reflects the alphabetic
variety used in the place it was discovered. Yet despite this, a pos-
sible origin from Halikarnassos would be consistent with the coastal
origin of the Carian alphabet of Egypt.
3. Despite some clear differences, the local writing most closely related
to the Egyptian alphabet is that of the coastal city of Kaunos.
4. Also, in one of the inscriptions from Iasos we find important par-
allels with the Egyptian alphabet (C.Ia 3). Of course, the problem
mentioned for C.Ha 1 (that the object could easily have been moved)
is again an issue, but in this case at least the place in which it was
discovered is certain.

Is therefore possible that the Carian alphabet found in Egypt comes

from certain western, mostly coastal areas (Myndos, Halikarnassos,
Keramos, perhaps also Iasos, and others), where an alphabetic variety
still not clearly attested in Caria was used. A further argument can be
given, but it must be used with a degree of caution: the strong possi-
bility that in the onomastic formulae of Memphis, some ethnic names
appear (see below p. 269–271). If we accept this hypothesis, some forms
become very meaningful: ksolb-, from Kasolaba, ≤ugli-, from Suangela
228 CHAPTER FIVE

(both cities near Halikarnassos), yjasi-, from Iasos, and kojol- from Kos.17
The possibility that alos ∞arnos may be the Carian form of the place
name Halikarnassos would also have some interesting consequences:
alos ∞arnos appears in an inscription from Memphis (E.Me 45), which
would support the idea that this place was one of the points of origin
of the Carians in Egypt, and also in C.xx 2, one of the inscriptions of
unknown origin that contains an alphabet very close to the Egyptian
alphabetic variety. However, these examples cannot be considered as
conclusive, and must be used with care.

6. The Common Origin of the Carian Alphabetic Varieties

The existence of a number of strongly differentiated alphabetic vari-
eties in Caria proper, and the hypothesis that the alphabetic variety
used in Egypt is not a sort of ‘Uralphabet’ but rather another local
variety, does not preclude the possibility of an original unity behind
the divergent Carian local writing systems. There is clear evidence in
favour of a common origin for these varieties: all the alphabetic varieties
share the same aberrant sound values assigned to the ‘Greek-looking letters’. In
other words, it is the unexpected use of the Greek letters that allows
us to speak of a common origin for the Carian alphabetic varieties.
Had the Greek letters retained the same sound values as in Greek, a
polygenesis of the diverse local writings could be conceivable. Conversely,
the fact that q Q is q, b B is b, m M is m, and so on, in all the Carian
alphabets can only be explained by attributing them a single origin.
The (apparently) non-Greek letters also show consistent values in the
different local writing systems, which reinforces the hypothesis of a mono-
genesis of the Carian alphabet. Even the exceptions can in general be
easily explained. Note the case of Kaunian t (T): this letter (or others
of a similar shape) represents “ in the rest of the Carian alphabets,
where t is used for t. But the Kaunian situation is not a chaotic one,
and can be explained in the wider context of Carian writing: (1) T and
t letters share enough formal similarities to consider a single origin;
(2) in Kaunos, / is used for “, a letter that seems to have been created
from T by means of a diacritic mark. The reason for the introduction

17
Cf. Adiego (2004:310).
 THE CARIAN ALPHABET 229

of this diacritic mark is, obviously, to differentiate this letter from

Kaunian T t. This implies that (1) in a earlier period, T or a very sim-
ilar letter was used for “ in the Kaunian alphabet, and (2) the need
for introducing a diacritic mark was caused by a sort of evolution of
the letter for t, which led to an excessive similarity between the letters
for t (> T) and for “ (already T). The starting point of Kaunian T t is
not attested but the abovementioned formal similarities between t and
T point to a t, or at least to a form equally similar to both t and T.
In conclusion, before Kaunian T and / there seems to have existed a
pair of letters that were very close to those present in the rest of the
Carian alphabets (t, T/F/f).18
We can now imagine a sort of Carian ‘Uralphabet’ that would include
at least all of the letters attested in all, or at least most, of the Carian
local varieties. The resulting Carian ‘Uralphabet’ could be the following:

Carian ‘Uralphabet’
a a xX ∞
dD d nN n
l l p p
ùW y zZ ≤
rR r IIYyÎÏ i
L2 l e2 e
qQ q w ÿ
bB5 b kK k
mM m & d
o o vÚ w
tT t 08 g ?
f F (> /) “ 199 z
s s jT_ j
H (?) c C (O ?) t
uU u $ ç Ø 4 (ÿ ? & ?) b

In this inventory only four letters, of more limited use, are missing:

18
The evolution that led to Kaunian T and / strongly recalls the process that
occurred in Latin regarding P and R: the first letter, originally G, changed to P, and
consequently, the original P become R.
230 CHAPTER FIVE

ñ ñ 1 (?)
% ã ? 6 ®

Obviously, this ‘Uralphabet’ is only a very hypothetical assumption,

and it may possibly be too rigid, insofar as some letters could have
been created after the differentiation of the local varieties, and could
have arrived to a wide number of alphabets as the result of diffusion.
Also, some or all of the four letters excluded from the hypothetical
‘Uralphabet’ could have existed in the original inventory of signs. But
irrespective of the exact number of letters present in the original alpha-
bet, the key fact is that the total number of different letters identified
to date is 34, and that they never appear all together in a single alpha-
betic variety. The highest number of different letters documented in a
local writing variety is 31 (in Memphis). This means that the Carian
alphabet, although it seems highly anomalous from a formal point of
view, given the strange phonetic values of a lot of signs, is much more
‘normal’ from a functional perspective: both the possible Carian orig-
inal alphabet and the different local varieties contain a number of let-
ters that do not differ greatly from either the Greek alphabet or from
the non-Hellenic alphabets of Anatolia (Phrygian, Lydian, Lycian, even
Sidetic). The former image of a writing system characterized by a huge
number of letters, which led to speculations about a semi-syllabic char-
acter, has now been clearly superseded thanks to our better under-
standing of the local differences and of the values of the letters.

B. THE ORIGIN OF THE CARIAN ALPHABET

As has been repeatedly mentioned, the most striking trait of the Carian
alphabet following its decipherment is that most of the alleged Greek
letters have an aberrant phonetic value in Carian. Strictly speaking,
only the letters a o u s show a good correspondence between form
and value from a Greek point of view. The rest of the ‘Greek’ letters
have unexpected values, and while in some cases these could be explained
phonologically (for example, the value /l/ for l), many other forms
cannot: t = /t/, n = /n/, m M = /m/, r R = /r/, etc. At first sight,
the Carian alphabet seems to have suffered a ‘metakharakterismós’ (to
use Gusmani’s expression, Gusmani 1994:120) which clearly contravenes
 THE CARIAN ALPHABET 231

the ‘stability principle’ present in the Lycian, Lydian and Phrygian

alphabets, according to which the letters for sounds common to both
Greek and the language that borrows the Greek alphabet are included
in the new alphabet, retaining their respective original forms and pho-
netic values.19
How, then, do we explain this clear anomaly in the Carian alpha-
bet? The hypothesis of a chaotic, arbitrary assignment of phonetic val-
ues to Greek letters cannot be ruled out, given that several parallels
exist in the adoption of Western writing systems by illiterate people in
modern times. However, this is absolutely the exception in the Anatolian
context. In addition, this hypothesis is a very frustrating one: if we
acknowledge a free attribution of phonetic values, the Carian alphabet
is rendered an inexplicable unicum. Moreover, the ‘chaos hypothesis’
does not explain why some letters did conserve their Greek phonetic
value (the abovementioned a o u s), and overlooks other possible
points of contact between Greek, Carian and other Anatolian alpha-
bets: for example the letter p for /p/, so closely linked to Lydian beta
b, whose value is also /p/.
It is precisely these possible points of contact that have led me to
develop an alternative hypothesis: that the Carian alphabet was origi-
nally also an alphabet of Greek origin, based on the same stability prin-
ciple as Phrygian, Lydian or Lycian (Adiego 1998b). The main difference
is that this first Carian alphabet, or the Greek model on which it was
based, underwent a strong process of cursivization that dramatically
altered the form of many letters. According to this hypothesis, Carian
t (‘qoppa’-sign) would come from a H, m M from a m, n N from a
n and so on. My hypothesis requires—and this is perhaps its weakest
point—that at a certain moment in the history of the Carian alpha-
bet, the strongly cursivized graphical system underwent a sort of ‘restruc-
turing’ to a system of ‘capital letters’, so to speak, and that to this end
Greek capital letters were again used as models, although in this case
only from a formal point of view and without consideration of their
phonetic values. A qoppa could inspire the new form for the /t/-letter,
a ny for the /m/-letter, and so on. This could also explain why the
letter H, a very poorly documented letter, and one whose phonetic
value is still unknown, has the shape of one of the most common Greek
letters.

19
For this ‘stability principle’ (‘principe de stabilité’), see Boisson (1994).
232 CHAPTER FIVE

If we follow to its logical conclusion the hypothesis that the Carian

alphabet was originally totally comparable to the Phrygian, Lydian and
Lycian alphabets regarding the stability principle, we must assume that,
like in Phrygian, Lydian, and Lycian alphabets, the majority of the
Carian signs that represent sounds existing in Greek originate from the
corresponding Greek letters. This assumption was developed in Adiego
(1998b), (2004:315–317). The following table represents, with some cor-
rections and updates, the hypothesis formulated in Adiego (2004) for
each Carian letter of the ‘Uralphabet’.

Letter Value Possible origin

a a Greek. a /a/
dD d Greek lD /d/
l l Greek B /l/
ùW y Originally a non-Greek value
Perhaps a modification of r /w/?
rR r Greek T /r/
L2 l Non-Greek value
Note in Hylllarima L, directly from Greek B /l/ plus
diacritic mark
qQ q Greek value? In any case, an origin from Greek t
seems likely
bB5 b Greek B (?). Note the archaic forms of beta , in
Crete
mM m Greek M (more exactly a M form)
o o Greek O
tT t Greek T
f F (> /) “ Non-Greek value
s s Greek M ‘san’ or S ‘sigma’
H (?) ? (most probably with a non-Greek value)
uU u Greek U
xX ∞ Non-Greek value
Modification of one of the tectal letters K /k/ q q x /kh/?
nN n Greek N
p p Greek B (cf. Lydian b /p/)
zZ ≤ Non-Greek value
Related to sampi-letter 3?
IIYy i Greek E /e/, E or EI /ej/ (> closed /e/)?
ÎÏ
e2 e Greek E, H /è/
w ÿ Non-Greek value
A modification of Carian e?
kK k Greek K /k/ or rather N /kh/?

(continued on next page)

 THE CARIAN ALPHABET 233

table (cont.)
Letter Value Possible origin
& d Non-Greek value (at least originally)
A ligature of Greek ll /dd/?
vÚ w Non-Greek value (?)
A modification of r /w/?
08 g ? Non-Greek value
199 z Non-Greek value (?)
jT_ j Non-Greek value
Perhaps related to Phrygian letter for /j/, y, Y
c C (O ?) t Non-Greek value
Related to sampi-letter 3?
$çØ4 b Non-Greek value (at least originally)
(ÿ ? & ?) A ligature of Carian bb /bb/?
Note in Hyllarima the use of B directly from Greek

It is clear that this hypothesis poses serious problems that make it

difficult to accept: it forces us to assign a very high chronology to a
cursivized writing system. No examples in favour of this old phase of
the Greek or Carian alphabets exist and consequently, the evolution-
ary processes proposed for each Carian letter, no matter how con-
vincing they may seem, are totally ad hoc.20 However, this is, together
with the hypothesis of a purely chaotic imitation of a Greek model,
the only solution that I am able to imagine, if one intends to prove
that the Carian alphabet comes from the Greek alphabet. Alternative
hypotheses would have to accept the intervention of other writing sys-
tems that could be responsible for the final form of the Carian alpha-
bet, and such a solution turns out to be an obscurum per obscurius, since
these supposed writing systems are unknown.

20
In any case, I consider that the parallel examples in other writing systems, although
they are not at all useful as evidence, do offer some support several explanatory hypothe-
ses on the origin of the Carian letters. This is the case of the South-Picenian letters
for t or q, see Adiego (1998b:68).
 CHAPTER SIX

PHONOLOGICAL FEATURES

A. THE PHONOLOGICAL SYSTEM

In the present section, I will try to illustrate the phonological system

of Carian. This goal can be achieved only in an approximate way, by
interpreting the graphemes of Carian on a twofold basis: from the cor-
respondences between Carian letters and how they are reflected in the
adaptation of Carian names in Greek and Egyptian, and from the rela-
tionships between graphemes within the Carian graphical system. The
degree of precision in describing the Carian phonological system can-
not be particularly high. In many cases, we can only attain a very
generic characterization of the phonological values of the graphemes.
I will try to avoid excessively Byzantine arguments. Firstly, I will attempt
to establish a descriptive table of possible Carian phonemes. A further
section will explore the Anatolian background of these phonemes.

1. Vowels and Semivowels

The inventory of possible graphemes for vowels and semivowels varies
in the different sub-corpora of Carian inscriptions. In general terms,
we can divide these into three areas: (a) the Egypto-Carian inventory,
(b) the Carian (except Kaunos) inventory, and (c) the Kaunos inven-
tory. Despite the differences between these three types, I believe that
some interesting generalizations can be traced, so that the resulting sys-
tems are coherent and can be easily inter-related.
(a) The Egypto-Carian inventory is the most complete. Here we find
9 different graphemes for ‘vocalic’ (in a broad sense, including possi-
ble semivowels) phonemes: a e i o u y j w ÿ.
(b) In the Carian inventory, the graphemes are reduced to 6: a, e,
i, o, u, y (for y, the letter in most of the alphabets is W; in Sinuri-
Kildara, V; in Mylasa, W).
(c) In the Kaunos alphabet, the inventory is limited to 5 different
graphemes: a, i, o, u, y.
 PHONOLOGICAL FEATURES 235

The most remarkable fact is undoubtedly that the Egypto-Carian sys-

tem displays a higher number of graphemes than the other two sys-
tems. How can this discrepancy be explained? During the decipherment
process, it became evident that the Egypto-Carian system contains
graphic alternations: i alternates with j, u alternates with w, and y alter-
nates with ÿ:

i/j u/w y/ÿ

∞i/∞j pnu≤ol/pnw≤ol mdayn/mdaÿn
(psm“kw)neit/(pd)nejt uksmu/wksmu-≤
(“ar)kbiom/kbjom-≤ upe/wpe
“arnai-≤/“arnaj-s u≤ol/w≤ol≤
wliat/wljat
yiasi/yjas[i-≤]

These alternations certainly cannot be explained as morphological, because

they appeared in different positions and in some cases, although not
all, no morphological differences between the words implied could be
observed. The simplest solution, at least in the case of i/j, u/w has
already been suggested in earlier works (see Adiego 1994a:47–48): j and
w seemed to be the graphemes for the semivowels corresponding to i
and u respectively. It is true that in some cases an interpretation of j
and w as semivowels is not absolutely clear, but in such cases, it is
plausible that the semivowels were used for their corresponding vow-
els on occasions. The advantage of this hypothesis is that it allows us
to explain the relationship between the Egypto-Carian and Carian sys-
tems in a straightforward way; given that the Egypto-Carian system is
chronologically older than the actual Carian inventories, a graphemic
reduction process could take place, so that the use of specific graphemes
for the semivowels /j/ and /w/ was abandoned, and the simple vocalic
letters /i/, /u/ were used in their place. This could be used to explain
a form such as u≤ol≤ in Hyllarima vs. w≤ol≤ in Egypt.
The situation is not so clear in the case of y vs. ÿ; while for i/j and
u/w, an opposition vowel (i, u)/semivowel ( j, w) can be established,
both from their distribution properties and from the survival of i, u in
the reduced systems of Caria itself, neither of these criteria is valid in
establishing the difference between y and ÿ. The number of appear-
ances is very similar, but the only clear example of alternation is not
sufficient to attribute to one or the other letter a vowel vs. semivowel
value, and in the case of the alphabets of Caria itself, while W continues
236 CHAPTER SIX

as a vocalic sign in many alphabets (including Kaunos), the letters V/W

of Kildara-Sinuri-Mylasa seems to be closer to w than to W. In any
case, I believe that the hypothesis of a vowel/semivowel opposition for
explaining both the existence of two graphemes in the Egypto-Carian
system and the reduction to one grapheme in the Carian proper and
the Kaunos systems is also the simplest solution for y/ÿ. The sole unre-
solved problem would be to ascertain which of the letters represents
the vowel sound, and which the semivowel. In the transcription system
adopted here, preference is given to W, transcribed y, but this must be
taken as a purely conventional solution.
According to this hypothesis, systems (a) and (b) coincide in that they
both consist of six different vocalic sounds, a, e, i, o, u, y, to which
three graphemes for the semivowels corresponding to the closest vocalic
sounds (i, u, y) are added in (a) ( j, w, ÿ). In the Kaunos system, where
there are no specific graphemes for semivowels, the most notable char-
acteristic is the absence of an e vowel.
As for the phonological value attributed to each vocalic letter, the
transcription system adopted here aims to offer a broad reflection of
the quality of the vowels:
a is used to adapt Greek a in Lusikrãthw = lysikrata-, `Ekata›ow =
ktai -. Conversely, a corresponds systematically to a sound transcribed
in Greek as <a>: ada = Ada; kilara = Kildara; par(a)- (in para-eym,
par-yri∞) = Para- (in Para-ussvllow, Par-ussvldow, Para-udigow; “aru≤ol
= Sarusvllow.
e is used for Greek h in the name OÈliãdhw (Carian uliade). In Greek,
Carian e is adapted systematically by means of <h>: mane = Manhw;
pleq-≤ = Peldhkow; qtblem-≤ = Kotbelhmow, Kutbelhmiw; somne-≤ = Svmnhw.
i is used to adapt Greek i: Lusikl∞w = lysikla-; Lusikrãthw = lysikrata;
Nikokl∞w = nik[—]la-; OÈliãdhw = uliade. Note also ‘Ipposy°nhw = i[—–]ini.
This latter example also demonstrates the use of Carian i to reflect
Greek e, but one must note that the example comes from Kaunos,
where no specific letter for e-sounds (Greek e, h) existed.1 Conversely,
Carian i/j is transcribed in Greek as i: ari“ = Ar(r)issiw; arli“ = Arlissiw;
ibarsi-≤/ibrsi-≤ = Imbarsiw, Imbras(s)iw; (“ar)kbiom/kbjom-≤ = Kebivmow;
pikre/pikra = Pigrhw.

1
The adaptation in Kaunos of Greek e by means of Car. i vs. that of Greek h by
means of Car. a seems therefore an attempt to reflect the different sound qualities of
(close) e and (open) h.
 PHONOLOGICAL FEATURES 237

o is represented by Greek v: (“ar)kbiom/kbjom-≤ = Kebivmow; ksolb-≤ =

cf. Kasvlãba; ktmno, ∞tmño = Ekatomnvw; msnord-≤ = cf. Masanvrãda;
plqo = Pel(l)ekvw; somne-≤ = Svmnhw; u≤ol (and compounds) = Ussvllow
(and compounds); yrqso-≤ = Urgosvw.
<u> is used to reflect Greek ou in OÈliãdhw = uliade. Conversely,
Carian <u>/<w> corresponds to Greek <ou>, <u>, <o>: quq, dquq =
Gugow, Idagugow; pun-w≤ol-≤ = Pon-usvllow; tñu-≤ = Tonnouw; wljat/wliat =
Oliatow/Uliatow; u≤ol/w≤ol≤ = Ussvllow.
For the phonological value of y/ÿ, the Greek-Carian bilingual inscrip-
tion of Kaunos C.Ka 5 clearly supports the argument for a /y/ (semi-
vowel /w/) sound; in this inscription, W y is used in the Carian adaptation
of the Greek names Lusikl∞w and Lusikrãthw (lysiklas[, lysikratas[ ). This
is consistent with the systematic use of Greek u for the transcription of
y/ÿ in the Carian names: “ayriq (E.Me 25) = Saurigow, yrqso (C.My 1) =
Urgosvw, etc. Unlike u, no examples of Greek ou, o for Carian y, ÿ are
documented in Carian personal names of Greek sources. Therefore, it
seems that the best explanation for the apparent proliferation of “u-
letters” in Egypto-Carian (a puzzling aspect of earlier works, see Adiego
1993a:273–275) is to assume that in Carian there were two [+ high],
[+ rounded] vowels, /u/ and /y/, differentiated by their back vs. front
articulation, respectively, and that in Egypto-Carian each one of these
sounds had a specific grapheme for the corresponding semivocalic glides
(/w/, /w/).
In Kaunian the vocalic system is reduced to five vowels: /i, a, o, u,
y/. There is no e, and this absence is apparent in the adaptation of
Greek names, where a is used for h: both Lusikl∞w and Lusikrãthw
appear as a-stems in lysikla-, lysikrata-. More obscure is the case of the
ethnic otonosn (accusative) for ‘Athenian’. The presence of o for Greek
h (’Ay∞nai, ’Ayhna›ow) seems inseparable from the apparent vocalic
metaphony that has converted the three vowels of the Carian form into
o. The trigger for this metaphony could be the final vowel: the Greek
à-stem would have been adapted as an o-stem, and this o would have
caused the metaphony (atano- > otono-, see below p. 259). It is irrele-
vant if the original form for Athens in Kaunian was in fact a Doric
form with a long a instead of h or not, because both à and è would
be adapted as a in Kaunian, as shown above by the personal names.2

2
Blümel (1998b:172–173) has argued convincingly against a Doric model for the
Carian word, pointing out, among other things, that Kaunian Greeks did not speak
Doric.
238 CHAPTER SIX

The Defective Notation of Vowels

One of the most difficult problems posed by Carian vocalism is the
apparently defective notation of vowels. This phenomenon can be
detected above all in the comparison of the Carian nouns with their
adaptation in Greek (1), but also in alternations within Carian docu-
mentation (2), and by comparing several Lycian (and Milyan) parallel
forms (3):

(1)
artmon—’Art°mvn
dquq—Idagugow
kbjom≤—Kebivmow
ksolb≤—(place name) Kasvlaba
ktais—‘Ekata›ow
ktmno, ∞tmño≤—Ekatomnvw
msnord≤—(place name) Masanvrada
ntro—cf. Neter-bimow, Lyc. Natr-bbijẽmi
parÿd∞≤—Paraudigow
plqo—Pel(l)ekvw
pleq≤—Peldhkow
pñmnn≤ñ—Ponmoonnow
qlali≤—Kolaldiw, Kulaldiw
qtblem≤—Kutbelhmiw, Kotbelhmow
qÿblsi≤—(ethnic) Kubliss//ow//
≤ugli≤—(place name) Souaggela
tñu≤—Tonnouw
ylarm-it—(place name) Ullarima

(2)
brsi—ibrsi≤—ibarsi≤
dw≤ol≤—idu≤ol≤
kbidn/kbdyn“
∞yrpai—∞yrapai≤
mdayn, mdaÿn—mwdon≤
pikrm≤—pikarm≤
pnu≤ol—punw≤ol≤ (Ponussvllow)
psm“k(wneit)—psma≤k—pisma“k/pisma≤k
sdi(sas)—sidi
“r(quq)—“ar(kbiom, etc.) (Greek Sar-)
 PHONOLOGICAL FEATURES 239

trqd-—trqude cf. Lycian Trqqñt-

ttbazi—ttubazi
urm≤—urom≤

(3)
sb—cf. Mil. sebe
ÿbt—cf. Lyc. ubete

To these examples we may add a number of forms that display chains

of consonants without vowels: adymd≤, a∞tmsk“, ankbu[, (idrayri)dsemd?bq,
pda∞m≤uñ, snn, sñaidlo, tbridbd≤, etc.
Some of these examples could be interpreted as the result of syllabic
liquids and nasals that in Greek must be represented with the aid of
a vowel. This explanation is valid for forms such as tñu-≤ vs. Tonnouw
or ñ vs. on in pñmnn-≤ñ vs. Ponmoonnow. Moreover, several internal alter-
nations such as pikrm-≤/pikarm-≤ or “r-/“ar- point to a syllabic liquid,
which develops an intrusive vowel in Carian (see below p. 262). But
this explanation does not account for many other forms in which the
interpretation of liquids or nasals as syllabic nuclei is less plausible; in
the case of msnord-≤ vs. Masanvrada, an interpretation as /µ.s–.o.ºd/
seems very unlikely. Also in examples like ntro, ntokris, it seems unlikely
that n can be a syllabic nasal in direct contact with t: a sequence -nt-
seems to have evolved into -nd- and is systematically noted by means
of a specific letter, d (trqd, trqud, cf. CLuw. Tar¢unt-). In other cases,
there are no liquids or nasals involved at all: sb vs. Lyc. sebe, ktº vs
Ekatº etc.
Yet conversely, if in several cases the inclusion of intrusive vowels
in Greek adaptations could be seen as an attempt to resolve the difficulties
of pronouncing some consonant groups in Greek (for example ºdqº →
ºdagº in dquq → Idagugow), this explanation is unfeasible when applied
to cases such as qlali- → Ko/ulaldiw, where kl was phonotactically
available in Greek. This explanation also fails to account for internal
alternations within Carian ( pnº/punº, dº/idº). Therefore, at least in a
considerable number of cases, we must accept the actual existence in
Carian of vowels that are not reflected graphically. The question of
how to explain this defective notation of vowels is indeed a very difficult
one. Can one assume that Carian writing shows the vestiges of a graph-
ical system wherein vowels were not noted? This hypothesis would have
fascinating implications for the origin of the Carian alphabet, since it
could imply that Carian writing was borrowing from a graphical
240 CHAPTER SIX

system very close to the Semitic model of the Greek alphabet, that is
to say from a purely consonantal system. But this hypothesis, attrac-
tive as it may seem at first sight, runs into serious obstacles. The use
of vowel signs is generally very consistent; no forms such as **u≤l or
even **≤l besides u≤ol or **kbjm besides kbjom are found, forms which
one would expect to find if the defective notation of vowels was a
purely graphical question. Moreover, in general, at least one vowel sign
is found in each independent Carian word, a particularity that war-
rants an explanation.3
Given that these consistencies are more important than occasional
alternations like dw≤ol-≤/idu≤ol-≤, a linguistic basis for the defective nota-
tion of vowels seems more likely. The possibility that the omitted vow-
els is a sort of schwa-sound can explain some cases, particularly those
in which a vowel a or e is found in the Greek adaptations. Note par-
ticularly an alternation like Ussaldvmow/Usseldvmow. Even in some
examples of Greek o/u vs. Ø in Carian, the back character of the schwa-
vowel could be attributed to a secondary backing caused by the pre-
ceding consonant; in all of these cases, the sound preceding the unrecorded
back vowel is q, a possible uvular sound (see below p. 244). However,
this explanation would make it difficult to explain i in dw≤ol-≤ vs. idu≤ol≤.
Perhaps then we are dealing not with schwa-like vowels, but simply with
non-stressed, short vowels (*a, *e, *i, *o/u) vs. stressed and/or long vow-
els. Finally, another factor that could account for the presence of vow-
els in Greek and their absence in Carian has been conveniently noted
by Tremblay (1998:119); it is possible that some Carian names were
borrowed by Greek before certain processes of vowel weakening or loss
took place in Carian. Tremblay (ibid.) put forward an important argu-
ment in favour of this hypothesis, suggesting that some indirect Greek
forms seem to be more archaic than the direct Carian equivalents (note
for instance the absence of original medial i in the ethnic ylarmit vs.
Greek Ullarima, if it continues the cuneiform place name Wallarima).4

3
Exceptions like sb are undoubtedly connected to the fact that this type of word
always appears attached to the following word (sb=polo, sb=ada, etc.), which would sug-
gest a proclitic nature. In other cases such as snn, the presence of syllabic nasals could
explain the absence of vowels.
4
Other examples given by Tremblay (ibid.) seem to me less compelling. Also the
attempt of elaborating a relative chronology (Tremblay 1998:117–118) is in my opin-
ion premature.
 PHONOLOGICAL FEATURES 241

The tendency to omit unstressed vowels (whether schwa-like vowels

or simply short vowels, without discarding an actual vowel loss) would
be a satisfactory explanation for the above statement that at least one
vowel per word is noted in Carian writing. It could also explain the
apparently morphological alternation found in the place name kbid-n
vs. the ethnic form in pl. acc. kbdyn-“, where stress movement would
be responsible for the i/Ø alternation: /’kbi.dn/→ /kb(i).’dy.n(V)“/.
In cases like ibrsi/brsi, trqude/trqd-, sidi/sdi, the notation vs. absence
of notation of the vowels seems inseparable from the type of conso-
nant that immediately follows the relevant vowel: b and d are at least
originally *mb, *nd respectively (see below). The presence vs. absence
of a full vowel is consistent therefore with an alternation between a
syllabic nasal µ, – and a Vm, Vn sequence (or a nasalized vowel)
(/imbrsi/, /ı̃brsi/ ~ //µbrsi/, /trqund-/, /trqũd-/ ~ /trq–d-/, /sindi/,
/sı̃di/ ~ /s–di/).
When w, j appear as a seemingly syllabic nucleus, there is some
doubt about their real value. Are they simply used instead of the purely
vocalic letters (i, u, respectively) or must they be interpreted as true
semivowels, omitting the vowel that accompanies them? Many exam-
ples are ambiguous, and point to either one interpretation or the other.
We must first of all note that in the case of j, the instances of an
(apparent) syllabic use are extremely scarce: generally, j appears with
a clear semivocalic function, immediately preceding and/or following
a vowel (wljat(≤), arjom≤, pdnejt, pjabrm, kbjom≤, kojol). In the entire Saqqâra
sub-corpus, only a single example of j in a syllabic position can be
found: ∞j in E.Me 36, which seems to be merely an uncommon use of
j instead of i (cf. the systematic spelling ∞i of this frequently used word
in the rest of the inscriptions of the sub-corpus). In the other Carian
inscriptions, I have found only two unclear examples in Thebes, an
equally complicated form in Silsilis, and the words pjdl in C.xx1 and
jzpemdane in C.xx2. All of these forms are ambiguous.
The situation of w is markedly different; the examples of this letter
in a syllabic position are far more frequent. Ultimately a single, com-
prehensive solution for all cases is simply impossible to find. Cases such
as wpe (E.Me 36, E.Me 41) seem to point to a simple graphical alter-
nation with u (cf. the more spread form upe for this word). Conversely,
the systematic use of w in the spelling of the (possibly) ethnic word
mwdon≤ (mwton≤ 1x), widely documented in Saqqâra, supports its inter-
pretation as a true semivowel followed (or preceded) by an un-notated
vowel.
242 CHAPTER SIX

The behaviour of the names of the u≤ol-family in the Saqqâra sub-

corpus also substantiates the use of w for w(V); in this case, the name
u≤ol, both in its solitary and compounded forms, appears with u when
in nominative case, but with w when in genitive: pnu≤ol (E.Me 19),
tdu≤ol (E.Me 24), “aru≤ol (E.Me 30) vs. w≤ol≤ (E.Me 12), punw≤ol≤ (E.Me
13), dw≤ol≤ (E.Me 35).5 A compelling explanation for this situation could
be that there was a stress displacement, with /’u≤ol/ vs. /w(V).’≤ol≤/.
In the genitives, the unstressed vowel would not be notated graphically,
and w would represent a true semivowel preceding it. This hypothesis
has a counter-example: in Thebes a nominative pnw≤ol is found. However,
in the Saqqâra sub-corpus the rule functions well and the examples
seem to go some way towards supporting this explanation. It is also
corroborated by the likely connection of u≤ol with the family of nouns
ending in -uassiw (Aktauassiw, Panuassiw, Saruassiw, cf. also Oa3a3iw),
where a *wa≤- basis could be reconstructed.
Also in the case of the names wljat, wliat, “arwljat-≤ and “r-wli≤, the
etymological connections proposed (see Chapter 11, s. v. wliat) point to
original *waliat-, *wali- forms.

2. Consonants
Voiceless Stops
The following five voiceless stops can be recognized in Carian:
p t/T xX kK q/Q
p t ∞ k q
Labial p, dental t and velar k do not pose any particular problems.
The phonological value of the corresponding letters has been firmly
established in previous works and it should be sufficient here to pro-
vide a few relevant examples:
p is used to adapt Egyptian p and Greek p in proper nouns: pisma“k
(and variants) = Psmtk, pdnejt = P3djNjt, pidaru = P¤ndarow. Conversely,
Egyptian p and Greek p reflect Carian p in Carian personal names:
paraeym = Prjm, pnu≤ol = Ponussvllow, pikre = Pigrhw, etc.
If the explanation of nproº (in nprosn≤) as a Carian version of the
Egyptian name Nfr-˙r (Nefervw) is accepted (Schürr 1996a:68, n. 18,

5
Adiego (1993a:273).
 PHONOLOGICAL FEATURES 243

Vittmann 2001:42), it could serve as a good example of an Egyptian

f adapted to p in Carian. See Chapter 11, s. v.
t is used for Egyptian t and for Greek t, and also y (/th/): pdnejt =
P3djNjt, ntokris = Nj.t-jqr (Nitvkriw), lysikrata-s = Lusikrãthw, otono-sn =
ÉAyhna-. This latter example confirms the absence of voiceless aspirate
stops in Carian, the corresponding Greek ones being adapted by means
of the simple voiceless stop. For the use of Greek t for Carian t in
personal names, cf. wljat (and variants) = Uliatow, Oliatow, tñu≤ = Tonnouw,
etc.
k is used for Egyptian k in the abovementioned personal names pisma“k
(and variants), ntokris, and for Greek k in the names lysiklas[ and lysikratas[
= Lusikl∞w, Lusikrãthw. k is used for k in the adaptation of Carian
nouns: kbjom≤ = Kebivmow, kilara- = Kildara/Killara, ksolb≤ cf. Kasvlaba.
Note also the equivalence of Carian kbidn ‘Kaunos’ = Lycian Xbide ‘id.’
(with Lycian <x> = /k/)
The exact values of ∞ and q are less clear. That ∞ represents a tec-
tal sound is easily deducible from the use of Egyptian k in the tran-
scription of the Carian name urs∞le≤, 3rskr. For this letter, a palatal value
(palatal stop [c] or palatalized velar stop [kj ]) can be envisaged, although
the argument is based only on comparative evidence: its use in the
(originally) relative pronoun ∞i, whose Anatolial proto-form is *kwis (Hitt.,
CLuw. kui“ ). It is plausible to imagine that Carian has undergone a
process similar to Milyan: the labiovelar stop has first been delabial-
ized (*k wis > *ki, with loss of the final s) and then palatalized before i
(*ki > /kji/ or /ci/, Milyan ki [ci]). In Lycian, /ci/ has eventually
become /ti/, completing the fronting process.6
The palatal character of ∞ is also recognizable in the striking alter-
nation ∞/k, shown by the name ktmno (Thebas) vs. ∞tmño≤ (Sinuri). Here,
if k and t were in contact, t could have caused the fronting of k to ∞.
Certainly, the Greek form of the name, Ekatomnvw, with a between k
and t, seems contradict the existence of a contact between the two
stops, but it is possible that the Greek form has undergone the ana-
logical influence of the goddess’ name Hekate, so that the a vowel
would be absent in the original Carian name.7 An alternative view, fol-
lowing on from Tremblay (see above) could be to assume that the

6
This explanation was formulated for the first time in Adiego (1995:29–31).
7
It is interesting to recall that Ekatomnvw can plausibly be seen as a pure Greek
name, as Neumann has repeatedly pointed out, see Neumann (1994:17).
244 CHAPTER SIX

Greek form reflects a more archaic stage than Carian, and that in this
latter language a syncope would have taken place, permitting a con-
tact between k and t. In any case, it seems more than a mere coinci-
dence that the same name appears in Lycian spelled as katamla, with
k [c], not x [k].
Apart from urs∞le≤ and ∞tmño-≤, there are no more clear examples of
onomastic identifications where ∞ would be included: the connection of
alos(d) ∞arnos(d) with the place name Halikarnassos remains very doubt-
ful. In any case, it would only be useful for confirming the tectal value
of the sound. As for p∞simt≤, although its connection with the well-
known Egyptian name Potasimto (P3-dj-Ór-sm3-t3wy, shorter form P3-
dj-sm3-t3wy, Greek Potasimto) suggested by Schürr (apud Ray 1994:205)
is a compelling theory, the phonetic details of the Carian adaptation
of the name remain obscure to me.
The tectal character of q is also beyond question; in the Greek adap-
tation of Carian names it is systematically represented by k or g: qtblem≤
= Kutbelhmiw, qlali≤, qlalis = Kolaldiw, Kulaldiw, quq = Gugow etc. In
the bilingual inscription E.Sa 2, the Carian name qyri≤ is adapted in
Egyptian as K3rr, with the use of the biliteral sign <k3> for ku ~ qÿº
(Vittmann 1996). It is also used in the possible Carian adaptation of
the Egyptian name Ny-k3w, Nek«w, niqau≤.
This letter appears in several of examples followed by o, u (triqo, plqo,
quq, trqude, etc.), and o, u are also the vowels used in Greek after k
when this latter corresponds to a q in contexts of defective vowel nota-
tion (qlali≤, qlali≤ = Kolaldiw, qtblem≤ = Kutbelhmiw, Kotbelhmow, per-
haps also qtblo = Kotobalvw, and this would seem to point to a back
articulation of q (uvular /q/?)
Occasionally, both k and q (there are no examples available for ∞ ) are
rendered in Greek through a voiced velar stop: pikre = Pigrhw, yrqso≤ =
Urgosvw, quq = Gugow, dquq = Idagugow. The two first examples can be
easily categorized as allophonic; k, q would tend to become voiced in
contact with the sonorant r (note the alternative spelling Pikrhw for the
second name, which confirms the rather spontaneous and irregular voic-
ing phenomenon). More delicate is the case of (Ida)-gugow, wherein the
spelling of the two voiced stops is systematic, making possible the exis-
tence of true tectal voiced sounds, not noted graphically in Carian.
An alternation ∞/q can be identified in the family of names in yriq/yri∞
(= Greek -urigow), note “ayriq vs. idyri∞-≤, paryri∞(-≤), and perhaps also in
ÿdiq/ÿd∞ (“aÿdiq-≤ vs. parÿd∞-≤ ) (for the possible common origin of yriq-
and ÿdiq- stems, see below pp. 262–263). It would be difficult to offer
 PHONOLOGICAL FEATURES 245

a clear explanation of this alternation if the rather distant sound val-

ues argued here (∞ palatal vs. q uvular) are accepted.

Voiced Obstruents(?)
bB5 dD
b /b/ [ b]? d /d/ [ ä]?
The Greek rendering of Carian names, along with additional evidence
from Egyptian and Lycian, confirms the point of articulation and the
voiced character of these two letters:
b is rendered by b: kbjom≤ = Kebivmow, ksolb≤, cf. (place name) Kasvlaba,
k≤atÿbr—Janduberiw, qtblem≤ = Kutbelhmiw, Kotbelhmow, qÿblsi≤ , cf. the
ethnic Kublisse›w, etc.
In addition to this evidence, we must also consider the use of Carian
b for Egyptian b in some personal names of Egyptian origin: ttbazi,
ttubazi = T3-dj(.t)-b3st.t (Tetobastiw), pdubez = P3-dj-b3st.t, Petobastiw,
Petoubestiw. The comparison with Lycian and Milyan reinforces this
phonological value; as well as the correspondence between the Carian
and the Lycian name of Kaunos, Car. kbid-, Lyc. Xbide, the lexical
equivalences Car. sb ‘and’ = Mil. sebe, Car. ÿbt = Lyc. ubete ‘offered’
are also very significant.
d is adapted as d in Greek: ada = Ada, dquq = Idagugow, dw≤ol≤,
idu≤ol≤ = Idussvllow, msnord≤ cf. place name Masanvrada, ardybyr≤ =
Arduberow. Conversely, Carian d reflects Greek d in the name uliade =
OÈliãdhw.
Carian d is used to render Egyptian d: pdnejt = P3djNjt.
A good example of the correspondence Carian d: Lycian d is the
word ted associated with Lycian tedi ‘father’.
It seems that there is no letter for /g/ [ g ]. See above for Greek g
corresponding to <k>, <q> in Carian.
It is possible that b, d were articulated as fricatives in intervocalic
position, as is the case in Lycian, but there is no direct evidence for
this type of articulation. Our suspicions are based mostly on the exis-
tence of the letters & d and Ø/4/B/ b: they represented, at least
&

originally, clusters of nasal + voiced stop (*/nd/, */mb/), precisely the

context typical for the articulation of voiced obstruents as stops vs. the
fricative realisation of these sounds in most other contexts (cf. the sim-
ilar situation in Lycian). But the doubts concerning the real sound value
of <d> and <b> in Carian (see below) are sufficient to cast doubt on
the fricative character of b, d.
246 CHAPTER SIX

It is also possible that in an initial position, the original /b/ and

/d/ have become voiceless, as in Lycian. However, the Carian docu-
mentation is not as clear as the Lycian evidence. Although examples
of b and d are scarce, and some of them can be interpreted as the
result of a lack of notation of the initial vowel (for instance, dquq =
Idagugow), there is not enough evidence to support the assertion that
all the examples of initial b and d must be interpreted in this way.
As for the indirect documentation, the number of forms with initial b
and d is also minimal, and in a considerable part of these a contact with
a sonorant r, l, n could be responsible for an allophonic voicing of a voice-
less stop: Bruajiw/Bruassiw, Brvlow (personal names), Br¤oula, Bridaw
(place names), and perhaps also Beryaw, Berrablviow, Dandvmow, Daru..ow,
Dersvmanhw, Dersvw, Dersv . . . tiw (personal names), Bãrgasa/Bãrgaza,
BargÊlia, Bolli.evn (place names). A good argument in favour of this
explanation is the alternation P-/B- attested by the name Bãrgasã/Pãrgasa.
The remaining examples of B- and D- would thus become very scarce
(personal names: Boivmow, Deibow, place names: Babein, Bvnitv, BubassÒw
and variants, Bvrand//a//, D°dmasa, Didassai, D¤duma—although this
is doubtful, as it could have been influenced by Greek—and DÊndason).
An alternative explanation for some of these words, to my knowl-
edge not proposed until now, would be to assume that the Greek ini-
tials B- and D- are in fact a reflection of Carian <b> (< *(V)mb) and
<d> (< *(V)nd ). This explanation is plausible at least in the case of
the name Dersvw, which could be compared simultaneously with
Andarsvw/Androsvw and with dar“ (if a PN). Also Brvlow, if from
*(V)mbrol-, could be connected to the family of names in (i)br- (cf. par-
ticularly para-ibrel-). The absence of an initial vowel in the Greek adap-
tation would correspond precisely to the situation in Carian, where we
find vis-à-vis forms both with an initial vowel (vgr. ibrsi-) and without
it (vgr. brsi-).
Arguing in favour of a devoicing process of b in an initial position,
we should finally mention a good indication found directly in Carian:
the internal alternation seen in Carian between piks- and dbiks, pik(a)rm
and dbikrm, where b is clearly etymological (bik- < PIE *bh èh2-).

Voiced Stops or Nasal + Voiced Stops

Ø, 4, (Hyllarima) B, (Mylasa) &
&

b d
/mb/ or /b/? /nd/ or /d/?
 PHONOLOGICAL FEATURES 247

As Schürr has demonstrated, practically all the interpretable examples

of the letters b and d find their origin in (respectively) *-mb- *-nd- groups
and are reflected systematically in Greek by -mb-, -nd-: (i)brsi, ibarsi =
Imbrassiw, Imbarsiw, trqd, trqude = CLuw. Tar¢unt-, lycian trqqñt (<nt>,
<ñt> = /nd/). We are left with the problem of ascertaining what these
letters actually represent in Carian. Three main hypotheses are plau-
sible:
(1) b and d denote true consonantal groups -mb-, -nd-, respectively.
(2) b and d denote pre-nasalized consonants (/mb/, /nd/), a type of
stop that exists in different world languages.8
(3) b and d denote only the oral voiced stops /b/ and /d/; like in
many other languages (from Lycian to Modern Greek or Spanish), /b/
and /d/ would appear as stops only in certain contexts, particularly
after nasals. In the remaining contexts, their articulation would be frica-
tive and denoted by b [b] and d [ä] (see above). The graphemes <b>
and <d> would therefore be used only to represent these stops, and
the nasal that precedes them would not be noted graphically.
With our current knowledge of Carian, it is impossible to choose
between these hypotheses. Preference could be given to (3) if we were
to find examples wherein b or d were used for a voiced stop not aris-
ing from a nasal + stop sequence. Unfortunately, no concrete exam-
ples of this kind have been found; the only form for which this explanation
has been suggested is pdaº in pda∞m≤uñ, compared in Adiego (2000:145)
with Lycian pddẽ ‘place’, from PIE *pedom. According to this interpre-
tation, d would represent a stop articulation of the dental in direct con-
tact with the stop p, as in Lycian (where a gemination has also taken
place). However, this connection is not conclusive enough to be used
as the basis of a definitive argument in favour of d = [d], and not
[nd] or [nd].

Liquids
rR l 6 L2L
r l ® l
/r/ /l/ /rj/? /l.l/?

8
Cf. Boisson (1994:219).
248 CHAPTER SIX

Four letters have been identified as representing liquids and nasals in

Carian.
r corresponds to Greek r and Egyptian r in the adaptation of per-
sonal names in Carian: lysikratas[ = Lusikrãthw, ntokris = Nj.t-jqr (Nitvkriw).
Conversely, both Greek r and Egyptian r reflect Carian r in Carian
names: kilara = Kildara/Killara, arli“(≤) = Arlissiw, arliom≤ = Arlivmow,
pikre≤ = Pigrhw, etc. paraeym = Prjm, arli“(≤) = Jr“3, “arkbiom = ”3rkbym,
urs∞le≤ = 3rskr.
l renders Greek l in lysiklas[, lysikratas[ = Lusikl∞w, Lusikrãthw, uli-
ade = OÈliãdhw, and l is used in Greek for Carian l: arli“ = Arlissiw,
arliom≤ = Arlivmow, lÿ∞se, lÿ∞si≤ = Lujhw, qlali≤, qlalis = Kolaldiw etc.
In Egyptian, where the graphical reflection of l was always problem-
atic (see Loprieno 1995: 31, 33), we find r for Carian l in 3rskr =
urs∞le≤, whereas the sequence ºrlº in arli“, arliom≤ is noted by means of
a single r ( Jr“3, Jrym). Although more precise conclusions about their
exact articulation (was r flapped or trilled, and was l more or less den-
tal?) cannot be established, it is clear that the letters r and l represent
the two basic liquids of the Carian phonological system.
The status of the two remaining liquid letters is less clear, and it is
significant that these signs are not found in all the Carian alphabets:
l is absent from the Thebes and Mylasa inventories, and ® is even less
widespread, having only been found in Egypt.
The use of ® is limited to the following words: ar®i“, me®≤, qdar®ou≤
and t®∞at(a)r≤. Both ar®i“ = Arrissiw and t®∞at(a)r≤ maintain the theory
of ® as a palatalized r /r j/, as a result of the contact with a palatal
sound (i, ∞).9 This explanation is less convincing in the case of qdar®ou≤,
whilst me®≤ does not offer any evidence either for or against this hypoth-
esis. In arri“, an assimilation process could be behind r® if we assume
that there is a connection with the more widespread name arli“. This
explanation would also be feasible for qdar®ou≤ if the word is related to
CLuw. ¢utarlà- ‘slave’, as suggested in Adiego (1995:24–25).
l is the letter used in Carian for the sound(s) transcribed in Greek
adaptations as ll and ld. Boisson (1994:216–217) offers a detailed
analysis of earlier proposals for interpreting this letter, and formulates
some possible values. While I believe that one cannot rule out the

9
It is worth noting that the new text of Hyllarima (C.Hy 1a) offers for the name
Arrissiw the form ari“, without any special sign for the (possibly peculiar) sound ren-
dered in Greek as -rr-.
 PHONOLOGICAL FEATURES 249

possibility that l represents a peculiar unitary sound, for which the

oscillating Greek spelling ll/ld would be a rough translation, it seems
to me preferable to interpret l as the simple notation of a geminate
sound, whose articulation was more dental than that of the Greek ll;
indeed this could explain the alternative spelling ld. That the sound
was very close to Carian l is demonstrated by the systematic use of l
in those alphabets (namely Thebes and Mylasa) that do not have a
specific letter for this geminate. Also, the use in Hyllarima of a dia-
critised Greek lambda for the sound (L) is a clear indication of an l
sound. The geminate nature of l (originally, at least) is supported by
its distribution: l never appears in an initial position, which probably
implies that it needed at least a vowel preceding it (/Vl.l/).

Nasals
m M n Nñ
m n ñ
/m/ /n/ ?
m and n represent the labial and dental nasal stops typical for many
phonological inventories of world languages.
m is used to adapt Egyptian m: pisma“k (and variants) = Psmtk.
Conversely, Carian m is reflected in Egyptian by m, and in Greek by
m: paraeym = Prjm, “arkbiom = ”3rkbym, kbjom≤ = Kebivmow, msnord≤ cf.
Masanvrada, etc.
n reflects Egyptian n and Greek n: -nejt, -neit = -Njt, ntokri≤ = Nj.t-jqr
(Nitvkriw), niqau≤ = Ny-k3w (Nexvw), nik[ ]la- = Nikokl∞w; and Greek n
transcribes Carian n: somne≤ = Svmnhw, pnu≤ol, punw≤ol≤ = Ponussvllow,
msnord≤ cf. Masanvrada, etc.
Schürr correctly established that ñ ñ, a letter absent from the Carian
alphabet of Egypt, also represents a kind of n, on the basis of the ono-
mastic identifications ∞tmño-≤ (2×) = Ekatomnvw and pñmnn-≤ñ = Ponmoonnow
in the Sinuri bilingual text (E.Si 2), and this has since been confirmed by
the new inscription of Hyllarima, where tñu≤ is found side by side with
Tonnouw. Note also the alternation ñ/n in ∞tmño≤ vs. (Thebes) ktmno.
Its absence from the Egyptian inventory is a little surprising, espe-
cially considering its wide distribution in most of the Carian alphabets
of Caria itself—it appears in such distant places as Hyllarima and
Kaunos—and the tendency of Egypto-Carian writing to contain letters
that the alphabets of Caria have lost (j j, v w, for example). Perhaps
the formal resemblance of ñ to z ≤ played a part in its disappearance.
250 CHAPTER SIX

Its exact phonetic value is difficult to determine. In most cases, it

can be attributed the value of a syllabic nasal (like Lycian ñ): pñmnn≤ñ,
ñmailomda, yri∞ñ, tñu≤ (note the Greek adaptation Tonnouw that could be
a reflection of /tn.nu/-), but its use in pda∞m≤uñ weakens the validity
of this interpretation. In any case, a functional difference between ñ
and n seems more probable than an articulatory difference: none of
the examples seems to support the interpretation of ñ as a nasal artic-
ulated as palatal, velar, or the like.

Fricatives
s fF/ zZ
s “ ≤
/s/ /“/ /ç/?
The exact value of the three fricative sibilants of Carian, and also their
origin (see below) is undoubtedly the phonological particularity of Carian
that has yielded most discussion.10 In this case, Greek adaptation of
Carian names proves to be largely useless, due to the existence in this
language of a single sibilant s /s/. The three Carian sibilants are sys-
tematically transcribed as s- /-s(s)-: “aru≤ol = Sarussvllow, arli“ =
Arlissiw, msnord-“ = Masanvrada, (i)brsi, ibarsi = Imbarsiw, Imbras(s)iw
etc. In the other direction, Greek s is adapted to s in Carian (lysiklas,
lysikratas).
More interesting, however, is the contribution of Egyptian, in which
Carian s is adapted to s and Carian “ to “: urs∞le-≤ = 3rskr, “arkbiom =
”3rkbym. As for the adaptation of Egyptian names, it is particularly
noticeable that ≤ and “ alternate in rendering the sound t /t“/ of the
Egyptian name Psmtk ( pisma“k, pisma≤k, etc.), in contrast with the use of
Carian t for the same Egyptian sound in tamou (see below). A possi-
ble use of Carian s for Egyptian s could be seen in p∞simt if the con-
nection of this form with Egyptian Potasimto (P3-dj-Ór-sm3-t3wy,
P3-dj-sm3-t3wy, Potasimto) were accepted.
From all this information, certain conclusions can be drawn: (1)
Carian s most probably represents the basic voiceless dental sibilant in
the phonological systems of the world, as shown by its use for transcribing
Greek s in the two Greek names quoted above; (2) “ is probably a
palato-alveolar voiceless fricative, which can be concluded from the use

10
See Hajnal (1998), Schürr (2001b), Melchert (2002).
 PHONOLOGICAL FEATURES 251

of Egyptian “ in ”3rkbym = “arkbiom; (3) As for ≤, its use in pisma≤k allows

two possible interpretations: it could be either a sound very close to
Carian “ (perhaps a palatal /ç/, like German ch in ich), given the alter-
nation pisma“k/pisma“k, or an affricate sound different from /t“/ (for which
the letter t already exists), corresponding roughly to the Egyptian affricate
t. If the latter were true, /ts/ would be a logical solution. In any case,
the interpretation as (palatal) fricative seems preferable; if ≤ were to
represent /ts/, we would expect to come across an occasional use of
Greek z in the transcription of the u≤ol-family of names, but in fact,
ss is systematically found.

Affricates
cC 1 9
t z
/t“/ /ts/ and/or /st/?
The attribution of the phonological value of t is based exclusively on
its use in the bilingual inscription E.Me 7, where it is used in the
Carian adaptation of the Egyptian name T3j-jm = w [‘amòu] (Tamvw,
Yamvw, Samauw, Samv#w). Further examples of this letter are extremely
scarce, and no satisfactory onomastic interpretations have been pro-
posed for the possible personal names in which it appears.
The value of z was established by Schürr on the basis of Carian
names that appear to contain the name of the Egyptian goddess Bastet
(B3st.t): tt(u)bazi, piub[a]zi (see Chapter 11, ss. vv.). Doubt remains as to
whether z represents a consonantal heterosyllabic sequence s + t (or
s + d or similar), or rather a phonetic result of this sequence (most
probably an affricate /ts/ or /dz/). No clear evidence can be found
in the other forms that contain this letter. In the new inscription of
Mylasa some possible onomastic identifications have been suggested (see
Adiego 2005) that point to both /st/ and to /ts/ as possible values of
z: qzali = Kostvlliw vs. myze cf. Mouzhaw. However, both identifications
are rather tenuous.

3. Letters of Uncertain Value

0 and its possible Kaunian form 8 can represent a nasal + voiced tec-
tal stop, but this assumption is based practically on a single piece of
evidence: the likely connection of ≤u0li≤ with the place name Souaggela.
This tectal value is also suggested by the sequence qrds8rdso[- in the
252 CHAPTER SIX

first line of C.Ka 2, in which a sort of figura etymologica formed from

the word qrds (cf. qrds, C.Ki 1, and qrdsol“ in the same inscription C.Ka
2) seems to exist. The rather conventional transcription g is adopted
here for these letters.
As for %, a letter absent from all the Egypto-Carian alphabets with
the sole exception of Thebes (and also E.xx 7 if the inscription really
comes from Egypt), we can draw on two rather weak sources: firstly
pr%idas in E.xx 7, for which Schürr has proposed a connection with
Bragxidai, the name of the priests of Apollon in Didyma (near Milet),
and secondly ]oml% in C.K 2, which may be related to C.K 5 uiomln.
The first case would seem to point to a nasal + tectal value, as for 0,
but if the second connection is reliable, perhaps % would in fact be
rather a type of nasal. Both contexts would then favour a nasal with
a tectal mode of articulation (for example, a velar nasal /fl/. Other
examples are not so clear, but some of them indicate a nasal rather
than a nasal + stop: the third line of C.Ka 2 shows a construction
[-]∞arlano% sb z≤arios% where % seems to be a morphological ending.
If a (tectal) nasal value is accepted, the forms could be interpreted as
singular accusatives in n (cf. lusikla-n in C.Ka 5) spelled with a tectal
nasal, for reasons we are unaware of. In any case, the possibility of a
/nasal + tectal/ value cannot be ruled out. In such a case, the difference
between 0 and % would come from the point of articulation of the
tectal, in other words, both letters would be the correlate of one of the
three voiceless tectal stops that exist in Carian (∞, k, q).
I provisionally and conventionally adopt the transliteration <fl> for
%, although this is not to say that I consider that attribution of a velar
nasal value to this letter certain, or indeed even preferable.
The letter O appears exclusively in the alphabetic variant of Kaunos.
Adiego (2002) proposed that it represents the letter c C t /t“/ of other
alphabetic variants. This proposal was based principally on the dis-
tributive properties of the sign, which appears in a final sequence -oO,
at least in the word punoO (C.Ka 2; the segmentation is guaranteed
thanks to the following word, otr“, equally segmentable in C.Ka 5). This
sequence can be compared to the endings in -ot from Hyllarima C.Hy
1, ºpususot, msot, to which we can now add muot, from the new frag-
ment of the inscription recently discovered.
An alternative approach, adopted by ”evoro“kin and tentatively sup-
ported in recent times by Schürr and Melchert, is to view O as a sort
of glide /w/, given its appearance between two o in most examples.
This explanation is complicated by the fact that in Kaunos there are
 PHONOLOGICAL FEATURES 253

three clear examples of ouo sequences, where it is difficult to interpret

u as anything other than a /w/ (ºouorº 2×, ºorouoº).
For this sign a transcription t2 is adopted.

4. Letters of Unknown Value

Strictly speaking, there remain only two letters for which no possible
phonological value can be suggested at the present: H and 1. Of these
two letters, the first is undoubtedly the most important, as it appears
in Egypto-Carian alphabets (Saqqâra 2×, Abydos 2×) as well as in in
Sinuri-Kildara (4×), in Stratonikeia (2×), and in Kaunos (7×). No expla-
nation has been found for any of the forms where H is used, not even
the two clear personal names from Saqqâra (E.Me 28 psHÿm[-]≤, E.Me
31 mHm≤ ). Most likely is that it had a consonantal value, given its inter-
vocalic context in forms such as ºaHuq[ (C.Ki 1), sb uHbit (C.Ka 2); this
explanation is consistent with its use in mHm≤, if a typical defective
vowel notation (ºm(V)H(V)º) is used.
As was noted in the chapter devoted to the Carian alphabet, the
apparently complementary distribution of H and c in the alphabets of
Caria itself is curious: where c appears, there is no H (as in Hyllarima)
and vice-versa (Sinuri-Kildara, Stratonikeia, Kaunos). One would be
tempted to interpret H as a local variant of c, a hypothesis also sup-
ported by the formal proximity of the two signs, but the existence of
both letters side by side in the Egypto-Carian alphabet, and the pos-
sibility that the Kaunian counterpart of c may in fact be O (see above),
make it difficult to argue the case further.
The situation of 1 is very different: it is only present in the Kaunian
alphabet, and only 9 examples (7 in C.Ka 2, 2 in C.Ka 4) can be
found. Purely for the sake of hypothesis, one could assume that it is a
particular Kaunian form for a letter existing in other alphabetic vari-
ants, in which case, the only remaining possibility would be to equate
it to Egypto-Carian 6 ®. However, this cannot be confirmed by any
of the instances in which 1 is used, and the possibility therefore remains
that the letter has a phonological value specific to Kaunian.
According to the drawings of ”evoro“kin, the still unedited graffiti
of Thebes seems to show two new signs: 0 0 and Z. Since I do not pos-
sess a definitive and accurate epigraphical edition of this new corpus,
I shall simply draw attention to this possible existence (see pp. 103–104).
Finally, very little (indeed almost nothing) can be said about the
strange ‘diamond-like’ sign (K, t ) that appears on two occasions (E.Th
254 CHAPTER SIX

28, E.Si 4), in both cases preceding the same word (bebint). Note that
the other two examples of the word are not accompanied by this sign
(E.Th 30, E.AS 7), making it highly unlikely that the sign was actu-
ally a letter.

5. Phonotactics
The defective notation of vowels makes it very difficult to draw an
accurate picture of Carian phonotactics, since the task of distinguish-
ing whether a sequence of consonants actually represents a consonan-
tal group, or whether in fact one vowel is graphically missing, is extremely
complicated. Accuracy is also compromised when resorting to the indi-
rect evidence in the Greek adaptation of Carian names, as there is a
risk of identifying certain characteristics as Carian when they in fact
belong to Greek adaptations of Carian onomastics. The only solution,
although far from perfect, is to combine both sources, but the results
are then incomplete and many gaps and uncertainties remain. Therefore,
in the following points, I shall limit myself to pointing out certain traits
that in principle can be definitively attributed to Carian.
1. Carian seems to share with other Anatolian languages the absence
of initial r-: there is no example of R- in Greek adaptations of Carian
personal and place names, and the only definite example of initial r in
Carian is found the name rtim, in the new inscription of Hyllarima (to
which rtmi of Tralleis can be added if a segmentation sdia rtmi is pre-
ferred to sdi artmi). In this case, a defective vowel notation (rtmi for
/artmi/) or a syllabic r produced by aphaeresis could be the cause
/ºtmi/ (for this latter possibility, cf. the similar explanation given by
Lycian rMmazata in Melchert 1994:297). In any case, it seems certain
that the possibilty of rV- at the beginning of a word does not exist in
Carian.
2. As stated above, there is no definite example of the letter l in
an initial position (Adiego 1993a:276). This restriction is consistent with
the geminate origin of the sound represented by l.
3. A similar tendency can be observed when considering d, as evi-
dence of (possible) initial d is very scarce. Our glossary only contains
three forms: dar“qemorms[, den, drual. We should also note that for two
examples a connection with two etymologically related function words
has been suggested (den as preposition comparable to Hitt. andan; drual,
segmented in d = rual, where d would be a preposition comparable to
Lycian ñte, see Chapter 11 ss. vv. for details). Also in this case—like l—
 PHONOLOGICAL FEATURES 255

the origin of d from a consonant cluster -nd- is a good explanation for

this restriction: in general, this cluster is preceded by a vocalic sound
(-Vnd-), which limits the appearance of d < -nd- in absolute initial po-
sition, which is then only possible if the vowel disappears secondarily
or is not noted graphically. A similar situation can be envisaged
for b. The (also scarce) examples of an initial b, when etymologically
clear, point to a preceding Vmb- sequence: brsi < ibrsi (also attested) <
*imbrsi-.
4. Note the following clusters in internal positions, documented both
in Carian and in Greek sources:
-bl- = -bl-: qÿblsi- = Kublisse›w
-b- = -mb-: ib(a)rsi = Imbarsiw, Imbrassiw
-kr- = -gr-/-kr- = pikre/pikra- = Pigrhw, Pikrhw
-∞s- = -j-: ly∞se/ly∞si- = Lujhw
-ks- = j: uksmu, wksmu-≤ = Omajamoaw
-mn-/mñ- = -mn-: ktmno-, ∞tmño = Ekatomnvw, somne = Svmnhw
-ñm- = -nm- = pñmnn- = Ponmoonnow
-rd- = -rd-: ardybyr- = Arduberow
-rl- = -rl-: arli“ = Arlissiw
-rm- = -rm-: armo- cf. Ermapiw
-rn- = -rn-? alos ∞arnos, cf. Alikarnassow?
-rq- = -rg-/-rk- : yrqso- = Urgosvw; trqude = Tarkonda[
-rt- = -rt-: artay- = Artaow, Arthumow
-rs- = ibarsi = Imbarsiw
-tb- = -tb-: qtblem- = Kotbelhmow, Kutbelhmiw
-d- = -nd-
-db- = -ndu-? kidbsi- cf. Kinduh?
Other internal clusters seems also to have existed: for instance -rk- in
“arkbiom, or -dr- (note idrayridsemd?bq, uodrou, uodryia[ in Carian and ÉIdrieÈw
in Greek sources).
5. For initial consonant clusters, the evidence is less certain: unlike
interior clusters, there are no clear examples appearing simultaneously
in both direct and indirect sources, with the exception of k≤atÿbr = Lyc.
Janduberiw (Zgusta KPN § 1061), which points to an initial k≤- (cf. also
Jermedu<be>row as an example of initial k + sibilant cluster). Some evi-
dence does exist of an initial cluster such as kb- (Carian kbidn = Lycian
Xbide), and—although very limited—for clusters of the type s + stop
(skdubrotoz≤, sqla, sqlumidun besides sp-, sk-: Skoaranow, Spareudigow).
Greek sources also contain several cases of stop + liquid beginnings
256 CHAPTER SIX

(br-, gl-, kr-, pl-, pr-, tr-),11 but there are no definitive examples in
Carian, in the sense that practically all the examples can be alternatively
interpreted as sequences of stop plus syllabic liquid (note for example
prflidas, prpwri∞, trqude) or as cases of defective vowel notation (qlali- =
Greek Kolaldiw, etc.).

B. OVERVIEW OF THE HISTORICAL PHONOLOGY OF CARIAN

The content of the following pages must be considered purely provi-

sional. Our present knowledge of Carian is such that we must be cau-
tious when attempting to sketch an overview of this kind. It must be
noted that most of the information that can be obtained from Carian
sources comes from Carian proper names, and onomastics is not always
a good tool for comparative research: firstly, because the etymological
interpretation of proper names can only be based on formal criteria,
since they lack an actual meaning; secondly, because we cannot be sure
that all the personal names correspond to the actual language, as some
of them could come from nearby languages, and others could betray
the conservation of old traits, modifications caused by analogical processes,
and other such characteristics. In any case, both problems can be
avoided to a certain extent when the volume of evidence and internal
consistencies makes it logical to interpret them as a true reflection of
Carian language. But I deem it necessary to issue this warning when
one comes to evaluating the information that follows.
I adopt the reconstruction of Proto-Anatolian (PA) as it appears in
the fundamental work of Melchert (Melchert 1994), which currently
represents the most useful tool for comparison.
It is impossible to give a complete account of the outcome of PA
sounds in Carian, with particular difficulties being found in the vocal-
ism. Evidence is extremely limited in many cases, due to the defective
vowel notation, and the complexity of the treatments of vocalic sounds
from PIE to PA and from PA to the particular Anatolian dialects. In
other cases, the evidence is simply nonexistent. Here I shall merely
indicate the more relevant aspects that can be identified from our cur-
rent knowledge of Carian.

11
There are, however, few examples.
 PHONOLOGICAL FEATURES 257

1. Vocalism
(1) One of the few traits that can be established is of great relevance
to the position of Carian, since it supports the theory that Carian
belongs to the group of “Luwic” dialects of Anatolian (Luwian, Lycian
and Milyan). This is the change of PA *è to /i(:)/. This is demonstrated
above all by the root pik-/-bik- < PA */be:H/- < PIE *bhè h2- ‘to shine’,
present in different Carian names (dbiks, pik(a)rm, pikre . . .).
(2) A unique characteristic of Carian is the emergence of a rounded
front vowel y, ÿ /y/ and its semivocalic counterpart /w/. This is a con-
ditioned change, but it is impossible to identify all the precise contexts
in which it takes place. However, at least a trigger for the fronting
seems clear: the immediate contact of original /u(:)/, /w/ with /i/,
/j/. We come across examples such as yiasi, yi≤∞ ?biks∞i, [—]ryin,
uodryia[, yjas[i≤], “ÿin≤ (2×), and ∞diye≤, siyklo≤, iÿkr≤ that point to */wi/
> /wi/, */uj/ > */yj/, */iw/ > /iw/, */ju/ > /jy/. Particularly strik-
ing is the Kaunian alternation yomln (C.Ka 4) / uiomln (C.Ka 5). It is
very likely that both spellings are attempts at representing /wi( j)o/-.
The absence of a specific grapheme for the semivowel /w/ in the alpha-
bet of Kaunos would explain the graphical oscillation. Note that ui is
only attested in Carian in this word, and that there are no examples
of wi, uj in the entire Carian documentation. It is also likely that other
examples of y, ÿ before a vowel other than i should be explained in a
similar fashion, as yomln: idyes≤, ∞aye, terÿez≤, ÿasd≤. Note particularly the
last example: it is tempting to bring it closer to yiasi, yjas[i≤] and to
reconstruct a protoform */wijas(V)nd/-.
For the remaining examples of y, ÿ, the case is not so clear, but the
influence of a near i, j can be envisaged: in the family of names in
ydiq/yriq (see below pp. 262–263), a sort of metaphony caused by the
i in the following syllable could be suggested. Even in ylarmit, one could
imagine a more distant assimilation, triggered by the final i, or rather
by the (un-notated) i present in the Greek form of the place name,
Ullãrima (*/ularim/º > /ylar(i)m/º. In the case of ethnics formed with
the Luwic suffix CLuw. -wanni, Mil. -wñni, kbdyn“, mdayn/mdaÿn, the con-
nection with [—]ryin cannot be overlooked, but the exact interpreta-
tion of y/ÿ in these forms remains unclear: is y/ÿ here used for /wi/,
like in yomln? Or rather has /wi/ coalesced into a single vocalic sound
/y/? A third possibility would be to ascribe the fronting process to
metaphony, by postulating a *-uni- > -yn(i) evolution.
258 CHAPTER SIX

(3) Another vowel that seems to have appeared secondarily in Carian

(at least for some words) is o. It appears to come from */a:/ and also
from */a/ in accented syllable (where */a/ would become a long vowel,
a change that is typical in Anatolian languages. Forms such as armo- (in
armotrqdosq, E.Hy 1a) < *armà-12 would also suggest this process. A sim-
ilar explanation could be given for ntro < *n(e?)trà (cf. perhaps for the
formation Lyc. kbatra < ºtr + à). In general, Carian names with -o could
be explained as original stems in - à: note particularly plqo and ksbo, where
-o is clearly accented. The latter form can be etymologically related to
the Lycian xahba, ‘grandchild’, also an à-stem. Finally, Kaunian otr“,
‘themselves’, as well as Lyc. atra-, HLuw. atra/i- also point to an accented
a > o, although the etymological origin of these words remains unclear.
This leads to an interesting explanation of -ol (Greek -vllow, -vldow)
names: Assuming Melchert’s explanation that Luwian -alla- forms come
from*-élo- with ’op’s law (*-élo > *-álla-), in Carian a lengthening of á
and subsequent change to o could have taken place *-élo- > *-állV- >
*-àllV- > *-o÷ll(V), spelled -ol (with l for, or the result of, geminated l,
see above p. 249). The forms with -el ( para-ibrel≤, ionel≤ ) ought to be
explained then as the result of a suffix with i-motion followed by
metaphony caused by i (see below): *-éli- > *-álli- > *-é(:)ll(i)- > *el.
In the case of -on in mwdon≤, a monophthongization process had been
suggested (see Adiego 1994:94) *mwdawn≤ > mwdon≤, but I now prefer
to look for another solution, given the form mdaÿn/mdayn, where the
monophtongization process has not taken place. Perhaps o here also
represents *à, which in this case comes from the contraction *ºa-wan- >
* ºa-an- (with loss of w) > *ºàn- > ºon-. In mdayn/mdaÿn, the lack of con-
traction can be attributed to a different vocalism of the suffix (*ºa-wen-),
whether original or originated by metaphony (*ºa-wen-i- vs. *ºa-wan-
as- in mwdon≤ ?). The other examples of possible monophthongi-
zation cited in Adiego (1994:49) (arliom≤ < *arliya-uma, etc.) are by no
means conclusive. It could also be argued that they come directly from
*à (> o).

12
That arma was an a-stem in PA seems the most likely explanation, although the
evidence is not certain: note Lyc. N304, 5 arMma, wherein the isolated context of the
form does not allow us to confirm that in Lycian it was also an a-stem (Melchert, DLL
s. v.).
 PHONOLOGICAL FEATURES 259

(4) The vocalism of the words en, ‘mother’, and ted, ‘father’, is a clear
indication of an umlaut process a > e / _.Ci similar to that of Lycian:13
PA sg. nom. *anna-s (cf. Hittte anna-“ ) > Luwic *anni-s (with i-motion,
cf. CLuw. anni-“ ) > carian en, like Lycian ẽni; PA *dáda-s > Luwic *tádi-
s (with i-motion, cf. CLuw. tàti-“ ) > Carian ted, like Lycian tedi. Cf. the
similar explanation above for -el.
(5) The form otonosn shows a strange ‘vocalic harmony’ in o. At least
the second and third vowels can be explained with *à > o (*ÉAyhn*a- >
*atànà-,14 cf. Lycian atãnaze/i ) > *atono-), but the first seems to be caused
by metaphony (*atono- > otono-).

2. Consonants
(1) The PA labial and dental voiceless stops remain unaltered in
Carian:
PA */p/ > Carian p /p/: para(eym), para(ibrel≤), Para- < *PA prò/prò:
cf. Hittite parà ‘forth’. Cf. pun-/pn- in punw≤ol≤, pnu≤ol, perhaps also in
punot2, < Luwic puna- ‘all’ (Lyc. punãma- ‘totality’, CLuw pùna- ‘all’).
*/t/ > Carian t /t/: trqud-e, trqd-os < PA *TºH–t- ‘Storm-god’ (CLuw
Tar¢unt-, Lyc. Trqqñt-). -t < -te or -ti < PA pret. 3rd sg. *-to or pres.
3rd sg. *-ti in ÿbt < *ubete ‘offered’ or < *ubeti ‘offers’, cf. Lyc. ubete.
(2) Like the other Luwic dialects, Carian presents a ‘satem-like’ treat-
ment of PIE, PA *∞, as can be seen in the demonstrative pronoun sa-
/sn- ‘this’ in sa, san, snn < PA *∞o- (Hitt. ka-, Luw. za-). Perhaps also sidi,
sdi ‘tomb’, if it can be connected with PIE *∞ei- ‘to lie’ (Lyc. sije-).
(3) Luwic also deals with *· > Ø in *·emro- ‘steppe’ > *imr- >*imbr- >
ibr-/br-: (i)br-si < *imbrV- < *·emr-, (para) -ibrel-≤ < *·emréli-.
(4) The only clear example of the treatment *kw is the (original) rel-
ative pronoun ∞i < PIE, PA *k wis (Hitt., CLuw. kui“, Lyc. ti, Mil. ki
[ci]). in this case Carian displays a process similar to Milyan: delabi-
alization and fronting before i.
(5) Forms like tedi < PA *dáda/i- or pik- (in pikre-≤, pikarm-≤, etc.) in
addition to dbiks and dbikrm point to an ‘unvoicing’ of voiced stops in
initial position, a process that has also taken place in parallel to this
in other Anatolian languages (cf. Lycian tedi, Lydian taadas).

