Bertrand Russell

Russell Society Home Page

About Bertrand Russell

About the Russell Society

The BRS Library

Society Publications

Russell Texts Online

Russell Resources

JOIN the Russell Society!

Officers and Organization

Contact Us

 What Is Truth?  (1906)*

By Bertrand Russell

Review of The Nature of Truth: An Essay. By Harold Joachim. Oxford: Clarendon Press, 1906

THE question “What is truth?” is one which every philosopher ought to face, although, unfortunately, since Pontius Pilate’s rather ill-timed introduction of it, it has become unfashionable to ask it. Mr. Joachim has done very well in undertaking a serious and careful discussion of the nature of truth. The advocates of any system of philosophy are too apt to assume its fundamentals as indubitable, and devote themselves to the mere development of consequences. This course is attractive, both because it is easy, and because it seems to achieve more in the way of positive construction. But, so long as disagreement on fundamentals persists, the development of consequences must appear as in the main waste labour to those who do not accept the premisses. Mr. Joachim’s book is valuable as an attempt to establish some of the fundamentals of the Hegelian philosophy; and, whether wholly successful or not, such an attempt is almost sure to be a help in defining the issues, and in suggesting ways of deciding them.

The book discusses three different theories of the nature of truth, and then proceeds to discuss error. The first theory of truth, which is the one the plain man would naturally adopt, is that truth consists in the correspondence of our statements or beliefs with the facts. This view is open to criticism from many points of view. Mr. Joachim criticises it on the grounds that the “correspondence” involved supposes a collection of distinct “facts,” which gives too atomic a view of the world, and that there is not really such a separation of judgment and outside fact as the theory supposes. In this criticism, he assumes that everything is modified by its relations to everything else, so that no two things are really independent, and that you cannot speak quite truly about anything without speaking the whole truth about everything. The assumption that everything is modified by its relations to everything else, being rejected by the second theory of truth which Mr. Joachim examines, is defended in the course of the examination of this theory.

The second theory (which is held by the present reviewer) maintains that truth is primarily a property of facts, which are something external to minds and to mind. “That the earth goes round the sun,” it says, is true, independently of whether any one thinks so, and independently of even the mere notion of its being thought. The belief that the earth goes round the sun, according to this theory, is true in a derivative sense, namely the sense that it is a belief in a fact; but the fact itself, the actual revolution of the earth round the sun, is something quite different from the belief in the fact.

This theory, as Mr. Joachim points out, stands or falls with the view that “experiencing makes no difference to the facts.” If I see a banker’s clerk descending from a bus, my seeing him does not turn him into a hippopotamus, but leaves him just what he would have been if I hadn’t seen him. This is denied by Mr. Joachim, on the ground that experiencing a fact is a relation to the fact, and that everything is modified by its relations. The view that everything is modified by its relations, is, of course, in one sense obviously true. But the sense in which it is assumed by Hegelians is not the sense in which it is obviously true. What they mean may, I think, be roughly expressed as follows. Suppose A is the father of B. Then, if you try to think of A without at the same time thinking of B, you are not really thinking about A at all, since paternity to B is part of A’s nature. You are thinking instead of an abstraction, in which you have omitted paternity to B, which is essential to the real A. Similarly, if A, instead of being a person, is some fact which B knows, you cannot think of A without at the same time thinking of B, since “being known to B” part of A’s nature. It follows, since everything is related, more or less, to everything else, that to think quite truly, you must think the whole truth; everything except the whole truth about the whole world omits something essential, and thereby fails to be quite true.

It is astonishing how far-reaching are the consequences of this logical doctrine as to relations. It leads straight to the view that nothing is quite real except the universe as a whole; that time, space, and matter, are unreal abstractions; and that there can be no reality which is not known to mind, though it may be unknown to this or that mind. I do not mean that Hegelians put forward this doctrine of relations as the foundation of their system, I mean that, when their arguments are analysed, they are all found to assume it, consciously or unconsciously. In Mr. Joachim’s book, the assumption is conscious ; but, unfortunately, nothing is done to support the assumption, except to show that its rejection leads to consequences which are incompatible with it – a characteristic which it shares with other assumptions.

The third view of truth, which Mr. Joachim considers better than either of the others, though not quite adequate, is the view of truth as coherence. This follows naturally from the doctrine that nothing is wholly true except the whole truth. Thus, a whole science is truer than any part of it; and so on. The essence of truth, in this view, is systematic coherence in an organic whole. The difficulty, to my mind, of accepting this as more than a criterion is that the meaning of coherence remains obscure, and that there is no evidence that there can only be one body of coherent propositions. One can imagine quite different worlds from the one we live in, just as coherent as this one.1 As a criterion, coherence, in some sense, is certainly invaluable; but when used as a criterion, there is always some nucleus with which other things have to cohere, the nucleus itself being accepted not only on account of its coherence with something else.

The last chapter, on Error, is very interesting, and is greatly to be commended for its candour. The problem of error, like the problem of evil, exists only for an optimist. If your theory proves that everything is good and true, it is awkward to have to add a postscript to say that after all some things are bad and false. The coherence theory of truth, however, provides an opportunity for error which is very convenient. Nothing, on this theory, is quite true except the whole truth; on the other hand, nothing is quite false. Thus error consists in supposing a partial truth to be quite true. “The erring subject’s confident belief in the truth of his knowledge distinctively characterises error, and converts a partial apprehension of the truth into falsity.” That is to say, error consists solely in rejecting the Hegelian doctrine that no single proposition can be quite true. So long as you do not reject this doctrine, and yet avoid thinking it quite true, you are safe. This reminds one of the Catholic doctrine that the sin against the Holy Ghost consists in the belief that a man can be saved without merit, i.e. in the acceptance of Justification by Faith, which is (or rather was) the distinctive dogma of Protestantism.

But it is difficult to make this theory of error fit the facts. Suppose A swears that B committed a murder, and B swears he didn’t, when in fact he did. If the jury, not being Hegelians, believe that A is speaking quite truly, they will surely be less in error than if they believe that B is speaking quite truly. Yet, on Mr. Joachim’s theory, both beliefs ought to be equally erroneous, since each is a half-truth. The jury, to avoid error, ought, whatever the evidence, to decide that B is half guilty and half not-guilty; and the judge ought to sentence him to be half hanged and half acquitted. But it will be unnecessary actually to pass sentence. For, if everything is a half-truth, everybody must be half hanged and half acquitted; and the operation of the law is superfluous.

These farcical deductions are, of course, avoided by the theory of degrees of truth. Although everybody is more or less hanged, some are much more hanged than others. This doctrine has difficulties of its own, especially the difficulty, which is fully faced and discussed by Mr. Joachim, that if no single truth is quite true, it cannot be quite true that no truth is quite true. This difficulty, inherent in the contention that only the whole truth is wholly true, leads Mr. Joachim to the admission, that the coherence theory is not quite adequate, and that his result is mainly negative. Nevertheless he says: “That the truth itself is one, and whole, and complete, and that all thinking and all experience moves within its recognition and subject to its manifest authority; this I have never doubted.”

The book is thoroughly self-consistent, and is a well-sustained endeavour to probe what is perhaps the most difficult of all problems. Agreement among philosophers is not to be expected; but it is desirable that each of the great types of philosophy should render its foundations explicit. And to this process Mr. Joachim has made a most useful contribution.


*  Bertrand Russell, “What Is Truth?” The Independent Review 9 (Jun 1906), 349-53  Review of Harold Joachim, The Nature of Truth, 1906

1 Mr. Joachim endeavours to show that this is impossible; but I fail to see any force in his argument (see p. 78)