The idea of ‘degenerate and, therefore, a Late phase of Indus’ was first put forward by N.G. Majumdar in the context of pottery found in the upper levels of two sites in Sind, Jhukar and Lohumjodaro. The Indus black-on-red style of pottery ‘continued but in a modified style’ and was accompanied by a coarse bichrome pottery which appeared for the first time. Majumdar further noted that at the site of Trihni this new pottery was found with a type of pots which had incised strokes at the shoulder and also occurred in the late level of Mohenjodaro. It is to Majumdar’s great credit that he understood that certain ceramic changes were taking place towards the stratigraphically upper or late phase of the Indus civilization. In 1938, Mackay noted that the brick masonry of Mohenjodaro during its late period was poor by the standard of the earlier periods, and ‘towards the end of the Late Period’ a particular excavated area (G section of DK mound) came to contain potters’ kilns, including one in the middle of its main street. That the potters could set up kilns in an area which was purely residential before was an indication that the civic standard had remarkably declined during this period.
We have indeed, come upon a striking example of the decay of an once honourable city, the cause of which we suspect to be the vagaries of the Indus rather than pressure by invaders, of whose existence we have, in fact, little positive evidence.18
The idea that Mohenjodaro fell to a group of invaders is associated with Mortimer Wheeler (1947) and is based on two sets of archaeological data. First, in the upper levels of the eastern mound at Mohenjodaro there are about thirty-three randomly distributed skeletons of males, females and children. These skeletons are not contemporary; in one case it has even been established that the cut mark which it bears did not lead to death. This massacre idea is a myth. Second, in different areas of the north-west there are a few archaeological objects which are of west and central Asiatic derivation. They were not found in well-defined archaeological contexts. There is no reason why they should be called contemporary and indicate incoming invaders. However, such objects were forcibly interpreted that way and along with the above-mentioned skeletons led to the hypothesis that the Indus civilization in general and Mohenjodaro in particular fell prey to a group of invaders. The scenario still has its appeal in certain quarters. Certain objects of north Afghanistan affiliation, which have been found in the Quetta and Mehrgarh area in the vicinity of the Bolan pass, are being touted as markers of a foreign infiltration towards the end of the Indus civilization. The protagonists of this view ignore the facts that these objects may well belong to the mature Harappan period and that such finds in the Oxus–Indus interaction zone are only to be expected in view of the close trading network all over the interaction area. Besides, one does not have to interpret cultural change in the Indus-Hakra plain with reference to situations in the Bolan pass.
The decline and transformation of the Indus civilization has to be understood in its own terms, and it is here that we have to focus closely on archaeological stratigraphy and changing cultural situation, to which Majumdar and Mackay drew attention long ago in the context of Sind. The same point has been highlighted by the current excavations at Harappa. Overcrowding of the site has also been noted here towards the late phase and the period which has specifically been called Late Harappan is Period 5 characterized by a pottery known as Cemetery H ware (so called because this was first found in a cemetery marked H in M.S. Vats’ early excavations at the site) and drains and burnt bricks of a smaller size. One notes that though the pottery changes, the elements of urban planning remain.
This phase may reflect a changing focus of settlement organization that was a pattern of the earlier Harappan phase and not cultural discontinuity, urban decay, invading aliens, or site abandonment, as has been suggested in the past.19
This continuity was first observed in Gujarat in the context of Rangpur. Period IIA of Rangpur is mature Harappan. Deterioration in the quality of pottery, absence of drains and baths and thatched roofs of houses suggest to the excavator ‘a decline in the prosperity of the surviving Harappans at Rangpur’ in Period IIB. Mature Harappan pottery further disappears in Rangpur IIC which, however, develops a new pottery type which is known, because of ‘the burnishing of the surface of the vessels after applying a thick red slip’, ‘lustrous red ware’. It is this pottery which becomes dominant in Period III: ‘the Lustrous Red Ware culture,’ as Rao showed at Rangpur, ‘was not an intrusion from elsewhere but a local development of the Harappa culture itself.’