13
Melchert (1994:296).
14
For the probable non-Doric origin of the form (according to Blümel), see above
p. 237. The non-existence of e in Kaunian is sufficient to explain the adaptation of
Greek h to a.
260 CHAPTER SIX

(6) In the sequences of Nasal + voiceless stop, this latter example

has become, as in Lycian (and probably in the rest of the Anatolian
languages), a voiced stop. The evidence is clear for dentals: *TºH–t- >
trq(u)d-, with *-nd- > d, see above.
(7) Traits (5) and (6) clearly indicate that the situation for Carian is
very similar to that of Lycian: voiced stops merge with voiceless stops
in initial position, become fricative in intervocalic position, and remain
as voiced stops only after a nasal, where voiceless stops merge with them.
(8) Carian liquids l, r and nasals m, n come from the respective liquid
and nasal sounds in PA and/or in Luwic:
r < *r: para(eym), para(ibrel≤), Para- < *PA prò/prò: cf. Hitt. parà
‘forth’, armo < PA *armà- ‘moon’, “ar-, “r- ‘upper’ (in “ar-u≤ol, “r-wli-≤,
etc.), cf. Hitt. “èr, CLuw. “arri, Lyc. hr-i, Mil. zri-
l < *l: wljat. Cf. Hitt. walliwalli- ‘strong, powerful’
m < *m: msn-ord≤. *msn- ‘god’: Cf. in the rest of Luwic languages:
CLuw. mà““an(i)-, Lyc. maha(na)-, Mil. masa-, Sidetic ma≤ara (pl. dat.)
‘gods’; *mu- (in uksmu-) = CLuw., muwa- Mil. muwa- ‘might, power’.
n < *msn-ord≤ < msn- ‘god’, see immediately above; -n acc. sg. end-
ing: lysikla-n, ork-n, etc. < PA *-n < PIE *-m; -yn- suffix for ethnic names
(kbd-yn-“ ‘Kaunians’) = CLuw. -wanni-, Lyc. -ñni-, Mil. -wñni-.
As for ®, l, ñ, see the remarks in pp. 248–250 above. l seems to
point to a geminate *-ll- (*-élo > *-állV- > -ol, see above). ® can be a
palatalized r, which in some cases could come from l (qdar®ou≤, cf. CLuw
¢utarlà- ‘slave’?). ñ comes from n at least in the ending -ñ of pñmnn-≤ñ
(acc. sg of a possessive adjective).
(9) PA ‘Laryngeal’ *H (PIE < *h2) appears as a tectal voiceless stop
k, q in Carian. The process is therefore parallel to that of Lycian (<x>
/k/ / <q>). Examples: for *H > k: kdou- (< *Hntawº-, cf. Lyc. xñtawat(i),
from PIE *h2ent-), pikre-, piks- (both containing PIE *bhèh2- > Luwic *piH-);
for *H > q: trqude, trqd- < Luwic *TºH–t- (CLuw. Tar¢unt- Lycian,
Milyan trqqñt-). If the word quq comes from PA *HuHo- ‘grandfather’
(> Lycian xuga), the lenition process seen in Lycian ( xuga = /kuga/) is
absent, at least graphically, in Carian. But the Greek form Gugow points
to a voiced articulation (the initial G- remains unclear).
(10) The existence of at least three fricative sibilants in Carian (s, “,
≤ ) poses a puzzling situation regarding their respective origins. It seems
that (at least some) results of the three sibilants come ultimately from the
single fricative voiceless dental fricative PA *s. Recent work by Melchert
(see Melchert 2002) has significantly clarified the matter, and new evi-
dence from the Hyllarima inscription seems to corroborate his views.
 PHONOLOGICAL FEATURES 261

Concerning Carian s, Melchert (2002) has stated that it reflects a

prehistoric simple *s in forms such as i[—]inis (C.Ka 5), ntros, “arnais,
etc., in which he claims to have identified genitive-dative forms coming
from PIE *-e/oso (Melchert 2002:309). This hypothesis has essentially
been confirmed by the new inscription of Hyllarima, wherein an analy-
sis of the ending -os in armotrqdos would suggest that it originates from
a genitive ending *-Vso, or from an adjective suffix *-Vso-.15 Another
possible new example of Carian s from *s is the name ksbo in C.My
1, if it is related to Lycian xahba ‘grandchild’ < PIE *h2onsu secondar-
ily converted to an à-stem.
As for “, the clearest examples for a conditioned origin are “(a)r- and
acc. pl. -“´(in otr“, kbdyn“, sarni“ ). The first example could be interpreted
as the result of a palatalisation process caused by direct contact with
r (cf. perhaps also zri-, not sri- in Milyan). One must therefore start
from *sri- (cf. Mil. zri-, Lyc. hri-, CLuw. “arri ) and to postulate *sri- >
*sº - > “º - (palatalisation) > “ar- (samprasàra»a):16 the second example is
the outcome of Luwic accusative plural *-ns (> -“ ). The parallelisms
with the use of z in Milyan are striking: there we find *sri- > zri- (zri-
gali, zriqali) and nom.-ac. pl. -z < *-ns (masaiz, Xbadiz, xuwasaz, etc.) vs.
Lycian hri, (pl. acc.) -s. Other instances of “ are not so clear (mol“, k“ow“ ).
Finally, in the case of “, Melchert proposes that Carian possessive-
genitive -“ continues the PIE possessive suffix *-asso- in its form -assì,
i.e. with i-mutation (Melchert 2002:311). From a phonological point of
view, the supposition is extremely compelling, given the likely palatal
character of ≤ /ç/ (see above), although some doubts do remain.17

3. Some Secondary Changes

In this brief section I offer a succinct analysis of several cases of sec-
ondary changes observable in the Carian documentation. Once again,
the material and results are far from conclusive, and must therefore be
viewed with a certain amount of caution.

15
The exact analysis of -s and its actual value in Carian is not relevant here, see
pp. 314–317 for the problems posed by the Carian s- ending.
16
I consider this interpretation of “ in “(a)r- as preferable to Melchert’s use of Lydian
serli-/selli- (note that Lydian <s> = /“/!) for explaining the palatalization. The devel-
opment of a secondary support vowel in contact with syllabic r is also visible in pikarm-
≤ vs. pikrm-≤ (from pikrº, cf. pikre-≤ ) and the Milyan testimony seems then more compelling
than the need for postulating that Carian “ar- comes directly from *ser-.
17
No explanation has yet been found for the form ib(a)rsi- if it comes from an -assì-
262 CHAPTER SIX

1. The family of names brsi/ib(a)rsi, (para-)ibrel, Greek Imbras(s)iw,

Imbarsiw, Imbarhldow, if the etymology from PA*·emro- is accepted,
would indicate a sound change *-mr- > -mbr- (<br>), a kind of epenthetic
development well known in other languages (cf. Greek êmbrotow < *amro-
tos < PIE *–-mºto-s, Spanish hombro [ombro] from Vulgar Lat. *umru(m)
< Lat. umerum).
Perhaps a similar process can be identified in the names Andarsvw,
Androssvw (cf. also dar“ ?): it is tempting to start from *narasà- (to be
related to Narasow, epithet of Zeus in Panamara, Caria)18 > *–rasò- >
*–drasò > * –dºsò > Andarsvw, Androssvw.
2. The examples mentioned above (1) point to a secondary charac-
ter of a in ibarsi- (Greek Imbarsiw), given that epenthetic b could only
appear in direct contact with the following r. Cf. also Andarsvw, dar“
if the explanation proposed here is accepted. It seems that in Carian
a kind of samprasàra»a could have taken place when r became syllabic,
and the entire process could therefore be the following: *imrasi- >*
imbrasi - > *imbºsi- > /imbarsi/º (ibarsi-). The same rule can be applied
to alternations such as pikrm-/pikarm-, and “ar-/“r- (cf. above for the effect
of this explanation on the merging of palatal “ in these latter forms).
3. It is plausible to imagine a common origin for the collection of
stems ÿdiq/ÿd∞-/yriq-/yri∞-/wri∞- if one accepts a progressive dissimila-
tion process. The following are all the forms that have been identified
thus far (E = in Egypt; C = in Caria):

(perhaps re-derived by -*iye in *-assiye-) suffix, because the phonological context is very
similar to that assumed for the origin of -≤. The possibility that these forms contain a
different suffix (*-∞o- or *-tyo-, see Melchert 2002:310 n.13) cannot be ruled out, but
it is a more ad hoc solution.
18
For this form a connection with CLuw. annara/i- ‘strong’ can be envisaged (Neumann
1994:22; see here p. 333).
 PHONOLOGICAL FEATURES 263

-d- -r-
Greek ÿdiq- ÿd∞- Greek yriq- yri∞- (wri∞-?)
yri∞-ñ (?) C
idyri∞-≤ C
Paraudigow parÿd∞-≤ E paryri∞(-≤) C
prpwri∞? E
rather prpÿri∞?
“aÿdiq-≤ E Saurigow “ayriq E
Senurigow
Spareudigow
Semeuritow?

It is possible that the original form of the stem was *yriq-/yri∞-,19 as

the non-compound form yri∞-ñ suggests. The forms with d are limited to
par(a)-, “a- and Spare-compounds, and for the two first, the correspond-
ing forms with r are also attested. If one assumes that “a-/Sa- is a vari-
ant of the well-known adverbial stem “ar-/“r-, originating from a loss of
r in intervocalic position (*“ar-yriq > “a-yriq, cf. perhaps Saussvllow besides
Sarusvllow), the resulting forms in d in compounds with par(a)-, “a(r)-
could be explained as a dissimilation r-r > r-d: par-yri∞- > *par-ydi∞ ( par-
ÿd∞-)-, *“a(r)-yriq (cf. “a-yriq) > *“a(r)-ydiq (“aÿdiq-).
4. Schürr has argued in favour of a change p > Ø from some alleged
examples of alternation (Schürr 1992:141). Perhaps the most compelling
example of this possible alternation would be the words for ‘stela’ (or
similar) in Memphis: upe/ue, although it is also possible that we are in
fact dealing with two different words. Less convincing is the name
paraeym vs. parpeym, where a different second element (ºeym/ºpeym) could
also be suggested.20 No other clear examples are known of, either in
Carian direct sources or in Greek ones.
5. Other examples of vowel and consonantal alternations, in Carian
direct sources or in Greek indirect sources, are far more occasional,
and not easy to explain. Note for example a/e in pikra/pikre, upa/upe,
which could be a morphological rather than a phonological alterna-
tion. For changes detectable in Greek sources, see Neumann (1994:18–19).

19
The alternation q/∞ remains unexplained (cf. supra pp. 244–245). As for y/ÿ, see
above pp. 235–236.
20
A further example, qarpsi- vs. qarsi-, offers a totally different context, and cannot
be compared to the two cases mentioned above, where p would be lost between vowels.
 CHAPTER SEVEN

ANALYZING CARIAN INSCRIPTIONS

Contrary to the customary order in grammars, where morphology,

together with phonology, precedes syntax, in the case of Carian it seems
methodologically more accurate to begin with the analysis of Carian
texts; our knowledge of Carian morphology depends on the way in
which the texts can be interpreted syntactically, and such an interpre-
tation remains in most cases controversial, to say the least. In the fol-
lowing pages I will try to analyze Carian texts from the ‘easiest’ to the
‘most difficult’, beginning with those that contain only very basic ono-
mastic formulae. The following step will be to analyze the inscriptions
of the Memphis sub-corpus, where we find more complex onomastic
formulae, but no recognizable verbal forms are attested. This analysis
will also allow us to identify some common nouns used in the formu-
lae of these funerary texts. The third section will offer an analysis of
some brief inscriptions (mainly on objects) that seem to contain forms
other than onomastic formulae. Finally, a few aspects of the interpre-
tation of the longer inscriptions, where there are serious difficulties of
analysis, will be briefly addressed.

A. BASIC ONOMASTIC FORMULAE

1. Inscriptions Consisting of Only an Individual Name

The briefest Carian inscriptions consist of a single word, which in the
great majority of cases can be confidently interpreted as an individual
name. As one can easily imagine, these types of texts are found among
the graffiti and also on some objects. In the graffiti, the name appears
without an ending, in a case that we can consider the ‘nominative’, a
logical explanation given that the inscription can be interpreted merely
as a type of signature: pisiri (E.Ab 1), piew (E.Ab 38), pla?t (E.Th 3),
wljat (E.Th 7), psma≤k (E.Si 7, E.Bu 5) etc. Only occasionally does the
 ANALYZING CARIAN INSCRIPTIONS 265

name appear with the -≤ ending, commonly interpreted as a ‘genitive’:

E.Ab 32 ∞arr≤, E.Ab 37 “arur≤.1
Also, in the so-called ‘pharaonic objects’ (excepting funerary stelae)
published in Masson-Yoyotte (1956) we come across inscriptions con-
sisting of only an individual name, in nominative or in genitive: wliat
(E.xx 2) pduba (E.xx 4,) vs. ionel≤ (E.xx 3).
The only possible examples of a form in another case—leaving aside
forms with no clear interpretation—are provided by the two identical
inscriptions on bracelets that contain the word kdu≤ol“, where the final
-“ could represent a different case ending.

2. Inscriptions Consisting of Only a Twofold Onomastic Formula

Despite the above examples, the most typical Carian onomastic for-
mula, as in other ancient Indo-European languages, seems to have been
a twofold one: individual name + father’s name. This latter is system-
atically expressed in Carian by the genitive ending ≤.2 This formula can
explain a number of Carian inscriptions consisting of two words, as
well as others in which a particle ∞i appears together with the two
names.
We can attempt a typological classification of twofold formulae on
the basis of the case of the first member and the presence or absence
of ∞i:
Nominative + Genitive
panejt iarja≤ E.Ab 2
ptn“e | ibarsi≤ E.Ab 3
“amow ltari≤ E.Ab 4, 5
plat | pals≤ E.Ab 7, 8, 9
piubez qurbo≤ E.Ab 10
untri | uantrpu≤ E.Ab 13
pdubez or≤ E.Ab 15
tamosi | inut≤ E.Ab 18
tamosi utnu≤ E.Ab 19
ninut | tamosi≤ E.Ab 20
ttubazi kattÿri≤ E.Ab 25
ialli | q∞blio≤ E.Ab 40

1
The interpretation of the ending -z (-≤ ) as a genitive was first made by Sayce
(1887[92] = 1893:141–142).
2
For the use of an ending -s- in C.Ka 5, see below p. 316.
266 CHAPTER SEVEN

ttbazi kt?tri≤ E.Ab 41

dÿbr | t®∞atr≤ E.Th 5
psma≤k ibrsi≤ E.Bu 4
Nominative + Genitive ∞i
pdnejt qÿri≤ ∞i E.Sa 1
pnu≤ol zmu≤ ∞i E.Me 19
(a) “enurt (b) p∞simt≤ ∞i E.Me 50
platt slaÿ≤ ∞i E.AS 5
s–ending (?) + Genitive ∞i
ap[---]ws a[rb]ikarm≤ ∞i E.Me 23
idmns | myre≤ ∞i E.Me 33b, cf. E.Me 33a
As the examples show, the twofold formula is typically used in graffiti,
particularly those from Abydos. In practically all of the quoted exam-
ples of graffiti, with the sole exception of E.AS 5, the formula consists
of the individual name in nominative followed by the father’s name in
genitive. E.AS 5 is the only one that contains the genitive accompa-
nied by a postclitic ∞i. In this and the other cases where the structure
N-Ø N-≤- ∞i appears (E.Sa 1, E.Me 19, E.Me 50), there is no visible
semantic difference when compared to the structures without ∞i. The
role of ∞i in all of these cases seems to be merely to connect the nom-
inal complement in genitive to the name it refers to. The most likely
explanation, as Hajnal (1997a) suggests, is that this construction pre-
supposes and/or comes from an elliptical word for ‘son’: pdnejt qÿri≤ ∞i
‘Pdnejt, the (son) of Qÿri- = ‘that of Qÿri-’.3 Given that from an ety-
mological point of view, an origin of ∞i from a Proto-Anatolian rela-
tive pronoun *kwis (< PIE *kwis) is a convincing interpretation, it is easy
to assume that behind pdnejt qÿri≤ ∞i and similar constructions, there
was an original meaning ‘Pdnejt, who (is the son) of Qÿri’.
As we can see, twofold formulae are very scarce among the funer-
ary stelae of Memphis, but the examples are very interesting in that
they show another pattern of the formula; as well as N-Ø N-≤ (3 exam-
ples), we can identify two instances with possible ‘s-endings’ (N-s N-≤,
both with the genitive followed by ∞i), but we should bear in mind the
possibility that we are in fact dealing with nominatives of s-stems (on
this problem, see pp. 314–317).
This twofold formula ‘individual name + father’s name in genitive’
is easily recognizable in longer inscriptions, which we will address below:

3
We can conclude that the second name represents the father’s name from the
Egyptian text that accompanies the Carian one, where P3-dj-Njt is mentioned as the
son of K3rr = qÿri-.
 ANALYZING CARIAN INSCRIPTIONS 267

in votive texts such as the inscription on a bronze lion (E.xx 7 uksi

wrm≤ ), or the inscription on a phiale (C.xx 1 “rquq qtblem≤), in the list
of names from Mylasa (C.My 1), and even in the mention of the two
satraps Idrieus and Ada ([--]ryin ∞tmño≤ sb ada ∞tmno≤ ‘[--]ryin of Hekatomnos
and Ada of Hekatomnos’).

B. THE STRUCTURE OF THE STELAE FROM MEMPHIS

The onomastic formulae limited to the two basic constructions men-

tioned above (‘Individual name’ or ‘Individual name + father’s name’)
are extremely scarce in the Memphis sub-corpus: to the examples men-
tioned above (four, and only for twofold formulae), we can add only
one other in which the individual name in the genitive is accompanied
by the father’s name and one of the formula words in Memphis, ue,
used to designate the object (‘stela’; on this word see Chapter 11, s. v.).
arli“≤ | psikro≤ ue E.Me 51
In general, the stelae from Memphis, insofar as they are integrally (or
almost integrally) preserved, show more complex onomastic formulae.
It is common to find threefold formulae, as well as another type of
formulae that includes more proper names.

1. Threefold Formulae
To the ‘individual name + father’s name’ formula, a third word in
genitive can be added. The simplest interpretation of these threefold
formulae would be to take the third name as the grandfather’s name
(papponym). As we will see below, this interpretation is certain in cases
such as the threefold onomastics formulae of the new inscription of
Hyllarima (C.Hy 1a), but in the Memphis sub-corpus, where this type
of formula is very frequent, (in fact, it can be considered the typical
onomastic formula of the funerary stelae) the situation is not so clear.
In Memphis we find the following three types of threefold structures:
N-Ø N-≤ N-≤, N-≤ N-≤ N-≤ and N-s N-≤ N-≤ (with or without ∞i after
the second and/or the third name). When the individual name is in
genitive, a word for ‘stela’ (ue, upe,4 wpe, upa) can appear (cf. above the

4
In one case (E.Me 26) accompanied by a demonstrative pronoun: upe sa ‘this stela’
(on sa, see Chapter 11, s. v.).
268 CHAPTER SEVEN

example of E.Me 51), so that the use of the genitive simply indicates
that the stela belongs to the individual it mentions.
The inscriptions consisting of only these kinds of threefold formulae—
I leave aside for now the inscriptions of a more complex structure—
are the following:5
Nominative + Genitive + Genitive
uksmu | lkor≤ | mrsi≤ E.Me 2
tamou tanai≤ qarsio[-?] ? E.Me 7
irow | pikarm≤ | mwdon≤ E.Me 14
u≤ol | mi∞≤≤ kdûsi≤ E.Ab 35
Nominative + Genitive ∞i + Genitive
“aru≤ol pleq≤ ∞i : ≤ugli≤ E.Me 30
qorb | isor≤ ∞i | ≤ugli≤ E.xx 1
Nominative + Genitive ∞i + Genitive ∞i
uqsi | “rwli≤ ∞i | mwdon≤ ∞i E.Me 20
plqo | pikrm≤ ∞i | mwdon≤ ∞i E.Me 40
Genitive + Genitive + Genitive
ttbazi[≤] | piub[a]zi≤ | aor[≤] E.Me 1
pikre≤ ue “arwljat≤ msnord≤ E.Me 3
arli“≤ urs∞le≤ kidbsi≤ E.Me 15
[. . .]u≤ | upe sa | triel≤ | mrsi≤ E.Me 26
s[--]et≤ | [--] | ynemori≤ | mwdon≤ E.Me 29
me®≤ | somne≤ | t®∞ata[r]≤ E.Me 34
| or≤ | wpe | qdar®ou≤ | t®∞atar≤ E.Me 41
Genitive + Genitive ∞i + Genitive
arli“≤ : upe : arlio[m≤] ∞i : yjas[i≤] E.Me 9
wksmu≤ | wpe | lkor≤ ∞j qarpsi≤ E.Me 36
[--]j[-]≤ [-]owt≤ ∞i : msnord≤ E.Me 48
Genitive + Genitive + Genitive ∞i
punw≤ol≤ : somne≤ qÿblsi≤ ∞i E.Me 21
(a) ÿasd≤ | yi≤∞?biks∞i≤ (b) mwdon≤ ∞i E.Me 46
Genitive + Genitive ∞i + Genitive ∞i
“dtat≤ | upa | w | wet≤ ∞i | mwdon≤ ∞i E.Me 13
sanuq≤ | ue | pntmun≤ ∞i mwdon≤ ∞i E.Me 28
s–Ending + Genitive + Genitive ∞i
ntokris | dw≤ol≤ | mwdon≤ ∞i E.Me 35

5
I add to this list an example from Abydos (E.Ab 35), and also the inscription of
unknown origin (but clearly close to the Memphis stelae) E.xx 1, which is consistent
with this type of threefold formula.
 ANALYZING CARIAN INSCRIPTIONS 269

The possibility that the third word is not a papponym can be deduced
from at least three factors: firstly, the very large number of examples
of the word mwdon≤ in that position, a word for which there are no
parallels among the Carian personal names in Greek sources; secondly,
the iteration of some words in that very position (≤ugli≤, msnord≤, mrsi≤ );
thirdly, the fact that these words never appear as a first name in any
onomastic formula, and even their appearance as a second name or
patronym is limited to the word t®∞atar≤, which appears in E.The 5 as
a clear patronym (dÿbr t®∞atr≤ ). These distributive properties do not
favour the interpretation of most of these forms as simple personal
names functioning as papponyms.
An alternative interpretation that was envisaged some years ago is
to classify them as ethnic names.6 This possibility is very clear in the
case of mwdon≤; there are few doubts that mwdon≤ is the genitive cor-
responding to the nominative mdayn, mdaÿn also found in the Memphis
sub-corpus, and following the discovery of the Kaunos bilingual inscrip-
tion, -yn/-ÿn has been confirmed as a suffix for the formation of eth-
nic names (kbd-yn-“ ‘Kaunians’).7 The exact meaning of mwdon≤ is a
different question, for which I refer to the Glossary (Chapter 11), where
a discussion of the various proposals of interpretation is offered. For
now it is sufficient to state that it could simply mean ‘foreigner’ or, if
it refers to a concrete place, that this must be the main point of origin
for Carian mercenaries in Egypt, given the high number of occurrences.
For some other third-position words, the possible connections with
well-known place names have not gone unnoticed (see Janda 1994:
174–176; Melchert apud Adiego 1995:20; Adiego 2004:310): ≤ugli≤ <
Souaggela, kidbsi≤ and/or kd!usi≤ < Kinduh, yjasi[≤] < Iasow, msnord≤ <
Maosanvrada, ksolb≤ < Kasvlaba, qÿblsi≤ < Kubliss/ow/. There is also
the interesting possibility of recognizing certain suffixal formations: -si-
in kidbsi≤, kd!usi≤, and also probably in mrsi≤, qarpsi≤; -i- in yjasi[≤] and

6
In Adiego (1993:212), the hypothesis formulated by Meriggi (1980) of interpreting
mwdon- as an ethnic name, was already taken into consideration, but the first author
to propose that a number of third names in onomastic formulae could be interpreted
as ethnic names was in fact Janda, see Janda (1994:174–176). The idea was also taken
up by Melchert (apud Adiego 1995:20) and further developed in Adiego (2004:309–310).
That I did not mention Janda’s work in this latter paper is a regrettable oversight, for
which I ask forgiveness.
7
For mwdon“ as an ethnic name, see Adiego (1993:212). For integration of mdayn/
mdaÿn—mwdon“ into the same paradigm and for the identification of the Luwic ethnic
suffix, see Melchert (1993:82–83).
270 CHAPTER SEVEN

qÿblsi≤. For -si-, a connection with Lycian -zi- is likely. For -i-, cf. Lyc.
-i( je)- (in Tr◊mmili( je)- ‘Lycian’), and above all the Carian form ylarmit,
which seems to contain an -i- suffix attached to the place name *ylar(i)m(a)-
= Hyllarima in order to express the meaning ‘Hyllarimean’. In the
case of msnord≤, ksolb≤, no suffixation can be recognized, perhaps due
to the defective vowel notation. It is rather puzzling that none of these
forms, leaving aside mwdon≤, show the typical -yn-/-on- ethnical suffix,
but this could be due to chance (cf. the great variety of ethnical suffixes
also present in Lycian).
A new argument can be added to this evidence for ethnic names:
there are three inscriptions in which a N-Ø N-≤ N-Ø formula is found,
which means that the third name can hardly be a papponym, since it
agrees in nominative case with the first name. Moreover, in two of the
three cases, the third word can be interpreted as an ethnic name:
“ayriq | parpeym≤ ∞i yiasi E.Me 25
idmns | myre≤ ∞i | mdayn ∞i E.Me 33a
triqo : parma≤≤ ∞i klorul ∞i E.Me 6
In the first example, read thus,8 we find the word yiasi, which cannot
be anything other than the nominative form, referred directly to the
individual name “ayriq, of the word that appears in E.Me 9 as yjas[i≤ ],
and which we have just connected to the place name Iasow.
E.Me 33a is equally clear: here, the third position in nominative is
occupied by mdayn, which has also already been interpreted as an eth-
nic name (corresponding to genitive mwdon≤). In this same inscription,
the onomastic formula appears iterated, but for the second time (E.Me
33b) without the ethnic name (idmns myre≤ ).9
In the case of klorul, a similar explanation can be envisaged, although
it is necessary to admit that there are no parallels in the place names
of Greek sources for a place name *k(V)loru- or similar.10
A further example of a possible ethnic name in nominative is offered
by E.Me 44:
(a) apmen “rquq≤ kojol ∞i
(b) mwton≤ ∞i

8
On this reading, see p. 54.
9
For the problem of idmns (s-stem nominative or rather a stem with s-ending?), see
below pp. 314–317.
10
An alternative analysis would be to think of a type of title referring to triqo.
 ANALYZING CARIAN INSCRIPTIONS 271

Leaving aside the particle ∞i, the structure of E.Me 44a is identical to
E.Me 25, E.Me 6 and (with the caveat of -s in idmns) E.Me 33a. The
last form therefore also appears in nominative. A good connection can
be established here with the name of the island of Cos. The exact par-
allelism with the name of the inhabitants of this island in Greek sources,
K≈Ûoi, is astonishing. For -l, note also a possible similar suffix in kloru-l.
As for (b), the genitive mwton≤ must be related to the father’s name
“rqurq≤. This specification has been added outside of the basic formula,
but connected with it by means of the agreement.11
To sum up, there seems to be good evidence for interpreting a great
number of third words in three-fold formulae as ethnic names. However,
cases like t®∞atar≤ prevent us from extending this interpretation to all
the examples of three-fold formulae. As the examples from Hyllarima
show, a threefold structure consisting of name + patronym + pap-
ponym also existed in Carian, and some examples from Egypt could
correspond to this kind of structure.
A more complicated question is which part the ethnic name must
be attributed to: in the inscriptions with the structures N-Ø N-≤ N-Ø
(ethnic name), N-Ø N-≤ N-≤ (ethnic name), or N-s N-≤ N-≤ (ethnic
name), there is no ambiguity: in the first case, the ethnic name refers
to the deceased, in the second and third, to the father of the deceased.
But in N-≤ N-≤ N-≤ (ethnic), the structure is ambiguous, as the last gen-
itive could refer to the first one or to the second one. The existence
of both possibilities, as demonstrated by the unambiguous formulae
above, does not help to resolve the problem. It is possible that the
different uses of ∞i also mark differences in the structure, but it is gen-
erally difficult to tell which functions this particle is bearing when it is
used.

2. Stelae for Women

At least in two cases, the illustration in the stela makes it clear that
the deceased was a woman: in E.Me 12 and in E.Me 13:
pjabrm | u≤ol≤ | mwdon≤ ∞i kbjom≤ | m[no≤] E.(Me 12)
“dtat≤ | upa | w | wet≤ ∞i | mwdon≤ ∞i (E.Me 13)

11
Note, however, that this overall analysis of E.Me 44 raises questions that are still
unclear: the individual to which kojol is referred bears an Egyptian name (apmen), which
is rather strange (we must accept, then, that this individual, coming from Kos, adopted
a new, Egyptian, name). Moreover, this intepretation implies that while apmen came
from Kos, his father was a mwdon-.
272 CHAPTER SEVEN

E.Me 13 does not show any special characteristics. We have already

included it in the list of Genitive—Genitive—Genitive inscriptions above.
Like other inscriptions of the same kind, the role of mwdon≤ is ambigu-
ous, and could complement either the first or the second name: wet≤
could be the father’s name, but it is also possible that it represents the
husband’s name (see below). The w after upa is unusual: is it a mistake,
as Masson suggests, or does it represent a word, either complete or
abbreviated?
E.Me 12 offers a more interesting structure. It presents a four-fold
onomastic formula, pjabrm- u≤ol≤- mwdon≤- kbjom≤ and a last, incomplete
word, whose integration as m[no≤] seems suitable using the example of
another four-fold inscription, E.Me 16, which we will come to imme-
diately after. The meaning ‘son’ for mno≤ has been already proposed
on several occasions (see Chapter 11 s. v. for further details), and seems
to be the simplest and most logical solution, especially in the light of
certain funerary inscriptions of Caria (see below p. 289 and ss.). There
seems to be some connection between the fact that the deceased is a
woman and the presence of an atypical four-fold formula. A good solu-
tion would therefore be to interpret u≤ol≤ mwdon≤ kbjom≤ m[no≤ ] “of U≤ol,
the mwdon, the son of Kbjom” as the name of the husband.
A comparable analysis can be proposed for E.Me 16, which presents
a structure very similar to E.Me 12, as has been mentioned above:
irow | pikra≤ ∞i semw≤ | mno≤ mwdon≤ ∞i (E.Me 16)
irow is a name of Egyptian origin, where it appears documented as
both masculine and as feminine (see Chapter 11 s. v.). If we choose to
interpret it as feminine, the structure and analysis of the text is iden-
tical to E.Me 12, the sole difference being the position of mwdon≤ ∞i,
which in E.Me 12 precedes the filiation formula with mno≤. Therefore,
here we can translate it thus: “Irou (f ), (husband) of Pikra, the son of
Semw, the mwdon-.” It is impossible to decide whether the different
placing of mwdon≤ ∞i is due to the fact that here, unlike in E.Me 12,
mwdon≤ refers to the father’s name, semw≤, and not to the husband’s
name, pikra≤: the accumulation of genitives renders the overall struc-
ture ambiguous.
The following inscription also contains the co-occurrence of the name
irow and the word mno≤:
irow≤ : psHÿm[-]≤ pttu≤ : mno≤ (E.Me 27)
 ANALYZING CARIAN INSCRIPTIONS 273

In this case, it is a threefold formula, and a superficial approach would

suggest that it be grouped with the other Genitive-Genitive-Genitive
formulae. But the presence of mno≤ after the third name seems to chal-
lenge this simple interpretation. If: psHÿm[-]≤ pttu≤ was a sequence of
patronym and papponym of irow–, we would instead expect the word
for ‘son’ to follow psHÿm[-]≤ , not pttu≤. But if in this case irow≤ is a
female name followed by the onomastic formula of the husband, the
placing of mno≤ makes sense: ‘Of Irow (f.), (the husband) of PsHÿm[-],
son of Pttu’.

3. Inscriptions with ted and en

Hajnal and Schürr independently proposed that the Carian words for
‘father’ and ‘mother’ could be recognised in the forms ted (E.Me 38)
and en (E.Me 32) respectively. The correspondences with Lycian and
Luwian are striking (CLuw. tàti“, Lyc. tedi; CLuw. anni“, Lyc. ẽni, see
Chapter 11, ss. vv.) and the contexts in which they appear favour this
analysis as kinship terms:
“ÿin≤ | upe | arie?≤ ∞i ted (E.Me 38)
iturow≤ | kbjom≤ | ∞i en | mw[d]on≤ ∞i (E.Me 32)
The structures of both inscriptions are very similar, and the differences
easy to explain. In both cases, the name of the deceased appears in
genitive; in E.Me 38 it depends on the word for ‘stela’ (here upe),
whereas in E.Me 32, as in many other cases, the word for ‘stela’ is
elliptical. The construction of the kinship noun with ∞i is identical in
both texts: N-≤ ∞i ted, N-≤ ∞i en, and the interpretation, clearly paral-
lel: ‘who (is) the father of N’, ‘who (is) the mother of N’.
The construction with the particle ∞i presents very interesting char-
acteristics. First of all, the interpunction in E.Me 32 seems to be more
than a mere coincidence; unlike the most widespread type of con-
struction with ∞i, characterized by the postclitical position of this parti-
cle (it appears systematically attached to the preceding word), in this
case, ∞i ‘hangs’ on the following word en. This prosodic discrepancy
seems to correspond to a deeper, syntactic, difference. It must be noted
that in both examples, ted and en must be analysed as nominatives,
which implies that the construction with proclitic ∞i constitutes a true
relative clause, despite the absence of an express verb, with ∞i and
ted/en in nominative (as the syntax of the relative clause demands) vs.
the respective antecedents in genitive. This differs from the postclitic
274 CHAPTER SEVEN

uses of ∞i: in an inscription like the above-mentioned E.Me 16, repeated

here:
irow | pikra≤ ∞i semw≤ | mno≤ mwdon≤ ∞i
Here, mwdon≤ is an ethnic name referring to semw≤. If the construction
of postclitic ∞i were similar to that of proclitic ∞i, we would expect
*semw≤ mdayn ∞i, i.e. both ∞i and mwdon≤ in nominative. Instead, mwdon≤
agrees with the antecedent of ∞i, semw≤. As Hajnal (1997a) has con-
vincingly demonstrated, this construction is a further evolution of con-
structions of the proclitic type; from the use of elliptical constructions
N1-(≤) N2-≤ ∞i (*mno) ‘(Of ) N1, who (is the son) of N2, the position of
N2-≤ was extended to receive predicates and appositions syntactically
attracted by the antecedent.
The existence of two uses of ∞i—as proclitic, introducing what can
still be considered a relative clause, and as postclitic, with a function
closer to a simple article—allows us to interpret the structure of E.Me
17 with a certain precision:
“arnai≤ upe | quq≤ bem≤ ∞i mdaÿn (E.Me 17)
Regardless of whether “arnai- is in this case a female name—if so, quq-
could be interpreted as the husband’s name, as in E.Me 16, 32 and
38—or a masculine one—which would imply that we are dealing with
an individual name + patronym + papponym formula—the most rel-
evant fact is that bem-≤ (gen.) ∞i mdaÿn (nom.) is exactly parallel to
“ÿin-≤ . . . ∞i ted and iturow-≤ . . . ∞i = en, in contrast to the abundant exam-
ples of postclitic constructions (mwdon“ = ∞i (irow pikra≤ = ∞i semw≤ mno≤
mwdon“ = ∞i E.Me 16, etc.).
This analysis of ∞i mdayn/mdaÿn vs. mwdon“ = ∞i encounters some prob-
lems in the interpretation of two inscriptions containing mdayn = ∞i:
(a) ta“ubt≤ / kuari≤b/ar | ≤en / niqau≤ / ptnupi
(b) idmuon≤ / ∞i | mdayn ∞i (E.Me 18)
(a) idmns | myre≤ ∞i | mdayn ∞i
(b) idmns | myre≤ ∞i (E.Me 33)
The difficulties posed by E.Me 33 are not so serious: we can assume
that idmns is an s-stem nominative form, and that mdayn depends directly
on idmns. The case of E.Me 18 is more difficult, but in this inscription
it is the overall sense of the text that remains unclear. In any case,
idmuon≤ ∞i mdayn ∞i, added by a different hand, must refer to a name
in part (a), as shown by the presence of ∞i after the two words. The
 ANALYZING CARIAN INSCRIPTIONS 275

closest name is ptnupi, in nominative form, which allows us to interpret

mdayn ∞i as an ethnic name referring directly to ptnupi, in which case
the construction of the postclitic ∞i would be regular (mdayn is also a
nominative, like ptnupi ).
It must be noted that the hypothesis envisaged here about the exis-
tence of two different uses of ∞i—a postclitic one introducing not only
genitive but also attributive and appositional words, and with attrac-
tion to the case of the antecedent, and a proclitic one in a construc-
tion closer to the original structure of relative clauses—is based on
evidence that remains scarce, and must therefore be considered as
merely provisional.

4. Other More Complex Funerary Inscriptions

Leaving aside those already analysed (E.Me 12, E.Me 16 and E.Me
44), and the very difficult stela E.Me 18, mentioned above, there are
only three other inscriptions from Memphis that include complex for-
mulae with more than three names:
arjom≤ : ue : mwsat≤ : ∞i : mwdon≤ : ∞i tbridbd≤ : ∞i (E.Me 42)
(a) lÿ∞si≤ | upe | “rquq≤ ∞i | ksolb≤ (b) arliom≤ | mno≤ ∞i (E.Me 43)
[--]qarm≤ : q[---]≤ ∞i : pdubi≤ mno≤ [mw]don≤ ∞[i ---]w≤ord≤ ∞i (E.Me 10)
E.Me 42 can be compared to other typical inscriptions that present
the structure N-Ø N-≤ N-mwdon≤ (E.Me 13, 28, 31), the sole difference
being the addition of a fourth name. The simplest interpretation is to
view this as the papponym, but expressed formally as a “patronym of
the patronym”: ‘Of Arjom the stela, who (is the son) of Mwsat, the
mwdon-, the (son) of Tbridbd-.”
E.Me 43 is perhaps less complex than it appears, as one can in fact
speak of two different formulae. In E.Me 43a, we find a typical three-
fold formula N-≤, N-≤ ∞i N-≤, where the third name seems to be an
ethnic, as already seen: ‘Of Lÿ∞si- the stela, the (son) of ”rquq-
Kasolabean” (with the ambiguity of the exact referent of the ethnic
name: ”rquq or Lÿ∞si?). E.Me 43b, arliom≤ mno≤ ∞i, literally translated
as ‘of Arliom, who (is) the son’ makes complete sense when considered
as a further addition to E.Me 43a. Therefore, the stela would be ded-
icated to both Lÿ∞si and his son Arliom. An satisfactory overall trans-
lation would be, “Of lÿ∞si- the stela, the (son) of ”rquq- Kasolabean.
(And) of Arliom, the son”.
276 CHAPTER SEVEN

As for E.Me 10, the lacunary character of the text makes it difficult
to understand. The most remarkable fact is that this inscription appar-
ently contains a five-fold formula, which represents an exceptionally
long structure. This unusual characteristic, and the presence of mno≤
after the third name, leads us to the possible conclusion that we are
once again dealing with a stela for a woman; aside from the last name,
the structure is very similar to E.Me 16, the only difference being the
genitive vs. nominative case for the first name, although this is largely
irrelevant:
[--]qarm≤ q[---]≤ ∞i: pdubi≤ mno≤ [mw]don≤ ∞[i] ]w≤ord≤ ∞i (E.Me 10)
irow pikra≤ ∞i semw≤ mno≤ mwdon“ ∞i Ø (E.Me 16)
In this example, a woman, [--]qarm, would therefore be mentioned as
the wife of Q[---], the son of Pdubi≤, the mwdon-. As in the case of the
irow inscription, the function of mwdon- is ambiguous, insofar as it could
refer either to pdubis or to q[---] (see above for E.Me 16).
If this interpretation is correct, the last name [---]w≤ord≤ must be
either the name of the father of pdubi- (i.e., the papponym of q[---])
or, less likely, the ethnic name of pdubi- if [mw]don- refers to q[---].

5. The Rest of the Inscriptions from the Memphis Corpus

The formulae and structures analyzed in the preceding pages account
for most of the inscriptions from the Saqqâra corpus. Only a few inscrip-
tions have not yet been cited and/or analyzed. Several of these are so
fragmentary that they cannot be included in any of the structures
already mentioned (E.Me 22, E.Me 37, E.Me 39, E.Me 52–E.Me 66).
The remaining examples will be commented on briefly in the follow-
ing paragraphs.
terÿez≤ | upe | nuol∞[---]sarmrol∞yt (E.Me 4)
Although the beginning of the inscription points to a typical N-≤
upe . . . formula, no parallels can be found for the final part. Unfortunately,
the gap in the middle of the text complicates the interpretation still
further. The final -ol in ?]sarmrol would be consistent with an indi-
vidual name in -ol (= -vllow) or with a possible ethnic name, like
kojol. This interpretation would leave the last three letters as an inde-
pendent word ∞yt. Could this be a verb? The final -t recalls the pos-
sible third singular ending -t in ÿbt (see p. 281–282). This would mean
 ANALYZING CARIAN INSCRIPTIONS 277

that a typical formula ‘Stela of Terÿez . . .’ would be followed by the

name of the dedicant, ‘?]sarmol made/dedicated’. This is, however, a
very hypothetical solution.12
psm“kwneit≤ | ue | naria≤ | ≤ugliq | sarl? (E.Me 5)
The characteristics of E.Me 5 are to some extent comparable to those
of E.Me 4. The inscription begins in a strictly ordinary way (N-≤ ue
N-≤ ‘Stela of Psm“kwneit, (son) of Naria-’), but the following two words
move away from the usual formulae; ≤ugli- has already been already
mentioned, appearing in two other inscriptions as third name, and this
interpretation as a possible ethnic form is corroborated by its resem-
blance to the place name Souaggela. But the most surprising feature
of the present inscription is that we would expect to see a form in gen-
itive ≤ugli≤, according to the typical formula N-≤ N-≤ N-≤. With this in
mind, is ≤ugliq merely a mistaken form intended to be ≤ugli≤ (id est,
q q erroneously used for z ≤ ), or is the final -q linked in some way
to the presence of the final word sarl?, a hapax? The existence of the
particle -q, perhaps of connective nature, is now confirmed thanks to
the new inscription of Hyllarima (armotrqdos=q), but its function in this
latter text unfortunately remains unclear.
The difficult reading of the last letter (sarl? is a proposal made by
Schürr; an alternative reading is sara) makes an analysis of the inscrip-
tion even more complicated.
(a) war[---]t[------]i[---]≤ | mdaÿn
(b) [--15--]a[-]i≤ | mdaÿn (E.Me 11)
The only reason for citing such a fragmentary, largely illegible inscrip-
tion is the repeated presence of mdaÿn. As the illustration on p. 43
shows, this stela represents a male-female couple, and each line of the
inscription, situated behind each member of the couple, seems to cor-
respond to the individuals drawn on the stone. We can imagine that
the structure of both onomastic formulae was identical: N-Ø N-≤ . . . mdaÿn,
i.e. individual name in nominative + father’s name (and also grandfa-
ther’s name?) in genitive + ethnic name mdaÿn in nominative. These
are the only examples in which mdaÿn appears without ∞i.

12
Schürr (1992:155) tries to connect the peculiar content of this inscription to the
fact that the stela in which it appears is a ‘stèle de donation’, representing the Pharaoh
Apries making an offering to the god Ptah.
278 CHAPTER SEVEN

tdu≤ol / kbos | “amsqi[. . .? (E.Me 24)

The inscription, and consequently the formula, is almost complete: only
a few signs can be missing in the final part of the last word. The most
notable feature of this inscription is that the individual name in nom-
inative is followed by another name ending in -s. Interpreting kbo-s as
a dative of dedication, which would be a good argument for -s as a
true dative ending, seems unlikely, not only because of the resulting
exceptional structure (“Tdu≤ol to Kbo”), but also because of the par-
ticular disposition of the text in the stela; tdu≤ol, situated at the very
top of the stela, and inscribed in longer letters, must be the name of
the deceased. The simplest solution is to consider kbos an attribute or
apposition in nominative to tdu≤ol. It could certainly be a title accom-
panying the name of the deceased, but an interesting connection also
arises if we interpret it as an ethnic name: the coin legend kbo, which
Konuk has convincingly classified as a place name, identifying it with
Keramos (Konuk 2000b). It is true that kbo may only be the initial let-
ters of the place name, unknown in Greek sources, but the possibility
that kbo is a complete form cannot be discarded: kbo-s could therefore
be an ethnic name corresponding to that place name. The suffixation
can be compared directly to (accusative) -s-n in otono-s-n ‘Athenian’,
from *otono- ‘Athens’. As for samsqi[, it could be the patronym (sam-
sqi[?. . .-≤].
wnuti≤ | kwar≤ mHm≤ ∞i | mwdon≤ [∞]i (E.Me 31)
The analysis of this inscription is hampered by the problematic sequence
kwar≤mHm≤, where it is not clear if we are dealing with a single name
or rather with two different names, both in genitive. The latter solu-
tion has been adopted here (p. 59), but the resulting structure is some-
what strange: it would apparently be a four-fold structure, where mwdon≤
appears as a fourth name, and not as third name, which it is more
typical.
Vittmann (2001:48–49) has offered a plausible solution to this prob-
lem and at the same time a very convincing interpretation of the first
word, wnuti≤. He suggests that wnuti- could be the adaptation of Egyptian
wnwtj, ‘hour-observer, astronomer’ (see Chapter 11 s. v. wnuti≤ ), so that
wnuti- would be a title referring to kwar-≤, the name of the deceased.
Thus the structure of the onomastic formula becomes an ordinary three-
fold one: ‘Of the astronomer Kwar-, who (is the son) of MHm-, who
(is the) mwdon-’ (once again with the ambiguity about the exact refer-
ent for mwdon-).
 ANALYZING CARIAN INSCRIPTIONS 279

[q?]lalis / [?]iam≤ ∞i / alos ∞arnos (E.Me 45)

In the two initial words we can easily identify a formula, Nominative
or “s-ending” + Genitive. As in other cases, it is difficult to decide
whether the name is an s-stem in nominative or an s-ending, in which
case the formula would have to be compared to the ntokri-s type (E.Me
35).
The pair of words alos ∞arnos is far more difficult to analyse. The
resemblance to the well-known place name Halikarnassos, first sug-
gested in Adiego (1990), continues to be an attractive—if problematic—
possibility. The word reappears in another inscription (C.xx 2, under
the form alosd ∞arnosd ), and I refer to the analysis of that inscription
(below p. 284). In this case, we cannot rule out the possibility of inter-
preting it as an ethnic name or even a place name.
tqtes | paraibrel≤ ∞i | mn[o-?] (E.Me 47)
The initial name could again be a nominative of an s-stem or an
s-ending form, tqte-s. The following sequence, paraibrel≤ ∞i mn[o-?], where
we find N-≤ ∞i mn[o-?] ‘son’, recalls the structures arie?≤ ∞i ted (E.Me 38),
kbjom≤ ∞i en (E.Me 32) mentioned above, and leads to an interpretation
‘tqte(s). who (is) the son of Paraibrel’. In this case, it is very likely that
the word ‘son’ appeared in nominative (*mno), a form that is still undoc-
umented.
loubaw | siral | pnld≤wl (E.Me 49)
This inscription is perhaps the most unusual example found in the
Carian corpus of Saqqara. Apparently a threefold formula, the second
and third names end in -l, which would be a morphological ending
without clear parallels. Moreover, no parallel can be found in the Carian
onomastics for any of the three alleged names. In any case, we must
bear in mind that the overall reading of the text remains very uncer-
tain (see p. 71 for remarks).

6. A First Summary
The previous pages have allowed us to familiarize ourselves with the
onomastic formulae, from the most basic to the more complex. Our
main sources of information were the funerary stelae of Saqqâra, where
it is common to find more complex structures than simply the indi-
vidual name, or of the individual name + father’s name. We have seen
280 CHAPTER SEVEN

the high frequency of threefold formulae, which contain a third name,

possibly an ethnic name, but also in some cases the name of the grand-
father. Even more complex structures can also be found, and in some
remarkable cases the increased complexity seems linked to the fact that
the deceased is a woman; in such cases, the onomastic formula that
accompanies the female individual name seems to refer to the husband
and his genealogy. In two exceptional cases, the deceased is mentioned
as the father or mother of another person (E.Me 38, E.Me 32).
Also typical in the Memphis funerary stelae is the absence of verb
forms. Only in the case of E.Me 5 could this possibility be envisaged,
but the fragmentary nature of the text prevents reaching a definitive
conclusion. The rest of the Saqqâra epitaphs are characterised by the
direct reference to the deceased in nominative or in an s-ending form,
or to the stele in an expressed or elliptical form, in which case the
name of the deceased appears in genitive, indicating possession. None
of these cases require the use of a verb form. Indeed, not even in the
constructions with ∞i is a verb used.
The absence of verb forms is counterbalanced by the presence of
certain nouns that belong to the Carian common lexicon: mno- ‘son’,
ted ‘father’, en ‘mother’, upe (and variants), ue ‘stela’, to which two
pronominal forms can be added: the abovementioned relative ∞i and
the demonstrative sa (in upe sa). As for the use of ∞i, we have identified
two different constructions: the first, closer to a use as relative, where
the nominal predicate introduced by ∞i is in the nominative while the
antecedent is in the genitive (type arie?≤ ∞i ted ), whilst the second is more
comparable to an article or connecting particle, where the complement
introduced by ∞i is attracted by the antecedent (type semw≤ . . . mwdon“
∞i ). Although the evidence is not sufficient to draw definitive conclu-
sions, we have been able to observe a clear correspondence between
the position of ∞i and each of these two functions: it seems to be pro-
clitic when used as ‘relative’, and postclitic when used as ‘article’.

C. ANALYZING BRIEF INSCRIPTIONS

In this section, I shall analyze a selection of brief Carian inscriptions

other that those mentioned in the preceding sections. I leave out those
graffiti containing more than simply onomastic formulae, given the
difficulties of reading present in most cases, which have not been ade-
quately edited (Thebes, Silsilis, Abydos). Speculating about the struc-
 ANALYZING CARIAN INSCRIPTIONS 281

ture and meaning of inscriptions for which we do not possess a reli-

able reading would be a very risky practice. In my view it is prefer-
able to limit the research to those inscriptions we can be confident of
reading correctly.

1. Inscriptions on Objects
Two ‘Pharaonic Objects’ (E.Me 8, E.xx 6) and the Use of sb ‘and’
We begin with two very brief inscriptions on so-called ‘pharaonic objects’,
which serve to illustrate the use of the coordinative conjunction sb
E.Me 8, a bilingual inscription on the base of a statuette of Apis,
consists of two parts. The text that appears in the first part, paraeym:
armon ∞i, is now interpreted without difficulties as ‘Paraeym the inter-
preter’, in direct correspondence to an identical formula that appears
elsewhere in Egyptian (see p. 41 and Chapter 11 s. v. armon). In this
case ∞i introduces an apposition to the personal name in nominative.
In the second part, two personal names in nominative are united by
sb, which has been unanimously interpreted as a coordinative con-
junction (‘and’), above all since the discovery of the Kaunos bilingual,
where it appears repeatedly with this function. E.Me 8b paraeym sb polo
can therefore be interpreted as ‘Paraeym and Polo’.
A construction similar to E.Me 8b, but in ‘s-ending’ case, is visible
in the inscription E.xx 6 on the basis of a statuette of Isis: “arnajs sb
taqbos ‘For/of ”arnaj and Taqbo’. It is probable, but impossible to
demonstrate, that these formulae of EMe 8b and E.xx 6, consisting of
a pair of names, represent the names of a husband and wife.
Three Inscriptions on bowls (C.xx 1, C.Ha 1, C.xx 2)
Three inscriptions on bowls constitute a type of small sub-corpus of
particular interest. In a very influential and decisive article, Melchert
(1993) offered an interpretation of one of these texts (C.xx 1), which
in my opinion remains essentially valid. We will begin thus with this
inscription, adopting in general terms the views expressed by Melchert:
“rquq | qtblem≤ | ÿbt | snn | orkn | ntro | pjdl? C.xx 1
In this inscription, an onomastic formula had already been identified:
“rquq qtblem≤ ‘”rquq (son) of Qtblem’. Melchert’s interpretation of the
remainder of the text is based on two fundamental and compelling
hypotheses: firstly, that ÿbt is a verb comparable both formally and
282 CHAPTER SEVEN

semantically to Lycian ubete, ‘offered’, and secondly that snn orkn is the
direct object of this verb, formed by a demonstrative snn (which would
belong to the same paradigm as san in the Athens bilingual inscription)
and a common noun referring to the phiale, orkn. Both elements appear
in the accusative singular, morphologically reflected by the ending -n
(sn-n ork-n). Melchert’s elegant interpretation of the five first words of
C.xx 1 is now supported by the discovery of the bilingual inscription
from Kaunos, which has confirmed the existence in Carian of an
accusative sg. ending -n and has dispelled the doubts about the value
of y/ÿ (close to u), thus assuring the equivalence ÿbt = Lyc. ubete.
The last two words remain problematic. Melchert has recently revised
his initial views on these forms. In his 1993 article he intepreted ntro
as a dative of the Carian name for Apollo. For pjdl, he proposed the
analysis as an apposition to snn orkn with the meaning ‘gift’, etymo-
logically related to Hitt. pài-/piya-, Luw. piya-, Lyc. pije- ‘to give’: -dl
would represent a suffix*-dhlo-). However, Melchert (2002:309–310)
denies the existence of Carian datives ending in a vowel and, follow-
ing a suggestion by Schürr, prefers to interpret of ntro as referring to
“rquq, with the meaning ‘priest of Apollo’: ‘”rquq (son) of Qtblem ded-
icated this bowl—the priest of Apollo as a gift’. Leaving aside for now
the discussion about the alleged “datives” in -s, the internal syntactic
reasons that Melchert adduces in refusing the interpretation of ntro as
dative are not particularly convincing: he states that pjdl cannot be an
apposition to the direct object if ntro is a dative, due to the separation
that this latter word introduces between snn orkn and pjdl. For this rea-
son, he classifies both ntro and pjdl as two ‘epexegetic, add on phrase[s]’
defining, respectively, the subject and the direct object, as his transla-
tion offered above tries to reflect. However, following this logic, I see
no reason why we should not consider only the word pjdl as an ‘epex-
egetic phrase’ (= ‘”rquq (son) of Qtblem dedicated this bowl to Apollo—
as a gift’). A further complication arises if we take ntro to be not the
Carian name for Apollo, but rather a derivative of it: it obliges us to
recognize the same derivative in E.xx 7, where an s-ending form appears
(ntro-s), which Melchert interprets as a dative. We must also assume
therefore that this latter inscription is dedicated not directly to Apollo
Branchid but to an unmentioned priest of Apollo Branchid, perhaps a
less satisfactory solution (see below p. 317).
In any case, despite these difficulties that depend to a great extent
on the crux about datives in Carian—on which see pp. 314–317 for
 ANALYZING CARIAN INSCRIPTIONS 283

further remarks—Melchert’s interpretation of the overall meaning and

the first part of the syntactic structure of C.xx 1 remains one of the
most brilliant contributions to the understanding of a Carian text, and
the consequences of the etymological connections it highlights are very
important.
The connection of C.xx 1 with another inscription on a bowl (C.Ha 1)
has not gone unnoticed:
smdÿbrs | psnlo | md orkn tÿn | snn (C.Ha 1)
Melchert (1993:81) correctly notes the presence of snn and orkn, the
same nominal phrase that appears as a direct object in C.xx 1. Given
that both inscriptions are engraved on similar objects, this parallelism
is consistent with the interpretation of snn orkn as ‘this bowl (acc.)’. But
the correspondence between the two inscriptions ends here, and it would
be very difficult to interpret the structure of C.Ha 1 using the exam-
ple of C.xx 1. The only clear form (apart from snn orkn) is smdÿbrs,
which is undoubtedly a personal name, although this is ambiguous due
to the final -s, which could be either an inflection mark (‘s-ending’) or
simply the final consonant of an s-stem. Forms such as k≤tÿbr, ardybyr-≤,
dtÿbr, Arduberow could point to the first solution, but the new personal
name ybrs-≤ from Hyllarima (C.Hy 1a), a true s-stem that seems to con-
tain the same root ybr-/ÿbr-, allows us to classify smdÿbrs as a pure nom-
inative form.
The presence of orkn . . . snn leads to the assumption that a transitive
verb must be also present in the inscription. Without entirely discard-
ing the possibility of omitting the verb, suggested by Melchert (1993:81),
it is advisable from a methodological point of view to pay careful con-
sideration to whether any of the three remaining words ( psnlo, md, tÿn)
could represent the verb of the inscription. If we look at each of the
three possibilities in turn, we will see that the choice here is not easy.
It seems almost certain that tÿn is a word in agreement with orkn snn,
given that it is situated between these two words; a different explana-
tion would imply that the noun of the recipient and the demonstrative
that accompanies it form a discontinuous noun phrase, a hyperbaton
that is not altogether impossible, but certainly quite unlikely.13 In the
case of md, we must not overlook its resemblance to the forms mda,

13
However, the interpunction that separates md orkn tÿn from snn suggests that this
possibility should not be ruled out altogether.
284 CHAPTER SEVEN

mdane, which have been interpreted as verbs in various studies (see for
instance Melchert 2002:308, n. 7). However, in my view, the forms mda
and mdane forms cannot be verbs, as we will see later (p. 324). This
leaves psnlo as the form most likely to be a verb. As evidence to sup-
port this hypothesis we could consider the resemblance of psnlo | md
to pisñoimda in the Hekatomnids’ decree from Sinuri (C.Si 2). In Adiego
(2000:141–143), this latter form was connected with the Anatolian verb
for ‘to give’ (Hitt. pài-/piya-, Luw. piya-, Lyc. pije-, etc.), see p. 304.
An alternative analysis of psnlo would be to take it as a personal
name; this possibility would lead us to accept Melchert’s hypothesis on
the absence of an express verb, or to attribute this function to md or
tÿn, the difficulties of which have already been mentioned. If pnslo is
a personal name, the resulting sequence smdÿbrs psnlo would create an
embarrassing dilemma: is smdÿbrs the indirect object and psnlo the sub-
ject, or vice versa? The dilemma is inseparable from the problem of
the Carian dative (see pp. 314–317).
Shown below is the third inscription on a bowl:
ÿ≤biks not : alosd ∞arnosd : jzpe mdane (C.xx 2)
Adiego (2000:153–155) offered an interpretation of this inscription. This
interpretation hangs on two fundamental hypotheses: that ÿ≤biksnot must
be segmented ÿ≤biks not and interpreted as a sequence Individual Name14
+ Verb (where not would be morphologically comparable to ÿbt in C.xx
1, see above), and that the well-known word mdane is not a verb, but
rather a sequence of particles and clitic pronouns. For no-t, a connec-
tion to Hitt. nà(i)-, CLuw. (reduplicated stem) nana- < * PIE *neyH–
‘to bring’ was cautiously put forward. Regarding mdane, an analysis
already proposed in former works was reintroduced: in -ne, a clitic
accusative pronoun -n- + a clitic dative pronoun -e attached to a base
md-, for which no explanation was given. The form jzpe was consid-
ered a proper name dative (‘to/for jzpe’). Finally, alos-d ∞arnos-d was
interpreted as an ablative instrumental (with -d = Luw. -ti, Lyc. -di ) of
the word (in fact a noun phrase) alos ∞arnos, also present in E.Me 45
and tentatively identified (in Adiego 1990a) as the Carian place name
Halikarnassos. This yielded a possible translation as ‘Ÿ≤biks brought it
to Jzpe from Halikarnassos’. Obviously, this proposal was based only

14
For ÿsbiks as PN, cf. yi≤{∞}bik≤ (E.Me 46a).
 ANALYZING CARIAN INSCRIPTIONS 285

on a set of hypotheses, and as such is far from conclusive. But the core
of the discussion must be retained here: not and mdane are the most
obvious choices for constituting the verb of the sentence, and either
choice has important repercussions for the interpretation of the other
inscriptions containing mda, mdane.
The inscription on a bronze lion (E.xx 7)
Another very interesting short inscription is found on the base of a
bronze statue representing a lion (E.xx 7):
ntros : prãidas / or“a / nu mdane : uksi wrm≤
ntro-s prãida-s shows an inflected form of ntro, the same word that appears
in C.xx 1, where it was interpreted as a god name (the Carian Apollo;
cf. however the alternative explanation as a derivative, ‘priest of Apollo’).
ntro-s is accompanied by prãida-s, whose final s seems to indicate an
agreement with ntro-s. Schürr proposed that this could be identified as
a Carian form connected to Greek Bragx¤dai, ‘the Branchids’—the
name of the family of priests consecrated to the cult of Apollo in the
sanctuary of Didyma near Milet (Schürr 1998:158)—but as attractive
as this theory is, it is dependent on the rather ad hoc assumption of
a phonological value close to ºgxº for the infrequent and obscure Carian
letter %. Lacking further evidence, I have chosen to adopt this assump-
tion here in purely conventional fashion (which explains the use of <ã>
for transliterating %, see p. 20). In any case, there is a general con-
sensus that prãidas constitutes an attribute or an apposition to ntros, with
which it would agree.
Melchert’s interpretation of the whole inscription (see Melchert
2002:308) is as follows: “Uksi, (son) of Ur(o)m, has now given it, the
or“a, to the priest of Apollo, the Branchid”. There are no particular
problems posed by analyzing uksi wrm≤ as a typical onomastic formula,
individual name + father’s name in genitive, and uksi as a nominative
and consequently, the name of the donor. The rest of Melchert’s inter-
pretation is more controversial: he claims that mdane is the verb, nu is
an adverb,15 or“a is the name of the object and finally ntros prãidas must
then be the indirect object in dative case. We have already seen (above

15
No comments are made about this word in Melchert (2002), but it is self-evident
that he assumes an etymological connection with PIE *nu- ‘now’ (Hitt. nu-, Lat. nunc,
Gr. nËn, etc.)
286 CHAPTER SEVEN

p. 282) that in this paper, Melchert abandons his earlier interpretation

of ntro/ntros as a god name in favour of the meaning ‘priest of Apollo’.
This change is a consequence both of his refusal to analyze ntro as
dative and of his assumption that the s-ending is a dative mark. The
result of this new analysis is in my opinion rather unsatisfactory: E.xx.7
becomes an inscription dedicated to an anonymous priest of Apollo,
whereas in C.xx 1 the verb ÿbt, ‘offered’, is not accompanied by the
name of the being to whom the object is offered. In my opinion it
would be simpler to interpret ntro/ntros as the god name (‘Apollo’), ntro
as dative and to search for an alternative explanation for the s-ending
in ntros pr%idas, but I leave the discussion of the forms with s-ending
until the following chapter.
The interpretation of mdane as a verb is also unconfirmed, and this
word will be analysed in the following chapter. In any case, if it is not
classified as a verb, the verbal form ought to be looked for in the pre-
ceding sequence of words: either nu, or“a, or less probably or“anu (with-
out discarding further segmentations). Of all these possibilities, nu seems
the most likely: final -u immediately preceding mdane reminds us of ew
mlane (/ ew lane 1×), a sequence that appears in some inscriptions of
Thebes, where it could also be a verb.
The ‘Tarhunt-inscription’ of Iasos (C.Ia 3)
The inscription of Iasos C.Ia 3 offers some very interesting forms but
unfortunately the overall structure and meaning remain unclear:
?] are“ | “anne mlne | siyklo≤ | “ann | trqude | ∞lmud [?
There is no question that the most significant word is trqude, interpreted
as a god name (Tarhunt, the Anatolian Storm God) in Blümel-Adiego
(1993), which has now been confirmed by the form armotrqdosq in
Hyllarima (C.Hy 1), where a dvandva Arma-Tarhunt (the moon-god
plus the storm-god) is easy to recognize.
Given that the inscription appears on a cratera, we can assume that
this object is offered to the god Tarhunt, which would imply that trqude
should be considered an indirect object in dative. However, the rest of
the text does not allow us to confirm this syntactic analysis; the only
other form that can be analyzed with any confidence is siyklo≤, a gen-
itive, probably of a personal name, which implies that it depends on
another personal name, as part of an onomastic formula. The most suit-
able candidate is ?]are“, the word that apparently begins the inscrip-
tion, but the discontinuity of the onomastic formula would then be
 ANALYZING CARIAN INSCRIPTIONS 287

somewhat surprising. Also unclear are the formal connection between

“anne and “ann and the resemblance of mlne to the sequence ºmln in
C.Ka 4 yomln C.Ka 5 uiomln, two forms which should probably be
interpreted as verbs (see below pp. 298–299, 301). Finally, the analy-
sis of ∞lmud[? is hampered by the uncertainty about the final part of
the word. Gusmani’s proposed complete form, ∞lmud[e], would give a
possible epithet for trqude.
The inscription of ”arkbiom (E.Sa 1)
The text of E.Sa 1, inscribed on the base of a reliquary for three
mummified reptiles, is the following:
“arkbiom : zidks mdane : ÿn[-?]/mo | den : tumn
The text is even more impenetrable than those of the inscriptions on
recipients analyzed above. It clearly begins with a personal name in
nominative (“arkbiom, also cited in the Egyptian part of this bilingual
text). This name seems to appear here as a complete onomastic for-
mula, with no mention of the father’s name. The inscription again con-
tains the verb or particle sequence mdane, a word ending in -o whose
complete reading is not possible ( ÿn[-?]mo), and a very obscure form
with no parallels, zidks. For the two last words, a tentative analysis was
made in Adiego (1995:21–23): given that the Egyptian text contains a
formula, ‘that Atum the great god may give life and health to ”úrkbym’,
I suggested that the Egyptian god name Atum could be reflected in
the form tumn at the end of the Carian text. This latter form would
therefore be an accusative in -n of a stem tum-. The somewhat unusual
presence of an accusative instead of a more typical dedication dative
was explained by analysing den as a type of preposition, etymologically
related to Hitt. anda, andan, Lyc. ñte.
Irrespective of this latter interpretation, the overall structure and
meaning of the inscription remain unclear. If we retain the hypothesis
that mdane is a particle sequence, the search for verbs can be limited
to the words zidks and ÿn[-?]mo, for which no parallels can be given.

