BD 161 .G2 Copy 1 LIBRARY OF CONGRESS 021 061 384 5 PROCEEDINGS OF THE SEVENTH ANNUAL MEETING OF THE AMERICAN PHILOSOPHICAL ASSOCIATION HELD AT CORNELL UNIVERSITY DECEMBER 27-28, ioopr TOGETHER WITH THE ADDRESS OF THE PRESIDENT THE PROBLEM OF TRUTH BY PROFESSOR H. N. GARDINER [Reprinted from the Philosophical Review, Vol. XVII, No. 2, March, 1908.] {Reprinted from the Philosophical Review, Vol. XVII., No. 2, March, 1908.] THE PROBLEM OF TRUTH. 1 IN speaking on the same subject as that selected for the Discus- sion that is to take place to-morrow morning, I do not seek to forestall the results of that discussion. Nor shall I attempt to deal with what to many may seem the more profound and signif- icant aspects of the problem, such as, for example, the relation of our finite knowing to absolute knowing, or the place which our particular truths must have in a final and complete meta- physical system. My aim is rather to set forth simply and clearly some of the more general considerations that ought, in my judgment, to be kept in mind when this subject is under debate. Now the first requisite in this discussion is surely a definite understanding as to what truth the discussion is about. * True ' and ' false' are adjectives like ' red ' and ' sweet' or ' good ' and ' bad,' and, like them, must be taken to qualify some object or ob- jects. But the objects they actually are taken to qualify are various, and hence an ambiguity in the conception of truth. We not only apply the terms to ideas, supposals, judgments, propositions, beliefs, and the like, but we also meet with true and false friends, true courage and beauty, false modesty and honor, and, alas, sometimes false dice, hair, and teeth. In this sense falsity may be itself a character of truth : " his faith unfaithful kept him falsely true." In the Hegelian philosophy we have another use of the term, according to which the higher category is truer than the lower, teleology is the truth of mechanism, spirit the truth of nature. We shall avoid at least one source of confusion if we 1 Delivered as the Presidential Address before the American Philosophical Asso- ciation at Cornell University, December 27, 1907. lI 3 114 THE PHILOSOPHICAL REVIEW. [Vol. XVII. agree, to begin with, that our concern is with the truth of propo- sitions. We assume that propositions are either true or false, or neither true nor false, or, in case a number of propositions are involved, are at once partly true and partly false, and that, in any case, regarding any intelligible proposition the question can be asked whether it is true or false, and in what way. If we agree to this, then certain not inconsiderable conse- quences would seem to follow. One, and most important, is that we recognize the truth we are talking about as a quality found in quite particular truths. For every proposition, whatever its range or comprehension, expresses and embodies a single, even if com- plex, truth, and the number of possible truths is as infinite as the number of possible propositions. This is not to say that truths are disconnected, or are, or relate to, ' independent entities,' or are merely externally connected in a series. Propositions hang together ; one truth implies, follows from, leads to another. Hence the possibility is not excluded that many truths may cohere together to form a system, and that all truths may ulti- mately appear as elements in one comprehensive system or realm of truth. But this last should not be dogmatically assumed at the outset in such a way as to prejudice investigation into the nature and conditions of particular truths. Not even the most resolute defender of an absolute system would maintain that such a system was even remotely attainable by man. 1 Not only have the propositions in common use little or no evident connection, but within the most organized forms of our knowledge, — the sciences, — principles of wide import in one department are totally ignored in others. Moreover, a system of truth is really, from the propositional point of view, a system of truths, and can- not, as such, be expressed or exhibited in any single proposition. Philosophers, as we know too well, often require for the expres- sion of their systems one or several pretty ponderous volumes. A true system would be one, all of whose propositions were true and also connected. Propositions about the system, however, are just as particular as propositions about its parts or about the 1 " It would be impossible that any man should have a world, the various provinces of which were quite rationally connected, or appeared always in a system." Bradley, Appearance and Reality, p. 367. No. 2.] THE PROBLEM OF TRUTH. 115 connections of its parts. At the outset, then, we ought, I think, in this discussion to recognize to the full the particularity of all propositional truths, and that whether they have to do with the more special or the more general aspects of their subject-matter. We ought, as far as possible, to avoid talking of truth ' at large ' ; and we ought equally to be on our guard against any bias in favor of a peculiar type of truth, as, for example, scientific as opposed to philosophical truth, or vice versa, or of either as against the episodical truths of every-day life. For if every propositional truth is particular, there is no prima facie reason for regarding one as more or less true than another, so far, that is, as it is true at all. Truths differ in value and significance ; some are trivial, some perhaps sublime. But, apart from special theory, there is no apparent reason why a proposition about even so trivial a circumstance as the present state of the weather, — which indeed may be important enough on occasion, — should not be as true as the truest propositions about such exalted objects as the existence of God, the constitution of the universe,, and the destiny of the human soul. The next point is, that the truth of any