Challenges the widely held thesis that scientific explanations are arguments by posing three questions that seem to raise difficulties for it: Why are irrelevancies harmless to arguments but fatal to explanations? Can events whose probabilities are low be explained? Or, to reformulate essentially the same question, is genuine scientific explanation possible if indeterminism is true? Why should requirements of temporal asymmetry be imposed upon explanations but not upon arguments? In addition to showing the untenability of the “third dogma,” this chapter (...) signals the development of a causal theory of explanation that will supplement the simple statistical‐relevance model of explanation advocated in earlier works by the author. (shrink)