The Lost River: On The Trail of Saraswati

by Michel Danino

  At the other end of the spectrum, we meet archaeologists who see no need for such a device, finding no trace of the Aryans’ arrival on the ground, and who therefore prefer to equate the Harappan with the Vedic. Dholavira’s excavator R.S. Bisht, for instance, finds that the city’s upper, middle and lower towns ‘temptingly sound analogous to three interesting terms in the Rig-Veda, viz., “parama” [upper], “madhyama” [middle] and “avama” [lower] . . .’85 ‘In the tripartite divisions of the Harappan city of Dholavira,’ he claims, ‘we find what the Vedic seers were trying to conceive of in supernatural and divine world’.86 Bisht also extracts from hundreds of passages of the Rig Veda numerous terms for villages and towns, fortifications, houses of various types, rooms, even doorways—a rich material vocabulary that belies the conventional view of early Vedic people as ‘pastoral’ or ‘nomadic’, and ignorant of anything remotely resembling city life.

  A few years ago, the scholar Bhagwan Singh went further and listed hundreds of Rig Vedic words related to trade, industry, etc.87 Even if one may often hesitate to accept his interpretations, as I do, once again the sheer diversity of the vocabulary is puzzling. In the field of governance, for instance, B.B. Lal asks, ‘What do we do with terms like rājan, samrāt, janarāj, rāstra, Samiti, sabhā, adhyaksa, etc. occurring in the Vedic texts, which clearly refer to categories of rulers, assemblies, etc.?’88 The spacious assembly halls at Mohenjo-daro and Harappa come to mind.

  Opposing such a perspective is the dominant view that the Rig Veda, apart from reflecting a non-urban environment, knows the horse (which is absent from Harappan iconography), but not writing (known to Harappans), and that linguistic considerations prevent us from dating the Rig Veda beyond the middle of the second millennium BCE. The debate—which is inextricably linked with the Aryan invasion or migration theory—has been raging for eight decades, and is not likely to end soon. It has generated a copious literature, including a few wide-ranging studies highlighting the complexity of the issue, and bringing in new disciplines such as anthropology, genetics and astronomy.89 A growing number of participants in the debate (myself included) have argued that the stark contrast between the Harappan and the Vedic, proposed by linguists, historians and other scholars, remains rooted in colonial misconceptions and misreadings of the Vedic texts. At the same time, it is fair to stress that there are complexities and potential pitfalls in trying to equate a popular culture, such as the Harappan, with a textual one, such as the Vedic.

  But my purpose in this chapter and the preceding one was not to enter into that labyrinthine debate; it was limited to offering a glimpse of the overall legacy of the Indus-Sarasvatī civilization: in whatever way we may wish to explain it, a manifold cultural continuum emerges between the Indus and the Ganges civilizations. As Lal puts it:

  Even today there is no walk of life in which we cannot discern the grass-roots features of this ancient [Harappan] civilization : be it agriculture, cooking habits, arts, crafts, games, ornaments, toiletry, religious practices or social stratification.90

  In no way does archaeology or literature justify the picture of a tradition ‘rudely and ruthlessly interrupted by the arrival of new people from the west’, as Piggot asserted, with the fabled Aryans in mind. Nor is the Gangetic culture ‘diametrically opposed’ to the Harappan culture, as Basham would have it.

  As a result, the old concept of a ‘Vedic night’ throwing open a gaping chasm between the two civilizations now stands rejected. According to Kenoyer, ‘It is clear that this period of more than 700 years was not a chaotic Dark Age, but rather a time of reorganization and expansion.’91 To be precise, ‘current studies of the transition between the two early urban civilizations claim that there was no significant break or hiatus.’92 Jarrige agrees: ‘This famous vacuum that was sometimes called Vedic night . . . has been filling up more and more thanks to numerous findings.’93 In his opinion, the considerable changes that followed the end of the Indus cities are to be understood ‘within the framework of a continuity with the preceding millennia, without any radical break of the sort too often proposed earlier’.94 Shaffer is equally categorical: ‘The previous concept of a Dark Age in South Asian archaeology is no longer valid.’95

  There is, however, a crucial element of the debate that is missing here: the Sarasvatī. The lost river has been summoned as a witness, and since in between she became the goddess of vāch, or speech, she deserves a fair hearing.

