The Truth About F. H. Bradley Stewart Candlish Mind, New Series, Vol. 98, No. 391. (Jul., 1989), pp. 331-348. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28198907%292%3A98%3A391%3C331%3ATTAFHB%3E2.0.CO%3B2-2 Mind is currently published by Oxford University Press.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/oup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact
[email protected].
http://www.jstor.org Fri Jun 22 07:32:17 2007
The Truth About F. H. Bradley STEWART CANDLISH Introduction
I.
Despite the efforts of some recent propagandists, most philosophers these days still do not read Bradley. T h e result is that they are content to work with a stereotype, whose sources and accuracy go unexamined. This stereotype is inaccurate, and most of us have got Bradley wrong. Perhaps this would not matter much if the error had no serious consequences. T h e problem is, though, that being wrong about Bradley, in the ways in which we are wrong about him, does have serious consequences. One is that we misunderstand our own history, and thereby misunderstand ourselves and the nature of what we are about. Another is that we are deprived of one of the opportunities to connect what we do in philosophy with what some of the philosophers of the past have done; we may, in consequence, fall under the impression that only we, and those obviously like us, are really doing philosophy, at least as it ought to be done. (This, a perennial temptation for philosophers, leaves the residual problem of explaining just what these other chaps, who after all were not stupid, were actually doing. But that problem is usually ignored.) But perhaps the most unfortunate consequence is that we may in our ignorance miss some interesting ideas which cannot be found in the modern philosophical canon. That there is in Bradley such an idea, in the form of a not currently recognized theory which was at least intended to be a competitor of currently recognized theories, is what I shall be arguing in this paper. 2.
T h e stereotype1
T h e elements of the Bradley stereotype are various, but their sources are not: by and large, one need look little further than Moore and Russell. Here are some of those elements, together with their sources. (I should add that the references I give for those sources are, mostly, merely typical, and should not be regarded as the sole places in which such claims are made. And I should add further that in them, particularly in Russell, Bradley is sometimes stigmatized only by implication, in the context of remarks applied generally to 'Hegelians'.) Bradley was a metaphysician in some bad sense of that word (Ayer); as
'
In this section I plunder an earlier paper of mine, 'Bradley's Logic and Bradley? Logic', Philosophical Books, 1984. Mind, Val. xcviii, no. 391, July 1989
@ Oxford University Press 1989
the other side of that coin, he was not a logician (the Kneales); he confined human knowledge to a subjective prison (Russell); he was an Hegelian (Russell); he was preoccupied with the 'higher', disdaining mere science (Russell); his metaphysical views were erected at least in part on the basis of elementary confusions, such as between predication and identity (Russell) or entailment and material implication (Moore); he was opposed to relations because he just assumed the only possible propositional form to be subjectlpredicate (Russell); his attempted reductio of relations treats them illegitimately as objects (any superficial reader); he believed that all relations are internal (Moore, Russell), and what this means is that apparent relations are reducible to properties (Russell) and that no relation holds contingently (Moore); this view of relations was for him an 'axiom' (Russell) or 'dogma' (Moore), hence, apparently, unargued; his resulting monism and idealism led him to reject, unlike Russell and Moore themselves, the common-sense view of the world as containing numerous objects (Russell) and as containing objects at all (Moore); he simply failed to see obvious facts ( M ~ o r e ) . ~
3. A further element o f the Bradley stereotype As with many stereotypes, the Bradley stereotype is not wholly inaccurate. That it is, however, extremely misleading is already well documented in a 'A metaphysician': A. J. Ayer, Language, Truth and Logic, and edn., London, Gollancz, 1946, p. 36. 'No logician': W. C. and M. Kneale, The Development of Logic, Oxford, Clarendon Press, 1962. This is a work which, as its title suggests, is written under the dominance of the idea of progress. Kant is discussed only grudgingly; Bradley, Bosanquet, and Hegel are not even mentioned. Thus an important element of the stereotype is formed by omission. Anyone who knows even only the titles of some of their books should wonder why they are ignored. 'Idealism's subjective prison': B. Russell, M y Philosophical Development, London, Allen and Unwin, 1959, p. 62. 'An Hegelian': Russell, Our Knowledge of the External World, rev. edn., London, Allen and Unwin, 1926, pp. 48-9 The evidence for this claim about Russell is best found not in isolated passages but in the general tenor of most of his remarks about Bradley. 'The disdain of science': Russell, M y Philosophical Development, p. 62. 'The confusion of predication and identity': Russell, Our Knowledge of the External World, pp. 48-9; and 'The Philosophy of Logical Atomism', p. 245, in his Logic and Knowledge, London, Allen and Unwin, 1956. 