Charles Hartshorne
Metaphysics can be approximately defined in a number of ways, including the following:
(a) The unrestrictive or completely general theory of concreteness
(b) The theory of experience as such
(c) The clarification of strictly universal conceptions
(d) The search for unconditionally necessary or eternal truths about existence
(e) The theory of objective modality
(f) The theory of possible world-states, or the a priori approach to cosmology
(g) The general theory of creativity
(h) The search for the common principle of structure and quality
(i) Ultimate or a priori axiology (theory of value in general)
(j) The inquiry into the conceivability and existential status of infinity, perfection, (unsurpassability), eternal and necessary existence
(k) The rational or secular approach to theology
Diverse as these formulae may seem at first glance, I believe that they all imply the same thing and differ only in emphasis and focus of explicitness. “Unrestrictive” says the same as “negates no positive or extralinguistic possibility,” and this is the same as “unconditionally necessary” (d). “Completely general” (a) or “strictly universal” (c) imply applicability both to what is and to what could be, and this coincides with what has no alternative and is simply necessary (d). It also coincides with what is eternal (d, j). The strictly, or unconditionally, necessary must be known, if at all, a priori (f, i) and be valid of all possible world states (f). What is inherent in experience as such (b) cannot be denied except verbally, and must be necessary and knowable a priori (d, f, i, j). “Inherent in experience as such” means exactly what it says; “inherent in human experience as such” would mean something else, and those who can see no great difference are probably not fitted for metaphysical inquiry. The most general factor in creativity (g) will be expressed in any possible creation; as the very principle of alternativeness it will have no objective or ontological alternative, but only a linguistic one, and thus be necessary.
“Theory of concreteness” may be compared to the Aristotelian “being qua being.” Collingwood bases his rejection of metaphysics, or his reduction of it to the making, and historical study, of “absolute presuppositions,” partly upon the argument that “being,” purely in general, can have no distinctive characteristics, and hence there is nothing significant to be said about it. But the theory of concreteness is not the bare theory of being, of what an “entity” is just as an entity. Rather it is primarily the theory of what a concrete entity is, as concrete rather than abstract. From this it appears that pure nominalism is metaphysically wrong, or at least antimetaphysical; for it destroys the distinction upon which metaphysics as such rests. Collingwood’s argument is cogent enough if “completely general” must mean “applies alike to the concrete and the abstract,” or without regard to any such distinction. What it should mean is, “applies in some way to the concrete as such, and in some presumably different way to the abstract.” A purely general theory of concreteness will include a theory of abstractness, but there will be a real distinction between the two. (Actually, Aristotle was not without resources for defending himself against Collingwood’s argument. Nevertheless, I believe we need a much sharper sense of the meaning of “concrete” than the Greeks or the Schoolmen achieved. The “fallacy of misplaced concreteness” is pervasive in the history of philosophy.)
There is another counterargument to Collingwood: it is idle to speak of being, or concreteness either, unless we can relate such a term to experience and knowledge. To generalize the object of thought or experience is to generalize thought or experience itself. And again, please note that “human experience” is restrictive, rather than a harmless redundancy. The unrestrictive notion of being or reality cannot be separated from the unrestrictive notion of experience. Nor—as Plato saw—can this generic notion of experience be separated from the purely general notion of value, the form of the Good. (Note definition i.) Thus at least one thing is true of any entity whatever, that it can be thought, experienced, and valued. Or, as Peirce put it, any concept is a potential contributor to the summum bonum, and how it can contribute is its meaning. In spite of Collingwood, all this is worth saying, and is true, rather than a mere presupposition.
