Volume 38, No.2 - Summer 1992
Editor of this issue: Antanas Klimas, University of Rochester 
ISSN 0024-5089
Copyright © 1992 LITUANUS Foundation, Inc.


Harvey E. Mayer
Defense Language Institute

Why is it no one properly appreciates an important work by Bulgarian scholar Ivan Duridanov? This book, Die thrakisch-und dakisch-baltischen Sprachbeziehungen, published by the Bulgarian Academy of Sciences in 1969, was mentioned only in part of a sentence as "incorporating the Baltic claims" to ties with Albanian by Eric P. Hamp in his "Albanian" in Volume Nine, Part Two of Current Trends in Linguistics, published in 1972. It seems to have been otherwise entirely ignored by Baltists. Yet it provides a new perspective to the understanding of Pre-Balts and Pre-Baltic. And it confirms some of my more recent notions of Early Baltic and Early Balts.

Surely my brand new view which I derive from Duridanov's book that the Thracians and Dacians were descended from Indo-European tribes which spoke forms of Pre-Baltic of sorts enhances the old idea I have had and, I am sure, others have had that the ancestors of the historical Balts moved around a great deal and in so doing spread Baltic linguistic items over an immense area. Duridanov's book allows us to expand that concept to include in a way even movement southward to the Balkans if we can accept the Thracians and Dacians as descendants of "Balts," or really, "Pre-Balts" of a sort. And hence I designate Thracian and Dacian and their Indo-European ancestor dialects, Pre-Thracian and Pre-Dacian, as "Southern Baltoidic," "Southern" with respect to their ultimate position as eventually more southerly than Baltic proper and "Baltoidic" to indiciate them as a class of "Baltic-like," if not exactly, Baltic dialects and then languages.

Duridanov does not even hint that Thracian and Dacian might have been varieties of Pre-Baltic. He merely suggests that certain strikingly parallel vocabulary items and structurings, and some of these, apparently, exclusive between these languages and the Baltic ones allow us to conclude that Baltic, Dacian, and Thracian in their early history once bordered on one another.

My evidence for stating that Dacian and Thracian are "Southern Baltoidic" is phonological. From the lexical items mentioned by Duridanov I can show that not only is the sequence initial ks- missing, but even metathesized, as Dacian examples show, to sk- in two of the same morphemes, skaud-'pain' and skuja 'pine,' which we find in Baltic, that is, Lithuanian. The Lithuanian forms skaudus 'painful' and skuja 'pine' show a metathesis of initial ks- to sk- (that is, not *šk-) which preceded the ruki law and was, therefore, Pre-Baltic since the ruki law assimilation of s to a preceding k to, at first, s, surely began operating in Late Dialectal Indo-European. Like the metathesis of initial ks- to sk- which we find direct evidence for only in Baltic and Dacian, the ruki law was an early attempt to reduce the possibility of h, that is, aspiration, to arise. These measures against aspiration were inspired by glottalization, itself a direct measure against excessive aspiration in Indo-European. And languages showing metathesis of initial ks- to initial sk- arose from those Indo-European dialects which had the heaviest early glottalization. These include the Baltic languages, Dacian, and, I say, Thracian. Note that the special correspondences between Baltic, Dacian, and Thracian in not only lexicon, but also in lexical structuring mentioned by Duridanov now take on particularly great significance when seen against the background of the special phonological parallels between them which I mention. I believe these phonological parallels, and particularly the underlying excessive early glottalization were the features which encouraged the unique syllabic consciousness of early Balts, Thracians, and Dacians and their ancestors necessary to keep alive these special lexical and word-building parallels in their languages. Normally, related languages and Indo-European dialects of long separation by large stretches of territory do not allow us to show nearly as many strikingly clear parallels in derivatives and compounds, and some of these so ancient that their meanings are somewhat uncertain.

An early scholar who classified Thracian which he incorrectly lumped together with Phrygian by its lack of initial ks-was August Fick who wrote in 1873 in the section called "Die grossen Nationen der Phryger und Thraker" in his Die ehemalige Spracheinheit der Indogermanen Europas (Göttingen) the following: "Es fehlen die eranischen Kennlaute: ks im Anlaut... im Phrygischen völlig." Considering initial ks-in New Phrygian kseune we can see clearly now that Phrygian was a language separated from Thracian by that very feature. Phrygian permitted initial ks-. Thracian, like Dacian and Baltic, did not. I believe this difference seen in the context of the Thracian, Dacian, and Baltic lexical parallels indicates that Thracian and Dacian are intrusive Baltoidic elements in the Eastern Balkans which, incidentally, lie due south of Prussia, Lithuania, and Latvia, a fact indicating the direct southerly route taken, I believe, by the Baltoidic Pre-Thracians and Pre-Dacians. These were surely intrusive elements since they were bordered by Pre-Greeks to the south, Pre-Illyrian-Messapic Albanians to the west, Pre-Phrygians to the east, and Albanoidic Pre-Slavs to the north, all of whose dialects were less heavily glottalized and, therefore, permitted initial ks-.

