Northwestern South Asia
The earliest agriculture in South Asia can be found along the western tributaries of the Indus River, at aceramic settlements like Mehrgarh (see Chapter 11 on Mehrgarh). Here, Southwest Asian crops, sheep, and goats have been discovered dating back to the seventh millennium bce.
This occurs in the context of aceramic Neolithic Mehrgarh, which was a modest sized and apparently sedentary village. The precise date at which the Neolithic begins here remains poorly established, with 7000 bce being a widely cited supposition.[503] The transition to Mehrgarh period ii and the ceramic Neolithic appears to be well dated to around c. 6000 bce, which is approximately coincident with the 8.2 ka event. This climate event saw an increase in aridity across the area, which is likely to have been a factor that led to the development of agriculture. For this period of >6000 bce up to c. 4000 bce, Mehrgarh remains essentially unique in providing details of subsistence and a continuous sequence of occupation, as no other contemporary sites have been systematically excavated. By c. 4000 bce, smaller Neolithic villages are known to have existed throughout the western borderlands, including theBannu basin, and these settlements re-emphasize the importance of domesticates of Southwest Asian derivation.9
There are perhaps as many as nineteen aceramic Neolithic sites (the Kili Gul Muhammad phase) that have been identified in western Pakistan,10 though some of the identifications are speculative (e.g. Rana Ghundai) and at least some sites listed as being villages have in fact previously been identified as Mesolithic/hunter-gatherer sites on the basis of the lithics found on the surface, and are effectively undated (e.g. Gul Shah Tup, Nekumshakh, Tup Takhtikhel, and Yarak in the Bannu region11). It is entirely feasible that some of the Mesolithic sites that have been identified were in fact used contemporaneously with the period i occupation at Mehrgarh and other likely and reputed aceramic sites.
Furthermore, given the evidence for continuity in huntergatherer subsistence in other parts of the subcontinent in later millennia (e.g. at Bagor, Adamgarh, and Tenmalai12), it is also possible that some hunter-gatherer subsistence may have occurred in later phases in the borderlands and other parts of western South Asia. Surveys in upper and lower Sindh have identified an abundance of sites that are characterized by a microlithic assemblage that is distinct from the Mehrgarh, Amri, Kot Diji, and Harappan assemblages,13 but again the precise dates of these sites are not clear.These early village sites were often situated on large alluvial fans, and such fan environments were also favoured by early farmers in various parts of West Asia, including many on the plains of southern Iran, such as the settlements at Rahmatabad and Shah-Maran/Daulatabad.14 It is possible that the adaptation of farming practices to such specific ecological conditions may have been a factor in the distribution of such early villages and the
9 Petrie (ed.), Sheri Khan Tarakai, 7.
10 G. Possehl, Indus Age: The Beginnings (Philadelphia: University of Pennsylvania Press, 1999), figs. 4.14 and 4.28, table 4.13.
11 F. Khan et al., ‘Prehistoric and protohistoric settlements in Bannu district', Pakistan Archaeology, 23 (1988), 99-148 (102); F. Khan et al., Explorations and Excavations in Bannu District, North-West Frontier Province, Pakistan, 1985-1988 (London: British Museum, 1991), 21-2.
12 D.K. Chakrabarti, India: An Archaeological History: Palaeolithic Beginnings to Early Historical Foundations (Delhi: Oxford University Press, 1999), 99, and The Oxford Companion to Indian Archaeology: The Archaeological Foundations of Ancient India (Oxford University Press, 2006), 97.
13 E.g. P. Biagi, ‘The Mesolithic settlement of Sindh (Pakistan): a preliminary assessment', Praehistoria, 4-5 (2003-4), 195-220, and ‘New discoveries of Mesolithic sites in the Thar desert (upper Sindh, Pakistan)', in E.
