Ontological Argument for the Existence of God

views updated


The Ontological Argument for the existence of God was first propounded by Anselm (c. 10331109), abbot of Bee and later archbishop of Canterbury, in his Proslogion (Chs. 24) and in his Reply to a contemporary critic.

He begins (Proslogion 2) with the concept of God as "something than which nothing greater can be conceived" (aliquid quo nihil maius cogitari possit, and other equivalent formulations). It is clear that by "greater" Anselm means "more perfect." (Sometimes he uses melius, "better," instead of maius, "greater": for instance, Proslogion 14 and 18.) Since we have this idea, it follows that "Something than which nothing greater can be conceived" at least exists in our minds (in intellectu ) as an object of thought. The question is whether it also exists in extramental reality (in re ). Anselm argues that it must so exist, since otherwise we should be able to conceive of something greater than that than which nothing greater can be conceivedwhich is absurd. Therefore "Something than which nothing greater can be conceived" must exist in reality.

In Proslogion 3 Anselm adds that "Something than which nothing greater can be conceived" exists in the truest and greatest way (verissime et maxime esse ); for whereas anything else can be conceived not to exist (and thus exists only contingently), "Something than which nothing greater can be conceived" cannot be conceived not to exist (and thus exists necessarily). For that which cannot be conceived not to exist is greater than that which can be conceived not to exist, and therefore only that which cannot be conceived not to exist is adequate to the notion of "Something a greater than which cannot be conceived."

Anselm explains (in his Responsio ) that by a being which cannot be conceived not to exist he means one that is eternal in the sense of having no beginning or end and always existing as a whole, that is, not in successive phases. He argues that if such a being can be conceived, it must also exist. For the idea of an eternal being that has either ceased to exist or has not yet come into existence is self-contradictory; the notion of eternal existence excludes both of these possibilities. This latter argument has been revived and developed in our own day (see below).

Many of the earliest manuscripts of the Proslogion contain a contemporary criticism (attributed in two of the manuscripts to one Gaunilo of Marmoutier) together with Anselm's reply. The criticism, summed up in the analogy of the island, is directed against Anselm's argument as presented in Proslogion 2. Gaunilo sets up what he supposes to be a parallel ontological argument for the existence of an island more perfect than any known island: such an island must exist, since otherwise it would be less perfect than any known island, and this would be a contradiction. In reply Anselm develops the reasoning of Proslogion 3. His argument cannot be applied to islands or to anything else whose nonexistence is conceivable, for whatever can be conceived not to exist is eo ipso less than "Something than which nothing greater can be conceived." Only from this latter notion can we (according to Anselm) deduce that there must be something corresponding to it in reality.

Perhaps the most valuable feature of Anselm's argument is its formulation of the Christian concept of God. Augustine (De Libero Arbitrio II, 6, 14) had used the definition of God as one "than whom there is nothing superior." The Ontological Argument could not be based upon this notion, for although it is true by definition that the most perfect being that there is, exists, there is no guarantee that this being is God, in the sense of the proper object of man's worship. Anselm, however, does not define God as the most perfect being that there is but as a being than whom no more perfect is even conceivable. This represents the final development of the monotheistic conception. God is the most adequate conceivable object of worship; there is no possibility of another reality beyond him to which he is inferior or subordinate and which would thus be an even more worthy recipient of man's devotion. Thus metaphysical ultimacy and moral ultimacy coincide; one cannot ask of the most perfect conceivable being, as one can of a first cause, necessary being, unmoved mover, or designer of the world (supposing such to exist) whether men ought to worship him. Here the religious exigencies that move from polytheism through henotheism to ethical monotheism reach their logical terminus. And the credit belongs to Anselm for having first formulated this central core of the ultimate concept of deity.