Duridanov separates Thracian from Dacian on phonological grounds. Dacian, he says, reflected Indo-European p, t, k and b, d, g as such and deaspirated Indo-European bh, dh, gh to b, d, g. These Dacian reflexes match Baltic ones. But Thracian, he says, like Phrygian changed Indo-European p, t, k to ph, th, kh, Indo-European bh, dh, gh to b, d, g, and Indo-European b, d, g to p, t, k. Spellings seeming to reflect b, d, g for expected p, t, k in Thracian words he blames on Dacian phonological influence. Admitting that spellings seeming to reflect p, t, k are the rule, he says that this resulted from weakly aspirated ph, th, kh. The truth, I believe, is that the far fewer spellings of Thracian words seeming to reflect ph, th, kh whether in Greek or Latin symbols were purely graphic resulting from Phrygian and Greek orthographic influences where aspirated stops seem to have existed. Fluctuations in spellings with symbols p/b, t/d, k/g in Thracian words, I think, represent real phonetic differences between these words and their Dacian counterparts which, from what Duridanov says, show no such fluctuation.

These differences are significant. They date the arrival of the Pre-Thracians to the Balkans as, I believe, significantly earlier than that of the Pre-Dacians, a fact coinciding with their geographic positions and other linguistic data such as Dacian's examples of skaud- 'pain' and skuja- 'pine' showing the Baltoidic metathesis of initial ks- to sk-, vocabulary items not attested in Thracian. Thus, Pre-Thracian reached the Balkans precisely after p(h), t(h), k(h), b(h), d(h), g(h) had all lost their allophonic aspiration, a feature characteristic of Central Indo-European dialects such as Baltoidic Pre-Lithuanian, Pre-Latvian, and Pre-Prussian and Albanoidic Pre-Illyrian-Messapic-Albanian and Pre-Slavic, but before glottalic b", d" g" (or p", t", k") had deglottalized either to b, d, g, the usual Indo-European change, or to p, t, k, the less common change. I say that intense contact with Pre-Phrygian influenced the Pre-Thracian choice of voiceless p, t, k as reflexes of glottalic b", d", g". The glottalic phonemes were the only Pre-Thracian stops left unaltered and were, therefore, the only ones subject to change in the manner of Phrygian ones on the arrival of Pre-Thracian to the Balkans.

The Grimm's Law-like sequences of changes, glottalic b", d", g" (or p", t", k") to p, t, k; W), d(h), g(h), to b, d, g; and p(h), t(h), k(h) to ph, th, kh with aspiration made phonemic, were possible only where early excessive glottalization had not taken place, that is, where, with weaker glottalization, aspiration was stronger, and certainly strong enough to survive. It was not strong enough to survive in Baltoidic which included Pre-Thracian. Thus Thracian really did show only one change, p, t, k from glottalic b", d" g" (or p", t", k"). It did not have the changes ph, th, kh from Indo-European p(h), t(h), k(h) expected by Duridanov, an expectation conditioned by Grimm's Law Indo-European studies. Therefore, Thracian orthographic representations really do reflect p, f, k, and certainly not a "weakly aspirated" ph, th, kh.

Since the Pre-Dacians left Baltoidic territory later than the Pre-Thracians, Dacian shows a sound system closer to Baltic with b, d, g from glottalic b", d", g" and the following extant examples of roots with initial ks- metathesized to initial sk-; Dacian Scaugdae from *Skaudgae from *Skaudgedae with *skaud-matched by Lithuanian skaudus 'painful' versus Slavic xudu 'bad' from earlier *ksoud- with no metathesis of initial ks- and Dacian Skuanes from *Skujaines with *skuja matched by Lithuanian skuja 'pine' versus Russian xvoja 'evergreen' from earlier '*ksuoi- also with no metathesis of initial ks-. As a parallel, all that can be found for Thracian is a word with an initial s-, presumably the reflex of Indo-European palatal k'- which had probably been preceded earlier by an initial s-. This earlier initial sequence, sk'- from earlier sk- had been metathesized, I believe from an earlier ks-. The Thracian word reflecting all this is, I believe, the tribal name Satrai which Duridanov compares with Sanskrit ksatra- 'dominion', Avestan, Old Persian kšaOura- 'dominion, empire', and Lithuanian šatrus 'alive, stem' in his 1976 book Ezik"t na trakite (Izdatelstvo "Nauka i izkustvo," Sofia). This is reminiscent of the Lithuanian word for "six," šeši, where initial š- reflects Indo-European palatal k' occurring after metathesis of ks- to sk- (and later sk- to šk-) and loss of initial s-, that is, š-.

