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



 The Nature of Truth  (1906)*

By Bertrand Russell

IN a recent work,1 Mr. Harold Joachim has examined at some length certain opinions held by Mr. G. E. Moore and myself. I propose first to discuss his arguments against these opinions, and then to consider briefly whether any arguments are possible which ought to appeal to both sides in such a debate.2 The difficulty, as regards the second question, lies in the fact that our differences are so fundamental that almost all arguments on either side necessarily begin by assuming something which the other side denies. As Mr. Joachim justly observes: “If an assumption is the basis of all Logic, then arguments directed against it appear, by a very natural confusion, to be eo ipso devoid of logical cogency. The assumption, in fact, gets established by a kind of ontological proof” (p. 39). This state of things is very unsatisfactory, and it is highly desirable to find, if we can, some common ground on which discussion is possible.

The assumption which Mr. Joachim sets out to refute is that “experiencing makes no difference to the facts” (p. 33). What, precisely, this assumption is to mean will appear more clearly in the course of the discussion. Mr. Joachim distinguishes two senses of the assumption, one of which, he says, is true but irrelevant, while the other, which alone is adequate to support the theory he is combating, “is false, conflicts with common-sense, and is in the end unmeaning” (p. 41). In the admissible sense, the facts are only independent of the experience of this or that particular person, but, not of “being experienced” in general. In the inadmissible sense, “the whole constituted by ‘the facts’ and ‘experiencing’ (in any sense of the term) is no genuine whole, but a mere external adjustment. The two factors are, or may be, related; but the relation. when, or as, it obtains, leaves each precisely what it was, viz., absolutely in itself and independent” (p. 41). We need not further consider the admissible sense, since, as Mr. Joachim contends and as I fully acknowledge, it is irrelevant to the discussion. As to the sense which he regards as inadmissible, his statement of it calls only for the remark that the whole which, as he puts it, “is no genuine whole,” is as much a genuine whole as any that this theory will admit; for it is a characteristic of the theory that its views as to the nature of whole and part are quite different from those adopted by the school to which Mr. Joachim belongs.

Mr. Joachim’s objection to this view comes to this: that if A is independent of B, A cannot be related to B; that, consequently, if the facts are independent of experience, they cannot be experienced; and nothing at all can be known. Thus he says (p. 42): “Greenness is, for the theory, an ultimate entity in the nature of things, which has its being absolutely in itself. How, under the circumstances, greenness can yet sometimes so far depart from its sacred aloofness as to be apprehended (sensated or conceived); and how, when this takes place, the sensating or conceiving subject is assured that its immaculate perseitas is still preserved – these are questions to which apparently the only answer is the dogmatic reiteration of the supposed fact.”

Mr. Joachim alleges that the plain man is on his side. I have been tempted to ask some plain man what he thought greenness was, but have been restrained by the fear of being thought insane. Mr. Joachim, however, seems to have been bolder. Considering the difficulty of finding a really plain man nowadays, I presume he asked his scout, who apparently replied: “Well, sir, greenness is to me the name of a complex fact, the factors of which essentially and reciprocally determine one another. And if you, sir, choose to select one factor out of the complex, and to call it greenness, I will not dispute about the term, for I know my place, sir; but as thus isolated, your greenness is an abstraction, which emphatically, in itself and as such, is not there nor anywhere.” At least, this is what I gather from the opinion of the plain man reported on p. 42. “Who shall say,” he concludes, “that his is the insight of a lying prophet, while yours bears the divine stamp of truth?” The answer to this question would require a whole treatise; for the present, therefore I will confine myself to Mr. Joachim’s contention 3 that mine is the insight of a lying prophet, while his bears the divine stamp of truth.

It is evident that the apparent force of Mr. Joachim’s argument lies in the use of such phrases as “sacred aloofness,” in which he assumes that if greenness is independent of experience, it cannot be related to experience. For I do not maintain that greenness, or anything else, has any “sacred aloofness”; I contend merely that there is such a thing, having various relations, among others relations of being perceived. Mr. Joachim’s argument, in fact, depends upon the assumption that, in any complex, the constituents of the complex are nothing; i.e., you cannot find an entity A, and say “A is a constituent of this complex.” This view is connected with the opinion that all relations are based upon the “nature” of the related terms. Thus he says (p. 43): “Clearly, the sentience to which greenness can be related is ‘vision,’ not ‘hearing.’ But we are to understand that this restriction is not based upon the nature of greenness as such, but is just a fact.” I do not know what the “nature” of greenness, as opposed to greenness, may be; it seems to be the ghost of the scholastic essence. This claim that relations are to be grounded in the natures of their terms is really a claim that all propositions are to be of the subject-predicate form, so that instead of saying “A and B have such-and-such a relation,” we should say “A has such-and-such a property and B has such-and-such another; these properties being part of the natures of A and B.” And as to the relation being “just a fact,” so, on the opposite view, is the “nature” of greenness; for why should it not have had a different “nature?”

