The Indo-European Homeland
The evidence of the oldest literary records of the Indo-European family of languages, the Rigveda and the Avesta, as we have seen, clearly and unambiguously depicts a movement of the “Indo-Iranians” from the east to the west and northwest.
And Central Asia and Afghanistan, which, according to the standard theory, is the route by which the Indoaryans migrated into India, turns out to be the route by which the Iranians migrated westwards and northwards.
This deals a body-blow to a very vital aspect of the theory which places the original Indo-European homeland to the northwest of Central Asia (ie. in and around South Russia), and it lends strong support to the theory that the Indo-European family of languages originated in India.
If, therefore, the scholars,, by and large, remain strongly resistant to the Indian homeland theory, it is not because the facts of the case rule out this theory, but because a defence of the standard theory has become a dogma with the scholars, and any scholar, particularly an Indian one, who pursues the Indian homeland theory is automatically held suspect as a fundamentalist or a chauvinistic nationalist.
So much so that any theoretical scenario which is loaded against the Indian homeland theory gains respectability; and some scholars go to the extent of deliberately projecting a blatantly false picture of the whole situation, calculated to place the Indian geographical area as far out of the geographical ambits of early Indo-European history as possible.
An example of this is the clearly fraudulent case presented by a Western scholar, Victor H. Mair, in a compilation, edited by himself, of the papers presented at the International Conference on the Bronze Age and Iron Age Peoples that was held at the University of Pennsylvania Museum of Anthropology and Archaeology (April 19-21, 1996).
Mair prefaces his presentation with a sharp diatribe against a wide range of what he calls “extremists, chauvinists, and other types of deranged - and possibly dangerous - persons (eg. those who locate the Indo-European homeland in such highly improbable, if not utterly impossible, places as the Arctic, along the Indus Valley, in the Tarim Basin, in China; nationalists and racists of various stripes; kooks and crazies who attribute the rise of Indo-Europeans to extraterritorial visitations, etc.)”.1
At the same time, he places himself in a beatific light by announcing that he himself is impelled to carry out “the search for the Indo-Europeans and their homeland”, and to “pursue it with enthusiasm”, because: “I perceive such an inquiry to be (1) intrinsically compelling. (2) innately worthwhile. (3) historically significant. (4) humanistically important. (5) devoid of political content. (6) scientifically solvable. (7) intellectually satisfying”, and dismisses scholars of a lesser breed with the pompous announcement: “If other people want to distort or pervert the search for their own purposes, that is their problem.”2
Mair proceeds to present his thesis, in a quasi-humorous vein, likening the spreading Indo-European family to a spreading amoeba.
And he presents his final conclusions, about the schedule of migrations and expansions of the Indo-European family, in the form of a series of nine maps, supposed to represent the situations in 4200 BC, 3700 BC, 3200 BC, 3000 BC, 2500 BC, 2000 BC, 1500 BC, 1000 BC, and 100 BC respectively.
We are concerned here only with his depiction of the Indian geographical area in these maps: incredible as it will seem to any scholar who is even generally acquainted with the facts of the Indo-Iranian case, Mair’s map for 1500 BC3 shows the undifferentiated Indo-Iranians still located to the north and west of the Caspian Sea!
Which western academic scholar in his right senses, and with any concern for his academic credentials, will accept that this depiction of the Indo-Iranian case in 1500 BC is even reasonably honest, or deny that it represents a most blatantly mischievous distortion of the facts?
It may be noted that Mair, pompously and sweepingly, claims that his maps “are intended isochronously to take into account the following types of evidence: linguistic, historical, archaeological, technological, cultural, ethnological, geographical, climatological, chronological and genetic-morpho-metric - roughly in the order of precision with which I am able to control the data, from greatest to least. I have also endeavoured to take into consideration types of data which subsume or bridge two or more basic categories of evidence (eg. glotto-chronology, dendrochronology, and linguistic paleontology).”4
An examination of the maps, even as a whole (and not just in respect of the Indo-Iranians) shows that Mair would be hard put to explain how his arbitrarily, and even whimsically, drawn-out schedule of migrations and expansions fulfils even any one of the above academic criteria, let alone all of them.
Mair claims to be interested, for a variety of noble reasons, in “the search for the Indo-Europeans and their homeland”; but it is clear that a “search” of any kind is as far from his intentions as possible, since his answer (South Russia) is already determined (although he does let out that his greater personal preference would have been to locate the core of the homeland “in Southern Germany, northern Austria, and the western part of what is now the Czech Republic”5, ie. in Hitler’s home-grounds), and all those who advocate any other solution automatically fall, in his opinion, in the same category as “kooks and crazies who attribute the rise of Indo-Europeans to extra-territorial visitations”!
Mair’s presentation can certainly be classified, in his own words, as among the presentations of “extremists, chauvinists, and other types of deranged - and possibly dangerous - persons”: doubly dangerous since scholars like him function on the strength of a monopolistic academic world which grants respectability to their most blatantly fraudulent efforts’ while shunning or condemning genuinely factual studies, among which we definitely count our own.
In such a situation, where any scholar, Indian or Western, who finds that the facts indicate an Indian homeland, has to struggle against a strong tide of prejudice in Western academic circles (not to mention the deeply entrenched leftist lobby in Indian academic circles), it is clear that establishing the truth about the original homeland is, practically speaking, an uphill task.
And the fundamental obstacle is the widely held belief that the science of LINGUISTICS has proved conclusively that the Indo-European homeland is located in and around South Russia, and, equally conclusively, that this homeland could not have been located in India: this belief, as we shall see in our Appendix One (Chapter 8) on misinterpretations of Rigvedic history, is so deeply entrenched in the psyche of all scholars, whatever their views, who examine the problem, that it appears to overshadow and nullify, in their perceptions, the effect of all other evidence to the contrary.
We will, therefore, primarily be examining, in this chapter, the linguistic evidence in respect of the location of the Indo-European homeland, and it will be clear that this evidence, wherever it indicates any geographical location, invariably points towards India.
We will examine the case for the Indo-European homeland as follows:
Archaeology and Linguistics.
The archaeological evidence has always been against the theory that there was an Aryan influx into India in the second millennium B.C., an influx so significant that it was able to completely transform the linguistic character and ethos of almost the entire country.
Even D.D. Kosambi, for example, admitted the fact even as he waxed eloquent on the Aryan invasion: “Archaeologically, this period is still blank… There is no special Aryan pottery… no particular Aryan or Indo-Aryan technique is to be identified by the archaeologists even at the close of the second millennium.”6
This is in sharp contrast to the situation so far as Europe is concerned. Shan M.M. Winn, for example, points out that “a ‘common European horizon’ developed after 3000 BC, at about the time of the Pit Grave expansion (Kurgan Wave #3). Because of the particular style of ceramics produced, it is usually known as the Corded Ware horizon. However, some authors call it the Battle Axe culture because stone battle axes were frequently placed in burials… The expansion of the Corded Ware cultural variants throughout central, eastern and northern Europe has been construed as the most likely scenario for the origin and dispersal of PIE (Proto-Indo-European) language and culture.”7
After a detailed description of this archaeological phenomenon, Winn notes: “Only one conclusion seems reasonably certain: the territory inhabited by the Corded Ware/ Battle Axe culture, after its expansions, geographically qualifies it to be the ancestor of the Western or European language branches: Germanic, Baltic, Slavic, Celtic and Italic.”8
However, this archaeological phenomenon “does not… explain the presence of Indo-Europeans in Asia, Greece and Anatolia”.9
This Corded Ware/Battle Axe culture represented the third wave of “the Pit Grave expansion (Kurgan Wave #3)” in the westward direction. Winn suggests that “an eastern expansion from the Caspian Steppe also occured at this time”,10 and tries to connect up the Tocharians with “the culture… known as Afanasievo… located in the Altai region… across the expanse of the Central Asian steppe to its ragged eastern boundary”,11 and the Indo-Iranians with the Andronovo culture which “covers much of the Central Asian steppe east of the Ural river and Caspian Sea”.12
However, he admits that these identifications are purely hypothetical, and that, even in hypothesis, and assuming the Andronovo culture to be Indo-Iranian, “it is still a hazardous task to connect the archaeological evidence… in the Central Asian steppe with the appearance of Iranian (Aryan) and Indic (Indo-Aryan) tribes in Iran, Afghanistan and India”.13
Consequently, he describes Indo-Iranian, archaeologically, as an “Indo-European branch which all the homeland theories we have reviewed so far have failed to explain”.14
The archaeological evidence for any Indo-European (Aryan) influx into India is missing in every respect:
a. There is no archaeological link with any other Indo European culture outside India.
b. There is no archaeological trail leading from outside into India.
c. There is no internal evidence in respect of any notable change in the anthropological or material-cultural situation in the northwestern parts of India, in the second millennium BC, which could be attributed to an Aryan influx.
In fact, the situation is so clear that a majority of archaeologists, both in India and in the West, today summarily reject the idea that there was any Aryan influx into India from outside in the second millennium BC. They, in fact, go so far as to reject even the very validity of Linguistics itself as an academic discipline which could be qualified to have any say in the matter.
This has created quite a piquant situation in Western academic circles. In his preface to a published volume (1995) of the papers presented during a conference on Archaeological and Linguistic Approaches to Ethnicity in Ancient South Asia, held in Toronto on 4th-6th October 1991, George Erdosy notes that the Aryan invasion theory “has recently been challenged by archaeologists who - along with linguists - are best qualified to evaluate its validity. Lack of convincing material (or osteological) traces left behind by the incoming Indo-Aryan speakers, the possibility of explaining cultural change without reference to external factors and - above all - an altered world view (Shaffer 1984) have all contributed to a questioning of assumptions long taken for granted and buttressed by the accumulated weight of two centuries of scholarship.”15
However, Erdosy points out, the perspective offered by archaeology, “that of material culture… is in direct conflict with the findings of the other discipline claiming a key to the solution of the ‘Aryan problem’, linguistics… In the face of such conflict, it may be difficult to find avenues of cooperation, yet a satisfactory resolution of the puzzles set by the distribution of Indo-Aryan languages in South Asia demands it. The present volume aims for the first step in that direction, by removing mutual misconceptions regarding the subject matter, aims, methods and limitations of linguistics and archaeology which have greatly contributed to the confusion currently surrounding ‘Aryans’. Given the debates raging on these issues within as well as between the two disciplines, a guide to the range of contemporary opinion should be particularly valuable for anyone wishing to bridge the disciplinary divide… indeed, the volume neatly encapsulates the relationship between two disciplines intimately involved in a study of the past.”16
The archaeologists and anthropologists whose papers feature in the volume include Jim G. Shaffer and Diane A. Lichtenstein, who “stress the indigenous development of South Asian civilization from the Neolithic onwards, and downplay the role of language in the formation of (pre-modern) ethnic identities”;17 J. Mark Kenoyer, who “stresses that the cultural history of South Asia in the 2nd millinnium B.C. may be explained without reference to external agents”,18 and Kenneth A.R. Kennedy, who concludes “that while discontinuities in physical types have certainly been found in South Asia, they are dated to the 5th/4th, and to the 1st millennium BC, respectively, too early and too late to have any connection with ‘Aryans’.”19
Erdosy and Michael Witzel (a co-editor of the volume, and a scholar whose writings we will be examining in detail in Appendix Two: Chapter 9) seek to counter the archaeologists in two ways:
1. By dismissing the
negative archaeological evidence.
We will examine their efforts under the following heads:
A. The Archaeological
I.A. The Archaeological Evidence
According to Erdosy, “archaeology offers only one perspective, that of material culture”.20 This limit renders the archaeologists unable to understand the basis of the linguistic theory.
Erdosy stresses that the theory of the spread of the Indo-European languages cannot be dispensed with: “The membership of Indic dialects in the Indo-European family, based not only on lexical but structural criteria, their particularly close relationship to the Iranian branch, and continuing satisfaction with a family-tree model to express these links (Baldi, 1988) all support migrations as the principal (albeit not sole) means of language dispersal.”21
But, according to him, the archaeologists fail to understand the nature of these migrations: they think that these migrations are alleged to be mass migrations which led to cataclysmic invasions, all of which would indeed have left behind archaeological evidence.
But, these “images of mass migration… (which) originated with 19th century linguists… exist today principally in the minds of archaeologists and polemicists”.22 Likewise, “the concept of cataclysmic invasions, for which there is. little evidence indeed… are principally held by archaeologists nowadays, not by linguists who postulate more gradual and complex phenomena”.23
It is this failure to realize that the “outmoded models of language change”24 of the nineteenth century linguists have now been replaced by more refined linguistic models, that leads to “overreactions to them (by denying the validity of any migrationist model) by both archaeologists and Hindu fundamentalists”.25
Thus, Erdosy, at one stroke, attributes the opposition of the archaeologists to the linguistic theory to their ignorance of linguistics and clubs them together with “polemicists” and “Hindu fundamentalists” in one broad category of ignoramuses.
But, it is not as easy to dismiss the views of the archaeologists as it is to dismiss those of “Hindu fundamentalists”.
It must be noted that the opposition of the archaeologists is to the specific aspect of the Aryan theory which states that there was an Aryan influx into India in the second millennium B.C., and not to the general theory that the Indo-European language family (whose existence they do not dispute) must have spread through migrations of its speakers: obviously the languages could not have spread through the air like pollen seeds.
But Erdosy puts it as if the archaeologists are irrationally opposed to the very idea of “the membership of the Indic dialects in the Indo-European family” or to the “family-tree model”. It is as if a scientist were to reject the prescriptions of a quack doctor, and the quack doctor were to retaliate by accusing the scientist of rejecting the very science of medicine itself.
The linguistic answer to the total lack of archaeological evidence of any Aryan influx into India in the second millennium BC, is to “postulate more gradual and complex phenomena”.
But, apart from the fact that this sounds very sophisticated and scientific, not to mention superior and patronising, does the phrase really mean anything? What “gradual and complex phenomena” could account for the linguistic transformation of an entire subcontinent which leaves no perceptible archaeological traces behind?
And it is not just linguistic transformation. Witzel admits that while “there have been cases where dominant languages succeeded in replacing (almost) all the local languages... what is relatively rare is the adoption of complete systems of belief, mythology and language… yet in South Asia we are dealing precisely with the absorption of not only new languages but also an entire complex of material and spiritual culture ranging from chariotry and horsemanship to Indo-Iranian poetry whose complicated conventions are still used in the Rgveda. The old Indo-Iranian religion… was also adopted, alongwith the Indo-European systems of ancestor worship.”26
In keeping with a pattern which will be familiar to anyone studying the writings of supporters of the Aryan invasion theory, such unnatural or anomalous phenomena do not make these scholars rethink their theory; it only makes them try to think of ways to maintain their theory in the face of inconvenient facts.
Witzel tries to suggest an explanation which he hopes will suffice to explain away the lack of archaeological-anthropological evidence: according to him, the original Indic racial stock had settled down in Central Asia, and had “even before their immigration into South Asia, completely ‘Aryanised’ a local population, for example, in the highly developed Turkmenian-Bactrian area… involving both their language and culture. This is only imaginable as the result of the complete acculturation of both groups… the local Bactrians would have appeared as a typically ‘Vedic’ people with a Vedic civilization.”27
These new “Vedic people” (ie. people belonging to the racial stock of the original non-Aryan inhabitants of Bactria, but with language, mythology and culture of the Indic people who had earlier migrated into Bactria from further outside) “later on… moved into the Panjab, assimilating (‘Aryanising’) the local population”.28
“By the time they reached the Subcontinent… they may have had the typical somatic characteristics of the ancient population of the Turanian/Iranian/Afghan areas, and may not have looked very different from the modem inhabitants of the Indo-Iranian Boderlands. Their genetic impact would have been negligible, and… would have been ‘lost’ in a few generations in the much larger gene pool of the Indus people. One should not, therefore, be surprised that ‘Aryan bones’ have not been found so far (Kennedy, this volume; Hemphill, Lukas and Kennedy, 1991).”29
What Witzel, like other scholars who suggest similar scenarios, is doing, is suggesting that the Aryans who migrated into India were not the original Indoaryans, but groups of people native to the areas further northwest, who were “completely Aryanised” in “language and culture”, and further that they were so few in number that “their genetic impact would have been negligible” and “would have been ‘lost’ in a few generations in the much larger gene pool of the Indus people”.
The scholars thus try to explain away the lack of archaeological-anthropological evidence by postulating a fantastic scenario which is totally incompatible with the one piece of solid evidence which is available to us today: THE RIGVEDA.
The Rigveda represents a language, religion and culture which is the most archaic in the Indo-European world. As Griffith puts it in his preface to his translation: “As in its original language, we see the roots and shoots of the languages of Greek and Latin, of Celt, Teuton and Slavonian, so the deities, the myths and the religious beliefs and practices of the Veda throw a flood of light upon the religions of all European countries before the introduction of Christianity. As the science of comparative philology could hardly have existed without the study of Sanskrit, so the comparative history of the religions of the world would have been impossible without the study of the Veda.”
Vedic mythology represents the most primitive form of Indo-European mythology: as Macdonell puts it, the Vedic Gods “are nearer to the physical phenomena which they represent, than the gods of any other Indo-European mythology”.30 Vedic mythology not only bears links with every single other Indo-European mythology, but is often the only link between any two of them (as we will see in Appendix Three, Chapter 10)
Does it appear that the Rigveda could be the end-product of a long process of migration in which the Indoaryans not only lost contact with the other Indo-European branches countless generations earlier in extremely distant regions, and then migrated over long periods through different areas, and finally settled down for so long a period in the area of composition of the Rigveda that even Witzel admits that “in contrast to its close relatives in Iran (Avestan, Old Persian), Vedic Sanskrit is already an Indian language”;31 but in which the people who composed the Rigveda were in fact not the original Indoaryans at all, but a completely new set of people who bore no racial connections at all with the original Indoaryans, and were merely the last in a long line of racial groups in a “gradual and complex” process in which the Vedic language and culture was passed from one completely different racial group to another completely different racial group like a baton in an “Aryanising” relay race from South Russia to India?
