II

The Palaeolithic Context

RESEARCH BACKGROUND

The modern phase of palaeolithic studies in India began in the 1930s when there were attempts to go beyond the succession and distribution of artefacts and consider the issues of related environment and geochronology. The most significant publication of this period was Studies on the Ice Age in India and Associated Human Cultures by H. de Terra and T.T. Paterson in 1939. This was based on their work (1935), in collaboration with Teil-hard de Chardin, from Kashmir to the Salt range, with a detailed focus on the Soan river valley, a tributary river valley of the Indus. Its background was provided by a few earlier discoveries in this region which drew the attention of de Terra who was a geologist.¹

The interrelationship between the stratigraphic profile of prehistoric tools and related environment was first highlighted in a major way in India by a study of tools collected in the eastern ghats of the Andhra region by L.A. Cammiade, who collaborated with M.C. Burkitt of Cambridge University in 1930. Their observations were the following:

A pluvial period or a period of heavy rainfall witnessing the formation of laterites on the east coast between the Krishna and the Palar—no sign of human presence.

Series 1—handaxes of quartzite—a long dry period denoting open plains—African parallels of some artefact types of this industry;

Series 2—a second dry period after a second pluvial phase which led to the formation of detrital lateritic and clay beds—more neatly made handaxes associated with a flake industry—raw material quartzite but some sandstone and chalcedony as well—this series of tools continues to occur in the river alluvium of a phase which is humid but not humid enough to suggest pluvial conditions;

Series 3—decrease in the rainfall leading to denudation—the presence of blunted back blades, burins, end-scrapers and ‘planing tools’ of a flinty nature such as Lydian stone;

Series 4—similar climatic condition—the large-scale occurrence of microliths.²

Map 1 Major Palaeolithic Distribution Area: 1. Las Bela, 2. Potwar Plateau, 3. Sukkur-Rorhi hills, 4. Milestone 101, 5. Kashmir and Ladakh, 6. Panjab Siwaliks, 7. Kutch, 8. Saurashtra, 9. mainland Gujarat, 10. Narmada valley, 11. Maharashtra, 12. Karnataka, 13. Andhra, 14. Tamil Nadu, 15. Palghat area of Kerala, 16. Orissa, 17. West Bengal and Bihar, 18. western Tripura, India and Lalmai hills, Bangladesh, 19. Son and Belan area, 20. western Nepal, 21. North-western Frontier Province, 22. west and east Rajasthan

Observations of more or less the same type were made in 1932 by Cammiade, Burkitt and F.C. Richards regarding a collection of tools near Madras.

However, the Terra—Paterson work is more significant because it went beyond such palaeoclimatic correlations and introduced a geochronological scale on the European model. Moreover, having established a geochronological scale for its main study area between Kashmir and the Salt range, it extended the framework, through archaeological comparisons, to include the Narmada valley in central India and the area around Madras in the south. The point is not whether they were successful or even logical in their quest. What is relevant is that they were believed to be successful and logical till the 1980s when the scientific basis of their groundwork was found to be wanting even in the context of the Potwar plateau where they primarily worked. Thus, it was a dominating influence in Indian palaeolithic research for a long time, and it is worth knowing, even at this chronological distance, what the work meant.

Basically, what they did was to postulate the existence of a number of tool-bearing terraces along the Soan river and correlate them to the already known Quaternary glacial cycle in Kashmir. The correlation was on the following basis: T-D = II glacial; T-1 = II interglacial; T-2 = III glacial; T-3 = III interglacial; T-4 = IV glacial; T-5 = recent. The last terrace, T-5, is immediately earlier than the present level of the river. Further, a complete parallelism was assumed between the Himalayan glacial cycle in Kashmir and the Alpine glacial cycle in Europe so that there was no difficulty in understanding the Soan valley succession of stone tools in terms of a global geochronological perspective.

In the Narmada valley of central India, four terraces were recognized by de Terra only in the tributary valleys but they were impossible to distinguish in the main valley where three different cycles of sedimentation were recognized. These cycles of sedimentation came to be known as the ‘lower group’, ‘upper group’ and ‘regur or cotton soil group’. In the present context, regur or black cotton soil is irrelevant. The lower and upper groups both show a gravel bed followed by a bed of clays and silts. The gravel bed or the boulder conglomerate of the lower group is more cemented than the gravel of the upper group. Similarly, the clay of the lower group is more intensely coloured and richer in concretions than the clay of the upper group. Thus, basically, it is a succession of cemented basal gravel, clay, finer gravel and clay, the whole being covered by regur or cotton soil. Two further points are noteworthy. First, there is no faunal difference between the two groups. Bos namadicus, Elephas namadicus, etc. being found in both groups. Second, Acheulian handaxes were found in the basal level of the lower group, thus making the Acheulian industry contemporaneous with the deposition of the basal conglomerate. The upper group yields rolled Acheulian industry but there is also an unrolled and apparently contemporaneous assemblage of flakes and discoidal and pebble cores of quartzite and trap, recalling the Late Sohan industry of the Soan valley. According to the Terra—Paterson scheme of correlations, T-1 and T-2 of the Soan valley with their Early Sohan industry are contemporary with the ‘lower group’ of the Narmada whereas the Soan valley T-3 and T-4 are contemporary with the Narmada ‘upper group’.

