Prisoner's Dilemma and Newcomb's Problem: Two Problems or One?

David Lewis argued that Newcomb's Problem and the Prisoner's Dilemma are "one and the same problem" or, to be more precise, that the Prisoner's Dilemma is nothing else than "two Newcomb problems side by side" (Lewis Philosophy and Public Affairs 8:235-240, 1979: 235). It has been objected that his argument fails to take into account certain epistemic asymmetries which undermine the one-problem thesis. Sobel (1985) acknowledges that many tokens satisfy the structural requirements of both problems, while questioning the generality of the thesis. Bermudez (Analysis 73:423-429, 2013), on the other hand, argues that there is a deeper structural conflict between the two problems in the sense that the epistemic requirements that give Newcomb's Problem its force are precisely those that prevent it from being a Prisoner's Dilemma. I argue that the epistemic asymmetry objections raised by Sobel and Bermudez fail to undermine the one-problem thesis. A different type of objection raised by Bermudez, one which relies on the contrast between parametric and strategic choices, fares no better. Although NP is a problem that has been primarily studied for its implications to decision theory, it contains enough game theoretic elements to justify the claim that a juxtaposition of two such problems can be thought of as a strategic game.