Summary
Evolutionists have argued for the role of coordinated physical aggression in shaping coalitional behavior among human males. When physical force determines access to important resources like food, territory, and mates, coalitions have a distinct competitive advantage over loners in conflicts ('stronger' coalitions similarly have advantages over 'weaker' coalitions). Historical, ethnographic, and even experimental evidence supports this hypothesis. Human females exhibit enduring bonds as well, however they almost never engage in coordinated physical aggression. I argue that women and girls engage in informational warfare, where the weapon is not violence but information, and where coalitions may indeed have a competitive advantage over loners. Below I briefly review relevant cross-disciplinary research: the primate socioecological model and how it relates to humans, empirical and theoretical work suggesting that reputation is an important mediator of access to contested resources, reputation manipulation via gossip, and my hypothesis that coalitions facilitate the ability to strategically manipulate reputations with gossip. I then discuss my experimental work exploring the social cognition of gossip, and my ethnographic work in real-world female coalitions: college sororities.
Background
Multidisciplinary research has shown that men everywhere tend to engage in coalitional, physically aggressive, between-group competition--'warfare'. Evolutionary accounts of this phenomenon argue that coalitions were more likely to prevail over lone individuals in physical aggression in ancestral environments. My primary research interest is in understanding, from an evolutionary perspective, female coalitional competition in humans. Coalitional, physically aggressive, between-group competition by women is nearly unheard of, historically and ethnographically. I have proposed, however, that female coalitions (i.e., dyads or cliques within communities) may in fact engage in a different type of coalitional competition. Female coalitions compete not physically but informationally, using gossip and other collaborative strategies to attack competitors' reputations and defend those of coalition members, a phenomenon I refer to as informational warfare. It is now widely recognized that women are not necessarily less aggressive than men, as long as indirect forms of aggression like gossip and ostracism are measured. What has not been recognized is that women also aggress in coalitions using indirect forms of aggression.
Females in many non-human primate species form coalitions to physically aggress against other females. The socioecological model of female primate sociality attempts to explain why females in some species establish strong, enduring bonds with other females, whereas females in other species have weak or absent bonds. Female relationships are influenced by the type and degree of within-group competition for food in a particular species: when food is abundant, evenly distributed, and low-quality (e.g., grasses and leaves), one animal's access to food is not limited by the efforts of another (i.e., there is no benefit to displacing a competitor from her feeding site); when food is patchy, scarce, and high-value (e.g., fruiting trees), individuals benefit from displacing other group members from their resources. Under the latter circumstances, cooperating to displace a competitor from her food is adaptive--and it is in those species whose ecology involves patchy, scarce, high-value food that females form the strong, enduring bonds that are often apparent in coalitional aggression over food and feeding sites. I apply this model to understanding human female sociality. At first glance, humans do not fit the model. For example, ancestral humans clearly subsisted on high-quality, patchily-distributed foods, yet it is males, not females, who engage in aggressive, coalitional competition. I argue, however, that the socioecological model does fit if we more closely examine the way food and other resources are allocated among group members in small-scale communities, and if we take into account the effects reputation has on this allocation.
Recent empirical studies in fields like evolutionary anthropology and experimental economics suggest that reputation--information about an individual based on his or her past behavior--may be an important variable in understanding the distribution of resources in small, kin-based societies. If access to resources provided by others were mediated by reputation in ancestral environments, then there should have been a strong selection pressure for the evolution of adaptations to manipulate one's own and others' reputations to one's own benefit. Such strategies involve the collection, analysis, and dissemination of reputation-relevant information, with the aim of increasing one's reputation relative to one's competitors (this activity closely resembles the culturally-constructed phenomenon of 'gossip'). Much recent research has shown that indirect forms of aggression like gossip are more typical of girls and boys. Other researchers have argued that females should emphasize this kind of aggression more than males because of sex differences in parental investment. I have argued that, in addition, females might incorporate reputational attacks more than males because of sex differences in vulnerability to reputational attacks, and sex differences in inter-group transfer (e.g., female exogamy is far more common across cultures than male exogamy). Within-group female competition over 'good' reputations is more likely to occur when women are heavily dependent on others for resources.
