Thracian Language and Greek and Thracian Epigraphy 1443813257 9781443813259 Compress

Thracian Language and Greek
and Thracian Epigraphy

Thracian Language and Greek
and Thracian Epigraphy
By
Peter A. Dimitrov
Thracian Language and Greek and Thracian Epigraphy, by Peter A. Dimitrov
This book first published 2009
Cambridge Scholars Publishing
12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
Copyright © 2009 by Peter A. Dimitrov
All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system,
or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or
otherwise, without the prior permission of the copyright owner.
ISBN (10): 1-4438-1325-7, ISBN (13): 978-1-4438-1325-9

To my children Vessela, Christina, and Georgi
TABLE OF CONTENTS
List of Illustrations ..................................................................................... ix
Preface and Acknowledgments................................................................... xi
Introduction .............................................................................................. xiii
Part I: The Evidence
Chapter One................................................................................................. 3
The Thracian Inscriptions
Chapter Two .............................................................................................. 21

Legends from Coins
Chapter Three ............................................................................................ 25

Evidence from Metal Vessels
Chapter Four.............................................................................................. 37
The Greek Inscriptions
Part II: Phonology and Morphology
Chapter Five ........................................................................................... 121

Introduction. Phonology
Chapter Six ............................................................................................. 129

The Components
Chapter Seven ......................................................................................... 139

Thracian Numerals
Chapter Eight .......................................................................................... 143

Analogy and Conclusion
Bibliography............................................................................................ 149
LIST OF ILLUSTRATIONS
Fig. 1-1. The Kyolmen Inscription
Fig. 1-2. The Ezerovo Ring
Fig. 1-3. The Gold Ring from Duvanlij
Fig. 2-1. and Fig.2-2. Coins of the Thracian Tribe of the Derroni
Fig. 3-1. The Silver Mug from the Bashova Mogila at Duvanlij
Fig. 3-2. Silver Phiale from the Rogozen Treasure, Inv. No. 22329
Fig. 3-3. Silver Phiale from the Mogilanskata Mogila Tumulus, Inv. No. B-68
Fig. 3-4. The Phiale from Kazanluk
Fig. 3-5. Gold Ring of Skythodokos
Fig. 3-6. Silver Phialae from Shipka
Fig. 3-7. The Silver Casserole
Fig. 4-1. Inscription from Kyustendil (Teres Garytinos)
Fig. 4-2. The Votive Plate from Kyustendil
Fig. 4-3. Inscription with Kozimases from Alexandrovo
Fig. 4-4. The Golden Plate from Sinemorets
Fig. 4-5. Graffito from Sboryanovo

PREFACE AND ACKNOWLEDGEMENTS
I owe the reader a short explanatory note.

First and foremost we should state that the material under study comes
from direct sources such as inscriptions on stone, clay, and metalwork.
Therefore, before starting to read the present study, we should keep in
mind that we will be facing a situation that is extremely complex. There
exists a methodological problem, originating in the past, which has caused
various misunderstandings. It is due to the volume of different entries
assembled in the goal to compile a thesaurus of the Thracian language.
Somehow, during the last two centuries, there was a whole set of methods
applied that were not in accordance with the progress made by linguistics.
For example, the choice made in assembling the two main corpora so far,
that of Tomaschek and Detschew, which present data from literary and
epigraphic sources. These data combined were not at all times convincing.
Sometimes controversial entries were included the interpretation of which
provoked long discussions. More attention was paid to details, which in
most cases were not concerned with the discussion of the whole body of
evidence.
One other point: whilst modern linguistics made a huge progress,
Thracian scholars stayed within the general Indo-European theory of the
Neogrammarians.
The method I used (as I explain in detail below) rests on the
description of Thracian onomastics obtained after phonological analysis,
because I am concerned with the fact that every single phonologically
attested form of phonemes and morphs is relevant. For, it helps to list all
possible forms of names thus showing all of the graphemes independently.
It was necessary to divide the material under study in sections in Part
I. The Evidence. For purely practical reasons, it was deemed useful (and
hereafter it will become clear from the presentation of the different
sources) to repeat at some places some entries as cross-references, as this
gives the study depth and facilitates the reader’s understanding of the
intricacy of the way the Thracian material has come to our knowledge. It
also provides to a larger extent the natural environment that fostered the
creation of such various forms.
An asterisk (*) is used before an entry in Chapter Four for
reconstructions of place names based on gods’ epithets.
xii Thracian Language and Greek and Thracian Epigraphy
In epigraphic texts the Leiden system of conventions is used.

Last but not least, it is worth reminding that any linguistic situation is a
part of the language development that has been termed “language change”
for language tends to loom ahead and we only later become aware of its
outcome.
Once, a modest Swiss scholar was able to realize that besides the
“history” of a language there is also a synchrony. While the historical
record of facts of a language is a long and complicated row of events
dutifully described and satisfactorily accounted for, synchrony is that
missing link that makes us witness the birth of the record yielding to its
origin. The name of the above mentioned scholar is Ferdinand de Saussure
from the town of Geneva whose doctrine of phonological system allowed
for the next generations to speak about linguistics as a science.
My debts of gratitude are to all those who helped me with their remarks,
observations, corrections, and guidelines. Above all, I am deeply grateful
to the late Professor Georgi Mihailov under whom I started my doctoral
dissertation back in the years. Also, I would like to thank many colleagues
and friends; without their support, advices, and encouragements I would
have long abandoned this project. I would like to acknowledge the expert
help and advices of Dr. Maya Vassileva, Senior Researcher at the
Bulgarian Academy of Sciences. My thanks also go to Dr. Thomas
Corsten, Professor at the University of Heidelberg, Professor Stephen V.
Tracy, former Director of the American School of Classical Studies at
Athens, and Dr. Jaime Curbera. Generous financial support from the
Mellon Foundation enabled me to spend three months as Senior Mellon
Research Fellow at the American School of Classical Studies at Athens
benefiting greatly of its research facilities. Ms. Davidoff and Ms. Todoroff
were instrumental in the technical aspects of the preparation of this book
for print and deserve acknowledgement for their endurance.
INTRODUCTION
Definitions
The present study of the Thracian language is based on an updated

collection of epigraphic material found in Bulgaria.
Our knowledge of the Thracian language comes from evidence that has
been assembled by modern scholars throughout the last two centuries from
two types of documents: literary documents derived from Greek and Latin
writers and inscriptions. Owing to the fact that epigraphy has furnished
less information, many scholars undertook investigations that were not
based but on one principle, to find any form of indication leading to any
possible information on the Thracian. Methodologically speaking, this
kind of “thesaurus at any rate” has produced only more confusion extant in
many writings to the present day. The multiplication of errors and
misunderstandings dating from Antiquity to modern times is responsible
for “rapprochements” of personal, god- and place-names in the vast area of
the Mediterranean region using onomastic material, glosses or
commentaries of ancient grammarians and lexicographers irrespectively. 1
Therefore, the updated collection of epigraphic material found in
Bulgaria is a kind of evidence that comes to reinforce our knowledge of
Thracian names and Thracian words and phrases sporadically discovered
and unveiled after years of scholarly work done by archaeologists,
historians, and linguists.
My goal as an author was to assemble the information from whatever
direct epigraphic sources offered it, i.e. coins’ legends or graffiti drawn on
precious metalwork (mainly phialae), and inscriptions on stones.2
In his Le dialecte éolien d’Asie (p.11-12), Hodot is differentiating the
two possible sources of his study by pointing out that the literary tradition
of the Lyrics rests on fragments and glosses that are “on le sait
indissociables” as far as the literary works pass through the intermediary
of grammarians, “les érudits de l’Empire”, and on the other hand the
papyri “conservant l’essentiel de l’oeuvre des Lyriques dependent de la
1
See Fol, 2000, 65-66.
2
For a thorough investigation based on direct documents see Hodot, 1990.
xiv Thracian Language and Greek and Thracian Epigraphy
“vulgate alexandrine”. It is imperative to underline what Hooker (1977,

14) said of Apollonius Dyscolus’ reporting on Aioleis: “By Aioleis he
means not ‘Aeolians’ as that term would be understood by a modern
student of the Greek dialects, but “the Lesbian poets as displayed in the
Alexandrian vulgate”.
The need of such an update is all the more urgent given the growing
interest in Thracian studies and language in particular, on the one hand,
and the abundance of misunderstandings and variae lectiones within the
context of the material used, on the other hand. In many articles and
books, scholars have been referring to numerous forms of names or words
without significant result due to the fact that a large number of newly
published materials is missing in D. Detschew’s handbook Die
Thrakischen Sprachreste, published in 1957. However, this is the one still
indispensable handbook that each and every student of Thracian is
counting on3.
Yet, over the years it became clear that corrections alone would not
suffice, and that a thorough investigation into the story of the appearance
of any evidence was necessary to prove that unique “ethnologische
Untersuchung über die alten Thraker” that W. Tomaschek offered to the
public first in 1893 and then in 18944. Actually, it is the cultural value of
the onomastic material that made the scholarly interest rise up and
eventually provoked the re-evaluation of the items studied by Tomaschek
and Detschew.
Thus, of many speculations on the etymology of a name or word,
nothing proved to be more important than the secure reading, the authentic
Thracian sound that can be found only in inscriptions5.
The one feature that makes this study different from those undertaken by
Tomaschek and Detschew emerges from that very secure reading, for it
gives us, along with many other important hints and insights through the
phonology of the Thracian, the historical perspective and the relative
chronology of a given language development. To this effect, Hodot
comments (1990, 12) that
“les documents du second type sont eux des documents directs: ce
sont les inscriptions gravées sur la pierre, incisées ou peintes sur
3
See K. Vlahov 1963, in an attempt to add up new entries to Detschew’s
handbook, and also I. I. Russu’s (1964) review.
4
W. Tomaschek, Die alten Thraker, Sitzungsberichte d. Akad. Wien, Bd. 128
(1893): I. Uebersicht der Stämme; Bd. 130 (1893): II. Sprachreste, 1. Glossen aller
Art und Götternamen, Bd. 131 (1894): 2. Personnenamen und Ortsnamen.
5
For an interpretation of the method see Hodot 1990, 12: the literary tradition on
Lesbian is “un témoignage médiatisé”.
Thracian Language and Greek and Thracian Epigraphy xv
la céramique, frappées sur les monnaies. Il est inutile de souligner

quel avantage sur les textes littéraires ou grammaticaux peuvent
présenter ces échantillons du dialecte qui nous viennent tout droit
des usagers eux-mêmes. Toutefois, pour que ces témoignages
épigraphiques et numismatiques soient pleinement utilisables, il
faut qu’ils puissent être datés, au moins approximativement, et
surtout que ne se crée pas á leur propos une nouvelle sorte de
‘tradition’, non plus manuscrite, mais imprimée, qui aboutirait au
même résultat : déformer l’aspect originel du dialecte”. 6
Yet another major difference lies in the effectiveness and clarity of the
data accumulated; whereas in the past, criteria for words of Thracian
language origin were based upon their being referred to as such by ancient
Greek and Latin authors, in this study they must qualify and be listed as
Thracian only if the inscriptions themselves indicate it.
In many cases, a predicament emerges as to what to opt for, and the
scholar is guided not just by explicit indications, such as e.g. the nation of
a Thracian mentioned in the inscription but also by the already established
tradition relying on the authority of scholars who have worked in the
Thracian field. I found the principles discussed in Beschevliew’s book
(1970) Untersuchungen ueber die Personennamen bei den Thrakern very
useful to this effect. The author discusses the subject in greatest detail
reviewing methodologies and approaches used by scholars in recent times.
Moreover, in this study the geographical distribution of the names is
limited to those territories, which are in the fringes of the Greek world.
The organization of the lemmata is another element differentiating the
present study. Etymological explanations have been included in few cases
and only to help understand cultural or historical value. In many cases,
out-of-date readings and variants were discarded from the list and about
four hundred new names have been introduced. Thus, relying on a
coherent body of entries, we were able to identify secure leads as to the
structure of the roots in Thracian. Whereas earlier ideas were based on
sometimes contradictory evidence, the disposition of the sequences is now
well documented through more than one occurrence. We can be certain
now (based on the entire collection), that stops rendered through or
do not really exist in Thracian, do not appear in the beginning of the
word, nor are they present in any other position. The few examples
containing or are due to foreign word-editing and are a good
example that some inscriptions were made by Greek nationals. One can
6
For the opportunities to follow the changes and the development of the Thracian
language see also Fol 2000, 66.
xvi Thracian Language and Greek and Thracian Epigraphy
compare Rolistene in one of the earliest original Thracian inscriptions on

the gold ring from Ezerovo: -stene is in fact a Greek loan word (= sthenes
“strong”). From forms such as -theithes, etc., we might assume that
regressive assimilations in foreign loan-words from Asia Minor took effect
at some point under the Late Roman Empire as the Thracians experienced
a strong Roman influence.
Again through that extensive evidence, we can be sure that the spirant s
was rendered with the Greek letter theta, except between vowels where z
was used.
The inscriptions from Rogozen provided a secure lead to the
mechanism of vowel prothesis in Thracian through ,
and .
There are many names that appear only in inscriptions, in their
“unedited” form, e.g. , , .7
The Thracians: A Brief Historical Overview

Thracian studies seem to have produced a solid record of topics
exhaustively examined by scholars over the past one hundred and twenty
years or so. Yet, the chronology that archaeology has produced through
adjustments taken from various sites as well as later interpretations prove
the existence of many points of indecision which need to be reformulated
with precision in describing the inert model of Thracian reality.
In this section, our concern is to provide essential information about
how it all began. Therefore, I am offering only a brief overview. In no
respect is it an exhaustive piece of history of Thrace, as some might
expect.
As tradition has it, the Thracians lived in numerous tribes spread over
the islands of the Aegean and up north to the Carpathian Mountains. The
islands of Euboia, Lemnos, Samothrace, and Thasos trace back a path of
historical and mythical memory of Thracian culture. Beyond the
mythology of movement of migrant populations and historical evidence,
we find Thracians settled in the lands north of the Mediterranean
according to the earliest literary and epigraphic evidence. Who were the
Thracians then? A simplistic and inaccurate answer to this legitimate
question would be to leave them unnamed, with indeterminate Indo-
European origin and culture. Yet, their language is termed “Indo-
7
We shall refer to these and many other examples in Chapter Five.
Thracian Language and Greek and Thracian Epigraphy xvii
European” by criteria that handbooks prefer not to discuss at length. The

several old scripts on stone or precious metalwork are referred to as
“inscriptions…, neither of which we understand”.8
Apart from the fact that in modern scholarship less attention is paid to
the Thracians and their history, politics has come to play a rôle by no
means indecisive in delaying and putting the investigation off. 9
Methodology was seriously influenced by approaches applied to standard
western-like colonialism.10
The Early History of the Thracian lands is derived from literary texts 11;
Homer provides the earliest and most abundant evidence.12 While
Hesiod’s scanty passages, along with those of the lyric poets Archilochus 13
and Alcaeus14, are imaginative and hopeful, the important descriptions by
Herodotus and Thucydides show the Thracians as numerous people about
quite a territory, and with established cultural traditions, warriors, allies,
inhabiting and dwelling in plains and mountains, along rivers and lakes. 15
We are also aware of the presence of newcomers around the end of the
Eneolithic Age and on the eve of the Bronze Age. 16 They occupied the
8
Beekes 1995, 23; 331.
9
Archibald 1998, 3: “Modern political divisions, geographical and ideological,
have almost to the present day conspired to prolong the isolation.”
10
Tsetskhladze 2006, 26-27: “books …link ancient and modern colonization and
‘colonialism’(s)”; Owen 2000, 139: “Greek colonization of Thasos, and indeed of
Thrace, is currently written from a wholly Hellenocentric and text-based
perspective, behind which lies an unspoken and pervasive comparison with
Western European colonolialism”. See more opinions in detail in Tsetskhladze
2006, 23-83. To this effct, see also Owen 2005, 5-7 etc.
11
See the comprehensive account of Mihailov 1991, 591-618, as well as Fol 2000.
12
Mihailov 2007, 9-42 [Homère comme source historique des états thraces].
13
Frg 79a D(iehl): κἀν Σαλμυδ[ησσ]ῶι γυμνὸν εὐφρονέστ[ατα
Θρήϊκες ἀκρό[κ]ομοι; Frg. 51 D(iehl): “ἄνδρας . . . ωλεῦντας αὐλὸν καὶ λύρην
ἀνήγαγεν ἐς Θάσον κυσὶ Θρέϊξιν δῶρ᾽ ἔχων ἀκήρατον χρυσόν, οἰκείωι δὲ κέρδει
ξύν᾽ ἐποίησαν κακά... ”; see also Homer, B 536-8, where he mentions the Abantes
in possession of Euboia with Halkis, Erethria, and Histiaia.
14
See frg.77 (Diehl, E. Anth. Lyr. Greac., I, Lepzig, 1936).
15
Archibald 1998, 3, using modern geopolitical terms, stresses on the importance
of the Balkan Peninsula whose “regions were closely linked to the eastern
Mediterranean through the periodic movement of people and through networks of
friendship, alliance, and exchange”.
16
Fol 2000, 63: newcomers appear at that time and it is not only the Varna
necropolis that is a direct record of them but archaeological finds from other sites,
as well.
xviii Thracian Language and Greek and Thracian Epigraphy
territories south of the Carpathians to the fringes of the Greek world.17 In

the 6th century BC we find the Thracians, under different tribe names and
in various places, settled in the territories that stretch roughly between the
ancient Euxeinos Pontos (Black Sea) to the plains of Axios (Vardar River)
and north of the Aegean up to the Transdanubian lands18. The Greek
historians noticed the political activity of the strongest Thracian kingdom,
i.e. that of the Odrysians. Approximately at that same time, the earliest
original Thracian inscriptions appeared. By the fourth century BC the
contacts and interactions with the Greek world (and language) were very
intensive.19 This was the time of the Odrysian political prime under king
Kotys I (383-359 BC), the period during which the most spectacular
Thracian tombs were built, as well as the most numerous inscriptions and
graffiti on metal vessels were produced.20 The Thracian lands were
involved in the politics of the Hellenistic world, while the armies of
Alexander the Great and his generals brought many Thracian mercenaries
as far away as Asia and Egypt. Some local peculiarities in the use of the
koiné can be traced. Thracians served as soldiers in almost all provinces of
the Roman Empire, after the Thracian lands had been incorporated in the
Roman Empire and became provinces: Moesia (later Superior and
Inferior) and Thracia in the first century AD.21 Despite the spreading of
Latin as an administrative language, Greek continued to be widely used. A
major part of the corpus of the Greek inscriptions found in Bulgaria
originated from Roman Imperial times, thus offering opportunities to leads
to Thracian language development. A large number of the 1 st - 3rd centuries
AD votive inscriptions are dedicated to the Thracian Heros (Horseman)
whose local epithets provide rich material for the present study.
Perhaps we should start the discussion by laying down the evidence
and leave the dubious interpretations aside for now. Because of gaps here
and there, it is hard to determine which evidence is to be considered most
relevant for that period.
I would like to stress upon the rare occurrence of artifacts in tells being
dated to the Late Bronze Age compared to those coming from necropoleis
which are more numerous.
17
Fol 2000, 63: “These are located in the lands to the south of the Carpathians, on
the Haemus Mountain and the Rhodopes, as well as in Hellas itself.”
18
Mihailov 19912, 591.
19
“In the period reaching the end of the 4th century BC, these relations led to
recorded translations”: Fol 2000, 67.
20
Archibald 1998, passim; Marazov 1998, passim; Fol 2000, 67.
21
Velkov and Fol 1978, 46-52.
Thracian Language and Greek and Thracian Epigraphy xix
Scholars are advancing the idea of a continuous habitation in nearly all

of Thrace without breaks for the Late Bronze Age going into Early Iron
Age (late second/early first millennium BC). 22 They speak of Thracians
being present in the region all the way from as early as the Late Bronze
Age.23
To place the Thracians in the context of the Eastern Mediterranean, we
will follow the new lead offered by the study of W. Burkert. 24 Basically,
this is a re-appraisal of the ‘provincial seclusion’ of civilizations in the
Early Iron Age.25 With more publications supporting the idea of the
relations that had existed between ancient Greeks and the East in Anatolia,
the discovery of the Hittite language and culture and later on the
decipherment of the Bronze Age Linear B tablets, it became clear that a
somehow mixed Bronze Age culture, a Bronze Age koine, had been in
existence26.
Already in the Bronze Age Greeks had moved beyond the Balkan
Peninsula on to the Aegean islands, Crete and Cyprus, along the coast of
Asia Minor, and even southern Italy and Sicily.27 The civilizations of the
East in Egypt and Mesopotamia, and westward Palestine, Syria, and
Anatolia were by far more advanced and interaction with them in the
second millennium, during the Late Helladic period, took place at a
different pace.
Roughly by the year 1200 BC all of these civilizations were affected
by catastrophic waves of fires, floods and destructions and succumbed to
obscurity.
In the East, the important Phoenician cities of Tyre and Sidon, with
south of them the Philistines, and up north the New Hittite Kingdom and
the Phrygians, seem to have no recollection of bloody and aggressive
battles along their way of establishing themselves in the fringes of the
collapsed world of the Hittites. There is an apparent cycle of cataclysms
happening: cultural interactions fluctuated in a way we can glean from
literary and archaeological sources.
Back to Thrace, the set of gold vessels found at Vulchitrun (northwest
Bulgaria)28 dated to the mid-second millennium BC confirm the assertion
of continuity by adding data that speak of lavish signs of rulers in the
22
Archibald 1998, 32-34; Georgieva 2001, 84.
23
Fol 2000, 63.
24
Burkert 2004, 2: “Classical”… does not connote isolation.”
25
See Burkert 2004, 3 and Dornseiff 19592, 30.
26
Burkert 2004, 5: “There were contacts and interactions on all sides”.
27
Burkert 2004, 5.
28
Venedikov 1987.
xx Thracian Language and Greek and Thracian Epigraphy
Thracian lands. The tripartite vessel is considered a unique one with

evidencing special rites for mixing three different liquids.29 Mycenaean
gold vessels and vessels from Alaca Höyük are said to exhibit comparable
artistic traits thus paralleling the Vulchitran particularities. 30
The Northern Aegean appears to have been related to Mycenean
developments in the sense that the Hellenes of the second millennium
expanding toward the north might well have had visited the lands of the
ancient Thracians. Recent archaeological finds provide the evidence for
such an assumption as far as Linear A clay seals31 and a clay spool32 are
concerned. Mycenean ceramic sherds have been unearthed in Koprivlen
(southwest Bulgaria) near Gotse Delchev (previously known as
Nevrokop).33 Furthermore, slabs engraved with spiral-like decorative
motifs, were paralleled to gold objects found at Mycenae; a scene on one
of them representing a male body with a sun-boat34 conveys an eastern
Mediterranean theme.
Recent discoveries offer more evidence on the 11th - 9th century BC life
and culture in the Thracian lands.35 Most of this evidence comes from
burial mounds, while metalwork contributed to its chronological
specification. Early Iron Age finds display local variants: those would
differ in the Dobrudzha region (northeast Bulgaria), in Babadag further up
in the northeast, along the Black Sea coast, Insula Banului to the west and
Chatalka and Pshenichevo to the south.36 Archaeological record speaks in
favor of a continuous life over four centuries, from the 10 th to the 7th BC.
29
Venedikov 1987, 97-98 sees a similarity with Eleusinian mysteries” kykeon, a
potion, that is known ever since Homer’s Od. 10. 290, 316.
30
Venedikov 1987, 60-62, 79-84; Sherratt and Taylor 1989, 121, fig. 4.
31
Matsas 1995, 242.
32
Fol and Schmitt 2000, 56-62.
33
Alexandrov 2005, 47-49: Six fragments of imported Mycenaean pottery were
found comparable to that from Mycenaean centres and dated to the early/middle
Late Helladic IIIB.
34
In the vicinity of Razlog (southwestern Bulgaria), seven slabs and fragments of
at least ten more were found. See recently S. Ganeva’s article with bibliography of
earlier accounts: Ganeva 2005, 147-51.
35
Chichikova 1972, 1977, who started publishing Early Iron Age archaeological
materials (mainly pottery), was the first establishing Iron Age chronology for
Thrace; 1990; Stoyanov 1997, 82 published the newly excavated site of
Sboryanovo; Archibald 1998, 26-47 discusses chronological problems; see also
Georgieva 2001, 83-94 and Gotsev 1994, 43-68. Gergova 1989, 231-40 pointed
out to some specific traits of burial rites traced down in the transitional period
through Late Bronze Age.
36
Stoyanov 1998, 164.
Thracian Language and Greek and Thracian Epigraphy xxi
Considering the Thracians in Anatolia, K. Sams discussed the “open

lines of communication between Phrygia and Europe” and a “cultural
corridor”.37 Most probably as a tentative time mark of the Thracians’ south
migration and southeastward expansion over the Bosporos into Asia Minor
one could set the very end of the second millennium BC. The Gordion
archaeological record at least speaks of newcomers from different areas in
Thrace judging on EIA handmade pottery.38
Colonization
Because of its huge cultural impact, the Greek colonization waives of
various “comings” of Greek settlers, merchants, etc. into the lands
surrounding the sea, we have a enormous source of information indirectly
or more straightforwardly leading to the early presence of the Thracians
and their practices.
In terms of “earliest possible”, we need to better understand why we
should abandon the model of violent conquest and subsequent
“asymmetrical power relationships”39. There seems to be unanimity
among scholars40 that in those remote times, the Thracians were not just
passive observers of the ‘coming of the Hellenes’ up north; they were
partners in creating a new environment for socio-cultural and economic
development in the Aegean. It is very true, that the earliest literary sources
in Greek poetry about the inhabitants known ever since as the Thracians
were not exactly positive.41
Contacts between Greeks and Thracians were not necessarily on a
hostile note for the entire period of their early interactions. The verse of
Archilochos42 is to be interpreted in the light of modern scholars’
37
Sams 1994, 21 and 1995, 1147. See also: Vassileva 2005, 227-34 with
bibliography.
38
Sams 1995, 1147-59, 1994, 21; 176.
39
Owen 2005, 6: “the assumption that asymmetrical power relationship, drawn
along ethnic lines, existed in all ‘colonized’ areas from the Late Geometric and
Archaic periods onwards is one which still pervades much of the literature”; and
Owen 2005, 18.
40
See Tsetskhladze 1999 and 2006.
41
See Vassileva 2005, 227-34 with bibliography.
42
Frg 79a D(iehl): κἀν Σαλμυδ[ησσ]ῶι γυμνὸν εὐφρονέστ[ατα
Θρήϊκες ἀκρό[κ]ομοι; Frg. 51 D(iehl): “ἄνδρας . . . ωλεῦντας αὐλὸν καὶ λύρην
ἀνήγαγεν ἐς Θάσον κυσὶ Θρέϊξιν δῶρ᾽ ἔχων ἀκήρατον χρυσόν, οἰκείωι δὲ κέρδει
ξύν᾽ ἐποίησαν κακά... ”;
xxii Thracian Language and Greek and Thracian Epigraphy
perceptions about that as a major event in the Mediterranean world which

customarily ancient historians use to describe as “Greek colonization”. 43
There is a clear contamination as to who, when, and how participated
in the process of Greeks undertaking those periplooi.44 Thus, Archilochos
was misinterpreted45 for being studied from too narrow a point.
As I tried to explain elsewhere46 the third century BC Hellenistic poet
Apollonios of Rhodes worked on a project to describe (and revive) the
deeds of the glorious wanderers of the past in his Argonautika by way of
literary (poetic) techniques, where a substantial part is devoted to the
ethnography and geography of islands of the Aegean and especially that of
Lemnos. We have a clear patch of remote historic events related skillfully
enough to let us think of a movement, slow as it might have been, towards
the northern territories. The Lemnian women led by Hypsipyle 47 imply a
vision of developments going way back to the time of courageous people
sailing the wine-dark sea on the northern approach. There is even a good
and sound story as to the many generations of Thracian families, depicting
the way Thracians gradually started leaving Lemnos in order to gain
territories on the “Thracian coast” opposite of Lemnos. They left their
Lemnian wives with their sons on the island. This has been done in several
attempts to obviously colonize the “lands on the Thracain coast”. Every
time the Lemnian women were observing ships coming to the island from
their pyrgoi, they were uncertain as to who was coming, deprived of their
husbands at that time.
Apollonios of Rhodes seems to be very well informed as regards the
geography of those events, describing what had been the most eye-
catching mounts, their peaks, the coastal line along what is now
Chalkidiki, as well as the islands of Lemnos and Samothrace. Was it just
because in his time this was a socially calm and prosperous portion of the
oikoumene, and therefore he felt at ease to describe it, or rather this is how
it looked in the sources he used?
My intention, accordingly, is to include these problems, and relate
them to other theoretical approaches in the study of Thracian antiquities, in
43
Here see Tsetskhladze 1999, and 2006; Owen 2005.
44
Tsetskhladze and De Angelis 1994.
45
See n.42 above. Owen 2005, 19: The author is of the opinion that a fragment of
an inscription where Archilochos called the Thracians “dogs” was not correctly
restored (after a 1930s scholar), and that it has no justification; it has rather a
modern sounding.
46
Dimitrov 2006a.
47
Hypsipyle is not recorded in inscriptions. It is the name “of the one woman who
dwelled the high gate” judging on the meaning of the name in Greek.
Thracian Language and Greek and Thracian Epigraphy xxiii
which remnants of onomastics are interwoven in a remarkably complex

way.
A few words are needed to set the background of Thracian linguistics.
We dispose of no other resource but the inscriptions. Again, for practical
reasons inscriptions in the Greek language are considered in establishing
the inventory of Thracian onomastics. Latin forms of names coming from
Latin inscriptions are more of a supportive nature for two reasons: for
being late or because of the fact that Thracian settlements were following a
Greek tradition throughout the Roman imperial period, keeping Greek and
less so Latin in their everyday administrative activities, with only a few
exceptions.
The Greek inscriptions have a history of their own in Thrace proper.
Their geography may well be divided in three major parts: (1) Thrace as
part of present-day Bulgaria (2) Thrace as part of present-day Greece (and
all of the Greek territories in Antiquity), and (3) Thrace in present-day
Turkey along with Asia Minor where traces of Thracians were found in
inscriptions. This division alone makes for the innumerable difficulties in
presenting the material.48 The legends on a limited number of coins were
used in this study. Their complexity requires a separate way of
investigation, and the evidence they provide was utilized with caution. 49
With many inscriptions found on metalwork, it became even more
difficult to keep up with the line of equally treating all inscriptions and
positive data at yield. For epigraphy developed its own way50 and
therefore facts are hardly to be interpreted from a single point of view. For
example, we cannot be sure that the value (in terms of alphabet reform,
editing, and lettering) of an inscription found on the Athenian agora could
be attributed automatically to an inscription found in Thrace. The situation
is comparable to that found in similar fields, such as history of religion51.
48
Detschew’s Die thrakischen Sprachreste is a good example of the diversity of
facts that at times do not help to solve the problem. Doing away with some of them
is and always will be a problem of methodology in approaching the material.
49
See Dimitrov and Penchev 1984.
50
See Bodel 2001, 2, 10-15.
51
See to that effect among many other problems the one outlined by Archibald
(1999, 431): “Archaeology can be expected to play a far more significant role in
the interpretation of religious behavior, not simply because written sources are
meager but because it provides evidence which no ancient writer was capable of
giving.” And further on, p. 435, “our problem is how to define what was distinctive
about the religious behavior of Thrace and how mutual interactions with other
traditions affected this distinctiveness.”
xxiv Thracian Language and Greek and Thracian Epigraphy
Phonology and Phonetics of the Thracian names

based on the inscriptions found in Bulgaria
The scanty evidence from the chronological layers does not yield to
working with a wealth of examples, nor is it any easier to describe in an
exact way the clusters in their succession one upon another through the
time of transformations of the Thracian language.
The outline of the theory of the phonological level in the name system
is very important for practical reasons. The inscriptions on stone and
metal, or graffiti scratched on clay vessels, have increased in number,
presenting unconditionally precise records for the history of the Thracian
territories.
The history of research conducted and dedicated to the Thracian names
so far, speaks for itself, supporting the assertion that the theory of their
phonological level is essential to the understanding of the whole process.
The system of the names at its phonological level is extremely interesting
by its various developments, the latter being of high historical importance
in solving a range of dubious problems. Owing to the fact that the
phonemes in Thracian had not been the object of special interest for the
19th Century scholars, an impression was formed in modern scholarship
leading to the conclusion that the whole system is hypothetical and
therefore impossible.
However it may be, it has been proved wrong by Dimiter Detschew,
who undertook a passionate journey into the phonological and morphemic
structure of combinations beyond the etymological explanation, within the
general theory of the Indo-European phonology. In fact, this is the first
attempt in the history of Thracian language studies.
Due to the etymological approach, which later became a principle of
methodology, we have many valuable observations and achievements at
hand. Due to that same approach, there are also theoretical
misunderstandings, which hindered the ongoing investigations into the
system of the Thracian language. Having undergone research mainly by
Bulgarian scholars, the Thracian language has also become the object of
interest of foreign scholars, in regard to the study of the so-called
“peripheral cultures” within the Hellenic world, such as the Asia Minor
regions of Bithynia, and especially Phrygia. 52
52
See for example Brixhe and Panayotou 1994, Corsten 1990 and 2007, Gaertner
2001, etc.
Thracian Language and Greek and Thracian Epigraphy xxv
There are several investigations, for which I have substantiated the

necessity of theoretical outline, and have showed the lack of systematical
presentations of the Thracian language wealth. This problem is connected
to another range of developments, which only seem external, but in fact
are the core of the same phenomenon: Thracian names are present in less
than ten inscriptions in Thracian language, while thousands of them were
found in inscriptions written in Greek and Latin, the latter being used in
this study as supporting evidence because of their later origin. For
centuries Greek had been the language of civilization in the Hellenic
world, as well as in the adjacent transitional lands. Greek has been the
official language of the Thracian kings and dynasts’ courts and
administration until the time Latin took over. The several genuine
Thracian inscriptions were written in characters of various Greek-like
alphabets. Judging on the script used in the Kyolmen inscription we could
imagine an epichoric variant, not necessarily derived from a Greek
prototype.53
The method of presenting the material, adopted here, is based on the
approach of the material used. It consists of studying the chronological
layers of evidence. In the notion of chronology as part of the notion of
phoneme (providing the phonemic variants and therefore the history of the
language), the study is centered on the history of the linguistic
transformation as evidenced in the Greek inscriptions found in Bulgaria,
reflected in the history of the phonology of the Thracian names, as well as
the rich cultural environment. The information that comes along with each
text enhances the understanding of the layers. The latter, in the light of the
interdisciplinary analysis, open new and unexpected perspectives of
describing the culture that once flourished in these vast territories.
Secondly, with using this approach in order to study the phonology, i.e.
the theoretical value in the nature of the Thracian phonemes, only the
epigraphic evidence is being employed, and mainly that from Bulgaria.54
Furthermore, the method of our investigation lies in the philosophy of
the principles adopted: secure data with examples that can explain the
developments under study by reaching a solution.
The basis of this method is trying to comprehend the phonology of the
Thracian language. We take into consideration the concept of the
53
Inscribing or carving letters on metalwork betrays an Iranian parallel as
evidenced in Thrace and Persia. See to this effect Vassileva 1992-1993 and
Zournatsi 2000.
54
It should be noted here that on several occasions occurrences found in Greece,
Anatolia, and elsewhere around Thrace proper, were used as parallels to Thracian
forms.
xxvi Thracian Language and Greek and Thracian Epigraphy
phonological rule, for rules may change under certain conditions just like
the language changes are made functional within a system.
The method itself does not automatically guarantee results. 55 A number
of procedures might have been lost due to impossible reconstructions
within the Thracian language, e.g. lack of sufficient cases or insecure
readings.
In the categories of Thracian sounds, the interrelationships between
various classes56 have been proposed and later perceived as theoretical
entities that may change according to “sound laws” 57, an equivalence that
should be the same for e.g. Ezbenis and Asbenoi. In their analysis, the
fluctuation e/a is to be referred to the way these “sounds” were
pronounced (closed or open pronunciation); the shift of s>z is a later
development58, or a feature that is not marked by any specific conditions 59,
or the shift was conditioned according to its word-initial or intervocalic
position60. It is true that the intervocalic S normally changes to Z between
vowels, however in our example we observe the same opposition between
/s/ and /z/. If we take its chronology into consideration, S and Z are
synchronic (as they appear in our Evidence) and therefore this opposition
is irrelevant in regard to their morphophonemic involvement. There is a
piece missing in this easy-to-solve puzzle. And it is namely that we are not
dealing with sounds but rather with those ”unreal sounds” or
abstractions.61
The underlying PIE */w/ and its treatment in Thracian through the
Greek beta conditioned the shift. A plausible reason could be the word
initial varying between /a/ and /e/.62 Hence, the above-mentioned classes
may be distinguished for subclasses, called allophones,63 which can be
analyzed for distinctive features; sonority is the one in our case. This very
feature is crucial to the understanding of the phenomenon, for this “sound
55
Beekes 1995, 103.
56
See Lehmann 1993, 8 for a brief but succinct presentation of the phonological
theory.
57
See Dečev 1960; Georgiev 1983; Duridanov 1985.
58
Duridanov 1985, 108.
59
Dečev 1960, 162-63: “Es folgt daraus, dass im Thrakischen das ide. s teilweise
unverändert bleibt, teilweise zu z wird”.
60
Georgiev 1983, 1173-74: “Ide. s ist im Thrakischen erhalten geblieben…. Im
Anlaut vor Vokal und intervocalisch wird s oft zu z (wie im Deutschen)”.
61
Lehmann 1993, 12: “besides the perceived, articulated sounds, classes were
proposed that were labeled ‘phonemes’. . . The classes… are abstractions”.
62
See Dimitrov 1994.
63
Lehmann 1993.
Thracian Language and Greek and Thracian Epigraphy xxvii
change” that we identify as a principle has long been explained, and leads
us to believe that generally one specific trait per se is involved.64
No doubt, sounds cannot be phonemes for they belong to a different
category. Their functional analysis is also called phonetics.65 Only on the
surface could one register the phonetic units, as they belong to the
physiology of the articulation.
We now arrive at the subsequent conclusion: first, there is no /z/ in
Thracian as a continuant of PIE */z/ as the latter simply does not exist.
Second, /z/ in synchrony is just an allophone, a variant of the phoneme
/s/66. Third, there may be another condition involved, e. g. assimilation e-e.
Fourth, only the phonemic analysis with the appropriate distribution of the
phonemes can lead us through establishing the etymology of this Thracian
etymon of e/asba from PIE * h1ek’w-o-s. Fifth, this conclusion would not
be possible, if we were to disregard the information from our direct source
and the method of the phonological (phonemic) analysis. 67
With those considerations in mind, I welcome you to Part One.
64
Lehmann 1993, 78; Lehmann 1952, 3.
65
So Anttila 1989, 207-8: “Phonetics in this framework was called functional, and
an enormous amount of work has been directed toward the principles and
procedures for arriving at this level. These principles are generally known as
phonemic analysis, and the functional phonetic surface units as phonemes.”
66
So Bonfante 1937, 127-29: In his critic of Jacobsohn’s (Festschrift Kretschmer
1926, 72 ff.) he examines a large group of words contrasting them against their
correspondent cognates in other Indo-European languages to finally make valuable
observations on the phonological and phonetic development in Thracian. Among
other, he pointed out to the way Thracian sounds were represented through Greek
script.
67
There is no mention of Asbenoi in Detschew, for this essential item was
unknown to the literary tradition.
PART I:
THE EVIDENCE
CHAPTER ONE
THE THRACIAN INSCRIPTIONS
I. The Kyolmen Stone Inscription1