2. Funerary Inscriptions of Caria and Athens

A brief sub-corpus of funerary inscriptions other than those from Egypt
can be identified. It consists of eight inscriptions, seven for different
Carian places (C.Tr 1, C.Tr 2, C.Al 1, C.Eu 1, C.Ka 1, C.Ka 3, C.Kr 1)
and the bilingual inscription of Athens (G.1). The funerary character of
288 CHAPTER SEVEN

other inscriptions (C.Eu 2, C.Si 1) cannot be ruled out, particularly in

the first case, but they do not display the requisite lexical elements that
would allow us to include them in this typological classification.
These lexical elements are the words s(i)di, present in C.Tr, CTr 2,
C.Al 1, C.Ka 1, C.Te 1, and ≤( j)as in C.Eu 1, G.1. Although neither
of these appears in C.Ka 3, the context of the inscription, engraved
on the facade of a typical Kaunian rock tomb, is undoubtedly funerary.
The only bilingual text is found in the inscription from Athens. In
Adiego (1992a:32–33; 1993a:165–170) an interpretation was attempted,
based on the hypothesis that the single Carian line corresponds exactly
to the first line of the Greek text:
s∞ma tÒde Tur[
≤jas : san tur[
According to this inscription, Carian ≤jas would be equivalent to Greek
s∞ma, whereas san would be a demonstrative pronoun corresponding
to Greek tÒde. This latter equivalence indicates a very interesting ety-
mological connection: san could be related to Luw. za-, Hitt. ka- ‘this’
< PIE *∞o-, and would also be evidence of the ‘satem’ treatment of
PIE palatals in Carian, as in the rest of Luwic languages (cf. Adiego
1995:12). Melchert (1993:79–80) refined this comparison, explaining the
Carian final n of san by comparing with the Hitt. adverb kàn(i) ‘look
here, lo!’, eni ‘this, the aforementioned’, and combining san and snn
from the bowl inscriptions in a single paradigm.
The interpretation of ≤jas as a word for ‘tomb’, ‘funerary monument’
or similar is reinforced by its presence in C.Eu 1, where it appears—
in the slightly different form ≤as—accompanying a standard twofold
onomastic formula:
“as : ktais idyri∞≤ : mn[o-?]
Both ktais and idyri∞≤ admit good onomastic identifications: ktais recalls
ÑEkata›ow, a Greek name commonly attested in Caria, perhaps because
of its resemblance to the indigenous name ktmno (adapted in Greek as

16
In former works, I suggested that ktais, analyzed as nominative, could contain an
s-stem created from the Greek nominative, on which the Carian inflection of the loan-
word could have been formed (cf. Lyc. zeus- in dat. zeus-i, from Greek ZeÊw). But for
the reasons explained above, I believe that the use of nominative can be excluded in
this case.
 ANALYZING CARIAN INSCRIPTIONS 289

ÑEkatÒmnvw). As for idyri∞≤, it belongs to the same family of names as

Saurigow, Senurigow (on this family, see pp. 262–263). Once again, the
final -s in ktais can seem ambiguous (s-stem16 or case ending), but in
this case the simplest solution is to interpret it as a inflectional mark,
given the presence of a word referring to the monument, which rules
out the possibility of expressing the name of the deceased directly in
nominative. ktai-s must therefore be an ‘s-ending’ form, and, as in many
other cases, it can be translated as true dative = indirect object (‘tomb
for Ktai’) as well as possessive or genitive (‘tomb of Ktai’). idyri∞≤ does
not pose any problems: it is the name of the father of ktai, expressed
in the usual genitive in -≤. As for mn[o-], it is regrettable that the final
part of the inscription is broken. An integration mn[os] seems the most
likely solution, because this word must agree with ktais. The complete
text can thus be translated: ‘Funerary monument for/of Ktai, son of
Idyri∞’, without ruling out, however, the possibility that mn[os] could
actually be a juxtaposed word to ktais, which would give the following
meaning: ‘Funerary monument for/of Ktai (son) of Idyri∞ (and his) son.
Interpreting the inscriptions with s(i)di is more problematic, since
their structure is more ambiguous and the segmentation of the words
in the contexts where s(i)di appears is not always certain.
We can begin with C.Tr 2, a complete and brief inscription with
virtually no problems of segmentation:17
an sidi a-/rtmi pau≤ / parãaq?
However, the interpretation of this text is seriously impeded by the
difficulties of reading posed by the last sign of the last word. As noted
in p. 131, all the existing drawings of this lost inscription seem to point
to Q, but the well-known problem of distinguishing between z, Q and
also o make it unwise to automatically accept a reading as Q. Therefore,
we must also consider the possibility of an alternative reading with z
( parãa≤?),18 a theory that has been vehemently defended by Schürr.
The main advantage of the reading parãa≤ is that it results in a clear
threefold onomastic formula, N-Ø N-≤ N-≤: rtmi pau-≤ parãa-. Although

17
Only the segmentation sidi artmi could be contested, because one could theoreti-
cally separate sidia rtmi (for this latter, cf. rtim C.Hy 1a), but this alternative segmen-
tation does not affect most of the possible interpretations that will be considered here.
18
The third possibility—reading o, hence parãao?—seems to be excluded, given that
a final sequence ao in this position has no parallels and does not lead to any satis-
factory analysis.
290 CHAPTER SEVEN

there are no clear parallels for a name parãa-, the typical onomastic
element par(a)- can be easily recognized. Even a connection to prãidas
could be considered: if prãida-s corresponds to the name of the priest
family of Didyma Branchidai, parãa- could be the Carian name corre-
sponding to Greek Brãgxow. However, the problem is to syntactically
connect the onomastic formula rtmi pau-≤ parãa-≤ with the initial for-
mula an sidi. This latter seems to consist of a noun that would denote
the funerary monument, sidi, preceded by what is probably a demon-
strative pronoun an (cf. san in G.1): therefore, an sidi would mean ‘this
tomb’, or ‘this (is) the tomb’, or similar. The presence of this formula
seems little consistent with an N-Ø (nominative) rendering of the name
of the deceased. This sentence could only make sense if we interpret
an sidi as an accusative, and assuming an elliptical verb (‘made’).
An attractive alternative is to suppose that artmi is not an accusative,
but rather a dative, which would mean that the overall sentence must
be interpreted as ‘This tomb/this (is) the tomb for A., (son) of P., (grand-
son) of P.’ Developing this hypothesis further, one might wonder if -i
could be a true dative ending for a stem artm-. This possibility would
allow us to integrate artm-i and the name (in nominative) rtim from
C.Hy 1a in a single paradigm (nominative (a)rtim-Ø / dative art(V)m-i ).
However, the existence in Minor Asian onomastics of a large number
of names formed on the basis of the onomastic element art(e/i)m-, but
with different derivations, makes this paradigmatic connection of artmi
and rtim a very fragile theory.19 Consequently, it is currently impossi-
ble to decide whether artmi is a nominative or a dative, since there are
significant difficulties encountered by either solution.
The reading parãaq would give a new perspective. If parãa- is a per-
sonal name stem, -q could recall ≤ugli-q in E.Me 5, but this is an obscu-
rum per obscurius solution, given that the interpretation of ≤ugli-q in the
context of E.Me 5 is also very problematic (see above). A very different
way of analyzing parãaq was suggested in Adiego (1993a:263) and devel-
oped further by Hajnal (1995[97]:20). Taking as a starting point my
proposal of connecting parãaq with the Lycian verb prñnawa- ‘to build’,
very common in Lycian funerary inscriptions (ebẽñne xupa prñnawatẽ
X, . . . ‘X has built this tomb . . .’ and variants), Hajnal tries to connect
final -q with the 1st singular active preterite ending in Luwic languages:

19
See Zgusta (KPN § 108) and here Chapter 11 s. v. artmi.
 ANALYZING CARIAN INSCRIPTIONS 291

CLuw. -¢a, Lyc. -xa, -ga. The advantage of this hypothesis is that it
convincingly resolves the structural problems posed by the text; an sidi
would be therefore the accusative of a verb meaning ‘to make’ (or,
with a more specialized meaning, ‘to build’), and the problem of an
onomastic formula with an individual name in nominative would no
longer be relevant, as the text would read, ‘This tomb (acc.) Artmi,
(son) of Pau, I made’. This is undoubtedly a very attractive hypothe-
sis, but it is seriously weakened by the doubts surrounding the exact
phonological value of % <ã>.
sñis : sdisa-/s : psu≤ol≤ / mal≤ : mno≤ C.Ka 1
The main difficulty of analysing of C.Ka 1 lies in how to interpret the
initial formula sñis sdisas, where a word sdi, undoubtedly a variant form
of sidi with defective vowel notation, can be identified. In psu≤ol-≤ mal≤
mno≤, it is not difficult to recognize a twofold formula, with both per-
sonal names in genitive ( psu≤ol-≤, mal≤ ). The well-known Carian word
for ‘son’, mno-≤, also in genitive, could agree with the name of the
deceased and govern the other genitive, mal-≤ (‘of Psu≤ol, the son of
Mal’), but, as in C.Eu 1, an asyndetic construction must not be ruled
out: ‘of Psu≤ol (son) of Mal, (and) of (his) son.’
As for the initial words, sñis sdisas, I believe it very unlikely that sñis
can be a personal name, although not impossible. The hypothesis that
will be considered here argues instead that it is part of the reference
to the tomb. The main problem is how to reconcile sdisas with the
form an sidi in C.Eu 1. One could attempt the segmentation sdi sas
and recognize here the different pronominal stem sa-/san-/snn-, but the
final s would then remain unexplained. Without dismissing this possi-
bility, I wish to propose an alternative: if we intend to find here the
same pronominal stem of an, the only possible procedure is to use the
segmentation sdis as. This gives a sequence sñi-s sdi-s a-s, which seems
to contain a common ending -s for all three words. Could this ending
be a plural mark? If this were the case, sñis could be interpreted as a
plural form of the pronominal stem san-/sn- (sñ-is), and the overall sense
of sñis sdisas could be ‘These (are) the burials . . .’ or ‘These burials are
those . . .’, completed by the genitives that follow (‘. . . of Psu≤ol (son)
of Mal (and) the son’. If I must be honest, this latter translation is
somewhat forced, as it assumes an asyndetic construction, which is a
possible, but not the only, interpretation of the whole sequence (see
above). However, the link between the presence of sdisas and a collective
292 CHAPTER SEVEN

tomb is much clearer in the other inscription in which this form appears
(C.Kr 1):
qot2omu sdisa-
s?n≤ “odubr≤ or rather: mn≤ “odubr≤?
sb mno≤ knor
noril?ams or rather: norimams?
In this inscription, the existence of more than one burial can be deduced
not only from the clearest part of the inscription, “odubr-≤ sb mno-≤,
which can be translated ‘of ”odubr and the son’, but also from the
fact that the tomb clearly contained three burial chambers, as pointed
out by Olivier Henry (pers. comm.). Unfortunately, the overall struc-
ture of the inscription remains very unclear. It is possible that qot2omu
could be a personal name, which would allow us to recognize the ref-
erence to three individuals (qot2omu, “odubr-≤ and one son, mno-≤), but
the syntactical connection of qot2omu with the rest of the inscription,
and the sequence n≤ that appears after isolating sdis-as, remain unex-
plained. Note also that the reading of the final s in sdisas and the sub-
sequent segmentation are far from conclusive. If we accept (with some
reservations) the reading sdisas, we could tentatively envisage that qot2omu
is not a nominative functioning as subject, but rather a nominative of
appellation, designating the owner of the tomb. In this case, n≤ could
be a resumptive pronoun referring to Qot2omu: ‘Qot2omu. These tombs
(are) of him, of ”odubr, and of the son . . .’ Needless to say, this inter-
pretation is more a desideratum than a fact based on solid evidence.
As for knor norilams (or knor norimams), practically nothing can be said.20
The two remaining inscriptions that contain sdi- are too fragmen-
tary to attempt an interpretation. In C.Tr 1, two personal names, one
in nominative (artmon) and the other in genitive ( pau≤ ) can be identified,
but sidi appears followed by an unclear and incomplete sequence, amt[.
The case of E.Al is worse still, where only sdi a[-]mob[ is legible.
The last funerary inscription of this sub-corpus, although complete
and without reading problems, thanks to the recent autopsy made by
Diether Schürr, is very brief and contains no specific word to desig-
nate the tomb:
“oru≤ / ann ibrs≤ C.Ka 3

20
Only as a purely assonant connection can we resort to Lyc. nere/i-, a term of
relationship whose exact meaning is unknown (Therefore sb mno≤ k-nor noril?ams ‘and of
the son and(?) the nor nori- relatives’?).
 ANALYZING CARIAN INSCRIPTIONS 293

The segmentation proposed here (already suggested in Adiego 1996)

seems to me the most plausible, although other possibilities exist.21
According to this segmentation, we can identify two names in genitive
(“oru≤, ibrs≤) and the word ann, probably a demonstrative pronoun that
can be related to an (see above). The meaning of the inscription would
therefore be ‘This (tomb is) of ”oru, (son) of Ibrs’ or, simply, ‘That (is)
of ”oru, (son) of Ibrs’.
In conclusion, the study of this sub-corpus of funerary inscriptions is
seriously limited by the scarcity of materials, and the divergent for-
mulae employed—despite the coincidence of several formulaic words
(s(i)di, ≤(j)as, and some demonstrative pronouns—cannot be reduced to
a single, stereotyped model of expression.

3. The Longest Graffito from Abu Simbel (E.AS 7)

While a great number of the Carian graffiti from Egypt contain only
onomastic formulae (usually N-Ø or N-Ø N-≤), some graffiti are more
extended and probably also include common lexicon. However, this
material cannot realistically be analyzed; for some important sub-cor-
pora there are no updated and reliable editions, and in the case of the
Abu-Simbel and Buhen graffiti, for which we are lucky enough to have
Olivier Masson’s reliable editions, the reading difficulties are such that
in most cases they cannot be used with any degree of confidence. For
that reason, I will limit my analysis to a long graffito with certain par-
ticularities that have attracted the interest of scholars.
naz ∞i∞ | bÿ≤ | esak?dow“ | mÿqudem | pisma≤k | bebint | mo | ne |
psÿ“[|?] ai[-]iqom
The most striking feature of this inscription, engraved by a Carian sol-
dier during the Nubian campaign in 593/592 BC, is the suspicion that
the personal name pisma≤k alludes directly to the Pharaoh Psammetichus
II, under whose supreme command the military campaign was under-
taken; this would closely link the graffito to the well-known Greek
inscription, also from Abu Simbel, in which the same Pharaoh is men-
tioned.22 It is true that pisma≤k (and variants) is well documented as a

21
For instance, to isolate only two personal names, “oru≤ and annibrs≤. Also theoret-
ically possible is the segmentation anni brs≤.
22
Bernand-Masson (1957), inscription nº 1: basil°ow §lyÒntow §w ÉElefant¤nan
Camat¤xo, / taËta ¶gracan to‹ sÁn Cammat¤xoi tÇoi YeoklÇow / ¶pleon, ∑lyon d¢ K°rkiow
katÊperye, uÂw ı potamÚw / én¤h: éloglÒsow dÉ ∑xe Potasimto, Afigupt¤ow d¢ ÖAmasiw:
294 CHAPTER SEVEN

name often used by the Carians in Egypt, but the appearance of the
word esak?dow“ in the graffito is favourable to this hypothesis: kdowº
/kndowº/ (or its phonetic result) has been convincingly compared to
Lycian xñtawa- ‘to rule’ (cf. also xñtawat(i)- ‘ruler, king’ = CLuw ¢andawat(i)-
‘supreme authority, king’).23 Therefore, in esak?dow“ . . . pisma≤k, we could
attempt to find a meaning ‘king Psammetichus’ or similar. Both the
initial sequence esa- and the precise function of final -“ are problem-
atic, and different possibilities can be envisaged: ese- could be compared
with the Lycian preverb ese-, for which a meaning ‘with’ has been pro-
posed (see Melchert DLL s. v.). Perhaps in this instance it may have
a reinforcing function. As for -“, two hypotheses can be considered: it
could be a suffix attached to the verbal stem kdow- ‘to rule’ in order
to create a noun (cf. Lycian -za?), or it could be a nominative plural
ending. This latter theory is less consistent with an interpretation of
pisma≤k as the name of the Pharaoh, since he would appear at the same
level as the other alleged commanders. Moreover, in the rest of the
Carian graffiti there is no form that can be likened to the names of
military officials cited in the corresponding Greek inscription.
Ultimately, these doubts cannot be resolved, as the rest of the inscrip-
tion remains unclear. It is possible that both the identification of pisma≤k
with Psammetichus and the equation of kdowº = Lycian xñtawa- are
correct, but the overall interpretation of both words as ‘king Psammetichus’
may still be erroneous.

D. THE LONGER INSCRIPTIONS

Some of the Carian inscriptions found in Caria proper are more extended
than the rest. We may assume that their content can be broadly defined
as legal: honorary decrees, regulation of cults, lists of priests, and so
on. As the study of fragmentary languages shows, the longer texts are
the most difficult to analyze. A good example is our meagre knowl-
edge of Milyan vis-à-vis Lycian: having at our disposal a text as long

/ ¶grafe dÉ èm¢ ÖArxon ÉAmoib¤xo ka‹ P°leqow OÍdãmo. “When the king Psammetichus
came to Elephantina, this was written by those who sailed together with Psammetichus,
(son) of Theokles and who came beyond Kerkis, as far as the river permitted. Those
speaking foreign tongues were leaded by Potasimto, the Egyptians by Amasis. We have
been written by Arkhon (son) of Amoibikhos, and Pelekos, son of Eudamos”.
23
Adiego (1995:19–20).
 ANALYZING CARIAN INSCRIPTIONS 295

as the Milyan inscription in the Xanthos stela does not necessarily guar-
antee an understanding of the grammar and lexicon of this language,
but in fact implies quite the opposite. It is preferable to work with a
sizeable corpus of relatively brief texts—as is the case with Lycian—
than to be restricted to a single, very long, and generally impenetra-
ble text.
This is also the situation with Carian. Texts such as the Kildara
inscription are currently impossible to analyse. We can identify some
isolated words, but we know nothing about the context in which they
appear.
The only way of ‘getting inside’ such a text is through the existence
of a translation. Of the entire Carian sub-corpus of longer inscriptions,
the only inscription that is accompanied by even a minimally legible
text, and whose content corresponds undoubtedly to the Carian text,
is the bilingual inscription of Kaunos. But even in this case, the results
that we are able to obtain are very limited. Our analysis of longer
inscriptions will begin therefore with this example.
A problem that seriously impedes the interpretation of this type of
inscription is our current inability to identify verbal forms. Without
knowing which words represent verbs it is practically impossible to
devise any approach to analysing the structure of the text and the func-
tion of the common nouns that it contains. Even in the bilingual of
Kaunos, where the existence of the Greek inscription ought to help
identify verbal forms, no agreement has been reached on which of the
words must be identified as verbs. As we have already seen, the prob-
lem of identifying verbs also affects shorter inscriptions, and only with
a substantial increase in the Carian documentation available will we
be able to resolve this great problem.

1. The Kaunos Bilingual Inscription

We will begin with the only text that can be read alongside a paral-
lel Greek text. The possibility of comparing the two texts explains the
special attention that this inscription has received since its discovery.
Even the editors of the inscription, Frei and Marek, did not limit them-
selves to a simple epigraphical edition, but offered in addition a first
attempt to establish the parallel passages in both texts, paving the way
for further research.24 Following this initial study, other scholars have

24
See Frei-Marek (1997), (1998).
296 CHAPTER SEVEN

drawn up essays of interpretation for the overall text or for concrete

sections.25 In the present work, I follow in general terms the views
adopted in Adiego (1998a), completing the discussion by referring to
the works cited above. No real developments have appeared since then,
and the initial enthusiasm inevitably generated by the discovery of such
a unique document has clearly diminished, since the results that have
been obtained are relatively modest.
Of the 18 lines of the Carian inscription, the current possible inter-
pretations are in fact limited to the first 10, essentially for three rea-
sons: (1) these 10 lines correspond roughly to the conserved part of
Greek inscription, the rest of which has been lost; (2) lines 11–12 of
the Carian text is unfortunately very difficult to read because of the
crack in the stone, which creates a lacune that makes it very difficult
to continue the analysis of the text; (3) the last lines of the Carian text
are practically impossible to interpret, since most of the forms are hapax
legomena and display no external characteristics that could help their
understanding. The segmentation of words in these final lines is also a
very difficult task.
The identification of proper names that appear in both texts has
proved fundamental to an analysis of the comparable parts of the
inscriptions. The correct interpretation of sb as a coordinative con-
junction is also relevant, as it thus becomes an important tool for estab-
lishing the phrases and sentences of the Carian text. A third decisive
factor, already noted by the editors, is the existence of textual paral-
lelisms with the other large text from Kaunos (C.Ka 2).

Kaunos bilingual (lines 1–10) Greek version

kbidn uiomln i[---] ¶doje Kaun[¤]oiw §p¤ dhmio[u]-

inis drual nik[--] rgoË ÑIpposy°nouw: Nikok-
lan lysiklas[-?] l°a Lusikl°ouw ÉAyhna›o[n]
otonosn sb lys[ikl] ka‹ Lusikl°a Lusikrãt[ouw]
an lysikratas[-?] [ÉA]yhna›on proj°nouw e[‰nai k-]
otonosn sarni[“] a‹ eÈerg°taw Kaun¤v[n auto-]
mdot2 un sb undo[--] Áw ka‹ §kgÒnouw ka‹ [-----]
tl“ kbdyn“ sb b2o[--] n auto›w §[. . .
ol“ otr“ sb a∞t [ms]-
kmt absims sb [---]

25
See among others Hajnal (1997b), Melchert (1998), Neumann (1998).
 ANALYZING CARIAN INSCRIPTIONS 297

A quick look is sufficient to identify the correspondence of all the proper

names that appear in both texts. Even in the only case where the words
are not formally related—the reference to the Kaunians, reflected in
Carian by means of kbidn, kbdyn“—Lycian evidence (Xbide is the Lycian
form for Kaunos) dispels any doubt. All these correspondences have
been underlined in the table above.
The onomastic formulae appear clearly in accusative case (marked
with the ending -n: nik[--]lan, lys[ik]lan), which allowed the first editors
to correctly identify the words ending in -“ as plural accusatives. This
means that the words that present this ending (including sarni[“ ]) could
represent the Carian equivalent of the Greek accusative plural formula
proj°nouw . . . [k]a‹ eÈerg°taw Kaun¤v[n auto]Áw ka‹ §kgÒnouw. The fact
that the Carian formula contains more words that the corresponding
formula in Greek made it difficult to establish the precise connection
between the two texts, and the attempt the first editors led to mistaken
assumptions. The identification of the word otr“ with Lycian atra-/etli-,
‘person, self ’, made independently by various authors, has clarified the
situation decisively; it implies that the Carian passage corresponding to
the Greek formula proj°nouw e[‰nai k]a‹ eÈerg°taw Kaun¤v[n], which
immediately precedes [auto]Áw, must be identified in the more extended
construction, sarni[“ ] mdot2 un sb undo[--]tl“ kbdyn“ sb b2o[--]ol“, which
in turn precedes otr“.
The possible explanation for this extended formula in comparison to
the Greek version will be dealt with later. Before addressing this prob-
lem, we must focus our attention on what is undoubtedly the main
obstacle to interpreting these first lines of the Kaunos bilingual: the
identification of the possible verbal forms. The need to first locate the
verbs in the bilingual is clearly justified by the fact that some of the
forms appearing in the sequence sarni[“] . . .b2o[--]ol“ have entered into
the discussion about possible verbal forms.
The Greek construction of the sentence is clear: an impersonal verb,
‘it seemed good’, constructed with dative (‘to the Kaunians’) and an
infinitive clause as the subject of the impersonal verb (‘that Nikokles
(. . .) and Lysikles (. . .) were public guests and benefactors of the
Kaunians’). As Greek syntax demands, the subject and nominal pred-
icate of the infinitive clause are in the accusative case. Much discus-
sion has arisen as to whether the Carian text contains, or indeed can
contain, a construction of this kind, or whether the formula of the
proxeny decree is instead expressed in Carian in a very different way.
The only clear indication that the Carian text offers is the fact, mentioned
298 CHAPTER SEVEN

above, that the personal names of the honoured Athenian citizens are
in accusative. This implies at least a verb that either directly or indi-
rectly governs these accusatives, and all the proposals formulated to
date attempt to find verbs among the words of these first lines. Scholars
have adopted two differing approaches to the question; while some of
them have assumed that there is only a verb in personal form, with-
out the actual presence of an infinitive, others have claimed to recog-
nize an infinitive form, which would indicate the further existence of
a main verb or equivalent construction.
The options suggested as possible infinite or finite verbal forms are
uiomln, mdot2un, and un (2×; from segmenting mdot2 un and un do[--]tl“ ).
Although the verbal character of rual has also been considered, in the-
ory there is a certain consensus to interpret it rather as a noun, within
a formula equivalent to Greek §p‹ dhmio[u]rgoË ÑIpposy°nouw.
The first editors adopted the approach of postulating a finite verb
construction. In fact, they claimed to have identified two finite verbs,
uiomln and mdot1un (interpreting O t2 as a glide sound between o and
u). Both would be third person plural preterite verbs with the respec-
tive meanings ‘to decide’ and ‘to make, to invest as, to establish as’
(Frei-Marek 1997:29–30), and the accusatives would be dependent on
this latter verb.
Melchert (1998) suggested that the accusatives are directly constructed
with a finite verb. But unlike Frei-Marek’s proposal, he prefers to clas-
sify uiomln as a noun with the meaning ‘decree’ (kbidn uiomln ‘decree
of Kaunos’, with kbidn as plural genitive of the place name),26 and
although he agrees with the first editors in considering mdot1un (also
interpreted as mdo/w/un, with O t2 as a glide) to be a finite verb gov-
erning the accusatives, he analyzes it as a first person plural preterite,
‘we have established, we have install (as)’. A weak point of Melchert’s
interpretation is that the analysis of this word as a first person verb
depends on the assumption that O represents a /w/ sound, as he com-
pares an alleged ending -/wun/ with Hittite preterite first plural end-
ing -wen and with the corresponding Lydian ending -wn.
The proposal in Adiego (1998a) is very different. Here, the suggested
segmentation is mdot1 un (although without considerations about the
phonological value of O t2, which was studied in a subsequent paper,

26
For a place name kbid- as plurale tantum, Melchert reminds us of the similar
interpretation for the Lycian name of Kaunos, Xbide (Melchert 1998:37).
 ANALYZING CARIAN INSCRIPTIONS 299

Adiego 2002), and both this un and the un segmentable in undo[--]tl“

are taken as infinitives depending on the finitive verb uiomln, for which
Frei-Marek’s interpretation as ‘they decided’ is accepted. As for un as
infinitive, it is compared with the Luwian infinitive ending -una (aduna
‘to eat’) and the whole form is considered as the infinitive of a verb
‘to make’, by comparing Lycian a(i)-, Cuneiform Luwian à-/àya- ‘to
make’. The remaining word mdot2 is then interpreted as a complement
of sarni[“ ] that would therefore constitute a locution equivalent to Greek
proj°nouw. This hypothesis was refined in Adiego (2002), where argu-
ments were given in favour of O as the Kaunian form of the letter c
t, and for the interpretation of endings in -ot, -ot2 as plural genitives.
These hypotheses do not cover every theoretically possible interpre-
tation of each of the words involved, and it is impossible with our cur-
rent knowledge of Carian to choose one above the rest, or to simply
deem them all incorrect. In any case, all these analyses coincide in see-
ing sarni[“ ] (or sarni[“ ] mdot2) and undo[--]tl“ (or do[--]tl“) kbdyn“ as the
Carian equivalent to Greek proj°nouw [k]a‹ eÈerg°taw Kaun¤v[n]. This
interpretation seems to be one of the scarce certainties—leaving aside
the onomastic forms—that currently exist about the structure and mean-
ing of the Carian text. However, it does pose a serious problem, inso-
far as between kbdyn“ and otr“ = aÈtoÊw we are left with sb b2o[--]ol“,
which has no recognizable correspondence in the Greek part. It must
be remembered that this sequence led the first editors to make mis-
takes, because they assume that sb b2o[--]ol“ is the Carian equivalent of
Greek ka‹ §kgÒnouw ‘and the descendents’, and otr“ was therefore inter-
preted as a possessive (‘of them. their’). This intepretation was undoubt-
edly caused by the fact that b2o[--]ol“ is linked to the former text by
sb, which made it difficult to find an equivalent to aÈtoÊw, since it
could hardly be preceded by a coordinative conjunction.
The Lycian etymology for otr“ mentioned above showed that this
interpretation was mistaken, since otr“ must be the Carian word trans-
lated into Greek as aÈtoÊw. But this explanation, based on a very con-
vincing etymological approach, fails to account in any way for sb
b2o[--]ol“.
A provisional solution was suggested in Adiego (1998:23): b2o[--]ol“
could be another ethnic name coordinated with kbdyn“, referring to
ÖImbrow, the fortress near Kaunos27 that Bean (1953:22) identified with

27
Strabo XIV, 651.
300 CHAPTER SEVEN

the archeological remains on the summit of Ölemez Da<, the moun-

tain north of Kaunos. This hypothesis would fit well with the initial b2
of b2o[--]ol“, but the fact that the Carian word is incomplete makes it
less certain. Moreover, it is not unthinkable that sb is in fact coordinat-
ing b2o[--]ol“ not with kbdyn“, but directly with sarni[“ ] and undo[-]tl“ /
do[--]tl“, so that the Carian equivalent to the Greek twofold formula
proj°nouw [k]a‹ eÈerg°taw would be a threefold one, (sarni“ . . . (un)do[--]tl“ . . .
b2o[--]ol“ ).
The identification otr“ = aÈtoÊw makes it likely that the following
sequence, sb a∞t[ms]kmtabsims sb [, corresponds to Greek ka‹ §kgÒnouw
ka‹ [.28 In Adiego (1998a) an etymological connection for a∞t[ms]km(t?)
was proposed, based on the identification of mskmº with the Luwian
stem mas¢a- recognizable in ma“¢à¢it- ‘growth, prosperity’, and the sug-
gestion that a∞t- could be compared with Hittite katta, kattan ‘down’.
a∞t[ms]km(t? ) could therefore be a compound noun formally and seman-
tically comparable to ‘off-spring’, German ‘Nach-kommenschaft’ and
Spanish ‘de-scendencia’. As for the remaining ºtabsims or ºabsims, it was
suggested in the same study (Adiego 1998a:25) that this could be a
possessive, comparable to Lycian ehbi( je) < *ebesi( je)-. All these con-
nections are phonologically sound, but they fail to explain ºt º (if it does
not belong to the word a∞t[ms]km) and the final -ms.
From absims sb on, the possibilities of interpreting the Carian text
decrease dramatically: both the sudden interruption of the Greek part,
and the lacuna of almost an entire line (l. 12) make it impossible to
continue with the analysis. We can identify further possible instances
of the conjunction sb ‘and’, but it is impossible to identify which type
of syntactic elements it is coordinating. Among the words of this final
section of the inscription, special attention should be paid to the sequence
orouo, which recalls Lycian arawa- ‘freedom’, a term that is consistent
with the sort of privileges conferred on the proxenoi in these types of
decrees. However, the segmentation is far from certain, as it leaves an
isolated letter H between sb and orouo. Also of interest here are the pos-
sible presence of the word mno- ‘son’ under the form mnos, and the
final sequence aitusi, where we could identify a possible imperative aitu
or preterite ait of a verb corresponding to Lycian a(i)- ‘to do, to make’
(cf. Lycian aitẽ ‘they made’).
28
The integration [ms] is based in the parallel sequence a∞tmsk[ in C.Ka 1.
 ANALYZING CARIAN INSCRIPTIONS 301

2. The Kaunos Inscription C.Ka 2

[-(?)-ui?]omlã qrdsgrdso[-]i[
[-?-]r sb a∞mnnartnyr obsmns[
[-?-]∞arlanoã sb z“ariosã i∞[
[-?-]nudrma ∞yrpai sarni“ sb u[
[-?-]aH punot2 otr“bi sb a∞tmsk[m
[-?-]dbi _1aitk ouor gdb“laã1_i[-]
[-?-] sarni“ sb 1orsol“ sb uHbit
[-?-]bi qrdsol“ ait _1mali H∞it
[-?-]intnor ∞yrapai≤ umot2 oba
[-?-]_iurt obsmsmñ1ñ ouor mt1_yr
[-?-][-]abrun∞ur[-]“yn“Hynn sb vacat
[-?-----------------]tbsms _1mali[
[-?------------------]maH 1sb an[
[-?----------------------]ba vacat
C.Ka 2, the most extensive Carian inscription, remains a virtually
impenetrable text. The discovery of the Kaunos bilingual has allowed
us to establish some connections between the two texts, and it is cer-
tainly very significant that the word sarni“, which in the new bilingual
seems to represent either totally or partially the Carian expression for
‘proxenoi’, is also present in C.Ka 2. Other important correspondences
are the word otr“ = aÈtoÊw, the possible word for ‘descendence’, a∞tmskm(t?),
and the noun or verb that appears apparently to indicate that the
inscription is a decree (uiomln, [--]mln in C.Ka 2). All of these ele-
ments seem to suggest that the content of C.Ka 2 and C.Ka 5 has a
similar purpose, but these correspondences are not corroborated by a
comparable overall structure. In this respect, the absence of clear ono-
mastic formulae, which we would expect to find if dealing with the
concesion of proxeny to more than one individual (note the plural
sarni“ ), is very puzzling.
In the first line, after ]mln, a word qrds can be segmented, recalling
a similar sequence in Kildara (C.Ki 1). In this case it seems to be fol-
lowed by another word etymologically related to it, grdso[, which points
to a possible figura etymologica. Another derivative of qrds appears in line
8, followed by a word ait: qrdsol“ ait. The final -“ points to an animate
accusative (and perhaps also nominative) plural, and ait recalls, as
Melchert (1998:35) has pointed out, Lycian aitẽ ‘they made’, so that an
interpretation as ‘the qrdsol-s made’ or ‘they made (them) qrdsol-s’ seems
likely. The presence of qrds immediately after ]mln, the noun or verb
302 CHAPTER SEVEN

that initiates a decree, implies that qrds could be referring to the city
or to a city institution such as the assembly.29
sarni“ appears twice, in both cases followed by sb, ‘and’. In the first
instance, it is impossible to decide whether sb is joining sarni“ to another
plural noun, due to the crack of the stone. The second example is
much clearer: sarni“ sb 1orsol“ is undoubtedly a sequence of two accusative
(or nominative) plural nouns. The first editors of C.Ka 5 tried to com-
pare this latter word with C.Ka 5, l. 8–9 b2o[--]ol“, and proposed the
integration b2o[rs]ol“. Although this hypothesis is very attractive, it is
complicated by a serious difficulty: ÿ b2 seems to be formally closer to
± b (C.Ka 4) than to 1.30
The only form that can be identified as a possible proper name is
l. 4 ∞yrpai (nominative?), which reappears as genitive singular in l. 9
(∞yrapai-≤ ).
The clearest syntactic parallel between C.Ka 2 and the Kaunos bilin-
gual is l. 6–7, otr“bi sb a∞tmsk[m . . .]dbi = Bilingual otr“ sb a∞t[ms]kmt
absims ‘themselves and (their) descendants’. Unfortunately, it is impos-
sible to calculate the letters missing between a∞tmsk[ and ]dbi, but it
seems likely that =bi was attached to (a form of ) the word a∞tmskm(t?),
so that the formula in C.Ka 2 would present a ‘X= bi and X= bi’
structure: otr“=bi sb a∞tmsk[m . . .]d=bi. The exact function of this = bi
is unknown. It could be a particle reinforcing the coordination (‘both
them and (their) offspring’, ‘not only . . . but also . . .’), but it could also
be a postposition (‘for’, ‘with’?). Both interpretations permit an etymo-
logical connection with Lycian -ppi in hr-ppi ‘on; for’ (cf. for this form
Lyc. hri = CLuw. “arri ‘up; (on) top’). Carian =bi could be a generalisa-
tion of the lenited form of this particle instead of Lycian unlenited -ppi
(cf. the parallel process in Lycian -be vs. CLuw. -ppa, Melchert DLL
s. v.).

3. Sinuri’s Longer Inscription (C.Si 2)

(a) [--]ryin ∞tmño≤ : sb ada ∞tmño≤
eri : pisñoi mda : pñmnn≤ñ : pda-
∞m≤uñ ∞i “aoyr∞ri mt∞elã

29
For a purely theoretical etymology suggested in Blümel-Adiego (1993:94) see
Chapter 11, s. v.
30
That 1_ and ± are different letters (which rules out a triple equivalence 1_ = ±
= ÿ) is confirmed by their co-occurrence in C.Ka 4.
 ANALYZING CARIAN INSCRIPTIONS 303

ñmailo mda lrHñ : stspñ vacat

sm“s[--5--] sb añmsñsi mda
sm[--7--]a∞e[
∞[--8--]tuñdñ[
ñe-?-[
(b) pim[. . .]
Ha?[. . .]
The longer inscription from Sinuri, engraved after a long Greek text,
could be a true bilingual, but the Greek text is so badly damaged that
it turns out to be practically unusable as a means of understanding the
Carian one.
The greatest progress in the analysis C.Si 2 has been made by Diether
Schürr: he was able to identify in the first line of the text a reference
to the Carian dynasts Idrieus and Ada, sons of Hekatomnos ([--]ryin
∞tmño≤: sb ada ∞tmño≤ ). Moreover, Schürr argued convincingly that it
belonged to the same stela as a fragment of a Greek inscription also
discovered in Sinuri. This fragment adds very limited, but nonetheless
interesting information about the content of the Greek inscription pre-
ceding the Carian one, and it is certainly more valuable than the infor-
mation that can be obtained from the scarce remains of the Greek text
in the bilingual fragment. The Greek unified text contains two decrees
of the abovementioned dynasts, both conceding an ét°leia (tax exemp-
tion). A personal name, Ponmoonnow, is also mentioned in genitive in
the first Greek decree, and Schürr also found it in the Carian text
(l. 2 pñmnn≤ñ, alternative reading ?pñmun≤ñ). This could be the same
name that appears in later Sinuri inscriptions in the form Pormounow,
as the denomination of the syngeneia devoted to the administration of
the temple.
In Adiego (2000), I attempted to make some progress in the under-
standing of the text. Many of the proposals outlined in that study were
very hypothetical, and I will not repeat them in any detail in the pre-
sent book. Perhaps the main idea that should be retained is my insis-
tence on searching for verbal forms: it is logical to assume that some
verbal form must exist not very far from the mention of the Hekatomnids,
expressing the action carried out by the dynasts, in the same way that
in the second Greek text the names of the dynasts were very proba-
bly followed by a verb ¶[dvkan] (if Robert’s very plausible integration
is accepted). In Adiego (2000) it was suggested that both pisñoimda (l. 2)
and ñmailomda (l. 4) were verbs. In the first case, a connection was
304 CHAPTER SEVEN

proposed with the Common Anatolian verb for ‘to give’ (Hitt. pài-/piya-
CLuw. piya-, Lyc. pije, etc.); in the second, I suggested a rather kling-
klang resemblance to Lycian ◊mmaitẽ in the trilingual of the Letoon of
Xanthos, a third person plural verb with a possible meaning ‘to install,
to build’. In both alleged verbs, mda was interpreted as the same ele-
ment that appears as mdane in other Carian inscriptions, which I con-
sider to be a particle chain (see p. 324). As for the resulting sequences
pisñoiº, ñmailoº, o (in both forms) and i (in pisñoiº) were interpreted as
clitic pronouns (Adiego 2000:142):

Verb Direct Object Indirect Object -mda element Translation

pisñ- -o- -i- -mda ‘they gave-it-to him/to
them’
ñmail- -o- -mda ‘they established it/him’

These postclitic pronouns, attached to the verb, were syntactically com-

pared with the well-known ‘nasalized preterites’ of Lycian ( pijetẽ, m̃maitẽ ),
where the nasalized vowel comes from the univerbation of the verb
with a clitic pronoun ( pije-tẽ < *-to+om ‘he gave-it’).
The rest of the ideas contained in Adiego (2002) are merely hypo-
thetical. In the sequence pñmnn≤ñ: pda∞m≤uñ ∞i a construction with the
relative ∞i was envisaged, and two possible interpretations were sug-
gested: ‘Pñmnn- who (is) the priest’ or ‘the syngeneia of Pñmnn-’. Both
interpretations for pda∞m≤uñ were considered to be consistent with an
etymological connection of pdaº with Lycian pddẽn- and Hitt. peda-,
‘place’. For eri, I proposed a possible word for ét°leia, on the basis of
its resemblance to Lycian arawa-, ‘exemption (ét°leia), freedom’, but
without discounting the alternative view suggested by Neumann (apud
Adiego 2000:144) of eri as preverb identical to Lycian eri (cf. Hitt. ar¢a).
Leaving aside all these speculative attempts, the most conclusive and
remarkable morphological information provided by the bilingual inscrip-
tion of Sinuri is the form pñmnn≤ñ, which must surely be an accusativus
genitivi or, more correctly, an accusative formed on the basis of the (old)
possessive adjective, comparable to the Lycian forms in -hñ (Lusãñtrahñ).
The same analysis is probably also valid for pda∞m≤uñ, although the
presence of u between ≤ and ñ obliges us not to disregard alternative
interpretations.
 ANALYZING CARIAN INSCRIPTIONS 305

4. The Hyllarima Inscription (C.Hy 1)

The recently discovered bilingual stone in Hyllarima has proved to be
a fragment that fits well alongside what was until recently the only
known inscription from that location. Joining the two pieces together
has provided us with a complete version of the Carian part of the orig-
inal stela with practically no problems of reading, and has also helped
us to understand the Greek texts that accompanied the Carian inscrip-
tion. Of these, the only one that seems to be contemporary or very
close to the Carian texts is the brief list of ‘priests of all the gods’ in
column B:

Column A Column B
“asqariod dymda kdu≤opizipususot
muot armotrqdosq mol“ msot ylarmit
brsi ari“≤ brsi≤ flere›ew ye«n pãntvn:
mane : u≤ol≤ ÑErm¤aw Fan°v ÑErm¤adow:
rtim u≤ol≤ pur?i≤ flereÁw ye«n pãntvn:
ÑUssvllow ÉArrissiow
u≤bzol tñu≤ brsi≤
pau mane≤ ybr-
s≤
(A later Greek inscription containing (A later Greek inscription containing a
a list of priests of Apollon follows) priesthood sale follows)

(Recall that other, later Greek inscriptions appear on both sides of the
inscription, see p. 136, n. 25 for details).
Any attempt to analyse the Carian text is seriously hampered from
the beginning by the doubts about the exact order in which the first
two lines of both columns (in fact, the only part containing common
vocabulary, since the remaining lines contain only onomastic formulae)
should be read. As pointed out in Adiego-Debord-Varinlio<lu (2005),
three posibilities can be considered: (1) to read the first line of each
column and then the second line of each column (A1 + B1 +A2 +
B2); (2) to read the two first lines of column A and followed immedi-
ately by the two Carian lines of column B (A1 + A2 + B1 + B2); (3)
to read all of column A before reading column B, that is, to treat the
columns as (relatively) independent texts (A + B). The existence of a divid-
ing line bewteen the columns seems to support the last solution, but
given that the two columns are apparently lists of priests, it is strange
306 CHAPTER SEVEN

that an identical word for ‘priests’ cannot be found in each column.

This option would also raise another intriguing question: why are the
names of the priests in the first column in Carian and the names in
the second column in Greek, despite the fact that Carian is used for the
headings of both columns?
Both solution (1) and solution (2) imply that the two first lines of
both columns would be the common heading for the entire Carian
inscription. But there are no clear arguments for or against either solu-
tion: the first lines in both are for now impenetrable, and in the case
of the second lines, although they do contain words that can be inter-
preted (see immediately below), their structure does not allow us to
clarify whether or not they consitute a consecutive sequence.
From the two initial lines of both columns, the only forms for which
we can find a plausible interpretation are armotrqdosq and mol“ msot ylar-
mit. The first, armotrqdosq, is undoubtedly the most relevant contribution
of the new Hyllarima fragment: its interpretation is straighforward—
although the syntactic connection with the rest of the text remains
unclear—and it is of great interest from both a linguistic and a cul-
tural perspective. In armotrqdosq, it is easy to recognize two typically
Anatolian god names: armo-, the name of the moon good (Hitt.
Arma-), and trqd-, the storm god already documented in other Carian
inscriptions (Kildara, Iasos). As noted in Adiego-Debord-Varinlio<lu
(2005), the simplest explanation of armotraqd- is to see it as a dvandva
compound (“Arma (and) Tarhunt”), which would allude to worship
devoted to a divine paredria.31 As for the morpho-syntactical interpre-
tation of armotrqdosq, we can identify here an ending -os (assuming that
trqd- retained its original nt-stem), followed by q, perhaps some sort of
enclitic connecting particle (in Adiego-Debord-Varinlio<lu 2005, I pro-
pose an etymological link with the enclitic coordinative conjunction
CLuw. -¢a, ‘and’), although it is by no means clear what this -q is actu-
ally connecting.
Regarding mol“ msot ylarmit in column B, in Adiego (2002:17) it was
analysed as a formula corresponding roughly to Greek flere›ew ye«n
pãntvn ‘priests of all gods’, which would mean ‘priests (mol“ ) of the
Hyllarimean gods (msot ylarmit)’. Although this interpretation requires

31
Other solutions seem less convincing, see Adiego-Debord-Varinlio<lu (2005:616–617)
for details.
 ANALYZING CARIAN INSCRIPTIONS 307

us to make a series of (sometimes fragile) assumptions about the mor-

phology and semantics of the three words involved, at least in the case
of ylarmit, it seems plausible to consider a connection with the place
name Hyllarima, as suggested for the first time by Ray (1988:152).32
The interpretation of msot as, ‘of the gods’, is based on the possibility
of connecting it with the Luwic word for ‘god’: note particularly Milyan
masa- ‘god’, and also Lyc. maha(na)-, CLuw. mà““an(i)-. The stem also
appears in Carian onomastics: msnord-, Masanvrada (see Chapter 11
s. v. msnord≤ ). Less definite connections can be established for mol“ =
‘priests’,33 where moreover the character of the final -“ (an ending or
rather a derivative suffix) remains unclear (see p. 318), although the
new inscription from Mylasa (C.My 1) now offers another example of
mol“ in the first line, preceding a long list of individuals. Even if the
meaning ‘priests’ proved to be incorrect, the appearance of mol“ in the
heading of two inscriptions consisting of lists of onomastic formulae
would still be worth noting.
From a morphological point of view, this interpretation of mol“ msot
ylarmit assumes that mol“ must be in plural, presumably in nominative
(as in the Greek formula), although the possibility of another case can-
not be totally ruled out. Following the same interpretation, msot ylar-
mit must by contrast be a genitive plural, characterized by the ending -t.
On these morphological repercussions of interpreting mol“ msot ylarmit
as ‘priests of the Hyllarimean gods’, see p. 319.
The remaining lines of column A consist exclusively of five onomastic
formulae, engraved at different times (see Adiego-Debord-Varinlio<lu
2005). With the exception of the second formula, where only the names
of the individual and the father are recorded, the rest of the formulae
are threefold: individual name, father’s name and grandfather’s name.
It is interesting to note that the names cited in the Carian part coin-
cide with several names that appear in the Greek inscriptions on
the same stone; particularly striking is the fact that in one of the
priesthood sales, the daimones of an Arrisis son of Imbrasis, and a
Hermias son of Arrisis, are quoted among the other divinities mentioned.

32
Schürr (1992:146) had attempted to identify a name corresponding to Greek
Eemiaw in the Greek list of priests by isolating armit, but the value y /y/ of W, now
confirmed by the Kaunos bilingual, makes the connection between ylarmit and Hyllarima
clearly preferable.
33
Cf. perhaps Lyc. mle-, ‘sacrificial offering (??)’ and related words (particularly mla-
traza, mluhidaza, two priestly titles) in Melchert DLL, ss. vv.
308 CHAPTER SEVEN

Both Arrisis and Imbrasis are names that appear in the first Carian
onomastic formula (brsi ari“≤ brsi≤ ), and it is reasonable to assume that
we are dealing with the same individuals.34

5. Other Inscriptions from Caria

The new inscription from Mylasa (C.My 1) consists of an initial line
as the heading for a list of personal names. Therefore, only this first
line contains what seems to be common lexicon. As is the case of most
Carian texts without personal names, it remains practically impenetra-
ble:
idrayridsemd ?bq mol“ ty∞[
The most interesting characteristic is the presence of the word mol“,
also found in Hyllarima, which has been analysed as accusative or nom-
inative plural with the meaning ‘priests’ (see above). The presence of
a noun in plural fits well with the following list of personal names. In
Adiego (2005:92–93) it was also suggested that ty∞[ could be the Carian
adaptation of Greek TÊxh, the goddess of Fortune. Therefore mol“ ty∞[
could mean ‘priests of Fortune’, ‘priests for Fortune’, although this is
merely hypothetical. The preceding sequence idrayridsemdbq is very difficult
to analyze. If the reading of d is certain, perhaps a word or particle
chain, md (cf. md orkn), could be segmented. In idrayriº, we can attempt
to identify two stems idra- and yri- that recall, respectively, the place
name Idriaw and the yri-formant in personal names such as “ayriq, idyri∞≤,
Saurigow. Cf. also (without tectal suffixation) E.Th 26 yri≤ and possibly
E.AS 5 pnyri≤ru.
Little can be said about the rest of Carian inscriptions from Caria.
The Euromos inscription C.Eu 2 does not seem to be a funerary
inscription, insofar as none of the typical words for ‘tomb, burial’ can
be identified, as already noted. Neither can we identify any clear ono-
mastic formulae. It is possible that the first sequence omob∞i contains

34
See Adiego-Debord-Varinlio<lu (2005:614) for a rather speculative attempt to con-
nect the formula brsi ari“≤ brsi≤ with Arrisis, Imbrasis and Hermias, by assuming that
the two Arrissis are the same person (which implies a descendence Hermias, son of
Arrisis, son of Imbrasis), and that the Greek individual Hermias would the same as
the Carian one brsi, a supposition based in the Carian name of Hermes Imbrasow/
Imbramow.
 ANALYZING CARIAN INSCRIPTIONS 309

the relative pronoun (hence the segmentation adopted here). As for the
following words, perhaps a genitive temazi≤ can be segmented. In this
case, the d that immediately follows it could be compared with d in
i[---]inis=d=rual from C.Ka 5. Another word that may be present in
the text is armon, which cannot be clearly interpreted: it could be directly
compared to armon in E.Me 8 (‘dragoman, interpreter’, although this
meaning would seem somewhat out of place here) or be analyzed as
an accusative in -n of the moon-god armo (attested in C.Hy 1a). The
rest of the inscription is even more difficult to fathom.
In the shorter text of Sinuri (C.Si 1) at least a possible onomastic
formula can be identified:
(1) adymd“ : yri∞ñ : t[-]rsi : [. . .?]
tbe≤
(vacat)
(2) yri∞ñ : binq : sñaidlo
t[-]rsi : [. . .?]tbe≤ is convincing as a N-Ø N-≤ structure, and the incom-
plete second name can be compared with Thebes qutbe, Kuatbhw.
The rest of the inscription remains unclear, a situation that is fur-
ther compounded by the reading difficulties. In Adiego (2000), sñaidlo
was tentatively analysed as a verb, comparable to pisñoi=mda, ñmailo=
mda. ºaid º was directly linked to Lycian a(i)-, ‘to make’, and sñº was
compared to Lycian se=ññ(e). Leaving aside these more provisional
attempts, one must focus on the striking disposition of the two texts; a
relatively large gap has been left between the two, and the fact that
in the first text a word adymd“ that possibly contains a plural ending
(-“, cf. mol“ ) is followed by a single onomastic formula, makes it plau-
sible that the formula adymd“: yri∞ñ was conceived as the heading of a
list of personal names, and that this list was never completed. The
inscription from Hyllarima shows how different names were added to
the list on subsequent occasions, and this could also have been also
the initial aim of C.Si 1.
The inscription of Kildara (C.Ki 1), shown below, is even harder to
analyse than the preceding texts:
[. . . . . . .(.)]zolba∞a[..(.)] kil[
[. . .]uda[. . .] trqdimr qrds tazomd[
kilarad[-]ybzsdmHnmkda[-]aHuq[
iasoum
The only elements to have been isolated as recognizable words are: (1)
the two instances of the place name Kildara (l.1: kil[, l.3; kilara or
310 CHAPTER SEVEN

kilarad? -d could be a case ending or a clitic form d, comparable to

-d- in i[---]inis=d=rual); (2) the god name trqd (l.2, or better trqdimr as
a compound name?); (3) the word qrds, also in l.2, undoubtedly related
to qrds, qrdsol in C.Ka 2, and for which an ‘institutional’ meaning has
been proposed. The only clear evidence for the content of the Kildara
inscription is negative: there is no point of contact with the Greek text
inscribed immediately after it. The name of the individual honoured
as proxenos and eueregetes in the Greek inscription, Uss[vllow?] son of
Samvow, does not appear,35 and there is no sequence that can be com-
pared with the phraseology used in the proxeny decree of Kaunos
(C.Ka 5).
As for C.Ka 4, the fragmentary inscription from Kaunos that is
largely incomplete and has no clear context, it is impossible to even
attempt an analysis:
[. . .]u≤ou≤ ibrsdr[-]
[. . .]a yomlnr_1i
[. . .]dar1_ idym“
Only two elements have been recognized here: the form ibrsº, without
any doubt related to the family of names ibrs(i)-, brsi, and yomln, a
variant of the word uiomln in the first line of the Kaunos bilingual,
interpreted as a verb or noun pertaining to the semantic field of ‘decree,
decision, to decree, to decide’.
The two Stratonikeia inscriptions (C.St 1, C.St 2), as far as we can
currently tell, consist exclusively of onomastic formulae.

E. BALANCE AND SOME CONTROVERSIAL QUESTIONS

The sum of the preceding analysis of a great part of the Carian cor-
pus can seem at first sight somewhat disappointing: the results that can
essentially be considered conclusive are very scarce, and even some of
these are perhaps overly optimistic, since speculation has been unavoid-
able in several cases. Strictly speaking, our ability to analyze Carian
inscriptions is limited to the identification of proper names and ono-
mastic formulae, which allows us to understand only a very small area

35
However, one could be tempted to read ßAmous “ !amous instead of ÎAsouM
iasoum in the last line, in order to obtain the name of the father of Uss[ollos?], Samoos,
mentioned in the Greek inscription, see p. 142.
 ANALYZING CARIAN INSCRIPTIONS 311

of Carian grammar. Only in the case of the Kaunos bilingual can

the interpretation be developed to any extent, thanks to existence of
the Greek version, but even with this advantage we do not obtain the
results we would hope for. As for the rest of the texts, the inscription
of “rquq qtblem≤ (C.xx 1) is something of an exception, since the bril-
liance of Melchert’s interpretation has not been matched for other
inscriptions.
These limited results will be systematized in the next chapter, where
I shall focus on two problems that have arisen in the course of this
chapter, namely the s-ending and the alleged verbal forms mdane and
variants.
 CHAPTER EIGHT

MORPHOLOGICAL FEATURES

In this chapter I summarize the morphological traits that can be obtained

from an analysis of the inscriptions, and interpret them from a histor-
ical and comparative perspective. Needless to say, the resultant picture
will remain exasperatingly incomplete, and in many cases will be based
on extremely fragile evidence. It will however be sufficient to demon-
strate the clear relationship between Carian and Indo-European Anatolian
languages and, more specifically, Luwic dialects.

A. NOMINAL INFLECTION

Our present knowledge of nominal inflection is based mainly on the

information given by the personal names, to which we can also add
some common nouns. There are no examples of adjective inflection,
with the exception of ethnic forms, which can be interpreted as adjec-
tival complements (for instance otonosn in C.Ka 5). Only a few endings
can be established with some confidence.

1. Nominative Singular
The singular nominative ending is systematically -Ø, as appears clearly
in the onomastic formulae (u≤ol, arli“, ada, etc.). This also seems to be
the ending for common nouns, for instance upe, ue ‘stele’.
This zero ending can be directly compared with Lycian and Milyan
sg. nom. c. -Ø, and we can imagine it to have a similar origin: a PIE
*-s > Proto-Anatolian (PA) *-s (Hitt., CLuw, HLuw, Pal., Lyd. -/s/)
dropped in absolute final position. Words such as the abovementioned
upe, ue seem to be vocalic stems of the common gender, also with loss
of final -s.1

1
If they are not old eh2- stems with inherited ending -ø for the sg. nom.
 MORPHOLOGICAL FEATURES 313

2. Accusative Singular
An ending -n for the singular accusative has been clearly established
thanks to the Kaunos bilingual, confirming former interpretations which
led to the same conclusion.2 The personal names nik[—]la-n, lys[ikl ]a-n
show this ending for a-stems of personal names (Frei-Marek 1997:34,
48). Another clear example of -n accusative is orkn (‘bowl’ or similar).
The ending -n for singular accusative also appears in the pronominal
flexion (see below p. 320).
This ending is also comparatively transparent: PIE *-m > PA *-n.
Unlike in Lycian, where the ending in contact with a stem final vowel
a, e results in a nasalized vowel (noted by means of special letters <ã>,
<ẽ>), the Carian forms with -an could point to a conservation of the
nasal, although it is not impossible that a process similar to the Lycian
example has occurred, though not noted with the same graphical pre-
cision. In ork-n we find the same ending, but in this case after a con-
sonant, which could represent a syllabic nasal or merely a defective
vowel notation (/orkVn/).

3. Genitive Singular
The most characteristic ending for the genitive singular, confirmed only
by personal names, is -≤. Melchert (2002:311) has argued convincingly
that its origin lies in PA *-assì-, a possessive suffix with i-mutation that
serves to create adjectival forms in Luwic dialects (CLuw. -assì-, HLuw.
-asi-, Lycian -ahi, -ehi, Milyan -asi, -esi) that carry out the proper func-
tion of a genitive (cf. Lyc. ẽni mahanahi ‘mother divine’ = ‘mother of
the gods’). This etymological connection explains the presence of -≤,
most probably a palatal sibilant /ç/ resulting from the contact of s
with ì, and either dropped or not graphically noted in Carian.3
It is less clear if -≤ in Carian continues to act as an adjectival suffix,
agreeing in number and case with the word that complements it, or if

2
See Melchert (1993:80).
3
Melchert (personal communication) informs me that his views about the origin of
Carian -≤ have changed since the publication of Melchert (2002). He now proposes
that -≤ is cognate with Hierohplyphic Luwian -(a)si, and that both forms come from
the true PIE genitival ending, *-osyo. This interesting alternative hypothesis merits fur-
ther discussion that will not be included here. In any case, it does not alter the phono-
logical explanation, formulated above, of -≤ as a palatal sibilant resulting from the
contact of s with i.
314 CHAPTER EIGHT

in fact it has been reinterpreted as a pure case ending. A single form,

pñmnn≤ñ in C.Si 2, suggests that -≤ retains its adjectival value, as it shows
a further accusative ending -ñ. However, the example is too isolated
to be considered proof that the adjectival character of -≤ has been pre-
served: it could merely be the result of a secondary process of agreement,
as occurred in Lycian with -hñ. In any case, whatever the explanation
of pñmnn≤ñ, it is a good illustration of that fact that in Carian, like in
Lycian, the limits between genitive and possessive adjective constructions
were rather blurred.4

4. The -s Ending and the Problem of Carian Datives

Throughout the chapter devoted to the analysis of Carian inscriptions,
we have repeatedly mentioned the existence of a specific -s ending,
which has been interpreted as a dative ending by some authors (Schürr,
Melchert, Vittmann). Our analysis has led us to the provisional con-
clusion that the solution is perhaps not so clear. We shall now attempt
to analyse this ending more carefully and evaluate the plausibility of
interpreting it as dative.
Any analysis of the possible examples of s-endings is hampered from
the beginning by the difficulty of distinguishing between true s-ending
forms and simple Ø-nominatives of s-stems (whose existence is undeni-
able, given the presence of s-stems inflected in genitive: ibrs-≤, idyes-≤,
pals-≤, etc.). To quote a case already mentioned (see p. 283), a word
such as smdÿbrs (C.Ha 1) could for example be analysed as an s-ending
form of a stem smdÿbr-, comparable to k“atÿbr, dtÿbr ardybyr-≤, or as a
nominative of an s-stem smdÿbrs-, parallel to ybrs-≤. The examples of forms
ending in -s from the sub-corpus of Memphis also help to illustrate
this. Listed below are all the Memphis inscriptions that present these
types of words. I exclude only E.Me 39, where the reading of -s is not
certain. The inscription is also very fragmentary and hardly usable:
ap[—]ws / a[rb]ikarm≤ ∞i (E.Me 23)
tdu≤ol / kbos | “amsqi[. . .? (E.Me 24)
(a) idmns | myre≤ ∞i | mdayn ∞i / (b) idmns | myre≤ ∞i (E.Me 33)
ntokris | dw≤ol≤ | mwdon≤ ∞i (E.Me 35)
[q?]lalis / [?]iam≤ ∞i / alos ∞arnos (E.Me 45)
tqtes | paraibrel≤ ∞i | mn[o-?] (E.Me 47)

4
More problematic is the intepretation of the form pda∞m≤uñ, see the discussion in
Melchert (2001:311).
 MORPHOLOGICAL FEATURES 315

From all of these examples, the most likely to have the s-ending form
is undoubtedly ntokris; Vittmann (2001:52) has argued convincingly that
this name was taken directly from Egyptian, ruling out Ray’s theory that
it can be identified as a form that arrived in Carian from the Greek
Nitvkriw, the only way to explain the final s as a part of the stem (by
assuming it to be an s-stem arising from the Greek sigmatic nominative).
Also suggestive of an s-ending, if the reconstruction of the initial letter
is accepted, is [q?]lalis, given the existence of a name qlali-≤ (G 2),
Greek Kolaldiw, Kulaldiw.
The rest of the examples are ambiguous: ap[—]ws, idmns and tktes
allow more than one interpretation.5 Finally, kbos is very unlikely to be
a personal name in nominative of an s-stem, because it occurs pre-
cisely after the nominative tdu≤ol, i.e. the name of the deceased. However,
it could also be a title or an adjective accompanying tdu≤ol (an ethnic
name? see p. 278), so that it would also be in nominative. A similar
explanation could be given for the enigmatic alos ∞arnos of E.Me 45.
For convenience, I will assume that all these examples can be inter-
preted as s-ending forms. In fact, this will not affect the ideas to be
formulated about the value of the s-ending to any great degree, since
with the exception of kbos and alos ∞arnos, the context in which we find
the rest of the examples is similar to that of the clearest ones (ntokris,
[q ? ]lalis): as the first word of an onomastic formula—which implies that
they represent the name of the deceased—followed by one or more
personal names in genitive.
By analyzing the possibility that the s-ending recognizable in some
words could be a mark of a dative case, we can first of all clearly sep-
arate the formal aspects from the functional ones. Formally, we can
agree with the hypothesis formulated by Melchert (2002:309) that this
s-ending comes from a genitive ending PIE *-e/oso (like Lycian -a/-ehe,
cf. Adiego 1994c). An alternative view—a Luwic possessive suffix with-
out i-mutation *-asso —cannot be ruled out, although Melchert (ibid.)
is right in pointing out that there is no evidence to support the pro-
posal of an adjectival character for -s in Carian. In any case, both
solutions are equally attractive from a phonological point of view, as
both offer a straightforward explanation for the presence of a different

5
In the case of idmns, the name idmuon-≤ could favour an analysis as idmn- if both
forms belonged to the same paradigm, but the differences between idmn- and idmuon-
do not allow us to confim this hypothesis.
316 CHAPTER EIGHT

sibilant sound in this -s-ending (< *-e/oso or *-asso-) vs. the Carian gen-
itives in -≤ (< *-assì, see above).
If we accept the above arguments, the problem for Carian datives in
-s can be reduced to a purely functional one, the question being whether
there is evidence for a functional displacement of these old genitives
to the expression of a dative value. As Melchert points out, such a dis-
placement would not be particularly problematic because it is well doc-
umented in other languages. However, in my opinion, the evidence
currently at our disposal is not at all convincing.
The most radical version of this hypothesis—that -s has become exclu-
sively the mark for dative, contrasting with -≤ specialized as genitive/pos-
sessive—is disproved, as Melchert admits, by the formula i[—]inis=d=rual,
where -s cannot be anything other than a genitive or possessive gov-
erned by rual (= Gr. §p‹ dhmio[u]rgoË ÑIpposy°nouw). Moreover, the ono-
mastic formulae in the same inscription nik[—]lan lysiklas[-?], lys[ikl]an
lysikratas[-?] suggest that -s should be interpreted as a genitive or
possessive.
If one accepts the existence of genitives/possessives in -s-, the claim
of an s-dative must be substantiated by unequivocal evidence, and this
is hard to find, since all the examples of alleged datives can also be
interpreted as genitives/possessives:
a) All the examples of possible datives in -s in the Memphis sub-
corpus can also easily be interpreted as genitives or possessives: if one
observes the texts quoted above, a translation ‘for X’ or ‘of X’ sounds
equally acceptable.
b) The same analysis can also be applied to other examples: an inter-
pretation of “arnajs sb taqbos as ‘of ”arnaj and of Taqbo’ is as good as
a dative interpretation ‘to ”arnaj and to Taqbo’. Vittmann (2001:52–54)
has made the case for a dative value based on the fact that the Carian
inscription appears integrated in an Egyptian formula, ‘give life’, which
seems to require a dative. He cites the parallel examples of certain
Egyptian-Phoenician bilingual inscriptions where the same Egyptian for-
mula ‘give life’ is used, and where the personal name is introduced in
Phoenician by the preposition l, ‘to, for’. However, while the Phoenician
parallel may seem to provide a strong argument for seeing the Carian
sequence as syntactically integrated in the Egyptian sentence, the oppo-
site may also be true, namely that the Carian and Egyptian parts are
more loosely related: the Egyptian sentence might simply be a stereotyped
formula, and the Carian phrase a mere indication of possession. Even
the Phoenician example accommodates a similar loose relationship
between both parts, insofar as l is usually employed in Phoenician for
 MORPHOLOGICAL FEATURES 317

introducing the target of a dedication. Regarding armotrqdosq in Hyllarima,

the context does not allow us to ascertain whether the word is a gen-
itive or a dative, but the character of the inscription, which presents
a list of priests, seems more consistent with an interpretation as geni-
tive (‘priests of the god Armotrqd-’) than as dative (although a reading
‘priests for the god Armotrqd-’ or a generic reference to the establish-
ment of a cult ‘to Armotrqd-’ cannot be ruled out).
c) The -s-dative hypothesis adds an unnecessary complication to the
interpretation of ntro/ntros in their respective inscriptions (see above
pp. 285–286). The dativum dedicationis with the god name ntro, ‘Apollo’,
a simple and elegant interpretation that fits well with the overall sense
of the text, becomes a nominative with the meaning ‘priest of Apollo’
and, conversely, ntros prãidas must then be an anonymous priest of
Apollo to whom the object is dedicated instead of a direct reference
to the divinity. But even accepting this explanation, there is nothing
to prevent us interpreting ntros prãidas as a possessive, insofar as the
overall structure of the inscription remains unclear.
d) Attributing a dative role to the forms in -s means that the form
trqude from Iasos remains unexplained. It is true, as Melchert remarks,
that the overall sense of the inscription is not clear, but if trqude is
simply the name of a god—as armotrqdosq in the new inscription from
Hyllarima seems to confirm—it is difficult both formally and semantically
to assign it a value other than dative. It cannot be an accusative (as
we would expect a final -n), a genitive (there is neither -≤ nor ‘possessive’
-s) or a nominative (the subject of the entire text seems to be a personal
name, perhaps ]are“, on which the patronym in the genitive siyklo-≤ is
dependent).
To summarize, we can draw two conclusions: (1) there is evidence
for the retention of the original genitive/possessive value of -s, at least
in some clear examples, and (2) there is no conclusive evidence to
affirm that -s has adopted a dative value.

5. Other Possible Datives

Note that my opinion, which does little to support the existence of
s-datives, is accompanied by an attempt to recognize other words as
datives. In this book at least three possible candidates have been men-
tioned: in the previous pages I have referred to trqude and ntro. We can
add also rtmi, but it depends on the analysis of the inscription C.Tr 2,
which remains very controversial. From an etymological point of view,
ntro could be explained as a dative in -o < *-à comparable with the
318 CHAPTER EIGHT

(admittedly scarce) examples of datives in -a for a-stems in Lycian (see

Melchert DLL:x; note for instance Urebillaha). As for trqude, we can
envisage a PIE dative singular ending *-ei, as can be expected in an
nt-stem such as trqud- < *tºHnt-. Finally, rtmi could be a dative in -i of
an i-stem (*artimi-?, cf. rtim nominative in C.Hy 1a).

6. Nominative Plural
No clear examples of nominative plural can be cited at present. We
can only mention two possible forms:
• in C.Hy 1, and also in C.My 1, it has been proposed that mol“ could
be a common plural nominative with the meaning ‘priests’, although
other explanations are possible. If this hypothesis were true, a suit-
able interpretation of mol“ could be mol-“, with an “-ending formally
identical to that of the common plural accusative -“ (see below), a
trait that Carian would share with Milyan (Nom. pl. -z / Ac. pl.
-z) against Lycian (-Ø / -s). However, -“ could in this case be a
derivative suffix, so that mol“ ought to be interpreted as mol“-Ø, lead-
ing to the converse situation: a nom. pl. in -Ø nearer to Lycian
than to Milyan.
• in the funerary inscriptions of Kaunos and Krya, the possibility was
envisaged that sdis could be a nominative plural corresponding to
sidi (see p. 291). However, this interpretation is difficult to reconcile
with either of the two above hypotheses for mol“. We could attribute
the discrepancies to a differentiation between a Kaunian dialect and
a purely Carian one, but at this point it becomes a merely ad hoc
solution.

7. Accusative Plural
The Kaunos bilingual offers good evidence of -“ as a plural accusative
ending for the common gender: sarni“, (un)do[—]tl“, kbdyn“, b2o[—]ol“
make reference to the two Athenian citizens honoured in the decree,
who are mentioned in accusative. The simplest etymological explanation
of -“ as an accusative plural mark can be found in Lycian -s and Milyan
-z, both from Luwic*-ns (< PIE *-ns). The appearance of the (presumably)
palatoalveolar sound “ instead of the generic, unmarked dental sibilant
s recalls the Milyan use of z instead of s, irrespective of the exact nature
of z in Milyan. In both languages, the use of a sibilant other than s
can be interpreted as the outcome of the original cluster *-ns.
 MORPHOLOGICAL FEATURES 319

8. Other Possible Case Endings

1) In Adiego (2002) it was argued that some words ending in -t were
plural genitives. The basis of this hypothesis is the interpretation of mol“
msot ylarmit as ‘priests of Hyllarimean gods’, roughly corresponding to
the Greek text ‘priests of all the gods’ (see pp. 306–307). The hypoth-
esis was developed from the assumption that the Carian letter O was
the Kaunian variant for the sign t, and consequently some words end-
ing in O = -t were also interpreted as genitives.
If this hypothesis is accurate, we still do not know what the origin
of this genitive plural ending might be. Given that -t probably repre-
sents a dental-palatal affricate /t“/ or similar, a direct derivation from
PIE -om (> Old Hittite -an, Lycian -ẽ ) is impossible, which opens the
way for analogical process. In this way, the palatal character of -t could
be somehow related to the palatal ending -≤ of the genitive singular.
Assuming a plural nominative *-ns (see above), an analogical propor-
tion Nom. sg. *-s (>-Ø)—Genitive sg. *-si-s (>-≤ ): Nom. pl. *-ns (-“ ?)
→ Gen. pl. *-ns-is (>-) could be proposed (with a phonological process
*-nsis > *-nt≤ > -t).
2) The contrast between alos ∞arnos (E.Me 45) and alos-d ∞arnos-d
(C.xx 2) leads me to think that -d could be a true case ending in the
latter example. The (very hypothetical) interpretation of C.xx 2 out-
lined in Adiego (2000:153–155; see here p. 284) pointed implicitly to
an analysis of alos-d ∞arnos-d as ablative (‘from Halikarnassos’). The end-
ing could then be compared to Lycian -adi/edi CLuw. -ati, HLuw.
-ari-/ati, but this comparison is complicated by the fact that in most
cases of the origin or Greek transcription of Carian d that can be con-
trolled, this letter reflects an original -nd- group, which would not be
the case of -d from *-Vdi (see p. 247).
Other analyses are also possible: one could connect -d with the likely
preposition d (cf. Lycian ñte) in C.Ka 5 d=rual (see p. 254). However,
one ought to assume then that d became a sort of case ending, the
only possible explanation for the iteration alos-d ∞arnos-d.

B. PRONOMINAL INFLECTION

We can identify in Carian at least two demonstrative pronominal stems:

s(a)- and a-.
The only clear indication of a pronominal paradigm can be stated for
the pronoun sa, on the basis of sa E.Me 27, san (G 1), snn (C.Ha 1,
320 CHAPTER EIGHT

C.xx 1): sa and san seem to be singular nominatives according to the

context in which they appear, whereas snn is undoubtedly a singular
accusative. The difficulty is to establish the relation between the two
nominatives sa and san, for which there are two possibilities: the difference
between both forms could consist of a gender distinction (sa common
gender vs. san neuter gender, reflecting -Ø vs. -n respectively), or alter-
natively that san is a kind of extended form (this was the explanation
given in Melchert 1993:79–80) independent of the possible gender dis-
tinction. This latter hypothesis is supported by the accusative form snn
that seems to imply a s(a)n- stem to which the accusative ending -n
has been attached. However, it is also possible that in snn, both n actu-
ally constitute an accusative mark,6 thus the first interpretation can be
maintained. In this sense, there is a striking parallel between Lycian
inflection and the pronoun ebe-:
Nom sg. c. ebe-Ø cf. Carian sa-Ø < *-s
Ac sg c. ebẽñnẽ cf. Carian snn “long” forms of original *-n
Nom-Ac sg. nt. ebẽ cf. Carian san < *-n
However, perhaps this parallelism is merely an illusion: in the case of
the pronoun a-, we find ann in C.Ka 3, where it can hardly be a com-
mon accusative parallel to snn, ebẽñnẽ, since it appears in a context with-
out a verb. It might instead be interpreted as a (neuter?) nominative.
The interpretation of an in an sidi is not clear either, because it is not
certain whether in this inscription it represents an accusative or a nom-
inative. In the latter case, interpreting it as neuter seems inconsistent
with the proposed analysis of sdis as as common plurals (cf. supra). In
summary, the current inventory of pronominal forms is too incomplete
and their interpretations too ambiguous for even a minimally coherent
paradigm to be traced.
Besides these demonstrative pronouns, we can also refer to the (at
least originally) relative pronoun ∞i, whose etymology is unproblematic:
it comes from PIE, PA *k wis (cf. Hitt, CLuw. kui“, Lyc. ti, Mil. ki ). As
seen in Chapter 7, this pronoun became in Carian a type of connect-
ing particle, in a similar way to the Old Persian haya/taya, although
some examples of its use can be still identified as remaining very close
to its original relative value.

6
Cf. Neumann (apud Melchert 1993:80, n. 5), who compared snn with Lycian ebẽñnẽ.
 MORPHOLOGICAL FEATURES 321

C. VERBAL INFLECTION

The difficulties of identifying verbal forms were outlined in our analysis

of the Carian inscriptions. The only word I consider to be even a min-
imally reliable verbal form is ÿbt, for which Melchert suggested a very
plausible interpretation (= Lyc. ubete ‘he offered’). The search for other
possible verbal forms can take one of two directions, which although
not totally incompatible, are in some cases difficult to reconcile. On the
one hand, recognizing an ending -t, comparable to Lyc. -te (or perhaps
also -ti if the verb was a present, not a preterite), opens the search for
other forms also ending in -t that can be compared with these singular
endings or with other endings that in Lycian are also characterized by
a -t (for example, the third plural present and preterit endings -ti/-te
after diphthongs, such as ºai-ti, ºei-ti, ºai-te / ºei-te). In such cases, the
resemblance of the sound context (t after a vowel) regarding the sin-
gular ending could make it likely that the outcome in Carian is also t
(although this is an optimistic supposition, and disregards possible ana-
logical alterations). The results of such a search are however very limited:
in our analysis of Carian inscriptions (Chapter 7), only two forms have
appeared that, based on their context, can be interpreted as verbs with
these type of endings: (1) not in C.xx 2 (connected there to Hitt.
nà(i)-, CLuw. nana- < * PIE *neyH-, so that a meaning ‘he brings/he
brought’—3rd sg. present or preterite—could be suggested); (2) ait in
C.Ka 2, l. 8, where it appears in a sequence qrdsol“ ait: ait has been
tentatively connected to Lycian aitẽ, a 3rd pl. preterite form of the verb
a- ‘to made’ (see p. 301). To these two forms one can also provision-
ally add the word aitusi in C.Ka 5, if segmented ait + usi, although
problems arise when trying to establish the meaning of usi. An alter-
native view—to segment aitu + si—is an attractive solution as it yields
a possible imperative form (cf. -tu in Lycian), but it fails to analyse the
word si.
Other possible forms ending in -t could allow a similar interpretation.
A good example is 1aitk, also in C.Ka 2, if segmented 1ait=k), but the
verbal stem cannot be established as we are unsure of the phonologi-
cal value of the initial letter 1.
On the other hand, scholars have searched for verbal forms by means
of an internal and combinatory analysis of the Carian texts. In this sense,
there has been a certain consensus in regarding a sequence repeated
in different inscriptions as a possible verbal form: the sequence mdane
(Thebes mlane). The fact that this interpretation is based on internal
322 CHAPTER EIGHT

and combinatory grounds explains why it preceded the definitive deci-

pherment of Carian: it was ”evoro“kin who suggested that mdane, mlane
was a verb (see for instance ”evoro“kin 1977:124). Following the deci-
pherment, authors such as Hajnal, Melchert and Schürr have followed
this line of argument, and in the case of these possible verbal forms
they have also added some etymological connections to the functional
basis of the hypothesis. The most recent and comprehensive treatment
of the subject is an unpublished paper by Melchert (Melchert, mdane).7
The interpretation of mdane, mlane as a verb is supported above all
by evidence from three brief inscriptions found on objects, where mdane
appears to be the one verb of a sentence. By contrast, most of the
examples of mlane, from Thebes, are contextually far less clear. All the
examples are given below:
mdane:
“arkbiom : zidks mdane : ÿn-?-mo | den : tumn (E.Sa 1)
ntros : prãidas / or“a / nu mdane : uksi wrm≤ (E.xx 7)
ÿ≤ biks not : alosd ∞arnosd : jzpe mdane (C.xx 2)
mlane:
a?q≤baq ewm ≤emot / qtblo owdown[. . .]mwarudk≤o mlane (E.Th 10)
?-˚bjqmq ew mlane qeb≤t | u[. . .]ü≤q | qwsal | mqabaewleqo“oski.oms
(E.Th 12)
lÿ∞se | “i“≤ | mlan[-?] (E.Th 35)
dquq | ewmlane | tebot | gkem≤ (E.Th 44)
w. dbo≤kn ewá¬å»e ˚[ (E.Th 47)
(cf. also lane in: balewlane | “rbk˚[-]sal | (E.Th 49).
Of the three examples of mdane, the last is of particular importance,
since there is a possible conflict with the word not, which I have proposed
to be the verb of the sentence.8 In the other two instances, it must be
recognized that there is a reasonable chance of mdane being a verb,
although there is no reason not to look for a verb in some of the other
words that appear in both inscriptions, and that remain uninterpreted.
If mdane is a verb, from the inscriptions in which it appears we could
take it to mean, ‘to vow, to offer’ or similar (cf. Melchert, mdane).
As stated above, Theban graffiti with the variant form mlane are
much more obscure. Only E.Th 44, and perhaps also E.Th 35—but

7
I thank H. C. Melchert for allowing me to use this paper. Although I disagree
with Melchert’s views on mdane and related forms, I consider his usage and analysis
of the forms superior to Hajnal’s (see Hajnal 1997b:151–157), who arbitrarily sepa-
rates mlane from mdane and does not include mda forms in his study.
8
Indeed, there is no reason why the inscription should not contain two different
verbs, not and mdane
 MORPHOLOGICAL FEATURES 323

note the lack of a final letter—offer apparently complete and brief texts
in which mlane could be a verb. In the remaining examples, it appears
after a sequence ew, very close to the form ewm, which is also typical
in the Thebes corpus. This creates the possibility of a different seg-
mentation of ewmlane: ewm + lane, instead of ew + mlane. The example
E.Th 49 introduces even more confusion: the sequence ewlane, analyzed
as ew + lane, supports both ew and lane as isolable words in ewmlane. A
possible compromise, and perhaps the correct solution, would be to
isolate three elements: ew + m + lane. This would allow different com-
binations: ew + m, ew + lane, m + lane and ew + m + lane.9
The difficulties of analysing mdane as a verb begin when other, less
favourable factors become involved: besides mdane, the sequence mda
appears repeatedly in C.Si 2:
. . . eri : pisñoi mda : pñmnn≤ñ : pda/∞m≤uñ ∞i “aoyr∞ri mt∞elã / ñmailo
mda lrHñ : stspñ vacat / sm“s[-5-] sb añmsñsi mda (C.Si 2a)10
Moreover, in C.Ha 1, a sequence md can be easily segmented:11
smdÿbrs | psnlo | md orkn tÿn | snn
As I have argued (see Adiego 1994a:54–55, 2000:140), this mda/md
seems almost inseparable from mdane, a theory also expounded by
Melchert in his unpublished paper. In fact, he interprets md- to be the
lexical root of all these alleged verbal forms. Melchert even extends
the presence of this root to the form mln in C.Ka 5, unifying md, mda,
mdane, mln under a common root *mVld- (by assuming a *-ld- > -nd-
process in the forms with d ), which he compares with Hitt. mald-12
‘vow, solemnly pronounce’, but also ‘dedicate, give’. As for the different
endings of the md-family of words, Melchert (mdane) offers the following
analysis: mln as a preterite third plural from *mVld-onto;13 mda as a pre-
sent third singular (used as a preterite) with an ending parallel to Hittite
-ài in maldài; -ne would be an enclitic object pronoun (following Schürr
1996a:66).

