You are here

Chapter Three | God and Logic

‘But why drives on that ship so fast,

Without or wave or wind?’

‘The air is cut away before,

And closes from behind.’

—S. T. Coleridge, The Ancient Mariner.

The point which we reached at the end of the last lecture was this: that in conducting any argument for theism it is imperative to be clear and consistent about the definition which we are employing of the word ‘God’. Otherwise we shall be constantly confusing the two questions ‘Is there a God?’ and ‘What is God like?’; and we shall in all probability fall into the fallacy of implicitly assuming a minimal definition of God in arguing for his existence and a much more ample definition in discussing his nature. There can be no harm in adopting different definitions of ‘God’ in different arguments for God's existence, provided we are quite clear what we are doing and carefully check each against the others. Thus, in one argument we might define God as a being having the attribute X and, having proved the existence of such a being, go on to argue that in addition it possessed the attributes Y and Z. In another argument we might define God as a being possessing the attributes X and Y and, having proved the existence of such a being (a more difficult task than the former but perhaps a possible one), go on to prove that it possessed in addition the attribute Z. We should then have two arguments proving the existence of a being possessing the attributes X, Y and Z, though a further argument might be necessary to prove that there is only one such being. Alternatively, we might have two arguments, one of which proved the existence of God defined as a being possessing the attribute P and the other proving the existence of God defined as a being possessing the attribute Q, and we might further argue that any being possessing one of these attributes also possessed the other and that there was only one being that possessed either. Again, we might even have an argument that proved the existence of God defined as a being possessing the attribute L and find that, in the very act of proving its existence, we had also proved that it possessed the attribute M. I shall not attempt to catalogue all the possible variants; they would appear to be in principle unlimited. But I shall do my best to heed my own warning; and in discussing any argument for the existence of God I shall try to make it plain what is the definition of God that is presupposed to the argument. And I shall devote the rest of this lecture to a consideration of the ontological argument, which, after having fallen into apparently irreparable discredit, has recently made a remarkable recovery.

The ontological argument, or, in view of their diversity, it would perhaps be better to say, ontological arguments, attempt to prove the existence of God simply from the definition of the word ‘God’ or from the concept which that word connotes. Thus an argument such as that in Descartes'rd Meditation, which is based on the fact that the concept of God is found in the thinker's own mind, is not strictly speaking an ontological argument; on the other hand the argument in his Fifth Meditation, based on the content of the concept itself, is an ontological argument in the strictest sense. As a matter of history, arguments of this type spring from two chief sources: the Proslogion of St Anselm of Canterbury in the eleventh century and the Third of the Meditations of René Descartes in the seventeenth. It is the Anselmian argument that has received most attention in recent years as an important contribution to philosophical theology, while that of Descartes still holds its position in academic syllabuses for its historical interest but does not appear to be taken very seriously in itself.

The definition of ‘God’ from which Anselm begins is that of something than which nothing greater can be thought’, aliquid quo majus nihil cogitari potest;1 that from which Descartes begins is that of a supremely perfect being.2 These definitions are not identical, though it might not be difficult to argue that they are equivalent, that is to say that anything which corresponded to either also corresponded to the other. It is Anselm's with which we shall be chiefly concerned. There are two statements of the argument in the Proslogion, or, as some recent writers would insist, statements of two similar but importantly different arguments, in chapters II and III of that work respectively. Both claim to prove that God exists, but the former merely asserts that the non-existence of God is false, while the latter asserts that it is not only false but impossible.

Proslogion II, having defined God as something than which nothing greater can be thought, draws a distinction between a purely mental existence of any object—its existence ‘in the understanding’—and its existence in concrete reality; it then asserts that an object which exists both in the mind and in reality is ‘greater’ than an object, otherwise precisely similar, which exists only in the mind. It is then argued that ‘that than which nothing greater can be thought’ must exist in reality as well as in the mind, since if it existed only in the mind we could think of a being otherwise precisely similar which did exist in reality as well. We should thus have been led into a contradiction, having thought of something greater than something than which nothing greater can be thought. Thus the hypothesis that God exists only in the mind must be false, and it follows that God exists in reality. The argument is thus a reductio ad absurdum, proving a proposition—‘God exists in reality and not only in the mind’—by showing that the supposition of its falsehood leads to a contradiction. The argument in Proslogion III derives a similar contradiction from a logically weaker hypothesis, namely the hypothesis that God can be conceived as not existing; and the conclusion is correspondingly stronger. It is that the nonexistence of God is not merely false but is inconceivable. Philosophers have in general seen little difference between the two arguments and have tended to take Proslogion III as a restatement or, at most, a clarification or amplification of Proslogion II. We shall, however, see later on that for some modern philosophers the difference is quite fundamental.

Among recent literature on the subject I would single out two symposia as specially useful. The Ontological Argument from St Anselm to Contemporary Philosophers, edited by Alvin Plantinga, covers both Anselm and Descartes, before going on to G. E. Moore, William P. Alston, J. N. Findlay, Charles Hartshorne and Norman Malcolm. It includes Anselm's controversy with his contemporary Gaunilo and the criticism made by the later scholastic St Thomas Aquinas. To Descartes’ own exposition it adds his controversy with Caterus, Gassendi and others and the relevant later writings of Spinoza, Leibniz, Kant and Schopenhauer. The other book, The Many-faced Argument: Recent Studies on the Ontological Argument for the Existence of God, edited by John Hick and Arthur C. McGill, pays little attention to Descartes, but it contains several critical and interpretative studies of Anselm, including those by Karl Barth and Anselm Stolz, and a very useful survey by McGill, before going on to contemporary or quite recent material. When it is added that in October 1968 a special section of the journal Religious Studies was devoted to five essays on the ontological argument and that other issues have contained articles on it as well, it should be evident that interest in it is anything but dead.

