Mother Tongue Newsletter 10 (April 1990)

7/27/2019 Mother Tongue Newsletter 10 (April 1990)

1/64

H

T0

--G--... ..

ENEWSLETTER OF THE

ASSOCIATION FOR THESTUDY OF LANGUAGEIN PREHISTORY

Issue 10, April 1990-........--

- - ~ ---- - -- --- . -- - - - - - - - - ~ -


2/64

ASSOCIATIONFOR THE STUDYOF LANGUAGEIN PREIDSTORY

OFFICERS: Harold C. Fleming, PresidentAllan R. Bombard, Vice PresidentAnne W. Beaman, SecretaryMary Ellen Lepionk:a, Treasurer

The Association for the Study of Language in Prehistory(ASLIP) is dedicated to exploring all aspects of distantrelationship among the languages of the world.Membership dues (in U.S. dollars) are $10.00 for the UnitedStates and Canada, $16.00 for Latin America, and WesternEurope, and $18.00 elsewhere.Inquiries, manuscripts and news items for inclusion in MotherTongue, and applications for membership should be sent to:

Allan R. Bombard, Vice PresidentAssociation for the Study ofLanguage in Prehistory73 Phillips StreetBoston, MA 02114-3426U.S.A.----

-------------------


3/64

MOTHER TONGUE 10, APRIL 1990

TABLE OF CONTENTS

1. John C. Kerns: Review of Colin Renfrew, Archaeology and Language: ThePuzzle of Indo-EuropeanOrigins2. Vitalij Shevoroshkin: Comments to the Revised Version of Murtonen'sComments3. Vitalij Shevoroshkin: Comments on Barnhard's Supplement to MT: "Lexical

Parallels between Proto-Indo-European and Other Languages"4. Vitalij Shevoroshkin: The Russians are Coming (to Ann Arbor)5. Karl Krippes: Problems Concerning the Comparison of Korean with OtherLanguages6. John D. Bengtson: An End to Splendid Isolation: The Macro-CaucasianPhylum7. Allan R. Barnhard: Review of Vitalij V. Shevoroshkin and Thomas L.Markey, eds., 7Ypology, Relationship, and Time8. Stephen A. Kaufman: Computer Program Information. Oriental Words: NewProducts for Processing and Crunching Them


4/64


5/64


6/64

the essence of science, for anything less than this is a devolution to ideology wherebymistaken notions are perpetuated indefinitely.In his first two chapters (9-19, 20-41) Renfrew summarizes the main linguistic andarcheological works that have been brought to bear on the question of Indo-European

origins. He mentions suggestions by Otto Schrader (1890), Gustav Kossinna (1902), V.Gordon Childe (1915 to 1958), Marija Gimbutas (1956 to 1979), and Pedro Bosch-Gimpera(1940 to 1961). Of these, Schrader, Childe and Gimbutas have favored a South Russianhomeland, while Kossinna suggested a North-Central European origin. On the other hand,Bosch-Gimpera has favored a Central European origin dating from the earliest Neolithictimes; this view is closest to those held by Renfrew and me. Special attention should bepaid to Renfrew's concluding remarks (41 ), to the effect that while these studies involvedthorough investigations, they all have been based on assumptions dating from the time ofKossinna and Schrader. These assumptions need to be reexamined, and i f they prove to beunjustified (as they do), the conclusions based on them are likely to be mostly wrong.In his third chapter, "Lost Languages and Forgotten Scripts" (42-74), Renfrew

presents a survey of the history of Indo-European languages and scripts as elucidated byscholarship during the past two centuries. Purists and specialists can quibble on one pointor another, and should; for example, I consider his Table XI (74) to be in error in somedetails. But in view of the complication of such multifaceted phenomena, it is hard to seehow a better survey could be provided within a single chapter.

In his fifth chapter, "Language and Language Change" (99-119), Renfrew provides aconcisely stated review of work done by linguists, mostly Indo-Europeanists, with specialreference to the last quarter century. He discusses such matters as phonemiccorrespondences, the family tree model, Schmidt's Wave Theory, loan words and vocabularyinvention, semantic change, word order (whether VO or OV), glottochronology andlexicostatistics, and he briefly comments on each of these. In general, his conclusions arein accordance with views held by most Indo-Europeanists, who will find them conventional.But for anyone else, this chapter is a condensed educational gem.These two chapters seem to have been provided mainly for the benefit ofarchaeologists, who maybe expected to be inexperienced in Indo-European linguistics. Theyserve the purpose well.Chapter four, "Homelands in Question" (75-98), presents Renfrew's arguments againsta South Russian homeland for the Indo-Europeans. He raises three points. First, appealsto the protolexicon for evidence of a specific place of origin and mode of life have provedto be unreliable.Second, such widely spread (and relatively late) archeological phenomena as BellBeaker and Corded Ware are increasingly recognized as due far more to local culturaldevelopments than to invasions. While it is true that the wide geographical uniformity ofthese phenomena suggests far-reaching cultural influences, there is no reason to supposethat it was due to anything more than intertribal trading activities or visits by prominentindividuals, including technical specialists. Under the circumstances, it is highly probablethat the linguistic effects were superficial, consisting mainly in the borrowing of vocabularyacross tribal lines with only minor displacement of the languages themselves; certainly this

2


7/64

is overwhelmingly implied by the extant linguistic phenomena, at least in the Central, North,and West-Central parts of Europe -- the Eastern edge of non-Russian Europe, and nearbyparts of Russia and Asia, evince late linguistic displacements on an extensive scale, mostlyWest-to-East and North-to-South.Third, the facile assumption that pastoral nomads or mounted warriors from SouthernRussia were preadapted for economic exploitation of Central and nearby parts of Europe,with its relatively hilly and more densely forested terrain, has been thoroughly discredited.Furthermore, the pastoral-nomadic way of life itself is dependent on continuous agriculturalactivity to sustain it, and indeed could not have evolved except from an agricultural society.In Southern Russia this society was clearly the Tripolye culture, linked by the Cucuteniculture to earlierLinearPottery settlements on the upper Dniester in a process of adaptationtoward steppe conditions. Under the circumstances, it is highly probable that the languageof the Kurgan people was indeed Indo-European, but its origin was in the Linear Potteryculture of Central Europe, not in Southern Russia or Central Asia.I fully agree with Renfrew on these points.

In the sixth chapter, "Language, Population and Social Organization: A ProcessualApproach" (120-144), Renfrew introduces a word which is new to me. The AmericanHeritage Dictionary (Houghton Mifflin, 1980) does not list "processual", and in e b s t e r ~ NewUniversal Unabridged Dictionary (Simon and Schuster, 1979) the meaning is given as "in law,relating to some legal proceeding or judicial process", apparently in reference to someformalized legal procedure recognized as being especially applicable in a particular legalsituation. It appears that archeologists have recently adopted the word to refer tointellectual models of special social processes which are believed to have been significantin the development of certain archaeologically observable phenomena, with particularreference to the possibility of using some of these phenomena as plausible indicators oflinguistic replacement in prehistoric societies.Thus defined, processual models seem to be a generalization of a more restrictedclass, mathematical models, familiar to physicists and engineers. Anyone who has devisedsuch a model is aware that it always embodies simplifying assumptions needed to make itfeasible to draw conclusions from it. Consequently, no matter how rigidly these conclusionsmay have been implied by the model, the validity of the results as a solution to a problemin the real world is always in doubt. In the case of the more general class of processualmodels the doubt is even greater. Renfrew is aware of this.Under the beading of linguistic change within a given area, Renfrew cites threesituations: (1) Initial colonization of the area by human beings and their language, (2)Replacement of the current language by another brought into the area by settlers orinvaders from outside, and (3) Evolution of the language into new dialects and theseeventually into mutually related but separate languages. He recognizes that at least the twolatter processes were operative at one time or another in the course of Indo-Europeanprehistory. Although the area in questionwas not named, the region in and around CentralEurope was very likely uppermost in his mind.Renfrew then focuses on the second of these, linguistic replacement. He discussesseveral processual models, as follows:

3


8/64

1) Demography/Subsistence. Under this heading he includes the Wave ofAdvanceModel.2) Elite Dominance.3) System Collapse.4) Constrained Population Displacement.5) Sedentary/Mobile Boundary Shift.6) Donor/Recipient Population Systems.Renfrew presents historical examples for each of these. For me, the Wave ofAdvance Model is the one which holds greatest interest because of its applicability to theLinear Pottery expansion, which quickly established primitive agriculture and minorpastoralism in Central and most of Northern Europe from Eastern Holland to the WesternUkraine. With a potential relative population density of some 50 to 1 over the earlierMesolithic population, it is highly plausible that the languages of the earlier peoples wereswamped by the effect of sheer numbers. This does not mean that the Mesolithicpopulation necessarily became extinct. On the contrary, the survival of Nordic populations

in Northern Europe strongly suggests that their biological ancestors willingly adopted theLinear Pottery culture, including its IE-Proper language, and actively participated in itspenetration into their territory. Of course, this implies that the earlier languages of theNordic peoples were not Indo-European since IE-Proper seems to have been the onlydialect of IE encountered by these peoples, though their languages may possibly have beenrelated to IE as members of some other Nostratic family, now extinct. Unfortunately, thereappears to be no way we can verify this.Thus far, Renfrew has concentrated on demonstrating the inadequacy of previoussolutions to the problem of Indo-European origins. I find his arguments convincing, and asstated above, I have further arguments of my own based on linguistic considerations which

lead to the same conclusions.In his remaining chapters Renfrew offers his own suggestions for a solution, dealingwith the following topics identified by chapter number and title, as follows:7) "Early Language Dispersals in Europe" (145-177).8) "The Early Indo-Iranian Languages and their Origins" (178-210).9) "Ethnogenesis: Who were the Celts?" (211-249).10) "Indo-European Mythologies" (250-262).11) "Archaeology and Indo-European Origins: An Assessment" (263-290).These are grouped together since they unavoidably contain much that is speculative.