proposition must be judged with reference to its own unique meaning and intent. It means to assert something specific about something in particu- lar, whether the form of the proposition be particular or general. If it means to assert something about 'this,' it must not be con- demned because it does not assert something else, or because it tells you nothing about ' that,' or because it does not exhaust: the possibilities or attain the ideal of a fully unified knowledge. It may be quite true, for example, that a certain train is scheduled to leave the station at five o'clock, whatever may be true, in metaphysical reference, as to the nature of space and time or, in economic reference, as to the management of a railway system. But if this is so, then we cannot admit, from the propositional point of view, that doctrine of ' degrees of truth ' which asserts that every proposition is partly false because of the modification it would receive by supplementation and re-arrangement when brought into relation with other elements which, for the time being, have been left out of account. This assertion appears to Il6 THE PHILOSOPHICAL REVIEW. [Vol. XVII. rest on a different conception of truth, Judged by its own mean- ing and intent, a proposition may be true without being all that is true, and a truth that is only true about the whole need not be more wholly true than one that is about the meanest of its parts. Again, a proposition that is complex may contain more truth than another without on that account being any more true. It is plausibly objected to this, that truths are not independent, that they at least tend to systematic union. And this we have admitted. But then, it is said, as elements in a system, each truth must modify and be modified by all the others ; as a mem- ber of the system, it cannot remain what it was in isolation, it gets transformed, and the more so in proportion to the width and depth of its connections. And from this it follows, on the argu- ment, regarding ' Reality ' as a system one of whose aspects is a completely unified ' Truth,' that all truths, in the end, are 'error,' and that, for example, mathematics, the most exact of the sci- ences, is also, as the most abstract, the least ' true ' of all. 1 We escape this consequence, I think, by holding strictly to our prin- ciple that the truth of any proposition must be judged with refer- ence to its own unique meaning and intent, and by distinguishing between truth and its evaluation. A given truth does, indeed, suffer modification in being systematically connected with other truths, but such modification need not be at all one of the truth of the proposition, but only of the way the truth is held, under- stood, and appreciated. Thus the schoolboy may know only the isolated truths that 5 + 2 = 7 and that 5 x 2 = 10 ; but if he later comes to see that these truths are connected, that 5 + 2 = 7 because 5 x 2=10, and vice versa, that neither would be true if the other were false, or if, as a philosophical mathematician, he holds a theory of numbers which throws light on the nature and connection of these propositions, he certainly holds these truths in a different way, they have for him a different value ; but how has the truth of either proposition been itself affected ? That 5 + 2 = 7 is, I suppose, as true, neither more nor less, to the mathematician as to the schoolboy, though the former has so many more connected truths at command that it has for him a richer signification. For truths too, like sensible facts, have an 1 Bradley, Appearance ana' /\ea/i/r, p. 370. No. 2.] THE PROBLEM OF TRUTH. II 7 import beyond their own intrinsic quality. Of course, the propo- sition in question is quite true only relatively to the general character of its own number system ; but this was implied in its assertion. But in this reference its truth would not be in the least affected by the discovery, or invention, of a different number system, if that were possible, just as a truth in Euclid is not affected by the equally valid, though less serviceable, truths of other geometrical systems. The fact that one truth is not, as such, altered by its connec- tion with other truths, may appear perhaps in a still clearer light, if we take a case where, as things stand, there is no such connec- tion, and then imagine what would happen if such a connection were brought about. " This table is round," and " this table cost $500," are propositions which have no sort of logical con- nection ; and hence the truth of the one would, in so far, be unaffected by that of the other. But suppose that round tables were exceedingly difficult to make, and that, besides being rare for this reason, they were esteemed peculiarly beautiful. Then they would be objects desired of the rich and coveted by the connoisseur, and a connection between the shape and the price would be so definitely established that we should see at once that a true proposition about the one would involve a corresponding proposition about the other. But would either proposition be more or less true ? Would the table be any more or less round, or its price any dearer or cheaper ? The suggestion is manifestly absurd. The difference would lie not in the truth, but in the truth's evaluation. It being understood, then, that the truth we are talking about is truth of propositions, that every proposition is specific, and that its truth is relative to its intended meaning, we may now state the essential problems in regard to this kind of truth. They may be expressed in two questions : (1) What do we mean by calling any proposition true? and (2) How do we know that it is really true? Or, otherwise stated, (1) What is the nature of the claim we make for it when we call it true ? and (2) How is this claim either established or discredited ? The first question relates to the nature of truth, the second to its evidence. Il8 THE PHILOSOPHICAL REVIEW. [Vol. XVII. But