The publication of this inscription triggered a rapid increase of
scholarly activity.2 Relevant archaeological sources as well as other
indications emerging from the linguistic analysis make it clear that a
number of problems arose ever since its first appearance in print. This
state of affairs has been summarized in the last major account on the
Thracian language made by Claude Brixhe and Anna Panayotou.3 The
authors are concerned with the idea of the identification of the signs used,
and with the reading of some characters.4 It is not clear whether they
suggest deciphering the inscription as a whole or reading and interpreting
it. No one so far has done both to decipher and interpret the text
successfully, nor has anyone considered the entire evidence.5
The Inscription
1. The tombstone-shaped sandstone stele consists of two parts. The
large piece measures 148x68x7-11.5 cm. A tractor broke off the two
smaller pieces during agricultural activities in the fields of the nearby
1
The inscription is dated to the 6th century BC (Archeological Museum of Sofia,
Inv. No. 6558).
2
See Woudhuizen 2000-2001 for an extensive list of almost all publications to
present.
3
Brixhe and Panayotou 1994. (This publication for some reason is unknown to F.
Woudhuizen) While they are not providing any interpretation of this particular
inscription, they are of the opinion that the previous publications on the inscription
failed to produce reliable facts and decisive evidence to solve the many problems
arising from the very nature of the find.
4
Brixhe and Panayotou 1994, 187.
5
See e.g. Georgiev 1966; Schmitt-Brandt 1967; Woudhuizen 2000-2001;
Ancillotti 1986; Schmid 1987; Theodossiev 1997, for approaches which Brixhe
referred to as “inteprétation parfois délirante” or “étymologies plus “acrobatiques”
(Brixhe and Panayotou, 187, 193).
4 The Thracian Inscriptions
village of Kyolmen that consequently led to rescue excavations. The actual

length of the two sections of the inscription is as follows: (1) 71cm; (2)
103 cm; (3) 34 cm.
Fig. 1-1. The Kyolmen Inscription
2. The above mentioned three sections of the inscription go from left to

right, if we consider the position of the cutter, (referred to hereafter as (1),
(2), and (3), see Figure 1-1); if facing the stone, (2) is situated on the left-
hand side along the edge of the stone reaching its upper edge, and then
turning to the right, i.e. horizontally where several more letters are carved;
(1) starts from the right-hand side of the stele passing through the center
and joining (2); and finally (3), being a little off the center goes vertically
down, where nine signs are carved from the center and down:
.
The Thracian Inscriptions 5
Ex autopsias, as I saw it in the Museum of Sofia, starting up from the

right side, the inscription reads as follows:
(1)
(2)
(3)
Archeology-related considerations
1. The inscription was found in grave No. 1 in the center of tumulus
No. 1.6 The description of the grave construction and the grave offerings
match the archaeological situation described by Filov for the Kukova and
Bashova tumuli in Duvanlij.7
2. The four burial mounds,8 situated at a distance from one another,
seem not to belong to one and the same necropolis.9 One or more slabs
placed across or along the long side, covered the tombs.10
3. Dremsizova-Nelchinova is of the opinion that the two graves belong
to members of the Thracian aristocracy, as the offerings found in situ
6
Dremsizova-Nelchinova 1972, 207-208: The construction of the tomb was made
in the following way: a rectangular pit, one meter deep was dug into hardpan, lined
up with flat pieces of sandstone, partly cracked. Cremated bones covered with a
clay bowl were found in the middle of the tomb. There were also 45 fragments of
scale armor, parts of spears, a bronze clasp broken into three pieces, several iron
rings from a harness, and most importantly, a handle of a black-figured lekythos
and amphorae handles and body sherds.
7
Duvanlij is a 5th century BC Thracian tumular necropolis: Filow 1934, 62 and
187.
8
Dremsizova-Nelchinova 1972, 217.
9
The first two tumuli were part of the necropolis of the settlement, now in the
place called Ungi, five-six hundred meters away from it. The tombs found in three
of the tumuli (no tombs in the second tumulus) are of the same kind: rectangular
tombs oriented north-south, dug out in the rock 0.8-1 m deep, with walls reveted
with sandstone slabs. Analogous tombs can be seen elsewhere in the villages of
Kuylevcha, Ravna, Yankovo, and in the districts of Lovech and Teteven. The
Kyolmen tombs stand out with their more elaborate construction. The burials were
performed either through cremation (tumuli 1, 2 or 4) or through resting the bodies
in the ground (inhumation).
10
In grave No. 1 a middle-aged man was buried (two of his teeth were found in the
tomb) together with his armor and some weapons. Grave No. 2 (from that same
tumulus No. 1) belonged to a woman judging by the ornaments (a gold button and
a lekythos). An assumption is made that this is the beloved one, buried together
with her dead husband, a well-known rite of the Getae. See Dremsizova-
Nelchinova 1972, 218.
(precious and imported rare objects of art and luxury were placed within)
indicate. The dates of the grave offerings range between the 5th and 3rd
century BC. However, the burial in tumulus No. 2, being the oldest, is
dated to the end of the 6th century BC. Based on this fact, she assumed
that the existence of a necropolis from as early as the 6th century BC
accounts for the use of the inscribed stone.11
4. The rock bearing the inscription is a gray-yellowish sandstone that is
composed of thin, easily separable layers. Totev is of a slightly different
opinion in describing the grave. At least four stones were used to cover the
grave.12 The longer side of the inscribed rock (being 148 cm long and 68-
74 cm in width by either ends) falls short by at least 50 cm, and therefore
could not have served as a cover slab.
5. There is sufficient archaeological evidence produced by Totev,13 and
Dremsizova-Nelchinova,14 that the stele in question was found face down
on the top of the tomb. Most recent publications follow Dremsizova-
Nelchinova’s opinion that the inscribed stone had been worked on and
then re-used.15
The Reading of the Inscription

1. The letters used in
these inscriptions are
(three variants). In the
Greek inscriptions from Bulgaria, different signs for Greek letters have
been found. The same is true for those written in the Thracian language. In
this inscription we have a few S-like characters. We consider them
graphemic variants in spite of what has been adopted in previous
publications. It goes without saying that the signs, besides their usual
identification belong with those of Greek alphabet, are also identified
according to the sense they might make in the course of the reading.
Sometimes the text itself solves many such problems. We will follow this
path in exploring the reading.
11
Dremsizova-Nelchinova 1972, 218-28, see especially the French summary: 228-
29.
12
Totev 1965, 10: the grave measures 1.94 x 1.84 x 1.18 x 1.07 m, 1 m deep. The
slabs were placed across over the longer sides, thus oriented east to west, while the
grave itself was north-south oriented.
13
Totev 1965, 9-11.
14
Dremsizova-Nelchinova 1972, 207-29.
15
Dremsizova-Nelchinova 1972, 228.
2. The execution of the inscription betrays many turning points and

hesitations, thus allowing to read and separate the words. However, a few
points of clarification are necessary. Following the general discussion,
which is at times very laconic, we will establish the points scholars agree
upon:
1) As the evidence shows, the letters are of different size, between
2.5 and 6 cm high, and feature early 6th to the beginning of the 5th
century BC usage of Greek characters. Scholars differ in their opinions
as regards the epichoric alphabets used from that of Euboia to that of
Megara. For example, the second sign in the inscription is interpreted
either as or as . There is however one such sign in the short
inscription which reminds us of the Greek word , where the
sign > is to be read .
2) Yet the combination of two identical vertical signs that are
read differently by some16 but mostly N by Schmitt-Brandt, Georgiev
poses another problem.17
3) With all those difficulties in mind, there is no doubt that we are
dealing with a local Thracian inscription, which was written and set in
stone by a Thracian in the Thracian language. Despite all differences,
there is no other more plausible assumption but to say that a Greek
alphabet was used to write Thracian, at least as early as the 6th century
BC in the proper Thracian inland.18 For example, the six-stroke sigma,
seen in has been found for the first time used in an
inscription from “the valley of the Thracian kings” near Kazanluk that
was dated to the late 6th century BC.19 Following common logic alone,
it seems that it was easier to adopt a variant of the Greek alphabet that
was in circulation in the Thracian lands, rather than making the
difficult assumption that the engraver had used diverse Greek-based
characters that belonged to epichoric scripts.
16
Woudhuizen 2000-2001, passim.
17
Schmitt-Brandt 1967; Georgiev 1966; 1983, 1161-63.
18
Tsetskhladze 2002, 86-87, in his comments about the spread of the Greek among
the “local elites” is relying not on facts, but on interpretations. The alleged ars/art
as “common prefix in Thracian names” simply does not exist. There are two
epithets applied to Thracian deities that contain ars- or art- in compound names.
However, they are not prefixes. (See Chapter Four and Chapter Six.The
Components). Besides, his assumption that the Ionians influenced the Thracians, is
based on contradictory evidence: the inscription from Kyolmen is not in Greek, is
dated in the 6th BC, and was written with characters that do not seem to be
necessarily of Ionian origin.
19
Dimitrov 1995, 23-25.
3. Suffice it to say that some of the earliest Greek inscriptions yet

known come from Euboia, and the Euboians, who were Ionians, used
sigma instead of san (M), and that the same is true of the Boeotians and
some people from central Greece and Peloponnese including eastern
Argolid, Arcadia and Elis as well as Sparta, Rhodes, and some Greek
colonies in Italy; or Achaea and Epirus.20 After examining the stone it
seems that the engraver had had tremendous difficulties in carving the
characters. Therefore, his hesitating in orienting the inscription and its
characters appear as a major technical difficulty coming from the texture
of the stone; for limestone is soft but also fragile. Perhaps this is the reason
why he avoided carving the joining hasta of the sign and making
different S.
4. The identification of the sign n as word-divider, although preferred
by Beševliev, followed by Woudhuizen and Ancillotti, does not solve the
problem.21 On several occasions and especially in the case with the
sequence of there is clear indication that the cutter was capable
of distinguishing between and (rendered by )22 since he
painstakingly carved a horizontal hasta in . His attempt at writing the
resulted in carving a rather wide chunk off the soft limestone rock. On the
other hand, he was consistent not to make the connecting diagonal hasta in
any sign, as it was technically impossible. There is one more instance
where he successfully wrote , precisely in the short nine-letter
inscription .
5. Besides that identification there is also difficulty with the sign < ,
which occurs four times where three of the letters are oriented right to left,
and one in the opposite direction. In this case, if we decided to read
with the rest of the scholars as a Greek loan-word, we have solid
grounds to consider the other two characters being of the same nature, i.e.
signs for . Rests alone the left-to-right orientated character, which having
absolutely the same shape, and should be read as , based on the fact that
the cutter consistently shifted the position of the letters. One more
observation: There are three equal signs in the sequence
that would make sense only as N and I letters.
Here we should also consider the conclusions we can draw from the
distributional analysis; no other scholar has resorted to such analysis,
20
See Cook 1987, 9-10.
21
Beševliev 1965a; Woudhuizen 2000-2001; Ancillotti 1986.
22
See Tod 1985, 253-54 where in a 332 BC inscription from Eresus the sign is
frequently used for H or N.
although many seem to be much concerned with the linguistic aspect.

Naturally, they try to contrast their views by gleaning details on the so-
called archaic letters derived at times from too many epichoric Greek
alphabets to make their argument agreeable. Following the oldest opinion,
that of V. Georgiev, inconsistently did they try to read the -sign as either
Y or , except for Beševliev.
6. The analysis of the inscription should start with identifying
combinations of letters that make sense reasonably enough to get to
understand it. The first striking detail is the sign that has been identified as
N. We can say with some certainty that this is a negation. For this sign
occurs a total of eight times in this inscription. By identifying this sign as a
negation from nĕ, nē and nei,23 we come to the following observations on
its distribution: (1) before L twice; (2) before E twice; (3) before I once;
(4) before H once; (5) before B, and (6) in end-of-word (Auslaut) position
at the very end of the inscription.
7. Our distributional analysis shows; (1) That the -sign in the first
place appears before and only; (2) That the N-sign appears before
consonants ( ) and vowels ( ). Therefore, we can
consider of a stable phoneme but not a morpheme, bearer of meaning.
The -sign, on the contrary, exemplifies a universal and qualifies for an
independent morpheme, most probably to be connected with the negative
particle *ne (Note 21). It is exemplified in the short separate part of the
inscription where we most probably can read, based on the Greek
, a formulaic “do not harm/destroy”. Most scholars are
unanimous about it.24 Some seem not to be.25 On the other hand, although
in our catalog (Compare the following names: , IGBulg. III
1222, 3rd century AD, a personal name. Not in Detschew;
[ , IGBulg. IV 2286, 159 AD; according to Mihailov this
is a female name; , IGBulg. V 5883, 77 AD; * ,
IGBulg. V 5883, 77 AD, a place name; , IGBulg. I 77, 78,
23
See Pokorny 19943, 756.
24
Beševliev 1965a; Schmitt-Brandt 1967 and Georgiev 1966. Completely different
reading in Woudhuizen 2000-2001, 294.
25
Theodossiev 1997, 223: his “contribution” consists in repeating statements
y made by
already y others or just
j offering firsthand reversed readings, e. g.
for , etc. See also his reports on Thracian inscriptions,
composed in the same fashion with many mistakes and incorrect statements in
Teodosiev 1991.
epithet of the Heros, 1st century BC in an inscription from

Shumen, a female name. , epithet of Apollo, IGBulg. IV
1928, 2nd - 3rd century AD; , inhabitants of
, IGBulg. III 1771, 2nd - 3rd century AD; , IV
2142, 2nd - 3rd century AD, epithet of Hera; , an epithet of
Apollo, IGBulg. III 1832, 1st - 3rd century AD; Spinopara,
, a place name, IGBulg. IV 2192; ,
, IGBulg. III 1817, 1st - 3rd century AD), we don’t
have many examples with Thracian names starting with N, the distribution
is the same as in the one in the inscription. The inscription from Ezerovo
offers the following distributions: N+E; N+H; N+T. The only missing
examples are before L and before B. Based on this assumption we can
divide the text in words that eventually will make sense. So, the first word
we come up with is . However, another possible combination is to
divide in and , because is already another combination
that occurs twice here and once in the inscription on the gold ring from
Ezerovo, where ASN follows a personal name just like this is the case
here. Along with the interpretation of the Kyolmen inscription, scholars
became aware of the resemblance of some of its sequences with those
found on the inscription on a gold ring from Ezerovo.26 This has been a
fruitful surmise that on several occasions we have the combination of
or to be interpreted as a genuine Thracian word. It is a
relative pronoun, which might also be an indefinite pronoun, from *ios,
Old Indian yas, Greek hos and *ni, Prygian ios ni, Russian kto ni, etc.
“whoever”, see numerous examples in Pokorny and Haas.27 in both
inscriptions comes after a name (or at least this is our conviction that we
should read and , the first being a compound
name consisting of - and the Thracian version of Gr. , while
the second consists of the genuine Thracian contrasting with the
Greek graphemic version known ever since The Iliad as - or -
on the one hand, and ( ) to *dyaus, on the other).
8. The evidence is too scarce to ascertain that the final in
is for a nominative ending, for the two sigmas, here and at
26
Consider the inscriptions from Rogozen, as well as other inscriptions written on
metalwork in formulaic Greek, which gives us ground to say that the genuine
Thracian inscriptions follow the same pattern. See Mihailov 1987.
27
See Pokorny 1994, 283 and Haas, 75 with more reference.
the end of are different in shape, if at all those graphemic

variants display different relevant distribution. Were they to denote that
one was a final ending and the other part of sequence, remains to be
seen. However, we can assume that -// is inseparable.
9. And again, we should separate IL from ASN at least for the sake of
following the already adopted logic of separation of words. IL might well
be a pronominal stem “der, er” (for the L part see Pokorny 24-25 under al-
ol- and compare to old Bulgarian and Bulgarian “lani” (last summer, or
last year; from Bulgarian “leto” “summer” and “year”). Latin “ille” and
Umbrian “ulo, ulu”) So, we can say with some certainty that this first
portion of the inscription reads as follows:
, .
Another detail is the segment , Greek for “years”?
10. A third detail is the beginning of the second half of the inscription
where we can read ( ) ( ), well known from later Greek
inscriptions found in Thrace, e.g. , son of Seuthes III and
28
Berenike, ; Since this is an inscription in Thracian, we cannot
expect to have the Greek form as in EBPY (the stem known since Homer)
being to the river name but rather EBAROZE without
syncope.29
11. Judging on the very consistent consonant-vowel frequency in both
inscriptions, we can also assume that this was the Thracian language
pattern whose one feature is characterized by open syllables.30
Furthermore, I believe that the assumption that the inscription should start
to be interpreted from EBAROZE is correct.
There are three considerations to be made at this point:
(1) In a number of inscriptions the E-ending in a name indicates a
genitive form, e. g. on four silver phialae from Duvanlij;
28
, son of Seuthes III and Berenike, IGBulg. III 1731 and Elvers
1994, end of 4th BC and , IGBulg. 538 from Glava
Panega, 2nd BC (?).
29
( ) for (compare Homeric ). As the genuine form has
full-fledged shape with open syllables, so the syncopated form should be altered
where possible. Liquids allow for that phenomenon to occur, while sibilants do not
or at least show a different attitude and hence, no s-endings in a possibly
nominative case form.
30
See Aristophanes’ Birds 1615:
and 1628: .
(2) This idea is corroborated by the fact that -/ is

inseparable being one word, and therefore there is no reason to read
.
(3) If these considerations are correct, then we have one clear part of
the inscription starting with ( ) and ending by
.
From here we have to figure out the word divisions keeping in mind
that we now know the end of this first part. Several considerations come to
my mind:
(1) Deduced from tombstone inscriptions in Greek and Latin,
seems like another Greek loan-word for “years”, though still uncertain.
(2) If this is correct, we probably have to cut the first word after A, as
A was clearly added afterwards and hence the word corrected by the cutter
himself. So, we might have i.e. ○ with a negative
“n”.
(3) has been interpreted in many ways but mainly as a
perfect form = Gr. , “has lived” (Schmitt-Brandt, Georgiev).31
Now, we think that after the genitive form there should be
a verbal form, plus something that explains the idea of possession.
Whatever the meaning of -// , and
look symmetrical and might be linked.
Conclusions
1) There are suggestions that the stele is a tombstone of a fallen soldier
(dynast?) rather than a dedicatory inscription. The other two parts of the
inscription speak in favor of that idea.
2) Based on the archaeological evidence, the existence of a necropolis
from as early as the 6th century BC accounts for the use of the inscribed
stone.
3) The size of the stele points to the fact that it could not have served
as a cover slab.
4) There is sufficient archaeological evidence produced by Totev, and
Dremsizova-Nelchinova, that the stele in question was found face down on
the top of the tomb. Most recent publications follow Dremsizova-
Nelchinova’s opinion that the inscribed stone had been worked with and
then re-used.
31
Georgiev 1983, 1161-62; Beševliev 1965a, 321.
5) The text as regards the Thracian language leads toward a formulaic

interpretation of the inscription in accordance with other grave-related
texts coming from Greece and Phrygia.
In my opinion, had the stele been attributed a major sacred function, it
would have been placed elsewhere, e.g. on a top of a mountain. A soft
sandstone would have not survived even a several-mile-long voyage.
Practice shows that solid granite rocks had been re-used in construction.
There is no evidence so far of a fragile rock being re-used for religious
purposes. Let alone take the fact that ancient Balkan poeple had made
rock-cut dedicatory or other monuments to last for millennia, as they did,
for instance. One possible explanation would be that this rock has been
removed from the nearby settlement to adorn the grave of a Thracian
aristocrat, though a ritual performed on that occasion should not be
excluded.32
As for the meaning of the text, we put forward a plausible reading
based on phonological analysis. Our understanding is that we cannot go, at
this point, beyond the level of some certainty; crossing over into complete
uncertainty would be the same guessing.
II. The Ezerovo Ring

The inscription in scriptio continua was found on a gold ring from a
tumulus near Ezerovo, district of Plovdiv.33 The letters were incised in
eight irregular lines and the last on the rim of the ring. The find is dated to
the 5th century BC.
32
See Fol 2002, 326.
33
See Detschew 566 for a full account on the find, as well as on various opinions
expressed by nineteen other scholars.
Fig.1-2. The Ezerovo Ring
1. The inscription starts with , where the ending E is the

same as in DADALEME on the phialae from Duvanlij34, and appears as
connecting vowel in many names, e.g. Kerse-bleptes vs. its form Kerso –
bleptes in inscriptions from the Roman period. I take it this is the genitive
ending in Thracian. Compare Dios-pouris, etc. where Dios- is the genitive
form of Zeus in Greek adopted in Thracian, too. It has become clear ever
since the 6th BC from BLABA in the Kyolmen inscription and from the
component –stene from Greek “sthenes” “strong” that Greek loan words
were present in Thracian. As for Roli-, the personal name Roles will
account for it.
34
Filow 1934. The author considered –e as a genitive ending.
2. AS is a problem if not combined with NE because of the S which

normally is changed to Z between vowels. There are fluctuations in a large
number of inscriptions from the 1st century AD on, e.g.: ,
, , , , ,
In many cases, S seems to be originally in positions where the
allophone Z does not appear. At the end of a morpheme or where the
allophone Z would create a misunderstanding the S is preserved between
vowels, as long as they are not doubled with Z-forms. S- and Z- seem to
be among the most numerous word-initial letters in Thracian. Some of
them show the above mentioned vocalic variation, though most seem to
reflect the etymological situation since PIE.
Traces of possible assimilation or dissimilation played a decisive rôle
in shaping the forms. Greek (through the cultural process of Hellenization)
contributed to this effect as well.
Sometimes, this can be attributed to bilingual texts, e.g. in
.35
Under the circumstances AS should be interpreted as *so, anaphoric
pronoun.36 Whether NE should be added up or not, is a possibility. In a
previous scrutiny of the case37 it was pointed that AS could be part of
ILAS in ILASNE, where the latter could be a deictic. Thus, ASNE means
“of that same”, if put in Genitive.
3. Or we can cut NE and RENEA, a negative imperative (“do not
remove”) from IE *r/or/r.38 If we have made the right observation, A in all
the sequences, as in RENEA, TILTEA, NESKOA, RAZEA, would be
equal to “or”. Then, TILTEA would be “do not lift” < *tl°-to, a suffixed
zero-grade.39
4. Following the same pattern, NĒSKOA should be segmented in a
negative N’ ĒSKO and A. It should be mentioned here that we have one
example of this sequence among the entries in The Evidence (the Greek
inscriptions): Niskatourme. For niska, Pokorny 311 under *en “in”, PIE
*ni, Skt. ni-ja- ‘ingenuus’, Avestan ni-zənta-, Skt. ni-tya- “proper”,
“bestaendig, eigen” = Gallic(?) Nitio-broges “akin to, apparenté”,
5. RAZE would make sense because of the intervocalic Z (=S) and A
is “or”.
35
See Chapter Four, s. v.
36
Beekes 1995, 202-203, 205: “that, the (just named)”. See also Watkins 2000, 81
and Pokorny 19943, 978. Cf. Avestan aeśa “this”.
37
See Dimitrov 2003.
38
See Pokorny 19943, 326-327.
39
Watkins 2000, 90; Pokorny 19943, 1060.
6. DOMEANTI40< *demo, dm°ə, see Pokorny 199-200; Lat. domo,

Perfect domui; Watkins16, suffixed o-grade form *dom(ə)-o “tame”,
subdue’, ‘adamant’ < *n°-dm°-nt-.
7. LEZYPTA “ying (up and) down on his (her) back”, see Watkins 94
at 6a. and 11 a. and b. under upo- “under, up from under, over”, Avestan
upa “up to, at” in upasta-vaka “praise”, etc.
8. MI = “to me”
9. In HERAZHLTA, I see the following sequences: H = I and therefore
IERA (=Greek, ‘sacred’) and ZHLTA, a past particle in -t, plural for “gold
objects”.
As regards a tentative translation of the inscription that will allow to
get a working notion of this difficult text, I propose the following
translation line by line:
(belongs) To Rolistene (of that same)
Do not remove
Or do not lift
Or (and) akin
To the obeying
Lying down
To
me sacred gold objects
This rough text could be edited in this way: “(This ring belongs to me
here) Rolistene, do not remove or lift the gold objects (that are) sacred to
me, the obeying lying down.”
III. The gold ring from Duvanlij41

The find is dated to the 5th century BC. The ring is in bad shape and I
was able to get but the following reading of the letters incised on the oval
bezel: HYC vacat A on the upper side, and on the lower side,
an image of a horseman riding to the left being in the middle. For
(with the typical sign for Z) we have a parallel in Detschew
291 Mezanaia, a feminine name from Kavadarci, now in Former
Yougoslav Republic of Macedonia. Therefore, MEZENAI seems to be a
dative from *mez-/maz-, “big”, IE *magh- and suffixed with -a/eno-. As
40
Beekes 1995, 249 on the formation of participles: The participle was formed
with –ent, Skt. adánt “eating”, yant “going” (* H1i-ent-), Thracian dome-ant-i
“obeying”.
41
Filow 1934, 130.
for the other letters, due to the worn-out parts of the ring, I am not able to
give any reasonable sequence.
Detschew 291 proposed a different reading: he reads MEZHNΛI and
considers it a genitive from MEZHNΛIΣ. Georgiev, followed by
Duridanov, adopted a different way of interpreting it.42
Fig.1-3. The Gold Ring from Duvanlij
IV. Four silver gilt phialae from Duvanlij43

Four silver gilt phialae, dated to the 5th century BC, were found in
Bashova Mogila, Duvanlij, bearing one and the same inscription:
. As mentioned above after Filow and Detschew 110 this is
a genitive form. This assumption is made on the existence of a widely
spread hypocoristic personal name . Georgiev44 made a whole
sentence out of it. Duridanov accepted his opinion and so did Katičić.45
V. Bilingual text
In a number of Greek inscriptions there are names that to my mind are
written in Thracian. I will interpret one of those bilingual texts here.
, on a monumental tomb in Smjadovo,
district of Shumen, northeast Bulgaria is a recent discovery (2000) after a
rescue excavation a burial mound No 47, west of a Thracian settlement, 4th
- 3rd century BC. Ed. pr. in Gicheva, R. and Rabadhziev, K. (2002, 550-
42
Georgiev 1983, 1160-61; Duridanov 1975, 90-91.
43
Filow 1934, 63.
44
Georgiev 1983.
45
Katičić 1976.
557). In this case, we will divide the first word in three, GONI MA
SHZH. This looks like a banal example of a funerary inscription in the
accusative case. The editors of the inscription suggested a thoroughly
unacceptable interpretation. We will follow and apply our method of
internal reconstruction. First, I will argue the very structure of this whole
line, as it appears on the tomb. As we have already made it clear, the
Thracian /s/ between vowels changes to /z/. Therefore, one should expect
the sigma to be a Z. And this is our first argument that the sequence starts
from S on yielding a separate word . Then, we feel that we should
have two other separate words out of , where MA is easily
identifiable as possessive-pronominal stem from *ma- “my, mine”.46 As
for we see a dative ending –i (already evidenced in many
examples, e.g. , , etc.) from IE *gwn-i “woman,
wife”. Following this logic of interpretation, we can get back to the Greek
translation “Seuthes’ wife” and follow up on the developments in .
We already know that there is no theta in Thracian and that the Greeks
perceived the Thracian /t/ from IE */th/ as theta. The phoneme /t/ in its
turn was subjected to assibilation and hence the result /z/, e.g.
from *Di-, the dative form of the IE word for “light”. The development so
far takes us to Seout, Seth, Sez because the diphthong existed in the Greek
version of that royal name. But in fact, the original Thracian etymon was
Sese. We are not sure about the nominative ending in Thracian and we
prefer to consider that /s/ was not used as such. A good parallel is seen in
Sese47 in an inscription from Germania Superior: Sese Lenulae f.
D[a]nsala mil. ex coh. [IIII?] Thracum, etc.
VI. The Inscription from Sitovo

In the hills near the village of Sitovo near Plovdiv, an inscription was
found in the late 1920’s.48 Ex autopsias, L. S. Bayun suggested a new
reading and a tentative translation in a Phrygian dialect.49 This gave
46
Bader 1986, 115.
47
CIL 13, 7049, see Detschew 434.
48
See Guerassimova-Tomova 1989, 131-40, for a full account on the history of the
problem with the inscription. Note that the title of her article in the original
publication is in quotation marks and according to it “la formation rupestre près de
Sitovo n’est pas une inscription.”
49
She considers the inscription genuine and dates it between the 3rd and the 1st
centuries BC: Bayun and Orel 1991, 147-48.
grounds for the inclusion of the text in the Phrygian inscriptions.50 During
my visit to that place on a steep slope of the Rhodope mountain, on what
seems to be a rock-cut façade of a monument, I saw clear lines of an
inscription. However, after examining the script, I came to the conclusion
that this could not be a Thracian inscription.
50
Orel 1997, Th-01, 352-356.
CHAPTER TWO
LEGENDS FROM COINS1
1. #1 ( ), ca. 500-480 BC
2. #7-16 , 475-465 BC, from the Velikovo Treasure
Fig.2-1 Fig.2-2
Coins of the Thracian tribe of the Derroni
3. #2,3 ( ), (‘of the Tyntenoi”), 500-450 and 475-465 BC

4. #4,5 OPHΣΚΙΩΝ, 500-480 BC (However, see Dečev 345 s.v.:
, quoted by Head (19113, 195; see
also Youroukova 1992, 16), connecting the name to I, 268 in the Iliad,
where the Centaurs are called mountaineers . This is a
typical case of Volksetymologie, the Thracian name only seeming to be a
Greek word. In fact, the Greeks might have had in view that they (i.e. the
Oreskioi) were really dwellers of mountains. According to Yurukova
(1992, 16) the Oreskioi, who did not happen to be evidenced in any other
source, are a generalized appellation for those Thracians dwelling the
Paggaios mountain. The only two hoards of staters from the Bulgarian
lands, unfortunately dispersed, were discovered in the Upper Nestos
course near Nevrokop. In both cases, archaic staters of Thasos have been
discovered along with the Oreskioi stateres, similar in their iconography,
which according to Yurukova belong to the same geographical region. As
1
The list has been compiled according to Yurukova 1992.
22 Legends from Coins
both types of coins were in circulation in the Upper Nestos course, one
might assume that those were the territories of the Oreskioi; their mint was
not far from Thasos in the region of Paggaios.
5. #17,18 ΜΟΣΣΗΩ, (i.e. “(belongs) to Mosses”), 475-465 BC,
Detschew 325 quoting Head 19113, 200 gives and ΜΟΣΣΗΣ
(Mωσσεω and Moσσεω); cf. also Detschew 325
and IGBulg. III 1473.
6. #19 GETA BAΣIΛEY HΔΩΝΕΩΝ, 476-465 BC, see Yurukova
(1992, 25) for details. She claims that the coin and its legend are
absolutely authentic and the only sources about this Thracian king.
7. #22-30 , 445-435 BC. This is a coin of the first
Thracian king of the Bisaltae, who had minted coins in the area between
the lower Struma and lower Vardar.
8. #31-37 , last decade of the 5th century
BC, a type silenos/kantharos coin, see Yurukova (1992, 43-45).2
9. #38 and #39, #40-1, 40-2 ,
last quarter of the 4th century BC to 350 BC, Yurukova (1992, 47-48).3
10. #41,42,43 MHTOKO beginning of 4th BC.4
11. #44-1,44-2, 45-48 ( ) 387-383 BC.
12. #49-57 KOTY 382-359 BC.
13. #58-65 AMATOKO, i. e. Amotokos II.
14. #66, 67 351-350-347 BC, Teres II.
15. #68, 69 ( ) 356-352; the pi is shaped as gama;
minted near Thasos.
16. #70-72 ( ), i.e. ‘belongs to Kersobleptes’, 359-346
BC.5
17. #73-74 , 340-316 BC.
18. #75-105 ΣΕΥΘΟΥ, 320/315-316 BC.
19. #106 [ΒΑ]ΣΙΛΕΩΣ / [Σ]ΠΑΡΤΟΚΟΥ 281-279/277 BC.
20. #107-111 ΣΚΟΣΤΟΚΟΥ, 277-260 BC; 108, 109 - ΒΑΣΙΛΕΩΣ
ΛΥΣΙΜΑΧΟY ΣΚΟΣΤΟΚΟΥ.
21. #112 ΡΟΙΓΟΥ, first half of 3rd BC.
2
See also Dimitrov K. 1999.
3
Dimitrov K. 1999, 175-180, claiming that these belong not to Seuthes I, but to
Seuthes II.
4
Cf. the last reference on MHTOKO in Dimitrov K. 1999 with chronology;
Schoenert-Geiss 1985; 1987, 27, 51-52; Yurukova 1992, 53-56, 62.
5
The reading belongs to Yurukova. I think that the correct one should be with an e
after Kers-E-bleptes.
Legends from Coins 23
22. #119-120 ΒΑΣΙΛΕΩΣ ΚΟΤΥΟΣ, ca. 270 BC; reverse:

ΒΑΣΙΛΕΩΣ ΡΕΣΚΟΥΠΟΡΕΩΣ.
23. #121-122 …ΤΥ. . ΚΟΤΥΣ 171-167 BC; 122: [Β]ΑΣΙΛΕ[ΩΣ]
[ΚΟ]ΤΙ.
24. #133-139, 175-100 BC Mostis; [ΒΑΣ]ΙΛΕΩΣ ΜΟΣΤΙΔΟΣ ΕΠΙ
ΣΑΔΑΛΟΥ; 134 ΒΑΣΙΛΕΩ(Σ) ΜΟΣΤΙΔΟΣ.
25. #140-143 ΚΟΤΥΟC (sic!) ΧΑΡΑΚΤΗΡ; 142: ΚΟΤΥΟ(C), 143
[ΒΑΣΙΛ]ΕΩΣ [ΚΟ]ΤΥ[ΟΣ].
26. # 144-146 ΒΑΣΙΛΕ[ΩΣ] ΣΑΔΑΛ[ΟΥ], 48-42 BC.
27. #147-150 ΡΕΣΚΟΥΠΟΡΙΣ 48-43/42 ΒΑΣΙΛΕΩΣ
ΡΕΣΚΟΥΠΟΡΕΩΣ.
28. #151 on the reverse: ΒΑΣΙΛΕΩΣ RΟΙΜΗΤΑΛΚΟΥ, 10 BC-12
AD, many instances.
CHAPTER THREE
EVIDENCE FROM METAL VESSELS
Pre-Roman Times
1. If we go in chronological order, we should perhaps start this list with

an urn (now in the museum of Karlovo) found near the principal route
between Plovdiv and Karlovo inscribed with a two-line text Μιλητια /
Επιαιος (the second line in retrograde)1.
2. , on two silver-gilt phialae, a silver mug and a silver
rhyton from the Bashova Mogila at Duvanlij, Inv. Nos. 1515-1518. See
Chapter One.
Fig. 3-1. The silver mug from the Bashova Mogila at Duvanlij
1
Archibald 1998, 148, n. 65, Domaradzka, addendum to Domaradzki 1993, 57, n.
7. The inscription is said to be dated in the 6th BC. The script would then be one of
the oldest in the Thracian inland and comparable to that in the Kyolmen
inscription. Is that really the case? Non vidi.
26 Evidence from Metal Vessels
The Treasure from Rogozen2

The inscriptions are grouped according to the way they were written.
Mihailov distinguished between inscriptions where letters were made out
of dots (en pointillé) and inscriptions carved on the surface of the metal.
The different techniques used were found on a tray with a mythological
scene (see below No. 2). Two people worked on it (two hands?).
The treasure was found accidentally in a field near Rogozen, NW
Bulgaria, by a farmer working on his property. The vessels were buried in
two lots 50 cm deep into the soil. A total of 165 pieces made of silver,
weighing 19. 91 kilograms were discovered on that spot situated north of
the Haemus (Stara Planina) Mountain.
The inscriptions include:3
Inscriptions en pointillé
a. , on a silver phiale, Inv. No. NIM 22327. G. Mihailov
preferred this reading for a reading with - at the end. Although there is a
sign resembling that of a upsilon, the rest of the finds show a consistent
use of - and not a single instance of a - ending.
b. , on a silver-gilt phiale, Inv. No. NIM 22304.
c. Two instances of , on two silver phialae,
Inv. No. NIM 22342 and 22346. Argiske/Ergiske should be localized in
the lower Hebros valley and its name is related to the name of the small
river Erginos, a tributary of the Hebros River, cf. Mihailov (1987, 1, 9).
d. . On a silver phiale, Inv. No. NIM 22343.
For its localization, see above.
e. . On a silver phiale, Inv. No. NIM 22331. For
Apros (later Roman colony Apri), localized near the Propontis, see
Theopomp. FGrH 115, F 160 in St. Byz. 107. 5, Ptol. 3. 11. 7, Strabo, 7.
frg. 55, Plin. NH 4. 47, cf. Detschew 20.
f. . On a silver phiale, Inv. No. NIM 22330.
g. , on two silver phialae, Inv. Nos. NIM 22328 and
22340. Cf. No. 22 below.
h. on a silver phiale, Inv. No.
NIM 22329.
2
Mihailov 1987.
3
Concerning the date of the inscriptions: Mihailov 1987, 1, 10.
Evidence from Metal Vessels 27
The damaged rim of the vessel does not allow for a secure reading of
the delta in Disloias. According to G. Mihailov it is dubious, and he
considered an initial (lambda) a possible variant. I think that the
Thracian material offers every supporting evidence of phonological point
of view to interpret this character as a delta rather than a lambda. (For
further considerations to that effect see in the second part of this book
dealing with the phonological theory of the Thracian material)
Fig. 3-2. Silver phiale from the Rogozen treasure, Inv. No. 22329
i. , a silver phiale, Inv. № NIM 22347.

j. (also see below two other examples (the
phiale from Alexandrovo and that from in a private collection), a silver
phiale, Inv. № NIM 22345.
k. . On a silver phiale, Inv. № NIM 22341.
Because of the specific word formation of Thracian names, we can
assume that this name is a composite one, where its components are
* in -, - on the one hand, and , on the
other hand. (More information further on in Chapter Three and Chapter
Five)
l. , on a silver-gilt jug, Inv. No. NIM
22412.
m. , on a silver phiale, Inv. No. NIM
22344. For the variation Kerse-/ Kerso- in this name see further on in
Chapter Five.4 For the ending e- in Kerse- compare DADALEME on the

silver phialae from Duvanlij.5
Graffiti
n. seems to be added 5 cm aside to the left of, on a silver phiale,
Inv. No. NIM 22345. (see above under No. 10). These letters have been
carved (not dotted) by someone else.
o. was incised at the bottom of a phiale from the treasure of
Rogozen Inv. No. NIM . 22340. Cf. g. above.
p. A longer inscription beginning with , then or
(?) vacat vacat …, on the bottom of a silver jug,
Inv. No. NIM 22418. G. Mihailov is not certain about the reading. The
sequences do not yield to any known name or word.
q. On the same vessel that bears (see above b.) there is
also . G. Mihailov gave no explanation. Some suggested this to
be the Greek adverb for “indeed”. However, this seems not to be the case.
We will reconsider it later, in connection of the formulaic language of
other inscriptions, e.g. on a phiale from Duvanlij, district of Plovdiv
(ancient Philippolis) a short inscription reads , on a gold ring
from Ezerovo inscribed with Greek letters in the Thracian language
another one reads , etc.
The following inscriptions on metal vessels (4. - 7.) belong to the first
group A, made en pointillé.
Phiale from Vratsa