To discuss them in detail would be tiresome and space consuming, and my own conclusionsdiffer from Renfrew's in some respects. In brief, his suggestions can be summarized asfollows:First, he concurs with most Indo-Europeanists in rejecting Trubetskoy's suggestionto the effect that the similarities among the conventionally recognized members of IE aredue to a sustained process of convergence among languages which initially were_mutually4


9/64

alien. Accordingly, Renfrew concludes that the family tree model must be acknowledgedas having considerable validity for the deployment of IE in Europe, at least in its earlierstages. I agree with Renfrew, regarding Trubetskoy's suggestion as remarkably obtuse inview of the fact that although some convergences have undoubtedly occurred, they areoutweighed by the many radical divergences. Furthermore, most of the resemblances arein precisely those features that have been observed to be least subject to change.Second, Renfrew equates the Indo-Europeanization of non-Russian Europe with theintroduction and spread of primitive agriculture (and minor pastoralism) into the Balkansand through Central Europe; thus he regards the Linear Pottery expansion as a Northernextension of the Starchevo-Koeroes-Karanovo (here SKK) culture, and the Funnel Beaker andthe Tripolye cultures as later Northern and Eastern extensions of Linear Pottery. Again Iagree, partly for the archeological reasons he cites, but also in view of important linguisticconsiderations.Third, he derives the SKK culture of the Balkans from the earliest agriculturalsettlements in Greece, and these from an early farming and herding culture in WesternAnatolia. I agree that this seems plausible from the evidence he cites. But then he goeson to assume that the IE language likewise had its source in Anatolia. Here I disagree sincethe linguistic evidence is inconsistent with this assumption. And since IE is a linguisticphenomenon, the linguistic considerations must take precedence over the apparentimplications of the archeological phenomena, which in preliterate cultures are uninformativein regard to language.Fourth, he goes to considerable length to show that the Indus Valley culture may wellhave been Indo-Aryan in speech. He bases his argument on a semantic analysis of the RigVeda, seeking to establish that it shows no clear evidence of the commonly assumed Aryandestruction of a pre-Indo-European civilization. Still, although he may have been successfulin this, it does not suffice to prove his thesis. According to Renfrew, the civilization wasflourishing shortly after 3000 BC, which is rather early for the presence of Indo-Iranians inthis region. On the other hand, the proposition that the civilization was Dravidian in speechhas considerable plausibility in view of the fact that the early presence of Dravidian speakersin Northwest India is apparent from the many borrowings from Dravidian languages inClassical Sanskrit and some even in the Rig Veda, and by the residual presence of islandsof Dravidian speakers in that general region today (Burrow, T., 1959, 373-388); hence it isapparent that the Dravidians had considerable cultural power at that time and place.Furthermore, it is doubtful that an analysis of the Rig Veda can give an accurate picture ofearly Indo-Aryan life. For comparison, we can consider the Homeric epics and the Greekplays of the Classical period. Although their roots go back to the Greek Dark Age and thelatest phase of the preceding Mycenaean period, they give a highly romanticized picturewhich is very different from that presented by the Mycenaean records. Nevertheless, thematter seems irrelevant in regard to IE origins since even such a seemingly early date as3000 BC was late in the panorama of post-IE developments.Fifth, he regards the Celtic presence in Central and Western Europe as dating froman early period, at least as early as the beginning of the Central European Bronze Age(perhaps 2200 BC). I agree with this, in principle, provided we distinguish between Westernvarieties of pre-Celtic and the more particularized and developed Celtic of historic_al times.

5


10/64

Once we recognize the bearers of the Linear Pottery culture (in their early phase) as theundivided speakers of the IE-Proper dialect of Indo-European, we can see immediately thatit is highly probable that Western dialects of IE-Proper were spoken in the Netherlands andNorthern France as early as 5000 BC. From considerations of geography and the operationof Schmidt's Wave Theory, it is very likely that these dialects were more closely related tothe contemporary pre-Celtic (on the upper Danube) than to any of the other historicallyknown branches of IE. As I see it, the later presence of Celtic languages in the West andSouthwest is due to successive processes of Elite Dominance imposed on populations whichwere a lready somewhat similar to Celtic in speech. I believe Renfrew would agree with this.Sixth, he takes a highly skeptical view of the attempts of certain scholars, mainlyFrench (Joseph Vendryes, Georges Dumezil, Emile Benveniste), to show that the tripartitecaste system of kings and priests, military systems, and peasants and artisans evinced by suchwidely separated peoples as the Indo-Aryans, the Celts and the Italic speakers, together withtheir mythologies and legal systems and rules of poetic meter, had a common origin datingfrom the Common IE period. I agree with Renfrew totally in this matter, regarding thework of these scholars as fatally marred by their bias, due to their mistaken preconceptionsof early Indo-European life. In fact, it seems highly suspicious that the specific terms usedin these systems were almost entirely different in the separate branches; although some ofthe terms can be traced to the common lexicon, their specific application in these branchesis independent. As to the alleged similarities in poetic meter, these can be accounted forby the fact that the languages still had a common structure at the time the separate systemswere being developed (around 3000 BC, perhaps).

In this connection, we can consider the alleged IE root *reg-, meaning "king" or"powerful chief'. Since its reflexes are found in Latin, Celtic and Indo-Aryan, widelyseparated in historical times, it has been customary to regard the root as certifiably datingfrom the Common IE period. But to me, it does not follow. For it is clear from the extantlinguistic evidence that the existence of the speakers of these languages in their historicallyknown positions could scarcely date from earlier than 3000 BC, probably later; this is latein the pattern of post-IE developments. Prior to this, there was a period of at least 2000years during which the linguistic ancestors of these peoples lived in or near the generalCentral European area, from Austria to Southeastern Poland. During this period, importantdevelopments were occurring in this Central Region, as I call it, and among these was agradual intensification of intertribal conflict due to increasing population pressures heresince the best land had long been occupied (Milisauskas, 1978). Because of the continuallyrepeated encroachments, though each on a small scale, there developed a tendency for thetribes to drift away from this general area in an effort to get some relief, imposing theirlanguages on outlying neighbors by Elite Dominance probably reinforced by peasantsettlements. Accordingly, it is not surprising to find these peoples in the ir historically knownpositions by historical times. Contrary to common opinion, the fountainhead of post-IEaggression and migration was primarily in Central Europe, and only secondarily in outlyingregions such as Southern Russia. From a processual point of view, this development canbe seen as fueled by agriculture, which in the Central Region was relatively productive ascompared to regions farther East. This was because of a relatively high rainfall and mildclimate due to the influence of the Atlantic weather system. . _

6


11/64

With this as background, we can now understand the true nature and origin of thephenomena treated by the above scholars. Because of the developments in the CentralRegion here described, gradually accelerating during the fifth, fourth and third millennia, andbecause the linguistic ancestors of the Celts, Latins and Indo-Iranians were still situated inor near this region during much of this period, these peoples naturally shared in a commoncultural trend leading to a more intensively hierarchical social structure. As this developed,the tribal languages came to be no longer mutually intelligible, so the linguistic termspertaining to the new structure were applied mostly independently. Nevertheless, by theoperation of Schmidt's Wave Theory, there undoubtedly was some intertribal linguisticborrowing, and one example of this may well have been *rig-. This appears to be due toa vrddhi modification (i.e., a lengthening of the vowel} of the more ancient and widelyoccurring root *reg- with reflexes in our "right", "reckon" and "reach", and with more generalmeanings such as to straighten or align or regulate. The vrddhi development seems to haveoccurred in a particular branch of post-IE-Proper in the Central Region in response to theneed for a term for the gradually emerging institution of the conspicuous chief havingincreased powers of coercion and consequent wealth; the new term then spread to some ofthe other branches in this region. Although this apparently preceded the Satemdevelopment and the change of Celtic *e to*[, this is not significant since these were lateand parochial in the panorama of post-IE-Proper developments. Obviously, the occurrenceof *rile- in Common Germanic is due to late borrowing from Celtic after the Celtic vowelchange had occurred, indicating the late emergence of Germanic speakers from theirNorthern isolation in response to late Northward expansion by the Celts.

In view of this, it is easy to see why a coincidence of roots in Indo-Iranian andCentral or West European branches is by no means an adequate indication of inheritancefrom Common IE or even Common IE-Proper; the same remark applies in regard to asupposed similarity of traditions. I regret such a long discussion, but it seems justified inview of the characteristic confusion in discussions of this topic.

As expected, Renfrew's final chapter is a recapitulation. It also includes digressionson the evolution of several other linguistic families.This completes my review and commentary on Renfrew's book as it stands. A shortdiscussion will now be added, based mostly on linguistic considerations, which will confirmmost of his conclusions. Unfortunately, there are many facets to be considered, and because

of space limitations we cannot go into as much detail as the subject actually merits.Renfrew's purpose is twofold: first, to show the inadequacy of the presently populartheories of Indo-European origins, represented today mainly by Gimbutas' hypothesis, byexposing the falsity of their assumptions and the circularity of reasoning generally used to

sustain them, together with the shallow time depths and consequent anachronisms implied;and second and coequally, to show the high antiquity of the presence of post-IE speakersin the Central Region and especially in the Balkans. He is clearly right on both counts.Regardless whether we look at the matter superficially or in detail, the linguistic indicationsof this early presence are abundant.Superficially, it is immediately apparent from the pattern of river names in the

7


12/64

Central Region and the Balkans that these together comprised the true heartland ofIndo-European in Europe, for this seems to be the only region where non-IE names are notfound, while such names do appear in the Southern, Western and Northern fringes ofEurope (mixed with IE names) -- (Kerns, J.C., 1985, 132; Tovar, Antonio, 1977). Thus itappears that IE speakers must have dominated the heartland from an exceptionally earlyperiod, long antedating any activity by Gimbutas' Kurgan people. For comparison, we canconsider the situation in England and the United States, where English has long beenestablished, yet many river names are of non-English origin. There is a similar situation inFrance and Spain, where many river names date from pre-Latin times.Investigating in greater detail, we again find that the Central Region was the trueheartland of Indo-European expansion. It is here that we find the most ancient dialectaldifferences between the historically known branches, together with many indications ofprolonged mutual influence between them by Schmidt's Wave Theory, suggesting profoundlyancient geographical juxtapositions in nearly the same relative positions as evinced inhistorical times- with a few notable exceptions, easily accounted for. The most egregiousexception is Indo-Iranian, which shows clear signs of early intimate contact with Slavic(Burrow, T., 1959, 18-23) and also with Greek and Armenian, yet practically no evidenceof early contact with Tocharian - the eventual contacts of Tocharians with Iranian tradersand Indian Buddhists are irrelevant in this discussion since the late separation of lndic fromIranian had already occurred by this time.Incidentally, i f anyone believes that the early contact of Slavic with Indo-Iraniancould have occurred outside of non-Russian Europe, it should be borne in mind that thereare definite signs of Indo-Iranian contact with Finno-Ugric while both were still undivided(Collinder, Bjorn, 1977, 140-151), yet there are no signs of Slavic contact with anyFinno-Ugric speaking people until the Christian era. At the time of the contact ofIndo-Iranian with Slavic the Indo-Iranian leveling of the vowels e, o, a to a (long and short)had not yet occurred, but at the time of its contact with Finno-Ugric this leveling had indeedoccurred or was at least well under way. This clearly shows the chronological priority of theSlavic contact and that it existed in non-Russian Europe, and it graphically illustrates thelater Eastward movement of the Indo-Iranians.Another marked exception is Tocharian itself. According to G. S. Lane (LaneStudies, circa 1967-70, 76-87), it seems to show its earliest connections with the Northwesternbranch Celtic, and later with the Northeastern branches Baltic and Slavic (in their pre-Satemform, of course). In contrast, it shows no evidence of connections with its geographicalneighbor Indo-Iranian except at the much later period mentioned above. On the otherhand, Lane makes no mention of the fact that Tocharian does show a few specialagreements of vocabulary with Greek, unknown elsewhere; from circumstantialconsiderations, I consider this significant, for I regard the Tocharians and Greeks asrespectively Eastern and Southwestern extremes of relict populations of pre-Satem speakersof Tripolye origin, predating the Eastward Indo-Iranian movement by at least a millennium.The major remaining exception is Anatolian, which stands apart from the rest ofIndo-European (here called IE-Proper). It shows no sign of early special connections withany specific branch of IE-Proper in preference to any of the others, not even to Greek,Phrygian or Armenian, its nearest geographical post-IE speaking neighbors in -historical