Olijdam and R.H. Spoor (eds.), Intercultural Relations between South and Southwest Asia: Studies in Commemoration of E.C.L. During Caspers (1934-1996) (Oxford: Archaeopress, 2008), 78-85.14 C.A. Petrie and K.D. Thomas, ‘The topographic and environmental context of the earliest village sites in western South Asia', Antiquity, 86/334 (2012), 1055-67 (1057). pattern seen in the dispersal of farming practices, where specific habitats were preferentially chosen (essentially a form of habitat tracking), consistent with a dispersal of at least some farmers from Southwest Asia.
Cereals present in the aceramic Neolithic (naked six-row barley, domestic hulled six-row barley, wild and domestic hulled two-row barley, domestic emmer, domestic einkorn, and a free threshing wheat) represent several of the ‘founder crops of Southwest Asian agriculture',[504] though no pulses have been found in these early levels. The presence of some modern wild barley in parts of Pakistan and Afghanistan has fuelled speculation that there may have been local domestication at or near Mehrgarh.1[505] Recent genetic research on barley suggests multiple origins, including an eastern barley group that differs from the Levantine and main European barleys, and could have originated in Iran or somewhere further east.1[506] However, an archaeological sequence documenting a domestication process is lacking and the crop package that dominated the Indus region during the third millennium b ce is West Asian in character (pulses, flax, wheat, and barley) with a limited indigenous South Asian component. This points to the predominance of this early crop dispersal from the west.1[507]
The faunal assemblage of earliest Mehrgarh is dominated by wild species, including gazelle, wild sheep, deer, wild buffalo, wild cattle, and goats, which are the second most common species after gazelles.[508] However, significant numbers of goat kids in burials and the remains of relatively small subadult or adult animals in rubbish deposits is consistent with kill patterns in domestic herds and indicates that domesticated goats were present from the earliest levels.
Over the course of the aceramic Neolithic, sheep and cattle increasingly came to dominate the faunal assemblage and the animals represented decrease in size, which is a common marker used to identify domestication.[509] Sheep genetics, however, appear to preclude an independent South Asian domestication of sheep, unless a population domesticated in the east was replaced more recently by imported breeds.[510]In this northwestern region there are a few taxa that are plausibly local domestications that broadened the agricultural economy. The most important of these was undoubtedly zebu cattle (Bos indicus), which has long been argued to be a local domestication at or around Mehrgarh between c. 7000 and 4500 BCE.[511] This is important because cattle dominate most faunal assemblages throughout the subcontinent, and they may all derive ultimately from this domestication process with the incorporation through hybridization of additional wild genetic lineages. In addition, Mehrgarh provides early evidence for the exploitation and probable cultivation of Indian tree cotton (Gossypium arboreum), which was to become important in Bronze Age and later textile production.[512] Water buffalo (Bubalus bubalis) and sesame (Sesamum indicum) are further domesticates that had been added to subsistence regimes certainly by 2500 bce in at least some parts of the Indus region.
There appear to have been a number of distinct stages in the transition from mobile hunter-gatherer subsistence to sedentary farmer-herder subsistence, the subsequent dispersal of the agropastoral subsistence economy based on wheat, barley, sheep, goats, and cattle, and the progressive sophistication in craft production in western South Asia.[513] At a fundamental level, this process was characterized by a diffusion of practices into South Asia via the Iranian plateau or Afghanistan, initially into western Pakistan.
The precise pattern of this process is, however, unclear. It is not readily apparent whether there was demic diffusion, involving movement of agropastoralist populations; cultural diffusion, involving the adoption of agropastoralism by populations previously using other subsistence strategies; or nuanced combinations of the two. A disjunction in the physical anthropological record at Mehrgarh and differing phylogenetic signatures in lentils could suggest two waves of diffUsion/migra- tion, with the second around 4500 bce.[514] Evidence for long-range trade in exotic raw materials such as lapis lazuli clearly indicates that long-range contacts were maintained over several millennia, so there should be no doubt that this earliest phase of village occupation in South Asia was one where mobility of people and ideas could be widespread.The transition to the use of fired ceramics c. 6000 bce saw a number of other important and dramatic socioeconomic changes for the inhabitants of early villages in the northwest. Along with evidence for the first fired ceramic vessels, we see the first substantial mud-brick buildings, clear indications that specific craft activities were being carried out, and a marked development of the agricultural subsistence economy of the inhabitants.[515] Ceramics were slab-built and, as they do in West Asia, represent a change in food processing to focus on grinding and clay-oven baking, a contrast to other parts of Neolithic India.[516] Notable changes are evident in the subsistence economy in period iia at Mehrgarh, including the almost complete replacement of wild animals with domesticates, the majority of which were cattle (over 50 per cent of the assemblage), with lesser numbers of sheep and goats.[517] There were also increases in the number of settlements, the size of settlements, and the settled area over the preceding phase.[518] However, the raw treatment of the data may have masked a range of possible dynamics.
