(PDF) Adaptation to Variable Environment

Current Anthropology Volume 58, Number 1, February 2017 000 Adaptation to Variable Environments, Resilience to Climate Change Investigating Land, Water and Settlement in Indus Northwest India by Cameron A. Petrie, Ravindra N. Singh, Jennifer Bates, Yama Dixit, Charly A. I. French, David A. Hodell, Penelope J. Jones, Carla Lancelotti, Frank Lynam, Sayantani Neogi, Arun K. Pandey, Danika Parikh, Vikas Pawar, David I. Redhouse, and Dheerendra P. Singh This paper explores the nature and dynamics of adaptation and resilience in the face of a diverse and varied envi- ronmental and ecological context using the case study of South Asia’s Indus Civilization (ca. 3000–1300 BC). Most early complex societies developed in regions where the climatic parameters faced by ancient subsistence farmers were varied but rain falls primarily in one season. In contrast, the Indus Civilization developed in a speciﬁc environmental context that spanned a very distinct environmental threshold, where winter and summer rainfall systems overlap. There is now evidence to show that this region was directly subject to climate change during the period when the Indus Civilization was at its height (ca. 2500–1900 BC). The Indus Civilization, therefore, provides a unique opportunity to understand how an ancient society coped with diverse and varied ecologies and change in the fundamental environmental parameters. This paper integrates research carried out as part of the Land, Water and Settlement project in northwest India between 2007 and 2014. Although coming from only one of the regions occupied by Indus populations, these data necessitate the reconsideration of several prevailing views about the Indus Civilization as a whole and invigorate discussion about human-environment interactions and their relationship to processes of cultural transformation. Adapting to Variable Environments, Being Resilient haps no surprise that there is ongoing interest in the way that to Changing Climates humans caused and/or responded to environmental and eco- logical change in the past (Barnes et al. 2013; Diamond 2005; Given the considerable importance of climate, climate change, Staubwasser and Weiss 2006). Unquestionably, there is much and human/environment relationships in the present, it is per- to learn from the past about the success or failure of adap- C. A. Petrie is Reader in South Asian and Iranian Archaeology, C. A. I. French is Professor in Geoarchaeology, P. J. Jones and Centre, 29280 Plouzané, France). D. A. Hodell is Woodwardian Pro- D. Parikh are PhD students, and D. I. Redhouse is Computer Of- fessor of Geology in the Department of Earth Sciences at the Univer- ﬁcer in the Department of Archaeology and Anthropology at the sity of Cambridge (Downing Street, Cambridge CB2 3EQ, United University of Cambridge (Downing Street, Cambridge CB2 3DZ, Kingdom). C. Lancelotti is Postdoctoral Fellow in the CaSEs Research United Kingdom [

[email protected]

]). R. N. Singh is Professor in Group Archaeology and History of Science and Technology, A. K. Pandey is Draughtsman, and D. P. Singh is Research Assistant in the Ar- chaeological Lab in the Department of Ancient Indian History, Cul- ture and Archaeology at Banaras Hindu University (Varanasi 221 005, India). J. Bates is Trevelyan Research Fellow at Selwyn Col- lege and Research Fellow at the McDonald Institute for Archaeol- ogy, University of Cambridge (Downing Street, Cambridge CB2 3ER, United Kingdom). Y. Dixit is Marie Curie PRESTIGE and LabexMER Postdoctoral Fellow at the Ifremer-Brittany Centre of Plouzané, France Laboratoire Environnements Sédimentaires (BP70 Ifremer-Brittany q 2017 by The Wenner-Gren Foundation for Anthropological This content All use subject to University 000 tations to particular environments and ecological niches and the sustainability and resilience of responses to environmental pressures and climatic threats. Disentangling these dynamics is not, however, a straightforward process, and it is increas- ingly recognized that responses to environmental change are neither deterministic nor straightforward; particularly because environmental parameters, human behavior, and the interre- lationship between these two elements are inherently complex (McAnany and Yoffee 2009; Miller, Moore, and Ryan 2011). This line of thinking recognizes that humans and the envi- ronment are neither independent nor simple variables; rather, they are both complex and interlinked in what has been de- scribed as both panarchy and a social-ecological system (SES) that witnesses cycles of resilience and adaptive change (e.g., Berkes, Colding, and Folke 2003; Gunderson and Hollig 2002). Leslie and McCabe (2013:116) have noted that while the con- cepts of resilience and adaptive change have been explored conceptually, empirical analysis remains rare, at least partly because resilience is difﬁcult to measure in the context of com- plex socioecological systems. Archaeologists can play a unique role here as they are able to empirically investigate the before, during, and after of past instances of success or failure, thus furthering understanding of adaptation, resilience, and human response to climate change (cf. Mitchell 2008; Van de Noort 2011). Although archaeologists recognize that human behavior is nuanced and varied, much of the debate about the impact of climate change on ancient civilizations has tended to be sim- plistic, both in terms of empirical approach and conceptual grounding. Debate has been dominated by numerous attempts to correlate global-scale climate records and the timing of local-scale cultural transformations that are visible in the ar- chaeological record (e.g., deMenocal 2001; Haug et al. 2003; Staubwasser et al. 2003; Staubwasser and Weiss 2006), despite recognition that there is rarely direct evidence to link the two data sets (e.g., Aimers and Hodell 2011). As a result, inferences tend to be speculative and end up in “correlation equals causation” circularity. Furthermore, despite attempts contrary, there remains a fundamental disconnect between sci- entiﬁc approaches to understand global climate and the dy- namics of climate change, on the one hand, and humanistic approaches to understand how human populations perceive climate and respond to climate change on the other (Barnes et al. 2013). Part of the problem is that archaeologists often uncriti- cally look to distant climate data sets to interpret local cultural dynamics, while paleoclimatologists tend to uncritically look for cultural correlates to the climatic events that they observe (Aimers and Hodell 2011). Archaeologists are now increasingly interested in under- standing the ways that humans respond to change and the de- gree to which their societies and choices are sustainable and facilitate resilience (e.g., McAnany and Yoffee 2009; Miller, Moore, and Ryan 2011). This does, however, present signiﬁ- cant empirical challenges as there is growing consensus that in order to properly comprehend human adaptation, sustain- This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement there is a lack of consensus about Indus sociopolitical structure and organization (e.g., Kenoyer 1994; Possehl 1998; Wright 2010), this actually serves to emphasize that interpretation of socioeconomic structures can be challenging in the absence of texts that can be readily translated (e.g., Parpola 1994). It is clear that following a protracted period of village- based settlement, the urban phase of the Indus Civilization developed on the plains of modern Pakistan and northwestern India (ﬁg. 1A) during the mid-third millennium BC (∼4.6– 4.5 ky BP). It has been claimed that during this phase, Indus settlements were distributed across an area of ca. 1 million km2 concentrated around the river systems of northwest South Asia (Agrawal 2007:3; Possehl 2003:1). While this is an overesti- mation of the actual area occupied by Indus populations, our present understanding of settlement distribution suggests that the Indus Civilization was likely the most geographically ex- tensive of all the early Old World civilizations (cf. Agrawal 2007; Possehl 2003; Wheeler 1968:4). Present knowledge indicates that there was a constellation of four or ﬁve particularly large Indus settlements, which are usually described as cities (Mohenjo-daro, Harappa, Rakhi- garhi, Dholavira, and possibly Ganweriwala; Kenoyer 2008: 188; Petrie 2013:91). The inhabitants of these cities produced, used, and traded distinctive types of material culture, includ- ing painted pottery and ﬁgurines that were presumably made locally, and jewelry, standardized weights, and stamp seals that were made from raw materials typically obtained from medium- and long-range sources (the abundance of relevant specialist reports are reviewed recently in Wright 2010:148– 166, 182–203; also Coningham and Young 2015:211–223). In a landscape dominated by rural settlements, Indus cities appear to have been the exception rather than the norm (Petrie 2013). The substantial distances between the major centers (at least 280 km) have been used to suggest that they each controlled vast hinterlands (e.g., Kenoyer 1997:54, 1998:50, table 3.1). It is, however, also probable that large and medium-sized settle- ments played an important and, perhaps, independent rather than subordinate role in both interactive processes and socio- economic control structures (Petrie 2013:91, 94–95; Sinopoli 2015:322). To some extent, the pattern of material evidence seen at the city-sites has also been observed at large, medium, and small settlements, and this has led to the suggestion that there was marked uniformity in some aspects of Indus material culture (e.g., seals, weights, script; Agrawal 2007:7; Chakrabarti 1999: 179ff.; Kenoyer 2008:207; H. M.-L. Miller 2013; Wright 2010: 23, 327, 334). While similarities between some cultural ele- ments have been emphasized, variation in material and human behavior has been recognized for some time and is increas- ingly being acknowledged (e.g., Ajithprasad 2011; Joshi 1984; Meadow and Kenoyer 1997:139; Petrie 2013:91, 95; Possehl 1982, 1992, 2002; Weber, Barela, and Lehman 2010; Wright 2010:180ff.). This variation is particularly evident in subsis- tence practices (e.g., Vishnu-Mittre and Savithri 1982:215; Weber 1999; Weber, Barela, and Lehman 2010; Weber and This content downloaded All use subject to University Figure 1. Distribution of urban-phase Indus settlements urban-phase Indus settlements (B) and their relationship A color version of this ﬁgure is available online. This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement and Dales 1977; against: Lambrick 1964, 1967). We also now have detailed reconstructions of river shifts in Sindh, which demonstrate the movement of the main Indus channel be- tween 4000 and 2000 BC (e.g., Flam 1981, 1993, 1999, 2013; Jorgensen et al. 1993). Furthermore, remote sensing has sug- gested that settlement patterns in southern Punjab may have responded to the dynamics of the Beas River system (Wright and Hritz 2013). These reconstructions and other geomor- phological investigations also provide insight into the other major topic of hydrological discussion, the impact of the drying of the Ghaggar/Hakra River, which is often equated with the “lost” Saraswati River (e.g., Clift et al. 2012; Danino 2010; Ghose, Kar, and Husain 1979; Giosan et al. 2012; Lal 2002; Mughal 1997; C. F. Oldham 1874, 1893; R. D. Oldham 1886; Shinde et al. 2006; Stein 1942; Valdiya 2002; Wilhelmy 1969; Yash Pal et al. 1980; also Flam 1999, 2013). In northwest India, connections between climate change and river shift have been mooted (e.g., Giosan et al. 2012), and it has also been posited that neotectonic processes have been a factor in reshaping hy- drology (e.g., Puri and Verma 1998). A number of separate archaeological projects have applied multidisciplinary analysis of environmental parameters im- pacting Indus populations, incorporating geology, geomor- phology, and bioarchaeology (e.g., Sindh Archaeological Proj- ect: Flam 1981, 1993, 1999, 2013; Jorgensen et al. 1993; Mission Archéologique Française en Inde: Courty 1985, 1995; Courty, Goldberg, and Macphail 1989; Francfort 1985; Gentelle 1985; Harappa Archaeological Research Project: Amundson and Pendall 1991; Belcher and Belcher 2000; Meadow 1991; Weber 2003; Beas Landscape and Settlement Survey: Schuldenrein, Wright, and Khan 2007; Wright 2010; Wright, Bryson, and Schuldenrein 2008; Wright and Hritz 2013; Wright, Khan, and Schuldenrein 2002, 2005a, 2005b; Indus Project of the Research Institute for Humanity and Nature: Rajaguru and Deo 2008; Shinde et al. 2008; Weber, Kashyap, and Mounce 2011). How- ever, thus far, there have been only limited attempts to correlate and integrate the ﬁndings of these projects (e.g., Schuldenrein, Wright, and Khan 2007; Wright 2010: 25–44). There has also been only limited attention to proxy evidence for ancient cli- mate that is proximate to the Indus zone and/or can be con- nected directly to the archaeological record. Climate has long been considered an important parameter for understanding the Indus Civilization, starting from Mar- shall’s (1931:2; after Stein 1931) suggestion that there has been a signiﬁcant decrease in rainfall since the Indus period. In querying this interpretation, Raikes and Dales (1961) high- lighted the “importance of integrating all types of evidence and checking on the inferences drawn from them” (279). However, traditionally, there has been an underappreciation of the re- lationships between the environmental and cultural dynamics that were in action. As elsewhere, archaeologists considering the Indus case have tended to either under- or overemphasize the possible role of climate (e.g., Clift and Plumb 2008:205– 210; Giosan et al. 2012; Kenoyer 1997, 2008:186; Possehl 1997a; Shinde et al. 2006; Staubwasser and Weiss 2006). Fur- This content downloaded All use subject to University Figure 2. Distribution of urban-phase Indus settlements Global Köeppen-Geiger Climate Classiﬁcation Zones. This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement dynamics, the use of material culture, and subsistence practices in this region. Although coming from one of several regions occupied by Indus populations, these data necessitate the re- consideration of many prevailing views about the Indus Civi- lization as a whole, and this paper aims to further invigorate discussion about human-environment interactions and their relationship to processes of cultural transformation. Factors Inﬂuencing Environmental Diversity in Northwest South Asia As noted above, the area across which Indus Civilization pop- ulations lived spans an environmental threshold characterized by a zone of overlap between winter and summer rainfall sys- tems and steep rainfall gradients for both systems. This par- ticular location spans a range of distinct ecological zones, with modern Köppen-Geiger Climate Classiﬁcations (Kottek et al. 2006) ranging from areas of arid hot desert (BWh) to areas of arid hot steppe (BSh) and areas that are warm and temperate with dry winters and hot summers (Cwa; ﬁg. 2A, 2B). An im- portant consequence of this environmental context is that even without human interference, water is available from different sources at different times of the year, including winter rain (December–February), rain from the Indian summer monsoon (June–September), snowmelt from the Himalayas, and the surface and river runoff that results from all of the above. The lack of systematic and localized paleoclimatic data means that it is not yet possible to fully reconstruct the dis- tribution of rainfall at the time of the Indus Civilization (see below). To frame our understanding, however, it is instructive to look at modern rainfall patterns to gain some insight into the nature of rainfall variability across the same geographic region. Plotting annual rainfall averages calculated using global rainfall data for the period between 1900 and 2008 illustrates that over the past century, different areas in northwest South Asia received different amounts of rainfall during an average year, ranging from 0 to 1,000 mm (ﬁg. 1A, 1B). In addition, there is also variation in the seasonal distribution of modern rainfall (ﬁgs. 3A, 3B, 4A, 4B). The summer monsoon makes the dominant contribution to the average annual rainfall in many areas of the Indus zone, particularly those to although a signiﬁcant proportion of summer rain is lost through evapotranspiration. In contrast, the extensive areas of Punjab and Sindh that lie along the Indus and the rivers of Punjab receive rainfall in different intensities and at different times during the year. To further complicate matters, the historical record shows dramatic interannual ﬂuctuations in the intensity of monsoon rainfall, with years of particularly heavy rainfall resulting in ﬂooding and waterlogging interspersed with years of monsoon failure (Adamson and Nash 2013; Possehl 1999: 286–287; Sarma 1976). While this assessment of modern rainfall patterning is informative, it cannot be assumed that the seasonal rainfall fell in similar patterns in the past. On the basis of analysis of sediments at the mouth of the Indus River, Staubwasser and This content downloaded All use subject to University Figure 3. Distribution of urban-phase Indus settlements mean winter rainfall (1900–2008). A color version This content downloaded All use subject to University of Figure 4. Distribution of urban-phase Indus settlements mean summer rainfall (1900–2008). A color version This content downloaded All use subject to University of 000 It is notable that each of the Indus cities was supported by a different hydrological regime. Harappa, Ganweriwala, and Mohenjo-daro are in areas on the alluvial Indus plain that differ from each other in amounts of rainfall and proximity to major watercourses that provide water from both nonlocal rainfall and snowmelt in the Himalayas. Mohenjo-daro also has evidence for the extensive use of wells (Jansen 1993, 1994: 270), and examples are also known from elsewhere, including Harappa (Kenoyer 2008) and Dholavira (Bisht 2005, n.d.: 138–145). It is presumed that in each of these instances, the inhabitants exploited both river water and groundwater. In contrast, Rakhigarhi lies at some distance from known water- courses but is in the zone where both summer-monsoonal and, to a lesser extent, winter rainfall systems operate today. It has been proposed that Rakhigarhi lay on the course of a now ex- tinct watercourse, which is often referred to as the Drishadvati (Nath 1998; Suraj Bhan 1975:95–101; Valdiya 2002). However, no evidence for this watercourse is visible today on the surface (Singh et al. 2010b), and analysis of the satellite imagery sug- gests that only small-scale watercourses are preserved in the subsurface (Mehdi et al. 2016, ﬁgs. 2, 10). It is not yet clear where the water used by the inhabitants of Rakhigarhi origi- nated, though a combination of wells and ponds that collect monsoon runoff is a viable option (Petrie 2013). Dholavira is located in an area that today receives relatively limited rainfall but is close to two seasonal streams or runnels and has a system of dams that help channel water into a series of large stone- lined reservoirs and tanks, all of which presumably helped compensate for the unpredictable water supply (Bisht 2005, n.d.:138–169). Recognition of this diversity in settlement lo- cation and availability of water is essential for understanding both adaptations to different environments and responses to environmental challenges in the Indus context. It has long been hypothesized that there was variation in the subsistence practices used by Indus populations (e.g., Chak- rabarti 1988; Vishnu-Mittre and Savithri 1982), and this ﬁts with the theme of coping with diverse environments. Although primarily speaking about Sindh and Baluchistan, Fairservis (1967:10, 42, 1971:169–172, 228–232) argued that Indus farm- ers were adapted to the diverse environments that they in- habited, particularly in terms of the adaptating practices to the available water resources. Speaking more broadly, Possehl (1982, 1992) and Joshi (1984) have both posited the existence of ecocultural domains. More recently, models have been pro- posed for helping to identify Harappan agroecological zones, and several distinct ecozones have been identiﬁed (Weber, Barela, and Lehman 2010). However, robust evidence to sup- port these suggestions is not widely available. For instance, Wright (2010) has pointed out that the archaeobotanical evi- dence that informs us about Indus populations is “uneven and dependent upon limited excavation” (170). Indus agriculture is typically characterized as being domi- nated by the exploitation of a particular set of animals (pri- marily zebu, goat, sheep, and water buffalo) and a range of winter and summer crops (Meadow 1996; Weber 1999; Wright 2010:168–170). The exploitation of particular crops appears This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement fore, argued that while the archaeobotanical assemblages thus far published do demonstrate regional variation in subsistence practices (e.g., García-Granero, Lancelotti, and Madella 2015; Weber 1999, 2003; Weber, Barela, and Lehman 2010; Weber and Kashyap 2016), they have not (yet) provided convincing evidence from any single location for cropping in two seasons in anything approaching equivalent proportions (see below). They thus advocate the use of more precise terminology to characterize the variation that is observed (Petrie and Bates, forthcoming). Although Indus populations may well have selected spe- ciﬁc plant crops, the degree of variation in local environmen- tal conditions, vegetation, rainfall, and water supply would invariably have necessitated speciﬁc adaptations to farming practices for successful farming in different regions. These adaptations would likely have included a range of approaches to water supply (cf. H. M.