AMERICAN JOURNAL OF HUMAN BIOLOGY 23:445–457 (2011) Original Research Article Spousal Violence and Paternal Disinvestment Among Tsimane’ Forager-Horticulturalists 1 1 2 3 4 JONATHAN STIEGLITZ, * HILLARD KAPLAN, MICHAEL GURVEN, JEFFREY WINKING, AND BASILIO VIE TAYO 1Department of Anthropology, University of New Mexico, Albuquerque, New Mexico 87131 2Integrative Anthropological Sciences, Department of Anthropology, University of California at Santa Barbara, Santa Barbara, California 93106 3Department of Anthropology, Texas A&M University, College Station, Texas 77843 4Proyecto Tsimane’ de Salud y Antropologı´a, San Borja, Bolivia Objectives: We develop and test a conceptual model of factors influencing the likelihood of physical wife abuse. The paternal disinvestment model emphasizes that spousal conflict over resource use results from men’s attempts to increase individual fitness at a cost to the family (e.g., through pursuit of extramarital affairs). We propose that men use violence to control women’s responses to the diversion of resources away from the family: to quell women’s objec- tions to male disinvestment, maintain women’s parental investment, and to dissuade women from pursuing relation- ships with other men. Methods: Interviews were conducted among men and women to determine rates of violence and demographic and be- havioral covariates. Structural equation modeling and generalized estimating equations analyses were used to test pre- dictions derived from the model. We also collected data on frequent complaints in marriage and women’s perceptions of arguments precipitating violence. Results: Over 85% of women experienced physical wife abuse (n 5 49). Indicators of paternal disinvestment posi- tively covary with indicators of marital strife and with rates of wife abuse. The wife’s age, matrilocal residence, and presence of joint dependent offspring decrease the likelihood of violence through direct and indirect routes. Conclusions: Wife abuse is linked to the importance of paternal investment in human families, and is a means by which men control women’s responses to a dual reproductive strategy of familial investment and pursuit of extramarital sexual relationships. This framework is more general than traditional sociological and evolutionary perspectives empha- sizing patriarchy and men’s sexual jealousy, respectively. Am. J. Hum. Biol. 23:445–457, 2011. ' 2011 Wiley-Liss, Inc. Spousal violence is the most common form of family vio- interventions to lower its occurrence and mitigate its lence (Levinson, 1989), yet its frequency varies substan- harmful effects. tially within and across populations. Among selected non- This article makes two contributions. First, we develop Western countries, lifetime prevalence of physical violence a general conceptual model of factors influencing the like- by an intimate partner varies from 13 to 61% of ever-part- lihood that women experience physical violence in mari- nered women, with an annual prevalence varying between tal/cohabiting unions. The model considers the impor- 3 and 29% (Garcia-Moreno et al., 2006). In the United tance of paternal investment in human families. It argues States, 22% of women are physically assaulted by current that violence is used by husbands to control wives’ or former partners at some point in their lives (Tjaden and responses to men’s diversion of resources away from the Thoennes, 2000). For the majority (78%) of victimized family, often in pursuit of extramarital affairs. Second, we American women, violence occurs primarily before the rela- evaluate this theoretical approach using data on spousal tionship ends, with 28% of injured women seeking medical conflict and wife abuse, and their correlates among Tsi- treatment following the most recent assault (ibid). mane’ forager-horticulturalists of Bolivia. There are sev- In addition to acute trauma, long-term health conse- eral reasons why the Tsimane’ provide an interesting test quences for physically abused women include chronic of our model linking household demography and resource pain, gastrointestinal and gynecological problems, allocation strategies to spousal conflict and wife abuse. unwanted pregnancy, fetal loss, post-traumatic stress dis- They are traditionally matrilocal early in marriage, have order, and depression (for references see Alio et al., 2009; limited residential privacy and exposure to violence in Campbell, 2002; Heise et al., 1994a,b). Children of abused media, they do not own land and exhibit relatively low women also exhibit higher early mortality and low birth variance in socioeconomic status within and across sexes, weight (Jejeebhoy, 1998, Murphy et al., 2001), suggesting that children are indirect victims of spousal abuse. Spousal violence raises public health and