Whewell on Newton’s Rules of Philosophizing

  • Robert E. Butts
Part of the Boston Studies in the Philosophy of Science book series (BSPS, volume 155)


Throughout his scientific and philosophical career William Whewell had a deep concern for methodological issues. His own positive methodological theories were developed in the three editions of Philosophy of the inductive sciences.1 Also relatively well known is his exchange with Mill on a large range of topics touching on induction.2 But what is perhaps not so well known about Whewell’s interest in methodology is that it was a partly practical, not merely philosophical, interest. His search for methods showed an abiding concern for discovering the best ways of accumulating scientific knowledge; he wished to hand on to his followers a finished method no longer in need of philosophical justification. In addition, he wished to discover the method of teaching science and mathematics that would best further the actual pursuit of scientific knowledge. Both motives appear early in his writings, for example, in The mechanical Euclid of 1837.3 This does not mean that Whewell was unaware of the basic philosophical questions underlying acceptance of this or that method. His writings give ample indication that such questions in philosophy of science were of great importance to him.


Scientific System Inference Rule Inductive Inference Theoretical Entity Universal Gravitation 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Whewell, Philosophy of the inductive sciences, 1st ed., 2 vols. (London, 1840); 2nd ed., 2 vols. (London, 1847), identified here as PIS; 3rd ed. appeared as The history of scientific ideas, 2 Vol 15. (London, 1858), identified as HSI; Novum organon renovatum (London, 1858), identified as NOR; and On the philosophy of discovery (London, 1860), identified as PD.Google Scholar
  2. 2.
    Whewell, PD ch. XXII; Mill, 1843, Book II, chs. V, VI, Book III, ch. II.Google Scholar
  3. 3.
    Whewell, ME, Appendix “Remarks on mathematical reasoning and on the logic of induction”, pp. 143-182.Google Scholar
  4. 4.
    Whewell was active in university reform throughout his life at Cambridge. He was instrumental in introducing massive changes in the teaching of science and mathematics, and he introduced the first programme in moral science (philosophy). For details of his scientific and philosophical life, see Todhunter, 1876, 2 vols. His academic and personal life is discussed in Mrs. Stair Douglas, 1881.Google Scholar
  5. 5.
    Whewell, PD chs. XXII, XV, XVI, XXI.Google Scholar
  6. 6.
    Whewell, TSM. William Hamilton, review of TSM 1836a, pp. 409-455; Whewell, reply to Hamilton, 1836, pp. 270-72; Hamilton, reply to Whewell, 1836b, pp. 272-75.Google Scholar
  7. 7.
    Whewell, PD ch. XVIII. All numbers in parentheses in the text of the paper refer to pages in this chapter of PD.Google Scholar
  8. 8.
    Unfortunately, Whewell unfairly prejudices the case by construing “familiar” as “what is familiar to everyone prior to receipt of any scientific training.” The Master of Trinity College bristled at the suggestion that science took its start in, and accommodated itself to, ordinary experience. But clearly this view of science can by no stretch of philosophical polemic be applied to Newton.Google Scholar
  9. 9.
    The exchange between Whewell and De Morgan on the so called logic of induction is fascinating for what it brings out about Whewell’s own confusions. He invented a heuristic device which he called an “inductive table,” the point of which was to display formally the validity of inductive inferences. At the same time, however, he insisted that induction involved non-logical components. De Morgan bemoaned Whewell’s confused and stretched use of the term “logic” in this context, insisting that discovery and inference are two different things. In a letter Whewell replied, “... I do not wonder at your denying these devices [the inductive tables] a place in Logic; and you will think me heretical and profane, if I say, so much the worse for Logic” (Todhunter, above, n. 4, II, p. 417). I discuss these features of Whewell’s theory of induction at length in the Introduction (IV. Whewell’s theory of induction) to Butts, 1968.Google Scholar
  10. 10.
    This point seems heavily confirmed by passages in Cotes’ Preface to the second edition of the Principia (1819) in which he actually cites details of Newton’s metaphysics and the opposing metaphysics of the Cartesians. The following two passages will have to suffice as evidenceGoogle Scholar
  11. [Newtonians want]... to follow causes proved by phenomena, rather than causes only imagined, and not yet proved. The business of true philosophy is to derive the natures of things from causes truly existent....Google Scholar
