Поиск по сайту




Пишите нам: info@ethology.ru

Follow etholog on Twitter

Система Orphus

Новости
Библиотека
Видео
Разное
Кросс-культурный метод
Старые форумы
Рекомендуем
Не в тему

список статей


Female allies and female power: A cross-cultural analysis

Catherine Yanca, Bobbi S. Low

Received 28 March 2003; received in revised form 30 June 2003

Article Outline

1. Introduction

1.1. Hypothesis and predictions

1.2. Women and resource control

1.3. Women and power/authority

1.4. Attitudes about women's behavior

2. Methods

3. Results

3.1. How are female allies affected by ecological variables?

3.2. What are the effects of having female allies?

4. Discussion

4.1. The functions of female allies

Acknowledgment

References

Copyright

1. Introduction

In the context of human evolutionary history, male–female conflicts of interest arise over reproduction, who can control important resources, and who has power (e.g., Low, 2000, Ridley, 1993). Further, men exert an extraordinary degree of sexual control over women, compared to other primates (e.g., Smuts, 1992, Smuts, 1995). With regard to resource control and formal political influence within the group, men have undoubtedly had more power than women throughout human evolution. Evolved sex differences in resource acquisition and risk taking contribute to divergent male and female patterns (e.g., Low, 2000, chap. 4 and 6). In most circumstances, men stand to gain more reproductively than do women from intense striving and risk taking. Yet, women in most societies still can control some resources and have some influence on political outcomes, even if informally. There are a few examples of modern-day societies in which particular women have significant amounts of resource control and hold powerful political positions, but such societies are nonetheless as rare today as they have been throughout most of human evolutionary history (e.g., Low, 2000).

The rarity of societies in which women control resources and enjoy political power raises questions about the ecological and social predictors of such autonomy and status. Under what circumstances are women likely to have significant resource control and/or political power? Are the societies in which women have relatively more power associated with specific ecological factors (e.g., predictability of ecological variables) or social factors (e.g., prolonged periods of male absence)?

Cross culturally, most societies are “patriarchal”—patrilineal, patrilocal, and with men controlling both material resources and sexual access to women (e.g., Smuts, 1992, Smuts, 1995). Here, we are interested in societies in which women can overtly control resources, have political decision-making power, or both (cf. Low, 1990b, Low, 1992). Such societies are not only rare but are highly variable with respect to marriage systems, residency patterns, subsistence, and type of economy (e.g., Sanday, 1981, Whyte, 1978).

Here, we focus on women's resource control and power in polygynous societies in the odd-numbered half sample (N=93 societies) of the Standard Cross-Cultural Sample (Murdock & White, 1969) using many of Whyte, 1978, Whyte, 1979 codings on female status. We examine only polygynous societies for two reasons: first, in this sample, only 7 of the 93 societies are monogamous—too few to permit statistical tests within monogamous societies; and second, the variation among the polygynous societies offers the possibility of identifying important ecological and social predictors of women's resource control and power.

In polygynous societies, women may form coalitions with their cowives. Could the presence of any “female allies,” not simply sisters, contribute to women's ability to control resources and exert influence? Although significant resource control and political power may be rare, it seems likely that in cases in which women have allies, they are more likely to have significant resource control and power than when they are more isolated. Further, ecological conditions affect many aspects of life (e.g., Low, 1988b, Low, 1989, Low, 1990a) so that the likelihood of living with female allies may differ under varying ecological conditions. Here, we seek pathways that begin with ecological conditions, flow through sociocultural conditions, and influence female resource control and power.

1.1. Hypothesis and predictions

We assume that the important causal pathways begin with ecological conditions, flow through sociocultural institutions, and consequently affect female resource control and power. However, the statistical methods we use are necessarily correlational.

We predict that within polygynous societies, married women will have greater resource control and political activity when they have female allies because male–female conflicts of interest over resource use are potentially so much greater in polygynous than monogamous systems (e.g., Low, 2000).