Descartes's Argument

Anselm's argument was rejected by Thomas Aquinas in favor of the Cosmological Argument and as a consequence was largely neglected during the remainder of the medieval period. It was, however, again brought into prominence by René Descartes in the seventeenth century, and most subsequent discussions have been based upon Descartes's formulation. Descartes made explicit the presupposition of the argument that existence is an attribute or predicate which, like other predicates, a given x can meaningfully be said to have or to lack. He claims that just as the idea of a triangle necessarily includes among the defining attributes of a triangle that of having its three internal angles equal to two right angles, so the idea of a supremely perfect being (a different formula from Anselm's) necessarily includes the attribute of existence. Consequently we can no more think, without contradiction, of a supremely perfect being which lacks existence than of a triangle which lacks three sides.

Descartes considers the following objection: From the fact that in order to be a triangle a figure must have three sides it does not follow that there actually are any triangles; and likewise in the case of the concept of a supremely perfect being. His reply is that whereas the notion, or essence, of a triangle does not include the attribute of existence that of a supremely perfect being does, and that therefore in this special case we are entitled to infer existence from a concept.

Kant's Criticism

Descartes's version of the Ontological Argument had some important contemporary criticsfor example, Pierre Gassendi and Johannes Caterus (Johan de Kater)but the classic criticism is that of Immanuel Kant. This moves on two levels. First, leaving the argument's presuppositions for the moment unchallenged, he grants the analytic connection that Descartes had affirmed between the concept of God and that of existence. In the proposition "A perfect being exists" we cannot without contradiction affirm the subject and reject the predicate. But, he points out, we can without contradiction elect not to affirm the subject together with its predicate. We can reject as a whole the complex concept of an existing all-perfect being.

Second, however, Kant rejects the assumption that existence is a real predicate. If it were a real, and not merely a grammatical, predicate, it would be able to form part of the definition of God, and it could then be an analytic truth that God exists. But existential propositions (propositions asserting existence) are always synthetic, always true or false as a matter of fact rather than as a matter of definition. Whether any specified kind of thing exists can be determined only by the tests of experience. The function of "is" or "exists" is not to add to the content of a concept but to posit an object answering to a concept. Thus, the real contains no more than the possible (a hundred real dollars are the same in number as a hundred imagined ones); the difference is that in the one case the concept does and in the other case it does not correspond to something in reality.

russell's analysis

Essentially the same pointso far as it affects the Ontological Argumentwas made in the twentieth century by Bertrand Russell in his theory of descriptions. This involves an analysis of positive and negative existential propositions, according to which to affirm that x 's exist is to affirm that there are objects answering to the description "x," and to deny that x 's exist is to deny that there are any such objects. The function of "exists" is thus to assert the instantiation of a given concept. "Cows exist" is not a statement about cows, to the effect that they have the attribute of existing, but about the concept or description "cow," to the effect that it has instances. If this is so, then the proper theological question is not whether a perfect being, in order to be perfect, must together with its other attributes have the attribute of existence but whether the concept of an (existing) perfect being has an instance. This question cannot be determined a priori, as the Ontological Argument professes to do, by inspection of the concept of God. The nature of thought on the one hand and of the extramental world on the other, and of the difference between them, is such that there can be no valid inference from the thought of a given kind of being to the conclusion that there is in fact a being of that kind. This is the fundamental logical objection to the Ontological Argument.

Hegelian Use of the Argument

Prior to Kant, the Ontological Argument had been used by Benedict de Spinoza and Gottfried Wilhelm Leibniz. Since Kant, the form of it that he discussed has remained under the heavy cloud of his criticism. However, G. W. F. Hegel and his school put the argument to a somewhat different use. As Hegel himself expressed it, "In the case of the finite, existence does not correspond to the Notion (Begriffe ). On the other hand, in the case of the Infinite, which is determined within itself, the reality must correspond to the Notion (Begriffe ); this is the Idea (Idee ), the unity of subject and object" (Vorlesungen über die Philosophie der Religion, Vol. II, p. 479). Otherwise stated, Being itself, or the Absolute, is the presupposition of all existence and all thought. If finite beings exist, Being exists; when beings think, Being comes to self-consciousness; and in the reasoning of the Ontological Argument, finite thinking is conscious of its own ultimate ground, the reality of which it cannot rationally deny.