'The confusion of entailment and material implication': G. E. Moore, 'External and Internal Relations', pp. 300-1, in his Philosophical Studies, London, Routledge and Kegan Paul, 1922. 'The belief that all propositions are of subjectlpredicate form': Russell, 'The Monistic Theory of Truth', p. 142, in his Philosophical Essays, London, Allen and Unwin, 1966; the remarks in question are reprinted in Ch. V of M y Philosophical Development. This is a constant Russellian refrain, appearing as early as 1903 in The Principles of Mathematics, 5 212 and continuing in the 1924 'Logical Atomism' (p. 324 of Logic and Knowledge). 'The internality of all relations': Moore, 'External and Internal Relations', op. cit., p. 276; Russell, 'Logical Atomism', op. cit., p. 335. 'The reducibility of relations to properties': Russell, The Principles of Mathematics, 2nd edn., London, Allen and Unwin, 1937, 5 212. 'No relation holds contingently': Moore, 'External and Internal Relations', op. cit., p. 308. 'The axiom of internal relations': Russell, 'The Monistic Theory of Truth', p. 141; some of the remarks in question are reprinted in Ch. V of M y Philosophical Development. 'The dogma of internal relations': Moore, 'External and Internal Relations', op. cit., p. 280. 'The common-sense belief that there are numerous objects': Russell, 'The Philosophy of Logical Atomism', op. cit., p. 1 7 8 Just how far Russell shared this common-sense belief can be gauged from p. 190 of the same work. 'The belief that there are objects of all': Moore, 'The Refutation of Idealism', p. 30, in Philosophical Studies. 'The failure to see
The 'Truth about F. H. Bradley
333
number of recent publications, and I shall not bother to go over this ground again. But I doubt whether these publications will have much immediate effect, since the resistance of stereotypes to change is well known. In this section, I want to demonstrate the resistance of this stereotype in the face of explicitly recognized and conclusive contrary evidence, by looking at one further element which I omitted from the list in the previous section. This element is the myth that Bradley held a coherence theory of truth. I t is time that the record on this matter was set straight. There is no published work of his in which a coherence theory of truth is ever expressed; and neither have I found it in his unpublished papers. T h e belief that Bradley had a coherence theory of truth is simply false. And I want to make clear the status of this assertion. I t is not a matter of interpretation of some texts over which people might reasonably differ. There is simply no argument about it. All one needs to do is to read the relevant pages without preconceptions. Yet the belief that he did have such a theory-indeed, that it is characteristic of his philosophy, and he a characteristic exponent of the theory-is one of the most widely propagated myths in the Bradley stereotype.3 Many of the writings I have just cited in footnote 3 are widely used student texts, and are thus particularly influential. Some of these texts are very recent, and will presumably continue to be influential. Some of the writings are ostensibly works of scholarship, whose authors should have known better. (Mea culpa.) T h e case of Haack is a particularly good illustration of the power of a stereotype: unusually, among writers of student texts, she actually went into what Bradley said, apparently even to the point of reading him, and certainly sufficiently to realize explicitly that he is not a coherentist at all. the obvious': Moore, 'The Refutation of Idealism', op. cit., p. 13; 'External and Internal Relations' op. cit., pp. 289, 306-7. On this tendency in Moore, see Simon Blackburn, 'A metaphysical muddle', The Times Ltterary Supplement, No. 4426, 29 Jan.-4 Feb. 1988. 'Bradley's coherence theory of truth': S . Blackburn, Spreading the Word, Oxford, Clarendon Press, 1984, p. 235, and 'Is Epistemology Coherent?', in Manser and Stock (eds), The Philosophy of F. H. Bradley, Oxford, Clarendon Press, 1984, pp. I 55-6; S. Candlish, 'The Status of Idealism in Bradley's Metaphysics', Idealistic Studies, 1981, p. 252, n. 6, 'Idealism and Bradley's Logic', Idealtstic Studies, 1982, p. 254, and 'Scepticism, Ideal Experiment, and Priorities in Bradley's Metaphysics', in Manser and Stock (eds), op. cit., p. 245; J. Dancy, A n Introduction to Contemporary Epistemology, Oxford, Blackwell, 1985, p. 110; S, Haack, Phtlosophy o f Logics, Cambridge, Cambridge University Press, 1978, p. 86; D . W. Hamlyn, The Theory of Knowledge, Basingstoke, Macmillan, 1970, p. 123; A. R. Manser, Bradley? Logtc, Oxford, Blackwell, 1983, p. 133; D. J. O'Connor and B. Carr, Introduction to the Theory of Knowledge, Brighton, Harvester, 1982, p. 166 (Bradley is here identified only by implication); J. Passmore, A Hundred Years of Philosophy, 1st edn., London, Duckworth, 1957, p. 116; T. J. Richards, The Language of Reason, Sydney, Pergamon Press, 1978, p. 139; R. Walker, 'Spinoza and the Coherence Theory of Truth', Mind, 1985, p. 2; A. R. White, 'Coherence Theory of Truth', in Edwards (ed.), The Encyclopedia of Philosoph,y, Basingstoke, Macmillan, 1967, p. 132, and Truth, Anchor, 1971, p. 109; R. Wollheim, F. H . Bradley, and edn., Harmondsworth, Penguin, 1969, p. 167; A. D . Woozley, Theory of Knowledge, London, Hutchinson, 1949, pp. 125 and 150.