“Being” suggests a contrast to becoming, yet according to Collingwood’s argument, the generic notion of being would ignore this contrast. Whitehead has shown that it need not be so. His formula is: any entity, abstract or concrete, has this in common with every other, that it constitutes a “potential for every [subsequent] becoming.” Here we have, for the first time in intellectual history so far as I know, a definition of “being” in terms of “becoming.” Hitherto, apart from Bergson (ambiguously preceded by Hegel), the effort had almost always been to define becoming in terms of being. This could only result in a fallacy of misplaced concreteness. Becoming is the more concrete conception, and includes within itself all needed contrasts with mere being. In each particular instance of becoming, what has previously become does not then and there become; yet it does constitute a causal factor in the new, and in all subsequent, instances of becoming. This cumulativeness of becoming (first stated sharply by W. P. Montague and Bergson) is causality, the efficacy of the past in the present. And any strictly eternal entity constitutes a factor in every becoming whatsoever. Actual happenings, for instance our acts of thinking now, are what relate themselves (consciously or not) to the eternal, as well as to past, happenings. To try to put all such relations into the eternal is either to eternalize everything, and therefore in no distinctive sense anything, or to dissolve the eternal into the flux, whereupon again the distinctive meaning of both terms is lost. The unique relatedness or relativity of becoming makes it the inclusive or concrete form of reality; all else is abstraction.
If the reader asks, what does “concrete” mean, he has just been given an indication, at least, of the answer. The concrete is the inclusive form of reality, from which the abstract is an abstracted aspect or constituent. Again, the concrete is the definite, for to abstract from details or aspects is to conceive the indefinite. The number “two” is abstract because it is indefinite or noncommittal, for instance, between two apples, two persons, two ideas, two entities. (Extensional logic seems sometimes to obscure more than clarify such problems.)
Plato rightly regarded the Form of Good as more ultimate than “being” taken in abstraction from value. Any such abstraction would conflict with (b), for valuation enters into every possible experience, and hence value enters into every possible object of experience or thought. As theory of concreteness, metaphysics can abstract neither from value nor from experience or thought. It abstracts from this concrete thing, this value, this thought or experience, also from this structure or quality, but not from concreteness, value, thought, experience, structure, or quality in their pure generality. Materialism either has no theory of concreteness, or it tries to construct one while making some of the abstractions just mentioned. But this is impossible. Materialism is a covert dualism, and dualism is simply the refusal or failure to achieve a general theory of concreteness.
That metaphysics is theory of concreteness is implied by Whitehead’s description “the critic of abstractions.” The abstractions are criticized, not, as in science, because they are inaccurate to the facts, but because other equally general or even more general abstractions are left out of account, and thus the general meaning of “concreteness” is not brought out. In all human knowing there must be abstraction, disregarding of details, but it is one thing to disregard details and another to disregard aspects quite as universal as those taken into account. Physics abstracts from mind, and even more obviously from value and quality as contrasted to structure. Yet any entity, even in merely being mentioned, is in some minimal way being related to experience and assigned a value; also structures, that is relations of relations, presuppose entities distinguished in some other way than merely by their relations. How else if not by their qualities? What current physics abstracts from is therefore no detail, even in the broadest sense of “detail,” but a matter of principle.
The point about structure as the concern of science is not just that it is general, but that it is precisely or mathematically expressible (even though not without subtle qualifications which are commonly forgotten, but which Plato long ago [partly] saw). Also structure is intersubjectively observable in a sense or degree in which quality and value are not. Accordingly, to abstract from quality and value in science is reasonable; but not because we thus transcend details. Mere generality is not the goal of science, which is, rather, definite or publicly and accurately manageable generality. Value as such and quality as such, like experience as such, are as far from mere details as any scientist can get; but their intersubjective manageability is difficult or problematic.
“Concreteness” is that abstraction in which only details are set aside and even unmanageable generalities are given their due in whatever way possible. (Note especially formulae b,g, and i.) No study, unless metaphysics, is responsible for this abstraction. To deny metaphysics on principle is to guarantee that there will be no careful, disciplined theory of concreteness. The difficulties which cause science to set this notion aside form no absolute obstacle to such a theory, for in metaphysics we are seeking, not prediction or specific definiteness, but only generic understanding. We are looking, not for particular facts, but for the principle of factuality itself, the general status and value of facts as such, any facts.
How the secular approach to theology (k) leads in principle to all the metaphysical problems and suggests their solution I have tried to show in my The Logic of Perfection; also in Anselm’s Discovery.
