Control tactics and partner violence in heterosexual relationships

Transcription

1 Evolution and Human Behavior 30 (2009) Control tactics and partner violence in heterosexual relationships Nicola Graham-Kevan, John Archer University of Central Lancashire, Preston, Lancashire, United Kingdom Initial receipt 18 December 2008; final revision received 18 June 2009 Abstract This study investigated sex-specific predictors of violent and nonviolent mate guarding used by men (n=399) and women (n=951) in heterosexual relationships, using both self-reports and reports on partners. We found, contrary to some previous evolutionary assumptions, that men and women showed similar degrees of controlling behavior, and that this predicted physical aggression to partners in both sexes. We also predicted from evolutionarily based studies that men's and women's control and aggression would vary as a function of female fecundity and mate value (relative to peer group and to partner). Fecundity was associated with men's and women's controlling behavior, but not their physical aggression: relationships where the woman was fecund showed higher rates of control. According to partners' reports, men and women who had lower mate values showed more controlling behavior and (to a lesser extent) more physical aggression. There was no support for the prediction that higher mate-value partners would be guarded more than lower mate-value ones. The following limitations are discussed: the sample and method of data collection, and the lack of information on the women's hormonal status Elsevier Inc. All rights reserved. Keywords: Mate guarding; Relationships; Physical aggression; Injuries; Controlling behavior; Evolutionary; Sex differences 1. Introduction 1.1. Mate guarding, control and physical aggression in relationships Most applications of evolutionary theory to partner violence view it as an extension of mate guarding, which is widespread in nonhuman animals (Parker, 1974). In species that have internal fertilisation and require paternal investment, paternity certainty will be increased by mate guarding. This logic has been applied to men's violence to their partners by Wilson and Daly (1992, 1993, 1996), who have linked negative forms of mate guarding, for example directly seeking to control a partner's behavior, to physical aggression by men to their partners. Underlying this behavior is a proprietary male mindset, whose function is to maximise paternity certainty. While its evolutionary logic is sound, the same authors have also combined with two others, who have argued that partner violence is largely male to female (Dobash, Dobash, Wilson, & Daly, 1992), a Corresponding author. School of Psychology, University of Central Lancashire, PR1 3TQ Preston, United Kingdom. address: ngraham-kevan@uclan.ac.uk (N. Graham-Kevan). view typically associated with the patriarchal explanation of partner violence (Dobash & Dobash, , 1980, 1998). This view runs counter to a considerable body of evidence from industrialized Western nations (Archer, 2000, 2002, 2006; Dutton, 2006; Felson, 2002), indicating that both sexes contribute to partner violence. While it is clear that for a man to maximise his reproductive potential he needs to invest only in his own biological offspring and hence avoid being cuckolded, it is also the case that for a woman to maximise her fitness she needs to secure adequate provisions for herself and her offspring, which usually means monopolising the father's resources. Indeed, it is clear that both men and women show sexual jealousy (e.g., Felson, 2002; Mullen & Martin, 1994; White, 1981), which follows if both sexes have reasons to guard their mates. Buss (1988) took a wider view of mate guarding, in the form of mate retention tactics, a term that included both attempted control and the use of force, and positive inducements, such as making oneself more appealing to the mate, providing gifts and conspicuous displays of resources. Most of the tactics reported by American couples were of this type, although a minority involved threats, spreading rumors, or violence to a rival or the mate (Buss, 1988; Buss & Shackelford, 1997). Flinn (1988) also studied /$ see front matter 2009 Elsevier Inc. All rights reserved. doi: /j.evolhumbehav

2 446 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) mate guarding in both sexes, in his study of mate guarding in a Caribbean village. In the present study, we consider negative forms of mate guarding, controlling behavior and physical aggression, in both sexes. Since the view that much of men's partner violence can be explained by paternity uncertainty is prevalent in evolutionary thinking (e.g., Figueredo et al., 2001; Peters, Shackelford, & Buss, 2002; Shackelford, Goetz, Buss, Euler, & Hoier, 2005; Wilson & Daly, 1992, 1993, 1996), we first assessed whether controlling behavior is displayed more by men than by women, as would be predicted if paternity uncertainty were the dominant cause of partner violence (Hypothesis 1). We then considered whether activities that involve controlling the partner's behavior are related to physical aggression in males but not in females (Hypothesis 2), which again would be predicted if paternity uncertainty were the main cause of partner violence. Continual mate guarding would be extremely time consuming and seriously impede the guarder's ability to engage in other important activities, such as acquiring food. Therefore one would expect humans to have evolved sensitivity to cues indicating when a mate needs more or less frequent guarding (Buss, 1988). Such cues may be external to the relationship, such as the presence of rivals, or from within, such as women's fecundity, and men's and women's mate value and genetic capital. We now elaborate on two of these cues to