| VIII. The Subject of All Change (Cosmological Argument) | Index | X. Conclusion (Summary of the Arguments) |
Author: Charles Hartshorne
[p. 299]
Where would such an idea, say as that of God, come from, if not from direct experience?,. . . No: as to God, open your eyes — and your heart, which is also a perceptive organ —and you see him. But you may ask, Don’t you admit there are any delusions? Yes: I may think a thing is black, and on close examination it may turn out to be bottle-green, But I cannot think a thing is black if there is no such thing as black. Neither can I think that a certain action is self-sacrificing, if no such thing as self-sacrifice exists, although it may be very rare. It is the nominalists, and the nominalists alone, who indulge in such skepticism, which the scientific method utterly condemns.
Charles Sanders Peirce, in Collected Papers, Vol. VI
The ontological argument turns logically upon the unique relation between the possibility and the actuality, the “essence” and the “existence,” of God. With ordinary finite ideas the task of knowledge is to decide among three cases: (1) the type of thing conceived is impossible, and hence non-existent (e.g.,a moral being totally without “freedom”); (2) the type of thing is possible, but there is no actual example (a Euclidean space?) ; (3) the thing is possible, and there is an example (a speaking animal) . The ontological argument holds that with the idea of God only two of these three cases need be considered, since one of the three, (2), is meaningless. If, the argument holds, there exists no God, then there also can be no possibility of the existence of a God, and the concept is nonsense, like that of “round square.” If, further, it can be shown that [p. 300] the idea of God is not nonsensical, that it must have an at least possible object, then it follows that it has an actual object, since a “merely possible” God is, if the argument is sound, inconceivable. Where impossibility and mere unactualized possibility are both excluded, there nothing remains but actuality, if the idea has any meaning at all.
The ontological argument itself does not suffice to exclude the impossibility or meaninglessness of God, but only to exclude his mere possibility. Or, as Leibniz said, it must assume that God is not impossible. (We shall consider presently whether the argument can be extended so as to justify this assumption.) The inventor of the argument, ‘Anselm, took it for granted that the man with religious experience, to whom he addressed his discourse, though he may doubt God’s existence, will not easily doubt that in hoping that there is a God he is at least hoping for something with a self-consistent meaning. Now, given a meaning, there must be something which is meant. We do not think just our act of thinking. What we think may not be actual, but can it be less than possible — unless it be a selfcontradictory combination of factors, singly and separately possible? In short, when we think, can we fail to refer to something beyond our thought which, either as a whole or in its elements, is at least possible? Granting this, the ontological argument says that, with reference to God, “at least possible” is indistinguishable from “possible and actual” (though, as we shall see, “possible” here means simply “not impossible” and has no positive content different from actuality). Let us now present the reasons for the contention that “at least possible” and “actual” are indistinguishable in the case of the divine.
According to one theory of possibility, a given type of entity is possible if the most general features, the strictly generic characters, of existence or of the universe are compatible [p. 301] with the production of such an entity. Thus, there is no contradiction of the most general features of reality in the supposition that nature has really produced Mr. Micawber. There is contradiction of the details of nature (such as the detail that Micawber is a character in a novel written by a highly imaginative author) , but these may be supposed otherwise without destroying the meaning, the generic content, of “existence.” But the idea of God is the idea of a being everlasting in duration, and independent, in a certain aspect of his being (in his individual “essence”), from everything else. Such a being could not be produced, since he must then be both derivative and underivative, everlasting and yet not everlasting. To create the omniscient, one must endow him with a perfect memory of the past before he existed; to create the omnipotent, one must endow him with incomparably more power, a metaphysically different order of power, than that which created him. It is hardly necessary to prolong the discussion: no theologian holding either type-one or type-two theism has ever rejected that portion of the ontological argument which consists in the proof that God could not be a mere possibility; and (as we are about to show) it is demonstrable that in order to reject this proof one must construct a theory of possibility which would not be required for ordinary purposes, so that the tables may be turned upon those who accuse the argument of making God an exception to all principles of knowledge. The argument does make God an exception, but only in the sense that it deduces this exceptional status from a generally applicable theory of possibility together with the definition of God. Nothing else is required. The opposition, on the contrary, sets up a general principle which, but for God and the desire to avoid asserting his existence (as following from his possibility) , would be without merit.
[p. 302]
It might, however, be thought that “possible " need not mean the consistency of the supposition of the thing’s being produced, or of its coming into existence due to some cause. Only with one type of thing, it may be held, does “possible” mean this. With another type, consisting of things with universal extent in time, a thing either just always exists or just always lacks existence, either status being possible, although no temporal cause could conceivably effect the difference.
I submit that this is a view so paradoxical that it would hardly be considered at all but for two reasons. One is that it invalidates the ontological argument. The other is that it lends color to the supposition that the laws of nature discoverable by science are eternal laws, although their non-existence is logically possible, and although, as eternal, they could never have been produced, constituting, as they do, the very machinery of all production, the presupposition of all events. The alternative to this supposition about laws is the idea that the laws of nature with which physics deals are themselves produced by the cosmic process, the most general principles of which are beyond “law” in this sense. (There must be some sort of law governing the production of laws, but this higher law is of another order, and may be conceived as the aesthetic principle of the value of order as such, and of the no less. real value of a certain element of freedom and disorder, of surprise and novelty, as well as repetition and predictability.) On this view, nothing is possible and at the same time not actual unless at some stage of the cosmic evolution the forces were such that there is no contradiction in the idea of their having taken a turn which sooner or later would have led to the production of the thing in question. Thus, if nature had developed other habits —and who shall say she could not have? — other “laws” would have [p. 303] obtained. But clearly God could not be possible in this way, and he is the only consistently conceivable object which must be conceived as unproduced, a reality always existing or never existing or even capable of existing, either in essence uncaused or a mere nonentity.
The old objection that if a perfect being must exist then a perfect island or a perfect devil must exist is not perhaps very profound. For it is answered simply by denying that anyone can conceive perfection, in the strict sense employed by the argument, to be possessed by an island or a devil. A perfect devil would have at the same time to be infinitely responsible for all that exists besides itself, and yet infinitely averse to all that exists. It would have to attend with unrivaled care and patience and fullness of realization to the lives of all other beings (which must depend for existence upon this care), and yet it must hate all these things with matchless bitterness. It must savagely torture a cosmos every item of which is integral with its own being, united to it with a vivid intimacy such as we can only dimly imagine. In short, whether a perfect God is sense or nonsense, a perfect devil is unequivocally nonsense, and it is of no import whether the nonsensical does or does not necessarily exist, since in any case it necessarily does not exist, and its existence would be nothing, even though a necessary nothing. Clearly, again, an island is not in essence unproducible and self-sufficient. Of course one can arbitrarily put concepts together and suppose that an island which could never be destroyed and had never been produced would be better than one capable of production, since some form of eternal life might go on upon it, undisturbed by any possibility of an end to such a world. But it is not apparent what would make such a world an island, if the “waters” which “washed” it never wore its shores, and if it were not a part of the surface of a body in space [p. 304] surrounded by other bodies capable of smashing it to pieces, and were not composed of particles capable of ultimately separating, etc.
The question is if such a conception would in the end be distinguishable from the idea of the cosmos as the perpetually renewed body of God, that is, not an island in the least, but an aspect of the very idea of God whose self-existence is upheld by the argument. The question is, Cana possibility be real, unless it would, if actual, be an effect of a cause which is real, or the effect of a possible cause which, if actual, would itself be the effect of a cause which . . . (the series ultimately terminating in a cause which is real)? Otherwise, possibility is something wholly apart from actuality, something no experience could ever reveal or evidence support.
