by John M. Frame

[“Ontological Argument,” for IVP Dictionary of Apologetics.]

The most influential formulation of this argument (though he did not use the term ontological) is found in the first three chapters of the Proslogium of Anselm of Canterbury (1033-1109). He had earlier written aMonologium in which he considered many arguments for God’s existence. But then, he says, ‘I began to ask myself whether there might be found a single argument which would require no other for its proof than itself alone; and alone would suffice to demonstrate that God truly exists…; and whatever we believe regarding the divine being’ (Preface, p. 1). He says that he made an extensive search for such an argument, and, when he was almost ready to discontinue his quest, that argument ‘began to force itself on me, with a kind of importunity’ (p. 2).

The Proslogium, unlike the Monologium, is a prayer. In the first chapter, Anselm invokes God’s presence, confessing God’s incomprehensibility and his own sin. He concludes with these famous words, ‘I do not endeavor, O Lord, to penetrate thy sublimity, for in no wise do I compare my understanding with that; but I long to understand in some degree thy truth, which my heart believes and loves. For I do not seek to understand that I may believe, but I believe in order to understand [credo ut intelligam]. For this also I believe, — that unless I believed, I should not understand’ (pp. 6-7).

The second chapter begins the argument that seemed to force itself on Anselm. In accord with the resolution of his prayer, that he seeks to believe in order to understand, he begins with a Christian belief: ’And, indeed, we believe that thou art a being than which no greater can be conceived’ (p. 7).

Now Anselm recognizes that some do not believe in such a God, like the fool in Psm. 14:1 who ‘says in his heart ‘There is no God.’ Nevertheless, this fool at least understands the words ‘a being than which no greater can be conceived,’ so we may say that in a sense this being ‘exists in the [fool’s] understanding’ (p. 8). But if it exists in the fool’s understanding alone, and not in reality, then we can imagine a greater being, namely one that exists, not only in the understanding, but in reality. So then the being in the fool’s understanding is not really a being than which no greater can be conceived. So a being that truly meets Anselm’s definition of God, a being than which no greater can be conceived, must exist not only in the understanding, but also in reality. Therefore God must exist, by virtue of his very definition.

In Chapter 3, Anselm draws the further implication that this God ‘cannot be conceived not to exist’ (p. 8). That is, if God can be conceived not to exist, it would be possible for us to conceive of a still greater God, one that cannot be conceived not to exist. So God not only exists, he exists necessarily, as some later philosophers and theologians would put it. He doesn’t just happen to exist; he must exist. Once we know the meaning of God, as Anselm has defined it, we cannot conceive of him not existing. For it is greater, better for him to exist than not to exist, and to exist necessarily rather than contingently.

Then through the rest of the book, Anselm seeks to prove the traditional attributes of God using the same method: God is ‘just, truthful, blessed, and whatever it is better to be than not to be’ (p. 11).

We can simplify Anselm’s argument, for ease of reference: (1) God has all perfections; (2) Existence is a perfection; (3) Therefore, God exists.

At first glance, many immediately suspect a fallacy. I recall a party game in which one friend produced a proof that 1=2 and challenged us all to find out what was wrong with it. (Turned out there was a concealed division by zero.) Similarly, one suspects a bit of conceptual legerdemain in Anselm’s argument. Can it really be this easy to prove the existence of God? But it has not been easy for philosophers and theologians to show where the fallacy is located, if indeed there is one. Anselm’s contemporary Gaunilo, Thomas Aquinas, David Hume, Immanuel Kant, J. L. Mackie, and others have rejected the argument, but many philosophers down to the present have accepted versions of it: Descartes, Spinoza, Leibniz, Hegel and his followers, and twentieth-century thinkers Charles Hartshorne, Norman Malcolm, and Alvin Plantinga.

The Proslogium includes an Appendix ‘In Behalf of the Fool’ by the monk Gaunilo, with a response by Anselm. Gaunilo points out that it is doubtful that we can even conceive of God in our minds according to Anselm’s definition: for who can conceive of a being than which no greater can exist? And if we can reason from concept to reality as Anselm does, we could as easily prove the existence of a perfect island. ‘For if it does not exist, any land which really exists will be more excellent than it…’ (p. 151).

Anselm answers Gaunilo at considerable length. Although Gaunilo speaks ‘In Behalf of the Fool,’ Anselm knows he is ‘by no means a fool, and is a Catholic.’ So, he says, ‘I think it sufficient that I answer the Catholic’ (p. 153). In response to Gaunilo’s first point, Anselm replies that Gaunilo, being a Catholic, cannot deny that God is conceivable, for he himself conceives of God. Anselm rather brushes off Gaunilo’s analogous proof of a perfect island, but his basic reply is that such an island could not fit his definition of God, that than which no greater can be conceived. Only one being meets the terms of that definition, namely the God of Christianity. The rest of Anselm’s reply discusses various senses of ‘conceiving,’ ‘understanding,’ and ‘existing,’ and relations among these concepts.

Immanuel Kant thought that Anselm misunderstood the nature of existence, by treating it as a perfection of God. In Kant’s view, existence is not a perfection, not even a property. It is not, indeed, a “real” predicate, though it can occupy the predicate position in a sentence like ‘God exists.’ For existence, Kant said, doesn’t add anything to our concept of something. If you conceive of a nondescript car and add to it the color blue, your concept changes. But if you conceive of a car and then conceive that same car as existing, nothing changes, Kant argued, for it is, after all, the same car. As Kant puts it, ‘a hundred real dollars do not contain a penny more than a hundred possible dollars’ (Kant Critique of Pure Reason, p. 282). So Kant thinks Anselm has erred by making existence one of God’s attributes or properties.

