3 The Truth of Propositions (1926)

The account of the truth of propositions which I gave (Mind, N.S., 138) seems to me not only to have been relevant to Dr Schiller's treatment of the question in his criticism of Bradley, but also substantially to embody the fundamental objections to his theory as restated (Mind, N.S., 139) under the heading Judgments versus Propositions. I cannot, therefore, hope to avoid repeating, and perhaps labouring, points that I have already made, but it may be possible for me in further discussion to make the issues clearer. Dr Schiller dwells on the difficulties and errors that arise from confusing propositions with judgments, and the sort of truth we can attribute to the one with that which can be claimed for the other. I should agree that it is highly undesirable to confuse propositions and judgments, if by “judgment” is meant judging; but, if so, there is no question of the truth of judgments, since, as I said in my previous discussion, it is for what a person judges, and not for his judging, that truth is claimed. On the other hand, if “judgment” is taken to mean what is judged, then I should deny that there is any distinction between judgments and propositions, and so any question of different sorts of truth. If to argue along these lines is to raise “the issue of Realism v. Idealism”, I do not see that it is irrelevant to the points in Dr Schiller's discussion of Bradley which I was criticising. Since what is judged is commonly not in a psychic setting, “truth in a psychic setting” can hardly be claimed for it.

But I argued further that even if we consider the actual conditions of what is judged, we still cannot claim for it “truth under those conditions” or conditional truth; that, unless we can think of it as having unconditional or unqualified truth, we cannot call it true in any sense. It is not a question of the “formal independence of truth-claims”, but of the claim of independent truth which we make whenever we assert anything as a matter of fact, i.e., whenever we assert anything at all. I had argued that Dr Schiller's theory was sceptical, because it denied that assertions (things asserted) were to be taken as simple matters of fact, or the reverse, and so implied that no definite assertion could be made. But the point can equally well be put by saying that any theory of the kind is illogical, since it can only be upheld by the making of statements of supposed fact. Now such statements are propositions. It appears to me, therefore, that this line of argument was strictly relevant to the attempt to show that, so far from adherence to propositions being capable of inducing scepticism, scepticism can only be avoided by

  ― 21 ―
insisting on claims to unqualified truth and rejecting claims to “truth in a context”. If Bradley aimed at both absolute truth and truth under conditions, he would necessarily arrive at a negative result. What I am maintaining, as against Dr Schiller, is that the aiming at truth under conditions must have been responsible for this, and not the aiming at absolute truth.

This “absolute truth” is, of course, just the truth that we claim for what we definitely believe. There is no question of truths of which we can be eternally certain, of beliefs which under no conceivable circumstances could we give up. Any proposition whatever can be denied, i.e. can be conceived to be false; and we have all had experience of giving up beliefs which we once confidently held. But while we held them, we held them to be absolutely true; we could not then imagine that there would be circumstances under which we should give them up, since we took them to hold quite independently of us. Otherwise, we could never have had any need to give them up, nor, if we had come to think differently, should we have thought that we were previously wrong. The point is, then, that at any given time there are certain things that we believe and, in so doing, regard as matters of fact. And if there is anything about which we are uncertain, what we are uncertain about is whether or not it is a matter of fact; we do not regard it as a matter of “uncertain fact”. Experience has shown us that we make mistakes, but it could not show us anything at all unless we sometimes made no mistake. Thus the mere possibility of contradiction, the general consideration that “we may be wrong”, could never lead us to give up a particular belief that we held; only a belief in other propositions which disprove it, could do so. The very fact that contradictory views are held is sufficient to show that some beliefs are true. Now when any belief is true, what is believed (the proposition) is something that has occurred; and when a belief is false, it is still, in being believed, supposed to have occurred. To speak, on this basis, of “absolute” truth, while it may be said to add nothing to the notion of occurrence, at least emphasises the fact that we cannot speak of relative or conditional occurrences.