I may be told that “logical possibility” is simply selfconsistency and that no further reality than this consistency is required. But the reply is that the meanings whose consistency is granted must mean something, and this referent of the meanings is not the consistency but the presupposition of there being any meanings, consistent or otherwise. If a consistent meaning means something, but something not even possible, then it means something very odd indeed. If it means only its own consistency, then it is really meaningless.
Let us be empirical. I may think of any object of any color I choose; will it be denied that an object of this color is consistently conceivable as a production of “nature"? In fact, of course, objects of at least approximately the same color have been actually given in my experience. The step “from thought to reality” is merely the reverse reading of the step from reality to thought without which there is no thought, as the very logicians who attack the ontological argument on the ground that it seeks to “derive existence from a mere idea” would be the first to grant. We are always in contact with the forces which produce realities, [p. 305] and hence we can think both actual and possible objects. Or, in other terms, we can distinguish, in the reality some portion of which is always given to us, between the essential or generic features and the details, and can see that this distinction implies that mutually incompatible details are both or all compatible (separately, though not together) with the generic features. But God is not a detail, and only contradiction results from trying to make his possi bility conceivable in the fashion in which alone mere possibility is ever really conceived.
We may go further. The reason God is not a detail, whose existence would be one of two equally conceivable alternatives, is that he is really the content of “existence,” the generic factor of the universe. To conceive God is not to conceive what might exist, but what “existence” itself must be — if the idea of God is not meaningless. Either God is nothing at all, or all else that exists exists in and through him, and therefore contingently, and he himself exists (in his essence, though not in his accidents) solely in and through himself, that is, necessarily. The cosmological argument showed that only “God” makes clearly conceivable the flexibility of the generic features of existence by which alternative details of existence can, as alternatives, be real. Alternativeness is one way of looking at creativeness, and the essential or cosmic creativeness is the divine, and nothing else.
Thus to make God’s existence exceptional in relation to his conceivability is a result, not a violation, of the general principle of existence. Whatever is merely possible, this possibility as such is real, is other than nothing, only thanks to something which itself is not merely possible but is reality itself as self-identical, or as that which, being the ground of possibility, is more than merely possible. It is an implication of the idea of God that he is that ground.
At some point potentiality and actuality must touch, and [p. 306] at some point meaning must imply existence. God is the general, the cosmic and everlasting, the essential or a priori case of the unity of essence and existence, and he is this because he is supreme potentiality as existing power, a real agent who eternally does one or other of various pairs of alternatives which he “can” do. All meaning implicitly asserts God, because all meaning is nothing less than a reference to one or other of the two aspects of the cosmic reality, what it has done or what it could do — that is, to the consequent or primordial natures of God.
It has been objected to the ontological argument that existence is not a predicate, and hence cannot be implied by the predicate “perfection.” But if existence is not a predicate, yet the mode of a thing’s existence — its contingency or necessity of existence — is included in every predicate whatever. To be an atom is essentially to be a contingent product of forces which were also capable of not producing the atom, and doubtless for long ages did not do so. Again, contingent existence (the equal compatiDility with existence or its negative) is implied by such predicates as those describing a man. His weaknesses imply that it is not true that he is the master of existence, able to exist through his own resources. The strength of God implies the opposite relation to existence. “Self-existence” is a predicate which necessarily and uniquely belongs to God, for it is part of the predicate divinity. It is part of the nature of ordinary causes that they are themselves effects of causes which antedate them. It is part of the nature of supreme causality that it is coextensive in time with all causal action. (Not that God’s action is in no sense affected by causes, for the law of action and reaction may apply to God; but simply that God, as an individual, cannot have originated out of pre-existent individuals. His existence is uncaused, whether or not all his properties [p. 307] are. Or, otherwise expressed, his essential properties, being one with his existence, have no ground in other individuals; but he may be subject, in spite of the Thomists, to accidents whose explanation is in part to be sought in the accidents occurring in other individuals.) To be God is essentially to be the supreme productive force itself, unproduced and unproducible (except in its accidents) by any force whatsoever. Hence either God is actual, or there is nothing which could be meant by his possible existence. Thus that God’s essence should imply his existential status (as contingent or necessary) is not an exception to the rule, but an example of it, since the rule is that contingency or non-contingency of existence follows from the Kind of thing in question.
There is another way in which the argument illustrates rather than violates general principles. The argument is not that God’s individual nature implies his existence, while other individual natures do not. It may reasonably be held that every individual nature implies existence, and indeed is an existence. By regarding possibilities alone, one can never reach any truly individual character. Individuation and actualization are inseparable by any test, since individuals as such are known only by pointing. Description of contingent things gives always a class quality, unless in the description is included some reference to the space-time world which itself is identified as “this” world, not by description. But “perfection,” as we shall see presently, is the one description which defines no class, not even a “onemembered” one, but either nothing or else an individual. If, then, it is true, as it seems to be, that mere possibility is always a matter of class, then the perfect being, which is no class, is either impossible or actual — there being no fourth status.
But if every individual quality implies existence, must [p. 308] not all individuals exist necessarily? The answer is that contingency is not a relation of existence to a thing, but of a thing to existence. To say a thing might not exist is not to say there might be the thing without existence. It is rather to say there might be existence without the thing. To pass from the actual to what might be is to generalize, ultimately to refer to the uttermost generalities. It is the world (in its generic features) which does not imply its contingent inhabitants, not the inhabitants which do not imply the world with themselves as.its existing parts. They do imply it. Without it they, as individuals, would not be, even as possible. There is an unutilized possibility of individuals, but not an individuality of the unutilized pos ty. Mr. Micawber is a quasi-individual, with some of the aesthetic properties of an individual, but not an individual in the strict sense. He is a class, specific enough to simulate an individual for the purposes of the aesthetic illusion or “make-believe.”
The unique status of God is that no distinction can be drawn between any individual having perfection and any other. Every perfect being must have the same space-time locus (omnipresence) , and must know the same things — all there are to know. If there had been another world, the God of this our world would have known it, for the very possibility of another world can be related to God only as something he (not some other God) could have done or can still do. Hence “the perfect” is no class of possibilities, all of which might be unactual, but only an individual character belonging to nothing, not even potentially (for the only individuality that could be involved is already involved) , or else belonging to the one real perfect individual.
The necessary being is, then, that individual which existence implies, and which itself implies, not simply existence [p. 309] (for every individual does that) , but implies, through the identity of its generic with its individual character, that (so far as its primordial nature is concerned) there is in its case no separation between possibility and actuality, the class and the individual. In other words, “perfection” implies that existence itself necessarily contains a real perfection, or that existence, in its cosmically essential features, is perfection as existent, as the unity of being and possibility. Or, perfection implies that existence, any and all existence, implies the existence of perfection as its ground.
Again, to conceive a thing in two alternative states, actual and possible, is to conceive something common to these two states, as well as something different. But between the world with God and the world without God nocommon feature could be found. For the world with God is the world completely dependent upon the existence of God, for both its actuality and its possibility, and hence it follows that in the absence of God nothing of the world as it would be with God could be identified.
Doubtless these are all ways of construing the one simple principle: nothing but existent perfection could make perfection possible, or rather, perfection cannot have the dependent relation to other things implied by the status of mere possibility, but must have either the status of an impossible idea or pseudo-idea, or else must be simply actual, with no alternative of non-actual possibility at all.
