Journal of Philosophy 21 (1924): 337-347.


C. J. Ducasse
University of Washington

I. The thesis which the present paper is concerned to defend is that no ontological position can be either proved or refuted; and therefore that any ontological position which is meaningful is tenable, and that its adoption or rejection is thus in the end purely a matter of one's personal taste at any given time.

II. First, then, what, as a matter of logical form, is an ontological position? The answer is that it is a statement of the form "To be is to be X," or "To be real is to be X." Thus Berkeley's position that "To be is to be perceived or percipient," Hobbes' position that "To be real is to be body," Professor Taylor's position that "To be real is to be free from contradiction, " etc. In such statements, however, the terms "To be" or "Being" on the one hand, and, on the other hand, "To be real" or "Reality," have been used more or less indiscriminately. But the logical situations resulting from the use of the two terms are very different indeed, and each must, therefore, be examined separately.

III. Let us first examine the implications of assertions of the form "To be is to be X" -- of assertions, that is, where the subject is merely "To be," or "Being," without any qualifieations such as the addition of the word "real" would constitute. Then, within "Being," as thus maximally taken, must find room not only physical things, but also fictions, illusions, mere meanings, null-classes, appearances, etc. -- everything, in short, that can in any way be mentioned, except "Nothing-at-all." This maximal interpretation of "To be" once agreed upon, it at once follows that any definition that has "To be" for its subject must also have "To be" and no more, for its predicate; for otherwise what the definition would do would be to make Being-in-general mean one only of its own particular kinds. It would pack the whole building into one of the rooms of it. And this consequence is obviously not to be avoided by saying that if Being is infinite, then, as in other infinite collections, some parts of it are equal to the whole. For what is true of an infinite collection is only that a one-one correspondence can be established between the entities of the whole collection and the entities of some part of it; and this is a very different thing, indeed, from saying that the two collections are "equal" in the sense of being reciprocally composed of identically the same entities -- that between the "map" and the "object mapped" there is no difference at all -- which is, of course, not true. Philosophical mystics always have perceived clearly that since all determination is negation, when the subject is undetermined Being as such, the application to it of any determining predicate constitutes a flagrant contradiction; and that, on the other hand, the application to it of a strictly non-determining predicate yields a mere tautology.

To take the position that "To be is to be X" can, therefore, consist only in arbitrarily defining "nothing-of-kind-X" and "nothing-at-all" as being strictly synonymous terms; and this entails that any word used (whether affirmatively, negatively, or interrogatively) by the maker of such a definition must either connote "being X" or "being a kind of X," or else be wholly meaningless. Thus, if one has, for instance, taken the position that "to be at all is to be perceived or percipient," then, if one uses the word "Matter" in any way, either that word connotes "being perceived or percipient" or else it means nothing at all. If it means nothing at all, Materialism can not be refuted because it can not then even be stated. If on the other hand, "being perceived or percipient" is made part of the meaning of the word "Matter," then, with such a meaning for the word "Matter," Materialism and Idealism turn out to be precisely identical positions, or Materialism at most a special kind of Idealism! And if, on the contrary, the Idealist allows the word "Matter" to mean something else than "being perceived or percipient," then, in so doing, he commits ontological suicide, abandoning as he does his idealistic position, and embracing the new position that "To be is to be percipient, or perceived, or something else." But a precisely similar predicament arises from the position that "To be is to be Matter," or, indeed, from any other position of the form "To be is to be X," where "To be" is maximally interpreted. Therefore, no such position can refute or be refuted by any other such position -- which is the very thesis of the Liberalism that I am defending. Any such position, in practice, can consist only in wholly ignoring whatever is not-X. The holder of it may be compared to a metaphysical ostrich, who has buried his head in his favorite patch of ontological sand.

IV. But now, I may be told, what the Berkeleyan idealist, for instance, actually means is that the word "Matter" does have a meaning, which is, indeed, in the broad sense perceived, but that there are no cases at all of that meaning. In this answer, a distinction is obviously implied between two kinds of Being, namely, being an individual case of a kind, which is being real, and being a kind without cases, which is being unreal. This, then, brings me to ontological positions of the form "To be real is to be X," for which the form "To be is to be X" very possibly is but an ellipsis. Let us, therefore, now examine the implications of positions of the form "To be real is to be X."

