Re: The Art of Avoiding History, by Peter Staudenmaier
Posted: Tue Feb 06, 2018 5:57 am
Part 1 of 2
Aryan Origins: Brief History of Linguistic Arguments
by Madhav M. Deshpande
From: India: Historical Beginnings and the Concept of the Aryan: Essays by Romila Thapar, Jonathan Mark Kenoyer, Madhav M. Deshpande, Shereen Ratnagar
© Individual essays with respective authors
© This collection: National Book Trust, India
Published by the Director, National Book Trust, India
NOTICE: THIS WORK MAY BE PROTECTED BY COPYRIGHT
Introduction
The discussion of the terms 'Arya' and 'Anarya' is normally dominated on the one hand by linguistics and archaeology, and, now more increasingly in the context of politics of knowledge as reflected in colonial and post-colonial histories of South Asia. Even in the ancient and classical period which is dominated by the discourses of the Hindu Dharmasastras (and Epics) on the one hand, and the contesting traditions of Buddhism and Jainism, these terms have played an equally significant role expressing conceptions of linguistic, ethnic, moral and spiritual identity, purity, and superiority. I have discussed many of these issues in my previous studies (see: Bibliography). Here, I wish to review primarily the linguistic side of the Aryan question and describe where the linguistic arguments stand at present. In the literature linguistics, the term Aryan is used to refer specifically to those who were speakers of Indo-Iranian languages, the Indo-Iranian itself being a branch of the larger Indo-European language family. The term Arya as a self-referring ethnic term is seen only in the ancient linguistic materials found in Iran and India. The ancient linguistic material of the Indo-Iranian branch consists of the Vedic texts of ancient India, and the Avestan and Old-Persian texts of ancient Iran. Linguistically related material is found in the treaty of Mitannis in Anatolia dating to the 14th century BC, though the word Arya is not found in that region to my knowledge. Thus, in the careful linguistic use of the term Aryan, it is used to refer to speakers of Indo-Iranian languages. However, in less careful, but apparently quite extensive usage, the term is sometimes used to refer to the entire Indo-European language family. With the uncritical equation of language and race in the nineteenth century, the term Aryan came to assume a racial significance, and it was extensively used in that manner by western historians, by the Nazis, and still continues to be used that way by white supremacist groups in many parts of the world, cf. Day (1994). The term Aryan, influenced by its usage in Europe, also appears in racist sense in many Indian writings of the 19th and 20th centuries, cf. Deshpande (2005). In the 19th century India, the term also came to be used to refer to a revivalist purist Hindu identity by groups like the Arya Samaja, and it continues to vibrate with Hindu nationalist groups till today. This widely divergent and multi-valent usage of the term Arya in the available literature of the last two centuries has complicated our understanding of the historical situation, which needs a careful sorting of the various meanings of this term, before one can deal with specific historical reconstructions. I do not believe that we can explain the linguistic make-up of ancient South Asia, without the assumption of migration of the speakers of a branch of Indo-Iranian into the sub-continent and their gradual contact and convergence with speakers of other languages such as Dravidian and Austro-Asiatic. Such a migration does not entail a concerted invasion of the region, such as the invasion by the forces of Alexander the Great, but neither does this necessarily require the imperceptibly slow and non-warlike expansions of agricultural communities as suggested by the alternative Indo-European history of Colin Renfrew (1988).
I respect the caution expressed by Kenneth A.R. Kennedy (1995: 60): "Affirmations as emphatic as those voiced by the Allchins insure that the search for the Aryan presence in linguistic and archaeological sources will survive for some time to come. However, biological anthropologists remain unable to lend support to any of the theories concerning an Aryan biological or demographic entity within the contexts of linguistics and archaeology." Indeed, one should not go by the old identifications of linguistic, religious, and racial identities. However, Kennedy's conclusions are not really contradictory to linguistic research, if we understand that there is no correlation between racial and linguistic identities. People speaking the same language can belong to different racial backgounds, and people of the same racial background may speak different languages. If we dissociate linguistic, cultural, and biological identities, as we need to, then we can certainly find cases of distinct linguistic identities where one would be hard-pressed to find biological differences. The speakers of Indo-Aryan Marathi and Dravidian Kannada can hardly be distinguished from each other on biological grounds, and yet these two languages belong to different linguistic families. There is no reason why the Aryans and non-Aryans of the Vedic period or of Indus Valley could not be linguistically different without necessarily showing up biological differences.
The differences as expressed in the Rgveda by the terms anindra, "those who do not worship Indra," adeva "those who do not worship Devas," ayajyu "those who do not offer sacrifices," sisnadeva "those who worship phallic gods," muradeva "those who worship images of gods" etc. are cultural differences. The term mrdhravac and possibly anas are indicative of linguistic differences. While the outsiders can hardly tell the difference between a Sinhalese and a Sri Lankan Tamil, the differences are perceptible to the populations of Sri Lanka. Similarly, even if our biologists of today cannot distinguish between the skeletons of Aryans and Non-Aryans, there is no reason to deny to them the differences as they perceived; for a critique of Kennedy and others, see Ratnagar (1998) and (1999). When the Rgvedic poets invoke their gods to help the Arya community and destroy the Dasa/ Dasyus, we are dealing with ethnically and culturally perceived differences. Within the Rgvedic period, there appears to have been a gradual move away from the earlier violent conflicts between the Arya varna and the Dasa varna (cf. yo dasam varnam adharam, guha kah, RV II. 12.4) toward some sort of co-operative and collaborative relationship between these two varnas as hinted in RV 1.179.6 : ubhau varnav rsir ugrah puposa (cf. P.V. Kane, History of Dharmasastra, Vol. II, Pt. I, p, 25, fn. 53). In my publications (Deshpande, 1993a), I have dealt with the details of how a slow inter-ethnic convergence may have come about in ancient India, and how the ancient Indian terminology of bija-ksetra "seed-field" provides access to a cultural understanding of biology.
In this connection, we have a lot of terminological dilemmas to deal with. For example, when we use the word "Arya," what do we mean by such a word? We know that the word was used as a self-referring term only by the ancient Iranians and by the authors of the Vedic hymns. Kuiper (1991: 6) says: "[In the RV] 'Aryans' were in general those who maintained the world order by means of sacrifices and gifts .... " Quite apart from the now-unacceptable racial interpretation of this term, Kuiper's definition makes this term purely functional, and stretches it away from any association with a real or perceived ethnicity. On the other hand, consider Southworth's comments: (1995: 261): "Therefore, while it is appropriate to refer to 'speakers of Proto-Dravidian' or 'members of the proto-Dravidian speech community,' or 'speakers of Old Indo-Aryan,' or even (when justified) 'Dravidian-Old Indo-Aryan bilinguals,' use of labels such as 'the Dravidians' or 'the Aryans' is not justifiable in linguistic terms. In fact, their use in any context is problematic, since they can only be justified by demonstrating the existence of a group which both spoke the language in question and possessed particular attributes."
But before getting into a review of the purely linguistic arguments, it is useful to see the range of usage of these terms in premodern literature. For example, the term Arya referring to an ethnic identity occurs in the Rgveda, sometimes in conflict with other communities called Dasas and Dasyus, sometimes in alliances with them as in the famous War of Ten Kings (dasarajna), and sometimes even in a balanced complementarity. We can also usefully compare Manu's Lawbook with the warning issued by the linguist Franklin Southworth (1995: 261). A term like 'Arya,' as the usage of the Dharmasastras shows, was indeed used in linguistic, ethnic, and moral contexts, and that it also had extensions into designations of regions. Manu (10.57) says that one should determine the low-status of a person who looks like an Arya (aryarupam iva), but is an Anarya (anarya), on the basis of his actions (karmabhih svair vibhavayet). This verse shows that there was indeed a prototypical expectation of how an Arya person looks like, i.e. his physical features (aryarupa). In this context, the word rupa "form, appearance" does not appear to have any symbolic meaning. This indicates that physical features which were associated with a prototypical Arya were indeed not restricted to only those who were socially recognized as Arya. Manu, however does not specify the Arya physical features, though Patanjali's Mahabhasya (on P.2.2.6) does say that normally one would not recognize a person "dark like the heap of Urad lentils" (masarasivarna) sitting in the market place to be a Brahmin, and that a Brahmin is normally expected to look gaura "fair," sucyacara "with clean conduct," and pingalakesa "With tawny colored hair." Manu recognizes that confusion extended also to the use of languages. Manu (10.45: mukhabahurupajjanam ya loke jatayo bahih/mlecchavacas caryavacas te sarve dasyavah smrtah //) indicates that there are people all over the known world who speak Arya and Mleccha languages, but the use of language cannot be taken as a factor determining their social identity. Outside the four varnas born from the head, arms, thighs, and feet of the creator, everyone is a barbarian Dasyu, whether he speaks an Arya or a Mleccha language. This statement clearly says that speaking an Arya language does not make a Mleccha/Dasyu into an Arya. This gives us some sense of the baggage the term Arya carries with it, and the baggage that we must recognize before a clearer sense of history can begin to emerge.
Beginnings of Indo-European Linguistics
Glimpses of the beginning of the modern study of Indian languages opened with encounters of Europeans with India. P. Sassetti, who was in Goa between 1581 and 1588, noted that Sanskrit had many words in common with Italian "particularly in the numerals six to nine, in the names for God, snake, and many others," Lockwood (1971: 22). In 1767, P. Coeurdous asked the French Academy for the reasons for the striking similarities between Sanskrit, Greek and Latin, and he suggested that such similarities were relics of the primitive language of mankind after the confusion of tongues at Babel. He compares, for instance: Skt. danam 'gift,' devas 'god,' janu 'knee,' madhyas 'middle,' and vidhava 'widow' with Lat. donum, deus, genu, and vidua, and also the present tense of the verb 'to be' : Skt. Sg. 1 asmi, 2 asi, 3 asti, Pl. 1 smas, 2 stha, 3 santi, with Lat. sum, es, est, sumus, estis, and sunt. However, the credit for suggesting a common bond in the clearest terms goes to Sir William Jones, who declared in 1786 before the Royal Asiatic Society of Bengal in Calcutta:
It should be noted that while William Jones argued for a common ancestral bond between Sanskrit, Greek, Latin, as well as Germanic, Celtic, and Iranian languages, his words are important: "sprung from some common source, which perhaps no longer exists." As Trautmann (1997, 1998) has demonstrated in detail, Jones was still working under the Biblical ideas of dispersal of languages after Babel, and did not look upon Sanskrit itself as the source of all Indo-European languages. It was Friedrich von Schlegel (1772- 1829) who in his work Uber die Sprache und Weisheit der Indier published in 1808 declared Sanskrit to be the parent language, of which Greek, Latin, Persian, and the members of the Germanic branch were the more or less degenerate descendants, and more remotely connected were Armenian, Slavonic and Celtic languages. Schlegel based his conclusions on a comparative study of the vocabulary and grammatical structure of these languages. While the comparative method he initiated survived, his conclusions were soon overthrown. Laying the systematic foundations of comparative Indo-European philology, Franz Bopp in his Uber das Konugationssystem der Sanskritsprache in Vergleichung mit jenen der griechischen, lateinischen, persischen und germanischen Sprache (1816) defined the position of Sanskrit within the Indo-European family: "I do not believe that Greek, Latin and the other European languages are to be considered as derived from Sanskrit ... I feel rather inclined to consider them altogether as subsequent variations of one original tongue, which Sanskrit has preserved more perfectly." In 1833, Bopp published his monumental work on Comparative Indo-European grammar: Vergleichende Grammatik des Sanskrit, Zend, Armenischen, Griechischen, Lateinischen, Lithauischen, Altslawischen, Gothischen und Deutschen. While William Jones deserves the credit for announcing the discovery of the genetic relationship between Indo-European languages, it was in Germany under Bopp and his successors that the science of Indo-European philology progressed with great speed.
