Sample Material of UPSC Mains Philosophy (Optional) Study Kit

Topic: Philosophy of Religion (Divine Necessity)

TERENCE PENELHUM

THIS paper is a discussion of certain limited aspects of traditional theological doctrines of the necessity of the divine existence and attributes. In spite of being greatly indebted to a number of recent treatments of these themes, I still feel that something worthwhile remains to be said about them, and that certain important morals can be drawn from them regarding the relation of theistic belief to metaphysics and philosophical analysis. The traditional doctrines seem to me to be a paradigm case of the impossibility of presenting basic religious assertions as answers to metaphysical demands for explanation; but when the demonstrative trappings are carefully stripped away, what remains is seen to be of the very essence of theism, and any difficulties inherent in it are distinct from those associated with traditional attempts to demonstrate theism.

I shall be concerned throughout with general and not historical considerations; but those illustrative references I shall make will all be taken from the Thomistic system, since this is the most enduring and meticulous example of the doctrines I discuss, and is in one way or another normative for a great many religious thinkers. It is also considered by many to be the most modest in its metaphysical claims, but I hope to indicate that its relative agnosticism does not make it less open to the criticisms to which all forms of demonstrative theism are subject.

Click Here for UPSC Mains Philosophy Study Material

I

Apart from the Ontological Proof, demonstrations of God’s existence have taken the form of insisting that there is some fact about the world which requires explanation, and that no explanation short of a deity will do. The fact chosen will vary from thinker to thinker or chapter to chapter, but there is one clear division among those facts said to be puzzling in the way required: on the one hand there is the bare fact that there is a world at all, that anything exists whatever; and on the other there is the fact, or indefinitely large group of facts, that the world is as it is, that it contains the particular sorts of quality or relation that it does. Since Kant there has grown up a habit of thinking of all the demonstrations based on each of these as just one argument, and forgetting the great variety of arguments that have been offered of each sort; this may not be very harmful in the first case, but it certainly is in the second: the world is a very varied place, and its orderliness is not the only feature of it that has been picked out as requiring God to explain it, even though this is the feature emphasized in the traditional Argument from Design. To avoid any restrictions associated with long-standing titles, I shall talk of Existential and

Qualitative arguments

In both there are certain features of great importance:

(1) In each case the force of the argument depends upon undermining the tendency to explain the fact singled out by means of more normal explanatory procedures, e.g. scientific ones. The most effective method is to render these irrelevant by making the puzzle too general for ordinary procedures to solve it. In the Existential case, temptationsto explain the existence of a given object in terms of the causal action of another are circumvented by showing that such explanations are congenitally incomplete because they leave unanswered the general question of why anything exists at all. In the Qualitative case, if a certain natural feature were scientifically explained it could be said that the natural regularity of which it was thereby shown to be an example, or at least some all-embracing one of which it in turn was an example, required explanation; the fact of order itself might be picked on for this; or one could just ask why it was that we had the sort of world which, however naturally, gave rise to this or that feature.

(2) Although less perceptive theists have often not seen this, especially in the Qualitative case, the demands for extra-scientific explanation should logically force its users to be discontented with the mere reference to a higher being. If it is puzzling that anything exists, it should seem puzzling that God does; if a certain feature’s presence in the universe is puzzling, then it should seem puzzling that it, or that which gives rise to it, should be present in God. In neither case is the explanation complete. To complete it more has to be built in: the being referred to has to be one whose existence, or whose possession of the relevantattributes, is self-explanatory. The potentially endless series of ‘Why?’· questions has to end in an answer that covers not only the last ‘Why?’ but the next one too. So we have the doctrines of a being who necessarily is, or necessarily is what he is. It is very important to see that exactly the same theoretical move is involved whether it is said that the divine existence or nature is self-explanatory to us or merely in itself; what is essential is that the explanation should be said to lie in the divine being, even if we do not know it.