The evidence of transformation is clear in the Indo-Gangetic divide too. The transformation was initially worked out on the basis of stratigraphy and associated ceramic continuities and changes at sites like Mitathal and Siswal in Haryana. Siswal A is early Harappan, whereas Siswal B is mature Harappan and Siswal C late Harappan. There is no dramatic cultural discontinuity anywhere:
Settlements are established and abandoned and items of material culture change through time, but there is no indication that the region was ever abandoned completely by humans nor witnessed a dramatic influx of foreign cultural groups.20
Continuity and transformation are the fundamental features of the phase after the mature Harappans in all their distribution areas. The cultural situation varied from area to area. In the Kachhi plain, the earlier tradition continued with new elements at the site of Pirak, whereas in south Baluchistan the Kulli culture might have lingered on. At Jhukar in Sind only a new pottery style emerged ‘in association with the continuing mature Harappan tradition without any break or sudden change in cultural continuity’ but the Indus script was limited to potsherds and cubical stone weights became rare, among other things. Period 5 of Harappa—Cemetery H Ware culture—is best focused in Cholistan which has 50-odd sites of this period, including one c. 38 ha site, Kudwala. The entire area between the Sutlej and the Yamuna is dotted with late Harappan sites with a strong concentration in the upper Doab. Mud houses, faience ornaments, semi-precious stone beads, copper, ritualistic clay objects, terracotta toy-cart frames, kilns, fire-altars, storage pits, etc. are parts of the general late Harappan cultural complex in this region. There is no reason to think that its agricultural diversity and richness became any less during this period. In Gujarat, there is apparently a miscellany of situations, ranging from the small sites with cattle pens (cf. Kanewal) to the fortified ones like Rojdi and Bet Dwaraka. The use of script becomes less common in comparison with the earlier period but does not disappear. Agriculture remains firmly entrenched, as in the earlier period, and, if anything, may have even expanded its scope. The number of late Harappan sites in Gujarat, as elsewhere, is considerably more than the number of mature Harappan sites. In Gujarat at least the tradition of foreign trade continues; the discovery of a seal with a motif (‘whorl’ motif) common in the Gulf at Bet Dwaraka and the finds of Indus and Indus-related seals at Nippur and in Failaka are proof enough of this continuity of the mature Harappan tradition of external trade.
At the same time, there is a marked overall element of decline. The archaeological repertoire becomes much simpler; the use of script becomes very limited, and there was much less use of raw materials transported over long distances. While trying to explain this decline, one has to point out a major feature of the distribution of late Harappan sites between the Sutlej and the Yamuna. There was a remarkable shift of the focus of settlements towards the Doab during this period. We do not really know why this shift took place but the increased pace of the drying up of the Ghaggar–Hakra system seems to have been one reason. Another reason may have been the fact that the Harappans overstretched themselves. Their foundations in many areas of their distribution were not particularly deep. In Cholistan and some parts of Sind, west Punjab, Haryana and Gujarat they had an ancestry going back either to the Hakra ware period or the early phase of the early Harappans. Elsewhere in the Sutlej—Yamuna stretch the early Harappan phase did not antedate the mature Harappan by a long period; in the Doab there is no early Harappan at all, and in Gujarat there was a substantial group of microlith-using hunter-gatherers in the background.
… One has to admit that the Harappans eventually came to be rather thinly stretched on the ground, and the weakening of their political fabric was almost inevitable. They were swallowed up, as it were, by the much less advanced pre-agricultural groups of inner India.21
We have noted that there was a marked shift in the focus of settlements towards the Doab. Once the Harappans moved in that direction and settled in the upper part of the Doab, they were geographically bound to be drawn into the main cultural development of the Ganga valley. Similarly, from the geographical point of view their dense distribution in Gujarat in the late Harappan period means that their interaction with, or movement towards, south-east Rajasthan, Malwa and the Deccan becomes a distinct possibility. As we shall see in the next chapter, there was a straight movement towards the Deccan along the Tapti valley, and there is evidence of interaction with both south-east Rajasthan and Malwa. The Harappans, in fact, were moving into the flow of cultural development of inner India.