< > on a phiale from the mound of Mogilanska
Mogila in the same region as Rogozen (district of Vratsa). According to G.
Mihailov, 1987, the inscriber committed an error writing ET instead of EG
as it is clear from h.: . Actually, “errors” of that sort and others,
such as omissions of letters, occur often times.
4
The two differently spelled forms did not go unnoticed. Badian 1983, 55 and n.
13. and Archibald 1998, 232 made note of it. The latter attributes the first to
epigraphic texts and the second to literary ones.
5
Two silver-gilt phialae, a silver mug and a silver rhyton from the Bashova Mogila
at Duvanlij: Filow 1934, 63, no. 2, 65, no. 3, 67, no. 4.
Fig. 3-3. Silver phiale from the Mogil

Inv. No. B-68.
Phiale from Alexandrovo

anskata Mogila tumulus,
, on a phiale from Alexandrovo (district of
Lovech).
Phiale from a private collection

or or on a phiale (probably from
the region of Pleven, not far from Vratsa, private collection).
The Treasure from Borovo

, on a rhyton and jug from Borovo (see Ivanov
1982, 22-35, pl. 27, fig 12, a, b, c).
( or or ) on a rhyton from Borovo. , as the
inscription reads, is a misspelling committed by the inscriber.
Phiale from Agighol, Romania

, on a phiale from Agighiol (district of Tulcea in
northern Dobrudza, Romania) (see Berciu 1974, 57-59, fig. 14 (facsimile).
Phiale from Branichevo

, on a phiale from Branichevo, (district
of Shumen, northeast Bulgaria). G. Mihailov’s comment (Mihailov 1987)
about the last two words is that this was an addition made by someone else
in bigger characters.
All these inscriptions constitute a separate field in the interpretation of
Thracian onomastics. Although written in Greek, they give us an idea of
the formulaic language used in similar cases.
Of particular importance to the material under study are the genitive
forms. The proper (personal) names followed by a preposition and again a
name in the genitive is the most common formula that occurs with these
inscriptions. The second genitive is a place name. The meaning of those
genitives seems to be clear. G. Mihailov thinks that the first one is to
designate the owner of the vessel, while the second is assumed to be the
place it is coming from. G. Mihailov argues that the interpretation of the
second genitive (that of the place name) is not to be understood as the birth
place of the owner. He is adducing the text of f. to
support his translation, which is “(vessel) of Kotys, Satokos, etc. (proper
names in genitive) from Beos, etc.”
Mihailov made a good point observing that if the second genitive was
to be understood as the names of the towns which offered gifts to Kotys,
the prepositions or would have been used.6
Another interesting insight is the probability that the owner had a two-
fold intention by indications that not only these precious gifts were his
property, but also that the towns were part of his estate.
Looking at all these inscriptions, one cannot help noticing that
Didykaimos, Satokos, Kotys, Kersebleptes, Teres, and Amatokos, all
owners of precious vessels, could well be historical figures, members of
the Odrysian royal family, end of 5th- first half of the 4th century BC:
Satokos is known as the son of Sitalces; Kotys is the Elder Kotys, cruel
king that lived between 383 and 359 BC; Kersebleptes, his son, ruled
between 359 and 341 BC.
Ergiske was a town on the lower Hebros river; Apros(n) is not far from
there, in the Propontis region. Beos(-n) was not known, however the
Roman station of Beo-dizos or Be-dizos was situated near Apros(-n).
6
Mihailov 1987.
The phiale from Kazanluk

On the bottom of a small phiale found during excavations conducted in
the so-called “Valley of the Thracian Kings” near the town of Kazanluk,
there is the following text in scriptio continua:
, which I interpreted as follows:
“(vessel of) Dyntas (Dyntos?), son of Zeila(s)”.7
Fig. 3-4. The phiale from Kazanluk
The reading is based on autopsy. The signs of the letters were incised
with a tool. Several observations are to be made about the possible
reading:
(1) Whoever incised the signs on the bottom of the phiale is not to be
compared with his Greek counterparts as regards the form of the letters,
and the way the short text had been edited;
(2) Letters such as the “dzeta”, the hesitation in the orientation of
“delta”, the “sigma” at the end;
Besides, when deciding on the final reading, there are other
considerations, such as the historical and cultural traditions of the
inscriptions found in Bulgaria. This phiale is, in the first place, not to be
isolated from the rest of the inventory of Thracian phialae, as the artistic
form and the contents of the short texts show.
7
Dimitrov 1995, 23-25.
The possible objections raised will be addressed elsewhere, but I will

mention the most important here:
The case of the “irregular” after the name of ( ), which
should be in the genitive case: In many ways, the Greek inscriptions
found in Bulgaria, follow a local tradition of competence in using a
foreign language, and there are a number of parallel instances in that
sense.8 Also, one should take into consideration the natural milieu in
which these inscriptions were written and therefore should not venture into
looking for exotic names or reaching for uncertain solutions.
In the end, the main difficulty was primarily in identifying signs like
“I” for “dzeta”, and the reading of the characters after ; it looks like a
ligature (a ligature would be too strange to appear so early in the 6th -5th
century BC), but to my mind it is, within the text of the inscription, a
badly executed one after so many hesitations. The engraver was in fact not
a professional one but rather a stranger who used Greek characters and
language.
The difficulty of the reading is in the “ligature” of A and Y.
The name of the main character is interesting for us in several ways.
First, both names appear for the first time here. For ( ) we can say
that it is a compound made of - and ( ). The form with –s is not
secure and this nominative could be restored to . The distinction
between H and I (iota) make us think of at least one obvious conclusion.
Namely, that the stem ZH was productive ever since the 6th BC. There are
no examples of that time with Z for what seems to be PIE *dy-.
A gold ring
, Filow 1934,105, on a gold ring with an image of a
horseman, found in a 5th century BC burial. According to the practice in
Thrace (see the inscribed phialae from Rogozen and elsewhere), this is a
genitive case-form, translated as “belongs to Skythodokos”. This is a
perfectly good Thracian name with a second element –dokos/-tokos
(Detschew, 145, 462-463; Dečev 1960, 158-76; Beševliev 1965b, 13-14;
1-57; Brixhe and Panayotou 1994, 190, 198). J. Boardman, in his work on
8
Nominative for Genitive: Meisterhans 1900, 203:” In Rechenschaftsablagen tritt
hauefig fuer einem Genetiv, Dativ, Akkusativ oder praepositionalen Kasus in
freierer Weise des Nominativ ein: bei Appositionen: CIA II. 809, c, 170 (325 v.
Chr.); Hodot 1990, 116, n. 168 Nominatif employé pour le génitif?; 122 (n. 205)
Mayser 1935-1970, 193 lines 4, 6, 25; 206.
Greek gems and fingerings, advanced an explanation of the inscription

which reads, as follows: “an odd name which should mean “friend” or
“host of the Scythian” (Boardman 2001, 230). He is followed by other
scholars (IGDOP, 13-14; SEG 44, 1999, No. 871).
Fig. 3-5. Gold Ring of Skythodokos
Two silver phialae and a bronze helmet from Shipka,

from the Golyama Kosmatka tumulus
1.
2.
9
Kitov 2005a, 90; 2005c, 49, fig. 20, 51, fig. 22; 52
Fig. 3-6. Silver phiale from Shipka
Roman Empire Times

1. , , an epithet of the
Thracian Heros (Horseman) on a silver vessel from Brestnitsa, district of
Pleven, north-central Bulgaria (Petkov 1960), 2nd - 3rd century AD. Not in
Detschew.
2. , , made with gold inlay on two silver
casseroles from Kyustendil, ancient Pautalia, southwest Bulgaria. It has
been published by Manov (1994), 1st - 2nd century AD, where he read
. The name is related to Koul- and -ktios, -k(t)us, -kous see
above and Kout- and Cutiula, a personal name CIL 3, 7330,
from Thessaloniki on a relief of the Thracian Heros (Horseman). A secure
parallel is the epithet (see above) from Koul-kous-anos/-
enos. For the reading of this name, see the considerations in Chapter Four
below under . Not in Detschew.
Fig. 3-7. The Silver Casserole

CHAPTER FOUR
THE GREEK INSCRIPTIONS
, a name (probably male) appearing only once in an

inscription from Metodievo, northeast Bulgaria, see IG Bulg. II 771:
, 1st - 3rd century AD.
According to Mihailov the is not sure. See Detschew 2 for a division in
- and the alternating / , although there is no
evidence from inscriptions found in Bulgaria for an independent use of the
latter.1
-, IGBulg. V 5687, - (vacat). Not in Detschew.
, IGBulg. IV 2338, a strategia (military district), 3rd
century AD. See Detschew 7.
, IGBulg. IV 1927, 2nd - 3rd century AD, on a column
inscribed with names (a sacred catalog?); an epithet of .
* , see above. The name of the place, village or small town,
where had been celebrated.
, IGBulg. I, 322ter from Mesambria (3rd - 2nd century BC):
[ ] Mihailov; II 560 f. “reperta in Asclepeio ad Glava Panega,
nunc in museo Serdicensi conservatur (inv. 3883)”; II 845 (gen. );
III, 1204 3rd AD; , III 1690
b18, from Pizos, 202 AD; III 1735 from Viden, near ancient
Seuthopolis in the sanctuary of the Thracian Heros, 3rd century AD.
, a personal feminine name, IGBulg. IV 2338, around the
end of the 1st century AD.
1
On the phonological variants , , etc., see Chapter Five.
38 The Greek Inscriptions
, a personal feminine name, IGBulg. III 974, in the vicinity of

ancient Philippopolis: , 3rd century AD;
Detschew 9 , masc., incorrect reading.
vacat , an epithet of Asklepios, IGBulg. 5810, 2nd - 3rd
century AD. Not in Detschew.
, a personal name, IGBulg. IV 2233 line 2, probably from
the 3rd century AD. See further discussion about Ai- versus Oi- and the
names of a town in Homer Iliad, 6. 304 Steph. Byz. 487. 7:
Pliny NH 4. 42: Oesyma. Not in
Detschew.
, a feminine personal name from Odessos, IGBulg. I 187,
considered Oriental by Mihailov: …[[ ] …, cf.
Zgusta KPN § 57, 1-33.
Amazenes, in a funerary inscription Damanaeo Amazeni, from Cabyle,
150 AD, Mihailov 1961, 208 and Velkov 1991, 29.
, a personal male name in IGBulg. III 1732:
[ ] dated to the early 3rd
century BC. Cf. IV 2297 line 4 (near Sandanski) .
, a female personal name from Odessos, IGBulg. I, 178,
considered Oriental by Mihailov: “… …”, cf.
Zgusta KPN § 57, 1-33. Detschew 16 considers it Thracian.
, a personal female name from Odessos, IGBulg. I, 178,
considered Oriental by Mihailov. Detschew 16 considers it Thracian.
, a personal male name from the Asclepeion in Glava
Panega, IGBulg. II 532, 533, 534, from 2nd - 3rd century AD; IV 2217,
2218, 3rd century AD. This is the same person whose name is ( )
( ) [ ] , see IGBulg. IV 2338, end
of the 1st century AD, .
(cf. the phiale from Branichevo in Metal Vessels)
, a personal name from Lipnik, district of Razgrad,
northeast Bulgaria, IGBulg. V, 5269: . Not in
Detschew.
, an epithet of Apollo IGBulg. III 1723, and see the lemma,
1st - 3rd century AD.
(?), IGBulg. III 1074, 211 AD.
, vel - , . A silver phiale from the Treasure
from Rogozen, Inv. No. NIM 22331; , on a silver phiale
The Greek Inscriptions 39
from Rogozen, Inv. No. NIM 22330. For Apros (later Roman colony
Apri), localized near the Propontis, see Theopompos FGrH 115, F 160 in
St. Byz. 107. 5, Ptol. 3. 11. 7, Strabo, 7. frg. 55, Plin. NH 4. 47; cf.
Detschew 20 (Cf. Chapter Three).
, IGBulg. I, 281 bis, as a second name of a wife of a
Thracian. Not in Detschew. The variant occurs in I, 77 as a
priest’s name.
, a town, from an inscription found in Dionysopolis, now
Dobrich, northeast Bulgaria, IGBulg. I, 13, line 6. According to Mihailov
(see also Detschew 121) the name seems to be in its original
form, as there are many forms in - from – , the word in
Thracian for “(market) town” besides ; around 48 BC.
, in two instances off , on two silver
phialae, Inv. No. NIM 22342 and 22346. Argiske/Ergiske should be
localized in the lower Hebros valley and its name is related to the name of
the small river Erginos, a tributary of the Hebros River, cf. Mihailov 1987,
1, 9. See . Not in Detschew.
, IGBulg. III, 1206 from Batkun, district of Philippopolis.
Ethnicon: . Mihailov quotes an inscription from
CIL 6, 2799 line 27 “M. Aur. M. f. Fl. Diza vico Ardileno” (from
Philippopolis), which makes the ethnicon certain.
, in Manov 2008, 126, 2nd - 3rd AD, from the museum of
Blagoevgrad. Not in Detschew. Two short notes. There is a redundant
phoneme variant rendered through beta as opposed to - ou. Therefore, we
could make the assumption that the second component of this composite
name is perhaps - from *poris, *por.
, IGBulg. IV 2074 from Pautalia, 2nd - 3rd century AD.
, [ , an epithet of the Thracian Heros,
IGBulg. III, 1599 from Augusta Traiana, 1st - 3rd AD. Here belongs
* , the name of the place where this local god was worshiped. From
the name of the river ?
, IGBulg. IV, 2003, epithet of Hera from Dolni Lozen,
district of Sofia.
, vel – , IGBulg. V, 5011 line 9. 1st century BC - 1st century
AD. Place-name from Moesia inferior.
, IGBulg. III 1588 from
Karanovo but brought there from Augusta Traiana, see comments by
Mihailov: , an epithet to Zeus Sebazios, 202 AD.
, IGBulg. II 677, 678, an epithet of Zeus Sabazios from

Nicopolis ad Istrum.
, IGBulg. III 980, an epithet of Hera from Philippopolis, 3rd
century AD.
, , from an inscription on a column, see Dimitrov
2008 (forthcoming).
, vel , IGBulg. III 1662, an epithet of the Heros
from southern Thrace in Bulgaria.
, IGBulg. IV 2319, an epithet of the Heros from Melnik,
southwest Bulgaria.
, IGBulg. V, 5883, line 8, from Kresna, southwest Bulgaria (see
IGBulg. IV, p. 233) personal name that is considered Oriental by Zgusta
(Zgusta KPN § 117,2).
, Manov 2008, 124, not in Detschew.
, IGBulg. III 1116 from Batkun, district of Philippopolis, a
military district . Cf. also IV 2338 , another form
for that same military district.
, IGBulg. III 1393.
, a Thracian male name, IGBulg. II 856, 1st- 2nd century
AD. The is not certain, possibly ? The inscription is dedicated to the
same god as in II 804 and 855.
, an epithet of Apollo, IGBulg. III 1859, 1860, 155-156
AD.
A , an epithet of the Thracian Heros, IGBulg. I 353, near
Mesambria, 1st century BC; , V 5381, 3rd
century AD, an epithet of the Thracian Heros; II 802, 841, 3rd century AD;
V 5394bis, 1st - 3rd century AD, an inscription now in the Museum of the
town of Poltava (Ukraine). This inscription has been found on the North
Black Sea coast.
( ) , IGBulg. II 801, 3rd AD; III 1838, 1st - 3rd century AD.
, IGBulg. III 1622 line 2; III 1838; III 1597;
, V 5394, 3rd century AD, an epithet of the Thracian Heros.
, IGBulg. III 1593 line 18, 1st - 3rd century AD.
-, IGBulg. II 845; IGBulg. III 1522, 1528, 2nd - 3rd century AD;
V 5345 from Marcianopolis, 2nd - 3rd century AD.
, IGBulg. III 1524, 2nd - 3rd century AD.
, IGBulg. I 275 from Odessos, 2nd - 3rd century AD (?).

, IGBulg. I 374, 2nd - 3rd century AD.
, IGBulg. I 374, 2nd - 3rd
AD; , I 402, after 19 AD, a dedication to his daughter; I 281
bis, from Ezerovo, 2nd - 3rd century AD; I 378, after 19 A. D; II 542 from
Glava Panega; II 830 and 839; , IGBulg. II
846; III 1112, 2nd - 3rd century AD; III 1169, 2nd - 3rd century AD; III 1218,
2nd – 3rd century AD; III 1404 , 3rd century AD;
III 1529, 2nd - 3rd century AD; III 1598, 1st -
rd st rd
3 century AD; , III 1599, 1 - 3 century AD;
, III 1670, 1st - 3rd century AD; III 1703,
st rd
1 - 3 century AD; III 1690b 48, , 3rd century
rd
AD; III 1690c 21, 3 century AD; III 1690b 48
, 3rd century AD; III 1690 b 49;
III 1690c 63, 3rd century AD;
III 1690b 22, 3rd century AD;
, III 1690 b 56, 202 AD;
, III 1690 b 31, 3rd century AD; III 1690b
56, 3rd century AD; III 1690 d 23; 1690b26 ,
rd rd
3 century AD; , III 1690 b 68, 3 century AD;
, III 1690 c 64, 3rd century AD; III 1713,
1st - 3rd century AD; III 1832; , III 1806, 1st
rd st rd
– 3 century AD; III 1832, 1 - 3 century AD;
, III 1846, 1st - 3rd century AD; III

1852, 1 – 3 century AD; IV 2201, near Pautalia, 2nd- 3rdcentury AD; IV
st rd
2234 from Sportela (Rila) 2nd - 3rd century AD;

nd
, V 5300, 2 - 3rd century AD;
, V 5360 from ancient Marcianopolis, present-day
Devnya. This occurrence does not exist in Detschew; V 5365,
, 2nd - 3rd century AD; V 5368;
, in V 5380;
V 5652, from
Bizone, 3rd century AD; V 5824, from Daskalovo,
3rd century AD; , V 5380; V 5381;
, V 5394bis.
, IGBulg. V 5343, from Marcianopolis,

2nd- 3rdcentury AD.
, IGBulg. II 677 from Nicopolis ad Istrum, 2nd - 3rd
century AD.
, IGBulg. III 1734, 1st-3rd century AD; V
nd rd
5467, 2 - 3 century AD.
, IGBulg. I 278, 3rd century AD.
, IGBulg. III 1316, 3rd century
st rd
A. D; III 1627, 1 - 3 century AD and Böttger and Halloff (1991, 478-
481), from Karasura.
, IGBulg. II 838; IGBulg. II 841
; II 868; , IGBulg.
III 963, 2nd - 3rd century AD; III 1515 b 13, 14, 15, 27-28, 3rd century AD;
III 1659, 1st - 3rd century AD; , III 1690
column b line 25, 202 AD; , III 1806, 1st -
3rd century AD; III 1853, 1st - 3rd century AD; III 1880, 1st - 3rd century
AD; , V 5329, from Marcianopolis, 2nd - 3rd
nd rd
century AD; V 5613, 2 - 3 century AD; V 5334, 161 AD.
, IGBulg. V 5287, 5288, 5289, all from the sanctuary
of Draganovets, 2nd - 3rd century AD. Not in Detschew.
, IGBulg. V 5291, 5290, 5292, all from the sanctuary of
Draganovets, northeast Bulgaria, 2nd - 3rd century AD. Not in Detschew.
, IGBulg. V 5294, an epithet of the Thracian Horseman
from the Draganovets sanctuary, northeast Bulgaria, and possibly in V
5306 .. (vel . . ), see Mihailov. Not in Detschew.
, in , IGBulg. III 1816, from the
sanctuary of the Thracian Horseman near Lozen, district of Harmanli,
northwest of ancient Hadrianopolis. Not in Detschew.
, IGBulg. II 844, a Thracian female name. Not in
Detschew.
, IGBulg. III 1446, 3rd century AD.
, IGBulg. III 1074 c 12, 3rd century AD;
, III 1690 d 12, 202 AD; , III
1690 d 19, 202 AD; , III 1814, 1 - 3rd
st
century AD; IV 2074 from Pautalia, 2nd - 3rd century AD; IV 2079 from
Pautalia, 2nd - 3rd century AD: - or ; IV 1947 from
Serdica, 2nd -3rd AD; , V 5436; V 5475, 2nd - 3rd

century AD.
, IGBulg. V 5011, 1st century BC - 1st century AD.
According to Mihailov this place-name is Greek, however there is a more
plausible explanation through the so-called ‘Greek interpretation’ of
foreign (e. g. in this case Thracian) words, as it is e. g. Utlesbis about
which Dečev 39 s. v. is advancing the hypothesis of it being a corrupt
wrongly recorded version of Utesbis: ‘ja eine Dublettform des thrak.
sein kann’. Yet another possibility to explain this name is
through syncope, which occurs in many a case.2
, IGBulg. I 426 from ancient Apollonia (now Sozopol), 5th -
th
4 century BC, ethnicon to the Thracian tribe of the Apsinthioi, who
inhabited the territory north of Thracian Chersonnesos. Detschew 39
rejects the possible link of this name to the herb name , which
according to Hesychius has the form ( ) or according to
Etymologicum Magnum the form , and derives it from the root
ap- /ab- “water, stream” as in Apsus, a name for a river once on the
territory of present-day Albania. On that same ground, Detschew 39 thinks
that Apsus is a Thracian name. Mihailov, in IGBulg. I 426, is skeptical
about its Thracian origin. He had rather take it for Greek “optime Graecum
esse potest”. L. Robert considered it to be Thracian3. The Apsinthioi are
mentioned in Herodotus 6. 34, as subduing the Thracian tribe of the
Dologkoi in war. Strabo 7. frg. 58 provides information on the place name
(the land of the Apsinthioi) which in his time was changed to
( ), belonging to another Thracian tribe. The same
information is stated in Steph. Byz. 153. 8. For other occurrences see
Detschew 39. Whether Apsinthioi is the true Thracian form, or it should
be understood as a Greek version which was attested in this inscription of
the 5th - 4th century BC and corroborated by the literary evidence, will be
discussed later in Phonology. However, we have to state that this is so far
the only occurrence of the letter psi in an inscription. Detschew 39 was
not aware of this occurrence.
2
See also Duridanov (1976, 53). Perhaps our best supporting evidence comes from
the Thracian inscription from Kyolmen, see above Chapter One, as well as
Dimitrov 2003, 351.
3
Robert, L. 1959, 230, n. 1.
, IGBulg. II 761.
IGBulg. II 524.
, IGBulg. IV 2173 from Caristorum, 2nd - 3rd century AD.
, IGBulg. III 1690 c 40, 202 AD: its inhabitants
.
, IGBulg. III 1690 c 40, 202 AD: the inhabitants of
and III 1690 c 92, 202 AD.
vel , IGBulg. IV 2077, from Pautalia, 2nd - 3rd
century AD.
, IGBulg. II 714, an epithet of the Thracian Heros, 3rd
century AD.
vel , IGBulg. IV 2077, from Pautalia, 2nd - 3rd
century AD.
, epithet of Heros, IGBulg. II 706,707 from
Paskalevets, district of Tirnovo, central Bulgaria.
, in Manov 2008, 126-7, 2nd – 3rd century AD, from the
town museum of Blagoevgrad. Not in Detschew.
, in Manov 2008, 131, 3rd century AD, a male name
from Sandanski, southwest Bulgaria (see also Angelov 2003, 142). Not in
Detschew.
, IGBulg. I 48 line 7, 238 AD; II 566 from Glava Panega; II
590, 2nd - 3rd century AD; IV 1934,
from Serdica, 2nd - 3rd century AD; V 5836, 2nd - 3rd century AD;
1953, from Serdica, 2nd - 3rd century AD; IV 2078,
, from Pautalia, 2nd - 3rd century AD; Mihailov
considers this name to be either Thracian or Roman (Latin).
, IGBulg. I 51c 9, cf. II 542, 566, I 48, 133, 3rd century AD
(the latter not in Detschew). See also , i. e. “Bessus
for himself” in Marazov et al. 1996, no. 45; Kabakchieva 2000, no. 57. It
is hard to say whether Bessus is Thracian or Roman. End of third century
AD. See also G. Mihailov’s comment (Mihailov 1977, 346).
, IGBulg. IV 2350, from Novo Lyaski, on the Nestos
river, 2nd - 3rd century AD. Not in Detschew.
, IGBulg. I 440, 5th - 4th century BC. Not in Detschew.
, IGBulg. III 1488, 3rd century AD;

, III 1690 c 42, 202 AD.
, IGBulg. V 5328, 2nd - 3rd century AD. Not in
Detschew.
, a town, now Hisar north of Plovdiv, IGBulg. II 1474, 3rd
century AD.
, inhabitants of IGBulg. II 1474, 3rd century
AD.
, IGBulg. III 1516, 3rd century AD. Not in Detschew.
, IGBulg. III 1690 b 17;
, IV 2234, from Sportela (Rila), southwest Bulgaria,
2nd - 3rd century AD.
, IGBulg. III 1690 b 15, 202 AD;
, III 1690 b 19, 202 AD.
, IGBulg. III 1690 d 21,
202 AD.
IGBulg. II 512 from Glava Panega, northwest
Bulgaria, 2nd - 3rd century AD; , III 1714, 2nd -
rd
3 century AD; Detschew 165:
, cf. Detschew 439, IGBulg. I 354 quarter
, 1st century AD (?); I 334 novies, DOYTOYBEH4
from Mesambria, according to Mihailov for ?), a
female, 4th century BC; Mihailov after all is not fully persuaded for such a
conjecture. , IGBulg. II 555;
, II 694, 2nd - 3rd century AD; ,
rd
IGBulg. II 843, 3 century AD; III 1039, III
1402 – a horseman, III 1474; III 1090 b 24, b 22,
b25, 3rd AD; III 1516, whose father is ; ,
brother of III 947 line9;
III 1283, III 1398, , son of III
1005, III 1419, III 1542,
, who is a priest of the Syrian Goddess III 918,
4
See Dimitrov P. 2005, 61.
, who is a Greek magistrate III 1430,

III 1405, 3rd century AD, III
1439, III 1440, III 1441, III 881 line 8,
III 1342, III 1057, III 1496,
IGBulg. III 1465, Thracian (?) or Latin Cercius,
III 1199, 3rd century AD; , III 1350, 3rd
century AD; III 1332, III 951, III 1307,
st rd
III 1452; ,III 1559, 1 -3 century AD; , III
1627, 1st - 3rd century AD; ,III 1626, 1st – 3rd century AD;
III 1666, 222-231 AD; , III 1690 b 20, c 25, c
38 and d 42, 202 AD; III 1690 b 32, 202 AD;
, III 1690 b 34, 202 AD; ,III
1690 b 57, 202 AD; ,III 1690 b 65, 202 AD;III
1690 c 35 , 202 AD; , III 1690 c
41, 202 AD and IV 2274 from Sandanski, southwest Bulgaria, 150 AD;
, III 1690 c 22, 202 AD;
, III 1690 c 42, 202 AD; , III
1690 c 66, 202 AD; , III 1690 c 67, 202 AD;
, III 1690 c 69, 202 AD; ,
III 1690 c 70, 202 AD; III 1690 d 46, 202 AD; , III
1690 c 93, 202 AD; vel , III 1784, 1st - 3rd
century AD; , an ex-voto for
, and his wife , III
1803, 1st - 3rd century AD; , IGBulg. III 1804, 1st - 3rd
century AD; III 1833, 1st - 3rd century AD. , son of
IV 2253, from Ilindentsi, near the border with Greece, on the Nestos river,
2nd century AD(?); IV 2134 from Dolna Dikanya,
near Serdica, 2nd - 3rd century AD; IV 2233 and IV 2234,
, from Sportela, now Rila, southwest Bulgaria,
nd rd
2 - 3 century AD; IV 2311: , 198 AD; IV 2230,
, from Barakovo, near Blagoevgrad,
southwest Bulgaria, 2nd - 3rd century AD; IV 1984 from Potop, near
Serdica, 2nd - 3rd century AD; IV 2341 , (a priest?),
nd rd
from Nicopolis ad Nestum, 2 - 3 century AD; , IV
2232 from Sportela, 2nd - 3rd century AD; IV 2010 from Ivanyane, 2nd - 3rd
century AD; IV 2225 from Vaksevo, southwest Bulgaria, 2nd - 3rd century
AD; IV 2346 from Nicoplois ad Nestum,

2nd - 3rd century AD; IV 2261 from Ploski, southwest Bulgaria,
2nd - 3rd century AD; IV 2322 , from Lehovo, southwest
Bulgaria, 211 AD; , IV 2234; V 5589, 77 AD; 5861;
5883, 77 AD; , Manov not in Detschew.
, IGBulg. III 1690 c
43, 202 AD; , III 1690 c 70, 202 AD;
, V 5462, 2nd - 3rd century AD.
, IGBulg. IV 2121 epithet of Heracles, 1st - 3rd century
AD; Detschew 48 regarded it as derived either from * , a
place-name, or from * , a non-syncopated form of the tribe
name Bisaltai. The diphthong -ei- may well be a genuine Thracian one as
the diphthongs tend to keep their form despite the Greek graphemic
variants 5.
(adj.), from an inscription in ancient Pistiros. See Velkov
and Domaradzka 1994, 1-15; …. ; and
see below.
-, IGBulg. IV 2008, where the epithet is
probably , q. v., 2nd - 3rd century AD.
, IGBulg. III 1690 c66, 1st -3rd century AD.
, IGBulg. III 1344, a female dedicant to the Nymphs, 1st
nd
BC - 2 century AD.
, IGBulg. II 861, 3rd century AD.
, IGBulg. III 956, 1st - 2nd century AD.
, IGBulg. III 1455, 1472, 3rd century AD; V 5534, 211-
212 AD.
, IGBulg. III 1801, 1st - 3rd century AD; a rock
vel , IV 2019 app. cr. Not clear. Very dubious;
, a female name, IGBulg. III 1828, 1st – 3rd
century AD; (?) , from Saladinvo near Pazardjik,
Gočeva (1989, 113-115). Ed. pr. . However, see H. Plekett’s
comment in SEG 41, 1991, p. 200 at 608: he reads . Gočeva did
not date the inscription; the sanctuary being known to us by other
inscriptions, we could date this inscription to the 3rd century AD. See also
5
See Dimitrov 2007a.
, from Burdapa, III 1347, 3rd century AD. Another

graphemic variant of .6 , Kabakchieva (2000,
nd rd
121), 2 - 3 century AD. A rectangular ex-voto made by the dedicant
Bendis, wife of Dizza(s).
, wife of Seuthes III, Elvers 1994, 241-66, end of 4th century
BC, see also IGBulg. III 1731 from ancient Seuthopolis.
, IGBulg. III 1690 d 52, 202 AD.
, IGBulg. III 1690 c 17,
202 AD.
a Thracian tribe in the Rhodope Mountain region, localized
around Bessapara, as well as in the mountainous regions of ancient
Orbelos and Haimos: IGBulg. I 344, line 3, 1st AD (see Hdt. 7. 111).7 See
also on a votive stele (?) the personal name and then ,
which I believe should be restored to i. e. “Bessus for
himself”. It is hard to say whether Bessus is Thracian or Roman (Marazov
et al. 1996,116-18), end of third century BC.
: , inhabitants of Bessopara, IGBulg.
III 947, 3rd century AD.
, inhabitants of Bessopara, IGBulg. III 947, 3rd
century AD.
, IGBulg. II 856, 3rd century AD, an epithet of the Thracain
Heros; see , IGBulg. II 804, 3rd century AD and
rd
, II 855, 3 century AD.
, IGBulg. III 1542, an epithet of the Thracain Heros, 3rd
century AD.
, III 1690 b 15.
, IGBulg. III 1572 is a wrong reading for Teres, see
ad 5560=1572
, see below at .
, IGBulg. I 378, , V 5011.
(sic) (sic), IGBulg. V 5329, 3rd century AD
(?).
6
See also Masson 1988, 10.
7
Sarafov 1972, 115-20; 1974, 135-38; see Mihailov 1980, 13-18 for an opinion
different from that expressed by T. Sarafov.
IGBulg. I 12, between 16 BC and 14 AD, during the reign of

Rhoimetalkes I; II 719
, 3rd - 4th
century AD; II 769 , 1st - 3rd century AD;
, II 844 a Thracian female name;
III 1283; III 1005 , 3rd century AD; III
1057 , 3rd AD; III 1342 3rd
rd st
century AD; III 1649, 3 century AD ?; , III 1802, 1 -
3rd century AD; , IV 2322, 211 AD; , IV
st rd nd rd
2337, 1 - 3 century AD; , in Manov 2008, 126, 2 - 3 AD,
from the museum of Blagoevgrad. Not in Detschew.
-, an epithet of the Thracian Heros, IGBulg. III 1750.
, IGBulg. II 861, a Thracian male name, 3rd century AD.
(?) IGBulg. V 5883,
1st century AD. Mihailov regards it as Thracian name, but it seems that
this looks more to be the Roman name Vitulus.
, an epithet of the Thracian Horseman, IGBulg. III 1834
=V 5645.
, IGBulg. III 1516, b21, 3rd century AD.
, IGBulg. I 177ter, 185bis, 188, V 5067, 1st - 3rd century AD,
were considered by Mihailov Oriental (from Asia Minor, cf. bibliography
in volume I, p. 191) unlike Detschew 536 who thinks that it is Thracian.
However, Mihailov is of different opinion in volume V, p. 28 at 5054 and
at 5067, p. 32: “ quod nunc Thracium esse puto”; cf.
, V 5067.
: , IGBulg. IV 2216, 1st - 3rd
century AD.
, IGBulg. IV 2216, 1st - 3rd century AD.
(?) , from Saladinvo near Pazardjik, Gočeva LB
32, 113-115. However, see H. Plekett’s comment in SEG 41,1991, p. 200
at 608: he reads . Gočeva did not date the inscription; the sanctuary
being known to us by other inscriptions, we could date this inscription to
the 3rd century AD.
, an epithet of the Thracain Heros, IGBulg.II 714.
, Mihailov: ‘loco incerto in valle Strymonis medii’ IGBulg.
IV 2331, 144 AD: ‘…6. …. littera B, quae mutila est, non tamen certa est,
P?’. Cf. Detschew 79, 95.
, IGBulg. III 1690 b 19,

202 AD; III 1690 c 14, 202 AD.
, III 1690 c 69, 202 AD.
, IGBulg. III 1338, 1339 from Bourdapa, 2nd -
3rd century AD; , III 1341, 3rd
century AD; See , III 1844 below.
, IGBulg. III 1844, 1st - 3rd century AD.
, IGBulg. III 1340 , inhabitant of
Bourdopa, and , inhabitant of Bourdapa, III 1341. See
further on in Phonology about the different vocalic variation a/e, a/o, etc.
, inhabitant of Bourdopa, IGBulg. III 1340 and
, inhabitant of Bourdapa, 1341. See further on in
Phonology about the different vocalic variation a/e, a/o, etc.
, IGBulg. V 5264=II 744, 2nd - 3rd century AD, vel
, is considered now Oriental by Mihailov. Detschew 81
regards it as Thracian. We think that this is a genuine Thracian name
judging on the components / - and in other Thracian names.
, IGBulg. III 1631, 1st - 3rd century AD;
, IGBulg. III 1690 c 36, 202 AD.
st rd
, III 1599, 1 - 3 century AD; III 1690 b 48
, 3rd century AD; , III 1690
b 57, 202 AD; , III 1690 b 59, 202 AD;
, III 1690 b 60, 202 AD;
, III 1690 c 37 and c 52, 202 AD;
, III 1807, from the sanctuary in Lozen,
district of Harmanli, northwest of ancient Hadrianopolis 3rd century AD.
, IGBulg. III 1194, from Batkun, 2nd - 3rd century AD.
, a town whose inhabitants were the ,
IGBulg. III 1473, 3rd century AD.
, IGBulg. III 976, 1st - 3rd century AD;
, III 1814, 1st - 3rd century AD.
, a male name (?), III 1815, 1st - 3rd century AD.
, IGBulg. III 1343, from Burdapa, 1st BC - 2nd century AD.

, IGBulg. III 1690 b 55, 202 AD;
, III 1690 b 56, 202 AD;
, III 1690 c 68, 202 AD;
, III 1846, 1st – 3rd century AD.

, III 1593, 1st - 3rd century
AD; , III 1593, 1st - 3rd century AD;
, III 1690 b 70, 202 AD;
, III 1690 c 47, 202 AD;
, III 1690 d 28, 202 AD;
, III 1690 d 51, 202 AD;
, III 1690 e 94-95, 202 AD.
, IGBulg. III 1631, 1st – 3rd century AD.
(ς) , IGBulg. III 1355, 3rd century AD.
, IGBulg. IV 2251 from Gorna Gradesnica
near the border with Greece, 3rd century AD(?); Detschew 87
, a male.
, III 1430, 3rd century AD; Detschew 90
.
, IGBulg. V 5687.
, IGBulg. IV 2117, an epithet of Hera, 1st - 3rd century
AD.
, IGBulg. III 1690 b 13, 202 AD;
, III 1690 b 40, 202 AD; , III
1690 b 45, 202.
,
IGBulg. III 1829 inhabitant of Bosagura, 1st - 3rd century AD.
, ,
IGBulg. III 1829 inhabitant of Bosagura, 1st - 3rd century AD.
vel , IGBulg. V 5070. Not in Detschew.

