Search Talks

This talk follows on from Wayne Myrvold\'s (and is based on joint work with Myrvold). I aim (and claim) to provide a unified account of theory confirmation that can deal with the (actual) situation in which we are uncertain whether the true theory is a probabilistic one or a branching-universe one, that does not presuppose the correctness of any particular physical theory, and that illuminates the connection between the decision-theoretic and the confirmation-theoretic roles of probabilities and their Everettian analogs. (The technique is to piggy-back on the existing body of physics-independent decision theory due to Savage, De Finetti and others, and to exploit the pervasive structural analogy between probabilistic theories and branching-universe theories in arguing for a particular application of that same mathematics to the branching case.) One corollary of this account is that ordinary empirical evidence (such as observed outcomes of relative-frequency trials) confirms Everettian QM in precisely the same way that it confirms a probabilistic QM; I claim that this result solves the Evidential Problem discussed by Myrvold. I will also briefly discuss the relationship between this approach and the Everettian \'derivation of the Born rule\' due to Deutsch and Wallace.

Much of the evidence for quantum mechanics is statistical in nature. Close agreement between Born-rule probabilities and observed relative frequencies of results in a series of repeated experiments is taken as evidence that quantum mechanics is getting something --- namely, the probabilities of outcomes of experiments --- at least approximately right. On the Everettian interpretation, however, each possible outcome occurs on some branch of the multiverse, and there is no obvious way to make sense of ascribing probabilities to outcomes of experiments. Thus, the Everett interpretation threatens to undermine much of the evidence we have for quantum mechanics. In this paper, I will argue that the Everettian evidential problem is indeed one that Everettians should take seriously, and explain why, in order to deal with it successfully, it is necessary to go beyond existing approaches, including the Deutsch-Wallace decision-theoretic approach.

The most common objection to the Everett view of QM is that it \'cannot make sense of probability\'. The \'Oxford project\' of writers such as Deutsch, Wallace, Saunders and Greaves seeks to meet this objection by showing that the Everett view allows some suitable analogue of decision under uncertainty, and that probability (or some suitable analogue of probability) can be understood on that basis. As a pragmatist, I\'m very sympathetic to the idea that probability in general needs to be understood in terms of its links with decision; but I\'m sceptical about whether the Everett picture provides a suitable analogue of decision under uncertainty. In this talk I\'ll try to justify my scepticism.

Orthodox thinking about chance, choice and confirmation is a philosophical mess. Within the many-worlds metaphysics, where quantum chanciness engenders no uncertainty, these things come out at least as well, if not better.

In \'Everett Speaks\' I will detail Everett\'s involvement in operations research during the Cold War. He was, for many years, a major architect of the United States\' nuclear war plan. I will talk about his family life and his personal decline. We will hear a portion of the only tape recording of Everett in existence, in which Everett and Charles Misner talk about the origin of the Many World\'s interpretation--twenty years later at a cocktail party.

A fundamental question for Everettians is whether they can formulate a many-worlds interpretation of quantum theory which explains why, amongst all possible types of intelligent creature with all possible types of evolutionary and experimental history, we find ourselves among those whose histories apparently confirm Copenhagen quantum mechanics. Since the theory clearly allows that we could have found ourselves otherwise, the answer has to be probabilistic. Everettians then need to supply some account of how probability is or can be attached to an apparently deterministic theory. I argue that it cannot arise through any of the various notions of subjective uncertainty advocated by Saunders, Wallace, Vaidman and Greaves, among others, since none of these notions are valid. Nor could an adequate notion of probability be inferred from any account of the “caring weights” that we, as hypothetical rational Everettian agents, should use when considering the welfare of our Everettian successors – even if a unique rational strategy were to exist, which I argue is not the case. A proposition of the desired form could simply be postulated, but at the price of reducing the entire interpretation – probability postulate, preferred basis, and interpretation of basis states --to unsupported and maybe untestable hypotheses about consciousness. On the brighter side, I describe a new proposal for solving the measurement problem by proposing a definite mathematical structure for possible branching worlds, which appears to have the advantages claimed for Everettian ideas (in particular, respect of Lorentz invariance and of conservation laws), without the interpretational and scientific difficulties. I also note one subtle distinction between Everettian and standard accounts of evolution, which implies that (in principle, albeit not necessarily in practice) it could be possible to distinguish between the theories.

One of the most remarkable features of our quantum universe is the wide range of time, place, scale, and epoch on which the deterministic laws of classical physics apply to an excellent approximation. This talk reviews the origin of such a quasiclassical realm in a universe governed fundamentally by quantum mechanical laws characterized by indeterminacy and distributed probabilities. We stress the important roles in this origin played by classical spacetime, coarse-graining in terms of approximately conserved quantities, local equilibrium, and the initial quantum state of the universe. The discussion is carried out first in the decoherent (or consistent) histories formulation of the quantum mechanics of closed systems (most generally the universe) assuming spacetime geometry is fixed. This is an Everettian generalization of the usual Copenhagen text book quantum mechanics of measurement situations that assumes the quasiclassical realm. Conversely we isolate the assumptions and approximations necessary to derive Copenhagen quantum mechanics from the more general quantum mechanics of closed systems. We describe a further generalization of usual quantum theory that is necessary to deal with quantum spacetime and describe under what conditions it predicts our observed coarse-grained classical spacetime that is a prerequisite for a quasiclassical realm.

I analyze a series of common objections to Everett\'s Many Worlds Interpretation. I discuss which ones are unique to quantum mechanics, and which have nothing to do with quantum mechanics per se as they can also be debated in the context of other areas of physics

Probability is often regarded as a problem for the many-worlds interpretation: if all branches of the splitting wavefunction are equally real, what sense does it make to say that the branches have different probabilities? In the decision-theoretic approach due to Deutsch and Wallace, probabilities acquire a meaning through the preferences of a rational agent. This talk reviews the decision-theoretic approach to probability in classical physics and quantum mechanics and shows that its application to the many-world interpretation creates a new difficulty for the latter.

I will rehearse and try to sharpen some of the perennial worries about making sense of probabilities in Everettian interpretations of quantum mechanics, with particular attention to the recent Decision-Theoretic proposals of Deutsch and others

I will review the current state of the probability problem. My main focus will be on the attempts by David Deutsch and myself to provide a proof of the Born Rule starting from Everettian assumptions, but I will also attempt to locate these attempts within the more general framework of the probability problem.