EHBEA Day Two: Competitive Altruism and Competition for Partners; also a lot of masculinity.

Today was the second day of the EHBEA conference in Amsterdam. Unfortunately, I was forced to miss the keynote, but there were some very interesting presentations. In particular the work by Gilbert Roberts on competitive altruism is very promising, and it was great to hear about it. Arno Riedl also reported some very interesting (and related) work on competition for partners as a driver of cooperation.

Gilbert Roberts on Competitive Altruism

Gilbert Roberts of Newcastle University presented models and empirical evidence on competitive altruism, which is the idea that by behaving altruistically, individuals provide information to potential cooperation partners. Competition for more appealing cooperation partners then drives reputation-enhancing (altruistic) behaviour.van Vugt, M., Roberts, G., & Hardy, C. L. (2005). Competitive Altruism: Development of Reputation-based Cooperation in Groups. In R. I. M. Dunbar & Louise Barrett, Handbook of Evolutionary Psychology (pp. 1-28). Oxford, UK: Oxford University Press.

Competitive altruism has been tested in two-stage economic games. The first stage allows for building a reputation as a good cooperator (e.g. Public Goods Games, PGG). Then, during the second stage players can choose partners to play paired games with. As such, competitive altruism is in particular a rival theory to indirect reciprocity. Roberts reported findings that in the PGG, competitive altruism drives contributions more than indirect reciprocity does.

Based on these findings, Roberts proposed an agent-based model of reputation building with three decision points:

  • Stage 1: Build reputation or not?
  • – choose partner according to reputation or not –
  • Stage 2: Cooperate or defect on partners? (or don’t play)

In simulations, reputation and cooperation can evolve (as can choosing, but at lower levels), at least for some parameters. As the number of meetings increases, higher levels of reputation building are observed.

Child Mortality and Reproduction

Caroline Uggla of the University College London proposed to investigate why health technologies are not adopted in developing countries even when they are available based on parental investment theory, and bringing together human behavioural ecology and public health. Using data from the Demographic Health Surveys, she showed that parental investment theory predicts how several mother- and child-status variables predict whether children will receive curative and preventative treatments. For example, older mothers invest more in their children. However, some puzzles remain. The positive effect of mothers’ age, for instance, appears contradictory with the negative effect of birth order. Also, sicker children are more likely to get curative, but less likely to get preventative treatments, but it is not clear why.

Paul Mathews of the London School of Hygiene and Tropical Medicine (LSHTM) presented evidence from two studies on reproductive plasticity, i.e. the idea that reproduction is dependent on environmental factors. In two experiments, he found that priming subjects with their own mortality increases the ideal number of children in men, but not women; and increases the desire for childlessness in women, but not men (furthermore, priming subject with dental health increases their desire to remain childless – a control condition gone hilariously wrong).

Susan Schaffnit, also of the LSHTM, presented findings on parent-child proximity and women’s reproductive fitness in Europe. She set out to test hypotheses about two conflicting theories. On the one hand, cooperative breeding theory predicts that living with parents increases reproductive fitness. On the other hand, local resource competition theory predicts that being around relatives (although more often siblings than parents) decreases reproductive fitness. Schaffnit’s analysis of data from Europe found support for both hypotheses: living parents are associated with younger age at first birth, but living with them has a negative effect. Women who move away from home later also have a higher probability of remaining childless.

Facial Characteristics and Ultimatum Game Behaviour; Female Economic Dependence

I missed the first talk of this session, by Daniel Tayler on the excludability of public goods, and was unfortunately distracted during the third by Hannah Cornish on systematicity of culture.

Poppy Mulvaney of the University of Bristol reported on two studies investigating the effect of receiver’s facial characteristics on proposer behaviour in the Ultimatum Game. In a UK sample, formidability – associated with physical dominance – predicted offers, but trustworthiness did not. However, in a second sample in the United States, trustworthiness, but not formidability predicted fair offers. Notably, the two samples did not differ in their personality ratings for the faces, thus leaving open the question as to why different facial characteristics would predict proposer behaviour.

Michael Price of Brunel University proposed a model of morality judgements on promiscuity that includes female economic dependence. This hypothesis is consistent with the idea that opposition to promiscuity is to promote paternity certainty. Indeed, female economic dependence somewhat predicted attitudes towards both male and female promiscuity in a model also including sex, religiosity, and conservativism. Price argues that religiousity and conservatism are only proximate predictors, i.e. that their anti-promiscuous stance requires explanation, and sees evolution in an environment of high female economic dependence as an answer.

Arno Riedl on Competition for Partners as a Driver of Cooperation

Arno Riedl of Maastricht University presented experimental findings that competition for partners can drive cooperation among strangers. The underlying question of this study is what it needs to enforce the social norms of cooperation in societies with infrequent interactions. This was tested using a very interesting modification of the classic Prisoners’ Dilemma that allowed Riedl his collaborator Aljaz Ule (of the University of Amsterdam) to tease apart the effects of exclusion and partner choice.

Subjects played repeated PD games, paired up in groups of three. They all learn about the most recent interactin of each subject in the triplet. In each triplet, at most one pair of subjects can play the PD (which is strictly preferable over not playing), thus inducing partner competition. Riedl and Ule then varied the condition under which pairs are formed:

  • baseline: formed randomly (no partner refusal, no competition)
  • refusal: formed randomly, but each subject can refuse to play
  • competition: subjects indicate acceptable partner, if feasible a pair is formed

Cooperation rates in round one were very similar (43/43/38%), but diverged strongly across 60 iterations: While baseline and refusal condition cooperation fell to around ten percent in a monotonic decrease, it steadily increased to almost fifty percent in the competition condition. Refusal was rare throughout.

Interestingly, in the competition condition, two types of groups emerge: non-cooperative and fully cooperative groups. This difference appears to be driven by the exclusion of previous defectors. In the refusal condition, refusal was rare (2%). Non-cooperative groups in the competition condition also showed rather low levels of exclusion of previous defectors (15%), which markedly set them apart from cooperative groups, whose members refused to cooperate with defectors 50 percent of the time. Riedl concluded that community enforcement of cooperation norms is possible, but requires competition. The mechanism at work is the exclusion of revealed defectors.

Recalibration Hypothesis; Evolution of Masculine Faces

Lars Penke of the University of Edinburgh tested the recalibration hypothesis, which argues that the effects of genetic variation on personality (or behaviour more broadly) is mediated by morphology (e.g., a proposal that narcicissm is founded in objective attractiveness). Using several anthropomorphic measures and third-party rating, Penke found only scattered associations with personality traits. He concludes that the recalibration theory explains only little variation in social personality traits, and only in men. Previous studies were likely confounded, e.g. by using self-reported attractiveness (which might itself correlate with traits like narcicissm).

Iris Holzleitner of the University of St. Andrews presented a model linking morphological masculinity and attractiveness. Two contrasting hypotheses have been put forward to explain the evolution of masculine faces. One is that they are a handicap to signal health, and evolved due to intersexual selection (i.e. choosy women looking for healthy males). In contrast, masculine faces might have evolved under pressure of intrasexual competition as a cue to dominance. Holzleitner’s model, which takes into account facial masculinity (as well as height and weight), finds that masculinity has a significant effect on social dominance, but not health; and an effect on attractiveness only via social dominance. This suggests that masculinity evolved as an intrasexually selected trait.