be investigated here Women's reproductive value and fecundity Men can only be cuckolded when their partner is fecund and therefore we would expect cues to female fecundity to affect men's and women's mate-guarding behavior. Men who prevent their fertile partners having extra-pair copulations eliminate the chance of being cuckolded. A woman's reproductive value declines from young adulthood into middle age (Fisher, 1930; Wilson & Daly, 1993). Therefore, researchers have used women's age as a proxy measure for reproductive value and have generally found that this is associated with more mate retention tactics (Buss & Shackelford, 1997) and more spousal violence (Figueredo & McCloskey, 1993; Peters et al., 2002; Shackelford, Buss, & Peters, 2000; Wilson & Daly, 1993) being directed to her, although Figueredo et al. (2001) found no association between a women's age and her partner's aggression. Fecundity varies not only with age but also with pregnancy and lactation, and the time since the birth of the last child, since a lapse of several years signals reduced fertility where no contraceptives are used (Flinn, 1988). Therefore, in studying whether a women's fecundity affects inter-pair conflict and aggression in a Caribbean village, Flinn (1988) operationalized fecundity in terms of the woman being under 40 years old, and either having no children or having children over 12 but under 48 months; the alternative category ( infecund ) consisted of women who were pregnant or had an infant under 12 months old or were over 40 years of age. Flinn found that men spent more time interacting with, and showed more aggressive behavior toward, fecund than infecund partners. From this study, we derived Hypothesis 3, that men with fecund partners will use more direct mate-guarding behavior than will men with nonfecund partners. There is less evidence on how a woman's fecundity might influence her own mate-guarding behavior. Although Flinn (1988) found no difference between fecund and nonfecund women's aggression towards their partners, Buss and Shackelford (1997) did find a weak but significant negative association between a wife's mate retention tactics and her age, suggesting that more fecund women guarded more than less fecund women. From this, we derived the tentative Hypothesis 4, that fecund women will use more mate guarding tactics than women who are not fecund Mate value Symons (1995) defined mate value as the degree to which each [mate] would promote the reproductive success of [the other] who mated with them (p. 87). It comprises many different facets, such as physical attractiveness, personality and resources. Although these differ in their importance for men and women (Buss, 1989), the sexes seek mates that are in many ways similar to themselves (e.g., Botwin, Buss, & Shackelford, 1997). Figueredo and McCloskey (1993) reasoned that violence is not a preferred mate-guarding strategy, but is more likely to be used by competitively disadvantaged males, men who are low in mate value, who are less physically attractive, less socially competent, less sexually adequate and/or financially poorer than their potential rivals. Figueredo et al. (2001) suggested that, unlike competitively disadvantaged women, who could engage in short-term mating with higher quality men, such an option is not available to competitively disadvantaged men, as women gain nothing from copulating with them. Such men will therefore be at the greatest risk of cuckoldry and hence need to use more frequent mate guarding behavior. This reasoning leads to Hypothesis 5, that men (but not women) with lower mate value will use more direct, aggressive, forms of mate-guarding behavior than those with higher mate value. A further possibility concerns the mate value of the partner. The higher this is, the more attention they will attract from other potential mates. There is evidence that this may be the case for women: Haselton and Gangestad (2006) found that more attractive women were mate guarded more than less attractive women, based on daily reports. The present prediction is that members of both sexes who perceive their partners to have higher absolute mate value will use more mate guarding (Hypothesis 6). Finally, we investigated the possibility that the sexes differed in the characteristics of their partners that were more associated with mate guarding, men guarding more

3 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) when their partners were more physically attractive, and women guarding more when their partners were higher in career prospects (Hypothesis 7). This is based on many studies indicating that the sexes are differentially attracted by these two features (e.g., Buss, 1989), and specifically on the finding that men used more mate retention tactics when they perceived their wives to be physically attractive, whereas wives' used three specific mate retention tactics more when their husbands had a higher income (Buss & Shackelford, 1997) The present analysis To test the hypotheses set out above, direct mate guarding was assessed through a range of controlling behavior, physical aggression and injuries sustained, within heterosexual relationships. Controlling behavior was used as a measure of mate guarding for several reasons. First, as the acts we measured are nonviolent, they allow an assessment of mate guarding that should not be confounded by the vulnerability of their partner. Second, the use of partner violence is socially sanctioned in some social groups more than others, whereas controlling behavior is less likely to vary in approval. Third, controlling behavior is pervasive (Stark, 1995) and can effectively be used to dominate another's life. Violence cannot by its very nature be used continually, without serious injury and death resulting. Fourth, unlike Buss's (1988) taxonomy, the scale used to measure controlling behavior comprised individual behavior used by one partner to directly control the behavior of the other, rather than behavior used both towards partners and rivals. We assessed mate value by asking respondents to rate themselves and their partners on six dimensions, relative to people they know. A relative, rather than absolute, assessment was used, as in the sexual marketplace, it is a person's peers who are their competitors, not people who are icons of success or beauty. Figueredo and McCloskey (1993) came to the same conclusion, stating that disadvantage in sexual competition is relative to members of one's social class, rather than absolute (p. 374). In summary, this study sought to investigate several evolutionary psychological predictions in a sample of men and women not selected for high rates of partner aggression. 2. Methods 2.1. Participants Participants were recruited via an sent to students and staff at the University of Central Lancashire. Of those who responded 5% failed to complete the questionnaire and 1350 participants (399 men and 951 women) provided usable data. Their mean age was 25 years (S.D.=9.3; range years), and that of their partners 26 years (S.D.=9.7; range years) Measures Controlling behavior Controlling behavior was measured using a revised form of the Controlling Behaviors Scale (CBS: Graham-Kevan & Archer, 2003), the CBS-R (Graham-Kevan & Archer, 2005). The CBS was developed from information in the Domestic Abuse Intervention Project (DAIP: Pence & Paymar, 1993), which involved examples of controlling behavior consistently reported by both victims and perpetrators as being used by violent men against their partners. The CBS-R uses behavioral categories and does not involve any items of physical aggression. It produces an overall score and scores for each of five subscales: Economic abuse [four items: α=.45 (self) and.58 (partner)]; Coercion and threats [four items: α=.70 (self) and.72 (partner)]; Intimidation [five items: α=.62 (self) and.74 (partner)]; Emotional abuse [five items: α=.75 (self) and.81 (partner)]; and Isolation [six items: α=.84 (self) and.88 (partner)]. The α values were generally similar to, or slightly higher than, those for the original CBS (Graham- Kevan & Archer, 2003) Physical aggression A slightly modified version of the Conflict Tactics Scale (CTS: Straus, 1979) was used to measure the occurrence and frequency of physical aggression. The CTS has good internal consistency (Straus, 1990) and discriminates between different samples (Archer, 2000; Graham-Kevan & Archer, 2003; Graham-Kevan & Archer, 2003). Only the 14 aggression items were used, the first six relating to verbal aggression and the last eight to physical aggression. The response format was a five-point frequency scale (never; rarely; sometimes; often; always). The eight physical aggression items produced a physical aggression score for respondents and for their partners. Cronbach's α for selfreports were.87; for partner-reports, they were.90. Injuries were assessed by the mean of two items from Morse (1995). The first asked the frequency with which a partner had physically injured the respondent over the last year during conflicts, and the second how frequently this had required medical treatment: each was scored 0 (never) to 4(always) Other variables Length of the relationship was classified as: 1=less than 1 month; 2=1 to 6 months; 3=6 to 12 months; 4=1 to 3 years; 5=3 to 5 years; or 6=over 5 years. This was measured as previous research found it be related to attachment to romantic partners (Feeney, 2004), separation distress (Fraley & Shaver, 1998), dyadic functioning (Moore, McCabe, & Brink, 2001; Talmadge & Dabbs, 1990) and relationship conflict styles (Stith, Jester, & Bird, 1992). It was therefore included in the analysis as a covariate. Whether the respondent (for female reports) or their partner (for male reports) had been pregnant in the last 12 months, is presently pregnant, and if so how many weeks, was recorded.

4 448 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) Table 1 Means and standard deviations (S.D.) for men and women respondents, for acts of controlling behavior Men Women Mean S.D. Mean S.D. t d Economic Coercion Intimidation Emotional Isolation Total Economic Coercion Intimidation Emotional Isolation Total The top set of figures are for self-reports and the lower ones for partner reports (i.e., the men's values are derived from women's reports and the women's values from men's reports). pb.05. pb.01. Fecundity was dichotomised as follows (following Flinn, 1988): (1) relationships involving fecund women (339 men and 869 women), either where there were no children and the female partner was under 40 years old (310 men and 793 women) or where the woman had a child over 1 but under 4 years of age (29 men and 76 women); (2) relationships involving nonfecund women (91 men and 155 women), where the female partner is over 40 years old (61 men and 126 women), or is pregnant (15 men and 15 women), or the youngest child of the woman is under 1 or over 4 years old (15 men and 14 women). Since mate value was a dichotomised measure, it was used in a factorial analysis, with length of relationship (see above) as a covariate. The definition of nonfecund women corresponds to Flinn's (1988) infecundity. The accuracy of using a 4-year interval since the birth of a child would be dependent on the couple not using contraception. In the present study, contraceptive use was not recorded; however, it could be argued that it is the presence of children lower than this age that is the proximal cue to fecundity, on the basis that contraceptives are new and have not shaped humans' basic responses. Mate value was assessed