¹ See, for example, Matt. 11. 27 and 20. 23, Mk. 13. 32, Jn. 4. 34 and 17. 24.

² Several authors have rejected this interpretation of the Trinity – Relative Trinitarianism, or ‘RT’ – simply because it is incompatible with the Indiscernibility of Identicals. See, among others, David, Wiggins, Sameness and Substance (Oxford: Blackwell, 1980), pp. 37–42;Google ScholarThomas, Morris, The Logic of God Incarnate (Ithaca: Cornell, 1986), pp. 26–9;Google Scholar and Cornelius, Plantinga Jr, ‘Social Trinity and Tritheism’, in Feenstra, Ronald J. and Plantinga, (eds.), Trinity, Incarnation, and Atonement: Philosophical and Theological Essays (Notre Dame: University of Notre Dame, 1989), p. 40.Google ScholarInwagen, Peter van develops a monumental alternative logic for RT in ‘And Yet There Are Not Three Gods, But One God’, in Morris, (ed.), Philosophy and the Christian Faith (Notre Dame: University of Notre Dame, 1988), pp. 241–78.Google ScholarRichard, Swinburne believes that RT is excessively mysterious: see ‘Could There Be More Than One God?’, Faith and Philosophqy, V (1988), p. 240,Google Scholar note 19. I have argued that either RT is incoherent or it entails that all three persons of the Trinity became incarnate in Jesus; see ‘The Plight of the Relative Trinitarian’, Religious Studies, XXIV (1988), pp. 129–55.Google Scholar

³ Although the Social Theory of the Trinity is not always characterized this way – it is often taken to mean simply that the Trinity consists of three distinct, fully divine persons, which is formally consistent with the contention that there is some sortal other than ‘person’ under which Father, Son, and Holy Spirit are numerically the same – many recent writers have explicitly or implicitly formulated the Social Theory as I do. See, for instance, Morris, , The Logic of God Incarnate, p. 210et passimGoogle Scholar; Plantinga, Ibid. section II; Swinburne, Ibid. section IV. I shall therefore presume that my usage is (at the very least) acceptable.

⁴ My statement of this problem is slightly unusual – normally it is presented as a difficulty for belief in more than one omnipotent being. But some philosophers claim that almightiness and omnipotence are distinct properties, and that while almightiness is a theologically crucial characteristic of God, the idea of divine omnipotence is self-contradictory, or at least manifestly untenable from the Christian point of view. See, for example, Peter, Geach, ‘Omnipotence’, Philosophy, XLVIII (1973), pp. 7–20.Google Scholar While I believe that Flint, Thomas P. and Freddoso, Alfred J. have shown that these claims are mistaken – see their ‘Maximal Power’, in Freddoso, (ed.), The Existence and Nature of God (Notre Dame: University of Notre Dame, 1983), pp. 81–114Google Scholar – I will suppose that Geach's position is defensible. For whether or not this is the case, the problem I have stated will be just as serious for the Social Trinitarian. As we shall see, the difficulty arises because the Social Theory maintains, in accordance with theistic tradition, that the will of a divine being is essentially unimpedable – and both an essentially omnipotent and an essentially almighty being must have an unimpedable will.

⁵ I have borrowed (and somewhat abridged) this argument from Swinburne, Ibid. p. 230.

⁶ Brown, , The Divine Trinity (London: Duckworth, 1985), p. 293.Google Scholar

⁷ For the sake of simplicity, I shall assume throughout this paper (unless I specify otherwise) that all of the members of the Trinity are essentially temporal. Although this assumption is controversial, none of my arguments would lose any of their force if they were rephrased so that they are compatible with the doctrine of divine timelessness – though the result would often be rather cumbersome.

⁸ See ‘The New Paradox of the Stone’, Faith and Philosophy V (1988), pp. 283–90,Google Scholar especially pp. 285 ff. In fact, the authors of this paper say that one of them (unidentified) is not sure whether any individual has properties that are de re essential to it (see their note 3). They actually argue that it is possible for two omnipotent beings, be they essentially or accidentally omnipotent, to will contradictory states of affairs.

⁹ As an example of a very influential Christian theologian who construes almightiness in this fashion, here is Augustine: … God is called almighty for no other reason than that he can do whatever he wills and because the efficacy of his omnipotent will is not impeded by the will of any creature (Enchiridion XIV, 96).

¹⁰ I have presupposed for the sake of readability that simultaneous causation is logically possible: otherwise, nothing could satisfy this definition. However, even if simultaneous causation is logically impossible, my analysis could be repaired along lines suggested by Swinburne, Ibid. p. 238, note 7.