9
Given the difficulties regarding the segmentaton of the (possible) elements that
make up ewmlane, I enter this in the Glossary (Chapter 11) as a complete word.
10
Cf. also C.Hy 1a dymda, but in this latter case I am not sure of the segmenta-
tion dy mda.
11
ºmd º also appears as a sequence in other inscriptions (C.My 1 (?), C.Si 2, C.Ki
1: note also in the first word of C.Ha 1), but the contexts are too obscure (C.My 1,
C.Ki 1) or simply too unlikely (C.Si 2, C.Ha 1) to allow us to isolate md as a word.
12
As Melchert recalls, the connection mln = Hitt. mald- was already suggested in
Hajnal (1997:152).
13
Without discounting the possible interpretation as a present (< *mVld-enti).
324 CHAPTER EIGHT

In Adiego (2000:140), I argued against this interpretation of

md/mda/mdane/mlane as a verb, and I believe that the objections for-
mulated in that study are still valid: it would be very unusual not only
that the same verb was used in very different classes of texts (dedica-
tory inscriptions on objects, visitors’ graffiti, a public decree, the heading
of a list of priests . . .) but also that the same verb was used at least three
times in the same inscription (the decree from Sinuri, C.Si 1). Thus if
we were to accept this interpretation we would be obliged to acknowl-
edge both an excessively polysemic or general meaning for md-, and
an unrealistic lack of lexical variety in Carian. For this reason, I choose
to recognise in md/mda/mdane/mlane a functional word or a chain of
functional words. If one wants to retain the interpretation of md- as a
verb, the only remaining possibility would be to see it as an auxiliary,
lacking any lexical meaning (cf. Adiego, ibid.). I am however more
inclined to categorise md/mda/mdane as a chain of particles; in Adiego
(1994:54–55) I already introduced the idea of analyzing mdane as a
chain of introductory particles followed by two enclitic pronouns, -n-
(accusative) and -e- (dative), although I now believe that an analysis of
-ne as a unique enclitic pronoun, as suggested by Schürr and Melchert,
is perhaps more plausible. In any case, md would remain as particle
chain m-d—documented tel quel in C.Ha 1—where one could tentatively
imagine a connection with the Lycian introducing particle me for initial
m-. It is true, nevertheless, that this comparison is a mere desideratum,
since the behaviour of Lycian me vs. Carian m- would be totally different
in terms of the order of constituents: whereas Lycian me always appears
in the first or second position in a clause,14 Carian m- can seemingly
be placed anywhere.15
Besides mdane (and variants), other forms have been considered pos-
sible verbs on the basis of the context in which they appear. Perhaps
the most likely of these to in fact be a verb is uiomln, documented in
the Kaunos bilingual inscription (C.Ka 5), and possibly also in C.Ka
2 [ui?]omlã and C.Ka yomln, if these forms are interpreted as pure
variants of uiomln. It is possible that this word is the Carian form cor-
responding to the verb ¶doje in the Greek part of the bilingual, although
this does not necessarily mean that the Carian version had a literal

14
See Melchert (2004), s. v. me.
15
The interpretation of mdane as a chain m+da+ne, where m stands as a particle, is
consistent with the tentative analysis of the Theban form ewmlane as ew+m+lane (analy-
sis based on the independent existence of ewlane and ewm, see above p. 323).
 MORPHOLOGICAL FEATURES 325

correspondence to the Greek one (indeed, other possible analyses of

uiomln have been envisaged, see Chapter 11 s. v.). Of the different
attempts to explain uiomln, perhaps the most convincing is the pro-
posal made by Hajnal and Melchert to isolate a verb, mln, a preterite
third plural of a stem, ml- < *mVld-, so that mln could come from
*mVld-onto (cf. Melchert, mdane).
One of the advantages of this analysis is that it can be applied to
another of the words to have been identified as a possible verb (Adiego
2000:142, see here p. 304): pisñ (in the sequence pisñoi mda) in C.Si 1a.
In Adiego (2000), pisñ was also interpreted as a preterite third plural
with the meaning ‘they gave’, formed from verbal root pi -, equivalent
to Hitt., Luw. piya-, Lyc. pije- ‘to give’, followed by an iterative suffix
-s- and a personal ending -ñ. The form would therefore be almost par-
allel to the Lycian, tusñti, a present third plural of an iterative stem tu-
s-, created from tuwe-. Both the function and the meaning proposed
for pisñ are consistent with the context: the form appears following the
onomastic formula of the Hekatomnids Idrieus and Ada at the beginning
of a decree, so a suitable translation might be ‘Idrieus (son) of Hekatomnos
and Ada (daughter) of Hekatomnos gave . . .”.16 This would mean that
pisñ could contain a preterite third plural ending -ñ from *-onto, paral-
lel to -n in mln.
Other possible verbal forms (ñmail in C.Si 1a, cf. Adiego 2000:142;
nu in E.xx 7) have been suggested on very hypothetical grounds, and
they will not be discussed here.
To summarise, the results of the search for verbal forms in Carian
are currently very uncertain. We can establish, with the utmost cau-
tion, only three possible endings: -t as a preterite (or present) third sin-
gular ending, and -t and -n/ñ as preterite (or present) third plural
endings (the first after the diphthong -ai, the second in other contexts,
similar to those that in Lycian become -ñte/-ñti ).

16
In Carian: [—]ryin ∞tmño≤ : sb ada ∞tmño≤ eri : pisñoi mda . . . As for eri, it could be
a preverb or a noun (direct object of pisñ ?), see above p. 304.
 CHAPTER NINE

THE GENERAL VOCABULARY AND THE PROPER NAMES

A. GENERAL VOCABULARY

Our knowledge of Carian general vocabulary is unavoidably very lim-

ited; we have already seen (Chapter 2) that the indirect sources pro-
vide us with a very small number of glosses, and these in turn, despite
their seemingly reliable origins (they appear to date back to Greek
authors of Carian origin), remain uncertain insofar as none of them
has been identified in the Carian inscriptions up to now. As for the
direct documentation, the only texts that can be interpreted with any
degree of confidence are a number of brief inscriptions that contain
exclusively, or almost exclusively, onomastic formulae. It is true that
some of the longer texts must contain essentially Carian common words,
but they are in general impossible to analyse. In this necessarily brief
section, I will limit myself to collecting in a synthetic way the general
vocabulary that arises from an analysis of the inscriptions (see Chapter
7): further details about these forms and about others, which are inter-
preted more hypothetically, will be presented in the corresponding
entries of the Glossary (Chapter 11).
Given the typology of a great number of Carian inscriptions, it is
not surprising that one group of the common words whose meanings
we can establish belongs to the semantic field of kinship: we know the
Carian words for ‘son’, mno-, ‘father’ ted, and ‘mother’ en, although in
the case of these two latter terms, the interpretation is based solely on
the combination of an etymological connection with other Anatolian
dialects and the suitable, but unique, context in which they appear (E.Me
38 and E.Me 32, respectively). No other kinship terms have been iden-
tified directly in the Carian inscriptions, and we can only speculate that
the personal names quq-Gugow and ksbo-Xasbvw might also be the com-
mon words for ‘grandfather’ and ‘grandchild’, on the basis of the good
Lycian parallels xuga ‘grandfather’ and xahba ‘grandchild’.
Due to the funerary character of many Carian texts, a collection of
words has been compiled which refer to the funerary stele or, more
generically, to the tomb: upe/wpe/upa, ue, ≤jas/≤as, s(i)di. However, no
 THE GENERAL VOCABULARY AND THE PROPER NAMES 327

clear etymological connections can be established for any of these words

(see Chapter 11 ss. vv.) and it is impossible to specify the exact mean-
ing in each case. The only definite word to have been identified for
different objects is ork-, which appears on two bronze phiales in accusative
(orkn), hence it must mean directly ‘phiale’ or more generically ‘bowl,
receptacle’. Also secure is the identification of øtr- as the Carian word
for ‘person, self ’.
Apart from these few words, most other items of Carian general
vocabulary have been identified based on merely hypothetical analyses
of the texts, which are not necessarily accepted by all the scholars. For
example, I have argued for mso- as the Carian word for ‘god’. The
connection is formally suitable (cf. CLuw. mà“san(i)-, Lyc. maha(na)- and
particularly Mil. masa-), and has good parallels in Carian onomastics
(msnord-), but it relies on my own interpretation of C.Hy 1, where it is
found. This problem is particularly evident in the case of Carian verbs;
we have already seen (Chapter 8) how difficult it is to find verbal forms
in Carian, so practically all the possible examples of verbs are in fact
dependent on a concrete interpretation of certain Carian inscriptions,
which are in general very difficult to analyse. The clearest form, ÿbt
in C.xx 1, points to a verbal stem ÿb-, ‘to offer’, but this interpretation
is essentially based only on the good parallel ubete ‘offered’ in Lycian.
Similar situations arise in the case of other possible verbs, such as ait,
which would seem to indicate a stem a- ‘to make’ (= CLuw. à/àya-,
Lyc. a(i)-), or pisñ, compared (see above Chapter 8) to CLuw. piya-,
Lyc. pije- ‘to give’. Needless to say, in cases like this, there is a clear
danger of resorting to circular arguments: if we take the example of the
latter form, we can establish a meaning ‘to give’ for an alleged verbal
form pisñ only on the basis of an etymological connection, yet we then
believe it plausible to claim Carian has a verb pi- ‘to give’, which in
turn confirms a genetic relationship between Carian and the Anatolian
languages! Lacking more concrete evidence (bilingual texts, convincing
combinatory analysis and so on), these forms—usually hapax legomena
in uncertain contexts—must be used with great care, and must be
included in our analysis only as provisory and hypothetical possibilities.
On functional words (demonstrative and relative pronouns), I refer
to Chapter 8. To those forms we can add the clear identification of a
copulative conjunction: sb (to be connected to Mil. sebe, ‘and’).
328 CHAPTER NINE

B. PROPER NAMES

As stated repeatedly in this work, the area that yields the greatest vol-
ume of information is Carian onomastics: the sources, both direct and
indirect, provide a large number of proper names, currently the most
valuable information on Carian available.
Carian onomastics can be analysed in two different, but comple-
mentary and interrelated ways. Firstly, diverse stems and suffixes can
be identified through an internal analysis of the compounding and deriva-
tion mechanisms that clearly intervene in their formation. Secondly,
many of these names can be totally or partially compared to the rest of
Anatolian onomastics, both from the first and the second millennia B.C.
Given our scarce knowledge of Carian common vocabulary, the first
approach yields very limited results: we are only able to establish some
regular patterns, both in composition and derivation, from a purely
combinatory analysis of the onomastic materials, and the meaning of
the elements—lexical stems, suffixes—that we isolate cannot be established.
Yet despite its limitations, this method of analysis is not altogether
worthless, insofar as it allows us to identify some recurrent procedures
in the formation of proper names. A good example is the series of
stems that appear combined, giving rise to a number of very charac-
teristic compound proper names, as the table of the p. 330 is intended
to show. We find here a set of first elements (i)d-, par(a)-, pun-, “ar-,
etc. that in general never appear as independent words and that are
used alongside a set of second elements (quq, u≤ol, etc.), which are for
the most part also documented in a simple, non-compounded, form.
This implies a different functional nature of first vs. second elements,
a possibility that can be confirmed in at least those cases where an ety-
mological explanation is plausible (see below).
More difficult is to establish possible derivation procedures without
the use of comparative evidence. The segmentation of suffixes is not
always easy, and it is also difficult to specify in some cases whether we
are dealing with a suffix or a lexical stem in composition; for instance,
the useful list of suffixes, created by Blümel using only this approach,
and included in his corpus of Carian personal names (Blümel KarPN:
32–33), offers some elements that can confidently be considered (on
the basis of an etymological approach) lexical stems, and not suffixes
(-biw, -muhw).1 In any case, several suffixes can be well established from

1
It must be said, however, that Blümel, very cautiously, does not speak of ‘suffixes’:
 THE GENERAL VOCABULARY AND THE PROPER NAMES 329

directly and indirectly attested personal Carian names: -ol (Greek

-vllow/-vldow, -vlliw), -s(i)- (Greek -assiw), -om (Greek -vmow). As for
place names, the well known suffixes -(V)ssow and -(V)nda are abun-
dantly attested.
Comparison with the other Anatolian proper names offers more
promising results. As I explained in Chapter 2, the research on Anatolian
onomastics during the last century demonstrated not only the existence
of a clear linguistic unity to which Carian onomastics belongs, but also
the Anatolian Indo-European character of this linguistic unity. In other
words, an important part of this geographical and chronological ono-
mastic continuum can be confidently analysed etymologically, since a
lot of lexical stems, suffixes, and morphological procedures of the
Anatolian Indo-European family can be identified. Three excellent exam-
ples of this type of interpretation are the decisive book of Laroche on
Cuneiform Anatolian personal names (Laroche: LNH), the equally influ-
ential work of Houwink Ten Cate devoted to Lycian personal names
(Houwink Ten Cate 1961), and finally the contribution of Zgusta (Zgusta
1964b), which although cited less frequently, is nonetheless very valu-
able. Comparison of Carian proper names with the rest of Anatolian
onomastics therefore not only provides us with a greater capacity for
combinatorial analysis, but also allows us to establish the meaning of
many lexical and functional elements that intervene in their construction.
In the following pages, I will offer (in a non-exhaustive, but in my
opinion sufficiently representative way) examples of this latter approach
by collecting several lexical elements of Anatolian origin that can be
identified in the Carian name system, and which generally also appear
in the known inventories of proper names from the rest of Asia Minor.
This exposition is very similar to that presented in Adiego (1993a),
where Houwink Ten Cate (1961) was used as a model,2 but in this
case, the names directly attested in Carian inscriptions are also included.3

the list appears in a section entitled ‘Komposition und Wortbildung’, although true
suffixes appear mixed together with these lexical stems.
2
Names with no indication of origin must be taken as Carian.
3
The reasons for this inclusion are given in p. 15.
 Carian Compound Proper Names
330

ibrel- kbiom- -muhw quq- p/bik(a)rm- p/biks- wli/jat- -uassiw u≤ol/w≤ol ydiq-/yriq- Others
/kbjom- /yri∞-/-yd∞

Ø Imbarhldow kbjom- quq- pik(a)rm piks- wliat- u≤ol-/w≤ol-

Kebivmow Gugow /wljat Ussvllow
Uliatow,
Oliatow
Akta- Aktauassiw Aktaussvllow Aktadhmow
(i)d- dquq- dbikrm, dbiks- dw≤ol-, idu≤ol idyri∞- idmn(-s)
Idagugow dbkrm Idussvllow idmuon-
idyes-
par(a)- paraibrel- Par(a)ussvllow parÿd∞ paraeym-
paryri∞ parpeym
Paraudigow
Pan(a)- Panamuhw Panuassiw Panablhmiw
p(u)n- pnu≤ol-/punw≤ol pñnmnn-
CHAPTER NINE

-pnw≤ol-, pnu≤ol- Ponmoonnow?

Ponussvllow
“a- Saussvllow Saurigow
“aÿdiq-
“ayriq-
“(a)r- “arkbiom- “rquq- “arwljat Saruassiw “aru≤ol- “rwli-
Sarus(s)vllow
Others Ekamuhw, a[rb]ikarm kdu≤ol(-“)? prpwri∞?
Ejamuhw tdu≤ol- Senurigow
Xeramuhw psu≤ol-
Yussvllow
Karusvldow
Maussvllow
 THE GENERAL VOCABULARY AND THE PROPER NAMES 331

1. Theophores
A type of personal name very widespread in Anatolian onomastics is
the theophore: we find god names used directly as personal names,
dvandva compounds in which two god names are associated¸ compounds
consisting of a personal name and a lexical element, nouns derived by
suffixation from a god name, and so on (see the enlightening study by
Laroche LNH:282–295).

§ 1. 1 (Hitt., Luw.) Arma, Carian armo-

Arma is the Hittite and Luwian Moon-god (Laroche NDH: 80). This
god name is now also documented in Carian, in the dvandva form
armotrqd- from Hyllarima (C.Hy 1).4 Its use in the formation of proper
names is commonly seen in the languages of the second millennium as
well as those of the first millennium.
Carian PN Ermapiw < Arma + piya (§ 3.1) = Armapiya (Laroche LNH
nº 135).
Cf. Ermapiaw (Zgusta KPN § 355–20, Lycia), Armapiaw (KPN § 97–3,
Lycia, Cilicia), Armapia (fem., KPN § 97–4, South of Phrygia-Lycia).
a) Other instances (not in Caria) of Arma, either alone or in composition with
muwa-, nani-, etc.: Houwink Ten Cate (1961:132–134), In Hittite and Luwian:
Laroche (LNH:290)
b) The variant Erma- (besides Arma-, Arma-) of the Carian name and others,
is usually attributed to the analogical influence of the Greek divine name
ÑErm∞w (already Kretschmer 1896:361).
c) The Egyptian flavour of the name Ermapiw (Sayce 1887[92]:122) was con-
sidered by Masson as merely coincidental, because the name can be satis-
factorily analysed from an exclusive Minor Asian point of view (Masson
1959:167–170). However, Hornblower (1982:357, n. 35), though accepting
Masson’s views, wonders if in cases such as this a sort of homonymy could
influence the choice of the name.

§ 1. 2. Luw. Tar¢unt-, Lyc. Trqqas, Mil. Trqqiz, Carian trq(u)d- Cf. Hitt.
Tar¢u-
The Anatolian Storm god. The name has a good PIE etymology (*terh2-
‘to cross’, ‘to pass’: in Anatolian ‘to overcome’, *térh 2-u- from a
-u-present).

4
Although I prefer to interpret armotrqd- as a theonym representing a divine pare-
dra due to the context in which it appears, I do not completely rule out the possi-
bility that it could be a personal name.
332 CHAPTER NINE

The god name is directly attested in Carian: trqude C.Ia 3, ?/trqd/?

C.Ki 1, armo-trqd- C.Hy 1. At least in Iasos and Hyllarima, it is most
likely that we are dealing with the divine name itself.
Place name Tarkondar//a// deducible from the name of a syn-
geneia Tarkondare›w (Mylasa).
More dubious are Konodvrkond//a// possible place name, from a
name of phyle ≤ Konodvrkondevn in Mylasa, Otvrkond//a//, from a
name of a phyle ÉOtvrkonde›w.
a) Internal analysis of Tarkondar//a// is not clear: Tarkond+ar(a)-? (perhaps
better than Tarkon+dar(a), Adiego 1993a:28).
b) In Konodvrkond(a), Otvrkond(a), neither the internal structure nor the ori-
gin of v vocal are clear (the connection with Hitt. ¢anna- ‘grandmother’,
Adiego 1993a:28 is very hypothetical). For the second name, a prothetic
o- could be postulated (Adiego 1993a:28),

§ 1. 3. Lyc. natrº-, god name to be identified with Apollo

The existence of a god name natrº, equivalent to Apollo, can be deduced
from the well-known ‘translation’ ÉApollÒdotow of the indigenous name
Natrbbijẽmi in the Xanthos bilingual. The god name seems to appear
directly in Carian in the form ntro- (see Chapter 11, s. v.), and can
also be recognized also as intervening in the formation of several per-
sonal names from Greek sources. The clearest example is Neterbimow,
undoubtedly the same name as Lycian Natrbbijẽmi. Also of this family
are the names Nvtrassiw and, less clearly, Nutar.

§ 1. 4. CLuw. màssan(i)-, Lyc. maha(na)-, Mil. masa- ‘god’, Carian mso-

(?). Cf. also Sidetic masara = yeo›w
A clear isogloss shared by Luwian, Lycian, Milyan and Sidetic (the pre-
sumed ‘Luwic group’) in contrast to Hittite, Palaic and Lydian, is the
common word for ‘god’. While Hittite and Lydian have forms inherited
from PIE *dyew- (Hitt. “iu(n)-, secondarily thematic forms “iuna-, “iuni-;
Lyd. civ-), the Luwic dialects present a form masa-, mà““an(i)-/*masa(na)
(> Lyc. maha(na)-), masar(a/i)-, of unknown origin and with serious
formal problems (is masa- the original form, and ma““an(i)-, *masana-
massar(a/i)- secondary formations?).
Place name Masan≈rada, msnord-, possible ethnical name from this
city. This can be analysed as a compound name with a second ele-
ment equivalent to Luwian *(a)radu-, see below.
Place name Massvn//a// (name of a demos in Mylasa).
Massariw (Carian name of Dionysos), PN Massarabiw, seem to con-
tain an r-form of the stem, as in Sidetic. Massarabiw can be analysed
as *masara-bi- (-bi = -piya, see below § 3. 1).
 THE GENERAL VOCABULARY AND THE PROPER NAMES 333

2. Some Nominal Stems

§ 2. 1. CLuw. annara/i- ‘forceful, virile’, ànnari- ‘forcefulness, virility’,

etc. Cf. Hittite innarà-, ‘forceful’.
Perhaps this stem can be seen in the place names Naras//a// and
Naruandow5 and in the name (or title?) naria- (see Chapter 11. s. v.). Cf.
also Narbaw?
Cf. also the name Andarsvw, if the explanation of -nd(a)r- as a result
of *-nr-, given in p. 262, is accepted.

§ 2. 2. CLuw.(?) *(a)radu-
This element can be identified in two Arzawian names, Tarhundaradu
and Piyammaradu. The status of (a)radu- is unclear: it could be a god
name related to the name of the Luwian stag-god, Runt(iy)a- (a more
recent form of Kurunta-), but this connection is far from certain (cf.
Laroche LNH:282, n. 6). Schürr has claimed to identify this element
in the Carian place name Masanvrada, ethnic (?) msnord-, assuming
that the place name derives from a PN Masanvradow, the name of
the city’s founder according to the information given by Stephan of
Byzantium (Schürr 2002). Although the validity of this latter point must
be viewed cautiously (Masanvradow could be an artificial creation from
the place name), Schürr’s argument is in my opinion quite persuasive.
Schürr also proposes that the same element can be identified in the
Carian place name Parembvrda (analysed as Paremb-vrda, Schürr
2002:170), and that the possibility also exists that other Carian names
in Greek sources that include a sequence ardº, ordº belong to this fam-
ily of nouns (place name Ardur//a//, PN Ordomaw). Even the PN
Arduberow/ardybyr-, if analysed as Ard-uberow/ard-ybyr- (not Ar-duberow/ar-
dybyr-) on the basis of ybrs- (C.Hy 1), could contain the same stem.6

§ 2. 3. CLuw. arpa- ‘confusion, tumult, strife’, Lyc. erbbe- ‘strife, battle’7

The stem CLuw. arpa-, Lyc. erbbe- intervenes in the formation of proper
names, but it is only attested in those of the first millennium, as Houwink

5
Neumann (1998:185), by contrast, connects Naruandow to an alleged Hitt. naru-,
a type of plant, but this latter word is actually Akkadian (*narû, a type of malt, see
Tischler 2001, s. v.).
6
According to Melchert, “the element -arada may be analyzed as a derivative of
the word seen in Hittite ard- ‘companion’. A Luwic *arada- would mean ‘community’,
which seems appropriate for a placename. The further derived -aradu would again
mean ‘companion’ or similar” (pers. comm.).
7
Not ‘defeat’ (Adiego1993a:30), see Melchert DLL.
334 CHAPTER NINE

Ten Cate has pointed out (1961:147–148).

The basis for his association of -arba-/-arbh- with this Luwian word
is a passage in Milyan of the Xanthos stela: trqq[i]z : esetesi[k]e er[b]besi[k]e:
lusasi : esene[m]la (TL 44d 12–13). Houwink Ten Cate takes the same
view as Laroche, who associated erbbe with Luw. arpa-. As for esetesi[k]e,
CLuwian also offers a possible direct cognate: a““atta““i-, Gen. Adj. of
a““atti- documented as divine epitheton (dÀla“ a““atta““i“ ). It is clear there-
fore that er[b]besi[k]e and esetesi[k]e are genitive adjectives used as epi-
thets of god Trqq[i]z. The key to connecting erbbe-/arpa- with onomastic
forms is the Cilician proper name Trokoarbasiw (KPN § 1512–22)
where the same god name and one of his epithets appear together in
composition.
PN Arbhssiw, Arbhsiw. The same name in Cilicia: Arbasiw (Zgusta
KPN § 85–1).

§ 2. 4. HLuw. hasu- Lyc. xahba- ‘grandson’

PN ksbo, Greek Xasbvw. Both Carian and Lycian point to a secondary
-à stem (Carian -à > -o, cf. armo) from the u-stem attested in Hieroglyphic
Luwian (cf. Melchert DLL). Cf. also Kasbvlliw, apparently derived
from ksbo/Xasbvw.

§ 2. 5. CLuw. ¢àpa/i- ‘river’, *¢apài- ‘irrigate, water’

It possible that the sequence kb- (Greek Kebº, Kbº) that appears in some
Carian names (kbjom-Kebivmow and “arkbiom, kbdmu-, Kbondiassiw, Kbvdhw,
and the indigenous place names kbid- for Kaunos and kbo- for Keramos)
is the Carian result of this Luwic stem: at least from a phonological
point of view, the correspondences are appropriate, and in the case of
the two place names, a denomination ‘river’, ‘irrigated (land)’ or sim-
ilar is very suitable (cf. Schürr 2001a:64 for Keramos-kbo).

§ 2. 6. CLuw. ¢ù¢a-, Hitt. ¢u¢¢a-, Lyc. xuga-, Mil. xuga- ‘grandfather’

This word intervenes in the formation of the onomastics of the second
millennium: Huhanani (LNH nº 379), Huhhaziti (LNH nº 385), etc. It is
also documented alone in the Lycian name Kougaw (KPN § 717).
Carian: quq, Greek Gugow.
dquq, Greek Idagugow. The (first element, (i)d- Ida-, is well docu-
mented as a proper name, either alone (Ida KPN § 451–1, Eida KPN
§ 451–2, both in Lycia) or as the first element of compounding: Carian
idu≤ol-, du≤ol- = Greek Idussvllow, dbiks-, dbkrm-; Lyc. Eida-ssala,
Ida-zzala KPN § 451–10, cf. Salaw, Zzala KPN § 1358–1).
I now have little doubt that the name of the Lydian king GÊghw must
 THE GENERAL VOCABULARY AND THE PROPER NAMES 335

have the same origin. The problem posed by the phonetics (Lydian
does not conserve PIE laryngeal *h2, unlike the other Anatolian dialects)
can be overcome if we imagine the name to have a Carian origin. As
a result, the long discussion in Adiego (1993a:40–41) does not need to
be repeated here.

§ 2. 7. CLuw. im(ma)ra/i- ‘open country’

im(ma)ra/i- (only verified as common word under the form of Gen Adj
im(ma)rassa/i-) is the CLuw. word corresponding to Hitt. gimmara-. It
appears in the formation of a proper name already found in Cuneiform
sources: Immaraziti (Laroche LNH nº 450). In the onomastics of the
first millennium it adopts the forms imbr-, imbar-, etc., id est, with the
development of an epenthetic labial. A possible form without labial
could be the place name Imrougara (KON § 374, Eastern Phrygia; cf.
Zgusta KON:198–199). In Lycian direct documentation, the word seems
to appear as ipre. With this in mind, the proper name Ipresidi is very
interesting, as it could correspond exactly to Luw. Immaraziti.
Most instances of proper names containing this element from the
documentation of the first millennium come from Lycia and Caria.
Place name ÖImbrow (promontory near Kaunos), from the pure stem.
PN para-ibrel, Imbarhldow. This corresponds formally to CLuw.
im(ma)ralla-, adj. attested fragmentarily (im-ra-a[l ) and as a place-name
(URUImralla), see Melchert CLL s. v.)
PN ibrsi-/ibarsi-/brsi-, Imbrasiw, Imbrassiw, Imbarsiw. It corresponds
to the CLuw. Gen. Adj. im(ma)ra““a/i- or to a further derivation of this
word (*im(ma)ra““iya-).
ÖImbramow, var. ÖImbrasow, Carian name of Hermes, Steph. Byz. s. v.
ÖImbrow. ÖImbrasow is also the name of a river in Samos.
Schürr’s objections8 to connecting these forms with Luw. im(ma)ra-
are in my opinion not overly convincing.

§ 2. 8. Lyc. mere- ‘law’, maraza- ‘judge, arbitrator’

Panamara. Cf. also PN Mareuw, Marow.

§ 2. 9. Hitt. muwa-, Luw. *mùwa-, Mil. muwa- ‘might, power’

This word, common to both Hittite and Luwian, is one of the most
characteristic terms in the formation of proper names from all periods.

8
Schürr (1991[93]:171), (2001b:104–105).
336 CHAPTER NINE

Before the discovery of Hittite and Luwian it had already been identified
by Kretschmer (1896:332–333). It was Friedrich (1931) who connected
the Anatolian names of the first millennium with the names in -muwa
of the second millennium, and who identified muwa- as a common word
in Cuneiform texts.
Carian: MÒtulow, mythic founder of the Carian city Samylia (Steph.
Byhz. s. v. Samul¤a, Zgusta KPN § 976: not in Blümel KarPN) =
Muwatalli-, a name well documented in Cuneiform sources (Laroche
LNH nº 837).9
With first element pan(a)-/pun(a)-: pñmnn-, Greek Ponmoonnow, Panam-
uhw, Panamuaw, an inhabitant of Kos (Zgusta KPN:695), place name
Pounomoua.
Pormounow is most likely a more recent form of pñmnn-, since it
appears as the name of a Syngeneia in the temple of Sinuri, where
pñmnn-/Ponmoonnow is documented in an earlier inscription.10 The sound
change required is commonplace (nm > rm by nasal dissimilation).
Other compounds with muwa- as a second element: uksmu-/wksmu-,
a Carian name corresponding to Ouajamoaw (Zgusta KPN § 1141–2,
Isauria, Cilicia), Ouajamvw (Zgusta KPN § 11141–2, ibid.). Perhaps also
kbdmu-.
muwa- as the first element in mwsat-, corresponding to Lydian Mousathw
(Zgusta KPN § 987a), Pisidian Moushta, Moshta (all equivalent to
CLuw. Muwaziti- Laroche, LNH nº 840?). Less clear is myze-, Mouzeaw.

§ 2. 10. Hitt. peda-, Lyc. pddẽn- ‘place’

Some Carian place-names show a stem Phda-, Pida-, Peda- that can
be connected with Hitt. peda-, Lyc. pddẽ and, ultimately, with a Proto-
Anatolian *pédon ‘place’, from PIE *pédom (> Greek p°don, etc.). The
stem also appears in the place names from Cuneiform sources, Petassa
and Pitassa (See Laroche TA1, nº 49).
Phdasa, Pidasa (pl. nt.) (§ 1054–1). According to Blümel KarON,
s. v., at least two places bore this name: Phdasa near Halikarnassos,
and Pidasa in Grion.
Pedanass//ow//. For the reading, see Blümel (KarON:177, n. 56),
not Pidº (Adiego 1993a:43)!

9
It has been suggested that MÒtulow not only corresponds to the name Muwattalli
but is in fact also the name of a Hittite king, transmitted by the Greek sources. The
question has been raised in the more general problem of the historicity of the Trojan
War (Motylos was a king who received Paris and Hellena during their flight to Troy).
10
The equivalence of the two forms was already suggested by Robert (1945:98).
 THE GENERAL VOCABULARY AND THE PROPER NAMES 337

Pidossus § 1058–1, an island near Halikarnassos.

Cf. Perhaps also the personal name Pedvldow.
In Adiego (2000), I proposed that the same stem could be seen in
C.Si 2 pda∞m≤uñ/?, see Chapter 11, s. v.

§ 2. 11. Luw. *pi¢a- ‘luminosity, splendour, might’

This stem appears in Luwian in this form only in personal names (See
Melchert CLL), but various derivatives are verified in common lexicon
( pi¢amma/i-, pi¢a““a/i-, pi¢atta/i-).11 The etymological interpretation is
clear: pi¢a- comes from PIE *bheh2- ‘shine, glow’ (OInd bhà́ti ‘shines’,
Greek fa¤nv ‘make visible’ etc.). The stem had already been identified
by Houwink Ten Cate (1961:156–157) in the Anatolian onomastic sys-
tem, particularly in the Luwian area (at this point, however, the mean-
ing of the root was unknown).
Carian:
piks-, dbiks-; Pigassvw.
pikre-/pikra- = Greek Pigrhw (var. Pikrhw, Pitrhw?). This is a frequently
documented name (either in this form or through derivates, as Pigrassiw,
§ 1255–5, Pamphilia), not only in Caria but also in other countries of
Asia Minor: Lycia, Pisidia, Pamphilia, etc. The only possible cognate
in Cuneiform sources is Pi¢irim, the name of a Cilician king, Laroche
LNH nº 977, but this is an isolated form (cf. Adiego 1993a: 36–37).
pikarm-/pikrm-, dbikrm-/dbkrm-, cf. the Lycian name from Greek sources
Pigramiw, Pigramow (Zgusta KPN § 1255–1/2).
Pijvdarow (internal structure not clear).
It is very doubtful that the two Carian place-names, Peigelasow and
Piginda, belong to this group.

§ 2. 12. CLuw. pùna- ‘all, totality’, cf. also CLuw. pùnata/i- ‘all’, Lyc.
punãma- ‘totality’
In Anatolian onomastics from all periods, two elements, puna- and pana-
are repeatedly found. Laroche (LNH:32) considered them two variants
of a single form, the meaning of which was unknown at the time.
Nowadays, a quantifier meaning ‘all, totality’ or similar seems to be
well established for Luwian pùna- (and derivative pùnata-), and for Lycian

11
The last two only in the form of GenAdj, with -a““a/i- suffix attached to them
(Melchert CLL, s. v.).
338 CHAPTER NINE

punãma-. The same meaning can be determined for the onomastic com-
ponent. Whether pana- was actually a variant of puna- is less clear (pana-
does not appear as a common lexical item in Luwian). If this is the
case, the alternation must be very old, because both puna- and pana-
are found in the Cappadocian onomastics, from the beginning of the
second millennium.

Pùna- in Carian:
PN pñ-mnn- = Pon-moonnow, Por-mounow.
Place name Pouno-moua
For these forms, and particularly for Pormounow, see above muwa-.
pnu≤ol-/pnw≤ol-/pnw≤ol- Pon-ussvllow.
More doubtful: pnyri≤ru-.
It is possible, although very hypothetical, that the common noun also
appears in Carian direct documentation in the form punot (see Chapter
11 s. v.).

Pana- in Carian:
Pana-blhmiw. For the second element, cf. Lyc. -plemiw, -plemi, -pl m̃mi
in Sedeplemiw, Sedepl m̃mi, Esedeplẽmi (KPN § 1387–1/3).
Pana-muhw. For the second element, see above § 2. 9.
Pan-uassiw. For the second element, cf. Akta-uassiw, Sar-uassiw.
Place name Pana-mara (for -mara, see above § 2. 8).

§ 2. 13. CLuw. ura- ‘great’, HLuw. ura/i-, cf. Hitt. ura/i- ‘great’
This adjective appears in the Anatolian onomastics of both the second
and the first millennia (cf. Houwink Ten Cate 1961:164–165): Laroche
LNH: nº 774: Massanaura, nº 872 Nattaura, n 1431 Ura, 1437 Urawalkui,
etc. Oraw (Lycian; Zgusta KPN § 1100–1), Ouramoutaw (Cilician; Zgusta
KPN § 1169), etc.
In Carian it can be recognised in the personal names urom-, urm-,
wrm-, and in the place name Urvmow (converted to Eurvmow, Eurvpow
by the influence of Greek). However, the etymological connection of
this latter form to ura- is challenged by the existence of variants such
as Kuròmew and huròmew (see Blümel KON s. v. Urvmow) because the
initial k-/h- would then remain unexplained.

§ 2. 14. Hitt. *walli-, walliwalli- ‘stark, mighty’ (Cf. CLuw. wallant- ‘fit,
capable’)?
Some Carian names point to a stem that could be reconstructed as
*wala/i-:
 THE GENERAL VOCABULARY AND THE PROPER NAMES 339

wliat/wljat, Greek Uliatow, Oliatow (documented directly in Carian

as wliat/wljat)
Oaloalow (about this form, see Adiego 1993b).
At least in the second case, a direct connection with the redupli-
cated stem Hitt. walliwalli- seems very likely (< PIE *äuelH- ‘to be stark’).
For Uliatow, however, there is also a possible relationship with the
family of Hittite walli-, ‘glory’, CLuw. walli(ya)-, HLuw. wa/iliya-, ‘to
exalt’ (see Melchert CLL s. v.), from PIE *äuelh1– ‘to choose, to select’
(if we are not actually dealing with the same root).

3. Verbal Stems

§ 3. 1. Hitt. pài-/piya-, Luw. piya-, Lyc. pije-, Pal. pi(sa)-, Lid. bid- [pid-]
‘to give’
All the Anatolian languages in which the verb ‘to give’ is found con-
tain a similar form that has a clear common Proto-Anatolian origin.
Its use in onomastics is also common to all Anatolian linguistics and
very productive, above all in the formation of theophores: Tarhundapiya
(LNH nº 1267) = Tarkumbiou (gen.; KPN § 1512–13, Cilicia).
Ermapiw, cf. above arma-
Massarabiw cf. above massan(ì)-
As well as the simple form of the stem, we also find in Anatolian
onomastics forms that can be clearly denoted ‘Luwic’ participles in
*-mo/i- (cf. CLuw. -mma/i-, Lycian -me/i-): Laroche LNH: nº 980 *Piyama-
d
KAL, nº 981 Piyamaradu, Lycian Armadapimiw, Pisidan Kouadapemiw, etc.
In Carian: Neter-bimow = Lyc. Natr-bbijẽmi = ÉApollÒdotow (in the trilin-
gual inscription of Xanthos).

4. Adverbs
As previously mentioned, Carian contains some characteristic compound
names that were systematized in the table above (p. 330) only on com-
binatory grounds. A comparative approach serves to confirm that, at
least in two cases ( par(a)-, “ar-), the first elements of these compound
names can easily be connected to adverbial stems in other Anatolian
Languages. The use of adverbs as the first elements of compounds is
also verified by the rest of the Anatolian proper names, particularly in
Lycian (see Houwink Ten Cate 1961:172–175 and the forms cited
below).
340 CHAPTER NINE

§ 4. 1. Hitt. parà, CLuw. parì, HLuw. para/i ‘forth, away’, Lyc. pri
‘forth; in front’.
This adverb appears in Carian as para-/par-, Greek Para-/Par-:
para-eym, para-ibrel, par-ÿd∞- (= Greek Para-udigow), paryri∞-.
Para-ussvllow, Par-ussvldow. Probably also in the place name
Paremborda (Neumann 1988:191, cf. also above p. 333).

§ 4. 2. Hitt. “ara, “èr, Luw. “arra, “arri, Lyc. hri

“ar-/“r-, Greek Sar- in Carian:
“ar-u≤ol-, Greek Sar-us(s)vllow
“r-quq, “r-wli —
Sar-uassiw. For the second element, cf. Pan-uassiw.
The following forms might perhaps be considered as a variant of
Sar-: “a-, Greek Sa-:
“a-yriq- (= Greek Sa-urigow), Sa-usvllow
Note the parallel use of hri- in the Lycian personal name Hrixttbili
(besides simple Ktibilaw, cf. Melchert DLL, s. v. Hrixttbili ).

5. Lallnamen
As mentioned in p. 13, Lallnamen—hypocoristic names whose structure
recalls the language of children, characterised by open syllables, with
or without different patterns of reduplication—were identified by
Kretschmer (1896) as characteristic of the Anatolian onomastic system.
Hittite and Luwian evidence confirms the antiquity of such formations
in Anatolian (see the exhaustive analysis in Laroche LNH:239–246).
Carian is no exception: we can identify a considerable number of names
that can be interpreted this way, although mostly in Greek sources,
since only one Lallname, ada, is directly attested. I refer here to Appendix
C, where it will not be difficult for the reader to identify them in the
list of Carian names from indirect sources. Many of these Lallnamen seem
to obey simple structures of the type aC(C)a (Aba, Abaw Abbaw ada-Ada,
Adaw Appa) and reduplicated CaC(C)a (Nana, Nanaw, Papaw, Tata, Tataw,
perhaps also forms in -h, -o- and -v if these stem vowels are secondary:
Nanh, Nannh, Nannow, Nannow, Nannv), while others show somewhat
different patterns (Minnaw, Nonnow) or are derived from suffixes of un-
known value (Amiaw, Ammiaow, Nannixow, Papiaw, Tatarion, etc.).
 THE GENERAL VOCABULARY AND THE PROPER NAMES 341

6. Suffixes
Extensive evidence exists of the use of different suffixes in the forma-
tion of Anatolian proper names, but as Laroche (LNH:327–328) rightly
pointed out, their analysis is hampered by diverse factors: the impos-
sibility of explaining their actual origins and values, the risk of incor-
rect segmentations (“les dangers du découpage formel” in Laroche’s
words),12 and the great variety of stems to which they can be attached,
which prevents us from identifying the derivational mechanisms behind
the construction of such names. These problems are compounded in
the case of the indirect documentation, insofar as the Greek adapta-
tions may be masking a more complex situation. A good example of
this is provided by sigmatic suffixes; taking into account only Greek
evidence, one might be tempted to consider a single suffix -ssi- both
for Arlissiw and for Imbrassiw. Direct evidence, however, available for
these forms but not for others, indicates two different Carian suffixes,
-“- for arli“-Arlissiw and -si- for (i)b(a)rsi-Imbrassiw. Note the parallel
situation in Lycian, where -si- in Mollisiw and Triendasiw from indi-
rect sources, treated as a single suffix (-zi [sic]) in Houwink Ten Cate
(1961:183–184), in fact corresponds to two Lycian different suffixes,
-se/i- (Mullijese/i-) and -zi- (Trijẽtezi-).13 It is because of these doubts that
I shall limit myself to listing a very reduced number of suffixes, citing
only those cases where the identification of the suffix seems clear.

§ 6.1 Place name suffixes -assa-, -nda- (Cuneiform sources) = -(a)ssow,

-(a)nda (Greek sources)
These two suffixes intervene in the formation of a large number of
Anatolian place names, documented both in the second and the first
millennia. As we have seen (cf. p. 13), it was Kretschmer (1896) who,
systematizing former ideas, rightly concluded that the Anatolian place
names in -ss-, -nd- widely attested in Greek sources belonged to a sin-
gle linguistic family spoken in Anatolia. Later, the (re)discovery of the
Anatolian languages of the second millennium allowed us to confirm this
theory, thanks to the appearance of numerous place names in -assa-,
-nda-, as well as authenticating these suffixes as purely Indo-European

12
Laroche (LNH:328, m. 25).
13
It is even possible, if we accept Neumann’s analysis (Neumann 1978:64; cf. also
Melchert DLL, s. v.) that Mullijesi- is actually a compound name from *mulli+ esi ‘shall
be strong’, parallel to Aruwãtijese/i- ‘shall be high’ (*aruwãti + esi ).
342 CHAPTER NINE

Anatolian. This is not the place to address a further aspect of Kretschmer’s

views, namely the possible link between both suffixes and the suffixes
-ss-, -ny- that appear in the Greek linguistic area, a very complex ques-
tion that remains unresolved.14 The place names formed with these
suffixes are systematically collected in two key works by Laroche: TA1
(-assa- place names) and TA2 (-nda- place names), where they are also
interpreted linguistically in the framework of Anatolian.
Both suffixes are widely found in Carian place names from Greek
sources, which should come as no surprise, since the south-western part
of Asia Minor is one of typical areas for -assa- and -anda- place names
in Hittite sources (Laroche TA2:72). Given that the exact location of
many place names form Hittite sources is unknown, it is impossible to
ascertain which of these place names in -assa- and -anda- were situated
within the limits of Classical Caria, but at least in two cases we can
establish a connection, thanks to recent progress in the study of Hittite
geography: (1) the identification Millawa(n)da with Milhtow, an Ionian
city situated on the Carian coast, is now widely upheld as a result of
the definitive equivalence Lukka = Lycia, and the subsequent reorder-
ing of the map of Western Anatolia;15 (2) the proposal of identifying
Mutamutassa from Cuneiform sources with the important Carian city of
Mylasa (Carruba 1996:23, Hawkins 1998:27), which although subject
to a certain amount of disagreement, is in my opinion consistent with
the newly configured Hittite geography.
It seems unnecessary to quote here the wealth of examples of Carian
place names constructed with these two suffixes. I refer the reader in
this case to Appendix C, where a complete list of names attested in
Greek sources can be found. As for direct Carian sources, there are
no certain examples of this type of place name, the only possible exam-
ple being the contentious alos(d) ∞arnos(d) if it is equated to Halikarnassos
(see Chapter 11, s. v. alos). A detailed analysis of Carian place names
in -(a)nda- and -(a)ssow, with some very interesting suggested inter-
pretations, can be found in Neumann (1988).

14
See the recent treatment by de Hoz (2004), who rightly criticizes the tendency
to oversimplify the problem and argues against a genetic relationship between both pairs
of suffixes.
15
Cf. Melchert (2003:6): ‘. . . it no longer seems possible to deny the long proposed
identifications of Apasa with Ephesos and Millawanda with Miletos’.
 THE GENERAL VOCABULARY AND THE PROPER NAMES 343

§ 6. 2 Luwic participle suffix (CLuw. -mma/i-, Lyc. -me/i-)

Laroche categorised as original Luwian participles a number of per-
sonal names with a -mma/i- suffix in Cuneiform sources (for instance
Ayami/aimi, from Luw. à/-àya- ‘to make’, Laroche LNH:530), although
not all the names that contain these suffixes can or indeed should be
considered participles.
Names in -mow, -miw are abundant in the Anatolian onomastics of Greek
sources, and Carian is no different. However, as in the case of the sec-
ond millennium names, it cannot be automatically assumed that these
are original participles, as this would constitute an oversimplification.
Rather, it is more advisable to interpret them in this way only if their
internal structure is clear and the base to which the suffix is attached
can be confidently identified as a verbal stem, in a similar fashion to
the abovementioned name Ayami/Aimi. In assuming this more accurate
analysis, the possibilities of recognizing participles among the Carian
names with nasal labial suffix decreases dramatically. In fact, we are
left with only one Carian name that can be interpreted as a participle
with absolute certainty: Neterbimow, thanks to its clear correspondence
to Lyc. Natrbbijẽmi, a transparent name formed by a theophore (Natr-)
and a participle of the verb pije- ‘to give’ (see above § 1.3 and § 3.1).
The search for other participles in Carian names yields only hypo-
thetical results: it is plausible that names such as qtblem-/Kutbelhmiw/
Kotbelhmow, or those ending in -om (kbiom-Kebivmow, arliom-Arlivmow,
etc.) may come from participles, but without a clear identification of
possible verbal stems in the sequences preceding -m-, it remains indemon-
strable.

§ 6. 3. -alla/i-
Anatolian personal names characterized by a suffix -alla/i- are well
attested (Laroche LNH:329–330). As with -mma/i names, formations of
different origin and meaning have converged under an apparently sin-
gle -alla/i- suffix, which could also be valid in the case of Carian com-
parable forms.
This suffix appears in Carian with the form -ol (Greek-vll-/-vld-)
and perhaps also -el (Grek -hld-). I refer the reader to Chapter 6,
p. 258 for a more detailed phonological explanation of these Carian
forms; here it is sufficient to note that a vocalic change à > o, commonly
seen in Carian, serves to explain the very particular form -ol- taken
by this suffix in this Anatolian dialect. Examples of -vll-/-vld- in
Greek sources are extremely numerous (see examples in Appendix C),
344 CHAPTER NINE

and it is also well documented directly in Carian, particularly by the

family of u≤ol-Ussvllow names, which undoubtedly contain this suffix
even though the origin and meaning of the stem to which they are
attached (u≤ º) is by no means clear.16
For other possible suffixes recognizable in Carian place names (-ulia-,
-um-), and a tentative connection to Hittite-Luwian onomastics, see
Neumann (1988:389).

16
In Adiego (1993a:44–45), ”evoro“kin’s proposal of connecting u≤-/Uss- to CLuw.
wà“u- was accepted. Now I have serious doubts about this connection, insofar as it
does not explain why the peculiar sibilant sound ≤ appears here. The explanation of
Carian genitives in -≤ as arising from *-ºsi- (see p. 313) points rather to a *usi- stem
as a starting point, although without discounting other possibilities. Could u≤- < usi-
to be related to CLuw. u““a/i-, Lycian uhe/i- ‘year’?
 CHAPTER TEN

CARIAN AS AN INDO-EUROPEAN
ANATOLIAN LANGUAGE

Despite the scarcity of linguistic information obtainable from the analy-

sis of Carian inscriptions, several traits can be observed that clearly
place Carian within the Indo-European Anatolian family of languages.
More precisely, some of these traits situate Carian in the group of so-
called Luwic dialects, integrated by Cuneiform and Hieroglyphic Luwian,
Lycian, Milyan, and probably also Sidetic and Pisidian, which share a
series of phonological and morphological features that differentiate them
from other Anatolian dialects (Hittite, Palaic, Lydian). The present
chapter will be devoted to summarizing all of these traits found in
Carian. For this task, it is essential to use evidence drawn from ono-
mastics, in order to create a more complete picture of the Luwic char-
acter of Carian. I am aware of the risk involved in basing an argument
on the etymological interpretations of proper names, but I think that
a significant number of these interpretations can be confidently used
to demonstrate the proximity of Carian to Luwian, Lycian and the rest
of the Luwic dialects. I refrain from offering an exhaustive treatment
of this subject because I consider it to be more realistic, and also more
illustrative, to focus on a reduced, but very meaningful set of traits that
clearly establish the Anatolian—and particularly Luwic—character of
Carian.
Beginning with phonology, a good indication that Carian belongs to
Anatolian is the preservation of the PIE laryngeal *h2, a trait that
differentiates Anatolian—with the exception of Lydian—from other
Indo-European languages. As shown in Chapter 6, in Carian (and also
in Lycian and Milyan) this PIE laryngeal becomes a tectal stop (lenited
in some positions, at least in Lycian and Milyan), in contrast to Hittite
and Luwian, where it appears as a (velar?) fricative (Hitt. CLuw. ¢,
HLuw. h). The examples are taken from onomastic materials, but they
seem convincing: pikº/bikº in piks, dbiks, pikre (Pigrhw), all from PIE *bhèh2-
vs. CLuw. pi¢º (see p. 337), quq = Lyc. xuga, ‘grandfather’, vs. CLuw.
¢ù¢a-.
Specifically Luwic is the satem treatment of the PIE palatal voice-
less stop *∞ (CLuw., HLuw. z, Lyc., Myl. s, against k used elsewhere
346 CHAPTER TEN

in Anatolian). Evidence for this treatment in Carian (> s) is provided

by the demonstrative pronoun s(a)-: sa, san, snn, comparable to CLuw.,
HLuw. za- vs. Hitt. ka-. We can also consider as Luwic the loss of PIE
*· in (i)brº = CLuw. im(ma)ra/i- ‘open country’, against its preservation
in Hitt. gim(ma)ra- (< PA *·emro-).
In derivational morphology, a remarkable trait that once again sit-
uates Carian in the Luwic subgroup is the presence of an ethnic suffix
-yn-/-ÿn-, easily interpretable as the Carian counterpart of Luwian
-wanni-, HLuw. -wani-, Lyc. -ñni, Mil. -wñni-.
In inflectional morphology we can also identify the ‘i-Mutation’, i.e.
the insertion of -i- (probably from PIE *-ih2-) before the ending (replac-
ing the thematic vowel if this exists) in the nominative and accusative
singular and plural of the common gender nouns, a widespread phe-
nomenon in the Luwic subgroup. It is true that this -i- does not appear
as a result of the apparently defective vowel notation in Carian,1 but
its effects can be detected in the umlaut displayed by the vowel of the
preceding syllable in the word ted, ‘father’, for which an original a can
be postulated, therefore *tadi- > *tedi- > ted- (cf. Lyd. taada- ‘father’ vs.
Lyc. tedi- ‘id.’, also with umlaut). Another possible example of ‘i-Mutation’
can be considered for the Carian genitive -≤, if we accept Melchert’s
very plausible etymological explanation (from the possessive *-assì-, in
fact an ‘i-mutated’ form of the possessive *-asso-).
Melchert’s explanation of -≤ also implies that Carian employed this
sigmatic adjectival suffix to build nominal complements, which repre-
sents another clearly Luwic feature of Carian: it is in these languages
(Cuneiform and Hieroglyphic Luwian, Lycian) that we find this suffix
used as a replacement for the inherited genitive endings, either in all
cases or in the great majority, depending on the dialect.
Other Carian case endings are equally consistent with this theory.
While the acc. sg. -n informs us merely of the IE character of Carian
(-n < PIE *-m), the common gender acc. pl. -“, if really indicative of
a *-ns origin (as the preservation of the sibilant seems to suggest, vs.
its loss when it comes from *-Vs in the nom. sg.), allows us to see the
proximity of Carian to Luwic: *-i-ns is the ac. pl. c. ending behind
CLuw. -inz, HLuw. <-iza> (= -/inz/), Lyc. -is, Mil. -iz. The presence
of the Carian palatoalveolar (?) sibilant -“, instead of simple dental -s,

1
Regarding this problem, see above pp. 238–242.
 CARIAN AS AN INDO-EUROPEAN ANATOLIAN LANGUAGE 347

suggests a secondary change of s, perhaps due to the contact with n,

similar to the case of CLuw, HLuw, Mil. -z, irrespective of which sound
actually represents z in each of these dialects, a question that is the
subject of much debate.
Verbal morphology cannot really be used as evidence of Carian as
an Indo-European Anatolian language, since the insufficient direct exam-
ples noted in Chapter 8 ( ÿbt, ait, etc.) produce a clear risk of circu-
larity: we interpret them as verbs mainly on the basis of their formal
resemblance to Lycian (ubete, aite), so it is dangerous to use them as
evidence for the classification of Carian. As for indirect examples, the
name Neterbimow, a Carian version of the Lycian Natrbbijẽmi (see above
p. 343), is a good, although isolated example that suggests the exis-
tence in Carian of a participle suffix -m, comparable to CLuw. -mma/i-,
Lycian -me/i-, itself another example of a Luwic trait.
Several words of the scarce vocabulary from Carian inscriptions whose
meaning can actually be confidently established also show clear paral-
lels to the rest of the Anatolian languages: en ‘mother’ = Hitt. anna-,
CLuw. ànna/i-, Lyc. ẽne/i-, Lyd. ẽna-; ted ‘father’ = CLuw. tàta/i-, HLuw.
tata/i- Lyc. tede/i-, Lyd. taada- (vs. Hitt. atta-); armo- ‘Moongod’ = Hitt.
arma-, Luw. *Arma- (see Melchert CLL, s. v.), Lyc. arm̃ma-, cf. Lyd. armta-
‘belonging to Arma’. Specifically Luwic are mso- ‘god’ (?), cf. also ethnic
msnord- = CLuw. mà““an(i)-, Lyc. maha(na)-, Mil. masa-, Sidetic ma≤ara (dat.
pl.), and the form of the name of the Storm-god, Carian trq(u)d- =
CLuw. Tar¢unt-, HLuw. TONITRUS-hut- (= *Tarhu(n)t-) Lyc. Trqqñt-
vs. Hitt. Tar¢u-. Finally, a meaningful isogloss shared by Carian and
Milyan is the copulative conjunction Car. sb, Mil. sebe ‘and’ (cf. also
Lycian se ‘and’).
Therefore, although the number of Carian phonological, morpho-
logical and lexical traits that can be used for comparative purposes is
limited, they are in my opinion significant and consistent enough to
classify Carian as an Indo-European Anatolian language, closely related
to the so-called Luwic dialects (Cuneiform and Hieroglyphic Luwian,
Lycian, and Sidetic). This comes as no surprise: even before decipher-
ing Carian, the analysis of Carian proper names from indirect sources
had already allowed some scholars to identify Carian as an Anatolian
dialect, and to point out its affinities with Lycian and other southern
Anatolian dialects, but we can now, for the first time, actually confirm
this linguistic classification of Carian on the basis of direct testimonies
from the language.
 CHAPTER ELEVEN

CARIAN GLOSSARY

A list of the words found in the Carian inscriptions is presented here,

accompanied by a brief interpretation (whenever possible) and a short
bibliographical note. It should be taken into account that in the cases
of inscriptions without interpunction, the entry is the result of my seg-
mentation, and must therefore be taken as hypothetical and viewed
with caution.
The following is the order of the letters (in transcription) adopted here:
a b b d d e g i/j k ∞ l l m n ñ ã o p q r ® s ≤ “ t t u/w y/ÿ/ z H 1
Note that the letters for the pairs of vowel/semivowel sounds have been
put together. This allows us to clearly show the equivalences between
forms such as w≤ol≤ (Egypt) and u≤ol≤ (Caria proper, where a specific
letter for /w/ does not exist).
I exclude from the glossary the sequences consisting of one or two
letters that appear in contexts that are very fragmentary, isolated and
impossible to interpret. The coin legends are also excluded (see the
appendix by K. Konuk).

aba ?d ?
C.Ka 8

abrq∞[
E.Ab 14

absims
C.Ka 5
Pronominal form?

In Adiego (1998a:25) absiº is tentatively connected to Lyc. ehbi < *ebesi ‘his’ (there-
fore a∞t[ms]kmt absiº = ‘his (for ‘their’?) descendents’), but this hypothesis leaves
final ms unexplained.
 CARIAN GLOSSARY 349

ada
C.Si 2a
Nom. sg. in the formula ]ryin ∞tmño≤ sb=ada ∞tmno≤ ‘Idrieus of Hekatomnos
and Ada of Hekatomnos’.
Fem. PN: Ada (the Carian queen, Hekatomnos’ daughter and Idrieus’
sister and wife): Ada in Greek sources, a typical Carian feminine name
(cf. Aba, Alasta) documented particularly in the regions of Mylasa and
Stratonikeia (Blümel KarPN:9).

Schürr (1992:138), Adiego (1994a: 40).

adymd“
C.Si 1
Singular or plural nominative? Verbal form?

Cf. Adiego 2000:152. Final -“ could point to a nominative-accusative ending.

Melchert (apud Hajnal (1995[97]:18, n. 15) suggests that it could be a verb in
a reflexive construction (‘he made for himself ’): ad (‘made’, cf. Lyc. ade) + md
(particle) + “ (reflexive); y would be an anaptyctic vowel.
Cf. dymda in Hyllarima?

ait
C.Ka 2
Possible verbal form: 3. plur. pret. ‘they have made’ = Lyc. aite (Car.
ai- = Lyc. a(i) ‘to do, to make’?).

Melchert (1998:35). See also Adiego (1998a:25, 2000:140–141).

aitusi
C.Ka 5
Perhaps related to ait. A segmentation aitu could offer a good connec-
tion with Anatolian imperatives (cf. Lycian tãtu ‘they must put’): there-
fore aitu, ‘they must make’, ai- ‘to do, to make’ = Lyc. a(i) ‘id.’), but
the resulting final word si would remain unexplained.

Adiego (1998a:25).

ai[-]iqom
E.AS 7
The segmentation of the word is very doubtful.
350 CHAPTER ELEVEN

Reading and segmentation according to Schürr. Ray (1994:203), starting from

the older reading psÿ≤ainiqom, proposes comparing psÿ≤ainiº to the Egyptian PN
P3-sb3-¢‘j-m-njw.t, Greek Cousennhw, literally ‘the star arisen in Thebes’, although
he recognizes that no explanation can be given for the isolated three final
letters qom. Vittmann (2001:58) mentions Ray’s proposal without further
commentaries.

akymyduÿeryly[vacat]d
C.xx 3
A complete inscription whose segmentation into words remains unclear.
The meaning of the whole inscription is unknown.

On this inscription, see Schürr (2001c). Schürr points out that the abundance
of vowel signs present here is very unusual in Carian. He isolates a form aky-
mudu as a possible 3rd pl. imperative, with a < */-ndu/ ending comparable
to Hittite and Luwian corresponding endings. For akymy-, he proposes a con-
nection with Hitt. ak(k)- ‘to die’. All these proposals are formulated within the
framework of a very speculative interpretation of the possible content of the
text. In any case, a segmentation akymydu ÿeryl[ is likely, given the unusual
sequence ºuÿeº.
The syllabic iteration y . . . y recalls the similar situation in ardybyr≤ (as well
as dtÿbr, k≤atÿbr, smdÿbrs), so that the first y in these cases could in fact be sec-
ondary (epenthetic?): akymyduº < *akmydu-.
Could this inscription actually be a sort of alphabet model (with the names
of letters: a-ky-my-du, etc.)?

a∞akowr
E.AS 4
a∞mnnartnyr
C.Ka 2

Van den Hout (1999) claims to recognize here a clitic chain a=∞=m=n, to be
compared with Lycian parallel forms. Cf. his analysis of a∞t[ms]kmt.

a∞t[ms]kmt
C.Ka 5
a∞tmsk[m-]d
C.Ka 2
The respective integrations are dependent on each other, but given the
textual connections between C.Ka 2 and C.Ka 5, they seem to be well
founded. Some doubts persist, however, regarding the final letters of
 CARIAN GLOSSARY 351

each word. Perhaps we are dealing with two differently inflected forms
of the same stem.
In C.Ka 5, the corresponding Greek text makes the meaning “descen-
dants” a plausible interpretation.

See Adiego (1998a:24) and here on p. 300 for a very speculative attempt at
an etymological explanation (a∞t- = Hitt. katta, mskm-, to be related to Luw.
ma“¢a¢it- ‘growth, prosperity’, the overall sense of the word being ‘offspring,
Nachkommenschaft’).
A radically different approach is taken in van den Hout (1999): he suggests
analysing this form as a clitic-sequence, in which he underlines =ms=, inter-
preted as a pl. dat. of a 3rd sing. personal pronoun. As for the rest of the
elements (segmented as a=∞=t=ms=km), he suggests some possible Lydian
parallels.

a∞t[
C.Ka 5
It could be the same word as in the two preceding entries (a∞t[mskm . . .]).

alos
E.Me 45
alosd
C.xx 2
Attested in both cases in agreement with the word ∞arnos: alos ∞arnos,
alosd ∞arnosd. Tentatively connected with the Carian place name Halikar-
nassos (ÑAlikarnassÒw), but this raises serious formal difficulties.

Halikarnassos-identification already included in Adiego (1990:135–136). With

more reservations: Adiego (1993a:245–246), (1994a:40).
The duplicate ending -d points clearly to two different nouns (substantive
+ adjective or vice versa). In Adiego (2000:154) alos-d ∞arnos-d is interpreted
as an ablative singular (Carian -d = Luw. -ti ) ‘from Halikarnassos’; alos ∞arnos
in E.Me would be merely the name of the city in nominative (see here pp. 279,
315).
Other proposals: a salutation or wish formula (Gusmani 1979a:222), (1986:62);
Schürr (1992:153) interprets -s as a dative ending, which makes an analysis
of these forms as place names difficult.

amt [
C.Tr 1
352 CHAPTER ELEVEN

an
C.Tr 2
ann
C.Ka 3
These forms appear in two funerary inscriptions. In C.Tr 2, an is accom-
panied by sidi, a typical word found in funerary contexts. In C.Ka 3,
it appears preceded and followed by two personal names in genitive
(“orus and i brs≤ ). The simplest interpretation is to consider an/ann a
demonstrative that functions as adjective modifying sidi in C.Tr 2 (‘this
tomb(?)’), and as substantive governing the personal names in C.Ka 3
(‘this of ”oru (son of ) Ibrs’).

For this interpretation, see Adiego (1996:161) and here pp. 290, 320. Also,
Hajnal sees here a demonstrative (Hajnal (1995[97]:20, from */eno-/). Adiego
(loc. cit.) suggests other possible forms of the pronoun in ankbu“ and añmsñsi.
Schürr (1996c:158) proposes that ann C.Ka 3 designates the tomb, but given
an sidi of C.Tr 2, the interpretation as demonstrative seems more suitable.

ankbu“
E.Bu 1, E.Bu 2
Perhaps a title or a kinship term (in nominative), given the contexts in
which it appears.

See Schürr’s etymological attempt (Schürr 1996[98]:97–98] to connect ºkbu“

with the kinship terms kombow, kombion ‘grandson’, attested in several late Greek
inscriptions from Caria, and related to Hitt. kappi- ‘little’ by Neumann (1961:61).
anº would then be a sort of prefix modifying in some way the kinship term
(cf. epñ-nẽni in Lycian, from nẽni ‘brother’).

an[
C.Ka 2

añmsñsi
C.Si 2a
Perhaps to be segmented into añ msñsi. In such a case, a comparison
with Luwian anni“ ma“sana““i“, Lycian [ẽ]ni mahanahi ‘mother of the Gods’
would be striking. In any case, the reading, based only on the Robert-
Deroy edition of C.Si 2, is not absolutely certain.

Connection of msñsi to Luwian ma““ana““a/i-, Lycian maha(na)- suggested in

van den Hout (1999:39).
 CARIAN GLOSSARY 353

aor≤
E.Me 1 (aor[≤]), E.Bu 6
PN in genitive, Carian adaptation of an Egyptian name ( J-Ór literally
‘O(?) Horus’, [a˙òr], Greek Avw (?), see DNb:55).

Vittmann (2001:42).

apmen
E.Me 44a
PN in nominative, Carian adaptation of the Egyptian name Óp-mn (lit-
erally ‘Apis is perpetual’, *[˙apimèn], Greek ÑApimenhw, see DNb:781).

ap[---]ws
E.Me 23
PN. It is not clear if it is a nominative of a s-stem, or a ‘s-case’.

a?q≤baq
E.Th 10

a[rb]ikarm≤
E.Me 23
PN in genitive. If the form is correctly completed, it gives a Carian
name corresponding to Lycian (in Greek sources) Arpigramow (Zgusta
KPN §104–1). The name would therefore be a compound ar-bikarm-.
For the second element, cf. pikarm-/pikrm-.

Completed and connected to Arpigramow by Kammerzell (1993:214), this view

is commonly accepted: see for instance Schürr (1992:139), Adiego (1994a:31)
(with further remarks on the structure of the name).

ardybyr≤
E.Me 52
PN in genitive. Corresponding to the Carian name in Greek adaptation
Arduberow. A name belonging to the family of names in -(d )ybr-.

Connection to Arduberow already suggested by Ray (1982b:189), but seriously

hampered by his decipherment system (†argébér≤ ). See Adiego (1993a:225–226),
(1994a:40).
The doubts about the exact reading raised in Schürr (2001c:119)—who sug-
gests an alternative interpretation a | rdybyr≤ )—are not particularly convinc-
ing, and in any case do not alter this correct identification.
On this family of names, see Adiego (1993a:224–227).
354 CHAPTER ELEVEN

are“
C.Ia 3
Probably a PN in singular nominative (but a plural nominative cannot
be ruled out). It is possible that the beginning of the word is incomplete.

arie ?≤
E.Me 38
PN in genitive.

arjom≤
E.Me 42
PN in genitive. The connection with the PN arliom- is unclear: is it
perhaps a variant, parallel to arŕi“, ari“ (Arrissiw), as well as arli“
(Arlissiw). Independently of this possible connection, compare also arj-
with the Carian PNs Ariauow or Aridvliw.

Schürr (1992:134) suggests attributing the alternation arjom-/arliom to a dialec-

tal variation.

ari“
C.St 1
ari“≤
C.Hy 1a
PN in nominative (ari“ ) and genitive (ari“≤ ). Correspondence to the
Greek adaptation Arrissiw is confirmed by the evidence of the bilingual
inscription C.Hy 1, where the name Arrissiw appears in the Greek
part (although the individuals mentioned are not necessarily the same).
Given this identification, ari“ could be a variant spelling of arŕi“, q.v.
Both ari“-Arrissiw and arŕi“ could be variants of arli“-Arlissiw, as arjom
against arliom. See arjom for further remarks.

ar∞ila≤
E.Me 39
PN in genitive. Perhaps a Carian adaptation of the Greek PN ÉArx°laow
(Dor. ÉArx°law). For the use of ∞ for a Greek velar stop, cf. urs∞le-.

Greek origin already suggested by Ray (1994:202)]. For an interpretation as

an purely Anatolian name, see Adiego (1993a:240).
 CARIAN GLOSSARY 355

arliom≤
E.Me 9 (arlio[m≤]), E.Me 43b
PN in genitive. It is the Carian name that appears as Arlivmow in
Greek sources.

For the identification, see Adiego (1990:134; 1993a:231; 1994a:35).

arli“
E.Ab 24
arli“≤
E.Me 9, E.Me 15, E.Me 51
PN in nominative (arli“ ) and genitive (arli“≤ ), corresponding to Arlissiw
in Greek sources. Cf. moreover the Carian place name Arlissow. The
stem seems to be the same as arliom-. Note also the possible variants
ari“, arŕi“.

Connection to Arlissiw already stated in Faucounau (1984:236). Cf. Adiego

(1993a:230; 1994a, n. 3.3).

armon
E.Me 8a, C.Eu 2
In E.Me 8a, noun in nominative: ‘dragoman, interpreter’, corresponding
to Egyptian p3 w˙m ‘dragoman, interpreter’ in this bilingual inscription.
In C.Eu 2: function and meaning unknown (it would be very unlikely
for it to have the same meaning as in the other example).

For a detailed discussion of the problems posed by the exact meaning of

Egyptian p3 w˙m, see Vittmann (2001:50–52), who argues convincingly for the
sense ‘interpreter’, and dispels all the doubts raised in Masson-Yoyotte (1956)
and subsequent literature about this interpretation.
Janda (1994:180–182), starting from the alleged meaning ‘herald’, proposes
for armon- an origin from *ar(V)ma-wanni-, where *ar(V)ma- would correspond
to HLuw. ataman-/adaman-/ ‘name’, assuming a semantic change, ‘name’ >
‘determination, decision’. The meaning ‘dragoman’ corresponds even better to
the semantics of this explanation (perhaps ‘interpreter’: ‘who names the things
in another language’) but the sound changes assumed are ad hoc.
For armon in C.Eu 2, see above p. 309.

armotrqdos
C.Hy 1a
Most probably a dvandva-like compound formed by the divine names
armo- (Carian version of the Anatolian moon-god Arma-) and trqd- (the
356 CHAPTER ELEVEN

Anatolian Storm-God, Hitt. Tarhu-, CLuw. Tarhunt-, Lyc. Trqqñt-, see

below, trqude). A (possessive?) ending -os (= Luw. -a““a-?) has been added
to the compound stem armo+trqd-.

On this form, and for a different possible analysis, see Adiego-Debord-Varinlio<lu

(2005:616–617). The possibility cannot be discounted that a secondary vocalic
stem (trqdo-) may have been formed from the original nt-stem (as in the rest
of Luwic languages) trqud- (<*tarHunt-), so that the ending would be -s (like in
ntro/ntro-s).

ar®i“
E.Bu 1(ar[®]i“), E.Bu 2
PN in nominative. The identification with the Carian name from Greek
sources, Arris(s)iw, depends on the exact value of the letter transcribed
as ®, but this is now reinforced by the certain equivalence ari“ (q.v.)
= Arrissiw in Hyllarima (where there is no letter ® ). The connection
of all of these forms to arli“-Arlissiw is unclear (cf. also arjom- and
arliom-).