before we attempt to deal with these questions, we ought, I think, to enquire more. particularly, first, into the nature of the object to which the predicates 'true' and 'false' are applied, and the possession of which constitutes that object a truth or a falsity. We have agreed that our concern is with the truth of proposi- tions, but the truth of a proposition is clearly not resident in the mere form of the words. What is true, if true, and false, if false, — and also, it may be added, what is doubtful, possible, neces- sary, etc., — is, primarily, what is asserted. In what is asserted we seem to have the original locus of a propositional truth. If what is asserted is true, then, and only then, is the proposition true, and thereby whatever mental act, content, or attitude it expresses on the part of the individual making or holding the proposition ; and contrariwise, if it is false. Now to apply the adjectives 'true' and 'false' directly to what is asserted, we have, curiously enough, to change the form of the proposition. In the proposition something is asserted of something, something is declared to be or not to be, to happen or not to happen, or, in general, to be so-and-so characterized. If now what is asserted is to be itself characterized, if, for example, it is to be qualified as true or false, it must itself be expressed as the subject of another proposition having such a character as its predicate. And this, as especially pointed out by Meinong, is done by expressing the 'what' that is asserted by a sentence beginning with ' that,' or by some form of words equivalent to such a sentence. Thus in the proposition, " crows are black," what is asserted is that crows are black. The question we must now ask is, What is the logical import of such a //W-sentence ? A proper answer should throw some light on the meaning of truth. In dealing with this question, we may proceed in either of two ways : we may abstract altogether from the thinking process and consider only the logical character of what is asserted, or we may connect the latter with the process out of which the assertion issues and the attitude in which its truth or falsity is recognized, and seek to determine its position and character relatively to that. From either point of view, its most salient feature appears to be that of belonging to an ideal realm of meaning distinct from and, No. 2.] THE PROBLEM OF TRUTH. 119 in a way, opposed to concrete and actually existent fact. That this paper is white, is neither an existing thing, like the paper, nor a real predicate of existence, like the paper's whiteness. The white paper exists, but I cannot in the same way say ' that this paper is white ' exists. I do not mean that this truth can in no sense be said to be. It can be made the object of a reflective thought, it can be examined as such, it can be talked about and become the subject of other true or false propositions. Thus, if it is false that this paper is white, then that this is false, is true. The point is that what is asserted is always ideal, and is never identical in existence with the object that the assertion is about. This is true even in the case when the latter object is itself ideal. 'That 3 is greater than 2,' for example, is neither the number 3, nor the number 2, nor the greater magnitude of the one as com- pared with the other. This difference gives rise to the problem as to the relation of the two, the relation of the meaning to the fact meant, in which it is usual to find the defining character of truth. Leaving this for the present, I may here point to an im- portant consequence of the distinction. There is high authority for the doctrine that truth (and also error) is a content of predication qualifying reality, a doctrine which is developed in the assertion that perfect truth would be the universe. 1 But if our distinction holds good, either this is impossible, or it relates to another kind of truth than proposi- tional truth. For the truth that so-and-so, for example, that this paper is white, is neither the subject of the proposition, nor the predicate, nor any quality of the object taken as real, but some- thing quite different, namely, a truth about it. How is the case altered if for a particular finite object, like this paper, we substi- tute ' Reality ' or the universe ? For whether the content by which the subject of a proposition or judgment is qualified, — and you may interpret your proposition so as to make the ' real ' sub- ject anything you please, — whether this content, I say, be con- ceived as a simple quality, or as a complex of qualifying relations, or, again, be conceived in abstraction as an ' idea ' divorced from 1 "We must unhesitatingly assert that truth ... if for itself it were perfect, would be itself in the fullest sense the entire and absolute universe." Bradley, " On Truth and Copying," Mind, N. S., Vol. XVI, p. 170. 120 THE JFHILOSOPHTCAL REVIEW. [Vol. XVII. existence, or concretely applied as actually qualifying an existent thing, there is, I submit, a clear distinction to be drawn between any finite object, or reality at large, taken as the subject of predica- tion together with whatever it may be said to be or to have, and the truth (or falsity) that it is, or is of such a sort, or has such and such a character. The character of a being is one thing, and may be called an idea or the object of an idea, as we choose to define it ; but that a being has this character is surely not an identity, pure and simple, with the character itself. If, therefore, we assume that Reality is one whole of being with a definite structure, and that this structure, its defining content, is grasped in a single thought, this thought, I suppose, might be said to possess the world in idea. But unless the thought went on to actually predi- cate of Reality as its structure the content thought, it would not possess the truth that Reality was so