Early India and Pakistan to Ashoka c8: Chapter VI. The Ganges Civilization

Early India and Pakistan to Ashoka c8

Language: English

Chapter VI. The Ganges Civilization

c8.2

FROM THE INDUS BASIN with its spreading coastline we turn eastwards to the doāb or two-river country of the holy waters of Hindustan, the Yamunā or Jumna and the Gaṅgā or Ganges. The actual distance from the Sutlej, of the Indus system, to the Jumna at the foot of the Simla hills is less than eighty miles, but anciently the two valleys were a world apart. Here was a true partition between the submontane Punjab and the Great Plains of northern India. Hereabouts only a little more than 100 miles intervenes between the vast Thar or Indian Desert, which stretches towards the Arabian Sea, and those same Himālayan hills; and it was through this jungle-filled corridor that the invader from the mountains of the north-west had to force his way. The corridor was further narrowed, indeed almost closed, fifty-five miles north of Delhi, where the upper waters of the (former) river Ghaggar approached to within a narrow margin of the Jumna. Squarely in the midst of this shallow divide stands the town of Pānīpat which has on at least three occasions been the scene of a decisive battle.¹ Today the mound which rises in the midst of the little town is sixty feet high, and its crumbling sides exude the débris of close upon 3000 years.²

c8.3

The archaeology of the Ganges valley is in its infancy, and correlation between its material cultures on the one hand and linguistic or literary factors on the other may wisely be deferred. The composite individuality which, from an early date, distinguished this region from the Punjab on the west and Bengal on the east requires much further evidence and analysis for its proper understanding. No part of India has been changed more drastically by the encroachment of tillage on jungle, and the first act of the historical imagination must be to replant 'the dark and pathless forest', the mahāvana, in which Rāma and the Pāṇḍavas had their being. In this close and tangled country the intruding Aryans at the end of the 2nd millennium B.C. may have found little previous occupation of a settled kind; such at least is the implication of the few sites which have hitherto been probed to the subsoil. The invaders themselves, it may be suspected, here quickly dropped their dominantly pastoral economy in favour of one based on agriculture in fertile riverside clearings. Here the territorial magnate began at no great interval to supersede the traditional tribal chieftain and to build the nations and kingdoms of the Indian epics, of which this is the homeland. Here, in the Middle Country or *Madhyadesha*, developed the 'Inner' category of Indo-Aryan speech, as distinct from the 'Outer' category in the Punjab, Bengal and Assam. Here emerged the essential India, with its imperial dynasties and its great religious teachers. At present, all this achievement is strangely lacking in palpable roots.

c8.4

Fig. 26 From 'Gangetic' copper hoards: 1, 'anthropomorph' from Sheorajpur; 2-3, antennae swords from Fatehgarh; 4-5, harpoons from Sarthauli and Bisauli; 6, ring from Pandi; 7, hooked spear from Sarthauli; 8, axe from Sarthauli; 9, axe from Gungeria; 10, axe from Dunria; 11, double axe from Bhagra Pir; 12-13, bar-celts from Gungeria

c8.5

2 1 0 2 4 6 8 10 FEET

c8.6

WEST

c8.7

EAST

c8.8

PERIOD V

c8.9

PERIOD IV

c8.10

BURNT EARTH

c8.11

WALL 16

c8.12

IRON SLAGS

c8.13

MUD WALL NO. 10

c8.14

PERIOD III NORTHERN BLACK POLISHED WARE

c8.15

PIT C

c8.16

PIT B

c8.17

MUD WALL NO. 19A

c8.18

MUD WALL NO. 31

c8.19

PERIOD II PAINTED GREY WARE

c8.20

PIT E

c8.21

PERIOD I 'OCHRE-COLOURED' WARE

c8.22

NATURAL CLAY

c8.23

SAND

c8.24

Section through mound at Hastināpura

c8.25

Fig. 28 Distribution of Gangetic copper hoards and hoards of copper flat axes

c8.26

Most detached of the available evidence is the famous series of copper hoards which have been found on upwards of thirty-four sites between the upper Ganges and Orissa.⁴⁸ Something like six hundred objects is involved, comprising eight main types: (i) flat axes, usually of stocky sub-rectangular form with splayed edge; (ii) shouldered axes with a clear kink or set-back at the points where the curved edge meets the stem of the implement; (iii) 'bar-celts' or chisels, up to two feet long, consisting of a nearly parallel-sided bar and an expanded chisel-edge in one of the lateral planes, not central as in the axes; (iv) rings made by bending a rod of circular section until the ends meet; (v) harpoon-heads strangely like the Magdalenian or Azilian horn harpoons of the West, with bilateral barbs and a loop or projection for the attachment of a cord; (vi) spearheads (sometimes called swords) with strong mid-rib and often with a projection from the tang for the attachment of a cord; (vii) swords, usually with the hilt, of one piece with the blade, bifurcated like antennae; and (viii) strange anthropomorphic objects, up to eighteen inches long, which might be taken to represent a human form with straddled legs and incurved arms. Their purpose and whether indeed their anthropomorphism is intentional are alike unknown.

c8.27

⁴⁸ [Note text not present on page]

c8.28

Figs. 26 and 28

c8.29

Analysis shows the presence of copper and a small amount of nickel and arsenic, in proportions consistent with an Indian origin for the ore. A few of the implements are of bronze; one of the antennae-swords from Kallūr, in Raichūr district in the Deccan, contains 9.5 per cent of tin,* i.e. is of an alloy suited by strength, elasticity and toughness for its martial purpose. The nearest ancient copper-workings to the Ganges basin are in Rājasthān and in Singhbūm, southern Bihar; of these alternatives the latter, in view of the distribution of the hoards, is the likely source.