For instance, as noted above, it is entirely possible that groups subsisting wholly or principally by hunting and gathering persisted in the borderlands and other parts of western South Asia after the appearance of ceramics. New villages may thus have been established in regions that did not previously have them.During the fourth millennium bce, there were further significant socioeconomic developments for the early village societies living in the western parts of the subcontinent. It is during this period that there is the clearest evidence for the initial spread of the wheat-, barley-, sheep-, goat-, and cattlebased agropastoralist subsistence economy into different parts of the hill, piedmont, and plain zone along the edge of the borderlands, and ultimately into the Indus plains themselves.[519]
In terms of what we understand about the relative and absolute chronology of early village sites in western South Asia, it appears that it was only after the expansion of wheat-, barley-, sheep-, goat-, and cattle-based pastor- alism into southern and northern Baluchistan and also southern Khyber Pakhtunkhwa that settlements practising agropastoralism appeared on the Indus plains. Settlements such as Rana Ghundai (Loralai) and Periano Ghundai (Zhob) appear to have been settled before Sheri Khan Tarakai (Bannu basin) and Jhandi Babar A (Gomal plain), which share a similar material assemblage and were in turn earlier than settlements on the plains.[520] Our understanding of the earliest village settlements on the Indus plain comes from excavations at Jalilpur and Harappa,[521] and the Cholistan survey,[522] during which a total of ninety-nine sites characterized by Hakra ware ceramics were identified (including fifty-two that were interpreted as camp sites and forty-five as settlement sites, and two where there is evidence for craft production) with a suggested four-tiered site-size hierarchy. Hakra ware ceramics include the first wheel-thrown ceramics in the region, alongside handmade wares. At Harappa this period is recognized as the Ravi phase (3300-3000 bce) and is succeeded by the early Harappan period that culminated in the transition to the mature urbanism of the Harappan era around 2600-2500 BCE.
Unanswered questions remain about the possibility of a separate early food-producing tradition of the upper Punjab, Haryana, and the Indo- Gangetic divide. Evidence from several sites in northwestern India indicates the presence of crops that were absent from Baluchistan and Khyber Pakhtunkhwa. There is the possibility of a local transition to crop cultivation based on indigenous monsoon taxa, such as millets (Panicum sumatrense, Setaria pumila, and Setaria verticillata) and pulses (Vigna mungo and Macrotyloma uniflorum).[523] In both Rajasthan and Saurashtra in Gujarat, the local Mesolithic tradition gave way to ceramic-producing cultures around 3500 bce. These are usually classed as ‘Chalcolithic' cultures, due to the early availability of copper and links to the Indus region. Current evidence points to the dominance of West Asian-derived winter cereals and pulses with some input from the indigenous crops of ‘inner' monsoonal India at these settlements.[524] Unfortunately data are only available from the era of the Indus civilization (after 2600 bce) but they may point to an Indian savanna agriculture, quite different from the winter-crop wheat/barley traditions of the contemporary Indus valley. While this provides an argument for a local Neolithic tradition with local plant domestications, it is not empirically documented and remains undiagnosed on the ground in terms of sites or material assemblages. It is also unclear whether this tradition should be seen as connected to the southern Deccan development of farming (see below), especially as both regions share the same introduced livestock complex.