Posts tagged game theory
Altruism, put simply, is helping others or cooperating with others, even if it is costly to self. Of course, something like that cannot evolve, unless there are benefits too, associated with such acts of apparent selflessness.
Cooperation, to start with, can evolve based on three forms of reciprocity: direct, indirect and network. All are based on the fact that there re repeated interactions between group of people- dyads, triads or many more. Reciprocity can typically be measured in the lab using the repeated Dictator/ Trust game.
Direct reciprocity was proposed by Robert Trivers as a mechanism for the evolution of cooperation. If there are repeated encounters between the same two players in an evolutionary game in which each of them can choose either to “cooperate” or “defect”, then a strategy of mutual cooperation may be favoured even if it pays each player, in the short term, to defect when the other cooperates.
Here, in direct reciprocity A trusts/helps B and hopes that when time comes B will reciprocate/help A. The top-of-the-mind factor is whether or not to trust somebody and whether or not to reciprocate someone’s trust. Trust and exploitation may be relevant issues here. In the Dictator/ Trust game this trust/exploitation manifests as the amount that is split and given to the other person vs kept with oneself.
In the standard framework of indirect reciprocity, there are randomly chosen pairwise encounters between members of a population; the same two individuals need not meet again. One individual acts as donor, the other as recipient. The donor can decide whether or not to cooperate. The interaction is observed by a subset of the population who might inform others. Reputation allows evolution of cooperation by indirect reciprocity.
This is partially correct that reputation for being trustworthy helps in indirect reciprocity; however that is only true for the downstream version; for the upstream version feelings of gratitude/happiness/awe/elevation in persons receiving the help/ witnessing the act also lead to more pro-social behavior by those receiving help/ witnessing. Thus feelings of gratitude/ awe/elevation mediate this kind of upstream indirect reciprocity. See below for upstream and downstream variants.
Individual acts of indirect reciprocity may be classified as “upstream” or “downstream”:
Upstream reciprocity occurs when an act of altruism causes the recipient to perform a later act of altruism in the benefit of a third party. In other words: A helps B, which then motivates B to help C.
Downstream reciprocity occurs when the performer of an act of altruism is more likely to be the recipient of a later act of altruism. In other words: A helps B, making it more likely that C will later help A.
Before touching upon network reciprocity, I will take a quick detour about kin selection. I believe kin selection or inclusive fitness is also a type of reciprocity (that between related individuals sharing genes) and may be rechristened genetic reciprocity. After all if A is likely to help B because they share x % of genes, the reverse is equally true and applicable. And of course this is mediated by emotional attachment to the kid/kin.
As per one definition of kin selection:
A biological theory stating that a gene that causes an organism to exhibit behavior detrimental to its survival will increase in frequency in a population if that behavior benefits the organism’s relatives, which will pass the gene on to subsequent generations.
If I slightly change words form above definition, I can now define a neighbor selection process as a cultural theory stating that a meme that causes an organism to exhibit behavior detrimental to its survival will increase in frequency in a population if that behavior benefits the organism’s neighbors, which will pass the meme on to subsequent neighbors.
We are now ready to look at network reciprocity:
Real populations are not well mixed, but have spatial structures or social networks which imply that some individuals interact more often than others. One approach of capturing this effect is evolutionary graph theory, in which individuals occupy the vertices of a graph. The edges determine who interacts with whom. If a cooperator pays a cost, c, for each neighbor to receive a benefit, b, and defectors have no costs, and their neighbors receive no benefits, network reciprocity can favor cooperation.
Basically, what I understand from the above is that if you help your neighbors sometimes such that the cost is not too high but benefits to neighbors are high and if cost to benefit compares favorably with average number of neighbors/ neighborly interactions you have, then in the long run you will benefit and this form of cooperation can evolve. To me the effects are mediated by the number of neighbors or sociability of a person.
Of course, even if you have all these mechanisms in place, cooperation may not evolve, as you may have free-riders. One important mechanism that has evolved to keep the free-riders in check is that of punishment. And once punishment is part of the picture you don’t even need repeated interactions, one-off games may be sufficient. I call this phenomenon Direct Punishment. One way it has been measured is with the Ultimatum game.
