The Collected Papers of Bertrand Russell. Vol. VII: Theory of Knowledge: The 1913 Manuscript 
edited by Elizabeth Ramsden Eames and Kenneth Blackwell.
Allen and Unwin, 258 pp., £35, May 1984, 0 04 920073 9
Show More
Freedom and Morality, and Other Essays 
by A.J. Ayer.
Oxford, 182 pp., £15, June 1984, 0 19 824731 1
Show More
More of My Life 
by A.J. Ayer.
Collins, 224 pp., £12.95, September 1984, 0 00 217003 5
Show More
Show More

The seventh volume of Russell’s Collected Papers contains the core of a book which he never completed. He stopped working on it, probably because he felt that he could not honestly go on. He had hoped that Wittgenstein would approve of what he had been writing, but when they met in May 1913, Wittgenstein told him that it was all wrong, and, as Russell admitted to Ottoline Morrell, he did not know how to answer Wittgenstein’s objections. He never published the whole of what he had written, and it is likely that what he did publish was the part that did not particularly interest Wittgenstein, while he suppressed the part on which the fire of Wittgenstein’s criticisms had been concentrated.

The confrontation must have been a scene of a kind that seldom occurs in the history of philosophy – brief and decisive. Russell had been looking forward to collaborating with Wittgenstein, who had arrived in Cambridge the year before in order to learn the philosophy of mathematics from him. It soon became clear that, if they worked together, it would be as equals rather than as tutor and pupil. When Russell showed Wittgenstein the book that he was writing, Wittgenstein produced for his inspection a manuscript of a very different character, Notes on Logic. Russell’s book is obviously a product of the English intellectual establishment and the line of his argument is smooth, flowing and confident, while Wittgenstein’s writing is more staccato and inspirational, a disclosure of thoughts shaped by a different tradition.

They disagreed about a fundamental problem in the theory of language. Many words get their meanings by tagging things in the way that names tag them. How, then, is it that some combinations of these words make sense while others do not? It is easy enough to produce examples of both kinds of combination, but quite another matter to produce a theory about them. The situation is typical of philosophy: we are confronted by an end-game problem and there do not seem to be enough pieces left on the board for its solution. To theorise is to make an explanatory move, but in this case it does not look as if there is anywhere to go.

Russell was a realist, and his instinctive response to this sort of difficulty was to invoke a new kind of thing. Now if he had been doing science, there would have been an agreed way of establishing the existence of a thing of a new kind. But it was not science and he had to appeal to the same realist instinct in us. Of course, we must be careful not to go too far in multiplying entities: our sense of reality must be what he called ‘robust’. But the real question here is whether our so-called sense of reality is anything more than a shared taste for a certain style of explanation.

What has to be explained is the evident fact that only certain combinations of words make sense. It is immediately plausible, but no real advance, to suggest that the reason is that the things tagged by them can be combined in corresponding ways in reality. The difficulty begins when we ask how such possibilities exist in the nature of things. Russell’s first step is to bring in the concept of structure. When a possibility is realised, we have an actual structure. That is simply a geometrical way of saying that we have a fact. His theory begins to gather momentum when he points out, quite correctly, that the possibility of the structure existed before it was realised. He then calls this possibility a ‘form’, and he claims that nature contains not only things tagged by words but also forms, which he saw as a set of stencils into which things can be slotted.

Because this is not science, we need to be told something about the character and status of forms and about the way in which we become aware of them. This was the point of divergence between Russell and Wittgenstein in 1913. Russell believed that our awareness of them is very like our awareness of the things whose structured combinations exemplify them and he used the same term in both cases, ‘acquaintance’. The only difference that he saw was that acquaintance with a form involved the knowledge that there was at least one actual structure exemplifying it. Wittgenstein took the opposite view: our awareness of forms is nothing like our awareness of the things combined in the structures that exemplify them. To be aware of a form is to understand a possibility, which is something dynamic, but the ultimate constituents of the world are logical atoms, things which have no structure and, therefore, cannot be understood but only tagged with names and identified.

