The Great Philosophers: Ayer

Vs

OSWALD

O ^

HANFLING

Ayer

THE

GREAT PHILOSOPHERS

Consulting Editors Ray Monk and Frederic Raphael

Oswald Hanfling

A.J. AVER Analysing What We Mean

PHCENIX

A PHOENIX PAPERBACK First published in Great Britain in 1997 by Phoenix, a division of the Orion Publishing Group Ltd Orion House 5 Upper Saint Martin's Lane London, WC2H 9EA Copyright © Oswald Hanfling 1997 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted iii any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of the copyright owner. Oswald Hanfling has asserted his moral right to be identified as the author of this work. A catalogue reference is available from the British Library ISBN 0 753 80182 5 Typeset by Deltatype Ltd, Birkenhead, Merseyside Printed in Great Britain by Clays Ltd, St Ives pic

A.j. AVER Analysing What We Mean

'T^he trouble with philosophy, it is sometimes said, is JL that nothing ever gets settled. Questions posed by the ancient Greeks are still being asked today and there seems to be no agreement as to the right answers. Whether, or to what extent, these complaints are justified will depend on what the purpose of philosophy is taken to be. But one may feel that there is something radically wrong with the whole enterprise and that, if only one could put one's finger on the flaw, the subject could be transformed, or replaced by something more fruitful. Such a view was expressed by Ayer in the opening sentences of Language, Truth and Logic, together with a confident belief that he had found the flaw: The traditional disputes of philosophers are, for the most part, as unwarranted as they are unfruitful. The surest way to end them is to establish beyond question what should be the purpose and method of philosophical inquiry. And this is by no means so difficult a task as the history of philosophy would lead one to suppose. (LTL 45) What did Ayer mean by 'the traditional disputes of philosophers'? What is philosophy? According to at least one influential understanding, it is a quest for transcendent truths, beyond the reach of ordinary or scientific

inquiry. This conception of philosophy was one of the main targets of Ayer's critique. We may begin by criticizing the metaphysical thesis that philosophy affords us knowledge of a reality transcending the world of science and common sense ... One way of attacking a metaphysician who claimed to have [such] knowledge ... would be to inquire from what premises his propositions were deduced. Must he not begin, as other men do, with the evidence of his senses? ... [But] surely, from empirical premises nothing ... super-empirical can legitimately be inferred. (LTL45-6) 'The function of philosophy', he declared, 'is wholly critical'; 'it is an activity of analysis'; and the idea that philosophy is 'a search for first principles' was a 'superstition from which we are freed by the abandonment of metaphysics' (LTL 37, 62). With these and similar assertions the young Ayer embarked on a course of discussion that was designed to shake the philosophical^ establishment. As we shall see, and as Ayer would later be thefirstto admit, the book has many flaws; but few would deny that it was, and remains, a major contribution to philosophy. One may also be impressed, and indeed astonished, by its sheer virtuosity and the extent of its coverage - all the more so in view of the youthfulness of its author.

A 'SUCCES DE SCANDALE' /T A

began writing Language, Truth and Logicf, Ayer reported later, 'in the Christmas vacation of 1933-4

and finished writing it in July 1935, three and a half months before my twenty-fifth birthday ... The book enjoyed an immediate succes de scandale.'^ A new edition appeared in 1946, after which it 'approached the status of a best seller', with new impressions appearing 'almost annually for the next twenty-five years'.^ Ayer produced many other important books, but LTL remains the work by which he is best known. In a television interview of 1979, Ayer was asked what he now saw as the main defects of his youthful work. 'I suppose', he replied, 'the most important of the defects was that nearly all of it was false.' This, like some of the claims in the book itself, was an exaggeration. It is true that already in the second edition he conceded that 'the questions with which [the book] deals are not in all respects so simple as it makes them apear';^ but the general approach of that book remained with him throughout his life, as will be illustrated below. The main change is that the confident, and often over-confident, style of the early book is replaced by a cautious, painstaking investigation of issues he had earlier disposed of too quickly and easily. LTL will be the basis of the present

book, but from time to time we shall move forward to related discussions in later works. The writing of LTL came about in a rather accidental way. In 1931, Ayer tells us, he was given two terms leave of absence from his tutorship at Oxford. I proposed to spend them in Cambridge, leaming from Wittgenstein, but [my former tutor] Gilbert Ryle persuaded me to go to Vienna instead. He had met Moritz Schlick, the leader of the group of philosophers and scientists ... who entitled themselves the Vienna Circle, at some international congress ... and thought that it would be a good thing if I could discover what work was going forward under his auspices. For my part I had become engaged to be married and thought that Vienna would be a nice place in which to spend a honeymoon. I was married to Renee Lees on 25 November 1932 and we went almost immediately to Vienna.^ He rapidly learned enough German to enable him to attend the meetings of the Circle. To anyone acquainted with the philosophy of the Vienna Circle, its influence on Ayer will be quickly apparent. This is true both of its ideas, which included 'the elimination of metaphysics', and of the boldness of its style. But Ayer was also deeply influenced by the British empiricist tradition of Locke, Berkeley and Hume. (His appeal to 'empirical premises' and 'the evidence of the senses' is a sign of this.) There was, however, an important

difference between the new empiricism and that of previous philosophers: the new empiricism was about meaning rather than knowledge. The charge against 'the metaphysician' was not merely that his claims were unsupported by suitable premises: it was that they were meaningless, because they failed to satisfy certain conditions that must be satisfied if a statement is to have meaning. The new philosophy became known as 'logical empiricism' or, more commonly, 'logical positivism'. The first has the advantage of indicating the affinity with the empiricism of Locke, Berkeley and Hume, whose influence is apparent in Ayer's work; while the qualification 'logical' indicates the distinctive concern with logical analysis.

THE CRITERION OF VERIFIABILITY hat not all words or sentences are meaningful is

T

obvious if we consider such examples as the nonsense

rhymes of Edward Lear. But a sentence may be nonsens-

ical in less obvious ways. This is so, according to Ayer, in the case of claims about a super-empirical reality. But what criterion, if any, is there for distinguishing sense from nonsense? Ayer's reply to this question was one of the most prominent features of his book. He gave it as follows: The criterion which we use to test the genuineness of

apparent statements of fact is the criterion of verifiability. We say that a sentence is factually significant to any given person, if, and only if, he knows how to verify the proposition which it purports to express - that is, if he knows what observations would lead him, under certain conditions, to accept the proposition as being true, or reject it as being false. If, on the other hand, the putative proposition is of such a character that the assumption of its truth, or falsehood, is consistent with any assumption whatsoever concerning the nature of his future experience, then, as far as he is concerned, it is ... a mere pseudo-proposition. The sentence expressing it may be emotionally significant to him; but it is not literally significant. And with regard to questions the procedure is the same. We inquire in every case what observations would lead us to answer the question, one way or the other; and, if none can be discovered, we must conclude that the sentence under consideration does not, as far as we are concerned, express a genuine question, however strongly its grammatical appearance may suggest that it does. (LTL48)\ The criterion of vetifiability should be distinguished from the 'verification principle' of the Vienna Circle: 'The meaning of a statement is the method of its verification.' The criterion provides an answer to the question 'When is a statement meaningful?' or 'What kinds of statement are meaningful?', while the principle is a claim about what meaning consists in: it is an answer to the question 'What 8

