|SITE MAP Home Indian Lexicon Corpus of Inscriptions Artefacts Decipherment|
Sarasvati River: Key dates and a Linguistic Area
Back to Table of Contents
The civilization sites in Punjab, Rajasthan, Cholistan, Kutch and Saura_s.t.ra can be explained by the Sarasvati river as a navigable channel right from Ropar to Lothal, ca. 2500 B.C.
Maps and a 180 page document with bibliographical notes may be retrieved from the Sarasvati River website. The evidence summarised in this note is from three sources: (1) Earth Sciences; (2) Archaeology; (3) Linguistics. The first two sources offer more precise dates for the migration of Sutlej away from Sarasvati and for the piracy of Yamuna away from Sarasvati. The third source reconfirms the hypothesis that the NW India was a linguistic area ca. 2500 B.C. This will be validated in a separate website which will present the corpus of inscriptions and clusters of lexemes from the linguistic area.
(1) Earth Sciences
Aloys Arthur Michel, The Indus Rivers: A study of the effects of partition, 1967, Yale University Press, New Haven.
"(About 600,000 years ago) the Main Himalayas at first replaced the Tibetan Marginal Mountains as the recipients of the full force of the monsoon, and the south-flowing streams were able to maintain their courses by down-cutting as the Siwaliks began to arise. But when the Lesser Himalayas (the Mahabharat Range in Nepal) were elevated, less than 200,000 years ago, some of these rivers were ponded back behind them, forming huge lakes (Note by Kalyan: Hence,the name Sarasvati!!) Eventually, the water rose high enough to find outlets to the south (added, of course, by the work of the truncated rivers cutting back through the Lesser Himalayas), and the lakes were drained, leaving the level basins found today around Kathmandu and Pokhara in Nepal (Note by Kalyan: note, the placid lake in PaontaSaheb where Yamuna pirated Sarasvati) South-flowing streams cutting through the Siwaliks had no difficulty in maintaining their courses because that frontal range was never high enough to shut off precipitation in their source areas. Yet the uplift of the Himalayas, including the Siwaliks, is apparently still continuing, offset by rapid erosion of course, and earthquakes are by no means uncommon as a result... (p.25)... there would seem to be little doubt that the present, almost imperceptible watershed between the Ganges and Indus drainage is very recent in origin. Here the key seems to lie in the shifting or migration of stream beds across the alluvium of the plains, and key role to have been played by the Jumna and a former stream (possibly the legendary Sarasvati) the course of which is now marked by the bed known as the Ghaggar in the Indian Punjab and Rajasthan, and as the Hakra in Pakistan Bahawalpur, that parallels the Sutlej towards the Indus. The enormous amounts of detritus brought down by the Punjab rivers and the present affluents of the Ganges are more than sufficient to explain stream blockage and shifting without invoking tectonic forces, and capture of one stream by another is well-attested. The Beas, for example, was captured by Sutlej at the end of the eighteenth century. Its old course near Harike to the Chenab above Panjnad is well marked in the landscape of the southern Pakistan Punjab, with the town of Kasur and a series of villages still lining its 'banks'. The Ghaggar, which is used in part by modern canals and which has begun to flow again as water tables have risen, may very well represent the former course of such a truncated river. Spate suggests that it could have been fed either by the Sutlej, itself occupying a different channel, or by the Jumna. If it was the Jumna, then the Jumna clearly has been captured by the Ganges... in the broadest sense the Indus Plains may be regarded as one vast and fairly homogeneous aquifer, a sort of vast sponge, capable of absorbing runoff from the foothills as well as rainfall and seepage from the rivers and canals that cross them, and of transmitting this subterranean flow downslope to the Arabian Sea. (Kalyan's notes: Note the legends regarding the disappearance of Sarasvati underground-antah salila_ sarasvati_!) The water table or top level of this vast reservoir varies with distance from the foothills and from the rivers and canal, as well as with local alterations in the nature of the matrix, and it varies from season to season and year to year. Recent investigations in the Pakistan Punjab have been sufficiently detailed to allow preparation of contour maps showing depth to water table, and comparisons with older data from wells indicate its general rise since irrigation was introduced (cf. Greenman et al, Maps 11, 12, 16-20). Variations in the salt content of the groundwater have also been charter over much of Punjab... The groundwater reservoir apparently represents at least ten times the annual runoff of the Indus Rivers, and in many areas offers an additional source of irrigation water when tapped by tubewells. The control of the water-table level by means of pumping from wells or by drains is also essential to the success of the surface-water irrigation, for in many areas the salt-carrying groundwater has risen perilously close to the surface (pp. 27-28)(Kalyan's note: see the situation in Pehoa-Prthu_daka-in Sarasvati Ghat and Brahma yoni near Vasis.t.a_s'ramam where the river becomes pra_ci_va_hini_; sarasvati is so named in the revenue maps of Haryana and also in Bharat Bhu_racana_, Survey of India maps.)