-L. Miller 2006, 2015; Petrie and Thomas 2012) and a spectrum of cropping strategies ranging between a heavy focus on winter or a heavy focus on summer crops, with the middle ground being made up of a nuanced array of strategies where different combinations of winter and summer crops were utilized according to local conditions and choices (Petrie and Bates, forthcoming; also Petrie 2013; see below). It is clear that the degree of ecological diversity encom- passed by the Indus Civilization and the variability of adapta- tion and response across that area is critical for understanding the developments of the Indus period. However, the diversity in socioecological systems can be characterized adequately only by detailed research in each of the relevant zones. This research has only recently begun to be carried out at a suitable resolution. In Pakistan, the most important contributions have come from the Sindh Archaeological Project (Flam 1993, 1999, 2013; Jorgensen et al. 1993) and the Beas Landscape and Settlement Survey (Schuldenrein, Wright, and Khan 2007; Wright 2010; Wright, Bryson, and Schuldenrein 2008; Wright and Hritz 2013; Wright, Khan, and Schuldenrein 2002, 2005a, 2005b), while in India, knowledge is advancing most overtly in Gujarat through the North Gujarat Archaeological Proj- ect (Balbo et al. 2014; García-Granero et al. 2016; Madella et al. 2010) and in Haryana/Punjab/north Rajasthan through the Land, Water and Settlement project and the earlier Mis- sion Archéologique Française en Inde (Courty 1985, 1995; Courty, Goldberg, and Macphail 1989; Francfort 1985; Gen- telle 1985). The evidence gathered by the Land, Water and Settlement project will be explored further below, covering ﬁve key areas: monsoon dynamics, the paleo-Ghaggar/Hakra, mon- soon ﬂooding, settlement dynamics, and variation in material culture and subsistence. Changes in Intensity of the Indian Summer Monsoon during the Holocene The signiﬁcant environmental variability of the Indus region and the apparent ﬂexibility of Indus populations in coping This content downloaded All use subject to University 000 part of the story, and it is the human response to this change in climate that is the critical element. For example, it has been suggested that decline in monsoon strength led to the diver- siﬁcation of the Indus crop assemblage through the adoption or intensiﬁed use of more summer crops such as millet and rice (Giosan et al. 2012; Madella and Fuller 2006; Wright 2010:321ff.). This reconstruction is perhaps overly simplistic, as it advocates the greater exploitation of summer crops at a time when the summer rainfall weakened, and it does not con- sider the different lengths of growing seasons required for millet and rice or the fact that each of these crops is suited to somewhat different ecological conditions. It has also been at least partly superseded by new evidence that gives fresh insight into the nature of environmental adaptation that Indus pop- ulations engaged in even before the development of urban centers (see below). The Role of the Paleo-Ghaggar/Hakra River in Northwest India Alongside the considerations of climate and climate change, there has been considerable discussion of the role of the paleo- Ghaggar/Hakra River in the origin, ﬂoruit, and transformation of the Indus Civilization. It is often suggested that this pa- leochannel provided an important source of water for Indus populations living in various areas (e.g., Danino 2010; Kenoyer 1997; Lal 2002; Mughal 1997; Tripathi et al. 2004; Valdiya 2002; Yash Pal et al. 1980). However, the lack of dates for the perennial ﬂow of water in this paleochannel and the lack of clarity about the source of that water means that this claim is largely speculation. Although traces of a paleochannel were observed on the ground in Rajasthan and Punjab in the nineteenth century, today it is primarily known thanks to a large linear feature visible on satellite imagery (Bhadra, Gupta, and Sharma 2009; Yash Pal et al. 1980). Analysis of sections exposed in wells and electrical resistivity surveys in various locations along the paleo-Ghaggar/Hakra in northwest India have suggested that this feature was one or more large relict river channels (Saini et al. 2009; Sinha et al. 2013; Mehdi et al. 2016). There have now been several attempts to date the ﬂow of perennial water through these paleochannels, in both Pakistan and India. There is growing consensus that the major paleo- channel ceased to be a perennial watercourse before the Holo- cene (Clift et al. 2012; Giosan et al. 2012; Lawler 2011:23; Saini et al. 2009). However, there is some evidence of water ﬂowing through various channels during the mid-Holocene (Clift et al. 2012; Giosan et al. 2012; Maemoku et al. 2012; Saini et al. 2009; Shitaoka, Maemoku, and Nagatomo 2012) and ongoing debate about whether the paleo-Ghaggar/Hakra was an earlier course of the modern Sutlej (Lawler 2011:23) or an earlier course of another river, perhaps the Yamuna (Clift et al. 2012). Despite being visible on satellite imagery, the fact that a large river channel is not visible on the ground in many areas This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement than seeking to explain the Cholistan settlement concentra- tion by proposing that summer monsoon rainfall is capable of supporting perennial river ﬂow, an alternative presented by a critical reexamination of the dynamics of the Cholistan settlement system. As published, the Cholistan survey data show that there were considerable numbers of sites occupied in the preurban, urban, and posturban Indus periods, each of which were up to ﬁve centuries in duration (Mughal 1997). As is true for many regions of the world, it is assumed that settlements were occupied for the entirety of each period. Unfortunately, we know very little about the lifeways of the people living in these settlements as no excavations have been published, and we have no robust data on local subsistence practices, geomorphology, or hydrology. We do know, however, that very small numbers of the Indus settlements were occupied in consecutive periods and that, in each period, settlements were concentrated in different parts of the survey area. Use of the Dewar algorithm (1991) to assess settlement contempo- raneity in the Cholistan data has suggested there is a reason- able statistical likelihood that as few as 5%–10% of settle- ments may have been occupied contemporaneously during the preurban, urban, and posturban periods (Petrie and Lynam, forthcoming). In contrast, using the same algorithm to anal- yze the data from the Rakhigarhi Hinterland Survey from northwest India reveals a signiﬁcantly high degree of contem- poraneity of occupation at settlements during the preurban, urban, and posturban periods in that region (75%; Petrie and Lynam, forthcoming). These data suggest that it may be a mistake to assume that the large numbers of settlements re- corded for each phase in Cholistan represent concentrated and dense settlement. Rather, Cholistan may have been charac- terized by an unstable settlement system with little continuity of occupation between periods at individual settlements, and only a subset of settlements may have been occupied at any one time. While this suggestion is provocative, we currently lack the data to determine whether it is sound, and the river systems of Cholistan undoubtedly require further detailed investigation. Instability in the Cholistan settlement system may have been a product of the operation of a braided river system, which would have been susceptible to the frequent small-scale avul- sions during the periods of ﬂooding that occur during mon- soon rains. Such an environment may have required settled populations to be relatively mobile in order to survive stantly shifting hydrology, and there may have been high pop- ulation mobility between settlement locales. Individual fami- lies or kin groups potentially spread their members between multiple settlements, and individuals or groups might have moved between settlements to access available water in times of shortage or stress. Such practices clearly have implications for our understanding of the degree to which Indus popula- tions were adapted to a diverse environment and the sus- tainability and resilience of those adaptations. This content downloaded All use subject to University 000 Figure 5. A, Geoarchaeology at Masudpur I showing an allu- vially thickened, organic sandy-silt-buried soil horizon beneath ca. 1 m of archaeological deposits. B, Close-up of a mi- cromorphological slide taken from this section. A color version of this ﬁgure is available online. This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement Figure 6. Distribution of urban-phase Indus settlements relationship to mean annual rainfall (1900–2008). are shown in white. A color version of this ﬁgure that lie along the Himalayan front and at the eastern edge of the plains (ﬁgs. 2B, 3B, 4B). Today these areas receive more than 300 mm of direct monsoon rain per annum (ﬁg. 4B), which suggests that they are likely to have received some rainfall even during periods of weaker monsoon, though this remains to be This content downloaded All use subject to University 000 properly understood. In particular, it needs to be determined whether and when speciﬁc habitats and environmental con- texts were being selected preferentially. This reassessment of the evidence for the distribution of settlements in northwest India suggests that the local Indus populations probably employed a range of approaches to land use, even before cities developed. Perhaps the key element is that for populations to have lived in such environmentally diverse areas, their agricultural systems must have been far more ﬂexible and adaptive to local conditions than is usually acknowledged. In some areas of northwest India, rainfall may have been sufﬁcient to grow crops without irrigation, while in others, various methods of low-cost irrigation or active water management (bunds, canals, etc.) may have been essential. It is thus likely that the ancient populations, in this area at least, made use of whatever water was available, whether it was from rainfall, runoff, and overbank ﬂooding or water ﬂow from streams and rivers (cf. H. M.-L. Miller 2006, 2015; Wright 2010:33–34). It is also likely that attempts were made to cap- ture and store water in ponds and tanks and to access under- ground water using wells, as is prevalent among modern pop- ulations (Petrie 2013; Singh et al. 2008). Although canal-based irrigation is frequently dismissed as a contributing factor in Indus farming practices, Chakrabarti (1988, 1999:327) has long argued that it must have played a critical role (cf. Francfort 1992; Gentelle 1985, ﬁg. 14). The identiﬁcation of evidence for irrigation (or its lack) should be a priority of future and similarly, the role of ponds and tanks requires vestigation, as both were potentially very signiﬁcant the Indus period. The Material Culture and Subsistence Practices of Village-Based Early and Mature Harappan Populations in Northwest India Although there is clear evidence for the widespread use of a range of distinctive material culture items and practices ing the urban phase of the Indus Civilization, it is the degree of material uniformity has been overstated 2013). When excavations at Indus settlements are published, it is the typically classic Indus material (e.g., seals, on-red decorated pottery) that is highlighted. However, of other cultural material is also recovered, and there eral instances where regionally distinct material, orated ceramic vessels and ﬁgurines, were produced locally. For example, excavations at the urban-phase mana in northwest India have shown that the population of this town-sized settlement predominantly used locally and distinctively decorated ceramic vessels (comprising the assemblage) and made relatively limited use of the more distinctive classic Indus ceramics well known from Rakhigarhi, Kalibangan, Harappa, and Mohenjo-daro (Uesugi 2011:179ff.). Excavations carried out as part of the Land, Water and Settlement project have deepened our comprehension of this This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement Figure 8. Classic Harappan ceramics from Farmana and local ceramics from Masudpur I and VII (after Uesugi 2009, 6.145, 6.161). A color version of this ﬁgure is available 2016; Petrie and Bates, forthcoming; Petrie et al. discovery conﬁrms earlier indications that these being used in this region (Saraswat and Pokharia 2002, 2003) but goes further by dating their exploitation using material culture indicators and direct accelerator trometry radiocarbon dates to the Early, Mature, and Late Harappan phases (Petrie et al. 2016). Millet appears to have been the dominant crop in all phases at both sites, the second-most abundant crop at Masudpur I, appearing in higher quantities and proportions than either wheat (Bates 2016; Petrie and Bates, forthcoming; Petrie tables S2, S3). These new dates conﬁrm that summer being used alongside winter crops before, during, and existence of the Indus urban center at Rakhigarhi, ferent to what is seen at Farmana (Weber, Kashyap, 2011). The excavations by the Land, Water and Settlement project thus conﬁrm that there was diversity in material between regions and between different types of Indus settle- ments within regions. They have also deﬁnitively that different subsistence pathways involving combinations of winter and summer crops were used in different areas and that there was marked diversity in the crop assemblage some regions before the Indus urban phase. They thus sug- gest that Indus populations in some regions were well adapted to living in diverse and changeable ecological and environ- mental conditions and were thus well placed to make sus- This content downloaded All use subject to University 000 Figure 10. Carbonized crop grains from Masudpur VII, includ- ing carbonized rice grains (Oryza; A) and carbonized millet grains (Echinochloa; B). This paper has outlined a wide range of new evidence that encourages the reconsideration of several aspects of the nature of the Indus Civilization, particularly the environmental and climatic context within which urbanism developed and, ulti- mately, declined. It is not yet possible to establish adequately how Indus populations responded to the change in rainfall patterns that affected the plains of northwestern South Asia ca. 2200–2100 BC. The evidence for climate change at a local scale indicates that there were clear changes to the patterns and intensity of summer rainfall in northwest India. Given the degree of environmental variation within the Indus zone and the range of adaptations to farming that were being used across it, it is likely that these changes in summer rainfall would have had a differential impact, with some regions feeling the change directly and perhaps acutely, while others would have been impacted indirectly, if at all. Ascertaining the nature of this differential impact is an obvious topic for future research. The new archaeobotanical data produced by the Land, Wa- ter and Settlement project shows that models arguing that the collapse of Indus urbanism was caused by a shift in the sum- mer monsoon (Staubwasser and Weiss 2006), which led to the diversiﬁcation of the crop suite used, including the widespread adoption and/or more intensive exploitation of rice and mil- let (Giosan et al. 2012; Madella and Fuller 2006), are overly simplistic. They are also potentially paradoxical, as it was in northwest India that there appears to have been a reduction in the quantity of the summer rainfall needed to water these summer crops and potentially aid the growing of winter ones. This content downloaded All use subject to University of Petrie et al. Investigating Land, Water and Settlement litical structure is evident. On the basis of current data, it appears that in Sindh, the city at Mohenjo-daro was signiﬁ- cantly depopulated during the ﬁnal urban phase and there was a reduction in the intensity of settlement in the region gen- erally (e.g., Joshi, Bala, and Ram 1984; Possehl 2002:212, 241, table 13.2). By ca. 1900 BC in Cholistan, the largest settlements were abandoned or reduced in size and almost all others were displaced (Mughal 1997:51–52), while in Punjab, major set- tlements, including Harappa, reduced in size (Kenoyer 2005, 2008; Wright 2010:310). Analysis of pathologies visible on skel- etons from Cemeteries R37 and H at Harappa, which span this protracted period of transition, has revealed evidence for vari- ous infections and diseases, including leprosy and tuberculo- sis, which indicate deteriorating health (Robbins Schug and Blevins 2016; Robbins Schug et al. 2013a, 2013b). The response in Haryana and Gujarat is visible in the abandonment of large settlements and a focus on smaller town or village-sized set- tlements. The review presented here highlights internal dynamics and frames them in relation to a changing climatic context, but we also know that other cultural dynamics were also at play within the Indus zone and the surrounding regions. These in- clude the deterioration of trade through the Persian Gulf, the increased evidence for contact with the populations of inner Asia (e.g., Bactria Margiana Archaeological Complex, or BMAC), and the establishment of new settlements in borderland areas whose inhabitants used distinctive material culture (e.g., Pirak, Jarrige, and Santoni 1979; see Wright 2010:228–230, 308–325). Precisely how all of these developments interrelate and, in turn, articulate with a weakening of the summer monsoon is as yet unclear, but it is possible that climate change intro- duced a degree of entropy into a very complex and interactive urbanized system, potentially creating unpredictable unpre- dictability. Large cities and high local population densities may have become unsustainable, but sustainability, resilience, and continuity may have been possible by embracing rural life- ways that saw the maintenance and dispersal of diverse ap- proaches to substance. The need to respond to climate change is only one factor that might have inﬂuenced Indus cultural transformation and the adaptation of Indus substance prac- tices to a range of ecological zones, and the resilience of these adaptations in the face of climatic and social change remain critical topics for future research. Acknowledgments The collaborative Land, Water and Settlement project laid the platform for the TwoRains project, which was awarded fund- ing by the European Research Council in 2015 and will run until 2020. The Land, Water and Settlement project was pri- marily funded by a Standard Award from the UK India Edu- in the Department of Humanities at Universitat Pompeu Fabra (c/Trias-Fargas 25-27, 08005 Barcelona, Spain). F. Lynam is Head of Interactive Development at Noho (46 South William Street, Dublin 2, DO2 EE02, Ireland). S. Neogi is Postdoctoral Research Fellow in the Department of Cultural Studies and Archaeology, Ludwig-Maximilians University, Munich, and Research Associate at Murray Edwards Col- lege at the University of Cambridge (Huntingdon Road, Cambridge CB3 0DF, United Kingdom). V. Pawar is Assistant Professor in the De- partment of History at Maharshi Dayanand University Rohtak (Rohtak, 124001 Haryana, India). This paper was submitted 23 XI 15, accepted 6 III 16, and electronically published 27 I 17. Research. All rights reserved. 