is meaning?' The criterion is, however, dedudhle from the principle. If meaning is identified with method of verification, then it follows that if there is no method of verification - if the statement is not verifiable - then it must be devoid of meaning. There is, however, no entailment in the other direction, so that one may put forward the criterion without commitment to the principle. This was the case with Ayer, though his terminology is not always consistent (he sometimes referred to his criterion as 'the principle of verification'). The verification principle, as advocated by members of the Vienna Circle, calls for separate discussion, but this will not be our concern. The first question that may occur to one on reading the above passage from Ayer is why, or whether, one should accept his criterion. That there is a connection between meaning and verification would hardly be doubted. A good way of finding out what, if anything, a person means by a given statement is, indeed, to ask what 'would lead him ... to accept [it] as being true, or reject it as being false'. And if the answer is 'Nothing', then we might perhaps conclude that, as far as he is concerned, the statement is meaningless. But is this necessarily so? One of those who refused to accept the criterion was Father Copleston, who held certain religious beliefs that would hot satisfy it. In a debate broadcast in 1949, Ayer tried to persuade Copleston of the error of his position. He invented the word 'drogulus' for this purpose. Suppose I say, 'There's a drogulus over there' and you

say ... 'What's a drogulus?' Well, I say, 'I can't describe what a drogulus is, because it's not the sort of thing you can see or touch, it has no physical effects of any kind, but it's a disembodied being.' And you say, 'Well how am I to tell if it's there or not?' and I say, 'There's no way of telling. Everything's just the same if it's there or it's not there. But the fact is it's there. There's a drogulus there standing just behind you, spiritually behind you.' Does that make sense? (ML 41-2) But Copleston stood his ground against this barrage. Given that 'drogulus' means a disembodied spirit, he replied, he would say that the statement offered by Ayer was 'either true or false, whether one can verify it or not'. The artificial 'drogulus', taken in isolation, may indeed strike one as meaningless; but Copleston, in picking up the phrase 'disembodied spirit', was appealing to a wider discourse which may be thought to give it meaning. A similar point may be made about one of Ayer's examples of metaphysics in LTL: 'The Absolute enters into, but is itself incapable of, evolution and progress' (LTL 49). 'One cannot', he commented, 'conceive of an observation which would enable one to determine' whether this is true; and until the author of the statement 'makes us understand how the proposition that he wishes to express would be verified, he fails to communicate anything to us'. Now a reader confronted by that example might indeed regard it as meaningless. But would this be for the reason given by Ayer? Ayer had taken the example 'at random' from a work by F.H. Bradley, without any 10

reference to the arguments in which it was embedded; but perhaps it would not seem meaningless to someone who had read those arguments - even if it failed to satisfy Ayer's criterion. In 1946 there appeared a second edition of Ayer's book, with a substantial introduction in which he modified several of its main tenets. In the following extract he conceded that metaphysics cannot be disposed of as easily as he had thought; butfirsthe addressed the fundamental question about the status of his criterion: what reason is there for accepting it? Is it true to what we actually mean by 'meaning'? In putting forward the principle of verificatiorr as a criterion of meaning, I do not overlook the fact that the word 'meaning' is commonly used in a variety of senses, and I do not wish to deny that in some of these senses a statement may properly be said to be meaningful even though it is neither analytic nor empirically verifiable. I should, however, claim that there was at least one proper use of the word 'meaning' in which it would be incorrect to say that a statement was meaningful unless it satisfied the principle of verification; and I have, perhaps tendentiously, used the expression 'literal meaning' to distinguish this use from the others ... Furthermore, I suggest that it is only if it is literally meaningful, in this sense, that a statement can properly be said to be either true or false. Thus, while I wish the principle of verification itself to be regarded, not as an empirical hypothesis, 11

but as a definition, it is not supposed to be entirely arbitrary. It is indeed open to anyone to adopt a different criterion of meaning and so to produce an alternative definition which may very well correspond to one of the ways in which the word 'meaning' is commonly used. And if a statement satisfied such a criterion, there is, no doubt, some proper use of the word 'understanding' in which it would be capable of being understood. Nevertheless, I think that, unless it satisfied the principle of verification, it would not be capable of being understood in the sense in which either scientific hypotheses or common-sense statements are habitually understood. I confess, however, that it now seems to me unlikely that any metaphysidan would yield to a claim of this kind; and although I should still defend the use of the criterion of verifiability as a methodological principle, I realize that for the effective elimination of metaphysics it needs to be supported by detailed analyses of particular metaphysical arguments.
terion is put forward as a methodological principle, for such a principle cannot be described (any more than a suggestion or a proposal) as true or false; and neither, therefore, can it be subject to the test of verifiability. But this still leaves the question: should we accept Ayer's proposal? Is his 'methodological principle' a good one - one that we ought to endorse? Ayer tells us that he would still 'defend the use of the criterion', but he does not say what the defence would be. Perhaps it can be argued, in favour of the criterion, that it is useful to distinguish the two kinds of statement on the lines suggested, even if the description of one of them as 'meaningless' is withdrawn. But how exactly is the criterion to be formulated and applied? This question is not as straightforward as it may seem.

THE MEANING OF 'VERIFICATION' n formulating his criterion, Ayer had to face a difficulty

I

about statements whose significance no one, including

himself, would wish to deny. Consider, for example, the case of general propositions of law - such ... as 'arsenic is poisonous'; 'all men are mortal'; 'a body tends to expand when it is heated'. It is of the very nature of these propositions that their truth cannot be established with certainty by any finite series of observations. (LTL 50) 13

This is SO; according to Ayer, because of a discrepancy between the scope of such propositions and the observations on which they are based. The number of observations is, necessarily, finite; but the statements in question are not restricted to afinitenumber of instances. ('All men are mortal' does not refer only to people observed so far.) Whether it follows that such statements 'cannot be established with certainty' is debatable, but Ayer, like many others, took it to be so. But then, if to verify is to 'establish with certainty', it will follow that these statements, not being verifiable, must be declared 'not literally significant', along with those of metaphysics. To acconmiodate such statements, Ayer introduced a distinction between 'strong' and 'weak' verification. 'A proposition is ... verifiable in the strong sense of the term, if, and only if, its truth could be conclusively established ... But it is verifiable in the weak sense if it is possible for experience to render it probable' (LTL 50). He would settle for verification in the weak sense. 'It is only if a negative answer must be given' to the question of weak verification 'that we conclude that the statement under consideration is nonsensical' (LTL 52). To make his position clear, he introduced the term 'experiential proposition'. Let us call a proposition which records an actual or possible observation an experiential proposition. Then we may say that it is the mark of a genuine factual proposition, not that it should be equivalent to an experiential proposition, or any finite number of experiential propositions, but simply that some experiential 14

propositions can be deduced from it in conjunction with certain other premises without being deducible from those other premises alone. (LTL 52) The 'certain other premises' would presumably be about suitable circumstances. Thus if 'arsenic is poisonous' is true, it will follow that if one observes ('experiences') Smith taking arsenic, one will also observe Smith becoming ill or dying. And this would suffice to enable the general statement to pass the test. There is another important difference between this formulation of the test and that with which we began. What was required there was, as we saw, that the person making a statement should know what observations 'would lead him ... to accept [it] as being true, or reject it as being false'. This is not satisfactory, because it does not distinguish between observations that really support the statement and those that do not. All sorts of observations might 'lead' someone to accept or reject a putative statement, including the 'nonsensical' statements of metaphysics. In the formulation just quoted, however, the subjectivity of 'leading someone to accept' a proposition is replaced by the logical relation of 'being deducible'. But, we may ask, in what sense would the relevant 'experiential propositions' be deducible? If I am in the right place at the right time to observe someone taking arsenic and dying, it does not follow that I will observe this. It is indeed obvious that we often fail to observe what is there to be observed. Perhaps the criterion could be reworded to take account of this point. 15