As with other philosophical and theological topics, much of the discussion has been concerned with the nature and purpose of the argument as its author St Anselm conceived it. This may at first sight be surprising, in view of the brevity and clarity of the argument itself, but its brevity may in fact provide the explanation; Anselm, so far from telling us exactly what was in his mind, leaves us to find this out for ourselves. It is therefore understandable that we do not all come to the same conclusion. Dr McGill, in his survey, distinguishes several schools of interpretation. The first, including both admirers and critics, sees Anselm as the supreme rationalist among Christian thinkers; he claims to prove everything by reason, even the deepest and most mysterious truths of the faith.3 Others see Anselm as, before all else, the Christian believer, who insists on the primacy of faith, whose desire to understand God arises from his love and hunger for God, and who, even in his most intensive arguments, always argues from the data of revealed faith for the satisfaction of loving faith. The most striking, and indeed paradoxical, presentation of this point of view is to be found in Karl Barth's work Fides quaerens Intellectum, published in German in 1931 and later translated into English. For Barth, what Anselm intended to do was to show by reason that reason could tell us nothing about God except that he is altogether out of the reach of reason. Others, such as André Hayen, while less extreme than Barth, see Anselm as a theologian rather than a philosopher; Hayen asserts that his argument with the ‘fool’ who ‘says in his heart “There is no God”’ is inspired by an apostolic love for all men which includes even the fool in its scope. Others, again, like Anselm Stolz, contend that the Proslogion is a work of mystical theology, that its presentation in the form of a prayer is not a literary device but is of its very essence and that it ‘represents an effort of the soul to raise itself to a kind of vision of God, by means of a dialectical investigation of the rationality of Christian dogmas’.4 Where experts disagree so widely it is temerarious for the layman to express an opinion, but I venture to suggest that the various views are not, with the possible exception of Barth's, incompatible. Anselm did not, as I see the matter, make the kind of distinction between faith and reason which we find, for example, in St Thomas, and which in some later theologians becomes not a distinction but an impenetrable barrier. He believed in God by faith, and the definition which he gives of God is, for him, only a clarification, in terms appropriate to discussion, of the truth about God as supreme and transcendent which is central to the religion of both the Old and the New Testament. But he believed by faith that God is supremely rational, and it therefore seemed obvious to him that, if one could only find out how to do it, it must be possible to prove the existence, and indeed the necessary and inevitable existence, of this supremely rational being. He believed that God had shown him how to do this, and he could never thank God sufficiently for it. I think, therefore, that one key at least to the Proslogion is to be found in the fact that it could never occur to Anselm that there was anything irrational or anti-rational about faith and revelation. Whether his arguments are rationally valid or not is, of course, another question.

Almost, if not quite, all of the many criticisms which have been made of Anselm's arguments accuse him of making an illicit transition from the conceptual or ideal to the actual realm. Even if I cannot think of God except as existing, the objection runs, this does not show that he actually exists. This holds equally against Proslogion II and Proslogion III; it holds also against the modified form of the argument in the Reply to Gaunilo, in which Anselm argues, first that his definition of God implies that God is an eternal existent, and then that an eternal existent cannot be a merely possible one. One way—it might be called the ‘classical’ way—of stating the objection is by saying that ‘exists’ is not a predicate; more precisely we might say that it is certainly not a predicate like other predicates.5

Grammatical predicates can indeed have a number of logical, epistemological and metaphysical kinds of function. If the horses in the field are brown, each one of them is brown; but if the horses in the field are numerous, it does not follow that each of them is numerous. More relevantly to our present concern, while most predicates refer to qualities of things and thus to components or aspects of their essence, ‘exists’ refers to the concrete actuality of things and not, except indirectly, to their essences at all. Many modern writers, as for example Étienne Gilson and Bernard Lonergan, have emphasised in this connection the radical difference between the apprehension of a concept and the affirmation of a judgment, and have pointed out that, while qualities are conceived as constituents of the essence of a being, its existence is affirmed in a judgment; and, we must add, not the judgment that its essence includes certain qualities but the more basic judgment that it is. Thus it is really misleading to speak of ‘existence’ at all; for the very shape of the word, parallel in its formation to ‘essence’, might suggest that it is a kind of quality or a complex of qualities. It is noteworthy that St Thomas Aquinas in his most characteristic works does not use the word existentia at all; he avoids its abstract flavour by using instead the verbal noun esse. It may be added that in a recent article in Religious Studies6 Mr M. J. A. O'Connor has claimed—as it seems to me, with justice—that Anselm's three arguments involve further fallacies of a logical character hitherto undetected. He points to the peculiarly psychological character of Anselm's definition of God—‘something than which nothing greater can be thought’—and remarks that a more natural definition would be ‘something than which nothing greater can be’. The difference is, he maintains, not trivial, for it leads Anselm to compare God's existence in the mind with God's existence in reality, as if the mind and reality were two locations in which God could exist; whereas in fact what exists in the mind is not God but the thought of God.7 ‘In point of fact’, writes O'Connor ‘Anselm deceives us into thinking he is treating G.C.B. [the greatest conceivable being] as a concept, when he is actually treating it as a given entity at different levels of ontology—mental and actual. From the supposed agreement of the unbeliever that G.C.B. exists at the one level he argues that it must also exist at the other.’8 And O'Connor shows how, for example, this inadvertence on Anselm's part leads him to suppose that he has proved that ‘G.C.B. cannot be imagined as existing and yet not actually exist’ when all he has in fact proved is that ‘G.C.B. cannot be imagined as existing and yet be known not to exist’—a very different proposition.