Clearly, the explanation offered by Witzel is totally inadequate, and even untenable, as an argument against the negative archaeological evidence.
I.B. The Linguistic Evidence
Erdosy speaks of the “disciplinary divide” between linguistics and archaeology.
And it is Michael Witzel whom Erdosy pits against the archaeologists whose papers are included in the volume: “Placed against Witzel’s contribution, the paper by J.Shaffer and D. Lichtenstein will illustrate the gulf still separating archaeology and linguistics.”32
We will not assume that Witzel’s papers in this particular volume represent the sum total of the linguistic evidence, but, since the volume does pit him against the archaeologists, let us examine the linguistic evidence stressed by him.
According to Erdosy, “M. Witzel begins by stressing the quality of linguistic (and historical) data obtainable from the Rgveda, along with the potential of a study of linguistic stratification, contact and convergence. Next, the evidence of place-names, above all hydronomy, is scrutinised, followed by an evaluation of some of the most frequently invoked models of language change in light of this analysis.”33
We have already examined Witzel’s “models of language change” by which he seeks to explain away the lack of archaeological evidence. We will now examine “the evidence of place-names, above all hydronomy”, on the basis of which Witzel apparently contests the claims of the archaeologists and proves the Aryan invasion.
Witzel does not have much to say about place-names. He points out that most of the place-names in England (all names ending in -don, -chester, -ton, -ham, -ey, -wick, etc., like London, Winchester, Uppington, Downham, Westrey, Lerwick, etc.) and in America (like Massachussetts, Wachussetts, Mississippi, Missouri, Chicago, etc) are remnants of older languages which were spoken in these areas.
But, far from finding similar evidence in respect of India, Witzel is compelled to admit: “In South Asia, relatively few pre-Indo-Aryan place-names survive in the North; however, many more in central and southern India. Indo-Aryan place-names are generally not very old, since the towns themselves are relatively late.”34
Witzel clearly evades the issue: he refers to “relatively few pre-Indo-Aryan place names” in the North, but judiciously refrains from going into any specifics about these names, or the number of such names.
He insinuates that there are “many more” pre-Indoaryan place-names in Central and South India, but this is clearly a misleading statement: by Central India, he obviously means the Austric-language speaking areas, and by South India, he definitely means the Dravidian-language speaking areas, and perhaps other areas close to these. So, if these areas have Austric or Dravidian place-names respectively, does it prove anything?
And, finally, he suggests that the paucity (or rather absence) of any “pre-Indo-Aryan” place-names in the North is because the towns concerned “are relatively late” (ie. came into being after the Aryan influx). This excuse is rather strange: the Indus people, alleged to be “pre-Indo-Aryans” did have towns and cities, but no alleged earlier place-names have survived, while the American Indians (in the U.S.A.) did not have large towns and cities, but their place-names have survived in large numbers.
Witzel goes into more detail in respect of the hydronomes (ie. names of rivers), but the results of his investigation, and even his own comments on them, are intriguing.
According to Witzel: “A better case for the early linguistic and ethnic history of South Asia can be made by investigating the names of rivers. In Europe river-names were found to reflect the languages spoken before the influx of Indo-European speaking populations. They are thus older than c. 4500-2500 BC (depending on the date of the spread of Indo-European languages in various parts of Europe). It would be fascinating to gain a similar vantage point for the prehistory of South Asia.”35
It is indeed fascinating. Witzel finds, to his chagrin, that “in northern India, rivers in general have early Sanskrit names from the Vedic period, and names derived from the daughter languages of Sanskrit later on.”36
Witzel tries to introduce the non-Aryan element into the picture: “River names in northern India are thus principally Sanskrit, with few indications of Dravidian, MuNDa or Tibeto-Burmese names. However, Kosala, with its uncharacteristic -s- after -o- may be Tibeto-Burmese (Sanskrit rules would demand KoSala or KoSala, a corrected form that is indeed adopted in the Epics).”37 Likewise, “there has been an almost complete Indo-Aryanisation in northern India; this has progressed much less in southern India and in the often inaccessible parts of central India. In the northwest there are only a few exceptions, such as the names of the rivers GangA, SutudrI and perhaps KubhA (Mayrhofer, 1956-1976).”38
Thus, there are four river-names which he tries to connect with “pre-Indo-Aryan” languages. But three of them, Kosala, SutudrI and KubhA are clearly Indo-European names (the hairsplitting about the letter -s- in Kosala is a typical “linguistic” ploy which we will refer to later on in our examination of linguistic substrata), and only GaNgA is generally accepted as a possible non-Indo-European name.
But the answer to this is given by Witzel himself: “Rivers often carry different names, sometimes more than two, along their courses. Even in a homogenous, monolingual country, such as Japan, this can be the case as names change as soon as the river passes through a major mountain range. In South Asia, to quote one well-known example, the BhAgIrathI and AlaknandA become the GaNgA. This increases the probability of multiple names from various languages for one and the same river of which only one may have survived in our sources.”39 (It may be noted that the Rigveda itself refers to the river as both GaNgA and JahnAvI).
Witzel cannot escape the “evidence of hydronomy” as he calls it, and he tries to explain it away by suggesting that “there has been an almost complete Indo-Aryanisation”40 of the river-names in northern India.
But his explanation rings hollow: “The Indo-Aryan influence, whether due to actual settlement, acculturation, or, if one prefers, the substitution of Indo-Aryan names for local ones, was powerful enough from early on to replace local names, in spite of the well-known conservatism of river-names. This is especially surprising in the area once occupied by the Indus civilization, where one would have expected the survival of earlier names, as has been the case in Europe and the Near East. At the least, one would expect a palimpsest, as found in New England, with the name of the State of Massachussetts next to the Charles River formerly called the Massachussetts River, and such new adaptations as Stony Brook, Muddy Creek, Red River, etc. next to the adaptations of Indian names such as the Mississippi and the Missouri. The failure to preserve old hydronomes even in the Indus Valley (with a few exceptions noted above) indicates the extent of the social and political collapse experienced by the local population.”41
Apart from anything else, does this last bit at all harmonize with the claim made elsewhere in the same volume (to explain the lack of archaeological-anthropological evidence of any invasion) that the “Indo-Aryanisation” of the northwest was a “gradual and complex” rather than a “cataclysmic” event?
Witzel starts out with the intention of pitting the linguistic evidence of place-names and river-names against the evidence of archaeology; and he ends up having to try and argue against, or explain away, this linguistic evidence, since it only confirms the archaeological evidence.
The long and short of the evidence of place-names and river-names is as follows:
The place-names and river-names in Europe, to this day, represent pre-Indo-European languages spoken in Europe before 2500 BC. The same is the case with Armenia: “among the numerous personal and place-names handed down to us from Armenia up to the end of the Assyrian age, there is absolutely nothing Indo-European.”42 And with Greece and Anatolia: “numerous place-names… show that Indo-Europeans did not originate in Greece. The same can be said for Italy and Anatolia.”43
On the other hand, northern India is the only place where place-names and river-names are Indo-European right from the period of the Rigveda (a text which Max Müller refers to as “the first word spoken by the Aryan man”) with no traces of any alleged earlier non-Indo-European names.
Witzel’s attitude towards this evidence is typical of the generally cavalier attitude of Western scholars towards inconvenient evidence in the matter of Indo-European origins: he notes that the evidence is negative, finds it “surprising” that it should be so, makes an offhand effort to explain it away, and then moves on.
And, later on, in his second paper included in the volume, he actually refers complacently to the whole matter: “in view of the discussion of hydronomy and place-names in the previous paper, it is also interesting that the Indo-Aryans could not, apparently, pronounce local names.”44
But, like it or not, the
evidence of place-names and river-names is a very important factor in
locating the Indo-European homeland in any particular area. And India,
and India alone, passes this test with flying colours.
We have already examined the evidence in the Rigveda which clearly proves that the original Indo-Iranian habitat was in India and that the Iranians migrated westwards and northwestwards from India.
We will now examine further literary evidence regarding the location of the original Indo-European homeland in India, under the following heads:
A. Tribes and Priests.
II.A. Tribes and Priests
The political history of the Vedic period is centred around the division of the various peoples who fall within its ambit into five major tribal groupings (not counting the TRkSis, who fall outside this tribal spectrum): the Yadus, TurvaSas, Anus, Druhyus and PUrus.
As we have seen, it is only one of these five tribal groupings, the PUrus, who represent the various branches of the Vedic Aryans, and it is only the PUrus who are referred to as Aryas in the Rigveda.
This brings us to the second division of the various peoples who fall within the ambit of the Rigveda: the division into Aryas (the PUrus) and Others (the Yadus, TurvaSas, Anus, Druhyus, etc.)
But there are two distinct words by which the Rigveda refers to these Others:
It is necessary to understand the distinction between the two words.
The word DAsa is found in 54 hymns (63 verses):
I. 32.11; 92.8; 103.3; 104.2; 158.5; 174.7;
The word Dasyu is found in 65 hymns (80 verses):
I. 33.4, 7, 9; 36.18; 51.5, 6, 8; 53.4; 59.6;
There are two distinct differences between the DAsas and Dasyus:
1. The first difference is that the term DAsa clearly refers to other tribes (ie. non-PUru tribes) while the term Dasyu refers to their priestly classes (ie. non-Vedic priestly classes).
[This is apart from the
fact that both the terms are freely used to refer to the atmospheric demons
as much as to the human enemies to whom they basically refer.]:
a. According to IV. 28.4, the Dasyus are a section among the DAsas.
b. The Dasyus are referred to in terms which clearly show that the
causes of hostility are religious:
Not one of these abuses is used even once in reference to DAsas.
c. The family-wise pattern of references to them also shows that the Dasyus are priestly rivals while the DAsas are secular rivals.
The Dasyus are referred to by all the nine priestly families of RSis, but not by the one non-priestly family of RSis (the Bharatas).
The DAsas are referred to by the Bharatas (X.69.6; 102.3) also; but not by the most purely ritualistic family of RSis, the KaSyapas, nor in the most purely ritualistic of MaNDalas, MaNDala IX.
d. The Dasyus, being priestly entities, do not figure as powerful persons or persons to be feared, but the DAsas, being secular entities (tribes, tribal warriors, kings, etc.) do figure as powerful persons or persons to be feared:
In three references (VIII.5.31; 46.32; 51.9), the DAsas are rich patrons.
In seven references, the DAsas are powerful enemies from whose fury and powerful weapons the composers ask the Gods for protection (I.104.2; VIII.24.27; X.22.8; 54.1; 69.6; 102.3) or from whom the Gods rescue the RSis (I.158.5). In three others, the word DAsa refers to powerful atmospheric demons who hold the celestial waters in their thrall (I.32.11; V.30.5; VIII.96.18).
In contrast, Dasyus never figure as rich or powerful enemies. They are depicted as sly enemies who incite others into acts of boldness (VI.24.8).
e. While both DAsas and Dasyus are referred to as enemies of the Aryas, it is only the DAsas, and never the Dasyus, who are sometimes bracketed together with the Aryas.
Seven verses refer to both Aryas and DAsas as enemies (VI.22.10; 33.3; 60.6; VII.83.1; X.38.3; 69.6; 83.1; 102.3) and one verse refers to both Aryas and DAsas together in friendly terms (VIII.51.9).
This is because both, the word DAsa and the word Arya, refer to broad secular or tribal entities, while the word Dasyu refers to priestly entities: thus, one would generally say “both Christians and Muslims”, or “both padres and mullahs”, but not “both Christians and mullahs” or “both Muslims and padres”.
2. The second difference is in the degree of hostility towards the two. The Dasyus are clearly regarded with uncompromising hostility, while the hostility towards the DAsas is relatively mild and tempered:
a. The word Dasyu has a purely hostile connotation even when it occurs in the name or title of heroes:
Trasadasyu = “tormentor of the Dasyus”.
On the other hand, the word DAsa has an etymological meaning beyond the identity of the DAsas. When it occurs in the name or title of a hero, it has a benevolent connotation:
DivodAsa = “light of Heaven” or “slave of Heaven”.
b. All the 80 verses which refer to Dasyus are uncompromisingly hostile.
On the other hand, of the 63 verses which refer to DAsas, 3 are friendly references (VIII.5.31; 46.32; 51.9); and in one more, the word means “slave” in a benevolent sense (VII.86.7: “slave-like, may I do service to the Bounteous”, ie. to VaruNa).
c. Of the 80 verses which refer to Dasyus, 76 verses talk of direct, violent, physical action against them, ie. they talk of killing, subduing or driving away the Dasyus.
On the other hand, of the 63 verses which refer to DAsas, only 38 talk of such direct physical action against them.
The importance of this analysis is that it brings to the fore two basic points about the rivalries and hostilities in the Rigvedic period:
a. The rivalries or hostilities were on two levels: the secular level and the priestly level.
b. The rivalries on the priestly level were more sharp and uncompromising.
Hence, any analysis of the political history of the Rigvedic period must pay at least as much attention, if not more, to the priestly categories as to secular or tribal categories.
II.B. The Three Priestly Classes
The basic tribal spectrum of the Rigveda includes the five tribal groupings of Yadus, TurvaSas, Anus, Druhyus and PUrus, and of these the PUrus alone represent the Vedic Aryans, while the other four represent the Others.
But among these four it is clear that the Yadus and TurvaSas represent more distant tribes (they are, as we have seen earlier, mostly referred to in tandem, and are also referred to as residing far away from the Vedic Aryans), while the Anus and Druhyus fall into a closer cultural spectrum with the PUrus:
a. In the PurANas, the Yadus and TurvaSas are classified together as descendants of sons of DevayAnI, and the Anus, Druhyus and PUrus are classified together as descendants of sons of SarmiSThA.
b. The geographical descriptions of the five tribes, as described in the PurANas, place the Yadus and TurvaSas together in the more southern parts (of northern India), and the Anus, Druhyus and PUrus together in the more northern parts.
c. The Rigveda itself, where it refers to the five tribes together (I.108.8) refers to the Yadus and the TurvaSas in one breath, and the Druhyus, Anus and PUrus in another: “yad IndrAgni YaduSu TurvaSeSu, yad DruhyuSu AnuSu PUruSu sthaH”.
But, the PUrus represent the various branches of the Vedic Aryans, and the Anus represent various branches of Iranians. It is clear, therefore, that the Druhyus represent the third entity in this cultural spectrum, and that it is mainly the Druhyus who will take us beyond the Indo-Iranian arena into the wider Indo-European one: appropriately, while the PUrus are located in the heartland of North India (U.P.-Delhi-Haryana) and the Anus in the northwest (Punjab), the Druhyus are located beyond the Indian frontiers, in Afghanistan and beyond.
The priestly categories, as we have seen, play a more important role in the rivalries and hostilities in the Rigvedic period than the secular categories.
In the earliest period, the only two families of RSis (from among the families who figure as composers in the Rigveda) were the ANgirases and the BhRgus, who were the priests of the PUrus and the Anus respectively. Logically, there must have been a priestly class among the Druhyus as well, but no such priestly class figures among the composers in the Rigveda.
The explanation for this is simple: the Druhyus were a rival and non-PUru (DAsa) tribe, hence their priests do not figure as composers in the Rigveda. Of course, the BhRgus, who were also the priests of a rival and non-PUru tribe, do figure as composers in the Rigveda, but that is because, as we have seen in the previous chapter, a section of BhRgus (after Jamadagni) aligned themselves with the Vedic Aryans and joined the Vedic mainstream (where, in fact, they later superseded all the other priestly families in importance, and became the dominant priests of Vedic tradition).
But since the Druhyus figure in the Rigveda, the name of their priestly class must also be found in the text, even if not as the name of a family of composers.
Since no such name appears, it seems logical that the name Druhyu itself must originally have been the name of this third priestly class: since priestly categories were more important for the composers of the Rigveda than the secular categories, and since the tribes for whom the Druhyus functioned as priests were an amorphous lot located far out on the frontiers of India and beyond, the name of the priestly classes became a general appellation for the tribes themselves.
Therefore, there were three tribal groupings with their three priestly classes:
PUrus - Angirases.
This trinary situation tallies with the Indo-European situation: outside of the Vedic and Iranian cultures, the only other priestly class of a similar kind is found among the Celts and the related Italics. While the Italics called their priests by the general name flAmen (cognate to Sanskrit brAhmaNa, “priest”), the priests of the Celts were called Drui (genitive Druad, hence Druids).
Shan M.M. Winn notes that “India, Rome, Ireland and Iran” are the “areas in which priesthoods are known to have been significant”;45 and he describes this phenomenon as follows: “Long after the dispersion of Indo-Europeans, we find a priestly class in Britain in the west, in Italy to the South, and in India and Iran to the east. Though these cultures are geographically distant from one another... they have striking similarities in priestly ritual, and even in religious terminology. For example, taboos pertaining to the Roman flAmen (priest) closely correspond to the taboos observed by the Brahmans, the priests of India.”46 Like the Indian priesthood, the curriculum of the “Celtic Druids … involved years of instruction and the memorization of innumerable verses, as the sacred tradition was an oral one”.47
After noting, in some detail, the similarities in their priestly systems, rituals, and religious and legal terminology, Winn concludes that the “Celts, Romans and Indo-Iranians shared a religious heritage dating to an early Indo-European period…”48
While the three priesthoods flourished only in these areas, they must originally have been the priests of all the branches of Indo-Europeans in the early Indo-European period. While the priesthoods themselves did not survive elsewhere, the names of the three priesthoods did survive in different ways. An examination of these words helps us to classify the various Indo-European branches into three groups:
1. PURUS: Indoaryan.
In the Rigveda, hymn VII.18, the DASarAjña battle hymn, refers to the enemy confederation once in secular (tribal) terms as “Anus and Druhyus” (VII.18.14), and once in what is clearly priestly terms as “BhRgus and Druhyus” (VII.18.6: the only reference in the whole of the Rigveda which directly refers to the BhRgus as enemies). Once, it may be noted, it also refers to the kings of the two tribal groupings as “KavaSa and the Druhyu” (VII. 1.8.12. Thus, even here, the general appellation “Druhyu” is used instead of the specific name of the king of the Druhyus).