There are other difficulties, however. Having introduced the requirement of 'weak' verifiability, Ayer commented that it 'seems liberal enough' (LTL 52) to accommodate laws of nature; but in his new introduction he had to concede (in view of an objection from Isaiah Berlin) that 'in fact it is far too liberal, since it allows meaning to any statement whatsoever' (LTL 15). To illustrate this point he invented a clearly nonsensical statement. The statements 'the Absolute is lazy' and 'If the Absolute Is lazy, this is white', jointly entail the observation-statement 'this is white', and since 'this is white' does not follow from either of these premises, taken by itself, both of them satisfy my criterion of meaning. (LTL 15) To avoid this embarrassing conclusion, he put forward an amended, rather complicated version of the criterion, but this will not be discussed here. It was, in any case, soon shown that the new version too was untenable. Further attempts were made by Ayer and others, but towards the end of his long career Ayer was still commenting on 'the continual failure of attempts to [formulate the criterion] in such a way as to find a middle ground between the over-strict requirement ... and the overindulgent licensing of gibberish'. He added, however, that he still did not want to 'discard the concept' of verifiability as a criterion of significance.^

16

VERIFICATION AND ANALYSIS

T

"^here is another, simpler way of objecting to the above formulation of Ayer's criterion, and this will take us

to one of the major themes of the book: analysis. Consider the simple conjunction The Absolute is lazy and

this is white', and suppose this were put forward as meaningful on the ground that it entails the observationstatement This is white'. Would Ayer have to concede that the statement as a whole is meaningful? It would seem that the proper way to deal with such examples is to separate the conjuncts, enabling the first to be described as nonsense and the second as meaningful. However, the required separation will not always be so easy. Take the case of statements about God, one of the major targets of Ayer's criterion. Suppose someone explained the existence of order and beauty in the world by reference to God's creation. Accommodating this to Ayer's formulation, he might claim that observation-statements to the effect that there is order and beauty in the world are deducible from the statement that the world was created by God. Would Ayer have to admit that this statement passes his test? We can see what the answer would be from one of his examples. His response was to separate out the meaningful part of the statement, as he saw it. Thus 'if the sentence "God exists" entails no more than that certain types of phenomena occur in certain se17

quences, then to assert the existence of a god' will mean no more than that - though, as he went on to observe, 'no religious man would admit that this was all he intended to assert' (LTL 152). Again, if a man tells me that the occurrence of thunder Is alone both necessary and sufficient to establish the truth of the proposition that 'Jehovah is angr/, I may conclude that, in his usage of words, the sentence 'Jehovah is angry' is equivalent to 'It is thundering'. (LTL 153-4) Here we see that the deducibility of observation-statements was not enough: what was needed was to analyse the statements under investigation, separating out what is meaningful from what is not. But whereas this seemed straightforward in the case of 'The Absolute is lazy and this white', it is not so in the case of statements about God, for of course the 'religious man', for one, would not accept the separations proposed by Ayer. The importance of analysis in Ayer's philosophy is more radical, however. Let us approach this by considering his choice of 'This is white' as a typical observation-statement. What did he mean by 'this'? He did not mean such things as tables and chairs: what he meant was a 'sensecontent' that the speaker was experiencing. The direct objects of observation, he held, are 'sense-contents' (or 'sense-data': he used various terms at different times) and not, as one might assume, such things as tables and chairs. Material things 'are logical constructions out of sense-contents' (LTL 86); and 'to say anything about [a 18

material thing] is always equivalent to saying something about [sense-contents] (LTL 185). What did Ayer mean by 'sense-contents'? In LTL, he gave little attention to this question. He introduced the term by reference to the 'ideas' of Locke and Berkeley; but since their use of this term is notoriously obscure and controversial, it hardly throws light on the difficulty. What is clear is that the term in question is to stand for something with which an observer is more directly related than he is with ordinary objects, so that his knowledge of the latter is either inferential (he infers that there is a table before him from the occurrence of suitable sense-contents) or - as in the passage quoted above - it is a matter of 'logical construction' (to say that there is a table before him is to say that sense-contents X, Y and Z are occurring or will occur). But Ayer did not explain what kinds of occurrences or entities X, Y and Z would be. In later writings he was more sensitive to the need for explaining and justifying his talk of sense-contents or sense-data. In The Problem of Knowledge he devoted a section to what he called 'the legitimacy of sense-data' (PK 105), in which he examined arguments for and against introducing such terminology, concluding that 'the argument from illusion' provides the strongest reason in favour of it. This argument proceeds from the premise that things sometimes seem to be otherwise than they really are (there seems to be an oasis before me, but it is only a mirage). It is then pointed out that the experience of seeing (say) a mirage is just like the experience of seeing 19

the real thing; and from this it is concluded that 'whenever anyone perceives, or thinks that he perceives, a physical object, it must then seem to him that he perceives something or other' (PK 104). The final step is tp 'pass from "it seems to me that I perceive x" to "I perceive a seeming-;^", with the implication that there is a seeming-;i^ which I perceive' (PK 105). It turns out, then, that the direct objects of perception - which in LTL he had called 'sense-contents' - are the 'seeming'-objects to which the argument from illusion appears to lead. One difficulty of this line of argument may be mentioned here. The claim that 'whenever anyone perceives ... a physical object, it must then seem to him that he perceives something or other' would not normally be recognized. On the contrary, 'it seems to him that he perceives a chair' would be held to imply that he does not really perceive a chair. Some argument is needed to defend Ayer's reversal of the logic of these expressions as normally understood. Let us turn to another aspect of the programme of LTL. The task that Ayer set himself was to show that all genuine statements could be analysed into statements about sense-contents. Whatever could not be so analysed must be consigned to some other category than that of genuine statement. (This, it turned out, was what the criterion of

verifiability really amounted to.) 'Our

remarks', he wrote, 'apply to all empirical propositions without exception, whether they are singular, or particular, or universal. Every [such] proposition is a rule for the 20

anticipation of future experience [of sense-contents]' (LTL 134). But why 'a riile for the anticipation', etc.? This brings us to phenomenalism - the view that apparently categorical statements about material things ('Here is a table', 'The table is white', etc.) are to be analysed in hypothetical terms ('If anyone looks this way, he will experience sensecontents X, Y and Z'; or 'If anyone were to look this way, he would', etc.). It is in this sense that the analysis into sense-contents must be understood: otherwise, it appears, it might have the absurd consequence of consigning ordinary ideas about tables and chairs to the class of metaphysical 'nonsense'. We normally think of such objects as enduring at times when they are not imder observation, but this belief cannot be tested by observation. It is here that phenomenalism, with its use of hypothetical statements, must come to the rescue. This explains how it is possible for a material thing to exist throughout a period when none of its elements are actually experienced: it is sufficient that they should be capable of being experienced - that is, that there should be a hypothetical fact to the effect that, if certain conditions were fulfilled, certain sense-contents ... would be experienced. (LTL 186) Statements about material things are to be analysed, accordingly, 'in terms of the hypothetical occurrence of sense-contents' (LTL 187). But^ we may wonder, is this really what we mean when we speak of tables and chairs? What, apart from avoiding 21