334
Stewart Candlish
Yet she could not resist going on treating him as one (cf. her pp. 89 and 94). There are some notable exceptions. David Holdcroft is one, recognizing that Bradley only ever offers coherence as the (not a) test of truth; but he makes the mistake of taking him to have nothing to say on the nature of truth at all. Another is Nicholas Rescher; but he takes Bradley to endorse a correspondence theory of truth. T h e person who comes closest to recognizing the facts here is James Allard, who expresses the same view as Rescher but correctly restricts the correspondence theory's sway in Bradley to The Principles of Logic (though he claims Bradley to have worked with a coherence criterion of truth at this early stage, which seems to me exaggerated at best). Even Allard, however, has nothing to say about Bradley's later theory of truth, the one he held to for most of his philosophical ~ a r e e r . ~
4. The source of the myth As far as I can tell, the principal guilty party is once again Bertrand Russell. Even if this is not so, Russell is certainly a major offender; and reference to him at this point will enable me to forestall one obvious objection. That objection is this. If, as I have implicitly just admitted, Bradley does offer coherence as the test or criterion of truth, does he not thereby hold a coherence theory of truth? T h e answer is that Bradley distinguished between the nature and the test of truth. And it cannot be alleged in defence of those who have propagated the myth that they have done so from within a philosophical standpoint where no distinction is drawn between test and nature. Russell, for example, in The Problems of Philosophy, espouses correspondence as the definition of truth (p. 128 and index); earlier (p. 122) he has rejected coherence as the definition-but, he says, 'we may often use it as a criterion' (p. 140). A theory of the test of truth is certainly a theory about truth; but it is not what those who have attributed the coherence theory to Bradley have usually had in mind. Russell, with the contempt for niceties that accompanied the contempt he seems always to have felt for the views of 'Hegelians' after he defected from their ranks, is often content to ignore individual variations of view and to smear the whole group on the basis of a rejection of the arguments of one individual. And I would hazard that this is what happened with Bradley and the coherence theory. I n the 1906-7 volume of Proceedings of the Aristotelian Society, Russell attacked what he called the Monistic Theory of Truth, as expressed in Joachim's 1906 book The Nature of Truth. (The original unified paper appears in revised form as the last two David Holdcroft, 'Holism and Truth', in Manser and Stock (eds), op. cit.; Nicholas Rescher, The Coherence Theory of Truth, Oxford, Clarendon Press, 1973, p. 23;James Allard, 'Bradley's Principle of Sufficient Reason', in Manser and Stock (eds), op. cit.
The Truth about F. H. Bradley
335
items in Russell's 1910 collection, Philosophical Essays.) O n page 136 of that 1910 version (in its 1966 reprint) he makes clear that he identifies this monistic theory with the coherence theory (as Joachim himself seemed prepared to do), and gives his standard objections to that theory. While he does not explicitly and unequivocally attribute the theory to Bradley at that point, he claims (ibid., p. 140) that the monistic theory of truth is equivalent to the axiom of internal relations (which he notoriously attributes to Bradley, time and again), and he goes on to refer constantly to Bradley in illustrative footnotes. Part I1 of this paper (the part which contains nearly all the references to Bradley) Russell chose to reproduce in full in his widely read M y Philosophical Development; while in an early footnote in part I he explicitly asserts that he chooses Joachim's book for discussion because it is typical: 'the best recent statement of certain views which I wish to discuss'. It is impossible to read this piece without getting the impression that Russell takes Joachim to be speaking for Bradley (something Joachim himself denied he was doing), and that Bradley is his real target. This technique of smearing by association seems to have become a habit with Russell. I n The Analysis of Mind (1921), on pages 266-7, the coherence theory is attributed to Joachim and to 'Hegelians'; the same occurs on pages 133 and 272 (in the Pelican reprint) of the 1940 An Inquiry into Meaning and Truth; it is mentioned, again with reference to 'Hegelians', on pages 172-3 of the 1948 Human Knowledge: Its Scope and Limits. In view of Russell's-and most other philosophers'-classification of Bradley as an Hegelian (despite his own explicit and published protestations to the contrary as early as 1 8 8 ~ ~a) reader , could be forgiven for assuming, as he was no doubt intended to do, that Bradley was a coherentist.
5 . Bradley on correspondence and coherence In order to understand what Bradley's views on truth are, it is essential at the outset to realize that it is a subject on which he changed his mind, and that the doctrines, however tentative, of the first (1883) edition of The Principles of Logic are not those of his later books; nor are they those of the additional material added by Bradley for the second edition of 1922 (page references to this work will always be to the second edition; I shall indicate, where relevant, in which edition the cited text originates). T h e difference between the early and the later views is illustrated by the fact that in the rather thorough index to The Principles of Logic there is no reference at all to Coherence (though there is a gesture in this direction in Book 111, Part I, 5 26); while on page 197 (text belonging to the first F. H. Bradley, The Principles o f Logic, Oxford University Press, 1883, Preface.
336
Stewart Candlish
edition) he categorically denies the doctrine, so characteristic of his later period, of degrees of truth: 'There are no degrees of truth and falsehood.' (However, to be quite accurate, I should add that it is at least arguable that Bradley was really committed to the doctrine of degrees of truth in 1883 even if he did not himself realize it; and that his rejection which I have just quoted appears to be based on confusing degrees of truth with degrees of assertion, as he himself later realized (ibid., p. 236, text belonging to second edition).) What, then, was Bradley's early theory of truth? It was the correspondence theory; he called it the 'copy' theory. This emerges in one of the notes added to the second edition (p. 591, n. I): The attempt, made at times in this work for the sake of convenience . . . to identify reality with the series of facts, and truth with copying-was, I think, misjudged. It arose from my wish to limit the subject, and to avoid metaphysics, since, as is stated in the Preface, I was not prepared there to give a final answer.
It cannot be claimed that Bradley mistook his own earlier views. Here are some remarks from the text belonging to the original edition (pp. 41-2; Bk. I, Pt. 11, $ 1): A judgment, we assume naturally, says something about some fact or reality. If we asserted or denied about anything else, our judgment would seem to be a frivolous pretence. We not only must say something, but it must also be about something actual that we say it. For consider; a judgment must be true or false, and its truth or falsehood can not lie in itself. They involve a reference to a something beyond. And this, about which or of which we judge, if it is not fact, what else can it be? . . . The actual judgment asserts that S-P is forced on our minds by a reality x. And this reality, whatever it may be, is the subject of the judgment. It is the same with objectivity. If the connection S-P holds outside my judgment, it can hardly hold nowhere or in nothingness. It must surely be valid in relation to something, and that something must be real. No doubt, as before, S-P may not be true directly of this fact; but then that again was not what we asserted. The actual judgment affirms that S-P is in connection with x. And this once again is an assertion about fact. There is a natural presumption that truth, to be true, must be true of reality. And this result, that comes as soon as we reflect, will be the goal we shall attain in this chapter.
