PROBLEMS OF THE LOGIC OF SCIENTIFIC KNOWLEDGE

PROBLEMS OF THE LOGIC OF SCIENTIFIC KNOWLEDGE

SYNTHESE LIBRARY MONOGRAPHS ON EPISTEMOLOGY, LOGIC, METHODOLOGY, PHILOSOPHY OF SCIENCE, SOCIOLOGY OF SCIENCE AND OF KNOWLEDGE, AND ON THE MATHEMATICAL METHODS OF SOCIAL AND BEHAVIORAL SCIENCES Editors: DONALD DAVIDSON, Princeton University JAAKKO HINTIKKA, University of Helsinki and Stanford University GABRIEL NUCHELMANS, University of Leyden WESLEY C. SALMON, Indiana University

PROBLEMS OF THE LOGIC OF SCIENTIFIC KNOWLEDGE Edited by P. V. TAVANEC D. REIDEL PUBLISHING COMPANY / DORDRECHT-HOLLAND

PROBLEMI LOGIKI NAUCNOGO POZNANIJA Published by Nauka, Moscow, 1964 Translated from the Russian by T. J. Blakeley ISBN-13: 978-94-010-3395-4 001: 10.1007/978-94-010-3393-0 e-isbn-13: 978-94-0 I 0-3393-0 1970. D. Reidel Publishing Company, Dordrecht, Holland Softcover reprint of the hardcover 15t edition 1970 No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher

TRANSLATOR'S NOTE This book could have been translated either by a logician with knowledge of contemporary Soviet philosophy or by a philosopher with some knowledge of contemporary logic. Because it was done by the latter, the translation was laboriously reviewed by at least seven experts charged with eliminating 'infelicities' (as one of them was kind enough to put it!) and egregious blunders in translation. The results will not always be uniform for it is hard for someone who is not devoted heart and soul to logic alone to see why one should prefer 'urn' to 'jar' when drawing probability marbles; or 'environment' to 'milieu'; or 'behavior' to 'comportment', etc. It is above all quite impossible to see why one should use the irretrievably vulgar expression 'one-to-one' instead of the legitimate 'univocal', especially when talking in the context of analogy. Other irregularities can be directly traced to the lack of a consecrated mathematical-logical nomenclature in Russian. We have resolved the conflicts by maintaining, with the help of the experts, the consecrated mathematical-logical terminology used in the non-soviet scientific world, while trying to convey the greater 'ontological depth' of the specifically dialectical-materialist view. 'Predlozenie' has been rendered as 'sentence', vyskazyvanie' as 'proposition', and 'utverzdenie' as 'statement' or 'assertion', unless the context indicated otherwise. 'Soderzatel'nyj' has been rendered as 'informal' where possible and otherwise as 'contentful'. In line with accepted usage, 'zakonomernyj' has been rendered as 'regular' or 'law-bound', according to context. In all cases where a decision had to be made, style has been sacrificed to logical clarity. But, uncomfortable neologisms, like 'deductification', 'always true', 'mass-phenomena', etc., have been avoided whenever a clear alternative was available. We have not tried to eliminate the lengthy enumerations, appositions and juxtapositions which are so typical of Soviet philosophical writing. The following list of inconsistencies on the part of the Soviet authors themselves will help to explain other obscurities. Sadovskij sometimes

VI TRANSLATOR'S NOTE uses vyvod and dedukcija as synonyms (=deduction); sometimes in the same paragraph he will use vyvod as the genus ( = inference), of which dedukcija and indukcija are species. To avoid any possible confusion, we have put 'inferable' for vyvodimyj, instead of 'deducible'. Demov's article is a study in the use and abuse of terminology. His main concept, 'inference by analogy' is sometimes vyvod po analogii, sometimes umozakljucenie po analogii: the two terms can be taken as synonymous but even the careless philosopher (not to speak of the careful logician) should note that umozakljucenie has psychologistic overtones. Demov uses kacestvo, svojstvo and priznak quite randomly: we have rendered them as 'quality', 'property' and 'characteristic', respectively; the resulting potpourri is his responsibility. Although we have bowed to the experts in translating Gorskij's odnoznacnost' as 'uniqueness', I am sure 'synonymity' would be more consistent with his philosophic history. It is also hard to imagine Gorskij using the term 'domain of individuals' and he most certainly does not intend the nominalism which seems to lurk behind these words. Since most of the symbolism is a mix of Cyrillic and Latin characters we have not transposed it in the translation. Some of the quotations of non-russian authors are taken from the original: as a rule, however, we have simply translated from the Russian - which offers the advantage of seeing the text in question through their eyes. In the bibliographies the blatant errors have been corrected (A. J. Ayer for A. Y. Ayer; H. Reichenbach for G. Reichenbach) but there are still many incomplete references - which is all too characteristic of Soviet bibliographical work. Our thanks are due to Professor Jaakko Hintikka and his staff for indicating weaknesses and mistakes in translation. Professor Edward M. MacKinnon of Boston College was kind enough to give the manuscript a final reading. The f(')llowing abbreviations have been retained, especially in the bibliographies: BSE = Large Soviet NT Encyclopedia i dr. = and others (et alii) ITEF = Institute of Theoretical and Soc. Experimental Physics = Scientific Works (periodic publications of scientific institutions) = works (collected or complete)