c8.30

In no instance has a hoard or distinctive hoard-type been found in a stratified deposit, and even the precise find-spot of most of the hoards is forgotten. At Rājpur Parsu, in the Bijnor district beside the upper Ganges, a hoard came from the periphery of a mound 'which is 5-7 feet high and covers an area of nearly 4 square furlongs'. Another hoard, from Bisauli in the Badaun district a little farther south, was ploughed up in a field that is 'almost flat without any signs of a mound'. At both these sites Mr. B. B. Lal dug exploratory trenches; at neither did he find further implements, but both sites yielded scraps of ill-fired thick ochre-washed ware, mixed at Bisauli with sherds of well-fired red-slipped ware with black designs. All the Bisauli pottery was near the surface—as indeed the hoard had been—but the groups of sherds were distinguished by the freshness of the red ware and the worn aspect of the ochre ware. Yet another hoard-site, at Bahādarabād eight miles west of Hardwār on the upper Ganges, has also produced ochre sherds. A comparable ochre fabric has been found by Mr. A. Ghosh farther west, in the Drishadvati valley of Bikaner, and has there been called, a little ambitiously perhaps, the 'Sothi Culture', from a place near the sub-divisional town of Nohar. More significantly, it characterized the lowest (earliest) occupation of the Ganges site at Hastināpura, to which we shall come in a moment. Little is at present known of this Ochre Ware; even the shapes of the pots have not been determined; and its association with the copper hoards is not proved. But a considerably more advanced exploration of the succeeding Painted Grey Ware levels in the Ganges-Jumna basin has failed to reveal in them any of the more distinctive hoard-types, and the impression for the moment is that these precede the Grey Ware and the full development of urban life in the region. In a chronology which is at present largely guesswork, I should say that that implies a date for the hoards before the 8th century B.C.; my Indian colleagues might prefer a somewhat earlier terminus.

c8.31

∗ Analysis made in 1945 by the Archaeological Chemist in India.

c8.32

What do the hoards themselves tell us? The flat axes with more or less expanded blades are the only possible link with the Indus Civilization, and the type is too generalized and widespread to support the view that the hoards represent 'the colonization of the Ganges basin by refugees and displaced persons from the Punjab and the Indus valley during the time of the break-up of the Harappā empire and the coming of raiders from the west'. The only important parallelism between the copper hoards and the Indus bronzes is the absence from both of the hafting-socket, which had been familiar in Mesopotamia and farther north since the 4th millennium. But the really distinctive elements of the hoards—the barbed harpoons, the spearheads with basal projection, the 'bar-celts', the so-called anthropomorphs, the shouldered axes—are not Indus types; and *per contra* the distinctively Indus blade with curved end is absent from the copper hoards. Analogies to the swords with antennae have been recognized in the Koban culture of the Caucasus and may be significant,44 but intermediate links are at present missing. In all the circumstances, the alternative theory that the hoards may be 'in fact traces of the Indo-Aryan migration' is equally difficult to sustain. Indeed, archaeologically our Ganges evidence is not ready for equation with any major historical event or speculation of the kind, and the Aryans in particular should, I suggest, be given a holiday for the time being. It is best to admit that no proto-Aryan material culture has yet been identified in India.

c8.33

Nevertheless, these hoards are not completely dumb. The axes, up to a foot in length and five or six pounds in weight, are excellent woodmen's tools. The barbed harpoons, based perhaps on bone or horn prototypes, proclaim extensive food-gathering in the fishful rivers near which they are found. That they were also used for hunting animals as formidable as the rhinoceros is shown, as B. B. Lal reminds us,¹⁴ by cave paintings of unknown date in the Mirzapur district of the Ganges valley, south-west of Banaras. The unbarbed spear-heads anticipate in general fashion the iron spears which the peaceful peasantry sometimes carry in these regions at the present day. Only the swords which occur on four of the sites imply a more military element, symbolic perhaps of rank rather than recurrent peril. The general inference is that the hoards represent semi-nomadic food-gathering communities, capable of clearing patches of jungle and perhaps (though this we do not know) of some sort of garden-agriculture, but living mainly by hunting and fishing. The likelihood fits well into far more modern pictures of tribal India.

c8.34

¹⁴ [Note text not present on page]

c8.35

It has been inferred that the deposition of the hoards 'itself suggests a time of insecurity and economic instability, and and may mean that the refugees from the Indus Civilization were not left undisturbed for long, as the Aryan invasions gathered momentum and pressed on, beyond the old frontiers of the Harappā kingdom and down into the Ganges valley'. This is to force the evidence beyond its warrant. The copper implements are shown, by their frequently specialized character and skilful casting and hammering, and by their distribution over 800 miles of jungle landscape, to have been the work of substantially whole-time experts who were probably also (as in other parts of the world) itinerant. The perils of one kind and another which must have beset them in their vagrancy and led to the occasional loss of their stock-in-trade were not necessarily more formidable than those which man and beast would normally impose upon the wandering craftsman and tradesman in the rough circumstances of the age. Once more it is unnecessary or premature to conjure up cosmic causes such as Aryan invasion to explain these sporadic casualties.

c8.36

As the evidence goes, then, we are confronted by a striking and individual development of metal-craftsmanship in the Ganges basin and the Orissa hills somewhere about 1000 B.C. amongst folk who were primarily hunters and probably lived in temporary villages in jungle-clearings. (Incidentally, none of the ochre sherds referred to above as possibly contemporary with the hoards has been found in association with built foundations.) Insufficiently large and wealthy to maintain specialists individually, these communities shared, it seems, in a nomadic craftsmanship for which there are many parallels. The sources from which this craftsmanship drew its initial inspiration are not clearly defined; it may be supposed that its ideas germinated in the metal traditions of the north-west (the Indus region and beyond), though it would not be surprising to discover trans-Himalayan sources in the north and north-east. This uncertainty, whilst emphasizing our present ignorance, also points the likelihood that much of the kind and quality of the hoards was due to the inventive skill of the Ganges coppersmiths themselves. Archaeology, in its search for ultimate origins, can readily overlook the possibility that those origins may lie very largely at our feet.

