Ancient South Asian cities in their regions

Carla M. Sinopoli

In the autumn of 1924, the history of urbanism in South Asia was lengthened by 2,000 years.^[288] Prior to this, scholars knew of the efflorescence of cities, states, and religious visionaries that had occurred in the Ganges Basin in the mid-first millennium bce.

The discovery that cities had existed in South Asia in the third millennium bce overthrew long-standing beliefs that history and “civilization” had come late to the region, and archaeological work immediately shifted from a focus on Early Buddhism to what came to be called the “Harappan” or “Indus Civilization.” However, in one respect, things did not change. Colonial scholarship had long looked beyond South Asia's boundaries for sources of political and cultural innovation in the region.

Since 1924, thousands of archaeological sites containing Indus materials have been identified in Pakistan, India, and the Arabian Peninsula, and excavations have been conducted at large urban sites and countless smaller settlements. In this chapter I begin by reviewing current knowledge of this first period of South Asian urbanism, situating the Indus cities in their larger regional landscapes. I follow by briefly addressing the end of the Indus tradition and the cities that followed more than a millennium later. In so doing, I explore two very different urban trajectories and urban landscapes of ancient South Asia - the first characterized by a small number of massive widely spaced cities that existed as islands of urbanism in a vast sea of villages; the second characterized by closely packed urban places in a “landscape of cities” (sensu Adams).

Splendid isolation: Indus cities

The Bronze Age cities of Mohenjo Daro and Harappa (and the lesser known Dholavira, Ganweriwala Thar, and Rakhigarhi) are justly renowned for their scale, dense urban architecture, and distinctive material culture. Research on these sites and the numerous smaller settlements and special­ized sites in the spaces between them has generated a vast literature on Indus chronologies, settlements, economies, long-distance interactions, and political structures.^[289] It is not my purpose here to comprehensively review this literature. Instead, I focus on a single issue: the very small number of cities that existed within the enormous geographic region over which Indus tradition sites are found and the considerable distances that separated these few urban places (Map 15.ι).

Map 15.1 Indus sites (after Thomas R.

This is not a new observation, and has been discussed by Kenoyer, Shinde et al., Wright, and others.^[290] Both Wright and Kenoyer have argued that Indus political organization is best understood under the rubric “city-states” - multiple autonomous polities centered on cities and their hinterlands, which together participated in shared “civilizational” understandings and inter­actions. Both have also observed that the Indus urban landscape was quite different than the densely packed cities that we generally associate with city­state organization, for example in Early Dynastic Mesopotamia, the Aegean, Mesoamerica, or Early Historic South Asia (see below). In those cases, cities/states were separated by small distances - measured in tens of kilo­meters or a few days' travel time. The effective territory of such polities was small, and each city was able to exert (relatively) effective economic and political control over its hinterland; and interactions among inhabitants of these neighboring cities were intimate and frequent.

In the Indus region in contrast, the five documented cities were separated by enormous distances, with the closest two, Mohenjo Daro and Ganwer- iwala Thar, located 280 kilometers apart, and Harappa and Mohenjo Daro separated by 600 kilometers. These distances did not preclude exchange and other interactions, which are well documented archaeologically. However, their scale is striking. Indeed, if one were to assume that each Indus city effectively controlled its surrounding territories and that there were no Indus sites not under the authority of an urban center, Kenoyer has esti­mated the territories of Indus “city-states” as being between 100,000 and 170,000 square kilometers.

Given the vast scale of the Indus landscape and the remarkably small number of cities, it is challenging to understand the roles these cities played as loci of political, social, economic, and ideological order within the larger Indus world.

Indus chronologies and the pre-urban setting

The urban phase of the Indus tradition spanned from 2600 to 1900 bce and followed upon a millennium of increasing economic and material elabor­ation, manifest in multiple, geographically discontinuous, interacting, regional archaeological traditions (largely defined on the basis of distinctive styles of material culture, particularly painted pottery). Throughout this “Early I Iarappan'' period (c. 3500-2600 bce),^[291] communities expanded into the rich alluvial plains along the Indus and Ghaggar-Hakra Rivers of modern Pakistan and northwest India.

It was during the pre-urban period that many practices that would come to characterize subsequent urban forms took shape. Population expansion into new environmental zones was accompanied by intensifica­tion and diversification in subsistence and craft production, with wide­spread evidence for increasingly sophisticated and specialized crafting technologies in a range of materials. And recent excavations at Harappa provide evidence for the beginnings of writing technologies - including the production of seals containing signs that appear to be related to later Indus scripts.^[292] In each of the Early Harappan traditions (for example, Ravi, Amri, Hakra, Sothi-Siswal, and the widely distributed Kot Diji), we also see evidence for inter-settlement variability - in scale, productive activ­ities, and settlement elaboration.

Settlement layout became increasingly formalized throughout this period, with evidence for large-scale construction projects including major plat­forms, fortifications, and flood-control features. Recent excavations at Harappa have documented the creation of massive enclosure walls and mudbrick platforms, evidence for functionally specialized neighborhoods, and an increasingly formal layout of space and routes of movement.

