Staying alive: Evolution, culture and women's intra-sexual aggression

Keywords

Abstract

Females' tendency to place a high value on protecting their own life enhanced their reproductive success in the environment of evolutionary adaptation because infant survival depended more upon maternal (rather than paternal) care and defence. The evolved mechanism by which the costs of aggression (and other forms of risk-taking) are weighted more heavily for females may be a lower threshold for fear in situations which pose a direct threat of bodily injury. Females' concern with personal survival also has implications for sex differences in dominance hierarchies because the risks associated with their formation in nonbonded exogamous females are not offset by increased reproductive success. Hence, among females, disputes do not carry with them implications for status as they do among males, but are chiefly connected with the acquisition and defence of scarce resources. Consequently, female competition is more likely to take the form of indirect aggression or low-level direct combat than among males. Under patriarchy, men have held the power to propagate images and attributions which are favourable to the continuance of their control. Women's aggression has been viewed as a gender-incongruent aberration or dismissed as evidence of irrationality. These cultural interpretations have "enhanced" evolutionarily based sex differences by a process of imposition which stigmatises the expression of aggression by females and causes women to offer exculpatory (rather than justificatory) accounts of their own aggression.

Staying alive: Evolution, culture and women's intra-sexual aggression

(1) Human males engage in aggression more frequently than females from about the age of two onward. Childhood differences are universal (Rohner 1976; Whiting & Edwards 1973). Adult differences measured by anthropological report (Brown 1991; Ember 1981) and by criminal statistics appear to be similarly universal. Simon and Baxter (1989) obtained homicide data from 31 countries for three time periods spanning 1962 to 1980. They calculated the percentage of female arrests and found no time or country in which female rates exceeded that of males. The mean percentage of female arrests was 10.56 per cent (s.d. = 5.55).

(2) The sex differential increases with increasing seriousness of the measure of aggression. Men in the United States commit 85.53 per cent of simple assaults, 87.31 percent of aggravated assaults and 88.5 percent of murders (Kruttschnitt 1994). Women's proportionate involvement in violent crime has remained stable over the last thirty years (Kruttschnitt 1993). At sub-criminal levels, recent meta-analyses have indicated that the sex difference is greater for physical aggression than for verbal or psychological aggression (Bettencourt & Miller 1996; Eagly & Steffen 1986; Hyde 1986; Knight, Fabes & Higgins 1996) and this is in agreement with prior narrative reviews of the literature (Frodi, Macauley & Thome 1977). The magnitude of these sex differences in psychological research have remained stable from the mid-1960s (Knight et al. 1996). The only form of aggression on which girls and women exceed boys and men is on measures of indirect aggression (gossiping and ostracising)--a fact to which I will later return.

(3) There is a high correlation between rates of male and female aggression across geographical areas. Rohner (1976) reported a correlation between male and female rates of aggression of r=.88 for both children and adults across a world sample of 101 societies. The correlation over 66 nations reporting criminal assault data to Interpol is .99. In England and Wales, rates of male and female violent crime correlate r=.98 over 43 police jurisdictions. In a study of 34 police reporting districts in Massachusetts, the correlation over region for male and female aggravated and simple assault was .90 (Campbell, Muncer & Bibel, under review).

(4) There is a high correlation between rates of male and female aggression over age. In both sexes criminal violence is most likely between the ages of 14 and 24 (Campbell 1995a) with the female peak occurring approximately two years earlier than that of males in line with females earlier attainment of puberty. The correlation between the sexes over age is .89 for assault and .99 for aggravated assault based on US data and .98 for UK crime statistics. Studies of aggression in children have also noted the remarkable similarity of the age curves (Bjorkqvist, Lagerspetz & Kaukiainen 1992; Eron, Huesman, Brice, Fisher & Mermelstein 1983)

Men's violence has already received considerable scrutiny by evolutionary psychologists (Daly & Wilson 1988). In species where one sex makes a higher parental investment than the other, the high investing sex is a resource for which the opposite sex competes. In humans as in many other species, females make a higher parental investment than do males. The reproductive strategies of women compared to men can be seen as reflecting concern with "quality versus quantity". Females reproductive success is constrained by the long period of gestation and lactation (and resources necessary to sustain these) required for each offspring, while males' success is constrained by the number of partners they can inseminate. In line with this, evidence suggests that humans have a pre-history of mild polygyny i.e. men tended to seek mating opportunities with multiple females. This is apparent in our universal sexual dimorphism, earlier male senescence and death, earlier female sexual maturity, longer male reproductive career, relatively large male sexual organs and a higher preferred rate of copulation by men (Daly & Wilson 1988; Mitani, Gros-Louis & Richards 1996; Oliver & Hyde 1993; Symons 1979). The fact that some dominant males will monopolise more than their fair share of females means that other males will face "reproductive death". Thus competition among males is high because the associated pay-offs in terms of reproductive success are high also. Dominance and resource-holding are linked among males. Both are part of the same evolved system used by males to attract females. As Wilson and Daly (1985: 60) note "... males are in competition for those resources, including feeding territories, nest sites and more intangible "resources" like political influence and social status that can be converted into reproductive opportunity, whether because they are directly attractive to females or because they help quell rival males". Daly and Wilson (1994) in a series of studies have argued that the higher rate of aggression in men is indicative of the crucial importance of status competition to male reproductive success. Males engage in dangerous confrontations and other forms of risky behaviour where the reward is an elevation in status in the local community-- the "young male syndrome" (Wilson & Daly 1985). Using homicide as their assay of aggression, they have attended particularly to apparently "trivial" altercations resulting in death among men and concluded that these incidents are principally about maintaining "face" when challenged by another male.

Wilson and Daly (1985: 88) note that "Women compete, too, and may even kill one another in the process but their lesser fitness variance means that they have little to gain, and at least something to lose by dangerous tactics". The following argument pursues this observation in more detail. While male aggression has been described as a "higher stakes, higher risk" enterprise (Daly & Wilson 1983: 92), I wish to emphasise that lower rates of aggression by women reflect not just the absence of masculine risk-taking but are part of a positive female adaptation driven by the critical importance of the mother's survival for her own reproductive success. In addition, using a co-evolutionary perspective, I will consider how patriarchal culture has distorted our understanding of women's aggression.

I will begin by describing the greater importance of personal survival for female inclusive fitness which render the costs of direct aggression greater for females than for males. I will argue that these differential cost-benefit outcomes were and are mediated by differences in fear of physical harm in men and women. The greater need for women to avoid serious physical injury has implications for the formation of dominance hierarchies. Though achievement of high rank may confer advantages in terms of resource access, the establishment of hierarchies and the pursuit of dominance within them is more costly for females than for males. Primate and human research bearing on sex differences in dominance hierarchies and status-seeking will be reviewed. Although females show less concern with status than do males, they must be concerned with securing resources. However such disputes are likely to be low-risk or indirect in form. Psychological and criminological studies pertinent to this argument will be reviewed. I will then consider how patriarchy awarded men the power to create and disseminate cultural images which enhanced a male monopoly on aggression by characterising female aggression as either an abnormal attempt to mimic male aggression or as evidence of psychopathology.

Evolutionary factors affecting form of female aggression.

1.1 Maternal investment and the need for personal survival.

In this section I will argue that the mother's presence is more critical to her offspring's survival and hence to her reproductive success than is the father's. This point is important because it forms the basis for the argument that females should be more concerned with staying alive than are men and this in turn accounts for their low-risk and indirect strategies of dispute resolution.

Lower fitness variance among females means that more females than males can expect to produce at least one child in their lifetime. Though a successful male can always out-reproduce a successful female, the principle difference between the sexes is the relative certainty of at least minimal lifetime reproduction. Furthermore, a female can be sure that her child will carry on average half of her genes. In humans (and in many birds and some monkeys) concealed ovulation means that a male can never be certain of his paternity. Contemporary data suggest that between 7 and 14 cent of infants are not fathered by the mother's partner and a woman is also more likely to conceive during an extra-marital affair than with her husband (Bellis & Baker 1990). A mother, unlike a father, can be certain that any sacrifices she makes to ensure her offspring's survival are not in vain from the point of view of genetic replication.

This is fortunate since women make a substantially greater contribution to parental investment than do males. At a purely biological level, male ejaculate is produced cheaply, quickly and constantly. Women however require twenty-eight days to move through a single reproductive cycle and if conception takes place contribute nine months to gestation. In hunter-gatherer societies (analogous to the circumstances in which 99 per cent of human history took place), lactation continued for up to four years. During this time, the mother probably carried the infant with her on gathering expeditions at a substantial cost in calories to her but with the benefits of continuous nutrition for the infant and contraceptive amenorrhea for the mother (Lee 1979). Hence the relative time costs to males and females respectively are measured in hours versus years and the metabolic costs are equally disparate.

Her motivation to ensure her child's survival is matched by the greater dependency of the infant on the mother rather than the father. The large endocranial size of our species together with a narrowing of the birth canal caused by bipedalism meant that infants had to be born relatively immature with a correspondingly longer period of dependency (Foley 1996; Lancaster & Lancaster 1983; Peccei 1995). In all societies, woman take primary responsibility for infant care (Ember 1981) and though this is doubtless a product of lactation, it extends beyond weaning and continues with solid food provision by the mother (Lancaster & Lancaster 1983). The primary attachment by infants is to the mother rather than the father (Kotelchuck 1976) and infants show greater fear of strange males than of strange females (Greenberg, Hillman & Grice 1977). There are no known cultures where mothers voluntarily abandon their children at the rate at which fathers do (Browne 1995) and mothers experience greater grief than fathers at the loss of a child (Zeanah 1989). The mother is the principle carer and protector of the infant.

A substantial threat to their infants' lives comes not only from predators and natural hazards but from immigrant males entering the group and from male "raiding" parties from neighbouring communities (Hrdy 1979; Wrangham & Peterson 1996). In many species of primate, males use infanticide as a means of bringing the mother back into oestrous, fathering their own offspring and dispatching those of rival males (Hausfater & Hrdy 1984). Females are highly sensitised to this threat and females' typical low-level aggression alters dramatically when a mother is faced with a strange and possibly infanticidal male. Smuts (1987, p. 407) has noted "This extreme female vigilance may explain why males often avoid infants and why they sometimes exhibit fearful responses when an infant approaches". The risk to infants from unrelated adult males has also been noted in humans. Stepchildren are sixty-five times more likely to be murdered than are children living with their two natural parents (Daly & Wilson 1988). In Britain fifty-two per cent of babies killed in the home were murdered by a step-father (Watson 1995). The mother plays a primary role in protecting the infant from such attacks.

The greater importance of the mother to the child's survival may also explain the phenomenon of menopause which is unique to humans. Let us suppose that in the environment of evolutionary adaptation, women gave birth to their fifth and last child at age 35. Women who died immediately thereafter reduced the last child's probability of survival relative to women who lived on for another five years (Peccei 1995). This would result in a small but consistent advantage to women who bore the genes for non-reproductive life after the birth of their last child and to the extent that this longer life had a genetic basis would pass it to her female children. The fact that men do not experience reproductive menopause and can continue to father children into old age suggests that their continued survival was less critical to their infants survival.

Male reproductive strategy also has direct implications for maternal care of infants. Polygyny is associated with earlier death among males. In part, this results from the dangers of male-male competition and from the generalised risky behaviour of young men. In England and Wales, the male-to-female ratio for deaths from external causes is at its most extreme at ages 15 to 24 years (3.93:1) compared to a more modest ratio of 1.72:1 for the childhood period of 5 to 14 years (Office of Population Censuses and Surveys, 1995). However males who survive past their peak reproductive years still die earlier than do females. Earlier male death is thought to result from the trade-off between survival and reproductive success in males. Testosterone, which energises male competitive behaviour, is associated with lower disease immunity and with higher rates of degenerative diseases (Folstad & Karter 1992). High rates of sexual activity in youth have been selected for despite the loss of longevity, again supporting the lesser importance of the father to infant survival. When males die, it is the female who must shoulder the full burden of infant care.