c8.37

From these *disjecta* we turn to the great urban sites which came into being in the Ganges valley during the 1st millennium B.C. and have sufficiently homogeneous elements to justify the name of the ‘Ganges Civilization’. Culturally, their common denominator is the presence successively of Painted Grey and Northern Black Polished Wares, described above (p. 26). Some of these sites, even in the dusty and ruinous condition in which we find them today, are truly stupendous. Overlooking the junction of the Ganges and the Jumna opposite Allāhābād is the immense mound of Jhusi, its riven sides oozing with sherds, including the famous Northern Black Polished or N.B.P. Ware of the middle and second half of that millennium. The mound has not yet been explored, but both its key-position and its size indicate its importance. And there are many others, ranging from the little fortified town of Bhiṭā, eighteen miles south of Allāhābād, to the mighty Kaushāmbi beside the Jumna, thirty-two miles to the west, or to metropolitan Mathurā (Muttra) itself, by the same kingly river. Bhiṭā is merely 400 yards square, though its mounds stand boldly above the plain. It may have been the Vichhī or Vichigrāma mentioned on certain sealings found on the site, but the scrappy excavations carried out there half a century ago were incompetent even for their age, and only the occurrence of N.B.P. Ware hints at the antiquity of the place. Kaushāmbi, on the other hand, has a circuit of some four miles and is one of the great sites of India. Before and after the time of the Buddha (c. 500 B.C.) it was the capital of the Vatsa Kingdom of the Purāṇas, and here stood one of Ashoka's edict-pillars (p. 174), now in Allāhābād Fort. Slight excavations were carried out on the site in 1937–8, and since 1948 its systematic exploration has been undertaken for Allāhābād University by Mr. G. R. Sharma. The results have not yet been published, save for interim notes.88

c8.38

The ramparts, over 30 feet high and armed at regular intervals by bastions, were built of mud revetted externally with a battered facing of baked bricks which, as excavated, presents an imposing frontage. They were built before the introduction of N.B.P. Ware, but overlapped the arrival of this fabric, and a central date in the 6th century B.C. provisionally meets the evidence. The facing of baked brick closely recalls that of the Harappā citadel (p. 97), but whether across the many intervening centuries a significant structural tradition had been preserved is more than doubtful; the idea of a sloping brick revetment is scarcely so highly specialized that independent invention is excluded. After the original construction of the Kaushāmbi walls, rectangular towers of baked brick were imposed upon the bastions, and the defences were renewed more than once. Within them, courtyard houses of the later centuries B.C. and possibly later still have been uncovered, together with the site of a large enclosed Buddhist monastery which has been identified with the Ghositārāma described by the Chinese traveller Hiuen Tsang, who came this way in the 7th century A.D. The monastery marked the site where the Buddha preached, and the excavator ascribes the first construction of the main stūpa or relic-shrine to the century of the Buddha's death. The general wealth of the city, in the midst of its fertile landscape and beside its great river-highway, is sufficiently indicated by the vast quantity of material, including many hundreds of terracotta figurines, which has been found here by the present excavators and by less orthodox predecessors.

c8.39

But of published work on the ancient sites of the doāb, the most significant is that carried out in 1950-2 by Mr. B. B. Lal at Hastināpura near an old course of the upper Ganges.⁴⁴ The excavation was designed mainly to reveal the culture-sequence, and was so far successful; the character and function of the buildings sampled at the successive levels remain to be investigated, as they should be, by wider digging. Five main phases of occupation were recognized. The earliest, Period I, was represented by a thin layer, never more than 1½ feet thick and often less, with no structures but with a few much-worn fragments of the Ochre Ware. Neither the shapes nor indeed the technique of the pottery could be ascertained. No implements were found, but the possible association of this ware with the copper hoards has been discussed above.

c8.40

Fig. 27

c8.41

There was no overlap with Period II, which contained the Painted Grey Ware described on p. 26, together with remains of mud or mud-brick walls, but with no clear evidence of the use of baked bricks. Copper was found in the form of an arrowhead, a nail-parer, an antimony rod and a few fragments; there were no stone implements and no iron objects, though in the uppermost levels lumps of iron ore were encountered. Two glass bangles from these strata seem to be the earliest yet recorded from India. The humped bull was represented both by terracotta figurines and by actual bones, which occurred in large numbers with those of buffalo, horse, sheep and pig. Whilst these represent considerable domestic herds and flocks, the bones of deer indicate that hunting supplemented farming; but more notable was the discovery of charred grains of rice which, with other rice recently recovered from Chalcolithic Nārdā Toli (p. 142), carry back rice-cultivation in the subcontinent many centuries behind the earliest date previously recorded (the 3rd century B.C.). Altogether, the remains of this period, dated roughly to 800-500 B.C., show a developed township with an economy based on mixed farming, and with adequately constructed houses of which at present little is known.

