Anthropic Explanation of Temporal Asymmetry in Everettian Quantum Mechanics

[Richard Corry]

Abstract. Our experience of the world is temporally asymmetric: coffee cools, glasses smash, people age and die, while the reverse processes never occur. One challenge for temporally symmetric theories, therefore, is to explain the asymmetry of experience. The usual solution is to posit special initial and/or final conditions for the universe. This solution, however, amounts to putting in the asymmetry by hand, and one may be left looking for a further explanation of the asymmetric boundary conditions. In this (still very rough) paper I explore another possible solution, namely that the asymmetry of experience can be explained anthropically, via an observer selection effect. In particular, such an explanation may be available in a purely symmetric version of the Everettian interpretation of quantum mechanics that makes use of the formalism of consistent histories.

[Robert Griffiths]

Two comments.

First. I think it only creates confusion to put consistent histories in the Everett/many worlds camp, since in one respect they are directly opposite, as I pointed out in my first paper in 1984. Quantum time development in consistent histories is stochastic; in the Everett approach it is deterministic. For a more recent perspective on consistent histories, see my presentation at this workshop.

Second, I would think that your anthropic argument works just as well in a classical world in which entropy is increasing. Quasiclassical frameworks are then useful in telling us that such arguments, whatever their validity, need not change if we believe the world is quantum mechanical, for classical mechanics is a quite adequate approximation to quantum theory in addressing questions of this sort.

Bob Griffiths

[Richard Corry]

Dear Bob (if I may?)
Thanks for your comments.
With regard to your first point, I didn’t mean to put the consistent histories approach in the Everettian camp. I see them as two separate interpretations of the formalism of quantum mechanics. My point is, rather, that many Everettians make use of some of the concepts and formalism from the consistent histories approach.
With regard to the second point, I agree that an anthropic argument for temporal asymmetry is possible in a classical world in which entropy is increasing. Indeed this is the approach that Boltzmann took. His idea was that given enough time we should expect low entropy regions of the universe to form purely through random fluctuation away from equilibrium. We can then employee anthropic reasoning to explain why we find ourselves on the upward entropy slope of one of these low entropy regions. As is well known, however, this anthropic explanation fails: entropy fluctuations down to a minimum equal to the entropy of the world we see around us now are overwhelmingly more likely than entropy fluctuations whose minimum entropy is anything like what we believe to have been the case in the distant past. The anthropic reasoning above, then, leads to the conclusion that it is overwhelmingly likely that the entropy fluctuation in which we exist bottomed out just moments ago, and all the records we appear to have of a lower entropy past are false. This conclusion is not only unwelcome, it is self-undermining since it takes away any justification for believing the theories we used to reach the conclusion.
The general lesson we can take away from this is that when providing an anthropic explanation it is not enough to show that some property X should be expected in some part of the cosmos and then use anthropic reasoning to explain why we find ourselves in such a place. The distribution of X must also be appropriate in some sense. In Boltzmann’s case the problem was that regions of sufficient low entropy were more likely to have false records of a lower entropy past than true. The anthropic explanation I am suggesting in the Everettian case does not appear to have this problem. While I still have trouble getting my head around the meaning of probabilities in EQM, there does not seem to be any suggestion that one set of consistent histories is “more probable” than another.

Cheers,
Richard.

[Author]

Posts