NOTES
1 Edens 1993; Hiebert 1994; Lamberg-Karlovsky 1986a, 1986b; Tosi 1986a, 1986b; Possehl 1996c, 1997b; Potts 1990; Mery 2000.
2 Kohl 1975.
3 Kohl 1975: 29—30.
4 Sauer 1952: 36.
5 Tosi 1986b; Potts 1990.
6 Possehl 1986a, 1997b; Weber 1990, 1991.
7 Cleuziou 1982.
8 Childe 1950.
9 Possehl 1997b: 96—97.
10 Possehl 1996c: 136—38, 143.
11 Peake 1928; Weisgerber 1984.
12 Heimpel 1982, 1987; Possehl 1996c: 135—36.
13 See Possehl 1996c: 136 for references.
14 Possehl 1986b.
15 Possehl 1996c: 138—45.
16 Ratnagar 1981: 39.
17 Crawford 1973.
18 See Possehl 1996c: 183—84 for illustrations and documentation.
19 Kenoyer 1984.
20 Meadow 1979: 297.
21 For example, Pabumath, Y. M. Chitalwala, personal communication 1988.
22 Belcher 1991: 114.
23 Cleuziou and Tosi 1994.
24 See Collon 1996 and Possehl 1996c: 148—52 for other reviews.
25 Possehl 1996b: 150—52.
26 Parpola 1994b.
27 Reade 1979.
28 Reade 1979.
29 Reade 1979: 8—23; pls. 9 and 10.
30 Chakrabarti 1982.
31 Mackay 1943: pl. XCIII, no. 14.
32 Chakrabarti 1982: 265.
33 Kohl 1975.
34 See Bokonyi 1985 for a recent discussion.
35 Frankfort 1936: 434, 1956: 19, fig. 9, date based on Reade 2001.
36 Dales 1968a: 19—21; Possehl 1994, 1996c: 164—65.
37 Possehl 1994.
38 Possehl 1996c.
39 Dales 1968a; Possehl 1996c: 167.
40 See Edens 1993: 334—338 for discussion and illustrations.
41 Carter, in press.
42 Blackman, Mery, and Wright 1989
43 Wright 1991: 80.
44 Possehl 1996c: 168—71.
45 Possehl 1996c: 173; Possehl 1986b: Mehi II.I.2a.
46 Gadd 1932; see also Parpola 1994b.
47 Srivastava 1991.
48 Mackay 1937—38: 343, 1943: 148, pl. 15.
49 Possehl 1996c: 176—77.
50 Possehl 1996c: 177.
51 Vats 1940: pl. CXXV, 1; Woolley 1934: pl. 159b; Mackay 1929: 169, pl. XLIII, 1—8.
52 Vats 1940: 390, pl. CXXV, 34, 36; Mackay 1937—38, vol. II, pl. C, 3.
53 Mackay 1937—38: pl. CXX, 27; Santoni 1984: 53; Schmidt 1937: pl. LII, H 2710 and H 3247; Pottier 1984: 149; Wheeler 1947: 80.
54 See Possehl 1996c: 179—80.
55 Parpola 1994a: 246—49; Possehl 1996c: 180—81.
56 Possehl 1996c: 165—66; Rao 1985: 549, fig. 17.
57 Woolley 1934: pl. 141b.
58 Pande 1971.
59 Pande 1971: 314.
60 Tosi 1983.
61 Tosi 1983: 309—16.
62 Scheil 1925.
63 Crawford 1973.
64 Basaglia 1977; Hakimi 1997; Santoni 1984; Gupta 1979: vol. II, 194—204; Masson 1988; Hiebert 1994.
65 Tosi 1977: 47—48.
66 Francfort 1989.
67 Tosi 1988: 66.
68 Shaffer 1987.
69 Masson 1988: pl. XXIX, nos. 1—2.
70 Mackay 1937—38: pl. CXXVIII, no. 15.
71 Masson 1988: 8.
72 Sarianidi 1998: fig. 17, nos. 8, 9, and fig. 21, nos. 6, 7.
73 Casal 1961: fig. 36.
74 Dupree, Gouin, and Omer 1971; Tosi and Wardak 1972.
75 Hiebert 1994.
76 Hiebert 1994.
77 Hiebert 1994: 161—63.
78 Hiebert 1994: 164.
79 Hiebert 1994: 176.
80 Kennedy 1995.
81 Marshall 1931i: pl. CLVIII, no. 1; Mackay 1937—38: pl. C, no. 3; pl. LXXIV, nos. 18—19; pl. CIII, no. 15; pl. CXX, no. 27.
