Wittgenstein was unreliable as an historian of philosophy. When he criticised other philosophers he rarely gave chapter and verse for his criticism, and on the rare occasions on which he quoted verbatim he did not always do justice to the authors quoted. I will illustrate this first in the comparatively unimportant case of Augustine and then in the more serious case of Frege.

Wittgenstein's Tractatus contains a wide range of profound insights into the nature of logic and language – insights which will survive the particular theories of the Tractatus and seem to me to mark definitive and unassailable landmarks in our understanding of some of the deepest questions of philosophy. And yet alongside these insights there is a theory of the nature of the relation between language and reality which appears both to be impossible to work out in detail in a way which is completely satisfactory, and to be bizarre and incredible. I am referring to the so-called logical atomism of the Tractatus. The main outlines of this theory at least are clear and familiar: there are elementary propositions which gain their sense from being models of possible states of affairs; such propositions are configurations of names of simple objects, signifying that those simples are analogously configured; every proposition has its sense through being analysable as a truth-functional compound of elementary propositions, thus deriving its sense from the sense of the elementary propositions when this view is taken in conjunction with the idea that the sense of a proposition is completely specified by specifying its truth-conditions. In this way the Tractatus incorporates in its working out a philosophical system analogous to the classical philosophical systems of Leibniz or Spinoza which are regarded by many people, in a sense rightly, as the prehistoric monsters of philosophy which are not to be studied as living organisms, but studied as the curiosities of human thought. And we may here agree that in the end we must simply reject a philosophy which incorporates such features as its postulation of simple eternal objects, or of a possibility of an analysis of a proposition which was presented as a pre-condition for the propositions that we ordinarily utter to make sense, and yet the specific form of which we are unaware of, and so on.

A fundamental notion of the Tractatus is that of the repetition of an operation. The operation specially mentioned is the simultaneous negation represented by the Sheffer stroke. ‘If an operation is applied repeatedly to its own results, I speak of successive applications of it. … In a similar sense I speak of successive applications of more than one operation to a number of propositions’ (5.2521).

I take as my text propostion 4.0312 of the Tractatus:

The possibility of propositions is based on the principle that objects have signs as their representatives.

My fundamental idea is that the ‘logical constants’ are not representatives; that there can be no representatives of the logic of facts.

Practically the same words occur (with two additional sentences) in Wittgenstein's Notebook for 25 December 1914, where Miss Anscombe translates them:

The possibility of the proposition is, of course, founded on the principle of signs as going proxy for objects.

Thus in the proposition something has something else as its proxy. But there is also the common cement.

My fundamental thought is that the logical constants are not proxies. That the logic of the fact cannot have anything as its proxy.

In the early years of this century the debate as to the nature of judgment was a central issue dividing British philosophers. What a philosopher said about judgment was not independent of what he said about perception, the distinction between the a priori and empirical, the distinction between external and internal relations, the nature of inference, truth, universals, language, the reality of the self and so on.

Tractatus, 5.62 famously says: ‘… what the solipsist means is quite correct; only it cannot be said but makes itself manifest. The world is my world: this is manifest in the fact that the limits of language (of that language which alone I understand) mean the limits of my world.’ The later part of this repeats what was said in summary at 5.6: ‘the limits of my language mean the limits of my world’. And the key to the problem ‘how much truth there is in solipsism’ has been provided by the reflections of TLP, 5.61.

Wittgenstein always thought that he had not been understood, and indeed that it was very unlikely that many people ever would understand him. Russell not only failed to understand Wittgenstein's later work; according to Wittgenstein himself, Russell profoundly failed to understand even the Tractatus. Professor Anscombe says even she did not understand him, and that to attempt to give an account of what he says is only to express one's own ordinariness or mediocrity or lack of complexity. Certainly, most people acquainted with the Tractatus, when that work was Wittgenstein's only published book, gave it what now seems a quite crass positivistic interpretation. Wittgenstein's own preface to the Tractatus, despite its last sentence, does not help. He does tell us that the whole sense of the work is that what can be said can be said clearly, and what we cannot talk about we must consign to silence: but this does not make it clear that what we cannot talk about is all that is really important. Even when one has realised all this, however, one is aware mostly of one's failure to understand; and that if one did get any distance in understanding the last sixth of the Tractatus, the process would be extremely difficult, and the results quite astonishing.