The words Druh/Drugh/Drogha occur throughout the Rigveda in the sense of “demon” or “enemy”. (The word BhRgu, for obvious reasons, does not suffer the same fate.)
2. ANUS: Iranian, Thraco-Phrygian, Hellenic.
a. Iranian: In the Avesta, in Fargard 19 of the VendidAd, it is an Angra (ANgiras) and a Druj (Druhyu) who try to tempt Zarathushtra away from the path of Ahura Mazda.
The priests of the Iranians were the Athravans (AtharvaNas = BhRgus), and the words Angra and Druj occur throughout the Avesta as epithets for the demon enemies of Ahura Mazda and Zarathushtra.
b. Thraco-Phrygian: While the Armenians, the only surviving members of this branch, have not retained any tradition about any of these priestly classes, it is significant that one of the most prominent groups, belonging to this branch, were known as the Phryge (BhRgu).
c. Hellenic: The fire-.priests of the Greeks were known as the Phleguai (BhRgu).
What is more, Greek mythology retains memories of both the other priestly classes, though not in a hostile sense, as the names of mythical beings: Angelos (ANgiras) or divine messengers, and Dryad (Druhyu) or tree-nymphs.
3. DRUHYUS: Baltic and Slavonic, Italic and Celtic, Germanic.
a. Baltic and Slavonic: The word Druhyu occurs in the languages of these two branches in exactly the opposite sense of the Vedic Druh/Drugh/Drogha and the Iranian Druj. In Baltic (eg. Lithuanan Draugas) and Slavonic (eg. Russian Drug) the word means “friend”.
b. Italic and Celtic: While the Italic people did not retain the name of the priestly class (and called their priests flAmen = BrAhmaNa), the Celtic priests, as we have seen, were called the Drui (genitive Druad, hence Druid).
A significant factor, showing that the Celtic priests must have separated from the other priestly classes before the priestly hostilities became intense, is that the BhRgus appear to be indirectly remembered in Celtic mythology in a friendly sense.
The Larousse Encyclopaedia of Mythology notes: “whereas the Celtic Gods were specifically Celtic… the goddesses were restatements of an age-old theme”.49 And two of the three Great Goddesses of the Celts were named Anu and Brigit (Anu and BhRgu?). And while all the Goddesses in general were associated with fertility cults, “Brigit, however, had additional functions as a tutelary deity of learning, culture and skills”.50
The main activity of the Drui, as we have seen, was to undergo “years of instruction and the memorization of innumerable verses, as the sacred tradition was an oral one”.51 The fact that the Goddess of learning was named Brigit would appear to suggest that the Drui remembered the ancient BhRgus, in a mythical sense, as the persons who originally introduced various priestly rituals among them (a debt which, as we have seen in the previous chapter, is also remembered by the. ANgirases in the MaNDalas of the Early Period of the Rigveda). The BhRgus, by the joint testimony of Vedic and Celtic mythology, would thus appear to have been the oldest or most dominant and innovative of the three priestly classes.
c. Germanic: The word Druhyu occurs in the Germanic branch as well. However the meaning (although the words are cognate52 to the Russian Drug and Lithuanian Draugas) is more militant: Gothic driugan, “do military service” and ga-drauhts, “soldier”; and Old Norse (Icelandic) drOtt, Old English dryht and Old German truht, all meaning “multitude, people, army”.
The meanings of the word Druhyu as it occurs in the Celtic branch (“priest”), the Germanic branch (“soldier”, etc. or “people”) and the Baltic-Slavonic branches (“friend”) clearly correspond with the word in the Rigveda and Avesta, where Druhyu/Druh/Drugh/Drogha and Druj represent enemy priests, soldiers or people.
Thus, to sum up:
1. PUru (priests ANgirases): Indoaryan.
2. Anu (priests BhRgus/AtharvaNas): Iranian, Thraco-Phrygian, Hellenic.
3. Druhyu (priests Druhyus): Celtic-Italic,
II.C. The Anu-Druhyu Migrations
The evidence of the Rigveda, and Indian tradition, clearly shows that the Anus and Druhyus were Indian tribes.
If they were also the ancestors of the Indo-European branches outside India, as is indicated by the evidence of the names of their priestly classes, then it is clear that the Rigveda and Indian tradition should retain memories of the migrations of these two groups from India.
Significantly, this is exactly the case: the Rigveda and the PurANas, between them, record two great historical events which led to the emigration of precisely these two tribes from India:
1. The first historical emigration recorded is that of the Druhyus. This emigration is recorded in the PurANas, and it is so historically and geographically specific that no honest, student of the Puranic tradition has been able to ignore either this event or its implications for Indo-European history (even without arriving at the equation PUrus = Vedic Aryans):
The PurANas (VAyu 99.11-12; BrahmANDa III.74.11-12; Matsya 48.9; ViSNu IV.17.5; BhAgavata IX.23.15-16) record: PracetasaH putra-Satam rAjAnAH sarva eva te, mleccha-rASTrAdhipAH sarve hyudIcIm diSam ASritAH.
As Pargiter points out: “Indian tradition knows nothing of any Aila or Aryan invasion of India from Afghanistan, nor of any gradual advance from thence eastwards.”53 On the contrary, “Indian tradition distinctly asserts that there was an Aila outflow of the Druhyus through the northwest into the countries beyond where they founded various kingdoms.”54
P.L. Bhargava also notes this reference to the Druhyu emigration: “Five PurANas add that Pracetas’ descendants spread out into the mleccha countries to the north beyond India and founded kingdoms there.”55
This incident is considered to be the earliest prominent historical event in traditional memory: The Druhyus, inhabitants of the Punjab, started conquering eastwards and southwards, and their conquest brought them into conflict with all the other tribes and peoples: the Anus, PUrus, Yadus. TurvaSas, and even the IkSvAkus.
This led to a concerted attempt by the other tribes against the Druhyus. AD Pusalker records: “As a result of the successful campaigns of SaSabindu, YuvanASva, MAndhAtRI and Sibi, the Druhyus were pushed back from RAjputAna and were cornered into the northwestern portion of the Punjab. MAndhAtRI killed their king ANgAra, and the Druhyu settlements in the Punjab came to be known as GAndhAra after the name of one of ANgAra’s successors. After a time, being overpopulated, the Druhyus crossed the borders of India and founded many principalities in the Mleccha territories in the north, and probably carried the Aryan culture beyond the frontiers of India.”56
This first historical emigration represents an outflow of the Druhyus into the areas to the north of Afghanistan (ie. into Central Asia and beyond).
2. The second historical emigration recorded is that of the Anus and the residual Druhyus, which took place after the DASarAjña battle in the Early Period of the Rigveda.
As we have already seen in our chapter on the Indo-Iranian homeland, the hymns record the names of ten tribes (from among the two main tribal groupings of Anus and Druhyus) who took part in the confederacy against SudAs.
Six of these are clearly purely Iranian peoples:
a. PRthus or PArthavas (VII.83.1): Parthians.
One more Anu tribe, not named in the Rigveda, is that of the Madras: Medes.
All these Iranian peoples are found in later historical times in the historical Iranian areas proper: Iran, Afghanistan, Central Asia.
Two of the other tribes named in the hymns are Iranian peoples who are found in later historical times, on the northwestern periphery of the Iranian areas, ie. in the Caucasus area:
a. Simyus (VII.18.5): Sarmatians (Avesta = Sairimas).
b. Alinas (VII.18.7): Alans.
And the name of one more tribe is clearly the name of another branch of Indo-Europeans - non-Iranians, but closely associated with the Iranians - found in later historical times in the area to the west of the Iranians, ie. in Anatolia or Turkey: the BhRgus (VII.18.6): Phrygians.
Significantly, the names of the two tribes found on the northwestern periphery of the Iranian area are also identifiable (as we have noted in our earlier book) with the names of two other branches of Indo-Europeans, found to the west of Anatolia or Turkey.
a. Simyus (VII.18.5): Sirmios (ancient Albanians).
The DASarAjña battle hymns record the emigration of these tribes westward from the Punjab after their defeat in the battle.
Taken together, the two emigrations provide us with a very logical and plausible scenario of the expansions and migrations of the Indo-European family of languages from an original homeland in India:
1. The two tribal groupings of Anus and Druhyus were located more or less in the Punjab and Afghanistan respectively after the Druhyu versus non-Druhyu wars in the earliest pre-Rigvedic period.
2. The first series of migrations, of the Druhyus, took plate shortly afterwards, with major sections of Druhyus migrating northwards from Afghanistan into Central Asia in different waves. From Central Asia many Druhyu tribes, in the course of time, migrated westwards, reaching as far as western Europe.
These migrations must have included the ancestors of the following branches (which are not mentioned in the DASarAjña battle hymns):
3. The second series of migrations of Anus and Druhyus, took place much later, in the Early Period of the Rigveda, with various tribes migrating westwards from the Punjab into Afghanistan, many later on migrating further westwards as far as West Asia and southwestern Europe.
These migrations must have included the ancestors of the following branches (which are mentioned in the DASrAjña battle hymns):
The whole process gives a clear picture of the ebb-and-flow of migratory movements, where remnants of migrating groups, which remain behind, get slowly absorbed into the linguistic and cultural mainstream of the other groups among whom they continue to live, retaining only, at the most, their separate names and distinctive identities:
1. The Druhyus, by and large, spread out northwards from northwestern Punjab and Afghanistan into Central Asia (and beyond) in the first Great Migration.
A few sections of them, who remained behind, retained their distinctive names and identities (as Druhyus), but were linguistically and culturally absorbed into the Anu mainstream.
2. The Anus (including the remnants of the Druhyus), by and large, spread out westwards from the Punjab into Afghanistan in the second Great Migration after the DASarAjña battle.
A few sections of them, who remained behind, retained their distinctive names and identities (as Anus), but linguistically and culturally, they were absorbed into the PUru mainstream and they remained on the northwestern periphery of the Indoaryan cultural world as the Madras (remnants of the Madas or Medes), Kekayas, etc.
3. Further migrations took place from among the Anus in Afghanistan, with non-Iranian Anu groups, such as the BhRgus (Phryges, Thraco-Phrygians), Alinas (Hellenes, Greeks) and Simyus (Sirmios, Illyrians or Albanians) migrating westwards from Afghanistan as far as Anatolia and southeastern Europe.
A few sections of these non-Iranian Anus, who remained behind, retained their distinctive names and identities, but, linguistically and culturally, they were absorbed into the Iranian mainstream, and remained on the northwestern periphery of the Iranian cultural world as the Armenians (who, however, retained much of their original language, though greatly influenced by Iranian), and the Alans (remnants of the Hellenes or Greeks) and Sarmations (remnants of the Sirmios or Albanians).
The literary evidence of the Rigveda, thus, provides us with a very logical and plausible scenario of the schedule and process of migrations of the various Indo-European branches from India.
At this point, we may recall the archaeological evidence in respect of Europe, already noted by us. As we have seen, the Corded Ware culture (Kurgan Wave # 3) expanded from the east into northern and central Europe, and the “territory inhabited by the Corded Ware/Battle Axe culture, after its expansions, qualifies it to be the ancestor of the Western or European language branches: Germanic, Baltic, Slavic, Celtic and Italic”.57
The origins of the Kurgan culture have been traced as far east as Turkmenistan in 4500 BC.
This fits in perfectly with our theory that the seven branches of Indo-Europeans, not mentioned in the DASarAjña hymns, migrated northwards into Central Asia during the first Great Migration. Five of these, the five European branches mentioned above, later migrated westwards into Europe, while the other two, Hittite and Tocharian, remained behind in parts of Central Asia till the Hittites, at a much later date, migrated southwestwards into Anatolia.
These two branches, which remained behind in Central Asia, it is possible, retained contact with the Indoaryans and Iranians further south: the fact that Hittite mythology is the only mythology, outside the Indo-Iranian cultural world, which mentions Indra (as Inar) may be evidence of such contacts.
Even more significant, from the viewpoint of literary evidence, is the fact that Indian tradition remembers two important peoples located to the north of the Himalayas who are called the Uttarakurus and the Uttaramadras: “The Uttarakurus alongwith the Uttaramadras, are located beyond the HimAlayas. Though regarded as mythical in the epic and later literature, the Uttarakurus still appear as a historical people in the Aitareya BrAhmaNa (VII.23).”58
It is possible that the Uttarakurus and the Uttaramadras were the Tocharian (Uttarakuru = Tokhri) and Hittite branches of Indo-Europeans located to the north of the Himalayas.
The scenario we have
reconstructed from the literary evidence in the Rigveda fits in perfectly
with the linguistic scenario of the migration schedule of the various
Indo-European branches, as reconstructed by the linguists from the evidence
of isoglosses, which we will now be examining.
One linguistic phenomenon which is of great help to linguists in their efforts to chalk out the likely scenario of the migration schedule of the various Indo-European branches from the original homeland, is the phenomenon of linguistic isoglosses.
A linguistic isogloss is a linguistic feature which is found in some of the branches of the family, and is not found in the others.
This feature may, of course, be either an original feature of the parent Proto-Indo-European language which has been lost in some of the daughter branches but retained in others, or a linguistic innovation, not found in the parent Proto-Indo-European language, which developed in some of the daughter branches but not in the others. But this feature is useful in establishing early historico-geographical links between branches which share the same isogloss.
We will examine the evidence of the isoglosses as follows:
A. The Isoglosses
III.A. The Isoglosses
There are, as Winn points out, “ten ‘living branches’… Two branches, Indic (Indo-Aryan) and Iranian dominate the eastern cluster. Because of the close links between their classical forms - Sanskrit and Avestan respectively - these languages are often grouped together as a single Indo-Iranian branch.”59 But Meillet notes: “It remains quite clear, however, that Indic and Iranian evolved from different Indo-European dialects whose period of common development was not long enough to effect total fusion.”60
Besides these ten living branches, there are two extinct branches, Anatolian (Hittite) and Tocharian.
Of these twelve branches, one branch, Illyrian (Albanian), is of little use in this study of isoglosses: “Albanian… has undergone so many influences that it is difficult to be certain of its relationships to the other Indo-European languages.”61
An examination of the isoglosses which cover the other eleven branches (living and extinct) gives a more or less clear picture of the schedule of migrations of the different Indo-European branches from the original homeland.
Whatever the dispute about the exact order in which the different branches migrated away from the homeland, the linguists are generally agreed on two important points:
1. Anatolian (Hittite) was the first branch to leave the homeland: “The Anatolian languages, of which Hittite is the best known, display many archaic features that distinguish them from other Indo-European languages. They apparently represent an earlier stage of Indo-European, and are regarded by many as the first group to break away from the proto-language.”62
2. Four branches, Indic, Iranian, Hellenic (Greek) and Thraco-Phrygian (Armenian) were the last branches remaining behind in the original homeland after the other branches had dispersed:
“After the dispersals of the early PIE dialects,… there were still those who remained… Among them were the ancestors of the Greeks and Indo-Iranians…63
“Greek and Sanskrit share many complex grammatical features: this is why many earlier linguists were misled into regarding them as examples of the most archaic stage of Proto-Indo-European. However, the similarities between the two languages are now regarded as innovations that took place during a late period of PIE , which we call stage III. One of these Indo-Greek innovations was also shared by Armenian; all these languages it seems, existed in an area of mutual interaction.”64
Thus we get: “Greek Armenian, Phrygian, Thracian and Indo-Iranian. These languages may represent a comparatively late form of Indo-European, including linguistic innovations not present in earlier stages. In particular, Greek and Indic share a number of distinctive grammatical features……”65
The following are some of the innovations shared only by Indic, Iranian, Greek and Armenian (Thraco-Phrygian); features which distinguish them from the other branches, particularly the other living branches:
a. “The prohibitive negation *mE is attested only in Indo-Iranian (mA), Greek (mE) and Armenian (mi); elsewhere, it is totally lacking… and there is no difference in this respect between the ancient and modern stages of Greek, Armenian or Persian”;66 or, for that matter, sections of Indic (eg. the prohibitive negation mat in Hindi).
b. “In the formation of the Perfect also, there is a clear ‘distinction’ between Indo-Iranian and Armenian and Greek on the one hand, and all of the other languages on the other.”67
c. The “Indo-European voiceless aspirated stops are completely attested only in Indo-Iranian and Armenian… Greek… clearly preserves two of the three voiceless aspirated stops whose existence is established by the correspondence of Indo-Iranian and Armenian.”68 All the other branches show “complete fusion”69 of these voiceless aspirated stops.
d. “The suffix *-tero-, *-toro-, *-tro- serves in bell Indo-European languages to mark the opposition of two qualities, but only in two languages, Greek and Indo-Iranian, is the use of the suffix extended to include the formation of secondary adjectival comparatives… This development, by its very difference, points to the significance of the Greek and Indo-Iranian convergence… Armenian, which has a completely new formation, is not instructive in this regard.”70 But, “Latin, Irish, Germanic, Lithuanian and Slavic, on the other hand, borrow their secondary comparative from the original primary type.”71
e. “The augment is attested only in Indo-Iranian, Armenian and Greek; it is found nowhere else.”72 And it is “significant that the augment is not found in any of the other Indo-European languages… The total absence of the augment in even the earliest texts, and in all the dialects of Italic, Celtic, Germanic, Baltic and Slavic, is characteristic.”73
Hence, “the manner in which Italic, Celtic, Germanic, Baltic and Slavic eliminated the imperfect and came to express the preterite presupposes an original, Indo-European, absence of the augment throughout this group of languages. We thus have grounds for positing two distinct Indo-European dialect groups.”74
f. The division of the Indo-European branches into two distinct groups
is confirmed by what Meillet calls the Vocabulary of the Northwest: “There is
quite a large group of words that appear in the dialects of the North and
West (Slavic, Baltic, Germanic, Celtic and Italic) but
are not found in the others (Indic, Iranian, Armenian and Greek)… their
occurrence in the dialects of the North and West would indicate a cultural
development peculiar to the peoples who spread these dialects.”75
1. “Hittite, the first to separate itself, shares many isoglosses with Germanic and Tocharian.”76
2. “Celtic, Italic, Hittite, Tocharian and (probably) Phrygian share an interesting isogloss: the use of ‘r’ to indicate the passive forms of verbs. This feature… does not occur in any other Indo-European language.”77
3. Italic, Celtic, Germanic, Baltic and Slavonic, as we have seen, constitute one distinct group (in contradistinction to another distinct group consisting of Indic, Iranian, Armenian and Greek).