Polygyny affects not only a male's likelihood of death but his likelihood of desertion. In monogamous societies, men seek more pre- and extra-marital affairs than do women (Daly & Wilson 1988; Fisher 1993). Men's preference for youth and physical beauty in sexual partners means that as their wives age, younger women become increasingly attractive. After divorce, men are more likely to re-marry younger partners and to produce further children in their second marriages. Male desertion adds to the burden of parental investment taken on by the female. Women are far more likely than men to seek and gain custody of children in a divorce and state benefit for single parents are overwhelmingly paid to women.

Data from studies of survivorship among orphans support the thesis that maternal survival is more important than paternal survival. Among the forest-dwelling Ache of Paraguay (Hill & Hurtado 1996), maternal death increases age-specific child mortality rate by a factor of five compared to a threefold increase when the father dies. (Paternal death and parental divorce are about equally hazardous to infants). Where the mother died in the first year of the infant's life, the mortality rate was 100 per cent. In a study of seventeenth to nineteenth century Ostfriesland in Germany, Voland (1988) reports that maternal death increases the risk of dying before the fifteenth birthday by a factor of 1.4 compared to a paternal death. Again, the discrepancy in survivorship is most marked in the first year of life where maternal death doubles the risk of infant mortality relative to paternal death.

In summary, women have long faced the same evolutionary problem with regard to inclusive fitness. Biological factors, infant dependence and male reproductive strategies mean that the mother is more critical to the offspring's survival than is the father. If a mother wants her children to survive, then she must be equally concerned with her own survival. Because of this, we should expect that women would have evolved a psychology in which the costs of physical danger would have been weighted higher than that of a male.

1.2 Psychological mediation of personal survival.

If survival is more critical to reproductive success in females than in males, we should expect to see a lower involvement in forms of aggression that pose a risk to their bodily integrity. This can be conceptualised as assigning a higher weighting to the costs of any given agonistic encounter. For the psychologist, a key question is: By what mechanism do humans "weight" costs in an encounter? I suggest that the relevant mechanism is fear and that, given an equal degree of objective risk of harm, females will experience greater fear than will males. Fear has been invoked by others as an explanation of sex differences. Gray (1987) argues that differences in fearfulness among primates are a result of male dominance over females, thus ignoring the issue of sex differences in intra-sexual aggression. MacDonald (1995) has argued that while behavioural avoidance is generally higher in females, there may be important differences in the types of stimuli that elicit fear in the two sexes. The current argument pursues this by suggesting that females should be more fearful than males particularly when they perceive threat to their bodily integrity.

With regard to objects that do not pose a direct risk of injury, we would expect to see little evidence of sex differences. If women were more afraid of people in general, we would expect to see a greater prevalence of avoidant personality disorder in women but there are no reported sex differences in this diagnostic category (American Psychiatric Association 1994). At a non-clinical level we would expect to see sex differences in extroversion, which we do not (Costa & McCrae 1992). However when extroversion is decomposed into its contributory facets, women score higher on warmth, gregariousness, activity and positive emotions while men score higher on assertiveness and excitement seeking (Corbitt & Widiger 1995). Females are no more prone to generalised anxiety disorder (American Psychiatric Association 1994), social or school phobia than are men (Marks 1987). Nor are females more likely to fear intellectual or sensory novelty. The Openness to Experience factor of the "Big Five" (which measures curiosity, imagination and creativity) shows no sex differences (Costa & McCrae 1992). The Experience Seeking subscale of the Sensation Seeking Scale measures desire for unusual experiences that do not contain a component of physical risk (e.g. through music, art, travel and unconventional people). Zuckerman (1994) reports that in 15 out of 17 cross-cultural studies no sex differences on Experience Seeking were found and concludes "The lack of difference on ES suggests that while men are high on the more active forms of sensation seeking, women are just as open to novel experiences through the senses and lifestyle as men" (Zuckerman 1994: 101).

Phobic reactions are of particular interest since their foci are thought to correspond to specific dangers faced by humans during their evolution. Marks & Nesse (1997, p.64) argue that such "evolved defenses often seem over-responsive...because repeated false alarms may cost less than a single failure to respond when danger is present". We should expect that such hyper-vigilance would be higher in females than males if the present argument is correct. Females are indeed more prone than males to panic disorder with agoraphobia (American Psychiatric Association 1994), and to phobias about animals (including dogs, snakes, insects and mice), blood and injury, and medical / dental procedures (Marks 1987). In short, women have a greater likelihood of over-reaction to open spaces (where predation was more likely), closed spaces (with the danger of being trapped), potential predators and parasite carriers, and the sight of blood or tissue injury signalling possible death. Anxiety-sensitivity describes fear of the harmful consequences of anxiety-related sensations. Females are more fearful than males only of the physical (i.e. health) consequences of anxiety. Males showed the greatest fear on items that measures the psychological and social facets of fear (Stewart, Taylor & Baker 1997) as might be predicted from men's particular sensitivity to autonomy and status (see 1.4).

Sensation-seeking is an inverse measure of fear. Males exceed females on physically risky forms of sensation seeking (Zuckerman 1994) and these in turn correlate significantly with a variety of physically dangerous activities such as involvement in crime, dangerous sports, injury proneness and volunteering for drug experiments and hazardous army combat (Zuckerman 1994). There are very significant sex differences in mortality from traffic accidents (after controlling for miles driven) which seem to be attributable to men's more frequent speeding, tailgating and running red lights (Wilson & Daly, 1985). Rates of accidental injury on the street and in the home are higher among boys than among girls even after controlling for exposure times (Christopherson 1989). Such accidents are likely the result of greater willingness by boys to take risks involving potential physical injury (Ginsburg & Miller 1982). Serious drug use is more common among males than females with sex ratios of between 3 and 4 to 1 for amphetamine, opioid, hallucinogen, inhalant and phenocyclidine abuse (American Psychiatric Association 1994). Girls perceive the health risk posed by drugs to be more serious than do boys (Turtle, Jones & Hickman 1997). Despite the fact that men are more likely to be the victims of violent crime than are women (Bureau of Justice Statistics 1997), women report higher levels of fear of crime (LaGrange & Ferrero 1989).

Women also exhibit greater concern with their health than do men. The strongest predictor of preventive health care is gender (Harris & Guten 1979). Women rate the importance of health higher than men, know more about health issues and are more likely to track the status of their health (Umberson 1992; Waldron 1988). Women visit the doctor more often than do men (after controlling for gynaecological and obstetric visits) and more often go for preventive care. Umberson (1992), investigating why the health benefits of marriage are greater for men than for women, found that women are significantly more likely to control the health-related behaviour of their spouse than were men. The prevailing sociological interpretation--that such female solicitousness derives from "gender role socialisation" (Umberson 1992: 908) or "cultural influences" (Waldron 1988: 204)--is not of course incompatible with the present view that such roles themselves derive from evolutionary pressures.

In focusing upon higher levels of female fear in response to prospective aggression, we are in a better position to account for results from human experimental work. Eagly and Steffen (1986) and Bettencourt and Miller (1996) in meta-analyses of 127 laboratory studies found that women estimated the danger of the same aggressive encounter to be higher than did males and that sex differences in aggression were greater to the extent that the actor would be in greater danger from aggressing.

In summary, I have argued that females show higher levels of concern with survival than do males and that fear is a plausible proximal mechanism for this sex difference. Sex differences in fear appear in childhood and fear is trait-like in its stability over time (Gullone & King 1997). This suggests that women's higher level of fear does not depend upon the experience of parturition as we would expect if it is an evolved adaptation resulting from the fact that young ancestral females who engaged in high-risk aggression were less likely to survive and hence to achieve any degree of reproductive success. With regard to aggression, I emphasise that this is not an argument for greater female vulnerability--in a same-sex encounter females have no greater objective chance of injury or death than do males since both combatants are equally strong. Nor is it an argument for female non-aggression--the willingness of an animal (male or female) to engage in or to escalate an agonistic encounter is a function of anticipated rewards and costs. For a female the cost is usually higher than for a male so that only a very high payoff (such as the successful defence of her offspring's' life, see 1.1) would make lethal combat a successful strategy. The rewards for a male are usually higher than for a female because of the resource-status link. Successful combat in one sphere carries implications the another and both have implications for reproductive success (Smuts 1987).

1.3 Primates: Males, females and dominance hierarchies.

Males compete with one another for dominance and its associated resources because these enhance reproductive success. Females compete with one another for resources which can be converted to offspring and so enhance their reproductive success. (As we shall see later, one such resource may be males when females rely upon them for subsistence). The fact that women in evolutionary terms had less to gain from achieving status has implications for the extent and form of their intra-sexual aggression. In this section, I will consider evidence from primates concerning female dominance hierarchies.

Symons (1979) argued that because of polygyny, there was no reason for females to form dominance hierarchies since they conferred no advantage in terms of number or quality of copulatory partners. Primate studies have generally concurred that males do not show a strong or systematic preference for high ranking females (Loy 1971; Packer 1979; Small & Smith 1985). However female reproductive success is linked to successful competition for nutrition, spacing and safety which might be enhanced for a female high in a dominance hierarchy. Many commentators note that the relationship between dominance and reproductive success is weaker and less consistent for females than for males (de Waal 1982; Silk 1987; Wrangham 1980). A recent review by Ellis (1995) of 700 studies of this relationship in a variety of species concludes that "...establishing and maintaining dominance relationships is less consequential for females than for males. Presumably natural selection has favored the use of dominance as a vehicle for enhancing RS for males more than for females" (Ellis 1995: 290). The advantage of a dominance hierarchy is that, once formed, there is a reduction in the frequency of conflict within the group (de Waal 1982; Ellis 1995). However for females, the costs that must be incurred in the formation of an achieved dominance hierarchy are rarely outweighed by the payoffs. The price of injury to the female and to her current and yet-to-be-conceived children is high (Smuts 1987) relative to the modest advantages in terms of reproductive success.

How then can we explain the existence of female dominance hierarchies in a number of cercopithecines--chiefly macaques, savannah and gelada baboons and vervets (see Ellis 1995)? In these female-bonded species, where feeding patches are monopolisable and females remain in their natal group (Mitchell, Boinski & van Schaik 1991), dominance does not depend upon risky combat but is inherited from the mother (van Hoof & van Schaik 1992; Wrangham 1980). In matrilineal hierarchies "Dominance status is transmitted across generations from a mother to her daughters, with the result that female kin rank adjacently in the dominance hierarchy rather than distributing themselves according to individual attributes such as their age and size" (Chapais 1992: 29). Any attempt to challenge the hierarchy is risky and only rarely are subordinate females willing to take the risk of an outright attack on a high-ranking female. This hierarchical inertia is all the more remarkable because the pay-offs are exceptionally large--once a rank reversal occurs it can be maintained not just for the lifetime of the female and her kin but for several generations. Despite the high payoff for success, females are generally conservative, as I would predict given their lower fitness variance and high degree of parental investment. Matrilines are extremely stable and highly resistant to change (Hrdy 1981). Walters (1980) found that a female's rank at the time of her birth correctly predicted her adult rank in 97 per cent of cases and in a 400 day study of yellow baboons, Hausfater (1975) found not a single instance of an agonistically-induced change of status among females. When rank reversals do occur, they are often the result of maternal death (rather than attack frequency) which has a direct effect on the daughter's rank and an indirect effect via reduction in the daughter's kin alliance size (Mori, Watanabe & Yamaguchi 1989). Chapais' (1992) experimental examinations of female rank relations indicated that only in situations where the possibility of injury was extremely low was a challenge made--a finding he referred to as a "minimal risk strategy" (Chapais 1992: 44).