Anselm, we have just said, speaks of God, and not just the thought of God, as existing in the mind; as if the mind was itself a place in which God actually exists.9 It must now be recognised that, according to one school of interpreters, a sympathetic school at that, this is precisely what Anselm did mean. (We must note, of course, that by ‘existence in the mind’ here we mean not that existence of God in all things by ‘essence, presence and power’, as St Thomas puts it, which arises from the fact that he is their creator and sustainer,10 but an actual identification of the idea of God in the mind with God's own entitative being.) In McGill's words, ‘At the beginning of the argument,… Anselm is not looking at or thinking about some mental notion and wondering whether this is a true notion which represents a real object. Through this idea he is looking at and thinking about reality, specifically about the unique perfection of God's nature.’ ‘According to the realists, then, since Anselm begins with a direct cognition of God's reality, it is better not to call this an “idea” at all, but an awareness. Some say that he views it in this way because of his Platonic doctrine of ideas.… Other historians explain this realism as the product of a decisive religious experience.… According to the realists, therefore, those who deny the conclusion cannot be accused of a logical blunder. They simply fail to recognise the awareness which is being analysed.’11

A very different interpretation of Anselm which McGill lists asserts the purely psychological character of the idea of God in the mind and sees the saint as arguing to God as the cause of this psychological datum with its peculiar character of compulsiveness; in this, as he remarks, it would appear to have much in common with Descartes’ argument in his Third Meditation. A variant of this interpretation is found in Étienne Gilson, who sees Anselm as appealing to God's reality as causing not the idea as such, but the logical necessity which the mind discovers in analysing it; this, however, means for Gilson that the argument contains only the initial datum for a proof. Still another interpretation finds importance in the distinction, in Anselm, between the words ‘conceive’ (cogitare) and ‘understand’ (intelligere). ‘Cogitare… might be called the cognition of objective possibility, just as intelligere is the cognition of objective actuality.’ The fool in the argument understands the definition of God, because he has it in his mind, but when he tries to conceive God as non-existent he finds this impossible. Conceiving operates within a noetic limit, imposed by reality itself. ‘The logic simply demonstrates that this initial limit whereby the mind is not able to conceive of anything greater than God, involves other limits, such as its not being able to affirm (Proslogion II) or even to conceive of (Proslogion III) God's non-existence.’12

In contrast with such interpretations as these Karl Barth maintains that for Anselm the definition of God from which he begins is itself a datum of revelation, so that all the argument shows—but for Barth it is a great deal—is that revelation does not allow us to conceive of a non-existent God. I find Barth's argument tortuous and unconvincing, but, as I have discussed it at some length elsewhere,13 I shall not repeat myself here. It is also dealt with exhaustively by Henri Bouillard in his book The Knowledge of God. I think, however, that Barth is right in seeing Anselm's definition as emerging from the Christian revelation. I agree with Gilson that ‘the inconceivability of the nonexistence of God could have no meaning at all save in a Christian outlook where God is identified with being, and where, consequently, it becomes contradictory to suppose that we think of him and think of him as non-existent.’14 And I agree further with Gilson that, significant as it is, Anselm's attempt to pass from the realm of concepts to the realm of actuality is a failure, though a heroic one. One further school of interpreters to whom McGill refers see the argument of the Proslogion as a sequel to Anselm's earlier work the Monologium, in which he seeks to prove the existence of God from the existence of things in this world; they do in fact admit the inadequacy of the Proslogiona real predicate and can be a constituent element as providing arguments for God's existence, though it is difficult to suppose that Anselm would agree with them.

McGill's discussion of the purpose and character of Anselm's arguments is a magnificent example of erudition and of analytical ability; as he somewhat disconcertingly shows, almost any of the interpretations proposed comes into conflict with statements made by Anselm himself. Perhaps all these interpretations are over-subtle and the truth is simply, as I have suggested, that Anselm saw no conflict between reason and revelation and that it seemed to him to be the most natural thing in the world to look for a rational proof of the existence of God, as he looked for rational proofs of the Trinity and the Incarnation. That his thought, like that of other Christians down to the thirteenth century, moved naturally in a Platonic realm of ideas and essences is not surprising. Like some other achievements of the saints (not, I hasten to add, like all), it calls for our admiration rather than our imitation. I am, however, startled by the grounds on which, at the end of his brilliant survey, McGill expresses his hopes of the future vindication of Anselm. I have in an earlier lecture said something about the importance of language as a means of communication between human persons and as an instrument for the clarification of human thought. McGill, however, attributes to words an inherent character as bearers of meaning which is apparently quite independent of their utilisation as tools by human beings. After referring to the dominant view of language as derivative and expressive, he writes as follows:

Today, for the first time in centuries, a serious challenge is being raised against this subjectivistic theory of language. Words, according to Martin Heidegger, are not primarily the tools by which men express what is already in their heads. Rather they are the instruments of reality itself, the medium through which being discloses itself, using man's voice as its spokesman. Language is not about reality, it is reality in the state of unveiledness and in every statement it is the subject matter—not the subjectivity of the author—which addresses man's thought.

