THE METAPHYSICS OF LOGICAL POSITIVISM by GUSTAV BERGMANN LONGMANS, GREEN AND CO. NEW YORK LONDON TORONTO 1954 LONGMANS, GREEN AND CO., INC. JJ FIFTH AVENUE, NEW YORK 3 LONGMANS, GREEN AND CO. Ltd. 6 & 7 CLIFFORD STREET, LONDON W I LONGMANS, GREEN AND CO. 21 J VICTORIA STREET, TORONTO I COPYRIGHT • 1954 BY LONGMANS, GREEN AND CO., INC. ALL RIGHTS RESERVED, INCLUDING THE RIGHT TO REPRODUCE THIS BOOK, OR ANY PORTION THEREOF, IN ANY FORM PUBLISHED SIMULTANEOUSLY IN THE DOMINION OF CANADA BY LONGMANS, GREEN AND CO., TORONTO FIRST EDITION LIBRARY OF CONGRESS CATALOG CARD NUMBER J 4-J79O Printed in the United States of America VAN REES PRESS • NEW YORK THERE WAS a time when a philosopher had to write books, or at least one thick book, if he wished to obtain a hearing for what was then called his system. This age is of shorter breath. Today many philosophers are content to pursue their education in public by means of the papers they insert in learned journals. If such a one has reached middle age and is fortunate enough to have friends who encourage him, he is urged to publish a volume of essays. This is the case in which I find myself. I have for some time, and I fear quite recklessly, conducted my education in public; I have a few generous friends whose judgment I value; and I must confess that the mere thought of writing a book on first philosophy terrifies me. Some causes of this terror prob­ ably belong in the psychoanalyst’s study; one reason I think bears public statement. As things now stand, no philosopher, at least no analytical philosopher, can reasonably start afresh. All he can hope to do is to throw some new light on a few points within the tradition to which he owes most of his views, or, perhaps, to add a new twist to the method prac­ ticed by its masters. Thus, if he submits to the rules of the book-writing game, he will be forced to say once more many things that have been said before and much better than he can possibly hope to say them. To do this, outside the seminar room, is to waste one’s colleagues’ time as well as one’s own. As for our students, I believe it is our duty to send them first of all to the classics and to the great con­ temporary masters. So instructed, they, too, should be able to make sense of what is not exactly a textbook. This is not a collection of my papers on first philosophy but a selection from them. Nor is the order in which they are arranged chronological. This requires some comment. The papers fall into three groups. Taken together, the first six, of most recent origin, provide an outline of the views I now hold. The second group consists of the next three, which are the earliest included in this volume. Together with three other still earlier ones which I have excluded, they form a unit centered around the realism phenomenalism issue. The excluded papers are “Pure Semantics, Sentences, and Propositions” (Mind, 53, 1944), “A Positivistic Metaphysics of Consciousness” (Mind, 54, 1945), “Undefined Descriptive Predicates” (Philosophy and Phenomenological Research, 8, 1947). I omit them because for the most part they merely say very badly what I have since said again, a little less badly, in the six essays of the first group. I mention them because there I first struck out on my own, trying to free myself from the influence of Camapian positivism though not yet, alas, from its apparatus. Having said that much, as I believe I should, I wish to add, as I believe I also should, that this by now radical dissent has not at all affected either my gratitude or my admiration for Carnap. I still think of him as the outstanding figure in a major phase of the positiv­ istic movement. The third group consists of all the remain­ ing essays, some of them very short. These are in the main elaborations of themes struck in the first nine pieces. The arrangement within this last group represents a compromise between their subject matter and the order in which I re­ member having written them. The concluding essay differs from the rest. Quite nontechnical, it touches at least indi­ rectly on my philosophy in that broader sense in which everyone who is not himself an analytical philosopher speaks of a man’s philosophy. Thus it is, perhaps, not out of place at the end of a volume that is otherwise rather technical. Aside from a few editorial changes I have left the papers as they were originally written. This means that, taken singly, most of them would benefit from expansion. Taken together, they should throw a good deal of light upon each other. It also means that some, particularly the earlier ones, contain passages which I could now improve and which do not, as they stand, completely jibe with the rest. But I fear that, if I started revising, I would soon reach the point where the only proper thing to do is what I dare not do, namely, write a metaphysical treatise. So I shall persist in my ways, not from either arrogance or laziness but because I cannot help it, hoping that I shall learn