However, within themselves, these five branches link together as follows:
a. Italic and Celtic: “Comparative linguists have long been aware of the links between Italic and Celtic, which share a number of archaic features. These links suggest that the two branches developed together.”78 Among other things: “Vocabulary is identical in parts; this is true of some very important words, particularly prepositions and preverbs.”79
b. Baltic and Slavonic: “The general resemblance of Baltic and Slavic is so apparent that no-one challenges the notion of a period of common development… Baltic and Slavic are the descendants of almost identical Indo-European dialects. No important isogloss divides Baltic from Slavic… the vocabularies of Slavic and Baltic show numerous cognates - more precisely, cognates that are found nowhere else or cognates that in Baltic and Slavic have a form different from their form in other languages.”80
c. Italic, Celtic and Germanic: “The Germanic, Celtic and Italic idioms present… certain common innovational tendencies.”81 But, Italic apparently separated from the other two earlier: “Germanic, Celtic and Italic underwent similar influences. After the Italic-Celtic period, Italic ceased undergoing these influences and underwent others… Germanic and Celtic, remaining in adjacent regions, developed in part along parallel lines.”82
d. Germanic, Baltic and Slavonic: “Because Germanic shares certain important features with Baltic and Slavic, we may speculate that the history of the three groups is linked in some way.”83
To go into more precise detail: “The difference between a dative plural with *-bh-, eg. Skr.-bhyah, Av. -byO, Lat. -bus, O.Osc. -fs, O.Ir.-ib, Gr. -fi(n), and one with *-m-, eg. Goth. -m, O.Lith. -mus, Ol.Sl. -mU, is one of the first things to have drawn attention to the problem of Indo-European dialectology. Since it has been established, principally by A. Leskien, that there was no unity of Germanic, Baltic and Slavic postdating the period of Indo-European unity, the very striking similarity of Germanic, Baltic and Slavic which we observe here cannot… be explained except by a dialectical variation within common Indo-European.”84 It is, therefore, clear “that these three languages arose from Indo-European dialects exhibiting certain common features.”85
To sum up, we get two distinct groups of branches:
Group A: Hittite, Tocharian, Italic, Celtic, Germanic, Baltic, Slavonic.
Group B: Indic, Iranian, Thraco-Phrygian (Armenian), Hellenic (Greek).
No major isogloss cuts across the dividing line between the two groups to suggest any alternative grouping: the phenomenon of palatalization appears to do so, but it is now recognized as “a late phenomenon” which took place in “a post-PIE era in which whatever unity that once existed had broken down and most of the dialect groups had dispersed”,86 and we will examine the importance of this phenomenon later on.
Other similarities between languages or branches which lie on opposite sides of the above dividing line are recognizable as phenomena which took place after the concerned branches had reached their historical habitats, and do not, therefore, throw any light on the location of the original homeland or the migration-schedule of the branches.
The following are two examples of such similarities:
1. The Phrygian language appears to share the “r-isogloss” which is found only in the Hittite, Tocharian, Italic and Celtic branches. However:
a. The Phrygian language is known only from fragments, and many of the linguistic features attributed to it are speculative. About the “r-isogloss”, it may be noted, Winn points out that it is shared by “Celtic, Italic, Hittite, Tocharian and (probably) Phrygian”.87
b. Armenian, the only living member of the Thraco-Phrygian branch, does not share the “r-isogloss”, and nor did the ancient Thracian language.
c. The seeming presence of this isogloss in Phrygian is clearly due to the influence of Hittite, with which it shared its historical habitat: “Phrygian later replaced Hittite as the dominant language of Central Anatolia.”88
2. Greek and Italic alone share the change of Proto-Indo-European voiced aspirated stops (bh, dh, gh) into voiceless aspirated stops (ph, th, kh). Sanskrit is the only language to have retained the original voiced aspirated stops, while all the other branches, except Greek and Italic, converted them into unaspirated stops (b, d, g).
But this similarity between Greek and Italic is because “when Indo-European languages were brought to Mediterranean people unfamiliar with voiced aspirated stops, this element brought about the process of unvoicing”,89 and this change took place in the two branches “both independently and along parallel lines”.90 Hence, this is not an isogloss linking the two branches.
Therefore, it is clear that the two groups represent two distinct divisions of the Indo-European family.
III. B. The Homeland Indicated by the Isoglosses
The pattern of isoglosses shows the following order of migration of the branches of Group A:
Some of these branches share certain isoglosses among themselves which represent innovations which they must have developed in common after their departure from the original homeland, since the remaining branches (Indic, Iranian, Armenian and Greek) do not share these isoglosses.
This clearly indicates the presence of a secondary homeland, outside the exit-point from the original homeland, which must have functioned as an area of settlement and common development for the migrating branches.
The only homeland theory which fits in with the evidence of the isoglosses is the Indian homeland theory:
The exit-point for the migrating branches was Afghanistan, and these branches migrated towards the north from Afghanistan into Central Asia, which clearly functioned as the secondary homeland for emigrating branches.
As Winn points out: “Evidence from isoglosses… shows that the dispersal cannot be traced to one particular event; rather it seems to have occured in bursts or stages.”91
Hittite was the first to emigrate from Afghanistan into Central Asia, followed by Tocharian.
Italic-Celtic represented the next stage of emigration. The four branches developed the “r-isogloss” in common.
Germanic was the next branch to enter the secondary homeland, and it developed some isoglosses in common with Hittite and Tocharian.
The Baltic-Slavonic movement apparently represented the last major emigration. And its sojourn in the secondary homeland was apparently not long enough for it to develop any isoglosses in common with Hittite or Tocharian.
The five branches (Italic, Celtic, Germanic, Baltic and Slavonic, in that order) later moved further off, north-westwards, into the area to the north of the Caspian Sea, and subsequently formed part of the Kurgan III migrations into Europe. The Slavonic and Baltic branches settled down in the eastern parts of Europe, while the other three proceeded further into Europe. Later, the Italic branch moved towards the south, while the Germanic and Celtic branches moved to the north and west.
Meanwhile, the other branches (barring Indic), Greek Armenian and Iranian, as also, perhaps, the one branch (Illyrian or Albanian) which we have not taken into consideration so far, migrated westwards from India by a different and southern route.
The scholars, now, generally accept the evidence of the isoglosses, so far as it concerns the schedule of migrations of the different Indo-European branches from the original homeland, or the interrelationships between different branches. However, when it comes to determining the actual location of the original homeland, on the basis of this evidence, they abandon their objective approach and try to make it appear as if the evidence fits in with the particular homeland theory advocated by them, even when it is as clear as daylight that they are trying to fit a round peg into a square hole.
The homeland theory generally advocated by the scholars is the South Russian homeland theory. Shan M.M. Winn advocates the “Pontic-Caspian area” within this region as the particular location of the homeland.
An examination shows that the South Russian homeland theory (“Pontic-Caspian” or otherwise) is totally incompatible with the evidence of the isoglosses:
1. To begin with, it is clear that we have two distinct groups of branches, which we have already classified as Group A and Group B.
As per the evidence of the isoglosses, the branches in Group A are the branches which migrated away from the original homeland, and those in Group B are the branches which remained behind in the homeland after the other branches had departed.
At the same time, all the branches in Group A are found to the north of the Eurasian mountain chain (except for Hittite in Anatolia, but this branch is known to have migrated into Anatolia from the north-east), while all the branches in Group B are found to the south of the Eurasian mountain chain (the northernmost, Greek, is known to have migrated into southeastern Europe from the south-east).
The logical corollary should have been that the original homeland is also to the south of the Eurasian mountain chain, and that it is located in the historical habitat of one of the branches in Group B.
However, the scholars regularly advocate homeland theories which place the homeland in the area of one or the other of the branches in Group A.
2. The branches in Group A developed certain isoglosses in common after they had migrated away from the homeland. As we have pointed out, this makes it likely that there was a secondary homeland where they must have developed these isoglosses.
However, any homeland theory which locates the homeland in a central area, like South Russia or any area around it, makes the location of this secondary homeland a problem: the Tocharian branch is historically located well to the east of South Russia, the Hittite branch is located well to the south of South Russia, and the Germanic and Italic-Celtic branches are located well to the west of South Russia. It is difficult to think of a way in which all these branches could have moved together in one direction from South Russia before parting from each other and moving off in totally opposite directions.
It is perhaps to avoid this problem that Winn suggests that the isoglosses shared in common by these branches are not innovations developed by these branches in common, but archaic features which have been retained by otherwise separately migrating branches.
In respect of the r-isogloss, for example, Winn puts it as follows: “Celtic, Italic, Hittite, Tocharian, and (probably) Phrygian share an interesting isogloss: the use of ‘r’ to indicate the passive forms of verbs. This feature, which does not occur in any other Indo-European language, is probably an example of the ‘archaism of the fringe’ phenomenon. When a language is spread over a large territory, speakers at the fringe of that territory are likely to be detached from what goes on at the core. Linguistic innovations that take place at the core may never find their way out to peripheral areas; hence dialects .spoken on the fringe tend to preserve archaic features that have long since disappeared from the mainstream… Tocharian… was so remote from the center that it could hardly have taken part in any innovations.”92
However, it is more logical to treat this isogloss as an innovation developed in common by a few branches after their departure from the homeland, than to postulate that all the other, otherwise disparate, branches eliminated an original “use of ‘r’ to indicate the passive forms of verbs”.
3. What is indeed an example of the “archaism of the fringe” phenomenon is the phenomenon of palatalization.
Winn describes it as follows: “Palatalization must have been a late phenomenon; that is, we date it to a post-PIE era, in which whatever unity that once existed had now broken down, and most of the dialect groups had dispersed: looking at the geographical distribution of this isogloss, we may note its absence from the peripheral languages: Germanic (at the northwest limit of Indo-European language distribution); Celtic (western limit); Italic, Greek and Hittite (southern limit); and Tocharian (eastern limit). It is the languages at the center that have changed. Here, at the core, a trend towards palatalization started; then gradually spread outward. It never reached far enough to have any effect on the outlying languages.”93
Note that Winn calls it a “post-PIE era, in which whatever unity that once existed had now broken down, and most of the dialect groups had dispersed”, and that he locates every single other branch (except Indic and Iranian), including Greek, in its historical habitat. He does not specifically name Baltic-Slavonic and Armenian, but it is understood that they are also located in their historical habitats, since he implies that they are “the languages at the centre” (ie. languages in and around South Russia, which is, anyway, the historical habitat of these branches).
Indic and Iranian alone are not located by him in their historical habitats, since that would clearly characterize them as the most “peripheral” or “outlying” branches of all, being located at the extreme southern as well as extreme eastern limit of the Indo-European language distribution. And this would completely upset his pretty picture of an evolving “center” with archaic “outlying languages”, since the most outlying of the branches would turn out to be the most palatalized of them all. Hence, Winn without expressly saying so, but with such a location being implicit in his argument, locates all the other branches, including Greek, in their historical habitats, but only the Indic and Iranian branches well outside their historical habitats and still in South Russia, and keeps his fingers crossed over the possibility of the anomaly being noticed.
Here we see, once again, how the manipulation required to locate the Indo-European homeland in South Russia compels the scholars, again and again, to postulate weird and unnatural schedules of migrations which make the Indo-Iranians the last to leave South Russia, and which locate them in South Russia long after all the other branches, including Greek, are already settled in their historical habitats: a picture which clashes sharply with, among other things, the extremely representative nature of the Rigvedic language and mythology, the purely Indian geographical milieu of the Rigveda (and the movement depicted in it from east to west, as we have seen in this book), and the evidence of the names of places and rivers in northern India right from the period of the Rigveda itself.
The “late phenomenon” of a “trend towards palatalization” which started “at the core” and “then gradually -spread outward”, and “never reached far enough to have any effect on the outlying languages”, can be explained naturally only on the basis of the Indian homeland theory: the trend started in the “core area”, in north and northwest India, and spread outwards as far as the innermost of the branches in Group A: Baltic and Slavonic, but not as far as the outermost of the branches in Group B: Greek.
Incidentally, here is how Meillet94 depicts the interrelationships between the various extant branches (he does not include Hittite and Tocharian in the picture, but it is clear that they will fall in the same group as Germanic, Celtic and Italic). (Figure on next page.)
While the north-south axis clearly divides the non-palatalized branches in the west from the palatalized branches in the east (where we must locate the “core” area where palatalization started), the northeast-southwest axes neatly divide the branches into the three tribal groupings testified by Indian literary records, (click on next link).
4. More than anything else, the one aspect of the evidence of the isoglosses, which disproves the South Russian theory, is the close relationship between Indic or Indo-Iranian and Greek, which is not satisfactorily explained by any homeland theory other than the Indian homeland theory.
In dismissing Colin Renfrew’s Anatolian homeland theory, Winn cites this as the single most important factor in disproving the theory: “All the migrations postulated by Renfrew ultimately stem from a single catalyst: the crossing of Anatolian farmers into Greece… For all practical purposes, Renfrew’s hypothesis disregards Tocharian and Indo-Iranian.”95
Supporters of Renfrew’s theory, Winn points out, “have tried to render the Indo-Iranian problem moot. They argue that the Indo-Iranian branch was somehow divided from the main body of Proto-Indo-European before the colonists brought agriculture to the Balkans. Greek and Indic are thus separated by millenniums of linguistic change - despite the close grammatical correspondences between them (as we saw in Chapter 12, these correspondences probably represent shared innovations from the last stage of PIE).”96
Winn’s very valid argument against the Anatolian theory is just as applicable to the South Russian homeland theory, or any other theory which seeks to bring Indic and Iranian into their historical habitats through Central Asia: this involves an extremely long period of separation from Greek, which does not fit into the evidence of the isoglosses which shows that Indic and Greek have many “shared innovations from the last stage of PIE”.
Archaeology, for one, completely rules out any links between the alleged Proto-Indo-Iranians located by these scholars in Central Asia, and the Greeks: Winn, as we saw, tries to identify the Andronovo culture which “covers much of the Central Asian Steppe east of the Ural river and Caspian Sea”,97 with the “Proto-Indo-Iranians” during their alleged sojourn in Central Asia.
However, not only does he admit that “it is still a hazardous task to connect (this) archaeological evidence of Indo-Iranians in the Central Asian Steppe with the appearance of Iranian (Aryan) and Indic (Indo-Aryan) tribes in Iran, Afghanistan and India,”98 but he also accepts that these so-called Proto-Indo-Iranians in Central Asia have “no links with… south-eastern Europe”,99 ie. with the Greeks.
It is only the Indian homeland theory which fits in with the evidence of the isoglosses. It may be noted again that:
a. The evidence of the isoglosses suggests that the Indic, Iranian, Armenian and Greek branches, as well as the Albanian branch, were the last to remain behind in the original homeland after the departure of the other branches.
b. These (naturally, barring Indic) are also the same branches which show connections with the BhRgus/ AtharvaNas, while those which departed show connections with the Druhyus.
c. Again, all these branches form a long belt to the south of the Eurasian mountain chain, while the other (departed) branches are found to its north.
d. And, finally, these are the only branches which are actually recorded in the DASarAjña hymns as being present in the Punjab area during the time of SudAs.
We have, in our earlier book, examined the question of the historico-linguistic connections between Indo-European and other language families like Uralic and Semitic. These connections are projected by many scholars as linguistic evidence for the origin of the Indo-European family in or around South Russia, but the evidence, as we saw, fails to prove their point.
However, a more complex and scientific analysis of the linguistic connections between Indo-European and other families forms the subject of a paper by Johanna Nichols, entitled, significantly, The Epicentre of the Indo-European Linguistic Spread, which is part of a more detailed study contained in the two volumes of Archaeology and Language (of which the particular paper under discussion constitutes Chapter 8 of the first volume).
Nichols determines the location of “the epicentre of the Indo-European linguistic spread” primarily on the basis of an examination of loan-words from Mesopotamia and the Fertile Crescent of West Asia.
As she points out, loan-words from this region must have spread out via three trajectories (or routes):
“To Central Europe via the Bosporus and the Balkans, to the western steppe via the Caucasus... and eastward via Iran to western Central Asia…”100
“The first step in specifying a locus for the IE homeland is to narrow it down to one of these three trajectories, and that can be done by comparing areal Wanderwörter in the IE cultural vocabulary to those of other language families that can be located relative to one or another trajectory in ancient times.”101
Therefore, Nichols examines loan-words from West Asia (Semitic and Sumerian) found in Indo-European and in other families like Caucasian (separately Kartvelian, Abkhaz-Circassian and Nakh-Daghestanian), and the mode and form of transmission of these loan-words into the Indo-European family as a whole as well as into particular branches; and combines this with the evidence of the spread of Uralic and its connections with Indo-European.
After a detailed examination, her final conclusions about the locus or epicentre of the Indo-European linguistic spread are as follows: “Several kinds of evidence for the PIE locus have been presented here. Ancient loanwords point to a locus along the desert trajectory, not particularly close to Mesopotamia and probably far out in the eastern hinterlands. The structure of the family tree, the accumulation of genetic diversity at the western periphery of the range, the location of Tocharian and its implications for early dialect geography, the early attestation of Anatolian in Asia Minor, and the geography of the centum-satem split all point in the same direction: a locus in western central Asia. Evidence presented in Volume II supports the same conclusion: the long-standing westward trajectories of languages point to an eastward locus, and the spread of IE along all three trajectories points to a locus well to the east of the Caspian Sea. The satem shift also spread from a locus to the south-east of the Caspian, with satem languages showing up as later entrants along all three trajectory terminals. (The satem shift is a post-PIE but very early IE development). The locus of the IE spread was therefore somewhere in the vicinity of ancient Bactria-Sogdiana.”102
This linguistic evidence thus fits in perfectly with the literary and other evidence examined by us in this book, and with the theory outlined by us.
