Rescher Studies: A Collection of Essays on the Philosophical Work of Nicholas Rescher 9783110329094, 9783110328684

Citation preview

Robert Almeder (Ed.) Rescher Studies A Collection of Essays on the Philosophical Work of Nicholas Rescher

READING RESCHER Volume 2

Robert Almeder (Ed.)

Rescher Studies A Collection of Essays on the Philosophical Work of Nicholas Rescher Presented to Him on the Occasion of His 80th Birthday

ontos verlag Frankfurt I Paris I Ebikon I Lancaster I New Brunswick

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de. North and South America by Transaction Books Rutgers University Piscataway, NJ 08854-8042 [email protected]

United Kingdom, Ireland, Iceland, Turkey, Malta, Portugal by Gazelle Books Services Limited White Cross Mills Hightown LANCASTER, LA1 4XS [email protected]

Livraison pour la France et la Belgique: Librairie Philosophique J.Vrin 6, place de la Sorbonne ; F-75005 PARIS Tel. +33 (0)1 43 54 03 47 ; Fax +33 (0)1 43 54 48 18 www.vrin.fr

2008 ontos verlag P.O. Box 15 41, D-63133 Heusenstamm www.ontosverlag.com ISBN 13: 978-3-938793-97-8 2008 No part of this book may be reproduced, stored in retrieval systems or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use of the purchaser of the work

Printed on acid-free paper This hardcover binding meets the International Library standard Printed in Germany by buch bücher dd ag

CONTENTS Preface: ROBERT ALMEDER The Limits of Science, Realism, and Idealism: ROBERT ALMEDER

1

On Possibilities and Thought Experiments: DIDERIK BATENS

29

Coherentism and Coherence Truth in the Philosophy of Nicholas Rescher: BRYSON BROWN

59

Philosophical Anthropology: JAMES W. FELT

89

Pluralism and Consensus: LENN E. GOODMAN

105

Mediations, Wagers, and Existential Issues: Nicholas Rescher on Religion and Philosophy: JOHN HALDANE

121

Rescher’s Metaphysics: WILLIAM JAWORSKI

141

The Limits of Science Reconsidered: ULRICH MAJER

151

Conceivability: DIEGO MARCONI

169

Weird Worlds: ROBERT K. MEYER

197

Nicholas Rescher on Greek Philosophy and the Syllogism: JÜRGEN MITTELSTRASS and PETER SCHROEDER-HEISTER 211 Rescher’s Evolutionary Epistemology: JESÚS MOSTERÍN

241

Common Sense: JOSEPH PITT

253

Systematic Philosophy and Theoretical Framework on Nicholas Rescher’s A System of Pragmatic Idealism: LORENZ B. PUNTEL

261

Nicholas Rescher • Project Name

On Rescher’s View of Idealism (and Pragmatism): TOM ROCKMORE

287

Rescher on Arabic Logic: TONY STREET

309

Philosophy in the Future: AVRUM STROLL

325

Rescher on Explanation and Prediction: BAS C. VAN FRAASSEN

339

Nicholas Rescher on Scientific Progress: Science in the Face of Limited Cognitive and Technological Resources: THEODOR LEIBER and ROLAND WAGNER-DÖBLER

363

Rescher on Dialog Systems, Argumentation, and Burden of Proof: DOUGLAS WALTON and DAVID M. GODDEN

401

Rescher on Process Thought: MICHEL WEBER

429

How is Scientific Knowledge Economically Possible?: Nicholas Rescher’s Contributions to an Economic Understanding of Science: JAMES R. WIBLE

445

Possibility, Plenitude, and the Optimal World: Rescher on Leibniz’s Cosmology: CATHERINE WILSON