In (d) it is assumed that, as Aristotle held, “unconditionally necessary and “eternal” are equivalent. I hold that the common doctrine of the “timelessness” of factual or contingent truths is metaphysical, and precisely in the bad sense. This is one of the most glaring illustrations of the charge that “positivists” do not escape metaphysical commitments. Reichenbach and Felix Kaufmann almost alone, among antimetaphysicians, have avoided this trap.
Concerning (e): Aristotle suggested, in somewhat uncertain outlines, a theory of objective modality, but failed to see or adhere to some of the implications of this theory. Nothing much was done about this until Peirce proposed his view of time as “a species of objective modality” with explicit reference to Aristotle. And indeed, it was Aristotle’s theory to which Peirce returned. That time is the “schema” (Kant) of our basic conceptions was already seen by Aristotle, although neither philosopher realized that time or becoming explains eternity or mere being, not vice versa. Eternity is a function of time, not the other way around. The current view that modal concepts are merely linguistic or logical, rather than ontological, is due to not seeing that the essential structure of time or becoming is inherent in concreteness (and in factual truth) as such, and hence is a priori, and that this structure is modal and cannot be grasped in merely extensional terms. That “only propositions can be necessary” is a mistake, for the creative process (one aspect of which is time) is the concrete from which alone any abstractions can be abstracted, and alternative propositions are contingent only because and insofar as this process as such is free to realize what either alternative asserts. What is objectively necessary absolutely is that the creative process will continue to produce creatures, and thereby itself continue to be creative. There can be no alternative to alternativeness itself. Deterministic philosophies fail to understand this, for they rule out alternativeness, at least as an innerworldly principle, and leave at most a mysterious contingency of the world as a whole, to which conceivable alternatives are perhaps admitted, but no principle by which any such alternative could have been realized. Determinism attempts to abstract from the creative aspect of experience, as materialism attempts to abstract from experience itself. Neither can offer an intelligible theory of concreteness. Here Socinus, Lequier, Bergson, Peirce, Varisco, Montague, Whitehead, Wenzl, and many others before them, including Plato and some Buddhists and a Hindu sect, are not far apart. But this neoclassical tradition is largely unknown to many philosophers today.
The reader will note that, as I use terms, “empirical metaphysics” is an absurdity. “Empirical cosmology,” perhaps—but then scientists are probably the best people to attempt this. And without an a priori element it is doubtful if much can be accomplished. Einstein has said that we might empirically show the world to be finite, but we could not thus show it to be infinite. This I suspect is correct. Infinity, whether of the actual past or of actual space, is not an empirical concept. It must then be a metaphysical one. (I also suspect that space is necessarily finite.)
Similarly with the nonquantitative infinity, or perfection—unsurpassability in some value sense—which defines deity. In spite of Leibniz no one knows what a “best possible” world would look like, nor if he (or she) did could he know that our world had this supposed character. Nor can inference from a perhaps partly botched product yield a flawless cause of that product. By no merely factual argument, therefore, can we prove or disprove divine perfection in any religiously significant sense. But in my view, nothing metaphysical, nothing in any strict sense necessary, infinite, eternal, universal, ultimate, absolute, or all surpassing could be known empirically.
Did Hume and Kant demonstrate the impossibility of significant theistic arguments? Yes, but only if they also demonstrated the impossibility of metaphysics, of a priori evidence about existence. Many take them to have done this. I think otherwise. And I have been studying these writers with care for forty-five (and then some) years.
It might be thought that empirical evidence could at least disprove the existence of perfection. On the contrary, Anselm showed that any being whose existence would or could be contingent, that is, with a possible alternative which facts might have shown to obtain, must for that very reason lack perfection in any strict sense, “even if it exists.” Thus to try to conceive a factual disproof of theism is already to change the subject. Only an idol, a fetish, as Peirce says, could be subject to factual disproof. Perfection, divinity, is a nonempirical topic, and must be handled as such. Or else mishandled, as so often occurs, both in theistic and nontheistic circles.
