Jeffrey Bub

We present a translation of Poincaré's hitherto untranslated 1912 essay together with a brief introduction describing the essay's contemporary interest, both for Poincaré scholarship and for the history and philosophy of atomism. In the introduction we distinguish two easily conflated strands in Poincaré's thinking about atomism, one focused on the possibility of deciding the atomic hypothesis, the other focused on the question whether it can ever be determined that the analysis of matter has a finite bound. We show that Poincaré admitted the decisiveness of Perrin's investigations for the existence of atoms; he did not, however, anticipate the kind of resolution of which the second question is susceptible in light of recent developments

We define a family of ‘no signaling’ bipartite boxes with arbitrary inputs and binary outputs, and with a range of marginal probabilities. The defining correlations are motivated by the Klyachko version of the Kochen-Specker theorem, so we call these boxes Kochen-Specker-Klyachko boxes or, briefly, KS-boxes. The marginals cover a variety of cases, from those that can be simulated classically to the superquantum correlations that saturate the Clauser-Horne-Shimony-Holt inequality, when the KS-box is a generalized PR-box (hence a vertex of the ‘no signaling’ polytope). We show that for certain marginal probabilities a KS-box is classical with respect to nonlocality as measured by the Clauser-Horne-Shimony-Holt correlation, i.e., no better than shared randomness as a resource in simulating a PR-box, even though such KS-boxes cannot be perfectly simulated by classical or quantum resources for all inputs. We comment on the significance of these results for contextuality and nonlocality in ‘no signaling’ theories

It is generally accepted, following Landauer and Bennett, that the process of measurement involves no minimum entropy cost, but the erasure of information in resetting the memory register of a computer to zero requires dissipating heat into the environment. This thesis has been challenged recently in a two-part article by Earman and Norton. I review some relevant observations in the thermodynamics of computation and argue that Earman and Norton are mistaken: there is in principle no entropy cost to the acquisition of information, but the destruction of information does involve an irreducible entropy cost.

The aim of cognitive neuropsychology is to articulate the functional architecture underlying normal cognition, on the basis of cognitive performance data involving brain-damaged subjects. Glymour (forthcoming) formulates a discovery problem for cognitive neuropsychology, in the sense of formal learning theory, concerning the existence of a reliable methodology, and argues that the problem is insoluble: granted certain apparently plausible assumptions about the form of neuropsychological theories and the nature of the available evidence, a reliable methodology does not exist! I argue for a reformulation of the discovery problem in terms of an alternative characterization of relevant evidence in neuropsychology