In south India, three terraces were observed in the Kortalayar plain near Madras. These terraces were formed after a deposition of detrital laterite over a boulder conglomerate. No fauna was found in them but they contained palaeolithic tools. The better known sites of this region are Vadamadurai and Attirampakkam. At Vadamadurai the collected tools were divided into three groups based on patination and typology, whereas most of the tools from Attirampakkam display a well-developed and late Acheulian industry.

D. Sen and V.D. Krishnaswami were among the Indian assistants in the Soan valley work. Both began to publish their findings from the late 1930s onwards. From the stratigraphical point of view, an important publication of this period was by K.R.U. Todd who studied the palaeolithic industries of Bombay in 1932 and 1939. The section published by him in 1939 shows the succession of bedrock, lower clay, lower gravel, middle clay, upper gravel and upper clay. The lower clay is devoid of tools but the reddish-brown lower gravel contained handaxes of presumably early type. On top of this gravel he found fine, developed handaxes in association with flake implements. This gravel bed suggested pluvial conditions to him. The overlying ‘middle clay’ suggested ‘the setting in of milder conditions’, and although no tool was found within it, a blade industry including cores, blades and scrapers was found on top of this clay. The upper gravel itself did not apparently show any industry but on its top lay a working site of blade-and-burin industry. The upper clay contained ‘a more developed blade and burin industry’, and finally, a microlithic industry appeared. In the 1940s, H.D. Sankalia observed the sequence of prehistoric industries in Gujarat which was noticed much earlier by Foote; this was also the time when D. Sen reported the occurrence of a palaeolithic industry in the lateritic conglomerate of Mayurbhanj in Orissa. The available knowledge about prehistoric India was assessed by V.D. Krishnaswami, the newly appointed ‘prehistorian’ of the Archaeological Survey of India, in the third issue of its official organ, Ancient India, in January 1947. This was the beginning of palaeolithic research in independent India.³

In retrospect, the foundations of this research in post-1947 India were fundamentally laid down in the 1930s and 1940s. Some of the basic features of palaeolithic research in independent India—the succession of lower gravel-lower clay and upper gravel—upper clay deposits, the equation of the gravels with pluvial conditions, the equation of the deposits of clay/silt with drier conditions, the finding of tool-bearing terraces in the Siwaliks, the percentage calculation of chopper-chopping and biface elements, etc.—can be found in the writings published in the 1930s and 1940s. It is through them that the earlier research of Foote and his contemporaries were integrated into the main framework of knowledge about prehistoric India.

At the invitation of the Government of India, F.E. Zeuner of the Institute of Archaeology, London, visited India in 1949 ‘to explore the possibilities of developing research in the prehistory and geochronology of India’. He was assisted in the field by a number of people from the Archaeological Survey and the Deccan College. A number of publications ensued, and to assess this research work as a whole, one has to refer to the publication of H.D. Sankalia's Prehistory and Protohistory in India and Pakistan in 1963. First, the Terra–Paterson work in the Soan valley was extended to the Sirsa and Beas–Banganga valleys in the Himalayan foothills by building up a sequence of palaeolith-bearing terraces. In the river valleys of the Peninsular block, the emphasis was on building up a wet—dry succession of gravel and silt. A few new areas, such as Malwa, Rajasthan and Andhra, came in for detailed investigation, mostly by Sankalia's own students. One should also mention the publication of a monograph on the prehistoric excavations in Mayurbhanj, Orissa, by N.K. Bose and D. Sen of Calcutta University in 1948, and the work in the Singrauli valley in the Vindhyan area of Uttar Pradesh by V.D. Krishnaswami and K.V. Soundararajan of the Archaeological Survey in 1951. However, the palaeolithic studies of this period were still overshadowed by speculations regarding the distribution of handaxe elements and the so-called Sohan-type tools made on pebbles. The occurrences of a middle palaeolithic flake-blade industry and upper palaeolithic blade-and-burin industry should have been clear from the Cammiade–Burkiu work in Andhra and Todd’s work in Bombay. Although Krishnaswami in 1953 drew attention to the significance of these earlier findings, Sankalia tended to miss this significance and dwelt only on the middle palaeolithic or what he called in the 1963 publication ‘Middle Stone Age’.⁴