A. Schleicher (1821-68) initiated the reconstruction of prehistoric Indo-European forms. By comparing the forms recorded in the daughter languages, one can imagine how the word looked in the parent Indo-European language. By comparing words for horse, i.e. Ski:. asvas, Gk. hippos, also ikkos, arid Lat. equus, archac equos, and by assuming that Sanskrit has kept the vowels of the parent language, Schleicher postulated *akvas as the word for horse in the parent Indo-European language. Subsequent research pointed out that the reconstruction should rather be *ekwos, since it had by then become clear that the vowel system of Sanskrit, far from faithfully reflecting the prehistoric stage, had in fact undergone sweeping changes which resulted in the transformation of both original e and o into a. Gradually it became clear that in certain respects the languages of Europe were more archaic than Sanskrit. "The discovery of the Law of Palatals was naturally a blow to the prestige of Sanskrit which philologists had hitherto assumed to stand close to the parent tongue" (Lockwood 1971: 28). The results of research in this field were crowned with the work of Karl Brugmann (1849-1919) and B. Delbruck (1842-1922). While other works of this period are now mainly of historical interest, Delbruck's Grundriss der vergleichenden Grammatik der indo-germanischen Sprachen (1886-1916) in five volumes remains an unequalled source of factual information and is indispensable for serious work on Indo-European linguistics. The position of Sanskrit within the Indo-European family as defined since the work of Brugmann and Delbruck has changed little within this field, even though great strides have been made in many areas of historical and comparative linguistics.
In the initial period since William Jones's declaration, it is evident that there was no clear understanding of a distinction between what later came to be called Indo-Aryan languages from the class of Dravidian languages. Among the traditional Sanskrit Pandits, there was a general belief that all languages other than Sanskrit were somehow derived from Sanskrit which was described by Dandin as a Divine language (daivi vak). This was evident from the usage of descriptive terms such as tatsama "words identical with those of Sanskrit," tadbhava "words (transparently) derived from Sanskrit," and desya "regional" words whose derivation from Sanskrit was not transparent. This terminology was widely used not only to describe Prakrit and Apabhram sa languages which historically belong to the Indo-Aryan family, but also to describe the languages that belong to the Dravidian family. I have described elsewhere that in the traditional Mima m sa belief, even the so called Mleccha "foreign" languages were also ultimately derived from Vedic Sanskrit. Early western scholarship after William Jones also went through a phase that did not know distinct origins for Indo-Aryan and Dravidian languages.
19th Century Views: Gauda and Dravida Linguistic Areas
There is a traditional classification of Brahmins of India into two groups. The northern group consisting of five sub-groups is given the title Gauda, while the southern group consisting of five sub-groups is given the title Dravida; for a detailed analysis, see Deshpande (2002). The five sub-groups coming under the term Gauda are: Sarasvata, Kanyakubja, Gauda, Maithila and Utkala. The five sub-groups coming under the term Dravida are: Gurjara, Maharastra, Tailanga, Karnataka and Dravida. The early state of linguistic understanding in the post-Jones period is seen in the discussions of these groups and sub-groups by Colebrooke. In his 1801 article "On the Sanscrit and Pracrit Languages," Henry Thomas Colebrooke brings up this classification in the context of describing the linguistic geography of India. Without a hint that this is a classification of brahmins, albeit in different regions of India, Colebrooke (1801: 226) generalizes this into a classification of the "Hindu nations":
Colebrooke had changed the context of the general classification away from a classification of brahmins to a classification of "Hindu nations" based upon their affiliation with language and regionality. Here too, he detected certain contradictions. He had doubts about the appropriateness of the classification of the Gurjaras among the Dravidas, and the classification of Utkalas among the Gaudas. He says (1801: 229):
What is difficult for us to understand is the exact basis upon which Colebrooke is making these judgments, because, as Trautmann (1997: 147) points out: "He ... derives all major Indian languages from Sanskrit, excepting only those known to have come from the invasions of foreigners."
The fact that Colebrooke moved this classification into the area of linguistic geography set the course for the future Indological use of classification in the works of successors, where a clearer recognition of major differences between Indo-Aryan and Dravidian language families becomes evident. As Trautmann's research shows, The second discovery, that of Dravidian language family, is usually credited to a missionary, Bishop Robert Caldwell, whose classic work, A Comparative Grammar of the Dravidian or South Indian Languages, was published in 1856 .... This, however, is mistaken. Fully forty years previous to Caldwell's book, in the very year in which Bopp was laying the foundations of the comparative study of Indo-European languages in Europe (1816), Francis Whyte Ellis, Collector of Madras, published an elegant proof of the Dravidian language family. He showed that Tamil, Telugu, and Kannada shared a common stock of roots and forms of grammar that were different from those of Sanskrit, a finding he extended also to Malayalam, Tulu, Kodagu, and 'Rajmahali' or Malto, a tribal language of the Gangetic basin, far to the north" (Trautmann 2005: xxiii). The work of Ellis, and more widely the work of Bishop Caldwell, influenced the evolving understanding of the different language families in India, and a new understanding of the terms Gauda and Dravida as language groups in India began to emerge. In Caldwell's work (1856, 1974: 56-57), we see a contrast set up between "Gaurian Idioms" and "Dravidian Idioms." The languages included under the heading Gaurian are Hindi, Bengali, Marathi, Gujarati, and Sindhi. Under the heading Dravidian are listed Tamil, Canarese, Tulu, Malayalam, Telugu, Toda, Kota, Gond, Ku, Rajmahal, Brahui, and Oraon. The departure from the original context in the direction of a language-classification, since the writing of Colebrooke, is so complete in Caldwell that there is no oddity felt in listing Marathi and Gujarati among the Gaurian group. The terms are undergoing a process of redefinition and recontextualization. We see the same process in Rudolf Hoernle's 1880 work A Comparative Grammar of the Gaudian Languages. Hoernle (Introduction, p. i) explains his choice of the term Gaudian: "I have adopted the term Gaudian to designate collectively all North-Indian vernaculars of Sanskrit affinity, for want of a better word; not as being the least objectionable, but as being the most convenient one." Referring to the specific languages, Hoernle (ibid) says: "Seven languages of the Sanskrit stock are usually enumerated as spoken in North India, viz. Sindhi, Gujarati, Punjabi, Hindi, Bangali, Oriya, Marathi." The languages of Maharashtra and Gujarat, coming under the old panca dravidas, are now listed as "languages of the Sanskrit stock ... spoken in North India." There is a newly emerging notion of a linguistic divide between Indo-Aryan and Dravidian, and this divide is expressed by using the old terminology of Gauda vs. Dravida.
Early Indian Responses to Indo-European Linguistics
By the second half of the nineteenth century, there were several theories and counter-theories circulating in India; for details see Deshpande 2005. The newly emerging class of western educated Indians was absorbing the theories of Indo-European philology and Caldwell's notion of the separate Dravidian family, and was reacting to these new ideas in various ways. Early nineteenth century scholarly figures like Sir R. G. Bhandarkar, who was trained in Bombay and Pune under European professors, completely agreed with the conclusions of Indo-European philology of the time, and began to offer explanations of the emergence of various languages in terms of contact between Aryan and non-Aryan populations in India. Bhandarkar's contribution toward an understanding of the Aryan origins may be seen most profoundly in his Wilson Philological Lectures on Sanskrit & the Derived Languages delivered in 1877 in Bombay, and published in form of a book in 1914. Referring to the progress of Indo-European philology especially in Germany, Bhandarkar says: "The progress made within about fifty years is marvelous, and affords a striking instance of the intellectual activity of the Europeans. In the cultivation of Philology and the elaboration of this new science the Germans, of all other nations, have been most prominent, and have done by far the greater portion of the work," (Bhandarkar 1914:5). Bhandarkar's explanations built on the hypothesis of various migrations, contacts, and adaptations are seen in the following discourse: "Though [the speakers of Pali] heard conjunct consonants and the diphthongs ai and au pronounced by the speakers of Sanskrit, as correctly as the other letters which they did not corrupt, their organs were not fitted to utter them. These peculiarities may have been natural or acquired. If natural, the people who first corrupted Sanskrit into Pali must have belonged to an alien race which came into close contact with the Aryas and learnt their language .... And there is another instance in History of an alien race having treated the sounds of the language of a civilized community in just the same way. The Barbarians who overran Italy and developed the Italian from the Latin, showed the same inability to pronounce the Latin conjuncts, and assimilated them as our Pali ancestors did," (Bhandarkar 1914: 47). The process of the emergence of the Prakrits is accounted for by Bhandarkar by referring to the migration of the Aryas from "the land of the five rivers" to "the country known afterwards as Brahmavarta and Kurukshetra." This is the country about Thanesvar, where "they formed a consolidated community in which an aboriginal or alien race was incorporated and the language represented by the Pali was the language of that race," (Bhandarkar 1914: 88). The idea of the Aryas, the speakers of Sanskrit, coming into contact with non-Aryas, and such a contact leading to a degenerative transformation of Sanskrit into Prakrits is an idea not inherently alien to the Sanskritic tradition. But the same tradition does not admit any notion of history for this divine and eternal language, and here Bhandarkar's efforts to find historical origins of and developments in Sanskrit did not go well with his contemporaries. These were departures from the Sanskritic tradition. If the speakers of Sanskrit were Aryas, and if Sanskrit itself resulted from a process of transformation from its Indo-European precursors and underwent later transformations of its own, were the Aryas themselves subject to transformative processes? If the transformations of Sanskrit into Pali were caused by the alien speakers trying to learn the Aryan language, what was it that caused the transformations which resulted into the very existence of Sanskrit itself, and what caused transformations within the very history of Sanskrit? Such questions indeed raise unpalatable issues, and Bhandarkar's own wording suggests that he, as a Brahmin, was himself caught in the middle. Bhandarkar's wording would suggest a belief that the Aryas were not in contact with non-Aryans in "the land of the five rivers." Even admitting that the Indus Civilization was not excavated by this time, one still finds this belief difficult to accept, particularly in view of the fact that there are north western Prakrits in Asokan inscriptions, a fact which was known to Bhandarkar by this time. Bhandarkar finds that the northwestern Prakrits like Paisaci, appear to be truly Aryan. Perhaps then this was the language of an Aryan tribe that had remained longer in the original seat of the race, and was connected with the ancestors of the Teutons, so as to develop a phonetic peculiarity resembling theirs, and emigrated to India at a very late period and settled on the borders. Or it might be that the tribe came to India along with the others, but living in the mountainous countries on the border in a sort of rude independence, it developed this peculiarity of pronunciation .... Since under this supposition they could not have come in very close contact with their more civilized brethren of the plains, their language did not undergo some of those phonetic modifications which Sanskrit underwent in the mouth of the aboriginal races," (Bhandarkar 1914: 94). So the speakers of northwestern Prakrits were truly Aryan, uncontaminated by contact with the non-Aryans, but not as civilized as the speakers of Sanskrit who did come in contact with the "aboriginal races"? Bhandarkar seems to believe that the Aryans were not in contact with aboriginal races until they moved from "the land of five rivers" into the interior of India. Thus the linguistic deviations from Sanskrit in the direction of Pali and other Prakrits are caused by the contact of the Aryas with the non-Aryas, while the linguistic deviations from Sanskrit in the direction of the northwestern Prakrits and other Indo-European languages are to be explained by "isolation" at best. This creates an interesting tripartite division: the ethnically pure but less civilized Aryas of the northwest, the pure civilized Aryas of the "land of the five rivers," and the uncivilized non-Aryas of the rest of the Indian subcontinent.