, IGBulg. IV 1954 from Sofia, 2nd - 3rd century AD. Not in
Detschew.
, a dedicant to Asklepios, on a votive
plate, not far from Pautalia, present-day town of Kuystendil, southwest
Bulgaria, 3rd century AD.8 Not in Detschew.
Fig. 4-1. Inscription from Kyustendil
8
Dimitrov P. 2006, 247-251. I was able to inspect this votive plate together with
two other fragments in the spring of 2006 thanks to the courtesy of Ms. Lidia
Staykova-Alexandrova and Mr. Andrey Tonev, curators at the Regional Museum
of History in Kyustendil.
, a female name, , IGBul. IV 2349 from

Kovachevitsa near Nicopolis ad Nestum, 2nd - 3rd century AD. Not in
Detschew.
, IGBulg. III 1816, 1st - 3rd century AD, a name for the
Thracian Horseman; Detschew 100 s. v. thinks that this is the first name of
a deceased man who had been made a Heros and to whom in the aftermath
this inscription had been dedicated.
, an epithet of the Thracian Horseman as Apollo from
the sanctuary in Lozen, district of Harmanli, northwest of ancient
Hadrianopolis, IGBulg. III 1813, 3rd century AD.
, IGBulg. III 1810, an epithet of the Thracian Horseman
as Apollo from the sanctuary in Lozen, district of Harmanli, northwest of
ancient Hadrianopolis (across the border in Turkey), 3rd century AD.
, IGBulg. III 1807 and 1808, an epithet of the Thracian
Horseman as Apollo from the sanctuary in Lozen, district of Harmanli,
northwest of ancient Hadrianopolis, 3rd century AD.
, IGBulg. III 1690 d 27,
202 AD.
, IGBulg. III 1690 c 20, 202 AD. ,
st rd
, IGBulg. III 1811 inhabitant of Geupasa, 1 - 3
century AD.
, IGBulg. III 1690 d 12, 202 AD.
, IGBulg. I 14c line 22, 3rd century AD; I 47c line 26.
, IGBulg. III 1626, 1st - 3rd century AD; -, IV
2057 from Pautalia, 135 (138?) AD; Mihailov thinks this name might also
be Roman.
, from Saladinvo near Pazardjik, Gočeva, LB 32, 113-115;
the author did not date the inscription; the sanctuary being known to us by
other inscriptions, we could date this inscription to the 3rd century AD.
, IGBulg. III 1497, an epithet of the Thracain Heros, 3rd
century AD.
ῷ in the Ares Collection
1996, 43 by E. Penkova who has dated it to the 2nd century AD. As I have
seen on the ex-voto, a small upsilon is inserted in the middle of a rather
large sigma. To be sure, the first part of this compound is secure, and we
know it from other names starting with -.
, a Thracian tribe, IGBul. I 323, before 48 and around 55 BC.
, a town near the sanctuary in Lozen, district of Harmanli,

northwest of ancient Hadrianopolis, IGBulg. III 1811, 1st - 3rd century AD.
, IGBulg. III 1811, from the sanctuary in Lozen, district
of Harmanli, northwest of ancient Hadrianopolis (across the border in
Turkey), inhabitant of Geupasa, 1st - 3rd century AD.
IGBulg. III 1346, 3rd century AD;
, IV 2343 from
Nicopolis ad Nestum, 3rd century AD.
IGBulg. III 1829 inhabitant of Bosagura, 1st - 3rd
century AD.
, IGBulg. III 1811 an epithet of the Thracian Heros at
Geupasa, 1st - 3rd century AD.
, IGBulg. III 1690 b 58, 202.
IGBulg. IV1983 near Serdica, 2nd - 3rd century AD.
, IGBulg. IV 2251 from Gorna Gradeshnitsa, southwest
Bulgaria, 3rd century AD (?).
, (genitive ) IGBulg. II 501. Detschew 107:
2nd - 3rd century AD (?). Cf. II 554 the genitive
same as Detschew 107.
, vel , IGBulg. V 5264=II 744, 2nd - 3rd
century AD.
vel , inhabitants of the ancient
rd
IGBulg. IV 2236, 3 century AD.
. (. ) . (. ) III 1690 d 64,

202 AD.
IGBulg. II 868, 869, an epithet, 2nd - 3rd century AD;

, V 5385, 1st - 3rd century AD. Not
in Detschew.
, IGBulg. II 867 bis, an epithet, 2nd - 3rd century AD.
, IGBulg. II 752.
, on two silver-gilt phialae, a silver mug and a silver
rhyton from the Bashova Mogila at Duvanlij: See Chapter Three.
, a place-name, from , an ethnicon.
Velkov 1991, 26-27. See below, in
IGBulg. III 1794. Dodoparon, present-day Golyam Manastir, was situated
several kilometers south of ancient Kabyle. According to Mihailov in
1794, Dodoparon was a village with a near-by sanctuary of Apollo.
, an ethnikon, see above. Velkov 1991, 26-27.
, ( ) IGBulg. V 5180,
nd
2 century AD. Not in Detschew. cf. , IV 2153, - in
, etc.
, IGBulg. V 5328, 2nd - 3rd century AD. Not
in Detschew.
a place-name, from an ethnicon.
Velkov 1991, 26-27. See below, in
IGBulg. III 1794. Dodoparon, present day Golyam manastir, was situated
several kilometers south of ancient Kabyle. According to Mihailov in
1794, Dodoparon was a village with a near-by sanctuary of Apollo.
, IGBulg. III 1803, an ex-
voto for himself and his wife , 1st - 3rd century
AD.
, IGBulg. III 1188 (=V 5486), 3rd
century AD. Detschew 110 .
, passim. The name of this Roman province is being mentioned
in a number of Greek inscriptions in Bulgaria. Inasmuch as this is the
Latin word Dacia, its Greek variant, known from literary sources after 2nd
century AD, this not a reason to assume that the name is genuine Thracian.
However, see Detschew 111.
, IGBulg. II 853.
, IGBulg. V 5329, 3rd century AD (?); ,V

5288, 2nd - 3rd century AD. The form is not in Detschew, but see
,Detschew 114.
, IGBulg. III 1690 c 54,
202 AD.
IGBulg. III 1690 c 13, 202 AD.
, IGBulg. III 1690 d
30, 202 AD.
, IGBulg. III 1690 d 63, 202 AD;
, III 1690 d 50, 202 AD.
in (?) , from Saladinvo near
Pazardjik, Gočeva 1989, 113-15; the author did not date the inscription;
the sanctuary being known to us by other inscriptions, we could date this
inscription to the 3rd century AD. See SEG 41, 1991, 200 at 608. The
second name should be a genitive , from . Not in
Detschew.
, IGBulg. II 768, from Eski Djumaya, now Targovishte, 3rd
century AD. See also Derzelaj.
, IGBulg. III 1108, 3rd century AD, a dedicant to the Thracian
Heros .
-, IGBulg. III 1516, 3rd century AD.
, an epithet of Apollo, from ancient Karasura, Böttger –
Halloff 1991, 481-88, 3rd century AD. The editors have made mention of
the only name that comes closer, namely the place name (near
Edessa, Steph. Byz. 222, 4) and its inhabitants with a
typical suffix - although “insgesamt aber als Parallele wenig
überzeugt”, see also Detschew 122. Mihailov, relying on a photograph,
misread this epithet, see IGBulg. V 5591. See also SEG 41, 590. Not in
Detschew.
a female name, IGBulg. IV 2292,
from Laskarevo.
(sic) , IGBulg. V 5329, 3rd
century AD (?), v. . Not in Detschew.
, IGBulg. IV 2288, 165 AD.
9
See Galabov 1964, passim.
, IGBulg. IV 2292,
from Laskarevo.
, IGBulg. IV 1938, a personal name, after the 2nd century AD:
.
, IGBulg V 5918, 1st - 3rd century AD.
Not in Detschew.
III 1348, , 3rd century AD;
, III 1204, there is a spelling mistake in
Detschew 134: ; III 1452, , 3rd century
AD.
, IGBulg. III 1626, 1st - 3rd century
AD.
, IGBulg II 761, ? , a dedicator to
the Thracian Horseman from a .
, IGBulg IV 2214 line 12 from Dolistovo, southwest
Bulgaria, 2nd - 3rd century AD; Mihailov is hesitating between Greek
( ) and Thracian, adducing the examples from III 1004 and 1616; see
Detschew 141 “ ”.
,IGBulg. III 1283.
, IGBulg. IV 2291 from
Laskarevo, 2nd - 3rd century AD; Not in Detschew.
, IGBulg. IV 2231 from Kocherinovo, 2nd - 3rd century AD;
Mihailov considers – to be .
, IGBulg. V 5294, 2nd - 3rd
century AD.
, IGBulg. III 1514, 3rd century AD; Detschew 249
.
, IGBulg. IV 2321 from Zlatolist, southwest Bulgaria,
125 AD.
, a female name, IGBulg. IV 2119 from Gorna Sekirna,
district of Breznik (near Serdica), 2nd - 3rd century AD; IV
2321 from Zlatolist, southwest Bulgaria, 125 AD.
, a female
name, in Manov 2008, 136, 2nd - 3rd AD, found near Sandanski. Not in
Detschew.
, (a female name), IGBulg. IV 2312 from Vranja, 2nd

rd
- 3 century AD. Not in Detschew.
, IGBulg. III 1846, 1st

rd
– 3 century AD.
IGBulg. III 1598, 1st - 3rd
century AD.
, IGBulg. I 230 bis (genitive), one of the many
forms for the Thracian god, 2nd - 3rd century AD. See below at .
, a soldier on ex-voto, from Karanovo near ancient
Augusta Traiana, Studia in Honorem G. Mihailov (1995) 76-77; see also
SEG 45:847.
, IGBulg. I 47, 215 AD, the Great Thracian god; I 47 bis:
(genitive), 221 AD, Not in Detschew; I 48, Derzela (genitive),
238 AD, which Detschew 119 read incorrectly as and
created another Thracian name ; around 48 BC; I 230 bis
(genitive), 2nd - 3rd century AD; only one of these occurrences
is included in Detschew. See also , II 768.
, IGBulg. II 770, an epithet of the Thracain Heros, 3rd century
AD. , IGBulg. V 5329, 3rd century
AD. Not in Detschew.
, IGBulg. I 46, between 44-43B BC
and 2-3 AD on a list of priests from Odessos; II 761,
? , a dedicator to the Thracian Horseman from a
.
, an epithet of Hera, Ares Collection,121, 2nd - 3rd
century AD. A rectangular ex-voto made by the dedicant Bendis, wife of
Dizza(s).
, IGBulg. IV 1970, possible reading? Compare , in
etc.,
Corsten 1990, 178-180. The inscription is dated to the 2nd century AD.
, from Saladinvo near Pazardjik,

Gočeva LB 32, 113-115; the author did not date the inscription; the
sanctuary being known to us by other inscriptions, we could date this
inscription to the 3rd century AD.
, IGBulg. I 312, in a decree from Mesambria on a stele in the
temple of Apollo, 3rd - 2nd century BC. Father of a Thracian of the tribe of
the Astai.
-, IGBulg. V 5076.
,
, IGBulg. I 258, 2nd - 3rd century AD;
in Detschew 538.
, IGBulg. III 1516, 3rd century AD, see .
rd
IGBulg. III 1516, 3 century AD;
, III 1690 d 48, 202 AD.
, IGBulg. III 1690 b 60, 202 AD;
, III 1690 c 30, 202 AD;
, III 1690 c 44, 202 AD;
, III 1690 c 53, 202 AD and
, III 1690 d 32, 202 AD;
, III 1690 d 13, 202 AD.
, IGBulg. III 1690 b 48, , 3rd
century AD; III 1690 b 61, .
, IGBulg. IV 2104 from Pautalia, 2nd - 3rd century AD;
Mihailov saw a drawing of the copy of the inscription. See V 5781 where
is accentuated. It seems that is a nominative form instead
of an anticipated genitive.
, on a silver-gilt phiale from Rogozen, Inv. No. NIM
22304. Cf. Chapter Three.
, IGBulg. II 555.
IGBulg. V 5861, Mihailov or a compound with
-; see infra . See also , in the
Catalogue of the Ares Collection, 121, 2nd – 3rd century AD. A rectangular
ex-voto made by the dedicant Bendis, wife of Dizza(s); , in Manov
2008, 127, 2nd – 3rd AD, from the museum of Blagoevgrad. Not in
Detschew; , in Manov 2008, 136, 2nd - 3rd AD, found

near Sandanski. Not in Detschew.
vel IGBulg. III 1784,
vel , IGBulg. III 1784, 1st
- 3rd century AD; , IV 2134 from Dolna Dikanya,
near Serdica, 2nd - 3rd century AD; Detschew 132 wrongly put
IV 2226 from Frolos (?), 2nd - 3rd AD: The letter Z is
inverted.
, IGBulg IV 2116, , 3rd
century AD, from Breznishki Izvor, near Breznik in the district of Sofia.
, IGBulg. II 483; 753; , III 1116, 3rd
century AD; , son of , III 1351, 3rd century
AD; IV 2338, lines 1 and 14 ,
st
end of 1 century AD; IV 2341, , (a priest?), from
Nicopolis ad Nestum, 2nd - 3rd century AD; , 5826, 2nd - 3rd
century AD.
, IGBulg. III 1690 d 39, 202
AD.
, IGBulg. V
5380, from Marcianopolis, 2nd - 3rd century AD.
, , IGBulg. II 511 from Glava Panega,
2nd - 3rd century AD; II 523, from Glava Panega, a
dedication to Asklepios; II 560 from Glava Panega: ;
IGBulg. III 917, 3rd AD; III 1005; III 1010; III 1106,
3rd AD; III 1166, 3rd
rd
century AD; III 1175, 3 century AD;
III 1201 3rd century AD; III 1305, 3rd century AD; III
1310, 3rd century AD; III 1348, (genitive), 3rd century
AD; , III 1782, 1st - 3rd century AD; IV 2104
from ancient Pautalia , 2nd - 3rd century AD; IV 1922, 3rd
century AD; IV 1927; , IV 1938; , IV 2015,
a priest; , from Sportela, IV 2232, 2nd - 3rd century AD;
, IV 2067, from Pautalia (Kjustendil), 2nd - 3rd
century AD; , IV 2004, from Gorni Lozen, after
2nd century AD; , IV 2011, near Serdica, after 2nd
century AD; IV 2027 from Opitsvet, area of Sofia, ,

3rd century AD, the Z is inverted; IV 2130, 3rd century AD;
IV 1927, 3rd century AD; from Lehovo, southwest
Bulgaria, IV 2322, 211AD; ( ) , V 5858. On a
fragment of a stele found in Mezdra, northwest Bulgaria, published by S.
Dremsizova-Nelchinova (1999), there is a possibly a dedicant, 3rd
century AD. See also SEG 49, 1999, no. 896.
, IGBulg. V 5825, 1st -
rd
3 century AD.
IGBulg. I 27bis, from Dionysopolis, father of
quod vide infra, 3rd century B. C. (?). Not in Detschew.
(- ) in the Ares Collection,121, 2nd - 3rd
century AD. A rectangular ex-voto made by the dedicant Bendis, wife of
Dizza(s). Not in Detschew.
a soldier, II 541 from Glava Panega, 3rd century AD.
vel [ ] IGBulg. IV 1942 from Serdica, 1st
century AD?
[ ] vel IGBulg. IV 1942 from Serdica, 1st
century AD?
, IGBulg. V 5328, 3rd century AD. Not
in Detschew.
IGBulg. II 709. Not in
Detschew.
IGBulg. II 709. Not in
Detschew.
IGBulg. V 5281, 1st - 3rd century
, a physician, (genitive), IGBulg. III 1776, 1st - 3rd
IGBulg. II 800. Not in Detschew.
, , IGBulg. IV 2067, from Pautalia
(Kjustendil), 2nd - 3rd century AD. Not in Detschew.
(?) for in , IGBulg. II 738 2nd -
rd
3 century AD; , IV 1955 from Serdica, seems to be of pre-
Roman times (2nd - 3rd BC?). Not in Detschew.
IGBulg. II 738 2nd - 3rd century AD. In this case
Mihailov prefers the reading for Not in Detschew.
, IGBulg. II 737, 2nd - 3rd century AD.

There is an inscription in IPE IV № 432; Catalogus agonisticus
Pars B, Columna I, 23.
[
Latyshev dated it to the first half of the 3rd century BC “[…esse
referendum satis mihi videor demonstrasse in commentario prioribus
editionibus addito]”.
See also CIB, 670-676, № 1137, Б 1. 23:
This is on a marble stele, inscribed on both
sides in two columns, divided by a vertical line. Found in Anapa
(Gorgipia). According to Dumberg and Skorpil, the inscription was made
by two different cutters and probably in different periods. It is a list of
athletes, winners in competitions. Latyshev believes that there was only
one cutter to execute the inscription (on account of a decree from Gorgipia
to commemorate the winners by inscribing their names on stelae, then to
add other winners’ names, etc.) J. and L. Robert are of a slightly different
opinion, i.e. that those winners were students in the local gymnasion (cf.
CIB). The inscription is said to be dated in the 3rd century BC because of
the genitive in –ou of names in –es. See also in V 5827, which
belongs to another dedicant (of two), and is another form of
(see the respective lemmata). It is to be noted
here that theta stands for the genuine Thracian /d/ from *dh-, and is a good
example how the Thracian /d/ was pronounced differently from what was
perceived by the Greeks: therefore, voiced stops in Thracian are treatments
of IE *dh-, etc. Chronologically speaking, the range of different forms
gives the following perspective of evolution in Thracian: /d/ from *dh and
/t/ from *d. 10
, IGBulg. II 730, 218-222 AD or 222-235 AD.
IGBulg. IV 2121 from Rasnik, near Serdica, 2nd - 3rd
century AD.
IGBulg. V 5282. Not in Detschew.
, IGBulg. IV 2221 from Kamenik near Pautalia, 2nd - 3rd
century AD; , III 1593, 1st - 3rd century AD.
IGBulg. III 1392. Not in Detschew.
, IV 2214 line 12 from Dolistovo, 2nd - 3rd century
AD; Mihailov hesitates between Greek ( ) and Thracian, adducing the
10
On these developments see also Dečev 1960, 152-55.
examples from III 1004 and 1616. See also Detschew 141
“ ”, which is a wrong reading.
, IGBul. I 50, b line19, 4th - 3rd century BC; II 796,
; II 843 ;
III 1073, 3rd century AD; III 1190
3rd century AD; III 1197, 1371; , III 1398, III 1520;
III 1690 column b line 21, 202 AD; IV 2014
from Gurmazovo near Serdica, 2nd - 3rd century AD;
(dative), V 5282, 1st - 3rd century AD; V 5309,
in both V 5577, 5578, found together in Augusta Traiana, first
half of the 1st century AD; V 5592, first half of the 1st century AD,
V 5634, 1st - 3rd century AD.
IGBulg. V 5011, 1st century BC – 1st
century AD; see also in the catalog of the names found on
phialae supra and a parallel from Gorgipia on the Black Sea
coast.
, a female name from Bithynia, on a tomb stele found in
Hamzabey Köy near Inegöl. The inscription has been published by
Corsten 1990, 264.
, IGBulg. III 1616c, Dubious whether Greek or Thracian. 3rd
century AD. Compare in III 1006.
IGBulg. IV 2286 159 AD.
* where the were, IGBulg. V 5589. Not in
Detschew.
, IGBulg. V 5589, near Augusta Traiana, inhabitants of
* . Not in Detschew.
, III 1690 d 17, 202 AD.
IGBulg. III 1404 3rd century AD;
Detschew 142 “richtiger ( )”, however the ending is - as
Mihailov put it;11 V 5858.
from Kresna, southwest Bulgaria (now in a private
collection), Ivanov 2004, 83-86.
, , III 1690 d 20, 202 AD.
11
Mihailov 1943, 99-101.
III 1690 d 44, 202 AD.

IGBulg. V 5281, 1st – 2nd century AD. Not in Detschew.
Mihailov suggested that it be compared with Diurpaneus in Tacitus quoted
by Oros. 7. 10. 4. See also Detschew 150.
In we see clearly two elements:
(1) -
(2) -
The first element is easily identifiable as there are a number of
personal names beginning with - - -.
The second element presents itself in a very unusual form, which is
partly due to the presumably genitive ending12 in and the
homophonic beginning of - . Both ending and beginning sharing
the same graphemic configuration -ou, make us assume namely that the -
ou in – could also be either /o/ or /u/. While the treatment of the
Indo-European syllabic resonants in Thracian is said13 to be ur/or, ul/ol,
um, and un/in, we could posit a possible reconstructed form *r°p, from
which we could get to urp/orp. There are just two sequences that
exemplify the same configuration: Diurpagissa, CIL 3. 14547 (“cum
Diurpagissa filio”) from Moesia Superior, and Diurpaneus, a leader of the
Dacians, quoted in Tacitus (apud Oros. 7. 10. 4).
Yet, there is another possible treatment of the syllabic resonant: to be
substituted to l, as there are examples like Oroles: Oloros, Aloros/Arulos.
There is also a male name, IGBulg. III 1324, 3rd century AD,
from Varvara, near Pazardjik. On the ground of similar developments in
/ we could posit
IGBulg. II 830 3rd - 2nd century BC see , II

856, 2nd - 3rd AD.
see the controversial readings in Angelov 2003, 138-142
and Sharankov 2004, 200. Manov (2008, 131) accepts Sharankov’s
reading. On the photograph Manov offered, one can read
12
For the formation of personal names in Thracian see Georgiev 1983, 1155;
Dimitrov 1994; Boïadjiev 2000, 145- 50; e. g. Dadaleme vs. Kotyos on metal
vessels. See the above-mentioned instances , and , etc.,
from Chapter Three (Mihailov, 1987, 5-19); Brixhe 1994, 186, n. 33.
13
Dečev 1960, 161; Georgiev 1983, 1156, 1167.
restored to “ (?)”. This funerary stele is now in the

museum of the town of Sandanski. Not in Detschew.
, on an ex-voto made by a soldier to the Thracian Heros-
Rider, from Karanovo near ancient Augusta Traiana, Studia in Honorem
G. Mihailov (1995) 76-77; see also SEG 45:847. Probably to be dated
around the beginning of the 3rd AD, but not before 212 AD. This epithet of
the Heros appears for the first time. Not in Detschew. See and compare
a personal name.
a name of a town, IGBulg. II 727, 2nd century
AD; II 728, end of the 2nd century AD; II 730, II 731, II 732, the
beginning of 3rd century AD; II 733, end of the 3rd century AD; II 734, end
of the 3rd century AD; V. 5257:
IGBulg. I 438, 5th - 4th century
BC. This same name has another genitive case reading in IV
2286. Not in Detschew.
, on a silver
phiale from Rogozen (see Chapter Three), Inv. No. NIM 22329. Not in
Detschew.
, IGBulg. III 1690 b 54, 202
AD.
IGBulg. g III 1440 3rd century
IGBulg. III 1516 3rd century AD.
vel IGBulg. III 1516, 1517 3rd century AD; III 1317,
rd
3 century AD; III 1005 3rd century AD;
III 1626, 1st - 3rd century AD; III
st rd
1626, 1 – 3 century AD; III 1626, 1st - 3rd century
AD; , III 1690 b 14, 202 AD;
, III 1690 c 18, 202 AD; , III
1690 c 68, 202 AD; , III 1690 c
71, 202 AD; , III 1690 d 14, 202 AD; III 1690 d 16,
202 AD; , III 1690 d 17, 202 AD;
, III 1690 d 19, 202 AD;
, III 1690 d 29 and d 58, 202 AD;
III 1690 d 47, 202 AD; III 1690
d 49, 202 AD; III 1690 d 57, 202 AD;
, III 1709, 1st - 3rd century AD;
III, 1741, 1st - 3rd century AD;

IV 2286 159 AD; according to Mihailov is a
female name. V 5611; V 5827, 3rd
century AD; vel IV 2002, 1st - 3rd
century AD, between Kadijn most and Liljac; V 5611
, 2nd - 3rd century AD. (?) V 5827, 3rd
century AD, Not in Detschew. Gerasimova-Tomova reads =
Latin Atiles,14 but Mihailov second-guessed the reading of under
question.15 After inspecting the photograph, it is quite clear that is
the correct reading. The other name , which belongs to another
dedicant, is another form of (see the
respective lemmata).
, in IGBulg. V 5577,
first half of 1st century AD.
Not in Detschew.
IGBulg. II 675 see , II 724
2 - 3 century AD; III 1517, 3rd century AD.
nd rd
IGBulg. III 965, 966 (both from the immediate vicinity of

Plovdiv), III 1209 (?) from Batkun in the Asclepios’ sanctuary. In all three
occurrences the name is written . Pre-Roman times? III 1593,
st rd
1 - 3 century AD; III 1690 b 66 from Pizos.
vel , IGBulg. III 1767, 3rd century AD. Not
in Detschew.
IGBulg. IV 2014, from Gurmazovo near Serdica,
2nd -3rd century AD; according to Mihailov, the same person made a
dedication in III 965, 966 (both from the immediate vicinity of Plovdiv,
1209 (?) from Batkun in the Asclepios’ sanctuary. All three occurances as
. Pre-Roman times? , a female name,
IGBulg. III 1701, 2nd - 3rd century AD.
IGBulg. II, 771 “
”,1st – 3rd century AD.
IGBulg. II 724, 2nd - 3rd century
AD, see II 675
, IGBulg. III 1756, an epithet of Apollo, 3rd century AD.
14
Gerasimova-Tomova 1980, 69, no. 97; 115, no. 97.
15
SEG 30, 1980, 769.
IGBulg. III, 1759, 1st - 3rd century AD. Not in

Detschew.
, IGBulg. II 868, from Abritus, 2nd - 3rd century AD.
vel , IGBulg. IV 2202, 2286; III 1876, 1st –
rd
3 century AD.
IGBulg. I 168 from Odessos, wife of
th rd
and daughter off 4 - 3 century B. C. (?).
vell IGBulg. I 334 novies, b1, 4th century BC
(?). The reading is not certain. See for a different reading of this
line as ( ). Not in Detschew. See also
IGBulg. III 1464 3rd century AD. Not in Detschew.
IGBulg. II 744, a name of a Thracian’s wife, 2nd - 3rd
century AD.
IGBulg. g I 14 b 19, father of Aurelius Victor, after 212
, son of Moukaporis from Pizos, IGBulg. III 1690 c 32, 202
AD.
IGBulg. IV 1929:… […, from
Serdica, 2nd - 3rd century AD. Not in Detschew.
, a military district (strategia) IGBulg. IV 2338, end of
the 1st century AD. This name seems to be a compound made of *
and .
in both V 5577,
and V
5578, found together in Augusta Traiana, first half of
the 1st century AD. Not in Detschew.
, IGBulg. III 1690 c
47, 202 AD; III 1690 d 24, 202 AD;
IGBulg. III 1690 e 94-95, 202 AD.
, III 1690 d 20, 202 AD;
III 1690 d 57, 202 AD.
16
For a different opinion see Boyadzhiev 2002a.
, IGBulg. V 5162 from Serdar, northwest Bulgaria. On the

stone: ; V 5736, 5746 from Slivnitsa in the sanctuary of
Asclepius Limenos, 3rd century AD. Not in Detschew.
IGBulg. III 1690 column b line
16; III 1805, 1st - 3rd century AD; V 5462, 2nd -
3rd century AD.
(o) , IGBulg. IV 2192 in an inscription of a collegium
p
at a temple of Asklepios, 1st - 3rd century AD; IV 2336, pre-Roman (?).
Not in Detschew.
IGBulg. II 856, 2nd - 3rd
century AD.
IGBulg. III 1210 2nd - 3rd century
from Kazanlak, on a silver phiale, 6th - 5th century BC. See
Chapter Three. See also in V 5827, which belongs to another
dedicant (of two), and is another form of (see
the respective lemmata).
, IGBulg. III 1809 inhabitant of * , 1st
rd
- 3 century AD.
, IGBulg. III 1809 inhabitant of * , 1st - 3rd
century AD.
, IGBulg. III 1846, 1st -

rd
3 century AD.
IGBulg. III 1516, 3rd century
AD; , III 1690 c 17, 202 AD;
vel – III 1690 d 36-37, 202 AD.
, IGBulg. III 1690 b 58,
202 AD; , III 1690 c 22, 202 AD;
, III 1690 c 43, 202 AD;
, III 1690 c 45 and d 61, 202 AD;
, III 1690 c 49, 202 AD;
III 1690 c 56, 202 AD;
III 1690 c 58, 202 AD; III 1690 c 65, 202
AD; , III 1690 c 71, 202 AD;
, III 1690 d 13, 202 AD;

III1690 d 63-64, 202 AD.
, in IGBulg. III 1794. Dodoparon,
present day Golyam Manastir, was situated several kilometers south of
ancient Kabyle. According to Mihailov in 1794, Dodoparon was a village
with a sanctuary of Apollo nearby.
This epigram might be dated to the 2nd - 1st century BC.
in Detschew 161. See and compare above ,
an adjective from Dadopara, i. e. an inhabitant of Dadopara (near ancient
Kabyle) an epithet of the Thracain Heros. Velkov 1991, 7-53.
, IGBulg. III 967=III 1893, 3rd century AD.
wife off IGBulg. 738, 2nd - 3rd
, from the inscription of Kyolmen, a personal name. See

Chapter One.17
, IGBulg. III 1690 b 41, 202 AD.
, son of Seuthes III and Berenike, IGBulg. III 1731 and
Elvers 1994, 241-66, end of 4th century BC. This occurrence is Not in
Detschew. Compare the legends on coins, as well, in Chapter Two, # 44-1,
44-2, 45-48 ( ) 387-383 BC. [ a
nominative singular, in Manov 2008, 126, 2nd – 3rd AD, now in the
museum of Blagoevgrad. Not in Detschew.
IGBulg. II 538 from Glava Panega,
2nd century BC (?).
( ) 387-383 BC in Chapter Two. Legends from
Coins.
IGBulg. IV 2104, 1st - 3rd AD; IV
st rd
2229, 1 - 3 century AD; , IV 2338, end
of the 1st century AD, and IV 2338, end of the 1st
century AD; IV 2343 from Nicopolis ad
Nestum, 3rd century AD.
17
Dimitrov 2003.
(. ) . (?), Uncertain reading IGBulg. III 974, 3rd century AD;

Detschew 178 ;
, IGBulg. V 5328, 3rd century AD; Mihailov recalls
from II 719, which “non est legendum , ut in commentario
Decevum sequens scripsi: non Graecum, sed Thracium esse nunc
apparet”. On the account of the system of Thracian diphthongs, the name
is regarded as genuine Thracian, as Thracian diphthongs were preserved in
inscriptions as late as the 4th century AD, see Dimitrov 2007a. Not in
Detschew.
IGBulg. V 5380, from
Marcianopolis, 2nd - 3rd century AD Not in Detschew.
IGBulg. V 5380 from
Marcianopolis, 2nd - 3rd century AD Not in Detschew.
, IGBulg. V 5281,1st - 3rd century AD, epithet of the
Thracian Heros. Not in Detschew.
IGBulg. II 804 (sic),
2nd - 3rd century AD; II 848.
, IGBulg. I 357, 358, ann epithet of the Thracian Heros, which
in other inscriptions appears as an epithet of Apollo, 2nd - 3rd century AD.
IGBulg. II 858.
IGBulg. II 771, an epithet of the Thracian Heros, 3rd
century AD.
IGBulg. V 5060, 3rd century AD; on the account of the system
of Thracian diphthongs, the name is regarded as genuine Thracian, as
Thracian diphthongs were preserved in inscriptions as late as the 4th
century AD, see Dimitrov 2007a. Not in Detschew.
, IGBulg II 553 from Glava Panega: Mihailov “fortasse
”. The inscription reads:
etc. See Detschew 167:
“Thrakische Aussprache des griech. .” Regardless of the fact
that seems to be a Greek word, the evidence is still extremely
important for us because it brings up the question with the genuine
Thracian pronunciation, but also that with the translation of a common
epithet for the one of many names the Thracian people called their god, the
Horseman. Therefore, this adjective should be translated as “the one God
on horseback”. As it seems to me, there is an obvious connection between
this name and Ez/sbenis, the latter derived from an etymon for “horse” in
Thracian.
IGBulg. IV 2103. The reading is dubious. Not in
Detschew.
, IGBulg. III 1690 b
62, quod vide infra. Not in Detschew.
, IGBulg. III 1517 line 39, 3rd century AD; III 1302
rd
, 3 century AD; III 1342 3rd
century AD; III 1690 b 55, 202 AD; and I
378, 19 AD, , son of
Rhoimetalkes II, who was a strategos; , I 77 from the
ancient Odessos, present-day Varna on the Black Sea coast, a name of a
priest, see also ; I 281 bis, as a second name of a wife of a
Thracian, 2 -3 century AD; IV 2069, 1st –3rd century AD (?); IV 2031;
nd rd
IV 2232, 1st - 3rd century AD.

? IGBulg. III 1593, 1st - 3rd century AD.
Obviously .
( ) , IGBulg. III 1739 from Viden, district of Kazanlak.
, IGBulg. III 1302 from Patalenica near Batkun, 3rd
century AD; III, 1739, 1st - 3rd century AD; IV 2116,
, 1st - 3rd century AD.
rd
AD III 1255 (?) 3 century AD. Not in Detschew.
IGBulg. II 744, 2nd - 3rd century AD.
, a female name, IGBulg. IV 2325, 1st - 3rd century AD.
IGBulg. II 836, 3rd
century AD; II 847, 2nd - 3rd century AD; III
974 III 1166 3rd century AD.
IGBulg. IV 2214 1st - 3rd century AD.
, IGBulg. IV 2234, from Sportela (Rila,
southwest Bulgaria), 2nd - 3rd century AD. In this same inscription
Epteikenthos is mentioned twice in line 6 and line 7. While in line 6 the
first four letters are under some doubt as to their upper halves but were
read clearly as EPTE by the editor G. Mihailov, line 7 has Ep(t)eikenthou
thus confirming beyond any doubt the reading of this name on line 6.
Nevertheless, Mihailov considered the latter not a mistake of the workman
(‘quadratarius’), but is of the opinion that it reflects a linguistic peculiarity
(“de phaenomeno linguistico agi videtur”). Detschew 167 followed the

reading of Ernst Kalinka (1906, 214) reads [Et]eikenthos on line 6, and
Eteikenthou on line 7. Not in Detschew.
, IGBulg. III 1342 from ancient Burdapa, 3rd century
AD Detschew 167: “ ( ) aus
Saladinovo” relying on V. Dobruski, 1897, 123 n. 1 fig. 1.; IV 1922 on a
column inscribed with names (a sacred catalog?); IV 2072, 1st - 3rd century
AD; IV 2214, 1st - 3rd century AD.
IGBulg. II 738, 2nd - 3rd century
( ) IGBulg. III 1335, 3rd century AD. Not in Detschew.
IGBulg. III 1341, 3rd
, IGBulg.g III 1690 b 35, 202; III
1690 b 38, 202 AD.
III 1057 3rd century y AD; III 1097 3rd
st rd
century AD; IV 2337, 1 - 3 century AD.
, , IGBulg. IV 1963 from ancient
Serdica now Sofia, after 2nd century AD.
, IGBulg. III 1733 is a second name of a wife of a
Thracian.
IGBulg. g IV 1959 (?); IV 2232,
IGBulg. IV 2153, 1st - 3rd
century AD.
, IGBulg. III 1690 b 51, 202 AD;
III 1690 c 58, 202 AD;
vel , III 1690 d 23, 202 AD;
III 1690 c 93, 202 AD.
, a female name, IGBulg. IV 1959.
III 1803, 1st-3rd century AD; III
1803, an ex-voto for , and his wife
, a female name, IGBulg. III

1701, 2nd - 3rd century AD; IGBulg. V 5475, 2nd -
3rd century AD.
on an ex-voto
dedicated to Asklepios for his daughter , IGBulg. III 1831, 1st
- 3rd century AD. The reading belongs to the editor princeps Katsarov
(1937, 284, n. 2 fig. 223; also J. et L. Robert Bull. Ep., 480 n. 214).
IGBulg. III 1166 3rd century
st rd
AD; -? III 1626, 1 - 3 century AD;
st rd
-? III 1626, 1 - 3 century AD;
III 1690 b 31, 3rd century AD;
III 1690 column b line 34, 202 AD;
, III 1690 b 46, from Pizos, 202 AD;
, III 1690 d 17, 202 AD;
, III 1690 d 21, 202 AD.
, IGBulg. III 1593, inhabitants of (see Chapter
Three), 1st - 3rd century AD.
III 1690 c 59, 202 AD.
Detschew 412 gave a different reading.
, see Chapter Three, on a
silver phiale, Inv. No. NIM 22344. For the variation Argiske/ Ergiske see
further on in Phonology. See also Dimitrov 1994, 86. Not in Detschew.
IGBulg. III, 1714 2nd - 3rd
century AD; IV 2343
from Nicopolis ad Nestum, 3rd century AD.
vel vel , III 1690 d 23,
202 AD. Not in Detschew.
, an epithet of Apollo, IGBulg. III 1766, 3rd century
AD.
vel , IGBulg. V 5070. Not in Detschew.