  {11}

  The Sarasvatī’s Testimony

  The rediscovery of the 1500 km-long bed of the Sarasvatī in the nineteenth century has far-reaching repercussions on our understanding of the origins of Indian civilization.

  Let us briefly survey the ground we have covered. The most ancient Indian text in our possession, the Rig Veda, reveals a keen awareness of the geography of the Northwest, an awareness expressed mostly through river names. Among them is the Sarasvatī, a mighty, impetuous river flowing ‘from the mountain to the sea’ and located between the Yamuna and the Shutudrī (Sutlej). Subsequent Vedic literature—the Mahābhārata, several Brāhmanas and Purānas among other texts—paint a consistent picture of the location, break-up and recession of the river. As a matter of fact, when nineteenth-century British officials explored the region, they found quite a few seasonal streams rising in the Shivalik Hills, but no major river flowing between the Yamuna and the Sutlej.

  They did record three things, however: a tradition that there was once a river flowing westward through the region (the ‘Lost River of the Indian Desert’); a broad, dry bed called the Ghaggar or Hakra, partly filled during strong monsoons; and, at the foot of the Shivaliks, a small stream called ‘Sarsuti’, flowing down from a tīrtha which tradition claims to be the source of the ancient Sarasvatī.

  In the same nineteenth century, these findings, correlated with the testimony of Sanskrit texts, convinced most Indologists that the bed of the Ghaggar-Hakra was the relic of the Sarasvatī. By the turn of the century, maps and gazetteers reflected that identification. The ‘lost river’ had been found again—the only question marks left had to do with the exact process of its decline, and whether it had reached the sea on its own or was a tributary of the Indus.

  To Max Müller and other Indologists of his time, the Sarasvatī’s disappearance was a milestone in the timeline of ancient India: because ‘the loss of the Sarasvatī is later than the Vedic age’, it provided in effect ‘a new indication of the distance which separates the Vedic age from that of the later Sanskrit literature.’1 In the next century, the twentieth, archaeology added a dramatic new piece to the puzzle by unearthing, along the dry beds of the Ghaggar-Hakra and its tributaries, numerous sites small and large, many of them belonging to a civilization named after the Indus. Changes in the distribution patterns of these sites both in India and Pakistan suggested that the Ghaggar-Hakra had started breaking up sometime in the third millennium, and that by 1900 BCE, much of its central basin had gone dry. A further confirmation of this was that some of the post-urban sites were found right on the dry bed of the Ghaggar.2

  The jigsaw puzzle might now be said to be complete, and we would be justified in affixing, with a satisfied flourish, a final curlicue at the end of the saga.

  But wait—there is a catch. Sarasvatī, standing in the witness box, looks as if she is about to make some awful confession.

  Let us contemplate the picture painted by J.M. Kenoyer in 1998:

  In the east, the ancient Saraswati (or Ghaggar-Hakra) river ran parallel to the Indus . . . Towards the end of the Indus Valley civilization, the ancient Saraswati had totally dried up and its original tributaries were captured by two other mighty rivers . . . The gradual drying up of the Saraswati river is an event documented both geologically as well as in the sacred Vedic and Brahmanical literature of ancient India . . . Many episodes of the Rig-Veda take place along the sacred Saraswati.3

  This is a fine summary of much that we have explored so far (and Kenoyer goes on to refer to Oldham and Wilhelmy). B
ut then, if ‘the ancient Saraswati had totally dried up’ towards the end of the Indus civilization, that is, around 1900 BCE, and if ‘many episodes of the Rig-Veda take place along the sacred Saraswati’, does it not follow that the said episodes took place before the end of the Indus civilization, while the river was flowing? There seems to be no escape from this conclusion.

  Yet, following Max Müller, all conventional history books and encyclopædias tell us that the Rig Veda’s hymns were composed by ‘Aryans’ who entered the subcontinent around 1500 BCE and pushed on towards the Yamuna-Ganges region, crossing it sometime between 1200 and 1000 BCE. Whatever the exact dates proposed (there are countless variations of this scenario), the said Aryans could only have settled in the Sarasvatī region after 1400 or 1300 BCE—centuries after the river had ‘totally dried up’, in any case. We are, therefore, asked to believe that the Aryans crossed at least five large rivers—the Indus and its four tributaries (see Fig. 2.2)—to settle down on the banks of a long, dry river, which they went on to extol as ‘mighty’, ‘impetuous’, ‘best of rivers’, etc. The proposition is incongruous in the extreme.

