Volume 27, No. 1 - Spring 1981
Editor of this issue: Antanas Klimas
ISSN 0024-5089
Copyright © 1981 LITUANUS Foundation, Inc.


 California State University, Fullerton

Let's talk about mythical languages. First there's Balto-Slavic. Then there's Common Baltic. You've surely heard about those. Then there's Illyro-Albano-Thracian. And finally there's Central Indo-European. That, incidentally, consists of Illyrian, Albanian, Thracian, Prussian, Latvian, Lithuanian, Slavic, and Iranian. I doubt that you've heard of these last two protolanguages because I made them up. Let's examine these hypotheses one by one.

If you trace the developments of the languages I call Central Indo-European back to their earliest stages, you will find that their phonemic systems matched. They all included the following sequences: p, b, t, d, k, g, k, g, s, m, n, I, r, j, v, long and short i, u, e, a, and long o. Yet no one, except me, has offered this hypothesis of Central Indo-European. Why? Because the languages involved don't have enough words in common. In fact, not even the Balkan languages I mentioned do. The situation there is so bad that specialists agree on only one thing — that Illyrian and Thracian are dissimilar. They don't agree at all on Albanian. Was it originally Illyrian, Thracian, or a type all its own? No one can say for sure.

As for Common Baltic, this is what we find. Prussian, Latvian, and Lithuanian do have many words in common. But these languages do not share one single phonological innovation resulting in a new phoneme that exists in all of them on the one hand, yet is not to be found in other Indo-European languages on the other. The same applies to Balto-Slavic. Still, people find Common Baltic and Balto-Slavic palatable as theories.

We now see that it was common vocabulary, and this includes grammatical endings, that lured linguists into creating the myths of Common Baltic and Balto-Slavic while the lack of it inhibited them from formulating the other two myths. But it took more than that to fool them in the first place and to cause them to perpetuate those myths in the second. It was the phonological shapes of large numbers of words in common found in the Baltic and Slavic languages and dialects which added the final deceptive touch.

To understand why this was and still is so, you have to know what is meant by the phrase: "the conservative nature of Baltic and Slavic". You should ask: "Conservative? With respect to what?" I say, conservative with respect, in particular, to Central Indo-European phonology, that phonology which is classifiable as Late Common Dialectal Indo-European. To clarify this, I say the following.

Linguists easily recognize Latin words in English as borrowings by their phonological shapes no matter what their vintage. But they mark early borrowings between Baltic and Slavic only with considerable difficulty, and cannot at all mark early borrowings between just the Baltic languages and dialects.

Perhaps now you can see why the Common Baltic and Balto-Slavic myths are so durable. They are supported by large look-alike vocabularies which defy analysis. You can hardly, if at all, tell which words are native and which are borrowings between Baltic and Slavic types of speech and between only Baltic types of speech. The conservativeness of phonology obscures and conceals this. Yet none of this seemingly "common" vocabulary points to either a Common Baltic or, even less, a Balto-Slavic protolanguage.

What, you should ask, do you look for as proof of Common Baltic or Balto-Slavic? You look for at least one special kind of phonological innovation resulting in a new phoneme. This innovation, thus, would regularly mark many words of every dialect as Baltic or Balto-Slavic as opposed to the same words without this innovation in other Indo-European dialects. In other words, we have proof of Proto-Germanic with the h in English heart, German Herz, Gothic haîrtô, Old Icelandic hjarta as opposed to the k in Latin cordis (genitive), Greek kardia, the s in Russian serdce, Armenian sirt, Latvian sirds, Prussian seyr from *serd-, or the đ in Lithuanian đirdis, all handed down from original Proto-Indo-European. All other intermediate Indo-European protolanguages have something like this. Common Slavic has several things like this, for example, y for long u elsewhere in Indo-European as in Russian myđ meaning 'mouse' versus Old High German mûs. Only the Baltic languages lack this common denominator. This means that "Common Baltic is a linguistic myth.

To repeat, there is not even one Common Baltic phonological innovation like the Germanic or Slavic ones. Aside from two questionable shifts found at analogy-producing morpheme boundaries dividing stems from grammatical endings, namely, the loss of jod between consonants and front vowels, which, incidentally, produces no new phoneme at all, and the debatable one of long and short a to long and short e to form the e-stems which, by the way, are morphemes and could have been merely borrowings in most Baltic dialects from another dialect or other dialects, the only phonological innovations we find are those which do not satisfy our conditions. Either they cover only some of the Baltic dialects, like đ in Lithuanian ones as a reflex of both palatal k and s under ruki law conditions, or they also can be found in other Indo-European languages, like the oppositions in phonemic pitch.

Those who favor the Balto-Slavic myth place a lot of emphasis on phonemic pitch in Baltic and Slavic. They admit that phonemic pitch had also existed in Greek. But they argue that in Greek it was confined to final syllables. They say that only in Baltic and Slavic can it be found internally in the word. But I say that none of this matters since the pitch opposition acute versus circumflex as something phonemic arose in Baltic and Slavic only after those types of speech had become separate entities. It is the relationship between phonemic pitch and the nature of ablaut in standard Lithuanian which makes this clear.

