[ Acknowledgements | Introduction | Part One | Part Two | Part Three | Appendix ]
For the sceptical view which he took of inductive inference, David Hume ever gave only one argument. That argument is the sole subject-matter of this book. My object is first to identify this argument, and then to evaluate it. The latter is what is attempted in Part Three (Chapters 5--9) of the book.
The main results of my evaluation can be summarized as follows. The sceptical conclusion of Hume's argument (I claim in Chapter 5 to show) is false. It rests (I claim in Chapter 6 to show) on a certain identifiable premiss which is false. Not all of my conclusions, however, are hostile to Hume's argument. Its true premisses (I claim in Chapter 7 to show) suffice to prove an important negative conclusion, though not a sceptical one, about inductive inferences. And what has historically been learnt from Hume's argument (I claim to show in Chapter 8) is of very great importance, even though it is partly quite opposite to what Hume intended to teach us.
But the evaluation of an argument requires, at least if it is to be of any interest either philosophical or historical, that the argument be first of all correctly identified. The detailed identification of what Hume's argument for inductive scepticism actually was, therefore, is attempted in Part Two (Chapters 2--4). These three chapters will, I think, be found the most contentious in the book; but the later, evaluative chapters nevertheless depend entirely for their interest on the accuracy of the account which I have given in Chapters 2--4 of what Hume's argument for inductive scepticism in fact was.
My identification of this argument (and in consequence my evaluation of it) involves identification of Hume's sceptical conclusion, as well as some of his premises, as being statements of logical probability. In Parts Two and Three, therefore, I needed to be able to assume a number of things about propositions of that kind. It seemed advisable to state these things all in one place, and preferable to do so before I made any mention of induction, or of Hume, or of inductive scepticism. This is what I have done in the only chapter of Part One. The remarks assembled there about statements of logical probability are not at all original, except in matters of emphasis. Their substance can be found in Carnap, or in Keynes, or in other writers in the Keynes-Carnap tradition. At least within that tradition, nothing in Chapter I is even controversial.
Hume's scepticism about induction is quite interesting enough, even considered in itself, to justify the present enquiry. But it is very doubtful, of course, whether in historical fact there exists any sceptical view of induction other than Hume's. (There could of course be other arguments for Humean inductive scepticism, beside the one that Hume himself gave for it). It is, at any rate, Humean scepticism, and no other, which has dominated subsequent philosophical reflection on induction, especially in the present century. Thus in 1921 J.M.Keynes could write, at the end of `Some Historical Notes on Induction', that `Hitherto Hume has been master, only to be refuted in the manner of Diogenes and Dr.Johnson' [1]. And in 1966 Professor Wesley Salmon still finds it natural, as indeed it is, to organize a book on The Foundations of Scientific Inference [2] in the form of a discussion of various attempted `answers' to Hume's inductive scepticism. Our inquiry has, therefore, by implication a much wider importance than would attach to a discussion of an argument of Hume, considered simply as such.
Still, it is no more than a historical fact (supposing it to be a fact) that inductive scepticism has been coextensive with Hume's inductive scepticism. The two are certainly not necessarily identical, and I wish to emphasize that it is the latter, not the former, which is the subject of this book. This limitation needs emphasis, because otherwise some of the claims made in the book will apt to appear more general, and more positive, than they really are. This is especially true of the argument advanced in Chapter 5 section (iii), and I ask the reader to observe that what I claim for that argument is not, for example, that it is a justification of induction; but just that it is a refutation of Hume's scepticism about induction.
[1] A Treatise on Probability (London 1921), p.273. (Referred to hereafter as Treatise, except where the full title is needed in order to prevent confusion of Keynes's book with Hume's Treatise).
[2] University of Pittsburgh Press, 1966.
[ Previous: Acknowledgements | Next: Part One | Table of Contents ]