The first implication is that if anything is not-X, it is not real; and to say this does not now mean that it is nothing at all, but only that it belongs to a realm of being other than that which was declared to be Real Being, thus, possibly, to the realm of existing illusions, appearances, or unrealities. In other words, the realm of Being-in-general is now thought of as divided into two sub-classes, one of them entitled to the name "Reality," and the other to the name "Unreality" or "Appearance." And the question now is, which one out of the various realms distinguishable within Being-in-general, is the one entitled to the name "Real"? Let us suppose this question answered, and the proposed answer to be that "Real being is being that is X." The question next arises, how can we find out whether or not that answer is correct? This obviously depends on what were supposed to be the data, and what the quaesitum, of the problem to which the assertion "To be real is to be X" purports to be a true answer. That problem might have been any one of four different ones:

1. It might have been: What entities are upon empirical examination all found to possess the character "Reality"? If this is the problem, the definition of the term "Reality" must be assumed agreed upon to begin with; and the answer to the problem is then to be ascertained by examining entities of various sorts within the realm of Being in general, and observing in each of them whether or not the character meant by the word "Reality" is present. The conclusion that "To be real is to be X" then means that it has been statistically ascertained that the character meant by the word "Reality" and the character "X" are always present together. But I do not think any philosopher has ever claimed his ontological position to have been arrived at in this way. Moreover, the clear understanding of the meaning of the word " Reality," which is necessary to attack the problem in this way, is not possessed. Far from being one of the data of the problem, it is on the contrary much rather itself the quaesitum.

2. But, secondly, the problem might have been the converse of the above; namely: What character is empirically found present in all and in none but the entities to which the name "Realities" is in fact applied? If this is the problem, then the concrete denotation, of the word "Realities," instead of as in the previous case the connotation of it, must be assumed clearly agreed upon to begin with: It must be assumed as agreed that this entity, and that entity, and that entity, etc., are all in fact called Realities, and that the problem is to discriminate and state abstractly the character in the possession of which all these entities agree, and which differentiates them from all the entities that are acknowledged not to form part of the denotation of the word "Realities." But it is, I presume, again obvious that no ontological position claims to be so grounded. And again, the denotation of the word "Reality" is no more definitely agreed upon than the connotation. The same individual entity, as a matter of fact, is classed as one of the realities, and also as not one of them, by different persons, and even by the same person on different occasions. Examination of a few concrete cases of the use of the term "Real" makes this evident. We say, for instance, that Spain is a real country, but Utopia is not real; that this table appears solid, but is really a mass of whirling electrons; that this table-top appears trapezoidal but is really rectangular, although for the painter it is really trapezoidal; that we have a real toothache, but only imagine a pain in our arm; that God alone is real, and that Spain, table-tops, material objects, toothaches, etc., are but empty conceits. We speak of what we really mean, really feel, really think, really imagine, really want, really are doing; of a man's real motive; of something being really worth the trouble; of love, or of beauty, as being the only reality of life; of the realization of our dreams, and of the fact that, after all, dreams are the only realities, etc.

3. But now, by this very array of cases of the use of the term "real," we find ourselves led to the idea that our problem is, after all, not concerned with the observation of nature (whether physical, psychical or other), but on the contrary with the observation of language as admittedly well used. Thus, thirdly, the problem might have been construed as having for its datum a purely verbal definition of Reality, announcing merely what users of admittedly good English have a priori proposed to signify when conferring on anything the status "real." The question would then be that of deducing a priori from that verbal definition, by the addition of suitable determinants, the various logically possible species of reality -- as, for instance, we deduce from the definition of the term "Parallelogram" the possible kinds of Parallelograms: squares, rhombs, etc. But a verbal definition of Reality, acknowledged to represent the meaning intended for the term by admittedly good English, is, unfortunately, not available; and our problem must therefore be the converse of the above, for that converse constitutes the only remaining possibility, which is as follows:

4. Given a number of expressions such as those listed above, that admittedly constitute good and meaningful English, and in which the term "real" is used, find out by induction from these expressions what definition of the word "real" could, in all of them alike, be substituted for the word itself without changing the meaning of any of the expressions. This, I say, is the problem which propositions of the form "To be real is to be X" attempt to answer, so far as they claim every one's assent. And now, what is the correct answer to that problem?