the allegations of 'metaphysics', is the advantage of representing the meaning of our statements in this way? According to Ayer, it 'serves to increase our understanding of the sentences in which we refer to material things' (LTL 91). He conceded that 'there is of course a sense in which we already understand such sentences ... as "This is a table", or "Pennies are round"', but, he argued, we may very well be unaware of the hidden logical complexity of such statements, which our analysis ... has just brought to light. And, as a result, [we] may be led to adopt some metaphysical belief, such as the belief in the existence of material substances or invisible substrata. (LTL 91 b) Here one might object that Ayer's concession - 'there is ... a sense in which we already understand' the statements in question - does not go far enough. We do not merely understand those statements: we regard them as having priority over statements about 'seeming'-objects, etc. Thus 'This is a seeming-table' would have to be explained by reference to 'This is a table' and not the other way round: the 'seeming' terminology is parasitic on the other. In a later work. The Central Questions of Philosophy, Ayer's views about phenomenalism had changed to a considerable extent. Here he defended the need for a vocabulary of 'sense-data', or something of that kind, in another way, by distinguishing between the 'appearances' of things and the things themselves - a distinction to which, he thought, a 'naive realist' could not do justice. 22

His argument in the following extract is not the 'argument from illusion'; it is an argument about the various and complex implications of such statements as 'Here is a table'. We think of the physical object as preserving its identity in the various guises in which it appears to us, but exactly how is this achieved? What is it that remains constant while its appearance varies? If the physical object is known to us only through its various appearances, in what way can we distinguish it from them? The naive realist ignores these questions ... He has no convenient vocabulary by means of which he can refer to the appearances of things, independently of the things of which we take them to be appearances. But if we want to discuss the relation of physical objects to their appearances, we do need such a vocabulary, and it is just this that the introduction of terms like 'sensible quality' or 'sense-datum' has been intended to provide. We may, indeed, not wish to be committed to all the implications which their use has been made to carry. Exactly how any such term is to be construed in order to be acceptable is a question which we shall have to examine. All that kam now suggesting is that something of this sort is needed ... We need only consider the range of the assumptions which our ordinary judgements of perception carry. To begin with, there are the assuroptions ... involved in characterizing anything as a physical object like a table. It has to be accessible to more than one sense and to 23

more than one observer and it has to be capable of existing unperceived. In addition, it has to occupy a position or series of positions in three-dimensional space and to endure throughout a period of time ... Neither is it only a question of the validity of these general assumptions. Our perceptual judgements are seldom indefinite, in the sense that we claim only to perceive a physical object of some sort or other. In the normal way, we identify it as a thing of some specific kind, and this brings in further assumptions, as, for example, that the object is solid, or flexible, or that it is not hollow. These further assumptions may relate to the purposes which the object serves, as when we identify something as a pen-knife, or a telephone: they may relate to its physical constitution, as in the identification of an object as an orange or an apple, which denies its being made of wax. They may presume on the deliverances of other senses, as when our descriptions of an object which we believe ourselves to be seeing or touching carries implications about the way in which it tastes or sounds or smells. But now can it seriously be maintained that all this can fall within the content of a single act of perception? Can my present view of the table, considered purely in itself as a fleeting visual experience, conceivably guarantee that I am seeing something that is also tangible, or visible to other observers? Can it guarantee even that I am seeing something which exists at any other time than this, let alone something that is made of such and such materials, or endowed with such and 24

such causal properties, or serving such and such a purpose? I think it evident that it cannot. But if these conclusions are not logically guaranteed by the content of my present visual experience, one is surely entitled to say that they go beyond it, and just this is what I take to be meant by saying that my judgement that this is a table embodies an inference. It embodies an inference, not in the sense that it results from any conscious process of reasoning, but just in the sense that it affirms more than can logically be entailed by any strict account of the experience on which it is based. What I mean here by a strict account Is one that is tailored to the experience, in that it describes the quality of what is sensibly presented/ without carrying any further implications of any sort. In the normal way, we do not formulate such propositions because we are interested not in the data as such, but in the interpretations which we have learned to put upon them. I cannot, however, see any logical reason why they should not be formulable. If I am right on this point, the naive realists are wrong in so far as they deny that our ordinary judgements of perception are susceptible of analysis, or deny that they embody inferences which can be made explicit. (CQ 79-81) It would indeed be wrong to deny that 'our ordinary judgements of perception ... embody inferences' of the kind mentioned, which, as Ayer makes clear, are various and extensive. But what are we to make of his claim that 25

those judgements 'are susceptible of analysis'? Are we to suppose that 'Here is a table' can be translated into a set of statements about the occurrence of sense-contents (or 'percepts') of the kind mentioned? If this is what he thought in LTL, it was no longer so now. I am not suggesting that physical objects are reducible to percepts, if this is taken to mean that all the statements that we make about physical objects, even at the common-sense level, can be adequately translated into statements which refer only to percepts. If . the demand for an adequate translation requires that the statements referring to percepts set out necessary and sufficient conditions for the truth of the statements about a physical object which they are meant to replace, I think it unlikely that it can be satisfied ... There is, on the one hand, the difficulty that a visuotactual continuant may be represented by an indefinite variety of percepts in an equal variety of contexts, so that if the percepts in question had not occurred, some others would have done as well, and, on the other hand, the objection that any description of a particular set of percepts will be bound to leave open at least the logical possibility that the observer is undergoing some illusion. But even if these difficulties could be met, there is another reason why I do not wish to adopt this position. The actual percepts that are presented to any observer, or even to the totality of observers at all times, are too scanty to answer to our conception of the physical world ... If the phenomenalist thesis is to 26

be at all plausible, it has to draw on possible as well as actual percepts, with the result that most of the propositions which render its account of the world will take the form of unfulfilled conditionals. They will state that if such and such conditions, which are not in fact realized, were to be so, then such and such percepts would occur. But apart from the obvious difficulty of giving a sufficient description of the conditions in purely sensory terms, I no longer think that such conditional statements are suitable to play this part. (CQ 106-7) With this abandonment of phenomenalism, what is left of the criterion of verifiability? 'The passage from percepts to physical objects', he now conceded, is not one 'of logical construction'. But how, in that case, are statements about physical objects to be understood? What is their meaning and how are they to be verified? 'The continued and distinct existence of physical objects' is, he concluded, 'simply posited'. He would now 'forsake phenomenalism for a sophisticated form of realism' (CQ 108). We may wonder what the younger Ayer would have said to this. Would he not have regarded the 'positing' of such objects as a clear example of metaphysics, 'devoid of factual content'?