Bhandarkar's student at Deccan College, B.G. Tilak (1856-1920) disagreed with Bhandarkar's moderate political views and reformist social views, but basically accepted the conclusions of Indo-European philology, with a twist of his own. His first publication on the subject of ancient history was the book Orion or Researches into the Antiquity of the Vedas, published in 1893. His second publication was The Arctic Home in the Vedas, published in 1903. His third and the last publication on the subject was the book Vedic Chronology and Vedanga Jyotisha [Containing also Chaldean and Indian Vedas and other miscellaneous essays], written in 1913 in the Mandalay jail in Burma and published posthumously in 1925. Trying to find history in the anadi "beginningless" Vedas was a departure from the Brahmanical tradition, but such a departure, facilitated even though it was by the introduction of Western education, was the path chosen by Tilak.
Tilak was in basic agreement with Bhandarkar on the conclusions of Indo-European linguistic theory, and had particular admiration for the work of Max Muller. He viewed Max Muller not only as a great scholar of the Veda, but as someone who pointed out to the west the importance of the spiritual contribution of India (cf. Tilak on Max Muller, Lokmanya Tilakamce Kesaritil Samkina Lekh, vol. 4, 1922, 124ff. Also see Max Muller's obituary written by Tilak in Kesari, November 6, 1900, ibid, pp. 573ff.). In his obituary of Max Muller in Kesari (ibid, pp. 575-6), Tilak approvingly points out Muller's contributions to Indo-European philology and comparative mythology:
However, while Tilak admired Max Muller, he believed that the purely philological approach of scholars like Max Muller was not sufficient to yield convincing results and that the astronomical method, though generally condemned by the philologists as being indefinite and unreliable, could be refined and used to make more exact predictions. In his own works based not so much on philology, but on astronomical evidence, Tilak argued that the ancestors of the Vedic Aryans, and possibly texts of the Vedas themselves came from the Arctic Home dating to a very ancient period of 8000 to 10000 Be. If one were to accept Tilak's dating of the Vedas and the ancestors of the Vedic Aryans, then the non-Sanskrit branches of Indo-European language family would all seem to be far younger than the Indo-Aryan branch. While this conclusion would make the Indian Aryans the senior, rather than junior, brothers of the colonial Western Aryans, and would boost the nationalist pride, Tilak did not systematically attempt to explain how he would derive all other branches of Indo-European from Sanskrit. Without explicitly claiming so, Tilak makes Sanskrit the oldest of all Indo-European languages, and hence a possible source of them all. As I have discussed elsewhere in detail (cf. "Arctic Home in the Vedas: Religion, Politics, and the Colonial Context," forthcoming), in spite of Tilak's political background, by the standards of the late 19th and early 20th century scholarship in the west, Tilak's geological and astronomical investigations had better scholarly foundations, than the openly Hindutva-motivated histories constructed by Savarkar, Golwalkar, and others.
However, in the nineteenth century Maharashtra, there was a continuing Brahmanical tradition that believed in the primordiality of Sanskrit and India as its homeland and holy-land, combined with a belief that all other languages were derived from Sanskrit through a process of degeneration (apabhramtsa-ization). This goes back to Manu's Lawbook that all populations of the world ultimately originate from the four primordial vannas created by God in India, through mixing of the original varnas and through migrations of various groups away from the Aryavartic center combined with loss of dharmic behavior (Manusmrti, chapter 10). Manu (10.43-44) lists Paundraka, Caudra, Dravida, Kamboja, Yavana, Saka, Parada, Pahlava, Cina, Kirata, Darada, and Khasa as original Ksatriyas who slowly became Sudras through loss of dharmic behavior and through non-contact with Brahmins (kriyalopat ... brahmanadarsanena ca, Manu 10.43). This Brahmanical view, combined with the appropriation of modern knowledge of the Indo-European family and the emerging sense of Indian nationalism, appears in the works of authors like Narayan Pavgee, a contemporary of Tilak. While Tilak argued for the Arctic Home of the Vedic Aryans, Pavgee argued for the Aryavartic Home of the Aryans, though he was willing to accept a theory that a branch of the Aryavartic Aryans migrated to the Arctic region and eventually returned to India. With the Aryavartic home of the Aryans, Pavgee was arguing against the notion that the Aryans came to India from outside. However, he was proposing that the Aryans spread across the world from the Aryavartic home, and that Sanskrit was the mother of all languages of the world.
The second volume of his Bharatiya Samrajya (1893) is titled: Aryalok va tyance Buddhi-vaibhav "The Aryan People and the Wealth of their Intelligence." The book begins with a section dealing with the original home of the Aryas (aryance mula-nivasa-sthana). Pavgee asserts at the very beginning that the Vedas are the oldest literature of mankind, and that the Vedas support the notion that northern India is the original home of not just the Aryans, but of the entire mankind (Pavgee 1893: 2). The Aryan family of languages originated in India, Pavgee (1893: 5) asserts, and it expanded westward from India through the regions of Iran, Greece, Italy, Spain, England, Germany, and Russia. Pavgee mentions the view of western philologists that the Aryans came into India from outside, but he does not support this view. In the ninth volume of his Bharatiya Samrajya (1900), titled Bharatakhandatil Nanavidha Bhasa "Various Languages in India," Pavgee has a section on the original land of the Arya language, i.e. Sanskrit. Pavgee asserts unequivocally that Sanskrit originated in the region of Aryavarta within India and is the mother of all Aryan languages. All languages such as Marathi, Hindi, Bengali, Gujarati, Iranian, Greek, Latin, German, English, and Polish were born from Sanskrit. They are daughters of Sanskrit (Pavgee 1900: 14). Various originally Arya groups left their religion and castes and left the original Arya homeland, and these eventually became the various branches of the Aryan language family. In support of this conclusion, Pavgee offers, not linguistic arguments, but a large number of passages from Smrtis and Puranas (1900: 16ff), a practice continued in nationalist history writing till today, see papers in Deo and Kamath (1993).
From Linguistic Reconstruction to Racist Theories in the west
Returning to the history of western scholarship during the nineteenth century and the early twentieth century, one needs to keep in mind one major factor, namely that the Indus Valley Civilization was discovered only in 1920s, and hence all previous scholarship of the Indo-European philology developed in the absence of any knowledge of the Indus Valley Civilization. While the discovery of the Indo-European language family was to a large extent a result of linguistic scholarship, the scholars who were engaged in this scholarship often used discourses that merged the distinction between language and ethnicity or race, (Day, 1994, Leach, 1990, and Trautmann 1997 and 1999). Beginning with William Jones, we find the use of terms like "nations" to refer to groups speaking particular languages, and words like "Semitic" and "Hamitic" combined both language and ethnicity, (Trautmann 1997 and 1998). With this mixture of terminology, the notion of a linguistic ancestor for the languages of the Indo-European family soon developed into finding an ancestral group of people, a nation or a race, that over time splintered into multiple sub-groups and these migrated from an original homeland to their eventual destinations where we find the descendant languages today. Today one can say with confidence that there is no necessary connection between language and ethnicity. The African blacks who speak French are linguistically as much Indo-European as white French speakers in France, just as the English-speaking peoples across the world speak an Indo-European language, though they belong to different races and ethnicities. Thus, linguistic identities are not necessarily related to racial or ethnic identities; the language of race was pervasively used by scholars in the nineteenth and twentieth centuries. In early writings, Max Muller used the word Aryan in racial meaning, even though he later recanted: "There are Aryan and Semitic languages, [but] it is against all logic to speak, without an expressed or implied qualification, of an Aryan race, Aryan blood, or Aryan skulls," Max Muller (1880, 4: 223). Edmund Leach (1990: 235) points out that by 1878 Max Muller was writing without equivocation about Aryans as if they were a racial group moving outward from some central Asian homeland by a process of conquest. In the 1890's, the linguist Sir George Grierson was explaining all Indo-Aryan dialect distributions in northern India as due to past military conquest, a view that was repeated without criticism in the 1969 printing of the Encyclopaedia Britannica, Leach (1990: 235). We see the same pattern in the works of leading European scholars like Hermann Oldenberg. Consider the following narrative from Oldenberg (1896, 1973: 39):
Such passages can be multiplied a hundred times indicating how wide the use of the racial terminology was. The racial terminology took different forms, e.g. the Aryans versus Semites in Germany, while in the context of colonial domains like India, the racial terminology made the relationships between the white Aryan British and the black Aryan Bengali ever more complicated -- sometimes positively, though most of the time negatively. Does this mean that one must not connect migrations of peoples to linguistic changes, for fear of involving ourselves in racialist thinking? I do not think so. Theorizing from where we stand today, we need to take into account all forms of human interaction that involves language to evaluate the phenomenon of linguistic relationships, genetic or otherwise, as well as phenomena like various forms of intensive and extensive bilingualism and diglossia. The social context of language in its widest sense needs to be taken into account to explain facts language, without insisting that a linguistic identity must be co-terminus with racial, biological, religious, or political identities. The different identities do indeed intersect, but do not logically entail each other. These various forms of identities do affect language use, and may be taken into account with due caution.
