264 Addenda and Corrigendaley and Son Valley, both on rivers flowing from Madhya Pradesh into Uttar Pradesh, have demonstrated stratigraphic and develop- mental relationships linking Upper Paleolithic and microlithic technologies. Although several sites have been excavated, infor- mation about environment, economy, or habitation features re- mains rare, except at Baghor I (24°ree;35'22"N82°ree;18'54"E), in north- ern Madhya Pradesh, where a possible shrine complex similar to those in the modern countryside has been found and dated to c. 10,000 B.C. At present, the Upper Paleolithic dates from c. 20,000 + to 10,000 B.C., which is later than comparable Eurasian developments. A complete range of traditional Old World "Stone Ages" is now definable for South Asia, although the chronology differs significantly from that for the rest of Eurasia.
II.1 (c). Microlithic/Mesolithic
Microlithic sites spanning the last 15,000 + years are distributed throughout South Asia. Evidence indicates microlithic groups en- gaged in three adaptive strategies: (1) hunting, gathering, and fish- ing; (2) trade and barter combined with hunting-gathering; and (3) limited food production (pastoralism and/or agriculture) combined with hunting-gathering. The first and oldest strategy reflects con- tinuity with Paleolithic developments, whereas the others devel- oped later, c. 7000–4000 B.C., and persist into the present.
Microlithic hunter-gatherers were responsible for the initial settlement of Sri Lanka, evident at Batadombalena Cave (6°ree;46'N80°ree;12'E). This occurred perhaps as early as 25,000 B.C. and certainly by 15,000 B.C. Similar sites are found throughout South Asia. Early examples outside Sri Lanka include Sarai Nahar Rai, Uttar Pradesh, 10,550–9550 B.C.; and Baghor II (24°ree;34'45"N82°ree;10'55"E) and Bhimbetka, in Madhya Pradesh, at 8645–7645 B.C. and 8070–7070 B.C. respectively. The discovery and documentation of extensive rock art in the Bhimbetka area has generated much discussion. Many of the stylistic depictions there of human behavior focus on hunting activities and may date to a very early period. At the same time, as already noted, the hunting- gathering strategy persisted for a long period of time and some scenes depict Early Historic and Medieval activities. A more pre- cise chronology must be developed for this rock art before it can have much archaeological significance. While hunting-gathering groups have persisted throughout South Asian history, most even- tually shifted to the other two strategies.
As early as c. 4000 B.C. hunting-gathering economies were combined with trade or barter by some groups to form a second livelihood strategy, depending on interaction with more complex social, economic, and technological groups such as the Harappan. Items such as pottery, metal objects, glass, and even coins in mi- crolithic, hunting-gathering sites indicate exchanges with more "complex" agricultural villagers, urban dwellers, and even state- level societies. Langhnaj in Gujarat and later occupations at Ba- gor, in Rajasthan, and Bhimbetka, in Madhya Pradesh, provide examples of this strategy. These interactions are based on hunter- gatherer groups exchanging labor and/or hinterland commodities for technologically complex objects.
Domesticated animals and/or plants at microlithic sites were considered, until recently, another dimension of the trade and barter strategy; but new data indicate this was not so in all circum- stances. For example, domesticated animals in Bagor's Period I and in early Adamgarh Cave levels now date to c. 5000 + B.C., much too early to reflect interaction with known local food pro- ducers. Excavations at Mahagara (24°ree;54'30"N82°ree;2'E), in Uttar Pradesh, and Kunjhun River Face (24°ree;31'15"N82°ree;11'15"E), in Madhya Pradesh, in the aforementioned Belan and Son Valleys, uncovered villages with circular, mud-coated bamboo structures where hunting-gathering was combined with the use of domesti- cated cattle and a few sheep or goats and plants, including rice. This initial stage of food production on the Gangetic Plain, the "Vindhyan Neolithic," may have begun as early as 6000 B.C.; however, most dates range between c. 4000 and 1500 B.C. with comparable groups persisting into historic times. Likewise, in the Indus River Valley similar groups had established settlements by 7000 B.C., initiating a sequence which extends into the Iron Age.
New Sources (in addition to those in the General Bibliography)
B. Allchin, A. Goudie, and K. Hegde (1978); D. P. Agrawal and B. M. Pande (1976); D. K. Bhattacharya (1989); N. C. Ghosh (1986); V. Jayaswal (1982); B. S. Karir (1985); V. N. Misra and P. Bellwood (1985); R. P. Pandey (1987); G. R. Sharma and J. D. Clark (1983); G. R. Sharma et al. (1980); M. J. Sharma (1982).
N.B. Many edited volumes listed for section II.2–5 contain ar- ticles relevant to the periods discussed here.