  Of the two scenarios, the first alone is plausible: the hymns that praise the Sarasvatī—and some of them are found in the oldest books of the Rig Veda4—must have been composed while the river was still flowing, which can be no later than the third millennium BCE.

  The contradiction between the two scenarios is vividly illustrated in a fine book on the Indus civilization, authored in 1997 by Raymond and Bridget Allchin. We have already quoted from it their ‘most moving experience’ at Kalibangan, where they gazed at ‘the flood plain of the Sarasvatī still clearly visible’ north of the settlement. Like Kenoyer, the Allchins acknowledge that the Sarasvatī

  continued to flow down to c. 2000 BC. The major reduction of sites in the Early Post-urban period (c. 2000-1700 BC) . . . strongly suggests that a major part of the river’s water supply was lost around that time; while the final settlement pattern of the Late Post-urban period indicates that the river was by then dry (i.e. by c. 1300-1000 BC).5

  Yet earlier in the same book, the Allchins state that the Sarasvatī is ‘recorded in the Rig Veda as a major river between c. 1500 and c. 1000 BC’.6 But how is that possible when archaeological evidence, in their own words, shows the river to have lost ‘a major part of its water supply’ between 2000 and 1700 BCE? This chronological impossibility is the result of the above phrase ‘between c. 1500 and c. 1000 BC’—artificial dates that rest purely on the old Aryan invasion theory, not on any physical evidence.

  Gregory Possehl notes the same contradiction and attempts to resolve it. We referred earlier (p. 149) to his long and useful discussion on the Sarasvatī’s evolution and the river’s chronology ‘actually founded in archaeological data’. In agreement with his colleagues, Possehl explains that ‘at the end of the third millennium the strong flow from the Sarasvatī dried up’.7 Aware of the Rig Veda’s eulogies of the river, he proposes that they may be merely ‘recollections’ of the time when the Sarasvatī was ‘a river of great magnitude’.8 Why ‘recollections’ is unclear: the text’s vivid descriptions of the river evoke anything but that. Possehl does not however press his own explanation, and comes to the crux of the whole issue:

  This [the drying up of the Sarasvatī towards the end of the third millennium] carries with it an interesting chronological implication: the composers of the Rgveda were in the Sarasvatī region prior to the drying up of the river and this would be closer to 2000 BC than it is to 1000 BC, somewhat earlier than most of the conventional chronologies for the presence of the Vedic Aryans in the Punjab.9

  But ‘somewhat earlier’ is quite a euphemism for a whole millennium or more. If the Vedic Aryans ‘were in the Sarasvatī region’ earlier than 2000 BCE, they must have entered the subcontinent between 2400 and 2200 BCE at the very latest, bringing us to a date that no proponent of the Aryan migration into India would be prepared to accept.

  In a word, the Aryan theory collides head on with a Sarasvatī drying up in the late third millennium.

  THE SARASVATĪ AND THE ARYAN PROBLEM

  While Kenoyer, the Allchins and Possehl cautiously refrain from concluding, Indian archaeologists such as B.B. Lal, S.P. Gupta, V.N. Misra, Dilip Chakrabarti,10 apart from other scholars, have proposed to put two and two together: the poets who sang the praise of the Sarasvatī lived on its banks while it flowed, therefore before ‘the end of the third millennium’, as Possehl just told us.

  Moreover, the Vedic rishis state that the river flowed ‘from the mountain to the sea’, which takes us back to an even earlier date: as we saw (p. 151), Possehl’s maps point to a break-up of the river during the Mature Harappan phase,11 in agreement with Mughal’s analysis of the pattern of sites in Cholistan (p. 149) and with two independent isotope studies; Fig. 6.8 reflects this clearly. That hymn from the seventh mandala of the Rig Veda must, therefore, have been composed before 2500 BCE—a whole millennium earlier than the conventional dates.