c8.42

The occupational strata of Period II had reached a height of 6-7 feet when a flood swept across the site and, for a time occupation ceased. When habitation was renewed (Period III) the Painted Grey Ware had given place to the Northern Black Polished Ware, ascribed above (p. 30) to the 5th-2nd century B.C., and baked bricks were now used alongside mud and mud-brick. Again, no details of house-plans were revealed in the trial-trenches, but baked-brick drains were encountered, and 'soak-wells' or 'soak-pits' of superimposed jars with perforated bases or of superimposed terracotta rings about 2 feet in diameter now begin to appear. Soak-pits and ring-wells of these types are characteristic of Indian towns from the latter half of the 1st millennium B.C. onwards.

c8.43

In Period III iron was regularly used for the first time: barbed and socketed arrowheads, chisels and sickles of this metal are recorded, although copper was retained for antimony rods, nail-parers and a variety of other purposes. Terracotta figurines of animals, especially the elephant, are better modelled than in the preceding period. Money now comes into use in the shape of punch-marked rectangular pieces of silver with symbols which include the 'crescent-on-hill' and the 'sun'; with them were a number of cast copper coins, both rectangular and round, bearing similar symbols. The indication is that of increasing and increasingly systematized trade.

c8.44

Within the limited area trenched, there was in Period III an accumulation of 5-9 feet, with evidence of 3-6 sub-phases; and it seemed that the Period had ended in fire. That the event occurred in or very little before the 2nd century B.C. is suggested by the facts that the N.B.P. Ware now ceased and that 5 coins of the rulers of Mathurā, attributed to that century, were found in the lower (earlier) strata of the succeeding Period IV. Both that Period, which is carried by coinage through a number of sub-phases into the 3rd century A.D., and a medieval occupation (11th-15th century), which followed after a long gap, lie outside our present scope.

c8.45

A not dissimilar sequence is indicated by the more imposing site of the ancient city of Ahichchhatrā, near Ramnagar in the Bareilly district of U.P. and close to a tributary of the Ganges.¹⁷ The city was the capital of the kingdom of North Pāñchāla which, in the Mahābhārata, was wrested from the king of Pāñchāla by the Kurus. It retained something of its metropolitan status until about A.D. 1100, when it was replaced by the modern Badaun. Today the deserted city looms majestically over the plain, with towering ramparts 3½ miles in circuit, the tall fragment of a medieval temple within them. Inconclusive excavation in 1940-4 showed two successive earthen ramparts below a stout wall of baked brick, and P.G. Ware is said to have been found below, as well as in, the earlier rampart. If so, the rampart should not be later than 500 B.C. and may be somewhat earlier. The excavators recognized nine main strata in the total life-span of the town, and observed that each stratum had its own individual plan, with differing building-alignments. As at Hastināpura, coins (square and round) came into use during the N.B.P. Ware period.

c8.46

Between them, the two sites of Hastināpura and Ahichchhatrā, imperfect though our knowledge be of both, proclaim the emergence of a comfortable and organized city-life in the Jumna-Ganges basin sometime in the first half of the 1st millennium B.C. To this period belongs the earliest settled occupation known to us in the Delhi region; at the Purāṇā Qila (Indrapat), which may fairly be described as the first Delhi, and Tilpat thirteen miles to the south. In 1955 a list of thirty-four Painted Grey Ware sites was published,89 and the number is constantly increasing. Nothing that we know contradicts the impression that this Ganges Civilization was as widespread as it was prosperous. We may recognize in it the general urban background of the Mahābhārata without too much labouring of detail: a picture of wealthy and jealous dynasties and polities, based upon a limitless and fertile soil and serviceable river-communications. About the middle of the millennium, a knowledge of iron-working spread through the region, doubtless introduced from Persia where iron-smelting had been familiar for five or six centuries. Its arrival meant no overwhelming revolution in the mode of life of the cities of the Indian plains. Some of their pots and pans underwent a technical change about this time; more important, the introduction of coinage, also from Persia, betrays a quickening of the commercial sense. But, once established, the Ganges Civilization endured through the centuries with a changelessness which the modern age has not altogether shaken. Ahichchhatrā may be dead, but Banaras swarms busily and immutably upon a buried past of similar antiquity.

c8.47

Of two other great cities of the Ganges basin—Rājgir and Pāṭaliputra—something must be said in a later context (p. 173).

c8.48

CHAPTER VII

c8.49

Early Civilization in Central India

c8.50

FROM THE GANGES BASIN it would be logical to move eastwards into Bengal or, turning the eastern corner of the central hills, veer southward into Orissa, where the coastal flats are an extension of the great northern plains. In Orissa two imposing sites have in fact been carefully sampled, but their urban origins are unlikely to go back beyond 300 B.C. At Shishupālgarh, near Bhubaneshwar, a part of a somewhat earlier settlement was powerfully fortified, apparently in the 3rd century B.C., with a clay rampart thirty feet high laid out on a remarkably symmetrical square plan.⁶⁹ The sides of the enclosure were three-quarters of a mile long, and in each of them, apart from posterns, were two magnificent projecting gateways, built of squared laterite blocks and equipped with flanking staircases. The unusual excellence of the masonry and the extreme regularity of the planning indicate administrative authority of a high order, and it might not be unduly rash to ascribe the fortress or fortress-town, as an imperial instrument of pacification, to the emperor Ashoka after his famous victory over the Kalingas of these parts about 264 B.C. Historically at least, the defences are most unlikely to have been constructed during the following half-century.

c8.51

⁶⁹ [Figure: Plates 29, 30]

c8.52

Seemingly about the same time at Jaugaḍā, district Ganjām, beside the Rishikulyā river in southern Orissa, a developed Iron Age culture and an earthen rampart twenty-five feet high were imposed upon a site previously occupied by villagers who had used stone axes of oblong section and, apparently, Black-and-red Ware (p. 161).⁷⁰ Here the association of the new work with Ashoka is perhaps encouraged by the presence of his fourteen rock-edicts in the immediate vicinity.