TRANSLATOR'S NOTE VII t. = volume UZ = Learned Papers (see NT) VF = Voprosy filosofii (Questions of Philosophy) Vyp. = issue ZETF = Journal of Experimental and Theoretical Physics December 23, 1966 Boston College

TABLE OF CONTENTS Translator's Note P. v. Tavanec and V. S. Svyrev, The Logic of Scientific Knowledge v 1 V. A. Smirnov, Levels of Knowledge and Stages in the Process of Knowledge I. Differences Between the Problems, 'Sensation-Thought' and 'Empirical-Theoretical' II. Basis of the Division of the Sentences of the Language of Science into Levels III. The Semantic System: Admissible Objects of Thought and Modes of Expression IV. Empirical and Theoretical Objects of Science V. Sentences Which Express Facts and Sentences Which Formulate Laws VI. Stages in the Process of Knowledge 1. Observation 2. Analysis of observation protocols and discovery of empirical connections 45 3. Prediction by means of facts and empirical connections 47 VII. Types of Explanation of Empirical Connections 48 VIII. Stages in the Process of Knowledge, II 50 4. Elaboration of the basic ideas and discovery of the fundamental relations, basic to explanation: formation of scientific theory 50 5. Deduction of some theoretical laws from other theoretical laws: development of theory 50 6. Explanation of scientific facts, i.e., discovery through empirical relations of corresponding theoretical models 51 7. The discovery of empirical connections through theoretical descriptions 51 22 22 23 26 29 33 41 42

x TABLE OF CONTENTS v. S. Svyrev, Problems of the Logical-Methodological Analysis of Relations Between the Theoretical and Empirical Planes of Scientific Knowledge 55 I. The Traditional Inductivist Approach to the Problem of the Relations Between Theoretical and Empirical Knowledge and its Limitations 57 II. Critique of the Neopositivist Approach to the Analysis of the Relations Between the Theoretical and Empirical Levels of Scientific Knowledge 64 III. Contemporary Logic of Science on the Relations Between Theoretical and Empirical Knowledge: The Connection of the Theoretical and Empirical Levels of Knowledge in the Structure of Hypothetical-Deductive Theory 69 IV. Contemporary Logic of Science on the Relations Between Empirical and Theoretical Knowledge: The Problem of the Establishment of Logical Correspondence Between Theoretical and Empirical Knowledge 79 A. A. Zinov'ev, Logical and Physical Implication 91 V. N. Sadovskij, The Deductive Method as a Problem of the Logic of Science 160 I. Introduction 160 II. Deduction and Deductive Inference 164 III. Deductive System and Deductive Theory 175 IV. Types of Deductive Systems 186 V. Problems of the Logical-Epistemological Analysis of the Deductive Sphere of Knowledge 199 G. I. Ruzavin, Probability Logic and its Role in Scientific Research 212 I. Introduction 212 II. Systems of Probability Logic 217 1. Various interpretations of the notion of probability 217 2. Systems of probability logic, based on the frequency interpretation of probability 221 3. Systems of probability logic which consider probability as some sort of logical relation between propositions 231

T ABLE OF CONTENTS Xl III. Probability Logic and Statistical Inference 245 IV. Probability Logic and the Problem of the Selection of Hypotheses 255 V. Probability Logic and the Problem of Confirmation of Hypotheses 258 A. 1. Uemov, The Basic Forms and Rules of Inference by Analogy 266 I. The General Schema of Inferences by Analogy 266 II. Traditional Analogy 273 III. Causal and Substantial Analogy 283 IV. Analogy of Consequence 291 V. Analogy of Correlation 297 VI. Functional-Structural and Structural-Functional Analogy 305 D. P. Gorskij, On the Types of Definition and Their Importance for Science 312 I. Preliminary Remarks 313 II. Types of Definition 315 III. The Problem of Definitions in Formal Systems 348 IV. On the Importance of Definitions in Science 365 A. L. Subbotin, Idealization as a Method of Scientific Knowledge 376 I. The Abstraction of Identity 376 II. Idealization 381 III. Some Methodological Considerations 387 A. N. Rakitov, The Statistical Interpretation of Fact and the Role of Statistical Methods in the Structure of Empirical Knowledge I. The Nature of Empirical Knowledge and the Principle of Verification 1. The schema of the process of verification 2. Substitution of the confirmation principle for the principle of verification 3. The principle of complete verifiability 4. Data of a single observation and the cardinal relation 5. The transition from data to statistical resume 6. The fact of science 394 395 396 398 399 400 403 405

XII T ABLE OF CONTENTS 7. The fact and the cardinal relation 406 8. Summary 408 II. The Statistical Nature of the Object and the Structure of the Construction of Empirical Knowledge 409 1. The object and objective object of investigation 409 2. The statistical nature of the empirical object 411 3. The components of the structure 413 4. The empirical situation and the single datum 413 5. Conditions for obtaining,d;. = 0. 414 6. Control-experiments 415 7. Investigation of the statistical nature of the set-up 416 8. Example of an experiment 418 9. The method of statistical modelling 419 10. The statistical nature of the so-called single observation 421 Index of Names 427