If it be thought suspicious that the ontological argument argues from a unique relation of God to existence (though one deduced from the normal relation plus the definition of perfection) , let it be remembered that, by definition, God’s relation to every question is unique. He is the unique being, unique because maximal, the only unsurpassed and unsurpassable being (in senses A and R). [p. 310] Naturally, God’s relation to existence is maximal also, that is, he exists under all possible circumstances, times, and places, in other words, necessarily. That which would exist, if at all, necessarily, cannot be non-existent and yet possible, for this would mean having existence as a contingent alternative, and a contingent alternative cannot be necessary. To object to this is to object to the idea of God, and not merely to the affirmation, “There is a being corresponding to the idea.”
If all individuals are contingent, then the whole of existence is contingent, and it might be that nothing existed, or it might be true (though nonsensical) that there was nothing of which any proposition would be true. Furthermore, what could constitute the identity of existence as such, if not an eternal and necessary individual manifested in all individuals? We human beings tend to carry our own personality with us in all our hypotheses, in so far as we say to ourselves, Suppose I were to experience so and so. This gives an aspect of identity by which we might try to define existence as such. But the definition would be solipsistic. Hence there must be some further aspect of identity, like ourselves in being a concrete existent, but unlike us in being able to constitute the unity, the all-embracing register of existence itself, without limitation upon conceivable variety and independence. This is what God is, the all-embracing register of existence, perfect in his flexible and tolerant (“merciful”) sensitivity to all experiences, who can see things as they see themselves, also as other things see them, and also as they are related without distinct awareness on the part either of themselves or of other imperfect things.
It is to the credit of the ontological argument that it has to be opposed by making an absolute disjunction between meaning and its referent, reality, or between universals [p. 311] and individuals, a disjunction at no point mediated by a higher principle. Only if there is one actual individual whose presence is universal, have universals an intelligible ground in actuality. Otherwise we have to relate mere universals and mere individuals by — what? Ordinary individuals, being non-universal in their relevance, cannot explain the identity of the universals as such. Aristotelian objections to disembodied universals can be sustained only if there be a universal embodiment, a “concrete” universal so far as present actuality is concerned, though a universal which is also (contrary to Hegelianism) abstract so far as the future and potentiality are involved.
Thus there is not from any point of view good reason to object to the exceptional status of God’s existence, every reason to welcome it as the completion of the theory of meaning.
It is often said (and with an air of great wisdom) that a “mere idea” cannot reach existence, that only experience can do that. But there is no absolute disjunction between thought and experience. A thought is an experience of a certain kind, it means through experience, even when it reaches only a possibility. A thought which does not mean by virtue of an experience is simply a thought which does not mean. Therefore, if we have a meaning for our thought of God, we also have experience of him, whether experience of him as possible or as actual being the question. It is too late to assert total lack of experience, once meaning has been granted. The only doubt can be whether the experience, already posited, is such as to establish possibility only, or existence also. But in the case of God no distinction between “not-impossible” and “actual” can be experienced or conceived. Hence we have only to exclude impossibility or meaninglessness to establish actuality.
[p. 312] Moreover, since God is conceived as all-pervasive of actuality and possibility, if we do not know God as existent, it cannot be because we have been denied some requisite special experience, since either any experience is sufficient, or else none could possibly be. Or, once more, either God is a meaningless term or there exists a divine being.
In still other words: either the idea of God is less than an idea, or it is more than a “mere idea” such as might designate an unactualized possibility, and isa direct awareness of an actual deity —as not only the mystics, but most theologians, have maintained. “Deity” may be nonsense, but a mere idea it cannot, without nonsense, be. To paraphrase Kant’s final remark on the subject, all disputation about this, the real, point of the ontological argument is labor lost, as much as disputation about arithmetic. To say God cannot be a mere potency and to say two and two cannot make five differ in the degree of clearness of the ideas involved, but not in the a priori, or (relatively) selfevident, character of the reasoning.
That the ontological argument is hypothetical we have admitted. It says, “If ‘God’ stands for something conceivable, it stands for something actual.” But this hypothetical character is often distorted out of all recognition. We are told that the only logical relation brought out by the argument is this: The necessary being, if it exists, exists necessarily. Thus to be able to use the argument in order to conclude “God exists necessarily,” we should have to know the premise “God exists.” This makes the argument seem ludicrous enough, but it is itself based on a self contradictory assumption, which says, “If the necessary being happens to exist, that is, if as mere contingent fact, it exists, then it exists not as contingent fact, but as necessary truth.” Instead of this nonsense, we must say, “If the phrase necessary being” has a meaning, then what it means [p. 313] exists necessarily, and if it exists necessarily, then, a fortiori, it exists.” The “if” in the statement, “if it exists, it exists necessarily,” cannot have the force of making the existence of the necessary being contingent — except in the sense that the argument leaves it open to suppose that the phrase “necessary being” is nonsense, and of course nonsense has no objective referent, possible or actual. Thus, what we should maintain is, “that which exists, if at all, necessarily,” is the same as “that which is conceivable, if at all, only if it exists.” Granting that it is conceivable, it then follows that it exists because it could not, being an object of thought at all, be a non-actual object. Or once more, the formula might be this: The necessary being, if it is not nothing, and therefore the object of no possible positive idea, is actual.
Kant confronted the ontological argument with a dilemma: either the argument is analytic, and then it begs the question by defining God as actual while pretending to derive his actuality from his mere possibility; or it is synthetic, and in that case actuality is added to possibility without warrant. But the argument consists in showing that “mere possibility” is meaningless with respect to God, and the inference is indeed analytic; but its premise is that God is not impossible, which — since God cannot be merely possible — leaves as the only case analytically allowed, that he is actual. It is not true that all things have two conceivable states, possibility and actuality. To assume this, and on the basis of this assumption to accuse the argument of begging the question, is for the accuser himself to beg the question. “Not impossible” or “conceivable” involves three different cases: (1) no such thing exists, but existence is or has been capable of producing it; (2) such a thing exists, but existence might have failed to produce it; (3) such a thing exists, and it is impossible that it should not [p. 314] have existed. In the third case, to conceive the thing as merely possible is to conceive nonsense. The argument for existence in that case will not be that the conception of this nonsense implies the thing’s existence, but that the impossibility of conceiving the thing not to exist leaves but two alternatives: the thing exists and hence can be the object of the conception of its existence, or the thing neither exists nor is a possibility of existence and the conception of it is nonsense, is the conception of nothing, unless of mutually contradictory elements (as is God conceived as sheer A). Thus we should not say that God is “possible therefore actual,” as though he were in one of two states because conceivably he is in another, but we should say that God is not impossible, i.e. inconceivable, and therefore he is in one of the three states given above, but since he is not a contingent but if anything a necessary or self-existent being, eternal, unproducible, etc., therefore only state (3) can apply. There is no question-begging, provided it be admitted that “God” stands for more than an impossibility. If he is not less than a possibility, he can only be more, that is, an existent. The general case of possibility, which does not decide between the three cases, may be called compatibility with existence or conceivability. God is either impossible, or he is compatible with existence in such wise that both the generic and the special aspects of existence imply his existence, but do not imply his nonexistence. Other things than God are conceivable if the generic aspect of existence implies neither their existence nor their non-existence, but is compatible with both.
The famous Kantian example of the hundred unreal dollars (as exactly one hundred, and as precisely dollars, even though they do not exist) is also beside the point, if our analysis is correct. With dollars the distinction between potentiality and full actuality is obviously meaningful; [p. 315] with deity itis, almost as obviously, irrelevant. This is not because existence is one predicate among others, which God, having the maximum fullness of predicates, cannot lack; but rather because the status of non-actualized potency of existence, that is, contingency, contradicts the predicate of deity.