from the following orientation paragraph: Compared to other people you know please rate yourself and your partner on the characteristics below, using the following scale. The response options were as follows: 0=very low, 1=low, 2=average, 3=high, 4=very high. Respondents rated themselves and their partner on six characteristics (physical attractiveness, personality, popularity, education, intelligence, and career or job prospects). The intercorrelations between self and partner characteristics indicated weak assortative mating for both men and women. The highest intercorrelations were for the same attributes and there was no evidence of women's physical attractiveness being more strongly related to men's status or resource potential than to other attributes. The items were combined to create single mate-value scores. Both men and women showed positive inter-item correlations between all six items, with none of the items reducing the overall α levels. The α values were as follows: men's self-reported mate value,.71; men's partners' mate value,.75; women's self-reported mate value,.68; women's partners' mate value, Results 3.1. Are control tactics shown more by men than by women (Hypothesis 1)? The means for the overall control tactics shown by men and women were not significantly different either for selfreports or for partner reports (Table 1), which is inconsistent with the view that controlling behavior is derived solely from male mate guarding. It indicates that control is exerted by both sexes. Table 1 shows the values for the subscales. There were significant sex differences (pb.01) in only two cases, and these were in the female direction Is controlling behavior related to physical aggression in men but not in women (Hypothesis 2)? Both men's and women's use of controlling behavior were positively correlated with their own use of physical aggression (Table 2). This was found for both self and partner reports, and is inconsistent with the view that controlling behavior is only linked to men's physical aggression (if paternity uncertainty were the sole explanation of partner violence). It indicates that in both sexes direct mate guarding is associated with physical aggression Is female fecundity associated with controlling behavior and aggression (Hypotheses 3 and 4)? Since fecundity and age are related (although not synonymous) for women, respondent's and partner's age were entered as covariates in two MANCOVAs to investigate the effects of fecundity and sex. In the first, the controlling behaviors (for respondents and partners) were the dependent variables: there was no multivariate effect of respondent's age [F(2,1398)=2.16, p=.116] or partners' age [F(2,1398)=2.43, p=.088] on controlling behaviors. In the Table 2 Correlations between controlling behavior and physical aggression Physical aggression Men's controlling behavior (n=399) Women's controlling behavior (n=951) SR PR SR PR SR PR SR=Self-reports; PR=partner reports. pb.001.

5 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) second MANOVA, respondents' and their partners' physical aggression were the dependent variables: there was no multivariate effect of respondent [F(2,1377)=0.49, p=.610] or partner's age [F(2,1377)=1.31, p=.21] on physical aggression. This indicates no effect of age independent of fecundity. To assess Hypothesis 3, that men with fecund partners will use more direct mate-guarding behavior than men with nonfecund partners, between-subjects MANCOVAs were conducted, with relationship length entered as a covariate (see Methods). Fecundity had a significant multivariate effect on men's reports of controlling behaviors [Wilks' Lambda F(5,411)=2.24, p=.05]: these were higher for men with fecund than with nonfecund partners. The length of relationship was a significant covariate [Wilks' Lambda F(5,411)=5.92, pb.0005], longer relationships being associated with more control. Men used significantly more economic [F(1,415)=5.06, p=.025], threatening [F(1,415)=7.13, p=.008] and intimidating [F(1,415)=7.20, p=.008] forms of controlling behavior when their partners were fecund. Partners' reports also showed a significant multivariate effect of fecundity on men's controlling behavior, in the same direction [Wilks' Lamba F(5,1000)=1.60, p=.019], the length of relationship again being a significant covariate [F(5,1000)=4.03, p=.017], longer relationships being associated with more control. Men used significantly more isolation [F(5,1005)=4.02, p=.045] when their partner was fecund. There were no effects of fecundity on men's (either self- or partner reported) use of aggression or infliction of injuries. The results therefore indicate that men with fecund partners used more of some forms of controlling behavior, but not more physical aggression. To assess Hypothesis 4, that fecund women will use more control tactics than women who are not fecund, between-subjects MANCOVAs were again conducted, with relationship length entered as a covariate. Fecundity had a significant multivariate effect on women's reports about their own use of controlling behavior [Wilks' Lambda F(5,1000)=3.73, p=.018]: fecund women used more controlling behavior than nonfecund women. The length of relationship was again a significant covariate [F(5,1000)=2.86, p=.014], longer relationships being associated with more control. Univariate analysis showed that fecund women used more emotional [F(1,1006)=2.55, p=.007] and isolation [F(1,1006)=15.38, pb.0005] controlling behavior. Men's reports about women's use of controlling behavior showed no significant multivariate effect of fecundity, although fecund women were reported to have used more emotional control than nonfecund women [F(1,1006)=4.83, p=.028]. There was no multivariate effect of fecundity on women's use of aggression, although