The Indus Civilization Read online

  INDUS SETTLEMENT PATTERNS AND SUBSISTENCE

  Introduction

  This section presents the environmental and human dimensions in the natural world of the Greater Indus region. In chapter 1, the cultural/natural regions of the Indus Civilization, called domains, were presented, along with the nature and history of the two principal rivers, the Indus and Sarasvati. What follows is a review of Indus settlement patterns and the subsistence practices of the Indus peoples.

  Settlement Patterns

  There are 1,052 settlements of the Indus Civilization known today.24 They range from tiny places one-tenth of a hectare or less in size, to Mohenjo-daro and Harappa, both nearly 100 hectares. No one has successfully demonstrated that the settlements of these people can be rationalized into a three- or four-tier system that is hierarchically arranged. While there are cities and towns, some of the best-known places are actually quite small (e.g., Lothal at 4.8 hectares, Chanhu-daro at 4.7 hectares). There is very good evidence for village farming communities. Many of the small sites, with thin scatters of pottery and no signs of permanent architecture, are interpreted reasonably as pastoral camps.

  The Indus “Breadbasket”

  During the Mature Harappan, the Sarasvati (Ghaggar-Hakra) River terminated in an inland delta near Fort Derawar. There are about 140 Mature Harappan sites in the vicinity of this delta, making it the most densely settled area of the time. There is actually an interesting succession of settlements in the Cholistan Domain. Dense occupation during the Indus Age begins with 82 Hakra Wares sites. During the Early Harappan, occupation drops, with only 20 Kot Dijian settlements. This is followed by the final “bloom” during the Mature Harappan, with approximately 140 sites.

  These data may indicate the variable flow of the Sarasvati. The flow would have been strong during the Hakra Wares Phase, dropped off during the Kot Dijian, and then strengthened again during the Mature Harappan. The final drying up may be documented by the progressive drop-off in settlement during the Posturban (Cemetery H, 40 sites) and Painted Gray Ware (14 sites) Phases,25

  Concluding Observation

  Indus settlements tend to cluster, with the open space between them occasionally filled with a few individual settlements. I have suggested that the open space was probably occupied by pastoralists, nomadic and otherwise.26

  The nine Indus sites that measure 50 hectares or more are given in table 3.1. Table 3.1 omits some large sites in Saurashtra because of lateral (spreading) stratigraphy. This also might be the case at Nagoor and Tharo Waro Daro as well. Also excluded are the giant sites in Bhatinda District (Lakhmirwala, Gurnikalan One, and Hasanpur Two).27 The raw data on their size indicate that they were 225 hectares, 144 hectares, and 100 hectares, respectively. Further investigation of these sites may illuminate why three cities developed so close together during the Mature Harappan. Perhaps lateral stratigraphy and/or the spreading of remains through agricultural activities is the explanation for the unusual sizes of these sites.

  Table 3.1 Mature Harappan sites of 50 hectares or more

  Site Province/State Estimated size (hectares)

  Mohenjo-daro Sindh 100

  Harappa West Punjab 100

  Rakhigarhi Haryana 80

  Ganweriwala West Punjab 80

  Dholavira Gujarat 60

  Nagoor Sindh 50

  Tharo Waro Daro Sindh 50

  Lakhueenjo-daro Sindh 50

  Nindowari Baluchistan 50

  Note: See Possehl 1999b: 727—845 for documentation.

  Subsistence

  The Indus peoples were mostly farmers and herders. Barley seems to have been the principal food grain, except in the Sorath Domain, where the people were cattle keepers par excellence who also raised goats, sheep, water buffalo, and a variety of crops. Cattle remains are consistently one-half or more of the faunal remains from Indus sites, no matter how the remains are measured (e.g., by bone count, bone weight, minimum numbers of individuals). Pigs may not have been domesticated, but pig remains and figurines document their use. The Indus peoples domesticated the chicken and kept several breeds of dogs28 and possibly house cats.29 Camels may also have been domesticated. Camel remains that have been found may be either the dromedary or Bactrian species. There is a certain amount of controversy over the domestication of the horse, which I believe results from the misidentification of the remains of the Indian wild ass.30 As far as I can tell, there are lots of asses documented at Indus settlements, but no domestic horses (Equus caballus).

