Does the granting of free will demand not only that moral evil must be possible, but also that natural evil exist? A positive answer to this question has been offered by Richard Swinburne, who has argued that an agent cannot have free will (in the relevant moral sense) without knowledge of how to bring about moral evil, and that this knowledge must arise, ultimately, from experience of natural evil. 1 By contrast, Steven Bo~r has claimed that neither the possibility of inflicting moral evil nor the existence of natural evil is required for the possession of free will. z
Recently, Paul Moser has criticised Swinbume's way of arguing for the connection between natural evil and free will, and has offered an alternative of his own, thus supporting the conclusion but not the argument. 3 Moser's alternative proposal hinges on the claim that natural evil is necessary for the acquisition of the concept of evil, without which we could not have free will, i.e., the ability to make free choices between good and evil. This view has some affinity with a rebuttal of BoOr which I have presented elsewhere 4 and which has since come under attack from Frank Dilley s and, most recently, Robert McKim.6 The aim of the first part of this paper is to offer a qualified defence of Swinburne against's Moser's objections. The second part is a response to McKim and, in particular, to an argument in his paper which, if sound, would undermine, not only my own method of rebutting BoOr, but both Moser's and Swinburne's positions also. Having responded to, and hopefully met this objection. I shall, in the final section, evaluate all three approaches insofar as they meet the challenge of the extent of natural evil in the world.
Swinburne's argument, as summarised by Moser, runs as follows: