INQUIRY AS INQUIRY: A LOGIC OF SCIENTIFIC DISCOVERY

Transcription

1 INQUIRY AS INQUIRY: A LOGIC OF SCIENTIFIC DISCOVERY

2 JAAKKO HINTIKKA SELECTED PAPERS VOLUME 5 1. Ludwig Wittgenstein. Half-Truths and One-and-a-Half-Truths ISBN Lingua Universalis vs. Calculus Ratiocinator. An Ultimate Presupposition of Twentieth-Century Philosophy ISBN Language, Truth and Logic in Mathematics ISBN Paradigms for Language Theory and Other Essays ISBN Inquiry as Inquiry. A Logic of Scientific Discovery ISBN

3 JAAKKO HINTIKKA Boston University INQUIRY AS INQUIRY: A LOGIC OF SCIENTIFIC DISCOVERY Springer-Science+Business Media, B.Y.

4 A C.LP. Catalogue record for this book is available from the Library of Congress. Printed on acid-free paper All Rights Reserved ISBN ISBN (ebook) DOI / Springer Science+Business Media Dordrecht and copyright holders as specified on appropriate pages within. Originally published by Kluwer Academic Publishers in SofIcover reprint of the hardcover 1st edition 1999 No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission from the copyright owner.

5 TABLE OF CONTENTS ORIGIN OF THE ESSAYS INTRODUCTION Vll IX 1. "Is Logic the Key to all Good Reasoning?" 2. "The Role of Logic in Argumentation" (with IJpo Halonen and Arto Mutanen) 47 "Interrogative Logic as a General Theory of Reasoning" 4. "What Is Abduction? The Fundamental Problem of 91 Contemporary Epistemology" 5. "True and False Logics of Scientific Discovery" " A Spectrum of Logic of Questioning" "What Is the Logic of Experimental Inquiry?" "The Concept ofinduction in the Light of the 161 Interrogative Approach to Inquiry" 9. (with IJpo Halonen) 183 "Semantics and Pragmatics for Why-Questions" 10. "The Varieties ofinformation and Scientific Explanation" "On the Incommensurability of Theories" "Theory-Ladenness of Observations as a Test Case of 241 Kuhn's Approach to Scientific Inquiry" 13. "Ramsey Sentences and the Meaning of Quantifiers" "Towards a General Theory ofidentifiability" 267

6 ORIGIN OF THE ESSAYS The following list indicates the first publication forums of the different essays included in the present volume (the forthcoming publication forum, if an essay appears here for the first time): 1. "Is Logic the Key to all Good Reasoning?," forthcoming. 2. "The Role of Logic in Argumentation," The Monist vol. 72, no. 1 (1989), pp (with Ilpo Halonen and Arto Mutanen) "Interrogative Logic as a General Theory of Reasoning," forthcoming in R. Johnson and J. Woods, editors, Handbook of Applied Logic, Kluwer Academic, Dordrecht. 4. "What Is Abduction? The Fundamental Problem of Contemporary Epistemology," Transactions of the Charles S. Peirce Society vol. 34, (1998), pp "True and False Logics of Scientific Discovery," Communication and Cognition vol. 18 (1985), pp "A Spectrum of Logic of Questioning," Philosophica vol. 35 (1985), pp "What Is the Logic of Experimental Inquiry?," Synthese vol. 74 (1988), pp "The Concept ofinduction in the Light of the Interrogative Approach to Inquiry," in Inference, Explanation and Other Frustrations: Essays in the Philosophy of Science, ed. by John Earman, U. of California Press, 1993, pp (with Ilpo Halonen) "Semantics and Pragmatics for Why-Questions," Journal of Philosophy vol. 92 (1995), pp "The Varieties of Information and Scientific Explanation," in B. van Rootselaar and 1.F. Staal, editors, Logic, Methodology and Philosophy of Science III, North-Holland (Elsevier), Amsterdam, 1968, pp "On the Incommensurability of Theories," Philosophy of Science vol. 55 (1988), pp "Theory-Ladenness of Observations as a Test Case of Kuhn's Approach to Scientific Inquiry," in PSA 1992, ed. by David Hull et ai., vol. 1, Philosophy of Science Association, East Lansing, MI, 1992, pp "Ramsey sentences and the Meaning of Quantifiers," Philosophy of Science vol. 65, (June 1998), pp "Towards a General Theory ofidentifiability," in Definitions and Definability, ed. by James Fetzer et al., Kluwer Academic, Dordrecht, 1991, pp All the previously published essays are reproduced here with the pennission of the respective copyright owners, if any. These pennissions are most gratefully acknowledged. Thanks are also due to the Editors of the volumes in which these articles appeared previously. vii