  Most Indus agricultural activities took place during the winter rabi season. The active floodplains and the areas directly adjacent to them were most intensely cultivated during the rabi season. Whether rice was a cultivar of significance during the Mature Harappan has yet to be determined.

  They grew dates and grapes and collected the Indian jujube. They were also great fish eaters, exploiting the rivers and lakes, especially in Sindh. Large fish vertebrae have been found at some Kutch Harappan sites. Salted and/or dried fish were traded over large distances during the Mature Harappan as documented by the presence of a marine species of catfish at Harappa.31

  Indus peoples apparently grew cotton for its fiber and perhaps for its oil. There is good evidence of the use of cotton cloth at Mohenjo-daro. Fibers were found in four contexts there (table 3.2). Additional information on cotton and a bibliography are also available in the references.32

  Table 3.2 Cotton at Mohenjo-daro

  Two silver vases, originally wrapped in a cotton bag Sahni 1931a: 194

  Cord wound around a copper blade Mackay 1937–38: 441, DK 8376

  Fabric adhering to razor Mackay 1937–38: 441, pls. CXVIII, no. 7

  Fine cord wrapped around a copper rod Mackay 1937–38: 441, DK 5844

  Gossypium-type pollen has been found at Balakot. This sample came from layer 4, which places it rather late in the Mature Harappan sequence. McKean argues as effectively as she can that this pollen resulted from cultivated cotton.33

  Lorenzo Costantini has observed that cotton was used at Mehrgarh during Period II. Several hundred charred seeds were associated with a fireplace in one of the compartmented buildings there.34 This was a provisional identification, included in the preliminary report for the 1978—1979 field season, and the work on these seeds needs further study. A single carbonized cottonseed was also found at Hulas, in the Late Harappan or Posturban Phase.35

  African millets appear in the Indus Civilization. The plants, with their Hindi-Urdu names, are sorghum or jowar, pearl millet or bajra, and finger millet or ragi. The importance of these plants is that they are summer grasses that prosper during the southwest monsoon, unlike wheat and barley, which are winter grasses that do not thrive as monsoon crops. The millets thus led to double or year-round cropping and were important, if not critical, additions to the prehistoric food supply.

  The appearance of these plants coincides with the beginnings, or at least the expansion, of significant maritime activity in the Arabian Gulf and Indian Ocean. It is proposed that an extension of this maritime activity took Indus sailors at least as far as the southern end of the Red Sea and possibly farther south along the east coast of Africa. It is in this environment that they came in contact with the millets, integrated them into their food supply, and eventually carried them back home to the Subcontinent.

  E. H. Aitken has noted that “famine is unknown in the Indus Valley”;36 however, scarcity does visit the desert areas. The productivity of the nineteenth-century agro-pastoral system produced so much grain that horses were fed with rice instead of grass. In Sindh, the traditional method of cultivation, prior to the large-scale canalization of the nineteenth century, was to find and plant, often without plowing, the fresh alluvium, especially in small side channels of the annual flood. The seeds of wheat and barley were often broadcast and left to mature on their own.37 Barley and wheat were sown in November and December and harvested in April.

  One of the unresolved issues is whether significant canal irrigation works were
associated with the Indus agricultural regime. The question is not whether they were able to dig ditches to drain swamps and move small amounts of water to and from their fields, but rather the bigger problem of moving water from the active river course(s) to different, higher, riverine environments, outside of the valley that would have been flooded naturally. This is an important, twofold issue. It has to do with the question of the Indus population’s capacity to gather and manage large labor forces, as well as the salinization of agricultural fields with its consequent drop in productivity.

