support your weight when you go up it. These are inferences from limited data to a general conclusion or a particular conclusion derived from one. The data remains limited even if it is all the data so far. Plainly data of the form Some A's have been or are cases of B' cannot entail All A's are B's'. What then are these inferences -- and how is it that they are justified? Ayer proceeds through this real and great problem throwing a clear light on every side and aspect of it. I know of no better exposition of it in the history of philosophy no better attempt to resolve it. It is at the level of the different attempt of Peter Strawson in his Introduction to Logical Theory (1952). It is superior to Karl Popper's supposed solution or dissolution, one of whose salient weaknesses is the idea that the logical problem of how we get to generalizations is somehow avoided if we take those generalizations as somehow only tentative, corroborated or in a plain way probable, rather than being proven truths despite the fact that we depend on them and indeed trust our lives to them in every minute of the day. However there is no more entailment between 'Some A's are B's' and 'That all A's are B's is corroborated than between 'Some A's are B's' and 'All A's are B's'. It needs adding that Popper's related idea that science actually aims not at the confirmation of generalizations but at their falsification is to say the least no more useful as pointed out by Ayer in passing in The Problem of Knowledge (pp. 73-4). The related subject of probability with its several competing theories is as well considered in Probability and Evidence. In this discussion as in the discussion of induction there is a greater care taken than in Language, Truth and Logic. Probability and Evidence is perhaps Ayer's most technical book. It readily repays the attention and persistence that its subject requires. This is as true of its concluding part on the troubled subject of such ordinary conditional statements as If the temperature of that water is raised another degree it will boil where the truth of the antecedent or if clause does not logically guarantee the conclusion. To come to a proper and full understanding of the truth of these instances of if P then Q' is to come to an understanding of the nature of scientific laws the nature of law like connections in the world. The Central Questions of Philosophy (1973) was published when Ayer was 63. It was put forward, partly, as an introduction to philosophy in the category where Russell's The Problems of Philosophy had reigned supreme. It takes up philosophical problems that the world and our existence in it force on us clearly not contrived or elusive or unformed problems. But this impressive book, the fourth of his finest ones is more than an introduction. It is an admirable retrospect of Ayer's own work indeed his own introduction to it. Yet more important, it takes matters forward. It returns to the concerns passionately written of in Language, Truth and Logic, including morality and religion it reconsiders or at any rate looks again at nearly all of Ayer's subsequent positions and supporting arguments. It therefore has to do with more subjects than have got into this introduction to his work. Some are closer to logic say the understanding of claims that two things are identical. Some are closer to descriptive metaphysics say the question of the way in which abstract entities like numbers can be said to exist. There is more on that crux of the philosophy of mind, the mind-brain relation. There are brief culminations of his thinking on the existence of real chance as against determinism, which oddly, like Peirce, he defended. There are thoughts on the meaning of life. Above all there is an advance in his thinking about reality, knowledge and perception. There is a unique departure from phenomenalism which does indeed have the disability of imprisoning each of us in a private world in favour of a realism if not naive realism. Ayer outlines a procedure of understanding whereby we begin not with private and subjective things but with neutral ones items that are not physical objects either. Through a procedure involving imagination rather tha deduction we come to what are called visuo tactual continuants, then to physical objects, and finally to private phenomena. Sense-data come last not first. He was proud of this line of argument rightly. Taking all of his career how great a philosopher was Ayer? You heard an opinion from me at the beginning of this introduction to his work. More needs saying. First some philosophical nativete needs to be put aside. It is possible to conjecture you will remember that Language, Truth and Logic raises and does not answer a very basic question about itself Is the Verification Principle a premise or a conclusion? You will remember too that the book itself does not provide a final statement of the principle. Are these dispiriting facts? Are they such as to put in question the achievement of Language, Truth and Logic?
One simple way to approach that question is by way of mighty Hume himself. Among his main contributions to philosophy one example was his analysis of causation. What it is at bottom for B to have been an effect of a he concluded as he is usually if not always understood is at bottom for all events like A to be followed by events like B. This like all of Hume's great contentions is at least disputed. Its enticing simplicity has not carried the day, and to my mind the contention can be disproved. The beginning of the disproof can be the fact that tonight is not the effect of today. On reflection, Hume's contention is also unclear in its idea of like events. We can also think of the equally great Kant and his doctrine of the unknowable noumenal world something nonetheless known to exist behind or under the ordinary world we experience. This is fundamental to much of his philosophy. It is certainly disputed, and certainly it calls out for clarification, indeed for rescuing from seeming self contradiction. Uses of the doctrine further for example in allowing for free will despite the determinism of the knowable phenomenal world are baffling we are tempted to say merely baffling.
The conclusion must be that great philosophy does not depend on being the discovery or defence of propositions made perfectly clear and known thereafter to be true. That conclusion also follows, of course, from the fact that pieces and whole traditions of great philosophy are plainly inconsistent. And yet philosophy does have to do with truth on large subjects. It is a kind of general logic aimed at truth. So the short story of its nature I think must be that great philosophy to some large extent consists in great mistakes greatly valuable mistakes enlightening and fertile mistakes in a progress towards truth. Ayer to my mind despite what can be said in objection to Language, Truth and Logic, played a large role in that progress by way of that book. Here is another fact relevant to the book's standing and that of its author. As you have heard, the book was published in 1936, republished after the war in 1946. After that there was little or no speculative metaphysics in the mainstream of English and American philosophy, whatever there was elsewhere. No doubt that report can only be made clear by moving towards and laying out a definition of speculative metaphysics. Merely to conjecture for a moment, a good definition will take speculative metaphysics to be speculation as to a transcendent reality probably speculation of an aspirational charactercertainly speculation that brings to mind the large metaphysicians of the past, mostly philosophical idealists. It will be at the other end of a scale from naturalisms. Speculative metaphysics will not include for example the analysis of conditional statements in terms of the mere assertion of possible worlds or a consideration of such fundamental ontological categories as those of particulars and propertiesor the idea that what it is to be perceptual aware of the room you are in is for the room in a way to exists. If it would be rash to the point of silliness to set out to explain the decline of speculative metaphysics only by Ayer's book it would also be rash to exclude it from the explanation or set of causes. The matter is not an easy one. It is hard to weigh causes hard to