So there is an 'official' correspondence theory in The Principles of Logic, for the temporary purpose of suiting the presuppositions of logic. (He later, rather nai'vely, complained about having been taken to 'advocate' it (Essays ) well as the official theory, the later view on Truth a n d Reality, p. ~ o g n ) .As is also lurking there in embryonic form at Book 111, Part 11, $4 15 and 16 (pp. 589-90; text belonging to the first edition). (The later theory appears explicitly on p. 704; but this is in a Terminal Essay added by Bradley for the second edition.) This ambivalence on Bradley's part should not
The Truth about F. H. Bradley 337 surprise us, for it is just one facet of a general problem with that book: the problem, that is, of determining how successful Bradley was, or even could be, in his stated aim of keeping metaphysics out of a book on logic, and confining himself to working within what he took to be the presuppositions of a lower science while leaving aside the question of the ultimate justification of those presuppositions (cf., e.g., Bk. I, pp. x-xi, p. 28, p. 49 et passim). For 'metaphysics' here we may read, without dangerous inaccuracy, 'his real and final views'. There is no need for us to adopt the extreme view of Manser's Bradley's Logic, that there is none of the later metaphysics in the first edition of The Principles of Logic at all, to be able to acknowledge that Bradley was trying to keep it out: thus, for example, one can see occasional surfacings of idealism in the language. (Nor, however, should we fly to the other extreme, Cresswell's view that idealism is everywhere presupposed in Bradley's work^.^) So much for Bradley's early correspondence theory. But what about coherence? As I have said, there seems to be nothing to be found about coherence at the time of the early theory. All that remains is for me to illustrate what he says about coherence at the time at which he held his later theory of truth. I t would be tedious to cite every reference he makes; for the time being, until you can check for yourself, you will just have to take my word for it that the burden of the following passage, from a paper attacking the view of Russell and Stout that judgements about our immediate experience are infallible, is typical: Prof. Stout denies, I understand, that coherence will work as a test of truth in the case of facts due to sensible perception and memory. Mr. Russell again has taken the same line. . . . What I maintain is that in the case of facts of perception and memory the test which we do apply, and which we must apply, is that of system. I contend that this test works satisfactorily, and that no other test will work. And I argue in consequence that there are no judgements of sense which are in principle infallible. ('On Truth and Coherence', Essays on Truth and Reality, p. 202; the paper first appeared in Mind, 1909.)
Coherence, then, is on Bradley's later theory the test of truth. On the face of it, his view of coherence differs from that of Russell in The Problems of Philosophy only in virtue of Bradley's asserting that it is the sole test. But what is his theory of the nature of truth? At this point we come to the currently unrecognized contender among the usual rivals for the post of the correct theory of truth. Max Cresswell, 'Reality as Experience in F . H. Bradley', AustralasianJournal ofPhilosophy, 1977; and 'Bradley's Theory of Judgment', CanadianJournal ofPhilosophy, 1979. For the reasons why we should not take this view, see the first two papers listed under my name in n. 3.
338
6.
Stewart Candlzsh
Bradley's l a t e r theory of t r u t h
We saw in the last section that Bradley criticized his earlier work for its assumption of a correspondence theory; so it is unlikely that we shall be able to come up with the entertaining conclusion that his later theory of truth is almost indiscernible from Russell's. And in fact he argues explicitly against correspondence. T h e correct conclusion, however, while less amusing than that one would have been, is much more interesting. Bradley is clear enough here to be allowed to speak for himself: The division of reality from knowledge and of knowledge from truth must in any form be abandoned. And the only way of exit from the maze is to accept the remaining alternative. Our one hope lies in taking courage to embrace the result that reality is not outside truth. The identity of truth knowledge and reality, whatever difficulty that may bring, must be taken as necessary and fundamentaL7 ('On Truth and Copying', Essays on Truth and Reality, pp. I 12-13; the paper first appeared in Mind, 1907. Arguments against correspondence are in the same paper. Cf. also p. 343 of the same volume, from 'On Some Aspects of Truth', which first appeared in Mind, 1911. There are also expressions of the theory on pp. 150 and 153 of Appearance and Reality.)
Bradley's later theory of truth, then, is that truth is identical with reality. (I shall have a little to say later of the place of knowledge in this equation.) We may call this the identity theory of truth. But does it even deserve the title of 'theory', let alone to be taken as a serious contender in the same race as the correspondence theory? For so far it is little more than a slogan; and it looks like a mad slogan too. T o see that it does figure as a solution to some serious philosophical problems; that it is not after all so different from its competitors; that it might even be correct. At the very least we can see that no other theory could be consistent with Bradley's metaphysics. But all this demands further investigation.
'
Another expression of it in Bradley's later writings occurs in the chapter of Essays on Truth and Reality entitled 'On Some Aspects of Truth' (first published in Mtnd, I ~ I I ) ,but this is the most unequivocal statement of the theory that I know of. I am grateful to T o m Baldwin for bringing this passage to my attention, and for forcing me to see the significance of it. Indeed, I owe the central idea for the presentpaper entirely to him, through both conversation and his kindly letting me see some of his so far unpublished written work. While listing acknowledgements, I should add that earlier versions of this paper were read in Perth, Sydney, Canberra, and Auckland, and that the resulting discussions were responsible for a number of improvements, as were the written comments of James Allard.
