AUTHOR'S PROOF JrnlID 10816_ArtID 9236_Proof# 1 - 19/12/2014 J Archaeol Method Theory DOI 10.1007/s10816-014-9236-5 31 2 Are Social Networks Survival Networks? An Example 4 from the Late Pre-Hispanic US Southwest 5 Lewis Borck & Barbara J. Mills & 6 Matthew A. Peeples & Jeffery J. Clark 7 8 # Springer Science+Business Media New York 2014 9 10 Abstract Archaeologists have regarded social networks as both the links through 11 which people transmitted information and goods as well as a form of social storage 12 creating relationships that could be drawn upon in times of subsistence shortfalls or 13 other deleterious environmental conditions. In this article, formal social network 14 analytical (SNA) methods are applied to archaeological data from the late pre- 15 Hispanic North American Southwest to look at what kinds of social networks charac- 16 terized those regions that were the most enduring versus those that were depopulated 17 over a 250-year period (A.D. 1200–1450). In that time, large areas of the Southwest 18 were no longer used for residential purposes, some of which corresponds with well- 19 documented region-wide drought. Past research has demonstrated that some population 20 levels could have been maintained in these regions, yet regional scale depopulation 21 occurred. We look at the degree to which the network level property of embeddedness, 22 along with population size, can help to explain why some regions were depopulated 23 and others were not. SNA can help archaeologists examine why emigration occurred in 24 some areas following an environmental crisis while other areas continued to be 25 inhabited and even received migrants. Moreover, we modify SNA techniques to take 26 full advantage of the time depth and spatial and demographic variability of our 27 archaeological data set. The results of this study should be of interest to those who 28 seek to understand human responses to past, present, and future worldwide catastrophes 29 since it isUNCORRECTED now widely recognized that responses to major human PROOF disasters, such as 30 hurricanes, were “likely to be shaped by pre-existing or new social networks” (as 31 reported by Suter et al. (Research and Policy Review 28:1–10, 2009)). 32 Keywords Social network analysis . E–Iindex. Migration . Resilience . Salado . 33 Persistence 34 L. Borck (*) : B. J. Mills School of Anthropology, University of Arizona, Tucson, AZ 85721, USA Q2 e-mail: [email protected] L. Borck : M. A. Peeples : J. J. Clark Archaeology Southwest, Tucson, AZ 85701, USA AUTHOR'SJrnlID 10816_ArtID 9236_Proof# PROOF 1 - 19/12/2014 Borck et al. Overview 35 Independence? That’s middle-class blasphemy. We are all dependent on one 3638 another, every soul of us on earth. (Shaw 1988, Act V, par. 244) 39 37 An enduring archaeological problem is why some regions were depopulated while 41 inhabitants remained in other areas for centuries and even millennia. When depopula- 4042 tion is sudden and arises at large spatial scales, it has often been called “collapse.” We 43 place this term in scare quotes because it has been interpreted to mean cultural 44 discontinuity—such as the Hohokam or Chaco “collapse” in the US Southwest or the 45 “collapse” of the Maya in Central America—without considering continuities in 46 indigenous populations through processes such as migration and decentralized social 47 reorganizations (see also the many critiques of the term, e.g.,Hegmonet al. 1998; 48 McAnany and Yoffee 2010;Middleton2012; Wilcox 2009). When an entireOF region is 49 depopulated, it frequently entails large-scale migrations. The underlying reasons for 50 these demographic shifts from one area to another have been central to understanding 51 human history across the globe (e.g.,Anthony1990;Cameron2013RO). 52 Another way to approach the problem is to look at why and how people are able to 53 persist in some regions over long periods. What makes settlementsP in these regions 54 more successful than others? Persistence may also be associated with the concept of 55 resilience (Holling and Gunderson 2002;Holling1973;Redman2005) because 56 communities that endure have withstood threats to their ongoing sustainability. 57 However, in some situations, persistence may be maladaptive rather than adap- 58 tive. For example, generational residential mobility can enhance the resilience 59 of particular individuals and groups (Varien 1999). Staying in one place may not be 60 the best decision because of negativeECTED changes in local, social, or environmental 61 contexts. Nonetheless, migrations are expensive and disruptive to the existing 62 social order and often result in migrants having lower status than the groups 63 they join (Mills 2007). 