This view has begun to liberate readers from the axiom that language expresses merely human ideas. We may expect, therefore, that eventually it will lead to a new approach to Proslogion II. When that occurs, interpreters will give full weight to Anselm's intelligere quod auditur, and will be able to understand why he thinks that nothing else is needed except uttering the words ‘that than which a greater cannot be conceived’.15

I am not surprised that Mr Thomas McPherson mildly remarks, in a review of The Many-faced Argument, that ‘this kind of approach seems to create more problems than it solves’. ‘The trouble with this’, he writes, ‘surely is that men utter falsehoods as well as truths, are obscure as well as lucid. It is no doubt true, if one chooses to put it that way, that reality expresses itself through man's voice. But so does unreality and confusion, and it is not always easy to tell the difference simply on the basis of the words we hear.’16 Such logolatry as this might well seem too great a price to pay for one's deliverance from the ghost-world of linguistic empiricism.

I have postponed to this point one group of thinkers to whom McGill makes brief reference, first because the symposium contains an essay by Aimé Forest which is devoted entirely to them, secondly because it seems to me to be of considerable importance, and thirdly because little attention has been paid to it in professional English-speaking philosophical circles. This is the school of so-called ‘reflexive philosophy’, in which the leading figures are Maurice Blondel, Jacques Paliard, Louis Lavelle and Ferdinand Alquté, not to mention Aimé Forest himself. In McGill's words:

This view is based on a fundamental distinction between what may be called ‘projective’ and ‘reflexive’ thought. In every moment of consciousness the mind reaches outward away from itself toward some object; it thus acts intentionally and projectively. At the same time, however, while grasping for things, the mind is also aware of its own operations, of its own intellectual, moral and emotional processes. In other words, alongside the knowledge which each man has of the world, he also possesses a reflexive knowledge of himself as an inner spiritual activity, as an élan, or dynamism, of consciousness and will.17

It is emphasised that this reflexive knowledge is not to be confused with introspection, in which the mind becomes the deliberate object of its own attention; indeed it seems to me that the choice of the word ‘reflexive’ is unfortunate and could be misleading. ‘In the latter [introspection] the mind turns away from the world and looks at its own mental processes as if they were objects. Reflexive knowledge, on the other hand, is what the mind learns about itself while in the process of thinking about objects. It accompanies, and does not replace projective thought.’18 It is thus, in the language of Karl Rahner and the other protagonists of transcendental Thomism, essentially unthematic and unformulated, though it can become formulated by subsequent thought. It is, so to speak, my knowledge of myself at the near end of my consciousness of the world around me and of my willing and acting within it. And it is something which the neobehaviourists such as Professor Ryle find it even more difficult to bring within their self-imposed limits than they find the facts of introspection. It is what the scholastics are referring to when they insist that we know ourselves not apart from, but in, the acts through which we know other beings.

Now it is the contention of the school of reflexive philosophy that in our awareness of ourselves we can also become aware of God as our creative ground and as the supreme reality; and it is held that it is the function of the ontological argument to enable us to recognise this. In Ravaisson's words, quoted by Forest: ‘From the interior and central point of view of reflection, the soul does not discern only itself, but also, at its foundation, the absolute from which it emanates.’19 For Blondel, with his basic metaphysical principle of ‘action’, the ontological argument complements and does not supersede the various forms of cosmological argument. In the meeting of the truth of being with the dynamism of the consciousness which not merely perceives but affirms it, action enables us to perceive not only what we are but what is at the root of our being. We do not perceive God himself, but we do perceive God in ourselves. Whereas Blondel's starting-point is our action, Jacques Paliard's is our self-consciousness, while for Louis Lavelle it is by reflection on our action that we form the idea of God, as enveloping our existence because he is the cause of it; it is an idea that is reflexive, not representative. For Ferdinand Alquié, too, the weight of the ontological argument is carried by reflection and not by representation: ‘the ontological argument consists in naming the presence which becomes manifest when the mind is aware of returning to itself.’20 Central to Alquié's thought, we are told, is the distinction between ‘ideas’ and ‘presences’. ‘The characteristic of an idea is that we can delimit it; it is in us in such a way that in principle nothing can prevent us from saying that we are its author. A presence, on the contrary, is not objective, if by this term we understand the character of that which lends itself to being determined. The ontological argument is the discovery of the absolute; it is necessary to affirm its existence because of the discovery that a presence cannot be reduced to the objectivity of ideas. By reflecting upon an interior presence, the spirit sees its limits and at the same time recognises what sustains its vitality.’21 It is emphatically asserted that this position is, in spite of appearances, utterly opposed to philosophical idealism. And Forest sums it up in the sentence: ‘Alquié's thought is the movement from Kant to St Anselm, from critical philosophy to spiritual philosophy.’22

Even from such a brief summary as I have been able to give here we might be led to wonder whether this very impressive and deeply devout movement of thought is really very closely related to St Anselm's. Forest indeed admits this. ‘One will hesitate’, he writes, ‘to acknowledge an accord between reflexive philosophy and St Anselm's teachings. We cannot say that these extremely original endeavours hold to the letter of the argument in the Proslogion.’23 His conclusion is that ‘the reflexive proof is authorised by St Anselm, rather than in agreement with him’—a somewhat enigmatic remark! His final remarks are, however, moving, and thought-provoking:

We may recall that at the end of his life Husserl was assessing the difficulties and possible failure of his religious philosophy. He said that his mistake had been to want to seek God without God. That is the temptation of idealism. More than any other experience, the Anselmian kind of self reflection is able to deliver us from it.24

We shall later on be considering a group of thinkers who have much in common with the reflexive school, when we discuss the cosmological approach to theism; this is the school of ‘transcendental Thomism’, of which Karl Rahner, Emil Coreth and Bernard Lonergan are leading exponents.