c8.53

⁷⁰ [Figure: Fig. 29]

c8.54

ŚIŚUPĀLGARH 1948 (ORISSA) WESTERN GATEWAY No 1 SITE SP IV

c8.55

Fig. 29 (Cf. pl. 19)

c8.56

In Bengal, too, on both sides of the present national frontier, intermittent excavation carried out on obviously important sites, notably Maināmati near Comilla,71 Mahāsthān in the Bogra district,72 and Tamluk (Ptolemy's Tamalites) at the head of the Hooghly estuary,73 has begun to present a similar picture. At Tamluk trial excavations have revealed a scanty assortment of 'neolithic celts and an ill-fired pottery' at the lowest level (compare Jaugaḍā, above), followed by a Gangetic culture (with cast copper coins) ascribed to the 3rd–2nd centuries B.C., and succeeded in turn by a deposit containing much Rouletted Ware of the 1st century A.D. Whether these occupations were continuous is not stated, but it would appear that here, as on the other sites mentioned, substantive settlement began in or about the 3rd century B.C., that is, in the Mauryan period. Incidentally, Mahāsthān has produced an interesting inscription on a limestone slab, recording the earliest known Bengal famine and the measures taken by the local authority to meet it by the issue of paddy from reserve stocks, which were to be replenished when opportunity offered. Both alphabet and language resemble those of Ashoka’s pillar edicts of about 250 B.C. (p. 174). Future exploration may show that civilization came late to the dense monsoon-ridden jungles of eastern India, and that these, like the Assam and Orissa hills down to modern times, had previously been the camping ground of small disarticulate societies in an essentially Neolithic condition.

c8.57

But south of the great plains, in a belt of central India which includes the Narbadā and the upper Godāvarī and extends roughly from Ujjain in the north to the borders of Mysore in the south, a number of town-sites have been dug into in recent years, and a consistent cultural sequence is beginning to appear. Of the full meaning of this sequence little can be said until vertical probing has been followed in its proper sequence by large-scale lateral clearance. For the time being we have to content ourselves with a mosaic of fragmentary indications.

c8.58

Nevertheless, the mosaic does already seem to make something of a pattern, and the pattern is this. There is in central India, at any rate outside Gujarāt, no recognizably urban phase which can be equated with that of the Painted Grey Ware of the plains, i.e. prior to 500 B.C. Future exploration may modify this view; but at present it would appear that along the great central rivers organized civic life began with the introduction of iron shortly after the middle of the 1st millennium B.C., following a phase of Chalcolithic village life in which copper was itself comparatively rare and costly—in other words, trade was rudimentary and the social horizon essentially a local one. If this provisional conclusion is supported, the spread of civilization in India assumes a logical shape: beginning with the fertilization of Middle Eastern ideas along the north-west coast and up the Indus valley about the middle of the 3rd millennium; then spreading southwards, on the one hand, as far as the Narbatā estuary, and overflowing from the upper Punjab on the other hand through the Pānipat-Delhi corridor into the great plains of the Ganges-Jumna doāb in the first quarter of the 1st millennium B.C.; and later, between the 5th and 3rd centuries B.C., percolating eastwards to the Ganges estuary and southwards amongst the Chalcolithic communities of central India. Farther south again, in peninsular India, it is likely enough that city-life was somewhat later still in its inception.

c8.59

Between the north and the centre, the main connecting link prior to the Iron Age appears to have lain in Gujarāt. Other lines of access doubtless contributed, and after the middle of the 1st millennium more central routes, such as that from the Jumna valley up the Chambal river and past Ujjain⁷⁴ took the lead. But as knowledge stands, it is in Gujarāt, and particularly in the Kāthiāwād peninsula, that a measure of continuity can be traced between the earlier civilization of the north-west and the central Indian complex. In this context the key site for the moment is Rangpur, near the head of the Gulf of Cambay. Archaeologically, Rangpur has had a chequered history. It was hailed in 1934 as the most southerly site of the Indus Civilization; in 1947 further exploration was thought to disprove its Indus association, but six years later renewed excavation replaced it on the Indus map.⁷⁵ Still insufficient investigation suggests that the earliest occupation was marked by crude microliths of jasper and agate, without pottery. This was succeeded by a settlement protected by a mud-brick wall over six feet thick and marked by a culture which may be described as a provincial variant of that the Indus Civilization. It included triangular terracotta 'cakes', faïence and steatite beads, a chert blade, and pottery with a peacock pattern, all allied to Indus types. Its thick red pottery, on the other hand, painted in black or chocolate with loops, dots, criss-cross, and horizontal and oblique lines, is less distinctively Harappān. It is to be expected that dilution or partial survival of this sort should occur near the periphery of the Civilization.

c8.60

This sub-Indus culture merged into a succeeding phase characterized by a lustrous red ceramic painted in black with stylized antelopes and less ambitious designs, and this was followed in turn by Black-and-red Ware of a kind which will be discussed later in connection with the South Indian megaliths (p. 161). Crude microliths still appear, but sherds of Northern Black Polished Ware indicate a date after 500 B.C. Substantially from the sub-Indus phase (latter half of 2nd millennium?) to the N.B.P. period, the occupation of the site seems to have been continuous.

