Heitzman, James;
The city in South Asia
Routledge, 2008, 283 pages
ISBN 0415343550, 9780415343558
topics: | india | south-asia | history | urban
Excavations at Mehrgarh provided convincing proof that the rise of agriculture was not the result of diffusion from the west but originated in South Asia (Jarrige 1995).
Few appreciate the antiquity of the agrarian and urban settlements represented by these mute mounds which people today plough and plant with crops, pillage for soil and bricks, or tread when they cross their doorsteps. Even the first modern historians of South Asia, while cognizant of the archaeological material lying near at hand, subordinated physical culture to textual analysis. During the nineteenth century, when British scholars oversaw the assembly of the historical disciplines, the model for archaeological reconstruction was to begin with references in textual sources and then correlate them with information gleaned through ground surveys and excavations. The most important chronological baselines were the ministries of Gautama, who established Buddhism, and his immediate predecessor, Mahavira, who promoted Jainism; information gleaned from the texts associated with their religious traditions suggested that these men lived around the middle of the first millennium BCE. The first Director of the Archaeological Survey of India, Alexander Cunningham (1814–93), enjoyed considerable success in following references from such textual materials, or from travel accounts such as that of the seventh-century Chinese monk, Xuan Zang, in order to identify specific mounds with places appearing in literary sources. The early results of this research clearly indicated that a society of considerable urban complexity and sophistication existed in what are today northern India, Pakistan, and Bangladesh about 2,500 years ago. ...
Early layer from Mehrgarh, c.7000 BCE. walls are
built with mud bricks. [source:wiki-commons]
The first structured excavations during the 1920s at Harappa and then at Mohenjo-daro yielded artifacts that resembled items uncovered during contemporary excavations in Iraq (old Sumer) dated to the third millennium BCE. These discoveries stimulated the growth of an entirely new field in South Asian archaeology dedicated to the study of what came to be known as the Harappan (after one of its most important sites), Indus Valley, or Sindhu-Sarasvati Civilization (after the two major rivers along which its sites clustered). The nomenclature of civilization applied to this extensive archaeological complex derives not only from its literate or proto-literate character but also from the indisputably urban character of its major sites.
Excavations during the 1970s and 1980s at Mehrgarh in Baluchistan provided convincing proof that the rise of agriculture and village life, evolving into more complicated forms of urbanization, was not the result of diffusion from the west but originated in the creativity of South Asian peoples (Jarrige et al. 1995). Mehrgarh lies at the foot of the Bolan Pass, along an ancient trade route allowing travel to the highlands of Afghanistan and Iran and also to the plains of southern Pakistan, with relatively easy access to a variety of ecological niches. Archaeologists discovered here a continuous record of habitation dating back to a pre-ceramic culture in the early-seventh millennium BCE which makes this site contemporaneous with Çatal Hüyük in Turkey (Mellaart 1967) and Jericho in Palestine (Kenyon 1970), several of the world's oldest known settlements. At that point the inhabitants of Mehrgarh were already constructing permanent structures of mud bricks including houses and storage facilities designed to preserve barley, gathered by semi-nomadic groups using flint sickles and relying on hunting for part of their subsistence. The subsequent story of this site includes the evolution of increasingly sophisticated ceramic technologies, domesticated animals and more permanent settlement structures. By the late-fifth millennium the inhabitants utilized copper, manufactured wheel-thrown pots, and produced artistically fired beads. By the fourth-third millennium BCE, permanent settlements and cultivation, possibly with irrigation facilities, had replaced traces of seasonal habitation. Wheat and grapes were important crops. Traces of large-scale pottery production suggest the presence of a multi-site marketing system. Large numbers of terracotta human forms appear in and around house sites. Extensive terracing and large retaining walls indicate a movement toward monumental public architecture. Trans-regional trade seems to have been important at Mehrgarh even from its earliest phases (revealed by the presence of shell ornaments coming from the ocean, 500 kilometers distant). The later phases of occupation link directly with nearby sites such as Nausharo yielding Harappan artifacts, thus providing a sequence from the very beginnings of settled life through the first urban urbanization in South Asia. Archaeologists have identified hundreds of sites where pre-Harappan village farming communities developed the complete range of technologies (including traces of monumental walls) that would give rise to Harappan urbanization. In addition to Mehrgarh four such artifact assemblages have been identified, each associated with their primary sites: Amri and Nal; Kot Diji;DambSadaat; Sothi and Siswal (see Map 1). Investigators now describe a pre-Harappan period beginning in the late-fourth millennium BCE, when changes were occurring within multiple cultures that prepared the ground for the complexity of the mature Harappan phase between 2600 and 1900 bce. The transition to the mature phase seems to have occurred rapidly, perhaps within only a few generations (Possehl 1999; 2002: 50–3).