The defect of this argument is that its conclusion is either trivial or excessively unclear. It is trivial if the reality of Being is synonymous with the existence of the sum of finite beings; but on the other hand, it is so unclear as to be scarcely interesting if Being is regarded as a metaphysical quantity whose distinction from the sum of finite beings cannot be explicated.

The use of the argument in early twentieth-century French "reflexive" philosophy (see bibliography) has affinities with the Hegelian use.

Contemporary Discussions

Discussion of the Ontological Argument has continued throughout the modern period and is perhaps as active today as at any time in the past. For there is perennial fascination in a piece of reasoning that employs such fundamental concepts, operates so subtly with them, and professes to demonstrate so momentous a conclusion.

Among theologians, attempts have been made to maintain the value of the argument, not as a proof of God's existence but as an exploration of the Christian understanding of God. Thus, Karl Barth regards the proof as an unfolding of the significance of God's revelation of himself as One whom the believer is prohibited from thinking as less than the highest conceivable reality. On this view Anselm's argument does not seek to convert the atheist but rather to lead an already formed Christian faith into a deeper understanding of its object. Again, Paul Tillich treated the theistic proofs as expressions of the question of God that is implied in our human finitude. They analyze different aspects of the human situation, showing how it points to God. Thus, the Ontological Argument "shows that an awareness of the infinite is included in man's awareness of finitude." This is in effect a Hegelian use of the argument.

hartshorne and malcolm

At the same time, some contemporary philosophersespecially Charles Hartshorne and Norman Malcolmrevived the second argument, or second form of the argument, found in Anselm's Proslogion (3) and in his Responsio to Gaunilo. As they have reconstructed it, this argument starts from the premise that the concept of God as eternal, self-existent being is such that the question whether God exists cannot be a contingent question but must be one of logical necessity or impossibility. A being who exists, but of whom it is conceivable that he might not have existed, would be less than God; for only a being whose existence is necessary rather than contingent can be that than which nothing greater is conceivable. But if such a necessary being does not exist, it must be a necessary rather than a contingent fact that he does not exist. Thus God's existence is either logically necessary or logically impossible. However, it has not been shown to be impossiblethat is, the concept of such a being has not been shown to be self-contradictoryand therefore we must conclude that God necessarily exists.

Hartshorne formalizes the argument as follows:

(1) q Nq"Anslem's principle": perfection could not exist contingently
(2) Nq NqExcluded middle
(3) Nq N NqForm of Becker's postulate: modal status is always necessary.
(4) Nq N NqInference from (2, 3)
(5) N Nq N qInference from (1): the necessary falsity of the consequent implies that of the antecedent (modal form of modus tollens )
(6) Nq N qInference from (4, 5)
(7) N qIntuitive postulate (of conclusion from other theistic arguments): perfection is not impossible
(8) NqInference from (6, 7)
(9) Nq qModal axiom
(10) qInference from (8, 9)

In this formalization q stands for (x )Px ("There is a perfect being" or "Perfection exists"); N means "analytic or L-true, true by necessity of the meanings of the terms employed"; and signifies strict implication.


The above argument seems to depend upon a confusion of two different concepts of necessary being. The distinction involved is important for the elucidation of the idea of God and represents one of the points at which study of the Ontological Argument can be fruitful even though the argument itself fails. The two concepts are those of logical necessity and ontological or factual necessity. In modern philosophy, logical necessity is a concept that applies only to propositions; a proposition is logically necessary if it is true in virtue of the meanings of the terms composing it. And it is a basic empiricist principle that existential propositions cannot be logically necessary. In other words, whether or not a given kind of entity exists is a question of experiential fact and not of the rules of language. On this view, the notion of a logically necessary being is inadmissible, for it would mean that the existential proposition "God exists" is logically true or true by definition. Anselm's principle, however, which is used as the first premise of Hartshorne's argument, was not that God is a logically necessary being (in this modern sense) but that God is an ontologically or factually necessary being, For, as noted above, Anselm was explicit that by a being whose nonexistence is inconceivable he meant a being who exists without beginning or end and always as a whole. (This is virtually the scholastic notion of aseity, from a se esse, "self-existence," that is, eternal and independent existence.) Interpreting "For God to exist is for him to exist necessarily" (prop. 1) in this way, we can validly infer from it that God's existence is ontologically either necessary or impossible (prop. 6). For if an eternal being exists, he cannot, compatibly with the concept of him as eternal, cease to exist: thus his existence is necessary. And if such a being does not exist, he cannot, compatibly with the concept of him as eternal, come to exist: thus his existence is impossible.