Kin are the likeliest allies, and of course in many societies, male kin assist and protect women (e.g., Smuts, 1992, Smuts, 1995). Here we suggest that unrelated female allies also matter. We hypothesize that women will have more access to female allies in polygynous societies in which women (1) are in sororal marriages, (2) are not captured from other groups, (3) can remain close to their families or live relatively close to their kin, and/or (4) are in marriages with few wives (low degree of polygyny). We first test whether ecological variables influence access to female allies (cf. Low, 1990a, Low, 2000), then whether when women have more (rather than fewer) allies, they will also (1) have greater resource control, (2) have greater political power and positions of authority, and (3) less often encounter negative attitudes about women and restrictive ideas about women's behavior.

1.2. Women and resource control

Resources garnered by women are typically used for family, offspring, and those who reciprocate (e.g., Hill & Hurtado, 1996, Low, 1990b; review by Low, 2000). In polygynous societies, cowives may be important reciprocators. If cowives are able to cooperate, collectively they may be able to gain more control over resources. Resource collection and control may afford polygynously married women greater lineage success through their sons if they are able not only to accumulate but also to pass on these resources, as among the Ashanti (although this circumstance is uncommon; Low, 1992, Low, 2000). However, conflicts among an individual woman, her husband, their joint offspring, and her cowives and their offspring can arise over allocation of resources as interests diverge Low, 1990b, Low, 2000, and divorce is not an uncommon outcome (Betzig, 1989). One mitigating factor is whether cowives are sisters Betzig, 1989, Burbank, 1987, Chisholm & Burbank, 1991, Murdock, 1949. A polygynously married woman may have little resource control or power in coalitions with nonsororal cowives because her reproductive interests are in competition with these cowives.

We predict that women will have greater resource control where they remain close to potential allies; for example, when cowives are sisters rather than nonkin. If true, we expect greater resource control in societies with sororal (rather than nonsororal) polygyny and societies with lower degrees of polygyny. For example, Cowlishaw and Mace (1996) showed that monogamous societies are frequently associated with absence of sex-biased inheritance, while strongly polygynous systems predominantly are associated with male-biased inheritance. Hartung (1982; also Clarke & Low, 2001) demonstrated that male-biased inheritance increases as degree of polygyny increases. Finally, Low, 1990b, Low, 2000 found that women's control over the fruits of male labor decreases as degree of polygyny increases.

1.3. Women and power/authority

Political power is reproductively profitable for men because it can translate into greater reproductive access or proxies thereof in modern contracepting societies (Pérusse, 1993, Pérusse, 1994; review in Low, 2000). In contrast, conditions in which women can translate “excess” resources (resources over and above the requirements for offspring health) are rare (Low, 2000). Interestingly, in the odd-numbered half sample of the Standard Cross-Cultural Sample, men alone hold formal political power in 70% of societies (Whyte, 1978). In just over 7%, both sexes have power; but men are still more powerful and/or more numerous in politics (Whyte, 1978; data are missing for the remaining 13%). Very few societies have been examined in which women's power is clearly equal to that of men, but the reliability of the codes has been challenged Low, 1990b, Low, 2000. The societies in which women's decision-making capabilities are centered on the same issues as men's are complex, matrilineal or double-descent systems with heritable resources (Low, 1990b), and these appear to be the main correlates of women's political power. Societies in which women hold political power can be “settled or migratory, matrilocal, or patrilocal include subsistence activities not generally associated with women's ability to control resources” and, of interest to us here, can be highly polygynous (Low, 2000).

In societies in which females have effective allies or are geographically close to such allies, we expect to find evidence of women in positions of power and authority whether these be formal or informal. In half of the societies studied, women can participate in community gatherings, even if segregated. While women in formal politics are rare, informal influence may be common (Low, 2000). Women in societies in which they have sororal cowives or remain close to their kin may, for example, have great power over kin networks and/or authority over those dwelling in the household.

1.4. Attitudes about women's behavior

Data exist on generally held beliefs or behavior related to how women are treated or expected to act (e.g., Whyte, 1978, Whyte, 1979). We predict that in societies in which cowives are sisters, and/or the couple lives closer to the female's than the male's relatives, attitudes about female premarital sex and extramarital affairs will be less constraining for women and more restrictive for men. Also under such conditions, we predict that emphasis on male machismo will be less than in societies in which females have fewer allies.

