New Directions in Bioarchaeology: Recent Contributions to the Study of Human Social Identities Kelly J. Knudson �� Christopher M. Stojanowski Published online: 11 April 2008 �� Springer Science+Business Media, LLC 2008 Abstract As a discipline that bridges the biological and social sciences, bioar- chaeology has much to contribute to a contextualized and theoretically sophisticated understanding of social identities. Here, we discuss the growing methodological sophistication of bioarchaeology and highlight new developments in osteological age and sex estimation, paleodemography, biodistance analysis, biogeochemistry, and taphonomy, particularly anthropologie de terrain. We then discuss how these methodological developments, when united with social theory, can elucidate social identities. More specifically, we highlight past and future bioarchaeological work on disability and impairment, gender identity, identities of age and the life course, social identity and body modification, embodiment, and ethnic and community identities. Keywords Archaeology Physical anthropology Osteology Human remains Introduction As an academic specialization, bioarchaeology coherently unites the disparate subfields of anthropology by combining aspects of both the biological and social sciences (Buikstra and Beck 2006 Katzenberg and Saunders 2000 Larsen 1997, 2002 Wright and Yoder 2003). As the field moves beyond its roots as descriptive osteology (Armelagos and Gerven 2003 Buikstra and Beck 2006 Stojanowski and Buikstra 2004, 2005) and continues to address important critiques of its basic K. J. Knudson (&) C. M. Stojanowski Center for Bioarchaeological Research, School of Human Evolution and Social Change, Arizona State University, PO Box 872402, Tempe, AZ 85287-2402, USA e-mail: kelly.knudson@asu.edu C. M. Stojanowski e-mail: christopher.stojanowski@asu.edu 123 J Archaeol Res (2008) 16:397���432 DOI 10.1007/s10814-008-9024-4

operational assumptions (e.g., Bocquet-Appel and Masset 1982 Cadien et al. 1974 Wood et al. 1992), bioarchaeologists continue to bridge evolutionary and social theory. As such, the study of human biological materials from archaeological contexts is in a key position to facilitate transdisciplinary research, to understand past and present populations through mortuary remains, and to make substantial theoretical contributions to the broadly conceived social sciences. The last of these pursuits is facilitated by an enhanced concern for greater historical contextualization of archaeological human remains and further articulation with social theory that complements and expands upon the biocultural health-related studies that anchored bioarchaeology���s developmental phase in the 1970s and 1980s (Baker and Kealhofer 1996 Cohen and Armelagos 1984 Larsen 1994, 2001 Steckel and Rose 2002). We agree with Goldstein (2006, p. 377), who writes, ������[c]ontext is everything.������ Context, in this sense, is broadly defined in relationship to operational issues of excavation practice as well as theoretical contextualization within an independent field of inquiry. We argue that sophisticated problem orientations are not possible without careful data collection in both the field and the laboratory, and that these data are most powerful when a research question with broad appeal is addressed. This, we think, is the primary defining characteristic of current bioarchaeological practice. Scholars are now addressing research questions not solely biological or evolutionary in orientation, and inferences about the social realm are more visible, paralleling recent scholarship in archaeology (see Casella and Fowler 2005 D��az-Andreu �� et al. 2005 Insoll 2007 Janusek 2004 Lucas 2004 Reycraft 2005 Wells 2001). In this article we argue that the study of identity is one emerging research theme that is uniquely approachable using bioarchaeological research. Building on Barth���s (1969) influential work on the situational nature of ethnic identities, increasing numbers of scholars argue that individuals adopt and manipulate any number of coexisting social identities over the course of their lifetimes. But just what is meant by identity? Identities can be both personal and communal, ascribed and achieved, manipulated and feigned. Gender, age, status, ethnic affiliation, and religion all represent forms of social identities with associated behavioral expectations and roles. Identities are about self-perception and self- promotion as well as constraints imposed by others. It is the process of social interaction within a matrix of intangible social identities that makes the human social world so complex. Identities are ������the process by which the person seeks to integrate his [or her] various statuses and roles, as well as his [or her] diverse experiences, into a coherent image of self������ (Epstein 1978, p. 101). Since nuanced identities are somewhat diffuse and difficult to encapsulate, we also outline what we explicitly reject as identity-based research in bioarchaeology. We argue that identity does not refer simply to origins, migration histories, or biological interconnections, as it may often be used. Rather, by focusing on the social construction of the human experience, the study of archaeological skeletal remains can make unique contributions to our understanding of social life in the past as well as those issues that plague the world today. We note that osteological indicators are both durable and plastic and therefore provide both mutable and immutable information about the identities people were signaling. Cranial and dental modification can be an overt stylistic display, dental pathology and paleodietary 398 J Archaeol Res (2008) 16:397���432 123