Sarasvati-Sindhu Research Centre has established in a technical monograph (by Dr. K. R. Srinivasan, ex-Director, Central Ground Water Board) that the central Sarasvati River basin in Rajasthan alone can support one million tube wells on a sustainable basis with recharge principally from the Rajasthan canal. Itihas Darpan (Hindi magazine) is bringing out a special issue on Sarasvati river.
Srinivasan, K.R., Paleogeography, Framework of Sedimentation and Groundwater Potential of Rajasthan, India-Central part of Erstwhile Sarasvati Basin, Group Discussion, Geological Society of India: Drainage Evolution of North-western India with particular reference to the Lost Sarasvati, December 1997, Baroda.
Sir A. Burnes, Memoir on the Eastern Nara Branch of the River Indus, giving an account of the alternations produced on it by an earthquake, also a theory of the formation of the Runn, Trans. RAS, III, 1834, pp. 550-88; Major F. Mackeson, Report on the Route from Seersa to Bahawulpore, JAS Beng., XLII, Pt.1, 1844, No. 145 to 153 recomming the conversion of Sarasvati river bed into a great road from the sea-coast in Sind to Delhi via Bahawalpur, Marot., Anupgarh, Suratgarh, D.a_bli, Ka_libagga_n., Bhat.ner (modern Hanumangarh), T.ibi and Sirsa.
Alex Rogers, 1869, A few remarks on the geology of the country surrounding the Gulf of Cambay, in Western India, Proceedings of the Geological Society, in: Quarterly Journal of Geological Society of London, Vol. 26, 1870, pp. 118-123. [Explains the remarkable presence of alluvium in the Gulf of Khambat thanks to the mighty Sarasvati river bringing down enormous amounts of detritus.]
Puri, V.M. and S.P. Verma, Glaciological and Geological Evolution of Vedi Sarasvati in the Himalayas, Paper presented in Delhi on 5 October 1997, Itihasa Sankalana Samiti (repr. in: Itihas Darpan, Special Issue on Sarasvati River).
Wilhelmy, Herbet, 1969, Das Urstromtal am Ostrand der Indusebene und der Sarasvi-Problem, Zeitschrift fur Geomorphologie, Supplementband 8: 79-93. [explaining the secular sequence of desiccation of Sarasvati River, first, the diversion of Sutlej westwards and second, the joining of Beas with Sutlej.]
Valdiya, K.S., River Piracy, Sarasvati that disappeared, Bangalore, Indian Academy of Sciences, Resonance, I, 5: 19-28, 1996. [explaining the river piracy-capture of the Yamuna by Ganga; explaining how through the Sarasvati River had flowed the combined molten glacier waters of Sutlej and Yamuna.]
Rao, S.M. and Kulkarni, A.M., 1997, Isotope Hydrology Studies on Water Resources in Western Rajasthan, Current Science, Vol. 72, No. 1, Delhi.