2. Methods

We examined the odd-numbered societies of the Standard Cross-cultural Sample (Murdock & White, 1969). These societies are geographically representative and include, within geographic areas, societies of different language groups and societies studied by a qualified ethnographer resident for a substantial period of time.

Whyte, 1978, Whyte, 1979 and Low (1990b) found that no single variable could be used to measure women's “status,” and that measures of female resource control and authority over others (outside the context of the “intimate household”) do not covary in ways that allow them to be combined usefully (Low, 1990b). We therefore have used multiple measures that fall into three categories: ecological variables, female ally variables, and resource control/political/behavioral variables (Table 1).

Table 1.

Description of ecological, female ally, and resource control, political, and behavioral variables

Variable name Variable description Variable categories Variable source
Ecological variables:
Pathogen stress Relative stress on a society as a result of pathogens within it 1=Low (0–8/21 total pathogen score, see Low) Low, 1988a, Low, 1988b
2=Medium (9–15/21 total pathogen score)
3=High (16–21/21 total pathogen score)
Animal husbandry Contribution of animal husbandry to subsistence 1=Low, <16% contribution Murdock, 1967, Murdock, 1981
2=Moderate, <36 contribution, >16%
3=High >36% contribution
Fishing Contribution of fishing to subsistence 1=Low, <16% contribution Murdock, 1967, Murdock, 1981
2=Moderate, <36 contribution, >16%
3=High >36% contribution
Female ally variables:
Distance from natal family What are the “relative” distances moved by the bride and groom away from natal families after marriage 1=Female generally moves farther Whyte (1978)
2=Male generally moves farther
Female capture Presence or absence of female capture in society 1=Society captures females Murdock and White (1969), Murdock, 1967, Murdock, 1981
2=Society does not capture females
Sororal polygyny Presence or absence of sororal polygyny in society 1=No sororal marriages in society Murdock and White (1969), Murdock, 1967, Murdock, 1981
2=Some or many sororal marriage
Degree of polygyny Relative measure of percent of marriages that are polygynous 1=Highly polygynous (>20% marriages) Murdock and White (1969), Murdock, 1967, Murdock, 1981
2=Low degree polygyny (<20% marriages)
Resource control variables:
Inheritance of property Who can inherit property of some economic value 1=Males only and male biased Whyte (1978)
2=Equal, female biased, females only
Marriage payment Payments are absent or only token versus bride price 1=Bride price and dowry Murdock and White (1969), Murdock, 1967, Murdock, 1981
2=Token bride price or no payment
Control of male labor Who controls disposal and use of fruits of labor by men alone 1=Males only and male biased Whyte (1978)
2=Equal, female biased, females only
Control of dwellings Who owns or controls the use of dwellings 1=Males only and male biased Whyte (1978)
2=Equal, female biased, females only
Political power variables:
Authority over children older than four Who has authority over postinfant unmarried children living at home 1=Males only and male biased Whyte (1978)
2=Equal, female biased, females only
Female kin power Relative control females have over their kin: “Kin Power Scale” 1=Low power in kinship contexts Whyte (1978)
2=High power in kinship contexts
Attitude/behavioral variables:
Emphasis on male machismo Is there a generally high value placed on males being aggressive, strong, and sexually potent 1=Marked and moderate emphasis Whyte (1978)
2=Little or no emphasis
Female extramarital affairs Are married women allowed to, and do they in fact, have extramarital affairs 1=Not allowed, rare and common Whyte (1978)
2=Allowed or very common

We tested for two kinds of associations between (1) ecological predictors (Table 1) and female allies and (2) between female allies and measures of female resource control and political power. We assume that some ecological factors affect many sociocultural arrangements, including the near presence of female allies and that the presence of female allies interacts with women's access to and control of resources, and political participation.

So many variables and so many potential relationships exist that we first explored a “full” conceptual model in a bivariate fashion, using all theoretically relevant variables (7 ecological variables, 4 ally variables, and 12 resource control/political/behavior variables). We used standard bivariate statistical analyses. However, because bivariate analysis hides covariance, we used multivariate regressions to measure direction and magnitude of relationships (SPSS version 10.1). The multivariate approach enabled us to incorporate multiple predictors in our conceptual model; first between the ecological variables and the ally variables, then between the ally variables and resource control and political variables.