I shall now make some remarks about a very different attempt to rehabilitate the ontological argument which has been made by Norman Malcolm and Charles Hartshorne. Their most important papers are included in Hick and McGill's volume, but mention should also be made of several full-scale treatises in which Hartshorne expounds and defends his argument at length, The Logic of Perfection, Anselm's Discovery and A Natural Theology for our Time. The special feature of this approach is that it very sharply distinguishes between the arguments in chapters II and III of the Proslogion and takes the latter as expressing Anselm's real intention, as well as being in fact valid. The difference between the two chapters, it will be remembered, is that the former claims to prove simply that God exists, whereas the latter claims to prove that he exists necessarily. To put the contrast in negative terms, the former claims to prove that God's non-existence is false, the latter that it is impossible. Most critics have taken Proslogion III as merely dotting the i's and crossing the t's of Proslogion II; after proving that God exists, they tell us, Anselm goes on to point out that, unlike other existents, he exists necessarily. Malcolm and Hartshorne, however, are in principle prepared to admit the criticism which accuses Proslogion II of wrongly treating ‘exists’ as a predicate and thus of making an illicit transition from the ideal to the real realm; though it would perhaps be fairer to say that they would admit the criticism if they thought that Proslogion II stood on its own feet without assistance from Proslogion III. The heart of their defence is that, whether or not ‘exists’ is a predicate and ascribes an attribute to God, ‘necessarily exists’ certainly is and does. Saying that God exists may not say what kind of being he is, but saying that he necessarily exists certainly does. Thus, existence may not be an attribute, but necessary existence is; and it is this latter that Proslogion III shows God to possess. For Hartshorne the argument is a rigid one in modal logic, that is to say the logic which considers propositions as not merely true or false, but also as necessary, contingent or impossible; in one place he sets it out in symbolic form in ten successive steps.25 His basic conviction is that to deny the existence of a being that has been proved to have necessary existence as an attribute is to utter a self-contradictory statement. I cannot see, however, that it makes any difference whether the attribute in question is existence tout court or necessary existence; in either case the most that can be validly argued is that if God exists he exists necessarily, but it does not follow that he exists. Hartshorne may be quite correct in arguing that necessary existence is included in God's essence, that is, in the kind of being that he is if he exists; but the existence by which an existing being exists in reality is not an essence or a constituent or aspect of an essence; it is an act. If God exists, we can, with St Thomas, identify his essence with his act of existing and say that it is his essence to exist; but until we know that he exists all that we can say is that if he exists this identification can be made. Hick makes the point by accusing Hartshorne of confusing two quite distinct notions, that of logical necessity and that of ontological or factual necessity. He writes as follows:

We have distinguished the following two concepts: (1) the logically necessary truth of a proposition, arising from the meaning of the terms employed in it; and (2) the factual necessity of a Being who exists eternally and a se. The two concepts are quite distinct; logical necessity is not a case of ontological necessity, nor vice versa. The necessary existence of an object, x, is defined as the existence of x without beginning or end and without dependence upon anything other than itself. The logically necessary truth of a proposition, p, on the other hand, reflects the circumstances that p is formed so as to be true by definition.… From the concept of God as ontologically necessary we can derive the analytic truth that if God exists, he exists eternally and a se, but we cannot deduce that it is a logically necessary truth that God exists, i.e., that the concept of an eternal Being who exists a se is instantiated in extramental reality. And yet this is precisely what Malcolm and Hartshorne try to do. They observe (rightly) that while ‘existence’ is not a real predicate (i.e., cannot figure as an element in the concept of a kind of being), ‘necessary existence’ in the ontological sense is a real predicate and can be a constituent element in the concept of deity. However, having established that ontological necessity (i.e., eternal existence a se) is a real predicate, they proceed as though what they had established is the quite different conclusion that logically necessary existence is a real predicate.26

And Hick is furthermore able to show the precise point in Hartshorne's statement at which the illicit shift of meaning is made; the plausibility of Hartshorne's argument depends on the fact that he has used the same symbol, ‘N’, to denote both kinds of necessity.

Substantially the same criticism as Hick's has been made by Dr David A. Pailin in an article published in Religious Studies.27 It is interesting to note that he defends the notion of ‘necessary existence’ against those modern philosophers who have asserted that the only proper application of the attribute ‘necessary’ is to propositions and that it is a misuse of the term to speak of a ‘necessary being’. My own comment on the question is that there is no harm in making both applications of the adjective as long as we do not fall into Hartshorne's fallacy of failing to distinguish between them. If the proposition ‘X exists’ is a necessary proposition, it does not seem unreasonable to describe X as a ‘necessary being’. I think that, as a matter of fact, most of the philosophers who object to the term ‘necessary being’ do so because they believe that no proposition asserting existence could be necessarily true; in this it seems to me that they are themselves failing to recognise the distinction drawn by Hick between logical and ontological necessity. There may, however, be something to be said for abandoning the term ‘necessary being’. Fr Norris Clarke has pointed out that it is far from universal in Christian usage. ‘St Thomas’, he reminds us, ‘never uses it as an attribute proper to God. This came in only through the Augustinian tradition stemming from Anselm.… St Thomas in no way deduces the existence of God from his essence, but rather defines his essence completely in terms of his existence: God is ipsum Esse Subsistens, pure subsistent act of existence.’ And again: ‘For St Thomas, “contingent” and “necessary” have quite different meanings from their now traditional use in modern philosophy, including modern scholasticism. For him, “contingent” meant simply any composite of matter and form, any generable and corruptible.… “Necessary” means just the opposite: any pure form that is not subject to substantial change. Both angels and God are such.’28

My conclusion is that, for all his ingenuity, Hartshorne has not been successful in his attempt to rehabilitate the ontological argument. There are indeed indications that Hartshorne's own thought on the question is not altogether finalised. Pailin quotes him as saying that he is now convinced that the second ontological proof is scarcely in Proslogion III but is rather in the Reply to Gaunilo. Pailin comments that ‘perhaps, in view of this, Hartshorne will in future not be too hard on those who failed to see in Anselm what Anselm himself never clearly distinguished’.29 And Hartshorne is certainly hard on his opponents, as anyone who has read his book Anselm's Discovery will admit. But before we leave him we should, I think, see in more detail what he really thinks Anselm himself failed to see.