Nichols’ analysis lovers three concepts:
1. The Spread Zone: “The vast interior of Eurasia is a linguistic spread zone - a genetic and typological bottleneck where many genetic lines go extinct, structural types tend to converge, a single language or language family spreads out over a broad territorial range, and one language family replaces another over a large range every few millennia…”103
2. The Locus: “The locus is a smallish part of the range which functions in the same way as a dialect-geographical centre: an epicentre of sorts from which innovations spread to other regions and dialects, and a catchpoint at which cultural borrowings and linguistic loanwords entered from prestigious or economically important foreign societies to spread (along with native linguistic innovations) to the distant dialects. If an innovation arose in the vicinity of the locus, or a loanword entered, it spread to all or most of the family; otherwise, it remained a regionalism. Diversification of daughter dialects in a spread zone takes place far from the locus at the periphery, giving the family tree a distinctive shape with many major early branches, and creating a distinctive dialect map where genetic diversity piles up at the periphery. These principles make it possible to pinpoint the locus in space more or less accurately even for a language family as old as IE. Here it will be shown that the locus accounting for the distribution of loanwords, internal innovations and genetic diversity within IE could only have lain well to the east of the Caspian Sea.”104
As we have already seen, the specific location is “in the vicinity of Bactria-Sogdiana”.105
“The central Eurasian spread zone (Figure 8.4), as described in Volume II, was part of a standing pattern whereby languages were drawn into the spread zone, spread westward, and were eventually succeeded by the next spreading family. The dispersal for each entering family occurred after entry into the spread zone. The point of dispersal for each family is the locus of its proto-homeland, and this locus eventually is engulfed by the next entering language. Hence in a spread zone the locus cannot, by definition, be the point of present greatest diversity (except possibly for the most recent family to enter the spread zone). On the contrary, the locus is one of the earliest points to be overtaken by the next spread.”106
Further, “the Caspian Sea divides westward spreads into steppe versus desert trajectories quite close to the locus and hence quite early in the spread.”107
3. The Original Homeland: “Central Eurasia is a linguistic bottleneck, spread zone, and extinction chamber, but its languages had to come from somewhere. The locus of the IE spread is a theoretical point representing a linguistic epicentre, not a literal place of ethnic or linguistic origin, so the ultimate origin of PIE need not be in the same place as the locus. There are several linguistically plausible possibilities for the origin of Pre-PIE. It could have spread eastward from the Black Sea steppe (as proposed by Mallory 1989 and by Anthony 1991, 1995), so that the locus formed only after this spread but still very early in the history of disintegrating PIE… It could have come into the spread zone from the east as Mongolian, Turkic, and probably Indo-Iranian did. Or it could have been a language of the early urban oases of southern central Asia.”108
Thus, the linguistic evidence fully confirms our theory of an original homeland in India, an exit-point in Afghanistan, and two streams of westward emigration or expansion.
Nichols does not advocate an Indian homeland, but:
a. She does accept that the Pre-PIE language could have come from any direction (east or west), or could have been native to south Central Asia (Bactria-Sogdiana) itself, since the linguistic data only accounts for the later part of the movement, and not the earlier one.
b. The later part of the movement, indicated by the linguistic data, is in the opposite direction (ie. away from India).
c. The literary evidence, as we have seen in this book, provides the evidence for the earlier part of the movement.
Nichols’ analysis of the linguistic data, moreover, produces a picture which is more natural, and more compatible with what may be called “linguistic migration theory”:
“As defined by Dyen (1956), a homeland is a continuous area and a migration is any movement causing that area to become non-continuous (while a movement that simply changes its shape or area is an expansion or expansive intrusion). The linguistic population of the homeland is a set of intermediate protolanguages, the first-order daughters of the original protolanguage (in Dyen’s terms, a chain of coordinate languages). The homeland is the same as (or overlaps) the area of the largest chain of such co-ordinates, i.e. the area where the greatest number of highest-level branches occur. Homelands are to be reconstructed in such a way as to minimize the number of migrations, and the number of migrating daughter branches, required to get from them to attested distributions (Dyen 1956: 613).”109
The theories which place the original homeland in South Russia postulate a great number of separate emigrations of individual branches in different directions: Hittite and Tocharian would be the earliest emigrants in two different and opposite directions, and Indo-Iranian, Armenian and Greek would be the last emigrants, again, in three different and opposite directions.
But the picture produced by the evidence analysed by Nichols is different: “no major migrations are required to explain the distribution of IE languages at any stage in their history up to the colonial period of the last few centuries. All movements of languages (or more precisely all viable movements - that is, all movements that produced natural speech communities that lasted for generations and branched into dialects) were expansions, and all geographically isolated languages (eg. Tocharian, Ossetic in the Caucasus, ancestral Armenian, perhaps ancestral Anatolian) appear to be remnants of formerly continuous distributions. They were stranded by subsequent expansions of other language families, chiefly Turkic in historical times.”110
It must be noted that the picture produced by the linguistic evidence analysed by Nichols fits in perfectly with the Indian homeland theory derived from our analysis of the literary evidence, but Nichols is not herself a supporter of the Indian homeland theory, and this makes her testimony all the more valuable.
Nichols suggests that there was a point of time during the expansion of the Indo-Europeans when “ancestral Proto-Indo-Aryan was spreading into northern India,”111 and that “the Indo-Iranian distribution is the result of a later, post-PIE spread”.112
How far does this fit in with the evidence analysed by Nichols?
The evidence primarily shows two things:
a. “The long-standing westward trajectories of languages point to an eastward locus, and the spread of IE along all these trajectories point to a locus well to the east of the Caspian Sea.”113
b. “The locus of the IE spread was therefore somewhere in the vicinity of ancient Bactria-Sogdiana.”114
The evidence shows “westward trajectories of languages” from a locus “in the vicinity of ancient Bactria-Sogdiana,” it does not show eastward or southward trajectories of languages from this locus.
Therefore, while Nichols’ conclusion, that the Indo-European languages found to the west of Bactria-Sogdiana, were the results of expansions from Bactria-Sogdiana are based on linguistic evidence, her conclusion that the Indo-European languages found to the south and east of Bactria-Sogdiana were also the results of expansions from Bactria-Sogdiana, are not based on linguistic evidence, but on a routine application of the dictum “what is sauce for the goose is sauce for the gander”. Also, perhaps, Nichols, who has no particular reason to believe that India could be the original homeland, finds no reason to go much further than is absolutely necessary in challenging established notions: as it is, she is conscious that the locus indicated by the linguistic evidence “is unlike any other proposed homeland”,115 and, therefore, she probably sees no reason to make it so unlike as to be provocative.
But the Indian homeland theory fits in perfectly with Nichols’ conclusion that the homeland lay along the easternmost of the three trajectories, the one which led “eastward via Iran to western central Asia,”116 since this same trajectory also led to India.
While Nichols’ detailed linguistic analysis brings into focus the geographical location of the original homeland as indicated by the relationship of Indo-European with certain western families of languages, some other scholars have also noted the relationship of Indo-European with certain eastern families of languages: we refer, in particular, to two studies conducted, respectively, by Tsung-tung Chang in respect of the Chinese language, and Isidore Dyen, in respect of the Austronesian family of languages.
A. The Chinese Language
Tsung-tung Chang, a scholar of Chinese (Taiwanese,) origin, has shown, on the basis of a study of the relationship between the vocabulary of Old Chinese, as reconstructed by Bernard Karlgren (Grammata Serica, 1940, etc.), and the etymological roots of Proto-Indo-European vocabulary, as reconstructed by Julius Pokorny (Indogermanisches Etymologisches Wörterbuch, 1959), that there was a strong Indo-European influence on the formative vocabulary of Old Chinese.
He provides a long list of words common to Indo-European and Old Chinese, and adds: “In the last four years, I have traced out about 1500 cognate words, which would constitute roughly two-thirds of the basic vocabulary in Old Chinese. The common words are to, be found in all spheres of life including kinship, animals, plants, hydrography, landscape, parts of the body, actions, emotional expressions, politics and religion, and even function words such as pronouns and prepositions, as partly shown in the lists of this paper.”117
This Indo-European influence on Old Chinese, according to him, took place at the time of the founding of the first Chinese empire in about 2400 BC. He calls this the “Chinese Empire established by Indo-European conquerors,”118 and identifies Huang-ti (the “Yellow Emperor”), traditional Chinese founder of this first empire, as an Indo-European (suggesting that his name should actually be interpreted as “blond heavenly god”, in view of his identity).
About Huang-ti, he tells us that he was a nomadic king who “ordered roads to be built, and was perpetually on the move with treks of carriages. At night he slept in a barricade of wagons. He had no interest in walled towns… All of this indicates his origin from a stock-breeding tribe in Inner Mongolia. With introduction of horse- or oxen-pulled wagons, transport and traffic in northern China was revolutionized. Only on this new technical basis did the founding of a state with central government become feasible and functional.”119
Further, “Huang-ti is mentioned also as the founder of Chinese language in the Li-Chi (Book of Rites). In the Chapter 23 chi-fa (Rules of Sacrifices),… we read: ‘Huang-ti gave hundreds of things their right names, in order to illumine the people about the common goods……’”120
In this way: “The aboriginal people had thus to learn new foreign words from the emperors. Probably thereby the Proto-Indo-European vocabulary became dominant in Old Chinese.”121
What Tsung attempts to do to Chinese civilization is more or less what invasionist scholars have tried to do to Indian civilization, and we can take his insistence that the first Chinese civilization was established by “Indo-European conquerors” with a fistful of salt. The logical explanation for the similarity in vocabulary is simply that there was a mutual influence between Old Chinese and certain Indo-European branches which were located in Central Asia in the third millennium BC or slightly earlier.
Basically, that is what his own hypothesis also actually suggests. According to Tsung: “Among Indo-European dialects, Germanic languages seem to have been mostly akin to Old Chinese… Germanic preserved the largest number of cognate words also to be found in Chinese… Germanic and Chinese belong to the group of so-called centum languages... The initial /h/ in Germanic corresponds mostly to /h/ and /H/ in Old Chinese.... Chinese and Northern Germanic languages are poor in grammatical categories such as case, gender, number, tense, mood, etc…”122
It is unlikely that this relationship between Germanic and Old Chinese developed in Europe, and nor does Tsung himself make such a claim. He accepts that “Indo-Europeans had coexisted for thousands of years in Central Asia… (before) they emigrated into Europe”.123
The influence on the Chinese language probably, according to Tsung, spread to other related languages later on: “Sino-Thai common vocabulary, too, bristles with Indo-European stems. In my opinion, these southern tribes were once the aborigines of Northern China, who immigrated to the south… Nevertheless they could not escape since then the influence of Chinese languages and civilization.”124
How far Tsung’s hypothesis will find acceptance is not clear. It is, however, a scholarly work by a Western academician (albeit one of Taiwanese origin) established in Germany, and it is being seriously studied in the West.
Such as it is, it constitutes further linguistic support for our theory that Central Asia was the secondary homeland for various Indo-European branches on their route from India to Europe.
B. The Austronesian Family of Languages
Isidore Dyen, in his paper, The Case of the Austronesian Languages, presented at the 3rd Indo-European Conference at the University of Pennsylvania in 1966, has made out a case showing the similarities between many basic words reconstructed in the Proto-Indo-European and Proto-Austronesian languages, as we have seen in our earlier book.
They include such basic words as the very first four numerals, many of the personal pronouns, the words for “water” and “land”, etc. And Dyen points out that “the number of comparisons could be increased at least slightly, perhaps even substantially, without a severe loss of quality”.125
Dyen is not, by any stretch of the imagination, a supporter of the Indian homeland theory; and in fact such a theory does not strike him even after he notes these similarities, since he points out that the distribution of the two families, and their respective homelands as understood by him, do not explain the situation. In his own words: “The hypothesis to be dealt with is not favoured by considerations of the distribution of the two families… The probable homelands of the respective families appear to be very distant; that of the Indo-European is probably in Europe, whereas that of the Austronesian is no farther west than the longitude of the Malay Peninsula in any reasonable hypothesis, and has been placed considerably farther east in at least one hypothesis. The hypothesis suggested by linguistic evidence is not thus facilitated by a single homeland hypothesis.”126
Dyen feels that the Indo-European homeland is “probably in Europe” and the Austronesian homeland “no farther west than the longitude of the Malay Peninsula”, and hence he finds that the “linguistic evidence is not… facilitated by a single homeland hypothesis”.
But, apart from the Indian homeland theory for the Indo-European family of languages, which Dyen ignores, there is also an Indian homeland theory for the ultimate origins of the Austronesian family of languages: S.K. Chatterji, an invasionist scholar, suggests that “India was the centre from which the Austric race spread into the lands and islands of the east and Pacific”,127 and that “the Austric speech… in its original form (as the ultimate source of both the Austro-Asiatic and Austronesian branches)… could very well have been characterised within India”.128
Therefore the linguistic evidence is “facilitated by a single homeland hypothesis” in the prehistoric past: the Indian homeland hypothesis.
Thus, any linguistic
evidence there is, in respect of connections between Indo-European and other
families in the Proto-Indo-European period, all point towards an Indian
homeland for the Indo-European family of languages.
As we have seen, there is plenty of linguistic evidence which clearly shows that the Indo-European family of languages originated in India.
We will now examine the linguistic “evidence” on the basis of which the linguists usually dismiss the Indian homeland theory, and in the name of which archaeologists are classified together with “Hindu fundamentalists”. Entire schools of scholars (as we shall see in our Appendix on Misinterpretations of Rigvedic History) are mesmerised into treating the external (to India) homeland and the Aryan invasion of India as linguistically established facts.
There are two main fields of linguistic study which have contributed to this misrepresentation of the linguistic situation:
a. The study of the so-called non-Aryan substrata in Indoaryan languages.
b. The study of the reconstructed Proto-Indo-European language, society and culture.
In this section of the chapter, we will examine the first of the two above aspects: ie. the so-called non-Aryan linguistic substrata in Indoaryan languages.
According to many linguists, the Indoaryan languages contain a large number of non-Aryan words, as well as grammatical and syntactical features, which appear to be Dravidian, or occasionally Austric - words and features which are missing in Indo-European languages outside India, and which therefore show that the Indoaryan languages were intruders into an area (North India) formerly occupied by speakers of Dravidian and Austric languages, who, in the course of time, adopted the Indoaryan speech forms. A special aspect of this argument is that names of Indian animals and plants, in Indoaryan languages, are alleged to be adopted from non-Aryan (Dravidian or Austric), thereby showing that the original Indoaryan speakers were not acquainted with the flora and fauna of India.
We have examined these claims at some length in our earlier book, and we will only summaries here our arguments given therein against them:
1. In respect of the grammatical and syntactical features common to Indoaryan and Dravidian, most of these features are also found in different Indo-European branches or languages outside India, so that the features in Indoaryan are not foreign to Indo-European and are more likely to be internal developments. And the modern Indoaryan languages do not necessarily represent a change from an originally Vedic like structure, since these modem Indoaryan languages are not, as popularly believed, descendants of the Vedic language, but descendants of other Indo-European dialects which we have called Inner-Indo-European dialects, whose grammatical and syntactical features may have been different from that of the dialects of the northwest and northernmost India, which produced Vedic and the ancestors of the extra-Indian Indo-European languages, and similar to the other non-Indo-European families within India (Dravidian, Austric), from pre-Vedic times.
2. The linguists classify words as non-Aryan not because they are recognizable loan-words from Dravidian or Austric (ie. words which have a clear Dravidian or Austric etymology and no Indo-European or Sanskrit etymology), but simply because they are words for which, in the subjective opinions of these scholars (who, in any case, are on a mission to hunt out non-Aryan words in the Indoaryan languages), the Indo-European or Sanskrit etymologies are “not satisfactory”.
In most cases, these words, or equivalent forms, are not even found in the Dravidian or Austric languages, and the scholars are therefore compelled to invent the “possibility of non-Aryan speeches (other than Dravidian, Kol and the later Tibeto-Burman), speeches now extinct, being present in India”,129 and being the source for these words. There is thus a clear predisposition to brand these words as “non-Aryan” by hook or by crook.
3. Most of the non-Aryan (Dravidian or Austric) etymological derivations sought to be postulated by the linguists for particular words are challenged or refuted by other linguists, who give clear Indo-European or Sanskrit etymological derivations for the same words; and it is clear that there is no consistency or consensus in the assertions of the linguists, beyond the basic dogma that there must be non-Aryan words in the Indoaryan languages.
4. Many of the derivations which the scholars try to assert from Dravidian or Austric are basically impossible ones, since, even apart from other considerations, these words contain phonetic characteristics which are inconsistent with those of the alleged source-languages. Thus words original to the Dravidian languages could not start with an initial cerebral or liquid (T, D, r, l), did not contain aspirate sounds (h, kh, gh, ch, jh, Th, Dh, th, dh, ph, bh) and sibilants (s, S), could not start with initial voiced stops (g, j, D, d, b) or have intervocalic voiceless obstruents (k, c, T, t, p), and did not contain obstruents + liquids (kr, pi, pr, tr, etc). And yet, the linguists regularly postulate a Dravidian origin for large numbers of words which contain these phonetic characteristics.
5. In the case of names of Indian plants and animals, the majority of them have been given Sanskrit etymologies, not only by ancient Sanskrit grammarians and etymologists, but even by modern Western Sanskritists like Sir Monier-Williams, etc. Linguists who are predisposed to reject these etymologies, without being able to give definite and indisputable alternatives, cannot be taken seriously.
6. Names of plants and animals which appear to have no clear or credible Indo-European or Sanskrit etymologies cannot be automatically treated as non-Aryan words (unless they have clear and indisputable Dravidian or Austric etymologies) purely on that ground, since the situation is identical in the case of words which are very clearly and definitely inherited Indo-European words.