It is in one sense not quite true that unreal dollars can have every predicate of real dollars. For if individuality or determinateness is a predicate, then there is no reason for ascribing this predicate to imaginary dollars. No one could show that an imaginary dollar has any exact shape or color, except by arbitrarily defining it to have, and this would start an endless process. (How to define a precise hue and shade of color?) The “real” might be defined as that which alone is definite, apart from human acts of definition, and transcendent of the utmost that these acts can attain. Now the predicate of deity is unique in that, as Kant himself notes, it is self-individuating a priori. Its definition determines the individual possessing the predicate sufficiently to distinguish it from all other individuals, actual or possible. What is left undetermined refers merely to the alternative accidents of one and the same God. In Whitehead’s terms, the definition determines God’s primordial nature, though not his consequent natures (it should best be put in the plural, for there is a new consequent nature every moment). But it is the same individual being which pervades all the consequent natures which have been or even might have been; just as the adventures of a man, including those he might have had if he, or his friends or enemies, had chosen differently, are all his actual or possible adventures, and no one else’s.
The notion that a predicate must be detachable from existence thus amounts to the idea that individuality is something added to qualities. And if by qualities is meant [p. 316] universal and limited ones, this is true. But absolutely particular qualities can hardly be found except as embodied in actual individuals, and absolutely unlimited properties, such as knowledge that is in no sense or respect less than knowledge, in no sense ignorance, can be found only in the unique and necessarily actual individual, God.
In arguing that there can be no contradiction in supposing a thing absent with all its predicates, but only in supposing a thing present with contradictory predicates, Kant is cither asserting that when we conceive a thing we can conceive it as neither actual nor possible, or else he is asserting that we can conceive God as merely possible. Neither contention should be granted.
Also fallacious is Kant’s attempt to prove the irrelevance of existence to perfection by arguing that if a possible thing had every perfection except one, the addition of existence would only make the thing exist with the same near-perfection which it had as possible. “The point, of course, is, once more, that nothing merely possible has any perfection in the strict sense, and that the transition from any imperfect value, however great, to perfection is not to be attained by adding another finite value, or by any addition. That one could add value forever in thought to the imperfect without reaching existence is only an aspect of the fact that one would never in this way reach perfection either. Unless the perfect is presupposed all along as that which the imperfect lacks and is measured by, then no transformation of the imperfect will define the perfect, which, as Plato said, defines both itself and the imperfect. The holy is not merely that which is without sin, but that which fully appreciates all interests; and we know sin as the deliberate failure to appreciate the totality of interests because we have some sense of this totality, some sense, however weak, of the holy. We do not get to God by adding to the idea [p. 317] of something else, but we do get to something else by subtracting from what we intuitively know of God.
The strongest point in Kant’s immensely influential but hardly too clear criticism of the argument is surely his contention that the argument’s premise, the non-impossibility of God, is not to be accepted lightly. But it is only Kant’s own subjectivistic system, generally rejected by those who in effect rely upon it as the base of his attack on the argument, that makes it seem sensible to him to suppose that a consistent idea might lack even a possible object.
One might of course hold that possibility depends on the constitution of the world, not on our thought. But then we have to ask, Would not another world constitution have been possible? The only ground for a negative reply must be that another world is inconceivable; so we come back to conceivability as ultimate criterion.
The fact that logicians have not elucidated these aspects of the theistic problem is certainly not a question of lack of ability, or of honesty. But logicians seem almost to have said to themselves: “We have, incidentally to our main concerns, disposed of the theologico-metaphysical proof par excellence. Nothing could be more satisfactory. Let us turn to more scientific concerns.” But if mere disposing of an argument is less instructive than understanding it, and if philosophical interests are not identical with those of natural science, then something further is to be desired.
It is said by logicians to be absurd to say, “The such and such (or the perfect) exists.” We must say, There is an x, an individual, such that it has a certain property. Thus: there is an x such that x is perfect (omniscient, etc.). Now the ontological argument merely holds that if this proposition is false, then perfection is imperfection. For if there isno perfect x, then perfection is either a meaningless term, [p. 318] or it means the mere possibility of perfection; but the mere possibility of perfection implies that perfection could come into being, or be produced or have its being derivatively from whatever it is that constitutes its “possibility,” and this amounts to saying that perfection could come into being as imperfection.
Why have logicians denied that we could ever infer from a predicate that something embodies that predicate? The ground appears to be the inductive one that most predicates do not imply existence; therefore we may suppose that none do. Such an inference obviously could not be conclusive. To clarify the matter we should consider carefully the relation between essence and existence in the most widely contrasting cases. By considering “redness” alone we certainly never could discover what things in the world are in fact red. But it may be going too far to say that the predicate redness is conceivable in complete detachment from red objects. If we imagine red, at least our psychological, and perhaps our physiological, state is somehow qualified by redness, and it is a moot point in philosophy whether redness literally does ever qualify anything except minds-and-bodies endowed with color vision. The quantitative properties ascribed by physics to things which we experience as red are distinguishable from redness as given. To ask whether anything is really red means in science whether anything really has these quantitative properties. And while these properties do not in the mere conception imply the reality of external objects precisely embodying them, it is nevertheless true that if we know what we mean by wave lengths and the like it is because we have experiences, and are organisms, which do illustrate in principle though not necessarily in detail what such quantitative aspects are like. In other words, the more fundamental aspects of predicates are always actualized somehow in the experience which refers to them.
[p. 319] The problem is not, whatever logicians may sometimes like to imagine, that of getting from mere disembodied predicates to actualities, but of getting from actualities, such as actual experiences (which include some portions of the actual environment as experienced) , to other real or possible experiences or portions of the environment. This is done by following the tracks of universals, generic features of actuality and possibility alike. By this means, predicates can be approximately (though only so) defined, even though they are not actualized. But if nothing like redness were actual here and now, say as the memory of a real red object, I could not here and now speak of the possibility of redness somewhere else. Nor could I do so unless the idea of a “place” were illustrated by the here and now. Logicians may claim that it is only a psychological, not a logical, necessity that essences should be illustrated in actuality. But the making of such a verbal distinction seems to correspond to no actual evidence.
The truth then seems to be that generic essences imply the reality of some instance or other, and that particular essences are only approximately identifiable apart from their instances. By combining a number of such approximate predicates one may conceive a possible something strikingly unlike anything actual. But simple and quite definite essences are apparently never knowable unless they are embodied.
Here lies a possibility of extending the ontological argument so as to overcome its hypothetical character. The basic paradox of the argument, overlooked by many of those who speak in generous terms about its paradoxicality, is that the inseparability of essence and existence in God seems to imply that it is no easier to know one th: other. As Aquinas, perhaps the best of all the crit the argument, pointed out, to have an intuition of the nature of God would be to have an intuition of his existence, [p. 320] so that any experience which furnished the premise for the argument would also, without the argument, furnish its conclusion. But for second-type theists, who admit not only an identity between the divine essence and the divine existence, but also an identity between the essence of the universe and the divine, the cosmological argument is open to the same objection. And the objection amounts to this, that a necessary truth can be deduced only from a necessary truth, since to know the contingent as such is ipso facto to know the necessary.
All theological truth is one, except in emphasis. All argument for God seeks simply to show that even in denying him we know him, that the conclusion in question is only a clearer way of seeing the premises — any premises which state what at bottom we all believe, or which explicate the most general aspects of any man’s thought. The experience from which we derive the premises for theological (or atheistic) argumentation must no less directly support the conclusion, but not for all that so clearly. We are seeing God in both cases if in either, but not in the same relations, and only when we have brought out a suffi cient number of these relations will we see that it is really God that we are dealing with. Deductive transition among these relations is for the purpose of enabling us to judge whether the entire system of ideas we are operating with really expresses what we intuitively know in all our experience and thinking. It is the same with deduction everywhere in philosophy.