As Kant saw in part, and Aquinas saw very clearly, the Qualitative question presupposes the Existential. One cannot finally answer the question of why the world is as it is without explaining why it exists, for nothing can have features unless it exists; though this is naturally only a part of the answer, since showing why anything exists is not enough to show why this sort of thing rather than that does. A theist who saw this would naturally begin by asking the Existential question, and then try to show that the being needed to answer it is implicitly enough to answer the Qualitative. This is in part the procedure Aquinas adopts: the Existential query is clearly basic in the Five Ways, but the Qualitative is answered by implication in the discussion of the Divine Attributes, and is, I think, operative also in the Five Ways themselves A word here about these famous arguments. They are regularly treated, since Kant, as though they are in effect one argument, called the Cosmological Proof, which turns out when stated to be identical with the Third Way, or argument ‘from possibility and necessity’. This is not confined to non-Thomists anxious to refute the arguments, E. L. Mascall, for example, says As I see it, the ultimate function of the Five Ways is to make it plain, by calling attention to five outstanding features of finite being, what the fundamental characteristic of finite being is. And that fundamental characteristic is its radical inability to account for its own existence.’

In other words, there is at bottom only one argument: since all the beings we know are such that their existence cannot be explained by reference to them, they must derive it from outside, and ultimately from a being whose existence does not require outside explanation, and who accounts for the existence of anything whatever. This certainly fits the Second and Third Ways well enough, and perhaps the First also; and it can be made to seem a reasonable reading of the Fourth and Fifth, which involve it. But it is surely over-simple to emphasize it to the exclusion of the Qualitative query. Aquinas is surely concerned not only with the contingency of the being of finite things, but also ‘E. L. Mascall, Existence and Analogy (New York: Longmans, Green. and Co., 1949), with that of their manner of being; in the Fourth and Fifth Ways he begins with the degrees of perfection things have and the order they exhibit, and not with their mere existence, which is more immediately his theme in the Second and Third. To borrow his language, just as the essence of finite things does not account for their existence, so the fact of their existence does not account for their essence. The ultimate answer required here is a being whose possession of the attributes necessary to cause those of finite things is self-explanatory, i.e. to be accounted for in terms of him. This explanation, however, would seem to have to be in terms of his existence. What else in him could his essence be explained by? So the two questions coalesce, the second vanishing into the first, perhaps, and we conclude with a being who must, by the sort of being he is, exist, and also must, by the mere fact that he exists, be the sort of being he is: one in whom essence and existence are identical. While granting Aquinas the credit for seeing that the two questions merge, and that the Existential is primary, I think it is mistaken to stress it to the extent of ignoring the presence of the Qualitative dimension from the very beginning.

II

Whether this historical excursus is sounu or not, we have two archetypal questions, which seem to have clear requirements for answers, and which tend to merge. Those who ask for them recognize that the required explanations are of quite a different order from scientific ones. We have to step out of such mundane explanatory processes and enter upon another, for they are inevitably never complete-this is the import of St. Thomas’s troublesome remarks about infinite regresses. The psychological mechanism of this transition is easy to share in, but its rationality is dubious from the start, just because the kind of process we step into is of such a different order from the one we step out of. For this same reason it cannot be regarded as the completion of the ordinary explanatory processes. This is easy to see when we reflect that the only reasons for stepping out of them into it, say, six finite causes back in the series rather than sixteen or six hundred are impatience or fatigue. Philosophers, contrary to common opinion, are prone to both, but they should not found arguments upon them. In the sense in which a Necessary Being explains, contingent causes do not explain incompletely, but are not attempts to explain at all. Sceptics might say that since the concept of explanation is formed by reference to contingent causes, the demands for explanation made in our questions are spurious; a question does not become intelligible merely because of the co-presence of a curiosity-feeling.

To this the metaphysical theist might reply that it is the spuriousness, and not the genuineness, of a question that need demonstrating, and that in the present case it is possible to explain quite clearly what sort of answer would satisfy the questioner. I would be the first to agree that when both sides insist that they are dealing with a unique case, considerations based upon this fact cannot be more than persuasive. But we do not need to rest content with these; nor do we need to consider the perplexing problem of whether there can be a question to which, in fact, there just is not any answer;2 we need merely to assert that there can be no genuine question where the only possible answer to the purported one is demonstrably absurd. And both the Existential and the Qualitative questions are like this.

III (a)

‘Why does anything exist?’ is a total question. There can be nothing not mentioned in the question to bring in to explain what is mentioned in the question. There are therefore three possible ways of answering it. We could say that every individual thing is self-explanatory; we could say that the totality of things is self-explanatory, and individuals explicable by reference to it; or we could say that one part of what exists is self-explanatory, and the rest explicable by reference to that part. Of these three only the last is of interest, since all three make use of the same crucial concept of a self-explanatory being, and the first two have extra problems of their own beside this. The objection to this form of argument centres on the self-explanatory being. Kant held that the argument from contingent to necessary being in the Cosmological Proof reduced that argument to the Ontological Proof, and in spite of the fact that he located this reduction in the wrong place, a his claim is correct.

First, what is wrong with the Ontological Proof? It amounts to the claim that ‘God exists’ is an analytic proposition. The standard, and correct, objection to this is that which Kant raised, viz. that to assert the existence of something is quite different from asserting what sort of thing it is, and to know that either assertion is true is not to know the other is. If someone came in unannounced and said, ‘It’s blue!’ we should not have much idea what he was talking about, but we would automatically know it was a visible physical object and not a philosophical theory or an Act of Parliament. But if he had come in and said, ‘It exists!’ we should know nothing about what sort of thing it was. Existence cannot vary in quantity or intensity, belong to some members of a class and not others, or be interrupted and then resumed. Moore has brought out some of the pecularities of the word ‘exist’ in a very well-known paper. From all this it follows that existence cannot be held to be a quality which a perfect being would have to have, since it is not a quality at all. So it further follows that no existential assertion can be analytic.

So much for the Ontological Proof. It is important to see that what refutes it is not a discovery about the structure of things, which might in a given case be different, but a logical discovery about the concept of existence, which sets it apart from other concepts; that no tautology can be existential is a consequence of this. Another consequence is the refutation of our Existential argument. For the distinctive character of the concept of existence precludes our saying there can be a being whose existence follows from his essence; and also precludes the even stronger logical move of identifying the existence of anything with its essence. These are the Anselmian error all over again. The only other way of explaining God’s existence by his essence would be by asserting a causal relationship between them, but this would run us into absurdities like saying that God would have to pre-exist himself, or that his essence would have to have something almost, but not quite, amounting to existence in order subsequently to express itself in being.

So there is no way in which the existence of any being could be held to be a fact explicable by reference to that being itself. Before passing to the Qualitative argument, there are two important sideissues to discuss. One is the argument of G. E. Hughes6 that ‘God exists’, though not analytic, might still be necessary, i.e. synthetically necessary; the other is the important historical claim that the Theistic position is further removed from the Ontological Proof than any position I have considered, and is therefore unaffected by what I have said.

(i) To say that ‘God exists’ is synthetically necessary is to run counter to fashionable views about necessity in propositions, but, as Hughes insists, one can be out of fashion and right. The difficulty for our present purpose is the notorious one for believers in synthetic necessity, of explaining the necessary character of the examples offered. If all necessary propositions are tautologies, this explains why we cannot deny them; failing an equivalent explanation, purported cases of synthetic necessity seem to have a merely subjective certainty. I do not see what sort of explanation could be had in the present case: certainly a Kantian type of explanation is unsuitable.

(ii) Since Aquinas differs from Anselm in holding that God’s existence has to be inferred from his effects and not from the mere concept of God, he is traditionally credited with having seen what was wrong with the Ontological Proof. He did see it was wrong, but not why it was, for he commits the same error himself. He says that we do not have the requisite knowledge of the divine nature to deduce God’s existence from it; but his own argument leads us from finite beings to a being whose existence does follow from his nature, and this entails that if we knew God’s nature we could deduce his existence from itand this is the mistake. To say that although God’s existence is self evident in itself it is not to us, is to say that it is self-evident in itself, and the error lies here. It is not our ignorance that is the obstacle to explaining God’s existence by his nature, but the logical character of the concept of existence. In order to introduce the morals I wish to draw from this, I shall discuss briefly a recently expressed view of Necessary Being put forward as the basis of discussion by R. L. Franklin. This is that a Necessary Being is just a being in whom the question ‘Why?’ stops, a being about whom it makes no sense to ask it. This looks to claim less than the Thomistic position, and Franklin claims that it offers an intelligible answer to the question of why anything exists, but not a demonstrable one. It is instructive to see why this is unsatisfactory: if we cannot ask why a given being exists there must be a reason why we cannot; if there were no reason, then we could ask this question-such at least is the assumption made in the initial stages of any version of the Existential argument. There are two possible reasons. The first is that the being in question is self-explanatory, which I have already tried to show to be an absurd reason. The second is that although the being in question is the cause of all other beings, there is no other being to be found which is the cause of it. In this case there would be one unexplained being, by hypothesis, and it would not answer the question of why anything existed. FrankIin says that ‘Why does anything exist?’ may not have an answer, but that if it does a Necessary Being (in his sense) would provide it. But a Necessary Being in his sense would not provide it, unless we went on to make it a self-explanatory being and thus reduced it to absurdity. It is not that something could provide an answer to the Existential question but maybe nothing does, but that nothing could. Now for the morals:

(i) It is absurd to ask why anything exists, because the only possible answers are in terms of the logically impossible notion of a selfexplanatory being. This is still logically impossible when it is softened by its user’s saying that we personally do not know the explanation.

