LITUANUS
LITHUANIAN QUARTERLY JOURNAL OF ARTS AND SCIENCES
 
Volume 40, No.1 - Spring 1994
Editor of this issue: Antanas Klimas, University of Rochester
ISSN 0024-5089
Copyright © 1994 LITUANUS Foundation, Inc.
Lituanus

DID THE "INDO-EUROPEANS" COLLIDE WITH "PRE-INDO-EUROPEANS"?

ALFRED BAMMESBERGER 
The Catholic University of Eichstätt

[This essay grew out of preparing a review of the volume Markey 1990. The main aim of what follows consists in outlining some methodological problems connected with research in Indo-European. Emphasis will be laid on the impact Lithuanian material can have in this context.]

It must be emphasized from the very beginning that Indo-European (abbreviated as IE) is in the first instance a linguistic term:

Indo-European is used to refer to the common ancestral language underlying a group of genetically related languages and reconstructed on the basis of those daughter languages. We do not know how many members of the language family ultimately died out without leaving any direct traces. We have documentation of some (Indo-European) languages like Hittite that died long before the birth of Christ. A number of (Indo-European) languages, among them Greek, Armenian, Albanian, Persian, Hindi, as well as English, German, Dutch, Swedish (Germanic), French, Spanish, Italian (Romance), Russian, Polish (Slavic), Irish, Breton, Welsh (Celtic), Lithuanian and Latvian (Baltic), are spoken to the present day. Indo-European means the ancestral language within a family of related languages.

The methods used in reconstructing Indo-European and the problems faced in this process will be outlined later; further details will be assessed in the ensuing sections. The material to be presented below has been taken from the storehouse of Indo-Europeanist lore, Pokorny 1959, which itself is a revision and reworking of the older so-called "Walde-Pokorny" (=Walde 1927-1932). A useful collection of material ordered on semantic principles is available in Buck 1949. Since the deep-cutting structural agreements between the various daughter languages cannot have arisen secondarily, we have to assume a common underlying proto-language. Forms reconstructed for this proto-language are labeled Indo-European and are preceded by an asterisk. For the general background and terminology my titles for 1984, 1989, and 1992 may be consulted.

The following list offers a selection of words found in the varius Indo-European languages meaning 'mother' (='female person who has given birth to a child'): Sanskrit matar-, Avestan matar-, Greek meter, Latin mater, Old High German muoter (precursor of Modern High German Mutter), Old English modor (precursor of Modern English mother). Old Irish mathir, Old Church Slavic mati.

These words have a number of phonetic features in common, and from the semantic point of view they share the meaning "female person who has given birth to a child'. From the phonetic point of view we can note that all the quoted forms have initial m-, a number of them have final -r-. and even those that lack the final -r exhibit it in inflexions; thus the genitive belonging to the Old Church Slavic mati is matere. The medial consonant is -t- in some cases, and it is usually assumed that in those languages where the medial consonant differs this is due to some secondary change. If we simplify the reconstruction process somewhat, we can arrive at a reconstructed stem form *matér- 'mother' for Indo-European. Alternatively the accent could have been on the first syllable, then we would reconstruct IE *máter-. The evidence on this point is conflicting.

Apart from innovations in the sound shape (in Greek, for example, the vowel -a- is kept in the Doric dialect, but in Attic -a- yields -e-), we must also reckon with the possibility of semantic innovation. The continuation of the stem *mater- in Lithuanian is moteris, whose meaning is "woman' in general. The semantic change is by no means difficult to explain. The semantic range of moter- *female who has given birth to a child' was extended to 'female' (in general). For 'mother' derivatives like motina are used in Lithuanian.

The corresponding term meaning 'male person who has begotten a child' shows a somewhat different distribution. The following items can be gleaned from Pokorny's collection: Sanskrit pitar-, Avestan pitar-, Greek pater, Latin pater, Old High German fater (precursor of Modern High German Vater), Old English faeder (precursor of Modern English father), Old Irish athir.