The illusion of an empirical test for the existence of perfection arises partly because we are used to looking to observed facts to tell us what exists, and partly because of subtle confusions embodied in the conventional ideas of omnipotence and omniscience, that is, of unsurpassable (or better, all-surpassing) power and wisdom. It has been supposed that a supremely free being could and must create creatures themselves quite without creative freedom— as though “creature” meant anything less than an individual instance of creativity other than supreme, and as though supreme creativity or decision making must or even could be a sheer monopoly of decision making. The fallacy of this notion can be seen a priori, and it would cause trouble no matter what facts of good and evil might be experienced. The error is logical; hence appeal to empirical facts is beside the point. A perfectly created and perfectly managed universe—in whatever sense this is conceivable-cannot mean a predestined or an absolutely controlled universe, for this is the same as no universe. Individuals (not alone human individuals) must in some degree be self-managed, agents acting to some extent on their own, or they are not individuals, concrete units of reality. This is inherent in the concept of concreteness. A definition of deity which violates this concept is illicit, regardless of facts. The literal meaning of “fact” involves making, creativity, as ground of the contingency of fact as such. I agree with Kant: empirical procedures to justify or falsify theism are one and all fraudulent—with all due respect to Mill, James, Brightman, Hick, and many others who have tried and are trying so valiantly to show the contrary.
But surely, you may be saying, concepts derive somehow from experience. Yes, but there are two classes of concepts, and two modes of derivation from experience. One set, the ordinary concepts, derive from special kinds of experience, and get their essential meaning from this specialization. The other set, the metaphysical concepts, derive from any experience in which sufficient reflection occurs, and they will be illustrated in any experience, even unreflective (if only for a reflective experience which is aware of the unreflective one). Such universal conceptions used to be called “innate ideas,” by which was meant something rather like what I have just described, though John Locke was perhaps not entirely to blame for misunderstanding this.
“No experience means no ideas”: who has ever denied that? (At most, classical theists perhaps denied it of God, who apparently, in their view, did not experience, but just knew.) The metaphysical ideas, however, are unique in applying to any experience, human, subhuman, superhuman; hence if some do not have them, it is not because they lack some particular sensory or perceptual experience to reflect upon, but because they do not reflect sufficiently, or in a sufficiently general way, upon those experiences which they have. As Leibniz cleverly put it, everything in the intellect comes from the senses— except the intellect itself! The intellect’s self-understanding is innate, a priori, or metaphysical. If Locke is supposed to have refuted innate ideas in this sense it is because many other accidental and irrelevant factors have been allowed to confuse the record.
Locke was particularly, and rightly, concerned about the supposed certainty of the innate ideas. But it was accidental and irrelevant if the rationalists identified “a priori” with “certain.” That any experience, to sufficient reflection, illustrates the metaphysical conceptions is no guarantee that a given individual at a given time has so reflected as to be “clear and distinct” or certain about his formulation of the conceptions. Formulation, verbalization, is an art, and a fallible one whose success is a matter of more or less; most animals and all infants have no such art. Who knows that we are not all partly infantile in this matter? Whitehead’s reference, in this connection, to our “ape-like consciousness” is perhaps a literary exaggeration, but otherwise it is seriously intended and appropriate. Mistakes are possible in mathematics and formal logic; is not metaphysics just that part of a priori knowledge in which clarity and certainty are least readily attained? Intellectual history suggests that this is so. And there is no reason, apparent to me, why lack of clarity or confusion must always be due to insufficient awareness of relevant facts. It may be due to insufficient, or biased, reflection. Accordingly, “a priori,” or “innate,” is one thing, and “certainly known” is another. If what a person most wants is certainty he or she might better turn to arithmetic, elementary logic, or even some parts of natural science, than to metaphysics.
It is arguable that there is but one metaphysical, innate, or strictly universal and necessary idea or principle, concreteness (containing internally its own contrast to abstractness), so that to speak of innate ideas in the plural is really to slur the distinction between our formulations of metaphysical truth and the truth itself, or the necessary nonrestrictive aspect of reality. The former, the ideas in the plural, are our contingent ways of trying to become conscious of the noncontingent ground of all contingency.