Constraints on regional survey due to intensive alluviation along the Indus and its tributaries do not allow for a good understanding of larger regional settlement patterns in many areas. Nonetheless, hundreds of Early Harappan sites have been recorded, and local settlement patterns are rea­sonably well documented in Cholistan and western India, with additional data from Baluchistan and around Harappa. Together with research on local material cultural traditions, these data point to a social and political land­scape in the pre-Harappan period comprised of multiple polities of varying scales that were linked by complex relations and interactions, including trade and exchange, population movement and expansion, and conflict. The spaces between these polities and regional traditions were likely filled by variously organized small-scale communities (including pastoral commu­nities) that may have played important roles in connecting dispersed regional centers.

Indus urbanism

Dramatic changes occurred across the larger Indus region beginning about 2600 bce.

Sites containing this Mature Indus (or Urban Harappan) material assem­blage occur over a vast area - with more than ι,ooo known sites distributed over more than ι million square kilometers. Within this territory, only five sites are considered “urban”: Harappa, Mohenjo Daro, Ganweriwala Thar, Rakhigarhi, and Dholavira. These cities were large and highly structured places - ranging from 50 to more than 200 hectares in estimated extent. They have been variously documented. Harappa and Mohenjo Daro were first identified in the mid-nineteenth century and large-scale excavations began following recognition of their antiquity in the 1920s. The others were first identified in the 1960s - Ganweriwala in 1962, Rakhigarhi in 1964, and Dholavira in 1967. Because of its location near the contested Indian-Pakistan border, there has been little research at Ganweriwala, though some excav­ations have recently been initiated. Dholavira, in contrast, was excavated by the Archaeological Survey of India from 1989 through 2003; to date, publica­tions are limited. Excavations have also occurred at Rakhigarhi, but only brief notes have been published and little information is currently available on its form or organization.

Mohenjo Daro, the most iconic of the Indus urban sites, is also its largest, at more than 200 hectares. It was excavated over many decades, with vast areas of well-preserved domestic and public architecture uncovered. High water tables and dangerous salinization have made the site vulnerable to erosion, and excavations have been banned for the last few decades. Non­destructive research and coring as part of conservation activities have, however, yielded important recent evidence, but much of our understand­ings of Mohenjo Daro derive from recent reanalyses of plans and excavation records and attempts to link material remains to excavated loci - a great challenge given the poor recording and gross stratigraphic units employed by early excavators.^[294]

Like other urban sites, Mohenjo Daro is characterized by multiple spatial zones - here, a high western mound, some 400 x 200 meters in extent and rising more than 18 meters above the plain (often referred to as the “citadel” mound) and a more expansive eastern mound some 1,100 meters north­south and up to 650 meters east-west (the “lower town”). Recent surveys have revealed extensive occupational deposits between and beyond these mounded zones. Several distinctive monumental structures were excavated on the western mound (Figure 15. ι). The names assigned by their original excavators remain attached to these structures: the Great Bath, the College, the Granary, the Stupa and Monastery. While the precise uses of these structures are uncertain, it is clear that they were sizeable, non-residential constructions, involving considerable labor investment and coordination. Also involving substantial labor were the massive mudbrick platforms and integrated hydraulic features constructed at various phases of the city's occupation. These platforms, which appear to have been constructed or

Figure 15.1 Mohenjo Daro citadel plan (after J. Marshall, Mohenjo-Daro and the Indus Civilization, 3 vols. [London: Arthur Probsthain, 1931]).

Figure 15.2 Mohenjo Daro HR area (after E. J. H. Mackay, Further Excavations at Mohenjo-Daro, 2 vols. [New Delhi: Government Press, 1937-8]).

augmented in discrete phases, were massive constructions that served as foundations for subsequent building phases - the largest are more than half a kilometer in length.

Platforms were also constructed on the eastern mound of Mohenjo Daro, the location of the city's densest residential architecture (Figure 15.2). Residences and workshops were organized in dense blocks bounded by large and small roads and footpaths. Houses varied in their interrelations, layout, and size, ranging from 90 to 183 square meters in floor plan. In addition, some very large residential structures have been documented; including HR ι, a structure of more than twenty-five rooms that also produced the famous “priest king” sculpture as well as fifteen seals and other elaborate artifacts of faience, steatite, alabaster, ivory, and semi­precious stone. Jansen and colleagues have documented two major patterns of household organization: clusters of residences oriented on a central space or courtyard and sharing a common well; and clusters consisting of large houses surrounded by smaller residences and, sometimes, associated workshops.^[295] These data point to significant variability in household organ­ization and scale, in the organization of urban spaces, and in social and economic statuses. In addition, several areas of specialized craft production have been documented in both of Mohenjo Daro's major mounds, with some suggestions that the production of certain categories of goods (for example, stoneware bangles) was under administrative control.