¹¹ This argument was inspired by a briefer one in William, Wainwright's ‘Monotheism’, in Robert, Audi and Wainwright, (eds.), Rationality, Religious Belief, and Moral Commitment (Ithaca: Cornell, 1986), pp. 302–3,Google Scholar note 21.

¹² Ibid. p. 232.

¹³ The Doctrine of the Trinity (London: Nisbet, 1943), p. 156.Google Scholar

¹⁴ Ibid. p. 283.

¹⁵ The Author of the Appeal, The Trinitarian Controversy Reviewed, etc. (London: 1760), p. 53;Google Scholar quoted in Hodgson, Ibid. p. 223.

¹⁶ See, for instance, Swinburne, ibid. pp. 231 and 236.

¹⁷ De Synodis 37. See also Hilary, , De Trinitate III, 4;Google Scholar Gregory of Nyssa, Contra Eunomium VII; and Augustine, , Contra Maximinum III, 7Google Scholar and De Diversis Quaestionibus 83, question 50.

¹⁸ Arius' writings survive only in quotations or summaries in the works of others, but it is easy to document all of the claims I have attributed to him. See, for example, the following passages in Athanasius: Contra Arianos 1, 5–6; De Synodis 15; and Epistola ad Episcopos Aegypti 12.

¹⁹ Ibid. p. 226.

²⁰ It must be admitted, however, that the traditional Christian should not feel entirely serene about rejecting monarchism. Ever since the Church codified its Trinitarian teaching in the fourth century, she has maintained that the Father is the fountainhead of the Son and the Spirit. Several clauses of the Creed of Constantinople, for example, make this explicit; and as Kelly, J. N. D. rightly says, this creed ‘is one of the few threads by which the tattered fragments of the divided robe of Christendom are held together’ (Early Christian Creeds, 2nd edn (London: Longman, 1960, p. 296)).Google Scholar Trinitarian republicans usually claim that theologically conservative Christians can abandon monarchism without diminishing the plausibility of the other central contentions of traditional Christianity. Unfortunately, I cannot pursue this matter here, but see Brown, Ibid. pp. 101–215.

On the other hand, since Trinitarian monarchism is incoherent, the Eastern and Western Churches can easily reconcile their ancient dispute concerning the procession of the Holy Spirit: both East and West should agree that the Spirit eternally proceeds from no-one.

²¹ Responsibility and Atonement (Oxford: Clarendon, 1989), p. 66.Google Scholar

²² Ibid. p. 64.

²³ ‘Could There Be More Than One God?’, pp. 228–29.

²⁴ For example, some of the later New Testament authors urgently warn congregations to beware of so-called Christians who deny that Jesus Christ is come in the flesh: cf. 1 Jn. 2. 18–25 and 4. 2–3, and also 11 Jn. 8. And these writers could scarcely have exaggerated the danger: as Kelly, J. N. D. notes, Gnosticism ‘came within an ace of swamping the central tradition’ (Early Christian Doctrines, 5th edn (London: Black, 1977) p. 142.Google Scholar

²⁵ Wisdom of Solomon 4. 7, 10–11.

²⁶ It was Edward Wierenga who first called this point to my attention. See The Nature of God (Ithaca: Cornell, 1988), pp. 132–3,Google Scholar note 35.

²⁷ My definition of an initial-segment conditional of freedom is roughly modelled on William Hasker's definition of an ‘initial-segment counterfactual’, though my concept differs somewhat from his. See Hasker, , God, Time, and Knowledge (Ithaca: Cornell, 1989), p. 32.Google Scholar

²⁸ If the members of the Trinity are temporal and simultaneous causation is logically impossible, then it will be tricky to specify just what it is for them to agree a decision-sharing proposal ‘immediately’. For (as I point out later) Trinitarian time would be dense – between any two instants of Trinitarian time there would be another instant. But we can get by without a more rigorous notion of immediacy: let us simply suppose that it means ‘in the smallest possible number of steps’.

²⁹ See, for example, Morris, , ‘The Necessity of God's Goodness’, in Anselmian Explorations (Notre Dame: University of Notre Dame, 1987), pp. 42–69Google Scholar (especially pp. 56–63). Philosophers who have tried to deduce God's essential perfect goodness from his other attributes include Swinburne, , The Existence of God (Oxford: Clarendon, 1979), pp. 97–102,Google Scholar and Keith, Ward, Rational Theology and the Creativity of God (Oxford: Blackwell, 1982),Google Scholar chapters 6 and 8.