0011-3204/2017/5801-0001$10.00. DOI: 10.1086/690112 downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Current Anthropology Volume 58, Number 1, February 2017 ability, and resilience, it is essential to consider local climatic and environmental conditions (e.g., Aimers and Hodell 2011; Dixit, Hodell, and Petrie 2014; Madella and Fuller 2006). In fact, it is arguable that an understanding of the local context is essential for establishing whether past human societies were willing, able, or, in fact, required to respond to global-scale pres- sures and potential threats. Focusing on the local context also allows for nuanced ex- ploration of the relationships between adaptation and resil- ience. While resilience can be viewed in terms of response to distinct step changes in climatic systems, behavior may al- ready be adapted to ecological regimes that are intrinsically variable during single years and between years, which may make them predisposed to resilience. This ﬁts neatly with what N. Miller (2011) has described as “predictable unpredictabil- ity,” where populations make use of subsistence and cultural strategies that are tailored to absorb and mitigate risk. This paper will explore the nature and dynamics of adap- tation and resilience in the face of a diverse and varied envi- ronmental and ecological context using the case study of South Asia’s Indus Civilization (ca. 3000–1300 BC), and although it will consider the Indus region as a whole, it will focus primarily on the plains of northwest India. Most early complex societies developed in regions where the climatic parameters faced by ancient subsistence farmers were varied but rain falls pre- dominantly in one season. The Indus Civilization stands apart from other early complex societies for a number of reasons, but the signiﬁcance and ramiﬁcations of the speciﬁc environ- mental context within which it evolved is not widely recog- nized outside of Indus research circles. Importantly, the geo- graphical spread of the Indus Civilization spanned a very distinct environmental threshold, where winter and summer rainfall systems overlap and steep rainfall gradients are also evident. It, therefore, provides a unique opportunity to under- stand how an ancient society coped with both diverse and varied ecologies as well as change in the fundamental and un- derlying environmental parameters. to the The Cultural and Environmental Context of the Indus Civilization The Indus Civilization was one of the great early complex societies of the Old World, and during its urban phase (ca. 2600–1900 BC), it spanned large parts of modern Pakistan and India (Agrawal 2007; Chakrabarti 1999; Fairservis 1967, 1971; Kenoyer 1998; Lal 1997; Marshall 1931; Possehl 2002; Sankalia 1962; Wheeler 1968; Wright 2010). The Indus Civi- lization has, however, been marginalized or excluded in much of the comparative literature on early complex societies, which is unfortunate, as it has much to contribute. For example, it has been argued that the Indus Civilization does not neatly conform to the prevailing models for early complex societies; for instance, the major Indus architectural structures that have been exposed do not match expectations of monumentality (e.g., Possehl 1998; although see Yoffee 2005:228–229). While from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 Kashyap 2016; Weber, Barela, and Lehman 2010), settlement systems (Petrie 2013), and the production and use of particu- lar categories of material culture, most notably ﬁgurines and ceramic vessels (e.g., Parikh and Petrie, forthcoming; Petrie 2013; Uesugi 2011). It has long been recognized that there is considerable vari- ation in climate, hydrology, and ecology across the extensive area in which Indus settlements are found (e.g., Agrawal and Sood 1982; Joshi 1984; Possehl 1982, 1992; also Chakrabarti 1999:153–160; Shinde et al. 2006; Wright 2010:166–170), but the speciﬁcs of this diversity and the degree to which it maps onto cultural variation has not been addressed in detail. En- vironmental factors undoubtedly placed speciﬁc constraints on cultural behavior and the choices open to the inhabitants of the various Indus regions, and it is arguable that compre- hending the ways in which humans interacted with diverse and potentially changing environments over time and across space is critical for understanding the rise, ﬂoruit, and decline of Indus urbanism (cf. Agrawal and Sood 1982; Petrie 2013). The underappreciation of the degree and implications of cultural and environmental variation across the Indus zone is particularly telling when it comes to explaining the decline and ultimate abandonment of the Indus urban centers. This process appears to have been accompanied by a reduction in settlement density in the western and central parts of the In- dus zone and an increase in the number of village-sized set- tlements in its eastern reaches (i.e., Haryana/Punjab and Gu- jarat; ﬁg. 1B). Indus urban decline has been referred to as a collapse or a transformation, and from the beginnings of re- search on the Indus Civilization, both natural and human factors have been invoked as likely causes (e.g., Allchin 1995; Marshall 1931; Possehl 1997a, 1997b; Ramaswamy 1968; Wright 2010). There is, however, no consensus as to which factors are the most signiﬁcant, and there have been substantial gaps in the evidence that might enable us to assess the process as a whole. These gaps include a shortage of focused research on the socioeconomy of the posturban and subsequent pe- riods, a lack of absolute dates, and little high-resolution cli- matic and environmental evidence directly from the region. Given these limitations in the evidence, it is perhaps un- surprising that there has been no agreement about the sig- niﬁcance of climate and climate change on the Indus Civili- zation. Some have argued that there is no conclusive evidence to show that there is any difference in annual rainfall patterns between 6000 BP and the present (e.g., Kenoyer 1997, 2008: 186; Possehl 1997a), while others have posited climate change as the primary cause for the collapse and/or transformation of the Indus Civilization (e.g., Clift and Plumb 2008:205–210; Giosan et al. 2012; Shinde et al. 2006; Staubwasser and Weiss 2006). Within the diverse zone occupied by Indus populations, environmental factors related to hydrology were certainly important, and shifting/drying rivers and ﬂoods have long been proposed as major culprits. There have, for example, been extensive arguments made for and against the impact of ﬂooding at Mohenjo-daro (e.g., for: Raikes 1965, 1968; Raikes from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). (including sites with Kulli and Sorath-Harappan material; A) and post- to the distribution of mean annual rainfall recorded between 1900 and 2008. from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 thermore, when climate has been invoked as a critical driver of social change, there has been a reliance on distant climate proxy data sets for support (e.g., Giosan et al. 2012; Staub- wasser and Weiss 2006), which is at least partly because of the lack of proximate proxy data that might inform us about its impact on the diverse local context. A range of climate proxy data is certainly available from various locations in the sub- continent, particularly from dry lakes in Rajasthan (e.g., Enzel et al. 1999; Prasad and Enzel 2006), and new proxy data sets continue to become available (e.g., Leipe et al. 2014; Prasad et al. 2014; Sarkar et al. 2015), but they are typically not proxi- mate to the Indus zone. Unfortunately, the highest-resolution proxy data currently available comes from regions far outside the Indus zone that are characterized by different weather sys- tems (e.g., Oman: Fleitmann et al. 2003; northeast India: Ber- kelhammer et al. 2012), while the more proximate data sets are either lacking in chronological precision or do not actually cover the critical period of the late-third and early second mil- lennium BC (see Madella and Fuller 2006:1287ff., ﬁgs. 2, 9; Possehl 1999:259–263, ﬁg. 3.112). Until recently, the most di- rect insights from within the Indus zone have come from modeling of the Intertropical Convergence in central Punjab (Wright, Bryson, and Schuldenrein 2008:42–43; see below). While top-down approaches that rely on distant proxy data provide broadscale context, they do not provide the level of bottom-up local-scale detail necessary to evaluate the nature of regional dynamics across a large and ecologically varied ex- panse. Arguably, such resolution is essential for establishing the nuances of local climatic and environmental conditions and whether human societies of the past were willing, able, or even required to respond to pressures and threats. Given that the Indus Civilization spanned a large and en- vironmentally diverse area, it is unlikely that climate change would have had identical or even comparable effects in all re- gions. Similarly, hydrological shifts that may have been dev- astating in one area might have had no direct impact in others or may even have been beneﬁcial. Furthermore, human be- havior was likely already adapted to ecological regimes that are intrinsically variable between seasons and between years (see Wright 2010:25–44, 312–313, 315–319). Comprehension of the interrelationships between past climate and environ- ment and human actions and reactions can result only from integrated approaches and collaborative research projects that seek to identify the interconnections between the archaeolog- ical evidence and the evidence for climatology, hydrology, sed- imentology, and even ethnography, which are fundamentally interrelated but are too often treated as independent data sets. This paper integrates research carried out as part of the Land, Water and Settlement project, which conducted col- laborative work in northwest India between 2007 and 2014 (http://www.arch.cam.ac.uk/rivers/). It reviews the evidence for environmental diversity in northwest South Asia, assesses the ramiﬁcations of recently obtained data on the ancient hydrology and climate of northwestern India, and presents new archaeological evidence relating to geomorphology, settlement from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). (A) and post-urban-phase Indus settlements (B) and their relationship to from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 Weiss (2006) suggested that the mid-Holocene was charac- terized by high intra-annual rainfall variability in an increas- ingly arid climate, but we have little comprehension of the nature of this variability on the ground. Wright, Bryson, and Schuldenrein (2008) have used macrophysical climate mod- eling to make predictions about the intensity of summer and winter rainfall at Harappa between 14,000 BC and AD 2000. They modeled a protracted period of reduced rainfall between ca. 2100–1600 BC, which corresponds to the period of Indus urban deterioration and was attributed to a reduction in both winter and summer rain. Wright, Bryson, and Schuldenrein (2008) make it clear, however, that it is not feasible to extrap- olate this record to other regions within the Indus zone (see also Balbo et al. 2014). It is important to remember that beyond rainfall itself, an abundance of perennial and ephemeral rivers and streams re- distribute water coming from the winter rains, snowmelt, and summer monsoon, and these all inﬂuence the hydrological systems of the Indus zone (Flam 1993, 1999, 2013; Jorgensen et al. 1993; Wright, Bryson, and Schuldenrein 2008). Further- more, in addition to rainfall and hydrology, there is variation in the underlying geology, soils, and geomorphology, and sim- ilar degrees of variation invariably existed in these elements in the past (e.g., Belcher and Belcher 2000; Schuldenrein, Wright, and Khan 2007). The available data thus indicate that the region inhabited by Indus populations was marked by considerable diversity in the distribution of winter and summer rainfall and variation in the quantity and intensity of rainfall in any one season in any one year. The Indus zone is thus “predictably unpre- dictable” in multiple ways. The variation in water supply com- bines with signiﬁcant variation in hydrology and soils to cre- ate a broad zone comprising numerous ecological niches. All of these parameters enabled and/or constrained the types and range of subsistence practices that were possible and thus frame our understanding of Indus adaptation and resilience to climate change and the relationship of these factors to Indus urban decline. Coping with Environmental Diversity the east, Within the broader context of overlying environmental vari- ability driven by climatic gradients, it is clear that Indus pop- ulations also occupied a diverse range of ecological niches or habitat zones. The Indus Civilization has long been regarded as riverine (e.g., Marshall 1931), and while many major In- dus settlements were located close to rivers (e.g., Harappa, Mohenjo-daro, Lothal), others were located in intermontane valleys (e.g., Dabar Kot, Periano Ghundai) or on alluvial fans (e.g., Nausharo and Ghandi Umar Khan) at the margins or in- side of what are today arid zones (sites in Sindh, Cholistan, and Gujarat), in areas that lack perennial rivers but are watered by monsoon rainfall (sites in Haryana and east Punjab), and even on islands (e.g., Dholavira; Petrie 2013; Petrie and Thomas 2012; Wright 2010:33–38). from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). (A) and post-urban-phase Indus settlements (B) and their relationship to of this ﬁgure is available online. 8 from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). (A) and post-urban-phase Indus settlements (B) and their relationship to of this ﬁgure is available online. 9 from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Current Anthropology Volume 58, Number 1, February 2017 to have been variable, and it has often been argued that two broad zones can be differentiated, with the predominant use of winter crops (rabi—wheat, barley, pea, lentil, chickpea) in some areas and the predominant use of summer crops (kharif—millet, rice, tropical pulses) being evident in others (Fuller 2006, 2011; Fuller and Madella 2001; Kajale 1991:173; Madella and Fuller 2006; Meadow 1996:398–400; Pokharia, Kharakwal, and Srivastava 2014; Weber 1999:818–822, 2003: 180–185; Weber and Kashyap 2016:9, 11, ﬁg. 1; Weber, Kash- yap, and Harriman 2010:36–37, ﬁg. 1; Wright 2010:169–170). It is also asserted that there was an increased use of summer crops from the beginning of the second millennium BC on- ward, and Wright (2010:43) has suggested that this agricul- tural diversiﬁcation may have been a response to ecological challenges. Variation in practices is typically presented through com- parison of Harappa in Punjab, which shows the predominant use of the winter cereals barley and wheat and the limited use of summer crops such as millets (Panicum) in what has been described as a complex multicropping system (Weber 2003: 181), and Rojdi and Babar Kot in Gujarat, which show an almost complete focus on summer crops (e.g., Reddy 1997, 2003; Weber 1991, 1999:816–818; Weber and Kashyap 2016; Weber, Kashyap, and Harriman 2010; Wright 2010:169–170; see also García-Granero et al. 2016). Winter and summer crops have been reported from several sites in northwest India, in- cluding rice and millet from preurban/Early Harappan pe- riod contexts at Banawali, Balu, and Kunal (e.g., Saraswat 2002; Saraswat and Pokharia 2002, 2003; Saraswat et al. 2000). It has, however, been argued that these attestations should be dis- counted because of a lack of quantiﬁcation in the ﬁnal pub- lications, and a lack of direct absolute dates (e.g., Fuller 2006: 13, 16; 2011; see Petrie et al. 2016). Furthermore, winter and summer crops are also seen at Farmana in northwest India, though rice is not present in the stratiﬁed contexts, but the signiﬁcance of this is difﬁcult to interpret as only presence and absence information for macro- and microbotanical re- mains are provided, alongside summative ﬁgures for seed density and ubiquity (Kashyap and Weber 2013; Weber and Kashyap 2016; Weber, Kashyap, and Mounce 2011, tables 11.1, 11.2). Given that we lack published quantiﬁed assemblages from most Indus sites where archaeobotanical analysis has been carried out, it is likely that interpretations based on contrasting Harappa and sites in Gujarat are too simplistic. The problems are partly related to coverage but also interpretation. For in- stance, as noted above, it has been argued that the cropping system at Harappa was a complex multicropping strategy (Weber 2003), which may have been a response to ecological challenges (Wright 2010:43). The published evidence that in- cludes quantiﬁcation (e.g., Weber 2003), however, suggests relatively restricted use of crops grown in the nondominant season. It could thus be argued that such low proportions of summer crops do not actually indicate extensive multicrop- ping (Petrie and Bates, forthcoming; Petrie et al. 2016). Petrie and Bates (forthcoming) and Petrie et al. (2016) have, there- from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 with this range of environments both form a critical backdrop to debates about the impacts of climate change. As noted above, until recently, however, debates about the impacts of climate change on Indus populations have been hampered by a lack of direct and proximate climate data. Proximate data is essential for establishing whether there was any local impact of globally detectable climate change on the plains of northwest South Asia during the Holocene. New proxy records have been collected from within the Indus zone as part of the Land, Water and Settlement project, and these inform understanding of variation in the climate affecting Indus populations. The most relevant is the climate proxy record from Lake Kotla Dahar in southern Haryana (ﬁg. 1A, 1B), which indicates that there were two distinct shifts in rainfall distribution and intensity during the mid-late Ho- locene that affected northwest India (Dixit, Hodell, and Petrie 2014). In the early Holocene, Kotla Dahar was a deep lake, implying regular and consistent rainfall input to offset evap- oration, which given its location, would have been primarily monsoonal (Dixit, Hodell, and Petrie 2014). The ﬁrst shift oc- curred at some point between ca. 4400 and 3760 BC (ca. 6400– 5760 BP), when there was a decrease in monsoon rainfall and a progressive lowering of the lake level. This initial shift is roughly coincident with the evidence for change from Did- wana (Zone D5) and Lunkaransar (Zone 3) lakes in Rajasthan, though there are no reliable dates for the transitions at either (Enzel et al. 1999; Madella and Fuller 2006; Prasad and Enzel 2006). The second of these changes is more directly