  Yet, not all early Sanskritists would have objected to such a date. Moritz Winternitz, a noted German Indologist, for example, chose to disagree with the dominant chronology:

  We cannot explain the development of the whole of this great [Vedic] literature if we assume as late a date as round about 1200 BC or 1500 BC as its starting point. We shall probably have to date the beginning of this development to about 2000 or 2500 BC . . .12

  This period is precisely the Mature phase of the Indus-Sarasvatī civilization—although Winternitz could not have known that, for he wrote the lines in 1907, long before the discovery of Harappa and Mohenjo-daro. In 1923, just a year before Marshall announced this discovery to the world, Winternitz was invited by Rabindranath Tagore to Vishwa Bharati; in the course of a lecture given in Calcutta, alluding to Max Müller and his followers, Winternitz made this extraordinary statement:

  I, for my part, do not understand why some Western scholars are so anxious to make the hymns of the Rgveda and the civilisation which is reflected in them so very much later than the Babylonian and Egyptian culture.13

  There is a key word here: ‘civilization’. Ever since Max Müller made it a dogma that the Rig Veda reflected a ‘primitive’, ‘nomadic’ and ‘pastoral’ culture, those labels have stuck, despite much contrary evidence provided by the text itself. In 1958, for instance, Sanskritist and linguist B.K. Ghosh, who accepted the mainstream invasionist view, nevertheless felt compelled to observe that ‘the Rgveda clearly reflects the picture of a highly complex society in the full blaze of civilisation.’14

  This leads us to a weighty question: we are told that the Harappan culture is pre-Vedic and therefore ‘non-Aryan’, and that the Aryans entered the Indus plains in the mid-second millennium BCE. We know where the latter lived and composed their hymns: apart from terms like saptasindhava, the hymn in praise of rivers (Fig. 2.2) shows that the whole of the Northwest was the Rig Vedic arena. But that also happened to be precisely the core Harappan territory—and only one civilization has been found there, not two: on the ground, no material culture has been found that might be identified with any ‘Aryan’ settlements. J.-M. Casal’s statement made forty years ago still holds good:

  Until now, Aryans have eluded all archaeological definition. So far, no type of artefact, no class of pottery has been discovered that would enable us to say: ‘Aryans came this way; here is a typically Aryan sword or goblet!’15

  Many scholars have, nevertheless, attempted to identify the elusive Aryans with this or that regional culture of Late Harappan times (Gandhara Grave of the Swat Valley, Pirak, Cemetery H, Painted Grey Ware, etc.), but besides their mutual incompatibility, all these hypotheses rest on arbitrary choices of what might constitute, on the material level, an ‘Aryan culture’. Besides, none of them covers the entire conjectured migratory path of the Aryans from Central Asia to the Ganges.

  In fact, two points have been broadly accepted by the archaeological world: the absence of any intrusive material c
ulture in the Northwest during the second millennium BCE,16 and the biological continuity evidenced by the skeletal record of the region during the same period, a continuity confirmed by over a dozen recent genetic studies.17 In other words, the arrival of the Aryans—an event which, we are told, would radically change the face of the subcontinent’s cultural and linguistic landscape—is completely invisible on the ground.

  If we add to this the cultural continuity discussed in the previous two chapters and the testimony of the Sarasvatī, the simplest and most natural conclusion is that Vedic culture was present in the region in the third millennium.

  At this stage, it would be tempting to lay down the equation ‘Harappan = Vedic’ as the next logical step. To quote B.B. Lal:

  The geographical area covered by the Rigvedic people, as given in Rig-Veda 10.75 [the hymn to the rivers, see Fig. 2.2], lay from the Gangā-Yamunā on the east to the Indus on the west. A simple question may now be posed: Which archaeological culture occupied the above-mentioned region during the period prior to 2000 BC [when the Sarasvatī dried up]? The inescapable answer is: it was the Harappan Civilization.18

  Lal, however, pleads for patience till the Indus script has been conclusively deciphered. Moreover, we cannot cut the ‘Aryan knot’ without studying its tentacular ramifications into linguistics, archaeoastronomy, anthropology and genetics, comparative mythology and religion, besides a few other fields. Any proposed solution to the Aryan problem must satisfy, or at least answer, all those disciplines and their legitimate demands. We must leave that attempt to another study.19

  The purpose of our journey through the mists of the Bronze Age is to follow the evolution of the Sarasvatī river and delineate its twin role: it was not only ‘the true lifeline of Vedic geography’, in Renou’s words, but also, in those of V.N. Misra, ‘the lifeline of the Indus Civilization’ (along with the Indus).