Early India and Pakistan to Ashoka c7: Chapter IV. More Stones
Early India and Pakistan to Ashoka c7
- Language
- English
Chapter IV. More Stones
c7.2
THE PRINCIPAL INDUSTRIES described in the last chapter may have begun about 400,000 years ago and may have ended—to use an abrupt term for what was certainly a very long and ragged process—something under 100,000, perhaps nearer 50,000, years ago, with survivals, as in South Africa, down to relatively recent times. So far, our handaxes and pebble-choppers have offered little hint of succession or replacement, and to this sequel we must now turn, though without great hope of definition. The example of Africa shows how much can be achieved by a single generation of scientific field-workers, and India is equipping herself for a similar advance. Meanwhile the present somewhat dubious position may be summarized under two questions. First, what evidence is there for a distinctive Upper Palaeolithic phase in India, as in Africa and Europe? And, secondly, is there in India, again as elsewhere, a recognizable Mesolithic between the Old Stone Age and the New—if indeed there be a New Stone Age at all? This question involves, *inter alia*, the problem of the microlithic or small-stone industries which occur profusely in parts of the subcontinent and stoutly resist easy classification.
c7.3
First, then, the Upper Palaeolithic. In the north, as already noted (p. 40), there was a tendency in the Late Soan towards finer flake-implements, apparently without handaxes. This trend is appropriate to the Upper Palaeolithic but needs much further investigation; for example, Terrace 4 of the Soan series is at present an uncertain quantity, and surface sites of small flakes and pebble-tools on the Sil (tributary of the Soan) near Pindi Gheb and at Dhok Pathan nearby are without stratigraphical context. Like the flakes, the pebble-tools are smaller than the average Soan type, and include a new form, a kind of awl made on a small oval pebble flaked on both sides at one end to make a sharp point. Associated flakes are similar to Late Soan: a few are retouched, and one is a steep scraper with fine trimming. Faceting of the platform is rare. 'The age of this series', adds Paterson, 'is uncertain, but, to judge from typology and the state of preservation, it is undoubtedly fairly late and must be at the earliest contemporary with the Late Soan, possibly later, perhaps even of the fourth glacial age.'²⁹ In the Jhelum valley of Kashmir, de Terra and Paterson found 'great numbers of artificially flaked stones, among which were flakes and cores reminiscent of palaeolithic technique' including the Levalloisian; 'but in all these places it was certain that the flakes are associated with pottery-bearing layers of either neolithic or historic date'.
c7.4
That has not carried us very far—or, rather, it has carried us too far—and the three terraces which have been noted in the Beās and Bāngangā valleys, with their consistent assemblage of choppers and handaxes (p. 40), likewise take us at present no further in this problem. But in central India the evidence begins to clear. In the last chapter attention was drawn to a remarkable disc and blade industry described by Sankalia from the upper gravels of the Pravara, tributary of the Godāvarī.³⁰ This industry apparently overlies the bulk of the Acheulian handaxes which, with rare survivals just sufficient to suggest a cultural overlap, are, if the account is correctly interpreted, concentrated in Gravel I, the lowest, of the geological succession (above, p. 46). It does not fit easily into the current categories, and may provisionally be known as the 'Upper Pravara' until its position and affinities are more amply established. Its materials are agate, chert, chalcedony and jasper, and its forms constitute a miscellany of scrapers, blades and cores. The cores, to some of which the somewhat loose term 'Clactonian' is applied, include also very non-Clactonian examples worked all over by pressure-technique, with irregular wavy edge. Amongst the blades a type about an inch long with blunted back and retouched edge merges into an end-scraper in examples with more emphatic flaking; but there are also half-round scrapers and side-scrapers with extensive retouch, and indeed the blade-scraper series exhibits a wide range and an evolved tradition. A few flakes show Levallois characters with faceted striking-platform and some retouching round the edges. Two or three implements with beaked point may be described as burins—a relatively rare type in India.
c7.5
Fig. 12 "Series 2" from Nevasa, Pravara valley: 1, chert burin; 2, chert double-ended burin; 3, agate point; 4, chalcedony point; 5, quartz point; 6, pointed chert flake; 7, pointed chalcedony flake; 8, leaf-shaped chert flake; 9, chert scraper; 10, chalcedony awl. ½
c7.6
Figs. 12-13
c7.7
Microlithic elements occurred in the overlying silt, above a considerable intervening deposit devoid of artifacts; but the Upper Pravara as a whole showed no specifically microlithic trend, and a vaguely Aurignacian facies can be claimed for it without emphasis. It shows some analogy with earlier phases of the Kenya Capsian or Aurignacian, and may in fact be the forerunner of an equivalent Indian development. It is at least sufficient to encourage further search for a true Upper Palaeolithic phase in the Indian succession.
c7.8
A comparable hint of an Upper Palaeolithic horizon was detected by Commander K. R. U. Todd in a clay and gravel pit at Khandivli, 21 miles north of Bombay,³¹ though here again an admirable piece of pioneer work awaits consolidation and development. Upon the native rock, a deposit of bluish-brown clay, the Lower Clay, 3 to 50 inches thick contained rough tools and flakes 'somewhat recalling a Clacton industry' but occasionally showing more than one period of flaking. This industry approximates in kind to the pre-Soan. On the surface of the Lower Clay occurred another somewhat similar industry comprising scrapers, cores and choppers, sometimes in mint condition, with 'an early form of rostrate hand-axe'. Overlying the Lower Clay was the Lower Gravel, 1-6 feet thick, containing numerous implements of Abbevillian and Clactonian types in varying states of preservation. On the surface of the gravel were Clactonian flakes and Acheulian Fig. 13 'Series 2' from Nevasa, Pravara valley: 1, chert discoid; 2, chalcedony flake with prepared striking-platform; 3-5, chalcedony blades with blunted back and edge partly retouched; 6, tabular side-scraper, chert; 7-8, chalcedony end-scrapers; 9-10, chalcedony discoids. 1/2 handaxes in mint condition, with a number of unclassified flake-implements 'probably Acheulian in culture'. So far, all is reasonably plain sailing. In the present context it is the relationship of these lower layers to their successors that matters.
c7.9
1
c7.10
2
c7.11
3
c7.12
4
c7.13
5
c7.14
6
c7.15
7
c7.16
8
c7.17
9
c7.18
10
c7.19
Above the Lower Gravel was alternatively a further layer of clay known as the Middle Clay, or a sandy stratum representing probably the laterization of the upper part of the Lower Gravel. These deposits were themselves devoid of implements, but on their surface was a blade industry, including cores, blades and scrapers, with a few small handaxes made on flakes. On top of the sand or Middle Clay lay the Upper Gravel, about 3 feet deep, on which was found a working-floor with a blade and burin industry accompanied by little heaps of waste flakes of indurated shale. Over all were 18 inches of Upper Clay containing a more developed blade and burin industry, including polyhedral and angle burins and one of parrot-beak type. In and on the surface appeared the inevitable microlithic industry of these parts (see below, and Chapter 7).
c7.20
In this series, both the Lower and the Middle Clays may represent periods of pluviation, whilst the gravels may indicate phases of a semi-arid character,[*] but to equate the pluviations with the standard glaciations is, in the present state of knowledge, mere guesswork. At a venture, the Lower Clay might be matched with the penultimate glaciation, and, if so, is congenially associated with its pre-Soan Abbevillian industries whilst the Middle Clay may correspond to the last glaciation, which would thus be brought into a familiar terminal relationship with a blade, burin and scraper industry of a roughly Upper Palaeolithic facies. Beyond that oracular statement, even guesswork falters.
c7.21
In the south of India a partially equivalent Upper Palaeolithic seems to emerge as Series 3 of the Cammiade-Burkitt succession. In the Nandikanama Pass north of Madras (p. 48), on the surface of a red alluvial clay which is in turn sealed by red sandy soil—both layers which should be subsequent to the gravels containing the handaxe and flake-industries of Series 1 and 2—have been found a number of knife-blades of black (Lydian?) stone showing fine blunting down the backs, together with core-scrapers and burins mostly of the gouge type. Small crescents or 'lunates', however, of a normal microlithic type also occur in the Series, and, if in place, suggest that the industry may be too late to constitute an immediate follow-up of the Mid-Pleistocene Series, 1 and 2. On the other hand, Series 3 has an affinity with the so-called Capsian of Kenya, where lunates also occur with backed-blades, burins and end-scrapers. Incidentally, two radiocarbon dates for the partially comparable Upper Capsian of North Africa give c. 6500 and 5000 B.C.,²² and, if the Indian dating should prove to be of the same order or even somewhat earlier, an appreciable gap between Series 2 and 3 becomes virtually certain.
c7.22
∗ Todd thought that the Lower Gravel was laid down under pluvial conditions, but this is perhaps less likely than the view here suggested.
c7.23
In brief, the Upper Pravara and possibly the Khandivli groups are for the moment the only industries in India for which an organically intermediate position can be claimed with reasonable probability between the Acheulian-Levalloisian groups of the late Middle Pleistocene and the evolved microlithic 'spread' to which we must now turn.
c7.24
The small-stone (microlithic) industries of Europe, Africa, western Asia and India present, like the handaxes before them, a standard example of the problems of diffusion over a wide map. Their considerable degree of specialization and a vague but not entirely negligible approximation in their dating (as it seems) do indeed suggest some sort of cultural nexus in their distribution; but progress in the understanding of this remarkable possibility is dependent upon a closer analysis of the existing material and the acquisition of new. There is slight but accumulating evidence that Arabia formed a cultural, as indeed it forms a geographical, extension of the Sahara zone, and so may have constituted a central link between northern Africa and western Asia on the one hand and India on the other (cf. p. 58).
c7.25
There are many places in India, such as the banks of the Narbadā, where the most casual collector can readily fill his pockets with these small flakes and implements. Their material—often chalcedony, agate, or other semi-precious stones—catches the eye, and their minute size stimulates curiosity and acquisitiveness. Many of them are undoubtedly the remains of composite tools, of which wood or bone formed a part; for example, arrow-shafts with barbs of small sharp flakes such as lunates, inset and gummed near the point on opposite sides of the shaft. Most of them, on the other hand, can only be described as 'industrial waste'.
c7.26
Fig. 16
c7.27
The familiar lunates—crescent-shaped flakes from ⅜ inch to 1⅛ inches in length, with the natural edge of the flake as the cord and steep retouching to blunt the rear arc, to take the pressure of the thumb or to inset in wood or bone—are the most constant denominator of microlithic industries, whether in Africa or India. Splayed lunates may have been used as transverse arrowheads. Trapezoidal forms occur occasionally but are rare; and points which can fairly be likened to the numerous burins of north Africa and western Europe are very exceptional indeed, unless 'borers' are included. Triangles merge into angular lunates and scarcely deserve separate classification. Disc scrapers are relatively abundant but are not particularly characteristic of the microlithic complex. On the whole, the Indian microlithic groups make an indifferent showing beside those of Africa, Palestine and Europe. Attempts to subdivide them into categories such as 'non-geometric' and 'geometric' are a move in the right direction but are premature in the present state of the evidence. As elsewhere, a very large Fig. 14 Distribution of stone axes and microlithic industries proportion of the material consists merely of primary flakes, which in a few instances suggest use but are mostly waste.
