We were left with a serious problem at the end of the previous chapter. The game of Chicken reopens the compliance problem that we had managed to solve in the case of the Prisoner’s Dilemma in Part II. I have argued that in Chicken the substantively rational strategy is to acquiesce to some threats but Gauthier’s moral theory seems to proscribe unfair acquiescence and even demands that agents enforce this proscription by threatening acquiescent would-be co-operators. How might we close this new gap between rationality and morality? In this chapter I will work on this problem from both sides. First, the roster of strategies in the previous chapter had grown long and complex. It is really possible to build agents that can draw the needed distinctions? Recall that I postulated a less broad co-operator (LBC) with an ability to discriminate sources of threats. If this hyper-discrimination turns out to be procedurally problematic, we would be left with Gauthier’s proposed narrower co-operator (NRC) as both rational and moral, closing the compliance dilemma for the case of Chicken. If, on the other hand, this procedural tactic should fail and LBC is procedurally possible, I must fall back to a moral defence. As in the case of reciprocal co-operation, I will need to defend a principle, less broad cooperation, which has an obvious moral defect. LBC players look to be traitors to the moral cause; they trade with the enemies of impartial constraint. As the analogy to trade and war suggests, the issue is not morally clear-cut. I will defend LBC’s morally less strenuous strategy, her willingness to accept co-operative outcomes even if they are unfair and her refusal to use sanctions to defend the higher standard of fairness.
There are two new procedural problems posed by the move from Prisoner’s Dilemma to Chicken. First, Chicken requires finer powers