Historical Linguistics after the Discovery of the Indus Valley Civilization
With the discovery of the Indus Civilization in 1920s began a new effort to make sense of the Indo-European migrations and the Vedic texts. Early on, the destruction of the Indus cities was connected with the Vedic Aryans led by their God Indra destroying the forts of Dasas and Dasyus. If this version of history is accepted, Indus was pre-Vedic and non-Vedic and was destroyed by Vedic Aryans. This most likely makes Indus a non-Indo-European linguistic region, and hence for decades various scholars, like Emeneau and Parpola, argued for a Dravidian identity for Indus. This was also supported by the consideration that Brahui in Baluchistan is a Dravidian language, and further by the claim of David McAlpin that Dravidian and Elamite of ancient Iran may belong to the same language family. Even though gradually the opinion shifted to the position that the Vedic Aryans entered the Indus area after the Indus cities had already declined and abandoned, the predominant linguistic opinion still held that the Indus Valley was a Dravidian speaking region. Decipherments of the Indus Seals by S.R. Rao and Madhusudan Misra, among others have claimed a form of proto-Sanskrit as the language of Indus, and the Vedic character of Indus civilization, though there is not a wide acceptance of these proposals, cf. for a review of recent proposals, see: Ratnagar (1996), Witzel (2002) and Mahadevan (2002). More recently scholars Michael Witzel have suggested that the area was most likely a multi-lingual multi-ethnic region that may have included Dravidian, proto- or para-Munda, as well other languages (related to isolates like Burushaski), including forms of Indo-Iranian on the periphery. This suggestion seems to be most realistic to me, as it allows an open-ended examination of all available linguistic data.
As for the latest twist on the decipherment of the Indus seals, Farmer-Sproat-Witzel (2004) proposes that the Indus seal symbols do not codify a language: "Archaeologists have long claimed the Indus Valley as one of the four literate centers of the early ancient world, complete with long texts written on perishable materials. We demonstrate the impossibility of the lost-manuscript thesis and show that Indus symbols were not even evolving in linguistic directions after at least 600 years of use. Suggestions as to how Indus symbols were used are noted in nonlinguistic symbol systems in the Near East that served key religious, political, and special functions without encoding speech or serving as formal memory aids. Evidence is reviewed that the Harappans' lack of a true script may have been tied to the role played by their symbols in controlling large multilinguistic populations; parallels are drawn to the later resistance of Brahmin elites to the literate encoding of Vedic sources and to similar phenomena in esoteric traditions outside South Asia (Abstract, p. 19)." While a nonlinguistic interpretation of Indus symbols is a distinct possibility that needs to be taken into account, the initial reactions to the above proposal from scholars do not seem to indicate a wide acceptance. In any case, whether because the available attempts to decipher the linguistic value of Indus symbols have all failed to reach a convincing result, or because they are intrinsically nonlinguistic symbols, so far the symbols of Indus do not directly contribute to our linguistic understanding of history, which must be teased out only from an intensive mining of linguistic sources such as the Vedic and Iranian texts, and this brings us back to the work of scholars like Emeneau, Burrow, Kuiper, Mayrhofer, Southworth, Witzel, Parpola, and Lubotsky, among others. The substratum languages in the area covered by the Indus Valley Civilization probably include Dravidian, (Para-) Munda, as well as ancestors of linguistic isolates like Burushaski, plus some traces of the languages the Bactria Margiana Archaeological Complex, as well as early forms of Iranian Indo-Aryan in the border areas. Southworth (2005: 328-9) says: "The Indus language (group) is probably the oldest detectable linguistic stage, both in Panjab and Sindh. An unidentified western Austro-Asiatic language designated as Para-Munda probably functioned as a lingua franca of this area. Some of these language groups may have been present as early as the seventh or eighth millennium Be, when archaeologists record the beginnings of agriculture in areas bordering the Indus Valley. Indo-Aryan appears in the form Vedic Sanskrit, dateable to the early second millennium Be, probably preceded by outer Indo-Aryan. The arrival of Dravidian-speaking people in the area is difficult to date and is the subject of some controversy. Evidence of Dravidian borrowings in the Rigveda dates only from about 1200 Be, according to some sources, yet it is nevertheless possible that Dravidian languages were spoken in Sindh and Saurashtra (and perhaps even in Panjab) considerably earlier. Ultimately, only Indo-Aryan survives in these areas, though Sindh shows traces of outer Indo-Aryan." Also noteworthy are Witzel's two 1999 articles "Substratum Languages in Old Indo-Aryan" and "Early Sources for South Asian Substrate Languages," as well as the forthcoming article "The Languages of Harappa."
New Linguistic Approach in Mid-20th Century: "India as a Linguistic Area"
The most significant development of the middle of the twentieth century linguistic study of India was the idea of "India as a Linguistic Area" seen in the early works of Murray B. Emeneau and F.B.J. Kuiper, and eventually consolidated in the work of Nicholas Masica. Until the emergence of the notion of Linguistic Area, the only dominant model was the genetic model used in Indo-European and other forms of historical linguistics. The genetic model explains the common elements between genetically related languages as being a reflection of common ancestry. Where the genetically related languages differ, those differences may have developed either internally in those branches, or they may be due to contact with other languages of different regions etc. But what about a demonstrably large number of common features shared by languages that evidently do not belong to the same language family? Where do these features and their commonality originate? The new direction proposed by areal linguistics was that diffusion of linguistic features across genetically unrelated languages can occur through intense and prolonged contact. Given sufficient intensity of contact and depth of time, genetically unrelated languages of a given geographical region begin to exchange phonetic, syntactic, and other features with each other, and over time develop a significant set of such shared features. The research work of Emeneau, Kuiper, Masica and others has demonstrated how this process occurred in the Indian subcontinent so that languages belonging to different families like Indo-Aryan, Dravidian, and Austro-Asiatic came to share a large number of common linguistic features through diffusion, rather than common ancestry. Thus, genetic relationships between languages as well as processes of diffusion among languages of a given linguistic area are all working simultaneously to produce the resulting state of a language or languages, and these different strands in the make-up of languages need to be analytically separated in order to get a proper sense of linguistic history. Franklin Southworth (1974) had previously used the term "linguistic stratigraphy" for this sort of linguistic analysis, and his recently published book (2005) calls it "linguistic archaeology," where he discusses how one can begin to peel off or dig into various layers of features acquired by a language or languages.
By the beginning of the second half of the twentieth century, especially with the researches of scholars like Emeneau, Burrow, and Kuiper, who added serious consideration of Dravidian and Austro-Asiatic languages and their co-existence in the Indian subcontinent with languages of the Indo-Aryan branch of the Indo-European family, many issues were newly brought into a sharper focus. In his 1975 article "Substratum Influence on (Rg-Vedic) Sanskrit," Hans Hock conveniently summarizes the conclusions of this research (p. 76-78):
Aryan Origins: Brief History of Linguistic Arguments
by Madhav M. Deshpande
From: India: Historical Beginnings and the Concept of the Aryan: Essays by Romila Thapar, Jonathan Mark Kenoyer, Madhav M. Deshpande, Shereen Ratnagar
© Individual essays with respective authors
© This collection: National Book Trust, India
Published by the Director, National Book Trust, India
NOTICE: THIS WORK MAY BE PROTECTED BY COPYRIGHT
YOU ARE REQUIRED TO READ THE COPYRIGHT NOTICE AT THIS LINK BEFORE YOU READ THE FOLLOWING WORK, THAT IS AVAILABLE SOLELY FOR PRIVATE STUDY, SCHOLARSHIP OR RESEARCH PURSUANT TO 17 U.S.C. SECTION 107 AND 108. IN THE EVENT THAT THE LIBRARY DETERMINES THAT UNLAWFUL COPYING OF THIS WORK HAS OCCURRED, THE LIBRARY HAS THE RIGHT TO BLOCK THE I.P. ADDRESS AT WHICH THE UNLAWFUL COPYING APPEARED TO HAVE OCCURRED. THANK YOU FOR RESPECTING THE RIGHTS OF COPYRIGHT OWNERS.
Introduction
The discussion of the terms 'Arya' and 'Anarya' is normally dominated on the one hand by linguistics and archaeology, and, now more increasingly in the context of politics of knowledge as reflected in colonial and post-colonial histories of South Asia. Even in the ancient and classical period which is dominated by the discourses of the Hindu Dharmasastras (and Epics) on the one hand, and the contesting traditions of Buddhism and Jainism, these terms have played an equally significant role expressing conceptions of linguistic, ethnic, moral and spiritual identity, purity, and superiority. I have discussed many of these issues in my previous studies (see: Bibliography). Here, I wish to review primarily the linguistic side of the Aryan question and describe where the linguistic arguments stand at present. In the literature linguistics, the term Aryan is used to refer specifically to those who were speakers of Indo-Iranian languages, the Indo-Iranian itself being a branch of the larger Indo-European language family. The term Arya as a self-referring ethnic term is seen only in the ancient linguistic materials found in Iran and India. The ancient linguistic material of the Indo-Iranian branch consists of the Vedic texts of ancient India, and the Avestan and Old-Persian texts of ancient Iran. Linguistically related material is found in the treaty of Mitannis in Anatolia dating to the 14th century BC, though the word Arya is not found in that region to my knowledge. Thus, in the careful linguistic use of the term Aryan, it is used to refer to speakers of Indo-Iranian languages. However, in less careful, but apparently quite extensive usage, the term is sometimes used to refer to the entire Indo-European language family. With the uncritical equation of language and race in the nineteenth century, the term Aryan came to assume a racial significance, and it was extensively used in that manner by western historians, by the Nazis, and still continues to be used that way by white supremacist groups in many parts of the world, cf. Day (1994). The term Aryan, influenced by its usage in Europe, also appears in racist sense in many Indian writings of the 19th and 20th centuries, cf. Deshpande (2005). In the 19th century India, the term also came to be used to refer to a revivalist purist Hindu identity by groups like the Arya Samaja, and it continues to vibrate with Hindu nationalist groups till today. This widely divergent and multi-valent usage of the term Arya in the available literature of the last two centuries has complicated our understanding of the historical situation, which needs a careful sorting of the various meanings of this term, before one can deal with specific historical reconstructions. I do not believe that we can explain the linguistic make-up of ancient South Asia, without the assumption of migration of the speakers of a branch of Indo-Iranian into the sub-continent and their gradual contact and convergence with speakers of other languages such as Dravidian and Austro-Asiatic. Such a migration does not entail a concerted invasion of the region, such as the invasion by the forces of Alexander the Great, but neither does this necessarily require the imperceptibly slow and non-warlike expansions of agricultural communities as suggested by the alternative Indo-European history of Colin Renfrew (1988).