A cursory glance at these words immediately reveals that the reconstruction is less straightforward than with the word for 'mother'. To be sure we again have the final stem-final -r. And all languages that offer information on the accent indicate the accent on the vowel preceding the -r. We have the same distribution as in the word for 'mother' with regard to the medial consonant. The beginning of the word, however, exhibits considerable variety. It has usually been assumed that the initial consonant is *p-, so that the consonantal skeletal of the steam can be reconstructed as *p-t-r-. In Celtic (as represented by Old Irish) *p- was lost, and in Germanic it was shifted to *f-.

The vowel between -t-r- can be assumed to be the same as in *mater-. But clearly the vowel between *p-t- cannot have been identical with the one found in *mater-. On the basis of the attested material this vowel has traditionally been posited as a and termed "shwa" (the term is from Hebrew grammar). For the 'male who has begotten a child' we can thus reconstruct a form IE *fater-.

It will have been noted that neither Old Church Slavic nor Lithuanian words for 'father' have been mentioned so far. The Old Church Slavic word for 'father' is oticu (also found in Russian otec), and this has always been considered to have arisen in nursery language. With regard to Lithuanian tëvas 'father' it must be admitted that its origin is hardly clear although some scholars think that the word may be connected with *pater-. But this is hardly convincing. Buck (1949:103) stated laconically that *pater- "is completely replaced" in Balto-Slavic.

The details of the scholarly discussion need not detain us here. What needs stressing, however, is the type of difficulty and uncertainty we are faced with. We gain the impression that somehow the word 'mother' is attested more widely than that for 'father'. We might conclude from this that the word for 'mother' arose at an earlier stage of Indo-European than that for 'father'. One could be tempted to say that *ma- is a very basic nursery element which served as the linguistic material on which the term *mater- was built. *pe-, evidently somewhat less readily explainable, could then have arisen secondarily and could somehow mirror the pattern set by *ma-. Of course we must be very careful with all conclusions of this kind. We must always reckon with accidents and chance coincidences.

It is worth noting in this context that fourth-century Gothic regularly uses the word atta (certainly somehow related to Old Church Slavic oticu), and the clearly inherited term Gothic fader is attested only once: in Galatians 4:6 the translator had to render Greek abba ho pater, and he put down abba, fadar. It seems clear that atta was somehow not suited for the solemn context; perhaps it was felt to mean something like Daddy. But for Galatians 4:6 we might want to assert that the word for 'father' was unattested in Gothic. The word corresponding to Old High German muoter is not attested in Gothic at all, in its place we find aithai.

The preceding remarks were meant simply to show some of the hazards we encounter in reconstructing Judo-European or successive daughter languages. But one basic conclusion should become clear nevertheless: if the earliest attested Indo-European daughter languages give us glimpses into their linguistic system in the second millennium before the birth of Christ, then it is obvious that common Indo-European must have been in use at a time before 2000 B.C. How much before that date is practically unknowable. For those who read the relevant literature it will soon become strikingly clear that most Indo-Europeanists avoid dates with regard to the chronology of Indo-European. If any dates are ventured, it is normally said that Indo-European must have ceased to be a uniform proto-language around 3000 B.C. at the latest.

With this type of wording one can certainly agree. But this does not in any way tell us how much earlier the split-up of the ancestral language may really have occurred. And we have no way of knowing. My own view is that ther common underlying proto-language would seem to belong to a much earlier period. I would imagine that by around 3000 B.C. there were several daughter dialects in use. To my mind, the latest possible date for a reasonably uniform proto-language is around 5000 B.C., but perhaps the date should be pushed much further into the past. It goes without saying that to try to locate the Indo-European spoken before the year 5000 B.C. would be an impossible task.

If we simply consider the general situation, it will indeed become clear that the timeless reconstructions that are usually practiced in Indo-European grammar are misleading. The Buck 1949:103 lists 'father', 'mother', and 'parents' as three concepts which he apparently views as being on the same chronological level. But surely they are not. There is good reason for assuming that 'mother' is the oldest of the three since the process of bearing children certainly required a descriptive term much earlier than the male's role of implanting semen. In the case of 'parents' the recent character of many of the terms used in the individual Indo-European languages is fairly obvious. No common term for 'parents' can be reconstructed for Indo-European.