It is true that Russell allowed that there is a difference between awareness of forms and awareness of things. But, according to Wittgenstein, he did so in a way that involved a further mistake. For suppose that awareness of a form really did involve knowledge of a truth. Then because awareness of a form is the understanding of a possibility, no possibility could be understood unless it had been realised and encountered as a fact. But that would put the cart before the horse, verification before understanding and truth before meaning. To use Wittgenstein’s own analogy, it would be like saying that someone could see that a portrait was a good likeness before he understood what counted as a good likeness.

The ideas of the past are like the occupants of Etruscan tombs, apt to turn to dust as soon as light and air are admitted. How could anyone accept the naive realism which frames this old controversy? If that is a typical chapter in the history of analytic philosophy, it must surely have reached its term by now and we ought to look elsewhere for ideas, perhaps across the Channel.

In fact, the vicissitudes of analytic philosophy since then make an intricate pattern, and, if ideas really were a matter for choice, it would not be a simple one. Wittgenstein went on to write the Tractatus Logico-Philosophicus, in which he developed a theory of language immune to the objections which, according to him, destroyed Russell’s theory: sentences are pictures which mirror the world’s possibilities. Like Russell, he was thinking only of factual sentences and of people only as knowledge-seekers. Apart from the new dynamic theory of possibilities, his most striking innovation was his so-called ‘mysticism’. There are many roads which can lead philosophers to a point where what they really want to say turns out to be strictly ineffable. Wittgenstein got there by reflecting on the exigencies of theorising: sentences cannot explain the way in which they themselves work, because they depend on it and the explanation of a phenomenon must be independent of it.

His first book was already very different in character from anything that Russell ever wrote, but a much greater divergence was to come later. The point from which the new development started was the practice of naming. In the Tractatus things were static and the dynamism of the world found expression only in the combinations of the names attached to them. True, the possibilities were already inherent in the things, but they were only possibilities of combination with other things. The things themselves were always conceived as minimum points in reality marked by names, like axes of rotation marked by dots on a blueprint. The change began when Wittgenstein brought human behaviour within the scope of the theory by treating the attachment of words to things as a part of human life.

Hilary Putnam has reviewed Ayer’s Philosophy in the 20th Century against this background in a recent issue of Partisan Review. The pattern is complex but the main lines are easily discernible. Russell is one point of origin for contemporary analytic philosophy and Wittgenstein is another, but mainly as the author of his much more influential posthumous book, Philosophical Investigations, in which he included human behaviour in the equation and developed his later ideas about the attachment of words to things. Ayer stands in the direct line of descent from Russell and in Freedom and Morality he cites him more often than any other philosopher. Russell’s influence on his theory of perception is particularly clear. We receive sense-data and recognise them without any help from other people or from circumstantial evidence. Solo recognition is the unanalysed starting-point and its roots in human life are kept out of the equation. This, of course, is regarded as a mistake by philosophers working in the opposed Wittgensteinian tradition. So we can picture the situation as a square with Russell and Wittgenstein at the top corners and their followers below them.

The disagreement between the two philosophers in 1913 was set in the frame of a realism on which they both agreed, but the contemporary controversy is more radical, because it goes beyond realism. Wittgenstein’s transcendence of realism in his later work is subtle and it is best understood if we approach it from a certain distance. If man is the measure of all things, then at the points where he applies his yardstick there may be no telling whether the way things match up to it is the way they really are. This is a general doubt, which in the long history of philosophy has often been raised about sense-perception. But there is also the possibility of a more specific doubt: perhaps our sense-organs do not even give us constant reactions to the things on which they are brought to bear. If so, they are intrinsically unreliable instruments of measurement and they need to be calibrated. This is science and individuals can check their senses against those of the majority, and the majority can test the constancy of their collective reactions by checking whether predictions based on them continue to come true.

Wittgenstein first transferred the specific doubt from perceiving to naming and then transformed it. The transference is a simple move: language measures the world and at the points where we apply it we need to maintain constancy in our linguistic reactions to things. We do this in the same two ways. Each of us can check whether he uses a word to tag the same things as other people and collectively we can check whether our agreed tagging yields predictions that continue to come true.

Wittgenstein’s transformation of this idea is less easy to interpret. He is not first indulging a doubt and then allaying it in the manner of Descartes. Rather, he is suggesting that man as the measurer belongs to the same system as everything that he measures. Within this system it is possible to check one thing against another, a scientific procedure. Philosophical understanding is altogether different, because it can be achieved only by attempting the impossible, failing, and then appreciating the inevitability both of the attempt and of the failure. The philosopher, unlike the scientist, tries to break out of the system and to ask about it, taken as a whole, questions that can only be asked and answered within it. The scientist makes limited flights and lands unscathed, but the philosopher is like Icarus contemplating his burnt wings on the way down.

These developments add a new dimension to the portrait of man as a knowledge-seeker. Since the Renaissance philosophers had treated his search as a purely contemplative enterprise detached from life as it is lived. When Wittgenstein looked again at the practice of applying words to things, he took a different view. He gave up the idea that the identity of a thing is simply given to its namer and argued that he had to work for it. He took the practice of applying words to things, which is the beginning of knowledge, and restored it to its place in human life.

We may now look back and ask whether anyone who accepts these ideas will necessarily regard the realism that Wittgenstein originally shared with Russell as a naivety. The case for this verdict is that it is simple-minded to suppose it possible to break out of the system of measuring subjects and measured objects in order to find out if things really are as after our best efforts inside it we say that they are. To put the point in another way, if anyone really did discover a new method of measurement, it would merely be an extension of science. The reason why we cannot burst through the skin of the bubble is simply that we carry it with us.

But was the earlier realism in fact based on the assumption that philosophy could go on an impossible trip into intellectual space? Russell never actually claimed to be able to view our whole system from a standpoint outside it. He simply did not ask himself exactly where he stood as a philosopher. Wittgenstein soon became conscious of the importance of this question and his answer to it was that philosophical ontologies are not really factual and strictly cannot be expressed in factual language. We may treat this ‘mysticism’ as an evasion, but we cannot charge Wittgenstein with a foolish mistake.

The question here is, of course, the general one: do our best efforts fail to capture reality? The sophisticated answer which many philosophers would give today is that they do not fail, but only because reality is sought and found within the system. Outside it realism is naive, but so too are all rival theories, because there is no platform on which a judge could stand to assess them.

The specific question, whether our linguistic reactions to the world are constant and reliable, is more interesting and more difficult. It is always risky to make predictions in the history of ideas, but it may well be that some of philosophy’s best points of growth are to be found in this area. Putnam thinks so and he criticises Ayer for passing them by.

In Ayer’s inheritance from Russell there are two things which had been handed down the long line of British Empiricists: preoccupation with man as a knowledge-seeker and concern about the credibility of the results of his search. Now anyone with that concern ought to be interested, not only in the immediate deliverances of the senses, but also in the subject’s immediate reactions to them. One of Putnam’s points against Ayer is that, like many empiricists, he focuses on the material delivered by the senses and neglects the psychology of the subject’s reactions to it. Naturally Ayer does not forget that it has to be recognised, but he turns his back on the idea that, if recognition were uprooted, removed from its familiar human context and based on isolated confrontation with one’s own sense-data, it might no longer be a possible achievement.

It is surely true that philosophy ought not to insulate itself from science. Indeed, it is obviously hazardous for the philosophy of perception to ignore the experimental results of cognitive psychology. It is, therefore, a credible prediction that future progress in the philosophy of mind will be integrated with advances in psychology. That will not, however, be a complete change of direction of the kind that Wittgenstein demanded from sense-datum theorists. His critique was more radical and it was developed within philosophy rather than on the borderline between philosophy and science.

Wittgenstein starts from the claim that the practice of applying a word to a series of things is possible only when there is a serviceable criterion for including new things in the series. Now Russell and Ayer maintain that a person can describe his own sense-data before he has established any connections between them and things in the world or between himself and other observers of those things. Wittgenstein’s objection is that no such language could possibly be set up, because there would be no serviceable criteria of identity for types of sense-data classified in such complete isolation. Genuine attachment of words to things is possible only as a part of human life in its natural setting.