Informational Warfare Theory
Evolutionary theories of women's alliances and friendships have tended to emphasize benign functions for relationships like nurturing and caregiving (e.g., Taylor et al. 2000), or defensive functions, such as protection from male harassment (e.g., Smuts 1992; see Laidler and Hunt 2001 for examples). I have proposed that coalitions of women (i.e., 'cliques' within communities) might be better at reputational competition than lone competitors. First, coalitions can provide more eyes and ears through which to collect damaging, accurate information about competitors. The more individuals there are trying to collect information, the more likely it is to be found, especially when extensive observations are necessary for acquiring relevant information, or when information is difficult to come by because (because, e.g., people actively conceal it). Second, coalitions can more thoroughly and accurately analyze information, by, e.g., synthesizing various pieces of information into a complete account. Third, coalitions can provide more vectors (mouths) to strategically disseminate information to relevant parties. Fourth, gossip reported by more than one person may be more believable (because, e.g., multiple, independent sources for a particular piece of information greatly increase its reliability, as my experiments have shown). Finally, coalitions may protect individuals by providing alibis, evidence against accusations, and more effective threats. On his view, the socioecological model does explain female coalitional competition in humans, when the competition is not physical but reputational; coalitions help individuals achieve better reputations than their competitors, thereby securing relatively more contested resources. My experimental and ethnographic research builds on the existing behavioral evidence that people gossip in dyads or groups, particularly women and girls, and that a primary objective of such gossiping is to diminish the reputations of competitors.
Social Cognition Experimental Research
Several of my social-cognitive experimental studies have shown that gossip recipients are sensitive to subtle cues that would be expected to correlate with gossip accuracy (Hess and Hagen 2006). In one set of experiments, I explored cues of gossip veracity, finding that reiteration has a small but significant effect of increasing gossip believability; that noisy signals decrease believability; that multiple sources increase believability; that source independence increases believability; and that competition between a gossiper and her target decrease believability. In another set of experiments, I explored the effects of coalitions on the tendency of subjects to use informational aggression. In one experiment, subjects were less likely to say negative things about a female classmate when the classmate was described as having a close friend with her in the class. This friend protected the classmate from negative gossip by the subject even when the subject was in competition with the classmate. In another experiment, I found that if an informational aggressor had a coalition, subjects were less likely to act in a way that provoked an informational attack on themselves by the aggressor; however, this was only true for female subjects.
Ethnographic Research
In addition to the experimental work partially described above, for the past several years I have been conducting ethnographic fieldwork in Greek-letter sororities at a California university. Because sororities are institutions that are explicitly based on the establishment and maintenance of friendships among college women, because a sorority is a closed social group in which every member knows every other member, and because the resources over which individuals compete are easily identified, sororities are ideal for studying friend preferences, social networks, information transfer, house and individual reputation, categories of competition and conflict, strategies individuals and coalitions use to cooperate and compete, and what the outcomes of conflicts are.
Social science research on Greek life tends to focus entirely on the negative stereotypes associated with sorority and fraternity membership, such as excessive alcohol consumption, elitism, hazing, and eating disorders. Consequently, leaders in the Greek community and university officials are extremely reluctant to allow researchers into the community. The first one and a half years of this research involved establishing long-term relationships with student leaders in the university's Greek community, and with university officials who oversee the Greek community's affairs. I volunteered as the data analysis technician for Greek Affairs for two fall sorority recruitment events ('Rush'), and attended many Greek events as an affiliate of the Greek Office. During this time, I studied the structure of the Greek community, the formal and informal structure of Rush activities, the distribution of power and responsibilities among members of this highly-organized community, and the university's rules for organized social events in the Greek community (including the consequences of breaking these rules). Importantly, I was also able to observe the conflicts that houses had with one another during Rush.
In 2002 and 2003, I used several research methods to study the interactions of the women of one large sorority house, 'Delta Nu.' My methods included in-depth interviews, surveys (including a relational aggression instrument), and social network analysis. I have over 180 hours of recorded interviews with 54 of the 111 Delta Nu members. I obtained each informant's 'life history' from the year before she started college to the present, focusing on family dynamics, friendships before college, Rush experiences, perceptions of Delta Nu among other sororities, initial integration into the house, close friendship formation within the sorority, interactions with other sororities, organized social events, the impact of friendships on romantic relationships, reputations of oneself and other house members, examples of cooperative experiences and conflicts within and between sororities, and roles as elected and appointed house officers. I am currently analyzing the interview tapes for categories, correlates, and outcomes of conflicts (e.g., reputational shifts) in sororities. Interestingly, the majority of conflicts were not between-house conflicts over men (something heavily emphasized in prior ethnographic work), but within-house conflicts over power, rights, reciprocity, responsibility, money, and reputation. I have over 110,000 data points from these two seasons, and plan to write up analyses of them for publication in the near future.