c7.28
POLISHED POINTED-BUTT STONE AXES • MICROLITHIC INDUSTRIES ◦ SITES WITH BOTH • (MODERN CITIES ◦) MAIN DECCAN TRAP MILES 100 0 100 200 300 400
c7.29
Fig. 14
c7.30
On the map, the concentrations of microlithic sites are in central and southern India, with extensions up the west coast to the lower Indus. The Punjab and the northern plains are almost blank, though whether this is significant or rather due to inadequate exploration or to the covering of appropriate sites by aggradation cannot be affirmed. A similar doubt arises from the seeming absence of microliths from north-east India—East Bengal, Orissa and Assam. But the negation receives support from the fact that south-east Asia as a whole has failed, so far, to produce any microlithic industry truly comparable with that of India south of the Vindhyas.
c7.31
The dating of these Indian industries is at present full of doubt. Provisionally, a line of division may be drawn between (earlier) microlithic industries which are not clearly associated with pottery and (later) ones which are. True, the incidence of pottery is likely to be a variable and even accidental datum. In recent years a number of Late Stone Age cultures devoid of pottery have been identified sporadically in Asia: at Kila Gul Mohammed near Quetta in Baluchistan, at Jarmo in the foothills of northern Iraq east of the Tigris, at Jericho in Jordan, and farther west in Cyprus. At Jarmo radiocarbon tests give a date about 4700 B.C. or earlier. At Jericho similar tests indicate that the pre-pottery Neolithic was flourishing both before and after 6000 B.C. The Baluch example produced a radiocarbon date of about 3350 B.C. Broadly, it may be affirmed that there is no reason, as knowledge stands, to assume the existence of good kiln-baked pottery anywhere in the world much before 4000 B.C., or in India before 3000 B.C. Some at least of the pottery found with microliths in India is of the Bronze or even the Iron Age, i.e. well within the 1st millennium B.C. Indeed it is a fair generalization that, in the Deccan and probably also in peninsular India, the Iron Age, which began there not earlier than the 5th century B.C. and in some places appreciably later, is an alien structure founded directly upon a microlithic and chalcolithic substratum (see Chapter 7).
c7.32
Until recently, a classic instance of the variable relationship of pottery to microliths was supplied by Sankalia's well-known site in Gujarāt, at Langhnāj. There a considerable depth of dry wind-blown sand accumulated anciently, but, at an average depth of four feet from the present surface, the accumulation was interrupted by a faint layer of humus, indicating temporarily damper conditions. In the lower part of the sand above the humus, microliths were associated with potsherds of 'Neolithic' age or later. The humus land-surface itself and the sand below it were long thought to be completely devoid of pottery; but further examination of these layers in 1952 and 1954 is said to have produced 'very small quantities of extremely comminuted pieces of coarse, handmade and ill-fired pottery'.²² The few fragments found were exceedingly tiny, and, if they are truly *in situ* and have not merely percolated through the very flimsy humus-layer, it would be interesting to know whether we are here confronted with true pottery or rather with baked basket-linings such as have been noted in the Upper Kenya Capsian. The buried humus layer produced vast numbers of microliths (including a majority of primary flakes) of quartz, jasper and chert. With them were sandstone slabs flattened on one side and used for grinding; and mineralized human skeletons of 'modern' dolichocephalic type, found in a crouched position below the humus, are assumed to be related burials. The fauna included Indian rhinoceros, hog-deer, axis deer, nilgai antelope, black buck, bovines, mongoose, pig, horse, dog or wolf, tortoise and fish, but not, in the latest review of the material, any hint of domestication. Nor do the grinding-stones necessarily reflect agriculture, since they may equally have been used for wild grasses and spices. In any case, it seems likely that the economy of the settlement was dominantly that of the hunter. It should be added that, in the sand below the main occupation-level was found a part of the shoulder-blade of *Rhinoceros unicornis* bearing at least eight artificial pits on one side and suggesting use as an anvil in the fabrication of microliths.
c7.33
The general picture is that of a community squatting on the dry sandy soil during a period of moderate rainfall, and occupied in the manufacture of small blades of which many may be supposed to have been set in bone or wood as composite tools. The small pebbles available for this purpose may have been collected in the Sabarmati river, some eighteen miles away. The community lived mainly on the products of the chase, but they supplemented their meat with prepared grasses or herbs, which nature may have provided without artificial aid. Pottery in the accepted sense was either absent or, if sparsely present, was of the crudest sort.
c7.34
The stone industry from this land-surface included a great quantity of waste, due to the uncertain and irregular flaking of the material used. Fluted cores show that many of the blades broke off half-way down the core, but are small monuments of human persistency. Some cores were made from thick flakes, a practice noted elsewhere in central India. Of a hundred implements properly so-called (as distinct from mere flakes), the commonest is the lunate, backed along the arc either from one side or both. Irregular specimens, bulging nearer one end, may be described as points. These may have been used as arrowheads. For the rest, the series includes a few rather crude scrapers, and an oblique-ended point which might almost be described as a burin; but as a whole the industry is a poor one, with few specialized forms.