8

- - ~ - - ~ - ~ ~


13/64

times. This fact is significant, and I will refer to it later since it militates against Renfrew'ssuggestion that the undivided IE language was injected into the Balkans from Anatolia, fori f he were right, it is very likely that there would be clear signs of close early specialrelationship between one or more of these three languages and Anatolian -- and no suchsigns exist. Here "early" refers to intimate contacts dating from the seventh millennium BC;there were indeed contacts after the third millennium, but that is irrelevant in thisdiscussion. This means that the speakers of each of these three languages were intrusivein their historically known positions.Although these three exceptions are the ones that spring most immediately to mind,they are by no means the only ones. Other less egregious instances are the repeated Celticexpansions to the West along the upper Danube, the Italic speakers to the South into Italy,the Greeks Southward into Greece and the Aegean area, and the Phrygians and ArmeniansSoutheastward into or near Anatolia. What is significant is that in all cases the tribesemerged into areas where they had no close early linguistic relationship with thesurrounding population. On the contrary, the earliest association of the Italic speakers waswith the early Celts on the upper Danube in Austria or Hungary (Lehmann, R.P.M & W.P.,1975, 83), the Phrygians and Armenians with the Greeks in the Eastern Balkans (Mann,S.E., 1963, vi), and the lndo-Iranians with the Slavs in Eastern Europe and secondarily withthe Armenians and Greeks. On the other hand, Anatolian shows no early specialrelationship with any branch of IE-Proper individually, but only with the entire group as awhole -- meaning that IE-Proper was still undivided at the time of its fan-like deploymentin the Linear Pottery expansion Northward from pre-Anatolian speakers in the Balkans(5200-4800 BC).

These phenomena, considered together, cannot be explained unless we recognize thefact that the main fountainhead of post-IE-Proper expansion was the Central Region in theaftermath of the Linear Pottery expansion and its economic and cultural maturing, notSouthern Russia or any place in Asia or the Near East. Also, the true source of theAnatolian speakers was in the Balkans in the SKK culture, not in Anatolia or the Caucasusor elsewhere in the Near East. In fact, from a processual point of view, the movement ofthe Anatolians from the Balkans can be seen as a Southeastward expression of the samepattern of recession from the Central Region as evinced by the Indo-Iranians in theirEastward movement, and for the same reason. Consequently, it is not surprising that thesemovements seem to have occurred at about the same time ( third millennium BC), which islate in the overall pattern of post-IE expansions. Yet another example of this process maywell have been Etruscan, which may have represented a Western member of the SKKculture, emerging in Northwest Italy as a result of a Westward movement. Unfortunately,the Etruscan language is little known, despite heroic efforts by Italian scholars and theircolleagues, but i f it was not an actual member of Indo-European, it was apparently moreclosely related to it (especially to Anatolian) than to any other Nostratic family (Kerns, J.C.,1985, 149-154).Elsewhere, I have suggested that the IE language was first introduced into theBalkans around 6500 BC by immigrant farmers from North of the Black Sea who had beenacculturated to this mode of life by Northward influences from the Fertile Crescent (Kerns,J.C., 1985, 1988), for it seemed reasonable in view of the fact that I n d o - E u r o p e ~ shows

9

---- -----------------


14/64

clear pronominal, morphological and structural similarities to other linguistic families to theEast -- Uralic, Altaic, Eskimo-Aleut -- the whole comprising a part of a larger group offamilies called Nostratic, of Mesolithic origin, which also includes Afroasiatic,Elamo-Dravidian and certain Caucasian languages. However, Renfrew rejects such a sourcefor the Indo-Europeans, preferring to identify them with farmers in Western Anatolia whoproceeded to extend their economy and language into Greece, then into the Balkans andthe Central Region in successive stages. But i f this were so, the Greek language wouldnecessarily show characteristics intermediate between Anatolian and the rest of IE, yet inno way can Greek be regarded as such an early bridge. On the contrary, Greek representsa late Southward intrusion into its historically known position, like Phrygian and Armenian,as the linguistic phenomena clearly imply. Accordingly, I disagree with Renfrew on thispoint. Renfrew's suggestion is a backward projection of his Wave of Advance Model toinclude Greece and Western Anatolia as well as the Balkans and the Central Region. Ingeneral, this might or might not be valid, depending on circumstances. In this case, I thinkit is not. For in the Central Region the advance was aided by the riparian pattern of thesettlements, which initially developed mainly along the rivers and their tributaries since itwas here in these relatively lower elevations that the easily tilled and fertile loess soils hadbeen deposited. This tended to channel the advance into some fractal dimension betweenone and two (disregarding variations in elevation), facilitating its general Northwardmovement in several directions right across the continental divide and along new rivers onthe other side - and in fact, this Linear Pottery deployment was the most rapid of all(Milisauskas, 1978; Kerns, J.C., 1985, 158-161). In Greece and Southern Bulgaria, however,this channeling effect was probably attenuated by the nature of the terrain with itsmountainous character and transversely flowing streams. Thus the acculturation of thevigorous Mesolithic population on the lower Danube to an agricultural way of life need nothave extended to an adoption of the new language, suggesting that their own language mayalready have been an early version of Indo-European before the farmers arrived -- in fact,I have recently come to favor this solution. I believe Renfrew himself has suggested this asa possible alternative.

It should be noted that there is no compelling reason for assuming that the firstIndo-Europeans in the Balkans are to be identified with the first agriculturists there, anymore than to assume that they were horse-riding pastoralists led by aggressive chiefs, whichhas long been the dominating assumption; as Renfrew and I have shown, there are botharcheological and linguistic reasons for rejecting that assumption.Yet in a similar manner, we could be wrong in our own assumption. True, ourassumption has seemed plausible in view of the rapid spread of agriculture and minorpastoralism in the SKK culture and especially in its Northern extension in the Linear Potteryexpansion, thus accounting for the unitary character of the IE-Proper division of IE despitethe disparate appearance of its branches in historical times. In the latter case, I have calledsuch a phenomenon a "focal" expansion, wherein the advance begins in a restricted region(here the Koeroes area in Hungary) and proceeds so swiftly that it attains its final extent inthe course of a few centuries, a time sufficiently short that the original undivided languagehas had little occasion for change. _

10


15/64

Thus the proposition that the first Indo-Europeans in the Balkans are to be identifiedwith the first agriculturists in that region is indeed plausible - but only up to a point. Thedifficulty arises when we try to identify that region, adjacent to Europe, whence they came.Since the dominating plants and animals of the Neolithic cultures in Europe are not nativeto the region, it is natural to assume that they were brought in by immigrants from outside,and Anatolia is indeed the most plausible source. But Renfrew goes on to assume that theundivided Indo-European language was also introduced by these people. I disagree with thislatter assumption since it implies consequences which conflict with the linguistic facts.But my own assumption, that agriculture and the Indo-Europeans entered theBalkans together from North of the Black Sea, seems equally unsatisfactory to me at thistime. While it is plausible from a linguistic point of view, there seems to be little evidencefor agriculture North of the Black Sea at a date sufficiently early, the seventh millenniumBC. Milisauskas (1978) does show theDniester-Bug culture in the proper position in his Fig.45, but in his Fig. 4.3 he dates its beginning at about 5600 BC (calibrated radio carbon), toolate to serve as a source for the SKK culture (begins about 6300 BC in Bulgaria). True, itis possible that the early Neolithic in the Ukraine simply has not been adequatelyinvestigated and that earlier dates may yet be found there, but it is unwise to put much faithin this.

The alternative, then, is to drop the suggestion that the first Indo-Europeans in theBalkans are to be identified with the first agriculturists there. Thus we are left with theassumption that the pre-agricultural Mesolithic people in the Balkans were already speakingearly versions of Indo-European at the time the first agriculturists arrived, and that theysimply became acculturated to the new technology and economy without adopting thelanguage of the immigrants - except, of course, to borrow new terms relating to the newway of life. This is the only suggestion I can think of that does not imply contradictions withrespect to observation. Indeed, it harmonizes well with the Nostratic origin of theIndo-Europeans in the wide-ranging Mesolithic culture.Lately there has appeared an article in Scientific American magazine (March 1990)by Gamkrelidze and Ivanov (here G&I) which presents a prehistory of early Indo-Europeanwhich is radically different from previous suggestions. On page 111 they present a "familytree" which appears correct in some details, but grossly wrong in others since it impliesanachronisms. But their most startling presentation is their figure on page 112. Here theyidentify the original homeland of IE with a region immediately South of the Caucasus and

East of Anatolia (largely present-day Armenia), and they postulate a Westward extensionthrough Anatolia to account for the presence of Anatolian languages there, and a furtherextension across the Hellespont into Eastern Macedonia to account for the Greeks in theirhistorically known positions. In this, they seem to have influenced Renfrew since Renfrewrefers to G&I in this connection, and they themselves refer to Renfrew with approbation.But as I have said, the suggestion is untenable from a linguistic and chronological point ofview. But the most egregious suggestion of G&l is to derive the remaining branches of IEfrom a postulated Southeastward migration from Armenia, curving in a counterclockwisedirection around the Caspian Sea and Westward through Southern Russia into the EasternBalkans, successively emitting Iranian into Iran, Indic into the Indus Valley, and T_ocharian

11


16/64

into Central Asia, and presumably early versions of Slavic, Baltic, Germanic, Italic andCeltic into Europe via the Eastern Balkans.This suggestion is, of course, completely wrong. It implies many anachronisms andother contradictions with respect to observable linguistic facts. It is even inconsistent withtheir family tree , which itself implies anachronisms.I cannot go into details; it would require a book in itself. It would be well to severall connections with G&l, for it is clear that they have not adequately thought the problemthrough. On the other hand, I have no quarrel with their reconstruction of the IEconsonantal system.

I admire Renfrew deeply. More than most scholars, he thinks. He is one of the fewscholars who have made a serious effort to master two disparate yet inevitably associateddisciplines: European prehistoric archeology, and Indo-European comparative linguistics.His only lack is an inadequate appreciation of the seminal importance of the Central Regionas the primary fountainhead of post-IE expansions. Unfortunately, this fault seems to bewidely shared among scholars.

For many years I have hoped that other scholars would do what Renfrew has done.

Reviewer's address:John C. Kerns6580 Chambersburg RoadHuber Heights, OH 45424U.SA

REFERENCESBosch-Gimpera, Pedro. 1960. El Problema Indoeuropeo. Mexico: Direccion General dePublicaciones.Burrow, Thomas. 1959. The Sanslait Language. London: Faber & Faber.Childe, V. Gordon. 1926. The Aryans, a Study of Indo-European Origins. London: KeganPaul, Trench & Trubner.-----. 1957. The Dawn ofEuropean Civilization. 6th edition. London: Routledge & Kegan

PaulCollinder, Bjorn. 1977. Fenno-Ugric Vocabulary. 2nd Edition. Stockholm: Almqvist &Wiksell.Gamkrelidze, Thomas V., & Ivanov, V. V. 1990. ''The Early History of Indo-EuropeanLanguages". Scientific American, Vol. 262, No.3.Gimbutas, Marija. 1979. "The three waves of the Kurgan people into Old Europe".Archives suisses d'anthropologie generate 43.Kerns, John C. 1985. Indo-European Prehistory. Cambridge: Heffers & Sons.. 1988. "PIE Archeology & Linguistics". Diachronica, Vol. V, Nos. 1/2:181-205.Amsterdam/Philadelphia: John Benjamins. --.