, IGBulg. V 5918, 1st - 3rd century AD, cf. and -
. Not in Detschew.
a female name, IGBulg. IV 2303, 146 AD; IV 2311, 198 AD.
a female name, sister of a Roman soldier, Angelov 2003,

138. See also Sharankov 2004, 200 in his Corrigenda accentuating the
name as . This name is obviously Thracian, so no need of
accentuating. Not in Detschew.
vel IGBulg. I 130, 2nd - 3rd century AD, wife of
Flavius. Not in Detschew.
a female name, (dat.), IGBulg. IV
2112, 1st - 3rd century AD.
, IGBulg. III 1743 epithet of Apollo Heros. See also
. Not in Detschew.
, IGBulg. III 960, 961, an epithet of the Thracian Heros, 3rd
century AD. Not in Detschew. ,
, an epithet of Zeus, IGBulg. II 762 from Kardam,
northeast Bulgaria, 2nd - 3rd century y AD; III 1773, III 1886, 1st – 3rd
rd
century AD; V 5612, 3 century AD.
, IGBulg. IV 2216-2218, 1st-3rd century AD and on an
ex-voto from Kyustendil, southwest Bulgaria,18 epithet of Zeus (Not in
Detschew); on an ex-voto from Cabyle, 3rd AD, Velkov
1991, 24-25, Not in Detschew.
18
two other fragments in the spring of 2006, thanks to the courtesy of Ms. Lidia
Fig. 4-2. The Votive Plate from Kyustendil
, IGBulg. III 1656, 3rd century AD or later. In Detschew 427

slightly different form ending in –EI, see , IGBulg. III 1654,
3rd century AD (or later), III 1655.
( ) , a place-name, connected to below.
in
n IGBulg. III 1457, 3rd century AD; III
rd
1458 , 3 AD; III 1459; III 1460, III 1461, an
epithet of Apollo (Heros), 3rd AD.
IGBulg. II 506, wrongly in Detschew 178 ; see also
in that same inscription. , II 506 from Kunino,
northeast Bulgaria. Detschew 178: . , V 5067.
, IGBulg. IV 2175, after L. Robert (Revue de philologie ‘Les
inscriptions grecques de Bulgarie’, 33, 1959, 180, n. 5). Detschew 178
; see also Dimitrov 2007a for a full account on and
similia.
, IGBulg. IV 2048 from Krasno selo, a town of Sofia.
Not in Detschew.
IGBulg. III 1516, 3rd century AD. Not in Detschew.
IGBulg. III 973, a place name, see below. Not in

Detschew.
, an epithet of the Thracian Heros, IGBulg. III 973,
3rd century AD. Not in Detschew.
IGBulg. I 184, 2nd - 3rd century AD.
, IGBulg. IV 2336, 2nd - 3rd century AD.
, IGBulg. III 1516 a line 28 from ancient Cillae. Not in
Detschew.
, IGBulg. II 506 from Kunino, northeast Bulgaria.
Detschew 178: .
II 724 2nd - 3rd century AD.
, IGBulg. IV 2048 from Krasno selo, a town of Sofia.
Not in Detschew.
, IGBulg. II 763 from Popovo, northeast Bulgaria, 2nd - 3rd
century AD; III 1147 3rd century AD.
from Piperitsa, district of Blagoevgrad, in
Manov 2008, 108. Not in Detschew.
IGBulg. IV 2294, from Laskarevo, southwest Bulgaria, 3rd
IGBulg. IV 2286, 159 AD. Not in Detschew.
, IGBulg. IV 1930, ann epithet of Apollo, from Serdica, after
2nd century AD.
, a female name, together with a well-known
male name, in Manov 2008, 125, 3rd century AD, district of Blagoevgrad.
On a funerary marble stele, perfectly visible appear a bearded man and a
woman, and the above mentioned inscription was clearly carved and easy
to read. Though the name of the female person resembles the Latin Celsus,
there are two features that might deem convincing as to its Thracian
origin. First, the ending –us in Latin is rendered with -os in Greek, resp.
Thracian, and second, the ending –ous is utterly convincing as belonging
to a rare Thracian class of names. See Kortozous below. Not in Detschew.
( ) , IGBulg. III 1335, an epithet of the Heros, 3rd century
, IGBulg. V 5360 from ancient Marcianopolis,
present-day Devnya. Not in Detschew.
AD.
, IGBulg. III 1742, 1744, 1745 epithet of Apollo Heros; see

, III 1743, epithet of Apollo Heros. Not in Detschew.
, IGBulg. IV 2153, from Caristorum,
1st -3rd century AD. Detschew 182: .
(. ) , IGBulg. III 1473, 3rd century AD. Not
in Detschew.
a personal name on a silver phiale from Kazanlak, Dimitrov
(1995, 23-25). See Chapter Three. VIII. Not in Detschew.
, IGBulg. V 5703. Not in Detschew.
, in an inscription from Nikaia, No1308 in Şahin (1982).
, IGBulg. III 976, 1st - 3rd
century AD.
, a wife, IGBulg. III 1787.
, a Thracian priest, in IGBulg. II 719 from theregion of
Tirnovo:
3rd - 4th century AD. in Detschew 187.
, IGBulg. IV 2211, 1st - 3rd century AD.
IGBulg. I 47 c line 29, 221 AD.
, IGBulg. IV 2289, 1st - 3rd century AD. See also Boyadzhiev
2002, 79-81.
, IGBulg. V 5269. Not in
Detschew.
, IGBulg. II 711, 2nd - 3rd century AD.
, III 1201, 3rd century AD.
, IGBulg. II 673, 2 - 3rd century AD.
nd
IGBulg. II 843, 3rd century AD. Not in

Detschew.
, IGBulg. III 1893 = IGBulg. III 967.
, IGBulg. III 967, 3rd century
AD, should be corrected to see
IGBulg. III 1893.
, IGBulg. II 872.
, IGBulg. V 5067, 3rd century AD. Not in Detschew.
(v. , IGBulg. V 5067).
, IGBulg. I 171bis, 2nd - 3rd century AD, wife of some

Greek. Not in Detschew.
, IGBulg. I 479 ter, 2nd - 3rd century AD, a Thracian female
name.
, IGBulg. V 5061, 1st - 3rd AD; this name could also be of
Asia Minor origin, see Zgusta KPN, §§393-98 or belonging to Asia Minor
people that have moved to Thrace. Not in Detschew.
, IGBulg. V 5319. Uncertain reading. Name of a dedicant? See
Detschew 194.
, III 1690 b26 , 3rd AD;
IGBulg. V5326(=II 790), 2nd - 3rd century AD.
, , IGBulg. V 5287, 2nd - 3rd
, IGBulg. I 162, 2nd - 3rd century AD A Thracian male. See
IGBulg. III 1690 b 26
, IGBulg. II 765, 2nd - 3rd century AD
, IGBulg. II 738, a name of the daughter of
2nd - 3rd century AD, = - Not in
Detschew.
, IGBulg. I 24, an epithet of Herakles, 3rd century AD.
Not in Detschew.
IGBulg. IV 2338. Detschew 177:
, IGBulg. III 1108, 3rd century AD, a dedicator to
the Thracian Heros .
A few words should be said here before we proceed with the numerous
epithets of Asklepios found in the sanctuary of Batkun by Dimiter
Tsontschev.19 As G. Mihailov has stated (p. 117, IGBulg III), a templum
of Asklepios was found near the village of Batkun, where many
monuments of different kind were discovered, and amongst them
inscriptions on votive reliefs in Greek bearing the local epithet of
Asklepios, as the god was worshipped there as Zymdrenos, Zumudrenos
etc. (see below). Mihailov (l. c.) thinks that the local name of that place
was *Zim(i)dra and not Sindrina as Tsontchev and others thought.20
19
Tsontchev 1941.
20
Tsontchev 1941, 12.
Before Tsontchev started his excavations, in many places around Batkun,

e. g. Patalenitsa, Lyahovo, Varvara, and in the monastery of St. Peter and
Pavel (Paul), many votive monuments were found.
In his IGBulg. Mihailov made a number of amendments to the texts of
the inscriptions but kept most of Tsontchev’s reading as far as our lemma
is concerned.
, a place-name, Tsontchev (=Cončev) 40 and IGBulg. III
see below.
, IGBulg. III 1178 , 1239
Tsontchev 63, an epithet of Heros, 3rd century AD.
IGBulg. III 1184, Tsontchev 42. Detschew 196
added a rho - On the stone there is no rho. A nicely
carved inscription in stoichedon. The zeta is I, not Z. This occurs often
times in the inscriptions from that sanctuary. 1st - 3rd century AD.
, an epithet of Asklepios IGBulg. III 1134, Tsontchev
33, 1136, Tsontchev 34, 1194, Tsontchev 52, 1226, Tsontchev 46, 1230,
on the stone, Tsontchev 54 and Detschew 196 without
zeta, 1236 (Tsontchev 62), all from Batkun, 2nd - 3rd century AD.
Tsontchev 19 and IGBulg. III 1115, 1187
(Tsontchev 44) (a dedication of Seuthes, son of Kotys), 1228, (Tsontchev
61), 1232 (Tsontchev 47); 3rd century AD; an epithet of the Heros.
Tsontchev 59 and IGBulg. III 1203, 2nd - 3rd century
AD.
Tsontchev 38 and IGBulg. III 1174, 2nd - 3rd century
AD.
Tsontchev 46 and IGBulg. III 1189, 3rd century AD.
Tsontchev 20; Mihailov added a small epsilon, which
the carver added himself, IGBulg. III 1120, 3rd century AD.
Tsontchev 18 and IGBulg. III 1117, 1121 (Tsontchev
20), 1191 (Tsontchev 48), 1225 (Tsontchev 54), 1234 (Tsontchev 56); all
from Batkun, 2nd - 3rd century AD.
(Tsontchev 26 ), IGBulg. III 1153,
2nd -3rd century AD.
, Tsontchev 40 and IGBulg. III 1178, 1239
( Tsontchev 63), an epithet of Asklepios (Heros), 3rd
century AD.
IGBulg. III 1119 ( Tsontchev 19),

1133, 1137-1140, 1161, 1163, 1171, 1177, 1186 , same
style of writing as in 1184, 1195, 1224, 1227, 1229, 1233, 1237, 1238,
1277 ( ; 3rd century AD.
, IGBulg. III 1286 ( ) Mihailov, epithet of
Asclepios (Heros), 3rd century AD.
IGBulg. III 1126, 1128, 1135 (Tsontchev 34
nd rd
); 2 - 3 century AD.
IGBulg. III 1162 ( ), 3rd century AD.
(Tsontchev 56) and IGBulg. III 1193,
2nd - 3rd century AD.
Tsontchev 32 and III 1132, 2nd - 3rd century AD, an
epithet of Hygia and Telesphor, 3rd century AD.
, IGBulg. III 1112, an epithet of Hera 2nd - 3rd century AD.

Detschew 197: “unsichere Lesung”. Mihailov is not pointing out to any
difficulty with the reading. This inscription was found in Ellidere, now
Vetren Dol, near Asclepios’ sanctuary in Batkun.
, see Chapter Three.
-, IGBulg. II 845, an epithet of Thracian Heros, Mihailov:

?
, IGBulg. V 5328, 2nd - 3rd century AD; Mihailov is questioning
its Thracian origin.
, IGBulg. IV 1963 from ancient
Serdica now Sofia, after 2nd century AD; , IV
2004 from Gorni Lozen, near Sofia, after 2nd century AD; V 5827, 3rd
century AD: the name , which belongs to another dedicant, is
another form of (see the respective lemmata).
a male name, IGBulg. III 1324, 3rd century AD, from
Varvara, near Pazardjik. On the ground of similar developments in
/ we could posit
See above.
AD. Mihailov interpreted the latter as accusative plural feminine.
, passim.
] on an ex-voto, an epithet of Hera

(Marazov et al. 1996, 46, no. 52), 2nd - 3rd century AD. This is a new
epithet, which appears here for the first time. It seems to me that the initial
iota stands for eta, and therefore it should be restored to It is
again a compound name with a first component Erg- as in
and a second one / . We have all
reasons to say that initial iota is unusual for Thracian and that the sequence
iR- is all the more unlikely to appear. I believe the inscription was cut
around the end of the third century AD when graphemic variants of such
nature occur quite often thus betraying a phonetic [i] for [e]. To be sure, let
remind that vocal quantity in Thracian was lost as early as the earliest
inscriptions (6th - 5th century BC). Not in Detschew.
, a river in Thrace, IGBulg. I 13, 65, 320; V 5215, 3rd century
AD.
, inhabitants of Istria, IGBulg. I 388, 260-250 BC.
(.) , IGBulg. IV 2120, 2nd - 3rd century AD, an epithet of the

, a name of a Thracian town nearby the city of Sliven,
IGBulg. V 5636, around 146 AD; IGBulg. III 1731 and Elvers
1994, 241-266, end of 4th century BC.
, a name of a dedicant, IGBulg. III 1359, 3rd century AD.
, Tiberius Claudius , IGBulg. II 561, from
Glava Panega, 3rd century AD; II 874
V 5385, 1st - 3rd century AD. Not in Detschew.
was engraved at the bottom of a phiale from the

Rogozen treasure, masculine personal name, see no. 22 from Chapter
Three. Not in Detschew.
, , IGBulg. III 1760, from
Ovoshtnik, Kazanluk area, central Bulgaria, 2nd - 3rd century AD
, IGBulg. V, 5883, 77 AD, line 8, from
Kresna (see IGBulg. IV, p. 233), where is considered Oriental by
Zgusta, KPN, § 117. 2. Mihailov (SEG 32, 1982, 640) thinks that it is the
same as . Not in Detschew.
, inhabitants of Kallatis, see below.
, IGBulg. I 65, 307. See Detschew 223 with more
evidence: he argues that the Greek-like form might well be derived from
the name of a nearby lake, as it is pointed out in Etymologicum Magnum
s. v. . There is also a possibility of a Bithynian origin. See
(- ), IGBulg. V 5011, a town, perhaps another name for the
town off
a personal name and as presumably meaning
“beloved” in inscriptions from Kyzikos, see Detschew 225 and four
instances from Thasos (IG 12. 8. 307 and 309; 308 and 317).
, IGBulg. III 1385, an epithet of Heros, from Gara
Kricim, district of Pazrdjik, 3rd century AD. One could think of a place-
name from the Radomir-Kjustendil area related to it: present-day
Kandjulica, where Kandj-(=Kandsch-) is from *Kandj- and –ica is just a
common Slavic suffix. Wrong reading in Detschew 226: .
This votive tablet was dedicated to the Thracian Heros for a “canis
rabidus”, and hence speculations about the etymology of his epithet, see
Mihailov.
, IGBulg. V 5328, an epithet of the
Thracian Heros, 2nd - 3rd century AD. Not in Detschew.
, IGBulg. I 79, an epithet of Apollo Heros; I 284, 285,
286, 290; 2nd - 3rd century AD: , I 288;
287, 289, all occurances from Varna, ancient Odessos.
(- ), IGBulg. V 5011, a town, perhaps another name for
the town off Not in Detschew.
, IGBulg. I 53, uncertain first and second letters, 2nd - 3rd
century AD; II 484, uncertain reading, 2nd - 3rd century AD; III, 1312, 3rd
century AD; IV 2057, 135 AD; IV 2208 (?); V 5730. See .
* fromm , IGBulg. II 837, epithet of the

Heros, 3rd century AD. Not in Detschew.
, IGBulg. II 837, an epithet of the Heros from
rd
* 3 century AD. Not in Detschew.
, IGBulg. I 172, a female name, from Odessos.
a place name, or IGBulg. IV 2163,
see app. cr. where G. Mihailov is hesitating (“duobus modis suppleri
possis”) which is the correct one of either readings; see also IV 2050-57,
2162, 2164-2166, 2168 Not in Detschew.
, IGBulg. IV 2150-57, 2162, 2164-2166, 2168, an
epithet for the local gods and a place-name. IV 2163
(?).
IGBulg. I 378, after 19 AD. Not in Detschew.
, IGBulg. IV 1997, an epithet of Hera, 2nd - 3rd century
AD.
, IGBulg. III 1690 b 36, 202 AD.
, IGBulg. III 1806, 1st - 3rd
century AD.
in Manov 2008, 124, found in the district of
Blagoevgrad, 2nd - 3rd AD. Not in Detschew.
, an epithet
p of Asklepios, IGBulg. V 5798, 3rd century
, an epithet of Asklepios,
p IGBulg. V 5787-5789, 5793,
5800, 3rd century AD. Not in Detschew.
K( ) IGBulg. V 5792- 5794, 5797, 5838, epithet of
Asklepios, 3rd century AD. Not in Detschew.
IGBulg. V 5786 epithet of Asklepios, 3rd century
, IGBulg. IV 2253, 1st - 3rd century AD; V 5883, 77
AD; V 5890 from Kresna, 1st - 3rd century AD. Not in Detschew.
IGBulg. V 5785 = IV 2125, 5823, epithet of
rd
Asklepios, 3 century AD. Not in Detschew.
, an epithet of Asklepios, IGBulg. V 5801 ( )
, 5802, 3rd century AD. Not in Detschew.
, a place-name, IGBulg. IV 2338, end of the 1st century
AD. The name appears in Procopius De aedificiis 4. 11 as ,
which according to Mihailov s. v. is the same as . However,
the two places were not the same, one being in Haemimontos, and the
other in Nicopolis ad Nestum. For various speculations on its etymology
see IGBulg. IV 2338, commentary. Not in Detschew (p. 238
).
IGBulg. V 5329, from
Dobroplodno, northeast Bulgaria. Not in Detschew.
, a place-name of a mansio, Lat. Cillae, IGBulg. III 1519, 3rd
century AD.
, an inhabitant of , see above, IGBulg. III 1520, 3rd
century AD.
, IGBulg. I 354quater, an epithet of Asklepios, 1st century
AD. (?). Not in Detschew. See also Mihailov’s comments in V 5127 (=I
354quater): “Ex BIAB (i. e. Bulletin of the Bulgarian Institute of
Archaeology) deprompserunt Ann. Ép. 1965, 133; J. et L. Robert Bull. Ép.
1965, 256 (ubi falso et ) et 1972, 298 (ubi recte
), (SEG 24, 1969, 906).”
, an epithet
p of a local god, IGBulg. III 1774, 3rd century
, the name of a tribe (phyle), from the Ancient Theatre
in Philoppopolis; IGBulg. V 5412; cf. SEG 34, 1984, 712.
, an epithet of Apollo, IGBulg. III 919, and 1002, from
ancient Philippopolis, 3rd century AD.
an epithet of Apollo, IGBulg. III 917, 921, 998, 3rd
AD; V 5435, which belongs to the inscriptions series of III 917-928, from
ancient Philippopolis.
a spurious epithet of Apollo, accepted by Detschew
239. However see the lemma at IGBulg. IV 1929 from Serdica, probably a
completely different reading must be supplied.
an epithet of Apollo, IGBulg. III 918, 3rd century AD,
from the territory of ancient Philippopolis.
, IGBulg. II 510, 3rd century AD Found in the sanctuary of
Asclepios at Glava Panega. I. I. Russu, Dacia 11-12, 1945-47, 262, n. 8
considered it an error of the engraver for (cf. J. et L. Robert
(Bull. Épigr. 1950, 22, n. 49). on stone. See also Dimitrova 2002,
217. Mihailov hesitated whether the name is Thracian or Latin. Not in
Detschew.
The name is of Thracian origin based on the following observations: 1)

the genuine Thracian morph /kent/ alternates with /kenth/ as e. g. in
Kenthiaros (see above) , etc.; 2) on the other
hand the morph /enth/ alternates with /end/ and /ens/ due to the ambiguous
rendering of Thracian phonemes through Greek ones; 3) There is room to
assume that in the 2nd and 3rd century AD allophones appeared probably
due to conflation between the two phonological systems (Greek and
Thracian) when putting a Thracian word in writing. The phoneme /th/
alternates with /s/ in many instances. Here again one should keep in mind
that /th/ is not a Thracian phoneme, but rather a Greek misrepresentation.
4) The /nth/ morph appears always in well known names but not in rare
Thracian words (epithets, etc) that show a clear Thracian sound as e. g. in
Skelentene, Sonketene, Tiditene, etc., where the structure of the original
morph is preserved well into the Roman period.
IGBulg. V 5932, 1st - 3rd century AD. Not in Detschew.
See , Cerzos, Cersa in Detschew 229 for the first part of
this name.
, IGBulg. III 1408, 3rd century AD It
should be mentioned here that this feminine personal name stands among a
range of names such as from inscriptions from Amphipolis and
Olbia, from Gorgippia (now in Ukraine), a name of a
soldier, 300 BC, in IG II2 1956, Cerzus in Edonis, from Samos,
and Cerza, feminine (CIL III. 8921). For more details see
Detschew 229. See also *Karza, a reconstructed place name in
Paphlagonia, based on the Zeus’s epithet Karzenos (Zgusta KON, §442).
Duridanov (1987а, 42) considers *Karza as non-Thracian.
an epitaph from Sozopol, ancient
Apollonia, on the Black sea coast, 5th BC, IMB 4, 2002, 124, n. 20. See
also SEG 52, 681.
(?), IGBulg. IV 2196. See also no. 16 from Chapter
Two: ( ) and example m. from Chapter Three:
: the name of the Odrysian king who
reigned 359-346 BC. Not in Detschew.
, IGBulg. III, 1517, 3rd century AD. Not in Detschew.
IGBulg. III 1465, Thracian (?) or Latin
Cercius. Not in Detschew.
an epithet of Apollo worshipped by the

, IGBulg. III 1711, 1st - 3rd century AD and see
III 1712, an epithet of Apollo from Ezerovo, 1st - 3rd
century AD.
, IGBulg. I 94, 2nd -3rd century AD; see V 5037 = I 94
from Odessos. The form was corrected in the addendum of IGBulg. I.
IGBulg. I 270, 2nd - 3rd century AD Adjective to
* a town. Detschew 243 read it differently:
IGBulg. V 5790, 5791, 5799, epithet of Asklepios, 3rd

IGBulg. IV 2017, an epithet of Zeus. Not in
Detschew.
, a place name whose inhabitants were called ,
see below.
Böttger – Halloff 1991, 485-488, inhabitants of
Not in Detschew. For a slightly different reading of the
inscription see IGBulg. V 5591, where Mihailov, relying on a photograph,
misread two words: for and
for . On the other hand, his suggestion
that (an ethnicon) is in the vocative case (“acclamatio in
vocativo”), see his V 5677 = IV 1934 with the same use of the vocative,
seems to be correct. Böttger and Halloff (1991) are of a different opinion.
, epithet of Asklepios, IGBulg. V 5697, 2nd - 3rd century,
IGBulg. III 1805, 1st – 3rd century AD, epithet of Zeus. Not
in Detschew.
, in an inscription found on a
fresco in the tomb of Alexandrovo, near Haskovo, southern Bulgaria. See
Kitov 2004; 2005b, 28; Gerasimova 2003, 177-78; Sharankov 2005. There
is another example from Gazoros, near present-day town of Serres,
northern Greece, that reads .21 In this
compound, the first component / / is a allomorph of / / as in
. Not in Detschew.
21
For more details and bibliography see Mihailov 1980b, 15.
Fig. 4-3. Inscription with Kozimases from Alexandrovo.
, IGBulg. III 1514, 3rd century AD; see ,

IV 2213 = V 5867. The latter Not in Detschew.
, IGBulg. III 1321, from Byaga, near Peshtera, central-
southern Bulgaria, 3rd century AD.
(gen.), on a golden plate from
Sinemorets, on the Black Sea coast, near Tsarevo, south-eastern Bulgaria.
Probably second half of the 2nd century BC. Not in Detschew.
22
I am thankful to Mrs. D. Agre, Institute of Archaeology, Bulgarian Academy of
Sciences, for allowing me to examine the plate. A joint publication with all the
details is forthcoming.
Fig. 4-4. The golden plate from Sinemorets
, IGBulg. I 301, inhabitants of a , 3rd century AD

Not in Detschew.
, IGBulg. III 947,
from Philippopolis, 3rd century AD. Not in Detschew.
IGBulg. III 1559, 1st - 3rd century AD.
, IGBulg. I 46, between 44-43 BC and 2nd -
rd
3 century AD on a list of priests from Odessos; I 400, a female name; I
466, 2nd - 3rd century AD; I 469, possibly of the Byzantine period; I 51 bis
a line 9, 2nd - 3rd century AD; , who is a priest of the
Syrian Goddess III 918; III 953, 3rd century AD; III
971; Detschew 264 incorrectly: ; 3rd century AD . III 1080;
. III 1116, 3rd century AD; III 1187,
3rd century AD; . III 1289, 3rd century AD;
III 1293, 3rd century AD;
rd
III 1303, 3 century AD; III 1317, 3rd century AD;
III 1353, 3rd century AD; III 1362, 3rd century
rd
AD; III 1538, 3 century AD; III 1626, 1st – 3rd century
AD; [ ] III 1690 b 36, 202 AD;

III 1895 = 1289 = 1091bis; Detschew 313
IV 1972; IV 2103; IV 2129; IV 2192; V 5338, 1st
century AD; V 5466, 1st - 3rd century AD; V 5741, 2nd - 3rd century AD.
Cf. Chapter Two and Chapter Three.
, IGBulg. I 306, 2nd century AD. Not in Detschew.
, IGBulg. III 971; Detschew 264
rd
wrongly ; 3 century AD.
, IGBulg. IV 2282.
,
, IGBulg. III 1846, 1st
rd
- 3 century AD. Not in Detschew.
, , from Glava Panega, IGBulg. II 523.
, IGBulg. IV 1934, epithet of Asklepios; V 5785, 231
AD; V 5786, 241 AD. According to Mihailov one more sigma might be
added. However, on the photograph of V 5785 it is clear that the name has
only one sigma. With the photograph of V 5786, exemplifying only one
sigma on line 3, and with the previous lines missing, it is difficult to
accept Mihailov’s supposition that a second sigma should be restored to
. Here compare OYLKTIOY (sic) from an inscription on
two silver casseroles (Manov 1994, 89-92), 1st - 2nd century AD. Manov
says that “Die Inschrift lautet OYLK. TIOY” and further on he gives
OYLK(ou) TIOY: “(Das Gefäß gehört) Ulkos, (dem Sohn) des Tios”
(1994, 89-90). After inspecting the vessel, I still believe that a kappa
should be restored at the beginning because of the vessel bearing severe
fire damage and because of the technique of gold inlaid letters used.
We understand that Manov’s intention was to interpret the inscription
following an already established tradition in inscribing gold and silverware
from Thrace. However, he went too far in supplying a genitive ending,
thus cutting the one name in two in order to get to a very problematic Tios
to fit his theory of an Asia Minor name. The fact that there is an ivy leaf
separating [K]OYLK and TIOY should not be a reason to make two names
out of one. It is more economic to read the name as one entity. We should
understand that this kind of decoration occurs in many a case in
inscriptions (for a similar division of a name see IGBulg. III 1117:
(ivy leaf) ), just like the ivy leaf at the end of the one
discussed above. So, in conclusion, as there are no such Thracian personal
names as Oulkos and Tios attested so far, [K]OYLKTIOS seems to be in
line with It is true though that both a personal and a
place name Tios is found in Bithynia and elsewhere in Asia Minor.23

However, Oulkos would not make any sense in Thracian should it be
derived from *wlko-s. Not in Detschew.
(. ) , IGBulg. IV 2120, 2nd - 3rd century AD, an epithet of the
, made with gold inlay on two silver
casseroles. It has been published by Manov 1994, 89-92, 1st - 2nd century
AD, where he read . The name is related to Koul- and –ktios, -
k(t)us, -kous see above and Kout- and Cutiula, a personal name
CIL 3, 7330, from Thessaloniki on a relief of the Thracian Heros
(Horseman). A secure parallel is the epithet (see above)
from Koul-kous-anos/-enos. For the reading of this name, see the
considerations above under Not in Detschew.
from Manastir, district of Varna, on a
votive tablet (Haralambieva and Ivanov 1987, 74-83), 1st - 3rd century AD;
V 5367, 1st - 3rd century AD. Not in Detschew.
, IGBulg. I 23, 222-235 AD; II 835, 2nd - 3rd century AD
vel , IGBulg. III 1690 c 34, from Pizos, 202
AD.
IGBulg. III 947,
from Philippopolis, 3rd century AD. Not in Detschew.
, IGBulg. I 258, 2nd - 3rd century AD.
, IGBulg. III 1690 b 47, 202 AD.
, IGBulg. I 25, ethnicon for a Greek inhabitant of
Dionysopolis 4th - 3th century BC. Detschew 270 considered it a genuine
Thracian name. It is worth noting that this is the only example with Kuz- .
The place name Kyzikos had been known ever since Herodotus (4. 14),
and it makes us think of yet another interference of spirants (s, z) in Greek
and their treatment in Thracian. While Thracian has had a solid steady z
from I. E. *s, Greeks might had vocalized a Thracian intervocalic s and
hence Kyz-ik-os from *Ku(ou)s-ik-, etc.
a famous place name in Asia Minor, in Bithynia, see above
the ethnicon . Strabo’s testimony about this ethnically mixed
area speaks for itself - the region was inhabited by Bithynians
(=Thracians), Phrygians, Mysians, Doliones, a Thracian tribe, and around
Kyzikos by Mygdones and Trojans. Steph. Byz. 391. 4 quotes Homer to
23
Detschew 1976, 506; Zgusta 1964, 1558-5.
say that the Thracians Doliones were inhabitants of Kyzikos. IG 7. 523

from Boiotia, gives us a (acc. pl. ).
, IGBulg. II 557, 3rd century AD, an epithet of the
Thracian Heros.
IGBulg. III 1712, an epithet of Apollo from Ezerovo,
1st - 3rd century AD; see above.
IGBulg. III
st rd
1621, 1 - 3 century AD. Detschew 274
from Saladinvo near Pazardjik k Gočeva 1989,
113-15; the author did not date the inscription; the sanctuary being known
to us by other inscriptions, we could date this inscription to the 3rd century
IGBulg. III 1326 from Malo konare, near Pazardjik.
Mihailov considers it Asian or Thracian. Not in Detschew.
IGBulg. II 868, 2nd - 3rd century AD See Detschew 274
due to confusion in reading the inscription.
, IGBulg. III 947, near Philippolis, 3rd
daugther of wife off IGBulg. I 27 bis,
from Dionysopolis, 3rd century BC (?). Compare the second element -
in and . Not in Detschew.
(?) , IGBulg. IV 2026, the reading is not sure.
Not in Detschew.
IGBulg. I 51bis, 2nd - 3rd century y AD, probably from a
Bithynian Thracian origin. Not in Detschew.
epithet of Asklepios IGBulg. V 5699 = IV 2029, V 5700 -
5715, 5705; all from the sanctaury of Asklepios Limenos at
Slivniza near Sofia, 2nd - 3rd century AD. Not in Detschew.
(?) is just a possibility as Mihailov put it (see
under , IGBulg. III 1385), epithet of the Heros 3rd century
AD. See G. Mihailov’s comments: “Paene omnes litterae ligatae sunt…. ut

NK=LYK legere possis”. Not in Detschew.
IGBulg. IV 2297, 1st - 3rd century AD; see at

III 1732. Not in Detschew.
IGBulg. II 501, 3rd century AD Not in
Detschew.
a female name from Piperitsa, district of Blagoevgrad,
in Manov 2008, 109. Not in Detschew.
IGBulg. I 77, 78, epithet of the Heros, 1st century BC.
a female name, in Manov 2008, 126, 154 AD, from the
museum of Blagoevgrad. Not in Detschew See Monta below.
IGBulg. IV 2268, 2nd - 3rd century AD; IGBulg. IV 2283 bis
(?). Not in Detschew.
a female name, in Manov 2008, 130, 2nd - 3rd AD,
from the Rupite, district of Blagoevgrad, now in the museum of the
Archaeological Institute in Sofia. Not in Detschew.
a female personal name, belongs to the dedicant, cf.
Detschew 291, from Kavadartsi, near present-day Skopje. See the
sequence on the gold finger ring from the Arabajiiska Mogila
at Duvanlij (Filow 1934 and The Evidence).
IGBulg. I 345, 3rd century AD, a place in Thrace, the
ancient town of Mesambria.
a place-name, present-day Lovech, IGBulg. II 598.
IGBulg. IV 2177.
passim.
IGBulg. IV 2284.
IGBulg. V 5828, 2nd - 3rd century AD. Not in
Detschew.
IGBulg. IV 2214 1st - 3rd

st rd
century AD; IV 2214; 1 - 3 century AD; IV
2337, 1st - 3rd century AD; IV 2343 from Nicopolis ad Nestum, 3rd century
AD; , IV 2350, from Novo Lyaski, on the
Nestos river, near the Greek-Bulgarian border, 2nd - 3rd century AD.
IGBulg. IV 2337, 1 - 3rd century AD.
st
IGBulg. IV 2337, 1st - 3rd century AD.

IGBulg. III 1727, early 3rd
century BC. Not in Detschew. In Detschew 301 this same name is said to
be from an inscription found at Delphi.
IGBulg. III 1410, an epithet of the Heros 76 AD.
Detschew 302:
IGBulg I 400, 1st - 2nd AD; IV 2288, 165 AD.
(?), IGBulg. III 1516, 3rd century AD.
IGBulg. V 5867= 2213, 2nd - 3rd century AD. Not in
Detschew.
a female name, IGBulg. IV 2228, 2nd - 3rd century AD.
, IGBulg. III 1773, 1st-3rd AD; IGBulg V 5011, 1st BC -
st
1 AD.
IGBulg. IV 2015, a priest, 2nd - 1st century
BC; wrong reading in Detschew 165: [ ?].
a female name, IGBulg. IV 2228, 2nd - 3rd century
AD; Angelov 2003, 138 from Sandanski, southwest Bulgaria, 1st - 3rd
century AD.
IGBulg. V, 5883, 77 AD. Not in Detschew.
and , IGBulg. IV 2337, 1st - 3rd
century AD; IV 2141. Not in Detschew.
IGBulg. I 374, 2nd - 3rd
century AD; II 542 from Glava Panega; Detschew 36 incorrectly:
.
(?), IGBulg. III 1516, 3rd century AD Not in Detschew.
IGBulg. I 469 bis, 3rd - 2nd century BC. Not in Detschew
312.
. . from Glava Panega, IGBulg. II 576. Not in Detschew.

IGBulg. III
1457, 3rd century AD.
IGBulg. III 1690 b 54, 202 AD.
? IGBulg. III 1626, 1st - 3rd century AD;
AD.
AD; see and Not in Detschew.
IGBulg. IV 1922, 2nd - 3rd century AD, on a column
inscribed with names (a sacred catalog?); , II
539, from Glava Panega, 3rd century AD; II 677
; III 1079 2nd -
3 century AD; III 1516, 3 century AD; III 1396, 3 AD; III, 1752, 1st -
rd rd rd
3rd century AD; III 1895 = 1289 = 1091 bis;

Detschew 313 . Here belongs also the false reading of
Detschew 221 (relying on Heydemann’ s reading 1874, 292
no. 788). IV 2337, 1st - 3rd century AD; IV 2203; IV 2222;
IV 2213, 2nd-3rd AD; IV 2194; IV 1923.
on an ex-voto
- 3rd century AD. The reading belongs to the editor princeps Kacarov
(1937,, 284 n. 2 fig. 223; also J. et L. Robert Bull. Epigr. 52, 1939, 480 n.
214).
IGBulg. I 538 from
Glava Panega, 2nd century BC (?);
IGBulg. II 769 , 1st - 3rd
century AD; III 1073, 3rd century AD; III 1101,
rd
3 century AD; III 1199, 3rd century AD;
, IGBulg. III 1760, 2nd - 3rd century AD; III
1793, 1st - 3rd century AD; V 5805, 2nd - 3rd century AD;
, an ex-voto for himself and his wife
IGBulg. IV 2015.
IGBulg. II 754.
IGBulg. III, 1736, 1st - 3rd century
AD.
IGBulg. II 846; III
1106 3rd centuryAD; III 1516, 3rd century AD;
III 1583, 1st-3rd centuryAD; III 1593, 1st-3rd
century AD; ? III 1626, 1st - 3rd century AD;
? 1626, 1st - 3rd century AD;
(accusative) IGBulg. III 1633, 3rd century AD;
III 1690 b 66 from Pizos;
III 1690 column b line 27, 202 AD; III
1690 b 52; , IGBulg. III 1690 b 52, 202 AD;
IGBulg. III 1690 b 64, 202 AD;
IGBulg. III 1690 b 65, 202 AD; III 1690 c 35
; III 1690 c 32, 202
AD; III 1690 d 40, 202 AD; , IGBulg.
III 1690 c 44, 202 AD; III 1690 c 50, 202 AD;
III 1690 c 51, 202 AD;
III 1690 d 38, 202 AD;
III 1690 d 40, 202 AD and d 45;
III 1690 c 63, 202 AD;
III 1690 d 59, 202 AD; III 1690 e 15;
1736, 1st - 3rd century AD;
III, 1741, 1st - 3rd century AD; III 1773, 1st –3rd
centuryAD; III 1806, 1st - 3rd century
AD; , V 5300, 2nd - 3rd century AD;
V 5343, from Marcianopolis, 2nd - 3rd
century AD; V 5611; V 5394bis, 1st - 3rd century AD; V 5821, 2nd - 3rd
century AD.
IGBulg. V 5628, 1st-3rd century AD.
( ), IGBulg. IV 1922, 2nd - 3rd century AD, on
a column inscribed with names (a sacred catalog?); IV 2196
(?); II 543, from Glava Panega, 2nd - 3rd
century AD; III 1690 d 34, 202
AD.

AD.
IGBulg. II 511, from Glava Panega, after 1st century AD;
III 1317, 3 century AD; IV 2142, 2nd - 3rd century AD; IV 2151, 2nd - 3rd
rd
century AD; (dat. ) IV 2312, 2nd -

3rd century AD; IV 2314, 2nd - 3rd century AD;
IV 2349 from Kovachevitsa near Nicopolis ad Nestum,
nd rd
2 - 3 century AD; , IGBulg. V 5329,
from Marcianopolis, 2nd - 3rd century AD; , in
Manov 2008, 125, 2nd - 3rd century AD, district of Blagoevgrad. Not in
Detschew.
both in Manov
2008, 126, 154 AD, from the museum of Blagoevgrad. The nominative
ends in –es, judging on the name of his son in that same
inscription, which reads
ἱῷ…Not in
Detschew. in Manov 2008, 130,
2nd - 3rd AD, from the museum of Blagoevgrad. Not in Detschew.
in Manov 2008, 136, 2nd – 3rd AD, found near
Sandanski. Not in Detschew.
IGBulg. IV 2228, from Boboshevo, 3rd century AD.
Not in Detschew.
(dat.) IGBulg.
IV 2312, 2nd -3rd century AD; V 5926, 158 AD. Not in Detschew.
IGBulg. IV 2228, 3rd century AD. Not in Detschew.
IGBulg. III 1016, 3rd century AD. Not in
Detschew.
, IGBulg. II 711, 2nd - 3rd century
st rd
AD; III, 1725,1 – 3 century AD.
II 836, 2nd -3rd
IGBulg. I 354, 355, 363, dedicator to the Heros, 3rd
century AD.; II 556, from Glava Panega, 2nd - 3rd century AD;
II 694, 2nd - 3rd century AD;
II 836, 3rd century AD; II 843, 2nd - 3rd
century AD; II 854; III 951, 3 century AD; III 1089, 3rd century AD;. III
rd
1093, 3rd century AD; , III 1097, 3rd century

AD; III 1107, 3rd

century AD; III 1133, 3rd century AD;
, III 1430, 3 century AD; III 1367, 3rd century
rd
AD; III 1377, 3rd century AD; III 1405, 3rd

century AD, 3rd century AD; , III 1438, 3rd
st rd rd
century AD; III 1593, 1 - 3 century AD; III 1516, 3 century AD;
III 1621, 1st - 3rd century
AD; -? III 1626, 1st - 3rd century AD;
? III 1626, 1st – 3rd century AD; III 1630,1st - 3rd
century AD; III 1631, 1st - 3rd century
AD; III 1631, 1st - 3rd century AD;
III1690 b18; I
rd
II 1690 b 27, 202 AD; III 1690 c 64, 3
century AD; , III 1690 b 13, 202 AD;
, III 1690 b 14, 202 AD;
III 1690 column b line 20, 202
AD; III 1690 b 21, 202 AD;
III 1690 b 33, 202
AD; III 1690 b 37, 202 AD;
III 1690 b 38, 202 AD;
III 1690 b 39, 202 AD;
III 1690 b 40, 202 AD;
III 1690 b 41, 202 AD;
III 1690 b 43, 202 AD;
III 1690 b 44 and d 55, 202 AD;
, III 1690 b 46, 202 AD;
, III 1690 b 51, 202 AD;
III 1690 b 63, 202 AD;
III 1690 b 64, 202 AD;
, III 1690 b 19, 202 AD;
, III 1690 c 14, 202 AD;
III 1690 c 23, c 39, 202 AD;
III 1690 c 24, 202 AD; , III 1690 c
25, c 38, 202 AD; III 1690 c 26, 202
AD; III 1690 c 36, 202 AD;
III 1690 c 51, 202 AD;

III 1690 c 52, 202 AD and c 33;
III 1690 c 53, 202 AD;
III 1690 c 54, 202 AD;
III 1690 c 55, 202 AD;
III 1690 c 56 and d 43, 202 AD;
, III 1690 c 61, 202 AD; III 1690 c 62, 202 AD;
, III 1690 d 27, 202 AD;
, III 1690 d 28, 202 AD;
, III 1690 d 29, 202 AD;
, III 1690 d 30, 202 AD;
III 1690 d 32, 202 AD;
III 1690 d 34, 202 AD;
III 1690 d 36, 202 AD;
vel – III 1690 d 36-37, 202 AD;
III 1690 d 38, 202 AD;
III 1690 d 41, 202 AD;
III 1690 d 42, 202
AD; III 1690 d 44, 202 AD;
III 1690 d 58, 202 AD;
III 1690 d 63-64, 202 AD; III 1690 d
71-72, 202 AD;III, 1717, 2nd - 3rd century AD; III 1723, 1st -3rd century
AD; III, 1758, 1st - 3rd century AD; , III 1782, 1st
-3rd century AD; III 1805, 1st - 3rd century AD;
III 1806, 1st - 3rd century AD;
, III 1816, 1st - 3rd century AD;
τῇ , III 1846, 1st - 3rd century AD.

III 1690 b 32, 202 AD; IV 2233 and 2234,
from Sportela, 2nd - 3rd century AD; IV 2074; IV 2114; IV 2224; V 5805,
2nd - 3rd century AD; V 5275, V5863=2141 from Blateshnica, 1st - 3rd
century AD; V 5651; V 5665; V 5716; in Manov 2008, 109, found in
Piperitsa, district of Blagoevgrad, 2nd – 3rd AD. Not in Detschew.
IGBulg. I 257, epithet of the Heros, 3rd century AD.
Not in Detschew.
an inhabitant of a town, probablyy See IGBulg. V

5652, from Bizone, 3rd century AD. Not in Detschew.
, see below in IGBulg. IV 2195. Mihailov under question
considers it to be a place name based on the epithet of Dionysos. It can
hardly be a name of a town but rather an adjective connected to the local
cult of Dionysos. Not in Detschew.
, IGBulg. IV 2195, from Tavalichevo near
Pautalia, 2nd - 3rd century AD. Possibly from *meudh ‘lament’, see
Pokorny (19943, 743) Greek , Old Bulgarian mislu “thought’ and
*worg ‘secret rites, orgy’, Watkins 100 and Pokorny (19943, 1168). See
further on in Chapter Five under treatments of PIE *dh, etc. Mihailov did
not make any comment in his lists of names. The Thracian origin of this
adjective is secured by its formation, with the typical ending –enos (for
this part in adjectives see in Chapter Five) and the treatment and
development of both mut- and org- (from *worg, where *w after
weakening is lost). Not in Detschew.
IGBul.
III 1621, 1st - 3rd century AD, an epithet to the Thracian Heros.
IGBulg. V, 5883, 77 AD.

IGBulg. V, 5883, 77 AD.
IGBulg. IV 2286 159 AD; according to Mihailov
this is a female name.
IGBulg. III 1222, 3rd AD. Not in Detschew.
in
etc., from Bithynia, (Corsten 2007, 131-33),

2nd BC.
IGBulg. III 1347, 3rd century AD is another graphemic variant
of See also Masson 1988, 10.
passim, , IGBulg. V 5011, from Balchik
(Dionysopolis), 1st BC - 1st century AD; see also V 5030, where the name
is not explicitly mentioned.
, IGBulg. IV 2015
rd
IGBulg. III 1352, 3 century AD.
IGBulg. I 233 bis, 3rd century BC. Not in Detschew.
IGBulg. II 599 (= V 5197) from Kamenec;
Mihailov In his 5th volume under no. 5197 redirects us to V 5634; 1st - 3rd
century AD; Mihailov (with bibliography) established a connection with
the cult of Jupiter Okkonenos in Bithynia, and his dissemination later in
Thrace with the same epithet. Not in Detschew. See Detschew 340:
for the inscription from Kamenec, corrected by Mihailov.
For this reading we have to rely on Mihailov’s explanation. The stone was
not preserved (“lapis non est servatus, ut lectionem conferamus”). For
see Pleket (SEG 32, 1982, 679). There are two examples of
cited by Şahin (1982, 118 and 1201). See also
below.
IGBulg. II 718, from Nicopolis ad Istrum, transferred later
to Tirnovo and found there in some house, 1st- 3rd century AD. Mihailov
mentions it again in V 5248, where he restores it to . See his
reasons expressed above.
century AD; see , II 855, 3rd century AD and II
rd
856, 3 century AD.
an epithet of Dionysos, IGBulg. III 1881,
an epithet of a local Apollo, on a marble relief, a
fragment of an ex-voto (G. Kabakchieva 2000, 67). Dated to the 3rd - 1st
century BC by the editor G. Kabakchieva who reads Not in
Detschew.
, an epithet of Apollo, IGBulg. III 1713 bis, 1st - 3rd century

AD; Not in Detschew. See below.
century AD 855; see II 856, 3rd century AD and
rd
II 804, 3 century AD.
ἱῷ… in Manov 2008,

126, 154 AD, from the museum of Blagoevgrad. Not in Detschew.
IGBulg. I 430, 5th - 4th century BC. Not in Detschew.
an epithet of the Thracian Heros IGBulg. II 832,
833, 2nd - 3rd century AD.
IGBulg. II 511, 536, 540, 544, all occurrences from
the Roman period.
IGBulg. II 874,
, IGBulg. III 1817, 1st - 3rd century AD. There
are different opinions as to the origin of this name. Detschew 370
considers it Thracian; Mihailov considers it to be either Thracian or
Greek. Russu (Dacia NS 2, 1958, 538; 1967, 46) regards it as Greek rather
than Thracian.
IGBulg. I 76 bis, derived from , a
dedicator’s name, 3rd century AD.
IGBulg. III 1018, 3rd century AD; see . Not in
Detschew.
( ), IGBulg. III 884 a 23, 3rd century AD.
IGBulg. III 884 a 23, 3rd century AD; see ( ).
nd st
IGBulg. I 283, epithet of the Heros, 2 - 1 century B. C;
IGBulg. I 283 bis, 1st century BC. – 1st century AD, according to Mihailov
“ut videtur”.
IGBulg. IV 2251 from
Gorna Gradesnitsa near the border with Greece, 3rd century AD(?). Wrong
reading in Detschew 365.
, IGBulg. IV 2233, 1st - 3rd century AD. Not in Detschew.
IGBulg. I 289, reconstructed by Mihailov; II 848.

[ vel ], IGBulg. V 5082=289. Not in Detschew.
, a Thracian emporium, IGBulg. III 1690.
, IGBulg. V 5329, from Marcianopolis, 2nd - 3rd century
AD, Thracian or Oriental. and versus
and show the shift / Not in Detschew.
* , IGBulg. III 1445, 3rd century AD; see
. Not in Detschew.
, IGBulg. III 1445, 3rd century AD. Not in
Detschew.
, IGBulg. IV 2322, 211 AD.
IGBulg. III 1626, 1st – 3rd century AD; III, 1726,
st rd
1 - 3 century AD; , III 1802, 1st - 3rd century AD.
, IGBulg. IV 2337,1st - 3rd century AD. Not
in Detschew.
, IGBulg. III 1760, 2nd - 3rd century
AD.
, a male name, IGBulg. III 1804, 1st - 3rd
century AD.
IGBulg. II 850, 2nd -3rd century AD. Not in
Detschew.
IGBulg. III 933, an epithet of the Heros, 3rd century AD.
IGBulg. III 1516, a female name, 3rd century AD. Detschew
377
IGBulg. III 1796, a male name(?), 1st -3rd
century AD; Not in Detschew.
inhabitants of the lake Prasias, from an inscription in
ancient Pistiros. See BCH 118, 1994, 1-15 ….
…Not in Detschew.
(dat.) in IGBulg.
IV 2312, 2nd - 3rd century AD; V 5913= IV 2289, where is a
wrong reading for . IV 2230: ,
nd rd
from Barakovo, near Blagoevgrad, 2 - 3 century AD; IV 2278; IV 2281,
1st -3rd century AD; IV 2337, 1st - 3rd century AD IV 2346,
(dat. ), 2nd - 3rd century AD; IV 2252, 1st - 3rd
century AD; IV 2257; IV 2278, 1st - 3rd century AD;
IV 2286, 159 AD; , IV 2298; V

5918, 5921, 2nd - 3rd century AD; from Sandanski, southwest
Bulgaria, in Angelov 2003,
142; and from
m Pipe
ritsa, district of Blagoevgrad, in Manov 2008, 108-9. Not in Detschew;
, together with , a female name, in Manov 2008,
125, 3rd century AD, district of Blagoevgrad; the same name in Manov
2008, 127, 2nd – 3rd AD, from the museum of Blagoevgrad. Not in
Detschew; in Manov 2008, 130, 2nd – 3rd AD, from the
Rupite, district of Blagoevgrad, now in the museum of the Archaeological
Institute in Sofia. Not in Detschew.
, IGBulg. II 588, 704, 753, 754, an epithet of the
Heros, 3rd century AD; V 5805, 2nd - 3rd century
AD; III 978, 3rd century AD; IV
2304, 162 AD, Detschew 387 ; IV
2344, the reading belonging to G. Katsarov (1919-20, 10 fig. 6 = SEG 1,
1923, 295, cf. SEG 3, 1927-29, 112 add. et corr. ).
an epithet of Thracian
Heros on a siver vessel from Golyama Brestnitsa, district of Pleven, north-
central Bulgaria (Petkov 1960; Rousseva 2006, 61, no. 49), 2nd - 3rd
century AD.
IV 2289 1st - 3rd century AD; IV 2311, a female name,
198 AD; (dat. ) IV 2312, 2nd - 3rd
AD; IV 2322, 211 AD; IV 2329, 1st - 3rd century
AD; IV 2254, 2 - 3 century AD; IV 2257; IV 2273; IV 2275, 2nd - 3rd
nd rd
century AD; V 5886, 222 AD; V 5927, 3rd century AD (?).

(dat. ), IGBulg. IV
2330, 1st - 3rd century AD; Detschew 387 put it differently.
on an ex-voto
rd
- 3 century AD. The reading belongs to the editor princeps, G. Katsarov
(1937,, 284 n. 2 fig. 223; also J. et L. Robert Bull. Epigr. 52, 1939, 480 n.
214).
IGBulg. IV 2214 1st - 3rd century AD.

an epithet of Apollo, IGBulg. IV 1928, 2nd - 3rd
century AD.
a place-name IGBulg. III 1185, 3rd century AD.
from Saladinvo near Pazardjik k Gočeva (1989, 113-115);
the author did not date the inscription; the sanctuary being known to us by
other inscriptions, we could date this inscription to the 3rd century AD. Not
in Detschew.
IGBulg. IV 2338, end of the 1st century AD, a female
name.
in Nos. 23 and 28 in Chapter Two.
, IGBulg. IV 2025.
, IGBulg. II
796 , III 1829
inhabitant of Bosagura, 1st - 3rd century AD. (cf. in in
Chapter Two)
IGBulg. III 1198, 3rd century AD;
from Karasura, Böttger and Halloff (1991, 474-
88), 3rd century AD. According to the editors, the change ou to o occurs
more often than o to u: “in zweiten Bestandteil des Namens” (Böttger and
Halloff 1991, 474). However, this is not likely the case as o or ou are not
in fact the graphemic variants of a morph /o/ but most probably a genetive
ending in Thracian.24
from Karasura (Böttger and
Halloff 1991, 474-88), 3rd century AD. Cf. supra, Not in
Detschew.
a military district (strategia) IGBulg. IV 2338, end of
the 1st century AD.
24
Dimitrov 2003, 351. In a number of inscriptions the E-ending indicates a
genitive form, e. g. on four silver phialae from Duvanlij and
in the Kyolmen inscription, our earliest examples. In the case of
Kerso-bleptes vs. Kersebleptes the genitive ending morpheme shows the variation
o/e. In many cases, in compound names the Greek genitive ending morpheme -ou
appears. Judging on the chronology of this Greek ending one should expect -o to
p
be used in inscriptions from the 5th century BC and before, whilst -ou from the 4th
century BC on.
IGBulg. II 750, 3rd century AD;

II 868, 2 - 3rd century AD; III 884, 3rd century AD; III 933, 3rd
nd
century AD; III 1039; III 1132, 3rd century AD; III 1318
3rd century AD; V 5276, 2nd - 3rd century AD;; V 5394; V 5830, 2nd - 3rd
century AD; (cf. No. 29 from Chapter Two).
Not in Detschew.
IGBulg. III 1115, 1st - 2nd AD; IV 2015,
nd st
2 - 1 centuryy BC; (dat. ), IV 2112, 1st - 3rd
century AD.
IGBulg. III 1132, an epithet of Hygia and Telesphor,
3rd century AD. Not in Detschew.
IGBulg. III 1690 b
37, 202 AD.
IGBulg. I 378, a military district, 19 AD; II 743.
, IGBulg. III 1090, 3rd century AD; V 5011, 1st century BC

st
– 1 century AD; V 5638bis; , son of Seuthes III and Berenike,
th
III 1731 (Elvers 1994), end of 4 century BC; V 5011
; V 5638 bis , from Kabyle, (see also
J. et L. Robert Bull. Ép. 1976, n. 474 where they wrote which
was not accepted by Mihailov ). Velkov 1991, did not included it in his
list. The verification on the photograph showed clearly .
Judging on the letter-forms this short inscription should be dated in the
Roman Imperial period. See Corsten (2007, 133) with an example from
Bithynia.
IGBulg. III 1690 c 24, 202 AD.
Not in Detschew.
IGBulg. IV 2322, 211 AD. Not in
Detschew.
, , III 1690 d 14, 202 AD. Not in

Detschew.
IGBulg. I 382, 2nd century AD; III 947, 3rd
rd
century AD; III 1349, 3 century AD; III 1439, 1441, 3rd
rd st rd
century AD; III 1516, 3 century AD; III 1657, 1 - 3 century AD.
III 1690 column b line 33, 202 AD;
, III 1690 c 49, 202 AD;
st rd
III 1690 c 55, 202 AD; IV 1963, 1 - 3 century
AD; IV 2104, 1st - 3rd century AD; V 5306, 2nd - 3rd
century AD; III 1690 c 59, 202 AD; III 1690 c 60,
202 AD; , III 1690 d 17, 202 AD; III 1690 d 26, 202
AD; III 1690 d 47, 202 AD;
III 1690 d 59, 202 AD; III, 1746, 1st - 3rd century AD.
, IGBulg. II 510, 511,523, 539, 3rd century AD.
IGBulg. II 564, 3rd century AD.
IGBulg. II 532, 551, 552, 573; 3rd century AD.
IGBulg. II 525.
IGBulg. II 512, 536.
IGBulg. II 514, 521.
IGBulg. II 517.
IGBulg. II 537, 577, 587; 3rd century AD.
an epithet of the Thracian Heros, IGBulg. II 531,
540, 543, 554, 2nd - 3rd century AD.
IGBulg. V 5877= III 2246; Mihailov is
correcting it back from - to -; also see V 5901-5905 and 5909
all from the so-called “Sanctuary of God Salenos” (Sanctuarium dei
Saleni), around 230 AD. He thinks that the epithet mentioned
in IV 2305 and the one from V 5610 should be etymologically
connected and considered together with similar in meaning.
Not in Detschew.
, IGBulg. IV 2348. Not in Detschew, but see Detschew 412
at ( ) .
IGBulg. II 526.
IGBulg. II 530 3rd century AD.
, IGBulg. II 516.
IGBulg. I 7 ter, 2nd century BC; I 42 line 10, 3rd - 2nd
century BC.
, IGBulg. III 1794, a place name, end of 4th century BC.
IGBulg. I 154, 2nd - 1st BC, a female name.
in a dedicatory inscription from Kabyle.
The dedication is to the Thracian Heros. Velkov 1991, 26 dated it to the 3rd
AD. He wrote but supposed that the epithet could be
( ). Such abbreviations are not known from inscriptions found in
Bulgaria, and we think the correct reading is as we gave it here. Velkov’s
reading was accepted in the publication in SEG 42:651.
The epithet appears for the first time here. Not in Detschew.
It should be mentioned that Detschew 425 listed Sassa (CIL 3, 14355.
15), a female person of Dacian origin and Saza (CIL 3, 14406a: Aurelius
Saza), unsecure reading.
It seems to me that there is no connection between Sases, etc. and the
Thracian form Seusa (see below) of the name of the Thracian king
Seuthes.
, son of Seuthes III and Berenike, IGBulg. III 1731and Elvers
(1994), end of 4th century BC. Cf. Chapter Three. Not in Detschew.
IGBulg. V 5635= IV1777, 2nd - 3rd century AD. Not in
Detschew.
an epithet of the Thracian Heros, IGBulg. III
1654, , III 1655, 3rd century AD (or later). See .
vel , III 1690 d 23, 202 AD. Not in
Detschew.
IGBulg. III, 1588
from Karanovo but brought there from Augusta Traiana, see comments by
Mihailov: , an epithet to Zeus Sebazios, 202 AD. Not in
Detschew.
IGBulg. IV 2286, 159 AD. See
above.
IGBulg. III 1690 b 59, 202 AD, which
Detschew 437 thinks should be corrected to ; Mihailov,
however says that on the stone it is written .
, IGBulg. III 975, an epithet of the Heros, 3rd century

AD.
( ), a male name from Sandanski, southwest Bulgaria (Angelov
2003, 142). Not in Detschew.
, IGBulg. IV 2174.
, a female name, IGBulg. III 1532 - 1534 (the same Flavia), 3rd
century AD; III 1698, 3rd century AD.
* , a place name linked to the epithet of Ares, see below. Not in
Detschew.
(?),V 5610 from Prilepci near Kardzhali, southeast
Bulgaria, an epithet of Ares. Probably cognate to the toponym * .
Cf. SEG 37, 610, SEG 41, 597 and Chaniotis (1991, 301, no. 75), who
suggests that is the beginning of the name of the dedicant
rather than that of a second epithet of the god. Not in Detschew.
IGBulg. I 378, a military district, 19 AD.
, town, present day Sofia, passim.
IV 1902 from Serdica (present-day Sofia) around 180 AD:
… …; 1907 around 142 AD:
nd
… 1914, 2 century AD: 1917,
2nd century AD: 1919, 2nd century
AD: III 1316, 3rd century AD.
, IGBulg. IV 1947 from Sedica, 2nd -3rd century AD.
, IGBulg. IV 2011, near Serdica, after
2nd century AD; Detschew 437 thinks that in 1690 b line 59,
from Pizos, 2nd - 3rd century AD, should be corrected to , but
Mihailov says that on the stone it is written .
IGBulg. II 539, 1st - 3rd century AD; III
st rd
1187, 3rd century AD.; III 1593, 1 - 3 century
AD; III 1690 d 41, 202 AD;
III 1690 d 50, 202 AD; III
1690 d 65, 202 AD; III, 1719 bis, 1 - 3 century AD; III, 1727, 1 – 3rd
st rd st
century AD; , i. e. Seuthes III, a Thracian king, from ancient

Seuthopolis, IGBulg. III 1731(Elvers 1994), end of the 4th century BC; IV
2336, pre-Roman (?); IV 2346, (dat. ), 2nd

rd
-3 century AD. Cf. Chapter Two.
orr IGBulg. II 683, epithet of the Heros 3rd AD. Not
in Detschew.
(?) IGBulg. II 743, strategia, according to Mihailov, who
is filling the vacat.
IGBulg. IV 2230 ,
from Barakovo, near Blagoevgrad, 2nd - 3rd century AD. Not in Detschew.
a recent discovery on a gold signet-ring found in the
Dalakova tumulus north of Topolchane, district of Sliven. Dimitrov
2008a, 26-32. The inscribed text reads: ( in the genitive
case). The ring is dated to the 4th century BC. Not in Detschew.
, IGBulg. V 5436. Not in Detschew.
, IGBulg. III 1690 b 45, 202 AD. Not in Detschew.
, IGBulg. III 1593, 1st - 3rd century AD, an epithet of
Apollo.
inhabitants of , IGBulg. III 1771,
nd rd
2 - 3 century AD.
a dedicant, III 1885, 1st - 3rd century AD.
IGBulg. I 203 bis, of uncertain origin according to
Mihailov. Not in Detschew.
, IGBulg. III 1799; Mihailov thinks it could be Thracian; 1st -3rd
IGBulg. I 133, genitive
IGBulg V 5921, 2nd - 3rd century AD. Not in
Detschew.
/ / on a clay amphora from Sboryanovo,
northeast Bulgaria.25
25
For an account on that archaeological site see above in Introduction, Stoyanov
(1997).
Fig.4-5. Graffito from Sboryanovo
, IGBulg. IV 2100, an epithet of Asklepios.

, a town ( ) of the see below.
inhabitants of the town of the Skaptoparoi (now
Gramada), IGBulg. IV 2236 in a dedicatory inscription to Caesar M.
Antonius Gordianus, around 238 AD.
IGBulg. II 804 (sic)
2nd -3rd AD; IV 1930, a
personal name from Serdica, after 2nd century AD.
orr IGBulg. IV 2163, see app. cr.
where G. Mihailov hesitated (“duobus modis suppleri possis”) which is the
correct one of either readings; see also IV 2050-57, 2162, 2164-2166,
Not in Detschew.
, (dat. ), IGBulg. IV 2330,
st rd
1 -3 century AD; Detschew 455: .
in Manov 2008, 109, found in Piperitsa,
district of Blagoevgrad, 2nd - 3rd century AD. See above Not in
Detschew.
in Manov 2008, 127, 2nd - 3rd AD, from the museum of
Blagoevgrad. Not in Detschew.
of the , IGBulg. III 1711,

, IGBulg. III 1445, 3rd AD Not in Detschew.
vel , IGBulg. III 1690 column b line
12. Mihailov preferred the form with -d-, while Detschew 457 put the one
with -l-. As it shows from the picture it sems that it is rather a lambda than
a delta; 202 AD.
, IGBulg. IV 2252, 1st -3rd century AD. Not in Detschew.
vel , IGBulg. III 1690 column b line
12. Mihailov preferred the form with -d-, while Detschew 457 put the one
with -l-. As it shows from the picture it sems that it is rather a lambda than
a delta; 202 AD.
IGBulg. II 731, 235-238 AD.; III 951, 3rd century AD;
(genetive) III 1601, 1st-3rd century AD;
III 1626, 1 - 3rd century AD;
st
1626, 1st - 3rd century
AD; III 1690 d 48, 202 AD;
III 1690 d 49, 202 AD; III 1690 d
65, 202 AD; III 1690 d 60, 202 AD; IV 2281, 1st - 3rd century AD; V 5560
is the same as III 1572, genitive ;
st rd
III 1593, 1 - 3 century
AD; V 5565, same as previous; V 5592, first half of 1st century AD; V
5595; V 5717, V 5723, 2nd - 3rd century AD.
IGBulg. II 722, 3rd century AD.
vel vel , IGBulg. III 1690 c 12, 202 AD.
, a name of a town, IGBulg. V 5011, 1st century BC – 1st
, IGBulg. IV 2025, an epithet of Apollo.
, IGBulg. IV 2213 = V 5867 where a new reading is
adopted based on a new inspection in situ of the stone after the
archaeologist R. Georgieva was lucky enough to recover it; see also
Janakieva (1988); see further on in the tribe name , II 501,
765, 766. Not in Detschew.
IGBulg. I 162, 3rd century AD.
-, an epithet of the Thracian Heros, IGBulg. IV 2115.
III 1690 e 18, 202 AD.
, a genetive (Filow 1934, 105), see

Chapter Three. (Filow 1934, 105), on a gold ring with an
image of a horseman, found in a 5th century BC burial. According to the
practice in Thrace (see the inscribed phialae from Rogozen and
elsewhere), this is a genetive case-form, translated to “belongs to
Skythodokos”. This is a perfectly good Thracian name with a second
element – dokos/-tokos (Detschew 145 and 462-463; Dečev 1960, 158-
76; Beshevliev 1965b, 13-14; 1-57; Brixhe and Panayotou 1994, 190,
198). J. Boardman, in his work on Greek gems and fingerings, advanced
an explanation of the inscription which reads, as follows: “an odd name
which should mean ‘friend’ or ‘host of the Scythian’” (2001, 230). He is
followed by other scholars in his opinion (IGDOP, 13-14; SEG 44, 1999,
no. 871).
IGBulg. I 164, 354 quater, 1st century AD. (?); III 1406, 3rd
century AD.
IGBulg. V 5067,
3rd century AD.
, IGBulg. II 587, Detschew 203, different wrong
reading:
, IV 2142, 2nd - 3rd century AD, epithet of Hera.
from Kresna, southwest Bulgaria (now in a private
collection: Ivanov 2004, 83-86). According to B. Gerov (1961, 325), the
epithet may be derived from the name of the ancient village that is present
day Baikalsko. During a recent visit to that place, upon my inquiries I was
told by the local people that there is an epithet applied to that place-name,
which at some point was used as the name for a small locality with a
somehow large pond, still existing, upon which a temple of Hera once
stood. It is located in a small valley, which because of the pond was called
Chuklyovo Blato (= ? Pond). Therefore, Chuklyovo may well be the
translation for Hera’s epithet, as B. Gerov put it.
in Manov 2008, 109, found in
Piperitsa, district of Blagoevgrad, 2nd - 3rd century AD.
an epithet of the Thracian Heros, III, 1726, 1st – 3rd
century AD.
, IGBulg. IV 2139, an epithet of the Thracian Heros.
, IGBulg. IV 2120; IV 2192.
, IV 2314, 2nd - 3rd century AD.

, IGBulg. III 1447.
IGBulg. III 1293, 3rd century AD, an epithet of the
Heros; V 5610 from Prileptsi near Kurdzhali, southeast Bulgaria (SEG 37,
610, SEG 41, 597). A. Chaniotis (1991, 301, no. 75) suggests that
is the beginning of the name of the dedicator rather than that of
a second epithet of the god.
IGBulg. V 5329, from Marcianopolis, 2nd - 3rd century
IGBulg. III 963, 2nd - 3rd century
AD; III 1392; III 1690 b 68, 3rd
century AD; III 1690 b 70, 202 AD;
, III 1690 c 18, 202 AD; IGBulg. IV 2337, 1st - 3rd
century AD
IGBulg. I 27 bis, 175, 4th - 3rd century BC (?), whose wife is
both occurrences not in Detschew.
IGBulg. III, 1204 3rd century AD;
rd
III 1132, 3 century AD; III 1690 b 39,
nd st
202; III, 1730, 2 - 1 century BC; the Thracian paradynast at ancient
Kabyle, a contemporary (and relative?) of Seuthes III, III 1731 (Elvers
1994), end of 4th century BC. Cf. in Chapter Two.
an epithet of Apollo, IGBulg. III 1832, 1st - 3rd century
AD. The reading belongs to the editor princeps I. Velkov 1934, 464-65.
Detschew 477 is following him:
However Mihailov believes that the right reading should
be from the epithet with shifted to to .26
, a place name, IGBulg. IV
2192.
, IGBulg. IV 2065, an epithet of Asklepios. Mihailov,
eodem loco, thinks that this should be derived from a place-name
*Spinthopura, unlike Detschew 477, who derives it from .
IGBulg. IV 2232-2234, 1st - 3rd century AD. Not in
Detschew.
IGBulg. II 838, 3rd century AD; III 1804,
st rd
1 – 3 century AD.
26
See Mihailov 1943, 65 and Schwyzer 1934-1970, 204, 210).
, IGBulg. IV 2122, an epithet of Zeus and Hera. Not in

Detschew.
IGBulg. II 765, 766, epithet of the Heros, 3rd century
AD.
IGBulg. IV 2050, an epithet of Asklepios. Dečev 480
derives it from Latin stramen. However, Mihailov, eodem loco, recalls
Procopius (De aedificiis 4. 4), where is a fortified place.
, IGBulg. III 1690 column b line 24, 202 AD.
, IGBulg. IV 2194.
vel III 1690 d 62, 202 AD.
, IGBulg. IV 1916.
epithet of the Thracian Heros, IGBulg. V 5608. Not in
Detschew.
(genitive) IGBulg. III 1172, 3rd century AD;
III 1190, 3rd century AD.
IGBulg. III 1854, 3rd century AD or later?
, an epithet of the Thracian Heros, IGBulg. IV 2246, 2nd - 3rd
century AD; IV 2305, 221 AD.
, IGBulg. IV 2026, an epithet of Dionysos. Not in Detschew.

, IGBulg. IV 2110, Apollini Tadeno on an inscription found
in Kabyle, Velkov 1991, 23 (= Filov, B., IBAD 1, 1910, 227 (non vidi)
and An. épigr. 1911, n. 17), an epithet of Apollo, ca. 150 AD. See also
Detschew 487.
- (. ) , IGBulg. IV 2216, an epithet of Dionysos, 1st - 3rd
IGBulg. V 5328.
IGBulg. I 354 quarter, 1st century AD(?). ,
IGBulg. II 489, , 2nd - 3rd century AD.
, IGBulg. II 536, from Glava Panega, 3rd

century AD; III 1064, 3 century AD; III 1516, 3rd century AD;
rd
III 1626, 1st - 3rd century AD; III

1690 b 17; IGBulg. III 1690 b 44, 202 AD;
III 1690 b 22; III 1690
d 51, 202 AD; III 1690 d 55, 202 AD;
III 1734, 1st - 3rd century AD; III 1786, 1st - 3rd
century AD; . III 1806, 1st - 3rd century AD; IV
nd
1916, after 2 century AD: …;. IV 1934, after
nd
2 century AD; IV 2015, 2 - 1st century BC; IV
nd
st rd
2022, 1 -3 century AD; IV 2112 ; IV 2133; IV 2214
st rd
1 - 3 century AD, IV 2074
; IV 2234 .
-, an epithet of the Heros, III 1879, 1st - 3rd century AD.
IGBulg. I 291, epithet of the Heros, 2nd century BC. (?)
Not in Detschew.
IGBulg. V 5462, 2nd - 3rd century
vel - IGBulg.
rd rd
III 1341 3 century AD; III 1317, 3 AD; III 1667, , 1st - 3rd
century AD.
, whose son is , IGBulg. IV 2331, 144 AD, in
an uncertain place on the middle course of the Strymon River; whose
wife’s name was , q. v. at Mihailov: “loco incerto in
valle Strymonis medii’ IGBulg. IV 2331…’6. …. littera , quae mutila
est, non tamen certa est, ?”. Cf. Detschew 79, 95.
, epithet of Apollo, IGBulg. g V 5617, 5618, 5619, 5620,
5621, 2nd century BC( ?). Not in Detschew.
( ) ( ), IGBulg. III 1355, 3rd century
AD.
, IGBulg. IV 2331, 144 AD from an uncertain place on the
middle course of the Strymon River; his father’s name was
and his mother was , q. v. at Mihailov: “loco incerto in
valle Strymonis medii”; IGBulg. IV 23316: “littera , quae mutila est, non
tamen certa est, ?”. Cf. Detschew 79, 95. See also the genitive
formTeretos below.
IGBulg. II 507; II 564, from Glava Panega, 2nd - 3rd century
AD; , II 844; III 1111, 3rd century AD;
IGBulg. III 1210, 2nd - 3rd century AD; III 1517,
3rd century AD; III 1593, 1st - 3rd century AD; III
1593, 1st - 3rd century AD; III 1595, 1st - 3rd century AD; ,
III 1690 c 41, 202 AD; III 1690 c 46, 202 AD; IV 2274, 150
AD; , III 1690c 61, 202 AD;
III 1690 d 52, 202 AD; III 1690 d 54, 202 AD;
[ III 1627, 1st - 3rd century AD, from Karasura. Böttger
and Halloff (1991, 478-81) claim a different reading based on their having
re-discovered the fragment that was first published by Dechev (1934, 76-
77, No. 9, Abb. 66). On the delineatio, provided by Dechev and used by
Mihailov in III 1627 it is clear that the vertical hasta next to P in the
restored [ could only be part of H or I, the latter not yielding to a
plausible reading. Böttger and Halloff’s suggestion for
orr is not convincing. Therefore, despite the difficulties,
is preferable. , son of Seuthes III and Berenike, III 1731and
th
Elvers (1994), end of 4 century BC; IV 2053; IV 2149; , 1st - 3rd
century AD; , 150 AD; IV 2291 , 1st - 3rd
st rd
century AD; 2330,1 - 3 century AD; IV 2338 IV 2338,
end of the 1st century AD, 144 AD V 5560=1572, 5505-
66, 5624, 5796, 2nd - 3rd century AD; 5829, 5667, 5638bis.
, a dedicant to Asklepios, on a votive plate, not far
from Pautalia, present day town of Kuystendil, southwest Bulgaria, 3rd
century AD.27 Cf. Chapter Two and Chapter Three. a recent
discovery on a gold signet-ring found in the Dalakova tumulus north of
Topolchane, district of Sliven. Dimitrov 2008a, 26-32. The inscribed text
reads: ( in the genitive case). The ring is dated to the
4th century BC (Not in Detschew).
IGBulg. III 1418, an epithet of Hera 3rd century AD.
27
two other fragments in the spring of 2006, thanks to the courtesy of Ms. Lidia
IGBulg. III 1005, 1010; 1348, 3rd AD;

III 1633, 1st - 3rd century AD.
, IGBulg. IV 2337, 1st- 3rd century
AD; , and in Manov 2008, 108, found in Piperitsa,
district of Blagoevgrad, 2nd - 3rd AD. Not in Detschew.
, IV 2263, 158 AD. Not in Detschew.
IGBulg. V 5288, 2nd - 3rd century AD.
Not in Detschew.
an
n epithet of Hera IGBulg. II 719.
3rd-4th
century AD. Cf. IGBulg. III 1222, 3rd century AD.
and Traidacus IGBulg. III, 1741 bis, 1st - 3rd century AD.
Not in Detschew.
, , III 1690 c
71, 202 AD.
IGBulg. II 860; , IGBulg. III 1709, 1st -
3 century AD;V 5890 from Kresna (not in IGBulg. IV), 1st - 3rd century
rd
AD.
, IGBulg. II 737, 2nd - 3rd century
AD.
an epithet of the Thracian
Heros in n
(sic!), from Royak, district of Varna Gočeva (1989, 113-
115); the author did not date the inscription; judging by the forms we
could date this inscription to the 3rd century AD. The author having not
supplied photographs of any of the inscriptions discussed in her
publication, we can only guess if both forms should
be accompanied by iota subscript (sic), as it is obvious that both words are
in the dative case (another (sic) is to accompany for the
substituted for ), or not. She rejected the first element of the epithet to
be related to on the ground of Detschew’s 515 comment28. It
should be mentioned that Detschew, l. c., has in fact included this name in
his Thracian corpus as such; his comments about it are that it could be
compared with the Phrygian , which is doubtful to be the same
28
Gočeva (1989, 115): “Wir könen es wohl kaum mit vergleichen, das
Detschew für phrygish hält”.
name because of the geminated ‘t’. There are numerous examples of forms
of names alternating an omicron with an omega; it is well known that
vowel quantity in Thracian, rendered by the means of the Greek letters, is
hard to be evidenced ever from the 5th century BC on. So, most probably
the first element of this compound epithet is .

AD.
IGBulg. IV 2027 from Opicvet, area of Sofia,

, 3rd century AD. The Z is inverted. Not in
Detschew.
, an epithet of Apollo Heros, IGBulg. I 281;
see , an epithet of Apollo, III 1713 bis, 1st - 3rd century AD.
PART II:
PHONOLOGY AND MORPHOLOGY

CHAPTER FIVE
INTRODUCTION: PHONOLOGY
This journey into the realm of the Thracians has been devoted to a
study of the variative linguistic sign aiming to determine the system of
those structures that could be made visible on the surface of mostlly
onomastic material.
This undertaking may also help us apply the methods used in other
linguistic studies and solve problems of considerable quantity, magnitude
and complexity.
As stated in the introductory notes, the method is that of defining and
describing layers and clusters from synchrony to diachrony.1
I an not certain if Labov is right when he says that he was able to
obtain the most in systematic detail from changes taking place in the
Germanic and Romance branch of I.-E. However, I have been facing a
linguistic reality, which may present changes that could complete his
evidence as far as sound change is concerned. For the Thracian has been
the subject of many publications, out of which the number of articles is not
decisively overwhelming compared to that of major publications or books.
Brixhe and Panayotou,2 who recently have made a valuable comment
looking over events, material, and publications, and who have severely
criticized the lack of consistency among scholars concerned with Thracian,
are of the opinion that Thracian was subjected to too many influences
which accounts for the complexity of the linguistic problems themselves. 3
And indeed, Thracian variations seen through those “crosscutting
influences”4 may appear even harder to study. Principles were sought for
in many studies, but none were able to present a uniform image of what
Thracian was. Nor were Brixhe and Panayotou able to point to a right
direction to follow. Because of the fact that the Thracians did not develop
a script of their own, their skepticism led them to asking the question who
1
See Labov 1994, 600, following the general strategy of using the present to
explain the past.
2
Brixhe and Panayotou 1994, 179-203.
3
Brixhe and Panayotou 1994, 185.
4
Labov 1994, 600.
122 Chapter Five
the Thracians really were. Therefore, were there any Thracians? If their
skepticism is justified, the fact that linguists are dealing primarily with
glosses and onomastic material should discourage everyone. 5
Under what conditions were the remnants of the Thracian language
shaped differently in one case or another, remains to be seen. Clearly
defined cases, which account for the real situation, will lead to a new
description of the facts. Dečev6 attempted one after Tomaschek,7 then
Georgiev,8 Russu,9 and Duridanov.10 We will use facts and data discussed
there. We will be concerned not that much with matching their
interpretations, but rather making our observations of sound change and
word shape of Thracian visible.
We don’t know for sure what the value of all Greek letters was in
rendering Thracian, but we should establish the guidelines, working
hypotheses, based on the notion that Greek easily assimilates words of
non-Greek origin.
In dividing the material under study, we should say a few words about
conclusions made after comparing the temporal layers.
The Ezerovo ring and the Kyolmen stele together with some coins give
the following:
No sign was used to denote orr . The evidence from the rest
of the inscriptions corroborates this. The several examples in Detschew’s
Die Thrakischen Sprachreste cannot persuade anybody that they are of
Thracian origin.11 Nor would names like , where the composed
character psi has a different value, e.g. that in , which would
make perfect sense.12
If one compares the frequency of signs used in the inscriptions dated to
the 6th and 5th centuries BC, the following will be found.
5
Brixhe and Panayotou 1994, 190 is certainly wrong to put just the “noms
propres”, i.e. personal names.
6
Dečev 1960.
7
Tomaschek 1980.
8
Georgiev 1983.
9
Russu 1967.
10
Duridanov 1985.
11
Only a few of those will be found in this file there, e.g. , an ethnikon,
where the y stands for another combination in the genuine name of this Thracian
tribe (why not or ?). The are mentioned in
Herodotus (7. 122). See also Detschew 444-46 for further evidence on this name.
The - ending is also dubious, since from the few inscriptions in Thracian we
have at hand there are no such Thracian endings.
12
See Detschew 39 s.v. and the lemma for this name in this book.
Introduction: Phonology 123
Based on our preliminary work on the inscriptions, we have now in

hand numerous examples that point to developments that exemplify the
history of the evidence of the Thracian language. Those developments will
be our objective.13
It is now clear that the earliest stages of the Proto-Indo-European (PIE)
have become reconstructable. This has brought coherent credibility to the
(PIE) phonological system through a number of revised and new examples
from various IE dialects, especially Anatolian.14
The phonological system came to be portrayed as that of a language
which underwent shifts of accent and, accordingly, also shifts of the
vocalic system, e.g. of sonants and laryngeals.
The phonological system of Thracian is reconstructable through the
phonological system of PIE and IE.
At first we have to deal with data that do not have secure
correspondences, and therefore in reconstructing Thracian we will recur to
a presentation that is not fully accountable for all shifts and intricacies of
PIE. Nevertheless, internal reconstructions seem to be the only possible
approach that helps to exemplify stages of linguistic developments. Let us
look at some examples:
The frequently used personal name Bi-t(h)-us is attested as early as the
4th century BC and throughout a period of seven centuries comes to us in
the form of or , the latter attested in the 4th century BC.15
Since Greek koine used the grapheme EI for /ī/, we can posit*bī-t(h)-us
from IE*wī-thu-s. The other possible way is to posit*bhi-thus> where bhi
as ‘two’ looks theoretically plausible. Compare PIE* h2embhi, ‘around’:
Skt. abhí-tas, Gr. , lat. amb-itus, ‘going around’ for the first
component, and forms with –dh-i, Skt. ádhi, “on”, Gr. “where’?,
ἰ “at home”, with -ti: Skt. íti, “thus”, Gr. , “against”, for the
second component.
13
Brugmann 1904, 28-30: The “Indogermanist hat es demnach zu erster Linie mit
den ältesten geschichtlichen Perioden der Einzelsprachen zu thun. In erster Linie,
nicht ausschliesslich. Denn bei der trümmerhaften und die Sprache ihrer Lautung
nach nur in rohen Umrissen zeichnenden schriftlichen Überlieferung alter
Sprachen muss oft in später Zeit Auftretendes zur Erläuterung des zeitlich
Zurückliegenden herangezogen werden…”
14
Lehmann 1993, 137.
15
Compare (sic!) in LGPN, where the earliest mentions are from Thasos
and Eretria on the island of Euboia in the 3rd century BC.
124 Chapter Five
The two possible interpretations rest on secure examples, e.g.