analysis can be used to identify dietary choices, activity patterns on the human skeleton can indicate particular behaviors, and biodistance analysis can be used to infer kinship and mate choices. Bioarchaeological approaches also are beneficial because of the temporally sensitive nature of the archaeological record, which allows inferences that are transformational, and not simply historical, in nature. The combination of durability, plasticity, and temporal sensitivity can be used to reconstruct past social processes in a manner simply not possible using archaeo- logical or historical data sets alone. And, in many cases, it is the bioarchaeological data set that should be preferenced for example, with inferences about mate exchange patterns and mestizaje or macroscale migration processes, the former often was inferred from historical documents while the latter was reconstructed based on material culture. It also is clear that the importance and timeliness of social identities research will not be waning any time soon, and the integration of biological data into these discussions can be illuminating. Social pluralism based on ethnic, linguistic, or religious identities can lead to inequality with potential health disparities analyses of health in the past can help illuminate situations where inequality is most likely to lead to health degradation. Genocides often are enacted based on ethnic identities molecular or morphological analyses can help track the development of these distinct identities through time while the study of status and health can contextualize the potential reasons why such violence occurs. Given the continued focus on ethnic violence, tribalism, and nationalist movements in modern political discourse, understanding how identity groups form has repercussions well beyond anthropol- ogy. Finally, misperceptions and discrimination based on age, gender, or sexual orientation derive from facile, unexamined notions of the ������natural order������ of humanity examination of age and gender identities in the past helps define how fluid such categories have been in human history. Within this framework, we focus on specific analytical approaches that articulate with the study of past social identities: osteological sex estimation and gender identity age estimation and identities based on the life course, paleodemography, and identities of inequality and disability and biodistance analysis and ethnic or community identities. However, to more thoroughly summarize recent methodological developments in bioarchaeology, we also discuss recent advances in two areas that do not isometrically articulate with one specific identity category. First, improved biogeochemical methods allow us to study identity at different levels, from the relationship between paleodiet and gender and status identities to the role of population movement in the formation of ethnic identity. Second, a French approach to excavation that is gaining popularity beyond Europe, anthropologie de terrain, provides a way to reconstruct funerary rites in the absence of direct material analogs that can then be used in analogous fashion to studies of mortuary symbolism. In the first half of this article we review each of these methodological domains, highlighting significant improvements and new directions in applied research. In the second half we outline unique bioarchaeological contributions to the study of the social manifestations of identity in the past. J Archaeol Res (2008) 16:397���432 399 123

Current developments in bioarchaeological research methods Methodological developments in osteological sex and age estimation Sex and age estimates remain fundamental to all subsequent osteological analyses and articulate directly with the study of gender and age identities in the past. In addition, sex and age estimates are used to generate a comparative framework for paleopathological, dietary, and mortuary analyses, to provide the empirical basis of past population demographic parameters, and to mitigate nongenetic factors that introduce error into biodistance analysis. Although sexing and aging methods are well known, there is a continued need to improve existing practices. Based on recent publications, three key areas continue to receive attention and represent recognized deficiencies in bioarchaeological practice in age and sex estimation. First, methods need to be developed that can be applied to skeletal remains that are poorly preserved and fragmentary. To this end, researchers are focusing on two approaches: (1) observations that do not rely on whole-bone preservation for the application of a discriminant model (Frutos 2003 Safont et al. 2000), and (2) skeletal elements that tend to preserve well due to their size and density (e.g., Barrio et al. 2006 Case and Ross 2007 Lynnerup et al. 2006). Such approaches are particularly appropriate since the Native American Graves Protection and Repatriation Act (NAGPRA) legislation, and similar laws enacted in other countries (Seidemann 2004), has impacted problem-oriented burial mitigation. Many of the largest and best-preserved comparative samples are being repatriated, creating an academic climate in which sampling designs are more