Many conflicting answers have been proposed, but there is one that has already met the approval of a number of philosophers of otherwise divergent views, and which is thus remarkable among ontological positions in that it appears to have some claim to being not a rival of, and incompatible with, others, but on the contrary to stand to these others in the inclusive relation of genus to species. Moreover, it appears to be compatible with all such cases of the use of the word "real" as were mentioned above. That answer declares Reality to be that of which we must take account for our purposes. William James, Royce, and Taylor, at least, agree as to this. But, now, the immediate consequence of accepting this definition of Reality as a correct induction of the verbal meaning of the term as actually used, appears to be, again, such an Ontological Liberalism as I am defending. Thus Taylor says: "So far then, it might seem that 'reality' is a purely relative term, and that our previous choice of ultimate freedom from contradiction as our standard of reality was an arbitrary one, due to the mere accident that our special purpose in sitting down to study Metaphysics is to think consistently. Of course, it might be said, whatever game you choose to play at, the rules of that particular game must be your supreme reality so long as you are engaged in it. But it depends on your own choice what game you will play and how long you will keep at it. There is no game at which we all, irrespective of personal choice, have to play, and there is, therefore, no such thing as an ultimate reality which we must all recognize as such; there are only the special realities which correspond to our special individual purposes. You have no right to set up the particular rules of the game of scientific thought as a reality unconditionally demanding recognition from those who do not choose to play that particular game."[Elements of Metaphysics, p. 52.]

But Professor Taylor states this conclusion with such admirable clearness only for the purpose of disputing it. His answer is as follows: For one thing, what the law of contradiction declares is that if I think at once that A is B and is not B, my thinking is then not true; but, to think truly about things is to think of them as they really are; hence, to say that non-contradiction is a condition of true thinking is also to say that it is a condition of real existence. And as to the rest, he says: " The very recognition of the fact that any one individual purpose or interest can only get expression by accommodating itself to a definite set of conditions, which constitute the reality corresponding to that purpose, carries with it the implication that the world is ultimately a system and not a chaos, or, in other words, that there is ultimately a certain constitution of things which, under one aspect or another, is of moment for all individuals, and must be taken into account by every kind of purpose that is to get fulfillment." Such are Professor Taylor's objections, and if Ontological Liberalism is to be vindicated, a complete answer must be made to them. I now propose one, in three parts, as follows:

1. First, I admit at once Professor Taylor's contention that the law of contradiction is not merely a law of logic, but also a law of Being. At least, I admit it in the sense that nothing, absolutely nothing at all, not even an appearance, can both be and not be, or both be what it is and not be it. But obviously this furnishes us with no criterion of distinction whatever between an appearance and a reality in experience, when experience is defined as Professor Taylor defines it, namely, as "immediate feeling or apprehension." For immediate feeling or apprehension, which as such merely is but does not mean, or assert, anything at all, never contains any contradictions. No immediate feeling or apprehension ever both is and is not, or ever both possesses and lacks a given character, or ever as such is inconsistent with any other immediate experience that we have had. The examples that Professor Taylor gives of "contradictions in our experience," having necessitated the making of a distinction between appearance and reality, are not cases of contradictions in our experience as defined by him, at all. Thus, the contrasts of "the seeming stability of the earth with its real motion," of "the seeming continuity and sameness of a lump of solid matter with the real discontinuity and variety of its chemical constituents," of "the seeming friendliness of the hypocritical self-seeker with his real indifference to our welfare" (p. 2), are not in the least cases of contradictions in immediate feeling or apprehension. The contradictions truly involved are in the constructions or interpretations that we place upon our immediate apprehensions, -- in the assertions that we make about them. But those immediate apprehensions themselves must need have been free from contradictions, else they never would have existed at all; the law of contradiction in its ontological application assures us of just that. And that law, therefore, furnishes no criterion whatever for a distinction, within experience, between an appearance and a reality. What leads to that distinction are contradictions in our constructions about experience, and these constructions, although, indeed, often called experience, are not experience at all in the sense of the term expressly specified by Professor Taylor. Therefore, freedom from contradiction is a criterion of distinction only among constructions about experience. Such of them as involve contradictions, we may choose to say, shall be called unreal. But such a definition of unreality obviously is wholly relative to the purpose of thinking rationally, and, for some one not at a given time having that purpose, unreality simply does not mean that.

