1 I am grateful to Isabel Wollaston and Phillip Goodchild, who have commented on earlier drafts of this paper.

2 Philo describes these as ‘Epicurus’ old questions’.

3 There are a number of distinctions that can be made, both in the way ‘the problem of evil’ is presented and in the kinds of solutions attempted. The problem can be set up as a logical one (is the claim that God is perfectly good, omniscient and omnipotent logically compatible with the existence of evil) or as an evidential one (does the existence of a large quantity of pointless evil count as evidence against belief in God, rendering the proposition that God exists less probable). The problem can furthermore be presented atheistically (as something which counts either decisively or significantly against belief in God) or aporetically (as a puzzle to be pondered by believers). Those who attempt answers, finally, can try to give an account of God and the world that genuinely explains the existence of evil, or more modestly, can offer arguments to the effect that evil does not rule out or render improbable the existence of Cod. In Plantinga’s presentation, only the former are properly called theodicies: the latter, which he engages in, are ‘defences’. I am for the most part using the term theodicy in a broader sense, to cover all attempted answers to the so-called problem of evil. Cf. the Introduction of Marilyn McCord Adams and Robert Merihew Adams, eds., The Problem of Evil (Oxford: Oxford University Press, 1990) for a fuller discussion of some of the distinctions I have mentioned. For the most part such distinctions will not be central to this discussion. It might be supposed that the kind of defence Plantinga develops, which does not try to make any claims about the way the world actually is, but only about how, logically, it might be, is less vulnerable to some of the criticisms developed below, but it does not entirely escape them.

4 To put this more technically, philosophers of religion distinguish between ‘restricted’ and ‘expanded’ theism. Christianity is one kind of ‘expanded’ theism.

5 It might be objected that there are prominent examples of pre-Enlightenment Christian thinkers quite happy to discuss God in a similar abstraction–Thomas Aquinas in significant portions of the Summa Theologiae, for instance. It is beyond the scope of this piece to go into the question fully, but I think a case can be made that in spite of superficial similarities, Aquinas and others like him were in fact engaged in a very different kind of project–in terms of its context, purpose, presuppositions, and overall shape–than Enlightenment figures or most contemporary philosophy of religion. For a related argument, see Nicholas, Wolterstorff, “The Migration of the Theistic Arguments: From Natural Theology to Evidentialist Apologetics,” in Rationality, Religious Belief & Moral Commitment, edited by Audi, R. and Wainwright, W.J.. (Ithaca: Cornell University Press, 1986)Google Scholar. Cf. also Fergus, Kerr, Afier Aguinus: Versions of Thamism (Oxford: Blackwell, 2002)Google Scholar.

6 In the case of Surin the attack is limited to what he terms theoretical theodicies—by contrast the so-called practical theodicies he considers more legitimate. This is a distinction to which we shall return.

7 Cf. for instance ‘Horrendous Evils and the Goodness of God‘ in The Problem of Evil, and especially the book of the same title, Horrendous Evils and the Goodness of God (Ithaca and London: Cornell University Press, 1999)Google Scholar.

8 This is a paraphrased version of her somewhat more technical definition, given on p. 26 of Horrendous Evils und the Goodness of God.

9 ‘Horrendous Evils and the Goodness of God‘, pp.211-212.

10 Horrendous Evils, p. 27.

11 I use the word ‘usual’ here because Marilyn McCord Adams’ own positive proposals need to be exempted. Adams presents her work as a solution to the logical problem of evil, and so in some broad sense one might include it as a theodicy. She is insistent, however, in refusing to attempt to find a ‘morally sufficient reason why God would ... permit evils’ (pp. 53-54), but concentrates instead on making the case that God can ‘defeat’ horrors and so be good to individuals.

12 Cf. Kathryn, Tanner, God and Creation in Christian Theology (Oxford: Blackwell, 1988)Google Scholar for an argument concerning the characteristic modern distortion of traditional Christian ways of relating God’s sovereignty to creaturely agency.

13 It should be noted that philosophers of religion are well-aware that a free will defence depends on what is called an ‘incompatibilist’ view of freedom—on the assumption, that is, that freedom is incompatible with determinism—and that there are those who defend the opposite view of compatibilism. While the view that I am sketching here might be taken as a kind of compatibilism, it is not compatibilism in its usual philosophical form. One taking the view I have outlined would, or at least could, still be an incompatibilist within the realm of created causes, insisting that to be free means precisely not to be determined by any created cause. So although I may be proposing a kind of compatibilism, it is not compatibilism as usually conceived.