Adiego (1992a:34). See also Adiego (1994a:35).

artay≤
E.Me 22
PN in genitive, to be compared with Artaow, although the Greek adap-
tation as an o-stem could mean that the stem is not exactly the same.
Cf. also Arthumow (Zgusta KPN § 109, Blümel KarPN:11), which can
be analyzed as a stem, *artay-/artey-, followed by a suffix, -m-. The final
part of Arthumow recalls names such as paraeym- parpeym-. For art º/Art-
cf. also Artuassiw, Aryuassiw.

Adiego (1993a:231–232; 1994a:35).

artmi
C.Tr 2
PN in nominative(?). It could be the Greek goddess name ÖArtemiw,
directly used as personal name (cf. Lydian artimu≤, also attested both as
personal and god names), or a variant form of the Carian name Artimhw.

The problem posed by the relationship between the Greek divine name (of
Asian origin?) Artemis, and the family of Anatolian names collected in Zgusta
KPN § 108 (Arteimaw, Arteimow, Artimhw, etc.), is not at all clear, and cannot
 CARIAN GLOSSARY 357

be dealt with here (see remarks in Zgusta KPN:102 and Adiego-Debord-

Varinlio<lu 2005:611). Cf. also the Carian name rtim.
The ambiguous structure of C.Tr 2 does not mean we can dismiss the the-
ory that artmi could be a dative, see above p. 290.

artmon
C.Tr 1(art{ }mon)
PN in nominative. It is the Carian adaptation of the Greek name
ÉArt°mvn. Note the identical adaptation in Sidetic (artmon).
The Greek PN ÉArt°mvn is well documented in Caria. For instance,
it appears in the well-known inscription of Halikarnassos, SGDI 5727,
where it is used by individuals whose father’s name is clearly Carian
(ÉArt°mvnow toË Panamuv, ÉArt°m[v]na Territò, etc.).

For the identification, see Schürr apud Adiego (1994, 3.49).

a≤b≤t
E.Th 13

a[-]mob[
C.Al 1

a[--]a[----]om≤
C.St 1
The final part clearly indicates a PN in genitive (for the ending, cf.
names such as arliom-, kbjom-, etc.

bal
E. Th 49
PN?

baqgk[. . .]
C.Ia 5

bebi
E.Th 23
It seems to be an incomplete form of bebint, see the following entry.
358 CHAPTER ELEVEN

bebint
E.Th 28 (bebi.nt), E.Th 30, E.Si 4 (be?bint), E.AS 7
Word of unknown function and meaning. In E.Th 28 and E.Si 4 it appears
preceded by a diamond-like sign (t, K) whose function is uncertain.

Former readings of some of these testimonies raised doubts about the second
letter of the word (w ÿ rather than e e), but a unified reading, bebint, must
now be preferred (see Schürr 2001b:108).
Schürr has argued in different works in favour of an interpretation as a
verb (a third singular preterite form, see Schürr 1996a:65) with the meaning
‘to offer’ (Schürr 2001b:108). It is also seen as a verb by Hajnal (1995[97]:18):
a 3rd sg. present ‘he sends’ of the verbal root that he also recognizes also in
binq (q.v.).

bebnd
E.Th 6
Perhaps related to the preceding entry (bebint).

bejeym
E.Th 28
Perhaps a PN in nominative. As for the final part of the word, cf. the
PNs paraeym, parpeym.

be≤ol
E.Ab 23
PN in nominative. Although no directly comparable form is attested
in Greek sources, it recalls other Carian names with -ol (Greek -vllow,
-vldow).

betkrqit[-- . . .]
E.Si 5

bi
C.Ka 2 (2×)
C.Ka 5
Apparently a conjunction or particle, as it appears in C.Ka 2 follow-
ing an accusative otr“ (‘themselves’). In this same passage, it seems to
be used as a postclitic in correlation: . . . bi . . . bi. Its identification in
C.Ka 5 is less certain. Cf. also another example ]bi in C.Ka 2.

On the possible different interpretations of bi, see p. 302.

 CARIAN GLOSSARY 359

bid≤lemsa
E.Ab 30
Function and meaning unknown. Not a PN, because it precedes an
onomastic formula in nominative. Cf. perhaps [--]msal (E.Bu 1), which
also precedes a PN in nominative.

Hajnal (1995[97]:23) proposes the segmentation bid≤lem sa, where sa would be

a demonstrative (see sa). For the first word, he suggests that ºem indicates a
participle.

bij≤≤pe
E. Si 8
Probably a PN in nominative.

binq
C.Si 1
A verb? The reading of the last letter is not absolutely certain.

Hajnal (1995[97]:18): ‘ich schenkte’, 1st sg. pret. < */píäiannà-¢a/, to be com-
pared to Hitt, piyannài- ‘to send’. For the same verb also ºbint, → bebint as 3rd
sg. present.

bsis
E.Ab 30

b?s?ui∞am
E.Lu 5

bu∞y[-----]i[-----]i
C.Ka 5

bÿ“
E.AS 7
A word that apparently agrees with the following: (esak?dow“ ). A demon-
strative pronoun in accusative or nominative plural?

The idea of a demonstrative pronoun has been repeatedly defended by Schürr

(see Schürr 1996a:69, 2001b:98, 112): b- would come from PA *obó/i-( >
CLuw. apa-, Lyc. ebe-) with aphaeresis (as in HLuw. pa- besides apa-, or Lyd.
bi- [pi-]). The ending -“ can be compared with -“, the acc. pl. animate end-
ing in C.Ka 5.
360 CHAPTER ELEVEN

bÿta“
E.Si 6
PN in nominative?

banol
C.My 1 (2×)
PN in nominative. Perhaps the Carian name corresponding to Ibanvlliw
in its Greek adaptation.

For this possible connection with Ibanvlliw, see Blümel-Kızıl (2004:134), Adiego
(2005:85).

bem≤
E.Me 17
PN in genitive.

b2o[--]ol“
C.Ka 5
Noun in plural accusative. It appears in a long sequence that corre-
sponds to the Greek formula proj°nouw e[‰nai k]a‹ eÈerg°taw Kaun¤v[n].

Frei-Marek (1997:38–39) try to bring this word closer to 1orsol“ (completing

therefore [rs]). However, this solution implies that the initial letter of the word,
ÿ (here transcribed by b2), has to be likened to 1, contradicting the theory
defended in the present book, that ÿ is a letter for b (see also Adiego 1998a:23).
As for the meaning, it hardly can be ‘descendents’, as Frei-Marek suggests,
given the clear correspondence otr“ = aÈtoÊw. b2o[--]ol“ must rather be a word
at the same level as kbdyn“ (in kbdyn“ sb b2o[--]ol“ ). In Adiego a connection of
b2o[--]ol“ with Imbros, the citadel near Kaunos, is very tentatively suggested.
See pp. 299–300.

brsi
E.Th 26, E.Th 48, C.Hy 1a
brsi≤
C.Hy 1a (2×), C.St 1
PN in nominative (brsi) and genitive ( brsi≤ ). Carian name rendered in
Greek as Imbarsiw, Imbras(s)iw. Form with aphaeresis or no notation
of initial vowel that coexists with the full forms ibrsi-, ibarsi-. Identification
now assured thanks to the bilingual inscription C.Hy 1, where both
brsi and Imbrasiw appear (although not necessarily referring to the same
individuals).
 CARIAN GLOSSARY 361

Schürr (1991[93], passim). For brsi—Imbrasiw in Hyllarima, see Adiego-Debord-

Varinlio<lu (2005).

dbikrm
E.Th 19
PN in nominative. Can be analysed as (i)d- + bikrm. For the first ele-
ment, cf. d-quq- (Greek Idagugow), d-w≤ol/id-u≤ol- (Greek Idussvllow),
d-biks. For the second element, cf. pikrm/pikarm (cf. also Pigramiw, Pigramow
in Lycia, Zgusta KPN § 1255–1, 2).
Note below the ‘devocalised’ form dbkrm.

dbiks
E.Th 13
PN in nominative. A compound name formed by d- (see dbikrm above)
and biks (see piks[ ).

dbkrm
E.Ab 34
PN in nominative. A variant form of dbikrm.

dmo“bqs
E.Si 11
Perhaps a PN.

See the study dedicated to this word by Schürr (1996b). There, ºbqs is con-
nected with piks-/biks- and other forms that are derived from PIE *bhè́ hos (see
piks-).

dm-?-n
E.Th 34
It could be a noun in accusative singular, given the context (prna∞non
dm-?-n).

dokmmpint
E.Th 4

dquq
E.Th 44
PN in nominative. It corresponds to Idagugow in Greek sources, a com-
pound name (i)d- (see dbikrm-) + quq- (see quq-).
362 CHAPTER ELEVEN

For this identification, see Adiego (1993a:235; 1994a:35–36). Tentatively com-

pared in Adiego (1995:27, n. 9) with Milyan ddxug[, but the segmentation and
the interpretation of this latter form are doubtful (see Schürr 1996b:154).

dr“≤iem
E.Th 53
PN in nominative?

Initial dr“ º recalls dar“ º in dar“qemorms[. Note that in Thebes d is practically

absent, and its function could be assumed here by d.

∂saml-?-?-$o
E.Th 16

dtÿbr
E.Th 2
PN in nominative. It belongs to the family of names in -(d)ybr- (and
variants). Cf. ardybyr-, dybr, etc.

On this family of names, see Adiego (1993a:224–227)

dw≤ol≤
E.Me 35
PN in genitive. This is a variant with aphaeresis (or no notation of the
initial vowel) of the name now attested in its complete form in Mylasa
(idu≤ol-), which corresponds to Idussvllow. A compound name (i)d- (see
dbikrm-) + u≤ol≤ (cf. u≤ol-/ Ussvllow).

For this identification, see Adiego (1993a:241). Aphaeresis suggested by Neumann

apud Adiego ibid.).
Schürr (1996b:154) recalls, in addition to Idussvllow, the Carian name
Yussvllow.

dÿbr
E.Th 5
PN in nominative. A name belonging to the family of names in -(d)ybr-.

On this family of names, see Adiego (1993a:224–227)

 CARIAN GLOSSARY 363

dymda
C.Hy 1a
Meaning and function unknown. Cf. in any case adymd“ in C.Si 1.

dar“qemorms[
C.St 2
Part of an onomastic formula, like those that precede it in the same
inscription?

Schürr (2001c:119) proposes the segmentation dar “qem orm s[ and interpret-
ing “qem as a participle. For dar“, if a PN, compare with Andarsvw and prob-
ably also Dersvw (see p. 246).

den
E.Sa 1

Analysed as a sort of preposition (comparable to Hitt. andan) governing tumn

(= ‘for Atum’) in Adiego (1995:21–23). See here p. 287.

drual
C.Ka 5
Unclear form. It must correspond in some way to Greek §p‹ dhmio[u]rgoË
in the bilingual inscription C.Ka 5, but the precise analysis remains
unclear.

An attractive theory has been formulated independently by Hajnal (1997b:150)

and Melchert (1998:37; cf. also 2002:308) that the entire sequence i[--]inis d
rual can be analysed as ‘under the ship of Hi[pposth]enes’, with d < *nde =
Lyc. ñte ‘in(to)’, i[—]inis genitive/possessive, and rual as a noun correspond-
ing to the title dhmio[u]rgoË. Attempts to find an etymological explanation for
rual (Hajnal 1997b:151, from */erowà-/ ‘freeedom’, cf. Lyc. arawa-; Schürr
2001b:109–110, a similar explanation, but starting from a different analysis
of the sequence) seem somewhat rash.

emsglpn
E.AS 4
364 CHAPTER ELEVEN

en
E.Me 32
‘mother’ (nominative). It corresponds to Lycian ẽni, Lydian ẽna≤, CLuw.
anni“ ‘mother’. Carian vocalism points to an ‘i-mutation’ stem (*ani-)
and subsequent umlaut a > e caused by this i-suffix.

For this meaning, see Schürr (1996a:62), Hajnal (1995[97]:21–22). For i-muta-
tion and umlaut, see Schürr (2001b:97) and here on p. 259.

eri
C.Si 2a
Function and meaning unknown.

In Adiego (2000:143–144) it is interpreted as a noun with the meaning ‘tax

exemption, ét°leia’, functioning as direct object of pisñoi mda (q.v.). Moreover,
a connection with Lycian arawa- ‘tax exemption’ is suggested. Neumann (apud
Adiego 2000:144) envisages the possibility that eri could be a preverb identi-
cal to Lycian eri. All of these hypotheses are dependent on a particular inter-
pretation of C.Si 2 traced in Adiego (2000), cf. here on p. 304.

esak?dow“
E.AS 7
Very likely to be a compound noun that seems to contain the stem
kdow- (cf. kdou≤ in E.Bu 1), usually interpreted as the Carian word for
‘king’ (cf. Lycian xñtawat(i )- ‘king’, and the form KNDWÍ/KNDWS in
the Aramaic part of the Trilingual inscription of the Letoon of Xanthos,
which is perhaps a direct reflection of the Carian word).

See Adiego (1995:18–21) for a detailed discussion of this question. The most
compelling point of this interpretation is undoubtedly the coexistence in E.AS
7 of esak?dow“ and pisma≤k (= Psammetichus)—separated by the word mÿqudem—
that opens the possibility of interpreting them as ‘the king . . . Psammetichus’,
and of linking this inscription to the well-known long Greek graffito, also from
Abu-Simbel, in which this Pharaoh is also mentioned. But decisive formal
details of esak?dow“ remain unexplained: the value of initial esa-, and the precise
function of -“, which seems to agree with the word preceding esak?dow“, bÿ“
and which recalls the ending of plural accusative in C.Ka 5, which would make
the interpretation ‘king Psammetichus’ very difficult to argue (see here on
p. 294).
 CARIAN GLOSSARY 365

ewlane
E.Th 49
See following entry and ewmlane.

ewm
E.Th 10, E.Th 13
A typical sequence in Theban graffiti: Cf. sl∞maewm, which can perhaps
be segmented into sl∞ma ewm, the incomplete form ewm-?-?-?-?, and the
sequence ewmlane. Note also ew in the preceding entry ewlane.

See p. 323 for a possible segmentation into two elements ew+m.

ewmlane
E.Th 12, E.Th 44
One of the alleged ‘verbs’ in mlane, mdane, q.v. Note that the segmen-
tation is not clear (ew+mlane, ewm+lane or ew+m+lane?), see p. 323.

euml ?bnasal
E.Bu 2
Analysis, function and meaning are unknown. Cf. eypsal also in Buhen,
which would allow us to isolate an element sal.

Could be sal and adverb with the meaning ‘here’ (cf., for the ending, Lyc.
ebeli, and for the stem, the pronoun sa-/sn-).

ewm-?-?-?-?
E.Th 52
See ewm entry.

eypsal
E.Bu 6
See euml?bnasal above.

gkem≤
E.Th 44
PN in genitive

grdso[-]i[
C.Ka 2
Probably related to qrds, qrdsol, especially since it appears in a sequence
qrds grdso[-]i[ that could constitute a type of figura etymologica.
366 CHAPTER ELEVEN

gdb“laã1i[-]
C.Ka 2

i[---]inis
C.Ka 5
PN, probably with s-ending. This incomplete name must be the Carian
adaptation of the Greek Hipposthenes, the name of the demiurge in
the bilingual inscription C.Ka 5. Tentatively completed as i[poz]inis
(Frei-Marek 1997:31).

For this interpretation, see drual. However, an alternative segmentation i[---]ini

sdrual (thus Schürr) cannot be discounted. As for the missing letters, Frei-
Marek’s solution is quite good, but not the only possibility (the use of <z>
for Greek sy is impossible to demonstrate).

ialli
E.Ab 40
PN in nominative.

iarja≤
E.Ab 2
PN in genitive.

iasoum
C.Ki 1

See here p. 142 on the doubts about the exact reading and a very hypo-
thetical alternative reading.

ibarsi≤
E.Ab 3
ibrsi≤
E.Bu 4
PNs in genitive. Alternate forms of a name corresponding to Imbras(s)iw/
Imbarsiw in the Greek source. For further remarks, see the variant form
brsi-.

ibrsdr[-]
C.Ka 4
The initial sequence is undoubtedly related to the ibarsi-/ibrsi-/brsi-
family of words, but both the precise analysis and the segmentation
 CARIAN GLOSSARY 367

into elements are not clear. Perhaps we must segment ibr-s dr[ and
identify here a construction similar to i[---]ni-s drual (‘under the demi-
urge Hipposthenes’, according to the Greek translation) in C.Ka 5.

Connection to ib(a)rsi-family already noted in Schürr (1992:140). Schürr also

recalls Imbrow, the name of a citadel near Kaunos (Zgusta KON § 373–1,
Blümel KarON:168), and whilst in former works he suggested that ibrsº could
be the ethnic name of Imbros, following the discovery of the Kaunos bilin-
gual he prefers to take ibr º as the actual name of the citadel, and to inter-
pret ibr-s dr[u] as ‘(in?) Imbros, ‘the people’ (with *sdru- ‘people’ as the basis
for sdrual ‘demiurgos’). However, s is most likely to be an ending, both here
and in C.Ka 5.

ibrs≤
C.Ka 3
PN in genitive. The same name as ib(a)rsi-, or a name closely related
to it.

The doubts about the analysis lie in the absence of i at the end of the stem.
It could simply be the result of a defective notation, or perhaps the reflection
of a different suffixation (for instance -s- < *-so- against -si- < *-siyo-).

idmns
E.Me 33a
E.Me 33b
PN. Case unclear (nominative of an s-stem, or rather a stem idmn- with
a ‘dative’ -s ending?). It seems to be a compound whose first element
is (i)d-, cf, idu≤ol-/dw≤ol-/Idussvllow, etc.

idmuon≤
E.Me 18b
PN in genitive. No parallel forms in Greek sources. It could be ana-
lyzed as a compound name: id + muon-. For the first element, see the
preceding entry (idmns), whilst -muon- could belong to the family of
muwa- names.

Janda (1994:176) suggests interpreting this name as a derivative of the Carian

place name Iduma by means of the ethnic suffix Luw. -wanni-: ‘aus Iduma stam-
mend’.
368 CHAPTER ELEVEN

idrayridsemdbq
C.My 1
Heading of the long inscription of Mylasa, followed by the words mol“
tu∞[, and a list of onomastic formulae. This sequence, undoubtedly con-
stituted by more than one word, remains impossible to analyse.

See here on p. 308 for the sequence idrayriº.

idu≤ol≤
C.My 1
PN in genitive. Carian name rendered in Greek as Idussvllow. For its
analysis, see the variant form dw≤ol-.

Blümel-Kızıl (2004:137).

idyes≤
E.Me 63a
PN in genitive. Initial id º recalls the lexical element (i)d-/Id-, see dw≤ol-.

idym“
C.Ka 4
The final -“ could point to a plural accusative (and also nominative?).

A stem idym- recalls the Carian place name Iduma.

idyri∞≤
C.Eu 1
C.My 1
PN in genitive. A clear compound name id+yri∞-, still not attested in
Greek sources. For the first element, cf. (i)d-/Id- in dquq, idu≤ol, etc. The
second element is the well-known stem yri∞-/yriq- /ÿdiq-/ÿd∞- (Greek
-urigow, -udigow).

Cf. Blümel-Kızıl (2004:137), who reconstruct a possible Greek adaptation

*Idurigow. For a unifying explanation of yri∞-/yriq- /ÿdiq-/ÿd∞-, see pp. 262–263.

inut≤
E.Ab 18
PN in genitive.
 CARIAN GLOSSARY 369

ionel≤
E.xx 3
PN in genitive. It could contain the same stem as ‘Ionia’, ‘Ionian’.

For this connection, see Schürr (1991[93]:173), Adiego (apud Schürr ibid.). In
Adiego (1994:49, n.15) a detailed account of the possible process is given:
starting from *iona- < *iyauna- < *iyawana- ‘Ionian’ (cf. Lyc. Ijãna-), *-wana-
would be the ethnic suffix Luw. -wanni-, Lyc. -ñni, Carian -yn/-on.

irasa
E.Si 3
PN in nominative? The context is very unclear.

irow
E.Me 14, E.Me 16
irow≤
E.Me 27
PN in nominative (irow) and genitive (irow≤ ). Most likely to be a femi-
nine name in E.Me 16 and 27 (see pp. 272–273). As according to
Vittmann, a name of Egyptian origin: J.r=w (attested both as a mas-
culine and feminine name), phonetically [iròw] or [ j6ròw].

Vittmann (2001:45). Egyptian origin already suggested in Ray (1994: 202). The
Egyptian interpretation for irow now seems preferable to former attempts to
analyse it as Anatolian (cf. Ray 1982b:184, Melchert apud Adiego 1995:23,
Hajnal 1995[97]:27, n. 38), especially given the difficulties raised by the different
proposals of this type (see Adiego 1993a:247; 1995:23–24).

isor≤
E.xx 1
PN in genitive, of Egyptian origin: it is the adaptation of Ns-˙r (liter-
ally ‘(s)he belongs to Horus’, phonetically reconstructed *[6s˙òr]; Greek
perhaps ÉEsour, ÉEsouriw, etc., DemNb:685).

Vittmann (2001:50). Former attempts to connect it to the Anatolian place name

Isaura (Adiego 1993a:247) must be ruled out.
370 CHAPTER ELEVEN

iturow≤
E.Me 32
PN in genitive, whose feminine character is clear from the context (the
word en ‘mother’ refers to it). Carian adaptation of the Egyptian (both
masc. and fem.) PN Jr.t=w-r.r=w (DemNb:70) *[ j6turòw], Greek ÉIyorvw.

Ray (1994:202). See Vittmann (2001:45) for the phonetically reconstructed form.

iÿkr≤
E.AS 5
PN in genitive

jzpe
C.xx 2
PN? It immediately precedes mdane in the inscription.

In Adiego (2000:154) it is analysed as a PN in dative. Schürr (1996a:65), takes

Jzpe to be the name of the donor of the object (presumably in nominative).
Hajnal (1997b:150) interprets the entire sequence alosd ∞arnosd jzpe as a dat-
ing formula ‘zur Zeit der ?-schaft von Jzpe’, but he does not explain the pre-
cise morpho-syntactical status of jzpe.
All the attempts to find explanations for a possible PN jzpe are somewhat
speculative: Schürr (1996a:65, n. 14) compares jzpe with the Lydian PN i≤tubelm-
, and Hajnal (1997b:150, n. 14) recalls the Persan PN Vi“tàspa-/ÑUstãsphw.

kattÿri≤
E.Ab 25
PN in genitive

kbdmu≤
C.My 1
PN in genitive. Perhaps a compound name kbd+mu-. The first element
clearly recalls the place name kbid- ‘Kaunos’ (for the omission of the
vowel, cf. particularly the form of the ethnic name kbd-yn-“ ), although
it is not clear if the PN alluded directly to the place name or if it con-
tained the common noun from which the place name was created. As
for the second element, it seems to be the well-known Anatolian stem
muwa-, ‘strength, force’.
 CARIAN GLOSSARY 371

For the connection of kbd(+mu)- to kbid- ‘Kaunos’, see Blümel-Kızıl (2004:136),

who also recall the Carian name from Greek sources Kbvdhw. For a possible
Anatolian etymology of all these names in kbº, see p. 334.

kbdyn≤
C.Ka 5
‘Kaunians’ (in plural accusative). Meaning assured by the Greek part
of the bilingual C.Ka 5. An ethnic noun derived from the place name
kbid- ‘Kaunos’ (q.v.) by means of a suffix -yn- that corresponds to Lycian
-ñni-, Milyan -wñni-, CLuw. -wanni-.

Identification as ethnic name and comparison of -yn- with the Luwic suffix
already in Frei-Marek (1997:37, 50).

kbidn
C.Ka 5
Carian name of the city of Kaunos (cf. the Lycian form xbide), or, more
improbably, an ethnic name derived from it (cf. kbdyn“ ). Morphological
analysis is unclear.
Cf. the Lycian name of this city, Xbide, and the Aramaic adaptation
of the god name ‘Kaunian king’, KDWÍ/KDWS KBYD”Y.

For the reading of the last letter of the word, see Frei-Marek (1998:2). It is
possible that this city name was in Carian a plurale tantum (as is presumably
Lycian xbide), according to Hajnal (1997b:149), see also Melchert (2001:310,
n. 12). Both Hajnal and Melchert (loc. cit.) point to a genitive plural (-n <
*-òm), but while Hajnal imagines a true genitival value (‘decree of Kaunos’),
Melchert suggests that this genitive plural could become a dative-locative, like
in Lydian (therefore kbidn ‘in Kaunos’).
For the possibility that kbidn is an ethnic name (in nominative plural, from
*kbid-wen-is), see Adiego (1998a:20), (2002:19–20). For a possible etymology of
the names in kbº, see p. 334.

kbjom≤
E.Me 12, E.Me 32, E.Th 13
PN in genitive. Carian name adapted in Greek as Kebivmow. kbjom- also
appears in the compound name “arkbiom.

Adiego (1993a:232). Hajnal (1995[97]) suggests that this name would contain
a participle < */piìemmo/ì/- ‘given’, with an alleged change *^m > om. For a
possible etymology of the names in kbº, see p. 334.
372 CHAPTER ELEVEN

kblow≤
E.Th 46
PN in genitive.

kbokt≤
E.Th 2
PN in genitive.

See Adiego (1994b:252), where the name is analysed as a compound formed

by kbo- (cf. kbo-s, perhaps also taqbo-s) and kt (cf. ktmno).

kbos
E.Me 24
Ethnic name? (‘Keramean’?). It could be formed from a place name
kbo- (‘Keramos’? see Konuk 2000b) by means of an -s- suffix (cf. otono-
s-n from otono- ‘Athens’).

For this interpretation, based on the peculiar structure of E.Me 24, see above
p. 278. The interpretation was already suggested by Schürr (2003:116, n. 1).
For a possible etymology of the names in kbº, see p. 334.

kdou≤
E.Bu 1
Noun in genitive.

Could it be the Carian word for ‘king’? See esa?kdow“.

kdu. si≤
E.Ab 35
PN in genitive. It seems to contain the lexical element kd-, as esa?kdow“,
kdou≤, kdu≤ol“ (= Hitt., Luw. ¢ant-).

kdu≤ol“
C.xx 4, C.xx 5
PN? It seems to be a name of the u≤ol-family, but the final -“ could
be a plural ending (cf. kbdyn“, sarni“, etc.) Note also the element kd- (see
preceding entry).

Schürr (2001b:117) suggests that we are dealing with the plural of an adjec-
tive whose meaning would be ‘belonging to the king (= the god)’ (for the
meaning ‘king’, cf. esa? kdow“, kdou≤ ).
 CARIAN GLOSSARY 373

kdu≤opizipususot
C.Hy 1b
Sequence most probably consisting of more than one word, but impos-
sible to be segment with any confidence.

Note the presence of the sequence kdu≤ º, to be related either totally or par-
tially to esa?kdow“, kdou“, kdu≤ol“.

ken
E.Th 28

kidbsi≤
E.Me 15
Ethnic name (less probably PN) in genitive. It could be (at least orig-
inally) the ethnic name derived from a place name *kidb-, to be identified
with the Carian city Kinduh in Greek sources.

Hajnal (1998:90). Schürr has attempted to establish different connections for

this word, which he instead considers a PN: comparison with the Carian name
Kindacow (Blümel KarPN:16), name of the father of the founder of Masanvrada,
according to Stephan of Byzantium (Schürr 1991[93]:170); connection to Mil.
xñtabasi, poss. adj. of xñtaba ‘ruler’ (Schürr 1996b:152) and to the Cilician
names Kendebhw, Kendhbhw, Kendhbaw (Schürr 2001b:105), for these latter names,
see Zgusta KPN § 576).

kilarad [
C.Ki 1
kil[
C.Ki 1
Place name: the Carian city of Kildara/Killara (Kildara/Killara). In
the first example it is not clear if d belongs to the word.

The identification was made already by Kowalski (1975:79, 83), although his
transcription was still very unsatisfactory (krºźara).

klorul
E.Me 6
Ethnic name, title, or common noun in nominative. Hardly a PN, given
the structure of the inscription in which it appears.
374 CHAPTER ELEVEN

See above p. 270, where the inscription is analysed and this word is taken as
an ethnic name. Schürr (1992:135) instead suggests that it is a common noun
with the meaning ‘wife’, because he interprets the entire inscription triqo parma≤≤
∞i klorul ∞i as ‘Triqo (f.), the wife of Parma≤’.

knor
C.Kr 1

kojol
E.Me 44a
Ethnic name, title or sim. in nominative. Hardly a PN, given the struc-
ture of the inscription in which it appears.

Above p. 271, kojol is analysed as an ethnic name and tentatively related to

the name of the island Kos. This connection is hampered by the fact that the
name of the person to which kojol seems to make reference, is of Egyptian
origin. Schürr (1992:155) proposes interpreting kojol as a ‘title in -ol’ (like
nuolº, sarmrolº), and suggests connecting it with the Carian gloss ko›on, ‘sheep’.

kolt
E.Si 2

ko“m≤
E.Th 39
PN in genitive

kowrn[. . .?
E.Si 2

kow[?-?]
E.Th 24

krws
E.Th 39, E.Th 45 (krwß)
PN in nominative.

ksbo
C.My 1
PN in nominative. Carian name that corresponds to Xasbvw (all the
examples are from Mylasa) in Greek sources. Cf. also Kasbvlliw.

Blümel-Kızıl (2004:137).
 CARIAN GLOSSARY 375

ksolb≤
E.Me 43a
Ethnic name (less probably PN) in genitive (see above p. 269). In either
case, it seems to be related to the place name Kasvlaba from Greek
sources.

Schürr (1992:143), Adiego (1994a:36). As an ethnic name: Janda (1994:174).

k≤atÿbr
E.Th 2
PN in nominative. Name belonging to the -(d)ybr-family. Cf. in this case
the Lycian name Janduberiw (Zgusta KPN § 1061).

For this family of names, see dÿbr-.

k“mmsm[. . .]
E.AS 8

ktais
C.Eu 1
PN in ‘s-case’ (‘dative case’) or, less likely, in nominative. Carian form
of the Greek name ÑEkata›ow.

For the identification: Schürr (1992:154), Schürr apud Adiego (1994a:39,

1994b:252). In Adiego (1994a:39) ktais is interpreted as an s-stem created from
a Greek nominative (like Lycian zeus-, from ZeÊw), but in the present book an
analysis of -s as an ending is preferred (see pp. 288–289).

ktmn
E.Th 37
PN, perhaps an incomplete form of ktmno (see the following entry).

Adiego (1994b:251).

ktmno
E.Th 25
PN in nominative, corresponding to Ekatomnvw (Zgusta KPN § 325–1–3,
Blümel KarPN:13) in Greek sources. Note the variant ∞tmño-.

Adiego (1994b:251–252).
376 CHAPTER ELEVEN

kt ?tri≤
E.Ab 40.
PN in genitive. Perhaps related to the PN kattÿri≤.

kuari≤bar
E.Me 18a
An unclear form. Perhaps it must be segmented into kuari≤ bar, the first
word being a PN in genitive (cf. the following entry kwar≤ ). However,
bar would remain unexplained.

kwar≤
E.Me 31
PN in genitive.

kudtubr
E.Th 9
PN in nominative. It apparently belongs to the family of names in
-(d)ybr-, but note the use of u instead of y/ÿ (like “odubr-, q.v.)

kwri≤
E.Th 34
PN in genitive. Perhaps it contains the same stem as kwar-, kuari≤bar.

∞arnos
E.Me 45
∞arnosd
C.xx 2
See alos, alosd.

∞arr≤
E.Ab 32
PN in genitive.

∞aye
E.Ab 31
PN in nominative.

∞diye≤
C.St 2
PN in genitive.
 CARIAN GLOSSARY 377

∞i
E.Sa 2: qÿri≤ ∞i; E.Me 6 (2×): parma≤≤ ∞i, klorul ∞i; E.Me 8b: armon
∞i, E.Me 9: arlio[m≤] ∞i; E.Me 10 (3×): q[---]≤ ∞i, [mw]don≤ ∞[i], [---
]w≤ord≤ ∞i; E.Me 12: mwdon≤ ∞i; E.Me 13 (2×): wet≤ ∞i, mwdon≤ ∞i;
E.Me 16 (2×): pikra≤ ∞i, mwdon≤ ∞i; E.Me 17: bem≤ ∞i; E.Me 18b (2×):
idmuon≤ ∞i, mdayn ∞i; E.Me 19: zmu≤ ∞i; E.Me 20 (2×): “rwli≤ ∞i,
mwdon≤ ∞i; E.Me 21: qÿblsi≤ ∞i; E.Me 23: a[rb]ikarm≤ ∞i; E.Me 25:
parpeym≤ ∞i; E.Me 28 (2×): pntmun≤ ∞i, mwdon≤ ∞i; E.Me 30: pleq≤
∞i; E.Me 31 (2×): kwar≤mHm≤ ∞i, mwdon≤ [∞]i; E.Me 32 (2×): ∞i en,
mw[d]on≤ ∞i; E.Me 33a (2×): myre≤ ∞i, mdayn ∞i; E.Me 33b: myre≤ ∞i;
E.Me 35: mwdon≤ ∞i; E.Me 38: ∞i ted; E.Me 40 (2×): pikrm≤ ∞i, mwdon≤
∞i; E.Me 42 (3×): mwsat≤ : ∞i, mwdon≤ : ∞i, tbridbd≤ : ∞i; E.Me 43a:
“rquq≤ ∞i; E.Me 43b: mno≤ ∞i; E.Me 44a: kojol ∞i; E.Me 44b: mwton≤
∞i; E.Me 45 (2×): [?]iam≤ ∞i, yi≤{∞}biks≤ ∞i; E.Me 46b: mwdon≤ ∞i;
E.Me 47: paraibrel≤ ∞i; E.Me 48: [-]owt≤ ∞i; E.Me 50b: p∞simt≤ ∞i;
E.Me 57: ]i≤ ∞i; E.Me 58: ]s≤ ∞i; E.Bu 6: ursea∞k ∞i; E.xx 1: isor≤ ∞i;
C.Eu 2: omob ∞i; C.Si 2a: pda∞m≤uñ ∞i

∞j
E.Me 36
Originally a relative pronoun, turned into a particle for introducing
complements. From PA *k wis < PIE *k wis (Hitt., CLuw. kui“, Lycian ti,
Milyan ki ). Most spellings point to a postclitical usage, with the excep-
tion of E.Me 32: ∞i en, where it seems to be proclitic.

See Adiego (1993a:213–216) for a brief status quaestionis and for a functional
comparison of ∞i with the Old Persian relative constructions. Compared to
Lycian and Milyan relatives in Adiego (1994a:46). See Hajnal (1997a) for a
more detailed treatment and here pp. 273–275.

∞i∞
E.AS 7
Meaning and function are unclear.

Schürr (2001:98) compares it to the Lycian indefinite pronoun tike (cf. also
Milyan -kike). While the comparison is sound from a phonological point of
view, the presence of an indefinite pronoun in E.AS 7 depends on the over-
all interpretation of the inscription, a question that remains unresolved.

∞iqud
E.Si 1
Very probably a PN in nominative.
378 CHAPTER ELEVEN

∞lbiks≤
E.Th 33
PN in genitive. It seems to include the nominal stem -biks- (cf. piks[,
dbiks, ÿ≤biks, yi≤{∞}biks-, but there are no clear parallels for the result-
ing first element ∞l-.

∞lmud [?
C.Ia 3
Perhaps an epithet of the word that it follows, the GN trqude ‘Tarhunt’,
although other interpretations cannot be discounted (for instance, a ver-
bal form).

The possibility of a word in agreement with trqude would increase if we accept

Gusmani’s proposal of reading a letter e after ∞lmud [ (therefore trqude ∞lmude,
with the same ending for both words).

∞?mpi
E.Si 10

∞tmño≤
C.Si 2a (2×)
PN in genitive. It is the typical Carian name Ekatomnvw (cf. the vari-
ant form ktmno), that here makes direct reference to the well-known
Carian dynast Hekatomnos, the father of Maussollos, Artemisia, Idrieus
and Ada.

Schürr (1992:137). See also Adiego (1994b). Neumann has repeatedly argued
in favour of a purely Greek origin of the name (a hypochoristicon of an
*ÑEkatÒmnhstow: Neumann apud Schürr 1993:137, n. 6; Neumann apud Adiego
1994b:248; Neumann 1994:17), but the existence in Carian of a noun mno-
‘son’ means that we can analyse the name as a Carian compound kt+mno (for
kt- cf. Akta-ussvllow, kbo-kt-, etc. see Adiego 1994b).

∞toi
C.My 1
∞toi≤
PN in nominative (∞toi) and genitive (∞toi≤ ). Tentatively compared to
ktai-, ÑEkata›ow.

See Adiego (2005:90–91). For the reading ∞toi≤, see Blümel (2005:188).
 CARIAN GLOSSARY 379

∞yrapai≤
C.Ka 2
∞yrpai
C.Ka 2
Two forms apparently belonging to the same paradigm: genitive (∞ura-
pai-≤ ) and perhaps nominative (∞yrpai ). It is not clear, however, if we
are dealing with a PN. Note the divergent vocalisation ºrpº/ºrapº.

∞[-]urb≤
E.Ab 36
PN in genitive

∞[--8--]tuñdñ[
C.Si 2a

limtaoa
C.Ia 1

lkor≤
E.Me 2, E.Me 36
PN in genitive

loubaw
E.Me 49
PN in nominative. It appears in the contentious inscription E.Me 49.

lrHñ
C.Si 2a

ltari≤
E.Ab 4, E.Ab 5 (ltari[≤])
PN in genitive

lÿ∞se
E.Th 35
lÿ∞si≤
E.Me 43a
PN in nominative and genitive. However, it is not totally clear if we
are dealing with two forms belonging to the same paradigm (which
380 CHAPTER ELEVEN

would imply a vowel-stem alternation ºe / ºi-), or two different stems

derived from the same root. In any case, the name seems to corre-
spond to the Carian name Lujhw in Greek sources (Zgusta KPN § 836,
Blümel KarPN:18). Cf. also Lu-uk-“u, the name of an Egypto-Carian
in Borsippa.

Adiego (1990a:134). On Lu-uk-“u, see Eilers (1935:217), (1940:192), and Waer-

zeggers, Borsippa, where the Egyptian origin of the Carians of Borsippa is
stated.

lys[ikl ]an
C.Ka 5
lysiklas[-?]
C.Ka 5
PN in accusative (lysikla-n) and in ‘genitive/possessive’ (lysikla-s(-?)).
Carian adaptation of the Greek name Lusikl∞w. It is not clear if lysik-
las is a complete form (of a true genitive in -s) or if a further letter
must be added. In this latter case, the most suitable solution is lysik-
las[n], a possessive adjective in accusative for expressing the name of
the father of Nikokl∞w in the inscription.

Frei-Marek (1997).

lysikratas[-?]
C.Ka 5
PN in ‘possessive’ (lysikrata-s(-?)). Carian adaptation of the Greek name
Lusikrãthw. As in the case of lysiklas[ (see the preceding entry), it is
not clear if the word is complete (representing a true ‘genitive’) or
whether it must be completed (most probably as lysikratas[n]) in order
to obtain a possessive adjective in accusative.

Frei-Marek (1997).

mal≤
C.Ka 1
PN in genitive.
 CARIAN GLOSSARY 381

mane
C.Hy 1a
mane≤
C.Hy 1a
PN in nominative and genitive. Typical Carian name that appears in
Greek as Manhw.

Adiego-Debord-Varinlio<lu (2005:607).

manon
C.Eu 2

mañ“qaraH≤rl-?- [
C.St 2

maãtnor
E.Th 34

maqly≤[
C.St 1
Perhaps a complete form. If so, it would be a PN in genitive.

marariso[- . . .]
E.Si 1
Apparently the beginning of a PN.

mdayn
E.Me 18b, E.Me 33a
mdaÿn
E.Me 11a, E.Me 11b, E.Me 17
Ethnic name (or similar) in nominative, the genitive of which is mwdon≤
(q.v.). Different possible interpretations have been envisaged, none of
them definitive: ‘foreigner’, ‘Carian’, ‘inhabitant of Myndos (a Carian
coastal city)’. In any case, the word seems to contain the suffix -yn-/
-ÿn- equivalent to the Luwic suffix for ethnic names *-weno/i- (cf. in
Carian kbd-yn-“ ).

For the paradigmatic relationship with mwdon≤, see Melchert (1993:82–83). If

this relationship is, as it seems, correct, former attempts to explain mdayn/mdaÿn
as a word for ‘husband/wife’ (Merigi 1980:35b–36a) or ‘farewell’ (Ray 1982b:184)
must be ruled out ( pace Adiego 1993a:219–220).
382 CHAPTER ELEVEN

mdot2
C.Ka 5
Function and meaning are unknown. The segmentation mdot2 un adopted
here is not definite.

In Adiego (2002:17–18) the word is analyzed as a plural genitive correspond-

ing to mdayn/mdaÿn, and a meaning ‘foreigner’ is assumed for all of these
forms. The sense of sarni“ mdot2 in C.Ka 5 would therefore be ‘representatives
(sarni-“ ) of the foreigners’, corresponding to Greek prÒjenoi).
Melchert (1998) prefers to read mdot2un as a complete word and, assuming
a glide-value (/w/) for the letter O (here transcribed as t2), analyzes the form
as a preterite first plural with the meaning ‘we have established’, and with
md º to be connected with Hitt. midà(i)- ‘fix, fasten’.

md [. . .]
E.Me 52
Perhaps to be completed as md[ay/ÿn].

md
C.Ha 1
See below mdane

mda
C.Si 2ª (3×)
See below mdane.

mdane
E.Sa 1, E.xx 7, C.xx 2
Analysis of this form has been much discussed (a verb or a chain of
particles?). Cf. also the possible variant mlane in Thebes (see mlane, ewm-
lane, ewlane).

On mdane and the two preceding forms md, mda see discussion in pp. 321–324.

me®≤
E.Me 34
PN in genitive.

meÿqak
E.AS 8
 CARIAN GLOSSARY 383

mi∞≤≤
E.Ab 35
PN in genitive.

mlane
E.Th 10
Cf. mdane, and see p. 323 on the intricate relationship between this
form and ewlane, ewmlane.

mlan[-?]
E.Th 35
Cf. the preceding entry.

mlqi≤
E.Th 27
PN in genitive.

mlne
C.Ia 3
Connection to mdane/mlane and/or ºmln in uiomln, yomln is possible,
but far from certain.

mmn∞al
E.Th 21

mnos
C.Eu 1 (mn[os?]), C.Ka 5
mno≤
E.Me 10, E.Me 12 (m[no≤]), E.Me 16, E.Me 27, E.Me 39, E.Me 43b,
C.Ka 1, C.Kr 1
mn[o-?]
E.Me 47

Common noun in genitive (mno-≤ ) and in a ‘s-case’ (mno-s), ‘son’. Perhaps

in some way related to HLuw. nimuwiza-, ‘son’. The precise analysis of
mnos in C.Ka 5 is far from certain, given the unclear context in which
it appears.

Meriggi (1967:223), (1980:35a), Ray (1982b:184–185), Gusmani (1986:63),

Adiego (1993a:216–219).
384 CHAPTER ELEVEN

moa[-]lboror
C.Ka 5

moi
C.My 1
PN in nominative. Perhaps it corresponds to the Carian name in Greek
sources, Moiw.

Blümel-Kızıl (2004:134).

mol“
C.Hy 1b
C.My 1
Plural nominative with the meaning ‘priests’?

See above pp. 306–207 on this interpretation, based on the analysis of C.Hy
1b and its possible correspondence to the Greek text that follows it.

mplat
E.Th 11
PN in nominative.

mqabaewleqo“osk$ioms
E.Th 12
An impenetrable sequence, undoubtedly consisting of more than one
word. Note the isolable sequence ew, to be related to ew lane, ewm, also
in Thebes.

mqt jq
E.Th 4

mrsi≤
E.Me 2, E.Me 26
Ethnic name or PN in genitive.

mrsj[. . .]
E.Me 54
Very probably related to the preceding entry.

mslmnlia
C.Ka 5
 CARIAN GLOSSARY 385

msnord≤
E.Me 3, E.Me 48
Ethnic name or, less probably, PN. Clearly related to the Carian place
name Masanvrada (Zgusta KON § 782, Blümel KarON:174). According
to Schürr, msnord-/Masanvradº can be analyzed as msn + ord, with a
second element comparable to Luwian -aradu in Tar¢unt-aradu (Laroche
LNH n. 1268, Piyam-aradu (Laroche LNH n. 981), so that msnord- =
Luw. *Ma““an-aradu-

Connection to Masanvrada already mentioned in Adiego (1990a:136). Note

that the word was formerly read as †msnori≤, which conditioned some pro-
posals of interpretation. For the corrected reading, see Schürr 2001b:103,
Schürr 2002:166–168).
As for Schürr’s analysis, see Schürr (2002:165–168). He interprets msnord-
as a PN and instead prefers to compare it with the name of the eponymous
founder of Masanorada, Masanvradow (Steph. Byz. s.v. Masanvrada).
On aradu-, see above p. 333.

msot
C.Hy 1b
Genitive plural (?) or another case from a stem mso- or similar = ‘god’?

See above pp. 306–307 for this interpretation, based on the search for par-
allels between C.Hy 1b and the Greek texts that follow it.

mt1yr
C.Ka 2

mt∞elã
C.Si 2a

mudo[n]≤
E.Me 65
mwdon≤
E.Me 10 ([mw]don≤), E.Me 12, E.Me 13, E.Me 14 (mwdon!≤), E.Me
16, E.Me 20, E.Me 28, E.Me 29, E.Me 31, E.Me 32 (mw[d]on≤), E.Me
35, E.Me 40, E.Me 42, E.Me 46b
Genitive of the ethnic name (or sim.) mdayn/mdaÿn, q.v.
386 CHAPTER ELEVEN

Interpretation as ethnic name already suggested in Meriggi (1980:35). Identification

of -on- with the Luwic ethnic suffix -wanni- and connection to the Carian place
name Mundow already noted in Adiego (1990b:501–502). See also Adiego
(1993a:210–212), Melchert (1993:82–83), Janda (1994:173–174).

mwk
E.Th 22

mumn“tnse-?
E.Th 30

muot
C.Hy 1a

-ot recalls identical endings in C.Hy 1 (kdu≤opizipususot, msot, cf. also ylarmit). If
the analysis of this latter is accepted (see s. v.), muot can be also a genitive plural.

mwsat≤
E.Me 42
PN in genitive. Cf. the Lydian name Mousathw (Zgusta KPN § 987a).
Perhaps both names, and also the Pisidian names Moushta, Moshta,
correspond to Luwian PN Muwaziti (Laroche LNH 840), a compound
of muwa- Hitt., Luw. ‘strength, force’, and Luw. ziti ‘man’.

Adiego (1992a:32), (1993a:233), (1994a:36).

mwton≤
E.Me 44b
Variant form of mwdon≤, q.v.

mute≤
C.St 2
PN in genitive. Cf. the Cilician name Moutaw (Zgusta KPN § 989–2).
Behind muº, the Anatolian stem muwa- can be identified.

Adiego (1994a:36).

mÿqudem
E.AS 7
 CARIAN GLOSSARY 387

myre≤
E.Me 33a, E.Me 33b
PN in genitive.

myze
C.My 1
PN in nominative.

Tentatively compared in Adiego (2005:91) to the Carian PN in Greek sources,

Mouzeaw.

mHm≤
E.Me 31
PN in genitive.

m[-]sao[
C.My 1

naria≤
E.Me 5
PN or title in genitive. If it is a personal name, it must be the father
of psm“kúneit in the bilingual text E.Me 5, which implies that this man
had a double denomination, Egyptian W3˙-jb-r‘-nb-[ (in the Egyptian
part) and Carian Naria-. However, it could instead be a title of Psm“kúneit
(‘general’, ‘priest’ or similar). Possibly related in some way to the fam-
ily of place names Naras/a/, Narisbara, Naruandow (connected to
CLuw annarai-, ‘forceful, virile’ = Hitt. innarà- ‘forceful, violent’, all
derived from PIE *h2nè́r ‘man’).

For these and other possible examples of this stem in Carian onomastics, see
p. 333.

naz
E.AS 7

ne
E.AS 7

nid≤kusas
E.AS 8
388 CHAPTER ELEVEN

nik[--]lan
C.Ka 5
PN in accusative, to be completed nik[ok]lan, nik[uk]lan, or similar. It is
the Carian adaptation of the Greek name Nikokl∞w.

Frei-Marek (1997).

ninut
E.Ab 20
PN in nominative.

niqau≤
E.Me 18a
PN in genitive. Very likely to be a Carian adaptation of the Egyptian
pharaonic name Nechao/Necho (Ny-k3w, Greek Nexvw).

Adiego (apud Schürr 1996a:63, n. 11).

noril ?ams
C.Kr 1

not
C.xx 2
Verbal form (‘he brought’)?

This is the interpretation proposed in Adiego (2000:153–155), where not is

analyzed as a preterite third singular verb, to be connected to Hitt. nà(i)-,
CLuw. (reduplicated stem) nana- < * PIE *neyH- ‘to bring’, see above p. 284.

nprosn≤
E.Ab 16
Apparently a complete PN in genitive. However, Schürr has proposed
a segmentation npro + sn≤, interpreted as a PN (nominative) + PN
(father’s name in genitive). As for the assumed first name, he compares
it with the Egyptian name Nfr-˙r Nefervw (DNb:641), phonetically
[nefer˙ó] according to Vittmann. For sn≤, Schürr provides the same
form in E.AS 8 and allegedly in C.Kr 1 (read and segmented differently
in this case). However, although the explanation of npro is very sound
(see also Vittmann), the existence of a Carian name *sn- is doubtful.

Schürr (1996a:68, n. 18), Vittmann (2001:42).

 CARIAN GLOSSARY 389

»ßw˚n
E.Th 30

n≤
C.Kr 1

See a very hypothetical explanation (as pronoun) in p. 292.

n≤n[-]s“
E.SS 1

ntokris
E.Me 35
(Presumably) feminine PN in ‘s-case’, a Carian adaptation of the Egyptian
fem. name Nj.t-jqr (literally ‘Neith is perfect’), Greek Nitvkriw (DNb:628).
This was the name of a daughter of Psammetichus I.

Schürr (1992:152, n. 9), Ray (1994:202–203). Vittmann (2001:52–53) argues

convincingly (against Ray loc. cit.) that ntokris arrived in Carian directly from
Egyptian, without Greek intermediaries. This implies that in ntokris, the final
s must be a case ending, not a vestige of the Greek sigmatic nomimative (see
above p. 315).

ntro
C.xx 1
ntros
E.xx 7
Carian God name, assimilated to Greek Apollo, in dative (ntro) and in
s-case (ntro-s) or, less probably, a priest title (in nominative ntro and in
s-case ntro-s), derived from a god name ntrº = Apollo. Independently of
either interpretation, there is a general consensus that ntro- should be
connected to Lycian Natr- in the PN Natr-bbijẽmi, ‘translated’ in Greek
as ÉApollÒ-dotow.

On these two alternative interpretations (the latter defended in Melchert 2002),

see above p. 282. Carruba (2002) rejects the equation Natr-/ntro- = Apollo,
and argues for a generic meaning ‘god’ and for an Egyptian loanword in
Carian (from Egypt. ntr ‘god’).
390 CHAPTER ELEVEN

nu
E.xx 7
Function and meaning are unknown.

Melchert’s tentative translation of E.xx 7 (see Melchert (2002:308) implies the

assumption that nu is an adverb, ‘now’ (to be compared therefore with Hitt.
nu-). See above p. 286, where an intepretation as a verb is suggested.

nuol$∞[---]sarmrol∞yt
E.Me 4

Perhaps more than one word, but the lacuna does not allow us to iso-
late words in the sequence.

Schürr (1992:155) claims to identify two “titles in -ol”, nuol and sarmrol (cf.
also his similar analysis of kojol). While the repeated ending in -ol could be
good evidence for segmentation, no parallel forms, either to nuol or to sarm-
rol, are attested in Carian.
Final sequence ∞yt could point to a verbal form, see above p. 276.

n[-]eakrnanb
E.Si 3

ñmailo
C.Si 2a
A sequence containing a verbal form?

Interpreted as a verb ñmail + enclitic pronoun -o- in Adiego (2000:141–142),

where it is further connected to Lycian m̃maite ‘they have established’), see here
on p. 304.

oba
C.Ka 2

obrbi≤
C.My 1 (2×)
PN in genitive.

obsmns[
C.Ka 2
 CARIAN GLOSSARY 391

obsmsmñ1ñ
C.Ka 2

omob
C.Eu 2

or
C.Ka 6
PN?

orbá
E.Th 20

orkn
C.Ha 1, C.xx 1
Noun in singular accusative, of a stem ork- ‘phiale’, ‘vessel’, or similar.

Melchert (1993:80). Neumann and Edwin Brown (apud Melchert, ibid.) suggest
a comparison with Gr. Ïrxh ‘jar’, Lat. orca ‘butt, tun’. Cf. also Lat. urceus, urna
‘water-pot’ (all these words probably share a common origin).

or≤
E.Ab 15, E.Me 41, E.Th 36
PN in genitive. According to Vittmann, an adaptation of the Egyptian
name Ór ‘Horus’ (phonetically [hòr])

Vittmann (2001:42). However, cf. or in Kaunos.

or“a
E.xx 7

ort
C.Ka 5
ort[-]
C.Ka 5

oru
C.Ka 5
392 CHAPTER ELEVEN

otonosn
C.Ka 5 (2×)
Ethnic name in accusative, ‘Athenian’. Derived from a Carian form
*otono- for ‘Athens’ by means of a -s- suffix.

Much has been discussed about the precise nature of the -s- suffix. It could be
a possessive suffix (= Lyc. -h-; also Frei-Marek 1997:35), an ethnic suffix equiv-
alent to Lyc. -s- (Schürr 1998:161; similarly Hajnal 1997b:160, but resorting
to Lyc, -is-) or, less probably, an ethnic suffix equivalent to Lycian -zi (cf.
Hajnal 1997b:160–161, n. 32, where the difficulties of this explanation are
highlighted). All three hypotheses are envisaged in Adiego (1998a:18).

otr“
C.Ka 2, C.Ka 5
Pronoun in acc. pl., corresponding to Greek aÈtoÊw in the bilingual
inscription C.Ka 5. It can be connected etymologically with Lyc. atla-/
atra- ‘person, self ’ (cf. also HLuw. atra/i- ‘person; image; soul’.

This interpretation and etymological connection was established independently

by various scholars: Adiego (1998a:21), Hajnal (1997b:164; 1998:102), Melchert
(1998:34), Neumann (1998:20)—who very appositely recalls the correspondance
Lycian atru: Greek •autÒn in TL 25a, 4. See here on p. 297.

owdown[. . .]mwarudk≤o
E.Th 10
A chain probably containing more than one word.

The initial sequence owdown hardly seems to be separated from wdwn, q.v.

ow∞meb≤t
E.xx 5

ouor
C.Ka 2 (2×)

pals≤
E.Ab 7, E.Ab 8, E.Ab 9
PN in genitive.
 CARIAN GLOSSARY 393

panejt
E.Ab 2
PN in nominative. Carian adaptation of the Egyptian name P3-n-Nj.t,
Greek Panitiw, literally “the one of Neith” (DNb:385). Cf. the variant
form pneit)

Schürr (1992:152, n. 9), Adiego (1993a:254), Ray (1994:203 and n. 19). Cf.
also Vittmann (2001:58).

paraeym
E.Me 8a, E.Me 8b (para!eym)
PN in nominative. The name presents the well-known adverbial stem
para- as a first element of a compound (cf. para-ibrel, Para-ussvllow,
etc.). As for -eym, it recalls Arthumow. Compare also parpeym-.

paraibrel≤
E.Me 47
PN in genitive. A compound name consisting of para- (cf. above paraeym)
and ibrel (= Greek Imbarhldow).

Adiego (1994:36–37). On ibrel-, a stem derived from *ibr- = CLuw. im(ma)ra/i-,

see above p. 335.

pareÿs
C.Kn 1

parma≤≤
E.Me 6
PN in genitive.

parãaq ?
C.Tr 2

Adiego (1993a:263), Hajnal (1995[97]:20). See the discussion of the reading

and possible interpretations of this word in pp. 289–291.
394 CHAPTER ELEVEN

parpeym≤
E.Me 25
PN in genitive. It is not clear if we are dealing with a compound with
par(a)- as a first element, or whether in fact a stem parp- should be rec-
ognized. For the final part of the word, cf. paraeym-.

For parpº, see s. v. prpwri∞.

par≤olou
E.AS 1

paruos≤
C.My 1
PN in genitive. Cf. the Carian name (f.) Paruv.

Adiego (2005:91).

parÿd∞≤
E.SS 1
PN in genitive. It corresponds to the Carian name Paraudigow in Greek
sources. A compound name par- (cf. para-ibrel, Para-ussvllow, etc.) +
ÿd∞-

Adiego (1994:43). For the family of names containing the stem yri∞-/yriq-
/ÿdiq/ÿd∞-, see pp. 262–263.

paryri∞
C.My 1
paryri∞≤
C.My 1
PN in nominative and genitive. A compound name: par- + yri∞-. See
the preceding entry.

pau
C.Hy 1a, C.My 1
p ?au
C.My 1
 CARIAN GLOSSARY 395

pau≤
C.Tr 1, C.Tr 2
PN in nominative (pau) and genitive ( pau≤ ) corresponding to the Carian
name transcribed in Greek as Paow.

Adiego (1994:37).

pa[-]in[-]t≤
E.Ab 17
Apparently a PN in genitive.

pd[
E.Me 64a
Cf. the following entries.

pdnejt
E.Sa 2
PN in nominative. Carian adaptation of the Egyptian P3-dj-Njt, liter-
ally ‘the one whom Neith has given’ (Greek Petenaiyiw, Petenhyiw,
Petenht, see DNb:316).

Adiego (1992a:29–30).

pdtom≤
E.Bu 2
NP, a Carian adaptation of the Egyptian name P3-dj-Jtm, literally ‘the
one whom Atum has given’, Greek Peteyumiw, Petetumiw (DNb:294).

Schürr (1992:152, n. 9) and apud Ray (1994:205); Vittmann (2001:58).

pduba
E.xx 4
PN in nominative.

pdubez
E.Ab 15
PN in nominative. Carian adaptation of the Egyptian name P3-dj-B3st.t,
literally ‘the one whom Bastet has given’, Greek Petyubestiow, Petobastiw,
Petoubastow, Petoubestiw (DNb:303).
396 CHAPTER ELEVEN

Identification proposed in Schürr (1996a:60), where the word is still read as

†pdubÿz. Corrected reading given in Schürr (2000:172). See also Vittmann
(2001:42). Cf. also the corresponding feminine theophoric name ttbazi, ttubazi.

pdubi≤
E.Me 10 (p∂uüi≤), E.Ab 6
PN in genitive. Cf. pduba. It is possible that we are dealing with two
forms of the same paradigm, cf. the similar situation in (lÿ∞se/lÿ∞si≤ ).

pda∞m≤uñ
C.Si 2a
An accusativus genetivi in agreement with the PN pñmnn≤ñ that precedes it?

See Adiego (2000:144–148) for this morpho-syntactic analysis and for some
attemps at etymological explanations (particularly the hypothetical connection
of pd º with Lyc. pddẽ ‘place’, cf. also here on p. 304).

pjabrm
E.Me 12
PN in nominative. According to the illustration of the stela in which
it appears, the name is feminine.

pjdl ?
C.xx 1
Noun in apposition to acc. sg. orkn? Other interpretations are also
possible.

The meaning and function of this word have been much discussed. Melchert
(1993:80–81) interpreted it as a word ‘gift offering’ (from a *piyodhlom, to be
related to the stem CLuw. piya-, Lyc. pije- ‘to give’), a view followed here (see
p. 282). Totally different is the approach of Janda (1994:178), who preferred
to see here a verb comparable to Lyd. bill /pill/ (< *pid-l ) ‘he has given’.

pidaru[
C.St 2
PN, probably to be completed pidaru[≤ ] in genitive. A possible Carian
adaptation of the Greek name P¤ndarow.

Adiego (1994a:39–40).
 CARIAN GLOSSARY 397

piew
E.Ab 38
PN in nominative. Adaptation of the Egyptian name P3-n-jwjw (liter-
ally ‘that of the dog’), Greek Pieuw, Pihow, Pih# (Cf. DNb:349).

See Vittmann (2001:44) for the identification and for further details on the
Egyptian variants of the name.

pikarm≤
E.Me 14
PN in genitive. It is equivalent to the Lycian names in Greek sources
Pigramiw, Pigramow (Zgusta KPN § 1255–1/2). The name contains the
same stem as pikra-/pikre-, q.v., to which a m-suffix has been added. Cf.
also the variant form pikrm≤, and the compund name dbikrm, dbkrm
(d +bikrm = pik(a)rm-).

Adiego (1992a:36), (1993a:233), Kammerzell (1993:19, 22).

pikra≤
E.Me 16
pikre≤
E.Me 3
PN in genitive. It is not clear if we are dealing with a simple alternance
a/e or with two different stems, one in ºa- and the other in ºe. The name
appears in Greek sources as Pigrhw/Pikrhw, a very spread Anatolian name.
The name shows the stem pik- = Luw. *pi¢a-.

Adiego (1992a:36), (1993a:228–229). On the family of names built on this

stem, see p. 337.

pikrm≤
E.Me 40
PN in genitive. A variant, ‘devocalised’ form of pikarm≤, q.v.

piks[
C.St 1
A PN or part of a PN. The part conserved clearly contains the nom-
inal element piks-/biks- Cf. dbiks, ÿ≤biks-/yi≤{∞}biks-, derived from the
stem pik- = CLuw. *pi¢a-, as pikre-/pikra-, pik(a)rm-.

On this stem, see p. 337.

398 CHAPTER ELEVEN

pim[. . .]
C.Si 2b

pisiri
E.Ab 1
PN in nominative. Very likely to be an adaptation of the Egyptian
name P3-n-Wsjr, literally ‘the one of Osiris’ or P3j-Wsir, ‘this of Osiris’,
Greek Pisiriw.

Ray (1994:203); cf. also Schürr (1996a:61–62), Zauzich (apud Schürr, ibid.),
Vittmann (2001:58). This Egyptian interpretation of the name must be pre-
ferred to former attempts to connect this name with Anatolian proper names
(Adiego 1993a:248 and, with many reservations, 1994:43).

pisma≤k
E.AS 7, E.AS 3
PN in nominative. Carian adaptation of the Egyptian name Psmtk,
Greek Cammhtixow. See also the devocalised form psma≤k.

Identification already proposed in Kowalski (1975:91).

pisma[“/≤k . . .]
E.AS 4
The same name as the preceding entry. Doubts about completing “ or
≤ arise from the existence of psm“k(wneit), as well as pisma≤k.

pisñoi
C.Si 2a
Analyzed as a verb pisñ + clitics (-o-i ), in Adiego (2000:141), where pisñ
is identified as a preterite third plural of a root pi-, ‘to give’. See above
p. 304 for details.

p.iub[a]Ωi≤
E.Me 1
PN in genitive.
Probably the Carian adaptation of an Egyptian name whose final part
contained the name of the goddess Bastet (B3st.t), although no exact
parallels for the whole name can be found.

Schürr (1996:62). For ‘Bastet-names’ in Carian, cf. ttubazi-, ttbazi-.

 CARIAN GLOSSARY 399

piubez
E.Ab 10
PN in genitive. It seems to be a variant of the preceding entry.

p∞simt≤
E.Me 50b
PN in genitive. Tentatively compared by Schürr to the Egyptian name
P3-dj-Ór-sm3-t3wy (shorter form P3-dj-sm3-t3wy), literally ‘one whom
Horus, uniter of two lands, has given’, Greek Potasimto.

Schürr (apud Ray 1994:205). Not included by Vittmann in his list of Egyptian
names in Carian sources (Vittmann 2001:58–59). Although the use of this
Egyptian name among the Carians of Egypt fits well with Egypto-Carian envi-
ronment (Potasimto was the name of the commander of the Carian and Ionian
mercenaires in the Nubian campaign of Psammetichus II), and the similari-
ties between Carian and Greek adaptations are striking, the use of Carian ∞
for Egyptian d or dj is surprising (compare the use of Carian d or t for Egyptian
d or dj in other names that also include the Egyptian verb dj, ‘to give’: pdnejt,
pdtom, ptnupi, etc.).
An Anatolian interpretation, by connecting p∞siº with Pija-, Pije-, Pijo-:
Lycian Pijaw (Zgusta KPN § 1263–1), Pije-darow (KPN § 1263–2), Carian
and Lycian Pijodarow) is given in Adiego (1993a:248).

pla?t
E.Th 3
PN in nominative. If the reading plat is accepted, it could be a variant
of the name plat q.v. (l instead of l is typical in Theban inscriptions).

plqo
E.Me 40
PN in nominative. It appears in Greek sources as Pellekvw, Pelekvw.
Note particularly Peleqow in the Greek graffiti of Abu-Simbel.

Adiego (1993a:234), Schürr (1991–1993:170). On Peleqow = Pel(l)ekvw (against

former interpretations as a Greek name derived from p°lekuw), see Adiego
(1994a:37), Masson (1994b:140), (1995:175).

plqodse
E.Th 52
PN? It seems to include the name plqo, but final ºdse remains impossible
to analyse.
400 CHAPTER ELEVEN

plat
E.Ab 7, E.Ab 8, E.Ab 9
PN in nominative.

platt
E.AS 6
PN in nominative. Perhaps related to the preceding entry, but the final
-t remains unexplained.

pleq≤
E.Me 30
PN in genitive. It corresponds to the Carian name in Greek sources
Peldhkow.

Adiego (1993a:234), (1994a:37).

pneit
E.SS 1
PN in nominative. Variant form of panejt (q.v.), an Egyptian name.

Schürr (1992:152, n. 9), Adiego (1993a:254), Ray (1994:203).

pnld≤wl
E.Me 49

About the peculiar inscription where it appears, the reading of which is very
difficult, see p. 279.

pntmun≤
E.Me 28
PN (or title?) in genitive. According to Vittmann, it could be the adap-
tation of Egyptian p3 ˙m-n∆r n Jmn *[ p˙ent6n6mùn], literally ‘the Prophet
of Amun’. Among other possibilities, Vittmann suggests that it could
be a title (therefore sanuq“ ue pntmun“ ∞i mwdon≤ ∞i: “Stela of sanuq, who
(was) the ‘Prophet of Amun’ (a priest title), who (was) mwdon.”).

Vittmann (2001:46–47). The fact that the three names of E.Me 28 appear
in genitive makes the structure of the inscription very ambiguous. See Vitmann
(2001:47) for different possible analyses. Given this ambiguity, I prefer to inter-
pret the inscription as a N-≤ (PN)—N-≤ (father’s name)—mwdon-≤ (ethnic name
probably referring to the father).
 CARIAN GLOSSARY 401

pnu≤ol
E.Th 40
pnw≤ol
E.Th 27
pnu≤ol
E.Me 19
PN in nominative. It appears in its Greek adaptation as Ponussvllow.
Note the variant form for the genitive, punw≤ol≤.
A name of the u≤ol-family (q.v.).

Adiego (1990a:135). On the Anatolian origin of the element p(u)n-, see pp.
337–338.

pnu≤o≤
C.My 1
PN in genitive. No parallel form is found in Greek sources, but it is
possible that the word should be corrected as pnu≤o<l>≤, so that the
name would be the same as that of the preceding entry.

Adiego (2005:84).

pnyri≤ru
E.AS 5

pñmnn≤ñ
C.Si 2a
PN apparently in ‘accusativus genetivi’. The name appears in Greek
as Ponmoonnow (documented also in Sinuri).

Schürr (1992:138).

polo
E.Me 8b
PN (?) in nominative. Coordinated with the PN in nominative paraeym
by means of sb, ‘and’.

Melchert (1993:84) suggests that polo is a common noun representing a kin-

ship relation. Although this possibility cannot be ruled out, this tentative pro-
posal of meaning and etymology (‘son’, comparing Hitt. pulla- ‘child, son’) is,
as Melchert himself recognizes, very speculative.
402 CHAPTER ELEVEN

potko≤l≤ ?
C.Ka 8
If the reading of the final letter is accepted, it is apparently a genitive
(of a PN?).

p?owk
E.Mu 1
PN in nominative.

prna∞non
E.Th 34

A form in accusative sg.? It seems to agree with another word ending in -n

(dm-?-n).

prãidas
E.xx 7
A word with s-ending, in agreement with ntros ‘Apollon’, q.v. Tentatively
connected by Schürr with Bragx¤dai ‘Branchids’, the priestly family in
charge of the sanctuary of Apollo in Didyma, near Milet.

Schürr (1998:158).

prpwri∞
E.Th 46
PN in nominative. Apparently a compound name that can be seg-
mented as prp + wri∞. For the first element, cf. perhaps the Lycian name
Perpenduberiw (Zgusta KON § 1242–1) or even Carian parpeym-. The
second element seems to be a variant of ºyri∞, also present in other
compound names (idyri∞-, paryri∞-).

For a possible explanation of the alleged use of w instead of y, see p. 105.

psikro≤
E.Me 51
PN in genitive.
 CARIAN GLOSSARY 403

psma≤k
E.Th 11 (psma≤[k]), E.Si 7, E.Bu 4, E.Bu 5
psma≤k≤
E.Si 2, E.Bu 1
psma[≤/“k . . .]
E.Me 55
PN in nominative ( psma≤k) and genitive ( psma≤k-≤ ). Carian adaptation
of the Egyptian name Psmtk, Cammhtixow. Cf. the variant form pisma≤k.

psm“kwneit≤
E.Me 5
PN in genitive. Carian adaptation of the Egyptian name Psmtk-‘wj-Njt,
literally ‘Psammetichus in the arms of Neith’. Carian and Egyptian
forms appear together in the bilingual text E.Me 5.

Adiego (1992a:29–30).

psnlo
C.Ha 1

On the different possible analyses of this word, see p. 284.

psoir≤
C.My 1
PN in genitive.

psrkrte
E.Th 30

psu≤ol≤
C.Ka 1
PN in genitive. It belongs to the family of names in u≤ol/Ussvllow,
but the remaining first element ( ps-) is not clear.

psÿ“ [|?]
E.AS 7
Reading and segmentation (suggested by Schürr) are very doubtful.

See the entry ai[-]iqom for an alternative analysis.

404 CHAPTER ELEVEN

psHÿm[-]≤
E.Me 27
PN in genitive.

ptn“e
E.Ab 3
PN in nominative.

ptnupi
E.Me 18a
PN in nominative (?) of Egyptian origin: P3-dj-Jnp, Greek Petenoupiw
(literally ‘the one whom Anubis has given’ (DNb:27).

Schürr (1992:152, n. 9). Ray (1994:204). Ray (ibid.) also offers an alternative
Egyptian explanation, starting from *P3-dj-nfr (literally ‘the gift of the good
one’), but this name, as he recalls, is not documented in Egyptian.

ptnuq ?i ?
E.Ab 26
PN in nominative of Egyptian origin, according to the new reading
proposed by Vittmann: ptnuqi would be Egyptian P3-dj-‘nq.t (literally
‘the one whom [the goddess] Anukis has given’, DNb:294), phoneti-
cally interpreted as [pe†e‘anùqi].

Vittmann (2001:44). Vittmann does not rule out an alternative reading ptnuti,
which also has a good correspondence in Egyptian: P3-dj-(p3)-ntr (literally ‘the
one whom the god has given’), Greek Petepnouyiw, DNb:306 (phonological
reconstruction: [pe†e(p)nùte)].

pttu≤
E.Me 27
PN in genitive. Tentatively interpreted as an adaptation of an Egyptian
name, which is not in fact documented, *P3-dj-t3wy, literally ‘One whom
the two lands have given’.

Ray (1994:205); Vittmann (2001:58).

 CARIAN GLOSSARY 405

punm[-]≤
E.Me 65
PN in genitive. It seems to be a compound name with pun- as the first
element.

See Schürr (2003a:95), who goes a little further and compares it with Lyc.
Punamuwe, Ponamoaw (Zgusta KPN § 1288–1), and Punamu(W)aw (KPN § 1288–2,
Pamphylia); cf. also Puna-A.A (= *Punamuwa) in cuneiform sources (Laroche
LNH: 1050). These latter comparisons depend on a restitution punm[u]≤.

punot2
C.Ka 2

See Adiego (2002) for a connection with Luwic puna-, ‘all’, and for a mor-
phological interpretation as plural genitive.

punw≤ol≤
E.Me 21
PN in genitive. It is the same name as pnu≤ol (and variants, equivalent
to Ponussvllow), q.v.

puor≤
E.Bu 6
PN in genitive. As according to Vittmann, an Egyptian name adapted
in Carian: P3-whr (literally ‘the dog’), Greek Povriw, Pouvriw (DNb:181)
pronounced [puhór].

Vittmann (2001:41).

pur?i≤
C.Hy 1a
PN in genitive.

purmoruos
C.Ka 5

p[-]lu≤
E.Ab 33
PN in genitive.
406 CHAPTER ELEVEN

=q
C.Hy 1
Connecting particle? It appears after armotrqdos, q.v.

Adiego-Debord-Varinlio<lu (2005:617).

qan
E.Si 2

See the somewhat risky interpretation in Schürr (2000:171) as a word with

the meaning ‘dog’ and borrowed from Lyd. *kãn- (also ‘dog’), based on the
hunting scene drawn under the graffito E.Si 2.

qanor
E.Th 34

Connected with the preceding entry in Schürr (2000:172).

#q#arm≤
E.Me 10
PN in genitive.

qarpsi≤
E.Me 36
Ethnic name (less likely to be PN) in genitive.