7 INTRODUCTION This volume has been singularly frustrating for me to edit. I am also painfully aware of the reason why. I am trying to do far too much in a single volume. I am trying to reform, or at least show how to reform, much of our epistemology and our general ("formal") philosophy of science. Such a task would require a longer sustained argument than can be provided by a single book, in at least two respects. On the one hand, the basic ideas of the new approach should be explained and motivated and their nature should be spelled out, including the conceptual tools needed to express them. On the other hand, the value of the new approach should be shown by applying them to a number of important representative problems. There does not seem to be any a priori reason why such a task could be accomplished by a collection of studies like the present one. Admittedly, such a presentation imposes greater requirements on the reader than an integrated book. It is like presenting to the reader, not a ready-make picture, but rather a jigsaw puzzle to be completed by him or her. Unfortunately, the picture I would like to paint is too large to be presented to the reader even in the form of a jigsaw puzzle of ideas. For that purpose, my basic ideas should have been explained more fully, more studies should have been carried out than has been possible, and those that I am presenting here should have been carried out to a greater depth and their common denominators explained much more patiently. Hence what the reader will find here is an imperfect jigsaw puzzle from which many pieces are still missing. My explanation, if not an excuse, is that it would probably have taken another lifetime's work to present to you a completed picture of the kind I have in mind. What I must try to do in this introduction is hence to help the reader by indicating what the hoped-for overall picture is intended to look like, how the pieces displayed here hang together, how they find their place in the total structure, and what some of the missing pieces are like. I am fully convinced that the missing tiles do exist for some cases because I have already found them but cannot present them in this volume. Even though I referred to the approach advocated in the central essays of this volume as "new," its leading idea is not. This idea is as old as Socrates, and hence older than most of our familiar epistemology and logic. It is the idea of knowledge-seeking by questioning or, more accurately, of all rational knowledge-seeking as implicit or explicit questioning. I am using the phrase "inquiry as inquiry" to express the idea. For what my leading idea is is precisely an assimilation of all rational inquiry in the generic sense of ix

8 x INQUIRY AS INQUIRY: A LOGIC OF SCIENTIFIC DISCOVERY searching for information or knowledge to inquiry in the etymological sense, that is, to a process of querying, or interrogation. Needless to say, Socrates has not been the only philosopher to consider inquiry as inquiry. It would not be difficult to persuade a Collingwood or a Gadamer that each and every proposition figuring in rational inquiry can be thought of as an answer to a question. Unfortunately, neither one of these two philosophers provides us with any real argument for this position or anything like an explicit "logic of questions and answers," in spite of their prominent use of this very phrase. In the essay "What Is Abduction?" I put forward an argument to the effect that if a reasoner is rational, that is to say, if his or her argumentative steps can be rationally evaluated, then they might as well be considered as answers to his or her question. For the presuppositions of rational evaluation are the same ones that make it possible to think of any new item information as being introduced as an answer to a question. The main exception to the need of doing so is the one first noted by Aristotle, viz. an answer that is (as we are wont to say) logically implied by earlier answers (together with one's initial premises). The logic of epistemology is therefore the logic of questions and answers - and of question-answer sequences. But the earlier forms of the theory of question-answer sequences are not adequate to the task. This prompted the paper "Interrogative Logic as the General Theory of Reasoning" (written jointly with Ilpo Halonen and Arto Mutanen) where an outline of an explicit theory of question-answer logic is presented. Alas, even within this survey paper many jigsaw puzzle pieces are still missing. Among them are most of the ingredients of a state-of-the-art epistemic logic. For it quickly turns out that while several of the basic features of a theory of questions and answers can be discussed by means of ordinary first-order logic, eventually the epistemic element in questions and answers must be acknowledged. To study this component adequately the development of an adequate epistemic logic is needed. This development is one of the many component tasks in the intended total reform of epistemology that cannot unfortunately be represented in this volume. Only a brief sketch of such a logic is presented here. But what is the payoff of such an approach? Is it merely an old theory in new bottles? Perhaps the most important new insight that is yielded by the interrogative approach concerns the role of deductive logic in the context of an overall empirical inquiry. Needless to say, asking a question and receiving an answer (that is, an interrogative move) is radically different from a step logical deduction (logical inference move). However, from a strategic point of view the two steps are parallel, in the sense that the principles guiding the choice of the best questions to ask are (partly but importantly) analogous with the strategic principles guiding the choice of the best logical inferences one can draw from given premises. Perhaps this result is not so new an idea, either. What it shows is that strategically speaking deductive logic is in the last analysis the guide to all good reasoning. I have called such a conception the Sherlock Holmes conception of logic. It can be taken to