14 Nicholas, Lash, Believing Three Ways in One God (London: SCM, 1992)Google Scholar, p. 51.

15 There is an exception. As already noted above I have drawn on Marilyn McCord Adams’ criticisms of the theodicies of other contemporary philosophers of religion, but have in fact given no reason to reject her own positive proposals. However, while she is in some sense giving an answer to the problem of evil—offering a way of showing God to be logically compossible with evil—she avoids offering an ‘answer’ in the sense that others do, in that, as noted above, she refuses to answer the question of why God permits evil.

16 Terrence, Tilley, The Evils of Theodicy (Washington, D.C.: Georgetown University Press, 1991), p.250Google Scholar

17 Kenneth, Surin, Theology and the Problem of Evil (Oxford: Basil Blackwell, 1986), p.67Google Scholar.

18 Cf. the thorough and impressive argument in Thomas Weinandy’s book Does God Suffer? (Edinburgh: T&T Clark, 2000)Google Scholar To his arguments I would add the comment that when a contemporary theologian asserts that God suffers, and a traditional theologian asserts that God is impassive, they are not necessarily talking on the same plane, and so the one is not necessarily asserting precisely what the other denies. To maintain that God is impassive and beyond change is, arguably, to maintain that certain categories cannot be used to speak of God at all, rather than to paint a kind of picture of what God is like. It is a grammatical rather than a descriptive affirmation. And therefore it would be equally inappropriate, from such a traditional position, to describe God as static as it would be to describe God as moved. If this is right then contemporary theologians are not so much affirming what earlier theologians rejected, as breaking a grammatical rule to which earlier theologians believed they were constrained to adhere.

19 ‘as a companion in suffering God gave comfort where humanly there was nothing to hope for in that hell’, History and the Triune God (London: SCM, 1991), p. 29Google Scholar.

20 Moltmann does, it must be said, make this comment about God as a companion in suffering giving comfort in the broad context of ‘points which have emerged from Jewish and Christian discussion of theology after Auschwitz’.

21 For related criticisms of Moltmann’s highly influential use of a story of Elie Wiesel, cf. Metz, J-B, ‘Facing the Jews: Christian Theology after Auschwitz’ in Fiorenza, and Tracy, , eds, The Holocaust as Interruption (Edinburgh: T & T Clark, 1984)Google Scholar, cited in Theology and the Problem of Evil, p. 124, and also Marcel, Sarot, ‘Auschwitz, Morality and the Suffering of God’, Modern Theology 7(1991), 135–152Google Scholar.

22 Certainly in the context of human relationships such a move would often appear highly manipulative. Consider the parent who, accused of having done some very particular kinds of harm to his or her children, responds with a discussion of how much he or she has suffered for the children over the course of their lives.

23 Theology and the Problem of Evil, p. 67.

24 In Church Dogmatics III, 3 Karl Barth prefaces his treatment of ‘nothingness’ with a discussion of the brokenness of theology, and the fact that theology must be a report of the way things are which does not ‘degenerate into a system’ (Church Dogmatics III,3 p. 295). He himself seems in danger of doing so, however, in his subsequent discussion of the ‘right’ and ‘left’ hands of God, and in the notion that nothingness is that which God does not will, so that ‘what really corresponds to that which God does not will is nothingness’ (p.352).

25 I am not advocating the assertion of logically incompatible propositions, but rather the holding of a set of beliefs which, somewhat more broadly, we cannot make sense of: There may be some other perspective in which they all make sense together, but if so this is something of which we cannot even begin to conceive.

26 An evil action, an action which lacks some good it should have, comes from a deficient will, which is deficient because it ‘does not subject itself to its proper rule.’(Summa Theologiae I 49 a.1 ad 3). Whatever is good and has being-the will itself-is caused by God, but not its deficiency. Where then does its deficiency come from; why does the will not subject itself to its proper rule? We are given no answer.

27 Carl Jung considered the doctrine of privatio boni ‘a regular tour de force of sophistry’, Collected Works of C.G. Jung, Volume 11, p.313, cited in David, Burrell, AquiPlus: God and Action (London: Routledge and Kegan Paul, 1979)Google Scholar