defined. But if it should effect this predication, then this truth, that Reality was so de- fined, would be, as truth and meaning, quite distinct from the content predicated, and this even though it were itself included in it. I am not, of course, maintaining that it is possible to grasp the world's structure without judging, or denying, on the other hand, the possibility of a speculative grasp, or aesthetic expe- rience, of reality beyond judgment. I am only maintaining that the so-called ' truth ' embodied in the content of predication, though the universe were the subject and though its whole con- tent were exhausted in the predicate, would not be identical with- out difference with the truth of any possible proposition. And I accordingly deny that truth, in the propositional sense, is, prop- erly speaking, a defining quality of any real being at all. It is neither the subject nor the predicate of a judgment ; it is neither substantival nor adjectival. It is a form of ideality, but its own unique form. Viewed in se this form appears, in each instance of it, to be {a) objective, that is, something cognized, or to be cognized, as dis- tinct from the processes of cognizing on the part of any individual mind. Hence it may be treated, for certain purposes, inde- pendently, just as physical objects are treated independently in the physical sciences, without reference to the conditions of our No. 2.] THE PROBLEM OF TRUTH. 121 knowledge of them. It appears (ti) as universal, that is, as claiming recognition and acknowledgment on the part of all minds. But whether it is actually acknowledged or not by any particular mind, seems indifferent to it. Failure to acknowledge it may be due to ignorance or to mental incapacity. Hence it may be maintained that truths, as such, are independent of their recognition by any mind at all. Truth, on this view, would con- sist in an ideal relation between w T hat is theoretically capable of being asserted and the objective fact that the assertion, if made, would be about. So extreme a contention we may not now be prepared to admit ; but the recognition of even the relative inde- pendence of truth should serve, I think, as a salutary check on the tendency evident in recent discussion to interpret the prob- lem of truth exclusively in terms of the process by which the claims of our ideas to recognition as true are tested and estab- lished. The view referred to would mean, I suppose, at least this, that there are real facts in the world, and hence, ideally, truths about those facts which are unknown and some of which, from the very nature of the case, are incapable of becoming known by any finite mind. And this we seem compelled to admit. For not only is knowledge progressive, so that more facts and objects get known or better known, but an infinity of facts col- lectively known are unknown to any single mind, and an infinity of facts once collectively known become irrecoverably lost, namely, the personal experiences of the individuals that made up the succession of all the generations past. Moreover, no finite mind knows, or pretends to know, the world's infinite multi- plicity in all its details, nor the specific ground or grounds of its differences, nor all the implications of any one of its actual experi- ences. No one, however relative to our thought and purpose he holds the world to be, seriously believes that it is wholly plastic, that it is wholly made and remade by our volition, and that there is nothing, I will not say merely given, but given in any sense at all to be simply acknowledged, or that fact and truth only are as they are discovered by us. But if this is so, then the distinction between truth and recognized truth, as well as between truth and the process of testing and acknowledging 122 THE PHILOSOPHICAL REVIEW. [Vol. XVII. it, would seem to have theoretical importance, even though it should be held that what is truth for us cannot be determined concretely apart from the conditions under which it is known. 1 Relatively to the act and process of knowledge, the meaning that is capable of setting up a claim to recognition as true may be viewed in several ways. Primarily it is of the nature of a sup- posal. The ideal meaning may be simply entertained. So far, though the supposal be false, there is no error. If, however, it is accepted, there is judgment and belief, and the belief may be erroneous ; but if it is also accepted, so to say, by the object as tested by the criteria suitable to the case in question, there is true opinion and knowledge. Three distinctions, pointed out by Meinong, seem to be essential in the analysis of judgment. We distinguish (i) the act of judging, — a temporal event in the men- tal history of the individual ; (2) the object or subject-matter that the judgment is about, — this maybe anything you please, but it is at any rate something other than the thinking and the partic- ular thought that aims at the knowledge of it ; and (3) the thought or supposal as an ideal, but immanent, objective content, — what the object is thought as, and what is asserted in the proposition. Here the problem of truth concerns the relation of the 'immanent,' thought-possessed, but objective content of the supposal to the contrasted ' transcendent ' or quasi-transcendent object that the supposal's content means to be true of. Another way of viewing the matter is to consider the supposal, the content of meaning expressed in the that-sentence, as of the nature of an answer to a question, or the solution of a problem. 2 1 Besides objectivity and universality, it is usual to ascribe timelessness and unchangeability also to what is asserted, taken as true ; and these characters, inter- preted in a logical and not in a temporal sense, would seem to hold except in cases where the notion of time enters into the predication, and there the relations are pecul- iar. If the reference is to past time, the truth (