Early Village Farming Communities of the Northwest
Early in the 7th millennium B.C. microlithic groups established Mehrgarh (29°ree;28'N67°ree;41'E), a site on the Bolan River southeast of Sibi, in Baluchistan. These people constructed mud-brick, mul- tiroom buildings for habitation and storage; made stone tools for
Between c. 5000 and 4000 B.C. Mehrgarh grew in size and sev- eral substantial, mud-brick storage buildings were built. During this period pottery production shifted from handmade to wheel- made vessels, while ornaments and a copper ingot indicate contin- ued metallurgical activities. Domesticated cattle became an eco- nomic focus, accounting for at least 60% of all animals found, one of the highest frequencies for cattle in the ancient world. Cotton seeds, a date stone, and water buffalo have also been identified; but it is not certain whether they were of domesticated origin. At least one contemporary site, Kili Gul Muhammad, in the Quetta Valley of Baluchistan, is known and possible sites are reported from the Bannu Basin in the North-West Frontier Province.
By c. 3500 B.C. Mehrgarh covered almost 75 hectares and had become a regional craft center. The distinctive wheel-made, black- on-red decorated pottery characterizing this period is found at sites throughout Baluchistan (e.g., Kili Gul Muhammad, Rana Ghun- dai, Anjira) as well as in early occupations at Mundigak (south- eastern Afghanistan) and Amri (Sind). These ceramic similarities indicate cultural affililations among widespread groups, although the precise relationships are poorly understood. At Mehrgarh craft activities intensified and there were increasingly diverse ornaments made from bone, shell, semi-precious stones, processed steatite, and metal. Copper and, perhaps, bronze objects, such as pins and compartmented stamp seals, were manufactured. The first gold bead also dates from this period. Despite these craft activities, few or- naments were found in the burials at Mehrgarh. Such finds were mainly bracelets, necklaces, and headbands composed of pro- cessed steatite beads combined with infrequent semiprecious stone beads. These ornaments were differentially distributed among pri- marily adult burials, which would suggest that complex social sta- tuses were emerging. However, differences appear to have been based essentially on age and kinship.
Between c. 3500 and 3000 B.C., and perhaps as late as 2500 B.C., regionally distinct, but basically similar, styles of bichrome and polychrome pottery were manufactured. At Mehrgarh and in the Quetta Valley these ceramic changes were gradual and associ- ated with technological improvements such as the production of a very fine grey pottery. These regional ceramic styles, according to some scholars, reflect development of related and interacting so- cial or ethnic groups. Representative sites include, in addition to Mehrgarh, Kechi Beg (30°ree;05'N66°ree;55'E), in the Vale of Quetta; Nal, in the south-central Baluchistan; Amri and Kot Diji, in Sind; and Sothi (29°ree;11'N74°ree;50'E), in northern Rajasthan. Although so- cial units of some ethnic groups, such as the Kot Dijian, persisted into the 2d millennium B.C., many had, by 2500 B.C., developed traits (see Chapter II) linking them with the later Harappan Civi- lization. Apart from their pottery, little else is known about most of these sub-groups. In general, craft and economic activities again intensified. At Mehrgarh the number of stamp seals increased, fe- male figurines appeared, and grapes were utilized for the first time.
The extent of cultural relationships with groups in southern Af- ghanistan is difficult to determine. However, the limited bichrome and polychrome pottery at Mundigak and in early Shahr-i Sokhta occupations indicates some interaction. Between c. 2500 and 2000 B.C., and perhaps later, sites in southern Afghanistan (Mundigak, Deh Morasi Ghundai [31°ree;35'N65°ree;30'E], Said Qala Tepe [31°ree;35'N65°ree;31'E]) and the adjacent portion of Iran (Shahr-i Sokhta) reflect extensive interaction with Central Asian cultures. Some scholars argue that these Central Asian influences are also reflected in post-bichrome/polychrome Quetta Valley ceramics and in the final Mehrgarh occupations. During this period Shahr-i Sokhta be- came a major urban center and impressive public buildings were constructed at Mundigak. While the nature and intensity of inter- action with Central Asia has not been determined, most scholars recognize that previous interpretations focusing on actual migra- tions from that area are too simplistic. Equally problematic is the absence of evidence at these Afghan sites and Shahr-i Sokhta for interaction with the contemporaneous Harappan Civilization. Sometime during the 2d millennium B.C., these sites, and many in Baluchistan, appear, for unknown reasons, to have been aban- doned.
