¹ For an argument to this conclusion, see Crisp, Oliver D., ‘Augustinian Universalism’, in International Journal for Philosophy of Religion 53 (2003), pp. 127–45CrossRefGoogle Scholar.

² The best known historical example of this is, of course, Friedrich Schleiermacher, who argued for theological determinism and universalism. See Schleiermacher, The Christian Faith, trans. and ed. H. R. Macintosh and James Stewart (Edinburgh: T&T Clark, 1928), pp. 550–1. See also Reitan, Eric, ‘Sympathy for the Damned: Schleiermacher's Critique of the Doctrine of Limited Salvation’, Southwest Philosophy Review 18 (2002), pp. 201–12CrossRefGoogle Scholar. There is a lively discussion in the current biblical, theological and philosophical literatures on the viability of universalism. Representative samples include Bell, Richard H., ‘Rom 5.18–19 and Universal Salvation’, New Testament Studies, 48 (2002), pp. 417–32CrossRefGoogle Scholar; Keith DeRose, ‘Universalism and the Bible’ located at: <http://pantheon.yale.edu/~kd47/univ.htm>; MacDonald, Gregory, The Evangelical Universalist (Eugene, OR: Wipf & Stock, 2006)Google Scholar; and Parry, Robin and Partridge, Chris (eds), Universalism: The Current Debate (Carlisle: Paternoster, 2003)Google Scholar.

³ This claim should be distinguished from Barth's view that Christ is the Elect One and humanity as a whole is somehow derivatively elect ‘in’ him. My claim here is just that the set of the elect could comprise one member only: Christ. Then no one else would be elect; everyone else would be damned. See Karl Barth, CD II/2, p. 123, and Crisp, Oliver D. ‘On Barth's Denial of Universalism’, Themelios 29 (2003), pp. 18–29Google Scholar.

⁴ Compare Trevor Hart who cites with approval Richard Bauckham as saying, ‘Only the belief that ultimately all men will be saved is common to universalists. The rationale for that belief and the total theological content in which it belongs vary considerably.’ Themelios 4 (1979), p. 49, cited in Hart, ‘Universalism: Two Distinct Types’, in Nigel Cameron (ed.), Universalism and the Doctrine of Hell (Carlisle: Paternoster, 1992), p. 2. It is true that there are numerous arguments for the conclusion of universal salvation. But my interest is in the logical form of these arguments, not in the variety of arguments themselves. Different arguments for universalism share a common logical form.

⁵ Of course, universalism is ‘particularist’. A universalist could claim that salvation is particularly and effectively applied to all humanity. But theologians usually reserve the term ‘particularist’ for those who maintain that salvation refers to an elect less than the totality of humanity. This theological convention governs the way the term is used in what follows.

⁶ See Kvanvig, Jonathan L., The Problem of Hell (New York: Oxford University Press, 1993), p. 74Google Scholar, for these characterisations of necessary and contingent universalism.

⁷ Hick, John, Death and Eternal Life (London: Collins 1976), pp. 242–3Google Scholar, author's emphasis.

⁸ Some theological libertarians argue that God can foreknow the libertarian choices of human creatures, and bring about a world wherein his creatures make only those choices he foresees will fit with his divine plan. Were this true, the theological libertarian could respond to Hick that God can be sure of the foreseen outcome of creaturely choices even when those choices are contingent. But Michael Rea has recently called this reasoning into question. He argues that libertarians who are presentists about the nature of time (as most theological libertarians are) are faced with a dilemma: either the future is open (God does not know the future choices of free creatures) or no one is free in the way libertarians understand freedom. See Rea, Michael C., ‘Presentism and Fatalism’, Australasian Journal of Philosophy 84 (2006), pp. 511–24CrossRefGoogle Scholar. Whether Rea is right or not, a theologically libertarian contingent universalist is still subject to the problem of modal masking, enunciated by Kvanvig. And this counts against a theologically libertarian universalism being truly universalist.

⁹ For Augustinians in the Reformed tradition, this is only one aspect of what is called distributive justice, that is, divine justice as it bears upon human sin. The other aspect of distributive justice is renumerative justice, which has to do with the fixing of desert, by God. For standard recapitulations of this point, see Shedd, William G. T., Dogmatic Theology, 3rd edn, ed. Gomes, Alan (Philipsburg, NJ: Presbyterian and Reformed, 2003), pp. 293 ff.Google Scholar, and Berkhof, Louis, Systematic Theology (Edinburgh: Banner of Truth, 1958 [1939]), p. 75Google Scholar.