³⁰ Philosophers who have tried to show that essential perfect goodness is incompatible with other properties typically attributed to God include Nelson, Pike, ‘Omnipotence and God's Ability to Sin’, American Philosophical Quarterly VI (1969), pp. 208–16;Google ScholarBruce, Reichenbach, Evil and a Good God (New York: Fordham, 1982),Google Scholar chapter 7; and Davis, Stephen T., Logic and the Nature of God (London: Macmillan, 1983),CrossRefGoogle Scholar chapter 6. For an excellent critique of their arguments, see Morris, The Logic of God Incarnate, chapter 5.

³¹ Morris argues as follows: his intuitions about God generate without intentional contrivance an overall belief-set in which it makes sense that there should be such intuitions and that they should be, at least a core of them, reliable. For if… [such a] God exists and creates rational beings whose end is to know him, it makes good sense that they should be able to come to know something of his existence and attributes without the need of highly technical arguments, accessible to only a few (Ibid. p. 134; italics his). But imagine that, although God is not essentially perfectly good, he is in fact perfectly good: and call this sort of deity, for obvious reasons, a ‘Cartesian’ God. If a Cartesian God exists and creates rational beings whose end is to know him – and as long as he is in fact perfectly good, he will surely not want systematically to deceive or mislead us about his own nature – he would have just as much reason as Morris' God for providing these beings with a non-technical method for discerning the modalities of his most important attributes. And he could do so, of course, by implanting in his rational creatures a natural tendency to believe spontaneously that his almightiness is incompatible with essential perfect goodness – though to be sure, Morris is evidence enough that this tendency is not irresistible! Therefore, insofar as Morris' modal intuitions about God are ‘self-confirming’, so are the Cartesian's.

³² I find it strange that although Morris recognizes this point – see ‘The God of Abraham, Isaac, and Anselm’, Faith and Philosophy 1 (1984), pp. 177–87Google Scholar – he does not exploit it to argue in any of his writings for the necessary goodness of God, even though, as I have noted above, he believes that God's necessary goodness is consistent with the other divine attributes and yet all attempts to derive the former from the latter are failures.

³³ For an especially lucid version of this argument, see Mackie, J. L., ‘Evil and Omnipotence’, Mind LXIV (1955), p. 209.CrossRefGoogle Scholar

³⁴ See The Nature of Necessity (Oxford: Clarendon, 1974), p. 184.Google Scholar Plantinga's formulation of this response to the atheistic argument (Ibid. pp. 180–5) is standard.

³⁵ De Trinitate III, 2. I acknowledge a debt to the French translation by Gaston, Salet, published as La Trinité (Paris: Cerf, 1959).Google Scholar

³⁶ A Social Trinitarian who endorses this kind of argument must maintain that no member of the Trinity is self-sufficiently perfect. For each of them has at least one perfection, e.g. supreme charity, whose instantiation depends in part upon the existence of someone else. Indeed, each of them depends for at least one of his perfections on the will of at least one other being. For according to Richard, a person (call her ‘A’) is supremely loving only if she loves another person of equal worth (call her ‘B’). But B is equal in worth to A only if B loves A in return. So the perfect charity of a Trinitarian person depends upon the charity of another, and thus on a voluntary act of another.

Note, however, that the Social Trinitarian could still easily distinguish his view from monarchism. For although the supreme charity of any member of the Trinity would depend (in part) on at least one of the others' loving him, no Trinitarian person could generate another by loving him, or prevent another from existing at all by refusing to love him. Therefore, all of them remain equally sovereign.

It might be objected that this view is incoherent. Each member of the Trinity causally depends upon the volitions of another for one of his essential properties – the property of being supremely loving. But no member of the Trinity causally depends upon the volitions of another for his very existence. And of course an essential property of something is a property which it cannot lack. But if each of the members of the Trinity depends on another for a property that he cannot exist without having, how could each of them also be self-existent? Is it not a necessary truth that whatever is dependent on something for one of its essential properties is also dependent on that same thing for its very existence? However, while I believe that this principle holds true if the dependent entity is a created object, it quickly leads to absurdity if extended to the Trinity. For not only will the Son be dependent on the Father for his very existence, the Father will be dependent on the Son for his very existence. And since the relation of depending on something's causal activity for one's very existence is transitive, it will follow that the Son is dependent for his very existence on his own causal activity.

³⁷ Mark Wynn is responsible for a number of significant improvements in this paper.