(ii) This in itself does not prevent us from saying that the existence of everything in the universe except one is due to that one (though we would presumably believe in that one for reasons not related to our present argument). But it does prevent us from going on to ask why that one exists, for in this context that would be equivalent to asking why anything does.

(iii) But unless you assume independently that a given being has no cause, you can always ask why it exists, i.e. what caused it. If you do assume it has no cause, you ipso facto make it impossible to ask why it is there.

(iv) So there may be a being who is the cause of everything else, but if there is he cannot explain the baffling fact of existence. For it is logically impossible to explain everything. The Principle of Sufficient Reason is demonstrably false.

(v) So the fact that things exist cannot entail the existence of God; it could only do so if God were self-explanatory. Failing this, the ‘Why?’ questions would only come to a halt if we had independent reasons for holding that the being we had reached was uncaused. And it would be these independent reasons that would bring us to theism. And it would not be a theism that explained. Theism cannot explain any more than atheism can.

III (b)

We turn now more briefly to our Qualitative question. The search for the complete explanation of the presence in the universe of any property is bound to lead to the claim that there is a being who has it, or some higher cause or analogate of it, seIf-explanatorily. The objections in this case are similar:

(1) Let us call the property P. Our doctrine could mean that God has P because of some other properties he has, the relationship being causal. This clearly only pushes the problem back to those other properties themselves.

(2) The doctrine could mean ‘God has P’ is a necessary proposition. I will assume this to mean ‘analytically necessary’, since, as before, its being synthetically necessary might give it certitude, but not explanatory power. Now to say that ‘God has P’ is analytic does not solve our problem, which is that of accounting for the occurrence, in the whole realm of being, of P. If it is analytic that God has P, this just tells us that having P is part of what is meant by the word ‘God’, i.e. that no being would be accorded the title who lacked P. But this merely means that there is a connection between the concepts of divinity and P-hood. not why either the first or the second has instances (or even whether either has). To know that ‘Birds have wings’ is analytic is to know something important about the words ‘bird’ and ‘wing’, but nothing at all about why winged creatures came to be. The analyticity of statements about God has been thought to raise more problems than it does. C. B. Martin9 has claimed that since ‘God is good’ is analytic, God cannot be identified with the man Jesus, for ‘Jesus is good’ is synthetic. This can be resolved by using Martin’s own distinction between ‘God’ as a proper name, and ‘God’ as a concept. ‘God (concept) is good’ is analytic; but ‘God (proper name) is good’ is synthetic, and learned, if at all, by experience. In our present case the problem posed by the Qualitative query is that of explaining the fact reported by ‘God (proper name) has P.’ This is not explained by showing that ‘God(concept) has P’ is analytic, even though it is. Only one recourse remains. Since the divine possession of P cannot be explained by reference to the divine nature, it can only be explained by reference to the divine existence. Let us say, then, that God’s having P is a deducible consequence of the fact of his existence. This would only explain his having P if his existence were previously said to be necessary, but let us ignore this. What is said of P would apply, by parity of reasoning, to all the divine attributes; so we come once This qualifying phrase will be assumed, but not stated, in what follows. more to the identity of the divine essence and existence, with some kind of priority being accorded to the divine existence. The connection between the Existential and Qualitative arguments is closer than Kant recognized.

The objections are not hard to find:

(i) Quite apart from the difficulties involved in saying God necessarily exists, we have gone in a circle if we fall back on this here. For we began by explaining his existence in terms of his essence, and we now find ourselves explaining his essence in terms of his existence.

(ii) The logical character of the concept of existence is not only nough to render it inadmissible to infer God’s existence from his essence, but also renders it inadmissible to infer his essence from his existence—or, again, to identify them. The morals to be drawn are the same:

(a) It is absurd to ask why anything has P, for the only possible answers are in terms of the logically impossible notion of a being in whom the presence of P is self-explanatory, etc.
(b) This in itself does not prevent us from saying that the presence of P in any being in the universe except one is due to its presence in that one, etc. But it does prevent us from going on to ask why that being has P, for in this context that would be equivalent to asking why anything does.
(c) But unless you assume independently that a given being has P without cause, you can always ask why it has P, etc.