  There are some things that are known about irrigation in the Indus Age. At Mehrgarh, “The charred seeds of wheat and barley . . . grow only on irrigated fields, also were collected from the ashy layers” of Period II, according to L. Costantini.38 This is followed with evidence for the existence of a ditch of significant size that was filled with Mehrgarh Period IV trash.39 This seems to indicate a date of Togau or Kechi Beg Phase use.

  L. Flam proposes that there were three different forms of irrigation in Sindh, Kohistan, and along the Kirthar front during the Early Harappan and Mature Harappan.40 The first used the natural flooding of a hill stream to irrigate land. The second form, documented at the site of Kai Buthi, made use of small, shallow ditches to gently guide spring water out onto a flat area that was used for cultivation. The third irrigation practice used by the Sindhis involved check dams, either as low, linear mounds of earth across broad fields called “bunds” or as larger stone structures along streams called gabarbands. The gabarbands are proper dams made of stone and built about half way across hill torrents and small rivers. They are designed to capture both soil and water.41

  There was some diversity in the Mature Harappan subsistence regime. The people of the Sorath Harappan in Gujarat were heavily dependent on pastoralism and relied on a suite of hardy, drought-resistant plants.

  In all, nearly 80 different plant species were identified from Rojdi. Of these taxa, only a small number were found in all phases of occupation (Rojdi A, B, and C) and in amounts or in densities which imply significant use of the plant. A general plant-based subsistence system can be identified at Rojdi, which is maintained throughout the Harappan portion of the occupation. In this system, Rojdi was a food-producing settlement occupied throughout the year where cultivation, the use of domesticates, and various pastoral activities were being performed locally, and where, to a lesser extent, plant gathering and hunting were also being practiced. The bulk of cultivation was centered on the summer monsoon and involved millets, although winter cultivation increased in importance during the occupation. In addition, the plants being exploited never needed intensive human involvement and were all hardy, drought resistant species.42

  Baluchistan and the Northwest Frontier (the Kulli and Northwestern Domains) would have been places suitable for the cultivation of winter wheat and barley. The hills and valleys of this mountainous area are also good pastureland and would have been home to large numbers of domesticated animals, especially sheep, goats, and cattle. Transhuman pastoral nomadism would have linked the highlands to the lowlands in a symbiotic partnership.

  The abundant remains of cattle and of cattle imagery generally seem to inform us that pastoralism, sometimes involving nomadism, was an important part of Indus life.43 The Punjab, or Harappa Domain, would have been home to both pastoral nomads and farmers. The ethno-historical record informs us that life there prior to the excavation of canals during colonial times was a duality of pastoralism and agriculture. The pastoralists lived in the higher areas between the rivers, and the farmers settled in the river entrenchments,44 as seen in this nineteenth-century account:

  Thus the district may be divided into two distinct portions—one the cultivated portion or des, and the other the grazing tract or bar. [Des] comprises . . . the land . . . on the banks of the rivers to the south. [Bar] contains that large, uncultivated tract which [is between the rivers]. . . . In the des we find agriculturists of settled habits, with rights and property in the soil, and deriving their chief support from their cultivation; whilst the people of the bar are graziers, leading a nomad life; possessing little or no landed property, and subsisting more on the profits derived from their cattle than their land.45

  This account highlights a great pattern in the lives of the Indus peoples as well: the complementarity of settled agriculture and pastoralism. While there must have been a variety of occupations for the peoples of the Indus Civilization, most of them would have been farmers and/or pastoralists. They produced the food that sustained the nonagricultural people in the great cities and towns of this civilization.