The main conclusion we can draw so far is that the linguistic material tells us practically nothing about the social structure of the group(s) of human beings among whom the language was used.

Although Indo-European as defined in 1. is clearly and primarily a linguistic term, inquiries about the "speakers" of Indo-European are fully legitimate. But one must be aware of the problems faced in this type of investigation. If we want to pin down the identity of "speakers", we must of necessity define chronology and geography: We want to know within what time-span they occupied a definable piece of land. And here the difficulties are enormous.

The science of archaeology is concerned with digging ancient sites and analyzing the material unearthed. But all material objects are by definition mute unless they carry an inscription that we can in some way decipher: No matter how many bracelets and earrings of whatever given fabrication we find, we can never know what type of language the manufacturers and their clients spoke.

The following sections will incorporate material from archaeology. Above all, some fairly recent theories put forward by Marija Gimbutas will be presented and discussed.

The difficulties adumbrated above, although quite formidable, should not deter us from trying to delve ever deeper into the past. Markey 1990 offers stimulating papers from an interdisciplinary conference held at Bellagio (Lake Como, Italy) and sponsored by the Rockefeller Foundation. The two opening papers immediately present the main problems: T. Gamkrelidze, "On the Problem of an Asiatic Original Homeland of the Proto-Indo-Europeans", 5-14, and C. Renfrew, "Archaeology and Linguistics: Some Preliminary Issues", 15-24, argue from different angles for a Near Eastern origin of the "Indo-Europeans". But it seems to me that we can leave the ultimate origin of the "Indo-Europeans" as an unsettled question; our main concern here will be "Indo-Europeanization" of Europe. In contrast to Indo-European, definitely a linguistic term, "European" can only be defined within the framework of geography.

The centerpiece in our context is certainly Marija Gimbutas, "The Collision of Two Ideologies" (171-178), and the title already clearly indicates her approach. Since it is not possible to discuss many of the detailed questions dealt with by Gimbutas, I wish to concentrate on the issue raised in the title of this essay: Was there really a collision of the "Indo-Europeans" with "Pre-Indo-Europeans"?

It seems to me that this is not the type of question which would allow a clear "Yes" or "No" as an answer. But since Markey's volume in general and Gimbutas' paper in particular definitely answer in the affirmative, I wish to record some of my reservations here. My main aim is to make clear what the assumptions are on which we can base our conclusions, but I would also like to expose some fallacies in the argumentation presented.

The "Indo-Europeans" would of necessity be speakers of the reconstructed language Indo-European. Since Indo-European languages are now spoken around the whole globe, it goes without saying that the speakers of Indo-European must have moved a lot. We know of the spread of English in the Early Modern Period. Around the year 1400 (when Chaucer died) the number of speakers of English did not exceed that of the speakers of German or Spanish or French. Nowadays English is the most widely spoken Indo-European language in the world.

When Indo-European languages are attested for the first time, they are already spread out over an enormous territory. The German term for Indo-European is indogermanisch, and it aptly catches the two extremes of the area over which various Indo-European languages were spoken in the Early Modern Period, namely Indic in the far south, and Icelandic, a Germanic language, in the north. How the "Indo-Europeans" came to occupy such a far-flung territory is one of the major riddles in scholarly discussion.

This problem can basically be subdivided into two separate issues: on the one hand we want to know where the original home, the "Urheimat," of the "Indo-Europeans" was; on the other hand, we want to trace the migrations of the various groups that carried the individual Indo-European languages into the new habitats. Since the following discussions will be exclusively concerned with "Indo-Europeans in Europe," some very difficult issues can be omitted from consideration. It may be confidently assumed that Europe was not the original location of Indo-European, but this also means that we need not of necessity pin down that original location. Suffice it to say that Indo-European was "imported" into what we call Europe.