c7.35
No significant difference is noted between the microliths on the ancient land-surface and those found (in far smaller numbers) with potsherds in the overlying sand. Again, the vast majority of the specimens from this upper layer are no more than minute primary flakes. The associated sherds, though of far better quality than the scraps from the lower levels, have not been dated more closely than 'Neolithic-Iron Age'.
c7.36
Fig. 15 Microliths from Teri sites, Tinnevelly district. Nearly 1/2
c7.37
A similar vagueness often attaches to sherds found elsewhere with microlithic industries, and the only value which can be derived from the association is the implication that no very remote age is ascribable to the industries in question: be it repeated that most of them are probably not earlier than the 1st millennium B.C. True, in East Africa crude pottery, or something very like it (above, p. 73), occurs in the Upper Kenya Capsian which is basically Upper Palaeolithic but includes an appreciable microlithic element, particularly lunates from ½ to 1 inch in length. This industry has not yet been closely dated, though a North African analogy of about 6500-5000 B.C. has already been cited (p. 69). The Capsian is followed in Africa south of the Sahara by more whole-heartedly microlithic groups, of which the earlier Wilton has been compared with the latest of the four Cammiade-Burkitt series from South India—a profuse surface-mélange of pigmy tools including lunates, backed knife-blades, triangles, and core-scrapers. Similar microlithic assemblages, equally undated have been collected widely in the Mahi and Godāvarī valleys, the Vindhya Hills and elsewhere in the centre and the south, but only in two instances have they, at the time of writing, been associated with an informative stratigraphy devoid of pottery.
c7.38
The first is at Rangpur in Kāthiāwād where, beneath material related to the Indus Civilization, was a layer containing crude microliths of jasper and agate without pottery. The layer has not yet been adequately investigated, and all that can be said is that on this site microliths were in use not later than the latter half of the 2nd millennium B.C. The second instance is in the Tinnevelly District of the southern part of Madras State. Here the coastal plain is characterized by extensive sand-dunes or *teris*, whence the sites associated with them are known as Ṭeri sites. The sands are connected with fossil beaches of calcareous sands largely of organic origin. Farther inland several morphological steps representing successive lagoon-floors of a greater age have been observed. On these various steps, A, B and C horizons have formed, but in many places the loose A horizon has been blown away, so that on the surface of the cemented B horizon is now encountered a concentration of industries originally deposited on or in A. In the aggregate, the industries must extend back to a very appreciable antiquity. They have been studied by Professor Zeuner and Mrs. Allchin.³⁵
c7.39
Fig. 16 Microliths from Yerangal, Bombay
c7.40
These stratified Ṭeri industries contain lunates, transverse arrowheads, bifacial points showing pressure-flaking, backed blades, unifacial and bifacial discoids, a microburin, end- and side-scrapers, and an occasional pebble chopping-tool of greater size. The tools are of quartz and variously coloured cherts. On certain sites, association with black-and-red sherds of urnfield or megalith type and of the last three or four centuries B.C. is ascribed to the blowing away of the A horizon and the consequent re-deposition of the sherds at the lower level. There is little doubt that, as a whole, the Tinnevelly Ṭeri industry represents a pre-pottery or at least a non-pottery phase. The pressure-flaked bifacial implements are elsewhere absent from India but recur in Ceylon.
c7.41
We seem to have here the vestiges of a hunting or fishing people settled or temporarily encamped in the vicinity of an old coast-line 20-30 feet higher than today, sometime before the advent of polished stone axes (rare in any event so far south) and the pottery-using Chalcolithic. The phrase 'Mesolithic Period' is still too nebulous in India for definitive use, but it would appear likely that, under the impact possibly of microlithic elements received via Arabia (p. 70), in the Peninsula as in Africa a microlithic trend in the Upper Palaeolithic was subsequently so emphasized as to represent a distinctive stage in the Stone Age sequence. The general absence of a true flint or obsidian deprived the Indian tool-maker of the apt materials which encouraged his opposite numbers in Africa and Europe to perfect the fabrication of these minute implements, and, with occasional exceptions, it must be confessed that the Indian
c7.42
Fig. 17 Microliths from Brahmagiri. ⅟
c7.43
micro-industries rarely rise above the second-rate. K. R. U. Todd, however, who had an extensive knowledge of Indian microliths particularly in the Bombay area,³¹ came to the interesting conclusion that 'the [west] coastal microliths as a whole appear, from their form, patination and state of preservation, to be definitely of earlier date than the pygmies from the inland sites at Jubbulpore and the Vindhya Hills. Further, they include many forms not recorded in the inland sites, which seem to consist of lunates, blades and cores only.* If this observation is confirmed—and it is consistent with the scarcity of microliths farther east though a crude non-pottery industry has recently been recognized near Durgapur in West Bengal—it may be interpreted as a further pointer towards a western, and therefore south-west Asian or north-east African, source for the Indian series, and will once more illustrate our 'Indian Ocean culture-pool' (see above, p. 58).
c7.44
³¹ [Text of the note is not present on this page]
c7.45
Certainly at the end of the story, in and before the 3rd century B.C., the technical level was a low one. The pre-Iron Age culture at the inland site of Brahmagiri, which was of this period,** was fairly extensively sampled and produced only a crude flake-industry with a microlithic bias—small backed-knives, serrated blades, one crested flake, a beaked burin (?), side-scrapers, and an inferior lunate. These poor implements were associated with polished stone axes and occasional copper and bronze, and would seem to represent the microlithic tradition at its last gasp.