12


17/64

Kossinna, Gustav. 1902. "Die indogermanische Frage archaologisch beantwortet".Zeitschrift fi1r Ethnologie 34, 161-222, (reprinted in A Scherer. 1968. Die Urheimatder Indogermanen. Darmstadt: Wissenschaftliche Buchgesellschaft 25-109).Lane, G. S. circa 1967-70. Studies in Historical Linguistics in Honor of George ShermanLane. (= University of North Carolina Studies in the Germanic Languages andLiterature 58.) Chapel Hill: University of North Carolina Press.Lehmann, R. P. M. & W. P. 1975. An Introduction to Old Irish. New York: The ModemLanguage Association of America.Mann, S. E. 1963. Armenian and Indo-European. London: Luzac & Co.Milisauskas, Sarunas. 1978. European Prehistory. New York: Academic Press.Schmidt, Johannes. 1872. Die Verwandtschaftsverhliltnisse der indogermanischen Sprachen.Weimar: Bhlau.Schrader, Otto. 1890. Prehistoric Antiquities of he Aryan Peoples. New York: Scribner &Welford.Tovar, Antonio. 1977. Krahes alteuropliischen Hydronymie und die west indogermanischenSprachen. Heidelberg: Carl Winter Universitatsverlag.

13


18/64


19/64

OOMI1ENTB "t!o the R E V I S ~ D VERSION o f 1'1URTONEN' s COMMENTS (see MT 7, 8]DETAILED REI'1A.RK...: 3. M' s long passage is to ta l ly irrelevant. IS's reconstruction is based-on Dolg's comparison published in Jazyki A.friki 1966;additional- cognates - in Dkl, p.62-3 (see VS in MTW). Msimply ignoresthese data.M has misunderstood IS 's table (p.l53 0? his dict.):"CC" (andnot "CC-") does not mean that there are no cases where an additional element,; appears after CaC-in AA.s or Ns; M could easily see his mistake bylooking through the l i s t of IS 's roots in the diet . _ 6. No point of discussion: see VS in MT7, p.l5 ( 'bl ind 1 = secondary ) S. Original identity: Akk. ? < *9- in th is root (as in Hebr. b:Q.r < *b:Q.r< AAs (*bl}r) < * b a ~ (Dkl,p. 2 ) ~ IS considers q secondary; Tu has a pet-r i f ied prefix (Dkl: Tu a-ber


20/64

242. Wrong; Kartv.connection i s very much valid, as corroborated by Dolg.who now reconstructs Ns * ~ ( ? ) ~ V 'recognize,know' > AfAs:Sem. * ~ ? n 'know'(as in Arab. ~ ? n ) , ~ y ' see ' (Soq. e t c . ~ y L i d : ( y = j ] ) : see ReconstructingLanguages and Cul;tures, Bochum, Brockmeyer i989, p. 95 no.24;cf.VS in MT7.447 Correctness of IS 's reconstr. recently confirmed by Orel and Stolbova rvJa 1988,no. 5,p. 70): WChad. *l!' ayl- ''Jump' : -Ns *C'elV ~ f o r ~ : ! _ , s e e 10).50. Old Hebrew had a meaning ' l iquid ' as well , not jus t ' (grape) juice ' ;original meaning was ' m o i s t ~ a n d ' ro t ten wash' or the like;bothtypes held.51. L doubt that 'smoke' belongs here; as for 'smelr, there are a t leas ttwo root variants (one with an affr ica te , another one with a sibi lant : thisis valid for a l l language phyla, by the way). - As I said: don't ascribeto IS bl\mdemhe never committed: A.Faber (MT?,p.20) seems to think thatIS was so bad that he would combine, in one root, 'smell/odor' .and 'ur ine ' !56. A good set.-Contrary toM, IE central meaning is ' look af te r ' .22 Correctness of IS 's Ns and Sem.-Ham. reconstr. is excellently confirmed by S t o l ~ o v a (same book as Dk3, p.87): WChad. ~ a ~ ' to taste ( t r . )as in Hausa c'ame, Angas ~ a m etc. [for phonetics, cf. WCh. * ~ ' a m , ~ a n' think ' as in Hausa c'amma.ni, Angas ~ a n etc. (Angas = 'remind')].(WCh.a?)59 and 60. I don' t see any faul ts in IS 's reconstr. of b o t h ' ~ r a m m a t i c a l t tworas desPite despe.rate attempts to discredi t i t . B y , ~ t h e r ; t a y , 11' s frequently used argument that connections between phyla (families) are "hardly l i kely" because of geograph. distances ia.>absolutely i r relevant as long aswe speak about genetic relat ionship betw. very ancient languages - andth is is exactly what we do.61. See 59 and 60. !2 real difference in meaning (cf . ~ ; to n e a r ; ~ )~ . A l l wrong. Meaning ' cover is present in Sem. (see D ~ , p . ? ? ) ; i t iseasy to show that -n , -1 , -d are augments: to IS'' s example (Cush. :Beja dem-I i i l 'press ' ) Dk's .may be added now: Eg. s-dm; Ch.:Hausa damo- (id.,p.?6-?.);instead of checking IS 's phon. table on p.l51 and learn that IE - n ~ h - , U r .-rJ- f i t exactly AfAs (and Sem.) -m- [a rule confirmed by many setsj,M sais : "The IE Ural velar nasal does not ta l ly well with the SH lab.one".66. I f th is or that scholar does not have IS 's roots in hisjher rootl i i t , a p ~ a r e . n t l y , h e / s h e has not f o u n d t h ~ y e t . - As for l i s t ing Gidar d!deunder md(m), - i t seems,a whole lo t of imagination i s needed for i t .22 Again,a gross injust ice to IS. 'Be s i lent , si lence' is well represented ~ ~ 1 3 l a n ~ a g e ~ , i . e . , Alt, Kartv. and AfAs both Sem. and Cush

\ a l l this confirmedn Dk2 (p.71-2) with a d d i t ~ o n a l examples.- nd aga n, instead checking IS 's table (where i t is correctly stated that Alt.-rJ- faswell as Ur.-rJ- and IE -ngh-; see 63] f i t s Kartv.and AfAs -m-), M irn..p-l,.e.-sJS 1s phonetic negligence. Some crit icism ![M simply ignored VS in f l T 7 : 6 3 ,~ ~ e e _ 5 9 and 60 above.// 21 ~ u i t e valid! (cf.VS in M T 7 , p ~ l ~ , n 9 . 7 5 ~ .2. See VS in MT7. I stress: a l l sounds in Oromo du? find correspondences both in ~ AfAs and IE roots. Not only Oromo du? means 'd ie '(see D k 2 , p . 8 7 ) ~ M ignored,in my remark,points of a real s c h o l a r l ~ in te rest ; he extracted only elements which he could (as he thought) useagainst ~ S - F c : : _ ~ l h s _ ~ d u _ ' - , +tf.t }_ (C;,,,c_u$/..J Cc>ttflrm t connect/0 ,_,_ "'dwe/1-1- ('? ..... H). Bythe way, IE haR other meanings (besides ' d ie ' ) comparable with AfAs.81. Since Cush. *qd is secondary to *gd,IS, naturally, ci ted (undert?""J Sem. forms of the type qdl. (Extra-Ns sets include *gVd- 'back pt!)As for -1 as a posRible augment, see Dk (AfAs a u g m ~ n t -1, - r , -1 / r ) .82. Both Chad. and Cush have gir(-) ' f i r e ' and 'day' (IS; add Orel,StOlbova, Position of Cushitic,forthcoming in Protolanguages and Protocultures, Bochum).Hence AfAs 'daylight ' etc. is not secondary.8?. I t i s beyond my understanding why should one-rnvent implausibleexP!anations when we already have a quite plausible one.(See VS in MT7).

~ ~ . Why i s AfAs g(w)qr an "ar t i f ic ia l 11 combination of two roots i fM himself acknowledges them being cognates? Since we have forms l ike garin Cush we should reconstruct AfAs guQr which is the same as IS 's *gwQr90. Ns *gurHa 'antelope' > AfAs gurH id . i s f ine;cf .also 59 and 60.~ t J A>f"-.V ~ l t ' - "'(Chcul. *d/ J_ ~ ) i e ' ; "dul- tiz fC-Us/-.; (;atea.-SomaC (or,/ Sfol.t',1Po5/t;o;. c f C u ~ f , , )


21/64

3A confirmation(EGusb.:Sidama ~ - u m - , WGhad. gur-): O r e l - S t o l b ~ o n Gush.9l.Sem. ~ o r ( less l ikely) *gr; WGh.and Gush.*gwr show AfAs gwr.92_ .See IS s explanation of al ternat ionsat the end of this entry. I t i s dangerous to explain them by primordial non-differntiation of stOP.S.94.AfAs *gwl (gul) i s shown by Sem.and Ghad. Hitt . bas-, and unstable "hand "' ').Dolg.eliminates Alt.128. See my remark in ~ ? . Note also that negation belongs to the stablest units , which means tha t jus t one cons. ( l ike 1 here) i s enough to ide129.Wrong conception expressed several times by-M.Reconstruction *?t fSem. (A.Faber, MT?,p.20) seems correct. A very stable root (cf. 128); . l ~ f130.If one reeds attentively,one understands that the archaic meani gmer-ts preserved in Hittite-Luwian (=Anat.), the most archaic IE languagecf. also M.Kaiser, Lexical ArchAisms in Slavic (forthcoming in Bochum).132.Dolg.now reconstructs Ns *?(a][o] ' se t t le , stay, be' >AfAs *?is-'be, exis t ' > Sem. [ ? ] i ~ - ' h a v e ' , ' e x i s t ' , * ' y - i ~ u ' he / i t has' > ' the re i s?IQay ' there i s ' ; Berb.*-h1suH ' a r r i v e ' ; E G u s h . * ? V s / ~ 'be,stay'>'spend d a134.Profound misunderstandingby M ; see also 121 above.Dolg.close to I!!?.A good Ns root; confirmed by Militarev (AfAs ?ayVl) lnd Dolg.r;b.Dolg.n6w con.firms ISs reconstr-.-:-cush. "'?it- ' ea t ' etc.( Sem. ~ a l - 'heigEtT,ql(a)y 'ascend, go up';Berb.*-Hliy 'climb,ascend';Gush. ~ a l - 'moutain, highland' ( IS 's and Dolg. ' s i.nterpret . ' to cross . . 'was incor;re.t).138.IS's reconstr. i s now supported by Dolg. (A!As q1b > Sem. qtiJbb->> qubb- Chad. [wj?]uHb- etc. [ e ~ - g . , in Lamang w u ~ a ; l]AfAs,Alt), qUbE {>Krtv.) ["Blalek's rule"].Cf. VS in MT?.139.In a few papers I proposed 2 Ns roots: 'water ' with -a - and 'drink'with -e- ; now Dolg. reconstructs Ns q;)aK'u 'water ' (> AfAs q(a]k'W- >Sem. *?k 'k ' , (?)*qk'w/*qwk'1 Gush. q a k ' W ~ [=GGush.] etc.) and q[E]gU 'd140.M's argument for 140 L'burn (offer ing) ' ] as being related to 137 isas---r'Ollows: "a burnt offering naturally goes up in smoke";this might be amajor breakthrough in etymol. m e t h o d o l o ~ but I 'm s t i l l not buying it.NowDolg. provides additional data (Ns q(aJlV > AfAs:Sem.*qlw; Ural. *alV ' tosacrif ice ' + Alt; but Gush.is excluded: different roots),eliminating "?".14l . I have x (uvular), not x (velar) in my note.-Since Hitt . has h- 1 ~l}ast- (usually compared with 3 other IE 1-ges) I fee l i n t r i ~ e r l . : Nostr. ha stable"laryngeal", - exac t ly the one which i s preserved in[IE and] Ht.a

* ~ o e g . [wi{hoat re.f-rehciJ is { ; r - o ~ ft;s. (ifes: pfeo..se d.on'-1. use /n.. prt"n-t.