Harappa shares with Mohenjo Daro a spatial organization of a number of distinct mounds and/or walled zones. However, nineteenth-century brick­robbing associated with the construction of the Indian railroad resulted in considerable destruction of portions of the site, particularly in the western “citadel” mound (Mound AB). As a result, little is known of the architectural plan of this area or whether significant non-residential constructions, such as found at Mohenjo Daro, existed. There is evidence that an extensive enclosure wall or revetment enclosed the mound, but little can be said about what lay inside those walls.

What we lack in Harappa's Mound AB has been partly compensated by the important systematic excavations in other areas of the site by the Harappa Archaeological Research Project (1986-2001).^[296] Rigorous excav­ations and regional survey efforts have resulted in a fine-scale chronology that has allowed the team to trace the history of the city from the late fourth millennium pre-urban periods through post-urban phases, which has put to rest any lingering arguments that the city appeared or disappeared abruptly. Excavations in the Mound E (lower town) area to the east of Mound AB have exposed a massive enclosure wall and elaborate gate systems, as well as domestic spaces, routes of movement, and craft production areas, and

Ancient South Asian cities in their regions evidence of efforts to monitor and control access and movement into the walled urban space. In the region around Harappa, Rita Wright and col­leagues have developed methods to systematically document sites along water courses and drainages.^[297] Although constrained by conditions affecting site preservation and visibility, they have successfully identified a number of sites that point to a complex regional infrastructure of various size settle­ments and agrarian landscapes in Harappa's immediate hinterland.

The third reasonably well-documented (though not yet fully published) Indus urban site is Dholavira, located on Kadir Island in the Great Rann (salt marsh) of Kutch in modern Gujarat, India. Fourteen seasons of excavation directed by R. S. Bisht of the Archaeological Survey of India beginning in 1989 yielded evidence of a complex and sophisticated urban space. Visually, the city of some 50 hectares (not including an extensive cemetery to its west) is quite different in plan and spatial organization than Harappa or Mohenjo Daro - though, like both, it contains multiple distinctive walled zones and elevational differences that distinguish segregated urban spaces. The excav­ators have defined seven major chronological phases, with Indus tradition occupations spanning from c. 2650 to 1900 bce. Their data allow an assess­ment of changing urban organization over time - as the city grew from a small fortified settlement founded c. 2650 bce - just on the cusp of the time of explosive urban growth at other urban centers. The city reached its greatest extent c. 2500 bce, when the maximal enclosure walls and major water-control features were constructed. Assuring fresh water supplies to the city for consumption and subsistence production was essential to inhabiting the salt marshes of Kutch, and considerable labor was invested in channeling water into the city from the two seasonal monsoon-fed streams that flanked the city. This was accomplished through a sophisticated network of canals, elaborate interconnected reservoirs, and wells.

Within the city walls, walled enclosures bounded habitation areas and other specialized spaces. In the south-central area of the site was a heavily fortified and elevated enclosure containing residential and “public” architec­ture called “the citadel” by Bisht. This area came to be separated from

“middle” and “lower” town residential zones to its north by an open space, which the excavators suggest was used for public or ceremonial gatherings. Other features excavated at Dholavira include monumental gate structures flanked by stone columns, stone sculptures of the kind previously only known from Mohenjo Daro, and evidence for a signboard at the North Gate, with an inscription of ten large signs (each more than 25 cm high) in Indus script, the first evidence that some Indus writing was meant to be read by public audiences. Future publication of the excavation report and detailed plans will no doubt allow much more to be said about this important site.

As noted, little information is available on the two other large Indus urban sites, Rakhigarhi and Ganweriwala Thar, though survey around the latter reveals that a complex and integrated hierarchical settlement system surrounded the city. More than a thousand other Indus sites are known from surface documentation, and excavations have occurred at dozens of sites. Several well-known sites, like Kalibangan, Lothal, and Chanhudaro, are relatively well documented. While these share complexly organized spaces and categories of material culture, they are small compared to the urban sites (about 5 hectares), though sites of intermediate scales have been recorded.

Conceptualizing the Indus

In conceptualizing the larger Indus phenomenon, questions of scale rise quickly to the fore. The geographic extent of sites containing Indus material culture assemblages is enormous. The major centers - the cities - were no doubt important and influential loci of cultural production and symbolic and economic power and wealth. Spaced at distances of hundreds of kilometers, each was likely the center of its own world in the day-to-day existence of its inhabitants. Yet each shared fundamental material connec­tions and, presumably, associated values - as manifest in spatial logics, systems of weights, measures, and writing, and an array of shared quotid­ian and elite material forms. Many of these same material logics also played out in the numerous smaller settlements of the Indus tradition, though as an overlay superimposed on considerable local variability. The social, political, and economic actors and mechanisms that forged the relations that connected the disparate cities and the hundreds of commu­nities that lay between them are, however, more elusive to identify. Certainly, the movement of craft products produced at an array of large and small sites, and likely individual artisans, artisan communities, and merchants were an important nexus of interconnections. Pastoral

communities too likely filled the spaces between large centers and inter­acted in a variety of ways with rural and urban communities.