However, it does not follow from this that an eternal being in fact exists but only that if such a being exists, his existence is ontologically necessary, and that if no such being exists, it is impossible for one to exist. Hartshorne's argument can advance from proposition 6 to its conclusion only by assuming at this point that it has been established that the existence of God is (not, or not only, ontologically but) logically necessary or impossible. He can then rule out the latter alternative (prop. 7), and conclude that God necessarily exists (prop. 8) and hence that he exists (prop. 10). Thus, in propositions 16 "necessary" means "ontologically necessary"; in propositions 610 it means "logically necessary"; and proposition 6 itself is the point at which the confusion occurs. (The same illicit shift between the notions of ontological and logical necessity can be observed in Malcolm's version of the argument.)

The conclusion to be drawn is that the Ontological Argument, considered as an attempted logical demonstration of the existence of God, fails. In both of the forms that are found in Anselm, and which are still matters of discussion today, the flaw in the argument is that while it establishes that the concept of God involves the idea of God's existence, and indeed of God's necessary (in the sense of eternal) existence, it cannot take the further step of establishing that this concept of an eternally existent being is exemplified in reality.

See also Anselm, St.


For the general history of the Ontological Argument, see M. Esser, Der ontologische Gottesbeweis und seine Geschichte (Bonn, 1905); J. Kopper, Reflexion und Raisonnement im ontologischen Gottesbeweis (Cologne, 1962); and C. C. J. Webb, "Anselm's Ontological Argument for the Existence of God," in PAS 3 (2) (1896): 2543. On its alleged origin in Plato, see J. Moreau, "L'argument ontologique" dans le Phédon," in Revue philosophique de la France et de l'étranger 137 (1947): 320343, and J. Prescott Johnson, "The Ontological Argument in Plato," in Personalist (1963): 2434. The main passage in Augustine which may have influenced Anselm (in relation to the Ontological Argument) is De Libero Arbitrio, Book II, translated by J. H. S. Burleigh as "On Free Will," in Augustine: Earlier Writings, edited by J. H. S. Burleigh (London and Philadelphia, 1953).

Anselm's version of the argument occurs in Proslogion, Chs. 24, and in his exchange with Gaunilo. The Latin text is in Dom F. S. Schmitt's critical edition of the Opera Omnia, Vol. I (Edinburgh, 1945), pp. 101104 and 125139. The best English translations of the Proslogion as a whole are those of Eugene R. Fairweather in A Scholastic Miscellany: Anselm to Ockham (Philadelphia: Westminster Press, 1956); A. C. Pegis in The Wisdom of Catholicism (New York: Random House, 1949); M. J. Charlesworth, St. Anselm's "Proslogion" with "A Reply on Behalf of the Fool" by Gaunilo and "The Author's Reply to Gaunilo," with an introduction and philosophical commentary by Charlesworth (Oxford: Clarendon Press, 1965); and A. C. McGill in The Many Faced Argument, edited by John Hick and A. C. McGill (New York: Macmillan, 1967), where McGill, as well as translating Proslogion 24, has translated Anselm's exchange with Gaunilo, arranging the texts so as to present Anselm's replies in relation to the specific criticisms to which they refer. The translations by S. N. Deane in St. Anselm: Basic Writings, 2nd ed. (La Salle, IL, 1962) include, in addition to the Proslogion, Gaunilo's On Behalf of the Fool and Anselm's Reply. Anselm's dialogue De Veritate is also relevant; there is an excellent English translation by Richard McKeon in Selections from Medieval Philosophers, Vol. I (New York: Scribners, 1929).