27

STATEMENTS ABOUT THE PAST / / ^ ^ u r remarks apply to all empirical propositions without exception ... Every [such] proposition is a rule for the anticipation of future experience' of sensecontents (LTL 134). But what would he say about the difference between tenses? Are we to suppose that statements about the past are really about the future? This was indeed his view in LTL. 'Propositions referring to the past', he stated, 'have the same h3^othetical character as those which refer to the present and those which refer to the future' (LTL 134). Thus (to quote from a later work) 'all that could now be meant by saying that Caesar crossed the Rubicon was that if we were to look in such and such history books we should discover that their authors affirmed it' (CQ 24-5). He was aware that his view might seem strange, but, he assured the reader, for my own part I do not see anything paradoxical in the view that propositions about the past are rules for the prediction of those 'historical' experiences which are commonly said to verify them, and I do not see how else our 'knowledge of the past' is to be analysed. (LTL 135) He suspected those who objected to his analysis of being misled by an assumption that 'the past is somehow "objectively there" ..., that it is "real" in the meta28

physical sense of the terai' (LTL 135). But what makes the analysis objectionable, and indeed highly paradoxical, is not that the past is 'somehow objectively there': it is that it makes no allowance for the distinction between reality and report: between 'Caesar crossed the Rubicon' and 'Such and such history books say that Caesar crossed the Rubicon'. 'I do not see', he argued, 'how else our "knowledge of the past" is to be analysed.' Here, as elsewhere, the requirement of analysis is more stringent than the original criterion of verifiability. Asked how I would verify the statement about Caesar, or what empirical observation would 'lead me to accept it as being true', I could certainly answer by reference to a reputable history book, and in that way establish the credentials of the statement as meaningful in accordance with the original criterion. But, as we saw, that criterion had, in the face of certain difficulties, to give way to the requirement of analysis, which in its turn produced difficulties including that about the past. Here again Ayer showed himself more sensitive in the second edition. He no longer thought that statements about the past were really about experiences that would or might be had in the future, though he still maintained that such statements could 'be analysed in phenomenal terms'. The analysis, as now conceived, would be in terms of sense-contents that the speaker would have experienced if he had been present. A difficulty about this analysis was, he conceded, that 'these conditions never can be fulfilled; for they require of the observer that he 29

should occupy a temporal position that ex hypothesi he does not' (LTL 25). He did not, however, regard this difficulty as serious, since, he claimed, we could suppose that the observer occupied a different temporal position just as we can suppose that he occupies 'a different spatial position'. But is this true? 1 can suppose that 1 were now in Cambridge and not in Oxford, but can I suppose that I was present when Caesar crossed the Rubicon? Could someone who was present then have been meV Forty-odd years later Ayer was still trying to analyse statements about the past on the lines of LTL. Having reviewed a number of alternatives, he confessed that he found the position in which he found himself 'counterintuitive'. In the following passage he deals first with historical statements and then with personal memories about the recent past. It is difficult to believe that the statement that Wellington defeated Napoleon at the Battle of Waterloo entails the report of a [relevant contemporary] conversation, still less that it is equivalent to that report and a number of similar pieces of evidence. Should I say that my current statement that the leaves of the tree under which I am seated are rustling slightly in the wind will mean no more to me tomorrow than that I have such and such a memory, supported perhaps by a meteorological report?® The second of these examples involves difficulties of which Ayer seems not to have been aware. He assumes 30

that he will remember ('have a memory') of the leaves rustling. But it is quite likely that he will not. We all make thousands of more or less trivial observations that pass from our minds quite soon. In that case, must we say that they did not take place - or that it is meaningless to suppose that some did? On the other hand, suppose that Ayer does remember what happened the day before. In that case 'Ayer remembers that the leaves rustled' will be true. But is this a statement about a present or future experience? This is what Ayer must hold, in accordance with the phenomenalist analysis. (That is why he speaks, rather oddly, of 'having such and such a memory' rather than of remembering.) In LTL he had claimed that to say of a person that he remembers 'is to say merely that some of [his] senseexperiences . . . contain memory images which correspond to the sense-contents which have previously occurred in the sense-history' of that person (LTL 166). But one may remember that the leaves rustled without having such images; and conversely, one may have such images without remembering that the leaves rustled: one may be imagining this as opposed to remembering it. (A similar problem was raised, but not solved, by David Hume in the eighteenth century.) An alternative view of memory is that when we say 'Ayer remembers that the leaves rustled' we are not speaking of the present occurrence of images (or anything else); we are saying, rather, that (a) Ayer knows that the leaves rustled, and (b) he knows this because he heard them rustling at the time. 31

MYSELF AND OTHERS Just as I must define material things ... in terms of their empirical manifestations, so I must define other people in terms of their empirical manifestations - that is, in terms of the behaviour of their bodies, and ultimately in terms of sense-contents. (LTL 171) / T suppose', wrote Ayer, looking back on his career X towards the end of his life, 'that none of my philosophical preoccupations has given me so much trouble as the problem customarily ... described as that of our knowledge of other minds.'^ The problem is one that, unlike some of those discussed so far, has troubled many ordinary people from time to time. I regard other people as being conscious like myself, but can I really know that they are? What 1 observe are the physical manifestations of thoughts and feelings, but, it may be asked, what about the thoughts and feelings themselves? For Ayer the problem presented itself in a more acute form, owing to his reliance on 'sense-contents'. Having repeated that 'all empirical knowledge resolves itself on analysis into knowledge of sense-contents', he observed that a person's sense-contents are 'private to himself. But from these premises, he pointed out, it seems to follow that one is 32

logically obliged to be a solipsist - that is, to hold that no other people besides himself exist, or at any rate that there is no good reason to suppose [this]... For it follows from his prennises, so it will be argued, that the sense-experiences of another person cannot possibly form part of his own experience. (LTL 169) In his response, Ayer insisted that there is no fundamental difference between statements about material objects and statements about other people; and that, in each case, to suppose that there exists something beyond sense-contents would be to lapse into metaphysical nonsense, just as I must define material things ... in terms of their empirical manifestations, so I must define other people in terms of their empirical manifestations - that is, in terms of the behaviour of their bodies, and ultimately in terms of sense-contents. The assumption that 'behind' these sense-contents there are entities which are not even in principle accessible to my observation can have no more significance for me than the admittedly metaphysical assumption that such entities 'underlie' the sense-contents which constitute material things for me ... And thus I find that I have as good a reason to believe in the existence of other people as I have to believe in the existence of material things. For in each case my hypothesis is verified by the occurrence in my sense-history of the appropriate series of sense-contents. (LTL 171) 33

