Cooperation and the division of labour

Cooperation is vital for maintaining the integrity of complex life forms. In many cases in nature cooperation manifests it-self through constituent parts performing different, but complementary, functions. The vast majority of studies on the evolution of cooperation, however, look only at the special case in which cooperation manifests itself via the constituent parts performing identical tasks. In this paper we investigate a class of games in which the socially optimal behaviour has the property of being heterogeneous. We show that this class of games is equivalent to a region of ST space (the space of normalised two-player games characterised by the ‘sucker’ and ‘temptation’ payoffs) which has previously been dismissed. We analyse, through a simple group selection model, properties that evolving agents would need to have in order to “solve” this dilemma. Specifically we find that positive assortment on pure strategies may lower mean individual pay-off, and that assortment on mixed strategies will increase pay-off, but not maximise it