82 Mackay 1937—38: pl. CVI, nos. 3—5; pl. CVII, nos. 5, 15; pl. CXL, no. 35; pl. CXI, no. 59; pl. CXLI, no. 17, pl. CXLII, nos. 29, 130; Marshall 1931i: pl. CLV, nos. 36—47; pl. CLV, nos. 31—33; pl. CLVIII, no. 12;
83 Marshall 1931i: pl. CXL, no. 16.
84 Amiet 1986: fig. 202.
85 Marshall 1931i: pl. C, nos. 1—3.
86 Ardeleanu-Jansen 1991.
87 Vats 1940: XCI, no. 255; pl. CXXV, no. 36; pl. CXXXIV, no. 4; pl. CXXXVII, no. 8; pl. CXXXVIII, nos. 2—4; pl. CXXXIX, no. 84.
88 Jarrige and Hassan 1989.
89 Schmidt 1937: pls. LXI, LXII.
90 Masson and Sarianidi 1972: pls. 26—28; Fairservis 1956: 224, 229.
91 Dales 1977.
CHAPTER 13
The Transformation of the Indus Civilization
INTRODUCTION
The abandonment of Mohenjo-daro and many other sites of the Indus Civilization in the early second millennium is a well-known archaeological observation.1 This is a time that has come to be called the Posturban Harappan because the focus of change seems to have been centered on urbanization and the other features of the Indus ideology. There is essentially no habitation at Mohenjo-daro between approximately 1900 B.C. and the Kushan Period in the early centuries of the common era, when a Buddhist monastery and stupa were built on the Mound of the Great Bath. Many other Indus sites were abandoned at about 1900 B.C., especially in Sindh, Baluchistan, and Cholistan. The size and complexity of the city of Harappa was reduced at this time. The few people who lived there interred at least some of their dead in the famous Cemetery H. Judged by its small size and lack of signs of internal complexity, Harappa was not a city during Cemetery H times. Ganweriwala, Dholavira, and Rakhigarhi, the other Indus cities, also dissolved as urban centers at the end of the Mature Harappan.
There was an abandonment, or severe depopulation, of a number of important Indus settlements including Chanhu-daro, Kot Diji, Balakot, Allahdino, Kulli, Mehi, Nindowari, Nausharo, Kalibangan, Ropar, Surkotada, Dholavira, Desalpur, and Lothal. There was also a disruption in the Indus economy. The production of a wide range of special materials, many of which seem to have been luxury items, was curtailed (e.g., long barrel-cylinder beads, etched beads generally, inscribed stamp seals). The art of writing was no longer practiced. Long-distance trade was reduced, although the peoples of the BMAC were in evidence in the Greater Indus region.
The distribution of the human population in northwestern India and Pakistan shifted in a significant way. Sindh, the Kulli Domain, and Cholistan were largely abandoned by settled farmers, but there was an increase in the number of settlements in the Punjab, Haryana, western Uttar Pradesh, and northern Rajasthan. The Harappans in Gujarat remained remarkably stable. Posturban times witness a return to a cultural mosaic not unlike the one found during the Early Harappan.
There are a number of theories concerning the transformation of the Indus Civilization. The most important of these are reviewed here.