64 In a recent review of archaeological approaches to “collapse” and resilience, Guy 65 Middleton (2012) suggestsORR that too much emphasis has been placed on environmental 66 factors at the expense of social factors (e.g.,Diamond2005). Archaeologists working in 67 the US Southwest—where a high-resolution paleoclimatic record indicates long-term 68 periods of droughtNC—have recognized that climatic downturns are not sufficient expla- 69 nations forU depopulation and migration (Kohler 2010;Minnis1985;VanWest1994; 70 Wright 2010). As Glowacki (2010, p. 200) pointed out “[c]omplete regional depopu- 71 lation is not a predestined outcome of drought and poor agricultural yield.” Van West 72 (1994; see also Ahlstrom et al. 1995 and Billman et al. 2000) demonstrated that, 73 Q3=Q4 despite the regional depopulation, yields could have sustained some of the population 74 in the Mesa Verde region during the drought of the last quarter of the thirteenth century, 75 albeit at lower levels. Other scholars have highlighted the increased incidence of 76 violence during periods of drought, especially during the early twelfth and late thir- 77 teenth centuries (e.g.,Kuckelmanet al. 2002; Kuckelman and Martin 2012). Kohler 78 (2010) suggested that the crosscutting social institutions that were created after the 79 eleventh century, which enabled the integration of segmentary groups with high 80 residential mobility, produced societies that could no longer easily fission when 81 mobility and social flexibility was required in later periods. These religious institutions 82 AUTHOR'S PROOF JrnlID 10816_ArtID 9236_Proof# 1 - 19/12/2014 Are Social Networks Survival Networks? Q1 would have facilitated intra-regional interaction while at the same time, they may have 83 laid the ground work for future inter-regional conflict and factionalism (Glowacki 84 2010;Kaldahlet al. 2004). 85 What the above Southwestern examples share is an emphasis on the formation, 86 transformation, and/or dissolution of different kinds of social networks. They also 87 largely focus on specific regions. Another way of incorporating social factors into 88 models of regional depopulation is to understand these areas in the context of inter- 89 regional networks (Mills et al. 2013a, b, c). As Hill and his colleagues (2010)pointed 90 out regarding the depopulation of the Mesa Verde area, this region can only be 91 understood by looking at population changes at much larger scales (see also Duff 92 1998). Several previous examples from the northern Southwest illustrate the value of 93 looking at migration across regional boundaries and tie these, either explicitly or 94 implicitly, to large-scale social networks. The intensity and extent of social networks 95 has been widely acknowledged in the anthropological literature as providingOF access to 96 information and resources to minimize risk and uncertainty (Colson 1979;Dyson- 97 Hudson and Smith 1978;HalsteadandO’Shea 1989; Wiessner 1977). This is partic- 98 ularly the case in the literature on foragers but it also pertains toRO agriculturalists in a 99 setting of high environmental variation, such as is found in many periods in the North 100 American Southwest. These authors argue that access to informationP and resources 101 from diverse environments can help populations deal with unpredictability. 102 Importantly, the work of Cordell and colleagues (2007) and Rautman (1993)suggest 103 that the distance and direction of such external connections are important in under- 104 standing how social networks were used to integrate diverse groups. This prior work 105 also shows the importance of inter-regional social networks in creating pathways for 106 future migration. 107 In this paper, we investigate severalECTED aspects of Southwest regions to better under- 108 stand the social factors that, during A.D. 1200–1450, led to regional depopulation on 109 the one hand and regional persistence on the other. We argue that the application of 110 social network analysis (SNA) to archaeological data provides insights into the ways in 111 which interaction at multiple scales (or lack thereof) can explain regional depopulation 112 or persistence. We use SNAORR to look at both internal and external ties to understand the 113 social component of regional depopulations as well as to examine how long-distance 114 ties created viable alternatives to staying in place. We argue that regions with relatively 115 low population sizesNC and relatively closed social networks were especially vulnerable to 116 depopulation.U Those regions with relatively open social networks, or closed networks 117 and large population sizes, were more successful through time. 118 Social Networks and Regional Depopulation 119 SNA refers to a broad range of methods and models, originally derived from the 120 mathematical field of graph theory, for formally exploring the structure of relations 121 among some set of social actors (Wasserman and Faust 1994). The most common 122 means for analyzing and visualizing social networks involves producing network 123 graphs where actors (which can be individuals or larger groups/organizations) are 124 depicted as nodes and the social relationships among these actors are depicted as lines 125 (also known as ties or edges). SNA is particularly useful for understanding the structure 126 AUTHOR'SJrnlID 10816_ArtID 9236_Proof# PROOF 1 - 19/12/2014 Borck et al.