The second edition of Sankalia’s The Prehistory and Protohistory of India and Pakistan was published in 1974. The list of the areas with lower, middle and upper palaeolithic tools in this book shows a major geographical expansion of the subcontinental palaeolithic studies, virtually all of it in modern India, thanks to the initiative taken by Sankalia and his Ph.D. students. Interestingly, what could have been a breakthrough in the establishment of the early antiquity of palaeolithic tools in the subcontinent was not properly pursued either in this or in the later period. Sankalia’s discovery of a massive flake in the Second Glacial Boulder Conglomerate deposit of the Liddar valley at Pahalgam in Kashmir and his preference to put this flake in the earlier First Interglacial context⁵ were not borne out by the subsequent investigation of R.V. Joshi and others who, in 1974, put the Pahalgam and related Boulder Conglomerate deposit and its stone tools only in the middle Pleistocene.⁶ That theirs was not the last word on the date of palaeoliths in this region has been suggested by the two plus million years ago (mya) dating of tools in the Potwar plateau, Jammu and Ladakh. Among the other new researches of the period one may mention those in Singhbhum and West Bengal, Orissa beyond Mayurbhanj and new areas in Maharashtra, Karnataka, Madhya Pradesh, Gujarat, Andhra and Uttar Pradesh. The three-fold stratigraphic and typological classification of the palaeolithic period into three successive periods of lower, middle and upper palaeolithic was by then consolidated, with the most important evidence of the upper palaeolithic coming from the Kurnool caves and elsewhere in Andhra, the Belan valley of Uttar Pradesh and the site of Patne in the Chalisgaon area of Maharashtra. Further, in an appendix to his volume Sankalia could list a handful of radiocarbon dates from the middle palaeolithic contexts. In one sense, these two books by Sankalia show that there was considerable progress in Indian palaeolithic research between 1963 and 1974, especially in the areas of stratigraphy and typology. The stratigraphic and typological continuity between the three horizons of the Indian palaeolithic could no longer be doubted, and there was certainly a better grip over the typological range of the tools themselves. At the same time, it has to be noted that within the range of the Indian lower palaeolithic no acceptable evolutionary sequence of tool-types emerged. A.P. Khatri’s claim of an Oldowan-type pebble tool element at the base of the Narmada valley palaeolithic succession and A.K. Ghosh’s claim of an Abbevillian to Acheulian type of handaxe succession in Singhbhum in south Bihar could not be given a sound basis. In a different form Khatri’s claim has been rejuvenated in the context of later excavations in the Durkadi stream area of the Narmada valley, but Ghosh’s picture in Singhbhum was doubted by Sankalia himself and has had no support since then.

Since the mid-seventies palaeolithic research in India has branched out in several directions beyond stratigraphical and typological considerations. First, a few traces of primary occupation of palaeolithic people have been exposed through controlled excavations at Paisra in Bihar, Bhimbetka in Madhya Pradesh and Hunsgi in Karnataka. However, none of these excavations yielded any sign of organic material which could be used for a fuller reconstruction of the contemporary life. Second, there has been some emphasis on working out the geochronology of lower palaeolithic and middle palaeolithic horizons. Admittedly, the dates available are still very limited in number, and at least in one case the dates have not contributed to a clear picture. However, the availability of some absolute dates for the Indian lower palaeolithic and a handful of non-radiocarbon dates from the middle and upper palaeolithic as well, must be considered an important step forward. Along with this has developed a concern for geoarchaeology and a quest for precision in the typological analysis of tool assemblages. Both these developments are very welcome, but as far as geoarchaeology is concerned, the use of geological and geomorphological jargon and the presentation of archaeological data in various computer-derived forms do not necessarily announce the maturity of geo-archaeological approach to palaeolithic studies in India. Rigorous multi-disciplinary investigations in the field of sediment deposits have been conducted in a rather limited area in Rajasthan. Some amount of sea-level studies have been conducted in Gujarat. In both cases the archaeological correlates are clear. In the case of Kashmir, which has seen extensive multi-disciplinary palaeoclimatic investigations, the archaeological correlates seem to be elusive before the upper palaeolithic stage. A study of this kind was also undertaken in the context of the Son valley in Madhya Pradesh and the Belan valley of Uttar Pradesh. Under the circumstances, what has been really successful in the field of palaeolithic studies in modem India is the development of an ethnography-based settlement—subsistence approach. This has been applied with success in Andhra and Karnataka, and may be pursued in other areas as well. The core of this approach lies in enunciating the possible ways of exploiting the foodresources available in the wild on the model of the local hunter-gatherers and even peasants and trying to explain the distribution of sites in the study area on that basis. Obviously, this approach has great possibilities in a country like India where subsistence behaviour on various levels can still be observed. This type of study can be undertaken by archaeologists without a large outlay of resources and personnel which the multi-disciplinary investigations require.

The aim of this brief research review of the palaeolithic context was to offer a generalized academic perspective of the various types of research which have been conducted in this field. For the period before 1974 some detailed discussion was needed to show how much of the modern research is rooted in that period, especially in the writings of Cammiade and Burkitt, de Terra and Paterson, and Todd. The rest of this chapter will be devoted to a study of the post-1974 situation.⁷