from the critics of this volume what improvements and clarifications are most needed so that I can try to provide them over the years in some future essays. But there is one matter on which I feel I ought to say a word now, since I know that it has bothered some of my friends. Consider an awareness whose content is referred to by ‘This is a tree’. An awareness must be dis­ tinguished from its content. Yet, to be an awareness with a certain content and to be an awareness of a certain kind is one thing and not two. The predicate I proposed as a name of this kind is, in the illustration, ‘ ‘This is a tree’ ’. The aware­ nesses that exemplify it are named by particulars. The dis­ tinction is, of course, but a variant of the familiar one be­ tween type and token; like the latter, it adds in most contexts nothing to the analysis. Having once called attention to it, I proceeded therefore immediately to neglect it, using such predicates as ‘ ‘This is a tree’ ’ as if they were particulars. But I have come to think that there are contexts in which the distinction is helpful. However that may be, an accurate formulation will be found in the third essay. A table indicating where the pieces originally appeared will be found at the end of the volume. The necessary per­ missions by publishers and editors have been granted and are gratefully acknowledged. One of the essays was at this writing still in press. The editor of the journal in which it was to appear has permitted me to withdraw it. A careful subject index should help to bring out the continuity of theme and thought among the essays. By compiling such an index Professors May Brodbeck, of the University of Minnesota, and William H. Hay, of the Uni­ versity of Wisconsin, have further increased the large debt of gratitude which I owe them. G.B. Iowa City July 1, 1953 Logical Positivism............................................................... 1 Sem antics..................................................................................17 Logical Positivism, Language, and the Reconstruction of M etaphysics...............................................................30 Two Cornerstones of Em piricism ......................................78 Two Types of Linguistic Philosophy..............................106 Bodies, Minds, and A cts.......................................................132 Remarks on Realism .............................................................153 Sense Data, Linguistic Conventions, and Existence 176 Russell on Particulars............................................................. 197 Professor Ayer’s Analysis of Knowing..............................215 On Nonperceptual Intuition.................................................228 Conditions for an Extensional Elementaristic Language 232 A Note on Ontology.............................................................238 Logical Atomism, Elementarism, and the Analysis of V a lu e ................................................................................243 Comments on Professor Hempel’s “The Concept of Cognitive Significance” .................................................255 [ix] The Identity of Indiscemibles and the Formalist Definition of“ Identity” ................................................268 The Problem of Relations in Classical Psychology . 277 Id e o lo g y ................................................................................300 Author’s N o te ......................................................................... 327 I n d e x ......................................................................................329 LOGICAL POSITIVISM is a movement rather than a school, in the sense that those to whom the label is currently applied represent a broad range of interests and, on questions of common interest, often disagree with respect to what con­ stitutes the right answer or about the proper method to arrive at it. Thus a systematic and critical account, preceded by a minimum of historical remarks, becomes preferable to a simple narrative. In the case of a far-flung, complex, and still very active movement, a report of this kind must to some extent reflect the opinions of the reporter; in a sense, it is merely a proposal as to what body of doctrine may reasonably be called Logical Positivism. The difficulties one encounters in isolating a sufficiently comprehensive core of sufficiently detailed agreement are not unrelated to the circumstance that the writers who are, in fact, called Logical Positivists derive intellectually and, in most cases, also biographically from one of two centers, the Cambridge School of Analysis and the Vienna Circle. One might even say, though not without qualification, that what has come to be called Logical Positivism—the term appeared for the first time in 1930—is the result of the interaction that * V. Ferm, ed., A History of Philosophical Systems (New York: Philo­ sophical Library, 1950), pp. 471-82. Reprinted by permission.