Political structures and relations and ideologies have been notoriously difficult to characterize for the Indus region, for reasons both evidentiary and historical. Part of the challenge, I believe, lies in the historical privileging of Mesopotamia alluded to earlier and to a trait list approach to the understanding of both ancient cities and ancient states. Thus, scholars have focused on what Indus cities are not, or are not in reference to idealized models of contemporary late Early Dynastic (ED) Mesopotamian cities (particularly the city of Ur), which flourished more than a millennium after Mesopotamia's first cities were created. These comparisons typically high­light what the Indus cities lacked (that ED cities had): royal burials, elite iconographies celebrating institutions of kingship and violence, and unam­biguous temples and palaces.

While not wishing to elide the very real differences in historical trajector­ies in Mesopotamia and the Indus, I would like to challenge this long­standing and by now rote comparison by suggesting that the more relevant and interesting juxtaposition is between these first-generation Indus cities and first-generation Mesopotamian cities of the fourth millennium bce Uruk period. This comparison suggests that there may be more similarities in these regions during the early phases of urban creation than commonly acknowledged. Representations of leadership, power, and authority were indeed more prominent in Uruk iconography and texts (for example, the Warka Vase, seals, and sculptures) than in the Indus (where there are a number of representations of humans or deities battling fierce beasts or supernatural figures¹¹). However, royal cemeteries were absent in both regions during the early centuries of urban formation, and distinctive palaces,^{[298] [299]} while perhaps present, have not been definitively documented in either. While it is always difficult to argue from the absence of evidence, it is tempting to suggest that during the periods when urban forms and new political orders were first being created in both regions (likely a time of considerable social, ideological, and economic stresses), social differences may have been materially veiled rather than celebrated. Again, this is not to suggest that the nature and trajectories of Indus cities and the polities that underlay them were not different than in Mesopotamian city-states. These differences may have included, as Kenoyer has suggested, a far wider distribution of power and authority among diverse Indus communities and elites, such that there was not a single “royal” administrative hierarchy (as did eventually develop in Mesopotamia).^[300]

11

12

Early historic city-states

The Indus tradition, however we might best understand it, began to disinte­grate around 1900 bce. While settlement continued at some of the urban (and non-urban) sites, its scale and character changed dramatically - from highly organized urban centers to small non-urban communities. The widespread shared “civilizational” features of the Indus tradition - painted pottery forms and motifs, seals, writing systems, etc. - disappeared, as more localized regional traditions once again predominated across northern South Asia. Population, while not likely declining in overall numbers, was redis­tributed across the landscape, with the demographic center of gravity shifting eastward.

It was more than a millennium before urban places were again created in South Asia. This second urbanization encompassed portions of the greater Indus region, but its core was to the east in the alluvial plains between the Ganges and Yamuna Rivers. Scholars have debated how, or if, these later cities drew upon Indus tradition knowledge and memories or whether they were entirely new creations. While it is unlikely that all memories and knowledge of the Indus past were lost, this second period of South Asian urbanization followed unique trajectories and appears to have drawn on rather different and more diverse influences and populations.

Archaeologically, the prelude to Early Historic cities lay in the “Painted Grey Ware” (PGW) period from c. 1100 to 700 bce. The distinctive wheel- made pottery after which the period is named is generally interpreted as an elite serving ware and is reported in low frequencies (never more than io percent of total ceramics) from nearly 700 archaeological sites over an extensive geographic area. Within the Ganges Basin, the PGW period was one of dramatic population growth and incipient social differentiation, fueled by in situ growth and, almost certainly, in-migration. Settlements grew in both number and diversity, spreading along tributaries and major rivers, as populations began to fill in the limited areas of relatively open land in this marshy and densely forested region.

This period of filling in and emergent and consolidating social and political differentiation was followed by the “Northern Black Polished Ware” (NBPW) period, from c. 700 to 200/100 bce. It was during this span that enormous urban places were constructed and the emergent sociopoli­tical hierarchies of the PGW period were elaborated and cemented. The formation of these new cities and the states that governed them began in the core area between the Ganges and Yamuna Rivers, though similar and closely related “Early Historic” political formations and urban centers spread quickly across a broad expanse, extending east-west from the Gangetic Delta in eastern India and Bangladesh into what is today northwest Pakistan, and north-south from the Himalayan foothills of modern Nepal to central India.