Multivariate analyses were run as a series of logistic regressions because there were only 93 cases and the data were primarily ordinal (chosen over multinomial logit because dependent variables with three categories could be collapsed into dichotomies with little loss of information; Allison, 1999, Menard, 1995). Ecological variables (made up of three categories) were made into dummy variables, and all other variables were made dichotomous. For example, inheritance of property was collapsed from four categories (male biased, male preference, equal inheritance, and female biased) into two categories: male biased and preference; and equal and female-biased inheritance. Diagnostic multiple regressions revealed some multicollinearity in the ecological variables; for example, we found that contribution of agriculture, hunting, and gathering to subsistence displayed low tolerances with degree of polygyny. That is, these ecological variables show some multicollinearity with degree of polygyny. In addition, some of the variables suffered from limited data.

We therefore removed variables that did not exhibit significance under the bivariate analysis and/or had low tolerances under the multiple regression tests for multicollinearity. We then conducted a second set of logistic regressions (“reduced model”) using only the remaining variables presented here (Table 1). The reduced model is more robust than the full model: it does not suffer as severely from problems posed by limited data and multicollinearity. We nonetheless point out cases in which the results of reduced model were different from those of the bivariate analysis and/or the full model multivariate analysis.

3. Results

We examine first the relationships between ecological variables and female ally variables; then we examine relationships between female ally variables and female resource control/power variables. In each case, we move from variables not exclusive to polygynous societies (e.g., relative distance moved from natal families and female capture) to those exclusive to polygynous societies (sororal marriages and degree of polygyny).

3.1. How are female allies affected by ecological variables?

In the reduced model, only three ecological variables showed relationships with the four female ally variables (Fig. 1). The first female ally variable we consider is distance from natal family: when husbands move farther (compared to their wives) from natal families, wives are more likely to have allies. When animal husbandry was unimportant, husbands were likely to move farther from their natal families than wives after marriage. This relationship is not surprising: anthropologists have long known about the strong male bonding in pastoral/animal husbandry societies (e.g., Nisbett & Cohen, 1996).


View full-size image.

Fig. 1. Conceptual model of pathways of influence. Solid lines: P<.05; dashed line: .05<P<.09.


The reduced model also showed that in societies in high pathogen stress areas, women are captured more than in medium pathogen-stress areas Low, 1988a, Low, 1988b. The more wives or concubines are captured, as opposed to courted, the fewer allies women are likely to have because such captive women are displaced from their kin. In areas of high pathogen stress, the likelihood of female capture is high so wives typically come from different villages and do not already know each other (Low, 1988b).

Societies with high pathogen stress tend not only to have high rates of capture of women for concubines and wives, but also tend to have nonsororal polygyny (wives are not related, contrasting with the general pattern of preference across societies for sororal polygyny). When cowives are sisters, women have more resource control and more political power and/or authority than when cowives are unrelated. In the reduced model, low pathogen stress increased the odds of sororal marriages. In addition, sororal marriages were very likely where animal husbandry had no or little importance. We suspect, as with distance moved from natal family (above), this relates to the value of male–male coalitions in pastoral/animal husbandry societies (e.g., Nisbett & Cohen, 1996). These results were also clear in the bivariate analysis. Oddly, in the full model, societies in which animal husbandry contributed very little to subsistence showed decreased odds of marriages being sororal. We suspect this discrepancy exists because the full model suffers from both limited data and multicollinearity.

The greater the degree of polygyny, the more cowives a man may have, and thus the greater the potential conflict of interest among wives. Under the reduced model, degree of polygyny was significantly related to two ecological variables. Degree of polygyny was low in societies with both low and medium pathogen stress and high in areas of high pathogen stress. It was also high in societies in which the contribution of fishing to subsistence was low or moderate. Both relationships with pathogen stress and fishing showed up in the bivariate analysis (cf. Low, 1988b).