However, Kant admits that his financial position is better with real dollars than with possible ones. And we know that a real car is different from an imaginary one, and that a real unicorn, if it existed, would be different from an imaginary one. The real car may look the same as the one in our head, but it is certainly something different. Existence is, therefore, different from other properties and predicates in some ways, but not in the sense that it makes no difference to the objects that have it. Thus it seems that Kant’s objection to the ontological argument fails, though it has generated and continues to generate much discussion.

The most common objection to the argument, voiced by Aquinas, followed by many others, is that concepts in the mind imply only mental existence, never existence in reality. There can be no “leap” from mind to reality. This argument invokes our intuition that we can think of many things, like unicorns and leprechauns, that do not exist in reality, and it is hard to conceive of anything in such mental concepts that in itself could prove that these objects exist in the real world. Anselm, however, does not say that it is generally valid to infer realities from concepts. For him, this inference is valid in only one case, the case of God. It is not valid for our concepts of unicorns or perfect islands, only for that being than which no greater can be conceived.

Now if it is never possible to argue from the contents of the mind to the nature of reality, then we are in a bad way. In one sense, the contents of our minds (including the experience of our senses, our rational reflection, our memories, imaginations, and concepts) are all we are directly acquainted with. If we are never able to reason from any of these to conclusions about the real world, then we cannot know the real world at all; we are shut up to skepticism. Empiricists, rationalists, idealists and others propose various ways of drawing this inference. (Insofar as Kant denied its possibility, he implicated himself of the charge of skepticism.) But the inference must be drawn.

Anselm’s own inference may owe something to Plato, for whom objects of our experience are reflections of more perfect objects, Forms or Ideas. We have a concept of goodness, for example, though nothing in our experience is perfectly good. Therefore, Plato believed, there must be a Perfect Good in the real world, which serves as a model, criterion, or standard of goodness. Although we may well reject Plato’s idea that we know the Perfect Good from having experienced it in a past life, it still makes sense to assert that the highest criteria of truth, beauty, and goodness must exist in reality, not only in our minds. Else we could not measure these qualities except by a subjective (and therefore arbitrary) standard. If goodness, truth, and beauty exist, there must be an objective standard by which to measure them.

To say with Anselm that God is that than which no greater can be conceived is to identify God as the highest perfection, the standard and exemplar of all greatness, and therefore of all goodness, truth, and beauty, and whatever other perfections there may be. Without such a standard or exemplar, there could be no goodness, truth, or beauty in the world; that is, the world would be a chaos. So there is a kinship between the ontological argument and the transcendental argument (q.v.). Both argue that if God exists only in our minds, there is no truth or meaning, indeed, no being at all. The greatness of God, therefore, must necessarily exist.

The ontological argument, therefore, expresses for Anselm the heart of the Christian worldview. God is the source of all value, so his existence must be presupposed if we are to accept the existence of anything else. It is not surprising, then, that this argument arises in answer to prayer and is expressed in the language of prayer, and it is not surprising that when Gaunilo raises objections Anselm responds, not to the fool, but to the Catholic. As he says in his Preface, he is not trying to understand in order to believe, but to believe in order to understand. When he discovers in a deeper way who the God of the Bible really is, that than which no greater can be conceived, he sees an important reason why he must exist.

The trouble with the argument is that people with other worldviews try to use it too. The God proved by Spinoza’s version of the argument is very different from Anselm’s, a God identical with nature, Deus sive Natura. The same may be said of the Absolute of Hegel and the process God of Hartshorne. In part the differences lie in the fact that different worldviews differ as to what is great or perfect. For Anselm, it is a perfection for God create all things from nothing (Proslogium, pp. 10-11), but not for Spinoza. For Anselm, it is a perfection for God to be passionless (p. 11), but not for Hartshorne. The ontological argument necessarily presupposes a system of values. For Anselm, that system comes from his understanding of the Christian faith. In that sense, the argument presupposes the Christian revelation, which, again, should not be surprising in view of Anselm’s prayers and the credo ut intelligam.

As Thomas Aquinas says, not everybody would acknowledge God to be “that than which no greater can be conceived,” for some, he says, have thought that God has a body. (Summa Theologiae, First Part, Q. 2, Art. 1, Obj. and Ans. 2). Nor would some acknowledge that existence is a perfection, even given that it is a ‘real predicate:’ to many Buddhists, for example, annihilation is preferable to existence.

So the cogency of the ontological argument as an apologetic for the Christian faith depends on the cogency of the Biblical system of values, its notion of perfection. It is not a religiously neutral argument, but one that immediately assumes the truth it seeks to validate. Presuppositional apologists (q.v.) frankly acknowledge and defend that kind of circularity in apologetics. Others may reject the ontological argument for this reason. But they must ask whether other arguments are not circular in similar ways. Does the cosmological argument not presuppose a causal order such as we find in Scripture, but not in David Hume? Does the teleological argument work unless it understands purpose to be personal, rather than impersonal?



Anselm of Canterbury, St. Anselm: Basic Writings, ed. by S. N. Deane (La Salle, IL: 1962).

Aquinas, Thomas, Summa Theologica in Anton C. Pegis, ed., Introduction to St. Thomas Aquinas (N. Y.: 1948).

Barth, Karl, Anselm: Fides Quaerens Intellectum (Richmond: 1960). Expounds the theological presuppositions of Anselm’s argument.

Frame, J. M., Apologetics to the Glory of God (Phillipsburg, N. J.: 1994).

Kant, Immanuel, Critique of Pure Reason, abridged, ed., tr., int.  by Norman Kemp Smith (N. Y.: 1958).

Plantinga, Alvin, God, Freedom and Evil (N. Y.: 1973). Includes a contemporary reconstruction of the argument according to modal logic and possible worlds.