To show the equivalence of the various characterizations of metaphysics in greater detail would be to work out an entire system. Here I wish rather to indicate some relationships of metaphysics to other studies. In logic one speaks of entities on the first logical level, values of the variables for “individuals,” but one may or may not propose a general theory of what it is which assigns an entity to this level. One may say that one has in mind individual things existing in the spatiotemporal world order; or sense data; or sometimes, even individual numbers. But obviously numbers are abstract, compared to persons or physical things; moreover, events (in the most concrete sense, total unit-becomings) may also be taken as individuals, and then persons or physical things are more or less abstract by comparison, unless they are (wrongly, I hold) taken to be mere actual sequences of events. Metaphysics does what logicians would do if they gave serious thought to the a priori or strictly universal traits of their first-level entities. In addition logicians assume but do not adequately investigate the concepts of knowing or experiencing; this, in its purely generic sense, metaphysics does investigate.
That metaphysics deals with quality as well as structure distinguishes it from mathematics, which deals with structure only.1 Also mathematics, like logic, does not discuss the experience of structure, but only the structures experienced; thus it says nothing, except in an informal, semiofficial way, about the value aspect which no mathematical experience, let alone nonmathematical experience, can lack. Nor are mathematicians as such interested in subhuman experience, whose reality nevertheless comes somehow under the general concept of experience. They may perhaps discuss superhuman (divine) experience, but casually or at most semiofficially. Yet the pure concept of experience must either entail the logical impossibility or entail the logical possibility and in principle necessity of divine (in some sense maximal) experience, and until we know what the entailment is we do not have a wholly clear concept of experience as such.
That metaphysics seeks for necessary principles distinguishes it from physics and scientific cosmology. For, if these attempt to claim necessity for their principles, they are simply doing metaphysics. And no mere physical experiment can establish necessity. Whether and how it can be established is exactly the metaphysical question. It is not to be settled casually. But in recent decades that is how it has generally been disposed of.
To the eleven characterizations of metaphysics already given, I should like to add one more:
(1) The attempt to make nonexclusive or purely positive statements.
Consider the sentence “there are people,” meaning men and women. It seems purely positive. But in fact it has negative implications which are not merely verbal. Consider the forests which would still be standing had people not cut, burned, or bulldozed them down; the animal species which would still persist had men or women not exterminated them. True, people might have existed in less destructive ways, but this is a matter of degree; some destructiveness was inherent in the very existence of our species. Ordinary existence is perforce partly competitive, and only partly cooperative or symbiotic. It is this competitive “existence” which cannot be necessary, for it cuts off extralinguistic possibilities, and this is the very definition of contingency. No a priori science could tell us which extralinguistic possibilities are cut off, and hence none could tell us what, in the ordinary sense, “exists.”
But take such a statement as, There is something concrete. What extralinguistic possibility does it cut off? Surely not that there is something abstract. Indeed, unless extreme “Platonism” or extreme antinominalism is correct, there can be something abstract only as an aspect of something concrete. So the sole possibility which could be said to be annulled by the statement is, There is nothing, whether concrete or abstract. And what is this but mere verbiage? Certainly such a “state” of reality must, at best, be absolutely unknowable. I hold that what the statement “there is something concrete” excludes is purely linguistic, not an objective possibility at all. This is one mark of a metaphysical statement, that its denial is verbal only, not signifying anything beyond language.
Consider unqualified determinism, the doctrine that, in the actual world at least, there is an absolute order according to which the successor or outcome of any particular state of things is uniquely specified and alone causally possible, given that state. Is this statement partly exclusive, or is it wholly positive? It seems at first glance to be the pure affirmation of causal order, and so entirely positive. But this is deceptive. For in the absolutizing of the aspect of order, the aspect of creativity is equally absolutely denied. And creativity is positive. It does not mean merely that what happens is not fully specified by the causal conditions and laws; it means that there is more definiteness in reality after a causal situation has produced its effect than before. This increase or growth in richness of determinations is not an absence of something, it is a positive presence. True, there was first, if you like, an absence; but subsequently and forever after there will be the succession from absence to presence, and thus the richer conception has the last word. Growth is inherent in the very meaning of becoming. In Plato’s world there was (as W. P Montague said) no Newton or Einstein; to all modern science Plato’s having been was a factor. Otherwise history is language idling.