Although there are no preserved written documents from the second and first millennia bce, important sacred texts were composed throughout this timespan and are preserved in documents recorded centuries later. These texts - the Vedas and associated commentaries and exegeses, the Puranas and Shastras, and Buddhist and Jain literature - are foundational to today's Hindu, Buddhist, and Jain religious thought. The direct application of these (largely) proscriptive religious documents to the interpretation of political, social, economic, or even sacred orders or practices of the period is complex. Nonetheless, when critically analyzed, they provide key information about these early cities and an increasingly complex and hierarchical social and political landscape and its idealized social orders.^[301]

Texts believed to have originated in the mid-first millennium bce refer­ence a landscape of great cities, the centers of variously organized states. These were the sixteen mahajanapadas (great states) (Map 15.2).^[302] Both texts

Map 15.2 Early Historic mahajanapadas (after F. R. Allchin,, The Archaeology of Early Historic South Asia: The Emergence of Cities and States [Cambridge: Cambridge University Press, 1995]., p. 116).

and archaeological survey data point to a complex hierarchy of places and people, including the cities on which these states were centered (mahana- gara), and subsidiary towns (nagara), markets and trading centers (nigama/ putabhedana), agricultural villages (gama), as well as pastureland and forest. According to the texts, mahajanapadas took two forms: monarchies (rajya) and oligarchies (gana), the latter ruled by councils of elders elected for a limited term by members of the ruling lineage. The ganas appear to have been largely restricted to the Himalayan foothills and to have self­consciously differentiated themselves from monarchical states in the fertile Ganga-Yamuna heartland. Categories of people referenced in texts include the four varnas - Brahmanas, Kshatriyas, Vaishyas, and Sudras - the founda­tions of the later system of Hindu castes, and numerous distinct occupa­tional groups with specialized roles and ranked statuses.

By 6oo bce, a dense urban landscape existed over much of northern South Asia. In their core area, these city-states lay in close proximity, at most a few travel days apart, creating a landscape of cities far more familiar to compar­anda in other regions of the world than was the Indus landscape discussed above. These Early Historic city-states engaged in a wide range of inter­actions - including economic exchange, elite intermarriages, and frequent warfare and competition. They shared a common elite and sacred material culture - evident in high-status NBPW serving vessels (known from more than 1,500 sites), terracotta figurines, a variety of ornaments and craft products of ivory, ceramic, glass, copper, iron, and stone, and aspects of urban archi­tecture. Major polities issued coinage in silver and copper, and stone weights and seals and sealings indicate shared standards that were important in the long-distance and local exchange that bound the region together.

Many Early Historic cities also assumed significance as important sacred places, for Vedic deities, and as places in the lives of the Buddha and Mahavira, two religious leaders who lived during the tumultuous period of state formation and attracted a large number of followers both during and after their lives. The construction of sacred monuments likely occurred relatively early; however, additions by later rulers have obscured the earliest religious constructions while also likely contributing to the longevity of many early cities.

Archaeological research on Early Historic cities is difficult and, in general, our archaeological evidence is much less rich for the densely populated Ganges-Yamuna core region than for the larger Indus region. While dense human occupation and agrarian transformation of the swampy and densely forested Ganges-Yamuna Doab did not begin until the mid-second millen­nium BCE, once settled, populations expanded rapidly and dramatically along the tremendously fertile river plains. By Early Historic times, populations were already considerable and the largest cities were many times the size of the Indus cities. Today, this is the world's most densely inhabited region and three millennia of agriculture and settlement have destroyed or obscured countless archaeological sites. And unlike the Indus sites, few Early Historic cities were ever fully abandoned. Many continue to be occupied today, with their earliest levels lying beneath tens of meters of later deposition. As a result, only small areas associated with early phases of urban formation have been excavated and many sites known from texts remain unlocated. Only Taxila in northwest Pakistan and the small town of Bhita have had extensive horizontal excavations, both in the early twentieth century, before major archaeological attention shifted to the Indus period.

A further challenge in examining processes of urbanism is the discordance between the resolution of our long archaeological phases and the rapidity of sociopolitical transformations. Thus, the 500-600-year NBPW period encompassed momentous changes: beginning before the formation of Early Historic states and cities and encompassing their expansion and consoli­dation and the succeeding creation and collapse of South Asia's first imperial state - the Mauryan Empire. While efforts to refine the archaeological chronology are underway, we have few absolute dates or careful strati­graphic excavations to anchor the sequences, and the archaeological evi­dence is not yet up to the task of carefully documenting, much less allowing us to explain, the changes that we know were occurring. Nonetheless, in addition to the textual sources alluded to above, there are archaeological data from excavations and, in less densely populated areas further from today's river courses, from a small number of regional surveys, and there have been some important recent field projects by the Archaeological Survey of India and university researchers.^[303]

Perhaps the most dramatic evidence preserved at several first millennium urban centers are the remains of the earthen ramparts that enclosed many Early Historic cities. The scale of these earthworks, which were sometimes faced with wood or burnt bricks, and the labor involved in their production, were enormous. For example, the ramparts of the city of Ujjain in central India, measured 75 meters wide at their base, are preserved to 14 meters high, and were over 5 kilometers in length. Comparable ramparts (though only 40 meters wide at the base!) were constructed at Kausambi, Rajgir, Ahicchatra, Atranjikhera, Hastinapura, and Mathura. Bastions, fortified gates, and watchtowers confirm the defensive roles of these walls. However, their height far exceeded any conceivable defensive needs. In addition to their defensive role, these walls would been visible from long distances on the flat expanses of surrounding plains, and no doubt signaled the power and importance of the places they protected to travelers and residents alike.

