[In the following excerpt, Campbell asserts that critics of Anselm's ontological argument have misrepresented his point, which is simply to demonstrate “that it cannot be said that God is not.”]

The study of Anselm's Proslogion argument on the existence of God which I recently undertook¹ emerged out of a growing conviction that commentator after commentator had been guilty of serious misrepresentation of its structure. Traditionally, Anselm has been taken as presenting in Proslogion 2 the first version ever to be formulated fully of the ‘Ontological Argument’. The reasoning in Proslogion 2 was—and often still is—supposed to proceed from an alleged definition of God as something-than-which-nothing-greater-can-be-thought, through inferences designed to show that, unless he exists in reality, he would not be something than which nothing greater can be thought, to the conclusion that God exists.

In more recent years attention has swung (especially amongst philosophers) to the first half of Chapter 3, where a number of commentators claim to find a second, logically independent, and sounder version of the Ontological Argument. This ‘modal’ version is supposed to proceed likewise from the alleged definition, through inferences derived from an alleged premise that necessary existence is a perfection, to the conclusion that God exists.

My bold claim is that all these commentators are demonstrably wrong. It is not often in philosophical theology that one is justified in using language of such finality, but in this case the matter is provable using the apparatus of modern symbolic logic. Once all the text of Chapters 2 and 3 is taken seriously and literally, it is possible to extract a series of premises and sub-conclusions which together constitute a three-stage argument for two conclusions.

The two conclusions which Anselm explicitly draws at the end of this formally valid chain of reasoning are that God ‘so truly is that he cannot be thought not to be’ and that God ‘is something than which a greater cannot be thought’. Thus, far from this identification of God with ‘such a nature’ being a definition—it would be ill-formed if it were—which Anselm uses as a premise in some version of the Ontological Argument, it is demonstrably one of his conclusions. So the long tradition of interpretation which stretches from Aquinas to many contemporary commentators must be wrong. This startling result—it still surprises me—is what has led me to reappraise the character of Anselm's thought.

As I read him, Anselm begins Proslogion 2 with the prayer that he might be given understanding of two matters: that God is just as we believe, and that he is what we believe. He then says what we believe God to be: something-than-which-nothing-greater-can-be-thought. But since he is then struck by doubt, for the fool has said in his heart that God is not, he inquires whether there is such a nature. Chapter 2 finishes by drawing the sub-conclusion that something-than-which-a-greater-cannot-be-thought—which he does not yet understand to be God—exists in reality. That is Stage One of the overall argument. Stage Two, which occupies the first half of Chapter 3, then consists of a deduction that this same nature cannot be thought not to be. Here again God does not explicitly figure, nor is Anselm concerned with necessary existence as we might understand it.

In the second half of Proslogion 3 Anselm swings back into the language of address, but that does not stop him from reasoning. As I read him, he there presents two independent pieces of argumentation, both of which move him on to his two main conclusions. Together, these comprise what I call Stage Three. The first leg of it is based on the new premise that ‘if any mind could think of something better than you, the creature would rise above the creator and judge the creator, which is absurd’. The second, which is the neater and more compelling of the two, is based on the premise that ‘whatever else there is, except for you alone, can be thought not to be’. In the weak form that everything that is not identical to God can be thought not to be, this is sufficient, together with his two previous sub-conclusions, to entail his two main conclusions. Read this way the entire argument emerges with a logical tightness and coherence that it is difficult to dismiss.

On this construction, the argument proper begins with Anselm's sentence ‘when the fool hears what I say, “something-than-which-nothing-greater-can-be-thought”, he certainly understands what he hears’. This makes the point de départ of the argument Anselm's use of that indefinite description in a communicative way. Understood in this fashion, Anselm's concern is whether he can continue to use with understanding the language in which he articulates his belief. His own assessment of what the argument shows is that ‘no-one understanding what God is can think that he is not’.

In this regard, it seems to me to differ significantly from, for example, the Five Ways of Aquinas. The latter presents us with arguments which are supposed to be proofs in the sense that, if the arguments were sound, the conclusion that God exists would follow in a way which is thoroughly impersonal. That is, those arguments are based on certain facts about the natural world, and their cogency is in no way dependent upon what any person might say or think. In contrast, Anselm's argument crucially involves the believer who speaks of something-than-which-nothing-greater-can-be-thought, and the fool who overhears his prayer and who is able to think of such a thing. In so far as Anselm sets out from his own speech-act, an act of communication, and not from an impersonal proposition, neither is his conclusion impersonal. His entire enterprise is directed towards drawing out the implications of the language used in articulating his faith in order to determine whether it can be said with understanding.

Because that is the character of the argument, it seems to me that we should give a cautious and qualified assent to Anselm Stolz's claim that Anselm has not presented a proof of the existence of God,² even though the proposition ‘God exists’ is deducible validly from his premises. The purport of his argument, as I read it, is that the whole realm of discourse which admits the believer to speak of God as something-than-which-nothing-greater-can-be-thought rules out the possibility of denying with understanding the existence of God.