Consider the atheistic denial that there is any unsurpassable or divine being, or the seemingly positive assertion that any being could be surpassed by another. The statement seems either negative or positive according to how it is stated. But the negativity is essential, while the positive character is no more than verbal. For only a surpassable being can by its mere existence compete with, or exclude, anything in other beings; the unsurpassable is unsurpassably able to harmonize its own existence with that of others, and consequently denying its existence opens up no positive possibility whatsoever, any more than “there is nothing concrete” does so.
Leibniz was even more right than he knew in saying that philosophers err in what they deny, not in what they assert. But he himself often denied when he thought he asserted. His demand for a “sufficient reason” for particular or concrete things reduced creativity, even in God, to zero; his assertion that an individual always possesses all its adventures reduced becoming to a sheer illusion; his acceptance of classical theism, or the view that all possible value is actualized in God, entailed the implicit renunciation of his own correct doctrine of genuine competition among possibles, for it rendered this doctrine irrelevant or unintelligible from the divine standpoint, that is, the standpoint of truth. But Leibniz’s assertion that every individual (as distinct from aggregates of individuals) must have something like experience or perception was truly in accord with his doctrine of positivity. “Every individual (not identical with an aggregate of individuals) experiences” denies nothing but what is itself a mere denial, viz., “some singular individuals are absolutely insentient and mindless.” Of course aggregates of individuals need not themselves be sentient individuals, and the possibility of forming such insentient aggregates is an a priori aspect of the idea of sentient individuality. Leibniz saw, as no man had before him, and not many have after him, that an entity can show itself to be singular rather than a mere aggregate only by acting as a singular, but this is also the only way it can show that it is sentient, though perhaps not on the level of conscious thought. If we could ask an atom, “do you feel?” having no language, it could only answer—just as it does, in effect, answer—by responding to stimuli (changes in its environment), by withdrawing in some situations, advancing in others, in some cases reorganizing itself internally, dropping constituents or assimilating them, acting in an “excited” or in a “satisfied” fashion. Any evidence there logically could be for very low-level sentience there actually seems to be. And in what possible world could it be otherwise? How in any world could one know either that a seemingly inert mass like a rock was inert also in its imperceptible parts or individual constituents, or that it had no such constituents? I hold that in any world, one either would not know what the individuals were, or would know them as genuinely active—and so, by the behavioral criterion sentient—agents whose activities collectively produced the sensible effect of an inert mass. Agency, singular concrete reality, are one; until one encounters the former, one does not know the latter. “Insentient,” in behavioral terms, taken absolutely, has the same force as “individuals unidentified.”
What we need to do is to follow Leibniz’s prescription not only where he followed it, but also where he did not. Metaphysical truth is purely positive truth; all else is partly negative, or, if absurd and a mere confusion, wholly so.
When another great rationalist, Spinoza, said that all things follow from the essence of God which also entails its own existence, he was really saying that “metaphysical” has no distinctive meaning. Leibniz is by implication in the same difficulty. But these men fell into this trap not because they were rationalists, if that means believers in the possibility of a priori or nonexclusive and purely positive knowledge. No logical bridge leads from “there is a priori or necessary truth about reality” to “all truth about reality is a priori or necessary.” Quite the contrary, to regard all truth or knowledge as metaphysical is to fail to take the idea of the metaphysical seriously for it deprives this idea of any distinctive meaning. Exaggeration is often in effect denial.
If Hegel was wiser than the previous rationalists in the matter presently being discussed he hid his wisdom under elaborate dazzling systematic ambiguity such as only high genius could have conceived, but which can hardly serve the needs of our day or the future. In spite of George Lucas this is my belief. In some other subjects Hegel is more relevant.
The way to discredit nonexclusive or necessary truth is the very way its sponsors adopted when they unduly extended its scope, so that contradictorily it became exclusive or contingent after all. Nothing is more necessary to the future of “rationalism” in the good sense than the avoidance of this extravagance, which deflates rationalism’s claims by overinflating them until they burst. Metaphysical or nonrestrictive truth is very little, rather than very much; but that little is precious nonetheless, provided it be seen in its purity and not adulterated by restrictive elements.
