Language, Truth and Logic. A. J. Ayer. Gollancz. Second edition, 1946.

THIS is a delightful book, to which I can give the sincerest praise possible, namely that I should like to have written it myself when young. Like Hume’s Treatise, it is enthusiastic, iconoclastic, hopeful, and good-humouredly pugnacious; fortunately, unlike Hume’s Treatise, it did not ‘fall still-born from the press’. The Introduction to the new edition displays admirable candour, and shows that Mr. Ayer has spent the intervening years to good purpose: what has been lost of early élan is more than compensated by the gain in maturity.

As a first approximation, the book has my sympathy, but taken as a final statement it seems to me in some points to take an over-simplified view, partly as a result of polemical zeal. If, in what follows, I dwell upon disagreements, I wish the reader to remember the background of general agreement, and not to forget my admiration of Mr. Ayer’s very exceptional clarity.

Let us tackle first a very vital question, that of the connection of verifiability with significance or meaning. If a sentence which is not a tautology is not to be ‘completely senseless’, Mr. Ayer demands ‘not indeed that it should be conclusively verifiable, but that some possible sense-experience should be relevant to the determination of its truth or falsehood’. One may question both the criterion, and the use that is made of it. To begin with the definition of ‘significance’ of sentences: I should myself define it syntactically. I should say that a sentence is significant if its words are understood and they are put together according to correct rules of syntax. Many sentences that obey this rule are not capable of verification—for instance, ‘there are distant nebulae receding from us with a velocity greater than that of light’ or ‘Napoleon was unhappy in St. Helena’. The question of the definition of ‘significance’, however, may be regarded as purely verbal; more substantial is the question of the interpretation of Mr. Ayer’s definition.

In what sense, for example, is physics verifiable? Mr. Ayer argues that all sentences which seem to mention unperceivable physical objects can be translated into sentences mentioning only Mr. Ayer’s experiences. I do not think either that this is true or that it follows from his criterion. The argument for the probable truth of physics is that, whenever its principles lead to a verifiable conclusion, the conclusion is verified if the necessary steps are taken. It is held (somewhat optimistically, I admit) that if the principles were false the verifiable conclusions would probably sometimes be false. There is, therefore, sense-experience relevant to the truth of physical principles, but that does not mean that the principles themselves can be interpreted in terms of sense-experience. We look for simple laws, and in order to make laws as simple as possible we do not hesitate to assume occurrences which we not only do not, but cannot, perceive. I do not believe it possible to state the laws of physics without assuming such occurrences.

Mr. Ayer speaks with approval of Mill’s definition of ‘matter’ as a ‘permanent possibility of sensation’. But what is a ‘possibility’? It can only mean something that would occur if something else occurred which in fact does not occur. But how are we to know what would have happened if ___? This does not substantially mean more than what does happen when ___. And I think this trouble about possibility affects the whole conception of verifiability as opposed to ‘actual verification: when something has not been verified, there is no clear meaning to the statement that it was verifiable. To say that it was verifiable presumably means that it would have been verified if something had been the case that in fact was not the case, and no one can decide how much that did not happen we may imagine to have happened in defining ‘verifiability’. Altogether, I should say, ‘possibility’ is a vague and dangerous notion, which should not be allowed to intrude into such a fundamental matter as the question when a sentence is significant.

Let us take a more mundane example than the principles of physics. Consider the statement ‘Napoleon III had a father’. We all believe this, though no one knows who he was, and very likely no one ever has known. We believe it because we believe in certain causal laws. Similarly we believe that our sensations have causes, though as a rule these causes lie outside our experience. I do not see why anybody should make a difficulty about such inferences, if they accept the inferences as to our own experiences that are obtained by means of them. Mr. Ayer has difficulties about the minds of other people, which seem to me quite unnecessary. When I believe that so-and-so has a toothache, I do not mean that if I were a dentist and were to examine his teeth I should see a cavity. And empiricism does not demand that I should mean this, if it is admitted that a hypothesis may be rendered probable by the truth of all its verified consequences.

The statement that the world existed before there was life, whether true or false, seems to me clearly significant; so does the statement that there is a future life, although Mr. Ayer maintains that to assert and to deny a future life are alike meaningless. In this he seems to me to go beyond even his own criterion. Verification is always in the future, and if there is a future life we shall, in due course, have empirical evidence of it. He might have said that the denial of a future life was meaningless, for if we do not survive death we shall never have any evidence of the fact. But if we do survive death we shall have the same kind of evidence as for the statement ‘it will rain tomorrow’, which also cannot be verified at present. It is true that Mr. Ayer bases his argument on the view that ‘the soul’ is meaningless, in which I agree with him. But he defines personal identity in terms not involving ‘the soul’, and, so defined, the question whether we survive death is surely not meaningless.

For my part, I am an unrepentant realist, in a sense which Mr. Ayer would condemn. I believe that the universe existed for countless ages before there were percipients; I believe that tables and chairs and other people are not functions of my perceptions. Since I can understand the sentence ‘A is before B’, I can understand the sentence: ‘There were occurrences before I was born’. And I see no reason to interpret this sentence in a Pickwickian sense. And when I say ‘you are hot’, I do not mean that I can see the sweat; I mean that you are having a feeling which I am not having, though of a kind with which I am familiar.

However, I have been led into being more critical than I feel. Mr. Ayer is wholly rational, and I do not doubt that, if arguments against him are valid, he will admit their validity. This is a very rare merit.

*  Bertrand Russell, Horizon (London), 15 (January 1947), 71-2