G.Urban M. Jansen, Forschungsprojekt DFG Mohenjo-daro, Aachen, 1981:
"The archaeological evidence pertaining to the earliest or first settlement at Kot Diji which was originally labelled 'Kot Dijian' has become a basic frame of reference for reconstructing a relative chronology of the early sites of the fourth and early third millennia B.C. in the Greater Indus Valley, drained by the Indus and its present and former tributaries including the now dry Ghaggar-Hakra rivers... The earliest layers of the settlement outside the fortified area correspond in time with the middle Kot Dijian levels and were dated 2805-2885 B.C. No C-14 date is available for the upper Mature Harapan occupation at Kot Diji but its contemporaneous materials at Mohenjo-daro are dated between 2060 and 2583 B.C. Thus, the Kot Dijian occupation turned out to be at least 500 years older than the fully urbanized stage of the Indus Civilization as represented at Mohenjo-daro... Kot Dijian cultural materials, both chronologically and culturally constitute an Early Harappan or formative stage of the Indus Civilization... processes of urbaniztion in the Greater Indus Valley had started during the fourth millennium B.C. The evidence from several sites clearly demonstrated that the first settlement of Kot Diji was an integral part of the cultural phenomen which was wide-spread throughout the Greater Indus Valley with remarkable uniformity of essential elements which later characterized the Indus Civilization in fully developed or urban stage... In India, the areas originally drained by the Ghaggar-Hakra rivers and their tributaries in northern Rajasthan, East Punjab and Haryana yielded a succession of Kot Diji related sites among which Kalibangan, Siswal, Mitathal, Banawali, Bhundan and Manda were also excavated... Dr. Allchin also continued to ignore the new discoveries especially in Bahawalpur... the geographical extent of the Early Harapan settlements as revealed by intensive field works started at the beginning of this decade, almost duplicates that of the succeeding Mature Harappan swettlements... (the civilization) certainly did originate and develop in the Greater Indus Valley." (pp. 1-19)
Mughal, M. Rafique, 1982:
"On the Pakistan side, archaeological evidence now available overwhelmingly affirms that the Hakra was a perennial river through all its course in Bahawalpur during the fourth millennium B.C. (Hakra period) and the early third millennium B.C. (Early Harappan period). About the middle of the third millennium B.C., the water supply in the northeastern portion of Hakra, roughly between Fort Abbas and Yazman (near Kudwala) was considerably diinished or cut-off. But abundant water in the lower (southwestern part) of this stream was still available, apparently through a channel from the Sutlej; this is attested by the heavy clustering of sites in that area during the late third and early second millennium B.C. (Mature and Late Harappan periods respectively). About the end of the second, or not later than the beginning of the first millennium B.C., the entire course of the Hakra seems to have dried up and a physical environment similar to that of the present day in Cholistan set in. This forced the people to abandon most of the Hakra flood plain. A few Painted Grey Ware settlements, most of them smaller than four hectares in size, are located along the upper part of the Hakra river. These were sustained by a meager water supply reaching there with seasonal regularity from the Ghaggar... the presence of Hakra Ware sites on top of old,reddish-brown sand, as observed on the south and southwest of Derawar, sould seem to indicate that the Cholistan part of the Thar desert had already advanced close to Derawar prior to the fourth millennium B.C."
A.K. Sharma, The Harappan Horse was buried under the dunes of..., Puratattva, Number 23, 1992-93, New Dehli, Indian Archaeological Society.
See notes on Formation of Indian Languages
M.B.Emeneau, Linguistic Prehistory of India [PAPS98 (1954). 282-92; Tamil Culture 5 (1956). 30-55; repr. In Collected papers: Dravidian Linguistics Ethnology and Folktales, Annamalai Nagar, Annamalai University, 1967, pp. 155-171].
"In fact, promising as it has seemed to assume Dravidian membership for the Harappa_ language, it is not the only possibility. Professor W. Norman Brown has pointed out (The United States and India and Pakistan, 131-132, Cambridge, Harvard University Press, 1953) that Northwest India, i.e. the Indus Valley and adjoining parts of India, has during most of its history had Near Eastern elements in its political and cultural make-up at least as prominently as it had true Indian elements of the Gangetic and Southern types. The passage is so important that it is quoted in full: 'More ominous yet was another consideration. Partition now would reproduce an ancient, recurring, and sinister incompatibility between Northwest and the rest of the subcontinent, which, but for a few brief periods of uneasy cohabitation, had kept them politically apart or hostile and had rendered the subcontinent defensively weak. When an intrusive people came through the passes and established itself there, it was at first spiritually closer to the relatives it had left behind than to any group already in India. Not until it had been separated from those relatives for a fairly long period and had succeeded in pushing eastward would I loosen the external ties. In period after period this seems to have been true. In the third millennium B.C. the Harappa culture in the Indus Valley was partly similar to contemporary western Asian civilizations and partly to later historic Indian culture of the Ganges Valley. In the latter part of the next millennium the earliest Aryans, living in the Punjab and composing the hymns of the Rig Veda, were apparently more like their linguistic and religious kinsmen, the Iranians, than like their eastern Indian contemporaries. In the middle of the next millennium the Persian Achaemenians for two centuries held the Northwest as satrapies. After Alexander had invaded India (327/6-325 B.C.) and Hellenism had arise, the Northwest too was Hellenized, and once more was partly Indian and partly western. And after Islam entered India, the Northwest again was associated with Persia, Bokhara, Central Asia, rather than with India, and considered itself Islamic first and Indian second. The periods during which the Punjab has been culturally assimilated to the rest of northern India are ew if any at all. Periods of political assimilation are almost as few; perhaps a part of the fourth and third centuries B.C. under the Mauryas; possibly a brief period under the Indo-Greek king menander in the second century B.C.; another brief period under the Muslim kingdom of Delhi in the last quarter of the twelfth century A.D.; a long one under the great Mughals in the sixteenth and seventeenth centuries A.D.; a century under the British, 1849-1947.'