On the interpretation of Anselm, the most important older writings are Charles Filliatre, La philosophie de saint Anselme (Paris, 1920); Alexandre Koyré, L'idée de Dieu dans la philosophie de S. Anselme (Paris, 1923); Karl Barth, Fides Quaerens Intellectum: Anselms Beweis der Existenz Gottes in Zusammenhang seines theologischen Programms (Zürich, 1930; 2nd ed., 1958), translated by Ian Robertson as Anselm: Fides Quaerens Intellectum (Richmond, VA: John Knox Press, 1960); A. Stolz, "Zur Theologie Anselms im Proslogion," in Catholica (1933): 121, translated by A. C. McGill in The Many Faced Argument (see above); Étienne Gilson, "Sens et nature de l'argument de saint Anselme," in Archives d'histoire doctrinale et littéraire du moyen âge 9 (1934): 551; and Adolf Kolping, Anselms ProslogionBeweis der Existenz Gottes im Zusammenhang seines Spekulativen Programms: Fides Quaerens Intellectum (Bonn, 1939).

For more recent historical studies, see Spicilegium Beccense, Proceedings of the Congrès International du IXe Centenaire de l'arrivée d'Anselme au Bec (Paris: J. Vrin, 1959).

For the view that there are two different arguments in Proslogion, see Charles Hartshorne, introduction to St. Anselm: Basic Writings (see above); "What Did Anselm Discover?," in Union Theological Seminary Quarterly Review (1962): 213222; and The Logic of Perfection (La Salle, IL: Open Court, 1962), Ch. 2. See also Norman Malcolm, "Anselm's Ontological Arguments," in Philosophical Review 69 (1 (1960): 4162, reprinted in Malcolm's Knowledge and Certainty (Englewood Cliffs, NJ: Prentice Hall, 1963). St. Thomas's criticism of the argument occurs in his Summa Contra Gentiles I, 10 and 11, and Summa Theologiae I, 2, 1.

For the argument after Anselm, see O. Paschen, Der ontologische Gottesbeweis in der Scholastik (Aachen, 1903); John Duns Scotus, De Primo Principio, translated by E. Roche (St. Bonaventure, NY: Franciscan Institute, 1949); I. Hislop, "St. Thomas and the Ontological Argument," in Contemplations Presented to the Dominican Tertiaries of Glasgow (London, 1949), pp. 3238; P. L. M. Puech, "Duns Scotus et l'argument de saint Anselme," in Nos Cahiers (1937): 183199; A. Runze, Der ontologische Gottesbeweis, kritische Darstellung seiner Geschichte seit Anselm bis auf die Gegenwart (Halle, 1882); O. Herrlin, The Ontological Proof in Thomistic and Kantian Interpretation (Uppsala: Lundquistska bokhandeln, 1950); Robert Miller, "The Ontological Argument in St. Anselm and Descartes," in Modern Schoolman (19541955): 341349, and (19551956): 3138; and Alexandre Koyré, Essai sur l'idée de Dieu et les preuves de son existence chez Descartes (Paris: E. Leroux, 1922).

Descartes's own version of the argument occurs in his Meditations (V, but see also III). Spinoza uses the argument in his Ethics, Part I, Props. 711. Leibniz's version is found in his New Essays concerning Human Understanding, Book IV, Ch. 10 and Appendix X, and in his Monadology, Secs. 4445. Kant's criticism of the argument (in its Cartesian form) occurs in his Critique of Pure Reason, Transcendental Dialectic, Book II, Ch. iii, Sec. 4.