We have seen that impossible predicates are arbitrary complexes of predicates severally thrown together. Now the predicate of deity is not an arbitrary congeries of essences thrown together. As Descartes said in this connection, it is as extraordinary for its unity as for its richness. Yet the argument is here more complex than he recognized, [p. 321] since we have to consider the two aspects of God as brought out in second-type theory. Of the necessary or primordial aspect, what the older theologians said (as though it applied to God as a whole) needs only to be repeated. We can speak of a variety of properties, the usual divine attributes, but they turn out to be merely different ways of conceiving one unique property. Omniscience and omnipotence are not related like having hoofs and having horns. No sense can be made out of omniscience that does not imply omnipotence, any more than feeling can be explained apart from volition, or vice versa. The same applies to the relations of knowledge and goodness in God. To say God is good is only to say his action is decided in full awareness of all the interests affected, and this could no more be without omniscience than omniscience could be without it.
But the consequent or concrete nature of God is not simple in this fashion. It embraces all the positive predicates actualized anywhere. This follows from the primordial attributes themselves, since to be omniscient is to include in one’s experience all that is, whatever it be. How can we be sure that all actual predicates are compatible together in such fashion that they could constitute the content of one experience? As Leibniz said, not all things that are possible are compossible. However, at this point second-type theism escapes a paradox of traditional theism. It does not hold that all possible values are included as actual values in the being of God, but only that all actual ones are included as actual and all possible ones as possible. Now all actual predicates are certainly compossible, or they could not all be actual. The notion of the togetherness of things in God is only the most intelligible notion of their togetherness in “‘ existence,” remembering that the togetherness we are most concerned with is togetherness of life, experience, values, not a togetherness — of no direct [p. 322] consequence to anyone, by definition — of mere dead matter or neutral entities. How can there be a contradiction in the idea of a knowledge of all actuality and of all possi bility? Could we define all actuality, or all possibility, in any other way than as the content of experience as it would be if all vagueness or unconsciousness of reference were overcome with full and clear awareness? And how can the various lives in the world make one world unless the unity of the world is itself living and sensitive to value differences?
Thus it is hard to see how there can be any contradiction in the idea of God as conceived in second-type theism, though there are, as we have seen, plenty of contradictions in first-type theism, according to which God is a mind — without a body; a power — resisted and acted upon by no other power; a will —without change; a knower of the contingent — yet wholly necessary in all his nature; a love —totally unaffected by the fortunes of those it loves; and so on.
Of course, first-type theists may respond to such a challenge by asking if there is not a contradiction in the duality of natures ascribed to God by the new theism. It has been said that the primordial and the consequent natures are really two Gods. This I hold to be a complete misunderstanding. Do a man’s character, so far as constant, and the man’s concrete stream of experiences make two men? Then why should the distinction between the abstract identical aspect of God and his concrete diverse aspects (which are infinitely plural, except that one may speak of them all generically —as the “consequent nature” —when they amount merely to the necessity that there be some unique concrete nature for each moment of time) generate two Gods?
Yet the idea of God might be regarded as an arbitrary [p. 323] and hence perhaps self-contradictory compound in another way. Perfect knowledge, it might be said, is the result of the uniting of the ideas of knowledge and of perfection, or, again, “knowing all” is the union of knowing and totality. But it is not so simple. Perfection would be meaningless without knowledge. Totality is already implied in the idea of knowledge. The knower has, by the meaning of knowledge, a world to know. He may know only a part of it, but this part as such belongs to the whole which omniscience would know. That the part is known without knowing the whole implies that the part as such is not known altogether as it is, or in other words that even the part is not known without qualification. But to qualify knowledge of the part we must understand what knowledge without qualification or in its simplicity would be. The complex idea here is that of partial or limited knowledge. Evidence of this is seen in the history of thought, in which the first version of assertions is in the absolute form, while the relativity of human truths is only very painfully and late appreciated. Thus omniscience is not the result of qualifying thé idea of knowledge derived from our own knowledge but of removing the qualifications we have, still imperfectly, learned to make in our own knowledge by using the more or less subconscious vision of God. God is not fundamentally negative — the non-finite — but the non-imperfect, that is, the perfect knower, the knower simpliciter, the knower who is never less than knower, the righteousness which is never less than righteous.
But are there not ideal conceptions which measure existence, and yet whose literal actuality is not so much as possible? Thus perhaps “absolute equality,” or “absolute circularity,” are not even possible existents, but only abstract measures of what exists. Yet, though impossible, they are not self-contradictory. And may not absolute [p. 324] goodness, or perfect knowledge and power, be similar “regulative” ideas, without constitutive import?
This is the most persuasive way, so far as I can see, to formulate the atheistic hypothesis. It is plausible to say that an actual perfect circle is impossible, and yet the idea of perfect circularity is free from contradiction. It is also plausible to say that circularity is a general idea without intrinsic individuality, and hence it may be rather an implicit comparison of individuals than a reference to any one as it is or could be in itself. Perhaps the idea of an absolutely circular individual really is a contradiction, destroying the very idea of individuality which it presupposes. The idea of circles would still have objective reference through the fact that some things are more nearly circular than others, with sheer circularity the ideal limit of the series of more and more perfectly circular things.
But although geometrical ideas are perhaps ideals which it is the very nature of reality, as Plato thought, to embody imperfectly, there is a sense in which these ideas are literally actualized. Geometrical equality may never be absolute, but it is perfectly possible to experience precise equality in another form. ‘There may be just two horses and just two cows in a field, and the number of horses is then exactly, not approximately, equal to the number of cows. Counting may be absolute, provided there is sufficient discontinuity between the units to be counted, as well as an unmistakable similarity between the units. In measurement of length the units are without such discontinuity, and there are all degrees of the similarity in question. We have then to treat as equal lengths which do not for our observation involve a different number of units. Absolute equality would have to mean that for no possible observation, or not for omniscience, would the number of equivalent units be different. ‘Tue NEcEssarILy EXxIsTENT 325
[p. 325] Again, a perfect circle could be defined asa line at every point equidistant from a given point, or as a self-returning line which has everywhere the same shape. Clearly the same sort of problem as those just considered is involved, the same mixture of algebraic and spatial ideas, the former quite capable of literal embodiment, the latter problematic, and in their absoluteness inseparable from the idea of omniscience, of God.
This idea itself seems to be of another order. True, the idea of “all” in omniscience can be got at algebraically. One may speak of all the letters in the equation . But, as we have seen, omniscience is not the mere knowing of all things whereas we know only some, for not a single item of this all is known absolutely as it is except toomniscience. Omniscience is qualitative as well as quantitative. We know nothing with absolute distinctness. Yet our idea of things “as they really are” can apparently only be the idea of them as they are or would be to an ideally distinct experience, i.e., to God. Thus it is “God” that defines “actuality,” not vice versa.