22/64


23/64

5205. Again we have k' /k in AfAs (cf. 15?, 180, 19?); Starostin comparesNs--xutv ' small ' (Cf.same root in Amerind) with SC: N o r t h - C a u c a s . * k ~ t ' V' short ' [ inNs and SC, no.?5]; I wonder i f we have here one of AfAs:SC i s o . .forms since both languages have k- (St. considers AfAs,Ns,SC "sis ters") .208. does not say that -y- was in AfAs (his "'ql-);he explains why.~ . I wonder how 1boil ' and ' sunrise ' can evolve into i ' r i se ' ; i tseems quite clear to me that i t is the other way around (c t . ' top ' , 'be high2l l . I fa i l to see how acquire' can be primary,and 'create ' secondary;anI agree with IS, that the original meaning was 'give bir th ' (same in IE,Dr.212.Evidence of -r / -n being a secondary element in *kpr,*kpn is precisely in the fact that they i n t e r c ~ a n g e . N o t e also k/k' (c t . 180,205 etc.).Sof la t ly disagree. that Sem.should be dropped, - but even i f we were l e f t witBerb. and Chad. only, we s t i l l would have a fine Ns set .215.AfAs seems tenable; validi ty of AfAs meaning seems to be indirectlycon?irmed by archaic Drav. ' s inge' (present also in Uralic). - I fa i l tosee how ' twis t , turn ' i s compatible with IS 's material.216 2 Sets are much stronger than M presents them (accepting them).might be a misprint ( text shows that IS meant qa,f) . - I have afe g that M tends to mix this root and 1 9 6 . - ~ f A s -a.- ln ' W M . ~ k b . $ i C u r J . f k l a . i j ~ o n222.I f la t ly disagree with the idea abot *kp(p) 'bend' being related: thla t ter belongs to 92. For qjk, see 15?,180,196,19?,205,212.AfAs of 222 f i twell Ural. *kappa 'paw' and IE iepH- 'hoof' (as usually, palat . i of IE idue to Ns t ront V). Seems to belong to 190; might be descriptive ~ e x t r a ~224, 229.For "geograph. remoteness see 59 and 60 above.~ . IE shows an ancient heterocl. i e r - u / l e r - n - ~ e m . -n- > I E ? .~ - - ~ , -d are augments as in many other cases. ~~ . An excellent reconstr.Cf. 150,151.-Ex-Ns: SC, Amerind,etc.~ . T h e passage about ' s tay ' and ' res t ' i s one of M's captious o b j e c t i o238. Only in part . // 239: Wrong ( ' ca l l ' belongs to a different root).2.41.TAls re.am.st-r-. isc.o':f,.;;:;;;r(V.J-a..'M_,t13). Not ev_gry "'Hitt. ,Luw. i j ) , *h (> zero), ? (stocf.,e.g.,VS) On Lar!dgeals (in:Die Laryngaltheorie, H d b ~ C . Y i n t e r , ' 8 8 ) ; vowe*a,*e,o, unstresse *i,*u (stressed > *ai, au or the l ike) ; this mAlikelyrt:;:;:O"i+)


24/64


25/64


26/64

THE RUSSIANS ARE COMING! (to Ann Arbor] r. 0 1 .em rer:.-r: ...)Starostin, Helimsky and Mudrak are coming to the U ~ ~tbS (we await them in mid-Febr.) Those i n t e r e s t e ~ in their lecturesetc . , -p lease contact U. of Mich., Slav. Dept, A MI 48109-12?5 :Ben Stolz, Kathrine Rowenchuk,or V.S.; te l . 313-?64-5355 (note my ad-dress from Apr.l through Jul.31: cjo Prof. Y.Koch, Ensl.Seminar, RuhrUniversitat Bochum, .D-4630 B o c h ~ ~ ~ ~ t 1, \/.Germany. - v.s. JSergei Starost in reconstructed several liC [=North Caucas. ] daughterlanguages; (witn:S:Nikolaev) N\lestC, IfEastC, and proto-NO; he wrote anetymol. diet . of NC. He made a major reconstructive work on Yeniseian;he p11blished (in '89) a major book on Old Chinese phonology. He and I .Peiros re-reconstructed ST (= Sino-Tibetan) and wrote an etymol. diet .of ST. In '88, in Ann Arbor, Starostin compared Ns (= Nostratic) andSO (=Sino-Caucasian); t}lis l a t te r was reconstructed by Starostin andNikolaev on the basis of NO, STand Yeniseian. Starostin wrote.an innovative work on glottochronology (to appear in '90). His book on Altaicand the origin of Japanese i s forthcoming in Moscow 1 + Hist. Plion.efi'cs t4 Nc.Eugene Helimsky is a leading expert in Uralic (his book on t h i ~ subjec t appeared in '82). Among his many papers there are three where Helirosky shows, with immaculate logic, why ' ~ t i - A l t a i c i s t s " are wrong inthe i r f ight against Nostratic.Oleg Mudrak i s a br i l l ian t scholar of the youngest generation, specia l i s t in A1taic and other "Eastern" languages, as well as in writing systems. His reconstruction of Eskimo-Aleutian appeared in ReconstructingLanguages and Cultures (Bochum, Brockmeyer 1989); i t was followed by hisreconstruction of ChUkcbi-Kamchatkan ("Kamchukchee") in Explorations inLanguage Macro-Families ( ibid. ) . **-k"BOC1IDI1 BOOKS" in 19 0: Our second book appeared in Dec.' 89 (Explora-t ions ; see a ove , ut the third one is somewhat la te : i t is expected now a t the end of March: P r o t o - L a n ~ a ~ e s and Proto-Cultures (pa:&ersby Greenberg [on IE vowels and E U r a s i a ~ c , Menges [Altaic], Tyler LUraland Drav.], Bengtson [Sino-Caucas.], Oreland Starostin [Etruscan], Militarev [ A f r o ~ A s i a t i c , or Afrasian]; Kaiser 's translation of Ill ich-Svitycheadings from a l l 3 issues of his Nostr. die t . , with a Semantic Index(English toNs) and Dybo's phonetic tables); I t r ied to show, in an intr

d u c t ~ o n , that Salishan (and, apparently, other Almosan-Keresiouan)languages belong to SC, and not to Amerind (Salishan shows a l l stablest elements: pron. ' I ' , ' t h o ~ ' ; numer. ' 2 ' , ' 3 ' ; terms for body parts, etc . asclearly belonging to so; in part icular , being similar to the archaicNC; ,Salishan - SC sound correspondences are very precise).+ Pe.iros OIL 5. E.AthOur 4th book, Sino-Caucasian L a n g u a ~ e s , i s expected in April; i t includes long papers on SC b1 Staros t in , iko laev and Bengtson, as well asa long l i s t of (some 2,000) NC roots reconstructed by St. and Nik.OUr 5th book (tentative t i t le ;Prehis tor of Cultures and Lan a eswill have-more papers (Bla!ek's on ar ., an us ra . , re s an to-bova's on Afro-As., Peiros 's on Austric etc.) as well as Jim Parkinson'sindex to a l l 5 volumes. All 5 books carry materials from our '88 SympoX D O f i J U D I D ~ ~ sium on L a n ~ a ~ e and Prehis tor tO t h e r ~ ~ r , Lexical A r ~ a 1 s m s in Slavic: om Ns tCommon Slavic; a collection of papers on glottochronology; a-collectionon global l inguis t ic connections,-and a few more.

-/tc -Jr *Many are aware, I am sure, of important papers on brode comparison and e ~ p reconstruction (including Ns) by VKclav B L A ~ E K (Leningradska 342,Ff1bram VII, 26102 Czechoslovakia).Blalek i s an excellent scholar, but hhas no time to study (he is a school teacher, busy a ll days in school).wrote le t te rs to President Havel; Ac.of Sc.;Univ., urging to provide Bla!ek with a ILore appropriate job,-and I urge you to do the same. -V.S.


27/64


28/64

(1982), should allow for a clearer understanding of Proto-Korean, for now an obscure andmisused term.Nostraticists like Bombard and Hopper need no longer be dissuaded by the ProtoJapanese-Korean reconstructions of Martin (1966). These have been reworked by John B.Whitman (1985), who presents a smaller but more systematic set of cognates andcorrespondences with equal attention to vocalic and consonantal segments. It is hoped thatthis Ph.D. thesis will soon be published in order to provide a more solid foundation for thecomparison of Proto-Japanese-Korean forms with reconstructions from other languages andphyla, e.g., Prato-Nostratic *prJaHw-f*prJeHw- 'fire, flame, spark' (Bombard 1989:27, no.40), Proto-Japanese-Korean *plir 'fire' (Whitman 1985). This comparison harks back toEckardt's comparison of Homeric Greek 1Ciip 'fire' and Korean pulfpur id. (1966). HisIndo-European-Korean hypothesis, like Koppelmann's (1933b) comparison of Korean withIndo-European, Sumerian, Gilyak, and Ainu, has met only ridicule among academic circlesin Korea and the West In view of Bombard's promising work (1989), we owe it to ourselvesto recheck these earlier attempts.In examining the interrelationships of Korean, Japanese, and Ainu on the one hand,and the Paleosiberian languages on the other, there exists a problem of a "common Siberiansubstratum" (a phrase used by Janhunen 1977:128). Street (1983:198) remarks with greatinsight that: "I suspect that by comparing the lexicon of Gilyak, for example, with those ofKorean, Japanese, and Ainu, we might detect early loanwords (possibly involving chains ofborrowing) which would eventually supply us with formal criteria for distinguishing betweenborrowed and inherited lexical items. As matters stand now, the lack of such criteria formsa major stumbling block for comparisons of Ainu with Japanese, Korean, or Altaic." Hecites (1983:203, fn. 15) Proto-Esquimo *qajuva- 'ladle', Ainu kasu(p) id., Old Japanesekasipa 'container for liquids', Proto-Altaic *kazlbuga 'spoon' as being quite probably"involved in some chain of borrowings." Professor Street writes to me in a letter dated11/15/89 that this is one of the reasons why he no longer deals with Japanese-Altaicconnections.2