He hastened to asstire the reader that his analysis did not fail to do justice to the reality of other people. It must not be thought that this reduction of other people's experiences to one's own in any way involves a denial of their reality. Each of us must define the experiences of the others in terms of what he can at least in principle observe, but this does not mean that each of us must regard ail the others as so many robots. On the contrary, the distinction between a conscious man and an unconscious machine resolves itself into a distinction between different types of perceptible behaviour. The only ground I can have for asserting that an object which appears to be a conscious being is not really a conscious being, but only a dummy or a machine, is that it fails to satisfy one of the empirical tests by which the presence or absence of consciousness is determined. If I know that an object behaves in every way as a conscious being must, by definition, behave, then I know that it is really conscious... It appears, then, that the fact that a man's senseexperiences are private to himself, inasmuch as each of them contains an organic sense-content which belongs to his body and to no other, is perfectly compatible with his having good reason to believe in the existence of other men. For, if he is to avoid metaphysics, he must define the existence of other men in terms of the actual and hypothetical occurrence of certain sensecontents, and then the fact that the requisite sense34

contents do occur In his sense-history gives him a good reason for believing that there are other conscious beings besides himself. And thus we see that the philosophical problem of 'our knowledge of other people' is not the insoluble, and, indeed, fictitious, problem of establishing by argument the existence of entities which are altogether unobservable, but is simply the problem of indicating the way In which a certain type of hypothesis is empirically verified. (LTL 171-2) In the second edition he drew back from this confident dismissal of the problem. Though still inclined to the 'behaviouristic' analysis, he confessed: 'I own that it has an air of paradox which prevents me from being wholly confident that it is true' (LTL 26). The air of paradox is not hard to explain. What we are asked to believe is that when, say, I describe someone as having a toothache, I am speaking of his behaviour (i.e. of those of my sensecontents into which his behaviour is to be analysed) and nothing beyond that. The air of paradox becomes even more pronounced if we consider statements in the first person, as when I describe myself SLS having a toothache. Such statements are not made on the basis of sense-contents obtained from observing one's ovm behaviour. But what, in view of this, is the relation between the meanings of statements in the first person and statements in the third person? The difficulty was well described by Ayer in a later discussion. On the \^ew in question, he wrote. 35

the statements which I make about my feelings cannot have the same meaning for any other person as they have for me. [And therefore] if someone asks me whether I am in pain and I answer that I am, my reply, as / understand it, is not an answer to his question. For I am reporting the occurrence of a certain feeling; whereas, so far as he was concerned, his question could only have been a question about my physical condition. So also, if he says that my reply is false, he is not strictly contradicting me: for ail that he can be denying is that I exhibited the proper signs of pain, and this is not what I asserted. (PK 214-15) Turning to the problem again in his Central Questions of 1976, he favoured an answer that had been provided by Hilary Putnam, which he expounded as follows (beginning v^th a quotation from Putnam): 'Our acceptance of the proposition that others have mental states is both analogous and disanalogous to the acceptance of ordinary empirical theories on the basis of explanatory induction.' The main point of difference is that unlike empirical theories such as those put forward by scientists the theory that other people besides oneself have mental states is one that has no serious rival; in this respect it is like the theory that there are physical objects. What we establish inductively, on the basis of our knowledge of our mental states and our observation of other people's behaviour, is a set of special hypotheses about their mental states: the alternatives to these hypotheses are hypotheses 36

which account for the same behaviour in terms of different mental states rather than hypotheses which deny to others any mental life at all. (CQ 134) 'None of this'/ he conceded, 'puts the sceptic out of court.' If he is able to persuade himself that he has been cast into a world in which only he is conscious, there is no way in which he can be refuted. He will not even be at a disadvantage in accounting for the appearances, since he can believe that everything is and will continue to be as if other people had minds, though in reality they do not. (CQ 135) To this argument he had no answer except to say that 'this is not a theory which I myself find it necessary or useful to adopt'. A more radical difficulty, however, is about the meaning of the supposed 'theory' or 'set of special hypotheses', when these involve states of affairs that are, by definition, distinct from one's own sense-contents. According to Ayer, I can 'establish inductively' that other people have thoughts and feelings, including toothache. But what, given the constraints of verificationism, would 'toothache' mean in this context? The difficulty is similar to that which arose about our knowledge of physical objects as distinct from sense-contents. As we saw earlier, Ayer thought of these as being 'posited' by us, and in the passage just quoted he compares 'the theory that other people have mental states' with 'the theory that there are physical objects'. But in both cases he seems to have gone 37

beyond the bounds of meaningful discourse as he himself had defined it.

NECESSARY TRUTHS: MATHEMATICS AND LOGIC n objection commonly made against empiricism,

A

observed Ayer, is that it cannot 'account for our

knowledge of necessary truths' (LTL 96). Here was a category of statements which, though not empirical, could not be written off as metaphysical nonsense: they needed to be treated in a special way. He proposed to show that such statements are 'devoid of factual content' (LTL 105-6), since 'they merely call attention to linguistic usages' (LTL 106). He first made this claim about what he called 'analytic propositions'. If I say, 'Nothing can be coloured in different ways at the same time with respect to the same part of itself, I am not saying anything about the properties of any actual thing; but I am not talking nonsense. I am expressing an analytic proposition, which records our determination to call a colour expanse which differs in quality from a neighbouring colour expanse a different part of a given thing. In other words, I am simply calling attention to the implications of a certain linguistic usage. Similarly, in saying that if all Bretons are Frenchmen, and all frenchmen Europeans, then all 38

Bretons are Europeans, I am not describing any nnatter of fact. But I am showing that in the statement that all Bretons are Frechmen, and all Frenchmen Europeans, the further statement that all Bretons are Europeans is implicitly contained. And I am thereby indicating the convention which governs our usage of the words 'if and 'all'. We see, then, that there is a sense in which analytic propositions do give us new knowledge. They call attention to linguistic usages, of which we might otherwise not be conscious, and they reveal unsuspected implications in our assertions and beliefs. But we can see also that there is a sense in which they may be said to add nothing to our knowledge. For they tell us only what we may be said to know already. (LTL106) A similar treatment could be given, he thought, to mathematical statements. Our knowledge that no observation can ever confute the proposition '7 + 5 = 12' depends simply on the fact that the symbolic expression '7 + 5' is synonymous with '12', just as our knowledge that every oculist is an eye-doctor depends on the fact that the symbol 'eyedoctor' is synonymous with 'oculist'. And the same explanation holds good for every other a priori truth. (LTL113) Now the statement that an oculist is an eye-doctor may fairly be described as 'adding nothing to our knowledge'. 39

since the words are, indeed, synonymous. But can the same be said about mathematical equations? 7 + 5' and '12' are mathematically equivalent, but they cannot properly be described as 'synonymous'. Douze is synonymous with 'twelve', '7 + 5' is not. If '7 + 5 = 12' adds nothing to our knowledge, it is not because the expressions are synonymous, but because we are most unlikely to be ignorant of the truth of this statement. But how would this apply to less obvious mathematical truths? According to Ayer, his treatment of '7 + 5 = 12' would also hold good for such statements as '91 x 79 = 7189', even though here we have to 'resort to calculation ... to assure ourselves that [these expressions] are s)monymous' (LTL 114). In the case of such statements as this, however, it cannot even be claimed that they 'add nothing to our knowledge' (let alone that the expressions in them 'are synonymous'). While anyone acquainted with the meanings of 'oculist' and 'eye-doctor' must be aware that an oculist is an eye-doctor, it is quite possible for someone acquainted with the meanings of '91', 'x', '79', etc. to be ignorant of the product of these numbers. He may then learn what the answer is (thus 'adding to his knowledge'), either by being informed of it or by working it out for himself. Ayer was not unaware of this difficulty. He would need, he said, to 'explain how a proposition which is empty of all factual content can be true and useful and surprising' (LTL 97). He thought he could do so as follows. The power of logic and mathematics to surprise us 40

depends, like their usefulness, on the limitations of our reason. A being whose intellect was infinitely powerful would take no interest in logic and mathematics. For he would be able to see at a glance everything that his definitions implied, and, accordingly, could never learn anything from logical inference which he was not fully conscious of already. But our intellects are not of this order. It is only a minute proportion of the consequences of our definitions that we are able to detect at a glance. Even so simple a tautology as '91 x 79 = 7189' is beyond the scope of our immediate apprehension. (LTL 114) Now it is true that our intellects are limited in the way described by Ayer. But this seems to support the claim that he was concerned to deny: namely, that mathematical statements can add to our knowledge, and do not merely 'tell us only what we may be said to know already'. The treatment of logic and mathematics in LTL, like that of other topics I have discussed, continued to trouble Ayer in later years. His vacillation on the subject was recollected by Professor F.M. Quesada, who reported a conversation between them as follows: 'Seventeen years have elapsed since in 1951, Freddie, you attended the Lima Congress. Do you still think that logic and mathematics are what you thought at that time?' Ayer was silent for a few seconds ... Then, with slow and hesitating words, he replied: 'Well,... f guess I don't think now the same way 41