relevant, as it shows Kotla Dahar becoming completely ephemeral ca. 2200– 2000 BC (ca. 4100 5 100 BP) as a result of an abrupt weak- ening of the monsoon (Dixit, Hodell, and Petrie 2014). This shift in the monsoon is visible as a 300 5 100-year event in speleothem records in Oman (Fleitmann et al. 2003) and north- east India (Berkelhammer et al. 2012) and appears to match a change in levels of discharge from the Indus between ca. 4200 and 3600 BP (Staubwasser et al. 2003). The Kotla Dahar evi- dence indicates that the shift in the intensity and extent of monsoon rainfall speciﬁcally in northwest India was both dra- matic and protracted, resulting in an ephemeral lake that con- tinues to the present (Dixit, Hodell, and Petrie 2014). The degree to which the data from Kotla Dahar might be extrapolated to other parts of the Indus zone is debatable, and the relationship between climate and culture change remains ambiguous. It is nonetheless tempting to highlight correla- tions. The weakening in monsoon strength ca. 2200–1900 BC appears to correlate broadly with both the maximum extent of occupation at Mohenjo-daro and Harappa and the onset of Indus urban decline, though this was not a uniform process (e.g., Wright 2010:43). The chronological correlation between the data sets is, however, imprecise due to the limitations of radiocarbon dating techniques in terms of precision (Dixit, Hodell, and Petrie 2014; Staubwasser and Weiss 2006). The Kotla Dahar proxy record suggests that climate must be formally considered as a contributing parameter in the process of Indus deurbanization, at least in the context of the plains of northwest India. This is, however, inevitably only from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Current Anthropology Volume 58, Number 1, February 2017 demonstrates that there has been a considerable alluviation in the channel since perennial ﬂow ceased. The precise subsur- face architecture of the paleo-Ghaggar/Hakra and Punjab hy- drological systems, the date at which particular channels car- ried perennial water, and the source of that water continue to be debated, but there is a real possibility that the paleo- Ghaggar/Hakra did not carry water perennially during the Indus period. If this is true, it has profound implications for interpretations of the importance of this hydrological system for Indus populations, not least because it means that while water continued to ﬂow through the paleo-Ghaggar/Hakra season- ally, it was not a perennial river in the centuries before, during, or after the Indus Civilization. Taken together, the new data stand in contrast to a range of historical attestations to the existence of a mighty perennial river along this course in northwest India (e.g., Chakrabarti and Saini 2009; Danino 2010). It is, however, important to point out that the extant documentary records are unlikely to conform neatly to modern distinctions between ephemeral and perennial water ﬂow. What might have appeared as a mighty river in times of monsoon-induced spate may have been dry at other times of the year. If the water ﬂow in the modern Ghaggar is any indication, rivers in this environment can be virtually empty for much of the year and full to overﬂowing during the monsoon. Perhaps more importantly, the watercourse need not have been perennial to have been important for the ancient in- habitants. The new data suggest that the settlements along the course of the paleo-Ghaggar/Hakra such as Kalibangan, Ba- nawali, and Bhirrana were not sited to exploit a perennial river but to gain access to water via reliable annual monsoon runoff and overbank ﬂooding. Water was undoubtedly exploited for different purposes when it was available and captured and stored for use at other times, and it is likely that the paleo- Ghaggar-Hakra was important during the Indus period for reasons that are quite different to those usually claimed. Over- all, the likelihood that the paleo-Ghaggar/Hakra was not a pe- rennial river has important implications for the way in which Indus populations were adapted to a diverse and variable en- vironment and the type of responses that were needed when that environment changed dramatically. In addition to its implications for understanding Indus settlement systems in northwest India, the possibility that the paleo-Ghaggar/Hakra did not carry perennial water is partic- ularly signiﬁcant for understanding the evidence for extensive Indus settlement in Pakistani Cholistan. Giosan et al. (2012) have suggested that “reliable monsoon rains were able to sus- tain perennial rivers earlier during the Holocene,” which “ex- plains why Harappan settlements ﬂourished along the entire Ghaggar-Hakra system without access to a glacier-fed river” (3). However, the monsoon is unlikely to have provided a sustained source of water throughout the year; instead it pro- duces a charge between June and September to the hydro- logical system that may have otherwise been dormant. Rather from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 Geomorphological Evidence for Monsoon-Induced Annual Flooding possibility is The recent data from Kotla Dahar and the paleo-Ghaggar/ Hakra are congruent with the results of systematic geomor- phological analysis of the context of Indus settlements on the plains of northwest India by the Land, Water and Settlement project. Analysis of soil and sediment samples taken adjacent to settlements lying in two areas along the paleo-Ghaggar/ Hakra in central Harayana (Burj, Bhirrana, and Banawali) and northern Rajasthan (Dabli-vas Chugta and Kalibangan) has shown that, during the Holocene, the lower-lying areas in the landscape were probably more or less continually subjected to the slow, low-energy seasonal deposition of overbank ﬂood deposits composed of ﬁne sand and silt (ﬁg. 5A, 5B; French, Sulas, and Petrie 2014; Neogi 2013; Singh et al. 2010a, 2012). These sediments are composed of very ﬁne micaceous sands and silts, suggesting low energy water transport, and were pre- sumably deposited by runoff associated with monsoonal rains and riverine overbank ﬂooding, which lead to the seasonal ag- gradation of alluvium (Giosan et al. 2012; Singh et al. 2010a, 2012). This reconstruction appears to correlate with Clift et al.’s (2012) reconstruction of the mid-late Holocene Sutlej and Yamuna River drainage and Flam’s (1993, 1999, 2013) analysis of sedimentation in Sindh. Today areas along and adjacent to the paleo-Ghaggar are subject to ﬂooding and associated sedimentation during the monsoon. The new geomorphological evidence from the Land, Water and Settlement project suggests that this process has been active for some time and was undoubtedly important for Indus farmers in this region (French, Sulas, and Petrie 2014; Neogi 2013; Singh et al. 2010a, 2012). In turn, this has im- portant implications for considerations of the impacts of cli- mate change. The weakened summer monsoon in northwest India after ca. 2200 BC attested at Kotla Dahar would have resulted in, at minimum, a reduction in the intensity of that rainfall, which in turn will have decreased the amounts of an- nual overbank ﬂood-induced sedimentation and erosion. Mon- soon weakening will thus inevitably have had consequences for farmers relying on overbank ﬂooding to water summer crops and the concomitant stored soil moisture essential for the es- tablishment of winter wheat and barley. Settlement Distribution Data a con- To contextualize the new understanding of rainfall distribu- tion, climate change, hydrology, and geomorphology in north- west India, the Land, Water and Settlement project has also carried out extensive investigation of the settled landscape of this region. There is a sizable body of evidence for the distri- bution of preurban, urban, and posturban Indus settlements throughout Pakistan and northwest India (e.g., Joshi, Bala, and Ram 1984; Kumar 2009; Possehl 1999). These data have been used to build models of long-term sociocultural change from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Current Anthropology Volume 58, Number 1, February 2017 and highlight a potential shift of settlement toward the Gan- ges plains in the wake of the decline of the Indus urban cen- ters (Giosan et al. 2012; Joshi, Bala, and Ram 1984; Madella and Fuller 2006). The limitations of the core data set have, however, typically been overlooked. Detailed surveys in northwest India by the Land, Water and Settlement project have demonstrated that a signiﬁcant proportion of these data are fundamentally unre- liable. Both reconnaissance and detailed surveys have shown that there are signiﬁcant errors in the published locations of many sites, highlighted that the knowledge of site distribution and density is dictated by the intensity and extent of previous surveys, and established that large numbers of sites of all pe- riods have not been recorded (Pawar 2012; Singh et al. 2008, 2010b, 2011). These realizations have several important im- plications; for instance, it has frequently been stated that there is a close spatial correlation between the paleo-Ghaggar/Hakra and the distribution of Indus settlements and that there is a profusion of Indus sites along this channel in the area to the east of Kalibangan (e.g., Danino 2010; Lal 2002; Valdiya 2002). The Rakhigarhi Hinterland Survey (Singh et al. 2010b) and Ghaggar Hinterland Survey (Singh et al. 2011) and compila- tions of other extant survey data (Kumar 2009) have revealed, however, that there are actually relatively few sites that lie di- rectly along the course of the paleo-Ghaggar/Hakra for much of its course across northwest India (ﬁg. 6A, 6B). Despite its limitations, the extant survey data can be com- bined with the Land, Water and Settlement project results to show that urban and posturban Indus settlements were not speciﬁcally concentrated along any river channel but were, in fact, distributed across various parts of the plain. This distri- bution includes areas close to the paleo-Ghaggar/Hakra, areas adjacent to ephemeral watercourses and areas that have no relationship to any visible water sources, including the desert margin in northwest India (ﬁg. 6A, 6B; Petrie 2013; Singh et al. 2008, 2010b, 2011). Although there is no clear evidence for a large paleochannel in the vicinity of Rakhigarhi, the possibility that there is a subsurface channel in this area cannot be dis- counted, though its age and precise course—and, hence, its relationship to the ancient settlements—will only be recon- structed through a targeted study to this end. While there is general consensus that there was an increase in settlement in northwest India in the posturban Late Ha- rappan period, this conclusion is almost entirely based on in- ferences arising from the aforementioned unreliable survey data. Importantly, the Land, Water and Settlement project sur- veys have shown that there was no increase in the number of village-sized settlements in the central part of the plains during the posturban phase, which implies that there was no sub- stantial increase in the local population in these areas (Singh et al. 2010b, 2011). This observation suggests that if the per- ceived intensiﬁcation of village settlement in northwest India situated during the posturban/Late Harappan period is real, then it was concentrated elsewhere, most probably in the areas that are warm and temperate with dry winters and hot summers (Cwa) from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 (A) and post-urban-phase Indus settlements (B) in northwest India and their Major Indus sites and sites investigated by the Land, Water and Settlement project is available online. demonstrated. The cultural processes that led to this pattern of settlement have still not been examined systematically, and additional areas in Haryana and the broad region along the Himalayan front in both Pakistani and Indian Punjab need to be surveyed if the nature of settlement distribution is to be from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Current Anthropology Volume 58, Number 1, February 2017 regionality. At the smaller, village-sized sites of Masudpur I and VII (ﬁg. 7), which lie within the hinterland of Rakhigarhi, only region-speciﬁc styles of pottery were used during the ur- ban phase, and no classic Indus types were recovered from either the surface or the excavations (ﬁg. 8; Parikh and Petrie, forthcoming; Petrie, Singh, and Singh 2009). Other types of characteristically Indus material were present, however, in- cluding various types of beads and bangles (ﬁg. 9A, 9B), sug- gesting that the populations of these settlements remained connected to the interactive networks that linked Indus popu- lations more broadly. This evidence for regional variation sup- ports the suggestion that the widespread attestation of classic Indus material is actually a veneer that overlay a considerable degree of cultural diversity (Meadow and Kenoyer 1997:139; Petrie 2013). There is also evidence to suggest that a diverse crop as- semblage and, thus, diverse subsistence practices were being used in northwest India well before the posturban period. The combined macro- and microscopic analysis of material from systematically recovered samples collected at Masudpur I and VII have revealed evidence for the exploitation of both sum- mer and winter crops and, particularly, the preferential ex- ploitation of millet (both Echinochloa cf. colona and Setaria cf. pumila), rice (Oryza), and a range of tropical pulses includ- ing mung bean (Vigna radiata), urad bean (Vigna mungo), and horsegram (Macrotyloma cf. uniﬂorum; ﬁg. 10A, 10B; Bates research, focused in- during dur- arguable that (Petrie beads, black- a range are sev- including dec- and used site of Far- produced 80% of sites like Figure 7. Excavations being carried out in Trench XA1, Ma- sudpur 1. A color version of this ﬁgure is available online. from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 tainable and resilient decisions in the face of environmental change. The choices that Indus populations made in the face of such change all potentially revolve around the consolidation of the rural/agrarian baseline and include deurbanization (and decentralization), simpliﬁcation of craft practices, population displacement, and widespread adoption of diverse approaches to subsistence. Conclusions: Adaptation, Resilience, and Changing Perceptions of Indus Urban Decline There is much to learn from investigating the archaeology of human adaptation, resilience, and response to climatic and environmental change and the adaptive and resilience strate- gies that complex sociopolitical systems may have to engage in to survive. If we are to understand how humans coped with climate change, it is important to understand how they were adapted to particular environments and whether those adap- tations enabled populations to be resilient in the face of epi- sodes of climate change. For most ancient complex societies, water is a critical factor, and the availability of water and the way that it is managed and used provide critical insight into human adaptation and the suitability and resilience of sub- sistence practices. ﬁgs. 6.126, online. 2016). This crops were both relative mass spec- and rice is or barley et al. 2016, crops were after the which is dif- and Mounce culture both demonstrated within Figure 9. Indus material culture from Masudpur I and VII, in- cluding steatite, faience, and agate beads (A) and Indus-style ce- ramic bangles (B). A color version of this ﬁgure is available online. from 118.208.219.165 on January 28, 2017 21:03:54 PM of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Current Anthropology Volume 58, Number 1, February 2017 Rather than being forced to intensify or diversify subsistence practices in response to climatic change, the evidence from Masudpur I and VII for the use of millet, rice, and tropical pulses in the preurban and urban phases suggests that local Indus populations were already adapted to living in varied and variable environmental conditions before the develop- ment of urban centers. These environments are today marked by differences in ecology and are subject to considerable var- iation in rainfall patterns during individual years and between years, and similar patterns might reasonably be expected for the past. This pattern of ecological diversity and variable rain- fall reinforces suggestions that different strategies must al- ready have been adopted in different areas in response to dif- ferent ecologies (Petrie 2013; Singh and Petrie 2009; Weber, Barela, and Lehman 2010; Wright 2010). This variation in approaches to subsistence is also matched by a hitherto un- deremphasized diversity in the nuances of cultural practices that have been at least partly masked by the overt and widely used veneer of distinctive (or classic) elements of Indus mate- rial culture. Based on the work conducted by the Land, Water and Set- tlement project, we argue that it is this fundamental diversity in behavior, particularly in the proportional exploitation of winter and summer crops, that may have made it possible for populations in some areas to adjust to the dramatic weaken- ing in monsoon rainfall after ∼4200 BP/ca. 2200–2100 BC. We also argue that true insight into suitable strategies for surviving in variable environments that undergo change can come only by establishing the degree to which subsistence systems were adapted to local conditions and resilient to factors such as water stresses and the socioeconomic and political stresses that result from climate change. It will, however, only be possible to characterize the level of variation in subsistence practices across the Indus Civilization when evidence for the propor- tional exploitation of individual plant and animal species in a range of different regions is more widely available. The impact of climate change on the populations of the Indus Civilization more broadly will inevitably also reﬂect the level and nature of interaction between the populations living in different regions. Looking at a global scale, it is clear that the patterns of impact and response to climate change were ex- tremely variable (McAnany and Yoffee 2009; Miller, Moore, and Ryan 2011), and we should expect the same from Indus populations. Humans are unlikely to have been passive in the face of environmental change, and cities and civilizations did not simply disappear. Rather, populations adapt, adjust, move, die out, or thrive, depending on their circumstances. In the Indus context, we know that the ﬁnal phase of the urban period (the late Mature Harappan/Harappa 3C phase; ca. 2200–1900 BC) appears to be a phase of intensive inter- action, at least in terms of networks of raw material acquisition and redistribution (Law 2011). It is also apparently the period in which Harappa was most densely occupied (Kenoyer 1991: 57; 2005). It was, however, a period of transformation, such that by ca. 1900 BC, a very different socioeconomic and po- from 118.208.219.165 on January 28, 2017 21:03:54 PM Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). in Indus Northwest India 000 and 300 BC.” Smaller grants were also awarded by the British Academy’s Stein Arnold Fund, the Isaac Newton Trust, the McDonald Institute for Archaeological Research, and the Nat- ural Environment Research Council. The PhD research of Sayantani Neogi was funded by UKIERI, and that of Jennifer Bates was funded by the Arts and Humanities Research Coun- cil. We would like to thank a large number of individuals and institutions that have made this project possible. First, we thank A. Vaish, K. N. Srivastava, and Dr. G. S. Gupta (Director Generals, Archaeological Survey of India [ASI]), Dr. B. R. Mani (Additional Director General, ASI), and Dr. R. S. Fonia and Dr. S. P. Manik (Directors of Excavations and Explorations, ASI) for granting us permission to carry out the ﬁeld research. The collaborative agreement on which this project is based was signed by Prof. P. Singh, Vice Chancellor of Banaras Hindu University (BHU), and Dr. K. Pretty, Pro-Vice Chancellor of the University of Cambridge. We are also grateful to the current Vice Chancellors of BHU, Prof. D. P. Singh and Dr. L. Singh. We have been given abundant support by Prof. S. R. Dubey, former head of the Department of Ancient Indian History, Culture and Archaeology, BHU, and Prof. G. Barker, former head of the Division of Archaeology, University of Cambridge and Director of the McDonald Institute of Archaeological Re- search. We would also like to acknowledge the assistance, ad- vice and hospitality of Prof. A. Singh (Rohtak University), Dr. N. Ahuja (Jawaharlal Nehru University), Prof. M. K. Jones (Cambridge), J. R. Knox (formerly of the British Museum), and, particularly, Dr. R. Tewari (Director General, ASI, former head of Uttar Pradesh State Archaeology), Dr. K. S. Saraswat (for- merly of the Birbal Sahni Institute), and Dr. P. Joglekar (Deccan College, Pune), all of whom joined us in the ﬁeld. We would also like to offer grateful thanks to Profs. R. Meadow, K. Thomas, and R. Wright and to Dr. E. Cork for reading early drafts of this paper and offering a range of helpful suggestions, correctives, and safety ropes that helped us keep focused. We would also like to thank Prof. M. Madella, who has been a constant source of advice and direct support in several instances. Last, we would also like to thank the three anonymous reviewers who provided useful comments and criticisms. These readers may not com- pletely agree with the opinions expressed in the ﬁnal paper, but their comments have hopefully made our arguments more ro- bust and helped us make this a better and more rounded work. Comments William R. Belcher Division of Social Sciences, University of Hawai’i-West O’ahu, 91-1001 Farrington Highway Kapolei, Hawaii 96707, USA (