The areas enclosed by the ramparts reveal that these cities were sizeable - Kausambi, for example, is estimated to have covered about 50 hectares in 600 BCE and tripled in area, to more than 150 hectares, over the succeeding two centuries. Some cities were far larger. By the time of Mauryan hegemony in the late fourth and early third centuries bce, the total area of its capital Patilaputra (modern Patna) is estimated at more than 2,000 hectares.

Given the paucity of horizontal excavations, archaeological evidence tells us little about the internal organization of the cities. However, both texts and archaeological sources point to a high degree of economic specialization and social and spatial differentiation. Cities and towns were important loci of economic production and exchange, and traces of craft production workshops and crafting debris have been identified at a number of sites. Major roads led into the cities through large gates in the earthen and brick ramparts, with smaller roads and footpaths branching off from the major routes. Excavations suggest a considerable range in the size and layout of residences and in the formality of urban infrastructure. Where horizontal exposures exist, architecture is dense, further confirming that these cities were home to sizeable populations. Little is preserved of public or adminis­trative architecture from the earliest phases of these cities, though textual sources suggest their existence.

Our information is somewhat better from the second half of the NBPW period. Historical knowledge is also richer for this period, with more sources and more concordances among them, allowing us to trace the rise of South Asia's first empire.^[304] By the fourth century bce, the long-standing conflicts among cities culminated in the ascendancy of one: the mahajanapada of Magadha, situated along the Ganges on the eastern edge of the core zone of early state formation. The names and regnal years of Magadha's kings are reported in a number of sources, allowing a reasonable outline of royal chronology from the sixth through third centuries bce. Less is known about the state's political and economic organization, and how, or if, it differed from the peer polities that it eventually conquered. It is apparent that Magadha's Maurya rulers were effective both militarily and in forging political alliances, and, by the late fourth century, a series of military victories extended the territories of the kingdom over virtually all of north­ern South Asia. The most famous Mauryan king, Ashoka (268-232 bce), is renowned for expanding the empire into the peninsula and as an import­ant figure in the history of South Asian Buddhism. His inscriptions, found on columns and boulders over a wide region, are among the earliest preserved texts in South Asia. Today, Ashoka is also a national symbol of an enlightened ruler of a united India, and sculptures from the period appear on India's currency.

Much has been written about this Mauryan period and Ashoka. Here I limit myself to a discussion of what we know about urban sites of the period. Here too, our evidence is limited by problematic chronologies and the prominence of the Mauryas in the historical imagination. Thus, arch­aeological phases and certain texts are argued to be “Mauryan” with only limited supporting evidence. The limited archaeological data do not provide evidence that dramatic changes in urban organization accompanied the formation of the imperial polity, though urban scales undoubtedly con­tinued to increase. However, we do have somewhat richer evidence for city plans and constructions from the latter part of the NBPW period than are available from the first half (though how these specifically relate to any particular political dynasty is impossible to decipher).

Early excavations at Bhita and Taxila, mentioned earlier, and more recent work at Indor Khera,^[305] provide the best horizontal exposures of the period. Information on the Mauryan capital Patilaputra comes from limited excav­ations conducted in the late nineteenth and early and mid-twentieth centur­ies and the fragmentary writings of Megasthenes, Seleucid ambassador to the Mauryan court during the reign of Candragupta (c. 324-297 bce). Mega- sthenes described a vast and densely populated city some 12 kilometers long by 2.5 kilometers wide and enclosed by a wooden palisade flanked by a broad moat. Excavations have revealed traces of the palisade and the remains of a massive royal or administrative structure, with eighty massive stone columns arranged on a grid on a large platform. Other massive free­standing stone columns were transported long distances from their quarry site at Chunar^[306] to several Early Historic cities, and bear inscriptions and “edicts” of the emperor Ashoka, which figure among the earliest preserved texts from South Asia. Elegant sculpted sacred images also date to this period and, with coins, may be among the few definitive material markers of Mauryan hegemony.

Mauryan rule did not long outlast the death of Ashoka in 232 bce, and the empire collapsed some forty years later. However, as alluded to above, many of the cities of the period endured through this political

Ancient South Asian cities in their regions transformation and through the conquest, incorporation, and collapse of many successive states and empires over succeeding centuries and millennia. The resilience of these (always changing) urban forms and spaces is the result of the durability of a range of social, economic, and sacred institutions and mechanisms that were able to sustain urban life even in times of dramatic political upheaval. Understanding this durability will require us to separate the political from the urban to consider the broad range of social, economic, and sacred institutions and mechanisms that sustained and sup­ported historic South Asian urban spaces.^[307]

Discussion

In this chapter, I have briefly explored two urban trajectories in a single geographic region. Each was characterized by large, differentiated, and complexly organized urban places. They inhabited very different physical and cultural landscapes - from the widely dispersed largely rural landscape of the Indus region to the dense urban landscapes of the Ganges-Yamuna Basin. While often described as short-lived, the few cities and many more numerous smaller settlements of the Indus tradition thrived for some 700 years, before the Indus tradition declined and disappeared, until its rediscov­ery by archaeology millennia later. The duration of many Early Historic Indian cities continued much longer, many remaining vibrant centers of population long after the Mauryan Empire's fall and through numerous successive states and empires, and leaving a legacy that endures to the present.