fecund women did inflict significantly more injuries than nonfecund women [F(1,1006)=4.41, p=.036]. Overall, these results supported the hypothesis that fecund women use more controlling and aggressive mate guarding tactics than nonfecund women, in that they used more isolation and emotional control, and inflicted more injuries, than did nonfecund women Is mate value associated with controlling behavior and aggression (Hypotheses 5, 6 and 7)? It was predicted that men (but not women) with lower mate value would use more direct, aggressive, forms of mate-guarding behavior (Hypothesis 5). For self-reports, there was no relationship between men's mate value and their use of controlling behavior (r=.06), or physical aggression (r=.03) or inflicting injuries (r=.03). A similar lack of association was found for women. Based on partners' reports, men's mate value was negatively correlated with their controlling behavior (r=.29), their physical aggression (r=.12) and inflicting injuries (r=.12). Although small, these values were all significant (pb.01). Contrary to the prediction, there was also a relationship for women, based on partner reports: women with lower mate value showed more control (r=.25), physical aggression (r=.10) and more inflicting injuries (r=.15), all values again being significant. Therefore, reports about a partner's mate value and mate guarding supported the prediction made for men, but also indicated that it applies to women. Hypothesis 6 predicted that both men and women who have high mate-value partners would use more mate guarding than those with lower mate-value partners. There was no support for this prediction: indeed, women with higher mate-value partners used fewer acts of controlling behavior (r=.12; pb.01), and all other correlations were nonsignificant. There was no evidence either for Hypothesis 7, that men would guard more when their partners were more physically attractive, and that women would guard more when their partners had better career prospects. The physical attractiveness component of mate value was unrelated to partners' overall controlling behavior for both men (r=.01) and women (r=.03), and was unrelated to physical aggression for men (r=.01) and for women (r=.005). The career component of mate value was unrelated to partners' overall controlling behavior for men (r=.002) or for women (r=.04), and to physical aggression for men (r=.002) or for women (r=.009). These correlations were for mate value measures reported for the partner, and control or aggression measures reported for the self. For self-reported mate value measures and reports of the partner's control or aggression, values were again around 0 in all cases. The main finding from these analyses is that, based on partners' reports, both men and women with lower mate value showed more controlling behavior and, to a lesser extent, more physical aggression. There was no support for the prediction that higher mate-value partners would be guarded more than lower mate-value ones, or that this applied differentially to physical attractiveness in women and to career prospects in men.

6 450 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) Summary of analysis The preceding analyses investigated sex differences in control tactics, their association with physical aggression, and the effects of fecundity and mate value on direct mate guarding, as measured by controlling behavior and physical aggression. Overall, controlling behavior did not differ between the sexes and was associated with physical aggression in both sexes. It was found to be higher for both men and women when the woman was fecund, but there was little sign of a similar effect for physical aggression. There was a pattern suggesting that both men and women with higher mate value used lower levels of controlling behavior and aggression, but this was only found for partner reports. There was no evidence that a partner's mate value was inversely related to the use of controlling behavior or aggression. 4. Discussion The aim of the present study was to investigate several evolutionarily based hypotheses concerning sex-specific biological cues to mate guarding behavior. We first found that the overall frequency of controlling behavior was similar in both men and women. This supports a number of previous findings (e.g., Charles & Perreira, 2007; Felson & Outlaw, 2007; Stets & Pirog-Good, 1990; Walby & Allen, 2004) and is inconsistent with a view of control in relationships that is based only on male proprietary attitudes towards women, whose function is to control their partner's sexuality (e.g. Peters et al., 2002; Shackelford et al., 2005; Wilson & Daly, 1992, 1993, 1996). It is consistent with the view that both sexes have reasons to control the behavior of their partners. We also found that controlling behavior was associated with physical aggression in both sexes. Again, this supports previous findings (Claes, & Rosenthal, 1990; Dasgupta, 1999; Felson & Outlaw, 2007; Follingstad, Rutledge, Berg, Hause, & Polek, 1990; Stets, 1988) and is counter to an evolutionary view that the link between partner violence and control is exclusively male. Female fecundity was a significant predictor of men's controlling behavior, with more control being used when the woman was fecund than when she was not. This replicates previous findings using either age (e.g., Buss & Shackelford, 1997; Figueredo & McCloskey, 1993; Wilson & Daly, 1993) or a measure of fecundity (Flinn, 1988), indicating that the effect is a robust one, given the diverse range of measures and designs involved. Men's physical aggression was not related to fecundity in our sample. Flinn (1988) used a measure of mate guarding that included physical aggression along with other measures, such as arguing and staring at rivals. Therefore, it is not clear whether there would have been an effect of fecundity on men's aggressive behavior had it been separated from the other measures. Buss and Shackelford (1997) measured only violence to rivals and so their findings cannot be compared to the present ones. Figueredo and McCloskey (1993) found that a woman's age, which is related to her reproductive value, was negatively related to violence from her husband. However, their sample included women from a shelter reporting on their violent husbands. It may be the case that the association between the woman's age and her partner's violence was driven by the shelter sample data, as this group contained the youngest wives and (presumably) the highest levels of male violence. Indeed, when Figueredo et al. (2001) attempted to replicate these findings in a population not selected on the basis of the occurrence of physical aggression, they did not find the effect. There was evidence that fecund women used more isolation and emotional control, and inflicted more injuries upon their partners, than did nonfecund women. This finding is broadly consistent with Buss and Shackelford's (1997) finding that younger women guarded their mates more intensely than older women did, using more emotional manipulation and mate concealment. Flinn (1988) found no effect of fecundity on mate guarding for women, and he suggested that this could have been due to his measure of mate guarding. He suggested that women in his sample might have been discouraged from using overt mate guarding behavior because female-to-female competition was regarded as an embarrassment to men in the Caribbean. The present analysis used controlling behavior towards the partner as a measure of mate guarding, and these do not carry the same social stigma as verbal or physical aggression directed towards another woman (Campbell, 1993). Why fecund women would mate guard more than nonfecund women is not clear from an evolutionary perspective. There was a moderate correlation between women's and men's use of controlling behavior, suggesting that control, like physical aggression (Archer, 2000), tends to be mutual within relationships. Future research could seek to investigate this by comparing onesided and mutually controlling relationships. There was support for the prediction that lower mate value men would show more mate-guarding behavior, but only from their partners' reports. Such reports are generally likely to be more reliable than perpetrator self-reports (Archer, 1999). This was found for both controlling behavior and physical aggression. However, contrary to the prediction, the same relationship was found for women. Thus, our finding was consistent with that of Figueredo et al. (2001), that competitively disadvantaged men showed greater use of physical aggression to partners, but we also found that this applied to women. The finding that both sexes showed this relationship is inconsistent with the explanation Figueredo et al. put forward, and on which we based the hypothesis. They argued that competitively disadvantaged men would tend to be paired with competitively disadvantaged women, due to assortative mating. Whereas competitively disadvantaged women could always have access to extra-pair copulations from higher mate-value men seeking short-term mating

7 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) opportunities, no higher quality woman would gain anything from mating with a competitively disadvantaged man (Figueredo et al., 2001: p. 316). This could not explain why in the present study lower quality women mate guarded their partners to a greater extent than higher quality women. Attachment theory may explain this finding. Lower selfperceived attractiveness in women has been found to be related to anxious attachment styles. Anxiously attached adults have been found to report relationships in which the anxious partner feels that others are undependable and untrustworthy, and likely to abandon them (Bogaert & Sadava, 2002). We also predicted, from Buss and Shackelford (1997), that those with higher mate-value partners would use mate guarding more than those with lower matevalue partners, but we found no support for this prediction. The present study investigated several variables identified by evolutionary approaches to partner control and physical aggression. It departed from most previous evolutionary studies in that we considered women's as well as men's aggression and control, rather than assumed that control and aggression in relationships are mostly male to female. The evidence from our sample did not support this assumption and was consistent with previous studies showing that both sexes seek to control their partners' behaviour to a similar extent, and that such behavior is associated with physical aggression in both sexes. The analyses of reproductive variables provided support for some of the predictions derived from evolutionary principles. There is much scope for expanding this research in the future, to assess a wider range of activities associated with mate guarding and to investigate evolutionarily informed hypotheses about the function of female mate guarding, which has been neglected owing to the prevailing influence of the paternity uncertainty framework. Future studies could also seek to address limitations of the present study, such as the use of a sample from an industrialised Western nation, the method of data collection and the possible influence of menstrual phase. It is known that samples may differ greatly in the levels of both male and female partner violence, according to the gender relations in that culture or society (Archer, 2006): women from industrialized Western nations show less victimization and more perpetration than those from nations