This observation has serious consequences. In all probability we can contend that there must have been people living in Europe when Indo-European arrived. What I wish to discuss is how some scholars view the arrival of "Indo-Europeans."

Since Indo-European is basically a linguistic term, any extension of its use must be conducted with circumspection. It can by definition apply to a reconstructed language only, and, in theory at least, this language must be viewed as uniform and free of dialects. What we can reconstruct on the basis of the individual daughter languages is the ancestral form of the individual languages, which will by definition exhibit dialect variety. Whether we can ever agree on a completely uniform underlying Indo-European forebear is at best doubtful.

If we use Indo-European with regard to speakers of individual languages, this terminology becomes extremely imprecise. The individual languages would be better termed Pre-Greek, Pre-Italic, Pre-Germanic, and so on.

It is indeed meaningful to talk about contacts of "Pre-Greeks" with speakers of other languages. And it is definitely useful to discuss the migrations of the "Pre-Greeks," from wherever they may have come, to the Greece and its colonies that they occupied in historical times.

But it is extremely doubtful whether it makes any sense whatsoever to talk of contacts between "Pre-Greeks" and "Pre-Indo-Europeans" since the latter would refer to a totally amorphous group. Furthermore, we would certainly have to settle the chronology of the supposed "Pre-Indo-Europeans."

The "collision of ideologies" would presumably have to mean that there was a firm "Indo-European" ideology which could come into some sort of contact with other ideologies. And here the logic is extremely weak. The "Indo-European" ideology has often been characterized in rather stereotyped ways, as exemplified in the following passage:

The Indo-European culture, with its patriarchal and classed social structure and its religion of warrior gods, was superimposed on Europe in several stages: first in central Europe in 4500-3000 B .C., it spread to the north and south in the first half of the third millennium B.C., coming later to the Aegean and Mediterranean islands. From the third millennium on, hybridized cultures and mythologies emerged. 

(Gimbutas in: Markey 1990:171)

My main objections to this thesis are of a chronological nature. If we take Gimbutas' more recent threshold of 3000 B.C., it can be safely said that there is practically no reason for assuming an Indo-European in the shape of a reasonably uniform language at so late a date. It is practically certain that the split-up of the original stock must have taken place a good deal earlier. I would not want to venture a date/but by 5000 B.C. the ancestral language must have been spread over so wide an area that it is almost meaningless to term it Indo-European in the sense outlined above. By 3000 B.C. we could probably talk of Pre-Germanic or Pre-Celtic or Pre-Italic as being carried to Europe or as already being used there.

But the whole logic of Gimbutas' presentation is flawed by the inclusion of material which is apparently used in order to prove the opposite of what was shown in the last quotation. Thus we are told that, for example, the Lithuanians have preserved folklore elements down to modern times which perpetuate "Pre-Indo-European" features:

...a very similar custom was described in the writings of 18th century Lithuania in which a black suckling pig was ritually sacrificed to the Earth Mother at the time of blessing the seed in the fall. Such customs show the influence of the pig on the fertility of the fields. An intimate connection between suckling pigs and grain is seen in the clay sculptures studded with grain from the Linearbandkeramic and Cucuteni cultures, 5th millennium B.C. (Gimbutas 1991:229)

Gimbutas has a lengthy section on the "bear-mother" and rightly mentions that "to this day in Lithuania, a woman in childbirth is called a bear" (1991:226). This can only mean that the Lithuanians preserve matrilinear traits. The fact that the adoration of the mother goddess was strong in early times is also to be expected. The assumption that the "Christian Virgin Mary is a demoted version of this original deity" (1991:223) is plausible.

But clearly this can only mean that the cult of the mother goddess continued for a long time; and, in any case, neither the cult itself nor a possible abolition of it is in any direct way to be attributed to the "Indo-Europeans."

The general conclusion can only be that Gimbutas herself in a way demolished the case she built up; however, to my mind, this demolition points the way to a better understanding of the prehistoric processes.

