LITHUANIAN QUARTERLY JOURNAL OF ARTS AND SCIENCES
Volume 39, No.2 - Summer 1993
Editor of this issue: Robertas Vitas, Lithuanian Research & Studies Center
Copyright © 1993 LITUANUS Foundation, Inc.
SLAVIC, A BALTICIZED ALBANIAN?
Harvey E. Mayer
Hamp (1984) in his criticism of my article (Mayer 1981), after his inaccurate comment that Iranian, because of its vowel system, 'can surely be classified only with Indic',1 makes statements about Albanian which stimulate major investigation.
He says: 'It can be said to be related more closely to Baltic and Slavic than to anything else, and certainly not to be close to Thracian.'2
I ask: If so, is Albanian more closely related to Slavic than to Baltic? And if it is, was Pre-Slavic originally so closely tied with Pre-Albanian in Late Dialectal Indo-European times that both represented very minor variations of the same dialect as opposed to other dialects like Pre-Baltic, Pre-Iranian, Pre-Germanic, etc.?
Martynov (1981) says that Proto-Slavic is Italicized Proto-Baltic while Common Slavic is Iranicized Proto-Slavic. This is all based on lexicon. My suggestion that Slavic mainly reflects in its essence an original Pre-Albanian variant both phonologically and lexically with a huge superimposed Baltic lexical influence now extending into morphology rests on a broader and deeper linguistic base. With this, it offers some answers to some vexing questions.
Phonologically, Slavic and Albanian have the following notable ancient ties: 1. Th, TH to T, 2. Dh, DH to D (T = voiceless stop, D = voiceless stop, h = aspiration, H = laryn-geal), 3. s alternating with h (not true of Baltic), 4. reflexes of k' (g'(h) kept separate from those of the ruki law (not true of Baltic), 5. ks- to h- (not true of Baltic), a special, exclusive Albanian-Slavic reflex, 6. more cases of k', g'(h) to k, g than other satem languages including Baltic which shows sibilants instead (Shevelov 1965 for the Slavic).
Lexically, Slavic and Albanian correspondences minus Baltic ones outnumber Baltic and Albanian correspondences minus Slavic ones.3 This is striking when we consider that the opposite is true for Hittite and Tokharian.
Viewing (Pre-)Slavic as a variant of (Pre-)Albanian, considering their relative geographic positions as a constant, makes it easy for us to see just how Martynov's comments about Italic and Iranian influences can apply not only for Slavic, but for Albanian as well. Note that North Iranian contacts apparently influenced Slavic while South Iranian contacts apparently influenced Albanian.4
This (Pre)-Slavic-(Pre) Albanian view allows us also to suggest an answer to the question mentioned by Trubačev (1981) of Common Slavic's absence from the region of Old European hydronomy in which Baltic plays an important role, possibly even that of the center of its radiation. He says, "... and this is very odd because it contradicts the supra-language and supra-dialect character of the named hydronymic region and also contradicts all the old ties of Common Slavic with other Indo-European languages of Europe, and finally contradicts the theory of departure of Common Slavic from the heart of Common Baltic or its western branch.' This (Pre-)Slavic (Pre-)Albanian view might also allow us to state with considerable assurance our answer to the question of the origins of Albanian, especially since a similar problem with ancient Albanian place names in Albania occurs. The answer to both questions is probably that the (Pre-)Slavs and (Pre-
The possibility of special, close aboriginal (Pre-)Slavic-(Pre)Albanian dialectal ties indicates the Carpathians as a common ancestral home for (Pre-)Slavs and (Pre-)Albanians where they led a pastoral, migratory existence. This location surely originally put them out of reach of contact with Pre-Baltic.
This is evidenced by two ancient phonological differences: 1. Slavic's and Albanian's reflexes of k', g'(h) separate from those of the ruki law versus Baltic's early immediate merger of them into š/s, ž/z, 2. The special Slavo-Albanian reflex of ks-to h-: Russian dialect xinit' 'to condemn', Russian xilyj, xiloj 'sickly'; Albanian (h)unj, Shkoder ulj, ulem 1 belittle' where h-is from ks- if not kh- (Fasmer 1973; 236-8)5 versus Baltic's metathesis of original ks- to sk-t an ancient change predating the ruki law: Lithuanian skaudus 'painful' versus Slavic xudu 'bad' (Stang 1965:95).