c7.46
The relationship of the Chalcolithic and iron-using cultures of central India with microlithic industries will be further considered in Chapter 7.
c7.47
At present the term 'Neolithic' or 'New Stone Age' does not make much sense in India. Unless some of the microlithic cultures be so classified, no integral Neolithic culture has yet been isolated there. Blade-industries with a microlithic trend existed, as we have already seen, both before and after the introduction of metals. It is less certain that polished stone implements preceded some slight knowledge of copper or bronze. Before turning to them, however, something more must be said of the flakes and blades which culminated in the Chalcolithic cultures of northern and central India.
c7.48
Both at Sukkur and at Rohri, situated on opposite banks of the lower Indus in Sind, have been found core and flake industries showing a typological affinity with the Chalcolithic series but associated neither with pottery nor with metal. This might be explained by the evident fact that the sites were primarily workshops. On the other hand, there is also some geological reason for supposing that they antedate the Indus Chalcolithic, in that the hill-top soil associated with them disappears on the slopes under a covering of silt left by an Indus flooding at a considerably higher level than today and *a fortiori* higher than in Indus Civilization times. (In the early days of Mohenjo-daro the flood-plain of the lower Indus was at least twenty feet lower than today.) This has suggested but does not prove that the Sukkur-Rohri industries may be earlier than, and ancestral to, the Chalcolithic cores and flakes of the Indus valley, though not necessarily at any very great interval of time.
c7.49
Their material is derived from the Eocene limestone of the locality and is as true a flint as can be found in the sub-continent. The cores, flakes and implements occur on hill-tops, either on or close to the surface, and include both unpatinated examples and others showing degrees of patination which have been thought, without emphasis, to indicate a time-scale. Amongst much waste, the Sukkur series comprises broad and narrow blades, all somewhat crude and thick, with rare retouching down one or both sides and sometimes with a natural ‘back’ or cortex; flakes either with lengthwise or with convergent primary flaking; a few ‘plane’ scrapers with steep trimming; a few conical cores showing long, narrow, shallow flake-scars, like the cores typical of the Indus Civilization; and occasional axe-like cores bearing a superficial resemblance to Acheulian handaxes but differing from them in technique as in date. The Rohri sites produce large quantities of blade cores, mostly conical, and a multitude of crude and irregular blades which may be regarded as 'export rejects'. It is sufficiently evident from the careful preparation shown by some of the cores that blades of good quality were also made, comparable more nearly with the better examples from the Civilization. These better blades were presumably dispersed by trade.
c7.50
The great Civilization which flourished in the Indus valley and southwards to the Gulf of Cambay at the end of the 3rd millennium will be discussed in the next chapter, but here it may be observed that the common knife with which this Civilization widely supplemented its bronze tools was a long, parallel-sided blade of chert. Similar blades are found in some of the Baluch hill-communities which existed before or alongside the Civilization. The blades were rarely retouched; Colonel D. H. Gordon has noted that, of 1,758 examined at the two principal Indus sites, only 104 showed any secondary working; a few of these had been worked down both sides to form a point, some are backed, and some are notched at the base for binding to a handle. The peculiarly brilliant gloss produced on the edge by the cutting of corn has occasionally been detected. Some of the nuclei or cores were also polished and had probably been used as burnishers on metal. No characteristic microlithic forms are present. On the other hand, there are a few large (sometimes huge) square-sectioned 'axes' which may have been either hoes or tree-cutters.
c7.51
Parallel-sided chert blades of normal Indus type have been found as far south as southern Kāthiāwād, at the Indus sites of Lothal and Rangpur. It may be therefore that there is a cultural kinship between these blades and those which have in recent years been recovered from a considerable series of Chalcolithic sites in central India (northern and central Deccan): Tripuri, Nagda, Maheshwar, Nāvdā Toli, Prakāsh, Bahāl, Nevāsā, Jorwe, Nāsik, and Maski. At these sites the Chalcolithic phase, though not closely dated, is clearly post-Indus, and 1200–500 or 400 B.C. is probably a fair guide.* Nor, in spite of the occasional affinity of their flake-industry with that of the north, can it be described as significantly ‘Indus’; it has, as already pointed out, a strongly microlithic bias (doubtless emphasized by the limitations of the raw material available) and includes lunates and occasionally other microlithic forms, although the published ascription of the overall term ‘microlithic’ to an industry such as that at Maski, where chert blades up to 5½ inches long are included,28 is not to be commended. Further reference will be made in Chapter 7 to these central Indian cultures; meanwhile it may be suspected that the Maski blades, like those of Kāthiāwād, represent an intrusion from the blade-industries of north-western India into the essentially microlithic cultures of the Deccan.
c7.52
Finally, something must be said about the ground or polished stone axes and adzes which are amongst the most characteristic stone artifacts of India. In some regions of the Deccan or the Peninsula they can be picked up almost casually on a country walk. The sequel normally is that their find-spot is quickly forgotten, and the museums of India and Great Britain contain innumerable unidentified examples. Something like eighty localities can, however, be put on the map, and it is clear that, apart from a thin scatter in the north-west (sometimes out of context), the distribution lies south-east of a line drawn from Bombay to Kanpur (Cawnpore) on the middle Ganges and extends at least as far south as the Kāverī river. In other words, the axes belong essentially to eastern and southern India, but are absent from the extreme south and from Ceylon.