The' Truth about F. H. Bradley
339
7. Intelligibility of the identity theory of truth One of the major impediments to understanding it is that we are nowadays almost automatically inclined to think of truth as a relation between language, or more generally, representation, and reality. And the idea that language and reality might be identical is hard to take seriously, whatever some philosophers may have thought. Thus Anthony Manser, in his Bradley's Logic (p. 133), talks of Bradley's 'attempt to make language and reality coincide'. But this late twentieth-century way of thinking is utterly foreign to Bradley, for whom the question was always one of the relation of thought, not of language, to reality. Reflection on this point can help us to make the statement of the theory into something a bit more than just a slogan, by facing up to the question, how can the slogan 'the identity of truth and reality' express a theory of truth at all? I t seems, at best, circular. Clearly Bradley means that truth is the identity of some x with reality; and the question is, what is the value of x here? It is, I think, fairly clear that this value is the truth-bearer, whatever that is in Bradley's philosophical logic. T h e closest one can get to identifying this bearer in a couple of words is to say that it is thought or thought in the form ofjudgement. It might be said that the theory as I have expressed it is rejected by Bradley himself in Chapter XV of Appearance and Reality, where (p. I 52) after a rhapsodic passage about the Absolute, he says: But if truth and fact are to be one, then in some such way thought must reach its consummation. But in that consummation thought has certainly been so transformed, that to go on calling it thought seems indefensible.
I think the answer to this suggestion is that Bradley's identity theory of thought and fact is no less aptly named than a more recent identity theory in the philosophy of mind. In both cases it was always intended that one side of the identity statement should outweigh the other, and perhaps the identity theory of truth is best compared with the so-called 'disappearance' version of the identity theory of the mental and the physical. (And we can, too, point out that, as Manser showed in Bradley's Logic, it is possible to extract points in logical theory from Bradley's work without going the whole metaphysical hog.) T h e suggestion that the value of x is thought is, so far, perhaps not much of an improvement on Manser's suggestion that it is language. But it is an improvement, and we have now perhaps sufficient motivation to investigate further; so I shall suspend the question of intelligibility for the time being, and move to a consideration of the philosophical problems which the identity theory is brought forward to solve. For this purpose we can stop worrying about the value of x, and just think in terms of the identity theory's most general form: that the truth-bearer is identical with the truth-maker.
340 Stewart Candlish
8. The identity theory as a solution Judgements, on the traditional understanding of them which Bradley accepted (with certain heavy qualifications which do not affect the issue here), involve ideas. Now there is a set of problems clustering around the notion of judgement which Bradley sums up succinctly in a passage near the start of The Principles of Logic: How then are ideas related to realities? They seemed the same, but they clearly are not so, and their difference threatens to become a discrepancy. A fact is individual, an idea is universal; a fact is substantial, an idea is adjectival; a fact is self-existent, an idea is symbolical. Is it not then manifest that ideas are not joined in the way in which facts are? Nay the essence of an idea, the more it is considered, is seen more and more to diverge from reality. And we are confronted by the conclusion that, so far as anything is true, it is not fact, and, so far as it is fact, it can never be true. (Bk. I, Ch. 11,s 3. The passage occurs in the context of an attempt to prove that judgements do not consist in syntheses of ideas; but it is clear that Bradley took the problems to be not restricted to this false view of judgement.)
I t is not difficult to provide, for sensibilities which style themselves as modern, a version of this passage in terms of language or representations. T h e suggestion is, then, that because of the inherent limitations of symbolism, it is impossible ever to have a true judgement. In the Principles, Bradley offers an account of judgement which, within the framework of assumptions he thinks necessary for logic, is designed to overcome this set of problems. But in Appearance and Reality (e.g. p. 147), where he argues that this framework is metaphysically inadequate, he notoriously suggests that truth is an unachievable ideal, a position which it is not easy to see as other than self-defeating: if the statement of the position is not itself true, why should we take any notice o i it? I shall return to this point. As well as these problems with judgement that are identified in The Principles of Logic, there is a problem with truth which Bradley identifies in the 1907 essay 'On Truth and Copying'. H e approaches the matter through criticism of the correspondence theory of truth. There is in this essay a certain verificationist strain, which emerges explicitly at one point: On the one hand we must not use words that have no positive sense, and, with this, all reality that falls outside experience and knowledge is, to my mind, excluded. ('On Truth and Copying', Essays on Truth and Reality, p. I 18)
This verification infects Bradley's discussion of the correspondence theory, and indeed helps to explain why he calls it the copy theory, since
The
t ruth about F. H. Bradley
341
he appears to assume that it is a theory of the genesis as well as of the nature of truth, according to which truth has to be obtained by a process of copying reality. Not all of his objections to the theory, however, depend upon this assumption or cannot be easily freed from it, and indeed several of them (these can be found in the first edition of The Principles of Logic as well) amount to a kind of roll-call of the kind of problems for the correspondence theory to which only the theory of truth-functions eventually provided a solution (if even then). In summary, Bradley's arguments against the correspondence theory are these. First, judgements about the past and the future cannot be the result of copying (ibid., p. 107). Second, the very facts whose copying is supposed to give us truth are themselves 'the imaginary creatures of false theory', whose apparent independent existence is merely the result of projecting on to the world the divisions imposed by thought, whereas if thought is to be capable of truth those divisions must exist independently of thought itself (ibid., p. 108). Thirdly, '[dlisjunctive, negative and hypothetical judgements cannot be taken as all false, and yet cannot fairly be made to conform to our one type of truth', and neither can '[ulniversal and abstract truths' (ibid., p. 109). But then Bradley does something rather surprising. H e goes on to discuss the pragmatic theory of truth, but suggests that at bottom, both theories commit the same error. This error is that of defending a 'truth which is external to knowledge' and a 'knowledge which is external to reality' (ibid., p. I I I). T h e argument that this is an error seems to turn on the claim that a vicious circularity is involved in the supposition, since for 'p' to be true, it must be true also that 'p' is a copy ofp, or that believing 'p' is advantageous; and for 'p' to be known to be true, it must also be known that 'p' is a copy of p or that believing it is advantageous. (And so ad injnitum.) His conclusion is that this fundamental difficulty can be overcome only by embracing '[tlhe identity of truth knowledge and reality' (ibid., p. I 13). T h e suppressed premiss of the argument is, of course, that there cannot be an internal relation between truth, knowledge, and reality, so that only identity will do. And this is a premiss which we may well wish to reject. But given that Bradley had argued independently against the reality of all relations, not merely external ones, and that the Tractatus Logico-Phzlosophicus had not yet been written, we can allow that in this context it was legitimate for him to assume such a premiss. T h e final derivation of the identity theory of truth is worth quoting in full: Truth, if is contended, is not to be the same as reality. Well, if so, I presume that there is a difference between them. And this difference, I understand, is not to be contained in the truth. But, if this is so, then clearly to my mind the truth must so far be defective. How, I ask, is the truth about reality to be less or more than reality without so far ceasing to be the truth? T h e only answer, so far as I see, is
