| OCR Text |
Show Chapter XII. Classical Utilitarianism and the Free Rider 512 chance, by agreement or tacitly.28 The more points of similarity between the present coordination problem and its precedent, the more each party is justified in expecting the other parties to concur in solving it in a similar or analogous way. Thus a solution established by precedent gives each party reason to assume a certain pattern of predictable behavior in the others and to calculate his own conduct accordingly. But I now argue that Hodgson's dilemma implies that no such precedent could be established within the constraints of an ideal Act-Utilitarian society that requires the publicity of its first and only principle of conduct. Successful public adherence to the principle of utility would have to presuppose prior regularities of conduct that could not themselves be justified or determined by that principle. So neither such regularities nor the precedent on which they might be consequent would be forthcoming under conditions that publicly recognized the principle of utility as the only rule of behavior. Thus the choice of action for a consistent Utilitarian would in every case have to be made not between some two alternatives but among a nearly unlimited number of unweighted (or equally weighted) possibilities. I conclude, then, that Gibbard's and Lewis's reliance on the possibility of constructing a weighted, limited-alternative coordination matrix not only fails to address the problem Hodgson raises but indeed begs the question at issue. Absent some better solution to Hodgson's problem, this means that the Humean, utilitymaximizing model of rationality is incapable of morally justifying an ActUtilitarian conception of the good society. Suppose the alternatives for two rational Act-Utilitarians were in fact, say, between going to the tennis courts and staying at home. This would mean that each party, being fully rational, might simply construct the same coordination matrix, accurately working out the probabilities and the desirable risk for each alternative. Knowing the full Utilitarian rationality of both parties, each would rightly expect this process of reasoning to be fully replicated by the other, so that each would expect the other to arrive at the same solution which she herself did. A condition of this would of course be that each know and assign the same weights and probabilities to each alternative and also have good reason for assuming that the other did so as well. Thus the expectations of each party would, in this case, have to be known to the other: these expectations would derive from their mutually acknowledged rationality and accurate weightings and probability assignments to the positive consequences of each alternative. But because both Gibbard and Lewis give the parties access to the same precedent-setting information, each can justifiably expect the other to assess the alternatives in just this way. The consistent but independent assumptions utilized in Gibbard's and in Lewis's solutions thus amount to stipulating in advance the 28 Ibid., pp. 33-36. © Adrian Piper Research Archive Foundation Berlin |
