Decisions, Decisions, Decisions: Can Savage Salvage Everettian Probability?∗

Critics object that the Everett view cannot make sense of quantum probabilities, in one or both of two ways: either it cannot make sense of probability at all, or it cannot explain why probability should be governed by the Born rule. David Deutsch has attempted to meet these objections by appealing to an Everettian version of Savage’s rational decision theory. Deutsch argues not only that an analogue of classical decision under uncertainty makes sense in an Everett world; but also that under reasonable assumptions, the betting odds of a rational Everettian agent should be constrained by the Born rule. Deutsch’s proposal has been defended and developed by David Wallace, and in a different form by Hilary Greaves. In this paper I offer some objections to the Deutsch-Wallace-Greaves argument, focussing in particular on the supposed analogy with classical decision under uncertainty.  Fabulous at fifty? Our forties are often a fortunate decade. Understanding ourselves better, resolving old problems, we celebrate our fiftieth year with new confidence, a new sense of purpose. So, too, for the Everett interpretation, at least according to one reading of its recent history. Its fifth decade has resolved a difficulty that has plagued it since youth, indeed since infancy, the so-called problem of probability. Critics ∗A slightly shorter version of this paper is to appear in a volume edited by Jonathan Barrett, Adrian Kent, David Wallace and Simon Saunders, containing papers presented at the Everett@ conference in Oxford in July, , and the Many Worlds@ meeting at the Perimeter Institute in September, . See Price () for the original talk on which this paper is based. †Email: [email protected]