342 Stewart Candlish this, that reality has something which is not a possible content of truth. But here arises forthwith the dilemma which ruined us before. If such an outstanding element is known, then so far we have knowledge and truth, while, if it is not known, then I do not know of it, and to me it is nothing. On the one hand to divide truth from knowledge seems impossible, and on the other hand to go beyond knowledge seems meaningless. And, if we are to advance, we must accept once for all the identification of truth with reality. ('On Truth and Copying', Essays on Truth and Reality, p. "3)
One of the most permanent features of Bradley's philosophy is his hostility to abstraction: the idea that we do not speak the truth if we say less than the situation we are talking about would justify our saying, just as we do not speak the truth if we say more, or something entirely different. Again, I do not intend to assess this idea. (Though I shall add that it is not as silly as it may sound; and that James Allard has demonstrated this in a couple of fine papers.8) T h e identity theory of truth is perhaps the culmination of Bradley's work of this idea. T h e difficulty is to see what it means, and how Bradley can show us that it is genuinely a solution of the problems which prompt it. I t is perhaps not too hard to see it as at least a formal resolution of the fundamental difficulty of the circularity allegedly involved in the notion of a truth which is external to reality. But it is far harder to see how it is any more than a striking restatement of the set of problems identified in The Principles of Logic, problems which may be summarized as the apparent inadequacy of thought to reality. There are, of course, various remarks (I have quoted one) in Bradley to the effect that with the attainment of truth, thought commits suicide as it passes over into the Absolute, and that all is well once the Absolute is reached; but what this amounts to is inexpressible because of the inherent inadequacy of thought and symbolism, and such remarks have, by and large, inflamed critics rather than pacified them. But it is not my purpose here to defend the identity theory of truth, so I shall pass on to another matter.
9. The identity theory, the coherence criterion, and degrees of truth In order to see how the identity theory of truth may be more than just an historical curiosity, it will help to ask how it is related to those characteristically Bradleian doctrines, that the criterion of truth is coherence and that truth has degrees. Surprisingly, the theory and the criterion are more independent than one might imagine. What motivates the coherence criterion is not the identity theory but Bradley's monism. This is made clear by Bradley's own
* James Allard, 'Bradley's Argument Against Correspondence', Idealistic Studies, 1980, and 'Bradley's Principle of Sufficient Reason', in Manser and Stock (eds), op. cit.
The ,Truth about F. H. Bradley 343 procedure in an essay from which I have already quoted, 'On Truth and Coherence' (Essays on Truth and Reality, pp. 202-18), where he defends the coherence criterion of truth. In that essay Bradley considers the denials of Russell and of Stout 'that coherence will work as a test of truth in the case of facts due to sensible perception and to memory'. And he proceeds by connecting the claim that there are infallible judgements of sensible perception and memory with the claim that there are independent facts (p. zo3), presumably on the basis that the truth of the latter claim is a necessary condition of the truth of the former. Rejecting as he does the mutual independence of facts, he rejects the possibility of infallibility of judgement, arguing that we must not confuse sensory data themselves, raw and unconceptualized, with judgements based upon those data, and that any judgement stands to be revised in the light of subsequent experience. Nowhere in this paper is any appeal made to the identity theory of truth. In the light of this degree of independence, we can note that it might, for example, be possible to hold a correspondence theory together with some form of monism more restricted than Bradley's (in that there would be an ultimate distinction between truth-bearer and truth-maker) and a coherence criterion. T h e problem would be to work out a way in which some features of the Absolute might mirror others; but as the Absolute is not an undifferentiated whole, this problem might be soluble. ( I leave it to those with an interest in the genuinely arcane.) Alternatively, one could hold to the identity theory while abandoning monism and the coherence criterion: thus a logical atomist could in principle hold that an utterly isolated judgement is true just in case it is identical with the utterly independent fact whose specification provides the judgement's content. What about the doctrine of degrees of truth? This arises in Bradley's thought in response to the difficulties entailed by his idea, which is already emergent in The Principles of Logic, that no judgement can ever be absolutely true. As we saw in 5 8, that idea looks to be self-defeating. T h e doctrine of degrees of truth offers an immediate avenue of appeal here, since it can be suggested that the claim that no judgement can ever be absolutely true is more true than its denial, and hence rationally to be preferred to its denial. Now the claim that no judgement can be absolutely true has two sources in Bradley. One of them is the views of judgement which we find him expressing as early as The Principles of Logic, essentially that judgement is ideal while reality is substantial, and hence that judgement is intrinsically unfitted for capturing reality. T h e other source is the monism which never appears in full-blooded form in the Principles, flowering instead in Appearance and Reality. Were the problem with attaining truth is that it involves completion of an infinite process, as we have to get the whole truth about Reality into our judgement. Now in theory these differing sources can be separated; and thus one might, in
344 Stewart Candlzsh
principle, be motivated to adopt the doctrine of the unattainability of trutfi, and defend it by appeal to the doctrine of degrees of truth, without being burdened by what one might think of as the extravagances of monism; nor need one take on the added burden of defending the coherence criterion. But even if I am wrong about the amount of independence there is among these several Bradleian doctrines, what does seem to emerge without much difficulty is that the notion of truth can be seen as a limit notion, a regulatory ideal of judgement which is no more to be deplored in principle than is, say, the notion of a substance's electrolytic properties at infinite dilution. And it can further be recognized that someone might hold the identity theory of truth without recognizing or admitting either of the sources of difficulty which led Bradley to suppose that complete truth was impossible of attainment. 10.