(iv) So there may be a being who is the cause of all other beings having P who have it, but he cannot explain the baffling fact of P itself, etc.

(v) So the fact that P can be found cannot entail that there is a God who has P; it could only do so if God’s having P were selfexplanatory. Failing this, the ‘Why P?’-questions would only come to a halt if we had independent reasons for holding that the being we had reached had P without cause, etc.

IV

There can, then, be no metaphysical compulsion to believe in God; for the sort of metaphysical questions which would necessitate theism are spurious. This does not refute theism, however. It would only refute it if the sort of explanatory demand we have been considering were inevitably involved in belief in God. This has been argued by J. N. Findlay;l1 he says that the attitude of worship entails independence of all other beings, both in his existence and his possession of his excellences, and entails that he possesses these in the highest conceivable degree; if this is accepted, and it seems it must be, he claims we have to go on to say that God exists and has his excellences in some necessary manner; given the absurdity of this, God’s non-existence follows. Theism can be rescued from his argument. One can agree with Aquinas that it is no limitation on God that he cannot perform logical absurdities; and one can adapt this and say it is no limitation on him that he cannot be a logical absurdity; and that is what a Necessary Being, in the sense we have examined, is. I think the readiness of Findlay’s disputants to agree that God’s existence and excellence are necessary is due to a dangerous and crucial ambiguity in the terms ‘necessary’ and ‘contingent’ (an ambiguity almost, but not quite, recognized by Rainer). It is pedantic of philosophers to insist that these words only apply to propositions and not to things; but our previous discussion should show that they will mean something different in each case. We do not need to say precisely here what propositional necessity is: let us say that, roughly, a proposition is necessary if its truth can be known without reference to anything other than a clear understanding of what is said or implied in it; contingency in propositions is the absence of necessity (and of its contrary, viz. necessary falsehood or, if this is the same, self-contradiction). As applied to things or events, ‘contingent’ will mean ‘dependent’ or ‘caused’, one thing or event being contingent upon another; ‘necessary’ will mean ‘not dependent on any other’, and, in addition, ‘having others dependent on it’. A thing is necessary if it is indispensable. For want of a better phrase I shall call necessity in this sense ‘factual necessity’. To be a theist is to believe that there is a being, God, who is factually necessary, all other beings being dependent, contingent, upon him. But the assertion of this will be a contingent assertion, in the propositional sense (and not in the Thomistic sense that Rainer adopts, viz. contingent to us but necessary for God). God’s existence and nature are unique in the universe in being free of factual contingency, but the assertions of them share in propositional contingency with all other assertions of fact.

Theists believe that God exists, that he is supremely great and good, that no other beings would exist if he did not, and that all their multitudinous features have their source in him. I have denied none of this. It merely means that God is factually necessary, indispensable. I have denied he can be necessary in any other sense, that he explains either his own existence or nature, or, ultimately, that of other things. Since this ideal is bogus, God is not denigrated if he is not held to realize it. If there is a factually necessary being then, this fact, though the most important fact there is, could not be proved. ‘Why, then, be a Theist?’ Well, theism is older than the Cosmological Proof, and can survive it. Some have tried to present it as an ordinary empirical hypothesis, but this has seldom impressed. If theism is to be seen to be true at all, it looks as though this will have to happen through individual confrontation with what purports to be God’s self revelation. A person who accepts this and then proclaims it to others is not free from philosophical criticism just because he does not proclaim his belief as metaphysically necessary; for he still makes statements which do not always appear to others to have the relationship to evidence which all statements (they say) have to have; and if ‘metaphysics’ is defined in the required way, what he says will be metaphysical to them still. But the objections thus raised against theism when it is expressed in undemonstrated assertions by such a person, however strong they are, are distinct from those much stronger ones that can be brought against offering theism as the answer to the sort of confused question we have examined above; and the sort of ‘metaphysics’ that may be contained in them or spun around them is likely to be immune to the objections we have stressed, whatever its other defects may be. Philosophical analysis will not progress much in understanding religious assertions, especially their relationship to metaphysical speCUlation. unless these distinctions are carefully borne in mind.

Click Here for UPSC Mains Philosophy Study Material

<< Go Back to Main Page