What emerges in Mr. Joachim’s discussion is, as he himself points out, that the views he is attacking are only tenable on the assumption that relations are “external,” i.e., that there is no such thing as the “nature” of the related terms in cases in which these terms are simple, and that relatedness is no evidence of intrinsic complexity. This is the fundamental doctrine of the view which he is criticising; the opinion that “experiencing makes no difference to the facts” is merely a special application of this fundamental doctrine. Having brought the argument to this point, one expects to find reasons alleged against the doctrine in question, but strange to say, no reason whatever is given except that it seems incredible to Mr. Joachim. The curious thing is that, elsewhere, he protests against immediate inspection as a test of truth, holding that coherence in a system is both the test and the meaning of truth. Nevertheless, in this instance, although he admits later (e.g., p. 178) that the system resulting from his assumption is not completely coherent, so that nothing except immediate inspection is left to recommend it, he is content to regard his view as firmly and irrevocably established by the fact that he cannot imagine it false.

But to support this statement, I will quote some of the principal sentences in the pages (pp. 45-49) which deal explicitly with this fundamental point. “That any Simples should combine” is “an arbitrary irrational fact, if it be a fact at all…. How can you treat them as each absolutely simple and independent, and also as related to one another to form a complex?” After setting forth that, in my opinion, the same greenness and ‘precisely and numerically the same’ relations enter as constituents into an indefinite number of different complexes” (p. 47) he proceeds:

“In this account of the union of Simple Entities to form Complexes, I can see nothing but a statement of the problem in terms which render its solution inconceivable. If you tell me that a penny in my pocket is ‘the same’ coin as a penny in yours, I agree that in a sense this is true enough. But if for the penny you substitute a simple eternal entity, and then go on to maintain that this simple self-identical entity is both in my pocket and in yours, and also in no place and at no time, I can only protest that a simplicity of this kind is too deep for me to fathom. Nor does it make the least difference if you call your simple entity a ‘universal.’ And if, finally, you insist that the relation of the simple entity to the points of Space which are my pocket, is ‘precisely and numerically the same’ as its relation to the points of Space which are your pocket, I must admit that I am unable to distinguish a ‘precise numerical identity’ of this kind from numerical diversity” (p. 48).

I have quoted this paragraph in full, because it contains the whole of what Mr. Joachim has to say on the fundamental question at issue.4 He passes on immediately to other views, considering the view in question sufficiently refuted.

Before considering the main point, it seems necessary to clear up two misunderstandings. First, I should not say that a penny in your pocket was the same as a penny in mine, unless it was the same penny, i.e., unless you had taken it out and given it to me. For a penny is a piece of matter, and its identity consists in being composed of the same particles. But the qualities in virtue of which we call it a penny (qualities which it may lose without losing its material identity) are, I should say, numerically the same in so far as they are not different qualitatively. Secondly, I do not maintain that greenness (e.g.), is “also in no place and at no time.” I maintain that greenness can be considered without regard to the spaces and times in which it is, and that in so considering it we do not alter it; i.e., it is possible in thought to isolate it, and in so doing we merely disregard its relations without in any way mutilating it itself.

Coming now to the main question, what is the essence of Mr. Joachim’s contention? “In this account,… I can see nothing but a statement of the problem.” “A simplicity of this kind is too deep for me to fathom.” “I am unable to distinguish a ‘precise numerical identity’ of this kind from numerical diversity.” The difficulty is that none of these are of the nature of arguments. They are simply statements as to what Mr. Joachim can or cannot imagine. It is, of course, implied that what he cannot imagine is nothing; but this can hardly be taken to be one of the fundamental premisses of all philosophy. The importance of the point is very great; for, except certain historical portions, and a few paragraphs at the end in which fundamental doubts are discussed, there is hardly anything in the book which does not assume that all the constituents of a complex must be complex. It seems a pity, therefore, that no grounds whatever should have been alleged in favour of this view, except that to Mr. Joachim the opposite appears inconceivable.