c7.53
On the basis of form they have been tentatively divided into three groups, each subdivided into four sub-groups. Within the narrow limits of the present summary I do not propose to discuss the individual types or possible variations in their distribution. This has been attempted in part by Subbarao, Worman and Allchin,²⁸ but except in the Bellary region little exact information is forthcoming. Here it will suffice to differentiate the two underlying techniques. Technique 1 consists of (a) the rough hammer-shaping of the stone—commonly dioritic trap or basalt—followed by (b) the further reduction of unwanted excrescences by means of a 'fabricator' (rather like a stone finger) and sometimes by pecking, and (c) a final grinding and polishing of the cutting edge and often over half or less than half of the surface of the tool. The resulting transverse section is ovoid, trapezoidal or sub-triangular. Technique 2 consists of (a) the removal of flakes from faces and sides of the core so that an implement with almost flat faces and rectangular section is produced, and (b) in general the smooth polishing, with occasional pecking, of the whole or most of the surfaces. The commonest type in Technique 1 is an axe or, rarely, adze with a bluntly pointed (sometimes rounded) butt; the most distinctive type in Technique 2 is the tanged or shouldered adze or hoe.
c7.54
∗ Since this was written, Dr. Sankalia has informed me that a C14 analysis of burnt wheat from a late chalcolithic stratum of Nāvdā Tolī by the University of Pennsylvania has given the date 1336 B.C. ± 125.
c7.55
Fig. 14
c7.56
The pointed-butt axes were first put into a stratigraphical context in 1947 at Brahmagiri in northern Mysore, thirty miles south of Bellary. There, on an ancient town-site, the latest of a clear sequence of three main cultures was dated by its rouletted and russet-coated painted ("Andhra") wares (p. 31) to the 1st century A.D. Below and overlapping it was an elaborately iron-using 'megalith' culture (p. 162) which is unlikely to have arrived there much before the 3rd century B.C. In turn, overlapping and underlying this megalith culture was a Chalcolithic culture with two subdivisions, marked by abundant pointed-butt axes, crude microliths, rough pottery and rare scraps of copper and bronze. The overlaps in each case imply a partial contemporaneity between the new culture and the old, but equally in each case there was no organic cultural transition. We are dealing with three essentially distinct cultures of which the lowest is characterized by the axes and ended, as a cultural substratum, soon after 200 B.C. In other words, our stone axes with their copper-bronze accompaniment endured in the southern Deccan for something like three centuries after the use of iron had been established farther north. The contrast is not remarkable; there are much later analogies for an emphatically differential development in India.
c7.57
The Brahmagiri axe culture or cultures extended downwards through 8 or 9 vertical feet of occupation-material—not mere dump, much of it, but the legitimate accumulation of floors, hearths and structures over a very appreciable period. What this means in terms of time is anyone's guess. Five centuries are probably not a fantastic over-estimate, and about 700 B.C. is offered as a provisional and schematic date for the beginning of this particular settlement.
c7.58
The Brahmagiri evidence is not alone. It is matched by that obtained by Dr. Subbarao at Sanganakallu, close to Bellary, where there was an axe-culture accumulation 4½ feet deep (a considerable depth on this hill-top site) before the megalithic culture arrived upon the scene. Again the axes were associated with microlithic flakes, and again an overlap was observed between the two main cultures.
c7.59
Other evidence supports the active survival of pointed-butt or comparable axes into the latter half of the 1st millennium B.C.: for example, at Sonepur in the Gaya district of Bihār they occurred with good Northern Black Polished Ware (p. 30) and iron in a stratified series.⁴⁰ But the occasional presence of an isolated stone axe (for example, at Taxila) in a late occupation is not necessarily significant chronologically or geographically, since these distinctive objects are liable to have been picked up as curiosities or even used for cult-purposes at any period down to the present day. They can still be seen in South India as offerings or symbols at village shrines. Only when they are present in quantity or are associated with factory-flakes or with other objects of the same cultural stage can they fairly be regarded as utile tools. Whilst they may well have survived in this capacity in backwaters of the Deccan or the Peninsula to an appreciably later date, there is at present no clear evidence of their fabrication after the end of the 3rd century B.C. or thereabouts.
c7.60
At the other limit of the time-scale, there is no proof that these axes preceded the working of copper or bronze. True, a negative of this kind is difficult to establish since, at the best, metal was scarce amongst the axe-using communities. But, as we have seen, it was known to the axe-culture of Brahmagiri. Similarly, in the lowest post-palaeolithic phase* at Nevāsā in the Upper Godāvarī-Pravara basin polished pointed-butt axes and hammer-stones were accompanied by two chisels, a needle, a spear and four beads, all of copper or bronze, together with the microlithic (chalcedony) industry of lunates, triangles and backed blades which we are learning to associate with the axes.⁴¹ The pottery included distinctive spouted vessels with flaring rims, painted with geometrical and other designs in black on a fine re-slipped surface and comparable with the wares from other chalcolithic sites (Nasik, Jorwe, Nāvdā Tolī) in the northern Deccan. The absolute chronology of these sites has not yet been determined; but on some of them the Chalcolithic was succeeded by an Iron Age culture of the latter half of the 1st millennium B.C., and the general context is consistent with the evidence from Brahmagiri. Short of proof, it is likely that most of the polished pointed-butt and similar stone axes of India were made between 1000 and 200 B.C., with a preference for the latter half of that period.
c7.61
⁴¹ * Separated from the palaeolithic phases by deposits of clay and silt.