¹⁰ For more on this and the other comments on the three parts of the restricted elect view, compare Oliver Crisp, D., ‘Augustinian Universalism’ and ‘Divine Retribution: A Defence’, Sophia 42 (2003), pp. 35–52CrossRefGoogle Scholar.

¹¹ A qualification: some Augustinian theologians have claimed God could have forgiven sin according to his absolute power (roughly, his power, all things considered), but not according to his ordained power (roughly, his power given his decree to create the world). The early John Owen, and William Twisse, prolocutor to the Westminster Assembly, both held this view.

¹² It is sometimes thought that Augustinians believe only a mere remnant of humanity will be saved. This is not true of all Augustinians. Some are much more sanguine about the scope of election, believing the vast majority of humanity will be saved. See, for example, Warfield, Benjamin, The Plan of Salvation, rev. edn (Grand Rapids, MI: Eerdmans, 1975), pp. 97–8Google Scholar. Compare Paul Helm's essay ‘Are they Few that be Saved?’, in Cameron (ed.), Universalism and the Doctrine of Hell for more on this question.

¹³ This point is made in Crisp, ‘Augustinian Universalism’.

¹⁴ Some Augustinians, particularly some Calvinists, have claimed to be libertarians. But I shall not discuss this possibility here. Interested readers should consult William Cunningham's essay ‘Calvinism and the Doctrine of Philosophical Necessity’, in his The Reformers and the Theology of the Reformation (Edinburgh: Banner of Truth, 1967 [1862]).

¹⁵ I realise that some philosophers and theologians claim that the Bible does not espouse particularism. Most recently, Thomas Talbott has made this claim. However, I do not think that this is terribly plausible. For an example of a New Testament scholar who agrees with me in this matter, see I. Howard Marshall's essay, ‘The New Testament does Not Teach Universal Salvation’, in Parry and Partridge (eds), Universal Salvation.

¹⁶ Compare John Owen, The Death of Death in the Death of Christ, in The Works of John Owen, vol. 10, ed. William H. Goold (Edinburgh: Banner of Truth, 1967 [1850–3]), p. 242: ‘In brief; an intention of doing good unto any one upon the performance of such a condition as the intender knows is absolutely above the strength of him of whom it is required, – especially if he know that it can no way be done but by his concurrence, and he is resolved not to yield that assistance which is necessary to the actual accomplishment of it – is a vain and fruitless flourish’.

¹⁷ Reprinted in Peterson, David et al. (eds), Philosophy of Religion: Selected Readings (New York: Oxford University Press, 1996), p. 511Google Scholar. I say Rahner endorsed something like this in his version of inclusivism because it is not clear that his ‘Anonymous Christianity’ doctrine actually entails that those with no knowledge of Christ will be saved. He seems to think that there is a sort of subconscious or non-conscious knowledge of Christ that means some righteous pagans will be saved, even if that knowledge is never exercised in conscious faith in Christ in this life. I thank Revd Dr Richard Fermer and Dr Karen Kilby for pointing this out to me in conversation.

¹⁸ Objection: Karl Rahner's doctrine is hardly the same as the sort of Calvinism espoused by someone like Owen, and no Calvinist would espouse such a doctrine. Response: consider the case of the nineteenth-century American Presbyterian divine, William G. T. Shedd, who in his book Calvinism Pure and Mixed (Edinburgh: Banner of Truth, 1986 [1893]), ch. 12, states that a person might have a purely dispositional faith and yet be saved. This sounds remarkably similar to Rahner's suggestion, and Shedd is certainly a doctrinally orthodox Calvinist. He is also not a universalist.

¹⁹ See, for example, Shedd, Calvinism Pure and Mixed, ch. 12, and Jonathan Edwards, Miscellany 393 in The ‘Miscellanies’ (Entry Nos, a–z, aa–zz, 1–500), The Works of Jonathan Edwards, vol. 13, ed. Thomas A. Schafer (New Haven, CT: Yale University Press, 1994). There is considerable scholarly dispute over how to interpret Edwards on this matter. For representative examples of both sides of the dispute, see Morimoto, Anri, Jonathan Edwards and the Catholic Vision of Salvation (University Park, PA: Pennsylvania State University Press, 1995)Google Scholar and Bombaro, John J., ‘Jonathan Edwards’ Vision of Salvation’, Westminster Theological Journal 65 (2003), pp. 45–67Google Scholar.