  12. Therefore if it be made clear that the attraction of all bodies is a property actually existing in rerum natura, and if it be also shewn how the motions of the celestial bodies may be solved by that property it would be very impertinent for any one to object that these motions ought to be accounted for by vortices....Google Scholar
  13. Those rather have recourse to occult causes who set imaginary vortices, of a matter entirely fictitious and imperceptible to our senses, to direct these motions. (Cotes, 1819, pp. I, XXIV, XXIII).Google Scholar
  14. 11.
    Blake writes: “It may well be doubted... whether... [Whewell’s] interpretation does not exhibit more ingenuity in telling us what Newton should have meant by his statements, than success in informing us what he actually did have in mind when he wrote them” (1960, p. 134). Blake’s comment is fair, but it should be kept in mind that Whewell is not writing as a historian of philosophy of science he is writing as a philosopher deliberately endeavouring to extract illumination from Newton’s rules. This philosopher’s use of historic materials frequently infuriates historians; but it is a defensible kind of philosophizing and one at which Whewell was a master. Indeed as I am endeavouring to show in this paper, Whewell’s philosophical analysis of the rules does bring out the background of Newton’s understanding of them. Thus philosophy and history can converge on a single point, and my account may be taken as sample of just this phenomenon. Beyond this, it would take an extended essay in historiography to generalize from the sample.Google Scholar
  15. 12.
    In a number of works, Paul K. Feyerabend has exposed this feature of scientific systems, and has endeavoured to provide a philosophical solution of the problem of deciding between competing scientific systems. Whewell, it seems to me, anticipates a number of Feyerabend’s problems, and suggests some roughly similar lines of solution. In other respects, however, Whewell’s theory is a good example of the scientific conservatism that Feyerabend is combating. See Feyerabend (1965, pp. 145-260). See also Butts (1966, pp. 383-394).Google Scholar
  16. 13.
    See the Introduction to Butts (1968).Google Scholar
  17. 14.
    Whewell discusses the role of ideas and necessary truth in science in many of his works. The fullest discussions are in PIS, 1st and 2nd eds., pt. 1, and in HSI. See Butts (1965a, pp. 1-21); Butts (1965b, pp. 175-181); and Walsh (1962a, pp. 139-145).Google Scholar
  18. 15.
    One can probably make out a fair case for reading Newton Whewell’s way. However, a consideration of just what Newton (and Newtonians) meant by “phenomena” shows that Whewell’s view is greatly over-simplified. Indeed, Feyerabend has interpreted Newton’s phenomena (and the phrase “derived from the phenomena”) in a way that makes the Newtonian view almost compatible with Whewell’s own theory of induction. He writesGoogle Scholar
  19. Newton’s “phenomena,” which are the elements of the new “experience,” are not everyday facts pure and simple; nor are they an experience that has been cleared from prejudicial elements and left that way. They are rather an intimate synthesis of laws, possessing instances in the domain of the senses and certain mathematical ideas.... Actual experiment, which always depends on a large variety of irrelevant variables, may therefore illustrate the phenomenon; it cannot establish it.... Describing a phenomenon means stating a law. It is therefore not at all surprising that one can now obtain laws by a derivation “from the phenomena.” Nor is it surprising that many of the laws obtained are regarded as irrevocable. After all, they are based upon premises that are part of “experience” and that are therefore beyond reproach. (Feyerabend, above, n. 12, pp. 159-160)Google Scholar
  20. On this reading of Newton — which has much merit — Newton’s concept of phenomenon combines the two elements that Whewell insisted upon, ideas and sensations. The interpretation also implies that Newton’s system was committed to the explanation-confirmation symmetry that Whewell thought so important. See also Feyerabend (1970).Google Scholar
  21. 16.
    For an illuminating discussion of the difference between regarding a law (or a system of laws) as subject to further refinement, but not to further confirmation (or possible disconfirmation), see Kuhn (1965, chs. II–V). His concepts of “normal science” and “scientific paradigm” seem to me partially to capture what Whewell had in mind. Feyerabend interprets Rule IV in the same way as does Whewell (Feyerabend, 1970, secs. 11,12, nn. 9, 16, 33). Of course Whewell, unlike Feyerabend, agrees with the main features of the hypothetico-deductive method.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 1993

Authors and Affiliations

  • Robert E. Butts
    • 1
  1. 1.Department of PhilosophyThe University of Western OntarioCanada

Personalised recommendations