The Case of Thracians and Greeks in the north-eastern

Aegean

By Peter A. Dimitrov,New Bulgarian University (Sofia)

Center for Thracology, Bulgarian Academy of Sciences

Example of Late Roman Imperial Period Votive Plate

Despite the spread of Latin as 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 1st-3rd century AD votive inscriptions are dedicated to the Thracian Heros (Horseman) whose local epithets provide a rich material for the present study.

Garutinos, Teres Garutinos, a dedicant to Asklepios, on a votive plate, not far from Pautalia, present-day town of Kuystendilcentury AD. Not in Deev.

An Appliqu with a Thracian Warrior (4th BC) whose hair cut was mentioned by Homer where the Abantesfrom Euboia are said to be akrokomoi.

Phonology and Phonetics of the Thracian names from 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 changes 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, being unconditionally precise records for the history of the Thracian territories.

The method itself is not automatically a guarantee for good results. 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 classes have been proposed and later perceived as theoretical entities that may change according to sound laws 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 development or a feature that is not marked by any specific conditions or the shift was conditioned according to its word-initial or intervocalic position . 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 as 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 with unreal sounds or abstractions. 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 a/e.So, the above-mentioned classes, may be distinguished for subclasses, called allophones 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 in this sound change we see a principle that has long been explained and reads that generally one specific trait per se is involved.

No doubt, sounds cannot be phonemes for they belong to a different category. Their functional analysis is also called phonetics.[11] Only on the surface could one register the phonetic units, as they belong to the physiology of the articulation. We now may come to the subsequent conclusion: First, there is no /z/ in Thracian 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/ [12]. 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 * h1ekw-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.[13]

Ex autopsias, as I saw it in the Museum of Sofia, starting up from the right side, the inscription reads as follows:(1) ILASNLETEDNLEDNENIDAKATROSO (2) EBA:ROZESASNHNETESAIGEKOA(3) NBLABAHGN.

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

(vessel of) Dyntas(Dyntos?), son of Zeila(s)

The case of the irregular after the name of (s), 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.

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 using Greek

letters 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 (s)

we can say that it is a compound made of

- and - (s). The form with s is not secure and this nominative could be

restored to .

The distinction between H (eta) and I (iota) make us think of at least one obvious

conclusion. That is, 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 stone slab from Gordion, G-02 (6thcentury BC)

The inscription in scriptio continua was found on a gold ring from a tumulus near Ezerovo, district of Plovdiv. The letters were incised in eight irregular lines and the last on the rim ofthe ring. The find is dated to the fifth century BC.

ROLISTENEASNERENEATIL

TEANHSKOARA ZEADOMEANTILEZUPTAMIHE

RAZHLTA

A 5th BC Coin of the Thracian Tribe of the Derroni. It reads DERRO(nikon), (from the Velikovo Treasure).

A 4th century BC tomb at Alexandrovo, Southeastren Bulgaria

There is another inscription from Gazoros, near present-day town of Serres, Northern Greece, 158-159 AD that reads .

In this compound, the first component // is an allomorph of / o / as in . Not in Deev.

The graffito from the Alexandrovo tomb

The inscription on the faade on the tomb at Smydovo, north-eastern Bulgaria, 4thcentury BC.

/

- possessive-pronominal stem from *ma- my, mine

(Dat.) < IE * gwen-i woman, wife

/t/ /z/: - < * Di-, the dative form of the IE word for light

Seout, Seth, Sez Sese, the original Thracian etymon

Translation:

To my wife (woman) Sese (set up this monument); wife of Seuthes.

Cf., IGBulg. II 844, near ancientMarcianopolis:() , . .Aulousoukos, daughter of Bithys, (and) wife of Teres

The same name in its original form appears in a 3rd AD Latin inscription as SESE.

: (belongs to) Seusa(s), son of Teres

In2004aninscriptionwasfoundinafieldwithintheterritoryofthevillageofBrestovitza neartheancientPhilippolis.Itreads:

.

.

etc.

Ezbenis and Asbenoi; Bet-espios, Out-aspios

- fluctuation e/a- Thracian etymon of e/asba

from PIE * h1ekw-o-s h=a/e- Sb- in Anlaut Zb- or Sou-

(graphemic variants)

On an inscription from Komotini, Northern Greece, 3rd 2nd BC.What interests us is the sign for zeta (I).

(BelongstoCotys ofBeos (n)

A bronze helmet discovered in the Golyama Kosmatkatomb in 2005 with an inscription:

(belongs to) Seuthes

Seuthes III (?) (c. 330-300BC)

We already discussed the existing alternation thi- / si- in Thracian, exemplified here through and .

The connecting syllable ou- is to be explained through the analogical assimilation of the Greek language editing, i.e. s-ou from ti/thi andord- /urd- leveled to t/th/s.

This alternation already exists in Thracian names and at least in the original Thracian variant is only a /t/ replaced by Greek /th/ (theta), which through contamination may have been altered to /s/.

The name starts with a Z-. In Indo-European there is originally no /z/, and thus this phoneme is an allophone of the original /s/. Therefore, in trying to be consistent here, we should edit Zbelsourdos to Sbelsourdos.

By comparing these two variants, one cannot miss the obvious voicing of the S- into Z-before b. This is a classic case of regressive assimilation due to the voiced element as in b representing the glide /w/.

The Case of Thracians and Greeks in the north-eastern AegeanExample of Late Roman Imperial Period Votive PlatePhonology and Phonetics of the Thracian names from the inscriptions found in Bulgaria : (belongs to) Seusa(s), son of Teres