| OCR Text |
Show Rationality and the Structure of the Self, Volume I: The Humean Conception 157 Hence the agent can be interpreted as maximizing utility in her actions whether she in fact behaves consistently or not. Indeed, the concept of expected utility-maximizing preference revealed in behavior would subvert the imposition of consistency constraints on action, even were those constraints not subject to worries about intensionality. For if any action can be interpreted as maximizing the agent's expected utility in virtue of the fact that she performs it, then in particular any preference ranking of alternatives the agent makes at a particular moment can be interpreted as the outcome of consistent pairwise comparisons among all the alternatives available at that moment, and so as reflecting a consistent 24 ordering of those alternatives. Ramsey's result does not redress the vacuity of (U) because it fails to insure the consistency through time of the agent's preferences. As Donald Davidson observes, "The theory merely puts 25 restrictions on a temporal cross-section of an agent's disposition to choose." This means that any behavioral violation of Ramsey's consistency axioms can be understood, in accordance with the principle of charity, as a change in the agent's preferences instead. Hence these restrictions are vacuously inviolable - not because agents never in fact behave inconsistently, but because the 26 Ramsey-Savage concept of a simple ordering by itself is not sufficient to ensure transitivity of preference through time. ordinarily do not; and that all-things-considered judgments express not preference but normative preferability, which it ordinarily does not. Temkin's discussion is valuable because of its breadth, detail, and focus; but these terminological idiosyncrasies lend it a greater air of paradox than seems warranted. I suggest a way of avoiding intransitivities caused by preferences among multi-dimensional alternatives in Volume II, Chapter III.9. 24 I use the term "preference ranking" to refer to the result of a pairwise comparison between two given alternatives, and "ordering" to refer to the resultant relations of priority that obtain among all such alternatives consecutively ranked. One way of putting my point would be to say that the burden of interpreting the concept of maximizing utility is carried by the concept of a preference ranking, not by that of a simple ordering. This is why the impossibility of linearly representing intransitive preferences does not exclude their cyclical ordering. For an example, see the discussion of Cleopatra in Section 3.2, below. 25 Donald Davidson, "Psychology as Philosophy," in Essays on Actions and Events (Oxford: Clarendon Press, 1980), 235. Sen makes essentially the same point in "Rational Fools: A Critique of the Behavioral Foundations of Economic Theory," Philosophy and Public Affairs 6, 4 (1977), 317-44; see esp. 325. 26 As refined by Leonard Savage in The Foundations of Statistics (New York: Dover Publications, Inc., 1971), 17-21; also see R. D. Luce and Howard Raiffa, Games and Decisions (New York: John Wiley and Sons, Inc., 1957), 25-28. Tjalling Koopmans (in "Allocation of Resources and the Price System," in Three Essays on the State of Economic Science (New York: McGraw-Hill, 1957) dismisses the time-dependence problem © Adrian Piper Research Archive Foundation Berlin |