[email protected]

). 20 VII 16 cation and Research Initiative (UKIERI) under the title “From the Collapse of Harappan Urbanism to the Rise of the Great Petrie et al.’s article, at ﬁrst glance, appears to be related to Early Historic Cities: Investigating the Cultural and Geo- climatic change and understanding subsistence and settlement graphical Transformation of Northwest India between 2000 variation within a small region of the Indus Valley Civilization. This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 000 Current Anthropology Volume 58, Number 1, February 2017 However, the authors conduct a vast synthesis concerning sub- settlement pattern contemporaneity. Petrie and colleagues sistence (primarily agriculture) variation, material culture, set- present data that suggest the Ghaggar-Hakra River was not a tlement pattern, and hydrological regimes during the third perennial source of water. Additionally, the density of settle- millennium BC. The key focus for this study is diversity and ments along the area, at least in terms of the Cholistan region understanding how ancient peoples focused on adapting to a in Pakistan, were probably not all occupied at the same time; highly diverse and ﬂuctuating environment, both in the short in fact, Petrie et al. suggest that as few as 5%–10% of the term of generations and the longer term of almost a thousand settlements were contemporaneous. These two data sets sug- years of occupation in the region. The synthetic nature of this gest that many aspects of current presentations related to the article is hinted at in relation to the project title, Land, Water role of this river system to the Indus Valley Civilizations are and Settlement project; the research focus of this project now overly simplistic. continues with the interdisciplinary work of the TwoRains This article is data heavy and deserves a second or third project. reading to understand the implications of the research. Rec- While climate studies related to ancient societies have been ognition of the diversity of the ecological and hydrological an important part of archaeological research, these studies have setting is the ﬁrst step in beginning to understand the ﬂexible gained additional importance and a greater visibility within adaptations that many of the regional aspects of the Indus the ﬁeld due to the political and social signiﬁcance of under- Valley Civilization must have had. Petrie et al. state that some standing current and past climate change. Within the past populations within this large regional civilization were well 5 years, several studies have focused on the role of climate adapted to ecological change as reﬂected in subsistence strat- change and the decline of the Indus Valley Tradition (i.e., Giosan egies and water management, and thus we may need to rethink et al. 2012). This research looks at very large-scale global mon- our understanding of what it means for this civilization’s “end.” soonal models and applies these to a large, regional context such as the Indus Valley Civilization. While I disagree with some of the conclusions and applications of these large, global models, these studies have brought into the forefront the need for more Heather M.-L. Miller regional research. Anthropology, University of Toronto, Mississauga, 3359 Mississauga One of the most powerful aspects of this paper is its recog- Road North, Mississauga, Ontario L5L1C6, Canada (heather.miller nition of the amount of regional variation that exists in terms of @utoronto.ca). 1 VII 16 rainfall patterns, hydrological patterns, and monsoonal cycles across the area that encompasses the almost 1 million sq. km This paper makes a substantial contribution to the literature that is thought to represent the expanse of the Indus Valley on complex societies and environmental effects, using the In- Civilization. The authors’ review of the amount of diversity in dus Civilization case. The authors present both their own im- terms of settlement, material culture, and agricultural practices portant new data and an invaluable summary of similar col- that exist across this landscape is key to our understanding laborative projects from elsewhere in the Indus world. They these ﬂexible socioecological adaptations. However, only a few place these and other recent climatic studies in environmental sites possess the detailed information that allow for ecological and cultural context, employing approaches from both the and subsistence reconstruction. While most researchers rec- sciences and the humanities. Indus researchers, those looking ognize that there is diversity across this larger landscape, there to use the Indus as a comparative case, and scholars interested is still the pervading myth of cultural homogeneity within the in the relationship between environment and society broadly Indus Valley Civilization. By recognizing the more regional will all ﬁnd this paper of great use, especially as Petri et al. have aspect of various characteristics of the Indus Valley Civiliza- employed a clear, direct writing style. tion, the authors are able to address a much richer landscape Readers now expect the “but” explaining where I think this of socioecological adaptation, albeit for the speciﬁc region of research went wrong. It is not coming. It is encouraging to Haryana/northern Punjab area. see this presentation in a high-proﬁle anthropology journal of This multidisciplinary project focuses on creating a local- the excellent work done by Indus scholars on the environ- ized model for the Haryana/northern Punjab region using geo- ment and climate change over the past few decades (and more). logical, geomorphological, and archaeological data. By using It is even better to see a publication emphasizing the impor- speciﬁc regional data, the authors are ﬁlling out a detailed un- tance of considering human adaptability and traditions when derstanding of the environment for a speciﬁc region. This assessing the possible role of climatic change in social and cul- should set the stage and create a model for ecological research tural change, contributing this important anthropological per- for future directions. This study allows us to understand the spective to current debates in several of the highest-proﬁle gen- regional climatic changes that occurred in this area. However, eral science journals. Petrie et al. warn us not to extrapolate these data to the Indus I am particularly thankful to see that the authors stress the region overall. dual-rainfall nature of Indus climatic regimes, as well as the During the summary of data related to diverse hydrological multiplicity of environmental contexts found within the bor- patterns, there are two important points that should be em- ders of this expansive ancient society. Too many studies of phasized here: the hydrology Ghaggar-Hakra River and the ancient climate change for this region focus solely on changes This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Petrie et al. Investigating Land, Water and Settlement in Indus Northwest India 000 in the monsoon summer rainfall system and ignore the winter their animals and how this relationship between plant and rainfall regime. There is no question that the monsoon system animal food production systems might have changed through is and was critical for this entire region, and especially so for time, although this may be part of the work of the North Gu- the eastern areas within modern-day India, both for direct jarat Archaeological Project (Madella et al. 2010) or other rainfall and for runoff or collected water supply. The winter projects noted by Petrie et al. Similarly, ﬁsh and other wild rainfall regime seems to have also been critical for Indus ag- plant and animal foods would also be affected by climatic/ ricultural systems, though, at the least for western Indus re- environmental conditions; we know that ﬁsh were important gions and the greater Indus River ﬂoodplains in Punjab and for at least some Indus communities (Belcher 1998, 2003). Sindh, and we lose a great part of the potential ﬂexibility and Nonfood plant and animal products were likely critical aspects adaptability of ancient (and modern) agricultural systems if of Indus economic systems, particularly for ﬁber and cloth we do not consider both systems, as the authors emphasize. production, and climate-linked upsets in these production sys- The authors highlight the importance of examining local tems could have had devastating effects on exchange networks conditions as well as large-scale ones; through this, they pro- or tax/tribute systems required to sustain Indus social and po- vide pathways from such large-scale climate reconstructions to litical structures. an understanding of the role of human actions and reactions. As always, there is still plenty of work to be done. This does For example, looking at the local sources and methods of ag- not take away from the substantial contribution made by the ricultural water supply shows the complexity of water avail- researchers of the Land, Water and Settlement project and the ability for Indus farmers, beyond a simple rainfall-based cli- other projects discussed in this paper. Their example here of mate change model, as the authors note. Their discussion of how to assess the effects of climate change on plant-based ag- the paleo-Ghaggar/Hakra hydraulic regime is a useful update, ricultural systems, including an examination of existing meth- and I look forward to the forthcoming publication on the as- ods of adaptation and resiliency already practiced in the Indus sociated settlement patterns, particularly if they prove to be as tradition, have laid the groundwork for approaches that can be signiﬁcant as the reassessment of the patterns in northwest applied to all of the other production systems described above. India (Singh et al. 2010b, 2011). The authors are to be highly commended on the result. There is a common but signiﬁcant oversight here, found in most discussions of the effects of climatic/environmental change on the history of the Indus Civilization as a tradition and as a complex society: the tendency to look at only agri- Steven A. Weber cultural crops, more speciﬁcally only at plant foods, ignoring Department of Anthropology, Washington State University, the effects on animal and other plant production systems. (I Vancouver, 14204 Northeast Salmon Creek Avenue, Vancouver, am also guilty of this in the writings the authors cite, except Washington 98686, USA (

[email protected]

). 16 VII 16 for the weasely insertion of a line or a footnote indicating that animal food production is important but that I will not deal First, the authors are to be commended for taking existing with it.) Since the early days of Indus archaeology and espe- models regarding the unique nature of the diversity of Indus cially since the ﬁrst environmentally inclusive archaeological Civilization environment and building a more concrete un- projects of the 1970s and ’80s, researchers have noted that derstanding of the inﬂuence climate change had on agricul- animal husbandry was a major part of Indus food production tural practices and, subsequently, the evolution of this civili- systems. Mixed agricultural economies were certainly an im- zation. These are important issues where more data and more portant part of Indus food production, with crop farmers also thorough discussions are needed. The paper’s strength lies in keeping animals, perhaps in signiﬁcant numbers, as part of the its addition of new data (mainly from Haryana) and its syn- well-known mixed economy beneﬁts of animal and crop co- thesis of existing arguments regarding agricultural diversity production. But there must also have been pastoral commu- and the signiﬁcance of climate change, especially as it relates to nities at this time, communities more focused on animals than local environments. While these contributions are very valu- plants (e.g., Bhan 2011; Mallah 2011; Meadow 1996; Meadow able, the paper at times fails to adequately distinguish between and Patel 2003; Mughal 1997:59). Such communities could which ideas are completely new and novel and which argu- have followed a range of lifestyles, from seminomadic to pri- ments are, in essence, extensions of existing debates. I strongly marily settled, including the opportunistic growing of fodder. disagree with their conclusion that “robust evidence to sup- What would be the effects of climate change on such animal- port” the existence of ecocultural or agroecological domains is oriented communities? Were they even more dependent than not available. The importance of understanding the diverse farmers on rainfall for food production? What other strategies and varied ecologies of this civilization has been an accepted might they have had to deal with the “predictable unpredict- reality for some time as is the argument that the “proportional ability” of water and, therefore, fodder supply in these regions? exploitation of winter and summer crops” may have made it Aside from Reddy (1997, 2003) and Chase and colleagues possible for populations in some areas to adjust to climate (Chase 2014; Chase et al. 2014) for Gujarat and Thomas (2003) change. The contention that the cultural and environmental for Bannu in the far northwest, few Indus researchers have variation across the Indus zone is underappreciated, espe- modeled how such communities might have managed to feed cially when it comes to explaining the decline of the urban This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 000 Current Anthropology Volume 58, Number 1, February 2017 centers, is puzzling, as this is well recognized in the archaeo- Seeds have been individually dated from systematically col- logical record. The argument that to understand cropping at lected samples, and presently more than 200,000 seeds have any point in time in South Asia, due to its great regional en- been analyzed. The database is completely quantiﬁable. Such vironmental diversity, a local ecological approach was the a large number of seeds have taken many years to analyze, soundest was among the several ideas incorporated into Osada but the goal is to have the complete collection published and Toshiki’s project Environmental Change and the Indus Civi- available by 2018. While the various occupations at Harappa lization. This major project, supported by the Research Insti- (3300–1700 BC) show winter cropping as the most important, tute for Humanity and Nature (RIHN; 2004–2011), drew summer crops were being used throughout. During Period 3A several similar conclusions to those that appear here in this (2600–2400 BC) and again in Periods 4/5 (1900–1700), sum- paper and in other papers associated with the Land, Water and mer crops increased in use according to measures of ubiquity, Settlement project. For example, the RIHN project discussed density, and frequency. Based on this data, I have argued that how climatic stress was experienced in different regions of the communities like Harappa made themselves sustainable dur- Indus Civilization in different ways; in some areas, the effects ing periods of climatic stress because they were capable of har- were much more extreme and made it difﬁcult to produce the vesting both summer and winter crops. Unlike other large crops necessary to keep several large urban centers sustain- Indus cities, Harappa was not abandoned in spite of indica- able. One conclusion was that the civilization’s communities tions of stress and decline. that were based on diverse multiseasonal cropping (including In sum, we know a lot more about cropping and shifts in both summer and winter crops) with local processing and agricultural practices and adaptive abilities in different regions storage (often at the household level) were more adaptive to of the Indus Civilization than this paper acknowledges. Re- climate stress. gional cropping variability due to ecological variation has been One minor concern I have with this paper is the extensive going on for a long time in South Asia and continues today. use or reference to forthcoming publications. This came up Climatic shifts need to be understood at the local level if we more than 15 times in the paper, and while I appreciate the are to better comprehend the evolution of the Indus Civiliza- difﬁculty in the pace in which material gets published, it tion. It is only from sites where we have extensive horizontal means that the reader has to take the authors’ word for much exposures, with samples representing many distinct features of the new information without being able to judge its quality such as ﬂoors, storage bins, trash areas, streets, drains, burials, oneself. and so on that we can begin to appreciate the relationship of Finally, I am excited to see the authors’ efforts at construct- plants to the rest of the material record and thus to cultural ing a quantiﬁable and comparable archaeobotanical database. variability. Notwithstanding my caveats, the debate raised in This type of approach, until recently, has been relatively rare this paper concerning climate change, diverse ecologies, and for the Indus Civilization, although the need for such an ap- adaptability in the Indus Civilization is very welcome. proach is well understood and has slowly become accepted as the best approach (see Fuller and Weber 2006). Surprisingly, the article either misunderstands the results or challenges the value of the archaeobotanical data collected from Rojdi, Far- Rita P. Wright mana, and Harappa. Summer crops were not only clearly ev- Department of Anthropology, Center for the Study of Human ident at all three sites but were an important part of the agri- Origins, Rufus D. Smith Hall, 25 Waverly Place, New York, cultural strategy. Rojdi was one of the ﬁrst large quantiﬁable New York 10003, USA (

[email protected]