On the basic issue of the existence of God Hartshorne distinguishes four possible positions, to each of which he assigns names: (1) God's existence is logically impossible (positivism); (2) God's existence is logically possible, but in fact false (empirical atheism); (3) God's existence is logically possible and in fact true (empirical theism); (4) God's existence is logically necessary (neo-classical theism). The last is, of course, Hartshorne's view and, as he sees it, Anselm's as well. However, Hartshorne parts company with Anselm on the meaning of Anselm's definition. When God is defined as ‘unsurpassable’ (which Hartshorne takes as a convenient abbreviation for ‘that than which nothing greater can be conceived’) what is meant is that God can not be surpassed by anything other than himself, not that he cannot continually surpass himself in the sense of continually adding to his own perfection. Thus, while God's existence is necessary, his actuality (by which Hartshorne means the character of his existence at any moment) is contingent. His actuality is thus concrete, but his existence (that which is common to all the instantiations) is abstract. And, while the existence of any particular universe (or of any particular state of the universe) is contingent, the existence of some universe or other appears to be necessary.

Hartshorne gives special attention to an article by Professor J. N. Findlay which appeared in Mind in April 1948 and which caused some astonishment in the philosophical world.30 In it Findlay claimed to have produced an ontological argument for the wow-existence of God. Briefly, his argument was that only a being who was unsurpassable in every respect could satisfy the requirements of the God of religion. It must possess every excellence and possess all these excellences in a necessary manner. However, Findlay asserts, for people of what he describes as a contemporary outlook, it will be clear that not only does no such being exist but its existence is either senseless or impossible. ‘It was indeed an ill day for Anselm’, he writes, ‘when he hit upon his famous proof. For on that day he not only laid bare something that is of the essence of an adequate religious object, but also something that entails its necessary non-existence’, or, he adds in a footnote, ‘“non-significance”, if this alternative is preferred.’31

Findlay's argument does not seem immune to response and it has in fact received a great deal of criticism, not least, as one might expect, from Hartshorne. For Hartshorne, what Findlay has disproved is empirical, but not neo-classical, theism, and fifteen years after the first appearance of his article Findlay admitted as much, writing as follows:

I still think that it makes a valid point: that if it is possible, in some logical and not merely epistemological sense, that there is no God, then God's existence is not merely doubtful but impossible, since nothing capable of non-existence could be a God at all.… Professor Hartshorne has, however, convinced me that my argument permits a ready inversion, and that one can very well argue that if God's existence is in any way possible, then it is also certain and necessary that God exists, a position which should give some comfort to the shade of Anselm.32

Thus Findlay and Hartshorne are agreed that God's existence must be either necessary or impossible; it cannot be a merely contingent matter whether there is or is not a God. I would agree with this last statement, but I would add that the necessity or impossibility must be ontological, not logical. Thus I do not think that it can be derived from the ontological argument, which, in spite of its name, is an argument in logic and not in ontology. I do think God's existence follows from the cosmological argument, but that is matter for a later lecture.

The most original feature of Hartshorne's discussion is his assertion that the God to whom the ontological argument leads is not a changeless but a changing and continually developing God. If he is right, not only has Anselm's argument been consistently misunderstood by almost all philosophers after him—in his extremely polemical work Anselm's Discovery Hartshorne lists between forty and fifty of these, including such giants as Aquinas and Kant and ranging from Gaunilo to Bertrand Russell and Findlay—but even Anselm himself failed to understand what he had proved. There is indeed something paradoxical in the assertion that, while almost everyone since Anselm has radically misunderstood the nature of Anselm's argument, Anselm himself radically misunderstood the nature of God. It is a pity that Hartshorne's book appeared almost simultaneously with Dr M. J. Charlesworth's edition, with commentary, of the Anselm-Gaunilo writings, for it would have been extremely interesting to have the comments of either of these writers on the other. Charlesworth inclines to see Proslogion III as a complement to the basic argument of Proslogion II, added to defend the latter against the objection that it does not adequately exclude the possibility of God existing in a merely contingent way; while chapter II compares possible and actual existence, chapter III compares two kinds of actual existence, namely contingent and necessary, and is therefore logically superior to chapter II.33 This goes most of the way to meeting Hartshorne's thesis that Proslogion III contains the real heart of the argument, but Charlesworth would not, any more than Anselm, admit Hartshorne's view of a developing God, who, while unsurpassable by anything else, is continually surpassing himself.