If the Lithuanians preserve clear traces of matriarchy, then patriarchy definitely cannot be an exclusive feature of "Indo-Europeans" since without any doubt Lithuanian is an Judo-European language.

We must give up the global attribution of patriarchy, warfare, etc. to the "Indo-Europeans." It is quite obvious that in the Europe of the last millennia B.C. a good deal of conquest occurred. If we want to pin down these conquests to groups of human beings and define them on the basis of their language, then the only reasonable terminology would consist in calling them "Pre-Greeks," "Pre-Italics," "Pre-Celts," "Pre-Germanics," etc. Calling them "Indo-Europeans" is meaningless and confusing.

By bringing in heterogeneous material, Gimbutas is definitely digging her own grave. She conflates totally conflicting pieces of evidence.

Before proceeding with more general observations, I wish now to concentrate on one mainly linguistic piece of Markey 1990 in order to show the limits of reconstruction work in Indo-European. E.P. Hamp, "The Indo-European Horse" (Markey 1990:211-226); offers a number of important observations, both linguistic and non-linguistic on domesticated animals. The 'sheep' was domesticated around 9000 B.C., whereas the 'horse' belongs to a more recent period (around 6000 B.C.). Inherited words for both of these animals stil! exist in present-day Lithuanian. Lithuanian avis 'sheep' continues IE *owis (cf. Latis ovis etc.), and Lithuanian aðva goes back to IE *ekwa, which is the feminine answering to Latin equus 'horse'. It is the word for 'horse' that will be our main concern here.

If Hamp is right in saying that the 'horse' was domesticated around 6000 B.C., then this clearly means that around the year 6000 B.C. speakers of Indo-European needed some term they could use whenever they wanted to refer to the animal.

The important part of Hamp's paper is the etymological account he offers for Indo-European *ekwo-. This stem is reconstructed on the basis of Sanskrit asva-, Latin equus, Lithuanian aðva, Greek hippos and some further cognates. Although details need not occupy us here, it may be said that the basic connection suggested by Hamp, namely linking *ekwo- with Greek okus 'rapid,' is very plausible. But if this is so, then ekwo-, at the time when the form was created, can have meant only something like 'rapid' or, possibly, 'the rapid one'. This entails the conclusion that we are unable to reconstruct the word for 'horse' for Indo-European. If *ekwos meant 'rapid' or, possibly, 'the rapid one,' then clearly another term meaning 'animal' must originally have been used together with *ekwos 'rapid,' there must have been an expression like *ekwos X '(the) rapid animal,' where X clearly meant 'animal.' In the course of time "X" was omitted in the sequence "*ekwos X" '(the) rapid animal (~'horse'), and *ekwos 'horse' is due to elliptical use.

The main conclusion can only be that the linguistic analysis of the word for 'horse' hardly tells us anything at all about the everyday life of the persons who used this word. It cannot be "proved" in any way that the "Indo-Europeans" were the ones who domesticated the horse.

Although the procedure generally adopted in archaeology is understandable because it has a time-honoured tradition, I think we must voice our strongest objections from the linguistic angle. The underlying problem is that we clearly need terms for persons whose remains and traces we uncover. We want to give names to the producers of whatever we find. And we tend to name these persons somehow on the basis of the language or languages that were or still are spoken in the area where the finds occur.

In order to show the hazards (not to say, folly) of this type of procedure it must suffice to point out one modern parallel. If it is said that the "Indo-Europeans" invented the wheel because in the majority of Indo-European languages we find words meaning 'wheel' (Greek kuklos, Sanskrit cakra-, etc.) and having similar shape, then this would be just as meaningful as saying that the Japanese invented the automobile because words like Nissan, Mazda, etc. are used with reference to cars produced in Japan. Inventions need not be directly connected with the language the inventor speaks but are due to the inventor's ingenuity.

As was mentioned in the beginning of this essay, Indo-European is a linguistic term. It is legitimate to extend this term to other spheres, but one would wish that this should be done with care and circumspection. In Gimbutas 1991 the linguistic material is at times used in a way that can only be called annoying.