). 28 VI 16 archaeobotanical data sets collected in South Asia (Weber 1991). Evidence from the analysis of more than 14,000 seeds Petrie and his colleagues make an argument for the signiﬁcance showed that summer crops remained prominent even after the of climate change and diverse ecologies as factors in adapt- decline of the monsoons. At Farmana, phytoliths, starches, ability and resilience in the Indus Civilization. They also review and carbonized seeds were recovered and analyzed (Weber, the current results of their Land, Water and Settlement re- Kashyap, and Mounce 2011), and some of this material was search that will ﬁll in important gaps. While the authors appear individually dated to secure its age. The material shows that to agree that cropping patterns and water systems inﬂuenced while both summer and winter crops were in use throughout the growth and decline of the Indus Civilization, in spite of its occupation, the ubiquity and density of summer crops re- what they consider the poor quality of the current evidence, mained constant, while the winter crops declined during the they fail to consider the diverse social organization that enabled Late Period. Published data and recently completed analysis regionally distinct communities to adapt differently when con- indicate that summer crops were always important even dur- fronted with regional and global shifts in moisture systems. ing the mature occupation where they accounted for nearly True enough, we learn a lot about subsistence practices and 35% of the crops. Summer crop production never declined even ecological variability, but there is more to it than available with a decline in the monsoons. foods and climate. Cropping, processing, and management of The most interesting archaeobotanical data set, as well as plants and water directly and indirectly reﬂect social processes the largest from any Indus site, is from the city of Harappa. and social organization. As examples of the ways in which they This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Petrie et al. Investigating Land, Water and Settlement in Indus Northwest India 000 inﬂuence adaptability and resilience, I turn to good contextual existed at 3000 BCE, when the ﬁrst settlers came to the region, data for cropping patterns at Rojdi and Harappa and water but even then precipitation was marginal (Agrawal 2009). In management systems at Dholavira. spite of the challenges of low rainfall and a harsh environment, the city became one of the Indus’s major centers. Cropping Patterns, Diverse Ecologies, The ﬁrst settlers at Dholavira built a modest water system in which they channeled the ﬂow of an ancient runnel to pro- and Household Economies vide potable water and what appears to be a still-water sys- Rojdi, a 12-ha agricultural community in Gujarat, and Ha- tem (Scarborough 1993, 2003; Scarborough and Lucero 2010). rappa, a major center in the late and posturban phases of the Its construction at Dholavira is documented from the initial Indus, provide contrasting adjustments to climate change. These (preurban) period when it was cut into base rock for the stor- results are based on excavated and stratiﬁed levels collected age of potable water. Very likely, a small group of knowledge- from primary, secondary, and tertiary contexts, some of which able farmers built and managed the timing and maintenance are accelerator mass spectrometry dated. Of speciﬁc relevance of the ﬂow of water. Later in the urban period, rock-cut res- is the evidence for chaff and weed remains at both sites. Resi- ervoirs and dams were constructed from nonlocal stones that dues of chaff and weeds are indicative of the threshing, win- were procured from a distance, when the system was ex- nowing, grinding, and cleaning of grains (Weber 1996, 1999) in panded, perhaps due to a population inﬂux and expansion of households. At Harappa, they are present only in the latest the city. The new system was massive. Many more reservoirs phases of occupation, when the average area of the site dimin- and dams were constructed that provided water for a variety ished. During the urban phases, the city’s agricultural produc- of human needs, including drinking water at its highest tion was based on large-scale, seasonally based cultivation in reaches. As water cascaded through the terraced city of ap- which the processing of crops took place in ﬁelds (Weber 1999, proximately 100 ha, it reached the lower levels of the city 2003:181) and is suggestive of a community or centralized or- carrying potable and domestic water before being channeled ganization. In distinction, the presence of chaff and weeds in to agricultural ﬁelds (Bisht 1994, 2005, 2009). The complexity households reﬂects a political and social reorganization that may of this system and the substantial labor force needed to con- have resulted from ecological stress due to climate changes. struct and manage its ﬂow could have been managed and This pattern differed from the long-term settlement at Rojdi. monitored at each level by collective groups, as has been doc- Millet was a preferred crop, and this small-seeded and lower- umented elsewhere by a more centralized bureaucracy in view yielding grain was well adapted to a farming community with a of its complexity. The Dholavira data provide evidence for the lower population than Harappa (Weber 2007, 2010). Although signiﬁcance of water storage and the contribution of small- there were some changes in cropping patterns at Rojdi, the scale producers at Indus’s major centers. presence of chaff and weeds is consistent with household pro- Though Dholavira’s reservoir system did break down, we duction. Both Harappa and Rojdi continued to be occupied know from my discussion of Rojdi that populations remained after the critical 1900 BCE date invoked by Petrie and his col- at sustainable levels in interior settlements after Dholavira leagues. When the two settlements are compared, their ability was not longer inhabited. It was the dual presence of these to adapt to climate change differed. At Harappa, it was based very different agricultural strategies and the small-scale house- on a shift from a complex social organization based on a cen- hold production that contributed to the sustainability of set- tralized or communal agricultural system and a more sustain- tlement in Gujarat. As Weber and others have noted, these able and less complex social organization that was based in changes may be part of wider social process elsewhere (Fuller households. Unlike the major center at Mohenjo-daro, for ex- and Stevens 2012; Weber 1999, 2003). I look forward to the ample, Harappa continued to be occupied at least until 1700 BCE published results of the Land, Water and Settlement research (Wright 2010). At Rojdi and neighboring settlements, farmers and whether their results are consistent with ones docu- continued to be based on household production and the cul- mented at Dholavira, Rojdi, and Harappa. tivation of millets until at least 1700 BCE and possibly later (Weber 1998, 2007; Weber, Barela, and Lehman 2010). Water Management Reply The water source at Dholavira does not conform to the elab- orate systems envisioned by a model focused on irrigation. We would like to thank our four colleagues for their com- Rainfall at Dholavira is exceedingly low, and the landscape is ments, which in general are complimentary but also raise absent of a major river system.1 Better monsoon conditions speciﬁc issues and/or points of criticism. As Miller recognizes, our paper sets out to speak to a broad audience. We set out to 1. In a widely circulated video production of the archaeology at synthesize a range of evidence from the Indus case in a way Dholivira, a narrator speaks of a river system. In publications, Bisht is that hopefully allows us to emphasize how it speaks to themes explicit that the water drained off of a “runnel.” relevant to non-Indus as well as Indus scholars in both the This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 000 Current Anthropology Volume 58, Number 1, February 2017 sciences and humanities. We also stand by our aim to propel able climatic variation in the eastern Indus region during the the Indus example into broader debates, particularly those re- Holocene. We have also mapped modern climatic patterns in lated to adaptation and resilience. relation to Indus settlement distribution (ﬁg. 2A, 2B), and our All of the reviewers commend us for building on existing archaeological and archaeobotanical investigations at six sep- models of Indus environmental diversity to achieve a more arate settlement sites across this one region highlight the like- concrete understanding of the inﬂuence climate change had lihood that there was considerable variation in ecological and on its agricultural practices and evolution. However, Weber, human adaptation and response within this region in the and to some degree Wright, suggest that we do not appropri- past (Bates 2016; Petrie and Bates, forthcoming). In this re- ately acknowledge previous scholarship, particularly that docu- gard, we note that both Belcher and Miller highlight the sig- menting agricultural and environmental diversity in the In- niﬁcance of the local nature and amount of regional variation dus context. We disagree with this and a number of associated that we discuss. We thus emphasize that our work has different suggestions. objectives and outcomes to the RIHN project, and our con- Throughout our paper, we acknowledge that previous work clusions about adaptation and resilience are both novel and has addressed related themes and cite as much of this research robustly integrated. as is feasible, and we note Miller’s recognition that we highlight We certainly do not suggest that there are no robust data the range of “excellent work done by Indus scholars on the on archaeobotanical or palaeoenvironmental change available, environment and climate.” However, we also make clear that nor do we suggest that cultural and environmental variation there are important differences between our approach and re- across the Indus zone has not been recognized in the archae- sults and those of previous work, both in terms of theoretical ological record. We do, however, highlight that some of the framework and the empirical approach of investigating one available data sets are either partial or incompletely published. speciﬁc region in detail. We argue that the overall corpus of environmental evidence This leads us to respond to Weber’s particular claim that from the Indus zone that is published, including information Toshiki Osada’s Environmental Change and the Indus Civ- about proportions of summer/winter cropping, is not yet de- ilization project based at Research Institute for Humanity tailed enough to constitute a robust basis for building appro- and Nature (RIHN) drew several similar conclusions to those priately nuanced models of agroecological diversity or to test made by the Land, Water and Settlement project, especially their veracity. We also stress that there is likely to be consid- as he refrains from citing any supporting references. Osada erably more variation than is currently recognized. and colleagues have published a formidable array of papers Leaving aside our work, well-resolved archaeobotanical in- and edited volumes on their work, and there is no question that formation is typically available from one or a small number the RIHN project has been extremely important for Indus of archaeological sites in any one area across the Indus zone. archaeology in its investigation of sites in multiple regions. In The important data from Rojdi (Weber 1991, 1999), Babar many ways, the team’s key contribution has come via its col- Kot and Oriyo Timbo (Reddy 1997, 2003), and Harappa (e.g., laborative excavations at settlement sites, including Girawad, Weber 1999, 2003) can tell us only so much about Gujarat Farmana, Madina, and Kanmer, and impressively, ﬁnal pub- and Punjab, respectively, as both are ecologically and hydro- lications of these excavations were available within 3 years of logically diverse regions. Such limitations make consideration the completion of the ﬁeldwork. It should, however, be ac- of the impact of climate change even more challenging. Thus, knowledged that, to date, there has been no synthetic presen- while Wright (2010) and others have addressed the implica- tation of its results, and the project’s discussion of environ- tions of climatic and agricultural variation for processes of de- mental factors relevant to these sites is either preliminary or urbanization, we argue that we still lack enough fundamental limited (e.g., Lancelotti and Madella 2010; Pokharia 2012; data to ground those ideas and test the theories, and these Rajagur and Deo 2008; Weber, Kashyap, and Mounce 2011). limitations should be acknowledged. Further, the broader project outputs on climate and environ- In our view, the Land, Water and Settlement project pre- mental change have either not been proximate (lake data from sents an important contrast in approach, which has enabled us Nepal; Nakamura et al. 2016) or, in our opinion, adequately to demonstrate that northwest India in the Indus period is far resolved (e.g., a relatively limited number of optically stimu- more environmentally, agriculturally, and culturally diverse lated luminescence dates relating to river-shift investigations; than has ever been previously acknowledged. We argue that Maemoku et al. 2012; Shitaoka, Maemoku, and Nagatomo 2012). this level of intraregional diversity has ramiﬁcations for our We argue that far more proximate and nuanced analyses are understanding of the Indus zone as a whole—as well as the required to provide a well-grounded palaeoclimatic framework extent to which the available data permit adequate modeling for the Indus region as a whole, and more comprehensive anal- of socioecological systems, cultural diversity, and responses to yses are required to resolve important questions about ﬂuvial climate change. regimes. In a similar vein, we suggest that, overall, the Indus archae- The Land, Water and Settlement project has obtained prox- obotanical evidence base remains as yet insufﬁcient to resolve imate palaeoclimate data from Kotla Dahar (Dixit, Hodell, and important patterns of social and economic organization. We Petrie 2014) and two other locations across northwest India concur with Wright’s assertion that archaeobotanical analysis (Dixit et al. 2014, 2015), and these records attest to consider- offers extremely important insights into these issues, and she This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Petrie et al. Investigating Land, Water and Settlement in Indus Northwest India 000 rightly points out the importance of evidence of chaff and weeds also the critical role of the ancient winter rainfall regime, and for gaining insight into agricultural labor organization and the her point about the need for information on this weather sys- agricultural economy. With respect to information about local tem is well made. Clarity in this regard will come only from conditions and their relationship to agricultural cropping and palaeoclimate data sets obtained from Pakistani and Indian associated weeds, Weber’s (1992, 1999) detailed and compre- Punjab, which must be an objective for future research. hensive work at Rojdi and Reddy’s (1997, 2003) work at Babar Miller also rightly draws attention to the impact of climate Kot and Oriyo Timbo are exemplary and are published with change on nonfood crops, animal husbandry, and pastoral raw data, facilitating detailed (re)analysis by others. However, communities. Reduced availability of water is likely to have the work at these and the Land, Water and Settlement sites are affected the growing of ﬁber and oil products, including lin- the exception rather than the norm. As Weber notes, the full seed/ﬂax, mustard, and sesame, particularly as cotton and ﬂax assemblage from Harappa—without doubt, the most impor- are both thirsty crops. Further attention to both the role of tant Indus archaeobotanical assemblage from an urban site to nonfood crops in Indus economies—and how they may have date—is still under analysis. This includes the crop-processing been affected by environmental changes—is another clear av- and weed-ecology data that are essential for discussions of both enue for future research, as Miller suggests. The Land, Water the ecology and social organization of agricultural production. and Settlement project has considered animal exploitation, We ﬁrmly believe that when complete, the analysis of the full but this research is still ongoing and incorporates a spectrum archaeobotanical data set from Harappa will revolutionize our of archaeozoological and isotopic analyses, including analysis understanding of many aspects of social and economic life at of carbon, oxygen, and strontium in tooth enamel carbonate to this major city and its hinterland and potentially force further investigate animal diet and mobility. This research will com- reconsideration of the aspects of adaptation and resilience ex- plement the work on pastoralism mentioned by Miller and plored here. If we are to properly characterize agroecological contribute to an expansion of knowledge about Indus animal diversity, far more data needs to be collated for each area, in- economies, facilitating their integration into discussions of ag- cluding consideration of local-scale soil and weed ecology, and roecological and economic diversity, adaptation, and resilience. we believe that the Land, Water and Settlement project has Weber neatly encapsulates our core contention by reiter- made considerable strides in this regard in northwest India ating that environment and climate need to be understood at (e.g., Bates 2016; Neogi 2013; Petrie and Bates, forthcoming). the local level if we are to better comprehend the evolution of We also fully concur with Wright’s point that there is more the Indus Civilization. Bottom-up approaches that focus on to understanding the Indus than available foods and climate. local-scale data relevant for understanding climate and envi- While our focus is on the relationships between environment ronment are seeing increasing archaeological application, and and Indus society, we do emphasize the importance of recog- the need for such evidence to support modeling of human nizing human agency, human choice, and social processes in responses to climate change is a key point of our paper. More how Indus populations responded to diverse and changing detailed, multiproxy studies across more parts of the greater environments. We also agree with Wright’s point that diverse Indus zone are needed if we are to reach the stage of being able social organization may have supported or shaped human ad- to compare and contrast responses across different social and aptation to climatic shifts and discuss this in our reassessment ecological Indus settings. Our excavations in northwest India of settlement patterns in Cholistan and through our evidence have thus far been limited, but we have excellent stratigraphic for cultural diversity in settlement location and the use of ce- and chronological control, and our operations have been wide- ramic vessels in northwest India. spread in terms of geographical scale. Our work thus creates an Wright makes another important point about the diverse opportunity to start understanding variation across an indi- nature of water management across the Indus zone, and this vidual region. However, as Miller rightly notes, “there is still sits well with our view that water management is fundamental plenty of work to be done,” and we suggest that further research to understanding how Indus populations responded to a var- will increase emphasis on the nature of ecological diversity and iable and changing climate. The range of approaches to water human responses to that diversity within each of the large-scale management practised within northwest India are fully in domains, or ecocultural regions, within the greater Indus re- keeping with Wright’s points about the diversity of water man- gion. We look forward to being part of that ongoing process agement evident at Rojdi and Dholavira in Gujarat. We do re- through the TwoRains project. iterate, however, that there is a clear need to carry out further —Cameron A. Petrie, Ravindra N. Singh, Jennifer Bates, research into Indus water management practices as part of Yama Dixit, Charly A. I. French, David A. Hodell, efforts to understand how Indus societies responded to varied Penelope J. Jones, Carla Lancelotti, Frank Lynam, and changing environments across the diverse setting that they Sayantani Neogi, Arun K. Pandey, Danika Parikh, occupied. Vikas Pawar, David I. Redhouse, and Dheerendra P. Singh As Miller notes, many of the studies exploring ancient cli- mate change consider proxies relevant to the summer rainfall References Cited system but overlook the dynamics of the winter rainfall re- Adamson, George C. D., and David J. Nash. 2013. Documentary reconstruction gime. H. M.