More than 1,000 sites display the cultural characteristics that we associate with the mature Harappan Civilization. They exist in all parts of Pakistan south of Kashmir, in Gujarat, Haryana, Jammu, Punjab, and Rajasthan in northwestern India, and even along theAmuDarya in northern Afghanistan. Most of the locations remain small, constituting village sites; if the average size of settlements during the early Harappan phase was 4.51 hectares, it increased to only 7.25 hectares during the mature phase (Possehl 1999: 555). The largest sites, however, are quite big. They include Mohenjo-daro (with a city core of about 100 hectares, and suburbs possibly covering more than 200 hectares) in Sind; Harappa (more than 150 hectares) in the center of Pakistani Punjab; Dholavira (between 60 and 100 hectares) in Gujarat; Ganweriwala (82 hectares) in Pakistani Punjab near the border with Rajasthan; and Rakhigarhi (between 80 and 105 hectares) in Haryana (Smith 2006: 109). The region called Cholistan, lying in the southeastern part of Pakistan's Punjab province, is among the most thoroughly explored regions of the Harappan civilization. Today this is an arid ecosystem, a northwestern extension of India's Thar Desert, but ground surveys and aerial photography reveal that a major river system once flowed here. Between 3,500 and 1,300 bce there was enough water to support a flourishing complex of village farming communities evolving into a four-tier hierarchy of settlement sizes. During the first centuries of occupation, nomadic occupation roughly balanced the number of more permanent agricultural settlements. By the time of the early Harappan phase (3100–2500) permanent settlements comprised 92 percent of all extant sites while nomadic campsites had shrunk to only 7.5 percent. After 2500, during the mature phase, out of 174 identified sites (most of them small villages) 45 percent were purely industrial locations for the production of bricks, pottery or metal objects while 19 percent of habitation sites also included kilns for commodity production. During this phase Ganweriwala, the pinnacle of the settlement hierarchy, became two closely associated mounds, today rising to a maximum height of 8.5 meters above the plain. After 1900 there was a dramatic drop-off in the number of sites and purely industrial zones, suggesting that the settlement complex was collapsing with the disappearance of the river system (Mughal 1997). The Vedic hymns preserve several references to the existence and subsequent disappearance of the Sarasvati River, corroborating a theory of radical transformation in this ecosystem by about 1300. The Cholistan surveys reveal, therefore, a complete sequence: the rise of village farming communities; the elaboration of more complex settlement hierarchies; and occupational specialization leading to urban growth and subsequent decline.