human rights Contract grant sponsor: NSF; Contract grant numbers: BCS-0721237, BCS-0422690; Contract grant sponsor: NIA; Contract grant number: concerns worldwide, yet most theoretical models of 1R01AG024119; Contract grant sponsor: Latin American Institute Grant spousal conflict are developed and tested among an unrep- (University of New Mexico). resentative sample of the world’s population: industrial- *Correspondence to: Dr. Jonathan Stieglitz, MSC01-1040, Anthropology, ized, formally educated, and relatively wealthy Western- Albuquerque, NM 87131, USA. E-mail: [email protected] Received 15 September 2010; Revision received 7 December 2010; ers. A general theory that both explains why husbands Accepted 15 December 2010 engage in physical violence against their wives, and that DOI 10.1002/ajhb.21149 predicts the conditions under which violence is more likely Published online 5 May 2011 in Wiley Online Library (wileyonlinelibrary. to occur would be useful in the design of public health com). VC 2011 Wiley-Liss, Inc. 446 J. STIEGLITZ ET AL. they produce most of their own food, and they lack formal 1990; Rao, 1997), and attempts to control a partner’s patriarchal institutions. In addition, Tsimane’ society is social behavior (Ellsberg et al., 2000; Gage, 2005; Garcia- gradually undergoing integration into the market econ- Moreno et al., 2006; Shackelford et al., 2005; Tjaden and omy. Sporadic wage labor opportunities exist for men, but Thoennes, 2000). Frequency of marital arguments over usually not for women, and men exercise considerable the use of time and resources is also positively associated control over the use of the small amounts of money they with rates of violence (Hoffman et al., 1994; Suitor et al., acquire. The data show that in spite of the fact that poten- 1990). Many studies, however, only report empirical asso- tial aggressors face significant social costs, physical wife ciations and lack a cogent conceptual framework specify- abuse is extremely common. ing direct and indirect pathways through which these The article is organized as follows. The first section effects operate. In this article we attempt to integrate reviews existing literature on male physical violence these previous findings by adopting a perspective empha- against spouses. It shows that there are clear patterns in sizing demographic and behavioral factors expected to the data, but that no single framework integrates the dis- influence the costs and motivations to husbands of engag- parate results. The next section develops our conceptual ing in physical violence against wives. model. Part I discuss how parental investment theory informs our understanding of male–female aggression and introduces why consideration of paternal investment An explanatory framework for male violence against spouses is crucial to understanding human intersexual relation- ships. Part II focuses on the nature of marital conflict and Part I: Parental investment theory, sexual conflict, and coop- the conditions under which spousal violence is more likely eration. Among sexually reproducing species, males and to occur. Part III presents a path model for predicting the females have conflicting reproductive interests and usu- frequency of violence against women. We then introduce ally exhibit asymmetries in parental investment (Emlen the study population and our methodology. The next sec- and Oring, 1977; Maynard, 1977; Parker et al., 1972; Triv- tion presents the results. To conclude, we discuss the em- ers, 1972). Females produce larger, more energetically ex- pirical and theoretical contributions of the analysis to the pensive gametes and almost always invest greater study of intimate partner violence and human sexual rela- amounts of time and energy in offspring than males. tions, and their implications for public health. Female reproductive success is therefore limited by access to resources critical for reproduction whereas male repro- ductive success is constrained by access to fertile females. Factors increasing the likelihood of violence against women: Among most mammals, male investment in reproduction Existing literature and explanatory approaches is limited to courtship and copulation (i.e., mating effort), rather than provisioning and care (i.e., parenting effort), Early studies report positive correlations between vio- (Clutton-Brock, 1991). Mating opportunities may be lence frequency and either level of marital restrictions obtained by successful courtship or through coercive imposed upon women or men’s control of wealth (Lester, means involving aggression toward other males or unrec- 1980; Levinson, 1989). Other studies highlight historical eptive females. Sexual coercion may increase the likeli- relationships between patriarchal dimensions (e.g.,