Yet it would be a mistake to take this assessment of the force of the argument to mean that Anselm's investigation is thoroughly internal to a closed circle of faith. The first two stages of the argument are expressed in language even the fool can understand, and they can be examined for their cogency independently of any peculiarly religious premise. Rather, I suggest that we are now in a position to see how the opposed interpretations of Anselm's approach—rationalistic or fideistic—rest on a common exegetical mistake.

If the argument did proceed from the identification of God with this nature, from the alleged definition, the methodological dispute would turn on how that ‘definition’ was derived. If it simply states a necessary truth, either of linguistic usage or of quasi-Platonic ideas, and if his conclusion simply were the proposition ‘God exists’, the argument could legitimately be classified as thoroughly rationalistic. Alternatively, if the alleged definition were either a datum of revelation or an article of faith, the entire argument would inevitably be faith-dependent. Significantly, neither of these interpretations gives Anselm's ‘what I say’ any logical role in the argument. And both lose their purchase once it is recognised that this identification is not a premise but a conclusion.

Nevertheless, it might be objected, does not Anselm's ‘what I say’ mean that what he is doing is drawing out the implications of the language of faith, so that even on my account he is a kind of fideist? To which I answer: Yes, he is drawing out the implications of the language of faith; but No, it does not follow that he is therefore a kind of fideist. This is where the role of the fool is crucial. To see this, we need in our imaginations to reconstruct the scene of the Proslogion, for the work is not to be taken for a metaphysical treatise cast in the language of prayer just for rhetorical effect. It has an implicitly narrative element which is crucial to its structure.

Anselm is praying, addressing God. He confesses the belief as to what God is, one of the beliefs he is yet to understand: ‘we certainly believe that you are something than which nothing greater can be thought’. Then he is struck by doubt as he recalls the fool's denial. (It is hard to stop the mind wandering in prayer!) ‘But is there any such nature?’ he asks himself. To deal with this, he breaks off from the language of prayer, and switches to third-person grammar; the God he was addressing—‘You’—temporarily drops out of his speech. Instead he envisages that the fool has been eavesdropping on his prayer, and has actually heard him speaking of such a nature. Since the fool does not believe that there is a God to whom prayer could ever be addressed, an analysis of God-talk would be to no avail. But—and this gets to the heart of Anselm's methodology—the language of faith is not entirely esoteric. Believers both use distinctively religious referring expressions and use words which are part of the public language, shared and understood by believers and unbelievers alike. So even the fool understands what he hears when Anselm uses ordinary familiar words in speaking of such a nature. And it is those public words whose proper usage Anselm then proceeds to investigate.

Once he has established that there is such a nature, and that such a thing cannot be thought not to be, but only then, can be resume his prayer: ‘And this is you, O Lord our God.’ Then, continuing his prayer, he offers his two reasons for thus identifying such a nature with the God to whom he is praying. I can only think that it has been an unconscious prejudice against its being possible to think logically while one is praying which has prevented so many commentators from recognising that this is where Anselm's argument reaches its climax.

If one reads the text carefully, one can see how delicately Anselm has structured his argument through his oscillation between second- and third-person discourse. His conclusions are not impersonal factual propositions. But he arrives at them through consideration of those public words which can feature in the oratio recta even of the fool. Yet he shows no sign of thinking that his conclusions ought to convince, or even silence, the fool. He only adverts to the fool again in order to ask how the fool could have said what has emerged as, strictly speaking, not intelligible.

I suspect that part of our difficulty in grasping the exact character of Anselm's method is that, for complicated reasons which lie in our own intellectual history, we want justified general propositions to underpin our belief-structures. Anselm, wisely in my view, does not push his argument to that point. He is content to establish to his own satisfaction that it cannot be said that God is not. But that is not equivalent to, nor does it entail, that one should say that God is.

His same concern with what one says, with the commitments of one's discourse, can be seen in his Reply to Gaunilo, who presumed to speak On Behalf of the Fool. Anselm prefaces his Reply with the words:

Since it is not the fool, against whom I spoke in my tract, who takes me up, but one who, though speaking on the fool's behalf, is not a fool and is a Catholic, it will suffice if I reply to the Catholic.

That is, Anselm insists that what is said must be ‘owned’. That is the same point as I was alluding to above when I spoke of the oratio recta of the fool. In Anselm's view, claims must be owned; he will not indulge in idle speculation or intellectual games. Nor will a general analysis serve to compel belief. Analysis must start out from what is said, and all it can prove are the ontological commitments inherent in a domain of discourse. This, I believe, is why he said: ‘I do not understand in order to believe, but I believe in order to understand.’

Notes

1. cf. my From Belief to Understanding (Canberra: The Australian National University, 1976).

2. ‘Anselm's Theology in the Proslogion’, trans. by A. C. McGill in The Many-Faced Argument, ed. by J. Hick and himself (London: Macmillan, 1968).