Hakra, another regional Indus Valley social group (named for the ancient, but now dried-up, Hakra River), emerged between c. 3300 and 2700 B.C. Knowledge about Hakra is limited to Baha- walpur survey data and the early occupations at Sarai Kola, Jalil- pur, and Swat Valley sites. The distributions of Hakra and Kot Dijian sites overlap and some scholars argue that a close cultural
Current data indicate that sometime before 7000 B.C. food pro- duction was locally developed in South Asia. Subsequent cultural developments witnessed a proliferation of agricultural and pastoral groups associated with an intensification of craft activities. Cer- tainly, by 4000 B.C., if not before, cultural similarities and divers- ities argue for a series of related, but distinct, social or ethnic groups throughout Baluchistan and the greater Indus Valley re- gion. This complex cultural mosaic formed a background for the emergence of Harappan Civilization c. 2500 B.C.
Information about the Harappan Civilization has greatly in- creased over the past decade. More than 400 sites are known in northwestern South Asia and a single, remote site, Shortugai (37°ree;19'N69°ree;31'E), on the Amu Darya in northern Afghanistan, has been excavated. Consequently, an abundance of descriptive data concerning Harappan sites and artifacts has been gathered. While a revised description would embellish that of Chapter II and be more extensive, the two would not differ dramatically. On the other hand, theoretical paradigms concerning Harappan culture have changed significantly, altering our understanding of early South Asian cultural history. The focus here will be on these recent in- terpretive changes rather than new descriptive data. More specifi- cally, new interpretations concerning Harappan origins, the nature of Harappan sociopolitical organization, and the relationship to de- velopments in the Early Historic Period will be discussed.
Although no consensus about Harappan origins exists, few scholars still subscribe to early theories about diffusion from a western source. The two current paradigms are: (1) that Harappan culture reflects a gradual (c. 3500–2500 B.C.) evolutionary devel- opment from previous Indus Valley groups, especially Kot Dijian; or (2) that Harappan culture, with all its complexity, reflects a series of dramatic and rapid—even "paroxysmal," according to one scholar—cultural changes (c. 2600–2500 B.C.) affecting ear- lier Indus Valley groups. Excavated sites that might resolve this issue are rare. While certain characteristics link earlier groups with Harappan (see the discussion in Chapter II of Kot Diji and Kali- bangan), no gradual, cumulative, culture change sequence has been identified. At most sites, stratigraphic "discontinuities" separate Harappan from earlier occupations, a situation again documented by recent excavations at Bala Kot (25°ree;27'N66°ree;42'E), a Makran Coast site near Karachi. Only the older Amri excavations provide evidence of a possible direct transition between the earlier Amrian and Harappan occupations and the stratigraphic context suggests it took place within a brief period. Unfortunately, Amri's strati- graphic complexities, combined with an inadequate chronology for the site, leaves the issue unresolved. Current excavations at Har- appa, Nausharo (near Mehrgarh 29°ree;28'N67°ree;41'E), and perhaps Dholavira (23°ree;53'N70°ree;13'E, on an island in the Rann of Kutch), plus additional dates from late Mehrgarh occupations, may even- tually clarify this issue. Whether the change was gradual or "pa- roxysmal," it was completed by c. 2500 B.C. in the central Indus Valley. By c. 2400 B.C. Harappan settlements were also estab- lished in eastern Punjab and Gujarat. The cultural processes re- sponsible for the formation and spread of Harappan culture remain debatable. Explanations predicated on population growth, subsis- tence resource competition, external and/or internal trade, and a complex, aggressive sociopolitical organization continue to be dis- cussed.
Much has been written, based largely on assumption and spec- ulation, about Harappan sociopolitical organization. Since its dis- covery, Harappan culture has been characterized as an "ancient civilization" and assumed to have those traits associated with such civilizations found elsewhere, especially that of Mesopotamia. One major characteristic of ancient civilizations, and presumably of the Harappan, was a centralized, hierarchically structured sociopoliti- cal organization equivalent to that of a unified state. Discussions of Harappan culture frequently refer to a Harappan Empire or an ultra-conservative theocracy headed by a "priest-king," to men- tion but two common reconstructions. Recent research and in- terpretive rethinking about certain aspects of Harappan culture in- dicate that such previous reconstructions are no longer tenable for the following reasons:
(1) the "homogeneity" of Harappan material culture and settle- ment layout has been greatly exaggerated. Certainly, Harappan material culture demonstrates a unity unequaled in the ancient world; but synchronic and diachronic variations are definable in the ar- chaeological record. Moreover, some scholars now suspect this "unity" reflects a social context in which craft production was tightly controlled by related, but widely dispersed, kinship units which may have been developing organizational traits now asso- ciated with castes. There is, however, considerable debate about this hypothesis.
(2) After seventy years of research, no Harappan building can be conclusively designated as a temple or palace. Even the func- tional nature of what had been taken for "granaries" has been