Something like the foregoing argument for the ultimacy of the idea of God is what Berkeley was perhaps trying, not very successfully, to formulate. The argument does not in the least depend upon supposing that what we immediately know is only our own ideas, only states of ourselves. The conclusion reached is not that all objects are merely ideas, or rather states, of God, but that they are such states, whatever else they may be. Nor is the argument one from the “egocentric predicament,” for it reasons from the idea that one’s own self, no less than every other, is relative to a measure of reality which our own half-unconscious awareness cannot furnish for itself any more than for anything else, and which only a clear consciousness could furnish. Thus the reasoning puts the center of things in [p. 326] the divine Other, not in the human ego. We do not fallaciously argue: Objects depend on us, for they are our ideas; but still, they do not really depend upon us, for we cannot produce ideas at will, hence we may suppose they depend upon God. The argument is rather: The immediate object, which is chiefly the life of the living parts of the body, does truly depend partly upon us; our control over it, by voluntary shifts of attention, is real, but very limited and imperfect. The cosmos is a cosmos because it is in all parts subjected to a control in principle like ours over the body but without the defects which express the fact that we are each the mind of a human body, not of the cosmos. Berkeley neglected the mind-body relation and hence did not see the element of dynamic interaction between subject and object, hence regarded the object as purely “passive,” without life of its own, a mere idea of the subject, an egocentered entity. Hence he could find in immediacy no analogy for the power which his theory ascribed to God of producing ideas in the minds of his creatures. We immediately produce ideas, that is, states in the sentient units of our bodies, and they in us; that is, we are contributors to their states and they to ours. God is simply the maximal contributor or cause of ideas in all minds; he is not the sole cause, for he acts in partnership with all other individuals, but the supreme cause.
Nor is the argument that we know things only as known. (by us). It is rather that we know them only as known and as self-known, that is, we refer all contents of experience to more than one center of awareness, or focus of individuality, in abstraction from which things are merely — abstractions, not concrete entities, dynamic units, given or thinkable as such. One’s “own” individuality is only one of the foci in one’s own experience, which is immediately, though in our case not very distinctly, social. The only [p. 327] focus which is necessary to the idea of reality is the divine, which is always present as the point of reference for our sense of our own reality on the same terms as for the reality of anything else.
Thus we define reality in terms of the divine as the experience which is distinct and in a sense complete where ours is vague and partial, hence an experience which confronts no unknown objects — the unknown being not absolutely so (or it could not be spoken of) but the vaguely, poorly known. Yet this definition must be formulated with care.
Simply to say a “complete all-perfect experience” is to raise all the paradoxes of immutable perfection and determinism we have so often pointed out. But our own experience, in its a priori aspects, as always, gives us the clue. There are two senses in which our experience is incomplete and disharmonized, and only one of them is responsible for the sense of ignorance. It is ignorance that I do not know just what I was doing at this moment yesterday, that my memory of that time is somewhat vague and fluctuating; it is ignorance that I do not know just what the possible or probable reader is feeling as I write these lines, that, in short, my idea of “the people in the world who some months from now could possibly read a book of this sort” is so vague. I may some day meet the reader and find out at least something of what he was doing at this very time. In the future I will find out what is already a part of the past, including the tendencies already established for the future. Or in the future I may find out what could be known at any time, such as some truth of mathemati These are the only two sorts of ignorance, and both are discovered by finding that the incompletion of one moment’s experience can be removed at another moment, although it is part of the meaning of the incomplete experience that [p. 328] its completion could have come earlier. Hence the ignorance. But the incompletion which consists in the fact that the thing to be known is itself future, is in the portion of time which is incomplete in essence and not accidentally — this incompletion is not ignorance, and its removal is an addition to knowledge which defines an addition to reality. Omniscience is simply that mode of knowledge in which only this kind of addition to knowledge is possible, the kind identical with the transition, for the object itself, from futurity to presentness. Nothing is vague in the perfect knowledge except as this vagueness coincides with futurity. In our knowledge there is much vagueness which we know is vagueness about the present and past, not about the future only. That this is so is an immediate datum. I not only am vague about what I shall feel, I also am unclear about what I have felt. Vagueness of memory is as immediately known as anything about memory. At any rate this vagueness will hardly be denied. But present perception in our case is vague also, not merely or essentially in that it does not tell what is going to happen, but that it does not tell with distinctness what is happening now. My several sensations, visual, auditory, and the others, have certain qualitative characters, alike or different with respect to each other, and these characters and relations are not as definite to my consciousness as I know they must be in themselves. This vagueness is both given and inferable from what is given.
All we have to do to conceive omniscience is to banish all such vagueness from the idea of experience, but leave that vagueness which defines the futurity of what is future. What I“shall” do tomorrow, that is not only vague now, it must ever be vague. For when tomorrow comes, what is then experienced cannot be what I “shall do” but what I then am doing, and the distinctness of this is compatible [p. 329] with the vagueness of the other. If I say, “I shall do x,” and I do not do x, this contradicts my assertion; but if I merely say, “I shall do something or other within such and such vaguely defined limits, and what I do will not be vague but definite, though what definite thing it will be is vague,” then my assertion is not refuted by the definite thing, within the specified limits, that I actually do when the time comes. If, however, I say that the limits specified are as definite as they could be made in advance, this assertion might be known to be false, either by discovery of a general law of behavior shown by probable induction to obtain, or by direct intuition into the determinate aspects of the future.
Thus, if I knew the future to depend more upon my advance resolution than upon the resolutions or plans or past histories of any other being, and if I knew what these other plans or resolutions or past histories were, then this knowledge plus my consciousness of my own resolution, which would take account of all the other data mentioned, would be the future as given now, that is, as future. (And I should be God.) Yet I still would not know what the future would be when present, for the future is the “determined to be determined somehow” (within more or less narrow limits), not the “determined to be determined just precisely in such and such a way.”
Thus an experience is conceivable which would be clear about its unclarities, and would have no unclarities except such as constituted the futurity of the future. There would be left for this experience something to find out about other experiences only in the sense that these other experiences themselves had their futures to find out, that is, to make, to actualize, to get into determinate present form. All beings other than God have to go into the future to find out not solely about the future but about what, for themselves or other beings, is already present or past. [p. 330] That there are these other beings is part of what is vaguely given in the present. The non-ignorant knowledge gocs into the future only to find out about the future, that is, it finds things where they are in time; those of its data that have the mark of pastness or presentness need no further experiencing to complete, but only those that have the mark of futurity.
In this way it seems possible without circularity to define omniscience as a certain completeness and clarity of experience, and reality as the content of such an experience. If this is correct, then the ideal by which imperfect knowledge is judged is the idea of perfect knowledge, of God, not the idea of mere reality. “Things as they are in themselves” only amounts to “things as they would be to a sympathetic intuition whose incomplete meanings were exclusively futuristic.” When we complain that no one understands or knows us fully, not even we ourselves, do we not implicitly appeal to such a perfect sympathetic intuition as would enjoy all our feelings and experiences to date (most of which we ourselves have largely forgotten, or remember with almost infinite vagueness) and our future experiences just so far as these are implicated in the others but no further?
Does anyone think there is some perfectly mindless slate, called “truth,” or “reality,” or “the past,” upon which all past experience that is incompletely defined in our present human experiences is distinctly inscribed? Yet what else can an atheist think (of course subconsciously, for, may I repeat it, the main difference between theists and atheists can lie only in the proportion of their thinking about cosmic matters which is conscious, and in the harmony or conflict between the conscious and unconscious portions)?
A plausible counter-argument to the ontological is to [p. 331] say that it is precisely the perfect which we should expect not to exist. Heroes with nothing but merits and virtues, like villains with no redeeming touch, are unconvincing. We know at once that they are fictitious. The ideal is the soaring of aspiration beyond the actual, enhancing the good and abstracting from the evil.