Apart from rare attempts at revolutionary take-over, the vast majority of female-female aggression is "low key and chronic" (Smuts 1987:402) and involves "mild bickering" (Walters & Seyfarth 1987:308). Hrdy (1981: 106) notes "...females rarely inflict serious damage on one another in their quarrels". Seyfarth (1976) reported that among baboons the ratio of approach-retreat interactions to bouts of overt aggression was 20:1 for females and 1.7:1 for males. Thus, even in species which have a female dominance hierarchy, intra-female aggression tends to be "low risk" because dominance is inherited not achieved, take-overs are rarely attempted and most mundane resource disputes involve threat and withdrawal rather than injury.

In other species of primates, matrilineal dominance hierarchies are not found because females leave their natal group. Among these non-bonded females, there is little evidence of agonistically-achieved linear hierarchies. One such species is the chimpanzee, our closest phylogenetic relative. Like humans, female chimps make a huge maternal investment--gestation is eight months and lactation lasts five years. Offspring whose mothers die before they are about eight years of age often die even though older siblings may attempt to care for them. The mother-child dyad is the primary social unit and adult females spend approximately 60 per cent of their time foraging alone with their infant (Pusey, Williams & Goodall 1997). In wild populations, adult females do not form strong bonds and rarely support one another. They rarely rest in close proximity or engage in mutual grooming. Though females can be assigned to high, medium and low status by aggregating data on submissive vocalisation over several years (Pusey, Williams & Goodall 1997), dominance relations are "weakly-developed (egalitarian) and unstable" (van Hooff & van Schaik 1992: 362) and "dominance behaviour...is uncommon and is never observed between some dyads" (Pusey, Williams & Goodall (1997:829). Direct agonistic conflicts are rare. Studies of captive chimpanzees likewise conclude that there is little evidence of linear hierarchy. De Waal (1989:53) describes female relationships thus: "By contrast, the female hierarchy is rather vague. Since status communication is rare among females, it is difficult and almost useless to assign them positions on a vertical scale. The same is true of feral chimpanzee females". He notes that he witnessed not a single instance of female status ritual over a six year period. However, unlike feral chimps, these females form close affiliative bonds, based on personal preferences and shared history, which are extremely stable. Bonobo females also form close bonds with little very low rates of female-female aggression. Of 259 aggressive episodes witnessed by Kano (1992), only 3.5 per cent occurred between females.

Male chimpanzees remain in their natal groups and are male-bonded. Because they are genetically related we might expect to see, as we do in female-bonded species, strong kin alliance formation. Male chimps do show patterns of meat-sharing, association and grooming that suggest a less competitive relationship with other males. However, as van Hooff and van Schaik (1992: 367) conclude "Sacrificing a fertilisation has far more serious consequences than sacrificing a morsel of food". Males--even when related--continue to strive for dominance in the group and hence for individual reproductive success. They are willing to incur high risks in their competition for dominance, unlike the "minimal risk" strategy of female bonded species.

I suggest that it is the absence of a relationship between dominance and resources among females which best accounts for the pattern of female aggression in non-female-bonded species such as chimps and humans (see Mitchell, Boinski & van Schaik 1991). When food resources are in high demand (because of their scarcity or spacing), females may be forced to compete. But that competition is likely to be low-key or indirect to the extent that the costs of injury are great and the reward of success limited. I turn now to a consideration of the human literature on sex differences in hierarchical organisation and status-seeking.

1.4 Humans: Men, women and status.

Anthropological surveys of traditional human societies indicate that, like apes, we have generally favoured patrilocal residence which entails female transfer and loss of female kin bonding (Ember 1978; Foley 1987; Leakey & Lewin 1979; Murdock 1967; Rodseth, Wrangham, Harrigan & Smuts 1991). Van Hooff and van Schaik's (1992: 363) conclude that "...we consider the low rates of aggression and agonistic support and the weakly expressed dominance hierarchy as diagnostic of non-FB (female bonded) groups". Taken together, these facts suggest that women should be expected to show less evidence of dominance hierarchies than do men.

Studies of children's social organisation suggests that dominance is more central to boys than to girls (Archer 1993; Browne 1995; Geary 1996; Maccoby 1990; Maccoby & Jacklin 1987; Maltz & Borker 1982; Savin-Williams, 1980; Thorne 1994). Boys tend to play in larger groups involving rough-and-tumble play and zero-sum games (Lever 1978). When given a choice boys choose to compete rather than co-operate while girls show the opposite pattern (Ahlgren 1983; Boehnke, Silbereisen, Eisenberg, Teykowski & Palmonari 1989; Moely, Skarin & Weil 1979). When girls compete too vigorously against their peers, they are likely to be rejected by them (Hughes 1988). Conflict exchanges among boys are means of displaying verbal skills and maintaining status hierarchies (Goodwin 1982; Kochman 1983; Labov 1972; Maltz & Borker 1982). Influence attempts by boys involve giving direct commands while girls are more likely to use polite suggestion. The challenging and competitive style of boys is manifest in the dominance hierarchies which they form (Maccoby 1988; Omark & Edelman 1975; Savin-Williams 1977). Girls are more concerned with developing shared norms and cohesion within the group (Eder & Sandford 1988) and more frequently resolve conflict through discussion than do boys (Eder 1990; Maltz & Borker 1982). Collaborative interchanges are more common in girls' groups while domineering exchanges are more common in boys' groups (Leaper 1991). With regard to status, Goodwin (1990) found that girls criticised and rejected other girls whose behaviour suggested that they felt themselves superior to other group members. Savin-Williams (1980) concluded that, "In comparison to their male counterparts, female adolescents are considerably less likely to form stable and consistent groups. Groups that do form are likely to be cliquish (exclusive, intimate, intense) and small (usually pairs or threesomes). If a group's structure can be ascertained, then it is likely to be less structured than male adolescent groupings" (Savin-Williams 1980: 361).

Among adults, men score higher on traits associated with competition for status. Many studies (see Hoyenga & Hoyenga 1993) using the Bem Sex Role Inventory (Bem 1974) and on the Personal Attributes Questionnaire (Spence, Helmreich & Stapp 1974) confirm that men score higher on the masculinity / agency scale which includes competitiveness, autonomy and dominance and lower on femininity / communion which is composed of items such as warm, sympathetic and compassionate. Wiggins, using the related scales of dominance and nurturance, corroborates this sex differences (Wiggins & Holzmuller 1978, 1981). Williams and Best (1990) in a cross-national study of fourteen countries found that men were rated higher on adjectives such as ambitious, dominant and hostile. Feingold (1994) using national norms from standard personality inventories and found that men were significantly higher than women on assertiveness and lower on trust, tender-mindedness and gregariousness. This effect was invariant over age, educational level and nationality.

Male interest in rank is also apparent in the experimental literature on leadership. Men emerge as leaders more often than women where the focus is task leadership (problem solving) assessed by amount of task contribution and by acknowledgement of leadership by participants and observers (Eagly & Karau 1991). Men more often use an autocratic leadership style (i.e. discouragement of subordinates from participation in decision making) than do women (Eagly & Johnson 1990) and this is especially true where subordinates are male. When the style of leadership is autocratic, men receive more favourable ratings than do women in terms of the group's satisfaction with the leader and their competence (Eagly, Makhijani & Klonsky 1992). Devaluation of female leaders is most likely to occur where subordinates are other women or where their sex is not specified. As Eagly and Karau (1991) note these findings accord well with everyday stereotypes about gender. Men's leadership style focuses upon solving an immediate problem with much less concern for the social harmony of the group. Indeed men's willingness to behave autocratically may be taken as a measure of their desire for dominance at the expense of social bonding. (This is not to deny that men forge alliances with one another but these tactical instances of mutual cooperation are ultimately aimed at achieving dominance over others. Men more than women endorse political stances likely to accentuate rather than equalise differences between individuals and groups, see Pratto, Stallworth & Sidanius 1997.) Women's contribution to the group is more likely to be social-emotional e.g. showing solidarity and expressing agreement (Eagly & Karau 1991) and when they assume leadership roles they are more likely to employ a democratic style which downplays their personal status. Sociolinguists using naturalistic observation in the workplace have also found that women dislike female leaders who employ an authoritarian leadership style and that the most successful women managers are those who report that they avoid behaving like authority figures (Aries 1976; Statham 1987; Tannen 1996). In line with this, women are better at encoding happiness while men are better encoders of anger (Coates & Feldman 1996).

In summary evidence from both children and adults suggests that females are less competitive than males, show less evidence of hierarchical organisation, are less interested in achieving leadership within the group and are more concerned with maintaining relationships of mutuality and reciprocity. Although females are relatively indifferent to status, they are willing to compete directly and indirectly for resources as we shall see in the next sections.

1.5 Gender and aggression: Direct and indirect.

In this section I will consider psychological studies of the development of sex differences with special attention to the social meaning and form of aggression. Boys' aggression becomes increasingly motivated by issues of social status and self-esteem while girls aggression, being principally concerned with resource acquisition but not status, is more likely to take less physically dangerous and more covert forms.

At pre-school ages, the majority of disputes are about resources--access to or guarding of a toy or territory (Hartup 1974; Hay & Ross 1982). Reduction of the amount of available play equipment increases aggression, as does the addition of one or two new and desirable toys (Smith 1974 a,b). Two-year-old children show no sex differences in this form of resource oriented aggression (Coie & Dodge 1998; Cummings, Hollenbeck, Iannotti, Radke-Yarrow & Zahn-Waxler 1986). Hartup (1974) examined four to seven year old children, focusing upon the distinction between instrumental and hostile aggression. Instrumental aggression (attempts to "retrieve an object, territory or privilege", p. 338) corresponds to what I call resource disputes. Hostile aggression (responses to "frustrations which involve ego threats or threats to one's self esteem" p. 338) corresponds to what I call status-oriented aggression. Boys were significantly higher than girls on hostile but not on instrumental aggression. Boys express this status-oriented aggression almost exclusively to other boys and boys are significantly more likely to retaliate against such an attack (Barrett 1979; Parke & Slaby 1983). Physical prowess seems to be an important component of standing in boys' dominance hierarchy at younger ages (Geary 1996; Parke & Slaby 1983). Teacher-rated aggression of third and fourth grade boys, but not girls, shows a significantly positive relationship with self-concept (Feshbach & Feshbach 1986). Among boys but not girls, both aggression and altruism are positively correlated with positive affect and emotional expressiveness with peers (Cummings, Hollenbeck, Iannotti, Radke-Yarrow & Zahn-Waxler 1986). Rough-and-tumble play among popular children is positively associated with interpersonal problem-solving ability and shows a negative relationship with teacher-rated antisocial behaviour (Pellegrini 1988). In summary, among boys moderate degrees of aggressive reactivity to personal challenge seem to be associated with social standing and self-esteem.

It is important to emphasise that there is no expectation that girls will not struggle for resources. Charlesworth (1996) employed a paradigm where four children had to co-operate in order to watch a movie on a small screen. One child could watch provided that two other children agreed to operate the controls. This relegated the fourth child to a bystander position. In cross-sex groups, boys dominated girls. But in same-sex groups, boys and girls did not differ in the total amount of viewing time. However boys used significantly more physical behaviour and girls used significantly more verbal behaviour. Charlesworth concludes that the data confirm his prediction that "sex would not affect the ability to compete for resources". Though girls can and do compete when necessary, girls prefer cooperation (while boys prefer competition) and among adults men value competition more than women (see Hoyenga & Hoyenga 1993). To the extent that girls require resources they will struggle for them. But to the extent that girls demonstrate an evolved predisposition to attend to their own survival, they will choose low-risk means of doing so.