I respect the caution expressed by Kenneth A.R. Kennedy (1995: 60): "Affirmations as emphatic as those voiced by the Allchins insure that the search for the Aryan presence in linguistic and archaeological sources will survive for some time to come. However, biological anthropologists remain unable to lend support to any of the theories concerning an Aryan biological or demographic entity within the contexts of linguistics and archaeology." Indeed, one should not go by the old identifications of linguistic, religious, and racial identities. However, Kennedy's conclusions are not really contradictory to linguistic research, if we understand that there is no correlation between racial and linguistic identities. People speaking the same language can belong to different racial backgounds, and people of the same racial background may speak different languages. If we dissociate linguistic, cultural, and biological identities, as we need to, then we can certainly find cases of distinct linguistic identities where one would be hard-pressed to find biological differences. The speakers of Indo-Aryan Marathi and Dravidian Kannada can hardly be distinguished from each other on biological grounds, and yet these two languages belong to different linguistic families. There is no reason why the Aryans and non-Aryans of the Vedic period or of Indus Valley could not be linguistically different without necessarily showing up biological differences.
The differences as expressed in the Rgveda by the terms anindra, "those who do not worship Indra," adeva "those who do not worship Devas," ayajyu "those who do not offer sacrifices," sisnadeva "those who worship phallic gods," muradeva "those who worship images of gods" etc. are cultural differences. The term mrdhravac and possibly anas are indicative of linguistic differences. While the outsiders can hardly tell the difference between a Sinhalese and a Sri Lankan Tamil, the differences are perceptible to the populations of Sri Lanka. Similarly, even if our biologists of today cannot distinguish between the skeletons of Aryans and Non-Aryans, there is no reason to deny to them the differences as they perceived; for a critique of Kennedy and others, see Ratnagar (1998) and (1999). When the Rgvedic poets invoke their gods to help the Arya community and destroy the Dasa/ Dasyus, we are dealing with ethnically and culturally perceived differences. Within the Rgvedic period, there appears to have been a gradual move away from the earlier violent conflicts between the Arya varna and the Dasa varna (cf. yo dasam varnam adharam, guha kah, RV II. 12.4) toward some sort of co-operative and collaborative relationship between these two varnas as hinted in RV 1.179.6 : ubhau varnav rsir ugrah puposa (cf. P.V. Kane, History of Dharmasastra, Vol. II, Pt. I, p, 25, fn. 53). In my publications (Deshpande, 1993a), I have dealt with the details of how a slow inter-ethnic convergence may have come about in ancient India, and how the ancient Indian terminology of bija-ksetra "seed-field" provides access to a cultural understanding of biology.
In this connection, we have a lot of terminological dilemmas to deal with. For example, when we use the word "Arya," what do we mean by such a word? We know that the word was used as a self-referring term only by the ancient Iranians and by the authors of the Vedic hymns. Kuiper (1991: 6) says: "[In the RV] 'Aryans' were in general those who maintained the world order by means of sacrifices and gifts .... " Quite apart from the now-unacceptable racial interpretation of this term, Kuiper's definition makes this term purely functional, and stretches it away from any association with a real or perceived ethnicity. On the other hand, consider Southworth's comments: (1995: 261): "Therefore, while it is appropriate to refer to 'speakers of Proto-Dravidian' or 'members of the proto-Dravidian speech community,' or 'speakers of Old Indo-Aryan,' or even (when justified) 'Dravidian-Old Indo-Aryan bilinguals,' use of labels such as 'the Dravidians' or 'the Aryans' is not justifiable in linguistic terms. In fact, their use in any context is problematic, since they can only be justified by demonstrating the existence of a group which both spoke the language in question and possessed particular attributes."
But before getting into a review of the purely linguistic arguments, it is useful to see the range of usage of these terms in premodern literature. For example, the term Arya referring to an ethnic identity occurs in the Rgveda, sometimes in conflict with other communities called Dasas and Dasyus, sometimes in alliances with them as in the famous War of Ten Kings (dasarajna), and sometimes even in a balanced complementarity. We can also usefully compare Manu's Lawbook with the warning issued by the linguist Franklin Southworth (1995: 261). A term like 'Arya,' as the usage of the Dharmasastras shows, was indeed used in linguistic, ethnic, and moral contexts, and that it also had extensions into designations of regions. Manu (10.57) says that one should determine the low-status of a person who looks like an Arya (aryarupam iva), but is an Anarya (anarya), on the basis of his actions (karmabhih svair vibhavayet). This verse shows that there was indeed a prototypical expectation of how an Arya person looks like, i.e. his physical features (aryarupa). In this context, the word rupa "form, appearance" does not appear to have any symbolic meaning. This indicates that physical features which were associated with a prototypical Arya were indeed not restricted to only those who were socially recognized as Arya. Manu, however does not specify the Arya physical features, though Patanjali's Mahabhasya (on P.2.2.6) does say that normally one would not recognize a person "dark like the heap of Urad lentils" (masarasivarna) sitting in the market place to be a Brahmin, and that a Brahmin is normally expected to look gaura "fair," sucyacara "with clean conduct," and pingalakesa "With tawny colored hair." Manu recognizes that confusion extended also to the use of languages. Manu (10.45: mukhabahurupajjanam ya loke jatayo bahih/mlecchavacas caryavacas te sarve dasyavah smrtah //) indicates that there are people all over the known world who speak Arya and Mleccha languages, but the use of language cannot be taken as a factor determining their social identity. Outside the four varnas born from the head, arms, thighs, and feet of the creator, everyone is a barbarian Dasyu, whether he speaks an Arya or a Mleccha language. This statement clearly says that speaking an Arya language does not make a Mleccha/Dasyu into an Arya. This gives us some sense of the baggage the term Arya carries with it, and the baggage that we must recognize before a clearer sense of history can begin to emerge.
Beginnings of Indo-European Linguistics
Glimpses of the beginning of the modern study of Indian languages opened with encounters of Europeans with India. P. Sassetti, who was in Goa between 1581 and 1588, noted that Sanskrit had many words in common with Italian "particularly in the numerals six to nine, in the names for God, snake, and many others," Lockwood (1971: 22). In 1767, P. Coeurdous asked the French Academy for the reasons for the striking similarities between Sanskrit, Greek and Latin, and he suggested that such similarities were relics of the primitive language of mankind after the confusion of tongues at Babel. He compares, for instance: Skt. danam 'gift,' devas 'god,' janu 'knee,' madhyas 'middle,' and vidhava 'widow' with Lat. donum, deus, genu, and vidua, and also the present tense of the verb 'to be' : Skt. Sg. 1 asmi, 2 asi, 3 asti, Pl. 1 smas, 2 stha, 3 santi, with Lat. sum, es, est, sumus, estis, and sunt. However, the credit for suggesting a common bond in the clearest terms goes to Sir William Jones, who declared in 1786 before the Royal Asiatic Society of Bengal in Calcutta:
"The Sanskrit language, whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either; bearing to both of them a stronger affinity, both in the roots of verbs, and in the forms of grammar, could possibly have been produced by accident; so strong indeed that no philologer could examine them all three without believing them to have sprung from some common source, which perhaps no longer exists. There is similar reason, though not quite so forcible, for supposing that both the Gothick and the Celtick, though blended with a very different idiom, had the same origin with Sanskrit; and the old Persian might be added to the same family."
It should be noted that while William Jones argued for a common ancestral bond between Sanskrit, Greek, Latin, as well as Germanic, Celtic, and Iranian languages, his words are important: "sprung from some common source, which perhaps no longer exists." As Trautmann (1997, 1998) has demonstrated in detail, Jones was still working under the Biblical ideas of dispersal of languages after Babel, and did not look upon Sanskrit itself as the source of all Indo-European languages. It was Friedrich von Schlegel (1772- 1829) who in his work Uber die Sprache und Weisheit der Indier published in 1808 declared Sanskrit to be the parent language, of which Greek, Latin, Persian, and the members of the Germanic branch were the more or less degenerate descendants, and more remotely connected were Armenian, Slavonic and Celtic languages. Schlegel based his conclusions on a comparative study of the vocabulary and grammatical structure of these languages. While the comparative method he initiated survived, his conclusions were soon overthrown. Laying the systematic foundations of comparative Indo-European philology, Franz Bopp in his Uber das Konugationssystem der Sanskritsprache in Vergleichung mit jenen der griechischen, lateinischen, persischen und germanischen Sprache (1816) defined the position of Sanskrit within the Indo-European family: "I do not believe that Greek, Latin and the other European languages are to be considered as derived from Sanskrit ... I feel rather inclined to consider them altogether as subsequent variations of one original tongue, which Sanskrit has preserved more perfectly." In 1833, Bopp published his monumental work on Comparative Indo-European grammar: Vergleichende Grammatik des Sanskrit, Zend, Armenischen, Griechischen, Lateinischen, Lithauischen, Altslawischen, Gothischen und Deutschen. While William Jones deserves the credit for announcing the discovery of the genetic relationship between Indo-European languages, it was in Germany under Bopp and his successors that the science of Indo-European philology progressed with great speed.
A. Schleicher (1821-68) initiated the reconstruction of prehistoric Indo-European forms. By comparing the forms recorded in the daughter languages, one can imagine how the word looked in the parent Indo-European language. By comparing words for horse, i.e. Ski:. asvas, Gk. hippos, also ikkos, arid Lat. equus, archac equos, and by assuming that Sanskrit has kept the vowels of the parent language, Schleicher postulated *akvas as the word for horse in the parent Indo-European language. Subsequent research pointed out that the reconstruction should rather be *ekwos, since it had by then become clear that the vowel system of Sanskrit, far from faithfully reflecting the prehistoric stage, had in fact undergone sweeping changes which resulted in the transformation of both original e and o into a. Gradually it became clear that in certain respects the languages of Europe were more archaic than Sanskrit. "The discovery of the Law of Palatals was naturally a blow to the prestige of Sanskrit which philologists had hitherto assumed to stand close to the parent tongue" (Lockwood 1971: 28). The results of research in this field were crowned with the work of Karl Brugmann (1849-1919) and B. Delbruck (1842-1922). While other works of this period are now mainly of historical interest, Delbruck's Grundriss der vergleichenden Grammatik der indo-germanischen Sprachen (1886-1916) in five volumes remains an unequalled source of factual information and is indispensable for serious work on Indo-European linguistics. The position of Sanskrit within the Indo-European family as defined since the work of Brugmann and Delbruck has changed little within this field, even though great strides have been made in many areas of historical and comparative linguistics.