3.2. What are the effects of having female allies?

In the reduced model regression results, when husbands moved farther than their wives, there was little or no emphasis on male machismo (Table 2, Fig. 1). In societies in which husbands moved farther than their wives from their natal families, all the following became more likely: sex equal and female inheritance of property, sex equal and female control of dwellings, sex equal and female authority over unmarried children in the household, and (marginally) acceptance of female extramarital affairs (Table 2, Fig. 1).

Table 2.

Significant results of reduced model multivariate logistic regression

Ecological variables B Exp(B) Significance Ally measurements B Exp(B) Significance Other social variables
Animal husbandry 1 1.55 4.73 .021 Distance from natal family 2.88 17.87 .008 Inheritance of property
1.61 5.02 .025 Control of dwellings
1.61 5.02 .051 Authority over kids >4
1.49 4.46 .030 Female extramarital
2.13 8.44 .035 Emphasis on male machismo
Pathogen Stress 2 −2.42 0.09 .006 Female capture
Pathogen Stress 1 4.30 74.00 .024 Sororal 1.84 6.32 .066 Inheritance of property
Pathogen Stress 2 2.55 12.81 .076 2.22 9.26 .008 Marriage payment
Animal Husbandry 1 4.10 60.62 .018
Pathogen Stress 1 −1.74 0.18 .015 Degree of polygyny −3.68 0.03 .007 Inheritance of property
Pathogen Stress 2 −1.73 0.18 .008 −2.77 0.06 .009 Control of male labor
Fishing 1 1.10 3.01 .019 −2.38 0.09 .009 Female kin power
Fishing 2 0.76 2.14 .085

Female capture presents something of a conundrum. In the reduced model presented here, surprisingly, the presence or absence of female capture showed no relationship to women's control of resources or political power (Table 2, Fig. 1). In the bivariate analysis, in contrast, societies without female capture tended to have token or no marriage payments (P=.036) and relatively high female kin power (P<.001). All societies with female capture had low or moderate female kin power. Societies without female capture had little or no emphasis on male machismo (P=.024) and were marginally more likely to accept female extramarital affairs (P=.065). We suspect that collinearity is the source of these differences.

Sororal marriage societies tended to have no, or token, marriage payments and tended not to require bride price or dowry (Table 2, Fig. 1). Sororal polygyny marginally increased the odds of sex-equal and female-biased inheritance of property (Table 2, Fig. 1). The reduced model also showed that in highly polygynous societies, female-biased and sexually equal inheritance of property were rare (also found by Low, 1988b; Table 2, Fig. 1). In addition, more exclusive male control of male labor was likely in societies with high degrees of polygyny (Table 2, Fig. 1). Finally, women had less power in kin groups in societies with high degrees of polygyny, a relationship not detected by Low (1988b).

4. Discussion

We hypothesized that, within polygynous societies, ecological relationships influence access to female allies, which in turn influence women's resource control, political power, and prevalence of negative attitudes about women's behavior. While the multiple regressions show direction and magnitude of the relationships between the explanatory and dependent variables, they do not confirm the particular pathways of causality we imagined in our model. We are confident that ecological constraints are the primary explanatory variables that influence social variables. However, we cannot say with certainty that the female ally variables lead to the resource control/conditions. These two kinds of variables will obviously interact. We suggest that access to female allies (in this model, the result of proximity to kin) is likely to influence the amount of resource control and political power/authority, but we suspect that as levels of resource control or political power/authority change, they may in turn affect women's proximity and/or access to female allies.

Ecological relationships clearly do influence women's access to female allies. Women may have more allies in societies with low contribution of animal husbandry for two reasons: first, in these societies, husbands were more likely to move farther from their natal families than their wives; and second, sororal marriages are more common in these societies. Females may have more allies in many societies with high contribution of fishing to subsistence because the degree of polygyny tends to be lower in fishing than in other societies, and in some fishing societies, male absence is prevalent. In societies with low pathogen stress, female capture is rare, degree of polygyny tends to be low, and sororal marriages are more common; thus, women may have more allies in societies with low pathogen stress than in those with high stress Low, 1988b, Low, 1990a.