"Low-risk" may refer to variations in the severity of the attack and I have already noted that sex differences are greater for physical than for verbal aggression among both adults and children (Eagly & Steffen 1986; Hyde 1986). "Low-risk" may also refer to a preference for indirect rather than direct means of contest. Indirect aggression is likely to be of especial relevance to sex differences because, if boys are concerned with aggression as a means of achieving status (as well as securing resources), then status must be publicly demonstrated or, at the very least, must be publicly attributable to them. Indirect aggression by its nature seeks to conceal the identity of the attacker and should therefore be less appealing to boys. Indirect aggression refers to a form of social manipulation where the target is attacked circuitously and the aggressor can therefore remain unidentified. It involves acts such as shunning, stigmatising and gossiping. Girls are more likely to exclude newcomers than are boys (Feshbach 1969), to destroy their adversary's property or tell tales on them (Brodzinsky, Messer & Tew 1979) and to use tactics of ostracising and manipulating public opinion (Cairns, Cairns, Neckerman, Ferguson & Gariepy 1989). Girls are significantly higher than boys on becoming friendly with someone else as revenge, gossiping and suggesting shunning of another (Bjorkqvist, Lagerspetz & Kaukiainen 1992; Crick & Grotpeter 1995). Studies of school bullying also report that girls preferentially employ indirect strategies (Ahmad & Smith 1994). Female use of indirect aggression continues into adulthood. Bjorkqvist, Osterman and Lagerspetz (1994), investigating victimisation in the workplace, found that women more than men used indirect forms such as spreading false rumours and not speaking. The tendency for girls and women to employ indirect means is not associated with greater condemnation of the use of direct physical and verbal aggression by females (Osterman, Bjorkqvist, Lagerspetz, Kaukiainen, Huesmann & Fraczek 1994).

It is worth highlighting the analogy between humans and non-human primates with respect to indirect aggression. A number of primatologists have observed that among cercopithecines, higher status females engage in mundane harassment of lower-status females which can cause suppression of oestrus and abortion (Chapais 1992; Hrdy 1981; Smuts 1987). Such tactics diminish the reproductive success of the victim and elevate the material resources available to the victor and her offspring. They do this indirectly in the sense that such tactics involve no direct combat between the adult females and thus competitive success in this modality does not carry with it the dangers of possible death which would offset any gains achieved.

1.6 Gender and crime: Severity and form.

I have argued that women should show compete for scarce resources while showing less concern with status than do men and that their resource disputes should reflect low-risk strategies. This leads to a consideration of the criminological literature and to three predictions of sex similarity and difference.

(1) We should expect women to show similar response to acute resource shortage as do men despite the fact that the absolute level of their involvement in crime will be lower. As I have noted, correlations between men and women for violent crime over geographical regions are typically in excess of .80. Similarly high correlations are also found for total crime indices ranging between .84 (Steffensmeier 1980) and .90 (Simon & Baxter 1989). Despite men's higher absolute rate of crime, crime by men and women seems to broadly responsive to the same ecological conditions (Campbell, Muncer & Bibel, under review). These conditions are socio-economic indicators of relative resource shortage (Cohen & Machalek 1988). Although victimless crimes may be more evenly spread through the class structure, when only serious victimful crime is considered studies concur that it is significantly and negatively correlated with social class (Elliott & Huizinga 1983; Ellis 1988). The relationship between crime and income /class does not appear to be linear but rather crime is concentrated among that sector of society where resource shortage is acute as evidenced by high rates of unemployment and welfare dependency (Brownfield 1986).

(2) Sex differences should be smaller for larceny/theft than for robbery since the former reflects a low-risk strategy of resource appropriation and the latter a high risk strategy. Larceny / theft is indeed the crime in which women's involvement comes closest to that of men. It includes appropriation of others' resources without direct physical confrontation and subsumes credit-card and welfare fraud, shoplifting, writing bad cheques, non-payment of bills and surreptitious taking of others' property. From the viewpoint of the present analysis, this crime is an index of pure resource competition without any element of physical violence. Unlike violent crimes where the proportion of female involvement has remained remarkably constant between 1934 and 1979, petty property crime has increased dramatically from 7.1 percent of all arrests in the United States in 1934 to 31.6 percent in 1979--a period during which the proportion of single women in poverty grew. Steffensmeier and Cobb (1981) have linked the increase in female larceny involvement to women's increasing reliance on welfare. Theft by women is usually tied to economic need and occurs as part of their domestic responsibilities for providing for their children. It is a low-risk resource-expropriative enterprise in which there is no hostile confrontation with the "victim".

This stands in marked contrast to the crime of robbery where women's involvement has remained remarkably low--they constitute about seven percent of offenders (United States Department of Justice 1989). Robbery is the quintessential male crime, in which violence is used both to extract resources and to gain status (see Campbell 1993). Though women need resources as much, if not more, than men they do not seek the additional pay-offs of dominance over a better-resourced individual or a reputation in the community as a "hardman" (Katz 1988; Lejeune 1977). The thrill associated with moment of confrontation is an important attraction of robbery. Robbery is about exploitation and humiliation and robbers are acutely sensitive to this aspect of the crime (Allen 1978). The financial motive that spurs men to robbery is not borne out of desperation--only 18 percent of robbers say that they need the money for themselves or their families (Walsh 1986). Seventy nine per cent of robbers spend the proceeds on drugs, alcohol, clothes, cars and vacations. A core concern for robbers is the lavish and conspicuous disbursement of money which impresses and indebts others (Katz 1988). Women aggress and they steal but they rarely do both at the same time because the equation of resources and status reflects a particularly masculine logic.

(3) Women's age-assault relationship should be similar to that of men where males constitute a resource and where well-resourced males are in short supply. Despite their lower rate of absolute involvement, Female assault--like male assault--peaks in the years between 15 and 24. In a social species, other people may themselves constitute a resource. This is particularly true in the case of female mate choice--a desirable male brings with him resources necessary to the successful care of children. Women prefer males who have plentiful resources and who are willing to share them (Buss & Schmitt 1993). This preference is likely to be of particular importance when there is substantial variation among available males in resources, for example, in areas where a high proportion of males are unemployed, drug addicted or destined to spend time in prison and others (often involved in the illegal economy) have surplus income which is advertised in flashy clothes and fast cars (see Campbell 1995a). "Losers" are more than merely neutral with regard to resources--they constitute a potential liability for a woman in that their presence will reduce available welfare benefits while their lifestyle means that income destined for her or her children is likely to be spent on drugs or drink (Miller 1986; Taylor 1993). Under such circumstances, there may be competition for the best-resourced men. Campbell, Muncer and Bibel (under review) found that 73 per cent of assaults by women aged less than 24 years were against other women and that assault was positively correlated with rates of female unemployment and welfare dependency. Schuster (1983, 1985) using data from China and Zambia found that female aggression is principally driven by competition over scarce resources and often these resources are male partners. The intensity with which women are prepared to fight to secure high-status males is related to the degree of female economic and social dependence on men. Cross-cultural studies indicate that female-female aggression most often occurs between co-wives or between a woman and her husband's lover who are in competition for male resources (Burbank 1987, 1995; Lamphere 1974; Levinson 1989). Campbell (1995a) found that female adolescent disputes often centre upon three issues relating to successful mate choice: management of sexual reputation, competition over access to desirable males and protecting established relationships from take-over by rival females. Interestingly, the peak age for female assault occurs at ages 15-19 compared to the male peak at 20-24 reflecting girls earlier sexual maturity. The above account suggests that the rise in female aggression during adolescence, like that of males, is associated with mate selection. It is important to note that women's willingness to engage in direct competition for male resources is largely limited to simple assault. Female-female homicide is vanishingly rare (Daly & Wilson 1988) and women are far less likely to use weapons than are men (Burbank 1987). This is in line with the current proposal that even when resource shortage drives women to direct competition for mates, the level of aggression is substantially lower than that of men.

For males, status (and toughness where this quality is a determinant of status) is a route to desired resources, including females. Males seek public recognition of their status and Wilson and Daly (1985) have described how apparently trivial altercations can result in homicide when an opponent's acts are interpreted as a public challenge to a man's honour and when to back down is to accept that dishonour. This interpretation harmonises with criminological and ethnographic accounts of male violence (Felson 1978, Horowitz 1983; Katz 1988). For females public recognition of toughness of status is not important because high status, dominant or aggressive females are not especially preferred as mates. In resorting to intra-sexual aggression, women's aim is to secure a valuable male rather than to achieve status within her own sex.

2. Cultural factors affecting male and female representations of aggression.

2.1 Female aggression and patriarchal culture.

To this point, my argument has been concerned with the application of evolutionary theory to behavioural sex differences in aggression. I turn now to a consideration of the way in which culture has interpreted these differences, specifically how patriarchal institutions may have stigmatised women's aggression and have lead women to offer exculpatory accounts of their aggression. Culture is not considered here as an independent but complementary force "socialising" the frequency of aggressive behaviour in each sex (although such an analysis has been offered by others, see Low 1989). Rather I examine culture as ascribing different meaning and value to the same behaviour when it is performed by men and women. Failure to recognise the role of meaning in human behaviour was a much criticised aspect of sociobiology and the term "evolutionary psychology" carries with it an obligation to address not only behaviour but aggregate cultural processes which give meaning to it. It is unlikely the historically-variable content of explanations of complex social action are hard-wired (Sperber 1994). It is more probable that they are acquired from the community and this accounts for their tendency to alter over time e.g. possession by evil spirits is not generally offered as an explanation of aggression in contemporary western culture.

Co-evolutionary theorists have argued that culture is a second stream of transmission where the units of selection are memes rather than genes. While some have held that genetic and cultural transmission are parallel and independent forces, others have argued for their interaction. Lumsden and Wilson (1981) and Durham (1991) have suggested that one common form of interaction may be "enhancement" whereby the impact of socially transmitted memes is to exaggerate evolutionarily adaptive traits. In terms of sex differences, a cultural enhancement bias would tend to exaggerate the difference between men and women. In addition, Durham uses the term "imposition" to describe how those in power may enforce upon the less powerful a meme favourable to the continuation of that power. In this section I will argue that patriarchal culture has imposed a meme which exaggerates sex differences by equating female aggression with social or individual pathology.

Culture refers to socially transmitted information which may be technological (methods of termite fishing by chimpanzees, use of a personal computer by schoolchildren), social (gossip, rumour and symbolic linguistic grooming) or semiotic (the semantic interpretation and value that should be accorded to an event). It is this latter component that is of principal relevance to the present argument. Patriarchy defined as "a system of organisation in which the overwhelming number of upper positions in hierarchies are occupied by males" (Goldberg 1993, p. 14) is universal (Goldberg 1993; Low 1992; Whyte 1978). (This is not to deny that some societies are matrilineal or that many accord respect to the roles that women play. Nor is it to argue that patriarchy is desirable or unmodifiable). Smuts' (1995) offers an account of the possible evolutionary genesis of patriarchy which begins with the argument that female exogamy reduced female kin support and consequently the ability to resist male coercion. In the service of inter-group competition, males developed strong male-male alliances which could also be used to control females. Subsequent to the advent of agriculture and animal husbandry, surplus resources were controlled by a few powerful males and, since these resources were necessary for female reproductive success, female autonomy was further reduced by female-female competition. Finally, the evolution of language allowed men to create and propagate ideologies of male dominance and female subordination.

Anthropologists have documented many folk tales directly supporting patriarchy and discouraging female insubordination (Ortner & Whitehead 1981; Sanday 1981). In our culture the media and other social institutions propagate images and stereotypes that are prescriptive of the expected behaviour and demeanour of men and women, including the acceptability of aggression. Despite the evolutionary necessity for aggression in both sexes, men's aggression has been valorised while women's aggression has been treated as evidence of pathology under patriarchy. In many cultures, values of physical courage, endurance, strength, skill and honour are associated with male warfare (McCarthy 1994) and successful participation in war are associated with status and increased sexual access to women in, among others, the Sambia (Herdt 1982), the Yanomamo (Chagnon 1988), the Samburu (Spencer 1965) and the Masai (Saitoti 1986), and the Dodoth (Thomas 1965). This valorisation is not limited to inter-group hostility but is manifest in ethnographic accounts of assaults committed by young men (Athens 1980; Katz 1988; Polk 1994) and by quantitative studies of masculine values which have identified willingness to use violence as an important component of masculinity (Mosher & Sirkin 1984; Thompson & Pleck 1986).