2After reading a draft of this paper, Street (12/13/89): "Poppe (1960:13), like Ligeti -whom you cite p. 3 end of first full paragraph -- apparently disliked Ramstedt's connectionof tuksald with the rabbit word."Though I can' t remember Miller's stating this in print, it seems to me that his ideaabout the prefix came from (a) Ramstedt's ' rabbit/ jump' connection, and (b) occurrence ofthe OJ pair Fasir-u and ta-Fasir-u 'run'. (Unger had ta- as a 'prefix' here.)''The former OJ form is pA *pelyi- (Street 1981:646; where I didn't dare mentionAinu pas- - cas- (Patrie 1982:70)). For a while I toyed with setting up a pA *tablyV- 'run,jump, flee' ?? < *ta- + pelyi-; cf. Mo. tauli- - tayuli- 'chase, pursue', Chag. tavu.S- 'run,jump' (TMEN 2.616)."My favorite rabbit form is Kirakos thalpqa (Pritsak Fiirstenliste 56-7)."I once thought that Korean and Japanese might have incorrectly cut off what theythought was a ta prefix. I.e. Koguryo *wusiga < *(ta)blyi-gan, OJ *wusagi < *(ta)blyi-gay.''The whole thing is a mess!!" -

2

---------------


29/64


30/64

'to besmear with mud'. Nonetheless, Menges (1975:40-43) believes that the Proto-Altaic,Tungusic, Korean, and Japanese words for 'hare, rabbit' are cognate, and perhaps furtherrelated to Mayan tal id. and Tamil cevijan id. He also adds words of unknown originbelonging to the Siberian Russian dialects: u!kan 'Hase' (Arkhangelsk, Sibiren, OrenburgChkalov), ulan id. (Pskov). Without being in a position to comment on the significance ofthe Mayan and Tamil forms, I should point out that a consideration of the Gilyak and Ainudata is missing in Menges' comparisons. The Siberian Russian forms are undoubtedlyborrowed from Gilyak or another Paleosiberian language. Yet, we can say nothing definiteabout the relationship, genetic or diffusional, between Ainu osuke(p/x). Old Korean*usikam, and Old Japanese usagi. Whitman (1985) has found that Ainu casi 'stockade' isa loanword from Old Japanese, in turn a loanword from Middle Korean. The subjectmarker /i - / in Korean cas-i 'fortress', is one of the few clues as to the ultimate origin of theword (John B. Whitman, personal communication). I f the resemblance between the Ainu,Korean, and Japanese words for 'hare, rabbit' is a diffusional one, we have for the timebeing nothing by a hypothesis for a possible direction of borrowing which is based on anisolated example, 'fortress'. Here, the comparative method breaks down.

I have presented these examples in a concise manner to illustrate the probleminvolved in comparing Korean (and Japanese and Ainu) with other languages. Moredialectal information will be needed for Korean, as with any other language, to reduce thepossibility of chance resemblance (see 'whale' above). Although the inter-relationships ofKorean, Japanese, and Ainu will be a promising sub-field of research, the existence of acommon Siberian stratum must be carefully studied before jumping to conclusions aboutgenetic relationships. I am not a proponent of ''wholesale borrowing" to explain the Altaictheory, nor do I which to reject outright the Nostratic theory as a conglomerate ofdiffusional and chance resemblances. Instead, I am proposing that the Altaic and NorthAsiatic languages be studied more thoroughly in order to gain more insight into theprehistoric relationships of these and the Paleosiberian language.

REFERENCESBombard, Allan R. 1984. Toward Prato-Nostratic: A New Approach to the ComparisonProto-Indo-European and Proto-Afroasiatic. Amsterdam and Philadelphia: JohnBenjamins.---------- 1989. "Lexical Parallels Between Proto-Indo-European and Other Languages."Supplement to Mother Tongue 9 (November/December 1989).Choi, Hakkun. 1982. Hankwuk-e Pangenhak (Korean Dialectology). Seoul: Tayhaksa.Dettmer, Hans A 1983. Review of Patrie (1982). Monumenta Nipponica XXXVIll.3:331-

33.Eckhardt, Andre. 1966. Koreanisch und Indogermanisch: Untersuchungen iiber dieZugehorigkeit des koreanischen zur indogermanischen Sprachfamilie. Heidelberg:Julius Groos Verlag.Hopper, Paul J. 1984. Foreword to Bombard (1984).Hymes, Dell H. 1959. "Genetic Classification: Retrospect and Prospect". Anthropological4


31/64

Linguistics 1.2:50-66.Janhunen, Juha. 1977. "Samoyed-Altaic Contacts-- Present State of Research". Altaica.MSFOu XIX: 123-230. Helsinki: SUS.Janhunen, Juha and Songmoo Kho. 1982. "Is Korean Related to Tungusic?" Hankul177:179-90.

Kang, Kil Wun. 1982. ''Hankwuk-e wa ui Pikyo" ("A Comparison of Korean and Ainu").Emwun Yenkwu 11:1-90.--------- 1988. Hankwuk-e Kyey Thong Non (Issues on the Origin of the Korean Language).Seoul: Yeng Sel Chwulphansa.Kara, Gyorgy. 1965. "Le dictionnaire etymologique et Ia langue mongole." Acta OrientaliaHungarica XVIII:1-32.Kim, Han-kyo and Hong Kyoo Park, ed. 1980. Studies on Korea: A Scholar's Guide.Honolulu: The University Press of Hawaii.Koppelmann, Heinrich. 1933a. "Die Verwandtschaft des Koreanischen und der AinuSprache mit den indogermanischen Sprachen". Anthropos 23:199-234.------- 1933b. Die eurasische Sprachfamilie: indogermanische, koreanisch undVerwandtes. Heidelberg: Carl Winter.Kreinovich, Evgenij A 1955. "Gilijatsko-Tunguso-Man&rske jazykovie paralleli".Doklady i Soob!cenija Instituta Jazykoznanija 8:135-67.Krippes, Karl A 1989. "Remarks on Korean-Ainu Etymologies". Ural-Altaic Yearbook61:149-51.

-------- 1990. Review of Kang (1988). Diachronica VI:l.Lee, Ki-mun. 1963. "A Genetic View on Japanese". Chosen Gakuho 27.---------- 1964. "Materials on the Koguryo Language". Bulletin of the Korean ResearchCenter no. 20.Lewin, Bruno. 1973. "Japanese and the Language of Koguryo," Papers of the CIC FarEastern Language Institute (Ann Arbor) 4:19-33.

Martin, Samuel E. 1966. "Lexical Evidence Relating Korean to Japanese." Language42/2.185-251.Menges, Karl. 1975. Altajische Studien II: Japanische und Altajische. AKM XLI.#.Wiesbaden: Karl Steiner.Miller, Roy Andrew. 1971. Japanese and the Other Altaic Languages. Chicago and London:The University of Chicago Press.---------- 1983. Book notice on Patrie (1982). Language 59/2.447-48.Murayama, Shichiro. 1961. "Nihongo to Kokurigo to no kankei ni kansuru zentei-hokoku".Jutendo daigaku-taiikugakubu-kiyo no. 4.---------- 1962. "Nihongo oyobi Kokurigo no sushi". Kokugogaku no. 48.Naert, Pierre. 1962. "Contacts lexicaux aYnoux-gilyak". Orbis 11:199-229.

Paniflov, V. Z. 1973. "Nivkhsko-altajskie Jazykovie Svjazi." Voprosy Jazykoznanija 5:3-13.Patrie, James. 1982. The Genetic Relationship of the Ainu Language. Honolulu: TheUniversity Press of Hawaii.Poppe, Nicholas, 1960. Vergleichende Grammatik der altiiischen Sprachim. L VergleichendeLautlehre. Wiesbaden: Otto Harrassowitz.---------- 1972. "A New Symposium on the Altaic Theory". Central Asia tic Joumal-16:37-58.5

- - - - - - - ~ - ~ - - - - - - - ~ - - - ~ - ~ - - ~ - - - - - ~


32/64

Ramsey, S. Robert. 1978. Accent and Morphology in Korean Dialects. Seoul: ThapChwulphansa.Ramstedt, G. J. 1949. Studies in Korean Etymology. MSFOu XCV. Helsinki: SuomalisU grilainen Seura.---------- 1957. Einft1hrung in die altiiische Sprachwissenschaft. L Lautlehre. MSFOu 104.Helsinki: SUS.Rozycki, William Vincent. 1983. Mongol Elements in Manchu. Ph.D. thesis. Departmentof Uratic and Altaic Studies. Bloomington: Indiana University.Sameev, G. D. 1930. "Man'anro-mongolskie jazykovie paralleli." Izv. Akademij NaukSSSR 8.601-26, 9:673-708.Street, John. 1962. Review of Poppe (1960). Language 38/1:92-99.---------- 1974. On the Lexicon ofProto-Altaic: A Partial Index ofReconstructions. Madison:Wisconsin: The Author.---------- 1983. Reivew of Patrie (1982). Journal of the Association of Teachers ofJapanese17:192-204.Tsintsius, V. I. et al., ed. 1974-77. Sravnitel'nyj slovar' Tunguso-Man'cfurskix jazykov. 2vols. Leningrad: Nauka.Whitman, John Bradford. 1985. A Phonological Approach to the Comparison ofJapaneseand Korean. Ph.D. thesis. Department of Linguistics. Cambridge, MA: HarvardUniversity.