We predicted a positive relationship between female allies and women's resource control, political power, and prevailing attitudes about women's behavior. That is, in conditions in which women can gain resource control and power, we expect the attitudes of others, over time, to come to consider such control appropriate. The reduced model shows some relationships between female allies and resource control: women are more likely to inherit property and have control over dwellings in societies in which husbands moved farther from their natal families than their wives. Compared to nonsororally polygynous societies, in societies with sororal polygyny, women are more likely to inherit property, and marriage payments are more often token or absent. Women are also more likely to inherit property and to have greater control of male labor in societies with low degrees of polygyny.

Women's authority was related both to degree of polygyny and to the relative distance men versus women moved from their natal families. Women in societies in which husbands moved farther than their wives from natal families were more likely to have authority over unmarried children in the household. In societies with low degree of polygyny (fewer cowives), women may have more authority because they have more power in kin groups. Both cases reflect women in positions of power over the household or kin and not at higher level within the society.

The final aspect of our second prediction is that when women have more female allies, restrictive ideas about women's behavior should be relatively rare (cf. Smuts, 1995). Two measurements showed significant relationships in the reduced model. When husbands move farther than their wives after marriage, women are less likely to encounter a strong emphasis on male machismo and are likely to enjoy acceptance of extramarital affairs.

Our analysis suggests that interactions are complex among ecological influences, women's access to female allies, and women's resource control and “voice” or “power.” The end result is that both ecological and socioecological conditions appear to influence women's access to potential allies. Pathogen stress and mode of subsistence influence distance from natal family, degree of polygyny, and likelihood of female capture and of sororal polygyny. In turn, access or closeness to female allies shows significant relationships with female resource control, female positions of power and authority, and attitudes about female behavior. When polygynously married women remain close to their kin, have fewer cowives, and have sisters as cowives, they have more resource control and more authority.

4.1. The functions of female allies

We focused here on how women's resource control and power relate to access to female allies: female kin and nonrelated women (e.g., cowives) whose interests converge. In both human and nonhuman species, coalitions of females can be powerful. In some nonhuman primate species, females form coalitions that can act to assist males, resist male sexual advances, forage together, and harass lower-ranking females (e.g., Hemelrijk, 2002, Low, 2000, Moore et al., 2001, Packer et al., 1995, Parish, 1996, Smuts, 1992, Wasser, 1983). Such female–female cooperation and competition have reproductive impacts not only in avoiding male dominance (e.g., Low, 2000, Parish, 1996) but in within-sex competition for resources Pusey et al., 1997, Silk, 2002, Wasser, 1983. For example, unrelated females may attack the offspring of lower status females.

Human female–female coalitions tend to be familial and are seldom powerful outside the household Irons, 1983, Low, 1990b. Coalitions seem to serve the purpose of information exchange on resource gathering, childcare, and work related to subsistence. Here, we have shown that their impact may be more far reaching.

In the context of cowife cooperation, Irons (1983) discussed female choice models as an example of how, in Australian Aborigines, women may enter into polygynous marriages not because of the suitor's resources but rather those provided by her would-be cowives. Hames (1996) suggested that Yanomamô cowives may offer each other advantages such as shared childcare, reduced workloads, and more work accomplished per unit time. Among the Yanomamô, polygynous households receive more food resources from other households than do monogamous households (Hames, 1996). These advantages accrue more or less independent of the husband's economic activities and produce something similar to economies of scale (Grossbard, 1980). In such situations, women appear to be little concerned about consequences of competition with cowives for their husband's resources.

In contrast, women may be isolated and have competitive cowives in male-coercion forms of polygyny (such as female capture), resulting in conflict and jealousy among cowives. Of course, jealousy and aggression among cowives are not simply dependent on male coercion but are widespread cross culturally. A number of studies show that conflict among wives is reduced when cowives are sisters Betzig, 1989, Burbank, 1987, Murdock, 1949.