I had contended that to reject this view of the truth of propositions, “it would be necessary to show that we do not mean by ‘truth’ something which actually occurs”. But though this view is equally opposed to “correspondence” and “coherence”, Dr Schiller has not dealt with it in arriving at the conclusion that “the sense of ‘truth’ which a ‘proposition’ may fitly claim would seem to be either that which is involved in the ‘correspondence’ or that which is involved in the ‘coherence’ theory of truth”. I shall not attempt to elucidate these “claims”, since if it is possible for a proposition to claim to be a matter of fact, there can be no other sense of truth to consider. I still maintain that what is “proposed” or supposed in a proposition is a certain state of affairs, and that whoever believes the proposition takes that state of affairs to have actually occurred — as he indicates by the use of the copula “is”. When, for example, I invite anyone to believe that Bradley is sceptical, I ask

  ― 22 ―
him to consider the actual Bradley and whether in his works he displays the actual characteristic of scepticism. I might attempt to prove my point by argument, but what I should expect to be proved at the end of the argument would simply be the occurrence (truth or fact) of Bradley's scepticism. In general, then, when a person formulates a proposition, the copula indicates that he thinks something has occurred, and the terms (the different functions of which need not be considered here) indicate what he thinks has occurred. In other words, a proposition is something which can be thought to have occurred or not to have occurred. But thinking that something has occurred is simply judgment, in the sense of judging. Thus when we speak of judgment in the sense of what is judged, we are speaking about a proposition; and the proposition or judgment is true, when the supposed situation has occurred. There is no question here of how we know this, how we can be “sure of our facts”. It is sufficient that we have beliefs, and that this is what they mean; that believing something and believing that it has occurred are the same thing. It still seems to me that this line of argument suffices to dispose of any theory which takes truth to depend on adequacy or relevance, or, in general, to be a matter of degree. We cannot think of situations as more or less occurring or as conditionally occurring. There are, as I said, cases where we assert that A conditions B, but then we take this whole situation as an absolute fact. In short, there can be no intermediate stage between absolute occurrence and absolute non-occurrence.

The difficulties about verbal forms and ambiguity, which Dr Schiller again raises, seem to me to go no way towards showing that a true proposition ought to be said at best to “correspond to a fact”, instead of to be a fact. If a proposition were merely a form of words, and words had that arbitrariness which he appears to assign to them, I do not see how it could even correspond to a fact. And the same can be said, if for “proposition” we substitute “statement”. By a statement we mean something stated by means of words, and by literal truth and literal falsity we mean that the propositions so stated are either truths conveyed by words or falsehoods conveyed by words. Dr Schiller thinks that if I demand literal truth, I have made for myself a short cut to scepticism, since “literal truth is at most only verbal, and, for the purpose of determining real truth, quite inadequate. It is for example bowled over by the slightest hint of ambiguity”. Yet we all make statements in words and believe them, and hence believe that any one who denies them is wrong; and, as I said in my previous discussion, “no one will deny” (and Dr Schiller has not denied it) “that knowledge of actual occurrences is conveyed by means of words”. It is not at all necessary for me to maintain that misunderstanding never occurs, but only to point out that understanding occurs and to indicate what sort of thing it is that is then understood. There could not be understanding of statements, unless it were false that a statement “always has, in principle, a plurality of meanings”. The principle of verbal communication is that the set of words used in making a given statement should have only one meaning.

  ― 23 ―
This is the principle that we employ in learning any language, our own included. There are exceptional cases where the same word stands for several distinct sorts of thing, and in these cases, as I pointed out, misunderstanding is commonly removed by the use of other words. But for the most part the distinct things we have to know are a certain arrangement of words, an occurrence of a certain character and the fact that those who understand the language think of the latter when they see or hear the former. This may be a state of affairs which has not “passed unobserved by anyone”, but it has to be pointed out in order to reinforce my previous contention that misunderstanding of words and misunderstanding of things are on precisely the same footing. Persons may make mistakes about words, but this does not render it impossible for a language to be understood and thus for literal truths to be believed and communicated.

