In this paper, I develop some internal problems with Alvin Plantanga's proper functionalist epistemology. I focus on: (1) how we know that a belief is the result of proper function and the special difficulties this occasions for religious beliefs; (2) what a properly functioning person should believe in various circumstances, and (3) the problem of design – whether the claim that God designed us can be reconciled with the claim that He was subject to trade-offs, compromises and unintended by-products. These serious internal problems cast doubt upon proper functionalism's fruitfulness as a theory of knowledge.

Moses and Monotheism is Freud's last book on religion. It was published in its entirety only after his flight from Nazi-occupied Vienna. Moses is perhaps Freud's most controversial book on religion. It is both an apology and a curse. It is a critique of traditional Judaism (by way of an Oedipal analysis of a deified Moses), a defence of a modern humanistic Judaism (a Judaism of moral and intellectual values), and a bitter critique of Christianity (a religion not of the Father but of the sons, a failed religion, which expresses its failure in anti-Semitism). A defiant and rebellious book, one might say, one in which an old man rises to meet his fate and does so with surprising wilfulness and vigour. But Moses also reveals a dogmatic Freud defending a critique of religion that is intellectually flawed and politically misdirected.

In the past few years, the focus of arguments against theism has shifted. Where previously the existence of evil has been thought by many demonstrative of the impossibility of God's existence, now it is frequently purveyed as merely evidence against the existence of a Supreme Being. Even this more modest claim has been forcefully denied by William Alston and Peter van Inwagen. I argue that their arguments are not persuasive. Not only do they suffer logical flaws but, if accepted, actually have pernicious effects on the values of reasoning and religious practice.

I take up Richard Swinburne's point, in his Responsibility and Atonement, Ch. 5, that although the past cannot be changed, wrongdoers may change its significance by ‘disowning’ their actions through atonement, just as their victims may do so through forgiveness. I argue that the point can and should be pressed much more strongly than it is by Swinburne within the terms of his own discussion; and that it has a much wider significance, transcending that discussion, for there is a constant interplay between events, human actions, and our retrospective assessment of the past. Finally, I look tentatively at the question in an eschatological perspective.

God is conceived in the Western theistic tradition to be both the Creator and Conservor of the universe. These two roles were typically classed as different aspects of creation, originating creation and continuing creation. On pain of incoherence, however, conservation needs to be distinguished from creation. Contrary to current analyses (such as Philip Quinn's), creation should be explicated in terms of God's bringing something into being, while conservation should be understood in terms of God's preservation of something over an interval of time. The crucial difference is that while conservation presupposes an object of the divine action, creation does not. Such a construal has significant implications for a tensed theory of time.

The doctrine of Middle Knowledge presupposes that conditionals of freedom (statements of the form ‘If A were circumstances C, he would perform X’) can be true. Such conditions are, where true, not true in virtue of the truth of any categorical proposition. Nor can their truth be modelled in terms of comparative similarity of possible worlds, because the structure of possible worlds is a necessary one, whereas the connection between antecedent and consequent of a conditional of freedom is a contingent one. Lewis and Stalmaker are committed to ‘conditional fatalism’, the view that things only would go a certain way if they would have to go that way. Although commitment to conditional fatalism does not itself import a commitment to fatalism, it is hard to find a separate motivation for it.

The problem of infant suffering and death is one of the most difficult versions of the problem of evil, especially when we consider how God can be thought good to the infant victims by the infant victims. In the first portion of this paper, I examine two theodicies that aim to solve this problem but fail. In the final section, I argue that the problem can be better dealt with by maintaining not that God must redeem the suffering of such children, but that such children are not the sort of beings whose suffering God can or must redeem.

God is good, God is just, God is almighty: only a madman doubts this… Doubtless when their elders suffer these afflictions we are wont to say either that their goodness is being tested…or that their sins are being punished. But these are older people. Tell me what we are to answer about children!

St Augustine, in a letter to Jerome