Women's aggression occurs more rarely than does men's for the evolutionary reasons that I have outlined. It might argued that its very rarity suggests that women's aggression, more than men's, springs from pathological disturbance. If this is so, then the attributions of pathological gender-role deviance or mental illness are veridical. However relative statistical rarity in and of itself does not invariably result in a more pathological label being applied. The point can be made by considering an area of dysfunction less laden with social condemnation. Dyslexia is four times more common in males than in females, yet we do not therefore assert that men's dyslexia is normal and women's pathological. Daly and Wilson (1994, p. 263) have argued strenuously that despite the rarity of violence "dismissal of violence as pathology cannot be sustained". I argue that the critical distinguishing feature in our understanding of male and female aggression is that male aggression is valorised by a set of "warrior values" (McCarthy 1994) that render it gender-congruent even if illegal. But among women aggression is considered "doubly deviant" in the sense that it violates typical and expected sex-appropriate behaviour (DeLisi & Soundranayagam 1990; Heidensohn 1996; Williams & Best 1990) as well as criminal law.

Why women's aggression has been stigmatised more than men's can only be surmised. Feminist commentators have suggested that it may serve to maintain female dependence upon men for protection (White & Kowalski 1994), to exclude women from warfare and from consequent political power (Lerner 1986), to deflect attention from the fact that much female aggression is responsive to men's domestic abuse (White & Kowalski 1994) or to control the use of aggression by women against dependent children (Macauley 1985). Specifically, the condemnation of female aggression takes two principal forms: (1) Aggressive women are behaving like men either as a result of societal changes or personality abnormality (2) Aggression in women is the result of a permanent or temporary loss of rationality caused by mental illness or hormonal disturbance.

As an example of the former, the British press has recently reported the "arrival" of girl gangs. These girls have been labelled as "yobettes" indicating the media view that their behaviour is an attempt to emulate male "yobs". (A similar media concern with and analysis of female gang involvement has occurred in the United States, see Chesney-Lind 1993). The hypothesis that girls' assumption of masculine forms of aggressive behaviour is linked to changes in women's roles in wider society was first proposed by Adler (1975). Despite data which confirm that percentage female arrests for violence have altered little in the last forty-five years (Steffensmeier & Cobb 1981) and that female criminality is more common among women who assume and endorse traditional roles for women (Smart 1979; Weis 1976), the media continue to disseminate the view that women's participation in hitherto "masculine" arenas of activity will result in masculine forms of behaviour such as aggression. Prior to the women's movement, when social change could not be identified as a factor in girls' involvement in violence, theories dwelt upon the masculinity of female offenders but accounts were framed in terms of individual gender-role pathology (reviewed in Campbell 1992). Empirical studies using both direct ratings of masculinity-femininity as well as scales designed to tap male-typical agentic qualities (e.g. active, independent) have failed to find evidence that female delinquents have more masculine self-concepts than non-criminal controls (e.g. Norland, Wessel & Shover 1981; Thornton 1982) yet despite this, a continued attribution of masculinity remains evident in popular treatments of women offenders (Kirsta 1994; MacDonald 1991).

On other occasions, female aggression is regarded as evidence of irrationality or psychiatric disturbance. Late luteal phase or premenstrual dysphoric disorder has been used to explain women's irrationality and aggression despite three decades of unresolved doubt about its medical status, and hormonal and neurochemical correlates (Blumenthal & Nadelson 1988). Many studies fail to find any reliable change in irritability over the monthly cycle (e.g. Ruble & Brooks-Gunn 1979) and critical methodological problems have been noted in many others (Parlee 1973). PMS is listed in the appendix of the DSM-IV pending further empirical support for its existence and it is estimated that only 3 to 5 per cent of women may meet the proposed criteria for diagnosis. Self-report of premenstrual symptoms is more common among those who regard menstruation as debilitating (Brooks, Ruble & Clarke 1977), who are experiencing neurotic and emotional problems (Kashiwagi, McClure & Wetzel 1976) and who come from cultural groups which hold traditional views about women's role (Al-Issa 1982). The popular availability of the PMS diagnosis offers doctors and women a means of excusing their aggressive outbursts as a result of uncontrollable pathology. Despite the wealth of available research on the impact of testosterone in males (Archer 1991), male aggression is not routinely attributed by doctors to hormonal abnormality.

The tendency to pathologise women's violence more than that of men can be seen also in the criminal justice system. A woman appearing before a British court is twice as likely as a man to be dealt with by psychiatric rather than penal means (Allen 1987: Burns 1992). In Britain in 1995, women constituted only 3.7 percent of the prison population but 10.5 percent of those referred to special hospitals (psychiatric prisons). Given that women in the general population are more likely to receive a psychiatric diagnosis than are men, these figures could simply reflect women's generally greater vulnerability to mental disorder. However a reversal of the normal sex differential in diagnosis is evident. When assigned to a special hospital, women are nearly twice as likely to receive a diagnosis of psychopathic personality than are males despite the 3:1 male to female ratio in the community at large (American Psychiatric Association 1994). This suggests the possibility that the legal definition of "abnormally aggressive or seriously irresponsible conduct" is being applied differently to men and women. Allen's (1987) investigation of 129 London court cases concluded that "Reports on males frequently cite histories of criminal delinquency and sexual promiscuity, but hardly ever suggest that these indicate any medical abnormality. A woman who manifests these traits, however, may be labelled as a psychopath" (Allen 1987, p. 81). In the United States also, when psychiatric symptomatology and degree of violence are controlled, girls are more likely than boys to be assigned to psychiatric units rather than correctional facilities (Lewis, Shanok & Pincus 1982; see also Feinblatt & Gold 1976). If female aggression were driven by an ego-syntonic and treatment-resistant pathology such as psychopathy, we would expect to see higher recidivism rates for women then men. In fact, quite the reverse is true (Moseley & Gerould 1975; Norland & Mann 1984). Nevertheless, within mainstream prisons, more women's establishments provide psychiatric services and female prisoners are prescribed far more psychotropic drugs than are men (Camp 1974; Edwards 1986).

2.2 Sex differences in accounting for aggression.

The stigmatised nature of female aggression has implications for the form of account which men and women offer of their own aggressive actions. Accounts (or explanations) make use of the available social representations which society provides. Social representations (Moscovici 1984) are mental models of phenomenon that encompass attitudes, values, images and explanatory schema with regard to a particular topic. They are passed between people in the course of their interactions and through media communications. They serve to elaborate and explain commonly encountered experiences and so provide a short-hand to social interaction. Social representations stand as contemporary "theories" (Donald 1991) about many aspects of the physical and social world--including aggression. In western culture there are two popular social representations of aggression--instrumental and expressive which correspond to and derive from academic theories.

Expressive theories focus upon intrapsychic determinants of aggression within the individual. Expressive theories share a common concern with the build-up of noxious tension, stress or arousal and its consequent discharge through aggressive behaviour. Some emphasise the noxious motivational aspects (Berkowitz 1993; Bernard 1990; Dollard, Doob, Miller, Mowrer & Sears 1939; Lorenz 1966) while others emphasise failure of inhibition, self-control or cognitive mediation which allows the accumulated anger to be expressed in behaviour (Eysenck 1964; Gottfredson & Hirschi 1990; Zillmann 1979). Under an expressive theory, the gratifications of aggressive action are seen as primitive, evident and virtually instinctive. The goal of socialisation is to suppress and contain such behaviour.

Instrumental theories' focus upon the positive interpersonal consequences of aggression for the aggressor. Operant theorists construct their argument in the language of reinforcement and expectancies, either positive or negative (Bandura 1973; Zillmann 1979). Tedeschi, Smith and Brown (1974) suggest that the benefits of aggression inhere in the instrumental value of having one's needs met by others as well as the personal benefits of a sense of power and control. Black (1983) views violence as a means of informal justice used to remedy personal affronts and injustices by those who lack access to formal institutions of legal process. Impression management approaches share an emphasis upon aggression as a means of establishing and maintaining public "face" and private self esteem (Toch 1969; Wolfgang & Ferracuti 1967). Instrumental theories agree that aggression produces payoffs whether they be extrinsic (material or social benefits) or intrinsic (the gratification of power for its own sake).

These two forms of social representation correspond respectively to excuses and justifications which are used to both account for past acts and disclaim future acts (Antaki 1994). A justification is a form of account where the actor accepts responsibility for the act but denies the inherent blameworthiness of his actions. An excuse is where the actor accepts the blameworthiness of the act but denies full responsibility for it. It is apparent that instrumental representations are justifications ("Aggression is necessary to get through to some people") while expressive representations are excuses (" I believe aggression is always wrong. I believe that my aggression comes from losing control."). Given the taboo nature of female aggression women and the greater institutional power of men, women should be more likely to employ excuses than justifications.

In a series of studies my colleagues and I have found significant sex differences in the degree to which men and women interpret their own physical aggression as relatively instrumental versus expressive (Campbell & Muncer 1987; Campbell & Muncer 1994; Campbell, Muncer & Coyle 1992; Campbell, Muncer & Gorman 1993; Campbell, Muncer, Guy & Banim 1996). Males in general are more inclined to describe their aggression in instrumental terms, viewing it as a form of legitimate social control over others' misbehaviour. Females in general are more likely to view aggression in expressive terms, as reflecting a regretted loss of self-control caused by high levels of stress. The effect size (weighted unbiased estimate) over 1,674 subjects in twelve samples is d=.842. This significant sex difference remains when indirect forms of aggression are considered (Archer & Parker 1994) suggesting that females tendency to excuse rather than justify is independent of the form which aggression takes. Though these social representation are culturally acquired, this learning takes place as early as 8 years of age (Archer & Parker 1994). I believe that social representations are post-hoc rhetorical devices employed by socially situated individuals to explain their actions. By this I mean that the relative position of men and women in society in general and in relation to aggression in particular has an impact on the kind of account they are likely to offer. (Future research may reveal that females' tendency to excuse rather than to justify extends beyond the domain of aggression to other socially condemned and gender-role incongruent behaviours, see Archer & Parker 1994).

This emphasis upon the situated nature of accounting allows predictions about where and when men and women should show a reversal of the typical pattern. As mothers, women are charged with the control of their children and consequently are sanctioned to employ aggression legitimately as a form of discipline. A study of mothers' accounts suggests that in this area, they employ justifications when discussing routine child-control which are frequently framed in instrumental terms--"I had to show him who was the boss", "It was a battle of wills and I knew I had to win" (Campbell 1995b). Conversely, men are sometimes placed in positions of powerlessness where they must account for their actions. Court transcripts indicate that male defendants employ excuses ("I just lost control") rather than justifications, when giving evidence (Cody & McLoughlin 1988). Mandatory treatment programs for abusive husbands seek to reduce justification use ("She was asking for it") by stigmatising the target behaviour (Fagan & Browne 1994). This frequently results in a move toward excuse-giving ("But I was drunk"). A recent study which explicitly asked subjects to justify or to excuse an act of aggression demonstrated that both sexes employ endorse a significantly greater number of instrumental and expressive statements respectively than at baseline (Duckett, Lance, Pemberton, Raistrick, Campbell & Muncer, under review). Such situated but predictable variability suggests that modes of account-giving are a function of the degree of stigma associated with the act rather than essentially with the speaker's sex.

In summary, under patriarchy male aggression has been treated as a natural (if sometimes criminal) expression of male competitiveness. Men describe their involvement in aggression in justificatory terms employing an instrumental representation to warrant the use of aggression. Women's aggression has been rendered unnatural and treated as evidence of pseudo-masculinity or irrationality. Consequently women describe their involvement in aggression in exculpatory terms employing an expressive representation which denies their full responsibility for their actions.

3. Conclusions.

I have argued that women's aggression is likely to be principally concerned with scarce resources rather than with status and is likely to take low-key or indirect forms. This pattern flows from consideration of evolutionary constraints which have shaped contemporary female psychology. Culture accords meaning to actions and patriarchal control has resulted in a view of female aggression as unnatural. The taboo nature of female aggression causes women to employ exculpatory rather than justificatory accounts. Specifically women use an expressive representation of their own aggression.

