Transcript talk2
Summary of The Hitchhiker‘s Guide to Altruism: Gene-culture Coevolution and the Internalization of Norms Tim Johnson, Benjamin Scheibehenne & Guido Biele LIFE Seminar, Winter 2004/2005 How do prosocial norms evolve? • Gintis (2003) approaches this question with two conditions in mind: 1. The psychological mechanisms that facilitate internalization (such as guilt and shame) are passed from parents to their offspring via genes. 2. Social norms are culturally acquired by an individual from either her parents (“vertical transmission”) or her peers (“horizontal transmission”) Under these conditions do prosocial norms evolve? And, if so, how? How do prosocial norms evolve? • Three models, each with different assumptions, are developed to address this question: 1. The first model has unrealistic assumptions: For each individual, norms are beneficial and outweigh the cost of internalization. 2. The second model has slightly more realistic assumptions: Norms can be costly (e.g. Altruism) Altruism: A behavior that enhances others’ fitness at the cost of one’s own fitness 3. The third model has realistic assumptions: Norms are costly and individuals select whether or not they wish to internalize a norm Definitions • Individuals have a genotype (genetically inherited psychological capacity) that determines what kind of behavior an agent can culturally acquire; agents inherit their genotype from their parents • Individuals’ phenotype (culturally inherited behavior (social norms) determines what kind of behavior an individual actually shows; agents acquire their phenotype through either vertical transmission (from parents) or horizontal transmission (from peers) • Internalization is the acquisition of a phenotype through cultural transmission. The first model: The evolution of internalization Assumptions: • There are 2 Genotypes: 1 = capacity to internalize norms 2 = no capacity to internalize norms • There are 2 phenotypes: “C” = norm internalized “D” = no norm • This results in 3 Phenogenotypes (PGT): 1C, 1D, 2D The first model: The evolution of internalization Assumptions (cont.): • Internalizing a pro-self norm means paying a cost in order to receive a benefit from imposing sanctions on oneself when a norm is violated. • 1st Step is genetic evolution: A baby inherits its genotype from its parents (either 1 or 2). • 2nd Step is cultural evolution: Children acquire phenotype either from their parents or from their friends (C or D). Results of the 1st model • The author analyzed a dynamical system in which the transition probabilities from parental (familial) PGTs to offspring PGTs, and the frequency of the familial PGTs (as a function of a PGTs fitness/payoff) determined the proportions of PGTs in the following generation • In one stable environment, 1C evolves (internalization and norm) • 2D is also stable, but only when the following conditions aren’t present: – – – – the benefit of internalization is large the internalization of norms is not costly (in terms of fitness) parental transmission of the phenotype is biased towards internalization assortative mating exists When norms are beneficial, they evolve. But what happens when norms are costly?? The 2nd model: The evolution of altruism • Genotypes are identical as 1st model: 1 = capacity to internalize norms 2 = no capacity to internalize norms • Two new phenotypes are added: “A”= Altruistic norm “B”= Normless State (#2) • So, in total, there are four phenotypes: A, B, C, D • And nine phenogenotypes (of which only five are important to us): 1AC, 1AD, 1BC, 1BD, 2BD Results of the 2nd model • Pure populations that are stable: 1. 2. 3. 4. 5. 1AC 1BC 1AD 1BD 2BD Conditions of stability: – 1 is stable whenever the internalization of norms yields benefits; when internalization does not yield benefits 2 is stable. – Altruism is stable whenever the attractiveness of becoming an altruist (a culturally determined value) exceeds the costs of altruism Under plausible conditions, altruism and internalization will coevolve. Will this happen if individuals exhibit a bias toward proself norms?? The 3rd Model: What happens when successful phenotypes are copied? • What if the probability of copying another individual depends on the others’ fitness? • The model is the same as before but now people “envy” high fitness phenotypes. The higher the envy, the more likely an agent is to copy an individual with higher fitness • Results: 1A is stable so long as envy is not high. If envy is high (meaning there is a bias toward pro-self norms), then 1C becomes stable. • The conditions in which 2D is stable do not change Conclusion • The three models show how pro-social, altruistic norms can emerge without repeated interaction or reputation • Problem: Payoffs are not frequency dependent (i.e. the fitness of a PGT is independent of others behavior) • The persistence of altruistic norms results because cultural evolution is “faster” than genetic evolution: fitness incentives aren’t as “attractive” as the benefits of internalized cultural norms. • Masanori will elaborate on this fact in his presentation…