Mohenjo-daro, perhaps the most important archaeological site in South Asia, excavated repeatedly since the 1920s (Marshall 1931), provides a detailed view of a primate city during the mature Harappan phase, including monumental constructions within a ‘high’ town or ‘citadel’ to the west, and habitation areas in a separate ‘low’ town of greater extent to the east. The northeastern portion of the citadel mound rises 13 meters above the plain and is the location of a Buddhist monastery dating from the early-first millennium ce that was constructed on the more ancient tell. Earlier British scholarship, never far from a military interpretive framework, described massive brickwork visible on the edges of the citadel as the remains of fortification walls (Wheeler 1968: 29–31, 40, 72–7). Hundreds of soundings on the mound have indicated, however, that its height is not part of a fortification scheme, nor is it simply the result of accumulated debris, but is a planned feature; before occupation occurred, a mobilized workforce had constructed massive retaining walls for a giant foundation that raised the bases of buildings 5 meters above a plain flooded regularly by the nearby Indus. The citadel mound has yielded the most impressive architectural remains – all constructed of mud or burnt bricks – found in any mature phase settlement. These include the Great Bath, perhaps the most photographed of all Harappan monuments: a watertight pool measuring 12 meters north-south and 7 meters east-west, capable of retaining 140 cubic meters of water. Water came from a well in a room on the east and outflow went through a drain consisting of a 1.8-meter-high drain roofed with a corbelled vault. The water was accessible by a stairway on the north side of the bath. Brick colonnades that probably supported a wooden superstructure, including screens or window frames, surrounded the bath. A public street enclosed the entire complex. The ‘granary’ to its west consisted of a foundation measuring 27 meters north to south and 50 meters east to west divided into 27 brick plinths, in nine rows north to south and three rows east to west separated by passageways, that once supported wooden pillars or beams. An ‘assembly hall’ or market in the southern part of the citadel mound featured 20 brick foundations arranged in four rows of five each. More public buildings seem to lie under the Buddhist monastery. On the eastern mound or ‘low’ town, which also may have rested in part on a massive artificial foundation (Dales 1965: 148; Jansen 1978), the most extensive horizontal excavations of any Harappan site have revealed the plans of contiguous habitation blocks. Several boulevards running roughly north-south (including a wide thoroughfare named ‘First Street’ by the excavators) provided a means of walking directly through the entire settlement. Smaller, parallel lanes ran north-south for shorter distances, intersected by other smaller lanes running east-west – a generally cardinal orientation. Obvious differences in house sizes and layouts betoken distinct gradations in class or status. In the northeastern habitation block a large, coherent complex of rooms demonstrates an unusual sturdiness of construction, perhaps evidence of a major public building, while another complex was labeled a ‘palace’ by the excavator (Mackay 1938: 41–70, 148–51). Building interiors were invisible from the street, but entrance to a household compound would present the visitor in some cases with a courtyard surrounded by rooms, or in other cases with a simple cooking and sleeping area in one or several rooms. The remarkable preservation of house walls suggests that windows, if present, existed high on the walls; examples of artworks showing windows indicate that wood may have been the primary component in window construction and ornamentation (Kenoyer 1998: 57–8). Roofs, undoubtedly flat, were probably constructed of wooden beams (spans seldom greater than 4 meters) and mud plaster. Some remains of stairways at this and other sites demonstrate that people had access to their roofs. There is clear evidence of delineated bathing areas within homes, especially some of the larger units; outside, carefully constructed brick drains running through public space intersected with pipes coming from homes to draw sewage out of the city and into sumps that could be dismantled for periodic cleaning (Jansen 1993). Water supply for the entire city came from an estimated 700 wells. With the possible exception of the elaborate drainage systems, the plan of the residential area at Mohenjo-daro and the hypothetical reconstruction of house architecture demonstrate close similarities to housing types still used in this part of the subcontinent.
The plan of Harappa, where excavations started slightly earlier than at Mohenjo-daro, includes an alignment of two main mounds (excluding the area of the modern village, which lies above another section of archaeological remains) on massive retaining walls and basements pierced on at least one side by a gate. At Harappa one finds another ‘granary,’ or two blocks of parallel brick basements that seem to have supported a wooden superstructure covering an area of 46 by 17 meters. At the site of Kalibangan along the bank of the former Ghaggar River, once again we find two distinct mounds: a heavily fortified section in the west measuring 120 by 240 meters, containing possible ceremonial platforms; and a separate, fortified area to the east measuring 240 by 360 meters encompassing habitation areas. If Harappan urbanization up to this point seems to demonstrate a standard plan, juxtaposing high-status public space in the west with a lowerstatus but differentiated population living in the east, town planning from other sites produces a more variegated pattern. At Dholavira, for example, a larger enclosing wall with bastion-flanked gateways encompassed a ‘lower’ town and a walled ‘middle’ town in turn flanked by a smaller, fortified ‘castle’ and adjoining ‘bailey.’ At the smaller sites of Banawali and Surkotada, a single outer fortification surrounded two subsidiary divisions separated by bastioned walls with gates. It appears, then, that the Harappans were concerned about issues of security and mobilized labor resources to establish spatial segregation within their settlements, but the architectural alignment varied widely from one site to the next. There were shared concerns, but not standard templates, in the careful planning of their towns (Chakrabarti 2004).