6


33/64

February 1990

AN ENDTHE

TO SPLENDID ISOLATION:MACRO-CAUCASIAN PHYLUM

by John D. BengtsonTo th is day, the bel ief tha t Basque and Burushaski ~ eto ta l ly isola ted languages i s widely held. For example, in thecurrent edition of Encyclopaedia Brit tanica (1983), ProfessorLuis Michelena of Salamanca s ta tes that "Basque remains an i so-la ted language. proof of a genetic relat ion beyond reasonable doubt appears remote." The unsigned art ic le on Burushaskiin the same encyclopedia s ta tes that the la t te r language i s simi la r ly "not known to be related to any other language in the

world."In his Guide to the World's Languages {1987: 377), Merri t tRuhlen s t i l l l i s ted Basque and Burushaski as i sola tes , thoughnow he accepts the i r Dane-Caucasian af f i l i a t ion (Ruhlen 1989).In his review of Ruhlens volume, Harold C. Fleming (1987: 197-198) mentions the Afroasiatic hypothesis of Basque origins , versus the Dane-Caucasian hypothesis, and indicates his favor ofthe l a t t e r , while also including Burushaski in "Vasco-Dene". Inanother review of Ruhlens Guide, Vaclav B l a ~ e k (J989a:16) holdsthat "Burusaski i s probably also re la ted [to Sino-Caucasian =Dena-Caucasian], while the Sino-Caucasian aff i l ia t ion of Basqueand mainly that of Sumerian i s only hypothetical ." So while theDena-Caucasian hypothesis has recently won converts among seriouspaleo-l inguists , the general tone i s s t i l l , understandably, caut ious, implying tha t any genetic t ies to the "Splendid Isola tes"are remote a t best.The present writer i s proposing that Basque and Burushaskiare, in fact , fair ly closely re la ted to one another, and, together with (North) Caucasian, form a phylum-level group (taxonomical ly comparable to Indo-European) which I will here designate as"!-!aero-Caucasian". In th is view, Basque, Caucasian, and Burushaski are family-level groups diachronically comparable to , e.g . ,Albanian, Slavic, and Armenian, while Macro-Caucasian i s subordinate to the macro-phylum Dena-Caucasian(= Sino-Caucasian), andcoordinate vdth the other phyla (Sumerian, Sino-Tibetan, Yeniseian, Na-Dene) making up Dena-Caucasian.The evidence for Macro-Caucasian, of ~ h i c h only a f ract ioni s presented below, consists of numerous lexical isoglosses inthe most basic semantic f ields (pronouns, parts of the body,natural phenomena, basic descriptives, animals and plants) ; in -t imate paral lels have been observed in the inf lect ion of nouns 'pronouns, and verbs. The assessment of this evidence convincesme that the-r-elationship i s a t a phylum level , with a diachronicdepth of no more than six millennia.I f th is proposed Macro-Caucasian phylum i s indeed as closely kni t as i s claimed here, why has it eluded general recognit ion for so long? Obviously, Indo-European i s a very specialcase (Fleming 1987: 160f), and has benefited from several his tor ica l advantages, including a large number of geographically

c o n c e ~ t r a t e d (sub-)families, many with very early records a l lo ~ w h ~ c h has been s ~ b j e c t e d to huge amounts of scholarly ~ t t e n -t ~ o n . M a c r o - C a u c a s ~ a n , on the other hand, consists of three


34/64

widely separated branches, only one of which (Caucasian) has thebenefi t of rather scanty ancient records (Ratt i , Hurrian, Urartean). I am certain that i f we had records of the ancient predecessors of Basque and Burushaski, the general recognition ofMacro-Caucasian would long since have been a M.i accompli.Other difficul t ies had to do with Caucasian specifically.Early Dene-Caucasian researchers, such as Trombetti and Bouda,treated Caucasian as including Kartvelian as well as North Caucasian, and they were bewildered by the wide array of d i ~ e r g e n tforms from both families. Recent studies, particularly by theSoviet Nostratic school, have clearly shown the dist inct originsof Kartvelian (which i s placed in the Nostratic macrophylum) andNorth Caucasian (Sino-Caucasian = Dene-Caucasian). Any commonorigin for the two Caucasian families would then have to besought in the possible kinship of the Nostratic (Eurasiatic) andDena-Caucasian macrophyla. (Fleming 1987: 163!.)Another problem was resolved by the formulation, by SergeiA. Starostin and Sergei L. Nikolaev, of a reconstruction of proto North Caucasian. This, and their forthcoming etymologicaldictionary of North Caucasian, will be of immense benefit to thefurther development of Macro- and Dena-Caucasian studies.In addition, I think a real obstacle to establishing genetickinship for Basque and Burushaski has been thei r legendary statusas "Splendid Isolates". The idea of Basque as a lonely, evennoble, isolated language acquired a l i fe of i t s own in the collect ive mind of mainstream l inguist ics. Even a softening of thisabsolute would lead naturally to the assumption that any l inguis t ic kinship with Basque would have to be very remote. (Cf.Ruh1en 1987: 74-75.) The possibil i ty that these "isolates"might actually belong with one another or with other languagesin a commonphylum, l ike Greek and Celtic , was rarely entertained.The evidence for Macro-Caucasian presented here i s not in tended to be a definitive statement of proof. No doubt some wil lbe convinced, while others will demand more. This presentationi s admittedly fragmentary, but I hope it will be a stimulus tomy colleagues to investigate for themselves, and contribute tothe discussion for or against the hypothesis proposed here.* * * * *The following 77 etymologies represent a fraction of thelexicon common to two or more branches (families) of the MacroCaucasian phylum. I t will be seen that vocabulary i s shared bya l l three branches in some cases, ~ d by pairs of branches (Basque- Caucasian, Basque-Burushaski, Caucasian-Burushaski) in others.I have not found s ta t is t ica l ly signif icant evidence that any ofthese pairs especially close, i . e . , forming a subgroup to theexclusion of the other language.or the 77 etymologies, about 30 appear to be restr ic ted toMacro-Caucasian, while the res t are also shared by the higherlevel grouping, Dena-Caucasian. Even among the la t te r category,the Macro-Caucasian forms often show peculiari t ies of m o r p h ~ l o g yor phonology. Besides about 65 etymologies from the most basicsemantic fields, several etymologies involving domestic animals

- - --------- -- ----------- ------


35/64

would indicate that animal husbandry was well in place before thediaspora of l>iacro-Caucasian speakers. Other shared lexicon pointsto a homeland that was hil ly or mountainous, and a ll Macro-Caucasian peoples, to this day, are inhabitants of high alt i tudes.MACRO-CAUCASIAN PHONOLOGYThe following are some of the regular correspondencesin i t i a l consonants in the Macro-Caucasian phylum:

MC Basque: Caucasian: Burushaski:*ph p(h) *p(h) ph > pf*b b *b (*p) b*th t (h) *t(h) th*d d *d (*t) d*kh h *k(h) kh*g g * g (*k) g*x h *x h*qh k.(h) *q(h) qh*G g *G g/J*X h *X X

~ h ~ J ~ h 1i h*h 0/h *h h*s z *S s*s s ~ s*ch 6 *c(h) ch

~ h tz ~ ( h ) ~ hSOME GRAMMATICAL EVIDENCE

The following are some of the Macro-Caucasian nominalendings:

of

1. Bsq -k (ergative), - ik (ablative, parti t ive) : Cau *-k'V(abl . , par t . , ins t . ) :Bur -A!f-ek ( ins t . ) .2. Bsq - ! (da t ive) : NECau *- i / -z (dat .) : Bur - ! (genit ive).3. Bsq - ra (adit ive) : Cau *-rV ( loc.) : Bur -ar (al lat ive).4. Bsq - ta (locative) : Cau :tv ( la t . ) : Bur -Ate/-Et (adessive). - - __.._ __.._5. Bsq -en (genit ive), -n (inessive) : Cau *-nV (gen.) :Bur -.An!/-ensT-Ane (comitative, inessive). -6. Bsq - ~ ( ins t . ) : Cau *-s(e) ( ins t . ) .7. NECau *-cV ( ins t . , erg.) : Bur -ce/-c i ( ins t . , contactiv:-e).


36/64

MACRO-CAUCASI_U{: SOHE LEXICAL EVIDENCE

BASQUE: (North)CAUCASIAN: BURUSHASKI:

1. n i ' I , me' *ni NSC 1382.34-5-

h izu

thou' you (pl.)

* zo ' I '*Gufwv

thou(Dargwa 1iu)

(Lak zu). -

., ., .,~ I V ,3!_N_!A- NSC 139- - -B_-n"' B_-Drv ~ - m NSC 142~ - , S - , - k u - , - k o - NSC 143

NSC 1446. -haur self ' -kher?.8.910.11.12.13.14.15.16.17.18.19.20.21.22.23.24.

ze-r 'what' *sa (Dargwa se)mihi tongue' *melc 'c ' imuthur snout' *mari'V nose, beak'buru head', (Udi for hair ' ,be-phuru eyelash' Kurin f i r i mane')hertz ' tooth 'bi-hotz 'hear t '

be-sA-n NSC 146m a l ~ 'jaw NSC 1-mUltUr nostr i l NSC 23bur ' ha ir , -1-pur'eyelash' NSC 13

-hAse(-me)molar tooth(m-)os ' (her) heart

poto-rro vulva' *pet ' i hole, vulva' NSC 48s 149bi-zka-r 'back' (Abkhaz a-zkWa) -sga on one's back'NSC 39gal- tzar ' s ide 'saihets ' s ide 'he-gal 1 wing, fin'esku 'hand'u-kab-il ' f i s t ' *ggwa::pv 'paw'hanka thigh' *hianqVhoin""'hun root ' *9-"'in(i)-qwv heel,ankle'

-NSC 41-sorut bosom, sidegAl-gi NSC 32heskN hisk 1 wrist, backof t h e ~ h a n d N s e 29kAf claws, talons'(Blazek 1989b)

u-kondo 'elbowhun,v hun marrow,brain'gore tz ' dung'

*g'WaintVhWen?V

elbow, knee' sc 1.12sc 7blood' ?hAn- blood( in comp.)lurAs25. su 1 f i re ' *c 'a j! (Lak c'u) s i ' f i replace ' , NSC 65

~ u - t i n ' f i r e - s t o n e s ~26. i -zar star l A - ~ h a r morning s tar 'NSC 73


37/64

(Basque)27. hil(a) 'moon'28.29.30.31.32.33.34.35.36.

hauts 'dust 'i -bar valley'hertz 'cloud, sky'in- tz ig-ar ' froste-l.hu-r 'snow'tx i txer / ~ i ~ e r /' ha i l 'e-ur i ' ra in ' ,hur-alde ' f lood'

37. arra ts 'n ight '38 .39. biha-r tomorrow'40. goiz morning'e-&!,!! 'day'41.42.43.

hun"' hon 'good'zahar 1 old 1

(Caucasian) (Burushaski)hAlAns NSC 72

*wgmc'o 'moon'*XurtV 'foam, scum'

hasa"" hisa 'month' S 129

*G11erV ' s tone'*cowgiV ' s l ee t '

* ~ w i u V (Cecen luo)__,. ._ -*c'c'VrV ' ice ,to freeze'*lwer-tV rain '* ' ~ i r i n . i i i*PVsV 'n ight '*PekV 'dawn'*GWemtV 'day'

* h W i n - ~ V*/?I/oxwv 'good'

~ 'dust 'bar1.2roxuronc 'cloud,chAJUr- 'cold 'hio ' ha i l '

her-8.1 t ' ra in '

gunc 'day'gUn 'dawn'

*sW!rHo 'year, old ' char- 'ancient,times'

NSC 77NSC 81s 80fogNSC 86T 221sc 5.7NSC 85NSC 71T 62NSC 69NSC 116of formerBNC 3644.

