The validity of the handicap principle in discrete action-response games

The validity of the handicap principle has spawned much debate in spite of the existence of a formal treatment. Simple models constructed to further investigate the issue were able both to prove and to disprove some of its claims. Here I show with the aid of a more general model, which takes into account both assumptions presented in these previous simple models: (1) that the previous results are not in conflict since they can be obtained as specific cases of this general model; (2) that ESS communication need not use costly signals, that is, even under conflict of interest, the cost of a signal used by a high-quality individual can be zero (or even negative) provided that the cost for low-quality signallers is high enough; (3) that only the cost relative to the benefits of the interaction should be higher for worse signallers; and (4) that in a discrete model the differential cost is only a necessary but not a sufficient condition for evolutionarily stable reliable communication. Copyright 1999 Academic Press.