One might say that the seventeenth-century continentals (some of them) were not so much rationalists or metaphysicians as intellectual imperialists. Like other ambitious men they wanted to take over any not too strongly contested territory. And as the excuse for political or military imperialists has been that if one’s own country had not been the aggressor against some defenseless area another country would have been, so the excuse for the older rationalists was that empirical science had not yet learned to claim its own. Today the danger is in the opposite direction: now metaphysical territory is too weakly held, and empirical scientists or their admirers want to take over even that extremely colorless or most abstract knowledge, the knowledge of nonrestrictive truth. This is merely the old mistake, but now made by going too far in the opposite direction. Empire is always temporary and insecure.
Just as it did not really exalt metaphysics to offer to let it swallow up science, so it does not exalt the empirical method to offer to let it adjudicate for possible worlds, necessities, or the all-surpassing. Confusion weakens everything, strengthens nothing.
Another classical example of blurring the distinction between the genuinely a priori and the empirical is Kant’s attempt to make Euclidean geometry (in an unqualified fashion) a priori, on the ground that human cognitive powers require it. But they require at most the approximate validity of this geometry, and in this approximate form the requirement is shown by respectable empirical evidence to be met by the contingent structure of the universe. The idea that an absolutely precise requirement (with conceivable alternatives) could be justified by so vague a ground as our self-awareness of the forms of our perceptual intuitions is logically incongruous. Absolute precision requires a more definite ground than that.
I recall these classical cases to remind the reader that the collapse of certain false claims to a priori necessity is no proof that all such claims must be false. Exaggerated reactions to exaggerated pretensions will not get rid of exaggeration itself.
A word about how “God” is intended in definition k. I am deeply convinced that classical metaphysics mistranslated the essential religious issue by misdefining “God.” It thereby fell into a subtle form of idolatry, worshipping not the divine fullness but an abstraction called “the absolute” or the “infinite,” “unmoved mover,” or “most real being.” None of these, I am persuaded, is genuinely worshipful, though they can be subjects of intellectual amazement, wonder, or awe. I agree with one main conviction of current philosophizing, that people have been so eager to answer questions that they have failed to give proper heed to the way words are being used, and perhaps misused, in formulating these questions. “Is there a greatest or most real being?” Well, first of all does the notion of a “greatest being” make sense, any more than that of “largest possible number”? I myself think the answer should be, No, neither greatness nor “reality” is without qualification logically capable of a maximum. Is there a strictly immutable yet living being? I think the answer is, No, life in any sense, no matter how exalted, implies some form of real change or becoming—indeed some form of real growth. I think the history of natural theology from Aristotle, Philo, and Augustine to Hume and Kant has brought out rather clearly that if God must be defined in terms of these traditional paradoxes, then theism has no rational content. But I am equally persuaded that this way of defining God is mistaken anyway, both on purely religious, and also on philosophical, grounds. This mistake is one of the penalties we have had to pay for putting too much trust in the first form of rational metaphysical thought, the Greek. Great as its achievements were, it was at best a one-sided first approximation, and at worst a melange of calamitous errors.
There are several reasons why a metaphysician cannot sensibly proceed very far without considering how he is to handle the theistic question. First, the logic of the idea of God itself shows that the whole content of metaphysics must be contained in the theory of deity. Comte was not mistaken in this: metaphysics is no more and no less than natural theology (supposing the latter is possible at all). For (a) the existence of God cannot be construed as contingent, an accidental exemplification of metaphysical principles, but must be admitted, if at all, as necessary, as inherent in these principles having any truth at all; moreover (b), the metaphysical or necessary features of the reality with which God deals (whether by creating it, knowing it, or what not) can only be the same as the intrinsic or implied correlates of whatever is necessary in God’s being. That is, what God must have or must know is necessary, and that alone; anything God merely can have or know is contingent, and so a subject for science or faith, not for rational metaphysics. The metaphysical theory of deity is either nonsense or it is simply metaphysical theory—period. I reject the old distinction between general and special metaphysics; theology, not metaphysics, is more than mere theism. The only concession to make here is that the theistic meaning of metaphysical concepts may be left implicit, unformulated. But to that extent the concepts will be left vague or ambiguous, and the system too indeterminate to give much purchase to rational criticism.