c8.61

For the rest, the Chalcolithic of central India—a phase to which the term 'civilization' cannot at present be applied although villages were already merging into towns—lacks any clear initial link with the Indus, unless the occasional animal patterns on pottery and certain types of flake implements are an echo from the Indus borderland. Instead, within a generally coherent scheme, it shows a wide diversity of detail, based upon local tradition and initiative. The most northerly site of the series is at Nāgdā, on the east bank of the Chambal north-west of Ujjain, where the uppermost third of a partially natural mound ninety feet high represents three cultural periods.* The earliest, twenty-two feet thick, includes massive mud and mud-brick structures of which one may possibly have been the bastion of a rampart. Alongside a few scraps of copper are many microliths, mostly parallel-sided blades with one side occasionally retouched and the other (the working side) often serrated. A few blades show a crested ridge. The pottery, mainly red but sometimes cream-coloured, is painted in black or chocolate with concentric circles, wavy lines, hatched lozenges, 'sun-symbols', spotted deer, and peacocks. The date is roughly the earlier half of the 1st millennium B.C.; for the phase is succeeded without appreciable break by an iron-using culture with 'megalithic' Black-and-red Ware. This is followed by the limited introduction of baked bricks and by N.B.P. Ware, with more iron. A potsherd and a terracotta ball from the upper levels bear inscriptions of about the 2nd century B.C.

c8.62

Here we have, then, earlier phases marked by a microlithic industry and in part by a ceramic which is essentially local though perhaps, in its animal-decoration, influenced at long range by the cultures of the Baluchistan frontier; and a later, iron-using phase in which, as the N.B.P. pottery indicates, elements that certainly included a new trend towards urbanization found their way southwards from the developing civic centres of the Jumna-Ganges doab. The absence of Painted Grey Ware and the presence of iron closely followed by N.B.P. pottery shows that this penetration from the northern plains occurred not appreciably earlier than the 5th century B.C.

c8.63

A similar process is indicated by a dozen other sites which have been trenched in the great central region. Here examples must suffice. A site of great promise is that of Maheshwar, probably the Māhishmati of the Mahābhārata, on the northern bank of the middle Narbadā, with the confronting site of Nāvdā Toli on the south side.¹⁷ As the opposed sites suggest, Maheshwar was an established crossing which carried the highroad from Ujjain and the Ganges basin into the Deccan. Its mounds rise to a height of 100 feet and have produced four main cultures. Like many other central Indian settlements, this was founded on the fertile Black Cotton soil which no doubt yielded a tolerably easy livelihood to an undeveloped community equipped with microliths but lacking both metal and pottery and not yet, therefore, commercially-minded. The microliths are not of an elaborate kind; they include small to medium-sized scrapers of chert and jasper, made mostly on flakes with faceted platform, and irregular and sometimes fluted cores. After a break, of uncertain implication, a more advanced microlithic industry, with backed and crested blades, lunates and points and associated with copper flat-axes, chisels and hooks, introduces the now-familiar central Indian chalcolithic. Querns and slingstones appear, and an abundant pottery, some ochre-washed but predominantly red and often painted with parallel lines, hatched triangles, circles and foliage, occasionally with antelopes and dancing human figures. This deposit is from five to eight feet thick, and may be ascribed to the second quarter of the 1st millennium B.C. It is followed by a deposit no less than twenty feet thick containing a number of occupation-layers marked by large baked or mud bricks, soakpits lined with bricks or pottery-rings, iron implements (sickle, hoe, nails, arrowheads, spearheads), punch-marked coins, N.B.P. Ware and 'megalithic' Black-and-red Ware, a little of which had seemingly appeared also in the preceding deposit. Here once more we have the familiar complex of the second half of the 1st millennium B.C.: a mixture of dominant 'Ganges' elements with certain others of a more southerly kind. Above this deep deposit occurred a finely burnished red ware with 'sprinklers' of a type found at Chandravalli (Mysore) and in western India at Kolhapur, Baroda and elsewhere in layers dating from the early centuries A.D. From this crowning deposit, the 'megalithic' Black-and-red Ware was absent.

c8.64

¹⁷ [Note text not visible on page]

c8.65

Nāvdā Ṭoli, the suburban site across the river, is a group of four mounds about 350 by 200 yards in extent and ten feet high, Fig. 30 Pottery of the Chalcolithic phase from Nāvdā Ṭoli. 1/4 containing huts indifferently rectangular or round on plan and built of wattle and daub with plastered walls and floors. The suggestion of the excavators that it was essentially a boatmen's village (as indeed the name implies) is probably near enough to the truth. Nevertheless, the Chalcolithic levels immediately below a crowning Iron Age stratum with N.B.P. Ware of the 5th century B.C. or later produced distinctive pottery of a type not yet recognized elsewhere in India; notably, pedestalled goblets and channel-spouted 'tea-pots' which Dr. Sankalia has compared with the spouted pots of Hissar III, or of 'Cemetery B' at Sialk in central Iran. 'Cemetery B' was dated by Ghirshman to c. 1000-800 B.C.,⁷⁸ and a recent C14 dating from an equivalent deposit at Hasanlu in Azerbaijan gives the confirming date of 812 BC ± 130 years.⁷⁹ How significant the suggested resemblance with the (much simpler) Nāvdā Ṭoli pots may be remains to be seen, particularly if a recent C14 dating for a late Nāvdā Ṭoli deposit (1336 B.C. ± 125, see p. 83) is confirmed. Intermediate examples between Persia and Nāvdā Ṭoli are at present lacking. A wide range of food-grains included rice.