In the initial period since William Jones's declaration, it is evident that there was no clear understanding of a distinction between what later came to be called Indo-Aryan languages from the class of Dravidian languages. Among the traditional Sanskrit Pandits, there was a general belief that all languages other than Sanskrit were somehow derived from Sanskrit which was described by Dandin as a Divine language (daivi vak). This was evident from the usage of descriptive terms such as tatsama "words identical with those of Sanskrit," tadbhava "words (transparently) derived from Sanskrit," and desya "regional" words whose derivation from Sanskrit was not transparent. This terminology was widely used not only to describe Prakrit and Apabhram sa languages which historically belong to the Indo-Aryan family, but also to describe the languages that belong to the Dravidian family. I have described elsewhere that in the traditional Mima m sa belief, even the so called Mleccha "foreign" languages were also ultimately derived from Vedic Sanskrit. Early western scholarship after William Jones also went through a phase that did not know distinct origins for Indo-Aryan and Dravidian languages.
19th Century Views: Gauda and Dravida Linguistic Areas
There is a traditional classification of Brahmins of India into two groups. The northern group consisting of five sub-groups is given the title Gauda, while the southern group consisting of five sub-groups is given the title Dravida; for a detailed analysis, see Deshpande (2002). The five sub-groups coming under the term Gauda are: Sarasvata, Kanyakubja, Gauda, Maithila and Utkala. The five sub-groups coming under the term Dravida are: Gurjara, Maharastra, Tailanga, Karnataka and Dravida. The early state of linguistic understanding in the post-Jones period is seen in the discussions of these groups and sub-groups by Colebrooke. In his 1801 article "On the Sanscrit and Pracrit Languages," Henry Thomas Colebrooke brings up this classification in the context of describing the linguistic geography of India. Without a hint that this is a classification of brahmins, albeit in different regions of India, Colebrooke (1801: 226) generalizes this into a classification of the "Hindu nations":
"The five Hindu nations, whose peculiar dialects have been thus briefly noticed, occupy the northern and eastern portions of India; they are denominated the five Gaurs. The rest, called the five Dravirs, inhabit the southern and western parts of the peninsula."
Colebrooke had changed the context of the general classification away from a classification of brahmins to a classification of "Hindu nations" based upon their affiliation with language and regionality. Here too, he detected certain contradictions. He had doubts about the appropriateness of the classification of the Gurjaras among the Dravidas, and the classification of Utkalas among the Gaudas. He says (1801: 229):
"Considering situation of their country, and analogy of language and writing, I cannot hesitate in thinking that Gurjaras should be considered as the fifth northern nation of India, and the Uriyas should be ranked among the tribes the Dacshin."
What is difficult for us to understand is the exact basis upon which Colebrooke is making these judgments, because, as Trautmann (1997: 147) points out: "He ... derives all major Indian languages from Sanskrit, excepting only those known to have come from the invasions of foreigners."
The fact that Colebrooke moved this classification into the area of linguistic geography set the course for the future Indological use of classification in the works of successors, where a clearer recognition of major differences between Indo-Aryan and Dravidian language families becomes evident. As Trautmann's research shows, The second discovery, that of Dravidian language family, is usually credited to a missionary, Bishop Robert Caldwell, whose classic work, A Comparative Grammar of the Dravidian or South Indian Languages, was published in 1856 .... This, however, is mistaken. Fully forty years previous to Caldwell's book, in the very year in which Bopp was laying the foundations of the comparative study of Indo-European languages in Europe (1816), Francis Whyte Ellis, Collector of Madras, published an elegant proof of the Dravidian language family. He showed that Tamil, Telugu, and Kannada shared a common stock of roots and forms of grammar that were different from those of Sanskrit, a finding he extended also to Malayalam, Tulu, Kodagu, and 'Rajmahali' or Malto, a tribal language of the Gangetic basin, far to the north" (Trautmann 2005: xxiii). The work of Ellis, and more widely the work of Bishop Caldwell, influenced the evolving understanding of the different language families in India, and a new understanding of the terms Gauda and Dravida as language groups in India began to emerge. In Caldwell's work (1856, 1974: 56-57), we see a contrast set up between "Gaurian Idioms" and "Dravidian Idioms." The languages included under the heading Gaurian are Hindi, Bengali, Marathi, Gujarati, and Sindhi. Under the heading Dravidian are listed Tamil, Canarese, Tulu, Malayalam, Telugu, Toda, Kota, Gond, Ku, Rajmahal, Brahui, and Oraon. The departure from the original context in the direction of a language-classification, since the writing of Colebrooke, is so complete in Caldwell that there is no oddity felt in listing Marathi and Gujarati among the Gaurian group. The terms are undergoing a process of redefinition and recontextualization. We see the same process in Rudolf Hoernle's 1880 work A Comparative Grammar of the Gaudian Languages. Hoernle (Introduction, p. i) explains his choice of the term Gaudian: "I have adopted the term Gaudian to designate collectively all North-Indian vernaculars of Sanskrit affinity, for want of a better word; not as being the least objectionable, but as being the most convenient one." Referring to the specific languages, Hoernle (ibid) says: "Seven languages of the Sanskrit stock are usually enumerated as spoken in North India, viz. Sindhi, Gujarati, Punjabi, Hindi, Bangali, Oriya, Marathi." The languages of Maharashtra and Gujarat, coming under the old panca dravidas, are now listed as "languages of the Sanskrit stock ... spoken in North India." There is a newly emerging notion of a linguistic divide between Indo-Aryan and Dravidian, and this divide is expressed by using the old terminology of Gauda vs. Dravida.
Early Indian Responses to Indo-European Linguistics
By the second half of the nineteenth century, there were several theories and counter-theories circulating in India; for details see Deshpande 2005. The newly emerging class of western educated Indians was absorbing the theories of Indo-European philology and Caldwell's notion of the separate Dravidian family, and was reacting to these new ideas in various ways. Early nineteenth century scholarly figures like Sir R. G. Bhandarkar, who was trained in Bombay and Pune under European professors, completely agreed with the conclusions of Indo-European philology of the time, and began to offer explanations of the emergence of various languages in terms of contact between Aryan and non-Aryan populations in India. Bhandarkar's contribution toward an understanding of the Aryan origins may be seen most profoundly in his Wilson Philological Lectures on Sanskrit & the Derived Languages delivered in 1877 in Bombay, and published in form of a book in 1914. Referring to the progress of Indo-European philology especially in Germany, Bhandarkar says: "The progress made within about fifty years is marvelous, and affords a striking instance of the intellectual activity of the Europeans. In the cultivation of Philology and the elaboration of this new science the Germans, of all other nations, have been most prominent, and have done by far the greater portion of the work," (Bhandarkar 1914:5). Bhandarkar's explanations built on the hypothesis of various migrations, contacts, and adaptations are seen in the following discourse: "Though [the speakers of Pali] heard conjunct consonants and the diphthongs ai and au pronounced by the speakers of Sanskrit, as correctly as the other letters which they did not corrupt, their organs were not fitted to utter them. These peculiarities may have been natural or acquired. If natural, the people who first corrupted Sanskrit into Pali must have belonged to an alien race which came into close contact with the Aryas and learnt their language .... And there is another instance in History of an alien race having treated the sounds of the language of a civilized community in just the same way. The Barbarians who overran Italy and developed the Italian from the Latin, showed the same inability to pronounce the Latin conjuncts, and assimilated them as our Pali ancestors did," (Bhandarkar 1914: 47). The process of the emergence of the Prakrits is accounted for by Bhandarkar by referring to the migration of the Aryas from "the land of the five rivers" to "the country known afterwards as Brahmavarta and Kurukshetra." This is the country about Thanesvar, where "they formed a consolidated community in which an aboriginal or alien race was incorporated and the language represented by the Pali was the language of that race," (Bhandarkar 1914: 88). The idea of the Aryas, the speakers of Sanskrit, coming into contact with non-Aryas, and such a contact leading to a degenerative transformation of Sanskrit into Prakrits is an idea not inherently alien to the Sanskritic tradition. But the same tradition does not admit any notion of history for this divine and eternal language, and here Bhandarkar's efforts to find historical origins of and developments in Sanskrit did not go well with his contemporaries. These were departures from the Sanskritic tradition. If the speakers of Sanskrit were Aryas, and if Sanskrit itself resulted from a process of transformation from its Indo-European precursors and underwent later transformations of its own, were the Aryas themselves subject to transformative processes? If the transformations of Sanskrit into Pali were caused by the alien speakers trying to learn the Aryan language, what was it that caused the transformations which resulted into the very existence of Sanskrit itself, and what caused transformations within the very history of Sanskrit? Such questions indeed raise unpalatable issues, and Bhandarkar's own wording suggests that he, as a Brahmin, was himself caught in the middle. Bhandarkar's wording would suggest a belief that the Aryas were not in contact with non-Aryans in "the land of the five rivers." Even admitting that the Indus Civilization was not excavated by this time, one still finds this belief difficult to accept, particularly in view of the fact that there are north western Prakrits in Asokan inscriptions, a fact which was known to Bhandarkar by this time. Bhandarkar finds that the northwestern Prakrits like Paisaci, appear to be truly Aryan. Perhaps then this was the language of an Aryan tribe that had remained longer in the original seat of the race, and was connected with the ancestors of the Teutons, so as to develop a phonetic peculiarity resembling theirs, and emigrated to India at a very late period and settled on the borders. Or it might be that the tribe came to India along with the others, but living in the mountainous countries on the border in a sort of rude independence, it developed this peculiarity of pronunciation .... Since under this supposition they could not have come in very close contact with their more civilized brethren of the plains, their language did not undergo some of those phonetic modifications which Sanskrit underwent in the mouth of the aboriginal races," (Bhandarkar 1914: 94). So the speakers of northwestern Prakrits were truly Aryan, uncontaminated by contact with the non-Aryans, but not as civilized as the speakers of Sanskrit who did come in contact with the "aboriginal races"? Bhandarkar seems to believe that the Aryans were not in contact with aboriginal races until they moved from "the land of five rivers" into the interior of India. Thus the linguistic deviations from Sanskrit in the direction of Pali and other Prakrits are caused by the contact of the Aryas with the non-Aryas, while the linguistic deviations from Sanskrit in the direction of the northwestern Prakrits and other Indo-European languages are to be explained by "isolation" at best. This creates an interesting tripartite division: the ethnically pure but less civilized Aryas of the northwest, the pure civilized Aryas of the "land of the five rivers," and the uncivilized non-Aryas of the rest of the Indian subcontinent.