where women are less empowered. It is likely that the situation in low gender empowerment nations corresponds more to that in the ancestral environment (Archer, 2009). If this is the case, replication of the present study in a low gender empowerment sample would be informative, since it is possible that the full impact of paternity uncertainty is only shown in such a sample. The methods of data collection in this area of research are varied. Our own study involved both perpetrator and victim reports, using the Conflict Tactics Scale (CTS: Straus, 1979). As noted above, perpetrator reports are likely to involve more systematic underrepresentation than victim reports (Archer, 1999). Dobash et al. (1992) suggested, on the basis of a selective review of inter-partner agreement, that the CTS is inaccurate and misleading. A later meta-analysis of three types of evidence (Archer, 1999) found much greater agreement than that inferred by Dobash et al. Correlations between couples' overall scores were over r=.5, similar to those for spousal ratings of personality, values that are raised when attenuation by measurement error is considered (Moffit et al., 1997). The use of epidemiological data, in the form of crime surveys (e.g., Laroche 2005; Mirrlees- Black, Budd, Partridge, & Mayhew, 1998; Tjaden & Thoennes, 1998), involves more representative and larger samples than is possible with smaller scale studies such as the present one, but they generally have the limitation of involving fewer questions relevant to a particular topic. A further limitation of the current study is that we did not have information on the menstrual phase of the premenopausal women (and whether or not they were taking contraceptive pills). It is known that there are changes in women's behavior around midcycle and that these can be detected by men (e.g., Miller, Tybur, & Jordan, 2007). Several studies have reported more male attentiveness and proprietary behavior during the fertile midcycle phase (Gangestad, Thornhill, & Garver, 2002; Haselton & Gangestad, 2006; Pillsworth & Haselton, 2006). It would therefore be interesting to assess the extent to which these findings extend to the more coercive forms of control tactics investigated in the present study. The two more recent of these studies found an interaction between mate value and menstrual phase on the partner's attentiveness: this might have had a bearing on the present null results with regard to the partner's mate value. References Archer, J. (1999). Assessment of the reliability of the Conflict Tactics Scales: A meta-analytic review. Journal of Interpersonal Violence, 14, Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin, 126, Archer, J. (2002). Sex differences in physically aggressive acts between heterosexual partners: A meta-analytic review. Aggression and Violent Behavior, 7, Archer, J. (2006). Cross-cultural differences in physical aggression between partners: A social-role analysis. Personality and Social Psychology Review, 10, Archer, J. (2009). Does sexual selection explain human sex differences in aggression? Behavioral and Brain Sciences, 32, Bogaert, A., & Sadava, S. (2002). Adult attachment and sexual behavior. Personal Relationships, 9, Botwin, M. D., Buss, D. M., & Shackelford, T. K. (1997). Personality and mate preferences: Five factors in mate selection and marital satisfaction. Journal of Personality, 65, Buss, D. (1989). Sex differences in human mate preferences: Evolutionary hypothesis tested in 37 cultures. Behavioral and Brain Sciences, 12, Buss, D. M. (1988). From vigilance to violence: Tactics of mate retention in American undergraduates. Ethology and Sociobiology, 9, Buss, D. M., & Shackelford, T. K. (1997). From vigilance to violence: Mate retention tactics in married couples. Journal of Personality and Social Psychology, 72, Campbell, A. (1993). Out of control: Men, women and aggression. HarperCollins: Pandora.

8 452 N. Graham-Kevan, J. Archer / Evolution and Human Behavior 30 (2009) Charles, P., & Perreira, K. M. (2007). Intimate partner violence during pregnancy and 1-year post-partum. Journal of Family Violence, 22, Claes, J. A., & Rosenthal, D. M. (1990). Men who batter: A study in power. Journal of Family Violence, 5, Dasgupta, S. D. (1999). Just like men? A Critical view of violence by women. In M. E. Shephard, & E. L. Pence (Eds.), Coordinating community responses to domestic violence (pp ). Thousand Oaks, CA: Sage Publications. Dobash, R. E., & Dobash, R. P. ( ). Wives: The appropriate victims of marital violence. Victimology: An International Journal, 2, Dobash, R. E., & Dobash, R. P. (1980). Violence against wives: A case against the patriarchy. London: Open Books. Dobash, R. E., & Dobash, R. P. (1998). Rethinking violence against women. Thousand Oaks, CA: Sage. Dobash, R. P., Dobash, R. E., Wilson, M., & Daly, M. (1992). The myth of sexual symmetry in marital violence. Social Problems, 39, Dutton, D. G. (2006). Rethinking domestic violence. Vancouver: UBC Press. Feeney, J. (2004). Transfer of attachment from parents to romantic partners: Effects of individual and relationship variables. Journal of Family Studies, 10, Felson, R. B. (2002). Violence and gender reexamined. Washington, DC: American Psychological Association. Felson, R. B., & Outlaw, M. C. (2007). The control motive and marital violence. Violence and Victims, 22, Figueredo, A. J., Corral-Verdugo, V., Frias-Armenta, M., Bachar, K. J., White, J., & McNeill, P. L., et al. (2001). Blood, solidarity, status, and honor: The sexual balance of power and spousal abuse in Sonora, Mexico. Evolution and Human Behavior, 22, Figueredo, A. J., & McCloskey, L. A. (1993). Sex, money and paternity: The evolutionary psychology of domestic violence. Ethology and Sociobiology, 14, Fisher, R. A. (1930). The Genetical