2. The second point of my reply to Professor Taylor has reference to his assertion that "The very recognition of the fact that any one individual purpose or interest can only get expression by accommodating itself to a definite set of conditions . . . carries with it the implication that the world is ultimately a system and not a chaos." I answer that what Professor Taylor calls "the recognition of a fact" is only entitled to be called "the formulation of a hope." For when we say that Reality is that of which we must take account in order to fulfill our purposes, we are formulating, as has been shown, not the description of an observation, but the meaning of a word; and no verbal definition implies the existence of what it defines. Thus, what the definition of Reality in terms of purpose explicitly means is that no act of ours can bring about the fulfillment of its purpose unless the effect that we purpose will result from the same sorts of causes as in the past, and we know those causes, and can set them in action. Indeed, we hope that all those conditions are met, but we do not "recognize them as facts" in the sense of knowing them to be met. Thus, we do not know that the world is a system and not a chaos; we know at most that it has in the past exhibited some order; and -- so far as we are purposive, teleological, planning beings -- we hope it will continue to do so. For otherwise we shall either be unhappy and disappointed teleological beings, or we shall have to give up the life of conscious purpose, and get such values as we may from the life that has chaos for its reality, namely, the life of excitement and adventure. And this brings me to the third part of my reply.

3. It is this: Reality is not broadly enough defined as that of which we must take account for our purposes. We are something else besides purposive, planning, rational beings. The unforeseen, the new, the unpredictable, the irrational, the bewildering -- in short, the chaotic -- also has its values for us. These values are those of venture, of excitement, of strong feeling, of strange emotion, of intoxication; and all these are valued not as means, but immediately and intrinsically. And so far and so long as we do value these, it is chaos and not system which is for us Reality. "Real" must therefore be defined as the adjective which is applied by any person to anything when he wishes to signify that that thing has value for him then, positive or negative, intrinsic or instrumental. And when any one applies to anything the adjective "unreal," he means to assert that so far as a given experience of value by him at a given time goes, the thing in question makes no difference at all either mediately or immediately, and is thus virtually non-existent. Being Real is thus a purely relational character, a status and not a quality, which is conferred upon and taken away from things by each individual, according as his valuations do or do not involve those things, whether directly or indirectly. And this definition, I repeat, is put forward strictly as an induction from concrete occurrences of the adjective "real" in admittedly good English, and it is by them that its correctness is to be tested.

Professor Taylor, it is true, attempts to bar the way to any such Liberalism as follows from that definition, by saying that "feeling is essentially teleological"; not in the sense "that it necessarily presupposes conscious anticipation of its guiding end or purpose," which he characterizes as a "monstrous assumption" (although it is that in terms of which his entire discussion is at least worded), but in the sense that the feeling, e.g., to use his illustration, the pain that the success of another man in winning the love of a girl gives us, would not arise unless there had existed, possibly unawares, "a specific psychophysical tendency of an essentially forward reaching or teleological kind."

But aside from the fact that Professor Taylor here finds it necessary to rest his whole metaphysical case on the highly precarious psychological theory that "pleasure is essentially connected with unimpeded, pain with impeded, discharge of nervous activity," he is, what is worse, attempting the logically impossible feat of subsuming under the concept of purposiveness its very own opposites -- namely, blind causation, planless spontaneity. He can succeed only at the cost of taking out of the term "purposiveness" itself the very meaning owing to which purposiveness postulates system instead of chaos. For when he says, in the instance above, that the pain would not have arisen unless there had existed, possibly unawares, a tendency of a teleological kind, "teleological" can not then here be opposed at all to blind, or to mechanical, or to automatic -- and, therefore, means no more than conditioning, in the most inclusive and general sense of the term. That is, Professor Taylor's assertion then means only that pain would not have arisen if the conditions of it had not been present. But with purpose thus made to mean more than "feeling that has a cause," the assertion that reality that of which we must take account in order to fulfill our purposes, turns out to declare merely that "reality is anything that a feeling depends on." And this definition does not in the least imply that the universe is a system and not a chaos, for individual states of the universe might condition or cause individual feelings, without any order or regularity in this causation at all. And even if there were system in the relation of states of the universe to feelings, there might yet be no other order whatever among the states themselves of that universe. Thus it is the belief in the possibility of active fulfillment of purpose through the conscious choice of means, which postulates, i.e., hopes, that the objects which are available as means are systematically connected with the objects desired as ends. But the mere experiencing of pleasure or displeasure, or of any feeling, demands no such postulate at all.