The "Madonna" idealisation of women as devoid of competition or aggression has alienated women from their own nature (Miner & Longino 1987). These recent cultural inventions are being challenged by women scientists in papers like "The myth of the coy female" (Hrdy 1986) and "The myth of the nonaggressive woman" (White & Kowalski 1995). It is perhaps ironic that some of these myths have been supported by the feminist movement who have tried to insist that women's aggression is only ever responsive to male violence (see Campbell 1993). The idea that females could have survived without the motivation and ability to compete for scarce resources is, from an evolutionary viewpoint, untenable. Nonetheless it is a viewpoint that it congenial to the continuance of male protection and control of women.

References

Adler, F. (1975) Sisters in Crime. New York: McGraw Hill.

Ahlgren, A. (1983) Sex differences in the correlates of co-operative and competitive school attitudes. Developmental Psychology, 19, 881-888.

Ahmad, Y. and Smith, P.K. (1994) Bullying in schools and the issue of sex differences. In J. Archer (Ed.) Male Violence. London: Routledge.

Al-Issa, I. (1982) Gender, hormones and psychopathology. In I. Al-Issa (Ed.) Gender and Psychopathology. London: Academic Press.

Allen, H. (1987) Justice Unbalanced: Gender, Psychiatry and the Law. Milton Keynes: Open University Press.

Allen, J. (1978) Assault with a Deadly Weapon: The Autobiography of a Street Criminal. New York: McGraw Hill.

American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington DC: American Psychiatric Association.

Antaki, C. (1994) Explaining and Arguing: The Social Organisation of Accounts. London: Sage.

Archer, J. (1991) The influence of testosterone on human aggression. British Journal of Psychology, 82, 1-28.

Archer, J. (1993) Childhood gender roles: Social context and organisation. In H. McGurk (Ed.) Childhood Social Development: Contemporary Perspectives. Hillsdale, BJ: Erlbaum.

Archer, J. and Parker, S. (1994) Social representations of aggression in children. Aggressive Behavior, 20, 101-114.

Aries, E. (1976) Interaction patterns and themes of male, female and mixed groups. Small Group Behavior, 7, 7-18.

Athens, L.H. (1980) Violent Criminal Acts and Actors. London: Routledge & Kegan Paul.

Bandura, A. (1973) Aggression: A Social Learning Analysis. Englewood Cliffs: Prentice Hall.

Barrett, D.E. (1979) A naturalistic study of sex differences in children's aggression. Merrill-Palmer Quarterly, 25, 193-203.

Bellis, M.A. and Baker, R.R. (1990) Do females promote sperm competition? Data for humans. Animal Behaviour, 40, 997-999.

Bem, S.L. (1974) The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 153-162.

Berkowitz, L. (1993) Aggression: Its Causes, Consequences and Control. New York: McGraw Hill.

Bernard, T.J. (1990) Angry aggression among the truly disadvantaged. Criminology, 28, 73-95.

Bettencourt, B.A. and Miller, N. (1996) Gender differences in aggression as a function of provocation: A meta-analysis. Psychological Bulletin, 119, 422-447.

Bjorkqvist, K., Lagerspetz, K. and Kaukiainen, A. (1992) Do girls manipulate and boys fight? Developmental trends in regard to direct and indirect aggression. Aggressive Behavior, 18, 117-127.

Bjorkqvist, K., Osterman, K. and Lagerspetz, K. (1994) Sex differences in covert aggression among adults. Aggressive Behavior, 20, 27-34.

Black, D. (1983) Crime as social control. American Sociological Review, 48, 73-95.

Blumenthal, S. and Nadelson, C. (1988) Late luteal phase dysphoric disorder (premenstrual syndrome): Clinical implications. Journal of Clinical Psychiatry, 49, 469-474.

Boehnke, K., Silbereisen, R., Eisenberg, N., Teykowski, J. and Palmonari, A. (1989) Developmental patterns of prosocial motivation: A cross-national study. Journal of Cross-Cultural Psychology, 20, 219-243.

Brodzinsky, D.M., Messer, S.M. and Tew, J.D. (1979) Sex differences in children's expression and control of fantasy and overt aggression. Child Development, 50, 372-379.

Brooks, J., Ruble, D. and Clarke, A. (1977) College women's attitudes and expectations concerning menstrual-related changes. Psychosomatic Medicine, 39, 288-298.

Brown, D.E. (1991) Human Universals. New York: McGraw Hill.

Browne, K.R. (1995) Sex and temperament in modern society: A Darwinian view of the glass ceiling and the gender gap. Arizona Law Review, 37, 971-1106.

Brownfield, D. (1986) Social class and violent behaviour. Criminology, 24, 421-439.

Burbank, V.K. (1987) Female aggression in cross-cultural perspective. Behavior Science Research, 21, 70-100.

Burbank, V.K. (1995) Fighting Women: Anger and Aggression in Aboriginal Australia. Berkeley: University of California Press.

Bureau of Justice Statistics (1997) Sex Differences in Violent Victimization, 1994. Washington DC: Department of Justice.

Burns, J. (1992) Mad or just plain bad: Gender and the work of forensic clinical psychologists. In J.M. Ussher and P. Nicholson (Eds.) Gender Issues in Clinical Psychology. London: Routledge.

Buss, D. and Schmitt, D. (1993) Sexual strategies theory: An evolutionary perspective on human mating. Psychological Review, 100, 204-232.

Cairns, R.B., Cairns, B.D., Neckerman, H.J., Ferguson, L.L. and Gariepy, J.L. (1989) Growth and aggression: 1. Childhood to early adolescence. Developmental Psychology, 25, 320-330.

Camp, J. (1974) Holloway Prison. London: David and Charles.

Campbell, A. (1992) The Girls in the Gang (Second Edition). Oxford and Boston: Blackwell.

Campbell, A. (1993) Men, Women and Aggression. New York: Basic Books.

Campbell, A. (1995a) A few good men: Evolutionary psychology and female adolescent aggression. Ethology and Sociobiology, 16, 99-123.

Campbell, A. (1995b) Representations, repertoires and power: Mother-child conflict. Journal for the Theory of Social Behaviour, 25, 35-58.

Campbell, A. and Muncer, S. (1987) Models of anger and aggression in the social talk of women and men. Journal for the Theory of Social Behaviour, 17, 489-512.

Campbell, A. and Muncer. S. (1994) Sex differences in aggression: Social roles and social representations. British Journal of Social Psychology, 33, 233-240.

Campbell, A., Muncer, S. and Bibel, D. (under review) Female-female criminal assault: An evolutionary perspective.

Campbell, A., Muncer, S. and Coyle, E. (1992) Social representations of aggression as an explanation of gender differences: A preliminary study. Aggressive Behavior, 18, 1-14.

Campbell, A., Muncer, S. and Gorman, B. (1993) Sex and social representations of aggression: A communal-agentic analysis. Aggressive Behavior, 19, 125-136.

Campbell, A., Muncer, S., Guy, A. and Banim, M. (1996) Social representations of aggression: Crossing the sex barrier. European Journal of Social Psychology, 26, 135-147.

Chagnon, N.A. (1988) Life histories, blood revenge and warfare in a tribal population. Science, 239, 985-992.

Chapais, B. (1992) The role of alliances in social inheritance of rank among female primates. In A. Harcourt and F.B.M. de Waal (Eds.) Coalitions and Alliances in Humans and Other Animals. Oxford: Oxford University Press.

Charlesworth, W.R. (1996) Co-operation and competition: Contributions to an evolutionary and developmental model. International Journal of Behavioral Development, 19, 25-39.

Chesney-Lind, M. (1993) Girls, gangs and violence: Reinventing the liberated female crook. Humanity and Society, 17, 321-344.

Christophersen, E.R. (1989) Injury control. American Psychologist, 44, 237-241.

Coates, E.J. and Feldman, R.S. (1996) Gender differences in nonverbal correlates of social status. Personality and Social Psychology Bulletin, 10, 1014-1022.

Cody, M. and McLoughlin, M. (1988) Accounts on trial: Oral arguments in traffic court. In C. Antaki (Ed.) Analysing Everyday Conversation. London: Sage.

Cohen, L.E. and Machalek, R. (1988) A general theory of expropriative crime: An evolutionary ecological approach. American Journal of Sociology, 94, 465-501.

Coie, J.D. and Dodge, K.A. (1998) Aggression and antisocial behavior. In W. Damon and N. Eisenberg (Eds.) Handbook of Child Psychology. Volume 3: Social, Emotional and Personality Development. (Fifth Edition). New York: Wiley.

Corbitt, E.M. and Widiger, T.A. (1995) Sex differences among the personality disorders: An exploration of the data. Clinical Psychology: Science and Practice, 2, 225-238.

Costa, P.T. and McCrae, R.R. (1992) NEO-PI-R Professional Manual. Orlando, FL: PAR Inc.

Crick, N.R. and Grotpeter, J.K. (1995) Relational aggression, gender and social-psychological adjustment. Child Development, 66, 710-722.

Cummings, E.M., Hollenbeck, B., Iannotti, R., Radke-Yarrow, M. and Zahn-Waxler, C. (1986) Early organisation of altruism and aggression: Developmental patterns and individual differences. In C. Zahn-Waxler, E.M. Cummings and R. Iannotti (Eds.) Altruism and Aggression: Biological and Social Origins. Cambridge: Cambridge University Press.

Daly, M. and Wilson, M. (1983) Sex, Evolution and Behavior. Belmont, CA: Wadsworth.

Daly, M. and Wilson, M. (1988) Homicide. New York: Aldine de Gruyter.

Daly, M. and Wilson, M. (1994) Evolutionary psychology of male violence. In J,. Archer (Ed.) Male Violence. London: Routledge.

DeLisi, R. and Soundranayagam, L. (1990) The conceptual structure of sex role stereotypes in college students. Sex Roles, 23, 593-612.

de Waal, F.B.M. (1982) Chimpanzee Politics. New York: Harper and Row.

de Waal (1989) Peacemaking Among Primates. Cambridge, MA: Harvard University Press.

Dollard, J., Doob, L.W., Miller, N.E., Mowrer, O.H. and Sears, R.R. (1939) Frustration and Aggression. New Haven, Conn: Yale University Press.

Donald, M. (1991) Origins of the Modern Mind. Cambridge, MA: Harvard University Press.

Duckett, H., Lance, S., Pemberton, K., Raistrick, J., Campbell, A. and Muncer, S. (under review) Social positioning and sex differences in the representation of aggression.

Durham, W. (1991) Coevolutionary Theory. Stanford, CA: Stanford University Press.

Eagly, A.H. and Johnson, B.T. (1990) Gender and leadership style: A meta-analysis. Psychological Bulletin, 108, 233-256.

Eagly, A.H. and Karau, S.J. (1991) Gender and the emergence of leaders: A meta-analysis. Journal of Personality and Social Psychology, 60, 685-710.

Eagly, A.H., Makhijani, M.G. and Klonsky, B.G. (1992) Gender and the evaluation of leaders: A meta-analysis. Psychological Bulletin, 111, 3-22.

Eagly, A.H. and Steffen, V. (1986) Gender and aggressive behavior: A meta-analytic review of the social psychological literature. Psychological Bulletin, 100, 309-330.

Eder, D. (1990) Serious and playful disputes: Variations in conflict talk among adolescent females. In A. Grimshaw (Ed.) Conflict Talk: Sociolinguistic Investigations of Arguments in Conversations. Cambridge: Cambridge University Press.

Eder, D. and Sandford, S. (1988) The development and maintenance of interactional norms among early adolescents. In P. Adler and P. Adler (Eds.) Sociological Studies of Child Development. Greenwich, CT: JAI Press.

Edwards, S. (1986) Neither mad nor bad: The female violent offender reassessed. Women's Studies International Forum, 19, 79-87.

Elliott, D.S. and Huizinga, D. (1983) Social class and delinquency in a national youth panel. Criminology, 21, 149-177.