The influence of the identity theory of truth
It is widely recognized that Moore at one stage held to a theory in which 'the world is formed of concepts', the correspondence theory of truth is rejected, and a judgement's 'truth or falsehood cannot depend on its relation to anything else whatever, reality, for instance, or the world in space and time'.9 It is also recognized that Russell for a time held a similar view,'' before abandoning it for the theory so well-known that it is often, without qualification, just called his theory of judgement.'' But it is, I think, also often thought that this earlier theory is a sort of bizarre, even pathological, aberration whose outbreak was narrowly confined in both time and space. It is, however, clearly a version of the identity theory of truth.12 Russell seems to have got it from Moore. Whether Moore got it from Bradley, I do not know; it may no longer be possible to find out one way or the other. In any case, obviously the hold of this theory was not limited to the confines of turn-of-the-century Cambridge. But the influence of the identity theory, however mediated, is yet more widespread than this. It is lurking, disguised, in unexpected places, suggesting that the theory may be one of those philosophical views which are the extreme outcomes of apparently natural assumptions. Russell's well-known theory of judgement was first formulated in detail in 1906; probably its most widely read version is the revised one which G. E. Moore, 'The Nature of Judgment', Mind, 1899, pp. 182 and 192, respectively. Russell expresses the view in parts I1 and 111 of his 'Meinong's Theory of Complexes and Assumptions', Mind, 1904. a n 1957 For example, Geach in his Mental Acts, London, Routledge and ~ e ~ Paul, l 2 This was recognized by Passmore in Ch. 9 of A Hundred Years of Philosophy. But as was noted above, Passmore does not seem to have noticed the presence of the theory in Bradley. lo
The T ~ u t habout F. H. Bradley
345
appeared as 'On the Nature of Truth and Falsehood' in the 1910 Philosophical Essays. One of its distinguishing features is that it does not invoke ideas, because 'in this view [that "judgements consist of ideas"] ideas become a veil between us and outside things'; Russell thinks it 'fundamentally mistaken'.13 Instead, he regards judgements as containing the actual objects themselves, those things, for example Charles I and dying and the scaffold, about which we are judging. Some of the rudiments of this view, though of course the Theory of Descriptions has intervened in the meantime, were already present in The Principles of Mathematics (ยง SI), written, significantly, at a time when Russell had not yet repudiated the work of Meinong: T o have meaning, it seems to me, is a notion confusedly compounded of logical and psychological elements. Words all have meaning, in the simple sense that they are symbols which stand for something other than themselves. But a proposition, unless it happens to be linguistic, does not itself contain words: it contains the entities indicated by words. . . . T h e confusion is largely due, I believe, to the notion that words occur in propositions, which in turn is due to the notion that propositions are essentially mental and are to be identified with cognitions.
The notion that real things are constituents of judgements is, of course, a notion characteristic of the identity theory of truth. In its detailed version, Russell's theory of judgement is developed against the background of (what he took to be) the view of Meinong: that judgements 'have' unitary but complex objectives in virtue of which they are true or false. (I deliberately leave this vague, because the theory as Russell presents it for rejection is utterly incredible.) In effect, Russell is apparently reacting against the identity theory of truth. His grounds, in summary, are that it is impossible to believe in the existence of real objectives where a judgement (or proposition, as he now calls it) is false, and that this provides sufficient reason for not believing in them even where the judgement is true. In consequence, he opts for a correspondence theory of truth, in which a fact, to which a true judgement corresponds, is something quite independent of that judgement itself, and a replacement theory of judgement in which judgement is 'a multiple relation of the mind to the various other terms with which the judgment is concerned' (Philosophical Essays, 1966 reprinting, p. 15s). The full account of the theory, together with its version of the correspondence theory of truth, is given in the last couple of pages of the paper. I t is not unfairly summarized thus: when we judge that, say, A loves B, we have 'before the mind' the person A, the person B, and the relation of loving, in such a way that the relation is not present 'abstractly' but as proceeding from A to B. The judgement is true when l 3 B. Russell, 'Knowledge by Acquaintance and Knowledge by Description', Proceedings o f the Aristotelrun Society, 1910-11; the quotation is from p. 160 of the longer version contained in his Mysricism and Logic (1963 edn).