c7.62
Before the origin of the Indian axes is considered, something must be said of a more individual type of implement which CHITOR KAUSAMBI Ganges R. Godavari Yellow River Yangtze-Kiang HONGKONG 0 500 1000 Miles]
c7.63
STONE SHOULDERED HOES [after Dani
c7.64
Fig. 18 Distribution of stone shouldered hoes
c7.65
partially coincides with them in distribution. The tanged and shouldered hoe or adze, with angular or sloping shoulders, often carefully polished and with a single-chamfered edge in one of the broad planes, is a well-known though not a common type in certain parts of eastern India, as far south as the lower Godāvarī. The material is commonly chert or schist. In Assam the form occurs freely except in the Sadiya frontier area; but examples from the Gāro and Khāsi Hills are rough and irregular and suggest provincial imitations. Though rare in Bengal (an example has recently been found in Midnapur district) they reappear near the Bengal-South Bihār border in the Santal Parganas, Mānbhūm and Dhālbūm, and farther south in Mayurbhanj and elsewhere in Orissa. To the west, stray specimens have been found in the Banda district of southern U.P. and within the fortifications of Kaushāmbī (p. 128), and one is alleged to have come from as far afield as Chitor in Rājasthān. Generally, however, the eastern bias of the type is clear enough.
c7.66
Fig. 18
c7.67
Beyond India, Burma has produced good examples, and the type, highly finished or in the rough, is widely characteristic of south-east Asia. It is recorded from Malaya; in Indo-China it is included both in the so-called Bac-Sonian culture north of Hanoi and in the more southerly Somrong Sen culture of Cambodia and Annam. In no instance has an example been found by a competent excavator, but there seems little doubt that the users of the shouldered hoe were also potters and occasionally employed bronze or copper.⁴² In Tongking, allegedly associated remains were of the Han period (202 B.C.–A.D. 220). Farther north usually crude shouldered hoes occur sporadically over a wide stretch of China from Hong Kong to Hunnan and the Yangtse valley of Szechwan.⁴³ More doubtfully, they are reported from Manchuria and Japan. The type has been related, very speculatively, to bronze shouldered axes of the Yin dynasty (about 1300–1028 B.C.) in An-yang, south-west of Tientsin.⁴⁴ Certainly the more highly finished rectangular examples suggest a metal prototype, and it has been suspected that the stone was in some instances cut with a metal wire, a technique which still obtains in India.⁴⁵ The evidence as a whole points to a Chinese parentage for the shouldered hoe, with an origin perhaps in the earlier half of the 1st millennium B.C. and a southerly trend through the Chinese lowlands to Laos and Burma. When it arrived in India is at present unknown; no example has there been found in any significant association, although the Tongking evidence, if verified, might favour a date in the latter half of the millennium, i.e. somewhat later than the first of the polished axes. To link its arrival with that of Austroasiatic languages from south-eastern Asia, as has been attempted, is mere guesswork.
c7.68
To return now to the polished axes of India is in some measure to retrace the same ground. It can be stated without argument that these essentially east-Indian types were not derived, at any rate directly, from western Asia.¹⁶ On the other hand, they occur in Burma, although casual collecting has made it impossible to plot find-spots adequately on the map. More vaguely still, Laos has produced both axes and shouldered adzes. In Malaya pecked and partially smoothed axes, comparable with those from India though more round-butted, are a recognized type. In Tongking they were used by communities which were still, it seems, largely in the hunting stage, without systematic food-production, and a technical sequence with an increasing use of stone-polishing has been postulated though not proved. Farther north, they occur in Honan and Shantung, and it has been observed that some of the North Chinese examples of pecked axes correspond even more closely with certain of those of India than do the examples from Indo-China. The patchy state of present knowledge, and not least the absence of clear dating, prevents any close correlation of the Chinese and the Indian evidence; but, geographically, routes up the Yangtze river (for example), past Szechwan in the direction of Hunnan and the Burma passes seem possible for the axes, though scarcely for the shouldered hoes, which, unlike the axes, are absent from Yunnan. Alternatively, coastal routes to Indo-China are feasible lines of access to Burma, Assam and Bengal. At least the evidence is ample enough to suggest, if less forcibly than in the case of the shouldered hoes, an eastern origin for our Indian series, with a bias in favour of central China. Even so, the major exploitation of the round-butted or, more characteristically, point-butted axe occurred on Indian soil.
c7.69
Fig. 14
c7.70
There remains the most important question of all: what, in terms of living man, man upon the landscape, is the implication of the whole of this stone-axe phase? Of the skeletal types of Brahmagiri and elsewhere nothing can be said until the loitering Anthropological Department of India produces its long-overdue reports. Meanwhile, we know not the shape of any one of these wielders of the stone axe. But what can be inferred of their way of life?
c7.71
First, it can be affirmed that the great accumulation at Brahmagiri reflects an essentially static, food-producing population, which had lived in timber huts, some at least rectilinear on plan and reinforced by basic dry-stone walls. The surrounding countryside under natural conditions would be jungle, but their abundant stone axes would enable the villagers to make clearings for farming, probably with the assistance of fire.