To reach the Eastern Balkans, Baltoidic Pre-Thracians and Pre-Dacians had to pass through Albanoidic territory in the Carpathians. I believe they captured some Albanoidics and brought them to the Eastern Balkans as slaves. Some of these Albanoidics escaped westward into the mountains to hide. From these less hospitable, poorer West Balkan areas some of these escaped Albanoidics crossed the Adriatic to Italy and became known as Messapians. The rest remained in the Western Balkans and became known as Illyrians whose direct descendants, I believe, are the Albanians who, incidentally, have kept up their old tradition of wandering on to Italy. This scenario explains some of the Non-Romance, "native" lexical corespondences between Rumanian and Albanian. Some of these items are Thracian and Dacian words which the ancestors of the Albanians learned from their Baltoidic Thracian and Dacian masters.

The Dacians were the ancient Southern Baltoidics more likely to have occasionally made trips back to Old Baltoidic territory to the north as their name suggests. Duridanov says its conceivable tie to Lithuanian dakyti 'agitate, make a mess,' dvaknoti 'act rashly,' dvakas, dvokas 'smell' suggests that the name Dakoi means "mobile, restless people" ("fahrige, unruhige Menschen"). Dacians moving north on passing through intervening Albanoidic territory in the Carpathians surely picked up some Albanoidics to bring along as slaves. These became later known as Slavs. The emasculated nature of Slavic from the viewpoint of old Indo-European vocabulary, that is, the lack in Slavic of words like vyras, aner 'man' and smakras, the old masculine word for "beard," attests to the servile status of their ancestors, the Albanoidic Pre-Slavs.

Assuming all this to be true, the dearth of ancient Albanoidic place names, be they Illyrian, Albanian, or Slavic, is no surprise. Surely Pre-Slavs and possibly early Slavs escaped from Baltoidic masters where and when they could as did their Southern Albanoidic cousins from Dacians and Thracians. And these runaways were not likely to give names to prominent geographic features which might aid their Baltoidic captors in finding them. It is ironic that now the dominant languages in both the Balkans and the Baltic has for the last 300 years tended to be Slavic, that is, Albanoidic rather than Baltoidic. In the Balkans, since approximately the seventh century, A.D., Baltoidic Thracian and Dacian even ceased to exist while Non-Baltoidic Albanian still survives only because the ancient prehistoric ancestors of the people who speak it managed to escape and stay free from their form Baltoidic masters.

Now, to display their Baltoidic heritage, here are some of the Thracian and Dacian words given by Duridanov with Baltic counterparts:

Thracian (Suntus: Lithuanian place names Suntupių kaimas, Suntuokių vienkiemis
Thracian z(i)burul 'flashing light': Lithuanian žiburys 'spark' 
Dacian zuv-: Lithuanian žuvis, Latvian Dialect zuva 'fish' 
Dacian zuras, Zyras: Latvian zveruot 'flash,' žuret 'blink,' Lithuanian žiūrėti look at'
Dacian Naparis: Prussian water and place names Nauper(y)n, Panawpern, Pa-naupern, Po-nopern

The influence of Southern Baltoidic, that is, (Pre-)Dacian and (Pre-)Thracian and the tribes that spoke these dialects with, originally, the heaviest early glottalization seems to have been very large. These speakers, have inherited a special consciousness of form visible in the correspondences above, which, I believe, resulted from this earlier heavy glottalization determined the future developments of Albanoidic so that its speakers which they took south became, eventually, Messapians, Illyrians, and Albanians, while those which they took north became Slavs. These people spoke a dialect of Indo-European with weaker glottalization and, consequently, had a lesser consciousness of form. The substantial differences between Albanian and Slavic can surely be traced to these very active Baltoidics who separated the speakers of a more or less uniform ancestor Indo-European Albanoidic dialect. Those speakers, being less form conscious, were going to permit changes large enough to divide Southern from Northern Albanoidic and eventually make these new variants mutually unintelligible to their speakers. Thus, Baltoidic (Pre-)Thracians and (Pre-) Dacians started the creation of two different languages from basically one Albanoidic prototype. Otherwise, as for what was going to survive from Baltoidic, we can say that since contact with Northern Baltoidics, I believe, seemed to have been maintained by Southern Baltoidic (Pre)Dacians and, possibly, even (Pre-)Thracians after their moves to the Eastern Balkans, their influence, I suspect, was startling. Lithuanians and Latvians may well have ties with "cradles of civilization" which were far more immediate and constant than most specialists have imagined!