limited and less exclusive. Focusing on those aspects of the human skeleton most likely to preserve is therefore important. The second issue is the population-specific nature of many statistical models of sex assessment. A statistical formula derived for one population with a certain range of male and female metric variation cannot be blindly applied to other populations. Population specificity is particularly challenging for continuous-scale measurements (Ramsthaler et al. 2007 Ubelaker et al. 2002) but also applies to ordinal- or nominal-scale observational data (��uric �� et al. 2005 Rogers 2005 Walrath et al. 2004 Williams and Rogers 2006). In response, new mathematical approaches that avoid the pitfalls of published discriminant function models are starting to appear. Albanese (2007) presents a method based on using the overall sample mean to calculate the most accurate male-female sectioning point for a univariate variable. This approach avoids extrapolating published statistical models and is useful for bioarchaeological data sets that contain multiple individuals. An even more flexible approach using mixture models was presented by Kramer and Konigsberg (1999) and applied for the purposes of sex assessment by Konigsberg and Frankenberg (2007). Mixture analysis is a more general form of discriminant function analysis that is internally generated for a sample, does not rely on existing known-sex discriminant function models, and is well suited for large bioarchaeological data sets that contain multiple individuals. These novel approaches advance the field significantly more than additional case studies demonstrating the effectiveness, or ineffectiveness, of a particular published method in a specific population. Although 400 J Archaeol Res (2008) 16:397���432 123

traditional discriminant function approaches are required for forensic anthropology, there is no need to rely on those methods in bioarchaeological research involving large data sets rather than individual cases. A third challenge for osteologists is sex assessment of subadults. The inability to reliably estimate the sex of preadult individuals hinders analysis of gender and life course identities in past populations as well as comparative analyses of diet and health differences among the younger age classes. This has led to a partial invisibility of subadults in bioarchaeological research design. Recent efforts in sexing subadults have focused on the pelvis and the mandible, producing mixed results. For example, while Franklin et al. (2007) found no size-based mandibular differences between males and females, Loth and Henneberg (2001) noted significant differences in mandible shape, which Scheuer (2002) was unable to reproduce. Others have taken a more holistic approach in their validation studies. Sutter (2003) returned very strong results using eight different published features. Importantly, Sutter���s (2003) work used mummies for whom the sex was known without error, but this highlights a key problem for the future. The number of known-sex subadult individuals in comparative collections is exceedingly small this is the major impediment to progress. As with many traits for adults, there also is likely to be considerable population specificity of subadult sexual dimorphism. Odontometrics may provide a suitable solution and is already promising (e.g., Kondo and Townsend 2004). New developments in age assessment reflect a reaction to the critiques of paleodemography that emerged during the 1980s, in particular the poor correlation between biological and chronological age, especially for individuals older than about 50 years. As a result, recent scholarship on osteological age assessment has focused on identifying morphological variation that is more highly correlated with chronological age-at-death and on revising existing aging standards to identify markers that may increase the upper limit of estimated ages. There also has been some concern with interpopulation variation and the effects of different disease processes on rates of aging. For the most part, authors have focused on validating existing methods for the os pubis (e.g., ��uric �� et al. 2005 Hoppa 2000), auricular surface (e.g., Buckberry and Chamberlain 2002 Igarashi et al. 2005), and sternal rib ends (e.g., Kurki 2005 Schmitt and Murail 2004). Because of its beleaguered history, cranial suture closure has received less attention (Ginter 2005). Several authors have explored digital imaging rather than direct bone observation for pubic age estimation (e.g., Hutchinson and Russell 2001), reflecting a concern with in-field or remote archiving of aging information. Scholars also continue to explore the potential of cementum annulation, a destructive technique that can produce very accurate estimates of age based on regular deposition of cementum on the tooth root (Wittwer-Backofen et al. 2004). A potentially significant new method also has recently been proposed in the form of acetabular changes similar to those seen on the pubic symphysis and auricular surface (Rissech et al. 2006 Rouge-Maillart�� et al. 2004). This is particularly important because the identification of entirely new aging systems is a rarity. Subadult age estimation has received much less attention, in recognition of the fact that growth processes produce more accurate age estimates for subadults than the deterioration and joint breakdown used to age adults. J Archaeol Res (2008) 16:397���432 401 123