Similarly, Chisholm and Burbank (1991) found in southeast Arnhem Land that sororally polygynous marriages had significantly fewer cowives than nonsororal marriages. Whether this means that related wives are better able to counter their husband's interests (of marrying more women and diluting resources among wives and their possible children), or simply that the number of sororal cowives is limited by family size, the result is fewer and less competitive cowives in sororal marriages. Sear, Mace, and McGregor (2003) suggest that social pressures from a woman's father- and mother-in-law to bear many children may be important in patrilineal polygynous societies. Women in sororally polygynous marriages were younger than their nonsororal cowives when they bore their first children and had longer inter-birth intervals (Chisholm & Burbank, 1991); perhaps the longer IBIs reflect the ability of women in sororally polygynous marriages to control some aspects of their reproduction, although we know of no good tests. Furthermore, the children of sororal cowives survived to age five more often than did children of nonsororal cowives. Chisholm and Burbank (1991) suggested that although women in sororal and nonsororal marriages may have the same lifetime fertility, fertility has a lower cost for women in sororal marriages than for those in nonsororal. This is because sororal cowives can cooperate effectively in parental effort, rather than paying the costs of competition among cowives who are not related.

These results suggest that indeed ecological factors can have subtle, previously unsuspected, effects on sociocultural patterns. Here, we suggest that ecological factors such as subsistence type and pathogen stress can affect women's lives in many ways—resource control, political power—as a result of shifting the availability of female allies.

Acknowledgements

We thank Beverly Strassmann of the University of Michigan Department of Anthropology, for substantive comments, and Laura Klem of the University of Michigan Center for Statistical Consulting and Research, who helped us develop the models used here. Thanks to the members of Natural Resources 505, Human Resource Ecology, for critical comments and help on early drafts, and to our anonymous reviewers for help in the submitted manuscript.

References

Allison, 1999 1.Allison PD. Logistic regression using the SAS system: theory and application. Cary, NC: SAS Institute; 1999;.

Betzig, 1989 2.Betzig L. Causes of conjugal dissolution: a cross-cultural study. Current Anthropology. 1989;30:654–676. CrossRef

Burbank, 1987 3.Burbank V. Female aggression in cross-cultural perspective. Behavior Science Research. 1987;21:71–100.

Chisholm & Burbank, 1991 4.Chisholm JS, Burbank VK. Monogamy and polygyny in southeast Arnhem land: male coercion and female choice. Ethology and Sociobiology. 1991;12:291–313.

Clarke & Low, 2001 5.Clarke AL, Low BS. Testing evolutionary hypotheses with demographic data. Population and Development Review. 2001;27:633–660.

Cowlishaw & Mace, 1996 6.Cowlishaw G, Mace R. Cross-cultural patterns of marriage and inheritance: a phylogenetic approach. Ethology and Sociobiology. 1996;17:87–97.

Grossbard, 1980 7.Grossbard A. The economics of polygamy. In:  Simon JL,  Da Vanzo J editor. Research in Population Economics. vol. 2:Greenwich, CT: JAI Press; 1980;p. 321–350.

Hames, 1996 8.Hames R. Costs and benefits of monogamy and polygyny for Yanomamö women. Ethology. 1996;17:181–199.

Hartung, 1982 9.Hartung J. Polygyny and the inheritance of wealth. Current Anthropology. 1982;23:1–12. CrossRef

Hemelrijk, 2002 10.Hemelrijk CK. Self-organizing properties of primate social behavior: a hypothesis for intersexual rank overlap in chimpanzees and bonobos. Evolutionary Anthropology. 2002;(Suppl. 1):91–94.

Hill & Hurtado, 1996 11.Hill K, Hurtado AM. Ache life history: the ecology and demography of a foraging people. New York: Aldine de Gruyter; 1996;.

Irons, 1983 12.Irons W. Human female reproductive strategies. In:  Wasser SK editors. Social behavior of female vertebrates. New York: Academic Press; 1983;p. 169–213.

Low, 1988a 13.Low BS. Measures of polygyny in humans. Current Anthropology. 1988;29:189–194. CrossRef

Low, 1988b 14.Low BS. Pathogen stress and polygyny in humans. In:  Betzig L,  Borgerhoff Mulder M,  Turke P editor. Human reproductive behavior: a Darwinian perspective. Cambridge, UK: Cambridge University Press; 1988;p. 115–128.

Low, 1989 15.Low BS. Human responses to environmental extremeness and uncertainty: a cross-cultural perspective. In:  Cashdan E editors. Risk and uncertainty in tribal and peasant economies. Boulder, CO: Westview Press; 1989;p. 229–255.