In the Michaelmas Term 1968 I gave a course of lectures on the Philosophical Investigations. Until then nobody had lectured at Cambridge specifically on that book, though it had been in print for fifteen years and must by that time have been lectured on in nearly every other philosophy department in the English-speaking world. One reason why we were so slow is suggested by a remark that John Wisdom made after hearing Max Black give a lecture on the Tractatus in the early fifties. As we came out of the lecture room he said to me ‘That was a strange experience. I have a clear memory of all that from my early years in Cambridge. And yet in some ways it was like hearing a lecture on Spinoza.’

The last part of Wittgenstein's Blue Book consists of a discussion of Solipsism. In the course of that discussion there occur several remarks (extending over about a page-and-a-half) which are explicitly concerned with the concept of a person and with the criteria of personal identity. This section is replaced in the Philosophical Investigations by half a sentence which reads: ‘… there is a great variety of criteria for personal “identity”’. Wittgenstein has italicised the word ‘identity’, and has placed it in inverted commas: I don't quite know why he does this, but it might be a hint to the effect that there is something slightly suspect about the notion of personal identity.

There is a passage in Wittgenstein's Philosophical Investigations in which he compares an answer that may be given to a philosophical question about someone else's pain with an answer that may be given to a question about the meaning of ‘It is 5 o'clock on the sun’. Wittgenstein does not compare other answers that may be given to the two questions. And he does not compare the questions themselves in respect of what lies behind them – making them ones which we can, or cannot, easily ‘see through’ – or in respect of how they should be answered. Yet there is material in what he says elsewhere in the Investigations and in other of his later writings for a manysided and, I think, useful development of the comparison. Anyway, that is what I shall attempt in this lecture.

It is sometimes said that a human being has a soul, whereas animals and lifeless things do not. The distinction made is of significance probably for most religions. Although it sets man apart and places him in a unique category, it should not be taken to imply that there is no difference between what is alive and has sentience, apart from man, and what is lifeless and unconscious. This was Descartes' error. For he ran together several distinctions and equated the soul with consciousness.

Wittgenstein refers to psychophysical parallelism in this apparently prejudiced way in paragraph 611 of Zettel, in the course of a rather remarkable passage. It begins at 605 with the claim that ‘One of the most dangerous ideas for a philosopher is, oddly enough, that we think with our heads or in our heads’. Subsequent sections develop this remark in a way that demonstrates Wittgenstein's rejection of the view that thinking is any sort of process in the head, whether a physiological process or a matter of the operations of ‘a nebulous mental entity’. Indeed he appears to consider that these ontologically opposed alternatives have a common source, in that they both derive from the mistaken view that there must be a mediating process between psychological phenomena such as my present remembering and my experience of the remembered event (cf. Z, 610). If we find no suitable mediating physiological process, we are easily led to assume that there must be a process of a rather different sort, and hence we are led to believe in a ‘nebulous mental entity’. But this whole line of thought in fact depends on a ‘primitive interpretation of our concepts’, an interpretation which we uncritically made at the stage at which we assumed that there must be a process of some sort mediating between the phenomena. We are reminded of Wittgenstein's earlier remarks in Philosophical Investigations, I, 308:

Speaking is so closely associated with making noises that such descriptions as ‘silent soliloquy’ and ‘soundless monologue’ have an air of paradox. Yet people frequently say things to themselves in such a way that not even a close observer has any reason to think they have done so. It is therefore tempting to suppose that on such occasions a sequence of surrogate speech sounds is produced in the person's head which he alone hears or introaudits, as if what distinguishes silent inner speech from normal speech is that the word substitutes are conveniently hidden from all save their producer.

Wittgenstein's book On Certainty which was first published in 1969, eighteen years after his death, is a collection of notes which he composed during the last eighteen months of his life. As his editors explain in their preface, these notes, which were written at four different periods, are all in the form of a first draft. They are more repetitive than they no doubt would have been if Wittgenstein had been able to revise them. Even so, they are characteristically succinct and penetrating, and the argument which they develop is easier to follow than that of the general run of Wittgenstein's later work.

I shall start by considering the apparently paradoxical doctrines that Wittgenstein put forward about knowledge: they show how the concept of knowledge is, as he says, ‘specialized’. This is not, as I shall show, a very important issue in itself, but it leads on to other points, of more interest: how it comes about, for example, that ‘not all corrections of our beliefs are on the same level’. I shall then discuss the idea that we inherit a certain picture of the world that forms the background of our experiments and researches. This idea, which is not of course unique to Wittgenstein, is, however, developed with many fresh insights. I end with some discussion of Wittgenstein's reported views on religious belief, which should not, in my opinion, be regarded as part of his contribution to philosophy, the interest of them being, perhaps, more biographical than philosophical.