It is significant that wherever original sk-/ks- occurs, Baltic has sk-, Slavic has reflexes of either, while Albanian, where anti-Baltic drift is the strongest, has h- from an original ks- or, I believe, from a ks- via metathesis from an original sk- before a back vowel as in he from skoia 'shadow' (Fasmer 1973:602).
My present view of Slavic as a heavily Balticized Albanian, I believe, helps make my answers to all of Hamp's objections to my previous article's position against a Balto-Slavic protolanguage cogent despite the 'evidence' he cites for it including 'syllabic contrast between long and short or acuted and circumflex',6 'the remarkable agreement of Baltic and Slavic in the incidence of a double reflex of the Indo-European syllabic sonants,7 and in the lengthening of syllables before Indo-European mediae,8 and in the derivational vrddhi affecting i's and u's developed in Baltic and Slavic analyzed so fully by Leskien, and in several basic formational features of the verbal system.' In the end all of this involves borrowing, calking, and otherwise favoring of morphemes, even those occurring as doublets, of one sort (words) or another (affixes) over others via dialect or language clustering contact.
With Pre-Slavs seen as originally basically nomadic northern Pre-Albanian descending northward from the Carpathians into regions with sedentary, most likely, technically superior Baits, we can expect their dialect to have been strongly influenced by Baltic. A further sign of this ancient situation is the much greater degree of homogeneity of Slavic than of Baltic where more anciently sedentary dialects tend to show far less homogeneity than migratory ones. Albanian, with more homogeneity than Slavic, shows an even later onset of sedentary habits of its speakers who, having moved south, characteristically, seem to have replaced one mountainous region (the Carpathians) with another (the Balkans) to continue an earlier, somewhat migratory, nomadic type of living before finally becoming sedentary.
I am grateful to Hamp for his remarks about Albanian. They helped me identify Pre-Slavic as a variant of Pre-Albanian and support the notion that it was originally significantly different from Pre-Baltic under whose influence it later fell. In a way, Slavic resembles English, a powerfully Romanized Germanic language. But where the situation with English has been relatively easily identifiable, the analogous situation with Slavic has been very elusive. This is not surprising. The influence of conservative Baltic gave Slavic a conservative appearance. The influence of innovative neighboring dialects and languages (Celtic, for example) on Albanian gave it a much changed appearance. The immediate reaction has been: Albanian and Slavic look so different while Baltic and Slavic seem so much alike.
To counter this, I provide the reminder that the proposed Slavic-Albanian linguistic unity ended a long time ago, when Indo-European syllabic nasals n, m, n, m became vowels plus nasals. Before this, Pre-Slavic and Pre-Albanian had had different new neighbors. For, perhaps, centuries before this Pre-Slavic, and not Pre-Albanian, had had Pre-Baltic and Pre-Germanic, and for, perhaps, millennia after this Slavic, and not Albanian, had Baltic and Germanic as neighbors.
Fasmer, Maks. 1964-1973. Etimologičeskij slovar' russkogo jazyka, tr. by O. N. Trubačev. Vote. 1-4. Moskva: Izdatelstvo 'Progress'.
_. 1973. (Volume 4 of the above) 236-8, 253, 602.
Fraenkel, Ernst. 1962-1965. Litauisches etymologisches Worterbuch, Vote. 1,2. Heidelberg: Carl Winter Universitatsverlag. Gottingen: Vandenhoeck & Ruprecht.
Hamp, Eric P. 1980. Thracian, Dacian and Albanian-Romanian Correspondences. Actes du IIe Congres international de thracologie. Vol. 3, 57-60. Bucuresti.
_. 1984. On Myths and Accuracy. General Linguistics. 24.4.238-9.
Martynov, V.V. Balto-slavjano-iranskie jazykovye otnošenija. Balto-slavjanskie issledovanija 1980, 16-26. Moskva: Izdatel'stvo "Nauka'.
Mayer, Harvey E. 1981. Two Linguistic Myths: Balto-Slavic and Common Baltic. Lituanus. 27.1.63-8.
_. 1983. Zum Lexikon und der balto-slavischen Frage. San Antonio: Defense Language Institute. /To appear in Zeitschrift fur slavische Philologie./
_. 1987. Prussian, an Aboriginal a-Language? Lituanus. 35. 5. 66.
Shevelov, George Y. 1965. A Prehistory of Slavic: The Phonology of Common Slavic, 139-49. New York-Morningside Heights: Columbia University Press.
Stang, Chr. S. 1966. Vergleichende Grammatik der baltischen Sprachen, 95. Oslo-Bergen-Tromso: Universitetsforlaget.
