Kin Selection and Group Selection

In group-structured populations in which some other assumptions are satisfied, kin and group selectionist methods provide formally equivalent conditions for change. However, this only shows an equivalence between two statistical methodologies, and this is compatible with there being a real, causal distinction between kin and group selection processes. This chapter pursues a Hamilton-inspired, population-centred approach to drawing that distinction, on which the differences between kin and group selection are differences of degree in the structural properties of populations. The relevant properties are K, the overall degree to which genealogical kin interact differentially, and G, the overall degree to which the population contains stable, internally integrated, and externally isolated social groups. A spatial metaphor (‘K-G space’) provides a useful framework for thinking about these differences.