475

Rescher on Aporetics and Consistency: JOHN WOODS

493

Responses to the Contributed Essays: NICHOLAS RESCHER

513

Work by and about Nicholas Rescher

553

List of Contributors

573

Preface Robert Almeder

O

ver many years Nicholas Rescher has published many books on a wide range of standard philosophical topic. including Greek Thought, Arabic Logic, Formal Logic, Leibnitz, Conditional and Hypothetical Reasoning, Semantic and Pragmatics (or Dialectics), Possible World Theory, Epistemology, Evolutionary Epistemology, Explanation and Prediction, The Coherence Theory of Truth, Scientific Progress, Scientific Realism, Idealism and the Limits of Knowledge, Thought Experimentation, Cognitive Economy and Epistemetrics, Pragmatism, Process Thought, Ethics and Moral Philosophy, Social Philosophy, Philosophical Anthropology, Common Sensism, Philosophy of Religion and Metaphilosophy. Scattered over numerous publications, it is only in the present book that these many themes of Rescher’s work come between two covers. In the light of his productivity, it seemed to some of us that it would be a nifty idea if somebody engaged a group of willing and able philosophers to contribute to a collection of critical essays on the development of Nick’s systematic work over these last thirty years. More specifically, the plan would be to direct these selected philosophers to critically examine specific sections of Rescher’s later works in order that the core of what Nick has written both early and late would be the subject of focused, probing, and critical essays that would hopefully connect, compare, contrast, and generally lend illumination and understanding to the overall vision articulated in both the earlier works and the works in the last thirty years. Voila! A Festschrift to celebrate and honor Nicholas Rescher on his eightieth birthday for a life of continuing high-order philosophical and human achievement. I won the toss for the honor of being the editor of this volume. The fruits of that plan are here for everybody to see and hopefully to enjoy. Finally, of course, this collection is dedicated to Nick and Dorothy for whom one can be sure that the authors of the essays in this volume have a deep respect and not infrequently an abiding affection.

The Limits of Science, Realism, and Idealism Robert Almeder 1. INTRODUCTION

A

few years back I published an article on Nicholas Rescher’s views on the limits of natural science. [(LNS) 2000.40-59]. That article focused somewhat less on the realism/idealism discussion and more on the cluster of inter-related epistemological theses centering on Rescher’s views on the limits of natural science. This essay will concentrate more on Rescher’s realism and idealism, as his basic views on the limits of natural science have not changed substantially over the past thirty years although there seems to be a development by way of exfoliation and further clarification on the realism/idealism issue. This latter development coalesces in varying degrees in various publications including: [Cognitive Pragmatism, (CP) 2001.; Realistic Pragmatism, (RP) 2000.; Pragmatic Realism, (PR) 2003.; Human Knowledge in Idealistic Perspective, (HKIP) 1991.; Induction, (I) 1981; the trilogy, A System of Pragmatic Idealism, (SPI) 1992-93.; Coherence Theory of Truth, (CTT) 1973.] Before examining Rescher’s matured realism and idealism, however, it might prove helpful simply to review his views on the limits of natural science. Most of what follows on Rescher’s views on the limits of science is a condensation of things I already described, analyzed, and applauded in greater detail in an earlier piece [(LNS) 2000], and so there is little need to repeat in detail the more scholarly references and analysis in the earlier exposition. Other contributors to this volume, of course, may provide a slightly different and possibly more complete assessment of Rescher’s view on the limits of science. So, I will be mercifully brief on the basic position Rescher offers on the limits of natural science. After that, and in the end, I will try to locate where future discussion on realism and idealism might reflect favorably Rescher’s basic position as herein described. As I now see it, incidentally, Rescher’s most recent characterization of his realism in terms of a ‘myopic’ realism puts him on the side of the gods on the general issue. That’s the good news. When so understood, however, we may want to urge that his realism more effectively emphasize that

RESCHER STUDIES

while we have no reliable procedure for determining precisely which propositions in our language are in fact true, rather than fallibly but fully and warrantedly assertible (or estimated or approximated) as true, we do and must nevertheless attain to the truth indirectly in some myopic fashion by way of justifiably and sincerely accepting all such propositions as items of human knowledge, knowing full well that certainly some of them will disappear, like the proverbial dodo bird, into the dark emptiness of a rejected past. But if human knowledge requires demonstrable truth, we must re-construe truth solely in terms of warranted assertibilities relative to the evidential context available and the applicative and adaptive consequences by way of prediction and control that accepting such beliefs will imply. Hopefully, Rescher will not regard this conclusion as an unacceptable offering, as it still thrives on the warrantedly assertible belief that some of our fully justified beliefs must in fact instantiate in some way our ordinary alethic and platitudinous conception of truth because, as Rescher himself affirms, we would not otherwise have any way to account for the broad or general success of many of our cognitive endeavors except by appeal to mystery or miracle. But this conclusion also suggests that we need to abandon common sense, at least as an a priori or strongly reliable indicator or criterion for determining which beliefs are true rather than (like all other empirical and fallible beliefs which can be items of human knowledge) fully warrantedly assertible by all the rules of acceptance we are collectively empowered to accept by our own lights. In this we need not fear passing over into a dreadful transcendental idealism or, more specifically, the view that all properties of the world commonly experienced are purely linguistic in nature. We only need to refrain from asserting that what we are fully justified in believing to be so, is demonstrably so; and recognize that human knowledge is in fact simply what we would be fundamentally fully justified in believing to be so. Some of those beliefs will not be true in the alethic or platitudinous sense of ‘true, but we will need to treat all of them as true in that alethic sense knowing full well that some of them will not be true in that sense. But we are getting ahead of ourselves. More on this later. 2. RESCHER ON THE LIMITS OF NATURAL SCIENCE Rescher’s core position on the nature and limits of natural science includes the following basic items:

2

LIMITS OF SCIENCE, REALISM, AND IDEALISM

1. There are important empirically answerable questions about the observable world that we cannot answer by appeal to the methods of testing and confirmation proper to the natural sciences. Answers to the questions about the validity of the inductive method itself, or about the basic common sense beliefs upon which natural science rests at any time, for example, we cannot established directly and non-circularly by appeal to either deduction alone or the inductive methodology in natural science. 2. Our methods of induction are also limited in that even when one’s particular beliefs about the world turn out to be fully warrantedly assertible under the usual methods of testing and confirmation in natural science; such warranted assertibility is evidence neither of their certainty nor of any strong guarantee of their truth. They are truth approximations, and sometimes they are not true. In fact, for this reason, all empirical beliefs are fallible, and hence, however wellconfirmed, or confirmable, they are defeasible, and subject to revision or rejection in the light of future evidence or future collective decisions on future standards for evaluating the evidence. 3. Moreover, although the products of inductive methodology may be limited in the above ways, the methodology in general is more than justified as a methodology reliably providing knowledge of the external world because in the long run, and on the whole, it leads often enough to truth. Thus, for example, the limits of natural science do not imply a purely instrumentalist interpretation of empirical knowledge in general and scientific theories in particular. 4. The number of non-trivial empirically answerable questions that natural science can in principle answer at any given time is inexhaustible, or indefinitely large; and so natural science will never, even unto eternity, answer all the questions that it can in principle answer. There will always be ignorabimus questions. 5. Whatever progress natural science will make unto eternity by way of answering more and more non-trivial questions in time, the progress will be increasingly limited by economic forces and an expansion of natural resources so that, under the best of circumstances, there will be a logarithmic retardation in scientific progress as time advances.

3

RESCHER STUDIES

This retardation will result from the ever-increasing cost of the technology necessary for doing science in conjunction with ever increasing global demand for both finite and limited resources and the demand for social services outside science. 6. Scientific theories are limited in that while seeking to correctly and completely describe nature, they can never quite achieve as much although, to be sure, they approximate, estimate, or come close to doing as much; and finally, 7. As natural science progresses, it is inevitable that the technology required for solving social problems will become so increasingly complex, along with the complexity of social problems and the answers offered, that science will not be able to deliver the technology or answers necessary to solve the inevitable social problems that will emerge. Natural Science, then, in its applicative function, is limited in that it will not be able to provide adequately and indefinitely for important social progress and welfare because answers provided by technology become ever more complex, and the technology becomes ever more costly and difficult to control in the interest of promoting social welfare. Let me offer now a brief exposition of the reasoning Rescher offers for the first six of the above listed items. My conclusion shall be that while one might conceivably question some of the finer implications of Rescher’s thesis on the eternity of scientific progress, we can nonetheless applaud the impressive answers offered on the general question of the limits and scope of natural science. Thereafter we can reflect more on his realism and idealism and hopefully offer some clarification on his current position as it has emerged in the last few years. 3. SCIENTISM AND LEGITIMATELY ANSWERABLE QUESTIONS OUTSIDE SCIENCE For Rescher, it is a mistake to believe that the only legitimately answerable questions about the world of physical objects are those that can be answered by appeal solely to the methods of testing and confirmation in the natural sciences. Nor has he argued by way of a fallback position that