WHEELER’S ARYAN ARMY
The first coherent “theory” concerning the eclipse of the Indus Civilization was put forward by Wheeler.2 He held that the Rgveda could be read in part as an historical document and that the conflicts described there were between what he thought of as newly arrived Aryan warriors and the indigenous Indus peoples. This is stated most clearly in Wheeler’s report on his 1946 excavations at Harappa:
The Aryan invasion of the Land of the Seven Rivers, the Punjab and its environs, constantly assume the form of an onslaught upon the walled cities of the aborigines. For these cities the term used in the Rgveda is pur, meaning a “rampart,” “fort” or “stronghold.” One is called “broad” (prithvi) and “wide” (urvi). Sometimes strongholds are referred to metaphorically as “of metal” (aynsi). “Autumnal” (saradi) forts are also named: “this may refer to the forts in that season being occupied against Aryan attacks or against inundations caused by overflowing rivers” (Macdonell and Keith 1912: vol. 1, p. 538). Forts “with a hundred walls” (satabhuji) are menti
oned. The citadel may be made of stone (asmamayi): alternatively, the use of mud-bricks is perhaps alluded to by the epithet ama (“raw,” “unbaked,” Rgveda, IV, xxx, 20; II, xxxv, 6). Indra, the Aryan war-god, is puramdara “fort-destroyer” (Rgveda II, xx, 7; III, liv, 15). He shatters “ninety forts” for his Aryan protégé, Divodasa (Rgveda I, cxxx, 7). The same forts are doubtless referred to where in other hymns he demolishes variously ninety-nine and a hundred “ancient castles” of the aboriginal leader Sambar (Rgveda II, xiv, 6; II, xix, 6; IV, xxvi, 3). In brief, he “rends forts as age consumes a garment” (Rgveda VI, xvi, 13).
Where are—or were—these citadels? It has in the past been supposed that they were mythical, or were merely “places of refuge against attack, ramparts of hardened earth with palisades and a ditch” (Macdonell and Keith 1912: vol. 1, pp. 356, 539). The recent excavation of Harappa may be thought to have changed the picture. Here we have a highly evolved civilization of essentially non-Aryan type (Marshall 1931h: 110—112) now known to have employed massive fortifications, and known also to have dominated the river-system of north-western India at a time not distant from the likely period of the earlier Aryan invasions of that region. What destroyed this firmly-settled civilization? Climatic, economic, political deterioration may have weakened it, but its ultimate extinction is more likely to have been completed by deliberate and large-scale destruction. It may be no mere chance that at a Late Period of Mohenjo-daro men, women and children appear to have been massacred there (Mackay 1937—38: 94—95, 116—117, 172). On circumstantial evidence, Indra stands accused.3 (original citations included)
Although Wheeler succeeds in presenting the drama of his theory with “Indra stands accused” as its capstone, this theory can be critiqued from a number of perspectives: chronological, cultural, historical. The chronology of the Aryans in South Asia is far from settled.4 Although Aryans might have entered the Subcontinent well before the composition of the Rgveda, most scholars now date the codification of this liturgical text to about 1000 B.C., almost a millennium after the transformation of the Indus Civilization. Therefore, no one should assume that the Vedic Aryans were on the scene to be active participants in the transformation process examined here. Also, it has been argued here, along with Dales and Kennedy, that the skeletons found in the upper layers of Mohenjo-daro are actually hasty interments, not the remains of massacre victims. 5 Finally, the historical content of the Rgveda has not yet been fully, and convincingly, defined. There is a reasonable supposition that there is important historical information in this liturgical text, but just where within it, and how significant this history might be, is still unsettled.
There is some doubt that this theory is a product of Wheeler’s original thinking. Krishna Deva and B. B. Lal have told me that the “Aryan hypothesis” seems to have been developed on information Wheeler took from V. S. Agrawala, a member of the Harappa excavation team in 1946. Agrawala was the tour guide and lecturer for visiting dignitaries and spoke of the Rgveda and references in it to warfare and the conquest of towns and cities, speculating that this might have been a description of the demise of the Indus Civilization.
WEARING OUT THE LANDSCAPE
A second theory on the Indus transformation is ecologically based. The brick kilns of Mohenjo-daro and Harappa have been invoked as a part of a proposal that the peoples of the Mature Harappan were wearing out their landscape. Millions of baked bricks went into the building and rebuilding of Mohenjo-daro. Millions of tons of firewood went into the baking of them. With allowance for the arrival of floating timber from the upper reaches of the Indus, this implies a widespread deforestation of the surrounding region.6 This notion has been widely critiqued.7 Four hundred acres of gallery forest is all that would have been needed to rebuild Mohenjo-daro every 140 years.
Other natural causes have been put forward for the eclipse of the ancient cities of the Indus, including the vagaries of the Indus River.