A crop package of wheat, barley, lentils, peas, chick peas, and flax was well developed by the late-third millennium bce. Several species of millets that originated in Africa (sorghum or jowar, finger millet or ragi, pearl millet or bajra) seem to have reached South Asia by the early-second millennium bce. Other millets and fruits, the Bactrian camel and the horse arrived from Central Asia. At most sites, a substantial proportion of the population must have made their livelihood raising sheep, goats, and cattle, or interacted with nomadic or semi-nomadic groups specializing in the care of animals (Weber and Belcher 2003). The site of Lothal in Gujarat provides some clues concerning the organization of trade routes that made possible the movement of goods at the regional level or between South Asia and points west. This small center (4.8 hectares) seems to have served as a warehousing or trans-shipment point complete with an artificial harbor and quay. A seal found here came from Dilmun, the name used in Sumerian records for the area around Bahrain in the Persian Gulf. Sumerian sources describe trade with Dilmun but also mention Meluhha, a term denoting the Harappan cultural region. It seems that commercial groups from Meluhha settled down in Mesopotamia for extended periods during the late-third millennium. Archaeological work in Oman and between Dubai and Abu Dhabi has found further evidence of Harappan traders on site and the direct borrowing of Harappan ceramic, metallurgical, and seal-cutting technologies around 2300 bce. Meanwhile, on land routes, Shortugai in northern Afghanistan appears to have been a purely colonial project allowing Harappans access to the only known sources of lapis lazuli. The high level of artistry evident in the processing of luxury and functional items made from metal (gold, silver, copper-bronze, lead, electrum), semi-precious stones, and sea shells, often found hundreds of kilometers from the points of origin for their raw materials, testifies to the vibrancy of a manufacturing sector closely tied to wide-ranging commercial networks (Allchin, and Allchin 1997: 167–76; Tosi 1991). The large site of Dholavira, located on a rather unproductive island in the present Rann of Kutch, is best conceived as an entrepot linking traders to the regional economy of early Gujarat (Bisht 1989; Chitalwala 1993; Dhavalikar 1993). Terracotta models of ox-carts and graphic representations of riverboats, very similar to designs still in use, indicate control over the transportation of bulk goods.
During the mature Harappan phase a widespread standardization of technology, craft production, and administration occurred throughout the Harappan sphere of exchange. The concentration on water control and bathing in household construction (including intramural drains, vertical pipes, and chutes through walls to streets) and public drainage systems remained a defining feature of their urban planning. In regions where bricks were the main building material, builders almost universally applied size dimensions of 1:2:4. Widely distributed motifs appearing on pottery seem to have accompanied mass production for extended markets. The presence of a graduated series of weights, following in their lower denominations a binary system of 1, 2, 4, 8, 16, 32, etc., foreshadows a measuring system that enjoyed a very long history in South Asia and proves that a single means for conducting commercial transactions existed in multiple sites. Most important is the evolution of stamps decorated with geometric forms from the pre- and early Harappan phases into beautifully crafted seals that displayed detailed pictographs and a completely original script during the mature phase. Although 90 percent of the known 3,000 inscriptions come from Mohenjo-daro and Harappa (Allchin and Allchin 1997: 191–2), seals and clay seal impressions from many Harappan sites and from abroad demonstrate their widespread usefulness as commercial certificates. In some cases, the presence of obviously mythological or ritual scenes suggests that some seals had institutional or administrative uses. The remains of a signboard from Dholavira and a small collection of longer or cylindrical seals feature enough characters to suggest that the Harappan script was a functional representation of natural language, although part of it may have been an iconic system presenting logos, brands, or directional markers similar to highway signs today. In any case, the nature of the literary evidence consisting almost entirely of very short examples of script found on seals and impressions (perhaps in most cases representing the names of persons, organizations, or places) has frustrated attempts at decipherment. The combination of script and other types of standardization points toward a high degree of administrative control strongly suggestive of a state, or a grouping of city states, with clearly expansionist tendencies during the late-third millennium.