In denying this, Dr Schiller, though he claims to have made a clear distinction between the truth of propositions and that of judgments, has not really shown what he means by “the ‘truth’ (in general and in the abstract) of a ‘proposition’”, or what is the “verbal meaning” to which formal logicians are supposed to confine themselves. In general and in the abstract any word might mean anything, and thus on a merely verbal basis there could be no such distinction as that between true and false propositions. But formal logicians surely proceed on the basis of the fact that, since we are capable of knowing what words mean, we can believe or disbelieve statements in words, i.e. find them literally true or false. And Dr Schiller himself, when he objects to the assumption “that the mere hearing of a proposition, which once formulated what seemed a truth in the judgment of its first discoverer, will suffice to re-start the same process in any mind that hears it”, has implicitly conceded, in the words “seemed a truth”, all that they require for their purpose. What was it that it seemed, and what was it that seemed so? It was a state of affairs that seemed (was judged) to have occurred. If words do formulate such suppositions and if they ever enable a hearer to make the same supposition, then true and false propositions can be stated and understood without being “verbal”. But it does not appear, on Dr Schiller's theory of propositions, how a proposition could ever seem to formulate a truth.

It is equally difficult for him to account for the notion of “seeming true” in terms of his theory of the truth of judgments. He wishes to consider “what the actual judgment meant” and to “discount as irrelevant the formal truth-claim made by all judgments true or false”. But we cannot leave the claim aside without explaining, as Dr Schiller does not do, what it is that is claimed; and it can hardly be denied that the person judging means to make this claim. All that would remain, if it were removed, would be the material, the terms, of the judgment, and they do not by themselves make it a judgment or convey what it means. The truth a judgment claims, then, and the truth it may possibly have, is what is indicated by the copula; viz. occurrence. It is certainly not adequacy to a situation, for if anything was claimed as adequate to a

  ― 24 ―
situation, it would be its adequacy that was said to have occurred. This line of argument is in accordance with my previous criticism of “contexts”. I pointed out that in his original discussion (Mind, N.S., 134, p. 221) Dr Schiller had said that “the judgment means that in view of all the circumstances present to its maker's mind and judged relevant by him, he has judged it best to make his judgment”. That is to say, when a man judges that A is B, he has ipso facto passed the judgment, “Judging that A is B is the best I can do under the relevant circumstances”. The infinite regress involved in attempting to maintain this position is obvious. On the other hand, it is equally obvious that the man is supposed, in the second place, to make a judgment of fact — “Judging that A is B is the best”, etc. Similar criticisms may be passed on what Dr Schiller now says — “The making of a (bona fide) judgment thus ipso facto implies its maker's belief that it was” (my italics) “the best response to the circumstances which he could conceive”. If he can make a judgment of fact in the second place, there is no reason for turning his original judgment from a claim to truth in fact into a claim to “truth in a context”. Dr Schiller contrasts the discussion of the truth of a judgment, as a question of value, with that of its meaning, as a question of fact. But if there is such a thing as a matter of fact, surely that and that only is what we mean by truth.

I should therefore maintain that Dr Schiller has not made it clear what he means by the “truth of judgments”, any more than by the “truth of propositions”. It may be admitted that whether or not we pass a certain judgment will depend on our purposes, and again that there is a distinction between judgments which are relevant to a certain inquiry and those which are not. But this has nothing whatever to do with truth; a relevant judgment may be false, an irrelevant one true. It may be more to the purpose of a certain discussion to consider whether a penny stamp is scarlet or not than whether it is red or not, but this fact will not enable us to determine whether it is scarlet. Moreover, if it is scarlet, it is also red, while if it is red, it may not be scarlet. Thus Dr Schiller has said nothing to show that being relevant makes a statement true. He has correspondingly not shown how being “misapplied” could make a “proposition” false. “Even the best and ‘truest’ form,” he says, “will not be applicable to all situations”. But the only situation to which it could be supposed to be “applicable” is the situation which, it says, occurs. I can only assume that the question is of implication; that, for example, we should “misapply” the proposition “All men are mortal”, if we proceeded from it to believe that “All chessmen are mortal”. But what this meant would be that we supplied a minor premise “All chessmen are men”, which is false, and which is therefore capable of leading to false conclusions. It does not, accordingly, appear either that a true proposition can by itself imply false conclusions, or that one proposition can be an “application” of another.