The curious and discouraging thing about this dispute is that conversely I cannot see what the ‘problem’ is which I am supposed to be merely re-stating. That the same man, in the strictest sense of the word ‘same,’ should be both the son of one man and the brother of another, or that he should be the brother of two men, or that greenness should have a resemblance to blueness and also to yellowness – such facts do not seem to me to call for any explanation. The demand for an explanation seems to depend upon some supposed law of sufficient reason – upon the notion that everything must have a reason for being as it is and not otherwise. Such a view can be supported by theism or by any teleological philosophy, though even then it is difficult to assign a sufficient reason for God. But apart from some such assumption, I cannot see why we should expect a reason for everything. And in spite of many efforts, I cannot understand why it should be thought that relatedness implies complexity; and, unfortunately, Mr. Joachim, though he holds that there is a reason for everything, does not offer any reason for his opinion about relations.

The arguments in the pages we have been considering are, therefore, such as will only appear cogent to those who already admit the conclusions which the arguments are intended to prove. This leads to the further question: What arguments on such a question are conceivable, which do not assume the question already decided? I think the only possible argument of this kind, on all fundamental questions, is some form or other of the reductio ad absurdum. That is to say, a position can be refuted in the eyes of one who previously held it if, assuming it to be true, and using only inferences of a kind which it admits to be valid, the falsehood of some essential part of the position can be deduced. This, of course, assumes that whatever implies its own falsehood must be false; but this assumption is made by all philosophers. The Hegelian dialectic is in part an argument of this nature: the inadequacy of the thesis is shown by the fact that it implies the antithesis, which is inconsistent with it. Mr. Joachim himself supplies an argument of this type, by showing that, if coherence (in his sense) is the essence of truth, then it cannot be quite true that coherence (in his sense) is the essence of truth. And if he had intended to refute the view of truth advocated by Mr. Moore and myself, it would have been necessary to accept it provisionally, and to have shown that it led to consequences inconsistent with itself. This method is, of course, difficult, because it is difficult to realise the position of an adversary so clearly as to be able to avoid assumptions which he rejects. And as against Hegelianism, at least in a form which it frequently assumes among its disciples, the method seems essentially incapable of employment. For wherever a contradiction is demonstrated, the Hegelian has only to raise an altar “to the Unknown Synthesis,” and it is easy for him to show that, according to his philosophy, there must be a synthesis unknown to him as a finite being. Hegelianism is, therefore, not internally refutable, so long as it is content to admit that as yet it knows nothing at all – an admission which it is apt to make in its last chapters, but which never prevents absolutely certain knowledge that its adversaries are mistaken.

So far as Mr. Joachim’s book goes, it would appear that the views advocated by Mr. Moore and myself are also not internally refutable: at least, this book does not attempt such a refutation. This state of things, it must be admitted, is very unsatisfactory, and seems to render the progress of philosophy almost hopeless. And at the best, even when the reductio ad absurdum can be successfully effected, its result is purely negative: it merely disproves some opinion, without thereby showing that some other opinion is right. It seems to me that the only hope lies in a more careful scrutiny of the premisses that are apt to be employed unconsciously, and a more prolonged attention to fundamentals, in the hope that gradually the area of agreement may be enlarged. Hitherto it has been the fashion to extol construction at the expense of criticism and analysis, and until very recently most philosophers have considered it an essential part of their business to provide something that could be called a proof of the existence of God. Hence premisses have been accepted at haphazard, almost without reflection; and attention has been almost wholly concentrated on results. This habit is especially fostered, it seems to me, by the Hegelian philosophy, which denies that it has any premisses, and therefore leaves to its opponents the task of discovering what its premisses are. In a work on “the nature of truth,” one might have hoped to find some defence of the premisses; but at any rate it is a rare merit in Mr. Joachim’s book that it makes some of the premisses explicit, which is perhaps as much as a philosophical work can be expected to do.

B. RUSSELL


*  Bertrand Russell, “The Nature of Truth,” Mind, n.s. 15, no. 60 (Oct 1906), 528-33  Review of Joachim, The Nature of Truth, 1906

1 The Nature of Truth, Oxford, 1906.

2 The relevant portion of Mr. Joachim’s book is contained in pp. 31-50. The rest of the book avowedly assumes his opinion on the questions in dispute: it is only in these twenty pages that he seeks to establish his opinion. The views which I am defending will be found in MIND, n.s., no. 52, and in my Principles of Mathematics, especially §§ 55, 212-16, 424-8

3 It may be objected that Mr. Joachim’s position does not depend on mere “insight,” since he has a criterion for deciding between rival insights. My reply is that his criterion is established by assuming his view on a logically prior question, and that this assumption is unsupported by the criterion to which it leads.

4 There is also a note, p. 47n, which offers me a dilemma; but this provides no argument, since the second horn of the dilemma (which is the one I accept) is refuted merely by the question: “How can it [greenness] – a simple numerically identical entity – enter into different existent complexes?” No attempt is made to show that it cannot: the impossibility is regarded as self-evident