Bhandarkar's student at Deccan College, B.G. Tilak (1856-1920) disagreed with Bhandarkar's moderate political views and reformist social views, but basically accepted the conclusions of Indo-European philology, with a twist of his own. His first publication on the subject of ancient history was the book Orion or Researches into the Antiquity of the Vedas, published in 1893. His second publication was The Arctic Home in the Vedas, published in 1903. His third and the last publication on the subject was the book Vedic Chronology and Vedanga Jyotisha [Containing also Chaldean and Indian Vedas and other miscellaneous essays], written in 1913 in the Mandalay jail in Burma and published posthumously in 1925. Trying to find history in the anadi "beginningless" Vedas was a departure from the Brahmanical tradition, but such a departure, facilitated even though it was by the introduction of Western education, was the path chosen by Tilak.
Tilak was in basic agreement with Bhandarkar on the conclusions of Indo-European linguistic theory, and had particular admiration for the work of Max Muller. He viewed Max Muller not only as a great scholar of the Veda, but as someone who pointed out to the west the importance of the spiritual contribution of India (cf. Tilak on Max Muller, Lokmanya Tilakamce Kesaritil Samkina Lekh, vol. 4, 1922, 124ff. Also see Max Muller's obituary written by Tilak in Kesari, November 6, 1900, ibid, pp. 573ff.). In his obituary of Max Muller in Kesari (ibid, pp. 575-6), Tilak approvingly points out Muller's contributions to Indo-European philology and comparative mythology:
"The efforts of the western philologists were given a new turn by their recognition of the ancient character of Sanskrit and their study of Panini's general rules and their exceptions regarding the derivation of various types of words from stems and affixes. In the last thirty/ forty years, German scholars have produced an entirely new sastra 'science' of comparative philology. One must not forget that Panini's work is the foundation of this new sastra. ... The second consequence of the western study of languages was in the area of the study of religion. Before the birth of western philologists like Burnouf, Bopp, and Jones, the people in Europe were familiar with only one religious tradition that was born in the region of Palestine. ... This orientation of the western thinking on religion changed with their study of the Veda, Avesta, and the Buddhist Tripitaka, and with a great transformation of ways of thinking, the science (sastra) of comparative religion was born. . .. The notion that the civilized people in Europe and Asia must have originally belonged to the same human race and were speakers of the same original language gradually developed and was successively strengthened. It is necessary to understand this history to fully comprehend the true significance of Professor Max Muller's efforts, scholarship, and writings."
However, while Tilak admired Max Muller, he believed that the purely philological approach of scholars like Max Muller was not sufficient to yield convincing results and that the astronomical method, though generally condemned by the philologists as being indefinite and unreliable, could be refined and used to make more exact predictions. In his own works based not so much on philology, but on astronomical evidence, Tilak argued that the ancestors of the Vedic Aryans, and possibly texts of the Vedas themselves came from the Arctic Home dating to a very ancient period of 8000 to 10000 Be. If one were to accept Tilak's dating of the Vedas and the ancestors of the Vedic Aryans, then the non-Sanskrit branches of Indo-European language family would all seem to be far younger than the Indo-Aryan branch. While this conclusion would make the Indian Aryans the senior, rather than junior, brothers of the colonial Western Aryans, and would boost the nationalist pride, Tilak did not systematically attempt to explain how he would derive all other branches of Indo-European from Sanskrit. Without explicitly claiming so, Tilak makes Sanskrit the oldest of all Indo-European languages, and hence a possible source of them all. As I have discussed elsewhere in detail (cf. "Arctic Home in the Vedas: Religion, Politics, and the Colonial Context," forthcoming), in spite of Tilak's political background, by the standards of the late 19th and early 20th century scholarship in the west, Tilak's geological and astronomical investigations had better scholarly foundations, than the openly Hindutva-motivated histories constructed by Savarkar, Golwalkar, and others.
However, in the nineteenth century Maharashtra, there was a continuing Brahmanical tradition that believed in the primordiality of Sanskrit and India as its homeland and holy-land, combined with a belief that all other languages were derived from Sanskrit through a process of degeneration (apabhramtsa-ization). This goes back to Manu's Lawbook that all populations of the world ultimately originate from the four primordial vannas created by God in India, through mixing of the original varnas and through migrations of various groups away from the Aryavartic center combined with loss of dharmic behavior (Manusmrti, chapter 10). Manu (10.43-44) lists Paundraka, Caudra, Dravida, Kamboja, Yavana, Saka, Parada, Pahlava, Cina, Kirata, Darada, and Khasa as original Ksatriyas who slowly became Sudras through loss of dharmic behavior and through non-contact with Brahmins (kriyalopat ... brahmanadarsanena ca, Manu 10.43). This Brahmanical view, combined with the appropriation of modern knowledge of the Indo-European family and the emerging sense of Indian nationalism, appears in the works of authors like Narayan Pavgee, a contemporary of Tilak. While Tilak argued for the Arctic Home of the Vedic Aryans, Pavgee argued for the Aryavartic Home of the Aryans, though he was willing to accept a theory that a branch of the Aryavartic Aryans migrated to the Arctic region and eventually returned to India. With the Aryavartic home of the Aryans, Pavgee was arguing against the notion that the Aryans came to India from outside. However, he was proposing that the Aryans spread across the world from the Aryavartic home, and that Sanskrit was the mother of all languages of the world.
The second volume of his Bharatiya Samrajya (1893) is titled: Aryalok va tyance Buddhi-vaibhav "The Aryan People and the Wealth of their Intelligence." The book begins with a section dealing with the original home of the Aryas (aryance mula-nivasa-sthana). Pavgee asserts at the very beginning that the Vedas are the oldest literature of mankind, and that the Vedas support the notion that northern India is the original home of not just the Aryans, but of the entire mankind (Pavgee 1893: 2). The Aryan family of languages originated in India, Pavgee (1893: 5) asserts, and it expanded westward from India through the regions of Iran, Greece, Italy, Spain, England, Germany, and Russia. Pavgee mentions the view of western philologists that the Aryans came into India from outside, but he does not support this view. In the ninth volume of his Bharatiya Samrajya (1900), titled Bharatakhandatil Nanavidha Bhasa "Various Languages in India," Pavgee has a section on the original land of the Arya language, i.e. Sanskrit. Pavgee asserts unequivocally that Sanskrit originated in the region of Aryavarta within India and is the mother of all Aryan languages. All languages such as Marathi, Hindi, Bengali, Gujarati, Iranian, Greek, Latin, German, English, and Polish were born from Sanskrit. They are daughters of Sanskrit (Pavgee 1900: 14). Various originally Arya groups left their religion and castes and left the original Arya homeland, and these eventually became the various branches of the Aryan language family. In support of this conclusion, Pavgee offers, not linguistic arguments, but a large number of passages from Smrtis and Puranas (1900: 16ff), a practice continued in nationalist history writing till today, see papers in Deo and Kamath (1993).
From Linguistic Reconstruction to Racist Theories in the west
Returning to the history of western scholarship during the nineteenth century and the early twentieth century, one needs to keep in mind one major factor, namely that the Indus Valley Civilization was discovered only in 1920s, and hence all previous scholarship of the Indo-European philology developed in the absence of any knowledge of the Indus Valley Civilization. While the discovery of the Indo-European language family was to a large extent a result of linguistic scholarship, the scholars who were engaged in this scholarship often used discourses that merged the distinction between language and ethnicity or race, (Day, 1994, Leach, 1990, and Trautmann 1997 and 1999). Beginning with William Jones, we find the use of terms like "nations" to refer to groups speaking particular languages, and words like "Semitic" and "Hamitic" combined both language and ethnicity, (Trautmann 1997 and 1998). With this mixture of terminology, the notion of a linguistic ancestor for the languages of the Indo-European family soon developed into finding an ancestral group of people, a nation or a race, that over time splintered into multiple sub-groups and these migrated from an original homeland to their eventual destinations where we find the descendant languages today. Today one can say with confidence that there is no necessary connection between language and ethnicity. The African blacks who speak French are linguistically as much Indo-European as white French speakers in France, just as the English-speaking peoples across the world speak an Indo-European language, though they belong to different races and ethnicities. Thus, linguistic identities are not necessarily related to racial or ethnic identities; the language of race was pervasively used by scholars in the nineteenth and twentieth centuries. In early writings, Max Muller used the word Aryan in racial meaning, even though he later recanted: "There are Aryan and Semitic languages, [but] it is against all logic to speak, without an expressed or implied qualification, of an Aryan race, Aryan blood, or Aryan skulls," Max Muller (1880, 4: 223). Edmund Leach (1990: 235) points out that by 1878 Max Muller was writing without equivocation about Aryans as if they were a racial group moving outward from some central Asian homeland by a process of conquest. In the 1890's, the linguist Sir George Grierson was explaining all Indo-Aryan dialect distributions in northern India as due to past military conquest, a view that was repeated without criticism in the 1969 printing of the Encyclopaedia Britannica, Leach (1990: 235). We see the same pattern in the works of leading European scholars like Hermann Oldenberg. Consider the following narrative from Oldenberg (1896, 1973: 39):
"In the ancient times for instance, the Aryans of the Northwestern part of further India had not yet entered deep into the borderland by the use of force and were still the brothers and almost neighbours of the Zarathustrian Aryas of Iran, or rather of the Aryans who were opponents of Zarathustra. The situation changed in later times. Hinduism spread all along the peninsula with the Aryan character ever weakening, with the blood of the natives mingling in their blood stream in a never-ceasing continuity and with an infinite series of shades of complexion, ranging from the fair to the dark, observable in the populace. It will not do to mix up the old times with the modern times."
Such passages can be multiplied a hundred times indicating how wide the use of the racial terminology was. The racial terminology took different forms, e.g. the Aryans versus Semites in Germany, while in the context of colonial domains like India, the racial terminology made the relationships between the white Aryan British and the black Aryan Bengali ever more complicated -- sometimes positively, though most of the time negatively. Does this mean that one must not connect migrations of peoples to linguistic changes, for fear of involving ourselves in racialist thinking? I do not think so. Theorizing from where we stand today, we need to take into account all forms of human interaction that involves language to evaluate the phenomenon of linguistic relationships, genetic or otherwise, as well as phenomena like various forms of intensive and extensive bilingualism and diglossia. The social context of language in its widest sense needs to be taken into account to explain facts language, without insisting that a linguistic identity must be co-terminus with racial, biological, religious, or political identities. The different identities do indeed intersect, but do not logically entail each other. These various forms of identities do affect language use, and may be taken into account with due caution.
