Richard Otte

Recent discussion of the problem of evil has centered around what is known as the probabilistic or evidential argument from evil. According to this argument the evil in our world is evidence against the existence of God, even though evil is logically consistent with God’s existing. Based on this it is claimed it is irrational to believe one of the traditional theistic religions, unless there is overwhelming positive evidence to counter this negative evidence. One of the most important and widely discussed versions of this argument is due to Paul Draper.1 In this paper I will look at Draper’s argument and argue that he has made a simple fundamental error; as a result his argument is irrelevant to most theists. After discussing this error in Draper’s argument, I will discuss probabilistic arguments from evil from the perspective of confirmation theory. The error in Draper’s argument is easily made and could occur in any probabilistic argument from evil; looking at confirmation theory and probabilistic arguments from evil will provide insight into reasoning about evil and belief in God.

In The Miracle of Theism Mackie attempts to defend Hume's argument concerning the rationality of accepting a miracle on the basis of testimony. He does this by first offering a precise account of what miracles and laws of nature are, and then by claiming that this implies that any evidence for a law of nature is also evidence against the miracle occurring. I argue that Mackie has committed a simple logical fallacy. Given Mackie's account of miracles and laws of nature, it does not follow that our evidence for natural laws makes it unlikely that a miracle occurred.

Charles Chihara has presented a problem he claims Bayesian confirmation theory cannot handle. Chihara gives examples in which he claims the change in belief cannot be construced as conditionalizing on new evidence. These are situations in which the agent suddenly thinks of new possibilities. I propose a solution that incorporates the important ideas of Bayesian theory. In particular, I present a principle which shows that the change of belief in Chihara's example is due to simple conditionalization.