This conversation, held in 1968, was recalled by Quesada when writing his contribution to The Philosophy of A.J. Ayer in the late 1980s. Responding to that contribution, Ayer confessed: '1 have not yet wholly freed myself from [my] perplexity' on this subject." Nevertheless, having added some fresh arguments, he concluded: 'I adhere to the line that I took in Language, Truth and Logic all those years ago.'^^ 'A fairly competent mathematician', he observed, 'can tell at a glance that 37 + 37 = 74. He has to use pen and paper to discover that 37 x 37 = 1369.'^^ Yet, he argued, 'since multiplication is definable in terms of addition, it is absurd to put these propositions into separate categories'. But from the point of view of the programme of LTL, there is good reason for putting them into separate categories. The essential question there, as we saw, is whether mathematical statements can be informative whether they have 'factual content' or merely 'tell us what we already know'. According to the empiricism of Ayer and others, only empirical statements can be informative. But the informative capacity of '37 x 37 = 1369' is not put into doubt by the claim that multiplication is definable in terms of addition. (It may be worth pointing out that addition itself can be informative: this is so where the sums to be added are large or many.) A relevant distinction can be made by reference to knowledge of meanings. Let us take '2 + 2 = 4' as an example. Anyone who understands the meaning of this sentence must recognize that what it says is true. But this is not so when we turn to '37 x 37 = 1369': in thi§ case 42

knowledge of meaning does not entail knowledge that the statement is true, and that is why the statement can be informative. It is true that there is a way of arriving at this information which is different from that of arriving at empirical information, but this difference is not captured by Ayer's account. (I have preferred '2 + 2 = 4' to Ayer's '37 + 37 = 74', because recognition of the latter may depend, as he puts it, on being 'a fairly competent mathematician'.)

THE ANALYSIS OF 'STATEMENTS OF VALUE' / ' ^ h e r e is still', wrote Ayer in Chapter 6 of LTL, 'one A objection to be met before we can claim to have justified our view' - namely, that all genuine statements of fact are analysable into statements about sense-contents. 'It will be said that "statements of value" are genuine [statements], but that they cannot with any show of justice be represented' in this way (LTL 136). In response to this challenge, he admitted that 'ethical concepts' cannot be analysed in the approved way, but immediately explained this in a way which would leave intact his thesis about 'genuine statements of fact'. The concepts in question, he explained, 'are mere pseudoconcepts'. The presence of an ethical symbol in a proposition adds nothing to its factual content. Thus if I say to 43

someone, 'You acted wrongly in stealing that money/1 am not stating anything more than if I had simply said, 'You stole that money/ In adding that this action is wrong I am not making any further statement about it. I am simply evincing my moral disapproval of it It is as if I had said, 'You stole that money/ in a peculiar tone of horror, or written it with the addition of some special exclamation marks. The tone, or the exclamation marks, adds nothing to the literal meaning of the sentence. It merely serves to show that the expression of it is attended by certain feelings in the speaker. If now I generalize my previous statement and say, 'Stealing money is wrong,' I produce a sentence which has no factual meaning - that is, expresses no proposition which can be either true or false. It is as if I had written 'Stealing money!!' - where the shape and thickness of the exclamation marks show, by a suitable convention, that a special sort of moral disapproval is the feeling which is being expressed. It is clear that there is nothing said here which can be true or false. (LTL142) He considered a difficulty that had been raised by G.E. Moore. If ethical statements are neither true nor false, how is it possible that people disagree about them? He illustrated the difficulty as follows. If a man said that thrift was a virtue, and another replied that it was a vice, they would not ... be disputing with one another. One would be saying that he approved of thrift, and the other that he didn't; and 44

there is no reason why both these statements should not be true. (LTL 146) Ayer's response was to admit, and indeed insist, that 'it is impossible to dispute about questions of value'. What moral arguments are really about are always 'the facts of the case', and not the value that is to be put on them. When someone disagrees with us about the moral value of a certain action or type of action, we do admittedly resort to argument in order to win him over to our way of thinking. But we do not attempt to show by our arguments that he has the 'wrong' ethical feeling towards a situation whose nature he has correctly apprehended. What we attempt to show is that he is mistaken about the facts of the case. We argue that he has misconceived the agent's motive: or that he has misjudged the effects of the action, or its probable effects in view of the agent's knowledge; or that he has failed to take into account the special circumstances in which the agent was placed ... But if our opponent happens to have undergone a different process of moral 'conditioning' from ourselves, so that, even when he acknowledges all the facts> he still disagrees with us about the moral value of the actions under discussion, then we abandon the attempt to convince him by argument. (LTL 146-7) What would Ayer say about the endeavours of moral philosophers? Should their work be consigned to the waste-bin along with that of metaphysicians? No: there 45

was something to be done in moral philosophy, but it was not much. We find that ethical philosophy consists simply in saying that ethical concepts are pseudo-concepts and therefore unanalysable. The further task of describing the different feelings that the different ethical terms are used to express, and the different reactions that they customarily provoke, is a task for the psychologist. (LTL 148) Ayer-s account of moral statements (his talk of 'special exclamation marks' and 'a peculiar tone of horror', etc.) may strike readers as crude and perhaps even ridiculous; his position became known as 'the boo-hooray theory'. However, the view that moral statements are neither true nor false, that they are expressions of feeling and not statements of fact, has been widely held both inside and outside philosophy, both before and since Ayer. (That's a value-judgement' is sometimes thought sufficient to consign a statement to the realm of personal feeling as opposed to objective fact.) Now it is clear that moral statements are different from empirical ones; and if 'genuine factual statement' is defined to mean 'empirical', then it will follow that moral statements are not genuinely factual. But should this definition be accepted? Is it a suitable way of marking the difference between moral and empirical? In ordinary language we commonly speak of truth and knowledge with regard to moral statements. Thus a person who regrets what he did may say 'I knew it was wrong' or 'It's true that 46