346
Stewart Candlzsh
there is a corresponding complex object, A's loving B, and false when there is not. What are we to make of this theory? T h e question is not so much whether it is correct; its difficulties are now well known. Rather, it is whether it is even a possible theory. We may take its expression, with its phrases like 'before the mind', as explanatory metaphor, pending the nonmetaphorical account. But the fundamental trouble with the theory on this interpretation is that no further non-metaphorical account is forthcoming (and indeed it is hard to see how it could be). So we might, instead, try to take it as the unvarnished presentation of the theory itself. But if we do that, we are faced with this problem: if the actual objects are before the mind in their actual relation, with that relation actually relating them and not merely figuring abstractly, then how can this combination differ from the actual fact which is being judged to obtain? Correspondence between judgement and fact is impossible on Russell's theory, for judgement and fact coalesce. A correspondence theory of true could fit Russell's theory only if the mind manipulated tokens of the objects of judgement, not the objects themselves (as in Wittgenstein's version). With this coalescence of judgement and fact which Russell's theory necessitates, there re-emerges the identity theory of truth from whose supposed rejection the new theory of judgement sprang. T h e relation's not figuring 'abstractly' might mean, not that it actually relates A and B, but merely that its 'sense' or direction must figure in the judgement. This would allow for asymmetrical relations while resisting the coalescence of judgement and fact. But if this is all that it means, Russell's theory is vulnerable to one of Bradley's own criticisms of it, that it cannot account for the unity of the proposition. Why did Russell himself not recognize this tendency within his own theory? One reason, I conjecture, is that his theory at that time included the mind itself as a substantial ingredient in the judgement, which precluded his identifying judgement and fact judged. Another is this: because of the realism to which he (tentatively) held before 1919, the influence of the identity theory requires that in judgement the mind puts actual physical objects into actual relations. But Russell's theory of judgement was not intended to attribute psychokinetic powers to the mind. How did he avoid psychokinesis? T h e answer is surely that the phrase 'before the mind' has to be interpreted as 'acquaintance', and that Russell's tentative realism was a representative realism in which the objects of acquaintance are sense-data and their properties and relations, not the objects of the physical world. This topic deserves a paper to itself, so here I shall just leave it with two brief observations. First, the great aim of Russell's theory of judgement, to account for the possibility of false propositions, is usually presented, by Russell and commentators, as being achieved by the replacement of a single relation of the judging mind to the
The &uth about F . H. Bradley 347 proposition by a multiple relation of the mind to the objects of judgement. But, in my view, the matter is more complicated than that. I t is vital for the possibility of false judgement that there be some distinction between the physical objects about which some judgement is made and the constituents of the judgement, and so vital to Russell's theory that it be a representative realism, that is, a theory which makes some concession to idealism. And we may still wonder about judgements which concern only mind-dependent objects, where the object of acquaintance is in Russell's view once more the judged object itself. Secondly, it can be seen now just how suited the identity theory of truth is to absolute idealism. Once the individual judging mind is lost, so is the possibility of the distinction between its ideas and reality. Coherence was not enough. For Bradley, there was no alternative. I I.
Philosophical archaeology?
Is the subject I have been dealing with one of merely antiquarian interest? I have suggested that it is not: that we risk misunderstanding ourselves and our discipline if, thinking of ourselves as up to the minute, we dismiss it as something for Keepers of Antiquities. I shall finish by adding to this suggestion. One of the sources of the identity theory of truth was, as we saw, the problem of how truth can ever be attained when judgement (language, symbolism) necessarily abstracts from its subject matter. This feature of language has been stressed quite recently: A signal (structure, event, state) carries the information that s is F i n dzgztal form if and only if the signal carries no additional information about s, no information that is not already nested in s's being F. I f the signal does carry additional information about s, information that is not nested in s's being F, then I shall say that the signal carries this information in analog form. When a signal carries the information that s is F in analog form, the signal always carries more specific, more determinate, information about s than that it is F . (Fred Dretske, Knowledge and the Flow of Information, Cambridge, Mass., 1981, P. 137)
Dretske illustrates the point by contrasting someone's saying 'The cup has coffee in it', which carries that information in digital form, with someone's showing a photograph of the same situation, which necessarily conveys more information, for example concerning the quantity of coffee or the shape of the cup, and carries the information in analogue form. I t is not hard to imagine how philosophers in the early twentieth century might have reacted to the assumptions Dretske appears to make in this passage: the assumption that some state of affairs (or whatnot) is
348
Stewart Candlzsh
isolable in the world; the helping oneself to semantics in order to identify this state; the apparent assumption that pictures determine their own method of projection, and that they are assertions. Nevertheless, these drawbacks are irrelevant to the fact that the analogueldigital distinction helps us to grasp what Bradley is after when he claims that judgement abstracts viciously from the concrete whole, falsifies the original, separates in language what is not separate in reality. It is clear that the problems to which the doctrines of and about truth offered by Bradley are intended to be solutions-particularly the doctrine of the infinite degrees of truth-are problems arising from the digitality of language in a sense close to Dretske's. This digitality, which Bradley stressed and Russell ignored, is essential to all natural languages. Only its removal can obviate any necessity for degrees of truth-to take a simple example, by having a symbol for all the shades of red which, unlike the word 'red' itself, varies systematically to suit every shade. Unless language were to be impossible to learn, the only such symbol suitable to the task would be the colour itself. But this suggestion is tantamount to the identity theory of truth. Thus it is not clear that Bradley would need the doctrine of degrees for a language that met his requirements. Recent work on the Sorites Paradox has involved attempts to solve it by hypothesizing what are, in effect, degrees of truth, that is, continuum-many truth values between o and I (though the direction of increase is the reverse of Bradley's). The current discussion just might benefit from an examination of its roots in Bradley's work.
Department of Philosophy The University of Western Australia Nedlands Western Australia 6009
STEWART CANDLISH