UC Berkeley

Many species employ conditional strategies for reproduction or survival. In other words, each individual "chooses" one of several possible phenotypes in order to maximize survival or reproductive advantage given the specific ecological niche (e.g., Moran, 1992). Humans seem to employ at least one conditional reproductive strategy; each individual chooses between a more short-term or a more long-term mating strategy (Gangestad & Simpson, 2000), and as with non-human animals, their choices relate in part to an assessment of their own traits (Belsky, 1997; Schmitt, 2005). However, the selection pressures that individuals of a species can exert on each other are not restricted to mate selection; they can arise from many forms of social interaction (West-Eberhard, 1983; Wolf, Brodie, & Moore, 1999). Evidence does suggest that individuals are sensitive to characteristics of the self, friend, and environmental conditions when choosing friends (Fehr, 1996; Rose, 1985; Verbrugge, 1977), and that a person's economic, social, and environmental circumstances influence how they form and organize their friendships (Adams & Allan, 1998; Feld & Carter, 1998). Thus, in this dissertation I hypothesized that humans have evolved a coherent range of conditional friendship strategies: that we vary predictably in terms of the friendships we form, based on an assessment of our own traits, others' traits, and our own current needs. I also hypothesized that individuals are able to perceive and detect reliable signals (Cronk, 2005; Searcy & Nowicki, 2005; Smith, 1994) of these strategies in others, enabling them to choose friends whose traits are most desirable to them.

In the Chapter 1, I proposed a continuum of individual differences in friendship strategy, anchored on one end by those who use friendships for exploration (e.g., skill-building and networking) and on the other end by those who use friendships for intimate exchange (e.g., emotional support and intimacy). I created and refined a measure assessing this continuum, and found that men tended to report a stronger exploration strategy than women. I also found that people who had a stronger exploration strategy also had a more short-term mating strategy and were more extroverted, and that people who had a stronger intimate exchange strategy reported themselves to be more kind and generous; these results remained when controlling for gender. However, friendship strategy did not relate to socioeconomic status, age, attachment avoidance, relationship status, or presence of kin relationships. There was some evidence that friendship strategy was related to the number of friends an individual reported having and how close they felt to their friends.

The purpose of the study in Chapter 2 was to examine whether specific behaviors people displayed were related to their self-reported friendship strategy. I used videotapes of interactions between unacquainted participants, and hypothesized that people with a stronger exploration strategy would show behaviors that projected charisma, confidence, and social skills such as humorousness, whereas those with a stronger intimate exchange strategy would show behaviors that projected intimacy, compassion, warmth, and emotional availability. Results did not support this hypothesis; participants' friendship strategies generally did not relate to the behaviors that were assessed, which included gazing at the interaction partner, gesturing with the hands, laughing, and Duchenne smiles.

Further developing the ideas introduced in Chapter 2, the study described in Chapter 3 examined naïve viewers' reactions to the videotaped participants. I hypothesized that when asked to choose a hypothetical new friend from among the videotaped participants, viewers would detect signals of friendship strategy and would thereby choose someone whose self-reported friendship strategy more closely matched their own. Results did not generally support this hypothesis. However, findings did indicate that viewers might have been more uniformly influenced by signals of friendship strategy, as indicated by the finding that many viewers preferred videotaped participants with a stronger exploration strategy; this finding was stronger in men than in women.

Overall, the results of this series of studies suggest that people do characterize their friendship preferences along a continuum, and that these preferences do relate to personality and mating strategy, suggesting that this continuum may reflect a conditional strategy, with individuals "choosing" a friendship phenotype based on their own traits. However, there was not clear evidence of the reliable signals that might be necessary to help individuals use these conditional strategies effectively when actually choosing friends. It is possible that these signals were not detected because they do not arise in individuals' first interactions with each other.

This is the first research to study human friendship choice using the theoretical framework of conditional strategies, and it lays the foundation for new understandings of how humans form voluntary relationships. Future studies should attempt to discover whether reliable signals of friendship strategy emerge after repeated interpersonal interactions, or under different conditions of interaction, and they should include participants with a wider age range and more varied socioeconomic status. Despite their limitations, the studies in this dissertation propose and begin to test a novel theoretical framework for studying human friendship, and they suggest avenues for future work that should capitalize on the framework of evolutionary social psychology described here.