On the structure of the inscription, which suggests an ethnic name, see pp. 267–
271. Note also the suffix -si- or -i- that appears in other possible ethnic names
(kidbsi-, q ÿbsli-, yiasi- ylarmi-). Among the possible Carian place names that
could be connected with qarpsi- (see Appendix B), Karbasuand/a/, Karpasuand/
a/ is a good candidate.

qarsio[-?]
E.Me 7

qdar®ou≤
E.Me 41
PN in genitive.

See Adiego (1995:24–25) for an attempted etymological analysis (compared with

CLuw. ¢utarlà- ‘slave, servant’; cf. the PNs in Cuneiform sources ›u-da-ar-lá,
›u-du-ur-lá, ›u-u-tar-li, ›u-ut-ra-la-(a“), ›u-ut-ra-li-i“ (Laroche LNH n. 411).
 CARIAN GLOSSARY 407

qeb≤t
E.Th 12

qku
E.Si 6

q∞blio≤
E.Ab 40
PN in genitive.

qlali≤
E.Me 37
G 2
[q ? ]lalis
E.Me 45
PN in genitive (qlali-≤ ) and in ‘s-case’ ([q]lali-s). This name appears in
Greek sources as Kolaldiw, Kulaldiw.

Adiego (1993a:235)

qorb
E.xx 1
PN in nominative.

qot2omu
C.Kr 1

qrds
C.Ki 1, C.Ka 2
Word with a possible ‘institutional’ meaning, given its appearance in
two legal texts. Cf. the following entry and also grdso[-]i [.

In Blümel-Adiego (1993:94) this is tentatively compared with Hitt. gurta-, ‘castle,

citadel, acropolis’ (cf. also the place-name of the Phrygian-Carian borderland
Gordio/n/). Melchert (1998:35, n. 2) suggests connecting it with Mil. kridesi, a
place-name.

qrdsol“
C.Ka 2
Acc. (or nom.?) pl. of a stem qrdsol-, apparently a noun derived by
means of the suffix -ol- from qrds: ‘belonging to the qrds’(?).
408 CHAPTER ELEVEN

Blümel-Adiego (1993:94), Melchert (1998:35). The sequence it appears in, qrd-

sol“ ait, could be interpreted as ‘they have made them belonging-to-qrds’ or
‘the belonging-to-qrds ones have made.’ (cf. Melchert ibid., Adiego 1998a:22).

qtblem≤
C.xx 1
PN in genitive. Name corresponding to Kotbelhmow (Blümel KarPN:17),
Kutbelhmiw (Zgusta KPN § 771, Blümel KarPN:18).

Adiego apud Schürr (1992:142), Adiego (1993a:235), Melchert (1993:78).

qtblo
E.Th 10
If PN (in nominative), it must be the Carian name adapted in Greek
as Kotobalvw.

Schürr (apud Adiego 1994a:43). The doubts about its character as PN are the
result of the very unclear context in which the form appears (see the remarks
in Adiego ibid.).

quq≤
E.Me 17
PN in genitive. It is the Carian name that appears in Greek sources
as Gugow. Cf. also the compound names dquq, “rquq.

Adiego (1993a:235, 1994a:37). On the possible Anatolian etymology of quq,

see p. 334.

qurbo≤
E.Ab 10
PN in genitive.

qwsal
E.Th 12

qutbe
E.Th 8
PN corresponding to the Carian name in Greek sources, Kuatbhw.

Adiego (1993a:235), (1994a:37).

 CARIAN GLOSSARY 409

q ÿblsi≤
E.Me 21
Ethnic name or, less probably, PN. If an ethnic name, it clearly recalls
the place name Kublissow (Zgusta KON § 1296, Blümel KarON:171).

See somne≤ (name to which q ÿblsi≤ is referred) for the curious coincidence of
personal name and ethnic name in the sole example of Svmnhw in the Greek
sources.
Comparison with the ethnic name (but still taking it as a PN) in Adiego
(1993a:235). Analysis as an ethnic name (with the same connection) in Janda
(1994:174).

qyrbmudolo
C.Eu 2

q ÿri≤
E.Sa 2
PN in genitive. It appears adapted in the Egyptian part of E.Sa 2 as
K3rr.

For the identification of qÿri≤ with K3rr: Schürr (1992:135), Adiego (1993a:161).
Egyptian k3 (written by means of a biconsonantic sign) is used here to reflect
a syllabic sound /ku/, so *kurº (Vittmann 1996). Note that in Egyptian a vowel
/y/ did not exist, so that the use of /u/ for /y/ seems reasonable. The dou-
ble r is explained by Zauzich (apud Schürr 1996:68) as a graphic attempt to
emphasize that r did sound (the final r was not pronounced in Late Egyptian).

qzali
C.My 1
qzali≤
C.My 1
PN in nominative and genitive.

Connection with the Carian name of Greek sources Kostvlliw (Adiego 2005:91)
is very hypothetical!

q[---]≤
E.Me 10
PN in genitive.

rdudmm»≤
E.Th 42
410 CHAPTER ELEVEN

rqemw
E.Th 52

rsy
E.Lu 2

rtim
C.Hy 1a
PN in nominative. Cf. the Carian name Artimhw and, more generally,
the family of Anatolian names collected in Zgusta KPN § 108): Arteimaw,
Arteimianow, Arteimow, Arteimhw, Artimaw, Artimhw, etc.

Adiego-Debord-Varinlio<lu (2005:611). On the possibilty that rtim and artmi

(q.v.) could be part of a single paradigm, see above p. 290.

r1i
C.Ka 4

sa
E.Me 26
Demonstrative pronoun in nominative (‘this’), which appears following
the word upe ‘stela’: upe sa: ‘this stela’. From PIE *∞o-/*∞eh2- > PA *∞o-/
∞à- (> Hitt. ka-, CLuw., HLuw. za-). See also san, snn.

Adiego (1992a:33). Hajnal (1995[97]:23) suggests that the same form can be
recognized in the sequence bid≤lemsa.
On these pronominal forms, see pp. 319–320.

sa?awon
E.Mu 1
saawon
E.Mu 1

san
G 1
Demonstrative pronoun in nominative. It corresponds etymologically to
Hitt. ka-, CLuw., HLuw. za- ‘this’, see above sa.

Adiego (1992a:33). On the final n, see Melchert (1993:79–80) and here on

pp. 288, 320.
 CARIAN GLOSSARY 411

sanuq≤
E.Me 28
PN in genitive.

sarl ?
E.Me 5

sarni“
C.Ka 2 (2×), C.Ka 5 (sarni[“])
Plural accusative of a stem sarni-. This word (or the wider sequence it
appears in) corresponds to Greek proj°nouw in the Kaunos bilingual
(C.Ka 5).

Frei-Marek (1997:39). There are diverse etymological proposals, none of them

definitive: connection to CLuw. “arri ‘above; up; for (?)’ (Frei-Marek, ibid., fol-
lowed by Hajnal 1997b:164, hard to accept, because *“arni“ would be expected,
as they themselves recognize); related to CLuw. zar“iya- ‘safe-conduct, Gastrecht’
or sim. (therefore sarni-, ‘guarantor’, Neumann 1998:29). Schürr (apud Hajnal
1997b:164, n. 35) compares sarni- with Lyd. saretas ‘benefactor(?)’.
In Adiego (1998a:22) the equivalence to proj°nouw is seen in the whole
sequence sarni[“ ] mdot2 ‘representative of the foreigners’ (with mdot2 [gen. pl.]
related to mdayn/mdaÿn/mwdon-, interpreted as ‘foreigner’).

sb
E.Me 8b, E.Th 13, E.xx 6, C.Si 2a (2×), C.Ka 2 (8×), C.Ka 5 (8×),
C.Kr 1
Coordinative conjunction: ‘and’. When there is interpunction in the
text, it always appears attached to the following word, as a sort of
proclitic.
Cf. Lycian se and particularly, Milyan sebe (both ‘and’). From PIE
*∞e (cf. Venetic ke ‘and’), plus a reinforcing particle *-be?

For sb = ‘and’, see the explanation already offered in Schürr apud Ray
(1990b:129–130). Connection with Milyan sebe: Neumann (1993:296). For se-
< PIE *∞e: Adiego (1995:31–32).
412 CHAPTER ELEVEN

sdi
C.Tr 1, C.Al 1
sdisas
C.Ka 1
sdisas ?
C.Kr 1
Noun used in funerary contexts (therefore ‘tomb’, ‘stela’ or sim.) The
morphological analysis of these forms remains unclear. Cf. the variant
form sidi.

Connected with PIE *∞ei- ‘to lie’ (Lyc. sije-)?

semw≤
E.Me 16
PN in genitive.

seqqejewsk
E.Th 4

sidi
C.Tr 2
A variant form of sdi, q.v.

siral
E.Me 49

siyklo≤
C.Ia 3
PN (?) in genitive,

skdubrotoz≤
C.My 1
A sequence containing an onomastic formula PN-Ø PN-≤, the main
difficulty being the point of segmentation. The best solution seems to
be skdu brotoz≤, but other alternatives cannot be dismissed.

slaÿ≤
E.AS 6
PN on genitive.
 CARIAN GLOSSARY 413

sl∞maewm
E.Th 34

smdÿbrs
C.Ha 1
PN in nominative (of a s-stem) or in s-ending (if the stem is smdÿbr-).
The name belongs to the family of nouns in -(d)ybr-/-(d)ÿbr, cf. ardy-
byr, dÿbr, etc. A comparison with the Carian name in Greek sources
Zermenduberow by Blümel (1990:81) is attractive, but the lack of r poses
a problem.

sm“s[–5–]
C.Si 2a

sm[–7–]a∞e[
C.Si 2a

snn
C.Ha 1, C.xx 1
Acc. sg. of the demonstrative pronominal stem sa/san-, q.v.

Melchert (1993:79).

sn≤
E.AS 8

sñaidlo
C.Si 1

Tentatively interpreted as a verb (aid- ‘they made’, cf. ait) preceded by a pro-
noun or introducing particles and followed by clitics in Adiego (2000:152),
where even an exact correspondence with Lyc. sẽñnaitẽ is proposed.

sñis
C.Ka 1

See p. 291 for a possible interpretation as a demonstrative pronoun (related

to sa/san-, ‘this’).
414 CHAPTER ELEVEN

som[n ? ]e
C.St 1
somne≤
E.Me 21, E.Me 34
PN in nominative (somne if the reading is accepted) and genitive (somne≤ ).
Directly comparable to the Carian name in Greek sources, Svmnhw.

It is undoubtedly a matter of chance, but it is curious nonetheless to note that

somne- is followed in E.Me 21 by a possible ethnic name q ÿblsi-, while the
only example of Svmnhw in Greek sources is an individual belonging to the
Kublisse›w (inhabitants of Kyblissos), mentioned as witnesses in a treaty between
Mylasa and Kindya (Inscription of Mylasa, n. 11 in Blümel’s edition).

sqla
E.Si 4

sqlumidun
E.Si 4

srton[-]t [. . .?]
E.Ab 28

stspñ
C.Si 2a

sursiabk
E.Si 6

suso
E.Lu 2

s[--]et≤
E.Me 29
PN in genitive.

≤as
C.Eu 1
A variant form of ≤jas (‘tomb’, or similar).
 CARIAN GLOSSARY 415

≤dun
C.Eu 2

≤emot
E.Th 10

≤en
E.Me 18a

≤jas
G 1
Noun corresponding to Greek s∞ma in the bilingual inscription G 1.

Adiego (1992a:33). For the difficulties of relating ≤jas to the Carian gloss soËa(n),
see p. 10.

≤o≤niabkol
C.Eu 2

≤t≤
E.AS 8

≤ugliq
E.Me 5
≤ugli≤
E.Me 30, E.xx 1
Ethnic name or, less likely, PN, attested at least in genitive (≤ugli≤ ).
The exact analysis of ≤ugliq is unclear. If an ethnic name, it can be
connected with the Carian place name Souaggela, although the doubts
about the precise sound value of 0 g make this identification more
difficult.

Adiego (2004:310). Connection to Souaggela suggested to me independently

by Pérez Orozco and Melchert (both in pers. comm.).

≤umo
C.My 1

≤uni≤
E.AS 8
416 CHAPTER ELEVEN

≤ysñal[
C.St 1

“abd ?aikal
E.AS 2

“amow
E.Ab 4, E.Ab 5
PN in nominative. It corresponds to Samvuow, Samvow in Greek sources.

Adiego (1994a:38). Vittmann (2001:55–57) does not rule out the possibility
that “amow-Samv(u)ow could be a ‘Carianised’ form of the Egyptian name Ô3j-
jm=w (directly attested in Carian as tamou q.v.).

“amsqi[. . .?
E.Me 24

“ann
C.Ia 3
“anne
C.Ia 3

“aoyr∞ri
C.Si 2a

See Adiego (2000:148–149) for a very hypothetical attempt to connect it with

Xrusaor¤w, the oldest name of Stratonikeia.

“arkbiom
E.Sa 1, E.Me 56 (“ark[bi/jom . . .?]), C.My 1
PN in nominative. Transcribed as ”3rkbym in the Egyptian part of the
bilingual E.Sa 1. Not found in Greek sources (although the form could
be confidently reconstructed as *Sarkebivmow). Compound name formed
by the adverbial stem “ar- (/gr. Sar-) (= CLuw. “arri, Lyc. hri, Mil. zri;
cf. also Hitt. “èr ‘upon’) and the stem kbiom-, also attested as an inde-
pendent name (see kbjom-, Greek Kebivmow).
 CARIAN GLOSSARY 417

Curiously, the Greek reconstruction *Sarkebivmow was already given as the

possible form corresponding to Egyptian ”3rkbym in Masson-Yoyotte (1956:52),
when both the true form “arkbiom behind the Carian letters and the name
Kebivmow were still unknown.
An initial, and still imperfect transcription of the name appears in Kowalski
(1975:90). See also Ray (1981:153). For an analysis of the name: Adiego
(1993a:242).

“arnai≤
E.Me 17
“arnajs
E.xx 6
PN in genitive (“arnai-≤ ) and in ‘s-case’ (“arnaj-s). It is not clear if the
stem can be related to the adverbial stem “ar-. A connection to the
Carian name of Greek sources Sarnow is hampered by the doubts about
the reading of the name (alternative reading Parnow, see Zgusta KPN:449,
n. 6, Blümel KarPN:24)

Adiego (1993a:250)

“arnw≤
E.AS 3
PN in genitive. Perhaps formed on the same stem as “arnai-/“arnaj- (see
preceding entry).

Adiego (1993a:250)

“arpt≤
E.Ab 33
PN in genitive.

“arwljat≤
E.Me 3
PN in genitive. A compound formed by “ar- (cf. “arkbiom, “aru≤ol) and
wljat- (q.v.). No Greek adaptation of the name has been found to date
(a form such as *Sar-uliatow, *Sar-oliatow is the most likely possibility).

Adiego (1993a:242–243).

“arur≤
E.Ab 37
PN in genitive.
418 CHAPTER ELEVEN

“aru≤ol
E.Me 30, E.Ab 6, E.Ab 30 (“a[ru]≤ol), E.SS 1
PN in nominative. Carian name that appears in Greek sources as
Sarus(s)vllow. A compound name formed by Sar- (cf. “arkbiom) and
u≤ol (q.v.).

Ray (1981:155, 161).

“aru≤[. . .?
E.Ab 42
If complete, a PN in genitive. But it is more likely to be an incom-
plete form of the noun “aru≤ol (see the preceding entry).

“asqariod
C.Hy 1a

“aÿdiq≤
E.Ab 30
PN in genitive. A compound name “a- + ÿdiq, perhaps a variant of
the name in the following entry.

On this explanation, see pp. 262–263.

“ayriq
E.Me 25
PN in nominative. It corresponds to the Carian name from Greek
sources, Saurigow. A compound name “a- (~ “ar?) + yriq. For the first
element, cf. Sa-ussvllow. The second element is the well-known stem
yri∞-/yriq- /ÿdiq-/ÿd∞- (see idyri∞-, paryri∞-, etc.).

Adiego (1993a:250), (1994a:44). On this family of names, see pp. 262–263.

“a[--]i≤b?wn
E.AS 5

“dtat≤
E.Me 13
PN in genitive.
 CARIAN GLOSSARY 419

“enurt
E.Me 50a
PN in nominative. It probably corresponds to the Carian name Sanortow
in Greek sources.

Adiego (1993a:236), and with some doubts about the identification (1994a:43).

“i“≤
E.Th 35
PN in genitive?

“odubr≤
C.Kr 1
PN in genitive. It seems to belong to the family of names in -(d)ybr-/-
(d)ÿbr-, but in this case u, and not y, is used.

Could this name be the Kaunian version of k≤atÿbr (= Lycian Janduberiw)?

For a > o, cf. otonosn and the following entry.

“oru≤
C.Ka 3
PN in genitive.

Assuming an a > o change (cf. otonosn and the preceding entry), a compari-
son with the Carian name Sarow could be feasible (for the adaptation of a
Carian u-stem in Greek as a thematic one, cf. pau- = Paow).

“rb˚[-]sal
E.Th 49

“rquq
E.Lu 2 (“?rquq), C.xx 1
“rquq≤
E.Me 43a, E.Me 44a
PN in nominative (“rquq) and in genitive (“rquq≤ ). A compound name
formed by “r- = “ar-/Sar- (cf. “ar+u≤ol-Sar+ussvllow, “ar+kbiom, “r +wli-)
+ quq (= Gugow in Greek sources).

Adiego (1993a:243). A name †arquq (Adiego 1993a:241, 1994a:35) does not

exist: all the alleged examples are actually misreadings of “rquq.
420 CHAPTER ELEVEN

“rwli≤
E.Me 20
PN in genitive. Can be analysed as a compound “r- (cf. “r-quq) + wli-,
very probably the same stem behind wliat/wljat (q.v.). Moreover, wli-
can be directly compared to the Isaurian name Oualiw (Zgusta KPN
§ 1134–3/4). Cf. also Pisidian Oliw, Zgusta KPN § 1086–1.

See Adiego (1993a:243).

“ÿin≤
E.Me 38, E.SS 1
PN in genitive.

-ÿin- recalls -yin in [--]ryin (the Carian form of the dynast name Idrieus), where
it is analyzed as a form of the ethnic suffix -yn-/-ÿn (see s. v. [--]ryin

tab
C.Ka 5

tamosi
E.Ab 18, E.Ab 19, E.Ab 21
tamosi≤
E.Ab 20
PN in nominative (tamosi ) and genitive (tamosi≤ ). Vittmann suggests that
we can recognise here the Egyptian name Pt˙-ms (literally ‘Ptah is born’),
note particularly the old Akkadian adaptation, Ta¢ma““i.

Vittmann (2001:43). However, note that there is no further evidence to sup-

port pt > t in Carian adaptations of foreign names.

taqbos
E.xx 6
PN in s-ending, coordinated with “arnaj-s by means of sb, ‘and’.

ta“ubt≤
E.Me 18a
PN in genitive.

tazomd [
C.Ki 1
 CARIAN GLOSSARY 421

tbridbd≤
E.Me 42
PN in genitive.

tdu≤ol
E.Me 24
PN in nominative. It belongs to the u≤ol-family of names, but the first
element, td-, is not clear. It is perhaps to be related to ted-, ‘father’?

Schürr (2003b:69, n. 1) sugggests considering initial t as a mistake and that

instead we should simply recognize here the same name as dw≤ol-.

tebot
E.Th 28 [teb%ot], E.Th 44

tebwnqmw
E.Th 38

ted
E.Me 38
Common noun in nominative: ‘father’. Compare CLuw. tàta/i-, Lyc. tedi-,
Lyd. taada- ‘father’. Note the apparent umlaut *a > e as in Lycian, which
implies an original stem in -i- or with i-mutation (*tadi- > *tedi- > ted-).

Schürr (1996[98], already suggested in 1996a:68), Hajnal (1997a:210).

temazi
C.Eu 2

terÿez≤
E.Me 4
PN in genitive.

tkrabi≤
E.Me 37
PN in genitive.

tksr
E.Lu 7
422 CHAPTER ELEVEN

tmonks
E.Th 41

tñu≤
C.Hy 1a
PN in genitive. The name appears adapted in Greek as Tonnouw in one
of the inscriptions that accompanies C.Hy 1.

Adiego-Debord-Varinlio<lu (2005:612).

tqlow
E.Th 32

tqtes
E.Me 47
PN in ‘s-case’ (or rather a nominative of an s-stem?).

trdy≤
C.My 1
PN in genitive

triel≤
E.Me 26
PN in genitive.

triqo
E.Me 6
PN in nominative.

trqdimr
C.Ki 1
Sequence that contains the god name trq(u)d-, ‘Tarhunt’ (see following
entry and armotrqdos). Neither the internal structure (perhaps it must be
segmented into two words trqd imr or trqdi mr) nor the function and
meaning (a PN or place name, or the god name proper?) are clear in
this case.

See Blümel-Adiego (1993:94–95), where possible connections for imrº and mrº
are proposed (imrº = CLuw. im(ma)ra/i- [but note that the normal correspon-
dence of this latter word in Carian is (i)b r-!], ºmr = Lyc. mere- ‘laws’).
 CARIAN GLOSSARY 423

trqude
C.Ia 3
God name: trqud- = Hitt. Tar¢u-, CLuw. Tar¢unt-, Lyc. trqqñt-, the
Anatolian Storm God. Unclear ending: perhaps a dative?
See also trqdimr, armotrqdos.

Blümel-Adiego (1993:94). trqude as dative: Adiego (1994a:38, 50). The simplest

interpretation is to assume that trqude is the divine name to whom the cratera
where C.Ia 3 is inscribed was dedicated, (Blümel-Adiego 1993:95), although the
overall context of the inscription is unclear and makes it difficult to give a
precise analysis (Melchert 2002:310).

t®∞atar≤
E.Me 34 (t®∞ata[r]≤), E.Me 41
t®∞atr≤
E.Th 5
PN in genitive.

The attempt to compare t®∞at(a)r- with Tarhunt- and the Carian place name
Tarkondar/a/ (Adiego 1992a:34, 1994a:43; see also Janda 1994: 175, who
interprets the word directly as an ethnic name) is hampered by the unsuit-
ability of the sound correspondences, particularly if compared with trq(u)d-,
the usual form of Tarhunt- in Carian. The interpretation as an ethnic name,
feasible for the examples from Memphis, is questioned by the clear PN
(patronym) from Thebes, see p. 269.

tsial
C.My 1 (2×)
PN in nominative.

ttbazi
E.Ab 41
ttbazi[≤]
E.Me 1
ttubazi
E.Ab 25
Feminine (?) PN in nominative (ttbazi, ttubazi ) and genitive (ttbazi[≤]).
As suggested by Schürr, an adaptation of the Egyptian name T3-dj(.t)b3st.t
(literally ‘the one (fem.) whom (the goddess) Bastet has given), Greek
Tetobastiw.

See Schürr (1996a), for details about the reading of the inscriptions in question.
424 CHAPTER ELEVEN

tumn
E.Sa 1
Accusative of tum-, a Carian adaptation of the Egyptian god name Jtm
(Atum)?

For this interpretation, see Adiego (1993a:255) and particularly (1995:21–23).

tur[
G 1
Beginning of a PN. It corresponds exactly to the truncated name that
appears in the Greek part of the bilingual G 1 (Tur[ ).

Adiego (1992a:33).

tusol≤
C.My 1
PN in genitive. The final ol of the stem could correspond in Mylasa to
the typical ending -vll/-vld- in Carian names. However, the name
hardly can belong to the family of the u≤ol-/Ussvllow names, given
the use of s, and not ≤, as would be expected.

ty∞[
C.My 1

For a very hypothetical interpretation, see Adiego (2005:92–93) and here on

p. 308.

tÿn
C.Ha 1

See p. 283.

t [-]rsi
C.Si 1

tamou
E.Me 7
PN in nominative, an adaptation of the Egyptian name that appears
in the hieroglyphic part of this bilingual inscription as Ô3j-Óp-jm=w
(literally ‘may Apis take them’ *[‘i˙pimòw]. But the Carian adaptation
in fact corresponds to Ô3j-jm=w or Ô3j-n.jm=w, Greek Tamvw, Yamvw,
 CARIAN GLOSSARY 425

Samauw, Samv#w, in this case an abbreviated form (phonetically *[‘amòw])

of the abovementioned name (DNb:1348–1349).

Ray (1981:58). See Vittmann (2001:55–56) for a discussion of the relationship

between tamou and the pure Carian name “amow.

tanai≤
E.Me 7
PN in genitive. Ray suggests that it may be an Egyptian name: *Ô3-n-
n3-j˙w, literally ‘offspring of the (sacred) cows’.

Ray (1994:200). Vittmann (2001:56, n. 87) notes that the name is not attested
in Egyptian.

trel
E.Bu 1

w
E.Me 13
Complete word, abbreviated form, or a simple mistake? See p. 272.

uantrpo
E.Ab 12
uantrpu≤
E.Ab 13
PN in nominative and genitive. It is very likely, but not completely
certain, that both forms belong to the same paradigm (therefore with
an alternation -o / -u≤; the parallel with -e / -i≤ in lÿ∞se/lÿ∞si≤ is striking).

uarbe
E.Th 1
PN in nominative.

uarila[-]os≤
E.Ab 39
Apparently a PN in genitive.

w#ar [---]t[------]i[---]≤
E.Me 11a
Remains of an onomastic formula N-Ø . . . N-≤.
426 CHAPTER ELEVEN

%wdbo≤kn
E.Th 47

wdwn
E.Th 13

Cf. owdown[ (E.Th 10). Janda (1994:182–183) observes the striking resemblance
of wdwn/owdown[ to the Pisidian sequence oudoun, but this comparison can-
not be taken further due to the similarly unclear contexts in which Carian
and Pisidian forms appear.

ue
E.Me 3, E.Me 5, E.Me 28, E.Me 29 ([ue]), E.Me 42, E.Me 51
‘Funerary stela’, or similar. It seems to be similar or correspondent to
upe/upa, but the precise relationship between the words (if it indeed
exists) is not clear.

See below s. v. upe about upe/ue connection. As indicated there, Schürr’s

hypothesis of a loss of p (upe > ue; Schürr 1992:141; 1993:172) is attractive,
but ad hoc.

uejresi
E.Si 2

wet≤
E.Me 13
PN in genitive.

uiomln
C.Ka 5
[ui ? ]omlã
C.Ka 2
Cf. also yomln, very likely to be a variant.
Probably a verbal finite form, corresponding in some way to Greek
¶doje, ‘It seemed good’, in the bilingual inscription C.Ka 5. However,
the precise analysis remains unclear. An alternative view, suggested by
Melchert, is to analyze it as a noun with the meaning ‘decree’.

All the analyses take as a starting point the example of the bilingual C.Ka 5,
contextually more clear (note moreover that the value of the final letter of the
C.Ka 2 example, here transcribed as <ã>, is far from being assured).
 CARIAN GLOSSARY 427

Frei-Marek (1997:30) propose a third plural person of a past tense with the
meaning ‘to decide’, whose subject would be kbidn, interpreted as ‘the Kaunians
decided’. Both Hajnal (1997b:151–153) and Neumann (1998:30) suggest that
the verb must correctly be ºmln, and try to connect the verbal stem ml- with
different Hittite verbs: malai- ‘to approve’ (Neumann), mald- ‘solemnly pro-
nounce, vow’. Similar analysis and etymological connection given in Adiego
(2002:20).
Eichner’s interpretation (only partially deducible from references apud Tremblay
1998:117, 123) concurs in isolating mln as a verb. He translates kbidn uio mln
as ‘il plaît/plaisant à la communauté des Cauniens’ with uio interpreted as a
dative sg (‘communauté’).
For Melchert’s view, see Melchert (1998:37): kbidn uiomln ‘decree of Kaunos’
(kbidn: place name, plurale tantum, here in genitive). In a supplementary note,
Melchert suggests very tentatively a connection of uiomln with CLuw. wayam-
man-, ‘cry, howl’, cf. also Hitt. wiyài-, ‘to cry’, assuming a semantic develop-
ment comparable to Lat. proclamatio to ‘proclamation’. However, in Melchert
(mdane), Hajnal’s view is preferred: he isolates mln and analyzes it as a preterite
third plural from, *mVld-onto, to be related to Hitt. mald-.

uksi
E.xx 7
PN in nominative (?)

uksmu
E.Me 2
wksmu≤
E.Me 36
PN in nominative (uksmu) and genitive (wksmu≤ ). Cf. the Anatolian names
Ouajamoaw (Zgusta KPN § 1141–2, Isauria, Cilicia), Ouajamvw (Zgusta
KPN § 11141–2, ibid.). This compound name therefore contains as a
second element the well-known Anatolian stem muwa-, ‘strength, force’.
It is very likely that the individual alluded to could be the same in
both inscriptions.

Neumann apud Adiego (1993a:236). On the identification of both individuals

(both show the same father’s name), see Masson (1976:38), Ray (1982b:187).
See here p. 336.

ula[----]ol
C.St 1
428 CHAPTER ELEVEN

uliade
C.St 2
PN in nominative. Carian adaptation of the Greek name OÈliãdhw,
very widespread in Caria, probably due to its resemblance to the purely
Carian name wliat (see the following entry).

Adiego (1994a:39–40). On the name OÈliãdhw see Masson (1988b).

wliat
E.xx 2
wljat
E.Th 7
wljat≤
E.Mu 1
PN in nominative (wliat/wljat) and genitive (wljat≤ ). It is the Carian
name that appears in Greek sources as Uliatow or Oliatow. Note the
compound name “arwljat-.

For the identification, see Adiego (1992a:31). The stem of the name has been
connected with Hittite walliwalli-, ‘strong, powerful’ (also on the basis of other
Anatolian names: Walawala, Walawali, and particularly Carian Oaloalow, about
which see Adiego 1993b), cf. Adiego (1993a:238). See here p. 339.

umot2
C.Ka 2

un
C.Ka 5 (2×)

Tentatively analyzed as an infinitive in Adiego (1998a:22), see here on p. 299.

Segmentation in both cases is not guaranteed!

undo[--]tl“
C.Ka 5
Acc. pl. c. of a stem undo[--]tl-. It seems to correspond to Greek eÈrg°taw
‘benefactors’ in the bilingual inscription C.Ka 5.

Already proposed in Frei-Marek (1997:38), who added the important obser-

vation that the sequence ºtl- could represent a suffix of a nomen agentis com-
parable to Hitt. -talla-.
 CARIAN GLOSSARY 429

In Adiego (1998a:22) a segmentation un do[--]tl“ is suggested, taking un as

an iteration of un (infinitive: ‘to make’) that appears just before (see preced-
ing entry), so that only do[--]tl“ was the word corresponding to eÈrg°taw.

untri
E.Ab 12, E.Ab 13
PN in nominative.

wnuti≤
E.Me 31
Genitive. According to Vittmann, this is a title rather than a PN, given
the good correspondence to Egyptian wnwtj (*[w6nùti/e] or *[wnùti/e])
‘hour-observer, horoscoper, astronomer’. Vittmann rightly observes that
the absence of an article in the adaptation of the Egyptian title fits well
with the Egyptian syntactical practice, consisting of omitting the arti-
cle when the title precedes the noun that it qualifies (in E.Me 31, wnuti≤
is the first word of the inscription, followed by the personal name kwar≤ ).

Vittmann (2001:48–49). See here on p. 278. Vitmann is reasonably cautious

in suggesting this interpretation, but the correspondence Carian wnuti = Egyptian
wnwtj cannot simply be a matter of chance.

uodrou
C.St 2

uodryia[
C.St 2

upa
E.Me 13
upe
E.Me 17, E.Me 4, E.Me 9, E.Me 22, E.Me 26, E.Me 38, E.Me 43a,
E.Me 64a ([. . . u?]pe)
wpe
E.Me 36, E.Me 41
Common noun in nominative: ‘(funerary) stela’, or ‘tomb’. Its connec-
tion with ue, used in similar contexts, is not clear.
Perhaps somehow related to Lycian xupa ‘tomb’?
430 CHAPTER ELEVEN

It is clear that upe/upa, independently of its precise meaning, makes reference

to the object where the inscription stands (‘funerary stela’), or to its function
(‘tomb’): see Meriggi (1980:36), followed by Adiego (1993a:208). E.Me 26
clearly supports this explanation, where upe appears accompanied by the demon-
strative sa, ‘this’ (Adiego 1993a:209). Other interpretations (‘son’, Ray 1982b,
followed by Kammerzell 1993; ‘I am’, Ray 1990a:72; a demonstrative, Ray
ibid.) must be ruled out.
Schürr has argued in favour of a generic meaning ‘stela’ instead of ‘tomb’,
assuming that E.Me 4 is a ‘stèle de donation’, and that the Carian text does
not seem to contain a typical onomastic formula (Schürr 1992:155). However,
pace Schürr, this stela has clearly been re-used (there is no connection between
the Egyptian and Carian texts), and the Carian inscription poses serious read-
ing problems that do not allow us to identify the type of formula used.
For upe/wpe vs. ue, Schürr has proposed a purely phonetic explanation, by
resorting to an alleged tendency in Carian to drop p in intervocalic position
(Schürr 1993:172; however, the evidence adduced is not convincing).

uqsi
E.Me 20
PN in nominative. Perhaps a simple graphical variant of PN uksi-?

urm≤
E.Bu 1
wrm≤
E.xx 7
PN in genitive. Note also the “vocalized” variant urom≤ in E.Bu 2, which
possibly alludes to the same person as urm≤ in E.Bu 1.

On the possible connection of this name to Luw. ura-, ‘great’, an adjective

that enters into the formation of Anatolian proper names, see above p. 338.

ur#o
E.Th 34

urom≤
E.Bu 2
PN in genitive. See urm≤, wrm≤.

urq
E.Lu 6
 CARIAN GLOSSARY 431

ursea∞k
E.Bu 6

urs∞le≤
E.Me 15
PN in genitive. It appears transcribed as 3rskr in the Egyptian part of
the bilingual inscription E.Me 15.

Final º∞le of the stem has led us to imagine a Carian adaptation of a Greek
name in -kl∞w (Neumann ers. comm. suggested ÉOrsikl∞w). The use of palatal
∞ for Greek k recalls Lyc. k (also a palatal) in Perikle < Gr. Perikl∞w.

urt
E.Th 34
PN in nominative?

u≤bzol
C.Hy 1a
PN in nominative.

u≤ol
E.Ab 35
u≤ol≤
C.Hy 1a (2×), C.St 2
w≤ol≤
E.Me 12
PN in nominative (u≤ol) and genitive (u≤ol≤/w≤ol≤ ), corresponding to
the Carian name that appears as Ussvldow, Ussvllow in Greek sources
(Zgusta KPN § 1629–7/8, Blümel KarPN:27); u≤ol- enters in compo-
sition with a series of prefixes (pn-u≤ol, “ar-u≤ol, id-u≤ol, etc.)

For the identification u≤ol = Ussvllow, one of the most decisive steps in the
Carian decipherment, see Ray (1981:160). About the possible etymology of
the name, see here p. 344, n. 16.

usot
C.Ke 1, C.Ke 2

On the doubts about these inscriptions, see p. 150.

432 CHAPTER ELEVEN

utnu≤
E.Ab 19, E.Ab 21 (u?tnu≤?)
PN in genitive.

uHbit
C.Ka 2

u[. . .]ü≤q
E.Th 12

ya
C.Ka 8

ÿasd≤
E.Me 46a
PN in genitive.

Initial ÿasº recalls the ethnic name (?) yiasi-, yjas[i]- ‘Iasean’ (see s. v.); ÿ vs.
yi/yj- finds a good parallel in y≤biks vs.- yi≤{∞}biks-. As for ºd-, it can be inter-
preted as a nt-suffix.

ybrs≤
C.Hy 1a
PN in genitive. The stem ybrs- recalls the family of names in -(d)ybr-
/(d)ÿbr-, particularly smdÿbrs (see pp. 283, 314). It is possible that this
indigenous name was identified with the Greek name ÑUbr°aw (Zgusta
KPN § 1624), commonly used in Caria, in a process similar to that of
wliat-OÈliãdhw.

Adiego-Debord-Varinlio<lu (2005:613).

ÿbt
C.xx 1
Probably a verb: 3rd singular preterite or present of a verb ÿb- =
Lycian ube- ‘to offer’: If preterite, it would be completely equivalent to
Lycian ubete, ‘he offered’. If present, it would be from *ube-ti.

For this interpretation, see Melchert (1993:78–79). Melchert’s interpretation is

followed by, among others, Adiego (1994a:240) and Hajnal (1995[97]). An
alternative view was attempted by Janda, who prefers to consider pjdl as the
 CARIAN GLOSSARY 433

verb in C.xx 1; Janda (1994:179) suggests that ÿbt can be a particle chain or
a noun corresponding formally to CLuw. upatit- ‘landgrant’ (< *‘donation’.
The Carian word would retain this original meaning).
About the possibility that ÿbt can be a present, cf. Hajnal (1995[97]:17).

yiasi
E.Me 25
yjas[i≤ ]
E.Me 9
Ethnic name (?) in nominative ( yiasi ) and genitive yjas[i≤ ]. Connected
with the Carian place name Iasos (Iasow)?

Adiego (2004:310) and here p. 270.

yi≤{∞}biks≤
E.Me 46a
PN in genitive. If the corrected reading is accepted (see p. 69), it is a
variant form of ÿ≤biks, q.v.

ylarmit
C.Hy 1b
Ethnic name (in genitive plural?) referring to the Carian city of Hyllarima.

Connection of this word to the place name Hyllarima already noted in Ray
(1988:152). For ylarmit as genitive plural with the meaning ‘Hyllarimeans’, see
Adiego (2002:17).

ymezus[
C.St 2

ÿn-?-mo
E.Sa 1
Cf. ÿnsmsos, although the integration of s between n and m is by no
means clear.

ynemori≤
E.Me 29
PN in genitive.
434 CHAPTER ELEVEN

ÿnsmsos
E.Mu 1
E.AS 3
Possibly a title or adjective in nominative, where a suffix -os can be
identified (cf. kbos E.Me 24).

yomln
C.Ka 4
Perhaps a variant form of uiomln, q.v.

ÿpdnmw%d
E.Th 4

yri∞ñ
C.Si 1 (2×)
Final ñ (if the reading is accepted), points to a possible accusative. A
possible stem yri∞- would recall the family of names in yri∞-/yriq- /ÿdiq-
/ÿd∞- (see idyri∞-, paryri∞-, etc.).
Can yri∞ñ have any connection with the GN Sinuri?

yri≤
E.Th 26
PN in genitive.

Cf. pnyri≤ru. Could yri- be related to yri∞-/yriq-?

yrqso≤
C.My 1 (2×)
PN in genitive. It corresponds to the name adapted in Greek as Urgosvw.

Adiego (2005:90)

ÿrsbe
E.Ab 6
Unclear word. Cannot be a PN in nominative, given the context in
which it appears (preceded by a PN in nominative (“aru≤ol) and fol-
lowed by a PN in genitive (pdubi≤ ). Perhaps a title?
 CARIAN GLOSSARY 435

ÿ≤biks
C.xx 2
PN in nominative. Compare yi≤{∞}biks≤. The name is a compound whose
second element is biks (cf. piks-, dbiks- and see p. 337 for an etymolog-
ical explanation).

ÿsm
E.AS 9

yysmt≤oHa[
E.Ab 27

zidks
E.Sa 1
Sequence immediately followed by mdane. Function and meaning unknown
(an ‘s-case’ of a PN? A verb?).

zmu≤
E.Me 19
PN in genitive.

z“ariosã
C.Ka 2
Coordinated by means of sb to another word ending in -ã ([-]∞arlanoã).

Hd“qedormñs[
C.St 2

H∞it
C.Ka 2

Hnmkda[-]aHuq[
C.Ki 1

Horouo
C.Ka 5
Function and meaning unknown.

In Adiego (2000:144) attention is drawn to the good parallel between ºorouo

and Lycian arawa- ‘tax exemption, ét°leia’, but the connection is hampered
by the initial H (difficult to separate from orouo, given that Horouo appears
immediately after sb, ‘and’).
436 CHAPTER ELEVEN

Hosurz
E.Ab 28

[41]
1aitk
C.Ka 2
Function and meaning unknown. Perhaps k could be an enclitic ele-
ment, and the resulting form 1ait º could be compared with ait in the
same inscription, for which an analysis as a third plural person verb
has been proposed (see ait).

1mali
C.Ka 2 (2×)

1orsol“
C.Ka 2
Apparently an accusative plural, coordinated with sarni“ by means of
sb. Meaning unknown.

?-˚bjqmqew
E.Th 12
Function and meaning unknown. Perhaps we are dealing with more
than one word. Note the final ew (see ewm), and the sequence ºkbj º,
which recalls the PN kbjom≤, “arkbiom and place name kbidn.

?-ras
E.Ab 43
PN in nominative (?)

ACEPHALIC WORDS

[-]aH
C.Ka 2
Cf. a similar ending (]maH) in the same inscription.

[-]ars
E.Ab 36
Apparently a PN in nominative.
 CARIAN GLOSSARY 437

[-]bdo
C.My 1
PN in nominative. No parallels have been established.

[-]bi
C.Ka 2
Perhaps the same enclitic element that appears twice in the same inscrip-
tion, see bi.

[-]diurt
C.Ka 2

[-?]iam≤
E.Me 45
PN in genitive.

[-]intnor
C.Ka 2

[-]∞arlanoã
C.Ka 2
Coordinated by means of sb to another word ending in -ã (z“ariosã).

[-]nudrma
C.Ka 2

[-]obiokli≤
C.My 1
Onomastic formula consisting of a PN in nominative followed by a PN
in genitive, the difficulty being the isolation of the two names.

[-]owt≤
E.Me 48
PN in genitive.

[-]qo
C.My 1
PN in nominative. For the final, cf. plqo-Pellekvw.

[-]tmai≤[--]
E.Bu 3
438 CHAPTER ELEVEN

[-]untlau[-]
E.Ab 22

[--]abrun∞ur[-]“yn“ Hynn
C.Ka 2
A sequence of words whose segmentation is not clear.

Neumann (1998:26–27) proposes isolating a word ∞ur[-]“yn≤, which he inter-

prets as an ethnic name in plural (with the -yn- ethnic suffix, cf. kbdyn“ ) and
connects to the Carian place name Koar(r)end/a/, Koarenz/a/ Koaranz/a/,
Kvra(n)z/a/. Although the proposed segmentation is attractive (particularly
regarding the ending -yn-“ ), the relation with the mentioned place name is far
less compelling; note that the Carian word cannot be read integrally: also the
correspondence Car. “ : Greek -nd-/-(n) z- (in the different variants of the name)
is not at all satisfactory.

[--]e∞ld
E.Si 2

[--] j [-]≤
E.Me 48
Extant letters of a PN in genitive.

[--]msal
E.Bu 1
Last letters of an initial word in a graffito from Buhen. Final ºsal recalls
euml ?bnasal, eypsal, also initial words in Buhen graffiti. See euml?bnasal for
a proposed interpretation of ºsal.

[--]ryin
C.Si 2a
PN in nominative. Given the context in which it appears, it must be
the Carian name of the dynast Idrieus. It is clear that the Carian form
of the name was not totally coincident with the Greek one. One could
tentatively complete *[id]ryin and assume a PN formed on the name of
the Carian city Idriaw, *idr-, by means of the suffix for ethnic names
-yin (cf. kbd-yn-“, mda-yn/mda-ÿn). The resulting meaning, ‘inhabitat of
Idrias’, could roughly correspond to the Greek ÉIdrieÊw.

Schürr (1992:137–138).
 CARIAN GLOSSARY 439

[--]Hl∞sasot2
C.Ka 5
The final -ot2 recalls similar endings in this inscription and in C.Ka 2:
mdot2, punot2, umot2.

[--]w≤ord≤
E.Me 10
Final part of a PN or ethnic name in genitive. For the ending, cf. the
ethnic name (or PN) msnord-

[--] yt2
C.Ka 5

]a[-]i≤
E.Me 11b
Extant letters of a PN in genitive.

]allia
E.Ab 30

]bewmsmnwdiq
E.Th 38
A sequence that seems to contain more than one word. We could per-
haps isolate a word smnwdiq, probably a PN related to the family of
names in yri∞-/yriq- /ÿdiq-/ÿd∞- (Greek -urigow, -udigow) (but note here
the apparent use of w). As for ]bewm, it recalls the element ewm pre-
sent in other graffiti from Thebes (see the corresponding entry).

]btdeo
E.Th 14

]b ?e≤
C.Ia 6
Probably the final letters of a PN in genitive.

]dar1
C.Ka 4

]i≤
E.Me 57
Final part of a PN in genitive.
440 CHAPTER ELEVEN

]∞≤
E.Me 60
Very likely to be the final part of a PN in genitive.

]latmne≤
C.St 1
PN in genitive. It is possible that the name may be complete. In this
case, cf. perhaps the Carian place name Latmow (Zgusta KON § 696,
Blümel KarON:173). The segmentation ]la tmne≤ would also be an
attractive theory, as well as a comparison of this latter with the Carian
PN Tumnhw (Zgusta KPN § 1615, Blümel KarPN:26).

]maH
C.Ka 2
Cf. [-]aH in the same inscription.

]no≤ ?
C.Ka 7
Last letters of a PN in genitive?

]ois?ur?mlo
C.Ka 9

]pri
E.Ab 26
Probably the last letters of a PN in nominative.

]q≤si≤
E.Me 53
Final part of a PN or ethnic name in genitive.

]rbn“a[
E.Si 9

]r≤wk[-]“ [
E.AS 8

]r[—]tnit
E.Ab 29
 CARIAN GLOSSARY 441

]sel“
C.St 1 (3×)
Very likely to be the final letters of a PN in nominative.

]s≤
E.Me 58
Final letters of a PN (s-stem) in genitive.

]tbe≤
C.Si 1
It seems to be the final part of a word (perhaps a PN?) in genitive.

]tbsms
C.Ka 2

]ub“ÿ
C.Di 1

]uda[
C.Ki 1

]ue∞l
C.Ia 2

]u≤
E.Me 26
Last letters of a PN in genitive.

]u≤ou≤
C.Ka 4

]utr[
E.Me 59

] ybzsdm
C.Ki 1

]zolba∞a[. . .]
C.Ki 1
 APPENDIX A

CARIAN INSCRIPTIONS IN TRANSCRIPTION

E = EGYPT
C = CARIA (AND NEARBY AREAS IN LYDIA AND LYCIA)
G = GREECE

E = EGYPT
Sa = Sais E.Me 3
Me = Memphis pikre≤ ue
Ab = Abydos “arwljat≤ msnord≤
Th = Thebes
Lu = Luxor E.Me 4
Mu = Murwàw terÿez≤ | upe | nuol+∞[---]sarmrol∞yt
Si = Silsilis
AS = Abu Simbel
SS = Gebel Sheik Suleiman E.Me 5
Bu = Buhen psm“kwneit≤ | ue | naria≤ | ≤ugliq |
xx = Unknown origin sarl?

E.Sa = Sais E.Me 6

triqo : parma≤≤ ∞i
klorul ∞i
E.Sa 1
“arkbiom : zidks mdane : ÿn-?-
mo | den : tumn E.Me 7
tamou tanai≤ qarsio[-?]
E.Sa 2
pdnejt qÿri≤ ∞i E.Me 8
a. paraeym : armon ∞i
b. para!eym : sb polo
E.Me = Memphis

E.Me 9
E.Me 1
arli“≤ : upe : arlio
ttbazi[≤] | Ñiub[a]Ωi≤ | aor[≤]
[m≤] ∞i : yjas[i≤]

E.Me 2
E.Me 10
uksmu | lkor≤ | mrsi≤
œårm≤ : q[---]≤ ∞i : p∂uüi≤ mno≤
[mw]don≤ ∞[i ---]w≤ord≤ ∞i
444 APPENDIX A

E.Me 11 E.Me 20
(a) wår[---]t[------]i[---]≤ | mdaÿn uqsi | “rwli≤ ∞i | mwdon≤ ∞i
(b) [--15--]a[-]i≤ | mdaÿn
E.Me 21
E.Me 12 punw≤ol≤ : somne≤
pjabrm | w≤ol≤ | mwdon≤ ∞i qÿblsi≤ ∞i
kbjom≤ | m[no≤]
E.Me 22
E.Me 13 artay≤ : upe : [. . .
“dtat≤ | upa | w | wet≤ ∞i | mwdon≤
∞i E.Me 23
ap[---]ws
E.Me 14 a[rb]ikarm≤ ∞i
irow | pikarm≤ | mwdon!≤
E.Me 24
E.Me 15 tdu≤ol
arli“≤ kbos | “amsqi[. . .?
urs∞le≤
kidbsi≤ E.Me 25
“ayriq | parpeym≤ ∞i
E.Me 16 yiasi
irow | pikra≤ ∞i
semw≤ | mno≤ E.Me 26
mwdon≤ ∞i [. . .]u≤ | upe sa | triel≤ | mrsi≤

E.Me 17 E.Me 27
“arnai≤ irow≤ : psHÿm[-]≤
upe | quq≤ pttu≤ : mno≤
bem≤ ∞i md-
aÿn
E.Me 28
sanuq≤ | ue | pntmun≤ ∞i
E.Me 18 mwdon≤ ∞i
(a) ta“ubt≤
kuari≤b-
ar | ≤en E.Me 29¡
niqau≤ s[--]et≤ | [ue] | ynemori≤ | mwdon≤
ptnupi
E.Me 30
(b) idmuon≤ “aru≤ol
∞i | mdayn pleq≤ ∞i : ≤ugli≤
∞i
E.Me 31
E.Me 19 wnuti≤ | kwar≤ mHm≤ ∞i | mwdon≤
pnu≤ol [∞]i
zmu≤ ∞i
 CARIAN INSCRIPTIONS IN TRANSCRIPTION 445

E.Me 32 E.Me 44
iturow≤ | kbjom≤ | ∞i en | mw[d]on≤ (a) apmen “rquq≤ kojol ∞i
∞i (b) mwton≤ ∞i

E.Me 33 E.Me 45
(a) idmns | myre≤ ∞i | mdayn ∞i [q?]lalis
(b) idmns | myre≤ ∞i [?]iam≤ ∞i
alos ∞arnos
E.Me 34
me®≤ | somne≤ | t®∞ata[r]≤ E.Me 46
(a) ÿasd≤ | yi≤{∞}biks≤ ∞i
E.Me 35 (b) mwdon≤ ∞i
ntokris | dw≤ol≤ | mwdon≤ ∞i
E.Me 47
E.Me 36 tqtes | paraibrel≤ ∞i | mn[o-?]
wksmu≤ | wpe | lkor≤ ∞j
qarpsi≤ E.Me 48
[--]j[-]≤
E.Me 37 [-]owt≤
qlali≤ | [. . .] ∞i : msn-
tkrabi≤ ord≤

E.Me 38 E.Me 49
“ÿin≤ | upe | arie?≤ ∞i ted loubaw | siral | pnld≤wl

E.Me 39 E.Me 50
[. . .]s? | ar∞ila≤ (a) “enurt
mno≤ (b) p∞simt≤ ∞i

E.Me 40 E.Me 51
plqo | pikrm≤ ∞i | mwdon≤ ∞i arli““ | psikro≤
ue
E.Me 41
|? or≤ | wpe | qdar®ou≤ | t®∞atar≤ E.Me 52
[. . .] ardybyr≤ | md[. . .]
E.Me 42
arjom≤ : ue : mwsat≤ : ∞i : mwdon≤ : E.Me 53
∞i [. . .]q≤si≤
tbridbd≤ : ∞i
E.Me 54
E.Me 43 [. . .] mrsj[. . .]
(a) lÿ∞si≤ | upe | “rquq≤ ∞i | ksolb≤
(b) arliom≤ | mno≤ ∞i E.Me 55
[. . .] psma[≤/“k . . .]
446 APPENDIX A

E.Me 56 E.Ab 3
[. . .] “ark[bi/jom . . .?] ptn“e | ibarsi≤

E.Me 57 E.Ab 4
[. . .]i≤ ∞i “amow ltari≤

E.Me 58 E.Ab 5
[. . .]s≤ ∞i “amow ltari[≤]

E.Me 59 E.Ab 6
[. . .]utr[. . .] “aru≤ol | ÿrsbe | pdubi≤

E.Me 60 E.Ab 7
[. . .]∞≤ plat | pals≤

E.Me 61 E.Ab 8
[. . .]i plat pals≤

E.Me 62 E.Ab 9
[. . .]≤[. . .] plat pals≤

E.Me 63 E.Ab 10
(a) idyes≤ piubez
(b) m [? qurbo≤

E.Me 64 E.Ab 11
(a) [. . .u?]pe : pd[ ≤? / [. . .]it
(b) [. . .]mi
E.Ab 12
E.Me 65 untri uantrpo
u[. . .]m | punm[-]≤ | mudo[n]≤
E.Ab 13
E.Me 66 untri | uantrpu≤
---].[..u][. . .]p[-]n[---
E.Ab 14
E.Ab = Abydos abrq∞[. . .?

E.Ab 1 E.Ab 15
pisiri pdubez or≤

E.Ab 2 E.Ab 16
panejt iarja≤ nprosn≤
 CARIAN INSCRIPTIONS IN TRANSCRIPTION 447

E.Ab 17 E.Ab 30
pa[-]in[-]t≤ bid≤lemsa : “a[ru]≤ol : “aÿdiq≤
[. . .]allia : bsis
E.Ab 18
tamosi | inut≤ E.Ab 31
∞aye
E.Ab 19
tamosi utnu≤ E.Ab 32
∞arr≤
E.Ab 20
ninut | tamosi≤ E.Ab 33
“arpt≤ | p[-]lu≤
E.Ab 21
to[-]a[---]l E.Ab 34
tamosi u?tnu≤? dbkrm [-]kb?[

E.Ab 22 E.Ab 35
[-]untlau[-]| u≤ol | mi∞≤≤ kdûsi≤

E.Ab 23 E.Ab 36
be≤ol [-]ars, ∞[-]urb≤

E.Ab 24 E.Ab 37
[. . .] arli“ “arur≤

E.Ab 25 E.Ab 38
ttubazi kattÿri≤ piew

E.Ab 26 E.Ab 39
[. . .]pri | ptnuq?i? uarila[-]os≤

E.Ab 27 E.Ab 40
yysmt≤oHa[ ialli | q∞blio≤

E.Ab 28 E.Ab 41
Hosurz | srton[-]_[. . .?] ttbazi kt?tri≤
(or: . . . +t[-]nota/rs | za/rusoH/l?)
E.Ab 42
E.Ab 29 “aru≤[..?
[. . .]r[--]tnit
E.Ab 43
?-ras
448 APPENDIX A

E.Th = Thebes E.Th 13

dbiks | kbjoms | wdwn | sb a≤b≤t
E.Th 1 ewm
uarbe
E.Th 14
E.Th 2 ]q[. . .]btdeo
dtÿbr | kbokt≤
k≤atÿbr E.Th 15
Very uncertain reading!
E.Th 3
pla?t E.Th 16
∂saml-?-?-” (vacat) dy “a
E.Th 4
dokmmpint seqqejewsk | mqtjq E.Th 17
ÿpdnmwd. ku

E.Th 5 E.Th 18
dÿbr | t®∞atr≤ t n

E.Th 6 E.Th 19
bebnd dbikrm

E.Th 7 E.Th 20
wljat orbá ˚ r i“

E.Th 8 E.Th 21
qutbe mmn∞al

E.Th 9 E.Th 22
kudtubr mwk | te

E.Th 10 E.Th 23
a?q≤baq ewm ≤emot bebi
qtblo owdown[. . .]mwarudk≤o mlane
E.Th 24
E.Th 11 kow[?-?]
psma≤[k] [?
| nm[
mplat | o[ E.Th 25
ktmno
E.Th 12
?-˚bjqmq ewmlane qeb≤t | u[. . .]ü≤q E.Th 26
| qwsal | mqabaewleqo“osk)ioms brsi yri≤
 CARIAN INSCRIPTIONS IN TRANSCRIPTION 449

E.Th 27 E.Th 40
pnw≤ol | mlqi≤ pnu≤ol

E.Th 28 E.Th 41
bejeym | teb”t tmonks
K
beb)int ken E.Th 42
rdudmm»≤
E.Th 29
]ke E.Th 43
p
E.Th 30
bebint | psrkrte | mumn“tnse-? E.Th 44
»ßw˚n dquq | ewmlane | tebot | gkem≤

E.Th 31 E.Th 45
Very uncertain reading! krwß

E.Th 32 E.Th 46
tqlow prpwri∞ kblow≤

E.Th 33 E.Th 47
∞lbiks≤ *wdbo≤kn ewá¬å»e ˚[

E.Th 34 E.Th 48
sl∞maewm | urt | kwri≤ | prna∞non brsi
| dm-?-n | maãtnor | qanor | ur”
E.Th 49
E.Th 35 bal ewlane | “rb˚[-]sal |
lÿ∞se | “i“≤ | mlan[-?]
E.Th 50
E.Th 36 pn-?
\ or≤
E.Th 51
E.Th 37 p
ktmn
E.Th 52
E.Th 38 plqodse | ewm-?-?-?-? | rqemw | k-?[
]bewmsmnwdiq tebwnqmw
E.Th 53
E.Th 39 dr“≤iem
krws | ko“m≤
E.Lu = Luxor Temple
450 APPENDIX A

E.Lu 1 E.Si 5
ds-? betkrqit[-- . . .]

E.Lu 2 E.Si 6
rsy bÿta“ | sursiabk | dr[- . . .]
suso qku
“?rquq [. . .?
E.Si 7
E.Lu 3 psma≤k
Very uncertain reading!
E.Si 8
E.Lu 4 bij≤≤pe
?-?-[-]ms[-]ry-?-?
E.Si 9
E.Lu 5 [. . .]rbn“a[-- . . .]
b?s?ui∞am | oã?
E.Si 10
E.Lu 6 ∞?mpi
| urq
E.Si 11
E.Lu 7 dmo“bqs
tksr (or: tasr)
E.AS = Abu Simbel
E.Mu = Murwàw
E.AS 1
E.Mu 1 par≤olou
p?owk | wljat≤ | ÿnsmsos [. . .]oe
saawon sa?awon
E.AS 2
E.Si = Silsilis “abd?aikal

E.Si 1 E.AS 3
∞iqud | marariso[-. . .] pisma“k | “arnw≤ | ÿnsmsos

E.Si 2 E.AS 4
[--]e∞ld | wa | psma≤k≤ | a∞akowr | emsglpn | b[. . .]
uejresi | qan | kolt | kowrn[. . .? pisma[“/≤k . . .]

E.Si 3 E.AS 5
irasa | n[-]eakrnanb pnyri≤ru | iÿkr≤ | “a[--]i≤b?wn

E.Si 4 E.AS 6
[. . .]K bebint | sqlumidun | sqla platt
slaÿ≤ ∞i
 CARIAN INSCRIPTIONS IN TRANSCRIPTION 451

E.AS 7 E.Bu 6
naz ∞i∞ | bÿ“ | esak?dow“ | mÿqu- eypsal
dem | pisma≤k | bebint | mo | ne puor≤ | aor≤ | ursea∞k ∞i
| psÿ≤[|?] ai[-]iqom
E.SS = Gebel Sheik Suleiman
E.AS 8
nid≤kusas | meÿqak | sn≤ | ≤t≤ | ≤uni≤ E.SS 1
| k“mmsm[. . .] n≤n[-]s“|“aru≤ol
[. . .]r≤wk[-]“[ pneit|“ÿin≤
parÿd∞≤
E.AS 9
ÿsm [? E.xx = Unknown origin

E.Bu = Buhen E.xx 1

qorb | isor≤ ∞i | ≤ugli≤
E.Bu 1
[--]msal | ar- E.xx 2
[®]i“ | psma≤- wliat
k≤ | urm≤ | an-
kbu“ | trel
kdou≤ E.xx 3
ionel≤
E.Bu 2
euml?bna- E.xx 4
sal | ar®i“ pduba
pdtom≤
urom≤ | an- E.xx 5
kbu“ ow∞meb≤t

E.Bu 3 E.xx 6
[-]tmai≤[--] “arnajs | sb taqbos

E.Bu 4 E.xx 7
psma≤k ntros : prãidas
ibrsi≤ or“a
nu mdane : uksi wrm≤
E.Bu 5
psma≤k

C = CARIA (AND TRANSITIONAL NEIGHBORING AREAS OF LYDIA AND LYCIA)

Tr = Tralles (Lydia-Caria) St = Stratonikeia
Al = Alabanda (-Eski Çine) My = Mylasa
Hy = Hyllarima Si = Sanctuary of god Sinuri near
Eu = Euromos Mylasa
452 APPENDIX A

Kn = Kindye C.Hy 1
Ki = Kildara (a) “asqariod dymda
Ha = Halikarnassos muot armotrqdosq
Di = Didyma (Ionia)
Ia = Iasos brsi ari“≤ brsi≤
Ke = keramos mane : u≤ol≤
Ka = Kaunos rtim u≤ol≤ pur?i≤
Kr = Krya (Lycia) u≤bzol tñu≤ brsi≤
xx = Unknown origin pau mane≤ ybr-
s≤
C.Tr = Tralles (Lydia-Caria border)
(b) kdu≤opizipususot
mol“ msot ylarmit
C.Tr 1
sdi amt[
pau≤ C.My = Mylasa
art{ }mon
C.My 1
C.Tr 2 idrayridsemdbq mol“ ty∞[
an sidi a- tsial tusol≤ : moi m[-]sao[
rtmi pau≤ banol paruos≤ : p?au paryri∞≤
parãaq? qzali obrbi≤ : tsial obrbi≤
banol yrqso≤ : paryri∞ psoir≤
[-]bdo pnu≤o≤ : myze trdy≤
C.Al = Alabanda and surroundings “arkbiom qzali≤ : ≤umo kbdmu≤
skdubrotoz≤ : pau ∞toi≤
C.Al 1 (Eski Çine) [-]qo idyri∞≤ : ksbo idu≤ol≤
sdi a[-]mob[ [-]obiokli≤ : ∞toi yrqso≤

C.Eu = Euromos C.Si = Sancutary of Sinuri near

Mylasa
C.Eu 1
≤as : ktais idyri∞≤ : mn[os?] C.Si 1
adymd“ : yri∞ñ : t[-]rsi : [. . .?]
C.Eu 2 tbe≤
omob ∞i : temazi (vacat)
≤dun : ≤o≤niabkol yri∞ñ : binq : sñaidlo
armon qyrbmudolo
manon C.Si 2
(a) [--]ryin ∞tmño≤ : sb ada ∞tmño≤
C.Kn = Kindye eri : pisñoi mda : pñmnn≤ñ : pda-
∞m≤uñ ∞i “aoyr∞ri mt∞elã
ñmailo mda lrHñ : stspñ vacat
C.Kn 1 sm“s[--5--] sb añmsñsi mda
pareÿs sm[--7--]a∞e[
∞[--8--]tuñdñ[
C.Hy = Hyllarima ñe-?-[
 CARIAN INSCRIPTIONS IN TRANSCRIPTION 453

(b) pim[. . .] limtaoa | [

Ha?[. . .] om

C.Ki = Kildara C.Ia 2

]ue∞l | ∞ob[
C.Ki 1
[. . . . . . .(.)]zolba∞a[. . (.)] kil[ C.Ia 3
[. . .]uda[. . .] trqdimr qrds tazomd[ ?] are“ | “anne mlne | siyklo≤ | “ann
kilarad[-]ybzsdmHnmkda[-]aHuq[ | trqude | ∞lmud [?
iasoum
C.Ia 4
C.St = Stratonikeia n[. . .]
pr[. . .]
C.St 1 is[. . .]
]sel“ a[--]a[----]om≤
]som[n?]e brsi≤ ula[----]ol C.Ia 5
]latmne≤ ≤ysñal[ baqgk[. . .]
] ari“ maqly≤[
]sel“ piks[ C.Ia 6
]sel“ p[ [. . .]b?e≤

C.St 2 C.Ke = Keramos

u≤ol≤ uodrou u[
mute≤ ymezus[
∞diye≤ uodryia[ C.Ke 1
uliade pidaru[ uso-
mañ“qaraH≤rl-?-[ t
dar“qemorms[
Hda“qedormñs[ C.Ke 2
uso-
C.Ha = Halikarnassos t

C.Ha 1 C.Ka = Kaunos

smdÿbrs | psnlo | md orkn tÿn | snn
C.Ka 1
C.Di = Didyma (Ionia) sñis : sdisa-
s : psu≤ol≤
mal≤ : mno≤
C.Di 1
]ub“ÿ
C.Ka 2
[ui?]omlã qrds grdso[-]i[
C.Ia = Iasos [-]r sb a∞mnnartnyr obsmns[
[-]∞arlanoã sb z“ariosã i∞[
C.Ia 1 [-]nudrma ∞yrpai sarni“ sb u[
]la [-]aH punot2 otr“ bi sb a∞tmsk[m
454 APPENDIX A

[-]d bi 1aitk ouor gdb“laã1_i[-] C.Ka 8

[-] sarni“ sb1orsol“ sb uHbit potko≤l≤? aba?d?
[-]bi qrdsol“ ait 1_mali H∞it ya
[-]intnor ∞yrapai≤ umot2 oba
[-]diurt obsmsmñ1 ñ ouor mt1 yr C.Ka 9
[--]abrun∞ur[-]“yn“Hynn sb vacat [. . .]ois?ur?mlo
[------------------]tbsms 1mali [
[-------------------]maH sb an[
C.Kr = Krya (Lycia)
[. . . . . . . . . . . . . . . . . . . . . . . .]ba vacat

C.Kr 1
C.Ka 3
qot2omu sdisa
“oru≤
m?n≤ “odubr≤ or rather: sn≤ “odubr≤?
ann ibrs≤
sb mno≤ knor
noril?ams or rather: norimams?
C.Ka 4
[. . .]u≤ou≤ ibrsdr[-]
C.xx = Unknown origin (presumably
[. . .]a yomln1 r_i
from Caria)
[. . .]dar1_ idym“

C.xx 1
C.Ka 5
“rquq | qtblem≤ | ÿbt | snn | orkn
kbidn uiomln i[---]
| ntro | pjdl?
inis drual nik[--]
lan lysiklas[-?]
otonosn sb lys[ikl] C.xx 2
an lysikratas[-?] ÿ≤biks not : alosd ∞arnosd : jzpe mdane
otonosn sarni[“]
mdot2 un sb undo[--] C.xx 3
tl“ kbdyn“ sb b2o[--] akymyduÿeryl[vacat]d
ol“ otr“ sb a∞t[ms]-
kmt absims sb [---]
yt2 oru sb a∞t[----] C.xx 4
bu∞y[-----]i[-----]i kdu≤ol“
[--]≤ un moa[-]lboror
[--]Hl∞sasot2 ort C.xx 5
tab sb ort[-] sb Hor- kdu≤ol“
ouo bi mslmnlia
purmoruos mnos G = Greece
aitusi
G 1 (Athens)
C.Ka 6 ≤jas | san tur[
or
G 2 (Thessalonike)
C.Ka 7 qlali≤ | k?[
]no≤? (or rather: ]noñ?)
 APPENDIX B

CARIAN GLOSSES

êla ‘horse’ g¤ssa ‘stone’

bãnda ‘victory’
Stephan of Byzantium, s. v. MonÒgisa
Stephan of Byzantium, s. v. ÉAlãbanda: MonÒgisa , < pÒliw Kar¤aw > (. . .)
ÉAlãbanda, pÒliw Kar¤aw (. . .) kt¤sma g¤ssa går tª Kar«n fvnª l¤yow •rmh-
d¢ KarÚw ∑n , épÚ toË paidÚw aÈtoË neÊetai. ka‹ nËn toÁw plak≈deiw ka‹
klhye›sa toË gennhy°ntow épÚ malak≈deiw l¤youw g¤ssa l°gousi.
KallirrÒhw t∞w Maiãndrou, metå n¤khn
flppomaxikÆn, ka‹ klhy°ntow ÉAlabãndou, kÒon ‘sheep’
˜ §sti katå tØn Kar«n fvnØn flppÒnikow.
êla går tÚn ·ppon, bãnda1 d¢ tØn n¤khn Scholia ad Il. XIV, 255: tÚ d¢ prÒbaton
kaloËsin. kÒon (ms. ko›on) ofl Kçrew Ùnomãzousin,
˜yen K«w ≤ poluyr°mmvn.
Stephan of Byzantium, s. v. ÑUlloÊala: Cf. Eustathius, ad. Hom. Il. XIV,
ÑUlloÊala, d∞mow Kar¤aw (. . .) êla 255: K«w (. . .) fas‹ d¢ toÁw Kãraw
går ofl Kçrew tÚn ·ppon ¶legon, …w ka‹ oÏtv kale›n tå prÒbata, ˜yen ka¤ ı
prÒteron e‡rhtai. n∞sow K«w …w poluyr°mmvn.

g°la ‘king’
soËa or soËan ‘tomb’

Stephan of Byzantium, s. v. Souãggela:

Souãggela, pÒliw Kar¤aw, ¶nya ı
tãfow ∑n toË KarÒw , …w dhlo› ka‹
toÎnoma. kaloËsi går ofl Kçrew soËan
tÚn tãfon, g°lan d¢ tÚn basil°a.

1
Reading mãnda in the two best manuscripts of Stephan.
 APPENDIX C

CARIAN NAMES IN GREEK SOURCES

A. PLACE NAMES (§: ZGUSTA KON)

Aba § 2–1. Bridaw § 173.

Agan/a/ § 8. Bubassow, Boubassow, Bubastow,
Aggvr/a/ § 65 Boubastow § 177.
Agorhsow § 10. Bvnitv § 159–2.
Adhssow § 17–3. Bvrand/a/ § 181.
Ayumbra § 25. Gerga § 202–1.
Alabanda § 37–4. Gordio/n/ § 215–3.
Alikarnassow § 44–7. D°dmasa § 250.
Alinda § 44–11. Didassai § 263.
Alkizv § 45. Dundason § 281.
Allianoi § 44–4. Enn/a/ § 297.
Allvss/ow/ § 50–2 Eorma § 298.
Amnist/ow/ cf. § 820, Blümel Erezow § 302–1.
KarON:165. Ermapilow § 305–1.
Amow § 60. Yaruai § 335.
Amuzvn, Amuzvn/a/ § 61–1. Yasyar/a/ § 336.
Amunand/a/ § 61–2. Yembrihmow, Yembrimow § 338.
Andanow § 66–3. Yemhs(s)ow, Yemissow § 339–1.
Anyemi § 71. Yigrow § 343.
Apodes[ Blümel KarON:165. Yudonow § 351.
Arara § 85–16. Yuhssow § 352–1.
Ardur/a/ § 90–6. Yumbria § 353–3.
Arissoullh Blümel KarON:165. Yussanouw § 355.
Arlai/a/ § 95–1. Yvdasa Blümel (KarON:167).
Arlissow § 95–2. Iasow § 358.
Armel/a/ Blümel KarON:165. Idriaw § 363.
Armokodvka § 96–1. Iduma § 364–1.
Arnaso/w/ § 97–3. Imbrow § 373–1.
Arpasa Lat. Harpasa § 98. Io.d- § 378.
Artoub/a/ § 100–2. Kaduih § 403–2.
Asshsow § 108–3. Kalbisso/w/ § 413.
Babein § 122–3. Kalunda § 414–2.
Bargasa (var. Pargasa, Bargaza) Kandasa § 426–2.
§ 135–1. Kandhb/a/, Kendhb/a/ § 428 (s. v.
Bargulia (later var. Barbulia) § 135–2. Kanduba).
Barkok≈mh Blümel (KarON:166). Kanhbion § 430.
Boll- § 158. Kaprima § 436.
 CARIAN NAMES IN GREEK SOURCES 457

Karbasuand/a/, Karpasuand/a/ Kubassow § 636.

§ 439–5. Kubim/a/ § 639–1.
Karoura § 452–2. Kubisyih § 639–2.
Karu/a/ § 454–3. Kubliss/ow/, hublis/ow/ § 1396.
Karuanda § 454–1. Kullandow § 645–1.
Kasa § 455–2 (s. v. †Kasaio/n/). Kumniss/ow/ § 646.
Kasar/a/ § 455–7. Kumvr/a/ § 647–2.
Kastabo/w/ § 458–1. Kuogrissiw § 649.
Kasvk/a/ § 461–1. Kuon § 652–2.
Kasvlaba § 461–2. Kuprand/a/ Blümel (KarON:172).
Kasvsso/w/ § 461–4. Kurbasa § 651–2.
Ka.nar/a/ § 423–6. Kuw § 652–1.
Kebialea § 471. Kushr/a/ § 653–1.
Kelimara Blümel (KarON:169). Kusshliw § 653–2.
Kemhsso/w/ Blümel (KarON:169). Kvrai/vn/ § 659.
Kendhbocorow § 477. Labara § 665.
Kenendvlab/a/ § 479. Labraunda (var. Larabiunda,
Keni- § 480. Labranda, Labrainda, Labrauunda,
Kepranow § 481. Labrenda, Lambraunda Labraenda)
Keraskord/a/ § 486. § 666.
Kecaro/w/ § 496–1. Lagina, Lageina § 670
K¤dram/a/ § 501. Lagnvk/a/ § 671.
Kildar/a/, Killar/a/ § 510. Laras/a/ § 688–2.
Kinduh, later var. Kunduh § 518–1. Larb/a/ § 689.
Kisariw § 522. Latmow § 696.
Koarbvnd/a/ § 538–1. Leibo/w/ (rather a person or god name)
Koar(r)end/a/, Koarenz/a/ § 704.
Koaranz/a/, Kvra(n)z/a/ § 538–2. Leukoid/a/ Blümel (KarON:173).
Kodap/a/ § 541. Lec- (epiclesis ZeÁw Lecunow, Lecinow.
Kodouvka § 544. cf. also Lepsia, a Carian island) § 706.
Kozanata § 547. Lhcimandow (different variants for the
Koliorg/a/, Kolierg/a/ § 552. inhabitants’ name: Lefsimãnioi,
Koloura § 558–1 (not in Blümel!). Lefs¤mandoi, LefsimandÇew,
Komurion § 658. LefsumandÇew, LhcimandÇew) § 707.
Komvond/a/ § 568. Lidh § 712–1.
Kond- Blümel (KarON:170). Lobold/a/ § 716.
Konodvrkond/a/ § 575. Lor.os/a/ § 720.
Korell/a/ § 581. Lurisso/w/ § 731.
Kormoskvn/a/ § 583–4. Lvm/a/ § 736.
Korrit/a/ § 589. Lvndarg/a/ § 737.
Kostobalo/w/, Kvstobalo/w/ § 662. Lvndokvmh Blümel (KarON:173).
Kot- Blümel (KarON:171). Lvr/a/ § 738.
Kot/a/ § 593–1. Lvruma § 739.
Kourb/a/ § 607. Lvsso/w/ § 740.
Crusa § 631. Madnas/a/, M°dmasow, Medmassa/
Kruassow § 632–2. Mednassa § 748.
Kuarda § 635. Mali/a/ § 756–2.
458 APPENDIX C

Masanvrada § 782. Pasand/a/, Pasada § 1015.