This raises the question of distinctness and connectedness, what a proposition or group of propositions implies being something distinct

  ― 25 ―
from itself, though connected with it. Dr Schiller, in his criticism of my remarks on selection, takes no account of my statements that things which are connected are at the same time distinct, and that a number of distinct things may be taken together as constituting one thing (as articles of furniture constitute a suite) without being any less distinguishable. Had he done so, he could hardly have referred to my “assumptions that there is only one way of selecting and that its objects are all distinct and lying about waiting to be recognised”. I gave definite examples of different ways of selecting and of the embodiment of one thing in another (though the two are still different things). And I pointed out that though “the truth of the matter is independent of our consideration”, our interests may lead us to consider various combinations and components of things. I cannot therefore consider myself entirely responsible for the obscurity which Dr Schiller finds in my argument. I repeat that unless we can speak of things independently, speak of them, that is, as simply occurring and not as “conditionally occurring”, we cannot make intelligible statements (or have beliefs) at all. The fact that there are conditions under which a state of affairs occurs, does not make it occur conditionally; it occurs and can be known to occur even if its conditions are not known. As before, I can know that Bradley was sceptical without knowing under what conditions he became so, or under what conditions in general scepticism arises.

What I was primarily considering, in discussing “selection”, was whether knowing a distinct state of affairs in the above way could be so described; whether, as I said, selection “is an admissible description of judging”. And I concluded that we might be said to select occurrences, in that our interests and purposes led us to consider certain occurrences and not others. What I denied was that our selection of a predicate for a selected subject could possibly be prior to any question of truth, since predicating one term of another is stating what we suppose to be true; and that we required a “right” to select, since, though we may choose or select certain objects of interest to us, we do not choose to select. Before it can be said that we exercise a right to select, it would have to be shown what would happen if we did not; but this could not be shown — knowledge of what was “neither one thing nor another” being quite inconceivable. Prima facie, then, the practice of speaking about distinct things stands in no need of justification. There is no distinct thing called “the real” which can be spoken of as a “continuum”, and which requires to be broken up before any other thing can be spoken of. Unless things were first of all distinguished, we could never “find from experience that great masses of reality are in fact irrelevant to our purposes and may safely be neglected”; hence this cannot be given as a reason for selecting. We can certainly “select wrongly” in that we may take things to be distinct which are not so, or things not to be distinct which are so, but we cannot consider whether we are entitled to make distinctions or not, since any sort of thinking involves distinction. In accordance with this view it should be said that truth is to be sought not in concentration upon

  ― 26 ―
“parts”, but in consideration of things or description of occurrences.

I have tried to show the importance of the fact that such descriptions are either true or false. I admit that we want to know whether they are true or false, but we cannot come to a conclusion in any given case by considering “rights” or statements of logical theory; we can only do so by observation or, in general, by reference to propositions which we believe. And the nature of belief requires the rejection of any theory of distinct sorts or different degrees of truth; truth being simply what is represented by the copula “is” in the proposition. Any such theory, or any view which attributes different meanings to “is”, is inherently sceptical or illogical, since only by the use of the unambiguous “is” of occurrence (as I have shown in relation to Dr Schiller's view, in particular) could the theory be formulated at all. We must think of propositions, therefore, as capable of being unconditionally true; a consistent adherence to the treatment of them as merely verbal forms would not allow of any enunciation of belief, that is, of any “judgment”.