Recently Devine writes as follows: The CIA's prescriptions, in sum, are to mess with the prisoners' minds (and, if need be, their bodies) in order to _create_ the textbook prisoners' dilemma, to encourage each prisoner to "defect," to turn in his or her comrades. The textbook treatment _assumes_ that the prisoners are already atomistic individuals and there is _no need_ to plumb the depths of their psyches in order to make them that way. In fact, the vast majority of economists reject the need to study the depths of psychology at all; instead, they simply assume that people "maximize utility," having no sense of honor or morality, solidarity or self-esteem. It treats the fact that many if not most prisoners do not defect as a "paradox" rather than as showing up the severe limitations of the theory. COMMENT: There are two competing views of the non-repeating PD. While there is a standard argument that concludes that the dominant and rational strategy is to defect,(the Dominance argument) there is a so-called symmetry argument that can be traced back at least to Anatol Rapoport in FIGHTS GAMES AND DEBATES (1960). This argument has been elaborated by L Davis in Prisoners, Paradox, and Rationality. AMERICAN PHILOSOPHICAL QUARTERLY 14: 319-27.) A somewhat different argument but equally critical of the standard argument is McClennen E. Prisoners Dilemma and Resolute Choice. In Campbell R. and Sowden L.(eds) PARADOXES OF RATIONALITY AND CO-operation UBC Press 1985. The basic idea behind the symmetry argument is that each prisoner being rational and in the same situation will see that the dominance argument leads to a sub-optimal result FOR BOTH as contrasted with their both remaining silent. THerefore as rational agents they will resolutely choose, to use McLennen's term, not to defect in order to improve the result for them both. This avoids the paradoxical result of the traditional argument in which allegedly rational choice leads to sub-optimal outcomes for both participants. I agree with Davis and McLennen but not with Rapoport because Rapoport thinks that somehow altruism is involved. ALtruism is not involved at all. The argument works even for rational egoists who have no interest per se in the welfare of others. However, when it is necessary to co-operate to maximize the welfare of each it is rational to do so. I have an unpublished (but presented) paper that argues against the dominance argument, and also Parfit's(REASONS AND PERSONS 1984) critique of ethical egoism as self-defeating as shown by the PD. I think the PD shows no such thing. I also think that Gauthier(MORALS BY AGREEMENT 1986) is incorrect in thinking that unconstrained maximisers would choose defection as the rational strategy in PD contexts. Of course many of the things discussed in the literature as PD's have nothing to do with PD's. For example the position of polluters who have no motive to install pollution devices in a competitive market even though this might advance the public welfare. In these cases all polluters defecting from the welfare producing policy does not on the whole hurt them, so there is nothing at all paradoxical involved. One case that is like that of the PD is that involving manufacture of aerosols without fluorocarbon propellants. If everyone did it, it would cheaper for all manufacturers, and in that sense in all their interests, but no one manufacturer will do it because they do not want to give up the market that still exists for the earlier propellant. Manufacturers did not oppose a ban on fluorocarbon propellants since it would provide assurance that no firm could defect and capture a market niche. CHeers, Ken Hanly Cheers, Ken Hanly