-L. Miller (2006, 2015) has previously emphasized of monsoon rainfall variability over western India, 1781–1860. Climate the important role of summer rains for winter cropping but Dynamics 42:749–769. This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 000 Current Anthropology Volume 58, Number 1, February 2017 Agrawal, D. P. 2009. Harappan technology and its legacy. New Delhi: Rupa ———. 1999. India: an archaeological history. Delhi: Oxford University Press. and in association with Inﬁnity Foundation. [RPW] Chakrabarti, Dilip K., and Sukhdev Saini. 2009. The problem of the Sarasvati Agrawal, Dharma P. 2007. The Indus Civilization: an interdisciplinary per- River and notes on the archaeological geography of Haryana and Indian spective. Delhi: Aryan. Panjab. Delhi: Aryan. Agrawal, Dharma P., and R. K. Sood. 1982. Ecological factors and the Harappan Chase, Brad. 2014. On the pastoral economies of Harappan Gujarat: faunal Civilisation. In Harappan civilization. Gregory L. Possehl, ed. Pp. 222–229. analyses at Shikarpur in context. Heritage 2:1–22. [HM-LM] Warminster: Aris and Philips. Chase, Brad, David Meiggs, P. Ajithprasad, and Philip A. Slater. 2014. Pastoral Aimers, James, and David Hodell. 2011. Societal collapse: drought and the land-use of the Indus Civilization in Gujarat: faunal analyses and biogenic Maya. Nature 479:44–45. isotopes at Bagasra. Journal of Archaeological Science 50:1–15. [HM-LM] Ajithprasad, P. 2011. Chalcolithic cultural patterns and the Early Harappan Clift, Peter D., Andrew Carter, Liviu Giosan, Julie Durcan, Geoff T. A. Duller, interaction in north Gujarat. In Cultural relations between the Indus and Mark G. Macklin, Anwar Alizai, et al. 2012. U-Pb zircon dating evidence the Iranian Plateau during the third millennium BCE. Toshiki Osada and for a Pleistocene Sarasvati River and capture of the Yamuna River. Geology Michael Witzel, eds. Pp. 11–40. Cambridge: Harvard Oriental Series Opera 40(3):211–214. Minora, Vol. 7. Clift, Peter D., and R. Alan Plumb. 2008. The Asian monsoon: causes, history Allchin, F. Raymond. 1995. The end of Harappan urbanism and its legacy. In and effects. Cambridge: Cambridge University Press. The archaeology of early historic South Asia: the emergence of cities and Coningham, Robin A. E., and Ruth Young. 2015. The archaeology of South Asia: states. F. Raymond Allchin, ed. Pp. 26–40. Cambridge: Cambridge Uni- from the Indus to Ashoka, c. 6500 BCE–200 CE. Cambridge: Cambridge versity Press. University Press. Amundson, Ronald, and Elise Pendall. 1991. Pedology and Late Quaternary Courty, Marie A. 1985. Le milieu physique et utilisations du sol. In Prospec- environments surrounding Harappa: a review and synthesis. In Harappa tions archéologiques au Nord-Ouest de l’Inde: rapport préliminaire 1983–84. excavations 1986–1990: a multidisciplinary approach to third millennium H.-P. Francfort, ed. Pp. 11–31. Paris: Éditions Recherche sur les Civili- urbanism. Richard H. Meadow, ed. Pp. 13–27. Monographs in World Ar- zations, Mémoire No. 62, Traveaux de la Mission Archaeologique Française chaeology, No. 3. Madison, WI: Prehistory Press. en Inde No. 1. Balbo, Andrea L., Xavier Rubio-Campillo, Bernardo Rondelli, Miguel Ramírez, ———. 1995. Late Quaternary environmental changes and natural constraints Carla Lancelotti, Alexis Torrano, Mattieu Salpeteur, et al. 2014. Agent-based to ancient land use (northwest India). In Ancient peoples and landscapes. simulation of Holocene monsoon precipitation patterns and hunter-gatherer E. Johnson, ed. Pp. 105–126. Lubbock: Museum of Texas Technical Uni- population dynamics in semi-arid environments. Journal of Archaeological versity. Method and Theory 21(2):426–446. Courty, Marie A., Paul Goldberg, and Richard Macphail. 1989. Soils and Barnes, Jessica, Martin Dove, Myanna Lahsen, Andrew Mathews, Pamela micromorphology in archaeology. Cambridge: Cambridge University Press. McElwee, Roderick McIntosh, Frances Moore, et al. 2013. Contribution of Danino, Michel 2010. The lost river: On the trail of the Saraswati. Delhi: anthropology to the study of climate change. Nature Climate Change 3:541– Penguin. 544. deMenocal, Peter B. 2001. Cultural responses to climate change during the Bates, Jennifer. 2016. Social organisation and change in Bronze Age South Holocene. Science 292:667–673. Asia: a multi-proxy approach to urbanisation, deurbanisation and village Dewar, Robert. 1991. Incorporating variation in occupation span into life through phytolith and macrobotanical analysis. PhD thesis, University settlement-pattern analysis. American Antiquity 56(4):604–620. of Cambridge, Cambridge. Diamond, Jared. 2005. Collapse: how societies choose to fail or survive. London: Belcher, Wayne R., and William R. Belcher. 2000. Geologic constraints on the Penguin. Harappa archaeological site, Punjab Province, Pakistan. Geoarchaeology 15 Dixit, Y., David A. Hodell, Rajiv Sinha, and Cameron A. Petrie. 2014. Abrupt (7):670–713. weakening of the Indian summer monsoon at 8.2 kyr B.P. Earth and Plan- Belcher, William R. 1998. Fish exploitation of the Baluchistan and Indus Valley etary Science Letters 391:16–23, doi:10.1016/j.epsl.2014.01.026. traditions: an ethnoarchaeological approach to the study of ﬁsh remains. PhD Dixit, Yama, David A. Hodell, and Cameron A. Petrie. 2014. Abrupt climate thesis, University of Wisconsin-Madison. [HM-LM] change and the decline of Indus urbanism. Geology 42:339–342. ———. 2003. Fish exploitation of the Indus Valley tradition. In Indus eth- Dixit, Yama, David A. Hodell, Rajiv Sinha, and Cameron A. Petrie. 2015. nobiology: new perspectives from the ﬁeld. Steven A. Weber and William R. Oxygen isotope analysis of multiple, single ostracod valves as a proxy for Belcher, eds. Pp. 95–174. Lanham, MD: Lexington. [HM-LM] combined variability in seasonal temperature and lake water oxygen iso- Berkelhammer, Max, A. Sinha, Lowell Stott, H. Cheng, Francesco S. R. topes. Journal of Paleolimnology 53:35–45, doi:10.1007/s10933-014-9805-3. Pausata, and K. Yoshimura. 2012. An abrupt shift in the Indian monsoon Enzel, Y., L. L. Ely, S. Mishra, R. Ramesh, R. Amit, B. Lazar, S. N. Rajguru, 4000 years ago. In Climates, landscapes, and civilizations. Liviu Giosan, V. R. Baker, and A. Sandler. 1999. High-resolution Holocene environmen- Dorian Q Fuller, Kathleen Nicoll, Rowan K. Flad, and Peter D. Clift, eds. tal changes in the Thar Desert, Northwestern India. Science 284:125–128. Pp. 75–87. Geophysical Monograph Series 198. Washington, DC: AGU. Fairservis, Walter A. 1967. The origin, character, and decline of an early Berkes, Fikret, Johan Colding, and Carl Folke. 2003. Navigating social-ecological civilization. American Museum Novitates 2302:1–48. systems: building resilience for complexity and change. Cambridge: Cam- ———. 1971. The roots of ancient India. Chicago: University of Chicago bridge University Press. Press. Bhadra, Bidyut K., A. K. Gupta, and J. R. Sharma. 2009. Saraswati Nadi in Flam, Louis. 1981. The palaeogeography and prehistoric settlement patterns in Haryana and its linkage with the Vedic Saraswati River—integrated study Sind, Pakistan (ca. 4000–2000 BC). Ann Arbor, MI: University Microﬁlms based on satellite images and ground based information. Journal of the International. Geological Society of India 73(2):273–288. ———. 1993. Fluvial geomorphology of the lower Indus Basin (Sindh, Bhan, Kuldeep K. 2011. Pastoralism in Late Harappan Gujarat: an ethno- Pakistan) and the Indus Civilisation. In Himalaya to the sea: geology, geo- archaeological approach. In Linguistics, archaeology and the human past morphology and the Quaternary. J. F. Shroder, ed. Pp. 265–287. London: occasional paper 10. Indus Project. Osada Toshiki and Akinori Uesugi, eds. Routledge. Pp. 1–25. Kyoto: Research Institute for Humanity and Nature. [HM-LM] ———. 1999. Ecology and population mobility in the prehistory settlement Bisht, R. S. 1994. Secrets of the water fort. Down to Earth 2:25–31. [RPW] of the lower Indus Valley, Sindh, Pakistan. In The Indus River: biodiversity, ———. 2009. Paradigms of the Harappan engineering at Dholavira. In A resources, humankind. A. Meadows and P. S. Meadows, eds. Pp. 313–323. festschrift in honour of Professor D. K. Bhattacharya. V. K. Srivastava and Oxford: Oxford University Press. M. K. Singh, eds. Delhi: Palaka Prakashan. [RPW] ———. 2013. The Sindh Archaeological Project: explorations in the lower Bisht, Ravindra S. 2005. The water structures and engineering of the Harap- Indus Basin and western Sindh. In Connections and complexity: new ap- pans at Dholavira (India). In South Asian archaeology 2001. Catherine proaches to the archaeology of South Asia. Shinu A. Abraham, Praveena Jarrige and Vincent Lefévre, eds. Pp. I.11–25. Paris: Éditions Recherche sur Gullapalli, Teresa P. Raczek, and Uzma Z. Rizvi, eds. Pp. 91–106. Walnut les Civilisations. Creek, CA: Left Coast. ———. n.d. Excavations at Dholavira (1989–90 to 2004–2005). New Delhi: Fleitmann, Dominik, Stephen J. Burns, Manfred Mudelsee, Ulrich Neff, Jan Archaeological Survey of India. Kramers, Augusto Mangini, and Albert Matter. 2003. Holocene forcing of Chakrabarti, Dilip K. 1988. Theoretical issues in Indian archaeology. Delhi: the Indian monsoon recorded in a stalagmite from Southern Oman. Science Munshiram Manoharlal. 300:1737–1739. This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Petrie et al. Investigating Land, Water and Settlement in Indus Northwest India 000 Francfort, Henri-Paul, ed. 1985. Prospections archéologiques au Nord-Ouest de and complexity: new approaches to the archaeology of South Asia. Shinu A. l’Inde: rapport préliminaire 1983–84. Paris: Éditions Recherche sur les Abraham, Praveena Gullapalli, Teresa P. Raczek, and Uzma Z. Rizvi, eds. Civilizations, Mémoire No. 62, Traveaux de la Mission Archaeologique Pp. 177–193. Walnut Creek, CA: Left Coast. Française en Inde No. 1. Kenoyer, J. Mark. 1991. Urban process in the Indus tradition: a prelimi- ———. 1992. Evidence for Harappan irrigation, Eastern Anthropologist 45(1/ nary model from Harappa. In Harappa excavations 1986–1990: a multi- 2):87–103. disciplinary approach to third millennium urbanism. Richard H. Meadow, French, Charles A. I., Federica Sulas, and Cameron A. Petrie, 2014. Expanding ed. Pp. 29–60. Monographs in World Archaeology No. 3. Madison, WI: the research parameters of geoarchaeology: case studies from Aksum in Prehistory Press. Ethiopia and Haryana in India. Archaeological and Anthropological Sci- ———. 1994. The Harappan state: was it or wasn’t it? In From Sumer to ences, doi:10.1007/s12520-014-0186-3. Meluhha: contributions to the archaeology of South and West Asia in memory Fuller, D., and S. Weber. 2006. Formation processes and paleoethnobotanical of George F. Dales Jr. Pp. 71–80. Madison, WI: Prehistory Press. interpretation in South Asia. Journal of Interdisciplinary Studies in History ———. 1997. Early city-states in South Asia: comparing the Harappan phase and Archaeology 2(1):91–114. [SAW] and Early Historic period. In The archaeology of city-states: cross-cultural Fuller, Dorian Q. 2006. Agricultural origins and frontiers in South Asia: a approaches. Deborah L. Nichols and Thomas H. Charlton, eds. Pp. 51–70. working synthesis. Journal of World Prehistory 20:1–86. Washington, DC: Smithsonian. ———. 2011. Finding plant domestication in the Indian subcontinent. Cur- ———. 1998. Ancient cities of the Indus Valley civilization. Karachi: Oxford rent Anthropology 52(suppl. 4):S347–S362. University Press. Fuller, Dorian Q., and Marco Madella. 2001. Issues in Harappan archaeo- ———. 2005. Culture change during the Late Harappan period at Harappa: botany: retrospect and prospect. In Indian archaeology in retrospect II: pro- new insights on Vedic Aryan issues. In The Aryan invasion: evidence, politics, tohistory. S. Settar and Ravi Korisettar, eds. Pp. 317–390. Delhi: Manohar. history. L. L. Patton and E. F. Bryant, eds. Pp. 21–49. London: Routledge Fuller, Dorian Q, and Chris J. Stevens. 2012. Agriculture and the development Curzon. of complex societies: an archaeobotanical agenda. In From foragers to ———. 2008. Indus urbanism: new perspectives in its origin and character. farmers: Gordon C. Hillman festschrift. Andrew S. Fairbairn and Ehud In The ancient city: new perspectives in the Old and New World. Joyce Weiss, eds. Pp. 37–57. Oxford: Oxbow. [RPW] Marcus and Jeremy A. Sablof, eds. Pp. 85–109. Santa Fe, NM: SAR. García-Granero, Juan J., Carla Lancelotti, and Marco Madella. 2015. A tale of Kottek, Markus, Jürgen Grieser, Christoph Beck, Bruno Rudolf, and Franz multi-proxies: integrating macro- and microbotanical remains to understand Rubel. 2006. World map of the Köppen-Geiger climate classiﬁcation up- subsistence strategies. Vegetation History and Archaeobotany 24:121–133. dated. Meteorologische Zeitschrift 15:259–263. García-Granero, Juan J., Carla Lancelotti, Marco Madella, and P. Ajithprasad. Kumar, Manmohan. 2009. Harappan settlements in the Ghaggar-Yamuna 2016. Millets and herders: the origins of plant cultivation in semi-arid divide. Linguistics, Archaeology and the Human Past 7:1–75. North Gujarat (India). Current Anthropology 57(2):149–173. Lal, Braj B. 1997. The earliest civilisation of South Asia. Delhi: Aryan. Gentelle, Pierre. 1985. Paysages, environment et irrigation. In Prospections ———. 2002. The Sarasvati ﬂows on: the continuity of Indian culture. Delhi: archéologiques au Nord-Ouest de l’Inde: rapport préliminaire 1983–84. H.-P. Aryan. Francfort, ed. Pp. 33–42. Paris: Éditions Recherche sur les Civilizations, Lambrick, Hugh T. 1964. Sind: a general introduction. History of Sind Series 1, Mémoire No. 62, Traveaux de la Mission Archaeologique Française en Inde Sindhi Abadi Board, Hyderabad. No. 1. ———. 1967. The Indus ﬂood-plain and the “Indus” civilization. Geo- Ghose, Bhimal, Amal Kar, and Zahid Husain. 1979. The lost courses of the graphical Journal 133:483–494. Saraswati River in the Great Indian Desert. Geographical Journal 145(3):446– Lancelotti, Carla, and Marco Madella. 2010. Preliminary anthracological 451. analysis from Harappan Kanmer: human-environment interactions as seen Giosan, Liviu, Peter D. Clift, Mark G. Macklin, Dorian Q. Fuller, Stefan through fuel resources exploitation and use. In Linguistics, archaeology and Constantinescue, Julie A. Durcan, Thomas Stevens, et al. 2012. Fluvial land- the human past occasional Paper 10. T. Osada and A. Usuegi, eds. Pp. 129– scapes of the Harappan civilization. Proceedings of the National Academy 142. Kyoto: Research Institute for Humanity and Nature. of Sciences 109:1–7. Law, Randall W. 2011. Inter-regional interaction and urbanism in the Ancient Gunderson, Lance H., and Crawford S. Hollig. 2002. Panarchy: understanding Indus Valley: a geological provenience study of Harappa’s rock and mineral transformations in human and natural systems. Washington: Island. assemblage. In Linguistics, archaeology and the human past occasional Haug, Gerald H., Detlef Günther, Larry C. Peterson, Daniel M. Sigman, paper 11. Pp. 1–800. Kyoto: Research Institute for Humanity and Nature. Konrad A. Hughen, and Beat Aeschlimann. 2003. Climate and the collapse Lawler, Andrew. 2011. In Indus times, the river didn’t run through it. Science of Maya civilization. Science 299(5613):1731–1735. 332:23. Jansen, Michael. 1993. Mohenjo-daro: city of wells and drains, water splendour Leipe, Christian, Dieter Demske, Pavel E. Tarasov, Bernd Wünnemannb, 4500 years ago. Bonn: Berisch Gladbuch Frontius-Gesellschaft. Frank Riedel, and HIMPAC Project Members. 2014. Potential of pollen and ———. 1994. Mohenjo-daro, type site of the earliest urbanization process in non-pollen palynomorph records from Tso Moriri (Trans-Himalaya, NW South Asia: ten years of research at Mohenjo-daro, Pakistan and an attempt India) for reconstructing Holocene limnology and humane environmental at a synopsis. In South Asian archaeology 1993. Asko Parpola and Petteri interactions. Quaternary International, doi:10.1016/j.quaint.2014.02.026. Koskikallio, eds. Pp. 263–280. Helsinki: Sumalainen Tiedeakatemia. Leslie, Paul, and J. Terrence McCabe. 2013. Response diversity in social- Jarrige, Jean-François, and Marielie Santoni. 1979. Fouilles de Pirak Volume I: ecological systems. Current Anthropology 54(2):114–143. texte. Fouilles de Pakistan No. 2. Paris: Diffusion de Boccard. Madella, Marco, P. Ajithprasad, Carla Lancelotti, Bernardo Rondelli, Andrea Jorgensen, D. W., M. D. Harvey, S. A. Schumm, and Louis Flam. 1993. Mor- Balbo, Charles A. I. French, David Rodríguez, et al. 2010. Social and en- phology and dynamics of the Indus River: implications for the Mohenjo vironmental transitions in arid zones: the North Gujarat Archaeological Daro site. In Himalaya to the sea: geology, geomorphology and the Quater- Project—NoGAP. Antiquity 84:325. nary. J. F. Shroder, ed. Pp. 288–326. London: Routledge. Madella, Marco, and Dorian Q. Fuller. 2006. Palaeoecology and the Harappan Joshi, Jagat P. 1984. Harappa culture: emergence of a new picture. Puratattva civilisation of South Asia: a reconsideration. Quaternary Science Review 13–14:51–54. 25:1283–1301. Joshi, Jagat P., Madhu Bala, and Jassu Ram. 1984. The Indus Civilisation: a Maemoku, Hideaki, Yorinao Shitaoka, Tsuneto Natomo, and Hiroshi Yagi, reconsideration on the basis of distribution maps. In Frontiers of the Indus 2012. Geomorphological constraints on the Ghaggar River regime during Civilisation: Sir Mortimer Wheeler commemoration volume. Braj B. Lal and the Mature Harappan period. In Climates, landscapes, and civilizations. Swaraj P. Gupta, eds. Pp. 511–530. Delhi: Books & Books. Liviu Giosan, Dorian Q Fuller, Kathleen Nicoll, Rowan K. Flad, and Peter D. Kajale, Mukund, D. 1991. Current status of Indian palaeoethnobotany: intro- Clift, eds. Pp. 97–106. American Geophysical Union Geophysical Mono- duced and indigenous food plants with a discussion of the historical and graph 198. Washington, DC: American Geophysical Union, doi:10.1029 evolutionary development of Indian agriculture and agricultural systems in /2012GM001218. general. In New light on early farming—recent developments in palaeo- Mallah, Qasid Husain. 2011. Spatial analysis of the contemporary settlement ethnobotany. Jane M. Renfrew, ed. Pp. 155–189. Edinburgh: Edinburgh pattern in the Thar Desert: an ethnoarchaeological investigation. Pakistan University Press. Heritage 3:81–88. [HM-LM] Kashyap, Arunima, and Steven A. Weber. 2013. Starch grain analysis and ex- Marshall, John. 1931. Mohenjo-daro and the Indus Civilisation. London: Arthur periments provide insights into Harappan cooking practices. In Connections Probsthain. This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 000 Current Anthropology Volume 58, Number 1, February 2017 McAnany, Patricia A., and Norman Yoffee. 2009. Questioning collapse: human Petrie, Cameron A. 2013. South Asia. In Oxford handbook of cities in history. resilience, ecological vulnerability, and the aftermath of Empire. Cambridge: Peter Clark, ed. Pp. 139–173. Oxford: Oxford University Press. Cambridge University Press. Petrie, Cameron A., and Jennifer Bates. Forthcoming. Double-cropping, multi- Meadow, R. H., and J. Mark Kenoyer. 1997. Excavations at Harappa 1994– cropping and adaptation to variable environments in the Indus Civilization. 1995: new perspectives on the Indus script, craft activities, and city orga- Journal of World Prehistory. nization. In South Asian archaeology 1995. F. Raymond Allchin and Bridget Petrie, Cameron A., Jennifer Bates, Tom Higham, and Ravindra N. Singh. 2016. Allchin, eds. Pp. 139–172. New Delhi: Oxford & IBH. Feeding ancient cities: dating the adoption of rice, millet and tropical pulses Meadow, Richard, and Ajita K. Patel. 