c8.66

c8.67

c8.68

c8.69

c8.70

c8.71

c8.72

c8.73

c8.74

c8.75

Fig. 30 31

c8.76

Plate 31

c8.77

Fig. 30

c8.78

Seventy miles north of Maheshwar, beside the Sipra river, stands the classic site of Ujjain, one of the seven sacred cities of India, rival even to Banaras in Hindu esteem.⁸⁰ In the semi-history of the period, Ujjain was the capital of the kingdom of Avanti, known later as Mālwā, but the excavations in progress at the time of writing under the skilled direction of Mr. N. R. Banerjee have not yet clearly identified any corresponding 7th-century occupation. At present it would appear that, as a developed city, Ujjain began substantially with the impact of an iron-using culture of Gangetic type about 500 B.C., when an area a mile long and three-quarters of a mile wide was enclosed by a great clay rampart with a basal width of 200 feet and a surviving height of forty-two feet. The rampart was reinforced against the flow of the river by transverse barriers of Fig. 31 Chalcedony microliths of the Chalcolithic phase: 1-5, from Nāvdā Ṭoli; 6-9, from Jorwe; 10-11, from Pravarā Saṅgam; 12-13, from Nāśik. ⅖ interlocked timbers vaguely recalling the timber-work of the riverine site of Pāṭaliputra (below, p. 176), and was subsequently revetted in part with baked brick. Outside it, where necessary, was a ditch eighty feet wide and twenty-two feet deep. On the trunk-route from the Ganges to the Arabian Sea, the fortress town of Ujjain was clearly a power to be reckoned with during the last five centuries B.C. In later epochs, changing circumstance and, in particular perhaps, the development of sea-traffic, robbed it of some of its arterial importance, but the ancient site remains a formidable ruin.

c8.79

c8.80

c8.81

c8.82

c8.83

c8.84

c8.85

c8.86

c8.87

c8.88

c8.89

c8.90

c8.91

c8.92

In its earliest civic phase, represented by six feet of strata, it included stone and baked-brick structures associated with arrowheads of bone and iron, spearheads and choppers of iron, and evidences of copper-working. Some of the pottery was the Black-and-red Ware already encountered elsewhere; with it were a black-slipped fabric and sherds of unslipped and red-slipped vesiculated ware, some of it made without the wheel. In the body of the rampart two sherds of Painted Grey Ware—the southernmost examples known—are not likely to be much later than 500 B.C.

c8.93

In the succeeding phase, represented by an accumulation fourteen feet thick, buildings were of baked or unbaked bricks and were roofed with oblong tiles. Streets were made of rubble bound with clay. The soakpits built of pottery-rings or pierced pots, which were to be characteristic of Indian towns for a thousand years, now made their appearance. Oval clay-lined ovens, blackened cooking-pots, grinding-stones and mortars represent kitchen-equipment; iron spearheads and arrowheads, with bone arrowheads which included one ‘stained with the blood of a bird’, were the principal outfit of the hunter and the soldier. N.B.P. Ware was found in large quantities and in varieties which were partially perhaps local products, some of them diversified with bands of black or saffron paint. The coarse vesiculated ware of the preceding period continued.

c8.94

Indeed Phase II grew naturally out of Phase I under increasing Gangetic influence. Two ivory seals inscribed in an early Brahmi script of the 3rd or 2nd century B.C. marked the upper range of Phase II, but this in turn melted into Phase III, of the Shunga period and later, with which we are not here concerned.

c8.95

Less metropolitan in character than Ujjain were three towns which may be taken as typical of the region south-south-west of Maheshwar: Bahal in the district of East Kandesh beside the Girna, tributary of the Tapti;⁵¹ Nāsik in the district of that name north-east of Bombay; and Jorwe in the district of Ahmadnagar.⁵² The last two are in the upper Godāvari system, a region thickly occupied in the 1st millennium B.C.

c8.96

Bahal began with what may now be called the 'regulation' microlithic industry of central India: mostly parallel-sided blades of chalcedony, agate or jasper, sometimes serrated, with occasional lunates and trapezes. This microlithic occupation falls into two sub-phases. Phase IA is marked by thick grey hand-made bowls with flared rim, of a kind which occurs also farther south at Brahmagiri in northern Mysore, and a few sherds of thinner grey ware painted on the rim in red ochre. Phase IB produced fine red wheel-turned ware, painted in black with a rich variety of designs—horizontal bands with hatched diamonds, triangles, ladders, concentric circles, criss-cross and wavy lines, foliage, and rarely animals (antelope or horse?). A few sherds of lustrous red-slipped ware are reminiscent of post-Indus Rangpur (p. 138). In the upper levels are spouted vessels of the Nāsik-Jorwe type (below), and sherds of a burnished Black-and-grey Ware with oblique lines painted in white, akin to the 'megalithic' Black-and-red Ware. A shapeless lump of copper justifies the appellation 'Chalcolithic', but it is at least certain that the quantity of metal used in this Phase was negligible. The two sub-phases are thought to have been separated by a break in the occupation, and the ceramic changes are consistent with a shifting population.

c8.97

At Tekwāḍā, across the river from Bahal, four burials of Phase IB have also produced Black-and-grey Ware of ‘megalithic’ types.

c8.98

Again, it is with the arrival of iron in Phase II that the settlement first assumes a purposeful and urban aspect. So far as we know, houses were indeed still of timber, but their equipment now includes legged querns with grains of rice and millet, and the microliths are replaced by abundant tools and weapons of iron—spearheads, arrowheads, knives, sickles. N.B.P. Ware of the Ganges tradition once more marks the source of inspiration. The excavator, M. N. Deshpande, observes that iron here occurs some ten feet below the lowest N.B.P. sherd, recalling a less marked priority for iron also at Nāgdā (p. 138); but on sites relatively so remote from the Ganges *doāb* the utile metal may understandably have penetrated ahead of the less essential ceramic, and the two innovations may be ascribed to the same main source. Burnished ‘megalithic’ Black-and-red Ware also occurs in the form of rimless bowls and shallow dishes, and there are red globular jars with a gritty core. Massive ear-disks (for wearing in the extended lobes of the ears) of semi-precious stones recall others ascribed to the Mauryan period, roughly the 3rd century B.C., at Rājgīr and Pāṭaliputra.