FURTHER READINGS

Bisht, Ranvir S., “Dholavira: New Horizons of the Indus Civilization,” Puratattva 20 (1991), 71-81.

“The Water Structures and Engineering of the Harappans at Dholavira (India),” in CatherineJarrige and Vincent Lefevre (eds.), South Asian Archaeology 2001: Proceedings of the Sixteenth International Conference of the European Association of South Asian Archaeologists, Held in College de France, Paris, 2-6 July 2001, Paris: CNRS, 2005, pp. 11-25.

Blackman, M. James, and Massimo Vidale, “The Production and Distribution of Stone­ware Bangles at Mohenjo-Daro as Monitored by Chemical Characterization Stud­ies,” in Richard H. Meadow (ed.), South Asian Archaeology 1989: Papersfrom the Tenth

International Conference of South Asian Archaeologists in Western Europe, Musee National des Arts Asiatiques-Guimet, Paris, France, 3-7 July 1989, Madison, WI: Prehistory Press, 1992, pp. 37-43.

Chattopadhyaya, Brajadulal D., “The City in Early India: Perspectives from Texts,” Studies in History 13 (1997), 181-208.

Erdosy, George, “City States of North India and Pakistan at the Time of the Buddha,” in F. Raymond Allchin et al., The Archaeology of Early Historic South Asia: The Emergence of Cities and States, Cambridge: Cambridge University Press, 1995, pp. 123-52.

Urbanisation in Early Historic India, Oxford: British Archaeological Reports, 1988.

Fritz, John M., George Michell, and M. S. Nagaraja Rao, Where Kings and Gods Meet: The Royal Center at Vijayanagara, Tucson: University of Arizona Press, 1985.

Gaur, Ruby C., The Excavations at Atranjikhera: Early Civilization in the Ganga Valley, Delhi: Motilal Banarasi Das, 1983.

Ghosh, Amitav, “Taxila (Sirkap) 1944-45,” Ancient India 4 (1948), 66-78.

Jansen, Michael, and Giinter Urban (eds.), Interim Reports, Aachen: Forschungsprojekt “Mohenjo-Daro,” 1984, Vols. 1 and ιι.

Jayaswal, Vidula, From Stone Quarry to Sculpturing Workshop: A Report on the Archaeological Investigations around Chuna, Varanasi and Sarnath, Delhi: Agam Kala Prakashan, 1998. Kenoyer, Jonathan M., “The Indus Valley Tradition of Western India and Pakistan,” Journal of World Prehistory 5 (1991), 331-85.

Kenoyer, Jonathan M., and Richard H. Meadow, “The Ravi Phase: A New Cultural Manifestation at Harappa, Pakistan,” in Maurizio Taddei and Giuseppe de Marco (eds.), South Asian Archaeology 1997: Proceedings of the Fourteenth International Confer­ence of the European Association of South Asian Archaeologists, Held in the Istituto italiano per l’Africa e l’Oriente, Palazzo Brancaccio, Rome, 7-14 July 1997, Rome: Istituto Italiano per l’Africa e l’Oriente and Istituto Universitario Orientale, 2002, pp. 55-76

Lal, B. B., “Excavations at Hastinapura and Other Explorations in the Upper Ganga and Sutlej Basins,” Ancient India ιι (1955), 5-151.

Leonardi, Giovanni, Interim Reports, Rome: Forschungsprojekt “Mohenjo-Daro”, 1988, Vol. ιιι.

Marshall, John H., A Guide to Taxila, Delhi: Government of India Publications, 1936. Taxila, 3 vols., Delhi: Motilal Banarsidas, 1951.

Meadow, Richard H., and Jonathan M. Kenoyer, “Excavations at Harappa 2000-2001: New Insights on Chronology and City Organization,” in Catherine Jarrige and Vincent Lefevre (eds.), South Asian Archaeology 2001: Proceedings of the Sixteenth International Conference of the European Association of South Asian Archaeologists, Held in College de France, Paris, 2-6 July 2001, Paris: CNRS, 2005, pp. 207-24.

“Harappa Excavations 1998-1999: New Evidence for the Development and Manifest­ation of the Harappan Phenomenon,” in Ellen M. Raven (ed.), South Asian Archae­ology 1999: Proceedings of the Fifteenth International Conference of the European Association of South Asian Archaeologists, Held at the Universiteit Leiden, 5-9 July, 1999, Groningen: Egbert Forsten, 2008, pp. 85-110.

Mughal, M. Rafique, “The Geographical Extent of the Indus Civilization During the Early, Mature, and Late Harappan Times,” in Gregory L. Possehl (ed.), South Asian Archaeological Studies, Delhi: Oxford University Press, 1992, pp. 123-43.