Massvn/a/ § 787. Passala § 1018.
Mastaura § 788. Patarous/a/ § 1022–3.
Maunn/a/, Maiunn/a/ § 793. Pedanass/ow/, Pedanass/ow/ § 1028,
Messaba § 804–1. 1059–2.
Mhyasai § 806. Peig°lasow § 1031.
Milhtow § 809. Peldek-ìt- § 1035.
Mniesu/a/ § 819. Phgasa § 1053–1.
MÒboll/a/, Mogola, Mvgla § 822. Phdasa, Pidasa § 1054–1.
Mokold/a/ § 828. Piginda § 1058–1.
Monnara pl § 832. Pidvssow, Pidossus § 1059–2.
Monogissa § 833. Pisymoi § 1065.
Mosoun/a/ § 842–1. Pisiliw § 1066–1.
Mugissow § 858. Pisuh § 1066–4.
Mudon- § 859. Pitaon, Pitaium § 1067–3.
Muhss/ow/ § 863. Pladas/a/, Platas/a/ § 1068, 1072.
Mulas(s)a (very late form: Milasa) Plamow § 1070.
§ 861–1. Plarasa § 1071.
Mundow § 862. Pluar/a/ § 1080.
Murshl/a/ § 866–1. Poluara § 1083–1.
Mursileia § 866–2. Pounomou/a/ § 1093–1.
Mvss/on/ § 871. Prinassow § 1101–1.
Naras/a/ § 885–3. Proposs/ow/ Blümel (KarON:178).
Narisbar/a/ § 886. Pruondr- Blümel (KarON:178).
Naruandow § 888. Purindow § 1114–1.
NinÒh § 898–1. Purnow § 1114–2.
Nouik/a/ Blümel (KarON:175). Pustow § 1116.
Jerasso/w/ § 907. Salei/a/ § 1148–2.
Ogond/a/ § 912. Salmakiw § 1150–1.
Ol/a/, Oul/a/ § 925. Samn/h/ § 1153.
Oloss/iw/ § 930–2. Samulia § 1152–2.
Olumo/w/ § 932. Saranso/w/ § 1165–2.
Omb/a/ § 934. Sasanda § 1176–2.
Ondoura § 935. Siana § 1261–1.
Orbhla § 938–4. Sikim/a/ Blümel (KarON:178).
Oryondouvk/a/ § 941. Sind/a/ § 1219–2.
Orsubli/a/ Blümel (KarON:176). Sindhssow § 1219–4
Orsvll/a/ § 950. Sinuri § 1222.
Otvrkond/a/ § 958. Solo/a/ § 1244–1.
Ouasso/w/ § 966. Solvn/a/ § 1244–3.
Palgosvlda § 996. Sparz/a/ § 1255.
Panamara § 1000. Suana § 1261–1.
Pandaj/a/ § 1001–2. Suaggela, Souaggela, Sfaggela (later
Parableia (Parablia?) § 1005. Yeaggela, Theangela) § 1261–2.
Parembvrd/a/ § 1007. Suarbeu[ § 1262.
Parkall/a/ § 1009. Suista § 1267.
Parpar- § 1012. Surna § 1272.
 CARIAN NAMES IN GREEK SOURCES 459

Svbala § 1274. Ualvka § 1393.

Svssow Blümel (KarON:179). Uarbesu/a/ § 1394.
Tabai § 1277–1. Ubliss/ow/ § 1396.
Tabarniw § 1277–5. Ugas(s)ow 1397.
Talagr/a/ § 1284. Udai, Kudai § 1398–2.
Tapass/oi/ § 1294. Udis(s)ow § 1398–4.
Taramptow § 1295. Uyubir/a/ § 1400.
Tarban/a/ § 1297–1. Uissow § 1402.
Tarbetv (?, or rather a personal name?) Ulim/a/ § 1404–1.
Blümel (KarON:179). Ullarima § 1404–2.
Tarkondar/a/ § 1299. Ullouala § 1404–3.
Tarm/ow/ (?, or rather a personal Umess/ow/ § 1405.
name?) § 1300. Urvmow, Kurvmow, Eurvmow, Eurvpow
Tezhra (?, or rather a personal name?) § 1412.
§ 1310. Usarbid/a/ § 1414.
Teleseitiw § 1312. Ussome[ Blümel (KarON:182).
Telmhssow, Telmessow, Telmissow, Utarm/ow/ Blümel (KarON:182).
Telemessow, Telmisum § 1314. Xalkhtvr, Xalkhetorew Blümel
Temoesso/ow/ Blümel (KarON:180). (KarON:182).
Tendhba § 1318. Vlasha § 1443.
Termera, Telmera, Termera § 1320–2. Vndr/a/ § 1444.
Terssvgass/ow/ Blümel (KarON:181). Vnzvssuaso/w/ § 1445.
Tnussow § 1347. Vspraonno/w/ § 1447.
Traldeiw, Tralleiw, Trallis § 1361–1. ]akondia Blümel (KarON:182).
Trara § 1362. ]eadovka § 1450.
Trobaliss/ow/ § 1368. ]erra Blümel (KarON:182).
Truban/a/ § 1374. ]hvka § 1457.
Tuennesso/w/ § 1379. ]kermu.ion Blümel (KarON:182).
Tumnhssow § 1384–5. ]nirea Blümel (KarON:182).
Tumnow § 1384–4.

B. PERSONAL NAMES (§; ZGUSTA KPN)

Aba, Abaw, Abbaw § 1–1/2/3/5. Andarsvw § 59–4.

Abersi § 5. Appa, Apfia, Apfiaw, Apfion, Apfianow,
Ada § 15–1. Afia, Afion, Affion, etc. § 66.
Ayuasiw § 128. Apoukvw § 79.
Akarmomeldvw § 27. Arbhs(s)iw § 85–2/3.
Aktadhmow § 38–1. Arduberow § 86–6.
Aktauassiw § 38–2. Ariauow § 89–2.
Aktaussvllow § 38–3. Aridvliw § 89–4.
Alasta § 42–1. Arlissiw § 95–1/2.
Alganiw § 44. Arlivmow § 95–3.
Alleaw § 52–3. Arris(s)iw § 106–1/2.
Amiaw, Ammiaow, Ammh, Ammin, Ammeiaw, Arshliw § 107–12.
etc. § 57. Artaow Blümel (KarPN:11).
460 APPENDIX C

Arthumow § 109. Isedum.xow § 485.

Artimhw Arteimhw § 108–4/6. Isemenda . . . ow § 486.
Artuassiw, Aryuassiw § 110–1/2. Iublhsiw § 494.
Aruassiw § 111. I . . . uagow § 1679.
Arvsiw Blümel (KarPN:11). Kay.divn § 1680.
Atthw § 119–10. Kakraw § 509–1.
Beryaw § 162–1. Kalbalaw § 512.
Berrablviow Blümel (KarPN:11). Karama . . . ow § 531.
Boivmow § 178–4. Karjaw § 539.
Bruajiw, Bruassiw § 196–1/2. Karreiw § 540–3.
Brvlvw § 197. Karusvldow § 544.
Geiw § 210–3. Kasballiw § 546.
Glouw § 224. Kasbvlliw § 545.
Gugow Blümel (KarPN:12). Kashsiw § 547–1.
Dandvmow § 251. Kaifenh § 558.
Daru . . . ow § 254. Kbondiassiw § 566.
Deibow § 264. Kbvdhw § 567–1.
Dersvmanhw Blümel (KarPN:12). Kebivmow Blümel (KarPN:16).
Dersvw Blümel (KarPN:12). Keldnassiw § 573.
Dersv . . . tiw § 275. Kemptuw § 575.
Ekamuhw Blümel (KarPN:12). *Kendhbhw § 576–9.
Ekatomnvw (more recent variants: Ketambissiw § 593.
Ekatomnvn, Ekatomnow) § 325–1/2/3. Kindacow (KarPN:16).
Ejamuhw § 340–2. Kinjimow § 617.
Ermapiw § 355–21. Kit.essvw (KarPN:16).
Zermeduberow (var. Jermedu<be>row) Koboldvow (KarPN:16).
Blümel (KarPN:13). Koibilow § 652–1.
Zonzolow § 390. Koidvw § 653.
Yekuilow § 417. Kolaldiw, Kulaldiw (KarPN:17).
Youw Blümel (KarPN:14). Koldobaw § 660.
Yualdiw § 438. Kolvldow § 661.
Yussow § 445–1 Kondalow § 676–1.
Ibanvlliw § 450. Kondmalaw § 676–2.
Idagugow § 451–4. Kondo[ § 676–5.
Idakow § 451–5. Korollow (KarPN:17).
Idbelaw Blümel (KarPN:14). Korriw § 686–3.
Ideghbow Blümel (KarPN:14). Koshtiow (KarPN:17).
Idmamu . . . ow § 452–6. Kosinaw § 703.
ÉIdrieÊw § 453. Kostvlliw § 705.
Idubl[ § 454. Kotbelhmow (KarPN:17) cf. Kutbelhmiw.
Idussvllow Blümel (KarPN:15). Kotobalvw (KarPN:17).
Il[.]uthw § 1678. Kouldoiw § 727.
Imbarhldow § 467. Kourvn § 737–5.
Imbarsiw, Imbras(s)iw § 469–3/4/5. Ktouboldow § 761.
Imbrhw § 469–9. Kuaremow § 764.
Innivn § 471–8. Kuatbhw § 765.
Indow § 473–1. Kulaldiw (KarPN 18) cf. Kolaldiw.
 CARIAN NAMES IN GREEK SOURCES 461

Kutbelhmiw § 771. O(?sa)rthumow § 1114–3.

Kutpiw § 772. Oseaw § 1121.
Kvbhw § 774–1. Ouvkhw § 1180.
Kvglvw § 775. Pagadow § 1186.
Latarshw § 799. Paktuhw § 1193.
Lugdamiw § 834. Patkuvlliw Blümel (KarPN:21).
Lujhw § 836. Paktuiskow Blümel (KarPN:21).
Makow § 848–1. Panablhmiw § 1197–3.
Malosvow (KarPN:18). Panamuhw § 1197–6.
Manhw § 858–1. Panuassiw § 1198.
Manitaw § 864–1. Paow Blümel (KarPN:21).
Mareuw § 873–6. Papaw, Papiaw, Papow, etc. § 1199.
Marow § 873–13. Paparivn § 1200–2.
Massarabiw § 880–2. Paraskvw § 1203–3.
Matiw § 882–6. Paraudigow § 1203–5.
Mausvllow § 885–1–6. Paraussvllow, Paraussvldow,
Memakow Blümel (KarPN:19). Parussvldow § 1203–6/8.
Metebiw Blümel (KarPN:19). Pargistaw § 1205.
Mindrvn § 921. Parmumiw Blümel (KarPN:22).
Minnaw, Minnion, Minniw, Minnh, Parnow § 1207.
Minnivn, Minnow, etc. § 922. Paruv § 1212–2.
Miskow/Miskvw § 929–2. Paruinna § 1212–1.
Mohnnow § 941. Passidhrow § 1219.
Moiw § 942. Pedvldow § 1232.
Mokollhw § 944–1. Pelaow Blümel (KarPN:22).
Molhw § 946–1. Peldemiw, Peldemvw
Monnhw § 959. § 1234–1.
Mosraiow Blümel (KarPN:19). Peldhkow § 1234–2.
Mouzeaw § 980–2. Pel(l)ekvw § 1234–3/4.
Naduw § 1008a. Pelkisiw § 1235.
Nana, Nanh, Nanaw, Nannh, Nannion, Perbilaw § 1239.
Nanniw, Nannixow, Nannow, Nannv, etc. Perignaw/Petignaw § 1241.
Narbaw § 1013. Phdisaw § 1249.
Neterbimow Blümel (KarPN:20). Pigassvw Blümel (KarPN:22).
Nonnh, Nonnow, etc. Pigrhw § 1255–6.
Nutar Blümel (KarPN:20). Pijvdarow § 1263–3.
Nvtrassiw Blümel (KarPN:20). Pirvmiw § 1266.
Oaloalow § 1134–2 (cf. Adiego 1993b). Pisindhliw § 1268.
Oa3a3iw § 1145–8. Pisku[ Blümel (KarPN:23).
Obrokaw Blümel (KarPN:20). Pis.nvw Blümel (KarPN:23).
Oletaw § 1085–1. Pitakolow (not †Gitakolow, Zgusta
Oliatow § 1085–2 § 221!) Blümel (KarPN:23).
Olohtow Blümel (KarPN:20). Pittaw § 1270–1.
Opinaw § 1096–2. Plouw § 1277–1.
Ordomaw § 1104–3. Ponmoonnow Blümel (KarPN:23).
Oridhumiw Blümel (KarPN:20). Ponusvllow § 1189.
Ortassiw § 1114–2. Pormounow Blümel (KarPN:23).
462 APPENDIX C

Purkeaw/Purkehw Tarmow Blümel (KarPN:25).

Saggotbhriw Blümel (KarPN:23). Tarv § 1515–2.
Saggvw § 1369. Tata, Tatarion, Tataw, Tateiw, Tath,
Samassiw § 1361. Tatia, Tatiaw, Tation, etc. § 1517.
Sampaktuhw, Sambaktuw § 1364–1. Tausaw § 1520.
Samvow § 1367–1. Tendessiw § 1534.
Samvuow § 1367–2. Territow § 1538.
Sanamvw Blümel (KarPN:24). Tiaimow § 1533.
Sanortow § 1371. Tobororow § 1577.
Sarow § 1377. Tounobow § 1592.
Saruassiw Blümel (KarPN:24). Totoliw § 1598.
Sarussvllow § 1378–1/2. Trus(s)hw § 1608.
Saskow/Saskvw § 1381. Truvlhw/Truvlow § 1609.
Sassvmow § 1379–6. Tumnhw/Tumnow § 1615.
Saurigow Blümel (KarPN:24). Uarkelaw Blümel (KarPN:26).
Sausvllow Blümel (KarPN:24). Uyesmaw Blümel (KarPN:26).
Seikilow § 1390. Uyhw Blümel (KarPN:26).
Semeuritow Blümel (KarPN:25). Uliatow § 1627.
Senurigow Blümel (KarPN:25). Urgaw Blümel (KarPN:26).
Seskvw § 1410–1. Urgilow Blümel (KarPN:26).
Sesvlhw § 1411. Urgosvw Blümel (KarPN:26).
Sibilvw § 1416–1. Ussaldow Blümel (KarPN:26).
Siduatow Blümel (KarPN:25). Ussaldvmow/Usseldvmow § 1629–4/6.
Sidulhmiw § 1422. Ussiw § 1629–1.
Silbow § 1426. Ussisiw § 1629–2.
Skoaranow Blümel (KarPN:25). Ussvihw/Ussviow § 1629–3.
Spareudigow § 1466. Ussvldow, Ussvllow, Ussvlow
Sueskurebow § 1477. § 1629–7/8.
S[u]s[k]h[w] (?)Blümel (KarPN:25). Xasbvw Blümel (KarPN:27).
Suskvw § 1486. Xhramuhw (cf. § 1639).
Svmnhw Blümel (KarPN:25).

Acephalic Forms (Blümel KarPN 27–28)

]alvldow ]rouessiw
]anvrremow ]ruassiw
]ketaw ]teieow
]kokvw ]toldiw
]ldoudhw ]uassiw
]llv[do]w ]ujki
]omvliw ]vldow
]ramow ]vllow
]rgigougou ]vrlemiw
 CARIAN NAMES IN GREEK SOURCES 463

RIVER NAMES
Idumow Tischler (1977:66). Maiandrow Tischler (1977:93–94), Cf.
Indos, (var.) Lindow (an erroneus KON § 752.
form?) Tischler (1977:67) cf. KON Marsuaw Tischler (1977:96–97).
§ 375. Morsunow Tischler (1977:102).
Kalbiw Tischler (1977:69). Salmakiw (a source) Tischler (1977:128).
Kenivw (var. Kinevw) Tischler (1977:78). Telmedius(?) Tischler (1977:143–144)
Kitvn (var. Keitvn) Tischler Timelhw, Teimelhw Tischler (1977:143,
(1977:80–81) 148). Cf. KON § 1338.
Kubersow Tischler (1977:85–86), KON Ubando/w/ Tischler (1977:64) Cf. KON
§ 639–4). § 1395.

GOD NAMES IN LITERARY SOURCES

ÖImbramow, var. ÖImbrasow (= Hermes; Mãsariw (= Dyonisus; St. Byz. s. v.

St. Byz. s. v. ÖImbrow, Scholia vetera in Mãstaura)
Theogoniam v. 338, and Eustathius, ÉOsog«a (= Zenoposeidon; Strabo XIV,
Commentarii ad Iliadem XIV, 281) 659, Pausanias VIII; 10, 4).
 APPENDIX D

CONCORDANCES

A. PRESENT BOOK FORMER EDITIONS

EGYPT

Sais E.Me 32 M 24
E.Sa 1 MY L E.Me 33 M 25
E.Sa 2 MY M E.Me 34 M 26
E.Me 35 M 27
Memphis E.Me 36 M 28
E.Me 1 MY A E.Me 37 M 29
E.Me 2 MY B E.Me 38 M 30
E.Me 3 MY D E.Me 39 M 31
E.Me 4 MY E E.Me 40 M 32
E.Me 5 MY F E.Me 41 M 33
E.Me 6 MY G E.Me 42 M 34
E.Me 7 MY H E.Me 43 M 35
E.Me 8 MY K E.Me 44 M 36
E.Me 9 M 1 E.Me 45 M 37
E.Me 10 M 2 E.Me 46 M 38
E.Me 11 M 3 E.Me 47 M 39
E.Me 12 M 4 E.Me 48 M 40
E.Me 13 M 5 E.Me 49 M 41
E.Me 14 M 6 E.Me 50 M 42
E.Me 15 M 7 E.Me 51 M 43
E.Me 16 M 8 E.Me 52 M 44
E.Me 17 M 9 E.Me 53 M 45
E.Me 18 M 10 E.Me 54 M 45a
E.Me 19 M 11 E.Me 55 M 46
E.Me 20 M 12 E.Me 56 M 47
E.Me 21 M 13 E.Me 57 M 47a
E.Me 22 M 14 E.Me 58 M 47b
E.Me 23 M 15 E.Me 59 M 48
E.Me 24 M 16 E.Me 60 M 48a
E.Me 25 M 17 E.Me 61 M 48b
E.Me 26 M 18 E.Me 62 M 48c
E.Me 27 M 19 E.Me 63 M 48d
E.Me 28 M 20 E.Me 64 M 49
E.Me 29 M 21 E.Me 65 Abusir
E.Me 30 M 22 E.Me 66 Kammerzell
E.Me 31 M 23 *180
 CONCORDANCES 465

Abydos Thebas
E.Ab 1 Ab 1 F E.Th 1 Th 47 ”
E.Ab 2 Ab 2a F E.Th 2 Th 48 ”
E.Ab 3 Ab 2b F E.Th 3 Th 49 ”
E.Ab 4 Ab 3b F E.Th 4 Th 50 ”
E.Ab 5 Ab 3c F E.Th 5 Th 51 ”
E.Ab 6 Ab 4 F E.Th 6 Th 52 ”
E.Ab 7 Ab 5a F E.Th 7 Th 53 ”
E.Ab 8 Ab 5b F E.Th 8 Th 54 ”
E.Ab 9 Ab 5c F E.Th 9 Th 55 ”
E.Ab 10 Ab 6 F E.Th 10 Th 56 ”
E.Ab 11 Ab 7 F E.Th 11 Th 57+58 ”
E.Ab 12 Ab 8a F E.Th 12 Th 59 ”
E.Ab 13 Ab 8b F E.Th 13 Th 60 ”
E.Ab 14 Ab 9 F
E.Ab 15 Ab 10 F Luxor
E.Ab 16 Ab 11 F E.Lu 1 G 19
E.Ab 17 Ab 12 F E.Lu 2 G 21
E.Ab 18 Ab 13a F E.Lu 3 G 22
E.Ab 19 Ab 13b F E.Lu 4 G 23
E.Ab 20 Ab 14 F E.Lu 5 G 24
E.Ab 21 Ab 15 F E.Lu 6 G 25
E.Ab 22 Ab 16 F E.Lu 7 G 26
E.Ab 23 Ab 17 F
E.Ab 24 Ab 18 F Murwàw
E.Ab 25 Ab 19 F E.Mu 1 ¥aba (1971), nº 196
E.Ab 26 Ab 20 F
E.Ab 20 Ab 14 F
Silsilis
E.Ab 21 Ab 15 F
E.Si 1 Si 39 F
E.Ab 22 Ab 16 F
E.Si 2 Si 53 F
E.Ab 23 Ab 17 F
E.Si 3 Si 54 F
E.Ab 27 Ab 21 F
E.Si 4 Si 55 F
E.Ab 28 Ab 22 F
E.Si 5 Si 56 F
E.Ab 29 Ab 24 F
E.Si 6 Si 57 F
E.Ab 30 Ab 25 F
E.Si 7 Si 58 F
E.Ab 31 Ab 26a F
E.Si 8 Si 59 F
E.Ab 32 Ab 26b F
E.Si 9 Si 60 F
E.Ab 33 Ab 27 F
E.Si 10 Si 61 F
E.Ab 34 Ab 28 F
E.Si 11 Si 62 F
E.Ab 35 Ab 29 F
E.Ab 36 Ab 8 Y
E.Ab 37 Ab 9 Y Abu Simbel
E.Ab 38 Ab 15 Y E.AS 1 AS 1
E.Ab 39 Ab 26 Y E.AS 2 AS 2
E.Ab 40 Ab 27 Y E.AS 3 AS 3
E.Ab 41 Ab 28 Y E.AS 4 AS 4 + Lepsius
E.Ab 42 Ab 29 Y Kar 4
E.Ab 43 Ab 34 Y E.AS 5 AS 5
466 APPENDIX D

E.AS 6 AS 6 Gebel Sheik Suleiman

E.AS 7 AS 7 E.SS 1 72 F
E.AS 8 AS 8
E.AS 9 Lepsius Kar 2 Unknown origin, likely to be from
Egypt
Buhen E.xx 1 MY C
E.Bu 1 M 50 E.xx 2 MY I
E.Bu 2 M 51 E.xx 3 MY a
E.Bu 3 M 52 E.xx 4 MY b
E.Bu 4 M 53 E.xx 5 MY c
E.Bu 5 M 54 E.xx 6 4 ”
E.Bu 6 M 55 E.xx 7 Lion

CARIA

Tralles Stratonikeia
C.Tr 1 D 1 C.St 1 D 12
C.Tr 2 D 2 C.St 2 36*

Alabanda and surroundings Halikarnassos

C.Al 1 D 13 C.Ha 1 33*

Euromos Didyma
C.Eu 1 D 3 C.Di 1 21*
C.Eu 2 D 8
Iasos
Kindye C.Ia 1 20a*
C.Kn 1 D 6 C.Ia 2 20b*
C.Ia 3 38a*
Hyllarima C.Ia 4 38b*
C.Hy 1 D 7 + C.Ia 5 47*
Adiego- C.Ia 6 48*
Debord- C.Ia 7 Berti-Innocente 2005
Varinlio<lu 2005
Keramos
Mylasa C.Ke 1 39a*
C.My 1 Blümel-Kızıl 2004 C.Ke 2 39b*

Sinuri Kaunos
C.Si 1 D 9 C.Ka 1 D 14
C.Si 2 D 10 C.Ka 2 D 16
C.Ka 3 28*
Kildara C.Ka 4 30*
C.Ki 1 D 11 C.Ka 5 44*
C.Ka 6 45*
 CONCORDANCES 467

C.Ka 7 46* C.xx 3 40*

C.Ka 8 49* C.xx 4 41a*
C.Ka 9 50* C.xx 5 41b*

Lycia Greece
C.Kr 1 D15 G.1 D 19
G.2 42*
Unknown origin, likely to be from
Caria
C.xx 1 34*
C.xx 2 35*

B. FORMER EDITIONS ➝ PRESENT BOOK

For the Egyptian inscriptions the order adopted in Meier-Brügger
(1979b) is followed here.

EGYPT

Abu Simbel MY C E.xx 1

(Masson 1979) MY D E.Me 3
AS 1 E.AS 1 MY a E.xx 3
AS 2 E.AS 2 MY b E.xx 4
AS 3 E.AS 3 MY c E.xx 5
AS 4 E.AS 4 (+ Lepsius MY E E.Me 4
Kar 4) MY F E.Me 5
AS 5 E.AS 5 MY G E.Me 6
AS 6 E.AS 6 MY H E.Me 7
AS 7 E.AS 7 MY I E.xx 2
AS 8 E.AS 8 MY K E.Me 8
MY L E.Sa 1
Buhen MY M E.Sa 2
(M = Masson 1978)
M 50 E.Bu 1 ‘Leningrad Isis’
M 51 E.Bu 2 (” = ”evoro“kin 1965)
M 52 E.Bu 3 4 ” E.xx 6
M 53 E.Bu 4
M 54 E.Bu 5 ‘Lion’
M 55 E.Bu 6 (Masson 1976)
Lion E.xx 7
Gebel Sheik Suleiman
(F = Friedrich 1932) Memphis-Saqqara
GSS 72 F E.SS 1 (M = Masson 1978)
M 1 E.Me 9
‘Pharaonic objects’ M 2 E.Me 10
(MY = Masson-Yoyotte 1956) M 3 E.Me 11
MY A E.Me 1 M 4 E.Me 12
MY B E.Me 2 M 5 E.Me 13
468 APPENDIX D

M 6 E.Me 14 M 48b E.Me 61

M 7 E.Me 15 M 48c E.Me 62
M 8 E.Me 16 M 48d E.Me 63
M 9 E.Me 17 M 49 E.Me 64
M 10 E.Me 18 Abusir E.Me 65
M 11 E.Me 19 Kammerzell (1993)
M 12 E.Me 20 *180 E.Me 66
M 13 E.Me 21
M 14 E.Me 22 Silsilis
M 15 E.Me 23 (F = Friedrich 1932)
M 16 E.Me 24 Si 39 F E.Si 1
M 17 E.Me 25 Si 53 F E.Si 2
M 18 E.Me 26 Si 54 F E.Si 3
M 19 E.Me 27 Si 55 F E.Si 4
M 20 E.Me 28 Si 56 F E.Si 5
M 21 E.Me 29 Si 57 F E.Si 6
M 22 E.Me 30 Si 58 F E.Si 7
M 23 E.Me 31 Si 59 F E.Si 8
M 24 E.Me 32 Si 60 F E.Si 9
M 25 E.Me 33 Si 61 F E.Si 10
M 26 E.Me 34 Si 62 F E.Si 11
M 27 E.Me 35
M 28 E.Me 36 Thebes
M 29 E.Me 37 (” = ”evoro“kin 1965)
M 30 E.Me 38 Th 47 ” E.Th 1
M 31 E.Me 39 Th 48 ” E.Th 2
M 32 E.Me 40 Th 49 ” E.Th 3
M 33 E.Me 41 Th 50 ” E.Th 4
M 34 E.Me 42 Th 51 ” E.Th 5
M 35 E.Me 43 Th 52 ” E.Th 6
M 36 E.Me 44 Th 53 ” E.Th 7
M 37 E.Me 45 Th 54 ” E.Th 8
M 38 E.Me 46 Th 55 ” E.Th 9
M 39 E.Me 47 Th 56 ” E.Th 10
M 40 E.Me 48 Th 57–58 ” E.Th 11
M 41 E.Me 49 Th 59 ” E.Th 12
M 42 E.Me 50 Th 60 ” E.Th 13
M 43 E.Me 51
M 44 E.Me 52
M 45 E.Me 53 Abydos
(F = Friedrich 1932)
M 45a E.Me 54
(Y = Yoyotte apud Meier-Brügger
M 46 E.Me 55
1979)
M 47 E.Me 56
Ab 1 F E.Ab 1
M 47a E.Me 57
Ab 2a F E.Ab 2
M 47b E.Me 58
Ab 2b F E.Ab 3
M 48 E.Me 59
Ab 3a F excluded
M 48a E.Me 60
 CONCORDANCES 469

Ab 3b F E.Ab 4 Luxor
Ab 3c F E.Ab 5 (ESS 1998)
Ab 4 F E.Ab 6 G 19 E.Lu 1
Ab 5a F E.Ab 7 G 21 E.Lu 2
Ab 5b F E.Ab 8 G 22 E.Lu 3
Ab 5c F E.Ab 9 G 23 E.Lu 4
Ab 6 F E.Ab 10 G 24 E.Lu 5
Ab 7 F E.Ab 11 G 25 E.Lu 6
Ab 8a F E.Ab 12 G 26 E.Lu 7
Ab 8b F E.Ab 13
Ab 9 F E.Ab 14 Caria and other Locations
Ab 10 F E.Ab 15 D 1 C.Tr 1
Ab 11 F E.Ab 16 D 2 C.Tr 2
Ab 12 F E.Ab 17 D 3 C.Eu 1
Ab 13a F E.Ab 18 D 4 excluded
Ab 13b F E.Ab 19 D 5 excluded
Ab 14 F E.Ab 20 D 6 C.Kn 1
Ab 15 F E.Ab 21 D 7 C.Hy 1
Ab 16 F E.Ab 22 D 8 C.Eu 2
Ab 17 F E.Ab 23 D 9 C.Si 1
Ab 18 F E.Ab 24 D 10 C.Si 2
Ab 19 F E.Ab 25 D 11 C.Ki 1
Ab 20 F E.Ab 26 D 12 C.St 1
Ab 14 F E.Ab 20 D 13 C.Al 1
Ab 15 F E.Ab 21 D 14 C.Ka 1
Ab 16 F E.Ab 22 D 15 C.Kr 1
Ab 17 F E.Ab 23 D 16 C.Ka 2
Ab 21 F E.Ab 27 D 17 excluded
Ab 22 F E.Ab 28 D 18 coin legends
Ab 23 F excluded D 19 G.1
Ab 24 F E.Ab 29 20a* C.Ia 1
Ab 25 F E.Ab 30 20b* C.Ia 2
Ab 26a F E.Ab 31 21* C.Di 1
Ab 26b F E.Ab 32 22* excluded
Ab 27 F E.Ab 33 23* excluded
Ab 28 F E.Ab 34 24* excluded
Ab 29 F E.Ab 35 25* excluded
Ab 30 F excluded 26* excluded
Ab 8 Y E.Ab 36 27* excluded
Ab 9 Y E.Ab 37 28* C.Ka 3
Ab 15 Y E.Ab 38 29* excluded
Ab 26 Y E.Ab 39 30* C.Ka 4
Ab 27 Y E.Ab 40 31* excluded
Ab 28 Y E.Ab 41 32* excluded
Ab 29 Y E.Ab 42 33* C.Ha 1
Ab 34 Y E.Ab 43 34* C.xx 1
470 APPENDIX D

35* C.xx 2 48* C.Ia 6

36* C.St 2 49* C.Ka 8
37* excluded 50* C.Ka 9
38a* C.Ia 3 51* excluded
38b* C.Ia 4 Belli-
39a* C.Ke 1 Gusmani 2001
39b* C.Ke 2 [rock inscription
40* C.xx 3 from Labraunda] excluded
41a* C.xx 4 Innocente 2002
41b* C.xx 5 [‘tegola di Iasos’] excluded
42* G.2 Blümel-
43* excluded Kızıl 2004 C.My 1
44* C.Ka 5 Berti-
45* C.Ka 6 Innocente 2005 C.Ia 7
46* C.Ka 7 Adiego-Debord-
47* C.Ia 5 Varinlio<lu 2005 C.Hy 1[+D 7]
 APPENDIX E

COIN LEGENDS IN CARIAN

Koray Konuk

The following catalogue aims to include all known coins bearing let-
ters in the Carian script.1 These coins form an integral part of the writ-
ten record, and, in spite of their small size, they are of great importance,
as they throw light on various aspects of Carian society and its lan-
guage in particular. The material, collected over several years, comes
from a variety of sources.2 While many coins are without a provenance,
some have been found locally and are today housed near their find
spots in museums and private collections. This may provide useful infor-
mation for their attribution which in many cases remains a difficult
matter. Our purpose here, however, is to focus on the legends and
questions of attributions are only very briefly discussed, especially when
specific studies are available. The catalogue presents first coins which can
be ascribed to a mint, then coins whose attribution remains uncertain.
Coin legends are obviously related to inscriptions, but there are
significant differences between them. Each coin issue was produced in
thousands of specimens, even if today only a handful, in some cases
one or two specimens, are extant. They were struck with dies which
had to be individually engraved in negative. The engraver cut the mir-
ror-like image of the design (type) and the letters on the die which,
when struck on a piece of metal, appeared in positive. Working in neg-
ative could result in confusion in the direction and position of the let-
ters. Carian was inscribed in either direction, though more often from

1
Much of the discussion and many of the attributions presented here and elsewhere
were first made public in a paper read at the Royal Numismatic Society in January
1996 and entitled ‘Carian Coin Legends’ (hereafter ‘1996 RNS paper’), of which this
is an extended and updated version. I am very grateful to Professor Ignacio Adiego
for kindly including this appendix in his book and for his useful comments, and to
Richard Ashton for improving my text.
2
For nearly 70 years, Robinson 1939 remained the only comprehensive study of
coin legends in Carian (listing fewer than a dozen examples). Not only has our under-
standing of Carian dramatically increased, but our documentation has also quadrupled.
472 APPENDIX E

left to right, especially on later inscriptions, and it is not always clear

whether on coins the right-to-left direction was meant or was simply
the result of the engraver’s confusion. In the description of the obverse
and reverse of each coin, legends are first transcribed between brack-
ets as they appear on the coin. When needed, questions pertaining to
the direction and reading of the legends are discussed.
One specimen has been cited to illustrate each variety. I have tried
to select the best preserved specimen, which is not necessarily the one
illustrated in standard catalogues. The weight is followed, when known,
by the die-axis. All coins are silver unless otherwise indicated.

I. MINTS

Mylasa
For detailed discussion and attribution to Mylasa, see Konuk (forth-
coming [a]). The mint of Mylasa was first suggested in my 1996 RNS
paper and appeared in print in Konuk (1998a:22–26) in which the last
two letters of M5 (my) were read as the beginning of the ethnic of
Mylasa in Carian. See also SNG Kayhan, 833–840. All silver fractions
are on the Milesian standard.

M1 M
Obv. Forepart of lion left; on its shoulder, O; below, one foreleg left.
Rev. Two rectangular punches applied separately, one of which has
M1 (m).
Lydian (Persic) standard stater; c. 500 B.C.
Weber, 6448 (11.13g) = Naville 14 (1929), 378.
The Lydian-weight lion forepart issues were quite prolific and rank
among the earliest coinages of Caria. Their attribution is debated,
Kaunos and Mylasa have been suggested (for an overview, see
Konuk 2000a:172 and Konuk forthcoming [a]). Several phases
can be observed which span the second half of the sixth century
B.C. M1 comes late in the sequence of minting and is linked to
the issues which have various signs on the shoulder of the lion.
These are not letters but linear devices. M1 (m), engraved in one
of the two rectangular punches, is attested on a few dies and,
although the possibility cannot be entirely ruled out, it does not
 COIN LEGENDS IN CARIAN 473

appear to be a random mark. Subsequent issues attributed to

Mylasa carry letters only sporadically (see below), and most of its
early coinage is uninscribed. If M was intended as a Carian let-
ter, it would represent the earliest occurrence of the initial of the
Carian ethnic of Mylasa.

M2 M
Obv. Head of lion left; below, one foreleg left; dotted circle.
Rev. Bird standing left, wings open; below left, M2 (m); all within incuse
square.
Tetartemorion; c. 500–450 B.C.
Klein (1999:no 499) (0.24g) = Hauck & Aufhäuser 18 (2004), 250
= Hirsch 187 (1995), 423.
This issue, of which three specimens are known to me (one is a
probable die duplicate in the Bodrum Museum of Underwater
Archaeology, 1–20–78 [0.24g, 12H], the other is in Oxford,
Ashmolean Museum [0.29g; 12H]; for a similar but anepigraphic
early example, see SNG Kayhan, 939), is to be linked to a sub-
stantial series of Milesian-weight tetartemoria of the same type but
later style (e.g. SNG Kayhan, 940–948; many are housed in the
archaeological museums of Milâs and Bodrum). These have been
traditionally attributed to Miletos on the basis of the weight stan-
dard and obverse type. This early issue, however, depicts the lion
in a very different way. Here, as on the Kayhan specimen, the
lion’s head has a straightforward posture (as on M1), whereas on
Milesian issues, the lion is depicted as a forepart with its head
turned back and the foreleg reversed. The lion’s forepart evolved
on later issues and came to be modelled on the Milesian type. In
this case there is no doubt that M2 (m) represents the initial of
the Carian ethnic of Mylasa.

M3 M
Obv. Forepart of roaring lion right, head turned back; below, reversed
foreleg; linear outline of its back between jaws.
Rev. Facing head of a lion with its forelegs on either side, in upper
right corner, M3 (m); all within incuse square.
Tetartemorion; c. 450–420 B.C.
Bodrum Museum of Underwater Archaeology, 3–26–91 (0.35g;
07H).
474 APPENDIX E

Also part of an extensive series, only a handful of inscribed spec-

imens is known. Among these a variety of the same early style
has with a smaller M standing next to the middle of the right
foreleg of the lion (Bodrum Museum of Underwater Archaeology,
10–7–91 [0.45g; 06H]). As on the previous example, M3 (m) stands
for the initial of the Carian ethnic of Mylasa.

M4 m
Obv. As last.
Rev. Young male head (Apollo?) facing; in right field, M4 (m); all within
shallow round incuse.
Tetartemorion; c. 420–390 B.C.
Künker 62 (2001), 129 (0.20g).
A variety features the letter in the left field (Oxford, Ashmolean
Museum (0.28g; 06H). For later issues, see M9 and M10 below.
M4 is the early phase of a coinage which ends with the tetarte-
moria in the name of the satrap Hekatomnos inscribed EK and
EKA (see below).

M5 V mW
Obv. As last.
Rev. Facing head of a lion with its forelegs on either side; below, M5
(w my); all within incuse square.
Hemiobol; c. 420–390 B.C.
SNG von Aulock, 7807 (0.42g) = Troxell (1984:n° 1A).
The style of this coin looks later than M3 (same type). M, m and
m are the same letter in the Carian alphabet (m). The next letter
is somewhat problematic. Adiego suggests that W was V in the
alphabet of Mylasa. If that is the case, the occurrence of both let-
ters in the same legend calls for an explanation. Even though the
new inscription found near Mylasa (Kırca<ız) does not include the
letter W, M5 proves that it was part of the city’s alphabet. Two
explanations spring to mind: either these letters are not the same
and therefore have different values, or V is not a letter but a tri-
dent as suggested by Troxell (1984:250). When the trident expla-
nation was proposed (Troxell considered mW to be Greek letters),
it seemed the most likely solution, as the Carian letter V was
unknown in that shape (apart from the rare occurrence of V at
Sinuri and Kildara) and no evidence existed at that time that it
 COIN LEGENDS IN CARIAN 475

was part of the alphabet of Mylasa. However, tridents are not

normally depicted on coins in such a simple way with mere strokes
(as with a letter): there is usually some ornamentation (arrow-like
tips of the tines, often volutes departing from the base of the shaft)
even when the size is minute. Moreover, the occurrence of Ú and
V on M6–M10 below in the same position on the reverse strongly
suggests that V is a letter, for Ú is definitely a Carian letter, even
though it is not attested on the new inscription from Kırca<ız. I
would be tempted to suggest that Ú is a variant of V, the former
being perhaps an earlier form of the latter. If so, the value of Ú
(w) would represent the same value for V. Finally, what is the
value of W which is also absent from the new inscription? The
suggestion that W and V are the same letters cannot be main-
tained on the basis of M5. Adiego gives W, a rather common let-
ter, the value y. The two letters mW (my) would thus plausibly
represent the first two letters of the Carian ethnic of Mylasa
(Konuk 1998a:23). For a discussion of V and a possible attribu-
tion to Hyssaldomos, see M9 below.

M6 V
Obv. As last.
Rev. As last but M6 (w); the incuse is round.
Hemiobol; c. 420–390 B.C.
Pfeiler (1962:20, 2) = Konuk (1998:223, 130).

M7 Ú
Obv. As last.
Rev. As last but M7 (w).
Hemiobol; c. 420–390 B.C.
Private collection (0.51g; 09H).

M8 Ú
Obv. As last but M8 (w) on lion’s muzzle.
Rev. As last but anepigraphic.
Hemiobol; c. 420–390 B.C.
Muharrem Kayhan collection, MK1231 (0.50g; 03H).
The position of Ú on the lion’s muzzle reminds one of the Milesian-
type tetrobols, diobols and obols of the Carian satrap Hekatomnos
which have on the lion’s muzzle the Greek letters EKA, EK and
E respectively. For a possible attribution to Hyssaldomos, see M9.
476 APPENDIX E

M9 Ú
Obv. As last.
Rev. Young male head (Apollo?) facing; in the lower right field, M9
(w); all within shallow round incuse.
Tetartemorion; c. 420–390 B.C.
Private collection (0.23g; 12H).
The letter is sometimes placed just below the facing head (e.g.
New York, ANS, 1980.23.5 [0.22g] = Troxell [1984:no 2A], M9
misdiscribed as a trident). These tetartemoria are distinguished by
a different reverse type. Tetartemoria of the same types but later
style were struck by Hekatomnos who put the Greek letters EKA
or EK in place of Ú or V. In Konuk 1998a: 22–26, I suggested
that Ú and (M5–M6 and M10) V (w) might also be the initial of
the name of a Carian dynast preceding Hekatomnos. His father’s
name is Hyssaldomos and an attribution to him is quite likely.

M10 V
Obv. As last.
Rev. Young male head (Apollo?) facing, turned slightly left; in the lower
left field, M10 (w); all within round incuse.
Tetartemorion; c. 420–390 B.C.
New York, ANS, 1983.53.464 (0.23g; 12H)

M11 Ú
Obv. As last.
Rev. Male head (Apollo?) right; in the lower left field, M11 (w); all
within shallow round incuse.
Hemitetartemorion; c. 420–390 B.C.
Private collection (0.14g; 10H).

Kasolaba?
For a likely attribution to the mint of Kasolaba, see Konuk (forth-
coming [b]). Given the wide time-span during which these coins were
issued, probably over a century, the legend is bound to refer to an eth-
nic rather than a dynast. M20 and M21 bear three letters: a9o (azo),
which at first glance are difficult to match with an ethnic. A large num-
ber of these coins occur in the collections of the archaeological muse-
ums of Milâs and Bodrum, and several find spots have been recorded,
which fall in the area between Mylasa and Halikarnassos. By studying
 COIN LEGENDS IN CARIAN 477

the Athenian Tribute Lists and the recently discovered inscription from
Sekköy, Descat (1994:66–68) demonstrates that the city of Kasolaba
ought to be located in that area. The reading azo shows a remarkable
similarity to the Greek ethnic of Kasolaba. The omission of a guttural
initial in the Carian legend should not be surprising since examples of
ethnics like Kyromos / Hyromos / Euromos, Kydai / Hydai and
Kyblissos / Hyblissos in the same district testify that such variations
were frequent. Kasolaba is the Greek transcription of a Carian ethnic
whose native spelling remains uncertain. It has been suggested, how-
ever, that ksolbz (ksolb≤ ) found in an inscription from Egypt (E.Me
43) may have been the genitive form of Kasolaba in Carian. Whether
or not this is the case, the coins suggest that the Carian ethnic started
with azo. It would not be far-fetched to expect that a new Greek inscrip-
tion with the form Hasolaba may one day come to light. All coins are
Milesian-standard hemiobols.

M12 3
6
Obv. Head of ram right.
Rev. Young male head left; to the left, M12 (az); all within incuse
square.
Private collection (0.43g; 09H); c. 450–400 B.C.
M12 and M13 show a distinctive archaic style; the deep incuse
square of the reverse also indicates an early issue. The initial let-
ter, a on later issues, takes a rather odd shape on these early
examples.

M13 4
5
Obv. As last.
Rev. Young male head right; to the right, M13 (az); all within incuse
square.
SNG Keckman, 865 (0.29g; 06H) = Troxell (1984:no 8); c. 450–400 B.C.

M14 3
“
Obv. As last.
Rev. Young male head right; to the right, M14 (az); all within incuse
square
Hauck & Aufhäuser 15 (2000), 206 (0.52g); c. 450–400 B.C.
478 APPENDIX E

The shape of the initial letter establishes a link between the first
phase of this coinage and the later issues (M15 onwards).
Same types from the next coin onwards unless otherwise indicated:
Obv. Head of ram right.
Rev. Young male head right; on either side, M15 (za); M16 (za); M17
(za); M18 (az); M19 (az); M20 (azo); M21 (azo); all within square
or round incuse.

M15 8 A
Bodrum Museum of Underwater Archaeology, 31–9–85 (0.42g;
01H).
Only the incuse square variety is known. The letter on the right
also has the shape M on a specimen in Oxford, Ashmolean Museum
(0.43g; 12H). There is also a specimen with Q in the Bodrum
Museum of Underwater Archaeology, 9–19–91 (0.51g; 12H).

M16 8 l
Private collection (0.34g; 02H).
This specimen has a round incuse. The letter on the right also
has the shape À on a specimen in SNG Keckman, 870 (0.46g; 09H;
a die-duplicate is in the Ashmolean, Oxford).

M17 9 a
Muharrem Kayhan collection, MK 1236 (0.49g; 06H).
The incuse on the Kayhan specimen is square; the round variety
is also attested (e.g. SNG Keckman, 869 [0.46g; 09H]). A variety
with 8 is known (SNG Keckman, 877 [0.39g; 03H]).

M18 a 9
New York, ANS (0.41g; 06H) = Troxell (1984:no 9B).
On M18–M21, when the shape of the incuse can be determined,
it is circular, and on some very shallow.

M19 a (obv.)
a 9 (rev.)
Obv. Head of ram right; below, M19 (a).
Private collection (0.36g; 10H).
 COIN LEGENDS IN CARIAN 479

M20 a 9o
SNG Kayhan, 997 (0.38g; 12H).

M21 a 9o
Obv. Persian hero-king right, in running-kneeling position, holding dag-
ger in the right hand and bow in the left; groundline.
London (BM), CM 1999–10–7–1 (0.34g; 6H).

Keramos
For the attribution to Keramos, see Konuk (2000), based on the con-
vincing attribution to Keramos by Ashton (1998) of a slightly earlier
coinage of the same types bearing the Greek letters KE.

M22 kBo
Obv. Bull standing right.
Rev. Dolphin leaping right; below, M22 (kbo).
AE chalkous; c. 400 B.C.
Private collection (1.05g; 04H) = Konuk (2000: no 2).
An obverse variety of this type has the forepart of the bull (Konuk
2000: no 1). M22 (kbo) is the beginning of the Carian ethnic of
Keramos.

M23 _NW (obv.)

luo (rev.)
Obv. Bull standing right; in front, M23 ( jy or kse).
Rev. Dolphin leaping right; underneath, M23 (kbo).
AE chalkous; c. 400 B.C.
SNG Kayhan, 804 (0.91g; 09H) = Konuk (2003:no 74) = Konuk
(2000:no 5).
The reverse legend of M23 is upside-down. In the present form,
the letters resemble a Greek delta, an upsilon and an omicron.
The most likely explanation is a mistake made by a Greek die-
cutter who, working in negative on the die, was led into error by
his mother-tongue. The variety with the obverse legend (which
might be Greek) appears to fall at the end of the series after which
Carian legend chalkoi were superseded by Greek legend chalkoi.
480 APPENDIX E

Kaunos
For a detailed discussion of this coinage, see Konuk (1998b). The definite
attribution to Kaunos was first presented in my 1996 RNS paper. It
was based on the reading of the coin legends with the new values given
to the Carian script. The initial of the ethnic on M24 and M25 (k)
and the subsequent legends (M26–M28) giving a second letter (b),
resulted in kb which is the beginning of the native name of Kaunos:
Kbid-, known from the Lycian version of the trilingual inscription of
the Letôon. In summer 1997, the discovery of a bilingual inscription
in Kaunos (C.Ka 5) gave for the first time the ethnic of Kaunos in
Carian and the first two letters on the new inscription were the same
as on the coin legends.
The earliest coinage of Kaunos is anepigraphic and spans the period
c. 490–450 B.C. M24 (k), the first occurrence of the beginning of the
Carian ethnic of Kaunos, appears towards the middle of the fifth cen-
tury B.C. The use of Carian ends with the chalkoi in c. 370 B.C.; these
are followed by chalkoi bearing the first three letters the ethnic of
Kaunos in Greek.
The following coins are Aeginetic standard staters unless otherwise
indicated.

M24 k
Obv. Female deity (Iris?) with curved wings and outstretched hands
flying left, looking right; holding a kerykeion in right hand and a
wreath in left.
Rev. Granulated patterns on either side of triangular baetyl; above left,
M24 (k); all within incuse square.
Paris, BN, 703 (11.76g; 09H) = Konuk (1998b:no 90a) (c. 450–430
B.C.).
Variants with the granulated patterns in the shape of stylised birds
(Konuk 1998b:no 95 [c. 430–410 B.C.]) or bunches of grapes are
known (Konuk 1998b:no 96 [c. 430–410 B.C.]).

M25 k
n
Obv. As last.
Rev. Bunch of grapes on either side of triangular baetyl with M25 (n)
in its centre; above left, M25 (k); all within incuse square.
 COIN LEGENDS IN CARIAN 481

London, BM (11.27g; 12H) = Konuk (1998b:no 99bisx) (c. 430–410

B.C.).
n (n) is added to a die in whose original state the baetyl was
anepigraphic.

M26 k 5
Obv. As last.
Rev. M26 (kb) flanking triangular baetyl; all within incuse square.
Sotheby 27 Oct. 1993 (Zürich), 694 (11.57g) = (Konuk 1998b:no
100h) (c. 410–390 B.C.).
Variants with the granulated patterns in the shape of stylised birds
(Konuk 1998b:no 95 [c. 430–410 B.C.]) or bunches of grapes are
known (Konuk 1998b:no 96 [c. 430–410 B.C.]).

M27 k 5
J
Obv. As last.
Rev. M27 (kb J) flanking triangular baetyl; all within incuse square.
Lanz 38 (1984), 272 (11.56g; 07H) = Konuk (1998b:no 110c)
(c. 410–390 B.C.).
It is uncertain whether the third and final sign (J) is an actual let-
ter. The use of this sign as a letter is not attested in Carian inscrip-
tions where it is sometimes used as a dividing stroke. There is
however another coin legend (M33) which includes the same sign.
As with M33, its final position in the legend raises the possibility
that it was used as a letter and not as a separation mark. For
further discussion, see M33 below.

M28 k 5
Obv. Head of Apollo three-quarter facing right or left, with on some
dies, chlamys fastened at neck.
Rev. M28 (kb) flanking sphinx seated left.
AE chalkous.
Künker 61 (2001), 67 (1,27g); Konuk (1998b:no 118) (c. 390–370
B.C.).

M29–M30 have been tentatively attributed to Kaunos: see Ashton

(2003:39–40).
482 APPENDIX E

M29 k
Obv. Bunch of grapes.
Rev. M29 (k) within circle of dots.
1/16th (?) Aeginetic stater; c. 400–350 B.C.
Terzian collection (0.48g; 02H) = Ashton (2003:39, 1).

M30 k
Obv. Corngrain within circle of dots.
Rev. M30 (k) within circle of dots.
1/32nd Aeginetic stater; c. 400–350 B.C.
Ashton collection (0.35g; 02H) = Ashton (2003:39, 2b).

Telmessos
M31 i F
Obv. Head of Athena left in Attic helmet; in front, linear device l;
dotted circle.
Rev. Heracles fighting left with club, left foot placed on rock; along
the right edge, erbbinna in Lycian characters; on either side of
Herakles, M31 (i t) all in incuse square with dotted border.
Light Lycian standard stater; c. 420 B.C.
Munich, Staatliche Münzsammlung (7.87g) = Babelon 1910:n°
385.
This stater of the Lycian dynast Erbbina, part of his regular issues
from Telmessos, is the only example to bear a legend in Carian.
Various readings based on the changing values given through the
years to the letters have been proposed (er, ir, i“ and finally it).
New evidence from the bilingual inscription from Kaunos has led
Adiego (1998b:58–60) and Meier-Brügger (1998:45) to give F the
value t. They wonder whether i might not stand for the initial of
Erbbina in Carian and t for the initial of Telmessos (Telebehi in
Lycian). But the legend may also be transliterated as ti and stand
for the first two letters of the same ethnic.
 COIN LEGENDS IN CARIAN 483

II. UNCERTAIN MINTS

Mint A

M32 £c¢
ñÌ Bg
Obv. Naked male figure, with wings at shoulder and heel, in the kneel-
ing-running position advancing left, head and legs left, trunk frontal,
left arm raised and right arm lowered; groundline.
Rev. Bull standing right; above and below, in two lines, M32 (dta /
ñibr); all within incuse square.
Aeginetic standard stater; c. 450–400 B.C.
London, BM, CM 1934–0611–4 (11.62g; 09H).
For a detailed discussion of M32, see Konuk forthcoming (c). Only
one specimen of this type is known and, according to Robinson
(1939:270), was reportedly obtained near Fethiye (Telmessos). The
orientation of letters suggests a reading from right to left. The last
letter, partly erased, is ñ (ñ) rather than z (≤), even though the
latter is a more common ending used for indicating the genitive
form. The transliteration should thus be atd / rbiñ. As an inde-
pendent word, rbiñ shows a striking similarity with Erbbina, the
Lycian dynast who minted at Telmessos c. 420–400 B.C. (see pre-
vious coin).

Mint B
All the following coins are Aeginetic standard staters unless otherwise
indicated.

M33 J P sN
Obv. Naked male figure, with wings at shoulder and heel, in kneeling-
running position advancing left, head and legs left, right arm raised
and left arm lowered; above right wing, {; dotted groundline.
Rev. Lion standing left with head turned back; above its back, { divid-
ing M33 (J?psg); dotted groundline; all within incuse square.
London, BM (11.68g; 09H) = Robinson (1936:pl. 14, 8); c. 450
B.C.
484 APPENDIX E

There is an obverse variant with the winged male figure advanc-

ing right (Robinson 1936: no 11). I consider { to be a linear
device, perhaps even an object rather than a monogram of the
letters Uo or z. On some contemporary issues from a different
mint (Troxell 1979:pl. 31, 35), { is depicted much like an object
decorated with dots. J is omitted on the following examples, sug-
gesting that it is the last sign of the legend which, therefore, is
read from right to left (gsp J ?). The meaning of the last sign ( J) is
problematic. Its sole function on Carian inscriptions is to separate
words, and it is never used as a letter. The two other legends in
which J appears are M36–M37 and M27. In the case of M36–M37,
our sign may well have served its normal function of separating
words, but for M27, this explanation can hardly be valid since it
is the last sign of the legend, as it is in the present example. For
M27 it is tempting to consider J as the equivalent of a Greek iota
because this is the very sound which comes after kb (for kbid ),
although it must be admitted that I is used as the third letter of
kbi in the bilingual inscription from Kaunos. To sum up, M27
and M33 compel us to accept J as a letter even though its absence
as a letter from inscriptions remains puzzling. As the sign J only
appears on coin legends, it may well be a direct influence of coin
engravers being more familiar with engraving Greek legends (see
also M23). It is worth noting here that the letter P was known
solely on coins until 2004, when a new inscription found at
Hyllarima revealed for the first time the letter B on an inscription
from Caria, albeit with a different value.

M34 °s0
Obv. Naked male figure, with wings at shoulder and heel, in kneeling-
running position advancing right, head and legs right, left arm
raised and right arm lowered; above left wing, {; groundline.
Rev. Lion standing left with head turned back, right forepaw raised;
above its back, M34 ( psg); underneath legend, small {; dotted
groundline; all within incuse square.
Paris, BN (11.64g; 11H); c. 450 B.C.
The more angular shape of ° is further confirmation that this let-
ter is a variant of p with a 90° rotation. There is a reverse vari-
ant which shows the lion with both forepaws standing on the
goundline (Robinson 1936:4, 9ter).
 COIN LEGENDS IN CARIAN 485

M35 s0
P
Obv. Naked male figure, with wings at shoulder and heel, in kneeling-
running position advancing right, head and legs right, left arm
raised and right arm lowered; above left wing, {; groundline.
Rev. Lion standing left with head turned back, right forepaw raised;
above its back, { and M35 (sg / p) whose third letter is in the
lower left corner; all within incuse square with dotted frame.
New York, ANS, 67.152.457 (11.78g; 09H); c. 450 B.C.

M36 ? s N_ J s?d
Obv. Naked male figure, with wings at shoulder and heel, in the kneeling-
running position advancing right, head and legs right, trunk frontal,
left arm raised and right arm lowered; above left wing, {; dotted
groundline.
Rev. Lion standing left with head turned back, right forepaw raised;
above its back, { dividing M36 (bsj|sbd ); all within incuse square
with dotted frame.
Paris, BN (11.72g; 12H); c. 450 B.C.
There is an obverse variety with the male figure advancing left
(Robinson 1936:pl. 14, 16). M36 links the old winged male / lion
type to the new issues with lion forepart / male head type with
a different legend. The right-to-left direction proposed for previ-
ous coins is also suggested for this coin by the orientation of the
two instances of ?; hence M36 should be transliterated as dbs|jsb.
The orientation of d does not help much since it is inconsistent
between M36 and M37.

M37 £Hs J N_sH

Obv. Forepart of lion right, jaws open; both forelegs visible.
Rev. M37 (dbs|jsb) vertically in front of wreathed male head left; behind
neck, {; all within incuse square.
Stater: New York, ANS, 63.35.1 (11.18g; 07H); triobol: SNG Kayhan,
979 (2.93g; 11H); obol or corroded diobol?: Peus 212 (2000), 276
(1.29g). c. 430–400 B.C.
On the New York stater, first published by Thompson (1966:8),
the first and penultimate letters are clearly H, not (as on M36) ?.
All recorded triobols are die-duplicates and feature H as the first
letter and ? as the penultimate. M37 must of course be the same
486 APPENDIX E

legend as M36. Perhaps the coin engraver was confused by the

orientation of these letters and turned ? into H, which would make
it readable in both directions, although it remains puzzling that
? was not rendered { to be consistent with M36. For a similar
confusion, see M23 above.

Mint C
M38 7
Obv. Head of lion three-quarter facing left, within linear circle.
Rev. Head of bull left; behind neck truncation, M38 (z or symbol); all
within round incuse.
Milesian standard hemiobol; c. 400–350 B.C.
Muharrem Kayhan collection (0.42g; 10H).
It is uncertain whether the sign on the bull’s neck is a character
or a symbol (linear device), because a variety of the same series
(Hirsch 221 [2002], 267; another specimen from different dies:
Hirsch 226 [2003], 1421) carries another sign (h) which is clearly
a linear device of a type encountered on other non-Carian issues
(e.g. SNG Kayhan, 744). There is also a variety of M38 without
any letter or symbol (SNG Kayhan, 990).

put
1L
M39 z
Obv. Head of lion three-quarter facing left, within linear circle.
Rev. Head of bull three-quarter facing left; above, put ( put); behind
neck truncation, L (h); below, z (≤ ); on the neck, 1 (z or sym-
bol); all within shallow round incuse.
Milesian standard hemiobol; c. 400–350 B.C.
Muharrem Kayhan collection (0.49g; 03H).
All known specimens of M39 are off-flan and lack the letter h on
the right edge of the coin (e.g. SNG Keckman, 862; Klein 1999:
n° 503). The Kayhan specimen is the only one known to me with
a full legend. The direction of the legend follows the shape of the
coin: puth≤. The last letter (≤ ) is used in Carian to indicate the
genitive form. We may interpret the legend as (the coin of ) putl,
most probably the name of a local dynast. The reverse is flat
 COIN LEGENDS IN CARIAN 487

without any trace of an incuse; this would place M39 after M38
in the sequence of issues.

Mint D
Types as follows unless otherwise indicated.
Obv. Forepart of bull left.
Rev. Forepart of bull left; below its head, M40 ( p); M41 (s); M42 (db);
M43 (∞); M44 (∞); all within incuse square.
All coins are Milesian standard diobols, apart from M46 which
is a Milesian standard obol.

M40 p
Paris, BN, fonds général 3292 (2.14g) = Babelon (1910:no 1789
[listed under Samos]), pl. CL, 12 = Troxell 1984:no 12A.

M41 s
Hirsch 55 (1967), 2175 (2.19g) = Cancio (1989:83).

M42 }_
Cahn 60 (1928), 858 (2.01g) = Troxell (1984:256, 12B).
_ is another example of a Carian letter used on a coin which is
otherwise only attested in inscriptions from Egypt (see above, M32).
The orientation of letters suggests a right-to-left reading which
would give the transcription bd.

M43 x
Private collection (2.16g; 12H).
The style of M43 is later than the preceding examples.

M44 X
Obv. Confronted foreparts of two bulls, their horns crossed.
Rev. As last, M44 (∞).
SNG Kayhan, 958 (2.10g; 12H).
488 APPENDIX E

M45 ux
Obv. As last, but later style; heads of bulls three-quarter facing, horns
not crossed.
Rev. Forepart of bull left; above its head, M45 (u∞ ); all within incuse
square.
London, BM (2.12g; 08H) = Six (1890:239, 42; pl. 17, 9).
The transcription ∞u is confirmed by M43 and M44 where ∞ is
the initial. The variety with the two bull foreparts was a prolific
coinage, especially in its early phase which is anepigraphic and
features the bull foreparts in profile with their horns crossed as if
the letter x was meant. Since other letters are attested (see above),
it is uncertain whether X, x or ux should be regarded as the
first letters of an ethnic. But if the letter x can be construed from
the crossed horns, a Carian ethnic beginning with ∞ and ∞u such
as Kydai / Hydai and Kyblissos / Hyblissos is an attractive pos-
sibility; see Konuk (2003:n° 69).

M46 ^
Obv. Forepart of bull left.
Rev. Head of bull left; below, M46; all within incuse square.
Milesian standard obol.
SNG Kayhan, 974 (1.14g; 03H).
It is uncertain whether ^ is a character or a linear device (sym-
bol). It is otherwise unknown.

Mint E
M47 ¥
Obv. Bearded head right.
Rev. Forepart of bull left; on its shoulder, M47 ( y); all within incuse
square.
Private collection (1.24g; 12H); c. 400 B.C.

M48 !
Obv. As last.
Rev. Forepart of bull left; in lower left corner, M48 ( y); all within
incuse square.
Paris, BN (1.59g; 12H) = Babelon (1910:n° 2494); c. 400 B.C.
 COIN LEGENDS IN CARIAN 489

M49 Ú
Obv. Male head right; wreathed?
Rev. Forepart of bull left; on its neck, M49 ( y).
Berlin, Staatliche Museen—Münzkabinett, gift T. Wiegand 1354/
1931 (1.10g; 04H); c. 400 B.C.
Winzer’s attribution of this series to the mint of Mylasa under
Hekatomnos by interpreting M48 and M49 as the Greek initial
of his name and taking the bearded head on the obverse as his
portrait is speculative (Winzer 2005:13.1). The different orienta-
tions of the letter, especially M47, suggest Carian and not Greek.
There is also an early variety with the same type without letter
(private collection, 0.68g; 09H). The length of time needed for
the stylistic change from the early issue to M49 suggests a civic
coinage rather than a dynastic.

Mint F
M50 oul or luo
Obv. Forepart of lion right, head turned back; to the right, M50 (oul
or luo).
Rev. Square punch mark.
Aeginetic standard stater; c. 500 B.C.
Paris, BN (11.71g).
The legend on this series has been long regarded as Greek. It has
usually been read as ouB (BMC Ionia, xxxiv) or oBu (Six 1890:
223). The former reading prompted an attribution to a dynast of
Miletos, the latter to the town of Olymos in Caria. Either read-
ing depends upon whether the legend is meant to be read from
the inside or the outside. As Head rightly points out in BMC Ionia,
a reading from the outside is extremely rare on archaic coins,
which undermines Six’s attribution to Olymos. However, Head’s
attribution to a dynast of Miletos is no longer satisfactory in view
of the recorded provenances which clearly point to a mint fur-
ther south in Caria. In my 1996 RNS paper, I suggested that the
legend should be regarded as Carian. The occurrence of this series
in early hoards such as the Santorini find (IGCH 7) points to a
date of c. 500 B.C. When Carian staters of Aeginetic standard
bear a legend in the fifth century B.C., it is usually in the Carian
490 APPENDIX E

script. My reading of the third letter (the lowest on the coin) is

l, not B; the horizontal stroke, even though a little erased, being
clearly visible. We would thus have oul for oul. This conclusion
has also been reached by I{ık (2003:124–126) who proposes the
same reading. oul may be linked to personal names like Ouliades
(Herodotos 5, 37 mentions one Oliatos ruler of Mylasa, c. 500 B.C.)
or Oulios. When the legend is read from the outside, there is a puz-
zling similarity with M22 and M23 which give the beginning of the
ethnic of Keramos in Carian. At this stage, an attribution to Keramos
would be premature, but this may change with future finds. One
should not exclude, however, a retrograde reading which is rather
common on early Carian coins. We would have luo (luo).

Mint G
M51 orou
Obv. Forepart of winged human-headed bull right.
Rev. Female head right; behind, M51 (orou); all within incuse square
with dotted frame.
Aeginetic standard hemidrachm (triobol) and trihemiobol; c. 450–400
B.C.
Oxford, Ashmolean Museum (2.87g; 06H); trihemiobol: private
collection (1.25g).

M52 orou
Obv. Sphinx seated right.
Rev. Female head right; behind, M52 (orou); all within incuse square
with dotted frame.
Aeginetic standard; obol; c. 450–400 B.C.
Oxford, Ashmolean Museum (1.00g; 06H).
For a detailed discussion of Mint G, see Konuk (forthcoming [d]).
The legends of M51 and M52 are usually regarded as Lycian and
are attributed to a dynast by the name of Uvug or Uwug. A
Lycian origin, however, is far from certain. First, the weight stan-
dard is clearly Aeginetic with half staters (same type as M51 but
anepigraphic) weighing c. 5.80g. That standard was quite wide-
spread among Carian mints active in the fifth century B.C. Lycia,
on the other hand, had its own weight systems and is not known
to have ever used the Aeginetic standard. Another argument against
a Lycian origin is the obverse type of a winged human-headed
bull, a very unusual iconography for Lycia. On the other hand,
 COIN LEGENDS IN CARIAN 491

the depiction of the sphinx on M52 is very close to that of the

contemporary coins of Kaunos, see Konuk (1998b:119). The style
and the particular care given to the striking are also features not
typical of Lycian mints. A further important piece of evidence to
support a Carian origin is the occurrence of several of the half
staters (mentioned above) in an unpublished hoard from Caria
dating to the middle of the fifth century B.C. which included a
variety of Carian coins (e.g. from Kaunos, Knidos, uncertain mints).
Coins of Caria and Lycia did not usually circulate together and
it is no surprise therefore that not a single coin from Lycia is
reported in the hoard. All in all, the cumulative weight of evi-
dence points to a mint in Caria, orou being the name of a Carian
dynast of the second half of the fifth century B.C.

Mint H
M53 i
Obv. Persian hero-king right, in kneeling-running position, holding
transverse spear and bow.
Rev. Ship’s prow left; on its rail, M53 (i ).
AV Persian daric, c. 450–400 B.C.
Paris, BN, collection de Luynes (8.25g) = Six (1890:pl. 17, 13) =
Babelon (1910:pl. 97, 24).
For a brief discussion of this unique daric, see Konuk (2000a:179).
The tentative attribution to Salmakis near Halikarnassos, first sug-
gested by Six and followed by myself, appears now to be unfounded.
It was based on an incorrect association with bronze coins which
turned out to be from the Cypriot mint of Salamis. According to
the typology established for darics and sigloi, the Persian hero-
king holding spear and bow is type IIIb which ends c. 400 B.C.

Mint I

M54

M54
Obv. Boar advancing left; above, M54; double groundline; dotted border.
Rev. Triskeles ending with duck’s heads; floral ornament growing from
central ring; around, kuprlli in Lycian; all within dotted border
incuse square.
492 APPENDIX E

Lydian (Persic) standard stater; c. 450 B.C.

Paris, BN (10.85g) = Babelon (1910:no 253).
The reverse legend is clearly Lycian and names the dynast Kuprlli
as the issuer. The five-character legend of the obverse is, on the
other hand, quite problematic. It has been variously transcribed
but a better preserved specimen from the same pair of dies (SNG
von Aulock, 4156; 10.70g) has allowed some precision in the read-
ing of the legend. The right-to-left direction is suggested by the
orientation of the letters. Mørkholm-Neumann (1978:no M 301a)
described the obverse as Carian with a question mark, while oth-
ers have not hesitated in recognising Carian: e.g. Durnford (1991)
(his reading of M54 as the Carian ethnic of Xanthos is no longer
tenable as some of the letter values he used are not accepted
today), Cau (1999). The only other instance of Carian being used
on a Lycian issue is M31 listed above. On close examination the
first character cannot be o as previously thought but is probably
d (g) a character form already encountered on some varieties of
M33. The second letter S may be “ as seen on the alphabet of
Hyllarima. The letter _ is not attested in the Carian alphabet, but
it may be a form of g (r). the letter K is quite difficult to inter-
pret; it may be a form of k (k) or l (l )? The last letter may be
¢ (a). All in all, there are far too many uncertainties over M54
to even describe it as Carian, and at this stage, I prefer not to
speculate on the various ways of transcribing it. The weight stan-
dard appears to be Lydian (Persic), which has as its stater a dou-
ble siglos weighing slightly less than 11.00g. This is exceptional
for Kuprlli and Lycian mints in general which struck coins in
their own local standards. The decision to use the Lydian stan-
dard may well have something to do with the obverse legend.
 ABBREVIATIONS AND BIBLIOGRAPHY

ABBREVIATIONS

ABSA = The Annual of the Britisch School at Athens.

AArch = Acta Archeologica
Acta Ant. Hung. = Acta Antiqua Academiae Scientiarum Hungaricae.
AC = L’Antiquité Classique.
AfO = Archiv für Orientforschung.
AGI = Archivio Glottologico Italiano.
ANSMN = American Numismatic Society Museum Notes.
ArOr = Archiv Orientální.
ASBW = Archiv für Schreib- und Buchwesen.
ASNP = Annali della Scuola Superiore Normale di Pisa.
BB = Beiträge zur Kunde der indogermanischen Sprachen.
BCH = Bulletin de Correspondance Hellenique.
BiOr = Bibliotheca Orientalis.
BMC Caria = B. V. Head, Catalogue of the Greek Coins in The British Museum, Greek Coins
of Caria, Cos, Rhodes, & c., London, 1897.
BMC Ionia = B. V. Head, Catalogue of the Greek Coins in The British Museum, Greek Coins
of Ionia, London, 1892.
BSL = Bulletin de la Société Linguistique de Paris.
BzN = Beiträge zur Namenforschung.
Cahn = Adolph E. Cahn, Frankfurt a. M.
CFC = Cuadernos de Filología Clásica.
CNG = Classical Numismatic Group, Lancaster, Pa (USA)—London (UK).
Colloquium Caricum = W. Blümel – P. Frei – C. Marek (eds.), Colloquium Caricum. Akten
der internationalen Tagung über die karisch-griechische Bilingue von Kaunos 31.10–1.11.1997
in Feusisberg bei Zürich = Kadmos 37 (1998).
Decifrazione del cario = M. E. Giannotta et alii (eds.), La decifrazione del cario, Roma, 1994.
DNb = E. Lüddeckens – H. J. Thissen (eds.), Demotisches Namenbuch, Wiesbaden, 1980.
Eothen = F. Imparati (ed.), Eothen. Studi di storia e di filologia anatolica dedicati a Giovanni
Pugliese Carratelli, Firenze, 1988.
EpAnat = Epigraphica Anatolica. Zeitschrift für Epigraphik und historische Geographie Anatoliens.
Fs. Bloesch = Zur griechischen Kunst. Hansjörg Bloesch zum 60. Geburtstag am 5. Juli 1972,
Bern, 1973.
Fs. Friedrich = R. von Kienle et alii (eds.), Festschrift Johannes Friedrich zum 65. Geburtstag
am 27. August 1958 gewidmet, Heidelberg, 1959.
Fs. Grumach = W. Brice (ed.), Europa: Studien zur Geschichte und Epigraphik der frühen Agais.
Festschrift für Ernst Grumach, Berlin, 1968.
Fs. Neumann I = J. Tischler (ed.), Serta Indogermanica. Festschrift G. Neumann, Innsbruck,
1982.
Fs. Neumann II = M. Fritz – S. Zeilfelder (eds.), Novalis Indogermanica. Festschrift für
G. Neumann zum 80. Geburtstag, Graz, 2002.
Fs. Oberhuber = W. Meid – H. Trenkwalder (eds.), Im Bannkreis des Alten Orients, K. Oberhuber
zum 70. Geburtstag gewidmet, Innsbruck, 1986.
Gs. Kretschmer = H. Kronasser (ed.), MNHMHS XARIN. Gedenkschrift Paul Kretschmer, Wien,
1956–1957 (I–II).
Hauck & Aufhäuser = Hauck & Aufhäuser Privatbankiers, Munich.
494 ABBREVIATIONS AND BIBLIOGRAPHY

Hirsch = Münzhandlung Gerhard Hirsch Nachfolger, Munich.