For example, some of the items mentioned in her lists (Gimbutas 1991: 393ff.) are totally incomprehensible to the non-specialist and will therefore at best serve to ridicule Indo-Europeanists' reconstructions. Since she largely dispenses with diacritic marks, the forms differ considerably from the generally accepted graphic conventions. Quite a few readers and critics will say that Indo-European shows a very rapid pace of development. The phonemes u and i in vocalic function are noted as w and y, the so-called "laryngeals" are represented as H. Consequently the word for 'son' (reconstructed on the basis of Sanskrit sûnus, Lithuanian sûnus, etc. as IE *sûnus in traditional representation) appears in the shape *swHnws p. 395); many readers will wonder whether any human being was ever able to pronounce a word of this shape. Other errors like Germanic *zaizaz (p. 394 [recte *gaizaz]) may be corrected by specialists. In some cases it is hard to know what the author really wanted to say. The entry *(H)nsi (sic), which is intended as preform for Latin ensis, also offers "Germanic sxasxaz" (recte *saxsaz); this may be intended to mean that both Latin and Germanic have a word for 'sword', which of course nobody wuld have doubted anyway. Obviously this type of material needs thorough revision.

It must remain doubtful, however, whether any revision can lead to more persuasive results since, to my mind, the underlying assumptions are largely incorrect. Agreements in vocabulary items prove hardly anything with regard to the prehistory of the speakers of the respective languages. Thus the fact that practically all languages nowadays spoken can somehow render "telephone', 'aeroplane' and 'television' does not prove anything about the share the speakers of the respective languages had in bringing about these inventions. Cultural borrowings are common.

Nevertheless it seems that this totally imprecise use of the term Indo-European is definitely spreading. I will now comment on just one aspect of a recent book by the outstanding French Indo-Europeanist, Francoise Bader (Bader 1989), who maintains that the 'Indo-Europeans" were familiar with writing. Two quotations will indicate her position.

L'on peut être sûr, en effet, que les hommes de lange indoeuropéenne ont connu 1'écriture. (Bader 1989:17) Les Indo-Européens ont connu 1'écriture. (Bader 1989:66)

The reader may well wonder what this can mean. Surely no texts have been found that are in any way decipherable and can be assumed to represent Indo-European.

If we find some type of carving which is perhaps datable to 5000 B.C., it is meaningless to call this "Indo-European writing" since we would not even know what the language represented was like. "Writing" certainly means that a linguistically decodable message is represented by some graphemic signs.

Writing is of course of immense importance in the cultural history of mankind, and there is no doubt that any type of script is the result of a most complex historical process. It should be stressed most emphatically that, important though the history of writing certainly is, it is not immediately linked to the development of any language represented by whatever means on whatever object. This can be shown most immediately by referring to the fact that all the world's languages can be represented one way or another by using the Latin alphabet. The writing system based on some type of phonemic representation can be applied to any language. So far, we have no texts that can in any way be termed Indo-European; all we have are the asterisked forms resulting from the reconstruction work of modern scholars.

Since we don't know where the "Indo-Europeans" lived and cannot even say what millennium should be chosen for any search for their habitat, it is certainly useless to talk about "Indo-European" writing. I would have no objection to "Pre-Greek" writing, but by definition this would have to be in a language that ultimately surfaced as some form of Greek as we know it.

With regard to Indo-European languages in Europe we have two options. Either Indo-European was autochthonous in Europe (which seems rather improbable) or speakers of Pre-Celtic, Pre-Germanic, Pre-Italic, Pre-Greek, Pre-Baltic, etc. arrived at some time in the Pre-Christian millennia. The latter alternative strikes me as the more plausible one.