c7.72
In this context, Dr. C. von Fürer-Haimendorf has suggested the living analogy of the Reddis, who maintain today an atavistic mode of life in the hills flanking the Godāvari gorge of the Deccan. They stand midway between semi-nomadic food-gatherers and settled peasantry, occupying hamlets of no great substance, and they supplement wild plants and tubers by raising crops and breeding animals. ‘Their agricultural methods are extremely primitive: they clear and burn the jungle and then sow sorghum, small millets and pulses in the ashes, partly by broadcasting and partly by dibbling with the help of a digging-stick. No hoe is used. . . . Neolithic man in possession of a polished celt may well have practised a similar form of agriculture.’⁴ If anything, the Brahmagiri axe-people would appear to have been more advanced and static (or less decadent and harassed) than the Reddis; it may be suspected that they had already learned to supplement ‘slash-and-burn’ agriculture with some knowledge of crop-rotation. But the general picture of a limited agriculture based on jungle-clearance within the range of nuclear settlements is a convincing one, and gives a needed actuality to the somewhat tedious vestiges now recoverable from the soil. To this it may be added that a number of ‘cinder-mounds’ situated mostly between Bellary and Hospet have been identified by Professor Zeuner as ancient accumulations of cattle-dung, and in and about one of them I collected a dozen or more pointed-butt axes in the space of half an hour. In 1938 Sir Leonard Woolley had a similar experience; at the Kupgal near Bellary he ‘picked up, on and around a cinder-mound, two or three celts of basalt trap’; and as long ago as 1872 R. Bruce Foot found a ‘celt’ in a cinder-mound between Bellary and Gadag. The suggestion is that of a pastoral community in our axe period. We may fairly regard our axe-people as herdsmen anchored by a rudimentary agriculture.
c7.73
In summary, about 1000 B.C. or a little later, groups of simple farmers, not perhaps originally very numerous, infiltrated into the hill-jungles of north-eastern and eastern India, coming from the direction of Burma if not beyond. They were equipped with stone industries of Neolithic types familiar archaeologically as far afield as northern China; there already, however, the axe-makers may have had some knowledge of bronze, which is recorded historically to have been in use in the North well before 1000 B.C. In south-east Asia, as far as Malaya, the same stone equipment is liable to appear with or without associated metal, which may not be earlier there than the 4th or 3rd century B.C. Meanwhile stone axes, followed, as it seems, by the shouldered hoes, found their way into India via Burma; but to associate their arrival either with Austro-asiatic languages or with megaliths is to exceed the evidence. None of these stone implements has yet in India been associated stratigraphically with a megalith.
c7.74
In central India it would seem that the stone axes were superimposed upon, or commingled with, the microlithic flake industries which were already rooted in the centre and south of the subcontinent. In central India at least it is unlikely that, to any major extent, the microlith-folk were still altogether ignorant of agriculture and stock-keeping, which had been well understood in the north and north-west for many centuries. To them, in the forested valleys, the new jungle-clearing axes may have been economically a useful accession. Whether from remote China the stone axe or adze was accompanied by a memory of copper and bronze is more problematical. This knowledge may rather have been contributed afresh by the Chalcolithic cultures of India. So it may be that, after all, the bringers of the stone axes were in a truly Neolithic stage of culture when they first entered the subcontinent. If so, here at last is a momentary Indian Neolithic of a conventional kind. But archaeology has not yet confirmed this. Systematic evidence from Burma or Assam would be crucial.
c7.75
CHAPTER V
c7.76
The Indus Civilization
c7.77
THE WORD 'CHALCOLITHIC' occurred in the last chapter and will recur in its successors. It is a term which has been widely and not always very discriminately used by writers on Indian archaeology; applied alternatively to essentially stone-using communities to which copper or bronze are rare luxuries, or, less aptly, to copper- and bronze-using communities which retain a substantial though subordinate stone equipment. It may be accepted as an ugly utility-term; an alternative label, 'Protometallic', is scarcely more elegant or exact.
c7.78
Of the Chalcolithic phase in the Baluch hills and the Indus valley much has been written in accessible form,¹ and the barest summary must here suffice. In general terms, the situation is as follows, with the reminder that all dates prior to the latter half of the 3rd millennium (and many after it) are very insecure.
c7.79
In the 4th and 3rd millennia, if not before, the sullen borderland where the Iranian plateau drops tumultuously to the Indus plain sheltered a hive of hidden tribal and village societies, appropriately cellular but sharing a common standard of living and technology. In their recondite valleys they were diversely reaching the cultural optimum which their rugged milieu permitted. Their pottery, modified from region to region and age to age, was by 3000 B.C. of good quality, wheel-turned and well-baked, often thin and attractively painted with geometrical or semi-realistic motifs which betrayed both their own individuality and their cultural kinship with the societies of the great plateau behind them. They made stone implements amongst which simple chert blades predominated, and a little copper or bronze was beginning to come their way for use or ornament. Occasional contact between one community and another and with the outside world is manifest, but for the most part the tribes or regions, though not necessarily devoid of a nomadic element, were self-contained on the basis of the local crops which they learned to irrigate by the concentration of flood-water (with or without masonry dams),¹⁸ and the small flocks and herds which the children tended on the neighbouring slopes. Further integration, whether political or cultural, was barred by environment.
c7.80
Just so far, it may be claimed that the potential of these upland groups was not unlike that of early Greece, with its small hill-divided and independent states. There, however, the encompassing sea provided a sufficient measure of mutual contact to create, if not unity, at least an inter-territorial sentiment of Greek civilization and nationhood as against the outer 'barbarians' of alien speech and tradition. In the massif of Baluchistan, save at its southernmost end, this vitalizing link was missing. Instead, however, there lay along its flank the great plain of the Indus and its tributary or adjacent rivers, a plain at that time largely jungle-covered but serving, with its broad waterways, in some sense as an inland sea and as a passable means of lateral communication. Those two factors—the terminal ocean and the flanking plain—were together formative in the sudden emergence of the next cultural phase, that of the Indus Civilization, probably a little before 2500 B.C.
First 79 of 141 passages. Read on