Thus, Carl D. Buck points out: “In the inherited names of animals there is little to be said about their semantic nature, for in most of them, the root-connection is wholly obscure.”130 Likewise, in the few inherited names of plants common to various Indo-European branches, he points out that “the root connections are mostly obscure”.131 Specifically, even a universal Indo-European word like *kuon (dog) has a “root connection much disputed and dubious”;132 and the equally universal word *ekwo (horse) has a “root connection wholly obscure”.133
Therefore, unless it is to be assumed that the Proto-Indo-Europeans were totally unacquainted with any plants and animals at all, it must be accepted that the names of plants and animals in any language need not necessarily be derivable from the etymological roots of that language: these names are more likely to have been “at first colloquial or even slang words”134 which rose up from common speech into the standard vocabulary.
7. When the names of certain plants or animals in the Indoaryan languages are demonstrably Dravidian or Austric, this will be because the plants or animals concerned are native to those parts of India where Dravidian or Austric languages are spoken. Thus the Sanskrit word ela is certainly derived from the Dravidian word yela, since the plant concerned (cardamom) is native to Kerala, which is in the heart of the Dravidian language area. The South Indian plant was borrowed, alongwith its name, by the people of North India.
In such cases, it need not even be necessary that the plant must not be found in the area of the borrowers. If a plant which is native to both North and South India was first cultivated and popularised in the South, then it is possible that the South Indian name would stick to the cultivated plant, even in the North. Thus, the tea plant is native to both China and India (Assam, etc.), and the cultivated varieties of tea today include both Chinese tea and Assamese tea. But China was the first to cultivate and popularise the beverage, and even today, the plant is known everywhere, including in India (and Assam) by its Chinese names (cA/cAy, tea).
Therefore, when there is any Dravidian or Austric name for any plant in Indoaryan languages, it is due to the geographical origin or historical cultivation of the plant in a Dravidian or Austric area, and not because the original Indoaryan speakers came from outside into an originally Dravidian or Austric India.
8. The names of plants and animals which are native to North India are of Indo-European or Sanskrit origin even in the Dravidian languages of South India and the Austric languages of eastern India. Thus, the words for camel (Sanskrit uSTra), lion (Sanskrit siMha) and rhinoceros (Sanskrit khaDgI or gaNDa) are derived from purely Indo-European roots: the word uSTra, in fact, is found in Iranian (uStra).
But, the Dravidian words for camel (Tamil-Malayalam oTTagam, Kannada-Telugu oNTe, Toda oTTe, Brahui huch, etc.), lion (Tamil cingam, Telugu siMhamu, Kannada siMha, etc.) and rhinoceros (Tamil kANDAmirugam, Telugu, khaDga-mRgamu, Kannada khaDgamRga; note also the Sanskrit word mRga, animal, necessarily added to the basic name), are all derived from the Sanskrit words. Likewise, the Austric words for camel (Santali Ut, Khasi ut) and lion (Santali sinho, Sora sinam-kidan, etc.).
This would clearly not have been the case if the northwestern areas, native to the camel, lion, and (at least in the Indus Valley period) the rhinoceros, had originally been Dravidian or Austric, or any other non-Aryan language areas before the alleged advent of the Indoaryans.
9. In addition (this is a point not made in our earlier book), it must be noted that the linguists often reject the Sanskrit or Indo-European origins of words in Indoaryan languages, or they reject correspondences between Indoaryan words and words in other branches of Indo-European, on the flimsiest of grounds: even a single vowel or consonant in a word which, according to them, is not what it should have been according to the strict and regular rules of Sanskrit or Indo-European derivations, is sufficient for them to brand the word as probably or definitely non-Aryan.
Thus, the connection between Vedic VaruNa, Greek Ouranos and Teutonic Woden is rejected, inspite of the fact that the close similarity of the names is backed by close correspondences in the mythical nature and characteristics of the three Gods, on the ground that the derivations are irregular. Likewise, the connection between Vedic PaNi/VaNi, Greek Pan and Teutonic Vanir will also be rejected on similar flimsy grounds, although, as we will see in Chapter 10 of this book, the three are definitely cognate names.
On the other hand, linguists connecting up Indoaryan words with Dravidian or Austric words have no compunctioris about linguistic regularity or accuracy: thus T. Burrow (‘Some-Dravidian Words in Sanskrit’, in Transactions of the Philological Society-1945, London, 1946) derives Sanskrit paN (to negotiate, bargain) and paNa (wager) from “Tamil puNai, to tie; tie, bond, pledge, security, surety, Kannada poNe, bond, bail…” etc. If these are Dravidian words in Sanskrit, then the related Greek Pan and Teutonic Vanir are also Dravidian words in these languages.
It is not only in respect of Indoaryan words that the linguists indulge in such hairsplitting: even in respect of the Greek word theós (God), instead of accepting that the word is an irregular derivation from Indo-European *deiwos, the linguists insist that theós is unrelated to *deiwos, and try to suggest alternative etymologies for it, eg. “from *thesós (cf. théspharos, ‘spoken by god, ordained’), but root connection much disputed and still dubious”.135 Some linguists go further: “Mr. Hopkins… rejects all the proposed etymologies and suggests that… théos itself is a loanword from pre-Greek sources.”136 However, while this kind of hairsplitting is occasional in respect of Greek, it is a regular feature in respect of Indoaryan.
We have seen, earlier on in this chapter, how Michael Witzel, while admitting to the fact that the rivers in North India have Sanskrit names from the earliest recorded (Rigvedic) period itself, tries to suggest that at least three river names, KubhA, SutudrI and KoSala, are non-Aryan, on grounds of the suggested Sanskrit etymologies being irregular.
But this kind of argument is basically untenable: while there can be no doubt that there is such a thing as regular derivations according to definite phonetic rules of etymology and phonetic change, there can be irregular derivations also, since human speech in its historical evolution has not evolved strictly according to rules. Thus, the Latin word canis (dog) is definitely derived from Indo-European *kuon: according to Buck, the “phonetic development is peculiar, but connection not to be questioned”.137 Likewise, the modern Greek ikkos (horse) is definitely derived from Indo-European *ekwo, although, as Buck points out, “with some unexplained phonetic features”.138
Hence, it is clear that linguists seeking to reject Indo-European correspondences, or Sanskrit etymologies, of Indoaryan words, on the grounds of irregular phonetic features, are not being strictly honest, and their opinions cannot be considered conclusive in any sense of the term.
This was a brief summary of our main arguments in our earlier book.
An examination of the writings of the various linguists who have written on this subject, as part of the sustained effort to produce long lists of “non-Aryan” words which form a “substratum” in Indoaryan languages, shows that logic and objectivity play no part in this exercise: any word in Sanskrit or in the modern Indoaryan languages, which appears to be similar in sound to any Dravidian word with even a vaguely similar meaning, automatically represents a Dravidian word adopted by Indoaryan in the eyes of these scholars, even when most of such words have clear Sanskrit etymologies, and many of them, or similar words, are found in other Indo-European languages outside India as well.
An examination or comparative study of the works of these linguists has been undertaken by an American scholar, Edwin F. Bryant, in his paper Linguistic Substrata and the Indigenous Aryan Debate. The quotations to follow are based on the rough draft of the above paper, the final version of which was presented at the October 1996 Michigan-Laussane International Seminar on Aryan and Non-Aryan in South Asia: Evidence, Interpretation and Ideology. (Bryant is currently on the faculty of the Department of History, Harvard University, Cambridge, USA.)
Bryant finds that “all these linguists are operating on the assumption, based on other criteria, that the Aryans ‘must have’ invaded India where there could not have been a ‘linguistic vacuum’”, and that, beyond this shared predisposition, there is no consensus among them on any specific point. His examination of the works of different linguists shows “that they are not internally consistent, since the opinions of the principal linguists in this area have differed quite considerably. This problematizes the value of this method as a significant determinant in the Indo-Aryan debate…”.
The extent to which these linguists (all of whom are otherwise in agreement in the belief that the Indoaryans are immigrants into India from an original homeland in South Russia) differ in the matter is made clear by Bryant:
1. About the grammatical and syntactical features common to both Dravidian and Indoaryan, Robert Caldwell (1856) was the first to draw attention to many of them; but he rejected the idea that these features constituted originally Dravidian grammatical and syntactical elements (which surfaced in Indoaryan as a substratum): “whatever the ethnological evidence of their identity may be supposed to exist… when we view the question philologically, and with reference to the evidence furnished by their languages alone, the hypothesis of their identity does not appear to me to have been established.”
But, a hundred years later, M.B. Emeneau (1956) drew up a whole list of such grammatical and syntactical features, and added to them in his later studies (1969, 1974). F.B.J. Kuiper (1967) and Massica (1976) also added to the list. These linguists concluded that these features were definitely evidence of a Dravidian substratum.
However, H. Hock (1975, 1984) strongly rejected the idea that these features are due to a Dravidian substratum. He pointed out that most of these features actually have parallels in other Indo-European languages outside India, and therefore they were more likely to be internal developments in Indoaryan. Since then, several other linguists, all otherwise staunch believers in the Aryan invasion theory, have rejected the idea that these features are Dravidian features.
F.B.J Kuiper (1974), a staunch protagonist of the substratum theory, admits that “we cannot compare the syntax of the Rigveda with contemporaneous Dravidian texts. The oldest Dravidian texts that we know are those of old Tamil. They probably date from about the second century AD and are, accordingly, at least a thousand years later than the Rgveda.”
M.B. Emeneau himself, although he sticks to the claim that a Dravidian substratum explains the situation better, admits (1980) that it is not as easy as that: “Is the whole Indo-Aryan history one of self-development, and the complex Dravidian development triggered by Indo-Aryan, perhaps even New Indo-Aryan, influence, or, in the case of Kurukh, borrowed from New Indo-Aryan?… no easy solution is yet at hand.”
2. F.B.J. Kuiper (1991) produced a list of 380 words from the Rigveda, constituting four percent of the Rigvedic vocabulary, which he claimed were of non-Aryan (primarily Dravidian) origin. Earlier linguists were more cautious in the matter of Rigvedic vocabulary. M.B. Emeneau (1980), for example, hoped that the linguists would agree at least on one word mayUra, as a borrowing from Dravidian: “I can only hope that the evidence for mayuura as a RV borrowing from Dr. is convincing to scholars in general.”
But P. Thieme (1994) examined and rejected Kuiper’s list in toto, gave Indoaryan or Sanskrit etymologies for most of these words, and characterized Kuiper’s exercise as an example of a misplaced “zeal for hunting up Dravidian loans in Sanskrit”. In general, Thieme sharply rejects the tendency to force Dravidian or Austric etymologies onto Indoaryan words, and insists (1992) that “if a word can be explained easily from material extant in Sanskrit itself, there is little chance for such a hypothesis”.
Rahul Peter Das (a believer in the Aryan invasion theory), likewise rejects (1994) Kuiper’s list, and emphasises that there is “not a single case in which a communis opinio has been found confirming the foreign origin of a Rgvedic (and probably Vedic in general) word”.
Therefore, it is clear that claims regarding Dravidian loan-words in Vedic Sanskrit are totally baseless.
3. So far as the modern Indoaryan languages are concerned, also, the untenability of the whole exercise of hunting down non-Aryan words in Indoaryan can be illustrated by an examination of a detailed study conducted by Massica (1991), a staunch believer in the Aryan invasion theory (and who, in fact, concludes that his study confirms the theory), who examined a complete list of names of plants and agricultural terms in Hindi.
Massica’s study found that only 4.5% of the words have Austric etymologies, and 7.6% of the words have Dravidian etymologies, and, even here, “a significant portion of the suggested Dravidian and Austroasiatic etymologies is uncertain”. When we consider that the few words where an Austric or a Dravidian etymology can be proved probably refer to plants and agricultural processes native to South India or Eastern India, Massica’s study clearly contradicts his conclusions.
Massica, however, classifies 55% of the words as non-Aryan (other than Dravidian and Austric, and other than non-Indian names for non-Indian plants), but of “unknown origin”.
It is words of this kind which, as we have already seen, have led the linguists to postulate extinct indigenous families of non-Aryan, non-Dravidian and non-Austric languages in ancient India, which have disappeared without a trace, but which constitute the main non-Aryan substrata in Indoaryan. As T. Burrow notes, even the most liberal Dravidian and Austric etymologising may not serve in explaining words which (in his opinion) are non-Aryan, since “it may very well turn out that the number of such words which cannot be explained will outnumber those which can be. This is the impression one gets, for example, from the field of plant names, since so far only a minority of this section of the non-Aryan words has been explained from these two linguistic families.”
However, although the linguists are compelled to resort to these stratagems, they are not very comfortable with them. Emeneau (1980), for example, admits: “it hardly seems useful to take into account the possibilities of another language, or language family, totally lost to the record, as the source” for the supposedly non-Aryan words.
Massica himself, although he brands the words as non-Aryan on the ground that there are no acceptable Sanskrit etymologies, admits that “it is not a requirement that the word be connected with a root, of course: there are many native words in Sanskrit as in all languages that cannot be analysed”.
Bloch and Thieme emphasize the point that the names of plants need not be analysable from etymological roots, since most of them will be slang or colloquial words derived from the “low culture” vernaculars of the same language.
4. It is in Classical Sanskrit word-lists that we find many words which can be, or have been, assigned Dravidian or Austric origins. This has led the linguists to emphasise a theory first mooted by Burrow (1968), according to which there was a very small number of Dravidian and Austric words (or none at all) in the Rigveda, which grew in the later Vedic literature, reached a peak in the Epics and PurANas, and in the Classical Sanskrit word-lists, and finally dwindled in the Prakrits, and even more so in the modern Indoaryan languages. This situation, according to Burrow, depicts a scenario where the Aryan immigrants into India were new arrivals at the time of composition of the hymns, and hence hardly any indigenous words had infiltrated into the vocabulary of the Rigveda. As the process of bilingualism developed (involving both the local inhabitants of the North preserving some of their original non-Aryan vocabulary as they adopted the Aryan speech-forms, as well as post-first generation Aryans inheriting non-Aryan words as they merged with the local people), the number of such words increased in the language of the Epics and PurANas, and the Classical Sanskrit word-lists. Finally, when there were no more bilingual speakers left in the North, since everyone had adopted the Aryan speech-forms, the appearance of non-Aryan words in the Indoaryan languages ceased, hence the modem Indoaryan languages have few such words.
However, Caldwell (1856), who was the first to produce lists of words “probably” borrowed by Sanskrit from Dravidian, rejected this substratum theory. He noted that the words did not include the essential aspects of vocabulary (such as actions, pronouns, body parts, etc.), and consisted almost exclusively of words “remote from ordinary use”, and hence concluded that the Dravidian languages could not possibly have been spoken in North India at the time of the alleged Aryan invasion.
Bloch (1929), who rejected the substratum theory completely, pointed out that the Dravidian languages of the South, even at the level of common speech, contain a massive amount of borrowed Sanskrit vocabulary covering every aspect of life. But this is not explained as an Aryan substratum in South India. The natural explanation for these borrowings is that a relatively small number of Sanskrit-speaking individuals were responsible for them. Likewise, the Dravidian words in Sanskrit were reverse borrowings, being introductions of Dravidian words into literary Sanskrit by similar Sanskrit-speaking individuals from the South. Such words were only part of the Classical Sanskrit lexicon, and few of them percolated to the Indoaryan vernaculars. Thus, even popular Sanskrit words like nIra (water, Tamil nIr), mIna (fish, Tamil mIn), heramba (buffalo, Tamil erumai), etc. are not used in the modem Indoaryan languages, which use, instead, derivatives of the Sanskrit words pAnIyam, matsya and mahiSa respectively. Such words, as Bloch points out, were artificial and temporary introductions into literary Sanskrit, most of which (although it is likely that some of them became so popular that they replaced, or accompanied, original Sanskrit words, and percolated down into modern Indoaryan) either died out completely, or remained purely literary words which did not become a part of naturally spoken Indoaryan speech.
Massica, in his recent study (1991) already referred to, also notes that Dravidian words in Sanskrit are not found in present-day Indoaryan languages like Hindi. Clearly, these words do not represent a Dravidian substratum in Sanskrit, but a process of artificial adoption of vocabulary from regional speech-forms, both Aryan and non-Aryan.
5. Many linguists question the idea that there could be a Dravidian or Austric substratum in the Indoaryan languages of North India, even on the grounds of the likely geographical distribution of these two families in ancient times. In respect of the Austric languages, even a staunch supporter of the non-Aryan substratum theory like Burrow (1968) admits that the possibility of an Austric substratum is remote since “the evidence as it is so far established would suggest that these languages in ancient times as well as now were situated only in eastern India”. Massica (1979) and Southworth (1979) also reiterate this point.
R.P. Das (1994) points out that there is “not a single bit of uncontroversial evidence on the actual spread of Dravidian and Austro-Asiatic in prehistoric times, so that any statement on Dravidian and Austric in Rgvedic times is nothing but speculation”.
6. In fact, when words are similar in both Indoaryan and Dravidian, it is more natural to conclude that the Indoaryan words are the original ones. According to Thieme, “all the Dravidian languages known to us fairly bristle with loans from Sanskrit and the Aryan vernaculars. Dravidian literature in South India came into existence under the impulse and influence of Sanskrit literature and speech. Wherever there is a correspondence in the vocabularies of Sanskrit and Dravidian, there is a presumption, to be removed only by specific argument, that Sanskrit has been the lender, Dravidian the borrower.”
While Thieme is, of course, an opponent of the substratum theory, even so staunch a supporter of the substratum theory as Emeneau (1980) admits that it is “always possible, eg. to counter a suggestion of borrowing from one of the indigenous language families by suggesting that there has been borrowing in the other direction”.
7. Ultimately, therefore, the whole question of a Dravidian, or non-Aryan, substratum in the Indoaryan languages is a matter of dogma rather than scientific study.
R.P. Das (1994), for example, points out that there is little linguistic logic involved in the debate about the Dravidian or Austric origins of Indoaryan words: “Many of the arguments for (or against) such foreign origin are often not the results of impartial and thorough research, but rather of (often wistful) statements of faith.”
Bloch (1929), likewise, had earlier dismissed the Dravidian derivations which many linguists sought to force on Sanskrit words, as being not “self-evident” but “a matter of probability and to a certain extent of faith”.