Alexander Häusler, a German prehistorian, has dealt with the spread of the "Indo-Europeans" and the development of Indo-European on several occasions, and one quotation may suffice to indicate his position:

Herauskristallisierung der indogermanischen Einzelsprachen aus einem im Areal Zwischen Rhein und Ural spätestens bereits im Frühneolikthikum verbreiteten Sprach-kontinuum,—eine Konzeption, die ohne "Invasoren" und "unterworfene Völker" auskommt und sich somit nicht so gut vermarkten läbt wie manch andere spektakuläre Hypothese. (Häusler 1991:96)

All in all, this theory seems far more plausible than the assumption of any type of "clash'. This means, of course, that specious theories about the original habitat of the "Indo-Europeans" lose their basis. And, at all events, the ascription of special character traits to the "Indo-Europeans"  becomes practically meaningless. There should not be any doubt, however, that Indo-European refers to a clearly definable linguistic system.

If by "Indo-Europeans" we mean speakers of the common underlying language that ultimately surfaced in the individual daughter languages, it is clear that we can "date" these speakers to any period preceding the year 6000 B.C. Their social structure would have been the same as that found among speakers of other languages of the period.

Ultimately new ways of thinking and new social structures were to develop, but these are not primarily connected with the various languages which the respective speakers used.

The attribution of a patriarchal social structure and of a strong tendency to warfare to the "Indo-Europeans" does not really make any sense whatsoever. The introduction and development of patriarchy may have occurred in several language groups, and it is certainly not specific of "Indo-Europeans".

The assumption of war faring "Indo-Europeans" has a long tradition, but to my mind there is practically no evidence to back it up. The question is in the first instance of chronology. If we call the heroes of the Homeric poems "Indo-Europeans," then it is quite obvious that doing battle was a major pastime for them. But the same would clearly apply to all their antagonists. And certainly not all of the antagonists were "Indo-Europeans." Peace or war is not defined by the language spoken by the people involved.

All the same, the tradition of equating "Indo-Europeans" with war faring invaders stretches back to the nineteenth century. It will certainly be difficult to eradicate this view, even though its fallacious nature becomes obvious if we consider the spread of the Germanic languages in the modern world. The most widely used Germanic language is certainly English, with German in second place, and the others (Swedish, Norwegian, Icelandic, etc.) following. Danish has a relatively small number of speakers nowadays, yet it would be absurd to assume that the Danes did not know how to do battle. Clearly other criteria are also and, perhaps primarily, involved in the spread of language. Of greater importance than conquest are mercantile ingenuity and inventive achievements.

Many of our currently held assumptions are tainted by nineteenth-century attitudes. One pervasive theory assumes the purity of a race or a nation, and a one-to-one relationship between race and language has often been invoked. Whether anybody has ever followed up this idea to its logical conclusion is not really important here. But surely the logical conclusion would be that incest and inbreeding were the usual and exclusive practice in a group of human beings that managed to remain "pure" over many centuries. There is absolutely no reason to pursue such a line of thought further.

A major problem in dealing with questions of this type lies in the fact that the term Indo-European languages, we mean a group of genetically related languages, and in most cases we really are talking abut the ancestors of some of the individual Indo-European languages that have left written records or are still spoken nowadays. The confusion arising out of this use could be avoided if we stuck to terms like Pre-Greek, Pre-Lithuanian, Pre-Celtic, etc. If we wanted an umbrella term for the whole group, we might wish to use Proto-European languages or the like. This term would clearly indicate the relationship, but would at the same time mark the difference between the ancestral language and the descendent languages.

The question of Indo-European reconstructions is a particularly complicated one. We depend on equations among cognates, and normally something like three witnesses are considered enough for positing a given underlying Indo-European form. In some cases this yields useful results, in others the results are hilariously astray. If we want to have reconstructions that really do prove something about "Indo-Europeans," it seems to me that we should increase the number of witnesses required.

This procedure might yield useful results. Thus the verbal form IE *esti can be reconstructed on the basis of such bits of evidence as Latin est, German ist, English is, Lithuanian esti, Sanskrit asti and many more. We can therefore assume that it existed in the proto-language. However, soon doubts arise. Do we take it that IE *esti had such a vague meaning as 'is'? We would do better to assume that IE *esti should have meant something like 'occupies a place' or the like. So even here the reconstruction faces considerable difficulties.