Theory of Natural Selection. Oxford: Clarendon. Flinn, M. V. (1988). Mate guarding in a Caribbean village. Ethology and Sociobiology, 9, Follingstad, D. R., Rutledge, L. L., Berg, B., Hause, E. S., & Polek, D. S. (1990). The role of emotional abuse in physically abusive relationships. Journal of Family Violence, 5, Fraley, R., & Shaver, P. (1998). Airport separations: A naturalistic study of adult attachment dynamics in separating couples. Journal of Personality and Social Psychology, 75, Gangestad, S. E., Thornhill, R., & Garver, C. E. (2002). Changes in women's sexual interests and their partners' mate retention tactics across the menstrual cycle: Evidence for shifting conflicts of interest. Proceedings of the Royal Society of London, Series B, 269, Graham-Kevan, N., & Archer, J. (2003). Intimate terrorism and common couple violence: A test of Johnson's predictions in four British samples. Journal of Interpersonal Violence, 18, Graham-Kevan, N., & Archer, J. (2005). Investigating three explanations of women's relationship aggression. Psychology of Women Quarterly, 29, Graham-Kevan, N. G., & Archer, J. (2003). Physical aggression and control in heterosexual relationships: The effect of sampling. Violence and Victims, 18, Haselton, M., & Gangestad, S. E. (2006). Conditional expression of women's desires and men's mate guarding across the ovulatory cycle. Hormones and Behavior, 49, Laroche, D. (2005). Aspects of the context and consequences of domestic violence Situational couple violence and intimate terrorism in Canada in Quebec City: Government of Quebec. Miller, G., Tybur, J. M., & Jordan, B. D. (2007). Ovulatory cycle effects on tip earnings by lap dancers: Economic evidence of human estrus? Evolution and Human Behavior, 28, Mirrlees-Black, C., Budd, T., Partridge, S., & Mayhew, P. (1998). The 1998 British Crime Survey. London: Government Statistical Service (Home Office. Moffit, T. E., Caspi, A., Krueger, R. F., Magdol, L., Margolin, G., & Silva, P. A., et al. (1997). Do partners agree about abuse in their relationship? A psychometric evaluation of interpartner aggression. Psychological Assessment, 9, Moore, K., McCabe, M., & Brink, R. (2001). Are married couples happier in their relationships than cohabiting couples? Intimacy and relationship factors. Sexual and Relationship Therapy, 16, Morse, B. J. (1995). Beyond the conflict tactics scale: Assessing gender differences in partner violence. Violence and Victims, 10, Mullen, P. E., & Martin, J. (1994). Jealousy: A community study. British Journal of Psychiatry, 164, Parker, G. A. (1974). Courtship persistence and female-guarding as male time investment strategies. Behaviour, 48, Pence, E., & Paymar, M. (1993). Education groups of men who batter: The Duluth Model. Springer Publishing Company. Peters, J., Shackelford, T. K., & Buss, D. M. (2002). Understanding domestic violence against women: Using evolutionary psychology to extend the feminist functional analyses. Violence and Victims, 17, Pillsworth, E. G., & Haselton, M. G. (2006). Male sexual attractiveness predicts differential ovulatory shifts in female extra-pair attraction and male mate-retention. Evolution and Human Behavior, 27, Shackelford, T. K., Buss, D. M., & Peters, J. (2000). Wife killing: Risk to women as a function of age. Violence and Victims, 15, Shackelford, T. K., Goetz, A. T., Buss, D. M., Euler, H. A., & Hoier, S. (2005). When we hurt the ones we love: Predicting violence against women from men's mate retention. Personal Relationships, 12, Stark, E. (1995). Re-presenting woman battering: From battered woman syndrome to coercive control. Albany law Review, 58, Stets, J. E. (1988). Domestic violence and control. New York: Springer- Verlag. Stets, J. E., & Pirog-Good, M. A. (1990). Interpersonal control and courtship aggression. Journal of Social and Personal Relationships, 7, Stith, S., Jester, S., & Bird, G. (1992). A typology of college students who use violence in their dating relationships. Journal of College Student Development, 33, Straus, M. A. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics (CT) Scales. Journal of Marriage and the Family, 41, Straus, M. A. (1990). The Conflict Tactics Scales and its critics: An evaluation and new data on validity and reliability. In M. A. Straus, & R. J. Gelles (Eds.), Physical violence in American families: Risk factors and adaptations to violence in 8,145 families (pp ). New Brunswick: Transaction. Symons, D. (1995). Beauty is in the adaptations of the beholder: The evolutionary psychology of human female sexual attractiveness. In P. R. Abramson, & S. D. Pinkerton (Eds.), Sexual Nature, Sexual Culture (pp ). Chicago: University of Chicago Press. Talmadge, L., & Dabbs, J. (1990). Intimacy, conversational patterns, and concomitant cognitive/emotional processes in couples. Journal of Social and Clinical Psychology, 9, Tjaden, P., & Thoennes, N. (1998). Prevalence, incidence, and consequences of violence against women: Findings from the National Violence Against Women Survey. Washington, DC: U.S. Department of Justice. Walby, S., & Allen, J. (2004). Domestic violence, sexual assault and stalking: Findings from the British Crime Survey. London: Home Office Research, Development and Statistical directorate. White, G. L. (1981). Some correlates of romantic jealousy. Journal of Personality, 49, Wilson, M., & Daly, M. (1993). An evolutionary psychological perspective on male sexual proprietariness and violence towards wives. Violence and Victims, 8, Wilson, M. I., & Daly, M. (1992). The man who mistook his wife for a chattel. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind (pp ). New York: Oxford University Press. Wilson, M. I., & Daly, M. (1996). Male sexual proprietariness against wives. Current Directions in Psychological Science, 5, 2 7.