In the second part of his autobiography Ayer says that he has never been convinced by Wittgenstein’s argument. This is not the place to attempt an adjudication. It would turn on the analysis of acquiring and preserving the ability to react in specific ways to different types of input. Is it enough that the subject should just feel that he is reacting correctly or does he need an independent criterion of correctness? Ayer is satisfied with the first alternative: Wittgenstein insists on the second one.

Putnam, like many contemporary philosophers, believes that Wittgenstein’s late work may have a humanising effect on analytic philosophy, and that it may lead to a liberalisation of ontology to correspond to other human interests besides the acquisition of knowledge. He probably does not mean that analytic philosophy will be superseded by something completely discontinuous. Even dinosaurs could claim to have escaped extinction if they really did evolve into birds.

The argument for Putnam’s prognosis starts from the fact, deplored by Heidegger, that Western philosophy has been preoccupied for too long with knowledge, its acquisition and its organisation, and so with sense-perception, science, logic and mathematics. These are, on the whole, cold subjects, detached from what gives human life its quality, and through them analytic philosophy arrives at an ontology that has little to do with human feelings and values. Russell himself wrote some eloquent descriptions of the denudation of this particular philosophical scene. Surely the pressure from outside philosophy will soon produce some re-afforestation within it?

Some distinctions are, however, needed at this point. One response to the pressure would be to try to get back to the age of innocence before the divorce of knowledge from feeling. That seems to have been Heidegger’s intention. Another would be to accept the divorce but to try for a richer ontology. A more radical response would be to give up the grading implicit in the construction of ontologies.

It is hardly likely that a reaction of the first kind will be promoted by Wittgenstein’s later work, which is an extended study of the diversity of different modes of discourse. True, he studies them in their place in human life as it is lived and he tries to show us why, when we withdraw into philosophy, we so often confuse them with one another. But his ultimate aim is to bring us back to an understanding of their idiosyncrasies. It is worth adding that Philosophical Investigations is a study which is itself expressed in the theoretical mode, and inevitably so, because no other mode is available. A consistent refusal to recognise and exploit the diversity of modes that are often combined in real life would reduce philosophy to silence, enigma or symbolic action. If anything, this would be like Wittgenstein’s early work, but the similarity would only be superficial, because what Wittgenstein says in the Tractatus is explicit and direct and not visibly affected by its strict ineffability.

Then will analytic philosophy try for a richer ontology? That too would hardly be in line with Wittgenstein’s later ideas. He had come to regard ontologies as crude expressions of preferences between the different conceptual systems that we spin between subject and object. If that is all that they are, it would be foolish to expect much to come of the inquiry whether human values are part of the furniture of the world or merely projections of human interests. The third option, abandonment of the grading implicit in the construction of ontologies, is the real tendency of Wittgenstein’s later thought. A philosophy which respects the idiosyncrasies of our different conceptual systems is unlikely to treat the word ‘real’ as an indivisible prize.

Is this the only future for analytic philosophy? Here, for what it is worth, is a reason for doubting whether the landscape will be quite so flat. One does not need to have read Montaigne or Descartes in order to realise that philosophy is not forced to choose between the two extremes, narrow favouritism and supine tolerance of all the manifestations of the human spirit. In some areas the consensus that is the mark of objectivity is lacking and in some it is faked. It is, therefore, unlikely that a philosophy which abandoned the idea of a single, indivisible prize. Reality, Being or whatever, would immediately abandon all grading. Even in areas in which there is genuine agreement some grading might be appropriate. The theoretical separation of fact from value need not produce the result deplored by Heidegger and Russell, but it might well reveal that the contribution made by man the measurer is more secure when the consensus is about the world that is our stage than it is when the consensus is about the drama that we ought to play out on it.

So how much would really have changed? The firmness and congruence of our reactions would have taken over the role formerly played by degrees of reality, and philosophy would be wearing the clothes of anthropology instead of those of physics. But if philosophy is not science, why would this be a change for the better? Because, if Wittgenstein is right, philosophy would not be bogus anthropology. Its observations about our speech and behaviour would have to be true, but they would also have a deeper point. They would show us why we want to escape from our own conceptual systems and why we cannot succeed: an oblique message conveyed by a factual study of their point of origin, ourselves.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences