When W. D. Hamilton was a student he expressed an interest in exploring the problem of altruism in an evolutionary context. His struggles in getting anyone interested in the issue and supporting his study of the topic is extensively detailed in Narrow Roads of Gene Land. But he persevered and for his efforts he came up with the framework of inclusive fitness (John Maynard Smith’s term was kin selection). To a great extent it was a revolutionary model, formalizing what he been roughly understood verbally.
But could inclusive fitness explain the social structure we see around us? Hamilton attempted to extend the framework to humans in the 1970s, but that was not particularly fruitful. Other dynamics which emerged on the scene drew more from game theory. Again, John Maynard Smith loomed large, but Robert Trivers also introduced reciprocal altruism into the lexicon. These sorts of processes were much favored by thinkers such as Richard Dawkins because they are simple elementary strategies and relations that are tractable, and can be programmed dand simulated (or analytically explored).
Other researchers have different ideas and appeal to alternative traditions. David Sloan Wilson, along with E. O. Wilson, have been trying to revive models predicated on higher levels of organization. Though often termed “group selection,” the first Wilson correctly labels it “multi-level selection theory.” Though I am willing to agree that the pendulum swung too far in favor of individual level game theory and inclusive fitness in the last few decades, I do find David Sloan Wilson’s triumphalism a bit much (though his books are worth reading, and I think this is a personality issue with David, as he engages in the same triumphalism with economists).
A lot of work still needs to be done to explain social organization and behavior, even in social insects! With that, two preprints in biorxiv caught my attention.
First, Co-evolution of dispersal with behaviour favours social polymorphism. In it the authors model a system where there are dispersing individuals and sessile individuals, and show that cooperative behavior and the sessile morph and selfish behavior and the dispersing morph can persist as two alternative strategies. The paper makes the assumption that the sets of behaviors are caused by different genes which are linked, and show that low recombination is necessary to maintain the linkage. This does not seem genetically realistic.
The second paper is of a broader purview, Stags, hawks, and doves: Social evolution theory and individual variation in cooperation:
One of the triumphs of evolutionary biology is the discovery of robust mechanisms that promote the evolution of cooperative behaviors even when those behaviors reduce the fertility or survival of cooperators. Though these mechanisms, kin selection, reciprocity, and nonlinear payoffs to cooperation, have been extensively studied separately, investigating their joint effect on the evolution of cooperation has been more difficult. Moreover, how these mechanisms shape variation in cooperation is not well known. Such variation is crucial for understanding the evolution of behavioral syndromes and animal personality. Here, I use the tools of kin selection theory and evolutionary game theory to build a framework that integrates these mechanisms for pairwise social interactions. Using relatedness as a measure of the strength of kin selection, responsiveness as a measure of reciprocity, and synergy as a measure of payoff nonlinearity, I show how different combinations of these three parameters produce directional selection for or against cooperation or variation in levels of cooperation via balancing or diversifying selection. Moreover, each of these outcomes maps uniquely to one of four classic games from evolutionary game theory, which means that modulating relatedness, responsiveness, and synergy effectively transforms the payoff matrix from one the evolutionary game to another. Assuming that cooperation exacts a fertility cost on cooperators and provides a fertility benefit to social partners, a prisoner’s dilemma game and directional selection against cooperation occur when relatedness and responsiveness are low and synergy is not too positive. Enough positive synergy in these conditions generates a stag-hunt game and diversifying selection. High levels of relatedness or responsiveness turn cooperation from a fitness cost into a fitness benefit, which produces a mutualism game and directional selection for cooperation when synergy is not too negative. Sufficiently negative synergy in this case creates a hawk-dove game and balancing selection for cooperation. I extend the results with relatedness and synergy to larger social groups and show that how group size changes the effect of relatedness and synergy on selection for cooperation depends on how the per capita benefit of cooperation changes with group size. Together, these results provide a general framework with which to generate comparative predictions that can be tested using quantitative genetic techniques and experimental techniques that manipulate investment in cooperation. These predictions will help us understand both interspecific variation in cooperation as well as within-population and within-group variation in cooperation related to behavioral syndromes.
I haven’t dug into the formal models in the methods sections of either preprint, so I won’t say much more. But, I will offer that as someone who has long been interested in this field there is a surfeit and not enough data to test the models. It is time for someone ambitious to figure out how to make these areas more empirically testable.