HS = Historische Sprachforschung.
IF = Indogermanische Forschungen.
IGCH = M. Thompson – O. Mørkholm – C. M. Kraay (eds.), An Inventory of Greek
Coin Hoards, New York, 1973.
InL = Incontri Linguistici.
IstMitt = Istanbuler Mitteilungen.
JAOS = Journal of the American Oriental Society.
JCS = Journal of Cuneiform Studies.
JEA = Journal of Egyptian Archaeology.
JHS = Journal of Hellenic Studies.
KarON see Blümel (1998a).
KarPN see Blümel (1992).
KlF = Kleinasiatische Forschungen, Weimar, 1927–1930.
KON see Zgusta (1984).
KPN see Zgusta (1964).
Künker = Fritz Rudolph Künker Münzenhandlung, Osnabrück.
KZ = Zeitschrift für vergleichende Sprachforschung.
Lanz = Numismatik Lanz, Munich.
LNH see Laroche (1966).
MH = Museum Helveticum.
MM = Münzen und Medaillen AG, Basel.
MSS = Münchener Studien zur Sprachwissenschaft.
Naville = Naville & Ars Classica, Lucerne.
NC = Numismatic Chronicle.
NDH see Laroche (1947).
OLZ = Orientalische Literaturzeitung.
Or = Orientalia. Commentarii periodici Pontificii Instituti Biblici Nova Series.
ÖJh = Jahreshefte des österreichischen Archäologischen Institutes.
Peus = Dr. Busso Peus Nachf., Münzhandlung, Frankfurt a. M.
PP = Parola del Passato.
PSBA = Proceedings of the Society of Biblical Archaeology.
RALinc = Atti della Accademia Nazionale dei Lincei, Rendiconti della Classe di Scienze morali,
storiche e filologiche, Serie VIII.
RE = Paulys Real-Encyclopädie der classischen Altertumswissenschaft: neue Bearbeitung unter
Mitwirkung zahlreicher Fachgenossen; herausgegeben von Georg Wissowa, Stuttgart, 1894–1963.
REA = Revue des Études Anciennes.
REG = Revue des Études Grecques.
RHA = Revue Hittite et Asianique.
RPh = Revue de Philologie.
SEAP = Studi di Eggitologia e di Antichità Puniche.
SM = Schweizer Münzblätter.
SMEA = Studi Micenei ed Egeo-Anatolici.
SNG Kayhan = Sylloge Nummorum Graecorum, Turkey 1, The Muharrem Kayhan Collection,
Istanbul-Bordeaux, 2002.
SNG Keckman = Sylloge Nummorum Graecorum, Finland, The Erkki Keckman Collection in the
Skopbank, Helsinki, part I Karia, Helsinki, 1994.
SNG von Aulock = Sylloge Nummorum Graecorum, Deutschland: Sammlung von Aulock, 18 fasci-
cules, Berlin 1957–1981.
TA1 see Laroche (1957).
TA2 see Laroche (1961).
TSBA = Transactions of the Society of Biblical Archaeology.
VJa = Voprosy Jazykoznanija.
Weber = L. Forrer, Descriptive Catalogue of the Collection of Greek Coins formed by Sir Hermann
Weber M.D., vol. 3, London, 1929.
 ABBREVIATIONS AND BIBLIOGRAPHY 495

WO = Die Welt des Orients: wissenschaftliche Beiträge zur Kunde des Morgenlandes.
ZDMG = Zeitschrift der deutschen morgenländischen Gesellschaft.
ZPE = Zeitschrift für Papyrologie und Epigraphik.

BIBLIOGRAPHY

Adiego, I.-J.
(1990a) “Deux notes sur la langue et l’écriture cariennes”, Kadmos 29, 133–138.
(1990b) Studia Carica. Investigaciones sobre la escritura y lengua carias, y su relación con la familia
lingüística anataolia indoeuropea, Doctoral dissertation, Barcelona. (published as microfilm).
(1992a) “Recherches cariennes: essai d’amélioration du système de J. D. Ray”, Kadmos
31, 25–39.
(1992b) “Glosses i pseudoglosses càries en fonts gregues”, in: J. Zaragoza – A. González
Senmartí (eds.), Homenatge a Josep Alsina. Actes del Xè Simposi d’Estudis Clàssics, Tarragona
28–30 de novembre del 1990, I, Tarragona, 51–54.
(1993a) Studia Carica. Investigaciones sobre la escritura y lengua carias, Barcelona.
(1993b) “Sobre OALOALON SGDI 5727.d30”, Kadmos 32 (1993), 173–174.
(1994a) “Les identifications onomastiques dans le déchiffrement du carien”, Decifrazione
del cario, 27–63, “Considerazioni conclusive”, ibid. 239–240.
(1994b) “El nombre cario Hecatomno”, CFC (Estudios griegos e indoeuropeos) n. s. 4,
247–256.
(1994c) “Genitiu singular en lici i protoluvi”, Anuari de Filologia. Studia Graeca et Latina
17, 11–33.
(1995) “Contribuciones al desciframiento del cario”, Kadmos 34, 18–34.
(1996) “Comentarios a la nueva lectura de la inscripción caria 28*”, Kadmos 35, 160–163.
(1998a) “La nueva bilingüe greco-caria de Cauno y el desciframiento del cario”, Aula
Orientalis 16 (1998), 5–26.
(1998b) “Die neue Bilingue von Kaunos und das Problem des karischen Alphabets”,
Colloquium Caricum, 57–79.
(2000) “La inscripción greco-caria de los Hecatómnidas en el santuario de Sinuri”,
Kadmos 39, 133–157.
(2002) “Cario de Cauno punoO”, Aula Orientalis 20, 13–20.
(2004) “Los alfabetos epicóricos anhelénicos de Asia Menor”, in: P. Bádenas de la Peña –
S. Torallas Tovar – E. R. Luján – M. Á. Gallego (eds.), Lenguas en contacto: El testi-
monio escrito, Madrid, 299–320.
(2005) “La nueva inscripción caria de Milasa”, Kadmos 44, 81–94.
Adiego, I.-J. – Debord, P. – Varinlio<lu, E.
(2005) “La stèle caro-grecque d’Hyllarima”, REA 107 nº 2, 601–653.
Ashton, R.
(1998) “Keramos”, in: R. Ashton et alii., “Some Greek Coins in the British Museum”,
NC 158, 46–49.
(2003) “(vi) Kaunian Notes”, in: R. Ashton – P. Kinns, “Opuscula Anatolica II”, NC
163, 36–40.
Avishur, Y. – Heltzer, M.
(2003) “Carians as skilled masons in Israel and mercenaries in Judah in the early I
millennium B.C.E.”, Kadmos 42, 87–90.
Babelon, E.
(1910) Traité des monnaies grecques et romaines. Deuxième partie, description historique, t. 2, Paris.
Bean, G. F.
(1953) “Notes and inscriptions from Caunus”, JHS 73, 10–35.
(1954) “Notes and inscriptions from Caunus (continued)”, JHS 74, 85–110.
496 ABBREVIATIONS AND BIBLIOGRAPHY

Belli, P. – Gusmani, R.
(2001) “Una nuova iscrizione rupestre presso il Santuario di Labraunda in Caria”, PP
56, 33–41.
Bengston, H.
(1954–55) “Skylax von Karyanda und Herakleides von Mylasa”, Historia 3, 302–305.
Benveniste, E.
(1931) “Noms cariens”, RHA I. 2, 52–57.
Bernand, A. – Aly, A.
(1959?) Abou-Simbel, inscriptions grecques, cariennes et sémitiques des statues de la façade, Le
Caire, Centre de documentation égyptologique, Collection scientifique.
Bernand, A. – Masson, O.
(1957) “Les inscriptions grecques d’Abu Simbel”, REG 70, 1–46.
Berti, F. – Innocente, L.
(1998) “Due nuovi graffiti in alfabeto cario da Iasos”, Colloquium Caricum, 137–142.
(2005) “Graffito cario su piede di coppa attica”, Bollettino dell’Associazione Iasos di Caria,
11, 20–21.
Bertoldi, V.
(1948) “Souangela, Tomba del Re”, PP 3, 5–11.
Blümel, W.
KarON: see (1998a).
KarPN: see (1992).
(1988) “Epigraphische Forschungen in der Region von Mylasa”, in: VI Ara{tırma Sonuçları
Toplantısı, Ankara 23–27 Mayıs 1988, 261–264.
(1990) “Zwei neue Inschriften aus Mylasa aus der Zeit des Maussollos”, EpAnat 16,
29–43 Taf. 12.
(1992) “Einheimische Personennamen in griechischen Inschriften aus Karien”, EpAnat
20, 7–34.
(1993) “SGDI 5727 (Halikarnassos): eine Revision”, Kadmos 32, 1–18.
(1994) “Über die chronologische und geographische Verteilung einheimischer Personen-
namen in griechischen Inschriften aus Karien”, Decifrazione del cario, 65–86.
(1996) “Epigraphische Forschungen im Westen Kariens 1994”, in: XIII Ara{tırma Sonuçları
Toplantısı, Ankara, 261–264.
(1998a) “Einheimische Ortsnamen in Karien”, EpAnat 30 (1998), 163–184.
(1998b) “Karien, die Karer und ihre Nachbarn in Kleinasien”, Colloquium Caricum,
163–173.
(2005) “Problematische Lesungen in der karischen Inschrift aus der Region von Mylasa”,
Kadmos 44, 188.
Blümel, W. – Adiego, I.-J.
(1993) “Die karische Inschrift von Kildara”, Kadmos 32, 87–95.
Blümel, W. – Kızıl, A.
(2004) “Eine neue karische Inschrift aus der Region von Mylasa”, Kadmos 43, 131–138.
Bockisch, G.
(1969) “Die Karer und ihre Dynasten”, Klio 51, 117–175.
Bohl, F. M.
(1932–33) “Inschriften mit unbekannter Schrift aus der Leidener Sammlung”, AfO 8,
173–174.
Boisson, C.
(1994) “Conséquences phonétiques de certaines hypothèses de déchiffrement du carien”,
Decifrazione del cario, 207–232.
 ABBREVIATIONS AND BIBLIOGRAPHY 497

Bork, F.
(1930) “Die Schrift der Karer”, ASBW 4, 18–30.
(1931) “Die Sprache der Karer”, AfO 7, 14–23.
Brandenstein, W.
(1934a) “J F X C in den epichorischen Alphabeten Kleinasiens”, Klio 27, 69–73.
(1935a) “Karische Sprache” in: Pauly-Wissowa, RE Supplementband VI, col. 140–146.
(1936) “Streifzüge . . . Zwei Karische Ortsnamen”, Glotta 25, 32–35.
Brixhe, C.
(1996) Review of Decifrazione del cario in: BSL 91/2, 214–221.
Cancio, L.
(1989) “A New Satrapal Coin of the KIM—EKA Series”, SM 156, 83.
Carruba, O.
(1998) “Zum Stand der Entzifferung des Karischen”, Colloquium Caricum, 47–56.
(1999a) “Bildungen karischer Ethnika”, SMEA 41/2 (1999), 175–180.
(1999b) “Ar/w/wazuma”, Kadmos 38, 50–58.
(2000) “Der Name der Karer”, Athenaeum 88, 49–57.
Cau, N.
(1999a) “La legenda caria su una serie monetale del dinasta Kuprlli”, Studi Ellenistici
12, 9–17.
(1999b) “Una nuova lettura di alcune leggende monetali carie”, Kadmos 38, 43–49.
(2003) “Nuovi antroponimi indigeni nelle iscrizioni greche della Licia di età ellenistico-
romana”, Studi Ellenistici 15, 297–340.
Deroy, L.
(1955) “Les inscriptions cariennes de Carie”, AC 24, 305–335.
(1959) Review of Masson-Yoyotte (1956) in: Or 28, 101–102.
Descat, R.
(1994) “La géographie dans les listes de tributs attiques: Lepsimandos et Kasôlaba en
Carie”, ZPE 104, 61–68.
(1998) “La carrière d’Eupolemos, Stratège macédonien en Asie Mineure. Appendice:
Note sur une inscription caro-grecque de Caunos”, REA 100, 167–190.
Dorsi, P.
(1979) “Le glosse carie”, InL 5, 27–35.
Dressler, W.
(1966–67[68]) “Karoide Inschriften im Steinbruch von Belevi”, ÖJh 48, 73–76.
Durnford, S.
(1991) “An instance of the Lycian name for Xanthos in Carian script”, Kadmos 30, 90–92.
Eilers, W.
(1935) “Das Volk der karka in den epichorischen Alphabeten Kleinasiens”, OLZ 38,
201–213.
(1940) “Kleinasiatisches”, ZDMG 94, 189–233.
Eichner, H.
(1994) “Zur Entzifferung des Karischen”, Decifrazione del cario, 167–170.
Erbse, H.
(1986) “Zu der Ilias-Scholien (Curae secundae II)”, Hermes 114.4, 385–398.
ESS (= Epigrahic Survey Staff ).
(1988) Reliefs and Inscriptions at Luxor Temple, volume 2: The Facade, Portals, Upper Register
Scenes, Columns Marginalia, and Statuary in the Colonnade Hall (University of Chicago
Oriental Institute Publications).
498 ABBREVIATIONS AND BIBLIOGRAPHY

Faucounau, J.
(1980) “Réflexions sur le déchiffrement des inscriptions cariennes”, Klio 62.2, 289–305.
(1984) “A propos de récents progrès dans le déchiffrement de l’écriture carienne”, BSL
79/1, 229–238.
(1989) “A propos de la lecture des inscriptions cariennes”, Kadmos 28, 174–175.
(1994) “Remarks on the Carian Alphabet of Sinuri”, Decifrazione del cario, 233–236.
Franklin, N.
(2001) “Masons’ Marks from the Ninth Century B.C.E. Northern Kingdom of Israel.
Evidence of the Nascent Carian Alphabet?”, Kadmos 40, 107–116.
Frei, P. – Marek, C.
(1997) “Die karisch-griechische Bilingue von Kaunos. Eine zweisprachige Staatsurkunde
des 4. Jh.s v. Chr.”, Kadmos 36, 1–89.
(1998) “Die karisch-griechische Bilingue von Kaunos. Ein neues Textfragment”, Colloquium
Caricum, 1–18.
(2000) “Neues zu den karischen Inschriften von Kaunos”, Kadmos 39, 83–132.
Friedrich, J.
(1931) “Zu den kleinasiatischen Personennamen mit dem Element muwa”, Kleinasiatische
Forschungen I, 359–378.
(1932) Kleinasiatische Sprachdenkmäler, Berlin.
(1952) “Karer in Numidien?”, Or 21, 231–233.
(1965) “Ein wohl kleinasiatisches Tontäfelchen mit unbekannter Schrift”, Kadmos 3,
156–169.
Georgiev, V. I.
(1960) “Der indoeuropäische Charakter der karischen Sprache”, ArOr 28, 607–619.
(1966) Introduzione alla storia delle lingue indoeuropee (= Incunabula graeca 9), Roma.
(1975) “Ein Versuch zur Deutung der griechisch-karischen bilinguis”, Kadmos 14, 64–67.
(1981) Introduction to the History of the Indo-European Languages, Sofia. [210–214].
Gérard, R.
(2005) Phonétique et morphologie de la langue lydienne, Louvain-la-Neuve.
Goetze, A.
(1954) “The Linguistic Continuity of Anatolia as shown by its Proper Names”, JCS 8,
74–81.
Gosline, S. L.
(1992) “Carian quarry markings on Elephantine Island”, Kadmos 31, 43–50.
(1998) “Quarry, Setting and Team Marks: the Carian Connection”, Journal of Ancient
Civilizations, 13, 59–82.
Gusmani, R.
(1967) Review of ”evoro“kin (1965) in: AGI 52, 79–84.
(1975) Neue epichorische Schriftzeugnisse aus Sardis (1958–1971) = Archaeological Exploration
of Sardis, Monograph 3, Cambridge, Mass.
(1978) “Zwei neue Gefässinschriften in karischer Sprache”, Kadmos 17, 67–75.
(1979a) Review of Masson (1978) in: Paideia 34, 220–223.
(1979b) “Spunti per la decrittazione di segni carii”, InL 5, 193–197.
(1982) “Zum Karischen”, Fs. Neumann I, 193–197.
(1986) “Die Erforschung des Karischen”, Fs. Oberhuber, 55–67.
(1988) “Karische Beiträge”, Kadmos 27.2, 139–149.
(1990) “Karische Beiträge II”, Kadmos 29.1, 47–53.
(1994) “Kritisches und Autokritisches zu den Entzifferungsversuchen”, Decifrazione del
cario, 115–120.
 ABBREVIATIONS AND BIBLIOGRAPHY 499

Hajnal, I.
(1996) Die Entzifferung unbekannter Schriften: Drei Fallstudien—ein Szenario?, Bern.
(1995[97]) “Das Vokalsystem des Karischen: Eine provisorische Bestandsaufnahme”,
Die Sprache 37, 12–30.
(1997a) “«Indogermanische» Syntax in einer neuerschlossenen anatolischen Sprache:
Die karische Partikel -xi”, in: E. Crespo – J. L. García Ramón (eds.), Berthold Delbrück
y la sintaxis indoeuropea hoy, Madrid-Wiesbaden, 193–217.
(1997b) “Die karisch-griechische Bilingue 44* aus Kaunos: ein erster Augenschein”,
Kadmos 36, 141–166.
(1997c) “Definite nominale Determination im Indogermanischen”, IF 102, 38–73.
(1998) “, Jungluwisches‘ *s und die karische Evidenz: Versucheiner dialektologischen
Klärung”, Colloquium Caricum, 80–108.
Hanfmann, G. M. A – Masson, O.
(1967) “Carian Inscriptions from Sardis and Stratonikeia”, Kadmos 6, 123–134.
Hegyi, D.
(1998) “The Cult of Sinuri in Caria”, Acta Ant. Hung. 38, 157–163.
Heubeck, A.
(1959a) Review of Masson-Yoyotte (1956) in: Gnomon 31, 332–336.
(1967–68) Review of Otkup“‘ikov (1966) in: IF 72, 331–333.
(1974) Review of Zauzich (1972) in: BiOr 31, 95–97.
Hornblower, S.
(1982) Mausolus, Oxford.
Houwink Ten Cate, Ph. H. J.
(1961) The Luwian Population Groups of Lycia and Cilicia Aspera during the Hellenistic Period
(= Documenta et monumenta orientis antiqui 10), Leiden.
de Hoz, J.
(2004) “De cómo los protogriegos crearon el griego y los pregriegos lo aprendieron”,
in: P. Bádenas de la Peña – S. Torallas Tovar – E. R. Luján – M. Á. Gallego
(eds.), Lenguas en contacto: El testimonio escrito, Madrid, 35–56.
Innocente, L.
(1992) “Stato delle ricerche sul cario”, Vicino Oriente 8/2, 213–222.
(1992a) “Concordanze delle iscrizioni carie”, SMEA 30, 25–87.
(1994) “Note epigrafiche”, Decifrazione del cario, 101–110.
(1995) “The Oxford Paracarian Inscription”, Kadmos 34, 149–154.
(1997) Review of Adiego (1993) in: Or 66, 116–117.
(2002) “Tegola di Iasos”, Kadmos 41, 179–180.
I{ık, C.
(1998) “Die Ergebnisse der Ausgrabungen in Kaunos bis zur Entdeckung der Bilingue”,
Colloquium Caricum, 183–202.
Ißık, E.
(2003) Frühe Silberprägungen in Städten Westkleinasiens, Saarbrücken.
Janda, M.
(1994) “Beiträge zum Karischen”, Decifrazione del cario, 171–190.
Jeffery, L. H.
(1961) The Local Scripts of Archaic Greece, Oxford.
(1964[65]) “Old Smyrna: Inscriptions on sherds and small objects”, ABSA 59, 39–49,
pl. 5–8.
Jordan, H.
(1968) Review of Otkup“‘ikov (1966) in: OLZ 63, 125–130.
500 ABBREVIATIONS AND BIBLIOGRAPHY

Jucker, H. – Meier, M.
(1978) “Eine Bronzephiale mit karischer Inschrift”, MH 35, 104–115.
Kalinka, E.
(1901) Tituli Asiae Minoris, vol. I Tituli Linguae Lyciae lingua lycia conscripti, Wien.
Kammenhuber, A.
(1969b) “Karer, Karia, Karische Sprache”, K. Ziegler – W. Sontheimer (eds.), Der
Kleine Pauly III, Stuttgart, col. 118–121.
Kammerzell, F.
(1990) Studien zu Sprache und Geschichte der Karer in Ägypten, Diss. Göttingen.
(1993) Studien zu Sprache und Geschichte der Karer in Ägypten, Wiesbaden.
Klein, D.
(1999) Sammlung von griechischen Kleinsilbermüzen und Bronzen, Nomismata 3, Milan.
Konuk, K.
(1998a) The Coinage of the Hekatomnids of Caria (unpublished Oxford University D. Phil.
Diss.).
(1998b) “The Early Coinage of Kaunos”, in: R. Ashton – S. Hurter (eds.), Studies in
Greek Numismatics in Memory of Martin Jessop Price, London, 197–223, pl. 47–50.
(2000a) “Influences et éléments achéménides dans le monnayage de la Carie”, in:
O. Casabonne (ed.), Mécanismes et innovations monétaires dans l’Anatolie achéménide. Numismatique
et histoire, Istanbul, 171–183.
(2000b) “Coin evidence for the Carian Name of Keramos”, Kadmos 39, 159–164.
(2003) Karun’dan Karia’ya, Muharrem Kayhan Koleksiyonundan Erken Anadolu Sikkeleri. From
Kroisos to Karia, Early Anatolian Coins from the Muharrem Kayhan Collection, Istanbul.
(forthcoming [a]) “The Early Coinage of Mylasa”, NC.
(forthcoming [b]) “Kasolaba, a New Mint in Karia?”, in: S. Drougou – E. Ralli (eds.),
Essays in Honour of Ioannis Touratsoglou, Athens.
(forthcoming [c]) “Erbbina en Carie?”, in: F. de Callataÿ et alii (eds.), Liber amicorum
Tony Hackens, Louvain-la-Neuve.
(forthcoming [d]) “Orou, dynaste de Carie”, in: P. Brun (ed.) Anatolica, mélanges en l’hon-
neur de Pierre Debord, Bordeaux.
Kowalski, Th.
(1975) “Lettres cariennes: essai de déchiffrement de l’écriture carienne”, Kadmos 14,
73–93.
Kretschmer, P.
(1896) Einleitung in der Geschichte der griech. Sprachen, Göttingen.
(1929) “Eine neue karische Inschrift”, KlF 1, 318–320.
(1954) “Zu der karischen Zeile der SÁma-Inschrift aus Athen oben S. 67”, Glotta 34,
160.
Laroche, E.
LNH see (1966).
NDH see (1947).
TA1 see (1957).
TA2 see (1961).
(1947) Recherches sur les noms des dieux hittites, Paris.
(1957) “Notes de toponymie anatolienne”, Gs. Kretschmer II, 1–7.
(1961) “Etudes de toponymie anatolienne”, RHA XIX.69, 57–98.
(1966) Les noms des hittites, Paris.
Laumonier, A.
(1933) “Notes sur un voyage en Carie”, Revue archéologique II, 31–55.
(1934) “Inscriptions de Carie”, BCH 58, 291–380.
(1958) Les Cultes indigènes de Carie, Paris.
 ABBREVIATIONS AND BIBLIOGRAPHY 501

Leclant, J.
(1951) “Fouilles et travaux en Egypte, 1950–1951”, Or 20, p. 474; pl. LXIV, 37–38.
(1960) Review of Masson-Yoyotte (1956) in: RPh, 339–340.
Legrain, G.
(1905) “Inscriptions from Gebel Abou Gorâb”, PSBA 27, 129.
Levi, D. – Pugliese Carratelli, G.
(1961–62 [1963]) “Nuove Iscrizioni di Iasos”, Annuario della Scuola Archeologica di Atene
vol. 39–40, 573–632.
Loprieno, A.
(1995) Ancient Egyptian, Cambridge.
Martin Cary, J.
(1991) “Carian in Egypt: the Demotic evidence”, Kadmos 30.2, 173–174.
Masson, O.
(1953) “Textes cariens d’Egypte I”, RHA XII (55) 32–38 and pl. XII–XIV.
(1954) “Epigraphie asianique . . . L’épigraphie carienne”, Or 23, 439–441.
(1959a) “Notes d’anthroponymie grecque et asianique”, BzN 10, 159–170.
(1959b) “Documents énigmatiques à inscription pseudo-chypriote et pseudo-carienne”,
Fs. Friedrich, 315–321
(1967) “L’ostrakon carien de Hou-Diospolis Parva (38 Friedrich)”, Fs. Grumach, 211–217,
pl. XX.
(1969) “Les Cariens en Egypte ”, Bulletin de la Société Française d’Egyptologie 56, nov. 1969,
25–36.
(1973) “Que savons-nous de l’écriture et de la langue des Cariens?”, BSL 68/1, 187–213
(1973[74]) Review of Zauzich (1972) in: Kratylos 18, 38–43.
(1974) “Notes d’épigraphie carienne”, Kadmos 13, 124–132.
(1973[75]) “Un nouveau fragment d’inscription carienne de Kaunos”,” Anadolu 17,
123–131.
(1975) “Le nom des Cariens dans quelques langues de l’antiquité”, Mélanges linguistiques
offerts à E. Benveniste, Paris, 407–414.
(1976) “Un lion de bronze de provenance égyptienne avec inscription carienne”, Kadmos
15.1, 80–83.
(1977) “Notes d’épigraphie carienne III–V”, Kadmos 16, 87–94.
(1977[78]) “Karer in Ägypten”, Lexikon der Ägyptologie III, col. 333–337.
(1978) Carian Inscriptions from North Saqqâra and Buhen, Egypt Exploration Society, London.
(1979) “Remarques sur les graffites cariens d’Abou Simbel”, Hommages à la mémoire de
S. Sauneron II, Le Caire, 35–49.
(1988a) “Noms cariens à Iasos”, in: Eothen, Firenze, 155–157.
(1988b) “Le culte ionien d’Apollon Oulios d’après des données onomastiques nouvelles”,
Journal des Savants, Juillet–Décembre 1988, 173–183.
(1991) “Anatolian Languages”, in: The Cambridge Ancient History 2a ed. vol III, Part 2:
The Assyrian and Babylonian Empires and other States of the Near East, from the
Eighth to the Sixth Centuries B. C., Cambridge. 666–676, Bibl.: 855–860.
(1994a) “Les inscriptions cariennes du tombeau de Montuemhat (Thèbes), ” Decifrazione
del cario, 191–194.
(1994b) “La grande inscription grecque d’Abou-Simbel et le nom probablement carien
Peleqos”, SMEA 34, 137–140.
(1994c) Review of Adiego (1993a) in: BSL 89/2, 185–186.
(1995) Review of Kammerzell (1993) in: Kratylos 40, 172–177.
(unpublished [1994]) “Le carien: état de la question”, Communication du 18 juin 1994,
Société de Linguistique de Paris.
502 ABBREVIATIONS AND BIBLIOGRAPHY

Masson, O. – Yoyotte, J.
(1956) Objets pharaoniques à inscription carienne, Le Caire.
Meier-Brügger, M.
(1976) “Zum karischen Namen von Kaunos”, MSS 34, 95–100.
(1978) “Karika”, Kadmos 17, 76–84.
(1979a) “Karika II–III”, Kadmos 18, 80–88.
(1979b) “Ein Buchstabenindex zu den karischen Schriftdenkmälern aus Ägypten”, Kadmos
18, 130–177.
(1980a) Review of Masson (1978) in: Gnomon 52, 383–384.
(1980b) “Karisch. Eine Bestandsaufnahme”, XX. Deutscher Orientalistentag 1977 in Erlangen
(= ZDMG Suppl. IV), 88–90.
(1981) “Eine weitere ,parakarische‘ Inschrift?”, Kadmos 20, 76–78, pl. I–III).
(1983) “Die Karischen Inschriften”, in: Labraunda, Swedish Excavations and Researches II,
Part 4, Stockholm.
(1994) “Ein neuer Blick nach zehn Jahren”, Decifrazione del cario, 111–114.
(1998) “Zu den Münzlegenden von Kaunos”, Colloquium Caricum, 42–46.
Melchert, H. C.
CLL = Melchert (1993b).
DLL = Melchert (2004).
(1993) “Some remarks on new readings in Carian”, Kadmos 32.2 (1993), 77–86.
(1993b) Cuneiform Luvian Lexicon, Chapel Hill.
(1994) Anatolian Historical Phonology, Amsterdam-Atlanta.
(1998) “Carian mdoVun ‘we have established’”, Colloquium Caricum, 33–41.
(2002) “Sibilants in Carian”, in: Fs. Neumann II, 305–313.
(2003) Chapters 1, 2, 5 of H. C. Melchert (ed.), The Luwians, Leiden-Boston.
(2004) A Dictionary of the Lycian Language, Ann Arbor.
(mdane) “Carian mdane”, unpublished paper (1999).
Mentz, A.
(1940) “Schrift und Sprache der Karer”, IF 57, 265–280.
Meriggi, P.
(1963) “Karisch ÉIt≈ana, hier. het. itapana-”, Kadmos 2, 73.
(1966) “Zur neuen ‘para-karischen’ Schrift”, Kadmos 5, 61–102.
(1967) “Zum Karischen”, Fs Grumach, 218–228.
(1978) “Sulla scrittura caria”, ASNP, Cl. di Lett. e Fil. serie III, vol. VIII, 3, 791–803.
(1980) Review of Masson (1978) in: BiOr 37, 33–37.
Metzger, I. R.
(1973) “Eine geometrische Amphora im Rätischen Museum in Chur”, Fs. Bloesch, 74–77
Meyer, G.
(1886) “Die karier. Eine ethnographisch-linguistische untersuchung”, BB 10, 147–202.
Mørkholm, O. – Neumann, G.
(1978) Die lykischen Münzlegenden, Göttingen.
Mørkholm, O. – Zahle, J.
(1976) “The Coinages of the Lycian Dynasts Kheriga, Kherêi and Erbbina. A Numismatic
and Archaeological Study”, AArch 47, 47–90.
Murray, M. A.
(1904) The Osireion at Abydos, London.
Nahm, W.
(1969) “Neue Lesungsvorschläge zur Grotthus-Tafel”, Kadmos 8, 58–73.
Naumann, R. – Tuchelt, K.
(1963/1964) “Die Ausgrabung in Südwesten des Tempels von Didyma 1962”, IstMitt
13/14, 16 ss. [pl. 25: Carian graffiti from Didyma].
 ABBREVIATIONS AND BIBLIOGRAPHY 503

Neumann, G.
(1961) Untersuchungen zum Weiterleben hethitischen und luwischen Sprachgutes in hellenistischer und
römischer Zeit, Wiesbaden.
(1969a) “Eine neue karische Inschrift aus Chalketor”, Kadmos 8, 152–157.
(1969b) “Lykisch”, in: Handbuch der Orientalistik 1 Abt., 2. Bd. 1–2 Abschn., Lief. 2,
358–396.
(1978) “Spätluwische Namen”, KZ 92, 126–131.
(1984) “Zum Namen des Cheramyes von Samos”, Würzburger Jahrbücher für die Altertums-
wissenschaft Neue Folge, Band 10, 41–43.
(1988) “Beobachtungen an karischen Ortsnamen”, in: Eothen 183–191.
(1993) “Zu den epichorischen Sprachen Kleinasiens”, in: G. Dobesch – G. Rehrenböck
(eds.), Die epigraphische und altertumskundliche Erforschung Kleinasiens: Hundert Jahre Kleinaisatische
Kommission der Österreichichesn Akademie der Wissenschaften, Wien, 289–296.
(1994) “Zur Nebenlieferung des Karischen”, Decifrazione del cario, 15–25.
(1998) “Sprachvergleichendes zur Kaunos-Bilingue”, Colloquium Caricum, 19–32.
Nicholls, R. V.
(1971) “Recent Acquisitions by the Fitzwillian Museum, Cambridge”, Archaeological Reports
for 1970 –71, 75–76, no 26, fig. 16.
Ö<ün, B.
(1998) “Warum Kaunos?”, Colloquium Caricum, 175–182.
Otkup“‘ikov, J. V. [OÚÍÛÔ˘ËÍÓ‚, ˛. ‚.]
(1966) KapuÈcÍue ̇‰nucu AÙpuÍu. èp‰‚‡pumeθÌ˚e pÂÁyÎmam˚ ‰e¯uÙpÓ‚Íu. [Carian
Inscriptions of Africa. Preliminary results of a decipherment], Leningrad.
(1968) “O· ÓÚÌÓ¯eÌËË ÍapËÈcÍÓ„Ó aÎÙa‚ËÚa Í ÍpËÚÓ-ÏËÍeÌcÍÓÏy Ë ÍËÔpcÍÓÏy cËÎÎa·ap˲
[On the connection of the Carian alphabet with the Creto-Mycenaean and Cypriot
syllabary]”, Atti e Memorie del I Congresso Internazionale di Micenologia-1967, Roma,
426–432 and 433–437 [abstract in Italian: “Sul rapporto dell’alfabeto cario con il
sillabario cretese-miceneo e cipriota”].
(1989) “K‡pËÈcÍËÈ Ë „pe˜ÂcÍËÈ flÁ˚Í. „eÌeÚ˘ÂcÍËe Ë ÂÚÌÓ-ÍyθÚypÌ˚ ÓÚÌÓ¯ÂÌËfl [Carian
and Greek language. Genetic and ethno-cultural connections]”, Klio 71, 66–69.
Paribeni, R.
(1936) “Etimologie dalla lingua dei Cari (?)”, Rivista di Fiologia Classica 64, 292–293.
Pedley, J. G.
(1974) “Carians in Sardis”, JHS 94, 96–99.
Petrie, W. M. F.
(1901) Diospolis Parva. The Cemeteries of Abadiyeh and Hu, 1898–1899, London.
Pfeiler, B.
(1962) “Zur Münzkunde von Milet”, SM 46, 20–22.
Pisani, V.
(1967) Review of ”evoro“kin (1965) in: Paideia 22, 420–424.
Poetto, M.
(1984) “Nuove monete carie”, Kadmos 23, 74–75.
Pugliese Carratelli, G.
(1948) “Cari in Libia”, PP 3, 15–19.
(1974) “Un’epigrafe caria in Persia”, Gururàjamañjarikà (Studi in onore di Giuseppe Tucci) I,
Napoli, 163–166.
(1985[86]) “Cari in Iasos”, RALinc vol. XI, fasc. 5–6, 149–155.
Ray, J. D.
(1981) “An approach to the Carian script”, Kadmos 20.2, 150–162.
(1982a) “The Carian Script”, Proceedings of the Cambridge Philological Society 208, 77–90.
504 ABBREVIATIONS AND BIBLIOGRAPHY

(1982b) “The Carian inscriptions from Egypt”, JEA 68, 181–198.

(1983) Review of Masson (1978) in: JEA 69, 194–195.
(1985) “The Carian coins from Aphrodisias”, Kadmos 24.1, 86–88.
(1987) “The Egyptian approach to Carian”, Kadmos 26.1, 98–103.
(1988) “Ussollos in Caria”, Kadmos 27.2, 150–54.
(1990a) “An Outline of Carian Grammar”, Kadmos 29, 54–83.
(1990b) “A Carian Text: The Longer Inscription from Sinuri”, Kadmos 29, 126–132.
(1990c) “The names Psammetichus and Takheta”, JEA 76, 196–199.
(1994) “New Egyptian Names in Carian”, Decifrazione del cario, 195–206.
(1998) “Aegypto-Carica”, Colloquium Caricum, 125–136.
Robert, L.
(1945) Le sanctuaire de Sinuri près de Mylasa Première partie: les inscriptions grecques,
Paris.
(1949) “Décret d’une syngeneia carienne au sanctuaire de Sinuri”, in: L. Robert, Hellenica
VII, 59–68.
(1950) “Inscriptions inédites en langue carienne”, in: L. Robert, Hellenica VIII, section
I, 5–22, fig. 1–2, pl. I–VI, VII–X, XXVIII–XXX.
(1950b) “Le carien Mys et l’oracle du Ptôon (Hérodote, VIII, 135)”, in: L. Robert,
Hellenica VIII, section I, 23–38.
(1957) “Deux inscriptions d’Iasos”, REG 70, 362–375.
Robinson, E. S. G.
(1936) “A Find of Archaic Coins from South-West Asia Minor”, NC 5/16, 1–16.
(1939) “Coin-Legends in Carian Script”, Anatolian Studies presented to W. H. Buckler,
Manchester, 269–275.
Rochette, B.
(1997–98) “La langue des Cariens à propos de B 867”, Glotta 74, 227–236.
Roos, P.
(1972) The Rock-Tombs of Caunus, I The Architecture (= Studies in Mediterranean Archaeology
34; 1), Göteborg.
(1985) Survey of rock-cut chamber-tombs in Caria, Part 1: South-eastern Caria and the Lyco-Carian
borderland, 9, Göteborg.
}ahin, M. Ç.
(1973[75]) “Lagina’dan (Koranza) iki yeni yazıt” / “Two New Inscriptions from Lagina
(Koranza)”, Anadolu 17, 178–185 (Turkish) / 187–195 (English).
(1976) The political and religious structure in the territory of Stratonikeia in Caria, Ankara.
(1980) “A Carian and three Greek inscriptions from Stratonikeia”, ZPE 39, 205–213.
Salmeri, G.
(1994) “I Greci e le lingue indigene d’Asia Minore: il caso del cario”, Decifrazione del
cario, 87–99.
Sapir, E.
(1936) “kÊbda a Carian Gloss”, JAOS 56, 85.
Sayce, A. H.
(1874) “The Karian Inscriptions”, Transactions of the Royal Society of Literature, second
series, 10, 546–564.
(1887[92] = 1893) “The Karian language and inscriptions”, TSBA 9.1, 112–154.
(1895) “The Karian and Lydian inscriptions”, PSBA 17, 39–43.
(1905) “Lydian and Karian inscriptions”, PSBA 27, 123–128 Lám. I–II.
(1906) “An inscription of S-ankh-ka-ra. Karian and other inscriptions”, PSBA 28,
171–177
(1908) “Karian, Aramaic, and Greek graffiti from Heshân”, PSBA 30, 28–29.
(1910) “Karian, Egyptian and Nubian-Greek inscriptions from the Sudan”, PSBA 32,
261–268.
 ABBREVIATIONS AND BIBLIOGRAPHY 505

Säflund, G.
(1953) “Karische Inschriften aus Labranda”, Opuscula Atheniensia I, 199–205.
Schweyer, A.-V.
(2002) Les Lyciens et la mort. Une étude d’histoire sociale, Paris.
Schmaltz, B.
(1998) “Vorhellenistische Keramikimporte in Kaunos—Versuch einer Perspektive”,
Colloquium Caricum, 203–210.
Schmitt, R.
(1978[79]) Review of Masson (1978) in: Kratylos 23, 98–104.
(1980) “Karer”, in: Reallexikon der Assyrologie und vorderasiatischen Archeologie V, 423–425.
Schürr, D.
(1992) “Zur Bestimmung der Lautwerte des karischen Alphabets”, Kadmos 31, 127–156
(1991–1993) “Imbr- in lykischer und karischer Schrift”, Die Sprache 35.2, 163–175.
(1993) “Zu ]NAPOUKV SGDI 5727.b4”, Kadmos 32, 172–173.
(1996a) “Bastet-Namen in karischen Inschriften Ägyptens”, Kadmos 35, 55–71.
(1996b) “Zur karischen Felschinschrift Si. 62 F”, Kadmos 25, 149–156
(1996c) “Zur karischen Felsgrabinschrift von Kaunos (28*)”, Kadmos 25, 157–159.
(1996[98]) “Karisch ‘Mutter’ und ‘Vater’”, Sprache 38, 93–98.
(1998) “Kaunos in lykischen Inschriften”, Colloquium Caricum, 143–162.
(2000) “Lydisches III: Rund um lydisch ,Hund‘”, Kadmos 39, 165–176.
(2001a) “Zur Inschrift nr. 50 von Kaunos und zum karischen Namen von Keramos”,
Kadmos 40, 61–64.
(2001b) “Karische und lykische Sibilanten”, IF 106, 94–121.
(2001c) “Zur karischen Inschrift auf dem Genfer Kultgegenstand”, Kadmos 40, 117–126.
(2002) “Karische Parallelen zu zwei Arzawa-Namen”, Kadmos 41, 163–167.
(2003a) “Zur karischen Inschrift der Stele von Abusir”, Kadmos 42, 91–103.
(2003b) “Zum Namen des Flusses Kalbis bei Kaunos in Karien”, HS 116, 69–74.
”evoro“kin, V. I. [ò‚ÓÓ¯ÍËÌ, Ç. à.]
(1962) “ä‡ËÈÒÍËÈ ‚ÓÔÓÒ [The Carian problem]”, VJa 1962.5, 93–100.
(1963) “O ıÂÚÚÓ-ÎÛ‚ËËÒÍÓÏ ı‡‡ÍÚÂ ͇ËÈÒÍÓ„Ó flÁ˚͇ [On the Hittite-Luwian charac-
ter of the Carian Language]”, VJa 1963.3, 83–84.
(1964a) “On Karian”, RHA XXII 74, 1–55.
(1964b) “Aegyptisch-karische Inschrift am Sockel einer Isisstatuette (Leningrader Staats-
ermitage)”, RHA XXII 74, 57–65 and pl. I–IV.
(1964c) “Zur karischen Schrift und Sprache”, Kadmos 3, 72–87.
(1964d) “Karijskij jazyk; sovremennoe sostojanie de“ifrovki i izu‘enija” [La lengua caria;
el estado actual del desciframiento y la investigación] Problemy indoevropejskogo jazykoz-
nanija, Moskva, 18–39.
(1965) àÒÒΉӂ‡ÌËfl ÔÓ ‰Â¯ËÙÓ‚Í ͇ËÈÒÍËı ̇‰ÔËÒÂÈ [Studies on the decipherment
of the Carian inscriptions], Moskva.
(1968a) “Zur Entstehung und Entwicklung der kleinasiatischen Buchstabenschriften”,
Kadmos 7, 150–173.
(1968b) “Karisch und Lykisch”, Atti e Memorie del I Congresso Internazionale di Micenologia-
1967, Roma, 462–472.
(1968c) “Karisch, Lydish, Lykisch”, Klio 50, 53–69.
(1969a) “Zu den ‘späthethitischen’ Sprachen”, XVII. Deutscher Orientalistentag-1968, Würzburg
= ZDMG Supplem. I. Wiesbaden, 250–271.
(1969b) “Zur Erforschung der kleinasiatischen Onomastik”, 10. Internationaler Kongress für
Namenforschung II, Vienna, 341–350.
(1977) “Zu einigen karischen Wörtern”, MSS 35 117–130.
(1982–83) “Über den Lautwert des karischen Buchstaben y”, InL 8, 71–78.
(1984[86]) “Verbesserte Lesungen von karischen Wörtern”, InL 9, 199–200.
(1988) “Carian proper names”, Onomata 12, 497–505.
506 ABBREVIATIONS AND BIBLIOGRAPHY

(1991) “On Carian Language and Writing”, Perspectives on Indo-European Language Culture
and Religion. Studies in Honor of Edgard C. Polomé. vol. I (= Journal of Indo-European
Studies. Monograph Number 7). 116–135.
(1992) “On Carian Language and Writing”, Newsletter for Anatolian Studies Vol. 8/1, 3.
(1994) “Carian – Three decades later”, Decifrazione del cario, 131–166.
Shafer, R.
(1961) Review of Masson-Yoyotte (1956) in: RHA XIX. 68, 39–40.
(1965) “A break in the Carian Dam”, AC 34, 398–424.
Six, J. P.
(1890) “Monnaies grecques inédites et incertaines”, NC, 185–259.
Snodgrass, A. M.
(1964) “Carian armoures. The Growth of a Tradition”, JHS 84, 107–118.
Spiegelberg, W.
(1928) “Eine Ichneumonbronze mit hieroglyphischer und karischer Inschrift”, OLZ 21,
545–548.
Steinherr, F.
(1950–51) “Zu den neuen karischen Inschriften”, Jahrbuch der kleinasiatische Forschung 1,
328–336
(1955) “Der karische Apollon”, WO 2, 184–192.
(1957) “Der Stand der Erforschung des Karischen”, Proceedings of the twenty-second Congress
of Orientalists . . ., Istanbul, 1951, Leyde, 44–49.
Stoltenberg, H. L.
(1958a) “Neue Lesung der karischen Schrift”, Die Sprache 4, 139–151.
(1958b) “Die karische Grabinschrift von Kaunos”, AC 27, 108–109.
(1959) “Deutung karischer Inschriften”, ArOr 27, 1–4.
Sundwall, J.
(1911) “Zu den karischen Inschriften und den darin vorkommenden Namen”, Klio 11,
464–480.
(1913) Die einheimischen Namen der Lykier nebst einem Verzeichnisse kleinasiatischer Namenstämme
[= Klio 11, Beiheft], Leipzig.
Thompson, M.
(1966) “Some Noteworthy Greek Accessions”, ANSMN 12, 1–18.
Tischler, J.
(1977) Kleinasiatische Hydronimie. Semantische und morphologische Analyse der griechischen
Gewässernamen, Wiesbaden.
(2001) Hethitisches Handwörterbuch, Innsbruck.
Torp, A.
(1903) Die vorgriechische Inschrift von Lemnos = Christiania Videnskabs-Selskabs Skrifter,
hist.-fil. Kl. 1903, 4.
Tremblay, X.
(1998) “Controversa Carica”, Colloquium Caricum, 109–124.
Treu, M.
(1954) “Eine griechisch-karische Bilingue und ihre Bedeutung für die Geschichte der
karischen Schrift”, Glotta 34, 67–71.
Troxell, H.
(1979) “Winged Carians”, in: O. Mørkholm—N. M. Waggoner (eds.), Greek Numismatics
and Archaeology, Essays in Honor of Margaret Thompson, Wetteren, 257–268.
 ABBREVIATIONS AND BIBLIOGRAPHY 507

(1984) “Carians in Miniature”, in: A. Houghton – S. Hurter (eds.), Festschrift für /

Studies in honor of Leo Mildenberg, Wetteren, 249–257, pl. 40.
Tuchelt, K.
(1970) Die archaischen Skulpturen von Didyma (= Istanbuler Forschungen 27), Berlin.
Tzanavari, K. – CHRISTIDIS, A.-Ph.
(1995) “A Carian Graffito from the Lebet Table, Thessaloniki”, Kadmos 34, 13–17.
van den Hout, T.
(1999) “-ms(-): a Carian Enclitic Pronoun?”, Kadmos 38, 31–42.
Varinlio<lu, E.
(1986) Die Inschriften von Keramos, Bonn.
Vittmann, G.
(1996) “Zum Gebrauch des k3-Zeichens im Demotischen”, SEAP 15, 1–12.
(2001) “Ägyptisch-Karisches”, Kadmos 40, 39–59.
(2003) Ägypten und die Fremden im ersten vorchristlichen Jahrtausend, Mainz.
Waerzeggers, C.
(Borsippa) “Carians in Borsippa”, Iraq (forthcoming).
Winzer, A.
(2005), Antike Portraitmünzen der Perser und Greichen aus vor-hellenistischer Zeit (Zeitraum ca.
510–322 v. Chr), March-Hugstetten.
¥ába, Z.
(1974[79]) The Rock Inscriptions of Lower Nubia, Czehoslovak Concession, Prague.
Zahlhaas, G. – Neumann, G.
(1994) “Zwei neue karische Inschriften aus München”, Kadmos 33, 160–168.
Zauzich, K.-Th.
(1972) Einige karische Inschriften aus Ägypten und Kleinasien und ihre Deutung nach der Entzifferung
der karischen Schrift, Wiesbaden.
Zgusta, L.
KON: see (1984).
KPN: see (1964a).
(1964a) Kleinasiatische Personennamen, Prague.
(1964b) Anatolische Personennamensippen, Prague.
(1968) Review of ”evoro“kin (1965) in: ArOr 36, 153–154
(1970) Neue Beiträge zur kleinasiatischen Anthroponymie, Prague.
(1984) Kleinasiatische Ortsnamen, Heidelberg.
 TABLE I
THE CARIAN ALPHABET

Nº (Masson) Letters A1 Transcription Notes

1 a A ~ À (E ) a
3 dDG d
4 l l
5(+41) W ù VV y (formerly ù/ü)
6 r R r
7 L 2 L l
9 qQ q
10 b5B b
11 mM m
12 o o
14 tT t
15 f F S òT _ “
17 s s
18 H ?
19 uU u
20 ñ ñ
21 xX ∞ (formerly x)
22 nN n
24(+2) p p (¯) p
25 zZ ≤
26(+8) IíÎÏyìYI i
27 ee e
28 w ÿ (formerly w)
29+30 kK k
31 & d
32 vÚ w (formerly ú)
33(+34) 0 8? g
35(+36) 199 z (formerly z)
37 % ∆
38 j_ j (formerly í)
39 _1 ?
40(+23?) c C / O? t / t2?
42 6 ®
43–44–45 Ø ® 4 B &? b? b
46 ÿ b2?
1
Note that, for the sake of simplicity, the letters are given in a left-to-right direc-
tionality, which in some cases is not actually documented (for example, ® is only
attested as ç in right-to-left directionality).
 TABLE II
CARIAN SIGNS IN COIN LEGENDS

LETTERS
6 5 [ A l (!) a M Q À a M12–21, M32
B u (!) 5 } H(!) b M22–23, M26–28, M32, M36–37,
M41
£d_ d M32, M36–37, M42
gr r M32, M51–52
l l M50
W¥ y (or rather w?) M5, M23, M47–48
L l M39
Mm« m M1–5
o o M20–21, M50–52
tF t M31, M39
s s M33–37, M42
u u M39, M45, M50–52
n n M25
ñ ñ M32
xX ∞ M43–45
z ≤ M39
iÌ i M31–32, M53
p°P p M33–35, M39–40
k l (!) k M22–30
VÚ w (or rather y?) M5–M11, M49
N0 g(?) M33–35
T j M23, M36–37
3489 z M12–21
71 (z or symbol?) M38–39
c t M32
J letter or M27, M33, M36–37
separation
mark
^ letter or symbol? M46
510 TABLE II

LINEAR DEVICES
o M1
l M31
{ M33–35
h M38
7 1 (or letter?) M38–39
^ (or letter?) M46
 INDICES1

1. ANCIENT PROPER NAMES

Carian Proper Names in Indirect Sources

In Egyptian Arrissiw 135, 221, 248, 305, 354, 356
3rskr 19, 47, 194, 195, 243, 248, 250, Artaow 255, 356
431 Arthumow 255, 356, 393
Jr“3 42, 47, 194, 248 Artimhw 356, 410
Jrym3 42, 194, 248 Aryuassiw 356
K3rr 33, 198, 244, 266, 409 Aryuassiw 356
Prjm 41, 188, 194, 195, 242, 248, 249 Babein 246
”3rkbym 32, 191, 194, 195, 248, 250, Bãrgasa/Bãrgaza 246
251, 416, 417 BargÊlia 246
Beryaw 246
In Akkadian Berrablviow 246
Lu-uk-“u 380 Bolli.evn 246
Boivmow 246
In Greek Br¤oula 246
Aba 340, 349 Bridaw 246
Abaw 340 Bruajiw/Bruassiw 246
Abbaw 340 Brvlow 246
Ada 13, 140, 236, 245, 340, 349 BubassÒw 246
Adaw 340 Bvnitv 246
Aktadhmow 330 Bvrand/a/ 246
Aktauassiw 242, 330, 338 Gugow 198, 237, 244, 260, 326, 330,
Aktaussvllow 330, 378 334, 408, 419
ÉAlãbanda 8, 11, 12 Dandvmow 246
Alasta 349 Daru..ow 246
Alikarnassow 255, 351 D°dmasa 246
Amiaw 340 Deibow 246
Ammiaow 340 Dersvmanhw 246
Andarsvw 246, 262, 333, 363 Dersvw 246, 363
Androsvw 246, 262 Dersv . . . tiw 246
Appa 340 Didassai 246
Arbhs(s)iw 334 D¤duma 246
Arduberow 245, 255, 283, 333, 353 DÊndason 246
Ardur/a/ 333 Ekamuhw, Ejamuhw 330
Ariauow 354 Ekatomnvw 140, 237, 238, 243, 249,
Aridvliw 354 255, 289, 375, 378
Arlissiw 196, 248, 250, 255, 341, Ermapiw 255, 331, 339
354, 355, 356 Zermenduberow 413
Arlissow 355 Yuagg°l/a/ 12
Arlivmow 248, 343, 355 Yussvllow 330, 362

1
For Carian words, see Chapter 11.
512 INDICES

Iasow 270, 433 Maussvllow 330

Ibanvlliw 360 Minnaw 340
Idagugow 237, 238, 239, 244, 245, 330, Moiw 384
334, 361 MonÒgissa 8
Idriaw 308, 438 MÒtulow 336
Iduma 367, 368 Mouzhaw 251, 336, 387
*Idurigow 368 Mundow 386
Idussvllow 245, 330, 361, 362, 367, Nana 13, 340
368 Nanaw 340
Imbarhldow 198, 262, 330, 335, 393 Nannixow 340
ÖImbramow (var. ÖImbrasow) 8, 308, 335 Nannow 340
Imbarsiw 236, 247, 250, 255, 262, 335, Nannv 340
360, 366 Naras/a/ 333, 387
Imbras(s)iw 236, 247, 250, 255, 262, Narasow 262
335, 341, 360, 361, 366 Narbaw 333
ÖImbrow 8, 299, 335, 367 Narisbara 387
Karbasuand/a/ 406 Naruandow 333, 387
Karusvldow 330 Neterbimow 238, 332, 339, 343, 347
Kasbvlliw 334, 374 Nonnow 340
Kasvlãba 237, 238, 243, 245, 375 Nutar 332
Kbondiassiw 334 Nvtrassiw 332
Kbvdhw 334, 371 Jrmedu<be>row 255,
Kebivmow 33, 236, 237, 238, 243, 245, Oaloalow 339, 428
249, 330, 334, 343, 371, 416, 417 Oa3a3iw 242
Kildar/a/, Killar/a/ 141, 236, 243, Oliatow 237, 243, 330, 339, 428
248, 373 Ordomaw 333
Kinduh 255, 269, 373 Osogva 8, 169
Koar(r)end/a/, Koarenz/a/ Otvrkond/a/ 332
Koaranz/a/, Kvra(n)z/a/ 438 Panablhmiw 330, 338
Kolaldiw 238, 239, 244, 256, 315, 407 Panamara 335, 338
Konodvrkond/a/ 332 Panamuhw 330, 336, 338, 357
Kostvlliw 251, 409 Panuassiw 242, 330, 338, 340
Kotobalvw 244, 408 Paow 395, 419
Kotbelhmow 198, 236, 238, 244, 245, Papaw 340
255, 343, 408 Papiaw 340
Kuatbhw 309, 408 Pãrgasa 246
Kubliss/ow/, Kublisse›w 238, 245, Par(a)ussvllow 236, 330, 340, 393,
255, 269, 409 394
Kulaldiw 238, 239, 244, 248, 315, Paraudigow 236, 238, 263, 330,
407 394
Kuròmew 338 Parnow 417
Kutbelhmiw 160, 198, 236, 238, 244, Paruv 394
245, 255, 343, 408 Parembvrda 333, 340
Latmow 440 Pedanass/ow/ 336
Lujhw 182, 248, 255, 380 Pedvldow 337
Manhw 236, 381 Peigelasow 337
Mareuw 335 Pel(l)ekvw, Peleqow 237, 238, 293,
Marow 335 399, 437
Masanvrada 237, 238, 239, 245, 249, Peldhkow 236, 238, 400
250, 269, 307, 332, 333, 385 Phdasa 336
Masanvradow 333, 385 Pigassvw 337
Massarabiw 332, 339 Piginda 337
Massariw 332 Pigrhw, Pikrhw 31, 236, 243, 244, 248,
Massvn/a/ 332 255, 337, 345, 397
Mãstaura 8 Pidasa 336
 INDICES 513

Pidossus 337 Spareudigow 255, 263

Pijvdarow 337, 399 Svmnhw 236, 237, 249, 255, 409, 414
Ponmoonnow 140, 238, 239, 249, 255, Tãbai 9
303, 330, 336, 338, 401 Tarkonda[ 255
Ponussvllow 197, 237, 238, 243, 249, Tarkondar/a/ 332, 423
330, 338, 401, 405 Tata 340
Pormounow 303, 336, 338 Tatarion 340
Pounomoua 336, 338 Tataw 340
Samul¤a 336 Territow 357
Samv(u)ow 141, 310, 416 Tonnouw 237, 238, 239, 243, 249, 250,
Sanortow 419 422
*Sarkebivmow 33, 416, 417 Tralle›w, Tralde›w 18
Sarnoj 417 Tumnhw 440
*Sar-oliatow 417 TumnhsÒw 9
Sarow 419 Tur[ 164, 197, 288, 424
Saruassiw 242, 330, 338, 340 Uliatow 36, 237, 243, 330, 339,
*Sar-uliatow 417 428
Sarussvllow 196, 236, 250, 263, 330, Ñ`Ullãrima 196, 200, 238, 240,
340, 418, 419 257
Saskvw 175 `ÑUlloÊala 8
Saurigow 237, 263, 289, 308, 330, Urgosvw 237, 244, 255, 434
340, 418 huròmew 338
Saussvllow 263, 330, 340, 418 Urvmow 338
Semeuritow 263 Ussaldvmow 240
Senurigow 263, 289, 330 Usseldvmow 240
Seskvw 175 Ussvllow, Ussvldow 135, 141, 183,
Sinuri 434 195, 196, 237, 305, 310, 330, 344,
Skoaranow 255 362, 403, 424, 431
Souãggela 8, 11, 20, 238, 251, 269, Xasbvw 326, 334, 374
277, 415 Xeramuhw 330

Other Anatolian Proper Names

In Greek Kendhbaw 373
Arbasiw 334 Kouadapemiw 339
Armadapimiw 339 Kougaw 334
Armapia 331 Ktibilaw 340
Armapiaw 331 Mollisiw 341
Arpigramow 353 Mousathw 336, 386
Arteimianow 410 Moshta 336, 386
Arteimaw 356, 410 Moushta 336, 386
Arteimow 356, 410 Moutaw 386
Arteimhw 410 Janduberiw 222, 245, 255, 375, 419
Artimaw 410 Oliw 421
Artimhw 356, 410 Oraw 338
GÊghw 334 Oualiw 421
Dada 13 Ouajamoaw 255, 336, 427
Eida 334 Ouajamvw 336, 427
Eidassala 334 Ouramoutaw 338
Ermapiaw 331 Panamuaw 336
Ida 334 Perpenduberiw 402
ÖImbrasow 335 Pigramiw 337, 361, 397
Imrougara 335 Pigramow 337, 361, 397
Kendebhw 373 Pigrassiw 337
Kendhbhw 373 Pijaw 399
514 INDICES

Pijedarow 399 Tar¢untaradu 385

Ponamoaw 405 Ura 338
Salaw 334 Urawalkui 338
Sedeplemiw 338 Walawala 428
Tabhnoi 9 Walawali 428
Tarkumbiou (gen.) 339 Wallarima 240
Triendasiw 341
Trokoarbasiw 334 In Hieroglyphlic Luwian
Tarhu(n)t- (TONITRUS-hut-) 347
In cuneiform sources
Àla- 334 In Lycian
Arma- 306, 331, 347 Erbbina 177
Ayami/aimi 343 Esedepl‘emi 338
›u-da-ar-lá 406 Hrixttbili 340
›u-du-ur-lá 406 Idazzala 334
›u-u-tar-li 406 Ijãna 369
›u-ut-ra-la-(a“) 406 Ipresidi 335
›u-ut-ra-li-i“ 406 Katamla- 244
Immaraziti 335 Mullijese/i- 341
Lukka 342 Natr- 162, 332, 389
Massanaura 338 Natrbbij‘emi 238, 332, 339, 343, 347,
Millawa(n)da 342 389
Mutamutassa 342 Punamuwe 405
Muwatalli 336 SedeplMmi 338
Muwaziti 336, 386 Teleb(ehi) 177
Nattaura 338 Trij‘etezi- 341
Petassa 336 Trqqñt- (nom. Trqqas) 239, 247, 259,
Pi¢irim 337 260, 331, 347, 356, 423
*Piyama- KAL 339
d Urebillaha (dat.) 318
Piyamaradu 339, 385 Xbide- 203, 243, 245, 255, 297, 298,
Pitassa 336 371
Puna-A.A = *Punamuwa- 405 Zzala 334
Runtiya- 333
Tar¢u- 331, 347, 356, 423 In Milyan
Tar¢unt- 239, 247, 259, 260, 347, 356, Kridesi 407
423 Trqqñt- (nom. Trqqiz) 260, 331, 334
Tarhundapiya 339 Xbadiz 261

Greek Proper Names

In Greek ÑErm¤aw 135, 305, 307
ÉAy∞nai 237, 243, 259 ZeÊw 289, 375
ÉAyhna›ow 155, 203, 237, 296 ÉIdrieÊw 140, 255, 438
ÉApollÒdotow 332, 339, 389 ÑIpposy°nhw 155, 203, 236, 296, 298,
ÉAristokl∞w 164 316
ÖArtemiw 356 KarikÒn 2
ÉArt°mvn 238, 357 Kãttouza 9
ÉArx°laow 354 Kaun¤oi 155, 203, 296, 297, 299, 360
Bragx¤dai 20, 251, 285, 402 K≈Ûoi 271
Brãgxow 290 K«w 8
ÑEkata›ow 236, 238, 288, 375, 378 Lusikl∞w 155, 203, 236, 237, 243,
*ÑEkatÒmnhstow 378 248, 296, 380
ÑEllhnikÒn 2 Lusikrãthw 155, 203, 236, 237, 243,
ÑErm∞w 331 248, 296, 380
 INDICES 515

M¤lhtow 342 TÊxh 308

Nikokl∞w 155, 203, 236, 249, 296, ÑUbr°aw 432
380, 388 Fãnhw 135, 305
OÈliãdhw 236, 245, 248, 428, 432 Crusaor¤w 416
ÉOrsikl∞w 431
Perikl∞w 431 In Lycian
P¤ndarow 242, 396 Lusãñtrahñ (gen.) 304
SkÊlaj 164 Perikle 431

Egyptians Proper Names

In Egyptian Ô3j-Óp-jm=w 424
J‘˙(?)[ 47 Ô3[ = *Ô3-n-n3-j˙w 194, 425
J.r=w 369
Jr.t=w-r.r=w 370 In Akkadian
J-Ór 353 Ta¢ma““i 420
Js.t 128
Jtm 32, 424 In Greek
W3h-jb-r‘-nb-[ 38, 387 ÖAmasiw 293
B3st.t 251, 398 ÑApimenhw 353
P3j-Wsir 398 Avw 353
P3-whr 405 ÉEsour 369
P3-n-jwjw 397 ÉEsouriw 369
P3-n-Wsjr 398 Yamvw 251, 424
P3-n-Nj.t 393 ÉIyorvw 370
P3-sb3-¢‘j-m-njw.t 350 Nek«w, Nexvw 244, 249, 388
P3-dj-Jnp 404 Nefervw 242, 388
P3-dj-Jtm 395 Nitvkriw 243, 248, 249, 315, 389
P3-dj-‘nq.t 404 Panitiw 393
P3-dj-B3st.t 245, 395 Peteyumiw 395
P3-dj-(p3)-n∆r 404 Petenaiyiw 395
P3-dj-Njt 33, 198, 242, 243, 245, 266, Petenhyiw 395
395 Petenht 395
*P3-dj-nfr 404 Petenoupiw 404
P3-dj-Ór-sm3-t3wy 244, 250, 399 Petepnouyiw 404
P3-dj-sm3-t3wy 244, 250, 399 Petetumiw 395
P3-dj-st 39 Petyubestiow 395
*P3-dj-t3wy 404 Petobastiw 245, 395
Psm∆k 194, 242, 249, 250, 398, Petoubastow 395
403 Petoubestiw 245, 395
Psm∆k-‘wj-Njt 38, 198, 403 Pieuw 397
Pt˙-ms 420 Pihow 397
Nj.t 249 Pih# 397
Nj.t-jqr 243, 248, 249, 389 Pisiriw 398
Ny-k3w 244, 249, 388 Potasimto 244, 250, 293, 399
Nfr-˙r 242, 388 Pouvriw 405
Ns-˙r 369 Povriw 405
Ó3py 41 Samauw 251, 425
Óp-mn 353 Samw#w 251, 425
Ór 391 Tamvw 251, 424
T3-dj(t)-wsir 39 Tetobastiw 245, 423
T3-dj(.t)b3st.t 245, 423 Cammhtixow Cam(m)atixow 293, 398,
Ô3j-jm=w, Ô3j-n.jm=w 251, 194, 195, 403
416, 424 Cousennhw 350
516 INDICES

Old Persian Proper Names

In Old Persian In Greek
Vi“tàspa- 370 ÑUstãsphw 370

2. LANGUAGES

Carian glosses in Greek tàta/i- 259, 273, 347, 421

êla 8, 11 upatit- 433
bãnda 8, 11 ura- 338, 430
g°la 8, 11 u““a/i- 344
g¤ssa 8 wallant- 338
ko›on, kÒon 8 10 walli( ya)- 339
soËa(n) 8, 10, 415 -wanni- 257, 260, 346, 367, 369, 371
tãba 9 wayamman- 427
toussÚloi 9 wà“u- 344
tumn¤a 9 za- 259, 288, 346, 410
zar“iya- 411
Cuneiform Luwian (CLuw)
à/àya- 327, 343 Hieroglyphic Luwian (HLuw)
annara/i- 262, 333, 352, 387 ataman- 355
ànnari- 333 atra/i- 258, 392
ànna/i- 347, 259, 273, 364 hasu- 334
apa- 359 nimuwiza- 383
*arada-,* aradu- 332, 333, 385 pa- 359
arpa- 333, 334 para/i 340
*a““att(i)-, a““atta““a/i- 334 piya- 282, 284, 339
aduna (Inf.) 299 tata/i- 347
-¢a 306 ura/i- 338
¢ant- 372 wa/iliya 339
¢andawat(i)- 294 -wani- 346
¢àpa/i- 334 za- 259, 288, 346, 410
*¢apài- 334
¢ù¢a- 334, 345 Hittite (Hitt.)
¢utarlà- 248, 260, 406 anda 287
im(ma)ra/i- 335, 346, 393, 422 andan 254, 287, 363
im(ma)ralla/i- 335 anna- 259, 347
im(ma)ra““a/i- 335 ak(k)- 350
*im(ma)ra““iya- 335 ar¢a 304
kui- 243, 259, 320, 377 ard- 333
-mma/i- 339, 343, 347 atta- 347
mà““an(i)- 260, 307, 327, 332, 347 ¢anna- 332
mà““ana““a/i- 352 ¢ant- 372
ma“¢a¢it- 351 ¢u¢¢a- 334
muwa- 260, 335 innarà- 333, 387
nana- 284, 321, 388 ka- 288, 346, 410
-ppa 302 kàn(i) 259, 288
parì 340 kappi- 352
pi¢a- 337, 345, 397 katta 300, 351
piya- 282, 284, 304, 327, 339, 396 kattan 300
pùna- 259, 337 gim(ma)ra- 335, 346
pùnata/i- 337 kui- 243, 259, 260, 320, 377
“arra 340 gurta- 407
“arri 260, 261, 302, 340, 411, 416 mald- 323, 427
 INDICES 517

malai- 427 pije- 282, 284, 304, 325, 327, 339, 396
midà(i)- 382 -ppi 302
muwa- 335, 386 pri 40
nà(i)- 284, 321, 388 prñnawa- 290
naru- 333 punãma- 259, 337
nu- 285, 390 rMmazata 254
pài-/piya- 282, 284, 304, 325, 339 se 347, 411
peda- 304, 336 s=eññe 309
parà 259, 260, 340 s‘eñnait‘e 413
piyannài- 359 sije- 259, 412
pulla- 401 tãtu (imperative) 349
“ara 340 tede/i- 245, 259, 260, 273, 346, 347,
“èr 260, 340, 416 421
“iu(n)- 332 ti 243, 259, 320, 377
-talla- 428 tike 377
ura/i- 338 trMmili( je)- 270
walli- 339 tuwe-, tu-s- 325
*walli- 338 ube- 201, 239, 245, 259, 282, 321, 327,
walliwalli- 260, 338, 339, 428 347, 432
wiyài- 427 uhe/i- 344
xahba 258, 261, 326, 334
Lycian (Lyc.) xawa- 10
a(i)- 299, 300, 301, 309, 321, 347, 349 xbid‘eñne/i- 203
arawa- 300, 363, 435 xñtawa- 294
arMma 258, 347 xñtawat(i)- 10, 260, 294, 364
atãnaze/i- 259 xuga- 260, 326, 334, 345
atra/etli- 204, 258, 297, 392 xupa- 290, 429
-be 302 -za 294
ebe- 290, 320, 359 -ze/zi- 270, 392
ebeli 365 zeus- 288, 375
ehbi 348
ehbi( je)- 300 Milyan (Mil.)
‘ene/i- 259, 273, 313, 347, 352, 364 ddxug[ 362
epñn‘eni 352 erbbe/i- 334
erbbe- 333 erbbesi 334
eri- 304, 364 esetesi 334
ese- 294 ki 243, 259, 320, 377
hri- 260, 261, 302, 340, 416 -kike 377
hrppi 302 masa- 260, 261, 237, 307, 332, 347
ipre 335 muwa- 260, 335
kbatra- 258 sebe 201, 239, 245, 327, 347, 411
maha(na)- 260, 307, 327, 332, 347, 352 -wñni 204, 257, 260, 346, 371
mahanahe/i- 313, 352 xñtaba 373
maraza- 335 xñtabasi 373
me 324 xruwasaz (nom.-acc. pl.) 261
-me/i- 339, 343, 347 xuga- 334
mere- 335, 422 zri- 260, 261, 416
mle- 307
Mmai- 304, 390 Lydian (Lyd.)
n‘eni 352 armta- 347
nere/i- 292 artimu- 356
-ñni 204, 260, 346, 369, 371 bi- 359
ñte 254, 287, 319, 363 bil-l 396
pdd‘en- 247, 304, 336, 396 civ 332
518 INDICES

‘ena- 346, 364 Latin

i≤tube lm- 370 orca 391
kãn- 406 urceus 391
sareta 411 urna 391
serli-/selli- 261
taada- 260, 346, 347, 421 Old Persian
-wn 298 kºka- 1
haya/taya 320
Sidetic
artmon 357 Proto-Indo-European (PIE)
ma≤ara (dat. pl.) 260, 332, *bhèh2 - 246, 257, 260, 337, 345, 361
347 *dyew- 332
*h2e/owo- 10
Pisidian *h2ent- 260
oudoun 426 *h2n¶r 387
*h2onsu- 261
Proto-Anatolian (PA) *∞e 411
*anna-s 259 *∞ei- 259, 412
*armà- 258, 260 *∞o-/Keh – 288, 410
2

*/be:H/- 257 *kwe 182

*dáda-s 259 *kwi- 202, 259, 266, 320, 377
*·emro- 262, 346 neyH- 284, 321, 388
*Hàwo- 10 *–-mºto- 262
*HuHo- 260 *nu- 285
*∞o-/∞à- 259, 410 *pedo- 247, 336
*kwis 259, 266, 320, 377 *terh2 - 331
*obó/i- 359 *8elH- 339
*pédon 336 *8elh1- 339
prò/prò 259, 260
*TºH–t- 259 Akkadian
*narû 333
Greek
êggelow 11 Egyptian
êmbrotow 262 ‘3 32
aÊtoÊw (acc. pl.) 392 ‘n¢ 32, 41, 128
•autÒn (acc.) 392 ym 33
¶doje (aor.) 426 wnwtj 278, 429
eÈrg°taw (acc. pl.) 428, 429 p3 w˙m 41, 355
yeÒw 11 p3 ˙m-n∆r n Jmn 400
kombow, kombion 352 n 47
p°lekuw 399 n∆r 32, 389
proj°nouw (acc. pl.) 382, 411 s3 47
s∞ma 288, 415 snb 32
Ïrxh 391 dj 32, 41, 128, 399

Coptic
eiom 33
PLATES
 521

Map 1. Carian inscriptions in Caria.

522

Map 2. Carian inscriptions in Egypt.

PLATES 523

Plate 1
524 PLATES

Plate 2
PLATES 525

Plate 3
526 PLATES

Plate 4