It should be added that, in summarising his discussion, Hartshorne admits that one basic objection to any ontological argument may remain in force, namely the objection that it is a mere assumption that ‘greatest conceivable’ and ‘inconceivable as non-existent’ are themselves consistently conceivable. He does, however, maintain that the other objections are dissolved and that therefore the whole question needs investigation de novo.34 Whether his own doctrine of a developing God leads to fresh difficulties is not discussed. It is an interesting fact that Hartshorne has stimulated in the United States a revival of the process-theology which was started by A. N. Whitehead in 1929 but which, perhaps because of the extreme unreadability—and perhaps even more extreme unlistenability—of his Gifford Lectures Process and Reality, made little impact outside rather narrowly academic circles. Let other Gilford Lecturers be warned by his example! It is at least interesting to find the doctrine of a developing God, which has usually been associated with a down-to-earth no-nonsense empirical outlook, emerging from such a sophisticated and alembicated mind as Hartshorne's. One fact remains clear, his almost incredulous contempt for the failure of one great thinker after another to perceive what, as he believes, Anselm really meant and of Anselm himself to see what he had himself really proved.

I shall conclude the present lecture with some mention of a careful and very detailed discussion of the ontological argument which is given by Dr Alvin Plantinga in his book God and Other Minds, published in 1967. Having stated the argument in the form given to it by Anselm in Proslogion II, he then considers Kant's alleged refutation of it and in particular the objection usually stated in the form that ‘existence is not a real predicate’. As modern objectors of this type he instances G. D. Broad, A. J. Ayer and John Wisdom, but the three to whom he gives detailed attention are Broad, Jerome Shaffer and William P. Alston, the last of whom he discusses at considerable length. His conclusion is expressed in the following words:

We have no reason to believe, therefore, either that existence in reality cannot be predicated of a being presupposed to exist in the understanding, or that Anselm's argument necessarily involves predicating real existence of such a being. I think the conclusion to be drawn is that we do not yet have a general refutation of Anselm's ontological argument.35

And again:

No one has produced, it seems to me, a sense for the term ‘predicate’ such that in that sense it is clear both that existence is not a predicate and that Anselm's argument requires it to be one. Nor has anyone shown, it seems to me, that existential statements (or an appropriate subclass of them) are not necessary. Every general argument of this sort with which I am acquainted involves some unsupported premise that does not seem self-evident and that Anselm would scarcely be obliged to accept.36

Having thus, as he holds, refuted the refutations of Anselm's proof, Plantinga goes on to examine in detail the proof itself. ‘In this chapter’, he writes, ‘I shall not argue that no version of the ontological argument can possibly succeed, but only that none of the more obvious ways of stating it do in fact succeed.’37 Thus his final judgment on the ontological argument is agnostic; neither the objections to the argument nor the argument itself has yet been stated in a form that is logically waterproof. His examination is extremely minute and makes use of the refinements of modern logic. His basic point is, to state it very roughly, that, when Anselm compares a God, supposed for the purpose of the argument to be non-existent in reality, with a being precisely like God except that it has the added property of real existence, he is in effect attributing to this being both real existence and its absence, and thus introduces into his argument a self-contradictory proposition. Plantinga makes a number of suggestions of the way in which the argument might be patched up, but condemns them all as unsuccessful.

What we need [he writes] for a really thorough examination of this issue is a complete and accurate account of the predication of the properties of nonexistent beings. Unfortunately I am not able to give such an account. Nonetheless this last form of the ontological argument is as specious as the preceding one. No doubt there are other reasonable interpretations of this Anselmian argument; I can scarcely claim to have refuted the argument überhaupt. But until other interpretations are suggested, the verdict must be that the ontological argument is unsuccessful.38

Nevertheless Plantinga does not leave the matter there. He passes on to consider the rehabilitation by Hartshorne and Malcolm of Anselm's argument in Proslogion III, the argument that the existence of God is not merely a fact but a necessary fact, or, in other words, that God's non-existence is not merely false but self-contradictory and impossible. We have seen Hick's criticism of Hartshorne; Plantinga devotes his attention to Malcolm, and to much the same effect, for he accuses Malcolm of committing a fallacy in his use of modal logic. He sums up the matter as follows:

It is a necessary truth that if God exists, then there is a being who neither comes into nor goes out of existence and who is in no way dependent upon anything else. But from this it does not follow, contrary to Malcolm's argument, that the proposition There is a being who neither comes into nor goes out of existence and who depends upon nothing is necessary; nor does it follow that God exists is necessary. Malcolm's reconstruction of the ontological argument therefore fails.39

The upshot of Plantinga's investigation, which is conducted with scrupulous care and professional expertise, is thus that neither the ontological argument nor the usual refutations of it are strictly valid. And this would seem to represent the present state of the question. Gilson has seen the most significant fact about the ontological argument to be the fact that it emerged in a Christian setting, that is to say, a setting in which it is natural to identify God with Being.

Thinkers like Plato and Aristotle [he writes], who do not identify God and being, could never dream of deducing God's existence from his idea; but when a Christian thinker like St Anselm asks himself whether God exists he asks, in fact, whether Being exists, and to deny God is to affirm that Being does not exist. That is why the mind of St Anselm was so long filled with the desire of finding a direct proof of the existence of God which should depend on nothing but the principle of contradiction.40

As far as Anselm himself is concerned, perhaps the last word may remain with Hans Urs von Balthasar:

Anselm does not distinguish between the natural and the supernatural, knowledge and faith, between the profane and the sacred; for he learned by faith that reason too was created for the sake of faith, nature for the sake of grace, and that both form, by their interconnection, a single revelation of the incomprehensible love of the Trinity.41

  • 1.

    Proslogion, II.

  • 2.

    Meditations, III. It is disputed whether Descartes had first-hand knowledge of Anselm's argument; cf. M. Charlesworth, St Anselm's Proslogion, p. 6 and n. 2.

  • 3.

    Dr Frederick Sontag writes: ‘Anselm introduces a subtle distinction into his discovery that God's perfection requires an ultimately ineffable nature by asserting that, although our terms refer to a being beyond comprehension, the meanings of the terms used in describing God are in themselves fully comprehensible’ (Divine Perfection, p. 40).