45.mehe .-thin' * G V - m V ! ' ~ ' V T 235tx ik i / ~ i k i / ' small ' *331/k'k'/V ' short ciki ' small ' T 78

*hwo-c'Worv 'gray, - ~ i r - A g g o 'white-faced'yellow' -cor catt le) NSC 124*ccakkWV 'white, ~ I k - e r k 'yellow'46 . zuri 'white '47 . yellow' NSC 12548 . horitv hoii 'yellow *g' g' Vhwvrv 'gray,brown49. hartz bear' Dagestan *XIWVrcv' squirre l , marten' NSC 9050. * g w a ~ e 'dog, wolf' gA13hu ' jackal ' NSC 8851. hor 'dog' Nakh-Dagestan BNC 72a*xwar 'dog' NSC 87

- - - - - - - - - - ~ - - -52. ~ rat , mouse' *cwargg"'V weasel, charge ' f lyingmarten' (Kabard. squirre l ' NSC 92,faRWa mouse')53. khur-lo 'crane 1 *q q'iri-q'q'WV garu-yo 'wading'crane' bird, heron' NSC 9654. piro 'duck' p h a r i ~ t-.J pferis55. igelcv ugaraxo, *g'g'WVrV-q'q'V lUrkUnf'l 1orkUn(etc.) ' frog ' (Khinalug gurkor) "'gUrgUc NSC 95


38/64

(Basque)56. sahats 1 uillow 157.58.59. hur hazelnut60. intzaur nut61.62. i - tzar 'old ox63. a-bere"' a-bel' ca t t l e ' - -64. zezen bull '65. txahal /cahal /' ca l f , heifer '66. ezne ~ 'milk'67.68. bil-dots lamb'69.70. (Souletin)manexina ewe71. ilhe 'wool'72. zikhiro ' castratedgoat73. urde ' p ig , hog'74. ezt i 'honey'75. ar male'76. a-tso 'old woman'77.

(Caucasian)

*kalV stick,pole, tree*t'VImV 'kernel ofa :fruit '

~ w o r - i ' i ' V nutxiiwse(w)o1.' ap' V ' l eaf '(Adygh tirap)*c(w)arnv cow(Rutul zar)pov

~ v n H V*HinisWu cheese*wil-q'erv*HowohV 'ram;he-goat

* n a w ~ E 'sheep about2 years old*?&lXIV

(Burushaski)sAskgAl-t9r smallbranch'tumatv tumA;?' ' shel lof nut, stoneof f rui t ' NSC 105xUnzUr 'kernel ofwalnut NSC 106tAp NSC 104chior young ox,steer KL 3

chin-dar

T 97hAmsnzbi!il-is rv bel-isewe; sheep overtwo years oldhui5 sheep

m a m u ~ i 'lamb KL 10* c ' c ' ~ k V goat, kid' chigir she-goat'(Andi c 'ek ' i r ) KL 7*warr!''A' e*hwi-mi33ii(Avar hoc'c 'o,Ingush ~ )*Hir-k'wv 'man*33WVjV ' female'(Ratti zzuwa woman,wife 1 )*q(w)anV woman'( D a g e s t ~ n qhVmV)

hir N hir~ - u s " ' ; ; ; us 1 wife 1

KL 17NSC 117

NSC 131 .NSC 135

qUma 'concubine'NSC 134--


39/64

BNCKLNSCsscT

======

Abbreviations:Basque-North-Caucasian: ~ i r i k b a 1985Kui t urnaja leksika: Starostin 1985Notes on Sino-Caucasian: Bengtson 1989Starostin 1989Sino-Caucasian: Starostin 1984Trombetti 1926

REFERENCESBengtson, John D. 1989. "Notes on Sino-Caucasian." (Forthcomingin S h e v o r o s ~ ~ n 1990.)

B l a ~ e k , V ~ c l a v . 1989a. Review of A Guide to the World's LanguageS. Vol. 1 (Ruhlen 1987). Manuscript.19 9b. Notes on Sino-Caucasian, in a le t te r to the-----------author.Cirikba, V.A. 1985. "Baskskij i severokavkazskie jazyki." inDrevnjaja Anatolija, 95-105. Moscow: Nauka.Fleming, Harold c. 1987. "Toward a Definitive Classification ofthe World's Languages." (Review of Ruhlen 1987.)Diachronica IV: 1/2: 159-223.Nikolaev, Sergei L. and Sergei A. Starostin. 1987. "North Caucasian Roots." (Forthcoming in Shevoroshkin 1990.)Ruhlen, Merritt . 1987. A Guide to the World's Languages. Vol. 1:Classification. Stanford: Stanford University Press.1989. "Phylogenetic Relations of Native AmericanLanguages." Presented a t the Symposium on PrehistoricMongoloid Dispersals. Sapporo, Japan, December 1989.Shevoroshkin, Vitaly (Ed.). 1990. Sino-Caucasian Languages.

(Bochum Publications in Evolutionary Cultural Semiot ics , Ed. by Walter A. Koch.) Bochum: StudienverlagDr. Norbert B r o c k ~ e y e r .Starost in, Sergei A. 1984. 11Gipoteza o geneti!eskix svjazjaxsinotibetskix jazykov s enisejskimi i severnokavkazskimi jazykami." in Lin v i s t i ~ e s k a a rekonstrukci a id r e v n e j ~ a j a i s tor i ja Vostoka IV: 19-3 Moscow: Aka- demija Nauk, Ins t i tu t Vostokovedenija. (Translationby William H. Baxter, forthcoming in Shevoroshkin 1990.)1985. "Kul'turnaja leksika v o b ~ ~ e s e v e r o k a v k a z s k o mslovarnom ronde." in Drevnjaja Anatolija,.74-94.Moscow: Nauka.1989. "Nostratic and Sino-Caucasian." in Lingvisti

~ e s k a a rekonstrukci a i __drevne ~ a a i s tor i a VostokaI : 10 -124. Moscow: Akademija Nauk, Ins t i tu t Vostokovedenija.Trombetti, Alfredo. 1926. Le origini della l ingua basca. Bologna:Accademia delle Scienze dell ' Is t i tu to di Bologna.(Annotated version reprinted 1966.)-


40/64

--


41/64

Typology, Relationship, and Time. Edited by Vitalij V. Shevoroshkin and ThomasL. Markey. Ann Arbor, MI: Karoma Publishers, 1986. Pp. xliv, 120.

Reviewed by ALLAN R. BOMHARD, Boston, MassachusettsThis book came into being through a proposal by one of the editors (Markey)that the other editor (Shevoroshkin) "make a selection of what he considered the _!:>est

of recent pro and con Soviet work on or about Nostratic and that the two thentranslate, edit, and preface the resulting collection for publication". What finallyemerged in the collection is mostly a group of articles published in the Soviet Unionover the past two decades discussing the (mostly posthumously published) work ofV. M. llli6-Svityc (HnnHq-cBHTLiq) on Nostratic. To a far lesser extent, the workof A. B. Dolgopol'skij (,[(onronoJILCKHH) is also discussed.

The book begins with a Foreword, the first part of which appears to havebeen written by Markey alone, and the second part by Shevoroshkin alone. In thesecond and longest part, Shevoroshkin begins by giving a brief history of thedevelopment of Illic-Svityc's (and Dolgopol'skifs) ideas on Nostratic and expressesstrong support for the Nostratic Theory in general and for Illic-Svityc's work inparticular. He then makes three proposals of his own: (A) the Proto-Indo-Europeansystem of stops should be reinterpreted as *TH, *T, *D (from Nostratic *T', *T, *D,respectively), (B) Proto-Indo-European had "strong" laryngeals as well as "weak"laryngeals (the so-called "strong" laryngeals survived in Hittite/Luwian, while theso-called "weak" laryngeals were lost), and (C) the laryngeals did not affect thequality (timbre) of contiguous vowels. Let us look more closely at each of tlieseproposals.A. Shevoroshkin's ideas concerning Proto-Indo-European consonantism are not allthat different from the proposals made by Joseph Emonds (1972). Where he

runs into trouble is in trying to derive his revised system from Proto-Nostratic.One would like to know how the glottalized series became voiceless aspiratesin Proto-Indo-European without merging with the plain voiceless stopssomewhere along the way. When one tries to work through various scenariosto arrive at Shevoroshkin's revised Proto-Indo-European system from itsalleged Proto-Nostratic antecedent, one runs into roadblocks at every turn. In


42/64

REVIEW OF SHEVOROSHKIN AND MARKEY (1986) 2other words, you cannot get there from here (1).

B. On the surface, Shevoroshkin's theories concerning "strong" laryngeals and"weak" laryngeals in Proto-Indo-European appear intriguing. The problem isthat the data do not fit the theory (2).

C. In order to be able to judge Shevoroshkin's theories concerning whether or notlaryngeals changed the quality of contiguous vowels, one would have to knowwhat phonetic properties he would assign to the laryngeals he posits. As longas he operates with cover symbols and employs ambiguous terminology, it isnot possible to form an opinion one way or the other about the validity of hisproposals.-Finally, Shevoroshkin bitterly attacks the work of Bombard (1984) in highly

emotional, intemporate language that can only be described as embarrassing. Thediscussion of Bombard's work is characterized by outright misrepresentation. Onegets the impression that Shevoroshkin did not read Bombard's book throughcarefully or that, if he did, he did not understand what he read. Rather than engagein a lengthy rebuttal, the reader is invited to look at Bombard's book forhim/herself. As for the emotional nature of Shevoroshkin's attack on Bombard, wemay quote from Bertrand Russell (1976:116) and let it go at that:

I f an opinion contrary to your own makes you angry, that is a sign that youare subconsciously aware of having no good reason for thinking as you do. Ifsome one maintains that two and two are five, or that Iceland is on theequator, you feel pity rather than anger, unless you know so little aboutarithmetic or geography that his opinion shakes your own contrary conviction.The most savage controversies are those about matters as to which there is nogood evidence either way. Persecution is used in theology, not in arithmetic,because in arithmetic there is knowledge, but in theology there is only opimon.So whenever you find yourself getting angry about a difference of opinion, be

1) We would expect the developments to have been more as follows than asproposed by Shevoroshkin: Proto-Nostratic *T', *T[h), *D > Proto-Indo-European*T, *Th, *D. A typological parallel exists within Semitic, where Proto-Semitic *T',*T[h), *D have developed into *T, *Th, *D in the Neo-Aramaic dialect ofTiir-'Abdin.2) For an excellent survey of the Laryngeal Theory, cf. Lindeman 1987; cf.also Winter 1965.


43/64

REVIEW OF SHEVOROSHKIN AND MARKEY (1986) 3on your guard; you will probably find, on examination, that your belief isgoing beyond what the evidence warrants.

We can now consider, in turn, each paper in the collection:

V. V. Ivanov: "Proto-Languages as Objects of Scientific Description." (1980).This paper is divided into three sections. In the first section ('The Differencebetween a Proto-Language and a Mere System of Correspondences"), Ivanov beginsby outlining the methodology by which a system of correspondences is used toreconstruct a proto-language. He notes that correspondences may be the resuh of

borrowings. Such cases cannot be used to establish genetic relationship but, rather,result from prolonged contact between two or more languages, which may or maynot be otherwise related. Ivanov then considers two examples of correspondencesbetween grammatical systems which cannot be explained by language contact: (A)the similarity between the earliest secondary verbal endings reconstructed forProto-Indo-European and those assumed for Proto-Kartvelian and (B) the similarityof heteroclisis in neute

Mother Tongue Newsletter 10 (April 1990)

Documents

Transcript of Mother Tongue Newsletter 10 (April 1990)