²⁰ This is just what Shedd claims. Of course, there are other limit cases too. How can severely mentally handicapped people be saved, or children who die in childbirth? If the Augustinian believes that (at least some) such persons will be amongst the elect, then conscious faith in Christ cannot be a requirement for salvation, contra Owen.

²¹ Shedd, Calvinism Pure and Mixed, pp. 128–9, emphasis original.

²² Owen, Works, vol. 10, p. 249.

²³ The problem for Arminianism as Owen sees it is that if unbelief is a sin (as Arminians would suppose), Christ must atone for it. But then, his atonement is effective for all he came to save. There is not metaphysical room, so to speak, for the Arminians to say that Christ's death is potentially but not actually effective, because this equivocates on whether or not Christ has actually died for the sin of unbelief or not. If he has died for unbelief, as well as all the other sins of those for whom he came to die, then either all are saved (irrespective of whether they believe or not), or only some are saved – and this is brought about by faith. Hence, Owen retains the RF condition and a particular (limited) atonement.

²⁴ Neither Shedd nor Edwards maintain that dispositional faith is given to all humanity; they are Augustinian particularists, like Owen. But that does not prevent us from using the notion of dispositional faith in support of Augustinian universalism.

²⁵ See Swinburne, Richard, The Existence of God, rev. edn (Oxford: Oxford University Press, 1991), pp. 113ffCrossRefGoogle Scholar.

²⁶ Thomas, Summa Contra Gentiles, I. 81. 4. Katherine Rogers’ discussion of this issue is helpful. See Perfect Being Theology (Edinburgh: Edinburgh University Press, 2000), pp. 103–4.

²⁷ Summa Contra Gentiles, I. 38.

²⁸ For this way of thinking about divine metaphysical and moral goodness, see Morris, Thomas, ‘The Necessity of God's Goodness’, in Anselmian Explorations: Essays in Philosophical Theology (Notre Dame, IN: University of Notre Dame Press, 1987), p. 43Google Scholar.

²⁹ There are other aspects to divine grace that are morally relevant here. For instance, Robert Adams says that grace means ‘a disposition to love which is not dependent on the merit of the person loved’: ‘Must God Create the Best?’ repr. in The Virtue of Faith (New York: Oxford University Press, 1987), p. 56.

³⁰ I assume this for the sake of argument and because it seems that a number of writers in this area have similar sorts of intuitions about the number of people saved in a given possible world. But it might be that this intuition is false. For what does ‘objectively morally better’ consist in? Perhaps a world where there are n+1 people saved and some damned has comparable moral worth, or, at least, is of equal moral worth to a universalist world since such a world would be glorifying to God and this is what gives a particular possible world its objective moral worth in the first place.

³¹ For a defence of the claim that God has no obligations, see Thomas V. Morris, ‘On Duty and Divine Goodness’, in Anselmian Explorations. Morris claims that God is in a state of maximal ‘axiological goodness’ (p. 40), a state according to which an agent necessarily acts in accordance with moral principles and freely engages in acts of grace.

³² Robert Adams, ‘Must God Create the Best?’ in The Virtue of Faith, p. 53.

³³ Ibid., p. 53, slightly adapted for present purposes.

³⁴ In other words, each member of the elect is elect at every possible world in which they exist, hence transworld election or reprobation. The notion of ‘transworld’ depravity (and election) is borrowed from Alvin Plantinga's discussion of the Free Will Defence in The Nature of Necessity (Oxford: Oxford University Press, 1974), ch. 9. This notion has been taken up by libertarian philosophical theologians in dealing with the soteriological problem of evil. See, for instance, William Lane Craig, ‘“No Other Name”: A Middle Knowledge Perspective on the Exclusivity of Salvation through Christ’, Faith and Philosophy 6 (1989), pp. 172–8. The use made of the notion here is, of course, different from Craig's account, which is non-Augustinian in nature.

³⁵ See Geach, Peter, Providence and Evil (Cambridge: Cambridge University Press, 1977), pp. 148–9Google Scholar.

³⁶ Is this revision of condition (c) a departure from traditional Augustinian particularism? I do not think so. It is, if anything, a clarification of the nature of distributive justice on the Augustinian way of thinking. Some Augustinians may dissent from this revision. If so, they may still avail themselves of the modified Adamsian argument outlined above.

³⁷ Thanks to Daniel Hill, Robin Parry and especially Paul Helm for comments on earlier drafts of this article. An ancestor of this article was read at a conference on eschatology held under the auspices of the Research Institute for Systematic Theology, King's College, University of London, in 2004. I am grateful to those present who commented on that version of the article, particularly Stephen Williams.