Historical Linguistics after the Discovery of the Indus Valley Civilization
With the discovery of the Indus Civilization in 1920s began a new effort to make sense of the Indo-European migrations and the Vedic texts. Early on, the destruction of the Indus cities was connected with the Vedic Aryans led by their God Indra destroying the forts of Dasas and Dasyus. If this version of history is accepted, Indus was pre-Vedic and non-Vedic and was destroyed by Vedic Aryans. This most likely makes Indus a non-Indo-European linguistic region, and hence for decades various scholars, like Emeneau and Parpola, argued for a Dravidian identity for Indus. This was also supported by the consideration that Brahui in Baluchistan is a Dravidian language, and further by the claim of David McAlpin that Dravidian and Elamite of ancient Iran may belong to the same language family. Even though gradually the opinion shifted to the position that the Vedic Aryans entered the Indus area after the Indus cities had already declined and abandoned, the predominant linguistic opinion still held that the Indus Valley was a Dravidian speaking region. Decipherments of the Indus Seals by S.R. Rao and Madhusudan Misra, among others have claimed a form of proto-Sanskrit as the language of Indus, and the Vedic character of Indus civilization, though there is not a wide acceptance of these proposals, cf. for a review of recent proposals, see: Ratnagar (1996), Witzel (2002) and Mahadevan (2002). More recently scholars Michael Witzel have suggested that the area was most likely a multi-lingual multi-ethnic region that may have included Dravidian, proto- or para-Munda, as well other languages (related to isolates like Burushaski), including forms of Indo-Iranian on the periphery. This suggestion seems to be most realistic to me, as it allows an open-ended examination of all available linguistic data.
As for the latest twist on the decipherment of the Indus seals, Farmer-Sproat-Witzel (2004) proposes that the Indus seal symbols do not codify a language: "Archaeologists have long claimed the Indus Valley as one of the four literate centers of the early ancient world, complete with long texts written on perishable materials. We demonstrate the impossibility of the lost-manuscript thesis and show that Indus symbols were not even evolving in linguistic directions after at least 600 years of use. Suggestions as to how Indus symbols were used are noted in nonlinguistic symbol systems in the Near East that served key religious, political, and special functions without encoding speech or serving as formal memory aids. Evidence is reviewed that the Harappans' lack of a true script may have been tied to the role played by their symbols in controlling large multilinguistic populations; parallels are drawn to the later resistance of Brahmin elites to the literate encoding of Vedic sources and to similar phenomena in esoteric traditions outside South Asia (Abstract, p. 19)." While a nonlinguistic interpretation of Indus symbols is a distinct possibility that needs to be taken into account, the initial reactions to the above proposal from scholars do not seem to indicate a wide acceptance. In any case, whether because the available attempts to decipher the linguistic value of Indus symbols have all failed to reach a convincing result, or because they are intrinsically nonlinguistic symbols, so far the symbols of Indus do not directly contribute to our linguistic understanding of history, which must be teased out only from an intensive mining of linguistic sources such as the Vedic and Iranian texts, and this brings us back to the work of scholars like Emeneau, Burrow, Kuiper, Mayrhofer, Southworth, Witzel, Parpola, and Lubotsky, among others. The substratum languages in the area covered by the Indus Valley Civilization probably include Dravidian, (Para-) Munda, as well as ancestors of linguistic isolates like Burushaski, plus some traces of the languages the Bactria Margiana Archaeological Complex, as well as early forms of Iranian Indo-Aryan in the border areas. Southworth (2005: 328-9) says: "The Indus language (group) is probably the oldest detectable linguistic stage, both in Panjab and Sindh. An unidentified western Austro-Asiatic language designated as Para-Munda probably functioned as a lingua franca of this area. Some of these language groups may have been present as early as the seventh or eighth millennium Be, when archaeologists record the beginnings of agriculture in areas bordering the Indus Valley. Indo-Aryan appears in the form Vedic Sanskrit, dateable to the early second millennium Be, probably preceded by outer Indo-Aryan. The arrival of Dravidian-speaking people in the area is difficult to date and is the subject of some controversy. Evidence of Dravidian borrowings in the Rigveda dates only from about 1200 Be, according to some sources, yet it is nevertheless possible that Dravidian languages were spoken in Sindh and Saurashtra (and perhaps even in Panjab) considerably earlier. Ultimately, only Indo-Aryan survives in these areas, though Sindh shows traces of outer Indo-Aryan." Also noteworthy are Witzel's two 1999 articles "Substratum Languages in Old Indo-Aryan" and "Early Sources for South Asian Substrate Languages," as well as the forthcoming article "The Languages of Harappa."
New Linguistic Approach in Mid-20th Century: "India as a Linguistic Area"
The most significant development of the middle of the twentieth century linguistic study of India was the idea of "India as a Linguistic Area" seen in the early works of Murray B. Emeneau and F.B.J. Kuiper, and eventually consolidated in the work of Nicholas Masica. Until the emergence of the notion of Linguistic Area, the only dominant model was the genetic model used in Indo-European and other forms of historical linguistics. The genetic model explains the common elements between genetically related languages as being a reflection of common ancestry. Where the genetically related languages differ, those differences may have developed either internally in those branches, or they may be due to contact with other languages of different regions etc. But what about a demonstrably large number of common features shared by languages that evidently do not belong to the same language family? Where do these features and their commonality originate? The new direction proposed by areal linguistics was that diffusion of linguistic features across genetically unrelated languages can occur through intense and prolonged contact. Given sufficient intensity of contact and depth of time, genetically unrelated languages of a given geographical region begin to exchange phonetic, syntactic, and other features with each other, and over time develop a significant set of such shared features. The research work of Emeneau, Kuiper, Masica and others has demonstrated how this process occurred in the Indian subcontinent so that languages belonging to different families like Indo-Aryan, Dravidian, and Austro-Asiatic came to share a large number of common linguistic features through diffusion, rather than common ancestry. Thus, genetic relationships between languages as well as processes of diffusion among languages of a given linguistic area are all working simultaneously to produce the resulting state of a language or languages, and these different strands in the make-up of languages need to be analytically separated in order to get a proper sense of linguistic history. Franklin Southworth (1974) had previously used the term "linguistic stratigraphy" for this sort of linguistic analysis, and his recently published book (2005) calls it "linguistic archaeology," where he discusses how one can begin to peel off or dig into various layers of features acquired by a language or languages.
By the beginning of the second half of the twentieth century, especially with the researches of scholars like Emeneau, Burrow, and Kuiper, who added serious consideration of Dravidian and Austro-Asiatic languages and their co-existence in the Indian subcontinent with languages of the Indo-Aryan branch of the Indo-European family, many issues were newly brought into a sharper focus. In his 1975 article "Substratum Influence on (Rg-Vedic) Sanskrit," Hans Hock conveniently summarizes the conclusions of this research (p. 76-78):
As is well known, the South Asian subcontinent today constitutes one of the paradigm cases of linguistic convergence: Three major linguistic families of distinct origins (Dravidian, Indo-Aryan, and Munda), as well at least one language isolate (Burushaski) have merged into a sprachbund which further extends into the fringes of the neighboring Iranian and Tibeto-Burman language families. This sprachbund (= linguistic area) is characterized by an impressive array of shared features which are found throughout most or all of the individual sub-dialects of the area and which include a contrast between dental and retroflex segments [for example, t, d, dh, and n, versus t, th, d, dh, and n], a prevailing SOV (= Subject-Object-Verb) sentence structure, and the phenomenon of a 'cumulative' extended sentence structure characterized by the tendency to limit the occurrence of finite verbs to the rightmost clause; while all preceding (non-relative) clauses have the verb appear as a non-finite 'absolutive.' Considering the massiveness of this present-day convergence, it is only natural to assume that it must be a result of centuries, if not of millennia, of quite intimate bilingual contact. And it is equally natural to speculate on the period when the developments leading to his convergence must have begun. In terms of the history of Indo-Aryan, did they begin at the time of the earliest extant texts, the Vedic Samhitas, especially the oldest among them, the Rg-Veda? Or did they take their course only in a later, post-Vedic period?
It is natural to speculate on the direction of the convergence in its early stages. Was it a mutual convergence, as it is usually encountered in present-day India? Or was it mainly a unidirectional convergence comparable, say, to that between Czech and German?
... As early as 1833, Pott considered the dental! retroflex contrast of Sanskrit at least partly due to the influence of the 'autochthonous' languages (78), later specifying these as Dravidian (1836: 19,453). In 1856, Caldwell made similar claims (38). As time progressed, the number of linguists subscribing to this view and adding other features of (Rg) Vedic or post-Vedic Sanskrit considered the result of Dravidian influence steadily increased.
In 1921, Przyluski added another possible source for some of these phenomena, namely Munda. Levi (1923) added further weight to the assumption that Munda at one time must have been very influential in the South Asian area. As a consequence, it was for some time considered possible that Munda, as well Dravidian, may have contributed to the specifically "Indian" features of Sanskrit (such as the dental/retroflex contrast); cf. e.g. Bloch (1934: 53-4). As a matter of fact, Pizagalli (1929: 165-7) believed that only the Munda languages could be the source for Indo-Aryan retroflexion. In the more recent literature, however, the influence of Munda is considered to have been minimal at best, Dravidian being cited as the probable source language; cf. e.g. Emeneau (1956, 1962a, 197-4), Kuiper (1967a).
There is thus a long tradition of explicitly or implicitly answering the question concerning the beginning of the South Asian convergence to have been pre-Vedic (since the dental/retroflex contrast occurs already in earliest, Rg-Vedic texts). Moreover, at least for the early, Vedic period, there is a tacit agreement that the convergence was unidirectional, from Dravidian (or Munda) to Sanskrit. At any rate, except perhaps for the area of vocabulary, no attempt seems to have been made to identify any instances of early convergence in the opposite direction.
This view, however, has not remained unchallenged. There has been a small, but constant flow of linguistic opinion according to which, especially in view of our present ignorance about the full linguistic panorama of early India, it is hazardous to attribute convergence only to a Dravidian (and/or Munda) substratum: There have been other languages which could have been the source for at least some of the phenomena, cf. e.g. Bloch (1929: 731-2, 1934: 322- 4), Mayrhofer (1953: 233-4), Thieme (1955: 436-48). In light of the fact that even today we the isolated Burushaski in the South Asian sprachbund, this cautionary attitude seems to be justified.
Some linguists even have completely rejected the hypotheses advocating early substratum influence, and have claimed that the developments in question can be accounted for as native, without any need for assuming an outside, non-Indo-Aryan substratum; cf. e.g. Buhler (1864), Bloch (1925: 16, 29).
Finally, other linguists, accepting the view that (some of) the phenomena in question can be accounted for as regular, native developments, proposed that at most, the substratum language(s) accelerated or 'helped' these developments; cf. e.g. Konow (1903: 455, 1906: 279), Bloch (1929: 723-4, 1934: 53-4) (thus changing his earlier, 1925 view), Burrow (1955: 95-6).