  A SELECTION OF THE PRINCIPAL SITES

  Ninety-seven of the 1,052 known Mature Harappan sites have been examined by excavation. Most of the work has been done at the large places: Harappa, Mohenjo-daro, Dholavira. But at least one Indus city remains untouched: Ganweriwala in Cholistan. Mehrgarh, with its very long sequence of occupations, has been much deserving of the sustained work there. Table 3.3 gives a short account of those sites of the Indus Age with five or more seasons of excavation. Harappa is far and away the place that has been intensely excavated, largely due to the substantial work done there since 1986 by the joint American-Pakistani team. Moreover, excavation alone does not tell the entire story. Michael Jansen conducted intensive, rewarding research at Mohenjo-daro for eight years, but undertook no excavation.46 It is also interesting to see that some of the best-known places of Indus life are missing from this list. Chanhu-daro, for example, famous as it is, has been subjected to only two seasons of excavation. Kulli and Mehi were barely scratched by Sir Aurel Stein when he was there in 1928. On the other hand, nine seasons of excavation have been devoted to the comparatively little known Sanghol.

  Mohenjo-daro

  For me, Mohenjo-daro is the epitome of all Mature Harappan settlements. It is about 100 hectares in size and I suspect that it was founded in the Early Harappan—Mature Harappan Transition by a group of Harappan “true believers.” The activities occurring at Mohenjo-daro were the essence of Harappan life and ideology. While each Mature Harappan settlement has its own character as a settlement or an urban center, Mohenjo-daro symbolically represents a good deal of what it meant to be Harappan.

  Table 3.3 Sites of the Indus Age with five or more seasons of excavation

  Site Seasons of excavation

  Harappa 36

  Mehrgarh 17

  Mohenjo-daro 15

  Dholavira 12

  Rojdi 11

  Kalibangan 10

  Sanghol 9

  Lothal 8

  Allahdino 6

  Rangpur 6

  Nausharo 6

  Amri 5

  Banawali 5

  Hulas 5

  Prabhas Patan 5

  There is an entire chapter on Mohenjo-daro in this book (see chapter 11). This testifies my sense of the city’s importance in understanding the Harappan Civilization.

  Harappa

  Harappa is the old Mature Harappan city, with settlement reaching back into the Kechi Beg—Hakra Wares Phase (3800—3200 B.C.; see figure 1.6). The site is situated at a place where the entrenchment of the Ravi River broadens, not surprisingly, into an area where substantial agriculture is possible. Mohenjo-daro is 600 kilometers to the southwest. The apparent size of Harappa, taken from the mounded area and associated artifact scatter, is approximately 100 hectares.

  Archaeological deposits dating to the Mature Harappan have been found under alluvium around the city, and no one is certain of the city’s exact size—it may perhaps be as large as 200 hectares. With a population density of about 200 people per hectare, and 100 hectares settled at one time, total population would have been approximately 20,000.

  There are five principal periods of occupation at Harappa, as shown in table 3.4. There is an imposing high area on the west at Harappa, surrounded by substantial brick walls. It is generally called Mound A-B. Wheeler labeled it a “citadel,” just like the Mound of the Great Bath at Mohenjo-daro. The Mat
ure Harappan citadel is another archaic thought about these cities; but not much is known of Mound A-B.

  Mound F, to the north of A-B, has a series of interesting buildings. Construction here seems to have begun late in Period 3A or early 3B. The most imposing building on Mound F has sets of parallel walls laid precisely on either side of a central road or corridor and is thought of as a granary, although this has never been confirmed by charred grain, storage vessels, or other collateral evidence. The granary at Harappa seems to have been built late in Period 3B. There is a series of circular threshing platforms to the south of the building. Their function has been determined through careful excavation of the wooden mortars in their centers, which are associated with grain husks (figure 3.5).

  Mounds E and ET, to the south, have what might be the remains of an outer city wall of substantial proportions. The configuration of this wall suggests that the valley between Mounds E and ET is an artifact of ancient city planning.

  There are two cemeteries at Harappa designated R-37 and H. They are located in the same general part of the site. The one designated R-37 is the largest Mature Harappan cemetery known to us. The dead in this place were treated in a variety of ways. Some skeletons have been found in an extended, supine position inside wooden coffins, the way many contemporary Americans are buried. Cemetery H is a burial ground for Period 5 and contains pot burials and a variety of other interments in the earth (fractional, dismembered, and so forth).