Trubačev, O.N. 1981. Replika po balto-slavjanskomu voprosu. Balto-slavjanskie issledovanija 1980, 4. Moskva: Izdatel'stvo 'Nauka'.
Winter, Werner. 1978. The distribution of short and long vowels in stems of the type Lith. esti: vesti: mesti and OCS jasti: vesti: mesti in Baltic and Slavic Languages. Trends in Linguistics 4. Recent Developments in Historical Phonology, ed. by J. Fisiak, 431-46. The Hague-Paris-New York: Mouton.
1 Indie and Iranian once had five vowel systems of a, e, i, o, u, with e reconstructible from a in ca, ja from ke, ge, short o restorable by Brugmann's law of oR + vowel to aR + vowel, and long o retraceable from long a in formations alternating with those with a demonstrable erstwhile short o. With short o merging with short a before the merger of their long counterparts in Iranian, we arrive at the vowel system of a, e, i, o, u which I proposed for a conceivable Late Central Indo-European dialect including Iranian.
2 Hamp (1980) says that in Pre-Roman times palatalization of Indo-European labio-velars before front vowels occurred in Albanian but not in Thracian which proves Albanian could not have been Thracian. But can we be sure that the Latin and Greek symbols used circa Roman times represented unpalatalized velars? Note also that Hamp surely classifies
Lithuanian as dose to Latvian despite Lithuanian's kelti, kilti, gyvas, geltonas, etc. without the palatalizations Latvian shows in celt, cilt, dzivs, dzeltans ('raise, rise, alive, yellow'), etc.
3 In an article (Mayer 1987) I mention that examination of Fasmer 1964-1973 and Fraenkel 1962-1965 indicates that Slavic-Albanian-minus-Baltic native cognate roots outnumber Baltic-Albanian-minus-Slavic ones by approximately 1/3. Since we expect languages of long separation to match better in original lexicon with more conservative Baltic, this opposite situation with Albanian indicates special ties with Slavic as a unit separate from Baltic.
4 For Indo-European k', g'(h) North Iranian and Slavic coincide with s, z while South Iranian and Albanian coincide with th, dh.
5 This is less likely from kh. Note k, not h, in theke 'tip' from a possible kh.
6 In an article (Mayer 1983) and again in another (Mayer 1987) I explain the opposition in Baltic and Slavic of circumflex versus acute which replaced that of short versus long in tautosyllables as having arisen from a process of affixation initiated in one dialect or language and later spread to the others via calking. In these positions I saw long syllables arising from infixed vowel morphemes homophonous with already present root vowels, e.g., *varn- 'raven': *va-a-rn- 'crow'. Here, by analogy with long root vowels inherited from Indo-European as the long o in *do- 'give' the new long vowels arising through contraction were acute. The same sort of new long vowels in affixes were circumflex as the e, (e/ė) in Old Church Slavonic neseaste: Lithuanian nešėte 'you carried'. This new long-vowel-creating process involving positioning of morphological elements was syntactic and, therefore, as easily calked as lexicon. In origin it was not at all phonetic.
7 Evidence from Indie (r, ur, ir) indicates three possible reflexes of an Indo-European syllabic resonant (R):1. neutral R which later became aR in many dialects and daughter-languages, 2, labialized uR, and 3. palatalized iR. Where o tended to become a, labialized uR and/or palatalized iR tended to be generalized for morphological purposes to avoid confusion of TR(T) and ToR(T) reflexes, now both TaR(T). Through dialect and language clustering influences (calking, etc.), uR was stabilized as the reflex of R was stabilized as an alternate reflex of R in satem palatalizing Baltic and Slavic.
8 This lengthening of syllables before Indo-European mediae mentioned by Winter (1978) was not phonetic, but morphological in origin. Thus we find long e in Baltic and Slavic sed- 'sit down' matched by the same in Gothic setun 'they sat'. Short vowels occur in the same morpheme with shifts in meaning in Gothic sitan 'to sit' and Slavic šid- 'having gone', xod- 'going'. (Fasmer 1973:253 for the etymological connection.) Also, note the additional exceptions with (originally) short vowels: Slavic voda 'water', koza 'goaf, ogni 'fire', čeznoti 'to disappear', kogutu 'claw', stogu 'stack', stirženi 'pivot'; Old Russian: mulzu 'I churn butter'; Russian lizat' 'to lick'; Lithuanian: laigonas 'brother-in-law', luba 'ceiling board'.