The Plausibility of Theories

Citation preview

The Plausibility of Theories Stephen Toulmin The Journal of Philosophy, Vol. 63, No. 20, American Philosophical Association Eastern Division Sixty-Third Annual Meeting. (Oct. 27, 1966), pp. 624-627. Stable URL: http://links.jstor.org/sici?sici=0022-362X%2819661027%2963%3A20%3C624%3ATPOT%3E2.0.CO%3B2-2 The Journal of Philosophy is currently published by Journal of Philosophy, Inc..

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/jphil.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri May 18 09:08:07 2007

624

T H E JOURNAL OF PHILOSOPHY

tion of P that establishes a rational connection (e.g., a group of good reasons) between the existing theory, observations, problems, etc. in a domain of science at any stage of its history, and the degree of plausibility of a new hypothesis proposed in that domain; and (b) an interpretation of the history of science such that, given P, it is legitimate to say of a new hypothesis on every occasion (or on some occasions) not only that it was introduced at a particular stage, but also that it ought (or ought not) to have been introduced then; or, of a hypothesis that was not introduced at a particular stage, that it ought (or ought not) to have been. If I am right about these presuppositions, there are formidable difficulties in the way of making the evaluations proposed. At very best, it seems to me, we might hope to establish a rational connection between the state of a science at a particular historical stage and the plausibility of adopting certain kinds of hypotheses. But in view of the enormous number of eligible alternatives within each kind, I doubt whether the rational connection could ever be made "tight" enough to allow a degree of plausibility to be estimated for the adopting of a specific hypothesis within one of the kinds. Thus it is making too strong a claim to say that Liebig's specific hypothesis "had in its inception a degree of plausibility." The most that should be claimed is that the hypothesis of a vital force was generally plausible at that stage in the history of biology. A more restricted claim would be that the hypothesis of a vital force ought not to have been ruled out a priori because of being "metaphysical" or "nonempirical. " THOMAS A . GOUDGE UNIVERSITY OF TORONTO

T H E PLAUSIBILITY O F THEORIES

"

UDLEY SHAPERE'S paper raises both historical and philosophical issues. I t would be captious of me to pick holes in his history, since his reading of Liebig parallels so closely those given elsewhere by my wife and myself. Instead, let me follow up the methodological points he bases on his case histories: notably, his concluding doubt whether, after all, any hard and fast line can be drawn, or is worth drawing, between "genuinely scientific" and "metaphysical or nonsensical " ideas and hypotheses. For this purpose, Shapere's historical analysis demands only

D

'Abstract of a paper to be presented in an APA symposium, commenting on Dudley Shapere, "Plausibility and Justification in the Development of Science," this J O ~ N A 63, L , 20 (Oct. 27, 1966) : 611-621.

SYMPOSIUM: PLAUSIBILITY IN SCIENCE

625

one gloss. I n Descartes, and also-it seems-in Huygens, the term 'matter' refers to a n y spatially extended agency capable of producing physical effects : in this respect, the seventeenth-century scientists began by following the usage of the traditional Greek term soma. By Liebig's time, however, the term was restricted (following Newton and Dalton) to ponderable, corporeal "stuffs" characterized by mass. Applying nineteenth-century criteria, for example, electromagnetic radiation did not count as a "material substance"; but, by Descartes's standards, it should certainly have done so. I n considering "the reasoning presented by Huygens" against regarding light as involving "a transport of matter, " one should therefore ask whether in Cartesian terms the argument was in fact correct. (Note: one cannot afford to underestimate the continuous shifts of meaning affecting key words like 'matter' and 'substance', 'cause' and 'force', throughout the last 350 years of natural science.) I t is clear, by now, that "the extreme speed with which light spreads on every side" does not rule out light's carrying energy and momentum-even if not mass-whether by the classical Poynting vector or in the form of "photons." So Huygens' reasoning, though "universally regarded as respectable science," appears in retrospect to rest on an assumption that was as unproved in his time as in ours. This may help to explain why, despite Huygens' advocacy, the wave theory of light remained unfashionable until the nineteenth century. Shapere derives three morals from his case studies. (1) Reasons can be given, and judicially appraised, not merely for accepting a novel scientific theory (hypothesis, explanation) as justified, but alternatively-in a more preliminary way-for taking it seriously, as being plausible: this fact, he implies (and I agree), undercuts philosophical attempts to divide considerations about scientific discovery, allegedly belonging to "pragmatics," from considerations about scientific proof, which are the concern of "inductive logic." ( 2 ) We may reasonably claim plausibility on behalf of a new theory in cases where-for example-a sufficient analogy appears to exist with a form of theory that has already been successful in another domain of science. (3) Provided such an analogy exists, a novel theory cannot be debarred from consideration either on formal grounds alone (as vacuous or nonsensical) or on grounds of employing intellectually disreputable concepts (e.g., as vitalistic). Now, the opening clause of moral 3 is crucial. In reacting against apriorism and conformism, one may be tempted to insist that every new scientific idea is equally entitled to a hearing. Shapere's argument steers a middle course: plausibility has de-