"Though this refers to cultural and political factors, it is a warning that we must not leap to linguistic conclusions hastily. The early, but probably centuries-long condition in which Sanskrit, a close ally of languages of Iran, was restricted to the northwest (though it was not the only language there) and the rest of India was not Sanskritic in speech, may well have been mirrored earlier by a period when some other language invader from the Near East-a relative of Sumerian or of Elamitic or what not-was spoken and written in the Indus Valley-perhaps that of invaders and conquerors-while the indigenous population spoke another language-perhaps one of the Dravidian stock, or perhaps one of the Munda stock, which is now represented only by a handful of languages in the backwoods of Central India.
"On leaving this highly speculative question, we can move on to an examination of the Sanskrit records, and we find in them linguistic evidence of contacts between the Sanskrit-speaking invaders and the other linguistic groups within India...
"...the early days of Indo-European scholarship were without benefit of the spectacular archaeological discoveries that were later to be made in the Mediterranean area, Mesopotamia and the Indus Valley... This assumption (that IE languages were urbanized bearers of a high civilization) led in the long run to another block-the methodological tendency of the end of the nineteenth and the beginning of the twentieth century to attempt to find Indo-European etymologies for the greatest possible portion of the vocabularies of the Indo-European languages, even though the object could only be achieved by flights of phonological and semantic fancy... very few scholars attempted to identify borrowings from Dravidian into Sanskrit...The Sanskrit etymological dictionary of Uhlenbrck (1898-1899) and the Indo-European etymological dictionary of Walde and Pokorny (1930-1932) completely ignore the work of Gundert (1869), Kittel (1872, 1894), and Caldwell (1856,1875)... It is clear that not all of Burrow's suggested borrowings will stand the test even of his own principles..."
M.B.Emeneau, India as a Linguistic Area [Lang. 32, 1956, 3-16; LICS, 196, 642-51; repr. In Collected papers: Dravidian Linguistics Ethnology and Folktales, Annamalai Nagar, Annamalai University, 1967, pp. 171-186].
"'India' and 'Indian' will be used in what follows for the subcontinent, ignoring the political division into the Republic of India and Pakistan, and, when necessary, including Ceylong also... the northern boundary of Dravidian is and has been for a long time retreating south before the expansion of Indo-Aryan... We know in fact from the study of the non-Indo-European element in the Sanskrit lexicon that at the time of the earliest Sanskrit records, the R.gveda, when Sanskrit speakers were localized no further east than the Panjab, there were already a few Dravidian words current in Sanskrit. This involves a localization of Dravidian speech in this area no lather than three millennia ago. It also of course means much bilingualism and gradual abandonment of Dravidian speech in favor of IndoAryan over a long period and a great area-a process for which we have only the most meagre of evidence in detail. Similar relationships must have existed between Indo-Aryan and Munda and between Dravidian and Munda, but it is still almost impossible to be sure of either of these in detail... The Dravidian languages all have many Indo-Aryan items, borrowed at all periods from Sanskrit, Middle Indo-Aryan and Modern Indo-Aryan. The Munda languages likewise have much Indo-Aryan material, chiefly, so far as we know now, borrowed rom Modern Indo-Aryan, thogh this of course inlcudes items that are Sanskrit in form, since Modern Indo-Aryan borrows from Sanskrit very considerably. That Indo-Aryan has borrowed from Dravidian has also become clear. T. Burrow, The Sanskrit Language, 379-88 (1955), gives a sampling and a statement of the chronology involved. It is noteworthy that this influence was spent by the end of the pre-Christian era, a precious indication for the linguistic history of North India: Dravidian speech must have practically ceased to exist in the Ganges valley by this period... Most of the languages of India, of no matter which major family, have a set of retroflex, cerebral, or domal consonants in contrast with dentals. The retroflexes include stops and nasal certainly, also in some languages sibilants, lateral, tremulant, and even others. Indo-Aryan, Dravidian, Munda and even the far northern Burushaski, form a practically