“The Harappan Settlement Systems and Patterns in the Greater Indus Valley (circa 3500-1500 BC),” Pakistan Archaeology 25 (1990), 1-72.

Narain, Awadh Kishore, and T. N. Roy (eds.), Excavations at Rajghat, Varanasi: Banaras Hindu University, 1977.

Possehl, Gregory L., “The Date of Indus Urbanization: A Proposed Chronology for the Pre-urban and Urban Harappan Phases,” in Catherine Jarrige (ed.), South Asian Archaeology 1989: Papers from the Tenth International Conference of South Asian Archae­ologists in Western Europe, Musee National des Arts Asiatiques-Guimet, Paris, France, 3-7 July 1989, Madison, WI: Prehistory Press, 1993, pp. 237-44.

Indus Age: The Beginnings, Philadelphia: University of Pennsylvania Press, 1999.

Ratnagar, Shereen, Enquiries into the Political Organization of Harappan Society, Pune: Ravish Publishers, 1991.

Shaffer, J. G., “The Indus Valley, Baluchistan, and Helmand Traditions: Neolithic through Bronze Age,” in Robert W. Ehlrich (ed.), Chronologies in Old World Archaeology, Chicago: The University of Chicago Press, 1992, Vols. 1 and ιι.

Sharma, Govardhan Raj, The Excavations at Kausambi (1957-59), Allahabad: University of Allahabad, i960.

Shinde, Vasant, Shweta Sinha Deshpande, Toshiki Osada, and Takao Uno, “Basic Issues in Harappan Archaeology: Some Thoughts,” Ancient Asia i (2006), 63-72.

Sinha, Bindeshwari Prasad, and Lala Aditya Narain, Patilaputra Excavation 1955-56, Patna: Bihar Directorate of Archaeology and Museums, i970.

Sinopoli, Carla M., “On the Edge of Empire: Form and Substance in the Satavahana Dynasty,” in Susan E. Alcock, Terence N. D'Altroy, Kathleen D. Morrison, and Carla M. Sinopoli (eds.), Empires: Perspectives from Archaeology and History, Cam­bridge: Cambridge University Press, 2001, pp. 155-78.

Smith, Monica L., “The Archaeology of South Asian Cities,” Journal of Archaeological Research i4 (2006), 97-i42.

Spooner, David B., Excavation at Shaji-ki-dheri: Annual Reports of the Archaeological Survey of India 1908-09, Calcutta: Government Press, 1909.

Thapar, Romila, Early India: From the Origins to AD 1300, New Delhi: Penguin, 2002.

Tosi, Maurizio, Luca Bondioli, and Massimo Vidale, “Craft Activity Areas and Surface Survey at Moenjodaro: Complementary Procedures for the Reevaluation of a Restricted Site,” in MichaelJansen and Giinter Urban (eds.), Interim Reports, Aachen: Forschungsprojekt “Mohenjo-Daro,” 1984, Vol. ι, pp. 9-38.

Trautmann, Thomas R., and Carla M., Sinopoli, “In the Beginning was the Word: Excavating the Relations between History and Archaeology in South Asia,” Journal of the Economic and Social History of the Orient 45 (2002), 492-523.

Verardi, G., “Preliminary Reports on the Stupa and Monastery of Mohenjo-daro,” in MichaelJansen and Giinter Urban (eds.), Interim Reports, Aachen: Forschungsprojekt “Mohenjo-Daro”, 1984, Vol. ιι.

Vidale, Massimo, “Specialized Producers and Urban Elites: On the Role of Craft Production in Mature Indus Urban Contexts,” in Jonathan M. Kenoyer (ed.), Old Problems, New Perspec­tives in the Archaeology of South Asia, Madison: University of Wisconsin, 1989, pp. 145-56.

Waddell, Austine L., Report on the Excavations at Pataliputra (Patna), Calcutta: Bengal Secretariat Press, i996.

Wright, Rita P., M. Afzal Khan, and Joe Schuldenrein, “The Emergence of Satellite Communities along the Beas Drainage: Preliminary Results from Lahoma Lal Tibba and Chak Purbane Syal,” in Catherine Jarrige and Vincent Lefevre (eds.), South Asian Archaeology 2001: Proceedings of the Sixteenth International Conference of the European Association of South Asian Archaeologists, Held in College de France, Paris, 2-6 July 2001, Paris: CNRS, 2005.

Wright, Rita P., Joe Schuldenrein, M. Afzal Khan, and S. Malin-Boyce, “The Beas River Landscape and Settlement Survey: Preliminary Results from the Site of Vainiwal,” in Ute Frank-Vogt and H. J. Weisshaar (eds.), South Asian Archaeology 2003: Proceed­ings of the Seventeenth International Conference of the European Association of South Asian Archaeologists, 7-11 July 2003, Bonn, Aachen: Linden Soft, 2005, pp. ιoι-ιι.