Sharpening the Electromagnetic Arrow(s) of Time

The various “arrows of time” and their interrelations are the subject of a seemingly never ending discussion in the physics and the philosophy of science literature. While the discussion in recent decades has undoubtedly produced numerous advances in the details of our understanding, it is hard not to be discouraged by the overall lack of progress in reaching a consensus on key issues.1 In addition to ruing the lack of progress, one could also make two general complaints about the arrows of time literature. First, one could complain (as several authors have) that the talk of “arrows of time”suggests that what is at issue is the directionality of time whereas what is often at issue is not directionality but temporal asymmetries. This is a defect that could be corrected by a change of terminology, but since talk of “arrows”has become well entrenched, I will continue to use it. What is not a matter of terminology, however, is whether the various asymmetries are merely asymmetries in time or whether they constitute asymmetries of time itself. That there are di¤erences of opinion on this matter is not surprising. But what is disconcerting is the lack of agreement on how this matter is to be decided.2 Second, it is especially vexing that the typical way of stating the puzzle about various temporal arrows or asymmetries rests on a false presupposition. The general