But we must beware of a fallacious induction of the form: all contingent realities are imperfect, therefore there is no perfect and necessary reality. Of course, no contingent thing embodies sheer perfection, and therefore conceptions of such a thing which ascribe perfection to it are fictions. And God is not a hero with nothing but merits, he is not the actuality of our aspiration to be as much as possible all that a human being may be. This aspiration is in ultimate truth a soaring beyond the actual. What we seek in regard to it God also seeks for us and himself, because of our freedom not altogether attaining it. But in another aspect God is the strictly superhuman being by reference to whose immutable essence the limitations and the possibilities of man can be measured. He is not what we ought to be, or should like to be, but what it is nonsense to suppose our being, except when his perfection is conceived as limited by certain human properties. The hero is he who trusts his friend or bravely faces his enemy even when he does not know what he is thinking and doing; God knows what all are thinking and doing.
Nor is it true that we reach the divine ideal by abstracting from evil. God is not the being whose life is sheer joy and beauty, but the cosmic sufferer, who endures infinitely more evil than we can imagine. What we abstract from in conceiving God is that which is itself a kind of abstraction, namely, ignorance, lack of interest in the interests of others. God is the concrete unity of the world, not the selected catalogue of its good aspects. This unity as such is purely [p. 332] good ethically, in that it is strictly all-inclusive and does face fully the evil as well as the good, does not evade anything or fail to realize the full quality of things. One may abstract the mere property of inclusiveness or catholicity of interest, the lovingness of God, but in so selecting this wholly good aspect (the primordial nature) we are omitting not only all particular evils but also all particular goods, the whole consequent nature of God, and have merely the fact that God always loves everything, without any of the things he loves. Thus the selection involved in isolating the holiness of God is fair as between the particular goods and evils, simply omitting both, and leaving a mere form of goodness as such, or in general, as holding of all possible states of God. This generic goodness is purely good only from an ethical standpoint, for from the aesthetic it is both good and evil, since the general form of love has the two sides of rejoicing with the joy and sorrowing with the sorrows of others, or of promoting their welfare, and in a manner loving what they love and hating what they hate.
The ethical absoluteness of God is hardly analogous to that fictitiously ascribed to the hero, for it is much more like that unfailing sympathy which a man has for at least some portion of his own body, the portion in the case of man shifting from moment to moment, since man’s body is integral with a cosmos whose mind is not his. God is beyond this form of limitation, the whole cosmos being his body, so that all parts of it are alike his immediate associates in the mutuality which is the connectedness of things.
In any case sinlessness as applicable to man is not holiness as conceived of God, and the two are separated by an infinity. Yet the human holiness is indeed what remains as possible of the divine when we restrict the immediate and vivid sympathy for all to immediate and vivid sympathy [p. 333] chiefly for varying portions of that little part of the world which is a human body, and at most for other human or animal bodies and minds with which the given person’s body is in effective, relatively direct interaction.
The ontological argument as no longer merely hypo- thetical is then as follows: Any predicate is either itself embodied in actuality or is a special case or combination of predicates that are so embodied; the predicate of deity (in its essence or primordial nature) is not derivable from other predicates, is not a special case or an arbitrary combination, but the most original or universal predicate, from which the others are derived by limitation or specialization. It is the unity of the underivable or generic predicates. Now the original unity of our generic concepts cannot be empty of meaning or self-contradictory, for all of our generic concepts depend upon it. Hence the idea of God is a genuine and self-consistent idea, and since it is consistent only when taken as referring to an actual deity, a merely possible one being the same as an undivine divinity, the predicate of deity must exist in a real God.
In this form, in which both the non-impossibility and thence the actuality of God are proved, the argument is no longer merely the ontological, but includes a form of the cosmological as the first step. But Kant showed that the converse relation is also true, that the cosmological is complete only if the ontological is valid. The cosmological proof shows that there must be a necessary (everlasting) being, but this proves God only if the religious idea furnishes the sole way of construing necessary existence. And if it does so, why can we not infer existence straightway from the idea? To this it has been objected that there is a difference between the direction of inference from perfection to existence in the two cases. The cosmological argument says, there must be a self-existent being, every [p. 334] self-existent being is perfect, is divine, therefore there is a perfect, a divine being; the ontological argument, on the contrary, says that every perfect being is self-existent, and according to ordinary logic, all a is b does not imply that all b is a. But the implication does hold provided it be shown also that there can be but one b. To say that all a’s (self-existent beings) are b’s (perfect beings), but some b’s are perhaps not a’s, is to imply either that a is a nullclass (in which case the cosmological argument must be invalid) or else that there is a conceivable plurality of b’s (perfect beings) one of which may be a and another not a (not self-existent). But a plurality of perfect beings is not conceivable. Hence the cosmological argument cannot be valid unless the ontological is so.
A subtle objection here is that of Aquinas, who held that the ontological argument is indeed valid per se (or for God himself) , but that it is not valid for our knowledge, since the ontological inference from perfection to existence proceeds from the (to us) unknown to the to-be-known, whereas the cosmological inference from contingent existence to necessary existence as perfect proceeds from the known to the provisionally unknown, and thereby follows the true order of finite knowing. But this is open to a double objection. Kant brought out one of the objections when he denied that we know contingent existence as such. How do we know that the world is not self-sufficient apart from God? How do we know that things really are contingent, that they could be otherwise than as they are? Scholastics will reply, because things change, and if a thing does not remain in the state in which it is, that state cannot be necessary. Its non-existence occurs, hence it must be possible. But Kant replied, it is possible when it occurs, but is it possible at any other time? Perhaps events are necessary when they occur? To answer this we must analyze [p. 335] what is meant by time and change, by past, present, and future. According to much recent thought at any rate, Aristotle was right, though not radical enough, in holding potentiality an essential aspect of time. The future is neither not-being nor actuality, but real potency.
This line of thought is incomplete until we have generalized beyond all open alternatives to reach the common features of all times, to which there is no alternative because they are presupposed by the very idea of alternation —as its universally common traits, involved on both sides of each and every choice. This non-alternative factor or factors can be understood, the cosmological argument shows, only as the perfection of God in his primordial nature. But who could reach this conclusion, or understand it, unless he had already some intuition, which needed only awakening, of the nature of God? Perhaps nothing could be derived from the cosmological argument by a mind wholly unable to see force in the ontological, for the two are the same relation read in opposite ways, and this relation is one of partial identity in the content of two experiences, the secular and the religious, an identity such that to have no appreciable degree (at least potentially) of the religious must mean that one has no appreciable degree of the secular.
The artificiality of the separation of the cosmological and ontological arguments (as in Aquinas) seems all the clearer in view of the fact that the idea of God which Aquinas (and apparently Anselm and Kant as well) wished to prove is in reality, as we have seen, by all available tests, self-contradictory and impossible, so that the valid ontological argument with respect to it is that God is as conceived impossible. Hence necessarily there is no such being. Truly the saint, and Kant with him, did well to insist that no man knows the possibility of such a God a priori. [p. 336] That is compatible with one’s knowing its impossibility a priori! But the very procedures which define this impossibility also show what qualifications suffice to remove it and to produce an idea which survives every test of conceivability, or at least, whose inconceivability, if it be so, must be of a radically different and more obscure order.