4

LIMITS OF SCIENCE, REALISM, AND IDEALISM

whether anybody knows anything at all is a scientific question. In fact, he finds such positions strikingly self-defeating because, as he says, “To engage in rational argumentation designed to establish the impossibility of philosophizing is in fact to engage in a bit of it.” [(LS). 1984. 3-4, 15, 108, and 210; and on the self defeating nature of scientism see also “Philosophobia” in The American Philosophical Quarterly vol.39, #3 (July 1992. 301-2 ] Like Carnap, and most other classical pragmatists, moreover, Rescher advances the view that whether one’s methods of testing and confirmation in natural science are themselves epistemologically acceptable is simply a function of whether they ultimately guarantee us precise sensory predictions and control (or applicative success) which is the primary end of cognitive inquiry. Trying to determining inductively whether one’s methods are acceptable is viciously circular if one’s purpose is to establish the inductive justification of the inductive methods of natural science. [(LS). 1984, 12 ff; I, 1981]. Along with Aristotle, Rudolf Carnap, Wilfrid Sellars, and many others, Rescher has consistently asserted that the first principles and the methodology of natural science cannot be justified or established in any non-circular way by explicit and direct appeal to those very principles in question. [Sellars, 1968, 101-15, and Carnap 1980; See also Rescher’s response to Bonjour’s piece in Sosa 1979 and in Rescher 1982. 25ff. See also Bonjour 1985 222 ff]. And yet, for Rescher, there must be some solution to the problem of induction because, contrary to what the skeptic suggests, we obviously have much knowledge of the physical world, thanks to the use of inductive methods. How else can we explain successfully the remarkable predictive and applicative success of so many beliefs acquired by appeal to inductive methodology? We know, for example, that atoms exist and this latter bit of knowledge depends on a good deal of inductive inference. Nor can we establish the validity of inductive methods a priori, as some have erroneously suggested. (Bonjour 1985, 1998). Rather, in the end they can, and should be, established pragmatically, that is, by directly seeing whether, when adopted, they lead ultimately to the observed satisfaction of the primary goals of science. If they do, while this pragmatic form of justification may indeed be empirical in a general way, it is certainly not scientific in the sense of proceeding directly from standard test-

5

RESCHER STUDIES

ing and confirming to whether the primary cognitive goal has been satisfied. Like other pragmatists, Rescher is often more inclined to talk about the vindication or the validity rather than the justification of scientific induction.[(I) ch.3] The Humean skeptic, incidentally, may demand more justification for inductive reasoning and first principles. But Rescher’s consistent response is that, for various reasons, such a demand roots in a sterile Cartesianism feeding upon a faulty argument whose conclusion is that all beliefs about the world are truly doubtful and therefore in need of justification. (1981, 1984, 12ff.; 1978, 175-84.; 2000, 2001, 2003). Once this Cartesian skeptical conclusion is eliminated, we need not worry about the charge of there being no non-circular defense of induction or of basic principles because these basic beliefs will not be the product of beliefs directly justified by appeal to the methods of induction. We can simply start the system with beliefs free of all real or honest doubt, and out of that set of initial credibilities build non-circularly the edifice of natural science. We call those initial credibilities the deliverances of common sense and, as long as they are not honestly doubted, we have no choice but to regard them as certain and without any need for conscious justification in terms of other known propositions. Other philosophers have opined that pragmatic justification ‘obviously’ does not work as a non-circular way of answering legitimate questions about the adequacy of induction or of first principles. After all, some say, claims about the pragmatic adequacy of any system of beliefs are empirical claims themselves which can only be justified on inductive grounds. Rescher has replied, however, that there is nothing at all viciously circular about observing whether one’s methods of testing and confirmation, and the assumptions upon which they rest, lead generally to reliable predictions of sensory experience. [ Apart from his general defense of induction in (MP) and (I), see Rescher’s remarks in Sosa (PNR) 174; in Almeder (PR) 25 ff. ; and his response to Bonjour in both (SEK ) 222 ff. and also in (DPR).Ch.6. For a view similar to Rescher’s see also Wilfrid Sellars, (SM) 101, and the last page of Rudolf Carnap’s “Empiricism, Semantics, and Ontology. 1959”]. Along with other pragmatists, Rescher has argued more recently that all Humean criticisms of induction amount to blaming induc-

6

LIMITS OF SCIENCE, REALISM, AND IDEALISM

tion for not being deduction; and that is by implication to advance an indefensible Cartesian a priorism. [See (RP), (CP), and (PR)]. We will return to this discussion later. If Rescher is right on this way of justifying basic beliefs and inductive methodology, and arguably he is, there will be a way to defend the epistemological foundations of science without falling into rank subjectivism or skepticism based upon an infinite regress, while simultaneously avoiding vicious circularity and any appeal to a purely a priori common sense defense of foundational beliefs in natural science. Accordingly, while natural science has certain limits in that its foundational beliefs and methods cannot be established non-circularly by direct appeal to propositions established in natural science, nevertheless, it will not follow from all that that science is limited in that its beliefs at any level are based upon beliefs which ultimately cannot be objectively vindicated. The justification of the inductive methodology rests on the applicative fruits and precise predictions the method ultimately provides us when in fact we cannot plausible be held accountable for that success in prediction and application of our beliefs. But that scarcely implies that the inductive method cannot fail us sometimes by providing us with systemically confirmed beliefs that we come to reject for good reason at a later date. Induction is generally reliable, but that is not to say that confirmation under the inductive methodology will give us the truth every time. [(LS) 1984.78-79)] The attractiveness of Rescher’s pragmatic reply that the Humean problem is rooted in a sterile and arbitrary Cartesianism, and thus that the reply liberates pragmatists such as Rescher from the charge of begging the question against the Humean argument certainly seem a convincing way of furthering the vindication and validation of induction in the face of Hume’s argument. [See also (MP) 1977. 175-184]. At any rate, Rescher is certainly no naturalized epistemologist singing unreservedly the praises of natural science and announcing the death of philosophy or traditional epistemology. His saying that there are some answerable questions about the world that science cannot answer and that natural science cannot in any case provide a non-circular inductive justification of the methodology of natural science differs, of course, from saying that there will always be unanswered questions in natural science, or questions that science can answer but will not, for various reasons, succeed at doing unto eternity.

7

RESCHER STUDIES

Naturally, again, as is evident in his adoption of fallibilism, natural science is limited in that it cannot provide us with logically certain beliefs, rather than highly likely estimates or approximations of the truth about the physical world. But this only means that all knowledge about the physical world is defeasible; and there is a world of difference between defeasible knowledge and no knowledge at all [(MP) 1977. 201ff; (S) 1978. Ch.4]. More on this shortly.

4. THE LIMITS OF SCIENTIFIC METHODOLOGY In Methodological Pragmatism, (1977b) and elsewhere, Rescher takes a cue from C. S. Peirce and argues against another skeptical argument about the possibility of scientific knowledge. In so doing, he points to other limits of natural science without abandoning the claim that we have knowledge in science about the external world. The skeptical thesis in question consists in avowing that the methodology of natural science can fail in any given case to provide us with knowledge of the world. Some robustly confirmed theories or beliefs established under inductive methodology have turned out to be false, and so there is inductive evidence that they will continue to do so. For this reason, the real probability of error attending any claim offered by inductive methodology more than suggests that there is no attainable knowledge of the world under the method of the natural sciences. We need only admit that often in the past some of our most cherished and robustly confirmed beliefs have turned out to be disturbingly and surprisingly false. This argument itself, incidentally, is an inductive argument based upon the frequency of past failures among robustly confirmed beliefs in natural science. Rescher, in response, argues that while thesis pragmatism is unacceptable for such reasons and, if followed, would support the skeptic’s claims, nevertheless methodological pragmatism is sound. The latter amounts to asserting that the general methods of the natural sciences provide knowledge. They are selected-out by nature because they tend to be generally reliable in producing true beliefs in the long run, although, to be sure, in any given case the methods may fail us. In sum, and as we noted above, Rescher claims along with Peirce, that the validity of inductive inference rests