We can glimpse the daily lifestyles of the Harappans through the many personal artifacts they have left behind.... Figurines of women and ‘mother goddesses,’ more common than male figurines, often display elaborate headdresses and hair ornaments, some common to many sites, others more limited in scope and perhaps indicating differences in status or ethnicity. Some women wore turbans or head bands, others tied their hair in a double bun at the back or displayed long braids. Necklaces, chokers, and belts accompanied short skirts. Adult women wore anklets of steatite and bangles on their left arms; the presence of truncated cylindrical stone amulets, worn around the neck, may indicate married status. Men with high-status demeanors, like the so-called ‘priest-king’ of Mohenjo-daro, wore full-body robes draped over the shoulder with woven or printed designs. Men, like women, pulled their long hair into a double bun behind the head and tied it with microbead hair ornaments. Long braids, bangles, and skirts were common to men as well as women. Images of children include small terracotta images perhaps used in votive rites. Other images represent children playing with toys. We have quite a few examples of their toys in the form of terracotta tops and clay marbles, miniature cooking vessels and furniture, and numerous figurines of domesticated animals (bovines, sheep, goats, ducks, rabbits, a fighting dog with projecting collar, a begging dog with collar, bears with collars, or performing monkeys) and wild animals such as elephants, rhinoceroses, and tigers that seem to have lived in Pakistan during the third millennium. The unsophisticated clay disks of various sizes found in many locations resemble the pieces used in a game called pittu today, in which one player throws a ball to knock down a pile and another player must stack the disks in sequence by size before chasing everyone else in a game of tag. Adults indulged in more complex games using cubical dice exactly like today's versions or gaming boards that required the movement of game pieces or tokens (Kenoyer 1998: 122–67). Presented with an urban society demonstrating such sophistication and expansive energy, we must wonder why Harappan culture failed to expand further or develop even more complex forms. The archaeological record reveals a major shift around 1900 bce (when long-distance contacts with Mesopotamia seem to have ceased) away from larger urban sites and toward smaller settlements. The strong centralizing tendencies of the mature Harappan, represented by the writing system, standardized weights, and common ceramic styles, disappeared. In this ‘localization era’ (Kenoyer 1991: 370–2) a variety of artistic styles asserted themselves. Earlier catastrophic explanations for the end of urbanization that proposed human conquest or widespread geo-morphological change have given way to more nuanced regional perspectives. A hydrological explanation for Cholistan's eclipse is convincing but the simultaneous extinction of sites in Sind and Pakistani Punjab must stem primarily from socio-economic causes. The late Harappan phase also witnessed an increase in Indian settlements, indicating the possibility of major movements of population. An earlier survey in Haryana and Punjab revealed up to 160 mature-phase sites and 149 later sites; a more recent survey in Haryana alone revealed 71 mature-phase and 275 later locations, plus 130 in Uttar Pradesh (Francfort 1985: 61–2; Allchin and Allchin 1997: 214–15).
What was the urban legacy of the Harappan accomplishment? Some scholars have tried to push the chronology of the northeastern sites into the late-second millennium, when they (almost) intersect with archaeological materials tied to a second urbanization (Shaffer 1993). Others have traced possible paths of cultural influence emanating from Harappan sites in Gujarat toward central India. Once again the regional picture presents a variety of patterns. In zones directly adjoining the Harappan interaction space, such as eastern Rajasthan where copper and tin exploitation occurred very early, distinct assemblages represented by the large village sites of Jodhpura and Ahar indicate that a parallel growth of chalcolithic, non-urban cultures was occurring during mature and late Harappan times. Moving into Central India, we find a village-based Malwa culture, and moving south into Maharashtra we encounter Malwa characteristics giving way to a later Jorwe culture, in villages thriving during the second millennium bce. Simultaneous expansion of ‘neolithic-chalcolithic’ sites was occurring in a zone stretching from western Uttar Pradesh, where a village-based culture using ‘ochre-colored’ pottery evinces late Harappan characteristics, through a series of cultural assemblages stretching all the way to Bengal (Chakrabarti 1999: 205–61). One must consider these cultures as distinctly non-urban, although they laid the groundwork for agrarian regimes that would support early historical cities. Thus, while we must acknowledge that the Harappan settlements bequeathed a physical culture, resource base, and cultural styles that became standard for many later village farming communities and towns in South Asia, we must also accept evidence for the complete disappearance of urbanization for perhaps a millennium following the disintegration of the Harappan complex. ... remains one of the world's great archaeological conundrums...