Second, the history of metaphysics is so entangled in various ways of trying to adjudicate contingent, historical religious questions that it is difficult to abstract a religiously neutral residue from this history. This is what one should expect from the logic of the case and it is what we find. (Difficult is not the same as wholly impossible.)
Third, I see in the application of metaphysical categories to religion their chief use and one test of their adequacy. If a metaphysician can make sense out of this topic so much the better, but if not why do we need his system? Science and common sense (almost seem to) take care of factual (contingent) questions; art, or personal or group faith, can handle the rest, if there is any residuum. But only the metaphysician can clarify what it means intellectually to ask about God. I hold that much ancient Greek metaphysics, and I include even Kant as a kind of belated and sub-Platonic Greek, has failed this test. It has muddled almost as much as it has clarified the topic it chiefly sought to illuminate. Not that we owe it no thanks for this failure. If we see better on the same topic it is partly because we have these majestic examples of how not to define God. And such learning by mistakes seems the destiny of all nondivine consciousness. Moreover, the (allegedly) Greek way of defining God is not so much wrong as one-sided. Terms like absolute, infinite, independent, uncaused, do apply to God—but only on condition that the correlative terms, relative, finite, dependent, caused, also apply. Contradiction of course results if—but only if—no qualification such as “in some respect” is attached to the attributions. The classical error was one of overtrust in extremely simple ways of characterizing God. It was a kind of learned (extremely stubborn) simple-mindedness.
If the religious issue is as central in metaphysics as it seems to be, to attempt first to settle everything else (as though there were in metaphysics much else to settle) and only then to ask about “God” is to be in danger of begging all chief metaphysical questions. Hume and Kant did just that, in my opinion. Unwittingly assuming antitheistic postulates, they not surprisingly inferred the impossibility of a rational theism. But the reasoning can be reversed: since theism at the least deserves a hearing on its merits, the experiment should be made of provisionally rejecting every postulate which shows itself hostile to theism. And this includes (Hume’s) axiom, “the distinguishable is separable” or independent, for though God (and this is inherent in the religious idea) exists no matter what other individuals may or may not exist, and thus is indeed separable from them, God is also thought of as the power upon whom all else depends, and thus the “creatures,” though distinguishable from God, are certainly not separable from, independent of, (God’s) existence. Several other assumptions of Hume, and some of Kant or Bradley, must likewise be “put in brackets” while we consider what metaphysical system if any would harmonize with religious requirements. No other procedure is rational inquiry in this sphere, but rather is dogmatic rejection posing as judicial examination.
Such is my apology for intruding the theistic aspect into the definition of metaphysics. “Neoclassical metaphysics,” when its ideas are adequately explicated, is neoclassical natural theology, and vice versa. In several books I have tried to show at least in outline how from the mere idea of God a whole metaphysical system follows; one may also proceed in the opposite direction, and show how from general secular considerations one may arrive at the idea of God and a judgment as to its validity. But the two ways of proceeding differ only relatively and as a matter of emphasis. In thinking metaphysically at all one is already more or less close to an explicit thinking about God, and in thinking at all clearly about God one is already somewhat conscious of metaphysical principles. A priori knowledge, valid “for all possible worlds,” must coincide in content with the most abstract aspects of omniscience. It is what is common to all that God could create or know, and of course God knows these capabilities. H. Scholz (theologian become logician) has pointed out how the symbolic logician is in a certain (very abstract) sense seeing things as God sees them. A forteriori the metaphysician is doing this—if he succeeds (in talking coherently about universal necessities).
Note
1. See P. Bernays, German mathematician.
Source:
Charles Hartshorne, The Zero Fallacy and Other Essays in Neoclassical Philosophy, pp. 95-108.