V. Having thus, as it seems to me, fully met Professor Taylor's objections to the Liberalism that I defend, there remains yet to deal with a meaning of the term "to exist" narrower than that of "being something, anything at all," i.e., of "being other than nothing at all." It is the meaning in which we say that men exist, but centaurs do not. Both Mr. Russell (Introd. to Math. Phil. Ch. XVI) and Mr. Johnson (Logic I, V) make attempts to give an account of what "to exist," as so used, means. But in my opinion these attempts are unsuccessful, because they both presuppose as already understood the meaning of the term defined. Mr. Johnson's view that the extistent means "what is manifested in time or space" presupposes that we already understand the difference between the space-time occupied by men, and that occupied by centaurs, the first of these space-times being said to "exist" and the second not. And Mr. Russell finally appeals to an intuitive "sense of reality," and merely applies epithets to those in whom it functions otherwise than in himself. Moreover, both Mr. Russell's and Mr. Johnson's discussions appear to me tainted by the grave error of supposing that to speak of a "non-existent" object, and to speak of the description of it, are one and the same thing. But when I speak of the centaur, Chiron, I do not mean the same thing at all as when I speak of the description of Chiron to be found in books; no more so than when I speak of the man, Socrates, do I mean the same thing as when I speak of the description of Socrates to be found in Xenophon's writings. Socrates and Chiron are equally objects, but the one is usually regarded as real, and the other not.

In what sense is it then true that men do "really" exist, and that centaurs do not? My answer is that it is in terms of certain human interests or valuations that the distinction between such "real" and "unreal'' existence is to be defined: The world in which men exist is the real world for us so far and so long as we are teleological, planning beings, in search of means to ends. But the moment we become esthetic or emotional or feeling beings, tasting and living in the values of the present instant, then that is real which is an object of present feeling, and the centaur may then well be supremely real, and the grocer and his goods supremely unreal. It is true that when the poet or the dreamer is questioned, he admits that the grocer is real. But to question him is to cheat; for to question him is to make him take not the poet's but the thinker's position. It is to make him live for the moment not in terms of intrinsic values, but in terms of the instrumental value of truth. To question him is to beg the question, unless it be clearly borne in mind that the question is not whether he, now a thinker, regards the grocer as real, but whether he, while a moment ago a poet, did then regard the grocer as real. The confusion between these two questions may well be called "the thinker's fallacy," and it is owing to the committing of that fallacy that grocers are declared to be more absolutely real than centaurs. What is true is that grocers, and in general the objects of ordinary perception, are more often real than centaurs; and that just now for us they are real and centaurs are not, because just now we are thinkers and not poets.

VI. And now, by way of making as explicit as possible the meaning of Ontological Liberalism, I repeat that the adjective "real when applied by any one at any time to anything means simply th that thing is then valued positively or negatively by him either for itself or as a means; in short, that it is at the time not indifferent to him. And that a given entity is or is not indifferent to a give person at a given time is an ultimate and bare fact of that person's taste or nature, which can not be argued away and is beyond logic. Indeed, all argument begins from it. From this view of the meaning of Reality, all others are deducible by specification of the taste, the interest, the valuation of which each is a function. Therefore, no such special view is false and none is true, but each is tenable and each also avoidable. Moreover, although this conclusion about the meaning of Reality is, I believe, itself true in the sense of being the only one compatible with all the ways in which the word "Reality" is actually used in impartial good English, anybody is, course, free to declare that he individually has no need of, and does not admit, some of the special senses of the word "Reality" that English language does recognize, and thus that the term as used by him connotes more than is included in my definition. If any one so declares, I do not know of any way to force him to connote by Reality less than he chooses, nor of any but biased reasons for attempting it. He has a right, in a free country, to speak a language of narrower extension than that which most of us use. The Liberalistic position is incompatible with his only in the sense that one can not both choose to roam through the various species of a genus, and also to confine oneself to one of them exclusively. Which of the two does is a matter of one's taste. Any non-liberalistic position is logically tenable, provided the holder of it confines himself to a language narrower than the English of most of us. It is true that from the point of view of the rest of us, he is, as it were, linguistically color-blind. But from his point of view, the rest of us are linguistically "seeing things"! And if he and we are not satisfied with thus calling each other names, the question as to who has the true definition of "real" can then be decided only on the basis of whether he can lock us up, or we him.