... a novel cultural assemblage characterized, in part, by a pottery known to archaeologists as the Painted Grey Ware (early to mid-first millennium BCE). We know of several dozen settlements yielding this pottery that supported relatively large populations living in dense, contiguous spaces – places that we can describe as cities or towns – by 600 BCE at the latest. The elaboration of more complicated fortifications accompanied the appearance of a pottery known as the Northern Black Polished Ware (mid- to late-first millennium BCE) and punch-marked coinages issued by a variety of independent political units that culminated in the metal currency issued under the Mauryan Empire (thirdsecond centuries BCE). Textual materials (e.g. sections of the Epics) during this period project a coherent body of religious practice and philosophical inquiry, plus an expanding role for the state (Thapar 2002: 139–46).
One of the most impressive early historical cities in northern India is Kausambi, known from multiple literary sources as the capital of the Vatsa region. The location of the city on a bluff along the Yamuna River, about 45 kilometers from its juncture with the Ganga River, is one key to its strategic importance as a commercial hub. As George Erdosy has suggested (1988: 53–4), the site lay within a range of cultivable but inferior soils, the last major agrarian region on the southern edge of the interfluvial zone, but within striking distance of mineral extraction regions to the south. The primary excavator of Kausambi (Sharma 1969: 22) claimed that permanent occupation of the site began in the twelfth century BCE, with an initial mud fortification dating back to the eleventh century, but later investigators have more conservatively dated the first walls to either the seventh or fifth century bce (Erdosy 1988: 55–61; Chakrabarti 1995: 196–7). Today the site features a mound averaging more than 10 meters in height, formed from the detritus of centuries of occupation, appearing as a plateau that dominates the surrounding countryside. A wall 6.4 kilometers in length still encompasses the entire site. Individual towers are more than 22 meters high, providing exceptional views of remains of the city's moat and a separate ring of mounds about 1.5 kilometers away. An inner citadel or palace, built before the occurrence of Northern Black Polished Ware, lay in the southeastern sector of the settlement. Excavations uncovered a vast trove of artifacts and artwork, now located in the Allahabad Museum, bringing the intensive habitation of the city up to at least the fourth century ce. The dozens of smaller tells located throughout Allahabad District have been the object of an intensive surface survey by Erdosy (1988) who evaluated their size at different time periods based on the visibility of type artifacts in order to describe a four-tiered size ranking of settlements between 600 and 350 bce, evolving into a five-tiered ranking by 100 bce. At that time, although Kausambi exceeded the next largest site by a factor of nine, the number of the city's suburbs and fairly large intermediate-sized (6–50 hectare) sites in its vicinity was at its peak. Among the mid-sized neighbors was Bhita, a fortified town perhaps originally occupied in the fourth century bce. As one of the only early historical sites in South Asia that has undergone extensive horizontal excavation, Bhita is important for revealing a main thoroughfare stretching west from its southeastern gateway with flanking rows of shops backed by habitation areas (Marshall 1915).
The settlement pattern in Allahabad District stands in contrast to the situation in Kanpur District to the west, in the heart of the Yamuna-Ganga basin, where Makkhan Lal (1984) identified dozens of village farming communities that flourished between 200 BCE and 300 CE, but only seven sites sized between 6 and 11.25 hectares. Further to the northwest we have Ahichchhatra, known to literary sources as the archaic capital of the Panchala region, where a mound between 10 and 20 meters deep lies behind a defensive wall about 5.6 kilometers in length. Ahichchhatra seems to have constituted the linchpin of a separate regional hierarchy of sites that may have included Atranjikhera 80 kilometers away (a site 50 hectares in area with a mud rampart dating from ca. 500 bce) and Jakhera, 15 kilometers from Atranjikhera (a site 8 hectares in area with a mud rampart dating from the eighth-sixth centuries bce). These examples suggest that the patterns of urbanization varied according to developments within the 16 major regions (mahajanapada) mentioned repeatedly in early literary sources and represented by early coinage (Chakravarti 2000).