AN AVULSION OF THE INDUS RIVER: THE POSITION OF LAMBRICK AND MUGHAL
The Indus River in Sindh is a mature stream that frequently changes it course. Much river training has been done on the Indus in modern times, so these course shifts are not as apparent today as they were through the nineteenth and early twentieth centuries. But they are a powerful part of the history of Sindh. Two archaeologists have suggested that a dramatic shift in the course of the Indus led to the abandonment of Mohenjo-daro.8
An avulsion of the Indus River might explain the abandonment of Mohenjo-daro, but there was a good deal more happening in the Greater Indus region at the turn of the third millennium than the abandonment of this once great metropolis. This leads those who use an avulsion of the Indus and the abandonment of Mohenjo-daro as an element in understanding the transformation of the Indus Civilization to revert to some form of domino effect to account for the eclipse generally: “as Mohenjo-daro went; so went the Indus Civilization.”
Mackay also believed that Mohenjo-daro might have been abandoned because of an avulsion of the Indus River, which was noted by his wife, Dorothy, an anthropologist, in another context.9
Another theory has a clearer intellectual genealogy, coming from an American hydrologist, R. Raikes, joined by Dales, elaborating an idea first presented by M. Sahni.10
THE RAIKES/DALES DAM
The Indus River has reemerged as a dynamic factor in the process of Harappan cultural change, due largely to the work of R. Raikes.11 He hypothesizes, along with G. Dales, that the waters of the Indus River were impounded by a natural dam in the vicinity of Sehwan (figure 1.2) and that Mohenjo-daro and other sites were so disrupted that it led to, actually “caused,” the abandonment of the city and the eclipse of the Indus Civilization. To support this proposition, they claim that the normal behavior of a big river simply cannot account for the evidence of massive flooding at Mohenjo-daro.
There are a number of critiques of this hypothesis.12 Wasson has noted that the unconsolidated sediments of the Indus floodplain, which would have been the substance of the dam, have very little structural integrity and could not have withstood the static pressure of the impounded Indus River.
This general theory was first proposed by Sahni, who noted the presence of thick, bedded alluvium at two hillocks south of Hyderabad, Sindh.13 This alluvium contained freshwater shells, one of which suggested prolonged submergence in fresh water. Sahni thought that this thick alluvium might be accounted for by a dam, perhaps at more than one place, across the Indus.
The forming of a large natural dam across the Indus may at first seem so unusual as to be unrealistic. But in 1819 an earthquake caused the formation of a ridge called the “Allah Bund” in northern Kutch, along the Sindh border, which temporarily dammed the minor eastern courses of the Indus system (figure 1.2). Interestingly, the Allah Bund was breached by the first significant flood against it, sustaining Wasson’s thoughts that unconsolidated alluvium is not strong enough to impound the Indus or even its minor extensions.
Another point of disagreement lies in the consideration of historical process. Is it necessary to believe that just because the Indus River was impounded, the Indus Civilization came to an end? It might be a possibility, but it hardly satisfies the “sufficient and necessary” criteria for successful historical explanation. Even if the waters of the Indus River had been impounded, the historical consequences might have been different from those proposed by Raikes and Dales. The impounding of the river would have to be seen as an immediate cause of change that resonated with some deeper, structural aspect of the Mature Harappan sociocultural system. If the Indus Civilization succumbed to new and unpredicted riverine forces, the explanation sought by historians and social scientists lies not directly with geomorphologic matters, but with the internal structure of the Indus way of life. The “flaw” that would have led to such catastrophic sociocultural change is not to be found within the natural world of geomorphology but within the human context of the Indus Civilization, its society, and culture.
NEW DATA FROM
THE LOST SARASVATI
Archaeological research along the now largely dry beds of the ancient Sarasvati and Drishadvati Rivers has led to the discovery of a large number of archaeological sites.14 Mughal’s recording of over 170 Indus sites on the Pakistan side of the border represents one of the most impressive archaeological feats of this century in the Subcontinent. When placed within the context of the distribution of Indus sites generally, they lead to a consideration that Cholistan may have been the (or one of the) principal region(s) for Mature Harappan grain production, their “bread basket,” as it were. Mughal’s findings are the most amenable to quantification at the moment, and a long-term trend in settlement counts can be illustrated by using his materials. The data are given in table 13.1.
The Indus Civilization Page 41