2003. Prehistoric pastoralism in North- in the Indus Civilisation. Antiquity 90(354):1489–1504. western South Asia from the Neolithic through the Harappan Period. In Petrie, Cameron A., and Frank Lynam. Forthcoming. Revisiting the Dewar Indus ethnobiology: new perspectives from the ﬁeld. Steven A. Weber and algorithm to explore settlement contemporaneity and (in)stability. Journal William R. Belcher, eds. Pp. 95–174. Lanham: Lexington. [HM-LM] of Field Archaeology. Meadow, Richard H. 1991. Faunal remains and urbanism at Harappa. In Petrie, Cameron A., Ravindra N. Singh, and Ashok K. Singh. 2009. Investi- Harappa excavations 1986–1990: a multidisciplinary approach to third mil- gating changing settlement dynamics on the plains: the 2009 survey and lennium urbanism. Richard H. Meadow, ed. Pp. 89–106. Monographs in excavations at Masudpur (Hissar District, Haryana). Puratattva 39:38–49. World Archaeology No. 3. Madison, WI: Prehistory Press. Petrie, Cameron A., and Kenneth D. Thomas. 2012. The topographic and ———. 1996. The origins and spread of agriculture and pastoralism in environmental context of the earliest village sites in western South Asia. northwest South Asia. In The origins and spread of agriculture and pasto- Antiquity 85:1055–1067. ralism in Eurasia. David R. Harris, ed. Pp. 390–412. London: UCL. Pokharia, Anil. 2012. Floral remains: on the record of plant remains at Mehdi, S. Muntazir, Naresh C. Pant, H. S. Saini, S. A. I. Mujtaba, and Prabhas Harappan settlement at Kanmer, subsistence economy with a change. In Pande. 2016. Identiﬁcation of palaeochannel conﬁguration in the Saraswati Excavation at Kanmer 2005–06 —2008–09. J. S. Kharakwal, Y. S. Rawat, and River basin in parts of Haryana and Rajasthan, India, through digital re- Toshiki Osada, eds. Pp. 795–818. Kyoto: Research Institute for Humanity mote sensing and GIS. Episodes 39(1):29–38. and Nature. Miller, Heather M.-L. 2006. Water supply, labor requirements, and land own- Pokharia, Anil K., Jeewan Singh Kharakwal, and Alka Srivastava. 2014. ership in Indus ﬂoodplain agricultural systems. In Agricultural strategies. Archaeobotanical evidence of millets in the Indian subcontinent with some Joyce Marcus and Charles Stannish, eds. Pp. 92–128. Los Angeles: Cotsen observations on their role in the Indus Civilization. Journal of Archaeo- Institute. logical Science 42:442–455, doi:10.1016/j.jas.2013.11.029. ———. 2013. Weighty matters: evidence for unity and regional diversity Possehl, Gregory L. 1982. The Harappan civilisation: a contemporary per- from the Indus Civilization weights. In Connections and complexity: new spective. In Harappan civilisation: a contemporary perspective. Gregory L. approaches to the archaeology of South Asia. Shinu A. Marcus, Praveena Possehl, ed. Pp. 15–28. Delhi: Oxford & IBH and the American Institute of Gullapalli, Teresa P. Raczek, and Uzma Z. Rizvi, eds. Pp. 161–176. Walnut Indian Studies. Creek, CA: Left Coast. ———. 1992. The Harappan cultural mosaic: ecology revisited. In South ———. 2015. Surplus in the Indus Civilization: agricultural choices, social Asian archaeology 1989. Catherine Jarrige, ed. Pp. 237–244. Monographs in relations, political effects. In Surplus: the politics of production and the World Archaeology No. 14. Madison, WI: Prehistory Press. strategies of everyday life. Christopher T. Morehart and Kristin De Lucia, ———. 1997a. Climate and the eclipse of the ancient cities of the Indus. In eds. Pp. 97–120. Colorado: University Press of Colorado. Third millennium, BC climate change and Old World collapse. H. Nüzhet Miller, Naomi. 2011. Managing predictable unpredictability: agricultural sus- Dalfes, George Kukla, and Harvey Weiss, eds. Pp. 193–243. Berlin: Springer. tainability at Gordion, Turkey. In Sustainable lifeways: cultural persistence in ———. 1997b. The transformation of the Indus Civilisation. Journal of World an ever-changing environment. Naomi Miller, Katherine M. Moore, and Prehistory 11(4):425–472. Kathleen Ryan, eds. Pp. 310–324. Philadelphia: University of Pennsylvania ———. 1998. Sociocultural complexity without the state: the Indus Civili- Museum of Archaeology and Anthropology. zation. In The archaic state. Gary M. Feinman and Joyce Marcus, eds. Miller, Naomi, Katherine M. Moore, and Kathleen Ryan, eds. 2011. Sustainable Pp. 261–291. Santa Fe, NM: SAR. lifeways: cultural persistence in an ever-changing environment. Philadelphia: ———. 1999. Indus Age: the beginnings. Philadelphia: University of Penn- University of Pennsylvania Museum of Archaeology and Anthropology. sylvania Museum. Mitchell, Peter. 2008. Practising archaeology at a time of climatic catastrophe. ———. 2002. The Indus Civilization: a contemporary perspective. Walnut Antiquity 82:1093–1103. Creek, CA: Altamira. Mughal, M. Raﬁque. 1997. Ancient Cholistan: archaeology and architecture. ———. 2003. The Indus Civilization: an introduction to environment, sub- Lahore: Ferozsons. sistence, and cultural history. In Indus ethnobiology. Steven A. Weber and Nakamura, Atsunori, Yusuke Yokoyama, Hideaki Maemoku, Hiroshi Yagi, William R. Belcher, eds. Pp. 1–20. Lanham, MD: Lexington. Makoto Okamura, Hiromi Matsuoka, Nao Miyake, et al. 2016. Weak mon- Prasad, Sushma, A. Anoop, N. Riedel, S. Sarkar, P. Menzel, N. Basavaiah, soon event at 4.2 ka recorded in sediment from Lake Rara, Himalayas. R. Krishnan, et al. 2014. Prolonged monsoon droughts and links to Indo- Quaternary International 397:349–359, doi:10.1016/j.quaint.2015.05.053. Paciﬁc warm pool: a Holocene record from Lonar Lake, central India. Earth Nath, Amarendra. 1998. Rakhigarhi: a Harappan metropolis in the Saraswati- and Planetary Science Letters 391:171–182. Drishadwati divide. Puratattva 28:39–45. Prasad, Sushma, and Yehouda Enzel. 2006. Holocene paleoclimates of India. Neogi, Sayantani. 2013. Geoarchaeological investigations of Indus settlements Quaternary Research 66(3):442–453. in the plains of northwestern India. PhD thesis, University of Cambridge, Puri, V. M. K., and B. C. Verma. 1998. Glaciological and geological source of Cambridge. Vedic Saraswati in the Himalayas. Itihas Darpan 4(2):7–36. Oldham, Charles F. 1874. Notes on the Lost River of the Indian Desert. Calcutta Raikes, Robert L. 1965. The Mohenjo-daro ﬂoods. Antiquity 39:196–203. Review 59:1–27. ———. 1968. Kalibangan: death from natural causes. Antiquity 42:286–291. ———. 1893. The Sarasvati and the lost river of the Indian Desert. Journal of Raikes, Robert L., and George F. Dales. 1977. The Mohenjo-daro ﬂoods re- the Royal Asiatic Society 25:49–76. considered. Journal of Palaeontological Society of India 20:251–260. Oldham, R. D. 1886. On probable changes in the geography of the Punjab and Raikes, Robert L., and R. H. Dyson Jr. 1961. The prehistoric climate of Ba- its rivers: an historico-geographical study. Journal of Asiatic Society of luchistan and the Indus Valley. American Anthropologist 63(2/1):265–281. Bengal 55:322–343. Rajaguru, S. N., and Sushama G. Deo. 2008. Preliminary observations on Parikh, Danika, and Cameron A. Petrie. Forthcoming. Urban-rural dynamics Holocene ﬂuvial sediments around Kanmer, Gujarat, India. In Linguistics, and Indus ceramic production in northwest India: a preliminary analysis of archaeology and the human past occasional paper 5. Toshiki Osada and the pottery from Masudpur I and Masudpur VII. In South Asian Archae- Akinori Uesugi, eds. Pp. 1–3. Kyoto: Research Institute for Humanity and ology 2012. Lefèvre, Vincent, Benjamin Mutin, and Aurore Didier. eds. Nature. Indicopleustoi. Turnhout: Brepols. Ramaswamy, C. 1968. Monsoon over the Indus Valley during the Harappan Parpola, Asko. 1994. Deciphering the Indus script. Cambridge: Cambridge Uni- period. Nature 217:628–629. versity Press. Reddy, S. N. 1997. If the threshing ﬂoor could speak: integration of agriculture Pawar, Vikas. 2012. Archaeological settlement pattern of Hanumangarh Dis- and pastoralism during the Late Harappan in Gujarat, India. Journal of trict (Rajasthan). PhD dissertation, MD University Rohtak, Haryana. Anthropological Archaeology 16:162–187. This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). Petrie et al. Investigating Land, Water and Settlement in Indus Northwest India 000 ———. 2003. Discerning palates of the past: an ethnoarchaeological study of Singh, Ravindra N., Cameron A. Petrie, Charles A. I. French, Andrew S. crop cultivation and plant usage in India. Ann Arbor, MI: International Goudie, Sanjeev Gupta, R. Tewari, A. K. Singh, et al. 2008. Settlements in Monographs in Prehistory. context: reconnaissance in western Uttar Pradesh and Haryana, April and Robbins Schug, Gwen K., and Elaine Blevins. 2016. The center cannot hold: a May 2008. Man and Environment 33(2):71–87. bioarchaeological perspective on environmental crisis in the second mil- Singh, Ravindra N., Cameron A. Petrie, Charles A. I. French, Sayantani Neogi, lennium BCE, South Asia. In A companion to South Asia in the past. Gwen Arun K. Pandey, Danika Parikh, and Vikas Pawar. 2010a. Geoarchaeological K. Robbins Schug and Subhash R. Walimbe, eds. Pp. 255–273. Chichester: survey and excavations at Burj, Fatehabad, Haryana. Puratattva 40:94–101. Wiley-Blackwell. Singh, Ravindra N., Cameron A. Petrie, Vikas Pawar, Arun K. Pandey, Robbins Schug, Gwen K., Elaine Blevins, Brett Cox, Kelsey Gray, and V. Sayantani Neogi, Manisha Singh, Ashok K. Singh, D. Parikh, and C. Lan- Mushrif-Tripathy. 2013a. Infection, disease, and biosocial processes at the celotti. 2010b. Changing patterns of settlement in the rise and fall of Ha- end of the Indus Civilization. PLoS ONE 8(12):1–20. rappan urbanism: preliminary report on the Rakhigarhi Hinterland Survey. Robbins Schug, Gwen K., Kelsey Gray, V. Mushrif-Tripathy, and A. R. 2009. Man and Environment 35(1):37–53. Sankhyan. 2013b. A peaceful realm? trauma and social differentiation at Singh, Ravindra N., Cameron A. Petrie, Vikas Pawar, Arun K. Pandey, and Harappa. International Journal of Paleopathology 2:136–147. Danika Parikh. 2011. New insights into settlement along the Ghaggar and Saini, H. S., S. K. Tandon, S. A. I. Mujtaba, N. C. Pant, and R. K. Khorana, its hinterland: a preliminary report on the Ghaggar Hinterland Survey 2010. 2009. Reconstruction of buried channel-ﬂoodplain systems of the north- Man and Environment 36(2):89–106. western Haryana Plains and their relation to the “Vedic” Saraswati. Current Sinha, Rajiv, G. S. Yadav, Sanjeev Gupta, Ajit Singh, and S. K. Lahiri. 2013. Science 97(11):1634–1643. Geo-electric resistivity evidence for subsurface palaeochannel systems ad- Sankalia, Hasmukh D. 1962. The prehistory and protohistory of India and jacent to Harappan sites in northwest India. Quaternary International 308– Pakistan. Poona: Deccan College. 309:66–75. Saraswat, K. S. 2002. Balu (297400N; 767220E), District Kaithal. Indian Ar- Sinopoli, Carla. 2015. Ancient South Asian cities in their regions. In The chaeology—A Review 1996–1997:198–203. Cambridge world history III: early cities in comparative perspective, 4000 Saraswat, K. S., and Anil K. Pokharia. 2002. Harappan plant economy at BCE–1200 CE. Norman Yoffee, ed. Pp. 319–342. Cambridge: Cambridge ancient Balu, Haryana. Pragdhara 12:153–171. University Press. ———. 2003. Palaeoethnobotanical investigations at Early Harappan Kunal. Staubwasser, Michael, Frank Sirocko, Pieter M. Grootes, and M. Segl. 2003. Pragdhara 13:105–139. Climate change at the 4.2 ka BP termination of the Indus Valley Civili- Saraswat, K. S., C. Srivastava, and Anil K. Pokharia. 2000. Palaeobotanical and zation and Holocene South Asian monsoon variability. Geophysical Re- pollen analytical investigations. Indian Archaeology—A Review 1994– search Letters 30(8):1425. 1995:96. Staubwasser, Michael, and Harvey Weiss. 2006. Holocene climate and cultural Sarkar, Saswati, Sushma Prasad, Heinz Wilkes, Nils Riedel, Martina Stebich, evolution in late prehistoric-early historic West Asia-introduction. Qua- Nathani Basavaiah, and Dirk Sachse. 2015. Monsoon source shifts dur- ternary Research 66:372–387. ing the drying mid-Holocene: biomarker isotope based evidence from the Stein, M. Aurel. 1931. An archaeological tour of Gedrosia. Memoires of the core “monsoon zone” (CMZ) of India. Quaternary Science Reviews 123: Archaeological Survey of India 43. New Delhi: Government of India Press. 144–157. ———. 1942. A survey of ancient sites along the “Lost” Sarasvari River. Sarma, V. V. 1976. A hydrometeorological study of the rainfall trend in Geographical Journal 99:173–182. the Ghaggar catchment area. Indian National Science Academy 42(2):102– Suraj Bhan. 1975. Excavation at Mitathal (1968) and other explorations in the 108. Sutlej-Yamuna divide. Kurukshetra: Kurukshetra University. Scarborough, Vernon L. 1993. Water management in the Southern Maya Thomas, Kenneth D. 2003. Minimizing risk? approaches to pre-Harappan Lowlands: an accretive model for the engineered landscape. In Economic human ecology on the northwest margin of the Greater Indus System. In aspects of water management in the prehispanic New World. Research in Indus ethnobiology: new perspectives from the ﬁeld. Steven A. Weber and Economic Anthropology. Vernon L. Scarborough and Barry L. Isaac, eds. William R. Belcher, eds. Pp. 95–174. Lanham, MD: Lexington. [HM-LM] Pp. 17–69. Greenwich, CT: JAI. [RPW] Tripathi, Jayant K., Barbara Bock, V. Rajamani, and A. Eisenhauer. 2004. Is ———. 2003. The ﬂow of power: ancient water systems and landscapes. Santa River Ghaggar, Saraswati? geochemical constraints. Current Science 87 Fe, NM: SAR. [RPW] (8):1141–1145. Scarborough, Vernon L., and Lisa Lucero 2010. The non-hierarchical devel- Uesugi, Akinori. 2011. Pottery from the settlement area. In Excavations at opment of complexity in the semitropics: water and cooperation. Water Farmana: District Rohtak, Haryana, India 2006–2008. Vasant Shinde, To- History 2:185–208. [RPW] shiki Osada, and Manmohan Kumar, eds. Pp. 168–328. Indus Project. Schuldenrein, Joseph, Rita P. Wright, and Mohammed Afzal Khan. 2007. Kyoto: Research Institute for Humanity and Nature. Harappan geoarchaeology reconsidered: Holocene landscapes and envi- Valdiya, Kharag S. 2002. Saraswati: the river that disappeared. Hyderabad: ronments of the Greater Indus Plain. In Settlement and society: essays Universities Press. dedicated to Robert McCormick Adam. Elizabeth C. Stone, ed. Pp. 83–116. Van de Noort, Robert. 2011. Conceptualising climate change archaeology. Los Angeles: Cotsen Institute. Antiquity 85:1039–1048. Shinde, V., T. Osada, M. M. Sharma, A. Uesugi, T. Uno, H. Maemoku, P. Vishnu-Mittre and R. Savithri. 1982. Food economy of the Harappans. In Shirwalkar, et al. 2008. Exploration in the Ghaggar Basin and excavations at Harappan civilisation. Gregory L. Possehl, ed. Pp. 205–221. Warminster: Girawad, Farmana (Rohtak Districts) and Mitathal (Bhiwani District), Aris and Philips. Haryana, India. In Linguistics, archaeology and the human past occasional Weber, Steven A. 1991. Plants and Harappan subsistence: an example of paper 3. Toshiki Osada and Akinori Uesugi, eds. Pp. 77–158. Kyoto: Re- stability and change from Rojdi. New Delhi: Oxford & IBH. search Institute for Humanity and Nature. ———. 1996. Distinguishing change in the subsistence and the material Shinde, Vasant, Shweta Deshpande, Toshiki Osada, and Takao Uno. 2006. Basic records. The inter-play of environment and culture. Asian Perspectives 35 issues in Harappan archaeology: some thoughts. Ancient Asia 1:63–72. (2):155–163. [RPW] Shitaoka, Yorinao, Hideaki Maemoku, and Tsuneto Nagatomo. 2012. Quartz ———. 1998. Out of Africa: the initial impact of millets in South Asia. OSL dating of sand dunes in Ghaggar Basin, northwestern India, Geo- Current Anthropology 39(2):267–274. [RPW] chronometria 39:221–226, doi:10.2478/s13386-012-0012-6. ———. 1999. Seeds of urbanism: paleoethnobotany and the Indus Civiliza- Singh, Ravindra N., and Cameron A. Petrie. 2009. Lost rivers and life on the tion. Antiquity 282:813–827. plains—approaches to understanding human/environment interaction be- ———. 2003. Archaeobotany at Harappa: indications of change. In Indus tween the collapse of Indus urbanism and the rise of the Early Historic cities ethnobiology. Steven A. Weber and William R. Belcher, eds. Pp. 175–198. (the Land, Water and Settlement project). In Sarasvati River—a perspective. Lanham, MD: Lexington. Pp. 102–111. Conference Proceedings. Kurukshetra: Kurukshetra University. ———. 2007. Millets and their role in early agriculture. Pragdhara 16:1–18. Singh, Ravindra N., Cameron A. Petrie, Charles A. I. French, Jennifer Bates, [RPW] C. Lancelotti, A. K. Pandey, D. Parikh, and D. I. Redhouse. 2012. Survey ———. 2010. Does size matter: the role and signiﬁcance of cereal grains in and excavations at Dabli-vas Chugta, Hanumangarh District, Rajasthan. the Indus Civilization. Archaeological and Anthropological Sciences 2(1):35– Puratattva 42:133–147. 43. [RPW] This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 000 Current Anthropology Volume 58, Number 1, February 2017 Weber, Steven A., Tim Barela, and Heather Lehman. 2010. Ecological con- Wright, Rita P., and Carrie Hritz. 2013. Satellite remote sensing imagery: new tinuity: an explanation for agricultural diversity in the Indus Civilisation evidence for site distributions and ecologies in the upper Indus. In South and beyond. Man and Environment 35(1):62–75. Asian archaeology 2007. Dennys Frenez and Maurizio Tosi, eds. Oxford: Weber, Steven A., and Arunima Kashyap. 2016. The vanishing millets of the Archaeopress, BAR International Series 2454. Indus Civilisation. Archaeological and Anthropological Sciences 8:9–15, Wright, Rita P., Mohammad Afzal Khan, and Joseph Schuldenrein. 2002. doi:10.1007/s12520-013-0143-6. Urbanism in the Indus Valley: environment and settlement on the Beas Weber, Steven A., Arunima Kashyap, and David Harriman. 2010. Does size River. In Indus Valley civilisation: dialogue Among civilisations. M. A. matter: the role and signiﬁcance of cereal grains in the Indus Civilisation. Halim and Abdul Ghafoor, eds. Pp. 102–113. Islamabad: Crystal. Archaeological and Anthropological Sciences 2:35–43. Wright, Rita P., Joseph Schuldenrein, Muhammad Afzal Khan, and Susan Weber, Steven A., Arunima Kashyap, and Laura Mounce. 2011. Archaeobotany Malin-Boyce. 2005a. The Beas River Landscape and Settlement Survey: at Farmana: new insights into Harappan plant use strategies. In Excavations preliminary results from the site of Vainiwal. In South Asian archaeology at Farmana, District Rohtak, Haryana, India, 2006–8. Vasant Shinde, 2003. Ute Franke-Vogt and Hans-Joachim Weisshaar, eds. Pp. 101–111. Toshiki Osada, and Manmohan Kumar, eds. Pp. 808–825. Kyoto: Research Aachen: Linden Soft. Institute for Humanity and Nature. Wright, Rita P., Joseph Schuldenrein, Muhammad Afzal Khan, and M. Raﬁque Wheeler, R. E. Mortimer. 1968. The Indus Civilisation. Third edition. Cam- Mughal. 2005b. The emergence of satellite communities along the Beas bridge: Cambridge University Press. drainage: preliminary results from Lahoma Lal Tibba and Chak Burbane Wilhelmy, H. 1969. Verschollene Städte im Indusdelta. Geographische Zeit- Syal. In South Asian archaeology 2001. Catherine Jarrige and Vincent Lefévre, schrift 56(4):256–294. eds. Pp. I.327–336. Paris: Éditions Recherche sur les Civilisations. Wright, Rita P. 2010. The Ancient Indus: urbanism, economy and society. Case Yash Pal, Baldev Sahai, R. K. Sood, and D. P. Agrawal. 1980. Remote sensing Studies in Early Societies. Cambridge: Cambridge University Press. of the “Lost” Sarasvati River. Proceedings of the Indian Academy of Sciences- Wright, Rita P., Reid A. Bryson, and Joseph Schuldenrein. 2008. Water Earth and Planetary Sciences 89(3):317–331. supply and history: Harappa and the Beas regional survey. Antiquity 82: Yoffee, Norman. 2005. Myths of the archaic state: evolution of the earliest 37–48. cities, states and civilisations. Cambridge: Cambridge University Press. This content downloaded from 118.208.219.165 on January 28, 2017 21:03:54 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).

(PDF) Adaptation to Variable Environments, Resilience to Climate Change: Investigating Land, Water and Settlement in Indus North