The Hegelian use of the argument is to be found in Hegel's Vorlesungen über die Philosophic der Religion, edited by P. Marheineke (Berlin, 1832), Appendix: "Beweise für das Dasein Gottes," Vol. II, pp. 466483, translated by E. B. Speirs and J. B. Sanderson as Lectures on the Philosophy of Religion (London: K. Paul, Trench, Trubner, 1895), Vol. III. pp. 347367; and in Vorlesungen über die Geschichte der Philosophie, edited by K. L. Michelet (Berlin, 1836), Vol. III, pp. 164169, translated by E. S. Haldane and F. H. Simson as Lectures on the History of Philosophy (London, 1896), Vol. III, pp. 6267; Edward Caird, "Anselm's Argument for the Being of God," in Journal of Theological Studies 1 (1899): 2339; E. E. Harris, Revelation through Reason (New Haven, CT: Yale University Press, 1958); Paul Tillich, "The Two Types of Philosophy of Religion," in his Theology of Culture (New York: Oxford University Press, 1959); and R. G. Collingwood, An Essay on Philosophical Method (Oxford: Clarendon Press, 1933). Collingwood's use of the argument was criticized by Gilbert Ryle in "Mr. Collingwood and the Ontological Argument," in Mind, n.s., 44 (174) (1935): 137151. There is a reply to Ryle by E. E: Harris in Mind, n.s., 45 (180) (1936): 474481), and a rejoinder by Ryle in Mind, n.s., 46 (181) (1937): 5357. This discussion is reprinted in The Many Faced Argument (see above).

For the argument in recent French "reflexive" philosophy, see M. Blondel, L'action (Paris, 1893 and 1950), L'étre et les ètres (Paris, 1935), and La pensée (Paris, 1934 and 1948); J. Paliard, Intuition et réflexion (Paris, 1925); and A. Forest, "L'argument de saint Anselme dans la philosophie réflexive," in Spicilegium Beccense, pp. 273294, translated by A. C. McGill in The Many Faced Argument (see above).

Some important contemporary philosophical discussions are C. D. Broad, "Arguments for the Existence of God," in Journal of Theological Studies 40 (157) (January 1939); J. N. Findlay, "Can God's Existence Be Disproved?," in Mind, n.s., 57 (226) (1948): 176183, reprinted in New Essays in Philosophical Theology edited by A. G. N. Flew and Alasdair MacIntyre (London, 1958); Nicholas Rescher, "The Ontological Proof Revisited," in Australasian Journal of Philosophy 37 (2) (1959): 138148; William P. Alston, "The Ontological Argument Revisited," in Philosophical Review 69 (4) (1960): 452474; Norman Malcolm, "Anselm's Ontological Argument" (see above); Charles Hartshorne, The Logic of Perfection (see above); Jerome Shaffer, "Existence, Predication and the Ontological Argument," in Mind, n.s., 71 (283) (1962): 307325, reprinted in The Many Faced Argument (see above); and a discussion by R. E. Allen, Raziel Abelson, Terence Penelhum, Alvin Plantinga, and Paul Henle in Philosophical Review 70 (1) (1961): 56109.

On the notions of "existence" and "necessary existence," see Bertrand Russell, "The Philosophy of Logical Atomism" (1918), in his Logic and Knowledge, edited by R. C. Marsh (London: Allen and Unwin, 1956); G. E. Moore, "Is Existence a Predicate?," in PAS, Supp., 15 (1936), reprinted in Logic and Language, 2d series, edited by A. G. N. Flew (Oxford, 1959); William Kneale, "Is Existence a Predicate?," in PAS, Supp., 15 (1936), reprinted in Readings in Philosophical Analysis, edited by Herbert Feigl and Wilfrid Sellars (New York: Appleton-Century-Crofts, 1949); George Nakhnikian and W. C. Salmon, "'Exists' as a Predicate," in Philosophical Review 66 (4) (1957): 535542; John Hick, "God as Necessary Being," in Journal of Philosophy 57 (2223) (1960): 725734; and Terence Penelhum, "Divine Necessity," in Mind, n.s., 69 (274) (1960): 175186.

John Hick (1967)