solid bloc characterized by this phonological feature... Even our earliest Sanskrit records already show phonemes of this class, which are, on the whole, unknown elsewhere in the Indo-European field, and which are certainly not Proto-Indo-European. In Sanskrit many of the occurrences of retroflexes are conditioned; others are explained historically as reflexes of certain Indo-European consonants and consonant clusters. But, in fact, in Dravidian it is a matter of the utmost certainty that retroflexes in contrast with dentals are Proto-Dravidian in origin, not the result of conditioning circumstances... it is clear already that echo-words are a pan-Indic trait and that Indo-Aryan probably received it from non-Indo-Aryan (for it is not Indo-European)... The use of classifiers can be added to those other linguistic traits previously discussed, which establish India as one linguistic area ('an area which includes languages belonging to more than one family but showing traits in common which are found not to belong to the other members of (at least) one of the families') for historical study. The evidence is at least as clear-cut as in any part of the world... Some of the features presented here are, it seems to me, as 'profound' as we could wish to find... Certainly the end result of the borrowings is that the languages of the two families, Indo-Aryan and Dravidian, seem in many respects more akin to one another than Indo-Aryan does to the other Indo-European languages. (We must not, however, neglect Bloch's final remark and his reasons therefor: 'Ainsi donc, si profondes qu'aient ete les influences locales, elles n'ont pas conduit l'aryen de l;inde... a se differencier fortement des autres langues indo-europeennes.')"
The profundity of these observations by Emeneau and Bloch will be tested through clusters of lexemes of an Indian Lexicon, which relate to the archaeological finds of the civilization. These clusters of lexemes will be presented in a separate website together with the corpus of inscriptions of the civilization as aids to the process of deciphering the pictorials and signs on the inscriptions.
There is a Sarasvati River running parallel to River Banas and joining the Little Rann of Kutch. There is a Sarasvati River joining the River Luni near Pushkar, Ajmer. There is a Sarasvati (Haraqaiti) in Afghanistan, a tributary of the River Kubha. All these rivers may have been so-named reminscing the days of the civilization days of the Ghaggar-Hakra-Nara wadi's which constituted the Sarasvati River joining the Arabian Sea near Lothal.
That Sarasvati was a more important river than the Sindhu may be noted from the following observations of John Marshall (1931, pp. 1-6): "(Mohenjo-daro) stands on what is known locally as the 'The Island'-a long, narrow strip of land between the main river bed and the Western Nara lop, its precise position being 27.19N by 68.8E, some 7 miles by road from Dokri... Twelve centuries ago, when the Arabs first came to Sind, there were two great rivers flowing through the land: to the west, the Indus: to the east, the Great Mihran, also known as the Hakra or Wahindah. Of these two rivers, the eastern one seems to have been the more important... Major Raverty, the foremost authority on the subject, concluded that at the time of the Arab invasion the main channel of the Great Mihran flowed a line roughly coincident with the existing Eastern Nara canal, which was once an important rive rbed (i.e. it passed close by the city of Alor...flowed...west of Umarkot, and so the Rann of Cutch (then an estuary of the sea) and by the Kori creek to the Arabian Sea. Cf. Raverty, The Mihran of Sind, and its tributaries, JASB, Vol. LXI, 1892, pp. 156-508). According to him, the terminal course of the Indus, which flows by Mohenjo-daro, was then a subsidiary branch of the Mihran, buts course was not the same as at prsent... the existence of two important Chalcolithic sites of Mohenjo-daro and Jhukar, the one in the near vicinity of the Indus, the other of the Western Nara loop..."
"Griffin Vyse recalls observations that Alexander the Great had also sailed to the great lake and to the sea by this 'eastern branch of the Indus'...'the eastern or greater arm of the Mikran described by Rashid-ud-deen as branching off from above Mansura to the east, to the borders of Kutch, and known by the name of Sindh Sagara (Elliot, Vol. I, p. 49). This ancient river is also identical with the Sankra Nala which was constituted by Nadir Shah the boundary between his dominions and those of the Emperor of Delhi."
Back to Table of