Peter Millican Any argument which attempts to prove God’s existence a priori based only on His nature can be termed an “Ontological Argument”. Historically, however, the term is inextricably associated with the famous argument presented in Anselm’s Proslogion chapter II, and with the later variant advanced by Descartes in his fifth Meditation and subsequently developed by Leibniz. Some have claimed that Anselm’s argument was anticipated in the thought either of various classical philosophers (notably Aristotle, Parmenides, Plato, and Zeno of Cition) or of Augustine, but although there are indeed suggestive passages in their writings, Anselm’s explicit “proof” of God’s existence based on his Nature does appear to be a genuinely original discovery.
The scope of Anselm’s argument, and its place within his thought, have been much disputed, with some commentators, notably Barth, interpreting it not so much as an argument for God’s existence starting from a definition of what He is understood to be, but rather as an illumination of God’s existence, starting from a revelation of His nature. Such an interpretation, according to which the argument moves from faith to understanding rather than the reverse, corresponds well with the Proslogion’s original title “Fides quaerens intellectum”, but it threatens to render Chapter II, traditionally seen as the heart of Anselm’s argument, otiose – if one starts from the premise that God has revealed Himself as having a certain nature, then it is hard to see any point to an argument for God’s existence which starts from that apparently question-begging premise. However in Chapter III Anselm goes on to develop his line of reasoning further, arguing that God exists in such a way that His non-existence is inconceivable. Hence if Chapter II is seen not as a self-standing argument but rather as a preliminary for Chapter III, then these might be taken together as an exploration of the character of God’s existence starting from His revealed nature, without rendering Anselm’s reasoning trivial or question-begging. That said, the reason why Anselm’s argument has inspired fascination amongst so many generations of philosophers and theologians is precisely that it does appear to spell out a line of thought which even the unbeliever can follow, starting not from a substantial revelation but merely from the understanding of words, and leading inexorably to God’s real existence. Most have dismissed the possibility of such an a priori proof, but the fascination of the argument remains because its flaws have been so hard to identify.
Anselm begins his argument by considering the Fool of Psalm 14 who “says in his heart ‘There is no God’”, and imagines this Fool hearing and understanding the formula which Anselm puts forward as his own conception of God, namely, as something-than-which-nothing-greater-can-be-thought. His reasoning then proceeds through the following six steps:
(1) The Fool understands the phrase “something-than-which-nothing-greater-can-be-thought”.
(2) Hence something-than-which-nothing-greater-can-be-thought exists at least in the Fool’s mind.
(3) It is greater to exist in reality than to exist in the mind alone.
(4) So if that-than-which-nothing-greater-can-be-thought existed only in the Fool’s mind, then it would be possible to think of something greater (e.g. the same thing existing in reality also).
(5) But this would be a contradiction, since it is obviously impossible to think of something greater than that-than-which-nothing-greater-can-be-thought.
(6) Therefore something-than-which-nothing-greater-can-be-thought must exist both in the Fool’s mind and in reality also.
This short paragraph has generated an enormous volume of commentary and criticism. Conceptual points made against the argument include objections to its alleged neo-Platonic presuppositions; its conflation of the Fool’s understanding of a phrase with the existence of an entity “in the mind”; its attempted comparison between existing and non-existing things; and its treatment of existence as a property which adds to something’s greatness (this last criticism is particularly associated with Kant, though his criticism was directed towards Descartes and Leibniz rather than Anselm). More specifically logical objections include the argument’s slide from the thought-properties of an entity to the existence of a thought-entity with those properties; its use of the formula “that-than-which-nothing-greater-can-be-thought” which seems to presuppose that some entity is thus uniquely referred to; and the alleged illegitimacy of its move from claims within the realm of concepts to claims outside it. However the most influential kind of objection to the argument refrains from identifying any specific flaw, and aims only to show that somethingmust be wrong, on the ground that similar arguments can be contrived for implausible entities such as a most excellent island (suggested by Anselm’s correspondent Gaunilo) or a perfect Pegasus (suggested by Gassendi in criticism of Descartes). Such parodies, together with scepticism about the possibility of a priori proof, have led the vast majority of thinkers to concur in rejecting Anselm’s reasoning even if they have been unable to agree on where exactly it goes wrong. The argument’s notoriety is based not on its plausibility, but on this very lack of agreement.
The reason why Anselm’s reasoning has proved so difficult to pin down seems to be that his formula “something-than-which-nothing-greater-can-be-thought” is subtly ambiguous, depending on what is taken to be the logical scope of the phrase “can-be-thought”. To take two possible interpretations, “something-which-can-be-thought-so-great-that-nothing-can-be-thought-greater” would be a recognisably Godlike nature whose non-instantiation would not, however, be contradictory (that something can be thought supremely great does not imply that it is supremely great); whereas “something-which-is-so-great-that-nothing-that-is-greater-can-be-thought-of” must indeed be instantiated (since by definition the formula then refers to the greatest thing there is, which on Anselmian principles must be an existent thing), but need not be Godlike. Sadly there is no interpretation of Anselm’s formula on which it both guarantees reference to a Godlike nature and also guarantees that the nature in question is instantiated.
More recent types of Ontological Argument lack the subtlety of Anselm’s original and fail more straightforwardly (the best known are collected in Plantinga 1965, while Oddy 1995 provides an exhaustive survey). Descartes’ argument, that God is perfect by definition and so must possess all perfections including existence, succeeds at best in showing that the concept of God is of a being conceived of as existing – as Kant pointed out and Mackie has emphasised more recently, no contradiction is implied by the simple denial that this concept is really instantiated. Twentieth century proponents such as Hartshorne, Malcolm and Plantinga have accordingly turned away from the crude Cartesian pattern, and instead sought inspiration from the principle of God’s necessary existence enunciated in Proslogion III. However the “modal” form of Ontological Argument – that if God is understood as a necessary being then His existence, if possible, must be actual – has an unfortunate mirror-image, in that the sceptic can simply retort that on the same principles God’s non-existence, if possible, must be actual. Hence such an argument remains unable to convince even if its modal principles are accepted: anyone unsure about the existence of a necessary God should be equally unsure about such a being’s mere possibility. As with other versions of the Ontological Argument, therefore, its premises will appear true only to those who are already convinced of God’s existence.
To sum up, the Ontological Argument fails to establish the existence of God. But in Anselm’s version at least it remains a fascinating logical conundrum and can, perhaps, for the believer, serve as a focus for contemplation of the special character of God’s existence.
Anselm (1077-8), Proslogion, translated by M.J. Charlesworth in St Anselm’s Proslogion, 1965, pp.101-55
Barth, Karl (1931), Fides Quaerens Intellectum, translated by Ian Robertson as Anselm: Fides Quaerens Intellectum, 1960