<*wri-ja, “town” in e.g. etc. Also <* wrughia, a
gloss, see Detschew 87 “Emmerkorn, Roggen”.
For one thing, Greek beta was pronounced [v] from the koine period
on, and therefore a confusion arose from the original*w >[F] in Greek, and
[b]. This contributed to many uncertain readings in Thracian, as well. In
our genuine Thracian inscriptions, b is used twice: in EBAROZE and in
BLABE. F disappeared at an early stage in Greek, but there are examples
of /b/ for /w/. In Thracian we are certain of these two developments and of
one more, the late “ou”, a weakened labialized “beta” from the 2 nd century
BC down to the late Antiquity.
If the sequence bī comes from*wī, “apart, in half”, then t(h)i- or it(h)
could be a suffix.
On the other hand, from the available evidence we have many
instances of names starting with BI and none with DI (meaning ‘two’) +
thi-.
The following forms have the same sequence that seems to belong to
bi-, “two, apart, in half”:
, etc.
From the list it could be concluded that - coincides with the variant
Zi- ( ) creating a confusion because of the widely spread name for
Zeus in Thracian. Thus, , , etc. are the most common.
A similar conclusion can be drawn about Greek.17
Another interesting example is <* Bī-sal-ēt-ēn-os, a
non-syncopated form of , a Thracian tribe from the lower
Strymon River. According to Stephanus of Byzantium (170. 16) their town
16
Cf. a Lakonian name (Dubois 2000, 45). In an excellent presentation
of hippo- “horse” in Greek names, the author is puzzled by that name finding it
“not easy: rather than translating it as “expert in horses”, it would be attractive to
compare the Vedic compound aśvavid-, “who finds and supplies horses” …” etc. I
will agree with Mr. Dubois as to that part in his statement that it is not easy to
incorporate this particular name in the list. For it belongs to the Thracian list and is
once again an excellent example of how the Greek language processed foreign
names. I shall continue with this particularly instructive case and point to a Greek
parallel in LGPN (sic!) and which is its Greek counterpart.
17
, , , cf. LGPN.
and region was ; there was also a river , a poetic

name for Strymon.18 While is a 4th century BC form,
is dated to the 1st - 3rd centuries AD. Whether the true
original form had been used in the inscription is uncertain.19
The location of the Bisaltae’s dwellings near the gulf of Salonica might
help with identifying the morphology and meaning of this very name as
well. As Detschew pointed out, the second element seems to be identical
with the name of the tribe of the Haleti (Detschew 12; 71). The Haleti are
said to have dwelled beyond the Sapaioi. The Sapaioi (Detschew 421-422)
were dwelling upon the Aegean coast across from Thasos. Both the Sapaei
and the Haleti settled in the region of the legendary Paggaios, a mountain
not far from the Aegean. Now, the compound Bei-salē-t-ai is composed of
bī- “apart, far, separated”, “on the other side”, “against”, and Sal-/Sald-
suffixed with –t, “salt”, “sea” (Watkins 2000, 72; Beekes 1995, 177
<*sēh2-l-s, acc.*sh2-é/-m, Lith. sól-(ymas), “salt water”), which yielded
Bī-sal-ē-t-ēnos. The long ē before the suffix -t- should go back to eh1>ē.
However, the sound change is in many cases governed by a process known
as analogy. And analogy is a systematic change according to model. 20
Thus ē/ānus is historically speaking, incorrect.21
Detschew 413 says that “über - siehe Zald- in ”.
On page 412 under the entry he quotes Jokl and Tomaschek
(II 2.79), who deduce from*Saldae, again from IE*sal-d,
“Salt”. Under (Detschew 173) he thinks of “Kaltwasser” or
“golden wasser” based on a source called Glava Panega.
Georgiev’s22 detailed analysis led him to IE*gholto- whence Thracian
sald/t-, “Gold” or “golden” corresponding to Old Bulgarian zlato, “gold”.
If we have to follow his “Charakteristik”, we should expect for the
resonant l to become ul in Thracian. 23 For IE*s we expect to see s in
Thracian24 and z between vowels, while for IE*gh it is or . Therefore,
the z is the norm for “gold” according to Georgiev.
As seen from the previously stated opinions, Dečev and Georgiev
somehow are reluctant to accept sald/t as the word for “salt” in Thracian.
18
Tomaschek, I 58.
19
See also Detschew 70-71.
20
See Beekes 1995, 72-79.
21
Szémereny 1996, 27-28.
22
Georgiev 1983, 1153-54.
23
Georgiev, 1163, 1167: “sold aus ghlto, ‘Gold’”
24
Georgiev, 1173, 1174.
126 Chapter Five
Relying on another root,*ghel “Gold”, well attested in all major IE

languages, they derive sal-d/t from the latter and attribute it the meaning of
“gold” instead of “salt”.
However, is this the case? A look at some IE roots shows the following
developments in Thracian. Indeed, instead of PIE*gh- vs. Thracian s (-z),
as proposed by Georgiev and Dečev, in his Charakteristik, there is IE*gh >
g in Thracian, e.g. through in , from the suffixed
zero-grade*ghl˚- yellow, green25, etc.
There is an epithet from lim(o)-enos, IE *loi-mo, Greek
, , a personal name26 (Euboia), Lat. limus, Celtic*(s)li-
m-no “slimy”, IE*IeH-, “to flow”,28 in an epithet for Aklepios in a place
27
called “vartop” (= ‘whirlpool’), where the soil is humid and giving in. Cf.
Old Bulgarian ВЪРТЪПЪ, “garden”, or “cave”, or “whirlpool”, where the
cult of Asklepios Limenos was worshipped.
The extended oi-grade is redueced to -i- and suffixed with -m- and
e/anus (for the latter see above Beekes, 75). The comparison with the two
genuine Greek words with the same meaning is very informative. It
allowed for us to make an observation about the structure and the word
formation in Thracian. Besides, it gives the opportunity to translate this
epithet: it goes Asklepios “of the Humid (=Clay) Places”.
The PIE Laryngeals and Labiovelars in Thracian

Needless to say that in this study of the Phonology of Thracian based
on inscriptional material, internal reconstruction is of utmost importance in
understanding the older stages of development, i.e. the history of particular
periods. The alternations that have been registered so far in our inquiry
lead to at least one obvious conclusion, i.e. much is unknown and left
unaccounted for.29 The systematic approach is therefore the best one to use
in describing the phonological system and its pre-history.
We know that the vowel alternation or Ablaut is a complex device in
the morphophonemics of the Indo-European languages. The very system
of the vowel gradation with alternating long and short vowels within the
25
cf. glei, Pokorny 19943, 362.
26
See LPGN.
27
Watkins 2000, 80: (s)lei.
28
Watkins 2000, 47, lei, “to flow”.
29
No trace of phonological developments to deal with laryngeals were found in the
reconstructions of: Dečev 1960; Georgiev 1983; Katićič 1976, on the Thracian.
See however Duridanov 1987b and c.
same root, that are found in different morphological categories, shows the
rôle it played in differentiating the structure of the root in PIE.30
The labiovelars were single phonemes as a result of the interaction of a
velar and a labial sound in the form of [w]. Originally Latin supplied the
fact that before the group qu in verse a syllable normally contains a short
vowel.31 Therefore, k+w, etc. was not what the labiovelars represented but
kw, a single phoneme.32 On the other hand, it has been generally assumed
that the “satem” languages i.e. Thracian, Sanskrit, Avestan, etc., lost their
labiovelars in their complete merge with velars. 33
Looking at some Thracian examples such as the word for “horse”, the
word for “four”, and the word for “far”, reveals the following developments.
PIE*h¹ek’w-o-34 is in Thracian*esb- or esp- in Esbenis, Ezbenis, Bet-
espios, Out-aspios. There is a tradition35 in deriving Esbenis, <* esw
because of the assibiliation of k. Consequently,*w changes to /b/ and the
unstable /w/ in its turn changes to /p/. While*w through /b/ has, in the
beginning of the Roman period (1st - 3rd AD), started to be spelled ou, /p/
appears here as an allophone of /b/ from* /w/. This form appears also as
asb- in the name of a little town, whose kometai (inhabitants) are called
Asbenoi. This last inscription was found in south-central Bulgaria near
three small villages called Koynare, Malo Koynare, etc., in Bulgarian,
meaning “Horse” or “of those dealing with (or in) horses” or “breeders of
horses”.
In Thracian, the group Sb- in Anlaut is unstable and changes to Zb- or
to Sou- (all graphemic variants). It seems that the preceding vowel through
its fluctuation a/e makes for the confirmation of /sb/ /sp/. One plausible
explanation would be to derive it from PIE*h¹ek’ w-o-s and posit h=a/e. In
principle, Thracian might well have prothetic vowels from PIE laryngeals,
but in the case of*h¹ek’ w-o-s, as well as in Argiske, Ergiske, etc.,
<*h2erg’- ‘white, bright, silver’ suggests that the e/a alternation might be
due to something within Thracian itself. 36 The appearance of a- in Asbenoi
beside e- in Esbenis is more difficult.
30
See Melchert 1987; Rasmussen 1999; Oettinger 2001.
31
Szémerenyi 1996, 67: “which cannot therefore represent a consonant group”.
32
Opinions are divided. See Szémerenyi 1996, 67, n. 1 for detailed bibliography.
33
Szemerenyi 1996, 66-67: “The infinitive of OCS женон (IE*gwhen-) is гънати,
in which ъ can be explained only as a reflex of the labial element of the IE
labiovelar. These instances, few as they are, are sufficient to refute the thesis that
labiovelars had never existed in the satem languages.”
34
Watkins 2000, 23.
35
Dečev 1960, Georgiev 1983.
36
Dimitrov 1994.
128 Chapter Five
Another development of*asw- might well be Eisa- in Eisenos,

Eisatralis as a further continuation of sound changes due to ongoing
phonological rules, where Esb- loses its unstable beta (from*/w/) and a
compensatory lengthening occurs, thus yielding ei, a secondary
diphthong.37
Accordingly, for “four” from IE* kwet(w)r˚-, a non-suffixed zero-
grade,38 in Thracian we have , ,
,39 since PIE* kw gives k unlike the other satem languages
(as Thracian which is considered a satem language ), i.e. Avestan aśva
“horse”, Skt. çatvaras “four”, from*kw-el or rather*kw-l˚-ai-
“far”, Greek “palaios”.
So for Keilaidebenos from*keila and*deba ‘town, hill-fort’40 a
plausible translation could be ‘Old Fort, Old Town’.
In Gonimaseze I see IE*gwn˚-i “woman, wife”. This instance is secure
through the parallel translation guné in Greek, and in my opinion is
sufficient to refute the thesis that IE*gw in Thracian is represented by k.41
In , we see PIE*gwelh2- 42 “gland”, Bulgarian “желъд”,
(=Желъдово).
For PIE*gwh there is a good example in Germanos, Greek thermos.43
There are still many problems remaining, but for a language with
limited data like Thracian one cannot expect to find explanations that fit
perfectly.
37
Dimitrov 2007a.
38
Watkins 2000, 45.
39
Cf. the place name Chertigrad < Chetri-grad “Four-town” in Bulgarian, possibly
a Bulgarian “translation” in its turn of the Greek version of in
misunderstanding of the component in Thracian (= “town” in
Greek).
40
Cf. Watkins 2000, 19:*dheu-, and Pokorny 19943, 261.
41
Cf. Georgiev 1983.
42
Watkins 2000, 34.
43
I agree here with Georgiev 1983, 1173.
CHAPTER SIX
THE COMPONENTS
In the Thracian compound names a number of well-defined

components1 are being identified.2 Some of them are being constantly used
within the system of the word-formation, some are emerging sporadically.
Many of the latter have parallels in the neighboring lands, such as Greece
and Turkey. As already pointed out in A. Morpurgo Davies’ presentation,3
onomastic compounds do not necessarily follow the development of
lexical compounds. For example, Thracian shows a peculiar
4
sequence -ktn- vs. e.g. . The sequence with -ktn- is in fact a
shortened variant of . Gemination is another common
phonological process, which is known to have played a significant rôle5 in
shaping personal names. This is a rare phenomenon in Thracian, and if this
the case, it is certainly not connected to the structure and the nature of the
genuine Thracian names. Most of the occurrences of geminated
consonants are either from a later period (after 1 st century AD) in Thrace
proper or again from the same period in onomastic material that is
Anatolian, in general.
The place name Autoptistai is supposedly a Thracian compound but
from the list of the component elements that contain -pt- and -ist- none is
associated with Auto- nor is -ist-/-st- producing anything closer or similar
to that place name. As we have already mentioned and pointed to other
instances as regards “misinterpretation” and “misunderstanding” of
1
See Georgiev 1983, 1151-78.
2
In this part because of the nature of the material under study we deal with
phonemes and morphemes. The compound nouns contain the core of the Thracian
phonemic distribution. This is the testimony of the genuine “Thracian sound” that
ultimately comes to persuade us that the combination of the phonemes and their
distribution represent the Thracian language to the best of our knowledge with the
mediation of the Greek lettering and language logic.
3
Morpurgo Davis 2000, 18.
4
See Thracian Numerals under “Ten”.
5
Cf. Vottero 1985 and Morpurgo Davies,18.
130 Chapter Six
Thracian words, sometimes onomastic compounds are not easily

recognizable for they simply do not make sense. 6 Yet another example of
both gemination and misunderstanding is Aulosell, where the latter
“simply does not make any sense” 7.
Therefore it is important to determine the structure of the Thracian
names in stricter terms. Here are the following elements that are being
found using the method of internal reconstruction. 8
1. –aba in: Abezelmis, Sauthaba.
2. Ama- in: Ama-istas, Ama-siakis, Ama-tokos, Am-
iazenis.
3. –ages in: Tasages.
4. Ai- in: AIRAI-, Aidezuris, Aisumoes.
5. Apro- in: Apro(s/n).
6. Apta/e in: Apte-kenthos.9
7. Arg- / Erg- in: Argiske, Argedauon, Ergiske, ,
Ergisenoi.
8. Athi/thē/ti/si/sē in: Athe, Athioutike, Athys, Aithazie,
Anthienos, Athyparenos.
9. Ard/z/s- in: Ardilenos, Arzenos, Arseos(n), Arse[lenos]
10. Arei/Ari- in: Areibalos, Ar- mou-lene.
11. As- in: Asenos, -asdenos, Ases, Asos, Asoutike, As-
doules.
12. Asb-/ Ezb- in: Asbenoi, Ezbenis.
13. Athe- in: Athe-, Athioutike, Athyparenos, Athypara,
Athys.
14. Auza- / Duza-/ Diza- in : Auza-kenthos, Auzatralis,
Dyza-tralis, Diza-
15. Aula- in: Aulariokos, Aularkenos, Aula(r)henos,
Aulenis, Aulou- Aulou-denthes, Aulou- zanis, Aulou-zenes,
Aulou-zenes, Aulou-zenis, Aulou-zenis, Aulou-kenthos,
Auloukenis, Aulou-kenis, Aulou-menes, Aulou-poris, Aulou-
sadenos, Aulou-sades, Aulou-sedes, Aulou-soukos, Aulou-trais,
Aulou-tralis
16. Aulo- in: Aulo-sell
17. Ap- in: Ap-sinthi-os
6
Masson 1966, 253 sq. calls such names “noms irrationels”. Cf. Morpurgo Davies
2000, with extended bibliography.
7
Morpurgo Davies 2000.
8
The author is trying to present an exhaustive list without withdrawing important
items. However, it is conceivable that this has been done within reasonable limits.
9
I think that here belongs the “edited” form Autoptistai.
The Components 131
18. Ba- in: Bazoparenoi, Ba-tou-deis, Bazitenos, Bazobalis

19. Balos- in : Areibalos, Bazobalis, Dreibalos, Zaibalos,
20. Barg- in: Bargidenthes, Bargos/ Barsos
21. Bazo-/ Baso-/Basso-/Besso- in: Bazitenos, Bazobalis,
Bazopara, Bassos, Bassous, Bas-ta-keilas, Basta-kiles, Bas-to-
keilas, Besso-denthes, Bessoparenoi
22. Bde- in: Bdekyrenos.
23. Bi- / Bei-/ Bē- in: Beibous, Beithas, Beithes, Beithipes,
Beithykeilos, Beithys, Beithytralis, Beisaletenos, Bemarkos,
Betespios, Biartas, Bizyenos, Birou-, Bisdes, Bithikenthos,
Bithys, Bitoulousos, Byaiparene
24. Bel- in: Belanos, Bel-, Belos
25. Bendi- / -benzi- in: Bendizeta, Bendikos, Bendis,
Bendios, Bendiparenos, Debabenzis. See also Oendis.
26. Be-o- in: Beo(s) as far as it is slightly different from
Bi/Be and occurs in a place-name.
27. Berenikē.
28. Blō-sti- in: Blosta, Blosti.
29. Bolb- in: Bolbabria.
30. Bor-/Bour- in: Borkeithias, Bourgeilos, Bour-dapa,
Bour-dopa, Bour-depenos, Bour-theithes. See also the alternating
Por-/Pour.
31. Bos-/Bōs- in: Bousipara, Bosagyrenos.
32. Bra-si- in: Brasitralis.
33. Bria- in: Bolbabria, Briazenis, Kardibria.
34. Bri-/Brei- in: Brizenis, Breilos, Brilos, Brikeses,
Bri(n)kazenis, Bri(n)kainos, Bri(n)kazis, Bri(n)kazeis,
Bririsoulas, Bryzos, Brylouzies.
35. Bre(nt)- in: Brentoparenoi.
36. Brou- in: Brouthenis.
37. Bryl- in: Brylouzies.
38. Bou(s)- :Bouzaē. Whether it is identical with Bos- or not,
depends on the lack of examples to accept this only occurrence
with intervocalic z from*s. We might consider the fact that here
after bou- a grammaticalization of zē with a typical eta ending for
feminine, without iota subscript in the dative case, might have
occurred. The inscription is dated to 144 AD. The examples with
Bos- are from around 202 AD. From that same time is Buzēs,
which no doubt belongs to the same “Sippe”.
39. –gai-/-ge-/-ga-/-gei/-gu in: Bourgeilos, Gelou-para,
Geseilas, Gaipes, Garoulos, Geiston, Gezoteizis.
132 Chapter Six
40. –gar- in: Garoulos, Geizagarenoi, and Garytinos.

41. –ge-/ gi-/ -gē- in: Gei-ka-tienos, Geikeilas, Gei-ke-
thienos, Gei-ke-senos, Geigon, Gesienos, Geskouporis, Geta,
Getai, Gepepyris, Gaipes, Gyze.
42. –gel- in: Gelou-para.
43. –geilos in: Bourgeilos.
44. –geni- in: Geni-kelas, Geni-keilas.
45. Gep- in: Gepepyris.
46. –ger- in: Germanos.
47. –ges- / -geis- in: Drigesos, Kaigeisos, Kaigeis,
48. -gethes in: Zigethes.
49. –geu- in: Geupasa, Geupasenos.
50. Gil- in: Gilpyris.
51. Gi-ti-s.
52. Go- in: Go-a-zeira, Go-l-eis.
53. [G]our- in: [G]our-theithes.
54. Gre- in: Greseitai.
55. –gyr- in: Bosagyra.
56. -da-/-dai-/-de- in: Aidezyris, Bargidenthes, Bdekyra, de-
enos, Tera-deenos, Dada in Dadaleme, Daleporis, Dezalas,
.
57. –daba / Deba- in: Dabeis, Daba-topeios, Daba-topienos,
Dauterenos, Debabenzis, Karidaba, Keiladebenos, Keiladeenos,
Keiladeouenos, Keilaibeithys, Keilaideenos.
58. –dakos, -dokos (alternating with –tokos) 10 in: Trai-
dakos, Sparadoko, Skythodoko(s), Amatoko, Metoko, Saratoko.
59. Dar-/Der- in: Darzalas, Daros, Daroubyr-, Derzalas,
Derzelas, Derzis, Dernaios, Dernaes,Derronikon.
60. –dates in: Touridates.
61. Daezeris, Daezis, Daikosis, Daises, Dakia, Daleporis,
Daleporis, Dalaitralis, Daletralis, Dalezalis, Damalsos, Damas
62. -de - in: Dedegnos, Dezos.
63. –denthē- / denthe- / Dentou- in: Aulou-denthes,
Bargidenthes, Besodenthes, Bessodenthes, Denthaibaris,
Denthebaris, Dentousykos, Dentysykos, Kaldenthes, Kardenthes.
64. –dei- / Di- / -di- / Doi- / Dou- / -Dy- in: Batoudeis,
Deidas, Deides, Deidikyros, Deidis, Deidos, Deizas, Deizezenis,
Deizezes, Deios, Deisoros, Deitouzaipou, Diazenis, Diarzios,
10
See Brixhe and Panayotou 1994, 190 according to whom the “flottement” d/t is
due to the influence of Greek -dokos on the authenthic Thracian -tokos.
The Components 133
Diasenis, Diaskenthos, Diaskouporis, Diatralis, Dibeithys,

Didios, Didykaimo, Diernes, Diza, Dizazelmis, Dizazenis,
Dizakenthos, Dizalas, Dizalis, Dizaporis, Dizas, Dizatralis,
Dizdon, Dizza, Dizouzos, Dizoulos, Dikebalis, Dikebalos,
Dikedos, Dikenais, Dilaes, Dilzes, Dilles, Dimes, Dindas,
Dinedinis, Dineilos, Dinizetes, Dintas, Dyntos, Dinikenthos,
Dinios, Dinis, Dintizila, Dios, Dioskous, Diosoura, Diosporis,
Diouzes, Dioukeilas, Diouporis, Diourpos, Diporis, Disatralis,
Diskodouratera, Diskous, Disyros, Diostralis, Doidalsos,
Dooupyros, Doudes, Doules, Douspieris, Doutoubee, Dydes,
Dydios, Dyzatralis, Dylazenis, Dysyra, Dytouporis, Dytoutralis,
Thintas.
65. Deo- in: Deospouris.
66. Desa- in: Desakenthos, Desoukantene.
67. Dis- in: Disloias.
68. Do- in : Dodoparon, Doleporis.
69. –dopa, -dapa, -depa in: Bour-dapa, Bour-dopa, Bour-
depenos.
70. Dor- / Dour - in : Dorzas, Dorso(s). Dorzenthes,
Dorzizenis, Dorzinthes, Dorpanas, Dorsas, Dortazenos,
Dourazeris, Doroutelmis.
71. Dos- in : Dositralis.
72. Drei- in : Drei…ouzeris, Dreibalos, Drenis, Dres-,
Dresapaike, Drigesos, Drizenis, Dringestas, Dringistas .
73. Dro- in: Droles, Drodeges,
74. Dy- in : Dynt(os).
75. Ebar- / Ebr˚- : Ebaroze, Ebryzelmis, Ebrytelmios,
Ebrenis.
76. Asb-/ Esb- / -esp- / Ezb-/ espios in: Asbenioi, Ezbenis,
Esbenis, Betespios.
77. Eithi – / Eisa- in: Eithia, Eithis, Eithialas, Eisasemes,
Eisatralis, Eisenos, Eitizenis, Eitiosaros, Eitrizene, Eifis,
78. Epta - / Epte- / Apte- / Apta- in: Eptazenis, Eptaikethos,
Eptaikenthos, Eptaitralis, Eptalos, Eptaporis, Eptapys, Eptatralis,
Eptezenis, Epteikenthos, Eptekenthos, Eptethisis, Epteizenis,
Eptelos, Eptenis, Eptepyris, Eptesykos, Eptetralis, Eptekenthos,
Eptelos, Eptemeneka, Epteporis, Eptepouos, Eptepyris,
Eptesykos, Eptesyhos, Eptetralis.
79. Arg- / Erg- in: Argiske, Ergiske, , Ergisenoi.
80. –estra- in: Estrakeenos.
134 Chapter Six
81. za- / ze- : Zaein, Zaibalos, Zaika, Zaineila, Zaipa,

Zairdainos, Zamazis, Zauroosene.
82. –zalas in: Darzalas, Derzalas, Derzelas, Dizalas,
Dintizila.
83. ze/se in: Deizezes, Dilzes, Dilles, Diouzes, Sese, Zeilas,
Mosses.
84. –zelmis/-telmis in: Abezelmis, Dizazelmis, Doroutelmis,
Ebryzelmis, Ebrytelmios.
85. –zenis / -senis in: Aulouzenis, Breizenis, Briazenis,
Deizezenis, Diazenis, Diasenis, Dizazenis, Dorzizenis, Drizenis,
Dylazenis, Ebryzenis, Eitizenis, Eitrizene, Eptazenis, Eptezenis,
Epteizenis, Zenis, Zerazenis.
86. -zenthes in: Dorzenthes, Dorzinthes.
87. –zēta in: Bendizeta, Dinizetes.
88. -zie / -zi- in: Brylouzies, Derzis, Zamazis, Zeizis,
Zeizodynon, Zeindroumenos, Zeines, Zeinoporeis, Zeipas,
Zaisisos, Zeiseios, Zeiskoris, Zeisodynon, Zeitralis, Zias, Ziaelas,
Ziakatralis, Ziamarke, Ziametralis, Zigethes, Zieris, Ziles,11
Ziouris, Zipys, Ziseisos, Zisyras, Zoitha.
89. Zou- in: Zoulgeses, Zoulemes, Zoumama, Zoura-,
Zourazeis, Zourazeris, Zourazes, Zoures.
90. –zouresēs in: Trazyreses
91. –zeis/-zis12 in: Aithazie, Brinkazeis, Brinkazis,
Brinkozis, Dazis, Brylouzies, Zeizeis, Zeizis.
92. –zeris in: Daezeris, Dourazeris, Drei…ouzeris, Zeris,
Zerdenos, Zerklene.
93. Zyde- in: Zydeenos.
94. Zykou- in: Zykoudaises.
95. Zylm- / Zym- / Zymz- / Zyml- in: Zylmydrenos,
Zymdrenos, Zymzdrenos, Zymlydrenos.
96. –zyris in: Aidezuris.
97. -theithes in: [G]our-theithes.
98. –ipēs in: Beithipes, Daba-topienos, Gaipes, Zaipa,
Zeipas, Zipys.
99. –istas/ -isde in: Autoptistai, Ama-istas, Geiston,
Istrianoi, Istros, Karistoron, Karistorenoi, Mestas, Mesteikenthos,
Mestikenthos, Mestoes, Mestos, Mestulas, Mestupaibes, Mostis,
Peistos, Pistous, Sikinystyrenos, Bisde.
11
For a new reading of IGBulg. IV 2289, see Bojadzhiev 2002a.
12
See Boyadzhiev 2002b.
The Components 135
100. -k-/-ik-/-sk-/-isk- in: Argiske, Dioskous, Ergiske,

Hrgiske, Asoutike, Skostoko(s), Eskados.
101. ka-/ki- in: Amasiakis, Aptakenthos, Artakene,
Aulariokos, Auloukenis, Aulousoukos, Barkeithios, Kaino(s),
Kabyle, Kakous.
102. Kal- / Kar- in: Kaldenthes, Kardenthes, Kallatianoi,
Kallatis.
103. Kam- in: Kamoles.
104. Kanz- in: Kanzourenos.
105. Kapr- in: Kaprenos.
106. Kara- in: Karabasmos, Karbateis.
107. Kardi – in: Kardibria.
108. Kari- in: Karidaba, Karistoron, Karossos.
109. Kars- in: Karsenos, Karsimienos.
110. Kase- in : Kaseiporis.
111. –kel/-keil-/-kil- in: Bastakelas, Bastakiles, Bastokeilas,
Bo skeilas, Dioukeilas, Geikeilas, Genikelas, Genikeilas,
Keiladebenos, Keiladeenos, Keiladeouenos, Keilaibeithys,
Keilaideenos, Keilaiskenos, Kelepyris.
112. Kell- in : Kellai, Kellēnos.
113. Kemp- in : Kempēnos.
114. Kendi-/ –kenthos, -kethos in: Apte-kenthos,
Auzakenthos, Auloukenthos, Beithykenthos, Desakenthos,
Diaskenthos, Dizakenthos, Dinikenthos, Eptaikethos,
Eptaikenthos, Epteikenthos, Eptekenthos, Eptēkenthos, Kendi.
115. Kers(e)- in: Kersebleptēs.
116. Ko- in: Kotys.
117. Kozi-/ Kozei- in: Kozimasēs (or Kozeimasos from
Serres, Greece).
118. –kyr-/ kour- in: Bdekyrēnoi, Blēkouros, Keirpara.
119. –l-/-le-/-lē-/-lou- in: Arselēnos, Arsilēnos, Geseilas,
Garoulos, Da-lai-tralis, Da-le-tralis, Dalētralis, Dalēzalis,
Disloias.
120. ma-/me- in: Armoulēnē, Gonimasēzē, Mezēnai,
Roimetalkas.
121. –mama in: Zoumama.
122. Mark- in: Bēmarkos, Ziamarkē.
123. mas-es/-os in: Kozimases (or Kozeimasos from Serres,
Greece), Gonimaseze.
124. -menēka in: Eptēmenēka.
125. -mētralis in: Ziamētralis.
136 Chapter Six
126. Mos-(?) in: Mossēs 13, Mostis.

127. –mou-/-mo- in: Armoulēnē.
128. *nana, nena, Neinēnos, Neinē, Nēnas.
129. -nd- / –nt- in: Dindas, Dintas, Dyntos, Brentopara,
Desoukantēnē.
130. –para in: Athyparēnos, Athypara, Bousipara, Brentipara,
Dodoparon, Geloupara, Keirpara.
131. –pasa/pasē in: Geupasa, Geupasēnos.
132. Patr- in: Patraos.
133. –pe-/-pi- in: Beith-ip-ēs.
134. –pouo- in: Eptēpouos.
135. –poris/-polis / -pyris in: Aulouporis, Gēpepyris,
Gēskouporis, Daleporis, Deospouris, Diaskouporis, Dizaporis,
Dizaporis, Diosporis, Diouporis, Diporis, Dooupyros,
Dytouporis, Doleporis, Eptaporis, Ketriporios, Reskouporis,
Eptepyris, Eptēporis, Eptēpyris, Zeinoporeis, Kaseiporis,
Kelepyris.
136. Re- in: Reskouporis.
137. –er / –ro-/Roi-/-rou- in: Garoulos, Zieris, Ziouris
Rolistene, Roigos, Roimetalkas.
138. Sa-/ Saba- (in Sauthaba) 14, Sad-/Sath- in: Sadalas, -
sathēs, Satoko, Sauthaba.
139. Sara- in: Saratoko(s)
140. -selm-/ –telmios in: Ebrutelmios. See zelm/selm above.
141. –si-/-ti-/-thi- in: Gesiēnos
142. –so-/-sou-/-sa-/s- in: Brisoulas, Geseilas
143. skē-/sko- in: Skērēzis, Skokelēnoi, Orēskion, Argiskē,
Ergiskē, Hrgiskē, Asoutikē, Gēskouporis, Diaskouporis,
Diskodouratera, Diskous, Zeiskoris, Reskouporis, Skostoko(s),
Skythodoko(s).
144. Spar-/ Spara- in: Spartokos, Sparatokos.
145. –sykos/-syhos in: Dentousykos, Dentysykos, Eptesykos,
Eptēsykos, Eptēsyhos.
13
Here it seems to me that the expressive gemination is due to a graphemic variant
of the genuine Thracian*/s/, which in intervocalic position turns to /z/.
14
Because of the word-formation of Thracian names, we can assume that this
name is a composite one, where its components are * in ,
on the one hand, and , on the other.
The Components 137
146. –soura / – syros in: Disyros, Diosoura, Dysyra, Zisyras,

Zousyreithas, Tōtou-soura.
147. tag- in: -tagēnos.
148. tad- in: -tadēnos.
149. tazou- in: -tazy-ēnos.
150. –talde/ -ou in: -talde.ouēnos.
151. –talka- in: Roimetalkas.
152. tar- in: Tarsoziē.
153. tas- in: Tasēs, Tasagēs.
154. tata-/tatou- in: Tatazia, Tatouporis.
155. –telmios in: Doroutelmis, Ebrutelmios, Zourotelmis,.
See zelm/selm above.
156. teuta- in: Teutamas.
157. tēr-(a)- in: Tērēs, Tēradeēnos.
158. tidi- in: Tiditēnē.
159. tē-/ti- in: Tiouta, Tyntenon.
160. –toko- in : Amatoko(s), Saratoko(s), Satoko(s),
Mētoko(s), Skostoko(s).
161. ton- in: Tonos.
162. tosk- in: Toskos.
163. Tu-nt- in: Tyntenon.
164. touri- in: Touridatēs.
165. tra-i- in: Traidakos, Tra-zuresēs.
166. Tralis in: Beithytralis, Daletralis, Diatralis, Dizatralis,
Diostralis, Disatralis, Dositralis, Dyzatralis, Dytoutralis,
Eisatralis, Eptaitralis, Eptatralis, Eptetralis, Eptētralis, Zeitralis.
167. Ture- in: Tyrelsēs.
168. Tōtou- in: Tōtou-soura.
169. ōita.
170. ōrdi-/ourdi- in: ōrdianos.
171. pt- in: Autoptistai, Aptakenthos, Eptazenis, Eptaikethos,
Eptaikenthos, Eptaitralis, Eptake(n)thos, Eptakethos, Eptalos,
Eptaporis, Eptapus, Eptatralis, Eptezenis, Epteizenis,
Epteikenthos, Eptekenthos, Eptethisis, Epteizenis, Eptelos,
Eptenis, Eptepyris, Eptesykos, Eptetralis, Eptēkenthos, Eptēlos,
Eptēmenēka, Eptēporis, Eptēpouos, Eptēpyris, Eptēsykis,
Eptēsyhos, Eptētralis, Kerzo-bleptēs, Skeptoi, Skaptoparēnoi.
172. –bl- in: Blēkouros, Blōsti, -bleptēs.
CHAPTER SEVEN
THRACIAN NUMERALS1
“Two”
“Two” in Thracian is from *dwō(u),*dwei2, e.g.

IE
th
(4 century BC), Greek
, Lat. duplus. The IE* /d/ as in the word for “light”* djēus is in
many cases changed to /z/ after assibilation in Thracian. However, this is
not the case with the word for “two”.
“Three”
The word for “three” in Thracian seems to be a first component in the
following compounds:
The numeral belongs to PIE *trei- which in Thracian is treated either

with /d/ or with /t/, as e.g. in from *tre- with the grapheme
AI representing /e/. Chronologically, /d/ precedes the shift of /d/ to /t/ 3
which is well attested in examples from the Roman period and
geographically distributed in Scythia Minor and north of the Danube
River. I think that this is due to a dialectal feature and under the influence
of the Latin language in the northern territories of the Thracian lands. 4
The IE */t/ is in Thracian rendered with /d/, which phoneme was
pronounced as a strong [d].5 A good example here would be the form
1
For a different approach see Polomé 1986, 185-9.
2
Pokorny 19943, 229.
3
For a different opinion see Dečev 1960, 149 ff., Georgiev 1983, 1172-73 and
1183, Duridanov 1987c, 57-64.
4
See also Detschew 522 at Triballi.
5
See Brixhe and Panayotou 1994, 198-99, in their severe comment on the
Lautverschiebung: “En vérité, cette hypothèse est, avec l’évolution supposée de*o
en*a et les migrations tribales ou ethniques, l’un des outils les plus évidents de la
140 Chapter Seven
LETED and LED from the 6th century BC Kyolmen inscription as third
person singular ending.
“Four”
In the case with “four” from IE * kwet(w)r-, the non-suffixed zero-
grade6 finds its continuation in the Thracian language in the personal
names ,7 since PIE *kw gives
k unlike the other satem languages8 (as Thracian which is considered a
satem language ), i.e. Avestan aśva “horse”, Skt. çatvaras “four” etc.
From the examples in Detschew 243, not attested in the inscriptions
from Bulgaria, we observe the usual assibilation of /t/ to /d/ as in e.g.
. This also the case with the
name of the phyle in Philippopolis and the variants
, and hence , an epithet of
the Thracian Horseman. All these examples date to the Roman period, 1 st -
3rd century AD and are attested in inscriptions from Philippopolis. The
cluster –nd- comes from the original –tr- after /t/ was assimilated to /d/
(with a Greek treatment as e.g. the epenthetic -n- before –g-) owing to the
fact that Thracian in that region was subjected to strong Hellenization ever
since the 4th century BC or earlier.
“Five”
From IE *penkwe,9 assimilated to kwenkwe in Celtic and Italic, we have
ginka as the etymon for “five” in Thracian, e.g.
, most probably meaning
“fifth”, Greek “fifth”. Compare Greek Lat. quinque
“five”, quintus “fifth” and Old Irish cōic “five”.
“Ten”
manipulation linguistique et historique. . . ”. And, on p. 199 they speak of strong

voiced stops *b,*d,*g and strong voiceless stops *p,*t,*k comparable in their
pronunciation with those in present-day German. (See Philipp 1970, 56-57).
6
Watkins 2000, 45.
7
Cf. Chapter Five, footnote 39.
8
See further on in Phonology under “Labiovelars”.
9
Pokorny 19943, 808; Watkins 2000, 64.
Thracian Numerals 141
From the oldest form in PIE*dekk m˚10 we have different developments

in the IE languages, such as e.g. Old Bulgarian ДЕСАТЪ, Greek ,
Skt. daśa, Lat. decem from*dek-, as e.g. perhaps in Thracian
. The confusion in the Thracian names between the component

for Di- (=Zeus) and the already mentioned Dei-, or Dai- as different
graphemes, create difficulties, though we can clearly see the difference
between from*Ze- (to Zeus) on the one hand, and for the
land of the Dakoi, on the other hand. We argue that the Di- component in
above does not belong to Di- for ‘two’ since there is no -kedos
as a second component in our list and therefore, we cannot divide this
compound into Di- and -kedos, but should read Diked-os. Following this
logic we can isolate now a stem*deken-t-, suffixed zero-grade *dektn-, or
*deko-/e- and *dike- without the cluster -tn-.
Perhaps we need to add a cultural note here explaining the use of the
etymon for “ten”. Daikosis is a name of a phylarchos, Dekentos “tenth” =
“Decima” as a “third” name of Flavia Modesta, a dedicant,11 Dikebalos, a
Thracian priest and king, etc. The mythological and ritual meaning of
“ten” in Thracian Orphism has long been discussed.12
10
Watkins 2000, 15; Pokorny 19943, 191.
11
Mihailov in IGBulg. IV 2231 from Kocherinovo thinks that –kentos is in fact -
kenthos, which is a possibility. However, why not Di-kenthos? Again, De- for Di-
is not impossible, but one would normally expect an example with Dē- in the list.
12
Most recently see Fol 2002, 269-72 with the references to his previous works.
CHAPTER EIGHT
ANALOGY AND CONCLUSION
Analogy
Analogy is a process that plays an important rôle in shaping the
Thracian material. This kind of language change consists of developments,
which could be interpreted differently. 1 Sound change is deemed regular
and causes irregularity, while analogy is irregular and operates in conflict
cases. This interference among others in language practice impinges
sequences, pre-determines chronological boundaries and takes part in
transformations of morphological character. Analogy works on already
existing models.2
Analogy reconciles inexplicable variations. For example, from PIE
*deiwos in Thracian there are
For - we assume
<*Dū-s-, the same in ( ) from Dū- suffixed with -nt-,
d: later changed to Z, and thence many forms with Z, see The Evidence.
We have therefore two developments from PIE dei-w-o-s: 1) dī-w-os
and 2) dīs.
In some examples, - was preserved and was still productive in
producing Dios from Dīw-os after *w has been weakened in intervocalic
position disappeared.
From IE to Thracian:
, etc.
The two isolated - and - yield in Gen. (from Dei-s-os > dīs),
Dative Di (from dei-w-i → dī), Acc. (e.g. ).
1
Sturtevant 19684, 94; Anttila 1989, 94.
2
Szémereny 1996, 27: When it occurs, purely phonetic developments in
accordance with the sound laws are for the most part suppressed and obscured, e.g.
ē/ānos in Thracian: Galabov 1964, 1-64.
144 Chapter Eight
Conclusions
The aim of this study was to produce new and reliable results using a
new method. Having concentrated on phonological changes visible
through various cases, an attempt was made to describe developments and
work with synchronic facts. Linguistic change and reconstruction filled
much of the texture of the investigation. Taking into consideration the
complexity of the material, I adopted an approach to suit both the
presentation of new material and the use of synchronic linguistics. The
cultural and historical information of the updated material complemented
the efforts to achieve a proper description of the changes in the
phonological sequences. Since this study was undertaken in continuation
of Tomaschek’s ethnological outlook on the Thracians and their language,
reaching proper understanding of Thracian variations served as a
consolidation of the belief that the method of synchronic and diachronic
analysis is hopeful enough to provoke new studies in the future.
In concluding my work, it is important to reiterate my ultimate goal,
which is to give new material but also to contrast it with the cultural and
historical background. Thus, the strict and hopefully rigorous way of
considering only the direct evidence is justified in utilizing data from
inscriptions found in Bulgaria enhanced with some extraterritorial ones.
The opening of this chapter with the topic of analogy is a means to
conclude this study of so many irregularities and to describe properly what
has been achieved so far. I also chose to exemplify the models circulating
in Thracian for the last time showing a pattern of language development
that to my mind is informative and that worked slowly within and through
the adoption and use of the Greek alphabet.
This leads us toward a group of interrelated conclusions:
(1) The Thracian language was ‘made’ regular from the 6 th
BC on.
(2) The difference between the genuine Thracian texts and
the material derived from onomastics make us think of a major
break-up that occurred around the end of the 6th – 5th century BC.
Judging on the script of the Kyolmen inscription and that of the
Ezerovo ring one cannot help drawing the conclusion that they
were executed using two distinctive scripts (alphabets). Whilst
the first could be seen as identical in many ways with the one
used in the Old Phrygian, the latter is Greek.
(3) The cultural and historical character of the study helped
reach a conclusion that the stele of Kyolmen is a tombstone of a
fallen soldier (paradynast?) rather than a dedicatory inscription as
Analogy and Conclusion 145
regards the formulaic interpretation of the inscription in