Kant appears to be right then in considering that if the ontological argument has no force the cosmological has none either. But he fails to see that the ontological argument, if valid, does more than furnish a required final step in the cosmological argument. Its validity implies the validity of all the steps in the other argument. (Indeed, all theological reasoning is of one piece. To be certain that it is right or wrong at any point would be to see its correctness or incorrectness at all other points. The difficulty, if not the impossibility, is to be quite certain at any point. We can only be as clear and certain as we are able to be, so to speak.) For if, as the ontological argument assumes, perfection is conceivable, and if, as the argument shows, conceivable perfection implies existent perfection, as that which all existence implies, then the cosmological reasoning, which holds that existence implies perfect existence, could not be invalid. This does not mean that the ontological argument, in its hypothetical form, presupposes for its validity the acceptance of the cosmological. On the contrary, the ontological argument shows that the mere consistency of the idea of God implies the validity of the cosmological argument. Hence, if one had rejected the latter argument but felt that the idea of God cannot be meaningless, one would be forced by the ontological to reconsider the cosmological argument. Thus the ontological supports the cosmological inference not only, as Kant says, by furnishing one necessary step in it, but also by implying that the cosmological argument as a whole is [p. 337] valid, that existence implies the existence of God, unless God is a self-contradictory idea.
Our only reason for any conclusion is some form of experience, and the harmony of secular experience with religious (as yielding the idea of God) is surely a reason for increased confidence in both. To construe this harmony as confirming religious experience by means of secular experience is the cosmological way — which shows that secular experience is incoherent in its generic aspects unless God is taken to exist; to construe the harmony as confirming secular experience (as interpreted by the cosmological a gument) by means of religious experience is the ontological way — which shows that the religious experience is not even coherent in its chief conception or qualitative datum unless this quality belongs to an actual and not a merely imaginary being and the sort of being called for by the cosmological inference.
The cosmological argument says, the world is not even possible, and hence secular experience is nonsense, unless God is actual; the ontological says, God is not even possible, and religious experience is nonsense, not just illusion — and therefore by the cosmological inference (whose validity can be inferred from the ontological) all experience is nonsense — unless God exists.
Just as there are people who deny that “God " need have any rational meaning, so there are those who deny that “universe” or “existence as such” need have any rationally explicable content. In the one case the cosmological, in the other the ontological, argument fails. The final decision derives from the realization through reflective experience of the meanings in question, and of the impossibility of making skepticism in either direction a sincere philosophy. All men, it seems, must ultimately or at least obscurely feel the religious ideal as the referent of all comparisons [p. 338] between interests, presupposing an inclusive interest in interests which can only be God’s and not any merely human interest. (The very ideal of universal tolerance which by a strange result of false religion has come to seem to some the privilege of the irreligious is really but a disguised form of the divine ideal. No one is really fully tolerant except deity, or could fully realize what is meant by the command to appreciate the various actual interests without prejudice.) Again, all men feel themselves parts of a whole, a universe. These two inevitable references have the same referent, as the two arguments show.
The relationship of the two arguments also means that religious experience warrants the theological enterprise. For if there is no possible inference from the world to God, then there can be no God and the very idea is nonsense. Conclusive refutation of the cosmological argument would invalidate the ontological, and vice versa. The arguments show that each mode of experience contains the same implicit affirmations as the other, but each with its own focus of greatest distinctness, the religious experience in its most perfect form containing the maximum all-round explicitness of outline.
The final argument is: the existential or cosmic ultimate —the key conception at which the search for knowledge of the real arrives — and the ethico-religious or value ultimate are one, and the character of this one is in both cases intelligible as deity. Or: experience is adequately guided for either practical or theoretical purposes only by the religious idea, The chief obstacle to agreement on this point has probably been the failure on both sides to distinguish between different dimensions of value, according as they do or do not permit an absolute maximum, with the resulting implication that the value ultimate means the complete actualization of the ideal in unimprovable perfection, [p. 339] whereupon the ideal loses its essential function, as Dewey so well insists — as did James and others before him.
It is hard to see how there could be any strict independence of the secular and the religious arguments. To reflect upon the idea of God is ipso facto to reflect upon its relation to existence and to other ideas. The idea of knowledge, even unqualified knowledge, implies something to know; the idea of power, something over which power is exerted; the idea of good will, that of other interests toward which good will can be extended. To know what one means by God without thinking of what one means by finite minds and a world which they constitute is impossible. Hence to think about God is the same as to think about the world, except in emphasis, and the cosmological and ontological arguments can only be two ways of seeing the same relationships.
To start with the idea of a being worthy of infinite or religious regard and loyalty is to end with, or never to have been wholly without, the idea of the world as integrated by an all-loving power. To start with the idea of the world as, possessing necessarily some sort of unity and order is to end with, it is to have had from the beginning, the idea of an all-loving being as the full explication of this unity. The only “argument” is the identity of these two problems. God is “the world” understood, the world is “God” understood. In both cases we start with (1) the perfect being as vaguely given to intuition and (2) the perfect defined in a more or less definite concept; and we end by verifying the correspondence of the conception to the intuition. The agreement of concepts and percepts is of course the test of all truth. Religious experience provides at least a pseudopercept of the world whole of which we are parts. So does secular experience, but with emphasis upon details or upon abstract aspects like geometrical pattern. Reflection shows [p. 340] that the two percepts describe the same object. This coincidence between world-intuition and God-intuition, secular and religious experience, is the only proof for God. We may trust our idea of what God is because it proves to be simply the full explication of what all our general or cosmic ideas imply, so that even to conceive the untruth of the idea is nonsense, for it would be the untruth of the ideas by which the conception of this untruth, and of any truth or untruth, is made meaningful. The only possible argument for God must show that doubt of God is doubt of any and all truth, renunciation of the essential categories of thinking.
If the theistic arguments are sound, no one is really without faith in God, any more than there are absolute skeptics; but some persons may be in a state of verbal confusion as to their basic beliefs. And certainly there are signs enough of confusion in atheistic writings. There are also many confusions in theistic writings, but they are (a theist will hold) lapses into verbal atheism, while the confusions of atheism are lapses into real, and not merely verbal, theism. The atheist really believes in the integrity of nature as permitting inductions, while the (first-type) theist only thinks he believes in the “timelessness” of deity, that is, by implication, in its non-purposive character, its lack of social passivity, etc. The negative character of atheistic beliefs, some of which are implied by first-type theism, explains how it is possible to confuse them with real beliefs. To deny something may only mean that we are pushing our belief in it down into the subconscious; but to assert a positive predicate we must have something positive in mind and be aware that we do. The atheist means to say that “deity” is a sound that fails to refer to any object; he is denying the significance of the word. But all the time there may be that in his thought to which the word could be attached, and which if redeemed from its obscurity in his [p. 341] mind would turn out to have the positive properties of which theists (in their more precise moments) talk. We need not repeat Descartes’s mistake of supposing that beliefs and ideas which we have promised ourselves to suspend are really made inoperative in our thinking by that resolution, or that introspection can immediately disclose the depths of our own meanings.
The theist must maintain that a philosophy will eventuate in the affirmation of God, or in something either less definite than God or else theoretically incoherent and practically vicious.
Dewey, for example, comes fairly near to asserting theism. At times he does assert something vaguely like it. So does Santayana. Marx asserted something fairly definite, the dialectic of history and of the cosmic process, but this something was definitely wrong, at least in part, and when all that was wrong is removed, what is left is simply a vaguer equivalent of theism. There is no absolute presumption against vagueness. We may have to be vague. But there is a relative presumption against being vaguer than we have to be, and the only way to know the limits is to try more definite formulations until we are checked by disagreements, logical or experiential. Atheisms that are not really idolatry, a vicious form of theism, are deficient in clarity; theisms are often atheistic in some of their implications and hence deficient in consistency. The search for a more definite version of what atheists are trying to affirm, and a more coherent version of what theists have affirmed — these are the two lines of progress which do not imply hopeless stupidity in one side or the other, but admit that each party has been partially right, the one in refusing to regard almost utterly vague conceptions (or else obviously relative principles offered as absolute) , the other in refusing to regard contradictions or ambiguities, as the best that man can do in clarifying his most general ideas.