You got phonemic pitch in internal syllables in those languages once acute pitch there stopped being redundant. For this to happen, long vowels in certain positions had to shorten. They had to do this in diphthongs with falling sonority in closed syllables. Thus when long vowels were followed by i, u, r, I, m, n in tautosyllabic diphthongs and were shortened, the acute pitch which had been an additional, redundant property of theirs now became phonemic. This acute pitch alone differentiated the word it was in from a similar one with circumflex pitch, the pitch of diphthongs with original short vowels. Thus in Lithuanian we have kárti meaning 'to hang' with only rising pitch to distinguish it from karti, the vocative of kartis meaning 'bitterness', which has falling pitch. All the instances of this phonemic pitch arose not in syllables reflecting Indo-European ablaut, but reflecting secondary ablaut developed later by Slavic and Baltic languages, as well as by other tongues after they had existed awhile as entities separate from Common Indo-European.

Slavic, Lithuanian, Latvian, and Prussian developed so-called long grades, as did Sanskrit with its vrddhi in internal position in roots as morphological devices. The evidence shows that Lithuanian and Latvian did this after they were entities separate from Indo-European. It was the shortening of these vowels representing "Baltic" ablaut that gave rise to phonemic pitch in internal syllables, that is, in root or stem syllables in these languages. As for Slavic, the evidence is not as clear even though linguists refer to long vowels in analogical positions in Old Church Slavonic as "Slavic" ablaut.

In Lithuanian and Latvian we can clearly separate old ablaut from new ablaut, but only with respect to the reflexes of Indo-European o and a. The old alternation is between an original long o which appears as uo with acute pitch versus an a with circumflex pitch in Lithuanian as in Lithuanian dúoti meaning 'to give' versus dãvë meaning 'he gave'. The new alternation is between an original long a which remains a long a in Latvian but appears as a long o with acute pitch in Lithuanian versus an a with circumflex pitch in Lithuanian as in Latvian kâra and Lituanian kórë meaning 'he hung' versus Lithuanian kãria meaning 'he hangs'. In a closed syllable we find an acute pitch on the a in the infinitive kárti in Lithuanian. This arose from a shortened long a which had developed in Lithuanian after it had been a separate entity.

The oldest level of Indo-European ablaut is short e versus short o. Short a versus either of these was rare in root syllables. Consequently, later developments in the corresponding long vowels in root syllables in Common Indo-European included long e versus long o, and not long a. In Central Indo-European languages short o and a fell together to a. No one can say when this and other common phonological changes occurred. They very likely arose after they had all been independent entities. This is what the fundamental divergence in their vocabularies indicates. Later, vrrdhi formations occurred in the Baltic and Slavic languages. In Lithuanian and Latvian they arose clearly as long a- grades. This mirrored the shift in ablaut alternations in short vowels from e versus o in Common Indo-European to e versus a in Lithuanian and Latvian.

Thus we can see from Lithuanian and Latvian that phonemic pitch in internal root or stem syllables arose not from any reduction of long o to short a which would indicate something ancient. It arose from reduction of long a to short a which indicates something recent. How, we might ask, are Balto-Slavic correspondences in pitch of this sort supposed to prove a Balto-Slavic protolanguage?

For Balto-Slavic pitch agreements to indicate an inherited rather than borrowed origin, we should find, for example, something like a hypothetical Lithuanian *kúorë with the long o reflex uo plus an acute alternating with kárti, all of which should have counterparts in Slavic like a hypothetical aorist *kare in Old Russian versus a hypothetical infinitive *korót' in Modern Russian. Instead, what do we find? Things like acuted várna meaning 'crow' versus circumflexed varnas meaning 'raven' in Lithuanian corresponding with voróna versus vóron in Russian! How weak! These things at best represent parallel development. Probably, though, they are nothing more than calks and other borrowings. And this applies not only to these and similar borrowings, all really on a morphological level, and not, as it might seem, on a phonological one, between the Baltic and Slavic languages, but also between the Baltic languages alone.

Thus there is no firm evidence for either Balto-Slavic or Common Baltic as protolanguages. We will have to consider them, therefore, as myths. This is especially necessary since the attested dialects involved are conservative fundamentally with respect to Central Indo-European, a phonological type arrived at independently by all Indo-European dialects reflecting it.


H. Mayer. "Baltic Membership in the West-Satem Subgroup", presented at the 6th Conference on Baltic Studies, Toronto, 1978.
 „   "Die Divergenz des Baltischen und des Slavischen", ZslPh, 1978.
 „   "Die frühzeitige Eigentümlichkeit des Slavischen", to appear in ZslPh. 
„   "Kann das Baltische als Muster für das Slavische gelten?", ZslPh, 1976.
 „   "The Balto-Slavic Protolanguage and Pan-Slavism", submitted for reading to the Vth Conference on Baltic Studies, Stockholm, 1979. 
„   "The Function of the Concept 'Balto-Slavic' ", JBS, 1975. 
A. Meillet. Les dialectes indoeuropéens. Paris, 1908.