Low, 1990a 16.Low BS. Marriage systems and pathogen stress in human societies. American Zoologist. 1990;30:325–339.

Low, 1990b 17.Low BS. Sex, power, and resources: ecological and social correlates of sex differences. Journal of Contemporary Psychology. 1990;27:45–71.

Low, 1992 18.Low BS. Sex, coalitions, and politics in pre-industrial societies. Politics and the Life Sciences. 1992;11(1):63–80.

Low, 2000 19.Low BS. Why sex matters. Princeton: Princeton University Press; 2000;.

Menard, 1995 20.Menard S. Applied logistic regression analysis. Thousand Oaks, CA: Sage Publications; 1995;.

Moore et al., 2001 21.Moore AJ, Gowaty PA, Wallin WG, Moore PJ. Sexual conflict and the evolution of female mate choice and male social dominance. Proceedings of the Royal Society of London. Series B. 2001;268:1–7. MEDLINE | CrossRef

Murdock, 1949 22.Murdock GP. Social structure. New York: Macmillan; 1949;.

Murdock, 1967 23.Murdock GP. Ethnographic atlas. Pittsburgh: University of Pittsburgh Press; 1967;.

Murdock, 1981 24.Murdock GP. Atlas of world cultures. Pittsburgh: University of Pittsburgh Press; 1981;.

Murdock & White, 1969 25.Murdock GP, White D. Standard cross-cultural sample. Ethnology. 1969;8:329–369.

Nisbett & Cohen, 1996 26.Nisbett RE, Cohen D. Culture of honor: the psychology of violence in the South. Boulder, CO: Westview Press; 1996;.

Packer et al., 1995 27.Packer C, Collins DA, Sindimwe A, Goodall J. Reproductive constraints on aggressive competition in female baboons. Nature. 1995;373:60–63. MEDLINE | CrossRef

Parish, 1996 28.Parish AR. Female relationships in bonobos (Pan paniscus): evidence for bonding, cooperation, and female dominance in a male-philopatric species. Human Nature. 1996;7:61–96.

Pérusse, 1993 29.Pérusse D. Cultural and reproductive success in industrial societies: testing the relationship at proximate and ultimate levels. Behavior and Brain Science. 1993;16:267–322.

Pérusse, 1994 30.Pérusse D. Mate choice in modern societies: testing evolutionary hypotheses with behavioral data. Human Nature. 1994;5:255–278.

Pusey et al., 1997 31.Pusey AE, Williams J, Goodall J. The influence of dominance rank on the reproductive success of female chimpanzees. Science. 1997;277:828–831. MEDLINE | CrossRef

Ridley, 1993 32.Ridley M. The red queen: sex and the evolution of human nature. New York: Viking; 1993;.

Sanday, 1981 33.Sanday PR. Male power and female dominance. Cambridge: Cambridge University Press; 1981;.

Sear et al., 2003 34.Sear R, Mace R, McGregor IA. The effects of kin on female fertility in rural Gambia. Evolution and Human Behavior. 2003;24:25–42.

Silk, 2002 35.Silk J. Females, food, family, and friendship. Evolutionary Anthropology. 2002;11:85–87.

Smuts, 1992 36.Smuts BB. Male aggression against women: an evolutionary perspective. Human Nature. 1992;3:1–44.

Smuts, 1995 37.Smuts BB. The evolutionary origins of patriarchy. Human Nature. 1995;6:1–32.

Wasser, 1983 38.Wasser S. Reproductive competition and cooperation among female yellow baboons. In:  Wasser S editors. Social behavior of female vertebrates. New York: Academic Press; 1983;p. 349–390.

Whyte, 1978 39.Whyte MK. Cross-cultural codes dealing with the relative status of women. Ethology. 1978;17:211–237.

Whyte, 1979 40.Whyte MK. The status of women in pre-industrial society. Princeton, NJ: Princeton University Press; 1979;.

Population Studies Center, School of Natural Resources and Environment, University of Michigan, Ann Arbor, MI, 48109-1115, USA

Corresponding author.

PII: S1090-5138(03)00065-5

doi:10.1016/S1090-5138(03)00065-5



2007:11:12