  • 4.

    McGill, op. cit., p. 65.

  • 5.

    David M. Lochhead, in a vigorous criticism of M. J. Charlesworth's St Anselm's Proslogion (Religious Studies, II (1966), pp. 121ff), has defended Anselm from the accusation of taking ‘exists’ to be a predicate. Cf. A. Plantinga's discussion, pp. 55ff infra.

  • 6.

    Religious Studies, IV (1968), pp. 133ff.

  • 7.

    I have slightly paraphrased O'Connor's argument, but I hope not misleadingly.

  • 8.

    art. cit., p. 136.

  • 9.

    Wallace I. Matson, in The Existence of God (1965), condemns Anselm's argument on the ground that ‘exists in the understanding’ cannot validly mean more than ‘is understood’ (op. cit., pp. 49ff).

  • 10.

    S.Th., I, viii, 3.

  • 11.

    The Many-faced Argument, pp. 71ff.

  • 12.

    McGill, op. cit., pp. 86f.

  • 13.

    He Who Is, 2nd ed., Appendix.

  • 14.

    The Spirit of Medieval Philosophy, p. 59.

  • 15.

    op. cit., p. 110.

  • 16.

    Religious Studies, V (1969), p. 124.

  • 17.

    op. cit., p. 89.

  • 18.

    ibid., p. 92.

  • 19.

    F. Ravaisson, Rapport sur la philosophie contemporaine en France, p. 271, cit. Forest, op. cit., p. 277.

  • 20.

    Forest, op. cit., p. 293.

  • 21.

    ibid., p. 294.

  • 22.

    ibid., p. 295.

  • 23.

    ibid., p. 299.

  • 24.

    ibid., p. 300.

  • 25.

    Logic of Perfection, pp. 49ff; reprinted in The Many-faced Argument, pp. 334ff.

  • 26.

    op. cit., pp. 347f.

  • 27.

    ‘Some Comments on Hartshorne's Presentation of the Ontological Argument’, Religious Studies, IV (1968), pp. 103ff.

  • 28.

    ‘Analytic Philosophy and Language about God’, in Christian Philosophy and Religious Renewal, ed. George F. McLean, p. 55 and n. 13. Dr Anthony Kenny makes the same point:

    To say that God is a necessary being is not necessarily to say that ‘God exists’ is a necessary proposition: when Aquinas uses this description of God in the Summa Theologiae he means simply that God is imperishable. Moreover, it has not been shown that the necessity of necessarily true propositions derives from human convention; there is much evidence, in the recent history of philosophical logic, in the contrary direction. The notion of necessary being, therefore, has not been shown to be incoherent [The Five Ways, p. 2, cf. ch. iv passim].
    Kenny's full argument is in his essay in British Analytical Philosophy, ed. B. Williams and A. Montefiore, pp. 131 ff.

  • 29.

    art. cit., p. 105.

  • 30.

    ‘Can God's Existence be disproved?’; reprinted in Plantinga, The Ontological Argument, pp. 111ff.

  • 31.

    Plantinga, p. 120.

  • 32.

    Language, Mind and Value, pp. 8f, cit. Plantinga, p. 131.

  • 33.

    St Anselm's Proslogion, p. 73.

  • 34.

    Anselm's Discovery, pp. 301ff.

  • 35.

    op. cit., p. 63.

  • 36.

    ibid., p. 64.

  • 37.


  • 38.

    ibid., pp. 81ff. Dr Jaakko Hintikka, in an essay ‘On the Logic of the Ontological Argument’ (Models for Modalities (1969), pp. 45ff), maintains that there are senses, not purely grammatical, in which existence clearly is a predicate; nevertheless he holds that ‘Gaunilo, Aquinas, and Kant… appear to have been shrewder—or perhaps merely sounder—logicians than St Anselm and Descartes’ (p. 52). His definition is that ‘x is an existentially perfect being… if and only if it exists, provided that anything at all exists’ (p. 46). It is not clear to me that this is equivalent to Anselm's definition (i.e. that any being that satisfied Hintikka's definition would satisfy Anselm's and that any being that satisfied Anselm's definition would satisfy Hintikka's) or indeed that it is equivalent to any other of the classical definitions of God, though Hintikka clearly thinks that it is. He writes:

    Our argument was couched in terms of one particular attempt to define God as the existentially most perfect being—‘a being than which a greater (existentially greater!) cannot be conceived’. It can be shown, however, that no other characterisation along similar lines can succeed any better. By reviewing all the different characterisations that one may try to give of an existentially perfect being—or of any being, for that matter—in the sole terms of the predicates of identity and existence, the concept of knowledge, quantifiers, and propositional connectives, one can see that no one of them makes an essential difference to our attempts to prove the existence of a being so characterised.… Suffice it to say that it is a straightforward consequence of the adequacy of any reasonable system of epistemic logic that I know of. [p. 51]
    Hintikka's argument is largely expressed in symbolic terms.It may be interesting to add that, on one interpretation of the Löwenheim-Skolem theorem, even pure mathematics cannot be satisfactorily formalised without some appeal to the empirical realm (cf. J. A. Benardete, Infinity (1964), pp. 264ff).

  • 39.

    ibid, p. 94.

  • 40.

    The Spirit of Medieval Philosophy, p. 59; cf. my He Who Is, p. 35 and appendix to 2nd ed.; Existence and Analogy, ch. ii.

  • 41.

    Essays in Theology, II: Word and Redemption, p. 83.