While both Das and Bloch are opponents of the substratum theory (though believers in the Aryan invasion theory in general), Emeneau (1980), a staunch supporter of the substratum theory, himself admits that these derivations are “in fact all merely ‘suggestions’. Unfortunately, all areal etymologies are in the last analysis unprovable, are ‘acts of faith’.”
The “faith” in all these cases is the faith in the external (to India) origin of the Indoaryans (and Indo-Europeans), which Emeneau (1980) describes as “our linguistic doctrine which has been held now for more than a century and a half”.
Hence, after his examination of the claims and counterclaims of the linguists, Bryant reaches the logical conclusion that “the theory of Aryan migrations must be established without doubt on other grounds for research into pre-Aryan linguistic substrata to become meaningful. However, the ‘evidence’ of a linguistic substratum in Indo-Aryan, in and of itself, due to its inconclusive nature, cannot be presented in isolation as decisive proof in support of the theory of Aryan invasions or migrations into the Indian subcontinent.”
Finally, we come to that aspect of linguistic studies which first led the linguists to dismiss the idea of India being the original homeland, and which first created the impression, which persists to this day, even after this aspect of linguistic studies has now been recognized by serious linguists as a method which cannot be relied upon for arriving at any conclusions on the subject, that linguistics has “proved” the non-Indian origin of the Indo-Europeans. We refer to the study of the proto-language and of its geographical implications for the original homeland of the Indo-European family of languages.
The linguists have reconstructed the Proto-Indo-European language on the basis of definite phonetic rules of sound-change and development, applied to the words common to different Indo-European branches. Allowing for the fact that most linguists often tend to adopt a rigid and dogmatic approach to the subject (which, as we have already seen, leads them to indulge in hairsplitting, and to reject many obvious cognate forms, like Greek theos, or to only grudgingly accept some others, like Latin canis and modern Greek ikkos), and that it is often difficult to explain changes in vocabulary, which makes it necessary to be cautious in postulating original words (as has often been pointed out, as an example, all the modem Italic languages have words for “horse” derived from a Latin word caballus: eg. Italian cavallo, French cheval, Spanish caballo, Rumanian cal; while the actual Latin word for the horse was equus. If Latin had been an unrecorded language, and it had been required to reconstruct it on the basis of words common to its present day descendants, the word equus would never be reconstructed), the reconstruction of the Proto-Indo-European language may generally be accepted as a reasonably valid one, with some natural limitations.
However, this reconstruction has not been treated as a purely academic exercise, but as a means of pinpointing the geographical location of the original homeland. There have been two main methods by which the linguists have sought to use the exercise as a means of rejecting the idea of an Indian homeland. and, since their endeavours appear to have been so successful in mesmerising all and sundry and in effectively derailing all rational inquiry into the subject, it is necessary for us to examine these two methods:
A. Linguistic Paleontology.
Linguistic Paleontology is a method devised by nineteenth century linguists, by which they sought to reconstruct the geographical and socio-cultural environment of the Proto-Indo-European people on the basis of words common to different Indo-European branches.
On the basis of the few names of animals, birds and plants, and words indicating climate, common to different Indo-European branches, the linguists concluded that the Proto-Indo-Europeans lived in a cold environment, and were acquainted with a few plants/trees like barley, birch, pine and oak, and animals like horses, cattle, goats, sheep, deer, bears, wolves, dogs, foxes and otters.
The names of these plants and animals do not really pinpoint a specific area, since they are all found in a large area ranging from Europe to North India, covering almost the entire Indo-European belt. But the linguists concluded that the evidence of these names clearly excluded India from being the location of the original homeland, since the common names did not include names of plants/trees and animals which are specifically found in India (such as the elephant, etc).
However, this argument is clearly illogical: if the Indo-European languages outside India do not appear to have names for plants and animals which are found in India, but not found in the areas where these languages are spoken; then the Indoaryan languages also do not have names for plants and animals which are found in Indo-European areas outside India, but not found in India. The conclusion that can be derived from this is simply that Indo-European languages generally (but not always) retained Proto-Indo-European names only for those plants and animals which were also found in their new habitats: they generally lost the names for plants and animals which were found in former habitats but not in newer ones. This would naturally be the case, when we consider that the speakers of most Indo-European languages would generally be natives of their respective areas, who adopted the Indo-European speech from immigrant Indo-Europeans, and who would therefore be ignorant of, and unconcerned with, plants and animals native to the former habitats of the immigrants.
Therefore, linguistic paleontology stands largely discredited today as a method of reconstruction of the original geographical environment of the Indo-Europeans, or at least as a method on the negative testimony of which certain areas like India could be excluded from being the original homeland. As the eminent linguist Stefan Zimmer puts it: “The long dispute about the reliability of this ‘linguistic paleontology’ is not yet finished, but approaching its inevitable end - with a negative result, of course.”139
T. Gamkrelidze and V. Ivanov,140 two linguists who are supporters of the Anatolian homeland theory, have recently examined words in the Indo-European languages which were largely ignored or missed by the linguists in general, and they have arrived at the conclusion that Proto-Indo-European names definitely existed for some more animals such as the leopard (Sanskrit pRdAku, Greek pardos, Hittite parsana) and the monkey (Sanskrit kapi, Greek kepos, which they also link, with k/mute alteration, with Germanic and Celtic words like Old Norse api, Old English apa, Old High German affo, Welsh epa and Irish apa, “ape”), and even more significantly, the camel and the elephant:
1. The camel is native to West Asia and to Central Asia. There are cognate words for the camel in Tokharian *alpi, Old Church Slavonic velibadu, Baltic (Lithuanian) verbliudas, and Germanic words like Old Norse ulfaldi, Old English olfend, Old High German olbanta and Gothic ulbandus. A related word in Hittite, according to C.D. Buck, is ulupantas or ulpantas which appears to be used for “ox”.141
The word is similar to the Greek word elephas for elephant, which is the source for all the European names for the elephant. Buck suggests that this word is “based upon… Egyptian words… to be analysed as el-ephas, the second part, like Lat. ebur, ‘ivory’, from Egypt. Ab, ‘elephant, ivory’, but first part disputed”.142 He adds: “Hence also (though disputed by some) with shift to ‘camel’, Goth. ulbandus, ON ulfaldi, OE olfend, OHG olbanta……”143
The evidence of the Tokharian word, however, conclusively proves that this word cannot be a borrowing by Greek from Egyptian. A word so borrowed could never have been transmitted to Tokharian in Central Asia by any manipulation of any known theory of Indo-European origins and migrations; and the Tocharian word is clearly a related one since it contains both the elements, the “second part” of the word as well as the “disputed” first part.
Therefore, while it is very likely that there was a “shift” from an original meaning “elephant” to a new meaning “camel”, this shift took place in Central Asia and not in Greece. The cognate words for camel in Tocharian, Germanic, Slavonic and Baltic (and also Hittite, where there has been a second shift in meaning to “ox”) clearly prove that all these branches shared a sojourn in the camel lands of Central Asia.
2. The Greek word el-ephas is exactly cognate (again, only the second part of the word) with the Rigvedic ibhas. As we have already seen in our chapter on the Geography of the Rigveda, ibhas is just one of the four purely “Aryan” names (ibhas, sRNI, hastin and vAraNa) for the elephant in the Rigveda. Gamkrelidze and Ivanov point out that the Latin word ebur, “ivory”, is also cognate to the Sanskrit ibhas.
We thus have the evidence of three different branches of Indo-European languages for the elephant as an animal known to the Proto-Indo-Europeans. As the Proto-Indo-Europeans were not native to Africa, African elephants (not being domesticated) could not have been directly known to them (even as an imported animal) in any other proposed homeland, and the Asiatic elephant is not native to any area north or west of India, the implications of this evidence are loud and clear.
Incidentally, it is possible that the Egyptian word Ab for “elephant” or “ivory” is itself derived from Sanskrit ibhas. We have it on the testimony of the Old Testament of the Bible (I Kings 22.10; II Chronicles 9.21) that apes, ivory and peacocks were imported from India (the peacocks confirm that the land referred to is India, or a transit port on the way from India) into Palestine, and doubtless the same was the case in Egypt as well.
The Hebrew word for “ape” in the above references is qoph which is derived by linguists from the Sanskrit kapi; and, likewise, Buck accepts kapi as the “probable source of Egyptian qephi”.144 Significantly, the words for elephant in Arabic and Hebrew, fil and pil respectively, are clearly derived from the Sanskrit word pIlu for a male elephant, thereby indicating that it was the Indian elephant rather than the African one which was known in this region.
3. An animal whose name is common to almost all the Indo-European branches is the cow (Sanskrit go, Avestan gao, German kuh, Latin bOs, Irish bo, Lettish guovs, Greek boûs, Old Church Slavonic krava, etc), for whom the reconstructed Proto-Indo-European word is *gwou. It is clear that the cow was a very intrinsic part of the life of the Indo-Europeans, as is proved also by its dominant status in the culture, idiom and imagery of the oldest Indo-European texts, the Rigveda and the Avesta.
Significantly, different ancient civilizations (Sumerian gu, Ancient Chinese gou) appear to have borrowed the word from the Indo-Europeans. It is, therefore, quite likely that the Proto-Indo-European homeland was a primary centre of diffusion of cattle breeding.
It may be noted in this context that recent research by scientists at the Trinity College in Dublin has revolutionised ideas about the origins of the domestication of cattle. It was formerly believed that cattle domestication first took place in Anatolia, and then spread to the rest of the world; and the humped breeds of Indian cattle, known in the West as Zebu or Brahmin cattle, were believed to be descended from these Anatolian cattle.
However, the scientists “who examined the DNA of 13 breeds of modern cattle found that all the European and African cattle breeds shared the same genetic lineage. But the eastern types came from an entirely different source. By backtracking the number of mutations that must have occured, the scientists have also deduced that the two lines split more than 200,000 years ago; and since the two lines are still distinct, the simplest interpretation of the research was that there were two separate domestication events.”145
Thus, India, the centre of domestication of other species of bovids, like the buffalo and the gayal, was also the centre of domestication of the eastern or humped cattle.
And, to howsoever great or small an extent, this appears to strengthen the claims of India to be the location of the original homeland of the Indo-European family of languages.
This is corroborated by the fact that Sanskrit retains a distinctly different root word for “milk”, which appears to be older, and closer to the original Indo-European ethos, than the common word for “milk” found in almost all the other branches of Indo-European languages.
Many of the other branches have related words for “milk”: German milch, Irish mlicht, Russian moloko, etc. And even where they appear to differ in the noun form, they share a common word for the verb “to milk”: Latin mulgere, Old High German melchan, Greek amèlgo, Old Church Slavonic mlešti, Lithuanian milZti, Albanian mjellë, Irish bligim, etc.
Only Sanskrit and Iranian stand out in not having any word related to the above. Instead, we have Sanskrit dugdha, “milk”, derived from the root duh-, “to milk”, with related verbal forms duxtan, dušidan, “to milk” in modern Persian (though not in the Avesta).
The root duh-, found directly only in Sanskrit, and only secondarily in Iranian, appears to have deeper roots in the Indo-European languages. According to many linguists (although many others dismiss the derivation as simplistic), the Indo-European words for “daughter” (Sanskrit duhitar, Persian dukhtar, Gothic dauhtar, Lithuanian dukte, Old Church Slavonic dUšti, Greek thugater, etc.) are derived from the same root, so that the word basically means “milkmaid”, indicating that cattle-breeding was a primary occupation among the Proto-Indo-Europeans.
VI.B. Archaic Dialectology
The second significant aspect of the study of the protolanguage, on the basis of which an Indian homeland was rejected by the linguists, was that Sanskrit, in some respects, represents a phonetically highly evolved form of the original Proto-Indo-European: thus, for example, to quote the most common factor cited, Sanskrit is a “Satem” language, and in fact, alongwith Avestan, the most highly palatalized of the Satem languages. The original Proto-Indo-European language was a “Kentum” language, and some branches evolved into Satem branches by a process of palatalization of original velars (k, g) into palatals (c, j) and into sibilants (s, S). The Kentum branches thus represent an older form of Indo-European, and all the Kentum branches are found only in Europe - or so it was thought until the discovery of Tokharian in Chinese Turkestan; but this discovery was quickly sought to be absorbed into the western homeland theory by postulating an early migration of the Tokharians from the west into the east,
However, as we have already seen earlier on in this chapter, the phenomenon of palatalization, as also various other features which represent phonetic evolutions from the Indo-European original, are now accepted as innovations which took place in the heartland of the Proto-Indo-European homeland after the migrations of early branches which retained the original features.
As Winn puts it: “Linguistic innovations that take place at the core may never find their way out to peripheral areas, hence dialects spoken on the fringe tend to preserve archaic features that have long since disappeared from the mainstream.”146 Therefore, the fact that Sanskrit represents a phonetically evolved form of the Proto-Indo-European language, far from being a negative factor in respect of the idea of an Indian homeland, is a positive one.
In fact, there are three factors, in respect of archaisms, which add up to make a strong case for an Indian homeland:
1. Various evolved phonetic features in Sanskrit, as we have seen, particularly in the matter of palatalization of original velars, definitely point towards India as the original homeland.
2. At the same time, in respect of vocabulary, Sanskrit is the most archaic or representative language in the entire Indo-European family. As Griffith puts it in his preface to his translation of the Rigveda, in the language of the Rigveda “we see the roots and shoots of the languages of Greek and Latin, of Kelt, Teuton and Slavonian… the science of comparative philology could hardly have existed without the study of Sanskrit…”
As we have pointed out in some detail in our earlier book, the fact that Sanskrit has retained the largest number of Proto-Indo-European words, even when its phonetic and grammatical features continued to evolve, is strong evidence of an Indian homeland: the language of a migrating group may retain many of its original phonetic or grammatical features, even when these features are lost or evolved away in the language still spoken in the original area, but it is likely to lose or replace a substantial part of its original vocabulary (though it may retain many telltale archaic words) as compared to the language still spoken back home.
Warren Cowgill, in the Encyclopaedia Britannica, points out that this was the case with most of the ancient Indo-European languages: “In prehistoric times, most branches of Indo-European were carried into territories presumably or certainly occupied by speakers of non-Indo-European languages… it is reasonable to suppose that these languages had some effect on the speech of the newcomers. For the lexicon, this is indeed demonstrable in Hittite and Greek, at least. It is much less clear, however, that these non-Indo-European languages affected significantly the sounds and grammar of the Indo-European languages that replaced them.”147 The same was the case with the modern languages: “When Indo-European languages have been carried within historical times into areas occupied by speakers of other languages, they have generally taken over a number of loan-words… however, there has been very little effect on sounds and grammar.”148
3. Finally, and most significant of all, we have the fact that within India itself, certain isolated languages have retained archaisms already lost even in Vedic Sanskrit. There is no way in which the presence of these languages, which definitely represent remnants of extinct branches of Indo-European other than Indoaryan or even the hypothetical “Indo-Iranian”, can be incorporated into any theory of migration of the Indoaryans from South Russia to India.
There are two such languages, one of which is now accepted by the linguists as a remnant of an extinct Kentum branch of Indo-European languages, but in respect of the other, detailed research is necessary from a point of view hitherto unsuspected:
a. The BangANI language, spoken in the Garhwal region in the western Himalayas (in Uttar Pradesh) was brought into dramatic highlight by Clans Peter Zoller, a German linguist, in 1987 (as reported in our earlier book) when he announced the discovery of the remnants of an ancient Kentum language in the older layers of this language.
Zoller pointed out that BangANI contained three historical layers: “The youngest and most extensive layer is where BangANI shares many similarities with the Indo-Aryan languages of Himachal Pradesh and Garhwal. The second is an older layer of Sanskrit words where one can observe a strikingly large number of words that belongs to the oldest layer of Sanskrit, the Sanskrit of the Vedas. The third and the oldest layer in BangANI is formed by words that have no connection with Sanskrit but with the Kentum branch of Indo-European languages.”149
By 1989, Zoller had presented a full-fledged case, which created a furore in linguistic circles. An immediate reaction to it was a joint project, by an Indian linguist Suhnu Ram Sharma and a Dutch linguist George van Driem, which examined Zoller’s claims. According to these scholars, “Zoller’s BangANI findings not only had far-reaching implications for our understanding of the prehistoric migrations of ancient Indo-Europeans, they also appeared to violate much of what is received knowledge in historical linguistics.”150 Hence: “In 1994, we conducted fieldwork in order to verify these remarkable findings. The results of our investigation are presented here. On the basis of these results, it is our contention that no Kentum Indo-European remnants exist in the BangANI language.”151
Not only did these linguists reject Zoller’s findings, but they also levelled serious allegations regarding Zoller’s professional integrity: “In view of our findings, and in view of the manner in which Zoller presented his, the question which remains for the reader to resolve in his own mind is whether Zoller has fallen prey to the wishful etymologizing of transcriptional errors or whether he has deliberately perpetrated a hoax upon the academic community. In other words, was the joke on Zoller, or was the joke on us?”152
The above is an example of the vicious reactions evoked among scholars inimical to the Indian homeland theory, to any serious scholarly study which tends to, directly or indirectly, support, or even appear to support, this theory.
The matter did not end there. Zoller took up the challenge and issued a strong and detailed rejoinder to the allegations of van Driem and Sharma. Even more significant was a detailed counter study by Anvita Abbi and Hans Hock which not only conclusively demolished their “refutation” of Zoller’s findings, and conclusively proved that BangANI does indeed contain the remnants of an extinct Kentum language, but also clearly showed that it was Suhnu Ram Sharma and George van Driem who had attempted to deliberately perpetrate a hoax on the academic community.
The long and short of it is that BangANI is now accepted by linguists all over the world as a language whose oldest layers contain remnants of an archaic Kentum language, a circumstance which is totally incongruous with any theory of Indoaryan immigrations into India.
b. The Sinhalese language of Sri Lanka is generally accepted as a regular, if long separated and isolated, member of the “Indoaryan” branch of Indo-European languages; and no linguist studying Sinhalese appears, so far, to have suggested any other status for the language.