I ought to have done something else'. And in a court of law it may be important to establish whether the defendant knew he was doing wrong. In this matter, as in others, Ayer was more sensitive in later writings to the dangers of departing from the ordinary uses of words. 'Certainly', he admitted in an essay of 1949, when someone characterizes an action by the use of an ethical predicate, it is quite good usage to say that he is thereby describing it; when someone wishes to assent to an ethical verdict, it is perfectly legitimate for him to say that it is true, or that it is a fact, just as, if he wished to dissent from it, it would be perfectly legitimate for him to say that it was false. We should know what he meant and we should not consider that he was using words in an unconventional way. What is unconventional, rather, is the usage of the philosopher who tells us that ethical statements are not really statements at all but something else, ejaculations perhaps or commands, and that they cannot be either true or false. (Ml 231-2) In view of these concessions, one might wonder why, as stated at the start of his essay, he 'still wished to hold' the view in question. In the following passage he gave his reasons for doing so. The issue, as he recognizes, is of wider interest and is, indeed, of fundamental importance for philosophical inquiry. When a philosopher asserts that something 'really' is 47

not what it really is, or 'really' is what it really is not, that we do not, for example, 'really' see chairs and tables, whereas there is a perfectly good and familiar sense in which we really do ... it should not always be assumed that he is merely making a mistake. Very often what he is doing, although he may not know it, is to recommend a new way of speaking, not just for amusement, but because he thinks that the old, the socially correct, way of speaking is logically misleading, or that his own proposal brings out certain points more clearly. Thus, in the present instance, it is no doubt correct to say that the moralist does make statements, and, what is more, statements of fact, statements of ethical fact ... But when one considers how these ethical statements are actually used, it may be found that they function so very differently from other, types of statement that it is advisable to put them into a separate category altogether; either to say that they are not to be counted as statements at all, or, if this proves inconvenient, at least to say that they do not express propositions, and consequently that there are no ethical facts ... It is merely a matter of laying down a usage of the words 'propositions' and 'fact', according to which only propositions express facts and ethical statements fall outside the class of propositions. This may seem to be an arbitrary procedure, but I hope to show that there are good reasons for adopting it. And once these reasons are admitted the purely verbal point is not of any great importance. If someone still wishes to say that ethical statements are statements of 48

fact, only it is a queer sort of fact, he is welcome to do so. So long as he accepts our grounds for saying that they are not statements of fact, it is simply a question of how widely or loosely we want to use the word 'fact'. My own view is that it is preferable so to use it as to exclude ethical judgements, but it must not be inferred from this that I am treating them with disrespect. The only relevant consideration is that of clarity. (MJ 232-3) The qualification 'although he may not know it', in the second sentence, should be noted. What Ayer is saying is that philosophers have sometimes failed to recognize the true nature of their claims. The philosopher teUs us, say, that we do not really see chairs and tables. Now this claim is not empirical: it is not as if further observation might reveal that we do, after all, see chairs and tables. The philosopher's arguments, whatever they are, are meant to rule out the logical possibility of our seeing chairs and tables. But this, according to Ayer, amounts to 'laying down a new usage of words'; so that statements that are possible, and indeed quite common, in the ordinary usage are not permissible in the new one. The philosopher's claim may look like a statement of fact ('we do not really see ...'), but what he is really doing is to 'recommend a new way of speaking'. It is in this sense that Ayer wishes his conclusion about moral statements to be understood. What we must consider, therefore, is not whether that conclusion is true, but whether the new way of speaking, under which 'true' 49

and 'know' are improper, would be beneficial. (The question is similar to that which arose about accepting the criterion of verifiability as a 'methodological principle'.) The implication at the end of the passage is that the new way of speaking would have the advantage of clarity; but is this really so? What is impprtant is to bring out the

differences between

moral

and empirical

statements (which Ayer proceeded to do by means of a detailed example). Thus one might draw attention to the difference between 'How do you know?' in the cases of 'He took the money' and 'He acted wrongly'. But it is not clear what could be achieved by introducing a new way of talking which, in any case, could not displace the existing one.

50

CONCLUSION A yer's work, as we have seen, is open to various XXobjections, but this does not detract from its importance. Ayer himself would not have been daunted by such objections and perhaps he could have answered them in ways 1 have not anticipated. But, in any case, he was himself, as we have seen, constantly questioning and revising his earlier ideas. Just as one must admire the bravado of his early book, so one must be impressed, when reading his later work, by his cautious and painstaking treatment of the questions at issue, and his constant striving to do justice to alternative views before arriving at his own conclusion. Ayer produced a vast amount of writing in his career, only a fraction of which - though an important fraction could be presented here. A bibliography of his writings, running to seventeen pages, includes books, articles and letters to newspapers on many topics, both inside and outside philosophy. His output, starting with a book review in 1930, continued steadily until his death, and in a sense even beyond that: the article 'What I saw when I was dead' appeared in the Sunday Telegraph in August 1988! This was written after a 'near death' experience, but his real death occurred some ten months later.

51

ABBREVIATIONS used to refer to works by Ayer

CQ

The Central Question of Philosophy (Penguin,

Harmondsworth, 1976). LTL

Language, Truth and Logic (Penguin, Harmondsworth,

1971). MJ 'On the analysis of moral judgements', in Essays (Macmillan, London, 1954). ML

Philosophical

The Meaning of Life and Other Essays (Weidenfeld and

Nicolson, London, 1990). PK

The Problem of Knowledge (Penguin, Harmondsworth,

1956).

52

NOTES I am grateful to Peter Hacker for comments on an earlier draft. 1. B. Gower (ed.), Logical Positivism (Groom, London, 1987), p. 23. 2. L.E. Hahn (ed.), The Philosophy of A.}. Ayer (Open Court, La Salle, Illinois, 1992), p. 18. 3. Ibid., p. 7. 4. Ibid., pp. 15-16. 5. The words 'sentence', 'proposition' and 'statement', appearing in various combinations in Ayer's writings, may be puzzling to readers not familiar with the literature. There are philosophical reasons for distinguishing between them, but I shall not go into these. On the whole I agree with Ayer's view, in a later book, that 'statement' will do for most purposes in expounding his views, and will follow this usage where appropriate. 6. Hahn (ed.). The Philosophy of A J. Ayer, p. 302. 7. Hahn (ed.). The Philosophy ofAJ. 8. Ibid., p. 598. 9. Ibid., p. 467. 10. Ibid., p. 478. 11. Ibid., p. 483. 12. Ibid., p. 481.

53

Ayer, pp. 601-2.

C:oN.sui;nN(;

KUITORS:

R A Y M O N K A N D FRK.DERK: R A P I I A E I

AJ. Ayer 1910 - 1989 Ayer is best remembered for Lmignage, Truth mid Logic (1936), which introduced British and American readers to the logical positivism of the Vienna circle. Hanfling shows in this introduction to Ayer's work how he turned this philosophy into a form of British empiricism in the tradition of H u m e . According to Ayer, philosophy is an activity oi malysis. Metaphysical truths can be neither established nor refuted by philosophical enquiry: they are meaningless. In support of this claim, he deployed his 'principle of verifiability'. But he found it difficult to refine the principle 'in such a way as to find a middle ground between [an] over-strict requirement' which would disqualify perfectly ordinary statements as meaningless, and 'the over-indulgent licensing of gibberish' - including that of metaphysics. Oswald Hantling recently retired f r o m his philosophy professorship at T h e O p e n University. His publications include Logical Positivism, Essential Readings in Logical Positivis7ii, Wittgenstein's Later Philosophy and The Quest for Meaning.

Cover painting": Swinging by VVassiiy Kandin.sky, courtesy of the T i t e Gallery