626

T H E JOURNAL OF PHILOSOPHY

grees, and needs arguing for. This said, however, we need a more detailed analysis of the nature of arguments for plausibility and of the standards for appraising them. That task is forever being embarked on-e.g., by C. S. Peirce and, more recently, by Norwood Russell Hanson-but it has yet to be convincingly completed. This does not mean that a "critique of plausible reasoning" could not be completed, but we must be clear what we are entitled to require of it. I t calls ( I suspect) less for the invention of a new formalism, to be christened abductive logic, than for a more systematic application of Shapere's own case-study methods. Why should "case-study " be the appropriate method ? Speaking generally, we may distinguish ( a ) arguments that apply a given concept (principle, method of reasoning) to a particular instance, without the nature or legitimacy of this application being called in question, from (bl arguments by which the very applicability of such a concept (principle, method) is established, having initially been disputed. This distinction is worth making in general terms, since it can be drawn in many fields of argumentlegal or aesthetic, as well as scientific-and, in every field, the scope for straightforward computation is restricted to arguments of type a. I n arguments of type b, reasoning is commonly based on appeals to precedent, analogy, and the other devices of "case law." (Contrast, for instance, the calculations by which our income-tax liabilities are computed, when the basis of taxation is not in dispute, with the arguments by which the detailed applicability of the tax code is established in cases where the taxability of any item is questioned.) Logicians have dreamed of eliminating the messiness of caselaw arguments from natural science finally-tanzen usque recurret. Within a formalized theory such as quantum mechanics, for instance, arguments may be judged by their conformity to mathematical requirements; but, the moment the ultimate relevance of that particular theory is challenged, the debate falls back-necessarily, since what dse is there?-on appeals to precedent. ("You're running far ahead of the facts, like William Prout" or "Take care not to repeat Duhem's mistake, of laughing at J. J. Thomson and the electron.") That is inevitable, in science as much as in law. For both are continzcing enterprises : continuing, because forever incomplete, in that the match between our concepts and our experience can always be further refined. I n a "finally completed" theory, such as euclidean geometry and newtonian mechanics were once thought to be, one might legitimately suppose that questions about the applicability of the key concepts could be settled once for all.

I n real-life science, as in real-life law, we no longer expect to achieve such finality. The most significant word in the title of Shapere's paper is, accordingly, the word 'development'; for it carries a built-in reminder that, in science, today's best-established concepts may have their application called in question a decade hence. This being so, we should resist even Shapere's last concession-the watered-down distinction between 'science' and 'nonscience' by which a novel idea having as yet "too little rationale" can provisionally be labeled as "nonscientific." For the speculative phase in the development of a science (or in the solution of any scientific problem of real profundity) is legitimate, recurrent, and necessary; and to dismiss it as being "not really science a t all" is to concede too much to the conceptual conservatives. STEPHENTOULMIN BRANDEIS UNIVERSITY

ANALYTICITY AND CONCEPTUAL REVISION * I. INTRODUCTION

T

H E view that analytic propositions are those which are true in virtue of rules of use is basically correct. But there are many kinds of rules of use, and rules of some of these kinds do not generate truth. There is nothing like a grammatical analytic, t,hough grammatical rules are rules of use. So, this rules-of-use view falls short of being an explanatory account. My problem is to find what it is that is special about those rules of use which do generate truth. I shall argue that they are distinguished from others by their purpose rather than their content. Given their special purpose, one can explain how they generate truth. I t will follow that linguistic regularities, considered apart from the purposes of those who use language, fail to provide a basis for understanding analyticity. On my account of it, analyticity turns out to be a less important characteristic of propositions than necessity. This is because necessity, unlike analyticity, has its roots, not just in a contemporary system of usage, but in a wide family of systems of belief and usage. My efforts to deflate the philosophical value of the analytic will be summed up in the conclusion that analytic propositions can be contingent. I think this conclusion is behind the feeling that the propositions of logic and arithmetic are not * To be presented in an APA symposium of the same title, Dec. 27, 1966.