However, apart from the fact that Sinhalese has been heavily influenced not only by Sanskrit and (due to the predominance of Buddhism in Sri Lanka) Pali, but also by Dravidian and the near-extinct Vedda, the language contains many features which are not easily explainable on the basis of Indoaryan.
Wilhelm Geiger, in his preface to his study of Sinhalese, points out that the phonology of the language “is full of intricacies… We sometimes meet with a long vowel when we expect a short one and vice versa”,153 and, further: “In morphology there are formations, chiefly in the verbal inflexion, which seem to be peculiar to Sinhalese and to have no parallels in other Indo-Aryan dialects… and I must frankly avow that I am unable to solve all the riddles arising out of the grammar of the Sinhalese language.”154
However, not having any particular reason to suspect that Sinhalese could be anything but an “Indoaryan” language descended from Sanskrit, Geiger does not carry out any detailed research to ascertain whether or not Sinhalese is indeed in a class with the “other Indo-Aryan dialects”. In fact, referring to an attempt by an earlier scholar, Gnana Prakasar, to connect the Sinhalese word eLi (light) with the Greek hElios (sun), Geiger rejects the suggestion as “the old practice of comparing two or more words of the most distant languages merely on the basis of similar sounds, without any consideration for chronology, for phonological principles, or for the historical development of words and forms…”155
However, there are words in Sinhalese, of which we can cite only one here, which cannot be so easily dismissed: the Sinhalese word watura, “water”, is not only closely cognate to the Germanic words (which includes English “water”) and Hittite water, but it represents a form which is impossible to explain on the basis of Sanskrit or Indoaryan etymologies. Geiger himself, elsewhere, rejects an attempt by an earlier scholar, Wickremasinghe, to derive the word from Sanskrit vartarUka as “improbable”; and although he accepts the suggestion of another scholar, B. Gunasekara, that the “original meaning is ‘spread, extension, flood’ (M. vithar)… Pk. vitthAra, Sk. vistAra,”156 he notes that “vocalism a.u. in vatura is irregular, cf. vitura”.157
M.W.S. de Silva, in his detailed study of Sinhalese, points out that “Indo-Aryan (or Indic) research began with an effort devoted primarily to classifying Indian languages and tracing their phonological antecedents historically back to Vedic and Classical Sanskrit… Early Sinhalese studies have followed the same tradition.”158 However, Sinhalese “presents a linguistic make-up which, for various reasons, distinguishes itself from the related languages in North India… there are features in Sinhalese which are not known in any other Indo-Aryan language, but these features, which make the story of Sinhalese all the more exciting, had not received much attention in the earlier studies.”159
He also points out: “Another area of uncertainty is the source of the small but high-frequency segment of the Sinhalese vocabulary, especially words for parts of the body and the like: eg. oluva ‘head’, bella ‘neck’, kakula ‘leg’, kalava ‘thigh’, etc. which are neither Sanskritic nor Tamil in origin. The native grammarians of the past have recognized that there are three categories of words - (a) loanwords, (b) historically derived words and (c) indigenous words… No serious enquiry has been made into these so-called indigenous words”.160
In his preface, de Silva notes that “there is a growing awareness of the significance of Sinhalese as a test case for the prevailing linguistic theories; more than one linguist has commented on the oddities that Sinhalese presents and the fact… that Sinhalese is ‘unlike any language I have seen’.”161 Further, he quotes Geiger: “It is extremely difficult, and perhaps impossible, to assign it a definite place among the modern Indo-Aryan dialects.”162
But, it does not strike de Silva, any more than Geiger, that the reason for all this confusion among linguists could be their failure to recognize the possibility that Sinhalese is not an Indoaryan language (in the sense in which the term is used) at all, but a descendant of another branch of Indo-European languages.
From the historical point of view, “a vast body of material has been gathered together by way of lithic and other records to portray the continuous history of Sinhalese from as early as the third century BC.”163 in Sri Lanka, and “attempts have been made to trace the origins of the earliest Sinhalese people and their language either to the eastern parts of North India or to the western parts”.164
But de Silva quotes Geiger as well as S. Paranavitana, and agrees with their view that “the band of immigrants who gave their name Simhala to the composite people, their language and the island, seems to have come from northwestern India… their original habitat was on the upper reaches of the Indus river… in what is now the borderland between Pakistan and Afghanistan”,165 and quotes Paranavitana’s summary of the evidence, and his conclusion: “All this evidence goes to establish that the original Sinhalese migrated to Gujarat from the lands of the Upper Indus, and were settled in LATa for some time before they colonised Ceylon.”166
A thorough examination, with an open mind, of the vocabulary and grammar of Sinhalese, will establish that Sinhalese represents a remnant of an archaic branch of Indo-European languages.
The evidence of BangANI and Sinhalese (the one word watura itself) constitutes a strong case for an Indian homeland since it clashes sharply with any theory of Indoaryan migrations into India.
Basically, the confusion that we see in respect of Sinhalese studies is also found in the study of Indoaryan languages in general. And the root of all this confusion is the general theory which maintains that:
a. The “Indo-Iranians” represented a branch of Indo-Europeans who separated from the other branches in distant regions and migrated to Central Asia, and shared a joint “Indo-Iranian” phase there, before separating and migrating into India and Iran respectively.
b. The “Indoaryans” represented that section of the “Indo-Iranians” who entered India and composed the Rigveda during the earliest period of their sojourn in the northwestern parts of India, before expanding into the rest of India and giving birth to the ancestral forms of the present-day Indoaryan languages.
The linguistic evidence (even apart from the archaic evidence of BangANI and Sinhalese) totally fails to fit in with this theory:
1. “Indoaryan” and Iranian do not constitute one branch, but at least two distinct branches: Winn points out that there are “ten ‘living branches’... Two branches, Indic (Indo-Aryan) and Iranian dominate the eastern cluster. Because of the close links between their classical forms - Sanskrit and Avestan respectively - these languages are often grouped together as a single Indo-Iranian branch”.167 And he notes that these close links came about due to “a period of close contact between Indic and Iranian people (which) brought about linguistic convergence, thus making the two languages appear misleadingly similar”.168
As Meillet had long ago pointed out: “It remains quite clear, however, that Indic and Iranian developed from different Indo-European dialects, whose period of common development was not long enough to effect total fusion.”169
The evidence of comparative mythology (see Chapter 10) also disproves the common Indo-Iranian hypothesis. Rigvedic mythology is often the only connecting link between different other Indo-European mythologies, while Avestan mythology appears to have no links with any other Indo-European mythology other than that of the Rigveda itself.
The “period of common development” which brought about the “close links between… Sanskrit and Avestan” was of course the “period of close contact between Indic and Iranian people” in the Late Period of the Rigveda, as we have already seen in the previous chapter.
2. The Indo-Iranian hypothesis is also disproved by the fact that Iranian shares at least one isogloss with Greek and Armenian (fitting in with our classification of these three branches as constituting. the Anu confederation of the Early Period of the Rigveda) which is not shared by Sanskrit: “In three Indo-European languages, whose grouping is significant - Greek, Armenian and Iranian - the shift from s to h occured, not, as in Brythonic, at a relatively recent date, but before the date of the oldest texts. Moreover, in all three, the distribution pattern is exactly the same: h develops from initial *s before a vowel, from intervocalic *s and from some occurences of *s before and after sonants; *s remains before and after a stop.”170
This shift, which is universal in the three branches, is not found in Sanskrit and a majority of the Indoaryan languages, although a similar shift took place “at a relatively recent date” in some modem Indoaryan dialects of the northwest and west (Gujarati, etc.) and, significantly, in Sinhalese.
Another, minor, point where Greek, Armenian and Iranian share a common development, distinct from Sanskrit, is in “those cases in which a morphological element ends with a dental consonant and the following element begins with a t”.171 All the three branches show st while “Sanskrit regularly shows tt”.172
3. There is one isogloss which is found only in the three branches referred to above (Greek, Armenian and Iranian) and in Sanskrit, and in some modern Indoaryan dialects of the north and northwest (as far as the western dialects of Hindi), but not in the majority of modern Indoaryan languages: “the prohibitive negation *mE is attested only in Indo-Iranian (mA), Greek (mE) and Armenian (mI), elsewhere it is totally lacking… and there is no difference in this respect between the ancient and modern stages of Greek, Armenian or Persian.”173
But there is a difference in this respect between the ancient stage (Sanskrit) and a majority of the languages in the modem stage of what the linguists classify as the “Indoaryan” branch (except for modem western Hindi mat, etc.).
This could be because most of the Indoaryan languages lost this word; but it could also be because most of the modern Indoaryan languages are descendants of Indo-European dialects which never had this word, and were not directly part of the common culture developed by the PUrus (the Vedic Aryans) and the Anus (Iranians, Armenians, Greeks) in the northern and northwestern parts of North India, after the departure of the Druhyus. Their ancestral dialects were what we have (in our earlier book) called the “Inner Indo-European” dialects spoken in the interior of India.
4. This, at any rate, is certainly clearly demonstrated in the development of Indo-European l in “Indo-Iranian”: “all of Indo-Iranian tended to confuse r and l …. Every IE l becomes r in Iranian. This same occurence is to be observed in the Northwest of India, and, consequently, in the Rigveda, which is based on idioms of the Northwest.”174
So, is this an “Indo-Iranian” phenomenon? Apparently not: “On the other hand, initial and intervocalic l was present in Indic dialects of other regions. Numerous elements of these dialects were gradually introduced into the literary language, which became fixed in Classical Sanskrit. This explains the appearance of l in more recent parts of the Rigveda and its subsequent rise in frequency.”175
Meillet correctly observes that this is “an instance of concordance of Iranian with the Indic idioms closest to the area of Iranian and discordance with Indic idioms further to the East”.176
The concept of an “Indo-Iranian” branch is based on “the close links between their classical forms - Sanskrit and Avestan respectively”,177 which is the result of a “period of common development”,178 as we have already seen. This period of common development was before the separation of the Vedic and Iranian people.
But this conversion of the original Indo-European l into r is a phenomenon pertaining to this period of common development, and it is not shared by the ancient “Indoaryan” dialects to the east of the Rigvedic area. These dialects, therefore, represent a pre-“Indo-Iranian” phase of Indo-European, which is incompatible with any theory of an Indo-Iranian phase in Central Asia and Afghanistan before the separation of the Indoaryans and Iranians and the consequent migration of Indoaryans into India.
It is also incompatible with any theory of the origin of the “Indoaryan” languages from the Vedic language which forms part of this joint “Indo-Iranian” phase. Therefore, while the word “Indoaryan” may be used in the sense of “Aryan or Indo-European languages historically native to India”, it cannot and should not be used in the sense in which it is generally used: ie. to mean languages descended from a language (Vedic Sanskrit) which, or whose proto-form, shared a joint “Indo-Iranian” phase with Proto-Iranian.
5. The theory that the Indoaryan languages are descended from Vedic Sanskrit is not really corroborated by linguistic factors. As we have pointed out in our earlier book, S.K. Chatterji makes the following remarks about the Old, Middle and New phases of Indoaryan:
“The Aryan came to India, assuredly not as a single, uniform or standardised speech, but rather as a group or groups of dialects… only one of these dialects or dialect-groups has mainly been represented in the language of the Vedas - other dialects… (might) have been ultimately transformed into one or the other of the various New Indo-Aryan languages and dialects. The mutual relationship of these Old Indo-Aryan dialects, their individual traits and number as well as location, will perhaps never be settled… The true significance of the various Prakrits as preserved in literary and other records, their origin and interrelations, and their true connection with the modern languages, forms one of the most baffling problems of Indo-Aryan linguistics… and there has been admixture among the various dialects to an extent which has completely changed their original appearance, and which makes their affiliation to forms of Middle Indo-Aryan as in our records at times rather problematical.”179
Thus S.K. Chatterji unwillingly admits (although he tries to explain it within the framework of the invasion theory) that:
a. There were many different dialects, of which the language of the Rigveda was only one, and that the modern Indoaryan languages may well be descended from these other non-Vedic dialects.
b. The relations (within each chronological group: Old, Middle or New; as well as between different chronological groups) between Old Indoaryan (Rigvedic and Classical Sanskrit, as well as the “other” dialects or dialect groups) and Middle Indoaryan (Prakrits) and the present-day New Indoaryan languages are “baffling” and “problematical” and “will perhaps never be settled”.
The problem will certainly “never be settled” if examined from the viewpoint of an Aryan invasion of India which treats the Indoaryan languages as descended from the languages of people who migrated into India from the northwest after an “Indo-Iranian” phase in Central Asia and an Indo-European phase in South Russia.
As per our theory, Proto-Indo-European, and its earlier forms, developed in the interior of North India. In ancient times, it developed into various dialects, many of which expanded into the northwest and Afghanistan. The divisions of these dialects can be conveniently classified in Puranic terms (howsoever unpalatable it may sound to modern ears) with the dialects of the extreme northwest (which included the ancestral forms of most of the European languages, as well as Hittite and Tocharian) being the Druhyu dialects, the dialects further to their east (mainly the ancestral forms of Iranian, as also Armenian and Greek) being the Anu dialects, and the dialects in the northern parts of North India (Uttar Pradesh, Haryana, and nearby areas) being the PUru dialects (including Vedic). In the interior were other dialects which represented other Puranic groups: Yadus, TurvaSas, IkSvAkus, etc.
With the emigration of the Druhyus, and later the Anus, and the predominant position which the Rigvedic language came to occupy (after the Vedic cult spread all over India, incorporated all the religious systems of the land in the course of time, and became itself the elite layer of an all-inclusive Pan-Indian religious system) in India, began the phase of Indian history which the linguists and historians have interpreted as the “Indoaryan” phase.
The Rigvedic language heavily influenced all the other languages of India, including the languages descended from the remnants of the Outer dialects (Druhyu, Anu), those descended from the Inner dialects (Yadu, TurvaSa, IkSvAku, etc), and also the Dravidian and Austric languages in the South and East.
In turn, the literary forms which developed from the Rigvedic language, Epic and Classical Sanskrit, were heavily influenced by all the other languages (Indo-European, Dravidian and Austric). As Meillet, in a different context (already referred to), puts it: “Numerous elements of these dialects were gradually introduced into the literary language which became fixed in Classical Sanskrit.”180
And finally, as Chatterji correctly puts it: “there has been admixture among the various dialects to an extent which has completely changed their original appearance.”181
To sum up the whole question of the Indo-European homeland:
1. The evidence of archaeology completely disproves, or, at the very least, completely fails to prove, the non-Indian origin of the Indo-Europeans.
2. The evidence of the oldest literary records (the Rigveda and the Avesta) proves the Indian homeland theory from three distinct angles:
a. The evidence of comparative mythology.
b. The evidence of the internal chronology and geography of the Rigveda.
c. The direct evidence in the Rigveda about the emigration of identifiable Indo-European groups from India.
3. The evidence of linguistics, in some matters, is either ambiguous or neutral, and , in some others, definitely confirms the evidence of the literary records which indicate that India was the original homeland.
It is, of course, natural that entrenched scholarship, both in India and in the West, will find it hard to swallow all this evidence, and the conclusions which inevitably and unavoidably arise from it. Especially such scholars as have spent all their lives in ridiculing and rejecting the Indian homeland theory, or in “proving” or corroborating the theory of Aryan invasion or migrations into India.
And it will be particularly hard to swallow because it comes from an Indian - the type of Indian whom they would prefer to brand as a “Hindu fundamentalist”.
The following tongue-in-cheek excerpt from Antoine de Saint-ExupEry’s well known children’s storybook, The Little Prince, illustrates the situation:
“…the planet from which the little prince came is the asteroid known as B-612. This asteroid has only once been seen through a telescope. That was by a Turkish astronomer, in 1909. On making his discovery, the astronomer had presented it to the International Astronomical Congress, in a great demonstration. But he was in Turkish costume, and so nobody would believe what he said. …Fortunately, however, for the reputation of Asteroid B-612, a Turkish dictator made a law that his subjects, under pain of death, should change to European costume. So in 1920 the astronomer gave his demonstration all over again, dressed with impressive style and elegance. And this time everybody accepted his report.”182
The type of attitude satirized by Saint-ExupEry in this imaginary incident is very much a part of world scholarly tendency even today: anyone, Indian or Western, who writes anything, howsoever logical, in support of the Indian homeland theory, represents the “fundamentalist” in his Turkish costume, (or the odd Westerner with a misguided infatuation for this fundamentalism) who deserves only scepticism, ridicule and summary dismissal. Conversely, anyone, Western or Indian, who writes anything, howsoever incredible or ridiculous, in opposition to the Indian homeland theory, represents the “objective scholar” dressed “with impressive style and elegance” in European costume, who deserves a sympathetic hearing and due support.
But the case for an
Indian homeland is so strong, and the case for a non-Indian homeland so weak,
that, inspite of any number of academic dictators decreeing “under pain of
(academic) death” that the Indian homeland theory be abandoned without
serious examination, or with only perfunctory and determinedly sceptical
examination, the academic world will untimately be compelled, nevertheless,
to accept the fact that the Indo-European family of languages originated in
India, or, at the very least, to drastically tone down, or qualify, their
strident rejection of it.
15IASA, preface, p.x.
16ibid., preface, p.xi.
17ibid., preface, p.xiii.
18ibid., preface, p.xiv.
19ibid., preface, p.xii.
20ibid., preface, p.x.
22ibid., preface, p.xiii.
23ibid., preface, p.xv.
24ibid., preface, p.xiii.
32ibid., preface, P.xiii.
33ibid., preface, p.xii.
41ibid., p. 107.
80ibid., p. 59.
81ibid., p. 13.
82ibid., p. 1 5.
84IED, p. 149.
125IE & IE, p.439.
139“On Indo-Europeanization” in the Journal of Indo-European Studies, Spring 1990.
140IE& THE IE.
145The Economic Times, Mumbai, 16/10/96, news-item, “Descent of Cattle”.
147EB, Vol.9, p.438.
149The Times of India, Mumbai, 14/6/87, news-item “Bangani older than Sanskrit”.
150Indogermanische Forschungen, 1 0 1, Band, 1996, p. 107.