In the Ultimatum game, the second player can inflict costly punishment on first player by refusing to accept the division; this costly punishment is dyadic in nature and the aggression/hostility/vengefulness of the second player ensures that cooperation in even one-off encounters happens.
Basically instead of trusting and helping B, A starts by exploiting B and B retaliates by punishing A at cost to oneself.
Of course one can then surmise that there can be a phenomena of indirect punishment. This again may happen in two ways:
- Indirect punishment upstream: A is exploitative in nature; A exploits B; B punishes A, who then feels guilt/ gets reformed and stops exploiting C or even starts helping C.
- Indirect punishment downstream: A is exploitative in nature: A exploits B, B punishes A; B gets a reputation for being tough/competent and stops getting exploited by others say C or C may now even help B.
The Indirect reciprocity effects can be seen in Public goods/ trust game.
I will now take a detour and introduce the HEXACO model of personality which set me thinking about this in the first place.
HEXACO is an alternate personality model that is based on the same principles as the Big Five/FFM; i.e. it uses factor analysis of lexical terms in various languages to arrive at major personality traits.
The six factors are generally named Honesty-Humility (H), Emotionality (E), Extraversion (X), Agreeableness (A), Conscientiousness (C), and Openness to Experience (O). The personality-descriptive adjectives that typically belong to these six groups are as follows:
Honesty-Humility (H): sincere, honest, faithful, loyal, modest/unassuming versus sly, deceitful, greedy, pretentious, hypocritical, boastful, pompous
Emotionality (E): emotional, oversensitive, sentimental, fearful, anxious, vulnerable versus brave, tough, independent, self-assured, stable
Extraversion (X): outgoing, lively, extraverted, sociable, talkative, cheerful, active versus shy, passive, withdrawn, introverted, quiet, reserved
Agreeableness (A): patient, tolerant, peaceful, mild, agreeable, lenient, gentle versus ill-tempered, quarrelsome, stubborn, choleric
Conscientiousness (C): organized, disciplined, diligent, careful, thorough, precise versus sloppy, negligent, reckless, lazy, irresponsible, absent-minded
Openness to Experience (O): intellectual, creative, unconventional, innovative, ironic versus shallow, unimaginative, conventional
The factor H is a new factor not present in Big Five/FFM. The E though looking similar to N of big Five, is conceptually different; it no longer contains anger/hostility which are instead present in HEXACO A. similalrly there are important differences between HEXACO A and Big Five A. the other 3, C, O and X (extarversion) are similarly conceptualized and defined in both systems and have same loadings when tested together.
Ashton and Lee, the proponents of the HEXACO model, have themselves related evolution of altruism to these traits [pdf] and I am building on their work.
Basically as per them,
To begin, we have proposed that the Honesty- Humility and Agreeableness factors represent two complementary aspects of the construct of reciprocal altruism (Trivers, 1971). Honesty-Humility represents the tendency to be fair and genuine in dealing with others, in the sense of cooperating with others even when one might exploit them without suffering retaliation. Agreeableness represents the tendency to be forgiving and tolerant of others, in the sense of cooperating with others even when one might be suffering exploitation by them. (For a discussion of two broadly similar, although not identical, constructs, see Perugini, Gallucci, Presaghi, & Ercolani, 2003.) Presumably, high levels of Honesty- Humility are associated with decreased opportunities for personal gains from the exploitation of others but also with decreased risks of losses from withdrawal of cooperation by others. In a similar manner, high levels of Agreeableness are associated with increased opportunities for personal gains from long-run reciprocal cooperation with others, as well as with increased risks of losses from exploitation by others. (Note that we use the term altruism in terms of a dimension of altruistic versus antagonistic tendency, which involves both a willingness to help or provide benefits to others and an unwillingness to harm or impose costs on others.)In addition, we have proposed that Emotionality represents tendencies relevant to the construct of kin altruism (Hamilton, 1964), including not only empathic concern and emotional attachment toward close others (who tend to be one’s kin) but also the harm-avoidant and help-seeking behaviors that are associated with investment in kin (see also Lee & Ashton, 2004). Presumably, high levels of Emotionality are associated with increased likelihood of personal and kin survival, as well as with decreased opportunities for gains that are often associated with risks to personal and kin survival.