9 INTRODUCTION xi vindicate the great detective's claims to have solved his cases by "deductions," "inferences," or "logic." This idea is explained in easily accessible terms inter alia in the first two essays of this volume. The essay "The Role of Logic in Argumentation" is included partly for historical reasons, (being the first presentation of the analogy) even though it deals only with a part of the strategic analogy between questioning and deduction. Another main insight that the interrogative approach yields is the possibility of a rational and even logical theory of discovery. Or, rather, I should say "actuality" instead of "possibility." For the logic of question-answer sequences already contains as one of its components a logical theory of discovery. Indeed, a theory of pure discovery is obtained as an extreme case by assuming that all of the answers that an inquirer receives are true and known to be true. Then the problem of justification becomes redundant, and yet we can develop an interesting, rich theory of discovery whose centerpiece is the problem of optimal question selection. There may not be any mechanical rules for the optimal question selection but that does not make the problem of discovery any less rational or even logical. Philosophers used to maintain that only a "context of justification" can be dealt with by rational (discursive and logical) terms, whereas a "context of discovery" must be left to the mercies of intuition and serendipity. The true situation turns out to be an almost diametrically opposite one. A logic of pure discovery - that is, questioning with invariably true answers - has a beautifully clear structure. In contrast, a logic of justification will involve the complicated problem of evaluating different answers by different answerers, and will therefore end up being much messier than a logic of discovery. Once again, an epistemological discovery prompted by the interrogative approach is better in step with common sense than recent philosophical preconceptions. For undoubtedly you would for instance rather base your argument in a court of law on the answers of an unimpeachable witness than those of a witness whose responses have to be tested against collateral information. The "logic of discovery" that results from the interrogative approach depends essentially on the class of answers that the inquirer is in the position to receive. Most philosophers have apparently assumed that for a scientific inquirer all the rock-bottom answers must be thought of as particular propositions. This assumption has led to the inductivist and to the hypothetico-deductive models of science. In reality, it is nevertheless totally unrealistic, as is illustrated among other things by the possibility of putting questions to nature in the form of experiments. An answer to an experimental question is typically a functional dependence between two variables, which can only be expressed in terms of dependent quantifiers, and hence not a particular proposition. These matters are discussed in essays 5-6 below. Again, they are unfortunately only the first sketchy words on the subject, not the last ones.

10 xii INQUIRY AS INQUIRY: A LOGIC OF SCIENTIFIC DISCOVERY Thus the logic of the experimental method can be spelled out and studied by reference to the interrogative approach. A beginning of such a study is made in the essay "What Is the Logic of Experimental Inquiry?" It has already led to two further insights of major importance. First, most actual dependence relationships established experimentally are only partial ones, holding only for a restricted range of the values of the controlled variable. The problem of extending and integrating such partial generalizations is a frequent and important kind of scientific problem whose general nature and whose history is only now beginning to be visible. It is discussed in my essay 'The Concept of Induction in the Light of the Interrogative Approach to Inquiry." The other interesting insight is that the conclusiveness conditions of answers to questions (not only experimental questions) are not entirely empirical in nature, but contain a conceptual element. In the case of experimental questions this conceptual element amounts to knowing what the function is mathematically speaking that an experiment has yielded as its initial output. I have discussed this matter in my paper "Knowledge of Functions in the Growth of Mathematical Knowledge" forthcoming in a volume entitled The Growth of Mathematical Knowledge edited by Herbert Breger and Emily Grosholz (Kluwer Academic, Dordrecht). Another promising line of thought opened by the question-theoretical approach is to develop an interrogative analogue to the usual theory of definability. In the new theory it is not any longer asked whether (and how) certain concepts are fixed uniquely by their role in some given theory alone. It is now asked whether (and how) they are specified by the theory plus the answers the inquirer can receive. It turns out that this interrogative analogue to definability has been acknowledged, given a name, and studied by the practicing methodologists called econometricians and systems theorists. It is a generalization of what they call identifiability. Some of the basic features of a general logical theory of identifiability are discussed in the essay "Toward a General Theory of Identificability. " A part of the interest of the concept of identifiability is due to the fact that it shows that a concept can be empirically determined by other concepts and yet not reducible to them. In this respect it is similar to, but much more sharply defined, than the confused notion of supervenience that is undeservedly popular in these days. A major application of the ideas which have motivated the interrogative approach concerns the theory of explanation. I have studied this subject in recent years in cooperation with Ilpo Halonen. Our results are too extensive to be included in this volume. What we have done is to indicate the question-theoretical basis of this study in the paper "Semantics and Pragmatics for Why-questions." Even though the older paper "The Varieties of Information and Scientific Explanation" predates my interrogative approach to inquiry, it complements this approach by illuminating the role of the concept of information in the scientific process.

11 INTRODUCTION xiii Information is the most important "epistemic utility" which helps to determine the payoffs and hence the strategies of interrogative games. One cannot discuss general problems of the philosophy of science in these days without commenting on the claims of the so-called "New Philosophers of Science," such as Kuhn, Hanson, and Lakatos. In the course of years, I have become increasingly disenchanted with their New Philosophy of Science, even apart from the simple-minded uses of the ideas of Kuhn and Hanson by their self-appointed followers. The main concept Kuhn and his ilk are wielding are far to unsharp, not to say confused, to serve as means of reaching serious insights into scientific discovery and scientific process in general. In the essays on the incommensurability of theories and on the theory-iadenness of observations, I show by means of examples that the New Philosophers' claims concerning these two typical concepts are at best confused and superficial. This will make (I hope) the outline of the puzzle picture I am dealing with clearer to the reader. I trust that it will also show the reader how much needs still to be filled in within those outlines. But even so, the picture that is beginning to emerge shows several truly important things about the epistemology of science and of the science of epistemology.