accordance with other grave-related texts coming from Greece
and Phrygia. The second and the third part of the inscription
speak in favor of that idea.
It should be mentioned also that the proposed analysis in reading,
interpreting, and understanding the oldest evidence have yielded not only
theoretical conclusions but also practical translations. We know for sure
that the genitive case ending is –e; we have a notion of some pronominal
stems, e.g. il- and as-; some verbal forms with the negative partice ne have
been identified along with the PIE* -t singular ending.
Arguments were supplied to support the opinion that bilingual texts do
exist. In the case of Gonimaseze, I was able to produce theoretical, as well
as factual material about the structure of the sentence in Thracian within
the theory of the Indo-European phrase.
Chapters Two and Three in Part One led to the following
conclusions:
New stems and names were added to the general list. Our knowledge
was consolidated as to the chronological layers of the sequences. A good
illustration would be the name of the Thracian king Ebrytelmis on a coin
dated in 4th BC vs. the common Ebryzelmis from the Roman period (1st -
3rd AD). Thus the phoneme /t/ changed to /z/ through /s/. Just less than two
centuries before that, according to a terminus ante quem in the Kyolmen
inscription, the transition of /s/ to /z/ occurred. The immediate inference is
that after that time and around 300 BC the Thracian underwent a major
change. From a historical and cultural perspective, this is due to the
Hellenization of the Thracian hinterland. Hellenism through Macedonian
rule won the political battle.
With the inscriptions on phialae, many royal names were established in
their secure form. Thus, we know that the right form is Kersebleptes for
that Thracian king that ruled around 360 BC. New items such as e.g.
Disloias, Sauthabas, Apro(s), Argiske, Ergiske, Satokos, Sitalces’ son,
Beo(s), Geista(i), both the names of Teres and his son Amatokos,
Didykaimo(s); a new name of an aristocrat from 6th - 5th century BC,
Zeila(s).
The Greek inscriptions on stone furnished around 450 new names out
of the whole body of around 900 names contained in The Evidence; these
are personal- and place-names, epithets of local deities, etc. This is our
main source of information on any of the developments registered after 4 th
BC. This is the period of further consolidation of using Greek lettering.
I found here more arguments in restoring the “unedited” true Thracian
form of the name of king Seuthes. In , IGBulg. III
146 Chapter Eight
1690 b 59, 202 AD, which Detschew 437 thinks should be corrected to
and Mihailov is in favor of what he saw was written on the
stone, namely , we have the intermediate variant of Sese in the
first component of this compound (from the bilingual inscription
Gonimaseze) before it got changed under Greek influence to Seuthes.
Therefore, the development was from Sete to Sethe to Sese to Seze. This
documents the history of the development of the Thracian language which
was made possible through the evidence of inscriptions only. An
interesting fact about comparing Greek with Latin evidence are the forms
Sethe and Sese, almost synchronic, the first being dated in 202 AD and the
latter around 300 AD.
In Part Two several conclusions were drawn.
In Chapter Five, the immediate conclusion is that once the
phonological system is established through all extant examples one could
undertake a next step to use the compound components to start reading and
understanding the Thracian language.
There is a substantial amount of skepticism stemming from
unsystematic efforts to crack on understanding Thracian. However, now
within the limits of its elements being determined, the whole body of the
language unfolds and becomes visible.
The system of the vowels reveals the following.
PIE *a (with all pros and cons) yielded an /a/ alternating with /e/ under
certain conditions: (1) in word-initial before d/, /r/ as well as before /p/ and
/s/ (2) in word-final of first compound components after /s/, /t/, coinciding
with an /e/ that appears to be an ending in e.g. DADALEME, EBAROZE,
ROLISTENE, etc., probably a Genitive sg. one. Whether a/e was a remote
reflex of a laryngeal is still a surmise.
Beside the fluctuation a/e, there is a stable /a/ throughout the entire
period in the evidence.
PIE */e/ is represented through /e/, but also through /e:/ due to the fact
that we do not have any knowledge the vowel quatity.
It goes the same way about PIE */o/.
PIE */i/ is perhaps a phoneme that does not appear in word-initial
position except for Istros and Istrianos. The phoneme /i/ also represents
PIE */oi/. PIE *loi-mo in from lim(o)-enos.
PIE */u/ is represented by /u/. So is PIE */eu/ in e.g. *meudh-,
Thracian Mutorgenos.
The PIE */s/ is also represented by a spirant through Thracian /s/, and
its allophone /z/.
The Consonants:
PIE *p,*t,*k are p,t,k in Thracian.
Analogy and Conclusion 147
PIE *kw in IE *kwet(w)r- in Thracian

w
since PIE *k gives k.
PIE *b,*d,*g are b, d, g in Thracian.
PIE *gw is g in gwn-i “woman, wife” and in < PIE
w
*g elh- “gland”, Bulgarian “желъд”.
PIE *gh > g in Thracian’
PIE *gwh there is a good example in Germanos, Greek thermos.
The resonants, l, m, n, r in Thracian continued the PIE sonants but are
often times found in the following combination:
r is represented by ur in e.g. Aidezuris. Or a/er in Argiske, Argedauon,
Ergiske, , Ergisenoi. Or or-/our- in: Borkeithias, Bourgeilos,
Bour-dapa.
For l there is -ol- in: Bolbabria. Or -lō- in: Blosti; -yl- in: Brylouzies; -
ul- in: Garoulos;
-il- in: Gilpyris.
For the sonant m beside m there is am-/me in Amatokos vs. Metokos.
The glides j and w are represented through i(iota) and B (beta) or Ou-.
For more examples, see Chapter Five in Phonology.
In Chapter Six, the compound components exemplify the phonological
sequences and give us a good idea about the shape (the morphology) of
many Thracian words.
In Chapter Seven, Thracian numerals are being identified among the
components, and there are adequate examples for “two”, “three”, “four”,
“five”, and “ten”.
And finally, I am convinced that whatever Thracian and Phrygian

parallels exist, this is an important lead to follow. I would also like to
underline the need of further elaborations on the approach in the future. It
should be detailed and if one decides to incorporate data from different
regions, different ages, and different languages, one should keep in mind
that diachronic and synchronic go hand in hand and therefore, the
combination of information sources should rest on a preparation of
precisely defined methods in order to make a determination of to the
outcome.
BIBLIOGRAPHY
Abbreviations
ANRW Aufstieg und Niedergang der römischen Welt. (Berlin/
New York)
Bull. Ep. Bulletin Epigraphique
CIA Corpus Inscriptionum Atticarum (Berlin)
CIL Corpus Inscriptionum Latinarum
Dečev Detschew, D. 19762. Die thrakischen Sprachreste
(Wien).
EB Etudes Balkaniques (Sofia).
FGrH Die Fragmente der griechischen Historiker
GPN Hornblower, S and Matthews, E. (eds. ) 2000: Greek
Personal Names. Their Value as Evidence (Proceedings
of the British Academy 104) (Oxford).
GSUFF Godishnik na Sofiyskiya Universitet. Filologo-filosofski
fakultet.
HS Historische Sprachforschung (= KZ)
IAI Izvestiya na Arkheologicheskiya institut
IBAI Izvestiya na Bulgarskiya Arheologicheski institut
(Sofia).
IBAD Izvestiya na Bulgarskoto arheologichesko druzhestvo
(Sofia).
IG Inscriptiones Graecae
IGBulg Mihailov, G. 19702-1997: Inscriptiones Graecae in
Bulgaria Repertae. Vol. 1-5 (Sofia).
IGDOP Dubois, L. 1996 : Inscriptions grecques dialectales
d’Olbia du Pont (Genève).
INMV Izvestiya na Narodniya muzei Varna
IPE Latyshev, V. 1885-1901: Inscriptiones antiquae orae
septentrionalis Ponti Euxini Graecae et Latinae. 1-4
(Petropoli).
JIES Journal of Indo-European Studies (Boston)
KBN Struvve, V. V. 1965: Korpus bosporskikh nadpisey
(Moscow/Leningrad).
150 Bibliography
KZ Zeitschrift für vergleichende Sprachforschung auf dem

Gebiete der indogermanischen Sprachen. Begründet von
A. Kuhn (Göttingen).
LB Linguistique balcanique (Sofia).
LGPN Fraser, P. M. and Matthews, E. (eds. ) 1987- 2000: A
Lexicon of Greek Personal Names. (Oxford).
PZ Praehistorische Zeitschrift
TS Fol, A. and Venedikov, I. (eds. ) 1980: Trakiyskite
Svetilishta. Sanctuaria Thracum Monumenta Thraciae
antiquae. 2 (Sofia).
SEG Supplemmentum Epigraphicum Graecum
Zgusta, KPN Zgusta, L. 1964: Kleinasiatische Personennamen. (Prague).
Zgusta, KON Zgusta, L. 1984: Kleinasiatische Ortsnamen. (Heidelberg).
ZPE Zeitschrift für Papyrologie und Epigraphik
Adams, W. L. 2007: ‘Symmiktous Katoikias’ and the city foundations of

the Thracian frontier’. Proceedings of the 10 th International Congress
of Thracology (Komotini-Alexandroupolis, 18-23 October 2005),
(Athens), 3-12.
Alexandrov, S. 2005: ‘The earliest Mycenaean pottery imports in
Bulgaria’. In Bouzek, J. and Domaradzka, L. (eds. ), The Culture of
Thracians and their Neighbours. Proceedings of the International
Symposium in Memory of Prof. Mieczyslaw Domaradzki, with a
Round Table ‘Archaeological Map of Bulgaria’. BAR International
Series 1350. (Oxford), 47-49.
Ancillotti, A. 1986: ‘Una nuova lettura dell’iscrzione trace di Kyolmen”.
Archivio Glottologico Italiano 71, 1-14.
Angelov, D. G. : 2003 ‘New Greek Inscriptions from the Strymon Valley
(Province of Macedonia)’. ZPE 143, 138-42.
Anttila, R. 1989: Historical and Comparative Linguistics. (Amsterdam
studies in the theory and history of linguidtic science. Series IV,
Current issues in linguistic theory, v. 6) (Amsterdam).
Archibald, Z. H. 1998: The Odrysian Kingdom of Thrace. Orpheus
Unmasked (Oxford).
—. 1999: ‘Thracian Cult – from practice to belief’. in Tsetshladze, G. R.
(ed. ) Ancient Greeks west and east (Leiden/Boston/ Koeln), 427-68.
Bader, F. 1986: ‘Structure de l’énoncé indo-européen’. Bulletin de la
société de linguistique de Paris 81, 71-120.
Badian, E. 1983: ‘Philip II and Thrace’. Pulpudeva 4, 51-71.
Bayun, L. S. and Orel, V. E. 1991: ‘The Inscription from Sitovo’. Orpheus
1, 144-48.
Thracian Language and Greek and Thracian Epigraphy 151
Beekes, R. S. P. 1995: Comparative Indo-European Linguistics. An

Introduction (Amsterdam/Philadelphia).
Berciu, D. 1974: Contribution à l’étude de l’art thraco-gète (Bucuresţi).
Beševliev, V. 1965a: ‘Inschrift in unbekannter Sprache aus
Nordbulgarien’. Glotta, 43, 3-4, 317-22.
—. 1965b: Prouchvania vurhu lichnite imena u trakite (Sofia).
—. 1970: Untersuchungen über die Personennamen bei den Thrakern
(Amsterdam).
Boardman, J. 2001: Greek Gems and Fingerings. Early Bronze Age to
Late Classical (London).
—. 1999: The Greeks overseas. The Early Colonies and Trade (Fourth
Edition. London).
Bodel, J. (ed. ) 2001: Epigraphic Evidence (London).
Boïadjiev, D. 2000: Les relations ethno-linguistiques en Thrace et en
Mésie pendant l’époque romaine (Sofia).
Bonfante, G. 1937 : ‘Il carattere satəm del tracio e la trascrizione greca
della z (s sonora)’. KZ 64, 127-29.
Boyadzhiev, D. 2002a: ‘Ziles Pyroula (a new reading of IGBulg.
5913=2289)’. In Boshnakov, K. and Boteva, D. (ed. ) Jibilaeus V.
Sbornik v chest na Prof. Margarita Tacheva (Sofia), 79-81.
—. 2002b: ‘Was there a Thracian common noun Zia?’. In Gicheva, R. and
Rabadhziev, K. (eds. ) Ðéôýç. Studia in honorem Prof. Ivani Marazov
(Sofia), 282-88.
Böttger, B. and Halloff, K. 1991: ’Inschriften aus Karasura (II)’. Klio 73.
2, 474-88.
Brixhe, C. and A. Panayotou, 1994: ‘Le thrace’. In Bader, F. (ed. ) Les
langues indo-européennes (Paris), 179-203.
Brugmann, K. 1904: Kurze Vergleichende Grammatik der
Indogermanischen Sprachen (Strassburg).
Chaniotis, A. 1991: ‘Epigraphic Bulletin for Greek Religion’. Kernos 4,
287-311.
Čičikova, M. 1977: ‘Céramique thrace fabriquée à la main du VIe au Ier
siècles avant notre ère’. Thracia 4, 123-40.
Cook, B. F. 1987: Greek Inscriptions (Berkeley/ Los Angeles, CA).
Corsten, Th. 1990: ’Einige neue thrakische Namen’. Betraege zur
Namenforschung 25. 3/4, 261-66.
—. 1991: Die Inschriften von Prusa ad Olympum (Bonn).
—. 2007: “Thracian Personal Names and Military Settlements in
Hellenistic Bithynia”. In Old and New Worlds in Greek Onomastics.
Second Colloquium of the Lexicon of Greek Personal Names (Oxford),
121-133.
152 Bibliography
Dechev, D. 1934: ‘Antichni pametnitsi ot Bulgaria’. IBAI 8, 76-77.

Detschew, D. 19762. Die thrakischen Sprachreste (Wien).
Dečev, D. 1960: Charakteristik der thrakischen Sprache. (=LB 2), 145-
213.
Dimitrov, K 1999: ’A Bronze Coin of the Odrysean king Seuthes II’.
Godishnik na arkheologicheskia muzey Plovdiv 9, 175-80.
Dimitrov, K. and Penchev, V. 1984: Sevtopolis. 2. Antichni i
srednovekhovni moneti (Sofia).
Dimitrov, P. 1994: Paleobalkanskiyat Vokalizum (In Bulgarian =The
Palaeo-Balkan Vocalism) (Sofia).
—. 1995: “Deux noms nouveaux d’aristocrates thraces dans une
inscription de Kazanlak”. Orpheus 5, 23-25.
—. 1998: ‘Thrace and Phrygia. Linguistics and Epigraphics’. In Tuna, N.,
Akture, Z. and Lynch, M. (eds. ) Thracians and Phrygians: Problems of
Parallelism. Proceedings of an International Symposium on the
Archaeology, History and Ancient Languages of Thrace and Phrygia.
Ankara 3-4 June, 1995 (Ankara), 111-14.
—. 2003: ‘The 6th Century BC Inscription from Kyolmen, District of
Preslav, North-Eastern Bulgaria (Archaeological Museum of Sofia,
Inv. No. 6558)’. Thracia 15. In Honorem Annorum LXX Alexandri
Fol, 345-54.
—. 2005: ‘The Thracians and Their Neighbors’. Thracia 16, 59-65.
—. 2006: “Three votive Tablets from Kyustendil and Its Vicinity”. In
Proseedings of the Museum of History of Kyustendil, v.13
(Kyustendil), 247-251.
—. 2006a: ‘Lemnian Women’. In Festschrift for M. Chichikova (Sofia).
(forthcoming)
—. 2007a: ‘Problèmes d’épigraphie’. Thracia 17. Studia in honorem Cirili
Jordanov. (Sofia), 319-323.
—. 2007b: ‘The Thracian Language: Problems of Chronology’.
Proceedings of the 10th International Congress of Thracology
(Komotini-Alexandroupolis, 18-23 October 2005), (Athens), 130-134.
—. 2008 (forthcoming): ‘The village of the Horse Riders’. ZPE.
—. 2008a: ‘A 4th Century BC Thracian Gold Signet-Ring from the
Dalakova Tumulus (SE Bulgaria)’. Archaeologia bulgarica (Sofia) 12,
2, 26-32.
Dimitrova, N. 2002: ‘ Inscriptions and Iconography in the Monuments of
the Thracian Rider’. Hesperia 71, 2, 209-229.
Dobrusky, V. 1897 : ‘Inscriptions et monuments figurés de la Thrace’.
BCH 21, 119-40.
Domaradzka, L. Addendum to Domaradzki, M. 1993: ‘Pistiros: Centre

commercial et politique dans la vallée de Maritza’, Archeologia
(Warsaw) 44.
Domaradzki, M. 1993: ‘Pistiros: Centre commercial et politique dans la
vallée de Maritza’. Archeologia (Warsaw) 44, 35-57.
Dornseiff, F. 19592: Kleine Schriften I (Leipzig).
Dremsizova-Nelchinova, C. 1972: ‘Trakiyski mogilni pogrebenia kray s.
Kyolmen, Shumensko’. IAI 32, 207-29.
—. 1999: ‘Nahodki ot krepostta kray Mezdra’. In Gočeva, Z. (ed. ) Studia
in memoriam magistri prof. Georgi Mihailov. (Sofia).
Dubois, L. 2000: ‘Hippolytos and Lysippos’. In GPN, 41- 52.
Duridanov, I. 1985: Die Sprache der Thraker (Neuried).
—. 1987а: ‘Zur kleinasiatischen Toponymie’. LB 30. 1, 39-50.
—. 1987b: ‘Probleme der Thrakischen Sprache I’. LB 30. 2, 73-84.
—. 1987c: ‘Probleme der Thrakischen Sprache II’. LB 30. 3, 137-46.
Duridanov, I. 1989: ‘Probleme der Thrakischen Sprache III’. LB 32. 2, 87-
112.
Elvers, K. -L. 1994: ‘Der “Eid der Berenike und ihrer Söhne”: eine Edition
von IGBulg. III 2, 1731’, Chiron 24, 241-66.
Filow, B. 1934: Grabhügel aus Duvanlij in Südbulgarien. (unter
Mitwirkung von I. Welkow und V. Mikov) (Sofia).
Fol, A. 2000: ‘Said and unsaid’. In Fol, A. (ed. ) Ancient Thrace (Sofia),
63-73.
—. 2002: Trakiyskiyat Dionis. Kniga treta: Nazovavane i vyara (Sofia).
Fol, A. and Schmitt, R. 2000: ‘A Linear A Text on a Clay Reel from
Drama, South-East Bulgaria?’. PZ 75, 56-62.
Gočeva, Z. 1989 : ‘Neue thrakische Namen und Epithete’. LB 32, 113-15.
Gaertner, J. F. 2001: Boïadjiev, D. : Les relations ethno-linguistiques en
Thrace et en Mésie pendant l’époque romaine. Sofia. Review, Bryn
Mawr Classical Review 2001. 10. 14.
Ganeva, S. 2005: ‘An attempt at characterisation of the slabs from the
town of Razlog in the cultural-religious context of the Late Bronze Age
in Bulgarian Lands’ In Bouzek, J. and Domaradzka, L. (eds. ), The
Culture of Thracians and their Neighbours. Proceedings of the
International Symposium in Memory of Prof. Mieczyslaw
Domaradzki, with a Round Table “Archaeological Map of Bulgaria”.
BAR International Series 1350. (Oxford), 147-51.
Gergova, D. 1989: ‘Thracian Burial Rites of Late Bronze and Early Iron
Age’. In Best, J. and De V ries, N. (eds. ) Thracians and Mycenaeans.
Proceedings of the Fourth International Congress of Thracology.
Rotterdam, 24-26 September 1984 (Leiden / Sofia), 231-40.
154 Bibliography
Georgiev, V. 1966: ‘Die Deutung der altertümlichen thrakischen Inschrift

aus Kyolmen’. LB 11. 1, 9-23.
—. 1983: ’Thrakisch und Dakisch’. ANRW II. 29. 2, 1148-94.
Georgieva, R. 2001: ‘Thracian Culture of the Early Iron Age’. In Maritsa-
Iztok. Archaeological Research. 5 (Radnevo), 83-94.
Gerassimova, V. 2003: ‘Preliminary Study of the graffito from
Alexandrovo Tomb’. Izvestia na istoricheskiya muzey Khaskovo 2,
177-79.
—. 1980: ‘Nadpisi ot svetilishteto na Asklepii Keylayden’. In TS, 48-94.
Gerasimova 2004,
Gerov, B. 1961: ‘Prouchvaniya vurkhu zapadnotrakiyskite zemi prez
rimsko vreme. I’., GSUFF 54. 3, 155-407.
Gotsev, A. 1994: ‘Contacts and interactions across the Eastern Balkan
Range during the Early Iron Age’. In Gergova, D. (ed. ) Helis 3. 1.
Culture and Religion in Northeastern Thrace (Sofia), 43-68.
Gicheva, R. and Rabadhziev, K. (eds.) Pitu/h. Studia in honorem Prof.
Ivani Marazov (Sofia).
Gočeva, Z. 1989: ‘Neue thrakische Namen und Epithete’. LB 32, 113-117.
Guerassimova-Tomova, V. 1989: ‘L’inscription de Sitovo’. LB 32. 2, 131-
40.
Galabov, I. 1964: ‘Trakiyskite imena na –hnoj i –anoj i tekhnite problemi’.
Izvestiya na Instituta za bulgarski ezik 10, 1-64.
Haas, O. 1966: Die phrygischen Sprachdenkmäler. (Sofia) (=LB 10).
Haralampieva, A. and Ivanov, V. 1987: ’Novootkriti pametnitsi na
trakiyskia bog-konnik ot Varnenski okrug’. INMV 23/38, 74-84.
Head, B. V. 19113: Historia Numorum. A Manual of Greek Coins
(Oxford).
Heydemann, 1874: Die antiken Marmorwerke in Athen. (Berlin).
Hodot, R. 1990 : Le dialecte eolien d’Asie. La langue des inscriptions VIIe
s. a. C. IVe s. p. C. (Paris).
Hooker, J. T. 1977: The Language and the Text of the Lesbian Poets
(Innsbruck).
Ivanov, D. I. 1982: ’Trakiyskoto sreburno sukrovishte ot selo Borovo’.
Godisnik na muzeite ot Severna Bulgaria 8, 22-35.
Ivanov, M. 2004: ’Nadpisut na stratega Diouzihj Dhnosqe/nouj'.
Arkheologia 1-2, 83-86.
Janakieva, S. 1988: ‘Skokelenoi – ein thrakischer Göttername aus einer
wiederentdeckten Inschrift’. Thracia 8, 119-21.
Jouroukova, J. 1992: Hoards Treasures from the Bulgarian lands. 1
(Sofia).
Kabakchieva, G. (ed. ) 2000: Ares Collection. Catalogue. Private Antique

Museum ARES Foundation (Sofia).
Katsarov, G. 1937: ‘Arkheologicheski vesti. Antichni pametnitsi ot
Bulgaria’. IBAD 11, 283-87.
Kacarov = Katsarov, G. 1919-1920: ‘Antichni pametnitsi iz Bulgaria
(chetvurta statiya)’. IBAD 7, 1-14.
Katičić, R. 1976: Ancient Languages of the Balkans (= Trends in
Linguistics State-of-the-Art Reports 4-5) (The Hague-Paris).
Kitov, G. 2004: ‘Novi nablyudenia v Aleksandrovskata grobnitsa’.
Archeologia (Sofia) 1-2, 42-51.
—. 2005a: Dolinata na trakiiskite vladeteli. (Varna).
—. 2005b: ‘New Discoveries in the Thracian Tomb with Frescoes by
Alexandrovo’. Archaeologia Bulgarica 9. 1, 15-28.
—. 2005c: The Newly Discovered Tomb of the Thracian Ruler Seuthes III.
Archeologia Bulgarica 9. 2, 39-54.
Labov, W. 1994: Principles of Linguistic Change (Oxford / Cambridge).
Lehmann, W. P. 1952: Proto-Indo-European Phonology (Austin, TX
/London).
—. 1993: Theoretical Bases of Indo-European Linguistics (London / New
York).
Manov, M. 1994: ‘Eine neue Inschrift auf zwei silbernen Schöpfkellen aus
Thrakien’. Tyche 9, 89-92.
—. 2008: Razkopki i prouchvania. v.38. Bulgarian Academy of Sciences,
National Archaeological Institute and Museum. (Sofia)
Marazov, I. (ed. ) 1998: Ancient Gold: The Wealth of the Thracians.
Treasures from the Republic of Bulgaria. Catalog of the Exhibition in
USA. (New York).
Marazov, I., Penkova, E., Avramova, M., Konova, L. And Markov, N.
1996 : Ares Private Antique Collection (Sofia).
Masson, O. 1966: in Philologus 110, 253.
—. 1988: ‘Les noms théophores de Bendis en Grèce et en Thrace’.
Museum Helveticum 45. 1, 6-12.
Matsas, D. 1994: ‘Minoan Long-Distance Trade: A View from the
Northern Aegean’. In Laffineur, R. and w. -D. Niemeier (eds. )
Aegeum 12. 1. Politeia. Society and State in the Aegean Bronze Age.
Proceedings of the 5th International Aegean Conference, University of
Heidelberg, Archäologische Institut, 10-13 April 1994. (Liège and
Austin, TX), 235-47.
Mayser, E. 1935-1970: Grammatik der griechischen Papyri aus der
Ptolemaeerzeit I. 3 (Berlin).
Meisterhans, K. 1900: Grammatik der attischen Inschriften (Berlin).
156 Bibliography
Melchert, H. C. 1987: ‘The Reflexes of h3 in Anatolian’. Die Sprache.

Mihailov, G. 1943 : La langue des inscriptions grecques en Bulgarie
(Sofia).
—. 1980a: ‘La ville de Satra à Crète et la tribu thrace des Satres’. Klio 62,
13-18.
—. 1980b: ‘Inscriptions de la Thrace égéenne’. Philologia 6, 3-19.
—. 1987: ‘Les inscriptions dans le trésor de Rogozen’. LB 30, 5-19.
—. 19912 : ‘Thrace before Persian Entry into Europe’. In Cambridge
Ancient History 3. 2, 519-618.
—. 2007: Scripta minora (Sofia).
Morpurgo-Davis, A. 2000: ‘Personal Names and Linguistic Continuity’. In
GPN, 15-40.
Nijboer, A. J. 2005: ‘The Iron Age in the Mediterranean: A Chronological
Mess or “Trade before the Flag”, Part II’. Ancient West and East 4. 2,
255-77.
Orel, V. 1997: The Language of Phrygians. Description and Analysis
(Delmar/ New York).
Oetinger, N. 2001: ‘Varia Hethitica’. HS 114, 80-89.
Owen, S. 2000: ‘New Light of Thracian Thasos: A Reinterpretation of the
‘Cave of Pan’. JHS 120, 139-43.
—. 2005: ‘Analogy, Archaeology and Archaic Greek Colonization’. In
Hurst, H. and Owen, S. (eds. ) Ancient Colonizations. Analogy,
Similarity and Difference (London), 5-22.
Petkov, H. 1960: ‘Novootkrito sakrovishte ot s. Golyama Brestnitsa,
Plevensko’. Arkheologia 1, 25-28.
Philipp, M. 1970 : Phonologie de l’allemand (Paris).
Pokorny, J. 19943 : Indogermanisches etymologisches Wörterbuch
(Tübingen / Basel).
Polomé, E. C. 1986: ‘A Note on Thraco-Phrygian Numerals’ . JIES 14,
185-9.
Rasmussen, R. J. 1999: ‘Initial h3 in Anatolian: A Vote for Chaos’. In
Selected Papers on Indo-European Linguistics. Part 2 (Copenhagen).
Robert, L. 1959: ‘Les inscriptions grecques de Bulgarie’, In Revue de
Philologie 33.
Robert, J. and L. 1939: Bull. Ep. 52, 480.
Rousseva, R. (ed.) 2006: National Museum of History. Catalog. (Sofia).
Russu, I. I. 1958: Dacia NS 2, 538 .
—. 1962: Studii si cercetari de istorie veche 13, 233.
—. 1964: ‘K. Vlahov, ’Nachträge und Berichtungen zu den thrakischen
Sprachresten und Rückwörterbuch’. Annuaire de l’Université de Sofia,
Faculté de Philologie 57, 1963, 221-372. (Review)’. Cercetâri de

linguistica 9, 326-32.
—. 1967: Limba traco-dacilor (Bucaresţi).
—. 1969 : Sprache der Thrako-Daker (Bukarest).
Şahin, S. 1982: Katalog der antiken Inschriften des Museums von Iznik
(Nikaia) II. 2 (=IK 10. 2) (Bonn).
Sarafov, T. 1974: ‘L’étymologie du nom de la tribu thrace BESSI, Satrai’.
Thracia 2, 135-38.
—. 1972: ‘Une nouvelle étymologie du nom de la tribu thrace Satrai’. EB
2, 115-20.
Schmid, W. P. 1987: ’Zur thrakischen Grabinschrift aus Kölmen’, KZ 100,
351-57.
Schmitt-Brandt, R. 1967: ‘Die thrakischen Inschriften’, Glotta 45, 40-60.
Schoenert-Geiss, E. 1987: Die Muenzpraegung von Maroneia (Berlin).
—. 1985: ’Maroneia und die Thraker’. In Eirene 22, 39-53.
Schwyzer, E. 1934-1970: Griechische Grammatik (München).
Sharankov, N. 2004: ‘Inscriptions from the Strymon Valley: Corrigenda’.
ZPE 148, 198-200.
Sharankov, N. 2005: ‘A Greek Graffito in the Thracian Tomb near
Alexandrovo’. Archaeologia Bulgarica 9. 1, 29-35.
Sherratt, A. and Taylor, T. 1989: ‘Metal Vessels in Bronze Age Europe
and the Context of Vulchetrun’. In Best, J. and De Vries, N. (eds. )
Thracians and Mycenaeans. Proceedings of the Fourth International
Congress of Thracology. Rotterdam, 24-26 September 1984 (Leiden /
Sofia), 106-34.
Slavova, M. 1995: ‘Die thrakischen Frauennamen aus Bulgarien.
Nachtraege und Berichtungen’. Orpheus 5, 23-25.
—. 2005: Phonology of the Greek Inscriptions in Bulgaria (=Palingenesia
83) (Stuttgart).
Stoyanov, T. 1997: Early Iron Age Tumular Necropolis. „Sboryanovo” I.
(Sofia).
Sturtevant, E. H. 19684: Linguistic Change (Chicago / London).
Szemereniy, O. 1996: Introduction to Indo-European Linguistics. Fourth
revised edition (Oxford).
Die Thraker. 2004: Die Thraker. Das Goldene Reich des Orpheus. Kunst-
und Ausstellungshalle der Bundesrepublik Deutschland (Bonn).
Theodossiev, N. 1991 : ‘5th-4th BC Thracian Inscriptions : A Tentative
Problematization’. Numizmatika 1-2, 58-64.
—. 1992 : ‘5th-4th BC Thracian Inscriptions : A Tentative
Problematization’. Numizmatika 3-4, 34-47.
158 Bibliography
—. 1997: ‘On the Reading and Date of the Kyolmen Inscription’.

Indogermanischen Forschungen 102, 216-29.
Tod, M. N. 1985: Greek Historical Inscriptions (Chicago).
Tomaschek, W. 19802. Die alten Thraker I (Wien). W. Tomaschek, Die
alten Thraker, Sitzungsberichte d. Akad. Wien, Bd. 128 (1893): I.
Uebersicht der Stämme; Bd. 130 (1893): II. Sprachreste, 1. Glossen
allerart und Götternamen, Bd. 131 (1894): 2. Personnenamen und
Ortsnamen.
Totev, T. 1965: ‘Kamenna trakiyska grobnitsa pri s. Kyolmen’.
Arkheologiya 7. 4, 9-11.
Tracy, S. V. 1990: Attic Letter-Cutters of 229 to 86 B. C. (Berkeley, Los
Angeles, Oxford)
Vassileva, M. 1992-1993: ‘The Inscribed Thracian and Phrygian Phialae
in Comparative Context’. TALANTA 24-25, 161-66.
Venedikov, I. 1987: The Vulchitrun Treasure (Sofia).
Tsetskhladze, G. R. and De Angelis, F. (eds. ) 1994: The Archaeology of
Greek Colonisation. Essays Dedicated to Sir John Boardman (Oxford).
—. (ed. ) 1999: Ancient Greeks west and east (Leiden/Boston/ Koeln).
Tsetskhladze, G. 2002: ‘Ionians Abroad’. in Tsetskhladze, G. R. and
Snodgrass, A. M. (eds. ) Greek Settlements in the Eastern
Mediterranean and the Black Sea (BAR International Series
1062)(Oxford), 81-96.
Tsetshladze, G. R. 2006: ‘Introduction. Revisiting Ancient Greek
Colonisation’. in Tetskhladze, G. R. (ed. ) 2006: Greek Colonization.
An Account of Greek Colonies and other settlements overseas.
(Leiden/Boston), XXIII-LXXXIII.
Tsontchev, D. 1941: Le sanctuaire thrace près du village de Batkoun.
Sofia.
Velkov, V. and Fol, A. 1978: ‘Les thraces en Egypt Gréco-Romaine. La
literature et la méthode’. Pulpudeva 2, 46-52.
Velkov, V. (ed. ) 1979: Istoriya na Bulgariya 1. (Sofia).
—. (ed. ) 1991: Kabyle 2. (Sofia).
Velkov, V. and Domaradzka, L. 1994 : ‘Kotys I (383/2-359) et l’emporion
de Pistiros en Thrace’. BCH 118, 1-15.
Velkov, I. 1934: ‘Novootkriti starini’. IAI 8, 447-67.
Vlahov, K. 1963: ’Nachträge und Berichtungen zu den thrakischen
Sprachresten und Rückwörterbuch’. GSUFF 57, 221-372.
Vottero, G. 1985 : ‘Procédés d’expressivité dans l’onomasstique
personnelle de Béotie’ In La Béotie antique. Actes du 4ème colloque
international sur la Béotie antique, Lyon-Saint-Etienne 16-20 mai
1983. (Paris), 403-17.
Watkins, C. 2000. The American Heritage Dictionary of Indo-European

Roots (Boston/ New York).
Woodhead, A. G. 1959: The Study of Greek Inscriptions (Cambridge).
Woudhuizen, F. C. 2000-2001: ‘The Earliest Inscription from Thrace’.
TALANTA 32-33, 289-305.
Yurukova, Y. 1992: Monetni sukrovishta ot bulgarskite zemi. I. Monetite
na trakiyskite plemena i vladeteli (Sofia).
Zournatsi, A. 2000: ‘Inscribed Silver Vessels of the Odrysian Kings: Gifts,
Tributes, and the Diffusion of the Forms of “Achaemenid” Metalware
in Thrace’. AJA 104, 683-706.

Thracian Language and Greek and Thracian Epigraphy 1443813257 9781443813259 Compress

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thracian Language and Greek and Thracian Epigraphy 1443813257 9781443813259 Compress

Uploaded by

Copyright:

Available Formats

Thracian Language and Greek

and Thracian Epigraphy

This book first published 2009

Cambridge Scholars Publishing

12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK

British Library Cataloguing in Publication Data

Copyright © 2009 by Peter A. Dimitrov

ISBN (10): 1-4438-1325-7, ISBN (13): 978-1-4438-1325-9

List of Illustrations ..................................................................................... ix

Preface and Acknowledgments................................................................... xi

Introduction .............................................................................................. xiii

Part I: The Evidence

Chapter Two .............................................................................................. 21

Chapter Three ............................................................................................ 25

Part II: Phonology and Morphology

Chapter Five ........................................................................................... 121

Chapter Six ............................................................................................. 129

Chapter Seven ......................................................................................... 139

Chapter Eight .......................................................................................... 143

Fig. 1-1. The Kyolmen Inscription

Fig. 1-2. The Ezerovo Ring

Fig. 1-3. The Gold Ring from Duvanlij

Fig. 3-4. The Phiale from Kazanluk

Fig. 3-5. Gold Ring of Skythodokos

Fig. 3-6. Silver Phialae from Shipka

Fig. 3-7. The Silver Casserole

Fig. 4-1. Inscription from Kyustendil (Teres Garytinos)

Fig. 4-2. The Votive Plate from Kyustendil

Fig. 4-3. Inscription with Kozimases from Alexandrovo

Fig. 4-4. The Golden Plate from Sinemorets

Fig. 4-5. Graffito from Sboryanovo

I owe the reader a short explanatory note.

In epigraphic texts the Leiden system of conventions is used.

The present study of the Thracian language is based on an updated

“vulgate alexandrine”. It is imperative to underline what Hooker (1977,

la céramique, frappées sur les monnaies. Il est inutile de souligner

compare Rolistene in one of the earliest original Thracian inscriptions on

The Thracians: A Brief Historical Overview

European” by criteria that handbooks prefer not to discuss at length. The

territories south of the Carpathians to the fringes of the Greek world.17 In

Scholars are advancing the idea of a continuous habitation in nearly all

Thracian lands. The tripartite vessel is considered a unique one with

Considering the Thracians in Anatolia, K. Sams discussed the “open

perceptions about that as a major event in the Mediterranean world which

which remnants of onomastics are interwoven in a remarkably complex

Phonology and Phonetics of the Thracian names

There are several investigations, for which I have substantiated the

THE THRACIAN INSCRIPTIONS

I. The Kyolmen Stone Inscription1

village of Kyolmen that consequently led to rescue excavations. The actual

Fig. 1-1. The Kyolmen Inscription

2. The above mentioned three sections of the inscription go from left to

Ex autopsias, as I saw it in the Museum of Sofia, starting up from the

The Reading of the Inscription

2. The execution of the inscription betrays many turning points and

3. Suffice it to say that some of the earliest Greek inscriptions yet

although many seem to be much concerned with the linguistic aspect.

epithet of the Heros, 1st century BC in an inscription from

the end of are different in shape, if at all those graphemic

(2) This idea is corroborated by the fact that -/ is

5) The text as regards the Thracian language leads toward a formulaic

II. The Ezerovo Ring

Fig.1-2. The Ezerovo Ring