If we can agree that IE had a word meaning "I', this is perhaps also interesting, but not very revealing. I assume that pretty well every language has a word meaning "I".

Is there anything that is peculiar to the IE linguistic system? I think that the most obvious candidate is the system of IE numerals. It is indeed noteworthy that the numerals from 'l' to '10' can be reconstructed with high probability, furthermore there are agreements in the numerals even up to '100'. This would indicate that at an early stage the "Indo-Europeans" had developed a rather complicated system of counting.

Since many details in the preceding essay have been of necessity "negative," it might be right and proper to concentrate on at least one issue in which Lithuanian linguistic material allows us to penetrate deeply into linguistic prehistory.

In the second person singular of present-day Lithuanian all verbs end in -i in the present, preterite and future, and together with a preceding vowel -i forms a diphthong:

 
vedi 'you lead'
esi 'you are'
turi 'you have'
þinai 'you know'
buvai 'you were'
turëjai 'you had'
turësi 'you will have'
etc.

The reflexive forms of the type vediesi clearly indicate that the starting-point must have been of the order of Baltic *-ei. This marker immediately makes one think of -ei- in Greek fereis 'you carry,' ferei 'he carries.' I am not saying that the element -ei- can be accounted for without difficulty, but I do maintain that the agreement between Baltic and Greek can hardly be due to coincidence. This type of agreement is what historical linguistics draws its material from.

Historical linguistics is ultimately a peaceful and democratically orientated scholarly pursuit. It is meant to have a positive influence on the everyday life of the fellow human beings of its practitioners.

If we have to choose between 'yes' and 'no,' then the question asked in the title of this essay can, to my mind, be answered only in the negative. It would certainly be meaningful to say that "Creeks" collided with "Pre-Greeks," and for this type of clash we have enough historical evidence to warrant our saying so. But we certainly do not know that the "Indo-Europeans" collided with any "Pre-Indo-Europeans."

If I wanted to push my argumentation to the extreme, I could phrase my views as follows: if the "Pre-Indo-European Europeans" were peaceful people, then there was absolutely no need for the arriving "Indo-Europeans" (no matter where they came from) to be war-loving. Military action, deplorable wherever it occurs, always requires at least two antagonists. In early periods of human history there was probably much less antagonism than in the modern period. There is no basis for talking of any collision. It would probably be much better to assume a symbiosis of various cultural strata. The clear-cut dichotomy between warring "Indo-Europeans" and peaceful "Pre-Indo-European Europeans" is a fallacy.

References

Bader, Francoise: 1989. Langue des dieux, ou l' hermetisme des poetes indo-europeens. Pisa, 1989.
Bammesberger, Alfred: 1984. English Etymology. Heidelberg; Winter. 
Bammesberger, Alfred: 1989. English linguistics. Heidelberg: Winter. 
Bammesberger, Alfred: 1992. The Place of English in Germanic and Indo-European. In: Richard M. Hogg, The Cambridge History of the English language, vol. 1: The Beginnings to 1066. Cambridge: University Press. 
Buck, Carl Darling: 1949. A Dictionary of Selected Synonyms in the principal Indo-European languages. Chicago: University of Chicago Press.
Gimbutas, Marija: 19912. The Civilization of the Goddess. San Francisco: Harper; see my review of the volume in Lituanus 38 (1992).
Häusler, Alexander: 1991. Review of Wolfgang MEID, Archaologie und Sprachwissenschaft (1989). In Kratylos 36:95-98.
Markey, T. L. Markey, Greppin, John A.C. (editors): 1990. When Worlds Collide: Indo-Europeans and Pre-Indo-Europeans. The Bellagio Papers. Ann Arbor 1990: Karoma Publishers; Linguistica Extranea, Studio 19.
Pokorny, Julius: 1959. Indogermanisches Etymologisches Wörterbuch. Bern: Francke.
Walde, Alois: 1927-1932. Vergleichendes Wörterbuch der indogermanischen Sprachen, 3 vols., edited by Julius Pokorny. Berlin.