In the Dialogues Hume attaches great importance to an objection to the design argument which states, negatively, that from phenomena which embody evil as well as good there can be no analogical inference to the morally perfect deity of traditional theism and, positively, that the proper conclusion as regards moral character is an indifferent designer. The first section of this paper sets out Hume's points, and the next three offer an updating of Hume's objection which will apply to Swinburne's Bayesian form of the design argument. The final section concludes that Hume's objection, suitably developed, holds against most of the main theistic arguments, even in their Bayesian form.

In response to Michael Bradley, I summarize my account of the criteria by which the various data of natural theology increase the probability of theism and together make it probable. I explain the sense in which a simpler theory leaves less to be explained, justify my claim that God's perfect goodness is entailed by his other divine properties, and show that not merely is theism simpler than Bradley's ‘Epicurean hypothesis’, but that the ‘mixed’ data of natural theology are more to be expected given theism than given the ‘Epicurean hypothesis’.

Augustine is commonly considered the greatest early proponent of what we call the free-will defence, but this idea is deeply misleading, as Augustine grew increasingly dissatisfied with the view from an early point in his career, and his later explorations of the implications of his doctrines of sin and grace led him to reject free-will theodicies altogether. As a compatibilist, however, he continued to reject the idea that God is responsible for the advent of evil. His alternative was his often misunderstood claim that the primal sin had a ‘deficient’ cause, together with a version of what Alvin Plantinga has nominated the ‘felix culpa’ approach. Thus, Augustine was actually the free-will defence's first major Christian detractor, and by the end of his career he had become its greatest critic.

According to a widespread tradition in philosophical theology, God is necessarily simple and eternal. One objection to this view of God's nature is that it would rule out God having foreknowledge of non-determined, free human actions insofar as simplicity and eternity are incompatible with God's knowledge being causally dependent on those actions. According to this view, either (a) God must causally determine the free actions of human agents, thus leading to a theological version of compatibilism, or (b) God cannot know, and thus cannot respond to, the free actions of human agents. In the present paper, I argue that one can consistently maintain that God is not causally dependent on anything, even for His knowledge, without being committed to either (a) or (b). In other words, an eternal God can know the free actions of agents even if libertarianism is true.

Theists typically think it appropriate to pray to God in the hope that He will thereby intervene in affairs. On the other hand, such prayer is often held to be quite inappropriate for pantheists; a view endorsed by many pantheists themselves. This paper argues for the exact opposite of these positions. It is maintained not only that pantheism can make sense of petitionary prayer but that, despite initial appearances to the contrary, classical theism can not.

Theories of the human person differ greatly in their ability to underwrite a metaphysics of resurrection. This paper compares and contrasts a number of such views in light of the Christian doctrine of resurrection. In a Christian framework, resurrection requires that the same person who exists on earth also exists in an afterlife, that a post-mortem person be embodied, and that the existence of a post-mortem person is brought about by a miracle. According to my view of persons (the constitution view), a human person is constituted by – but not identical to – a human organism. A person has a first-person perspective essentially, and an organism has interrelated biological functions essentially. I shall argue for the superiority the constitution view as a metaphysical basis for resurrection.

Robert Oakes has argued that theism defeats the ‘doctrine of public-world fallibilism’. That is, Oakes has argued that theism supports infallibilism about public-world beliefs such as ‘There is an olive on the floor’, or ‘I have two hands’. Given the enormous discussion of radical scepticism in the recent epistemological literature, this argument is well worth investigating. In this short note, however, I argue that the argument Oakes presents is unconvincing. The truth of theism does not support public-world infallibilism.

This essay constitutes a brief response to Martin Blaauw's paper, ‘Divorcing theism from infallibilism’, Religious Studies, 43 (2007), 349–354, in which he raises a number of thought-provoking objections to an earlier paper of mine which appeared in this journal: ‘Theism and infallibilism: a marriage made in heaven?’, Religious Studies, 40 (2004), 193–201. In the following counter-response, I hope to show that the argument developed in my original essay manages to survive his objections. (All page references in the text are to Martin Blaauw's paper above.)