« PreviousContinue »
lhat Induction is simply the inverse method of Deduction. If Induction simply consists in framing hypotheses, deducing consequences from the hypotheses, and then comparing these consequences with individual facts for the purpose of verifying them by specific experience14, I grant that the procedure must, in most cases, be very untrustworthy. In my first Appended Note to my Section on Hypothesis, I have examined this account of Induction, which is virtually identical with that of Dr. Whewell. •In opposition to it, I maintain the following theses, which are explained and defended in the course of my work: i. That our inductions are not always preceded by hypotheses (and it might be added that even where they are, the hypothesis itself must rest originally on some basis of fact, that is to say, on some induction or other, however imperfect; for a hypothesis must always be suggested by something of which we have had experience); 2. That the mere verification of our hypotheses by specific experience is not sufficient to constitute a valid induction, unless the instances conform to the requirements of one of the inductive methods, or (as in the case of the fundamental laws of
"Vol. i. pp. 307, 308.
inductive reasoning) be coextensive with the whole experience of mankind. Induction, I maintain, may or may not employ hypothesis, but what is essential to it is the inference from the particular to the general, from the known to the unknown, and the nature of this inference it is impossible to represent adequately by reference to the forms of deduction.
Mr. Jevons' statement that 'induction is really the reverse process of deduction' I am wholly unable to reconcile with the following statements which occur in the very same page15: 'In its ultimate origin or foundation all knowledge is inductive,' and 'only when we possess such knowledge, in the form of 'general propositions and natural laws, can we usefully apply the reverse process of deduction to ascertain the exact information required at any moment.' When we compare these statements, the circle seems complete. A precedes B, and B precedes A. A depends for its validity on B, and B depends for its validity on A. No wonder that human reasoning affords us no 'certain' results!
In offering these criticisms on some fundamental points of difference between Mr. Jevons and myself,
15 Vol. i. p. 14.
I am far from denying the utility of many portions of his work, especially the chapters on the Methods of Measurement and on Hypothesis.
In the present Edition of this work, I have occasionally availed myself of the 'Inductive Logic' of Mr. Bain, a work which, though it does not, in my opinion, supersede Mr. Mill's Logic, supplies on some points a valuable complement to it.
In this, as in the last edition, I have to acknowledge the kindness of Professor Park of Belfast, whose corrections and suggestions have enabled me to make both my works more accurate and serviceable than they would otherwise have been.
*** It may be useful to the reader to be informed that the new matter introduced in the present edition occurs chiefly on pp. 7-9, 53, 218, 219 (Uniformities of Coexistence), pp. 200-202 (the Historical Method), pp. 217, 218 (the distinction between Inductio par Enumerationem Simplicem and the Method of Agreement), pp. 241-243 (the constant alternation in practice of the inductive and deductive processes), and pp. 292, 293 (the Argument from Universal Consent). Several minor alterations and additions have also been made.