Ellis, L. (1988) The victimful-victimless crime distinction, and seven universal demographic correlates of victimful criminal behaviour. Personality and Individual Differences, 9, 525-548.

Ellis, L. (1995) Dominance and reproductive success among nonhuman animals: A cross-species comparison. Ethology and Sociobiology, 16, 257-333.

Ember, C.R. (1978) Myths about hunter-gatherers. Ethnology, 17, 439-448.

Ember, C.R. (1981) A cross cultural perspective on sex differences. In R.H. Monroe, R.L. Monroe and B. Whiting (Eds.) Handbook of Cross Cultural Human Development. New York: Garland.

Eron, L., Huesman, R., Brice, P., Fisher, P. and Mermelstein, R. (1983) Age trends in the development of aggression, sex typing and related television habits. Developmental Psychology, 19, 71-77.

Eysenck, H.J. (1964) Crime and Personality. London: Routledge and Kegan Paul.

Fagan, J. and Browne, A. (1994) Marital violence: Physical aggression between women and men in intimate relationships. In A. Reiss and J. Roth (Eds.) Understanding and Preventing Violence Volume 2. Washington DC: National Academy Press.

Feinblatt, J.A. and Gold, A.R. (1976) Sex roles and the psychiatric referral process. Sex Roles, 2, 109-122.

Feingold, A. (1994) Gender differences in personality: A meta-analysis. Psychological Bulletin, 116, 429-456.

Felson, R. (1978) Aggression as impression management. Social Psychology, 41, 205-213.

Feshbach, N.D. (1969) Sex differences in children's modes of aggressive responses toward outsiders. Merrill-Palmer Quarterly, 15, 249-258.

Feshbach, S. and Feshbach, N.D. (1986) Aggression and altruism: A personality perspective. In C. Zahn-Waxler, E.M. Cummings and R. Iannotti (Eds.) Altruism and Aggression: Biological and Social Origins. Cambridge: Cambridge University Press.

Fisher, H. (1993) Anatomy of Love. London: Simon and Schuster.

Foley, R.A. (1987) Another Unique Species. London: Longman.

Foley, R.A. (1996) An evolutionary and chronological framework for human social behaviour. Proceedings of the British Academy, 88, 95-117.

Folstad, I. and Karter, A.J. (1992) Parasites, bright males, and the immunocompetence handicap. American Naturalist, 139, 603-622.

Frodi, A., Macauley, J. and Thome, P.R. (1977) Are women always less aggressive than men? A review of the experimental literature. Psychological Bulletin, 84, 634-660.

Geary, D.C. (1996) Sexual selection and sex differences in mathmatical abilities. Behavioral and Brain Sciences, 19, 229-284.

Ginsburg, H.J. and Miller, S.M. (1982) Sex differences in children's risk-taking behavior. Child Development, 53, 426-428.

Goldberg, S. (1993) Why Men Rule: A Theory of Male Dominance. Peru, Illinois: Open Court.

Goodwin, M.H. (1982) Processes of dispute management among urban black children. American Ethnologist, 9, 76-96.

Goodwin, M.H. (1990) He-Said-She-Said: Talk as Social Organisation among Black Children. Bloomington: University of Indiana Press.

Gottfredson, M. and Hirschi, T. (1990) A General Theory of Crime. Stanford, CA: Stanford University Press.

Gray, J.A. (1987) The Psychology of Fear and Stress. Cambridge: Cambridge University Press.

Greenberg, D, Hillman, D. and Grice, D. (1977) Infant and stranger variables related to stranger anxiety in the first year of life. Developmental Psychology, 9, 207-212.

Gullone, E. and King, N.J. (1997) Three-year follow-up of normal fear in children and adolescents. British Journal of Developmental Psychology, 15, 97-111.

Harris, D. and Guten, S. (1979) Health protective behaviour: An exploratory study. Journal of Health and Social Behaviour, 20, 17-29.

Hartup, W.W. (1974) Aggression in childhood: Developmental perspectives. American Psychologist, 34, 944-950.

Hausfater, G. (1975) Dominance and reproduction in baboons (Papio cynocephalus). Contributions to Primatology, 7, 1-150.

Hausfater, G. and Hrdy, S.B. (Eds.) (1984) Infanticide: Comparative and Evolutionary Perspectives. New York: Aldine.

Hay, D.F. and Ross, H.S. (1982) The social nature of early conflict. Child Development, 53, 105-113.

Heidensohn, F.M. (1996) Women and Crime. Second Edition. London: Macmillan.

Herdt, G.H. (1982) Fetish and fantasy in Sambia initiation. In G.H. Herdt (Ed.) Rituals of Manhood. Berkeley: University of California Press.

Hill, K. and Hurtado, A.M. (1996) Ache Life History: The Ecology and Demography of a Foraging People. New York: Aldine de Gruyter.

Horowitz, R. (1983 ) Honor and the American Dream. New Brunswick, NJ: Rutgers University Press.

Hoyenga, K.B. and Hoyenga, K.T. (1993) Gender-Related Differences. Boston: Allyn and Bacon.

Hrdy, S.B. (1979) Infanticide among animals: A review, classification, and examination of the implications for the reproductive strategies of females. Ethology and Sociobiology, 1, 13-40.

Hrdy, S.B. (1981) The Woman That Never Evolved. Cambridge, MA: Harvard University Press.

Hrdy, S.B. (1986) Empathy, polyandry and the myth of the coy female. In R. Bleier (Ed.) Feminist Approaches to Science. New York: Pergamon.

Hughes, L.A. (1988) "But that's not really mean": Competing in a cooperative mode. Sex Roles, 19, 669-687.

Hyde, J.S. (1986) Gender differences in aggression. In J.S. Hyde and M.C. Linn (Eds.) The Psychology of Gender: Advances through Meta-Analysis. Baltimore: Johns Hopkins University Press.

Kano, T. (1992) The Last Ape: Pygmy Chimpanzee Behavior and Ecology. Stanford, CA: Stanford University Press.

Kashiwagi, T., McClure, J.N. and Wetzel, R.D. (1976) Premenstrual affective syndrome and psychiatric disorder. Diseases of the Nervous System, 37, 116-119.

Katz, J. (1988) Seductions of Crime: Moral and Sensual Attractions of Doing Evil. New York: Basic Books.

Kirsta, A. (1994) Deadlier than the Male. London: HarperCollins.

Knight, G.P., Fabes, R.A. and Higgins, D.A. (1996) Concerns about drawing causal inferences from meta-analyses: An example in the study of gender differences in aggression. Psychological Bulletin, 119, 410-421.

Kochman, T. (1983) The boundary between play and nonplay in black verbal duelling. Language in Society, 12, 329-337.

Kotelchuck, M. (1976) The infant's relationship to the father: Experimental evidence. In M.E. Lamb (Ed.) The Role of the Father in Child Development. New York: Wiley.

Kruttschnitt, C. (1993) Violence by and against women: A comparative and cross-national analysis. Violence and Victims, 8, 253-270.

Kruttschnitt, C. (1994) Gender and interpersonal violence. In A. Reiss and J. Roth (Eds.) Understanding and Preventing Violence Volume 3. Washington DC: National Academy Press.

Labov, W. (1972) Language in the Inner City: Studies in Black English Vernacular. Philadelphia: University of Philadelphia Press.

LaGrange, R.L. and Ferraro, K.F. (1989) Assessing age and gender differences in perceived risk and fear of crime. Criminology, 27, 697-717.

Lamphere, L. (1974) Strategies, cooperation and conflict among women in domestic groups. In M. Rosaldo and M. Lamphere (Eds.) Woman, Culture and Society. Stanford: Stanford University Press.

Lancaster, C.S. and Lancaster, J.B. (1983) Parental investment: The hominid adaptation. In D. Ortner (Ed.) How Humans Adapt. Washington DC: Smithsonian Institution Press.

Leakey, R. and Lewin, R. (1979) People of the Lake. London: Collins.

Leaper, C. (1991) Influence and involvement: Age, gender and partner effects. Child Development, 62, 797-811.

Lee, R.B. (1979) The Kung San: Men, Women and Work in a Foraging Society. Cambridge: Cambridge University Press.

Lejeune, R. (1977) The management of a mugging. Urban Life, 6, 123-148.

Lerner, G. (1986) The Creation of Patriarchy. New York: Oxford University Press.

Lever, J. (1978) Sex differences in the games children play. Social Problems, 23, 478-487.

Levinson, D. (1989) Family Violence in Cross-Cultural Perspective. Beverly Hills, CA: Sage.

Lewis, D.O., Shanok, S.S. and Pincus, J.H. (1982) A comparison of the neuropsychiatric status of female and male incarcerated delinquents: Some evidence of sex and race bias. Journal of the American Academy of Child and Adolescent Psychiatry, 21, 190-196.

Lorenz, K. (1966) On Aggression. London: Methuen.

Low, B.S. (1989) Cross-cultural patterns in the training of children: An evolutionary perspective. Journal of Comparative Psychology, 103, 311-319.

Low, B.S. (1992) Men, women, resources and politics in pre-industrial societies. In J.M.G. Van Der Dennen (Ed.) The Nature of the Sexes. The Netherlands: Origin Press.

Loy, J. (1971) Estrous behavior of free-ranging rhesus monkeys (Macaca mulatta). Primates, 12, 1-31.

Lumsden, C. and Wilson, E.O. (1981) Genes, Minds and Culture. Cambridge, MA: Harvard University Press.

Macauley, J. (1985) Adding gender to aggression research: Incremental or revolutionary change? In V. O'Leary, R.K. Unger and B.S. Wallston (Eds.) Women, Gender and Social Psychology. Hillsdale, NJ: Erlbaum.

Maccoby, E.E. (1988) Gender as a social category. Developmental Psychology, 24, 755-765.

Maccoby, E.E. (1990) Gender and relationships: A developmental account. American Psychologist, 45, 513-520.

Maccoby, E.E. and Jacklin, C.N. (1987) Gender segregation in childhood. In H. Reese (Ed.) Advances in Child Development, Volume 20. New York: Academic Press.

MacDonald, E. (1991) Shoot the Women First: Inside the Secret World of Female Terrorists. New York: Random House.

MacDonald, K. (1995) Evolution, the five factor model and levels of personality. Journal of Personality, 63, 525-567.

Maltz, D. and Borker, R. (1982) A cultural approach to male-female miscommunication. In J. Gumperz (Ed.) Language and Social Identity. New York: Cambridge University Press.

Marks, I.M. (1987) Fears, Phobias and Rituals. Oxford: Oxford University Press.

Marks, I.M. and Nesse, R.M. (1997) Fear and fitness: An evolutionary analysis of anxiety disorders. In S. Baron-Cohen (Ed.) The Maladapted Mind: Classic Readings in Evolutionary Psychopathology. Hove: Psychology Press.

McCarthy, B. (1994) Warrior values: A socio-historical survey. In J, Archer (ed.) Male Violence. London: Routledge.

Miller, E.M. (1986) Street Woman. Philadelphia, PA: Temple University Press.

Miner, V. and Longino, H.E. (1987) Competition: A Feminist Taboo? New York: The Feminist Press.

Mitani, J.C., Gros-Louis, J. and Richards, A.F. (1996) Sexual dimorphism, the operational sex ratio and the intensity of male competition in polygynous primates. American Naturalist, 147, 966-980.

Mitchell, C.L, Boinski, S. and van Schaik, C.P. (1991) Competitive regimes and female bonding in two species of squirrel monkeys (Saimiri oerstedi and S. sciureus). Behavioral Ecology and Sociobiology, 28, 55-60.

Moely, B., Skarin, K. and Weil, S. (1979) Sex differences in competition-cooperation behaviour of children at two age levels. Sex Roles, 5, 329-342.

Mori, A., Watanabe, K. and Yamaguchi, N. (1989) Longitudinal changes of dominance rank among the females of the Koshima group of Japanese monkeys. Primates, 30, 147-173.

Moscovici, S. (1984) The phenomenon of social representations. In R. Farr and S. Moscovici (Eds.) Social Representations. Cambridge: Cambridge University Press.

Moseley, W.H. and Gerould, M.H. (1975) Sex and parole: A comparison of male and female parolees. Journal of Criminal Justice, 3, 47-58.

Mosher, D.L. and Sirkin, M. (1984) Measuring a macho personality constellation. Journal of Research in Personality, 18, 150-163.

Murdock, G. P. (1967) Ethnographic Atlas. Pittsburgh: Pittsburgh University Press.

Norland, S. and Mann, P.J. (1984) Being troublesome: Women on probation. Criminal Justice and Behavior, 11, 115-135.

Norland, S., Wessel, R. and Shover, N. (1981) Masculinity and delinquency. Criminology, 19, 421-433.

Office of Population Censuses and Surveys (1995) Mortality Statistics: Cause. London: H.M.S.O.

Oliver, M.B. and Hyde, J.S. (1993) Gender differences in sexuality: A meta-analysis. Psychological Bulletin, 114, 29-51.

Omark, D. and Edelman, M. (1975) A comparison of status hierarchies among young children: An ethological approach. Social Science Information, 14, 87-107.

Ortner, S.B. and Whitehead, H. (1981) Sexual Meanings: The Cultural Construction of Gender and Sexuality. Cambridge: Cambridge University Press.

Osterman, K., Bjorkqvist, K., Lagerspetz, K., Kaukiainen, A., Huesmann, L.R. and Fraczek, A. (1994) Peer and self-estimated aggression and victimisation in 8-year-old children from five ethnic groups. Aggressive Behavior, 20, 411-428.

Packer, C. (1979) Male dominance and reproductive activity in Papio anubis. Animal Behaviour, 27, 37-45.

Parke, R.D. and Slaby, R.G. (1983) The development of aggression. In E.M. Hetherington (Ed.) Handbook of Child Psychology Volume 4: Socialization, Personality and Social Development. New York: Wiley.

Parlee, M.B. (1973) The premenstrual syndrome. Psychological Bulletin, 80, 454-465.

Peccei, J.S. (1995) The origin and evolution of menopause: The altricality-lifespan hypothesis. Ethology and Sociobiology, 16, 425-449.

Pellegrini, A.D. (1988) Elementary-school children's rough-and-tumble play and social competence. Developmental Psychology, 24, 802-806.

Polk, K. (1994) When Men Kill: Scenarios of Masculine Violence. Cambridge: Cambridge University Press.

Pratto, F., Stallworth, L.M. and Sidanius, J. (1997) The gender gap: Differences in political attitudes and social dominance orientation. British Journal of Social Psychology, 36, 49-68.

Pusey, A., Williams, J. and Goodall, J. (1997) The influence of dominance rank on the reproductive success of female chimpanzees. Science, 277, 828-831.

Rodseth, L.T., Wrangham, R.W., Harrigan, A.M. and Smuts, B.B. (1991) The human community as a primate society. Current Anthropology, 32, 221-254.

Rohner, R. (1976) Sex differences in aggression: Phylogenetic and enculturation perspectives. Ethos, 4, 57-72.

Ruble, D.N. and Brooks-Gunn, J. (1979) Menstrual symptoms: A social cognition analysis. Journal of Behavioral Medicine, 2, 171-194.

Saitoti, T.O. (1986) The Worlds of a Masai Warrior: An Autobiography. Berkeley: University of California Press.

Sanday, P.R. (1981) Female Power and Male Dominance. Cambridge: Cambridge University Press.

Savin-Williams, R. (1977) Dominance in a human adolescent group. Animal Behaviour, 25, 400-406.

Savin-Williams, R. (1980) Social interactions of adolescent females in natural groups. In H. Foot, A. Chapman and J. Smith (Eds.) Friendship and Social Relations in Children. London: Wiley.

Schuster, I. (1983) Women's aggression: An African case study. Aggressive Behavior, 9, 319-331.

Schuster, I. (1985) Female aggression and resource scarcity: A cross-cultural perspective. In M. Haug, D. Benton, P. Brain, B. Oliver and J. Mos (Eds.) The Aggressive Female. Netherlands: CIP-Gegevens Koninklijke Biblioteheek.

Seyfarth, R.M. (1976) Social relationships among adult female baboons. Animal Behaviour, 24, 917-938.

Silk, J.B. (1987) Social behavior in evolutionary perspective. In B.B. Smuts, D.L. Cheney, R.M. Seyfarth, R.W. Wrangham and T.T. Struhsaker (Eds.) Primate Societies. Chicago: University of Chicago Press.

Simon, R.J. and Baxter, S. (1989) Gender and violent crime. In N.A. Weiner and M.E. Wolfgang (Eds.) Violent Crime, Violent Criminals. Beverly Hills: Sage.

Small, M.E. and Smith, D.G. (1985) Sex ratio of infants produced by male rhesus macaques. American Naturalist, 126, 354-361.

Smart, C. (1979) The new female offender: Reality or myth? British Journal of Criminology, 19, 50-59.

Smith, P.K. (1974a) Aggression in a preschool playgroup: Effects of varying physical resources. In J. deWit and W.W. Hartup (Eds.) Determinants and Origins of Aggressive Behaviour. The Hague: Mouton.

Smith, P.K. (1974b) Aspects of the playground environment. In D.V. Canter and T.R. Lee (Eds.) Psychology and the Built Environment. London: Architectural Press.

Smuts, B.B. (1987) Gender, aggression and influence. In B.B. Smuts, D.L. Cheney, R.M. Seyfarth, R.W. Wrangham and T.T. Struhsaker (Eds.) Primate Societies. Chicago: University of Chicago Press.

Smuts, B.B. (1995) The evolutionary origins of patriarchy. Human Nature, 6, 1-32.

Sperber, D. (1994) The modularity of thought and the epidemiology of representations. In L.A. Hirschfield and S.A. Gelman (Eds.) Mapping the Mind: Domain Specificity in Cognition and Culture. Cambridge: Cambridge University Press.

Spence, J., Helmreich, R. and Stapp, J. (1974) The Personality Attributes Questionnaire: A measure of sex role stereotypes and masculinity-femininity. Journal Supplement Abstract Service Catalog of Selected Documents in Psychology, 4, 42 (Number 617).

Spencer, P. (1965) The Samburu. Berkeley: University of California Press.

Statham, A. (1987) The gender model revisited: Differences in the management styles of men and women. Sex Roles, 16, 409-429.

Steffensmeier, D. (1980) Sex differences in patterns of adult crime, 1965-77: A review and assessment. Social Forces, 58, 1080-1108.

Steffensmeier, D. and Cobb, M.J. (1981) Sex differences in urban arrest patterns, 1934-1979. Social Problems, 29, 37-50.

Stewart, S.H., Taylor, S. and Baker, J.M. (1997) Gender differences in dimensions of anxiety sensitivity. Journal of Anxiety Disorders, 11, 179-200.

Symons, D. (1979) The Evolution of Human Sexuality. Oxford: Oxford University Press.

Tannen, D. (1996) Talking from Nine to Five. London: Virago.

Taylor, C. (1993) Girls, Gangs, Women and Drugs. East Lansing, MI: Michigan State University Press.

Tedeschi, J.T., Smith, R.B. and Brown, R.C. (1974) A reinterpretation of research on aggression. Psychological Bulletin, 81, 540-562.

Thomas, E.M. (1965) Warrior Herdsman. New York: Knopf.

Thompson, E.H. and Pleck, J.H. (1986) The structure of male role norms. American Behavioral Scientist, 29, 531-543.

Thorne, B. (1994) Gender Play: Girls and Boys in School. New Brunswick, NJ: Rutgers University Press.

Thornton, W. (1982) Self-concept as a mediating factor in delinquency. Adolescence, 17, 749-768.

Toch, H. (1969) Violent Men: An Enquiry into the Psychology of Violence. Chicago: Aldine.

Turtle, J., Jones, A. and Hickman, M. (1997) Young People and Health: The Health Behaviour of School-Aged Children. London: Health Education Authority.

:

Umberson, D. (1992) Gender, marital status and the social control of health behaviour. Social Science and Medicine, 34, 907-917.

United States Department of Justice (1989) Uniform Crime Reports. Washington DC: Government Printing Office.

van Hoof, J.A. and van Schaik, C.P. (1992) Cooperation in competition: The ecology of primate bonds. In A. Harcourt and F.B.M. de Waal (Eds.) Coalitions and Alliances in Humans and Other Animals. Oxford: Oxford University Press.

Voland, E. (1988) Differential infant and child mortality in evolutionary perspective: Data from the late 17th to 19th century Osfriesland (Germany). In L. Betzig, M. Borgerhoff Mulder and P. Turke (Eds.) Human Reproductive Behaviour: A Darwinian Perspective. Cambridge: Cambridge University Press.

Waldron, I. (1988) Gender and health-related behavior. In D.S. Gochman (Ed.) Health Behavior: Emerging Research Perspectives. New York: Plenum.

Walsh, D. (1986) Heavy Business: Commercial Burglary and Robbery. London: Routledge and Kegan Paul.

Walters, J. (1980) Interventions and the development of dominance relationships in female baboons. Folia Primatologica, 34, 61-89.

Walters, J. and Seyfarth, R.M. (1987) Conflict and cooperation. In B.B. Smuts, D.L. Cheney, R.M. Seyfarth, R.W. Wrangham and T.T. Struhsaker (Eds.) Primate Societies. Chicago: University of Chicago Press.

Watson, L. (1995) Dark Nature. London: Hodder and Stoughton.

Weis, J. (1976) Liberation and crime: The invention of the new female criminal. Crime and Social Justice, 1, 17-27.

White, J.W. and Kowalski, R.M. (1994) Deconstructing the myth of the nonaggressive woman: A feminist analysis. Psychology of Women Quarterly, 18, 487-508.

Whiting, B. and Edwards, C. (1973) A cross-cultural analysis of sex differences in the behavior of children aged three through eleven. Journal of Social Psychology, 91, 171-188.

Whyte, M.K. (1978) Cross-cultural codes dealing with the relative status of women. Ethnology, 17, 211-237.

Wiggins, J. and Holzmuller, A. (1978) Psychological androgyny and interpersonal behaviour. Journal of Consulting and Clinical Psychology, 46, 40-52.

Wiggins, J. and Holzmuller, A. (1981) Further evidence on anrogyny and interpersonal flexibility. Journal of Consulting and Clinical Psychology, 46, 40-52.

Williams, J. and Best, D. (1990) Sex and the Psyche: Gender Roles and Self Concepts Viewed Cross Culturally. Beverly Hills; Sage.

Wilson, M. and Daly, M. (1985) Competitiveness, risk-taking and violence: The young male syndrome. Ethology and Sociobiology, 6, 59-73.

Wolfgang, M. and Ferracuti, F. (1967) The Subculture of Violence. New York: Barnes and Noble.

Wrangham, R.W. (1980) An ecological model of female-bonded primate groups. Behaviour, 75, 262-300.

Wrangham, R.W. and Peterson, D. (1996) Demonic Males. New York: Houghton Mifflin.

Zeanah, C.H. (1989) Adaptation following perinatal loss: A critical review. Journal of the American Academy of Child and Adolescent Psychiatry, 467, 468-469.

Zillmann, D. (1979) Hostility and Aggression. Hillsdale, NJ: Erlbaum.

Zuckerman, M. (1994) Behavioural Expressions and Biosocial Bases of Sensation Seeking. Cambridge: Cambridge University Press.

I would like to thank the Psychology Department at the University of Durham who granted me a term of research leave during which this paper was written. I am grateful to Dr. K. Hoyenga and seven other anonymous reviewers who provided both theoretical comments and invaluable concrete references. Thanks also to Dr. Steven Muncer, Professor John Archer, Professor Robin Dunbar, Professor Kirsti Lagerspetz and many undergraduate students who have discussed these ideas with me.