c8.99

In Phase III the N.B.P. Ware continues but the Black-and-red drops out. The former may not merely have arrived late on this site, but may have continued somewhat later than usual if its apparent overlap with Sātavāhana coins of the 1st century B.C. is significant.

c8.100

Both Nāsik on the Godāvarī and Jorwe on the tributary Pravara work to rule in the microlithic phase;¹ indeed, Jorwe never got beyond that phase, though it mitigated its scraps of chert and chalcedony with six flat-axes and a bangle of low-grade bronze which is probably in fact a natural copper alloy. The presence of the copper axes here and at Maheshwar (p. 140)

c8.101

may represent a hang-over from the Indus or sub-Indus culture of the north-west coast; it is less remarkable than is their general absence from these microlithic sites, where their utility in the recurrent process of jungle-clearance must have been cardinal. It is a constant question how these communities, with their miserable microliths, coped with the luxuriant vegetation which certainly surrounded them, although their preference for sandy and relatively open spots such as Langhnāj is a partial explanation. In other parts of the world, microlithic communities have similarly sought a negative milieu.

c8.102

Nāsik is a mound about 800 by 400 feet with a height of 100 feet of which only the uppermost twenty feet are man-made. Its Phase I produced lunates, trapezes and two-edged blades, and it was observed that 60 per cent of the microliths were 'usable'; in other words, no factory site was included and, to that extent, the limited exploration suggests a measure of specialization or professionalization. With the microliths was ochre-washed and orange-coloured pottery, both wheel-made and handmade and occasionally painted with red or black bands. A distinctive type is that of a spouted vessel, sometimes painted after baking. This type, already noted at Bahal (p.145), is found also at Jorwe and at Nevāsā north-east of Ahmadnagar, and is probably related to a spouted vessel from the levels marked by stone axes and microliths at Brahmagiri in northern Mysore. A similar affinity may be recognized in the large round-bottomed vessels with high rims found at all three sites.

c8.103

The microlithic phase at Nāsik ended characteristically with the introduction of iron, though an intervening 'weathered horizon', of uncertain implication, is recorded. The Iron Age occupation was marked by iron spearheads, arrowheads, knives, choppers and caltrops, by four-legged saddle querns and pottery-ringed soakpits, and by N.B.P. and Black-and-red Wares. Traces of mud-walled houses were identified, but their plans are unknown. Above this phase come Russet coated (‘Āndhra’) wares and ‘sprinklers’ of red polished pottery, all characteristic of the 1st century A.D.

c8.104

It is unnecessary here to follow further the exploratory trenching which, over a wide area, has begun to make sense of the cultural sequence of central India without, as yet, throwing much light upon its social significance. That will follow, be it emphasized, when determined lateral excavation over a long period succeeds these easy and quickly rewarding forays.

c8.105

Meanwhile, with the possible exception of Ujjain, which may be found to have an earlier nucleus but owes its formidable urban character to advancing Gangetic interests about 500 B.C., the sites mentioned above, with others of their kind, have their origins in a widespread and often rudimentary microlithic culture—if the term culture can be so narrowed. Sometimes, as at Rangpur and Maheshwar, this primary culture or industry is thought to have been devoid at first of associated pottery; but at more than one site two to four successive phases of microlithic occupation have been marked by differing pottery-groups and have suggested a certain mobility in locally variant communities at a uniform economic and technological stage of development. In so far as generalization is feasible, we may say that there was in central India, in the earlier half of the 1st millennium B.C., a considerable village-population which used microliths, a little copper or low-grade bronze, and rather coarse painted pottery, often but not always wheel-turned, with a certain community of types that included large round-bottomed bowls with flaring rims and spouted, carinated ‘teapots’. This population extended from the Vindhya Range 600 miles southwards into northern Mysore.

c8.106

Towards the south of that area, the microlithic culture lasted perhaps until the 3rd century B.C. Towards the north, but occasionally as far south as Amarāvatī on the Kistna river, it was replaced in the 5th or 4th century B.C. by a vigorous iron-using culture which seems, in the main, to have come from the Jumna-Ganges doāb. With the arrival of iron, the microliths went suddenly and almost completely out of use. (Rangpur may provide an exception, but much further work is needed at this site.) Shortly afterwards that typical Ganges product, the steel-like N.B.P. Ware, came in, preceded or accompanied here and there by the more local 'megalithic' Black-and-red Ware of which more will be said in the next chapter. These features, supplemented by native industries, mark the real beginning of civic life in central India. Houses were sometimes still of timber, but mud, mud-brick and baked brick now appear, and pottery-lined soakpits represent a new measure of urban sanitation. Fortifications are not normal, or at any rate have not been widely recognized. It would appear that an undisputed *Lebensraum* was still available for most of the overflow from the northern plains and for the newly-enlarged and elaborated communities of the Deccan.

c8.107

In summary terms, the contrast between the first urbanization of the northern (Gangetic) plains and that of the Deccan is this. In the north, urbanization began in the developed Bronze Age, during the earlier half of the 1st millennium B.C.; and when, about 500 B.C., iron and the parallel innovations of coinage and a new ceramic technique appeared upon the northern scene, they were absorbed into the existing social system without any drastic change of personality. In the Deccan, on the other hand, the Iron Age descended from the north as a sudden imposition upon a Chalcolithic (largely microlithic) local tradition of altogether lower grade. There, the Iron Age began in the full sense as a revolution.

First 106 of 108 passages. Read on

Early India and Pakistan to Ashoka c8: Chapter VI. The Ganges Civilization

Chapter VI. The Ganges Civilization

Linked texts (26)