Abstract

William Rowe has put forward four popular evidential arguments from evil. I argue that there was already a prominent distinction between logical and evidential arguments from evil—the IN-IM-distinction, and that its adoption leads to two important results. First, all three non-Bayesian evidential arguments are actually not evidential but logical, while the Bayesian evidential argument genuinely evidential. Second, and most importantly, Rowe’s Bayesian evidential argument is redundant, in the sense that it has the same diculties his three non-Bayesian arguments have. His move from the three earlier non-Bayesian arguments to the Bayesian argument is futile.