8

LIMITS OF SCIENCE, REALISM, AND IDEALISM

on the fact that if we follow inductive methods carefully, we are likely in the long run to hit the truth more so than if we were to follow any other method, and that is the reason why evolution and natural selection favors the use of the scientific method . [(MP) 1977b, chapters 3 and 4; See also Almeder 1982. ff]. This means that while highly confirmed beliefs in the history of science may subsequently be found false, the inductive method (or the hypothetico-deductive method) is, with suitable provisos, generally reliable most of the time. So, for the skeptic who argues that truth-value revision in natural science shows that the inductive methods fails to provide knowledge in any given case, Rescher’s reply is that there is no justification for such a strong skepticism about the failure of scientific methodology. It is still the methodology that turns up beliefs most likely to be true and, more often than not in the long run, what is most likely to be true is true.[(MP) 1977b, 66ff, 99ff, 189, 197-202]. So, even though science may be limited in that not all robustly confirmed beliefs in science are demonstrably true at any given time, nevertheless, neither can it be the case that none, or very few, are true. And if the skeptic urges that knowledge is a function of certainty (of the Cartesian sort) Rescher’s response is again that such a criterion is too strong relative to what our deepest intuitions reveal about what we all know. Because of the reflections just noted, Rescher has frequently said that owing to the fallible deliverances of the inductive method on any specific claim that we can never be sure with regard to any particular claim that it is the very truth itself rather than our best estimation [(SR),(LS) 79, 84, (MP) 12ff and EI 236-7]. In fact in the earlier works, he often characterizes confirmed beliefs in science merely as estimates of the truth, that the truth in fact is never reached in science, and that therefore realism in science is a matter of intent rather than accomplishment that there is no difference between what we think to be so, and what is so (EI 223). For example, in the Limits of Science [(LS) 1984)] he says: Our science, as it stands here and now, does not present the real truth. The best it can do is to provide us with a tentative and provisional estimate of it (77-78).The standards of scientific acceptability do not and can not assure actual, or indeed, even probable or approximate truth.(79) In the same place he also goes on to claim that the history of science teaches us ( like the Preface Paradox) that “we know, or must presume, that ( at the synoptic level) there are errors though we certainly cannot say

9

RESCHER STUDIES

where and how they arise” ( 83). As we shall see, Rescher remains equally firm in later publications that science only provides us with estimates of truth, that we can never say with certainty that any particular proposition confirmed in science is true. How he ends up a scientific realist with this thesis in hand is by an appeal to what he ultimately calls ‘myopic’ realism. More on that soon enough. 5. SCIENTIFIC PROGRESS CONSTRAINTS

IN

GENERAL,

AND

ECONOMIC

Rescher has also argued that the number of nontrivial empirically answerable questions in natural science is limitless. No matter how many questions we can (and do) answer in natural science, there has always been a large number of unanswered questions and, indeed, typically, the answering of questions produces more questions. Hence, if one believes in induction, the future should be like the past in that there will always be interesting empirical questions to answer. In short, for Rescher, unto eternity scientific progress will continue by way of answering more and more questions in time. Hence, natural science is limited in not being able to answer all theoretically answerable questions. By implication, there is no substance to the claim that natural science could come to some end in answering all legitimately answerable questions. Those questions are indefinitely many. Interestingly enough, however, while this thesis implies that scientific progress will continue indefinitely long ( because there will always be nontrivial scientifically answerable questions) the progress that science takes will be drastically limited as time goes on such that, for all practical purposes, science will come to an end although not literally so. Let me explain. In Scientific Progress (1977a) and elsewhere, Rescher has urged that even though natural science will progress indefinitely long, there will be a logarithmic retardation in scientific progress. Such retardation will occur because there will be an ever-increasing economic cost of scientific research combined with the need to finance other human needs competing with scientific progress in a world of finite and decreasing resources and productivity (even in the presence of advances in technology). So, for Rescher, although scientific progress will enter into a stage of progressive deceleration, still, given an indefinite future, scientific progress will never stop in some final theory of the empirical world. In defending this thesis,

10

LIMITS OF SCIENCE, REALISM, AND IDEALISM

Rescher claims some dependence by way of inspiration on the important work done by Peirce on the economics of research and, of course, adds a few points Peirce left out. Here briefly is his basic argument that natural science is limited in crucial ways by economic forces. Given Planck’s principle that every advance in science makes the next advance more difficult and requires a corresponding increase in effort, ever-larger demands are made of the researchers in science as it progresses. If we combine this conclusion with the thesis that real qualitative progress in science is parasitic upon advances in technology, and that these advances in technology are in turn ever more costly, then we shall need to conclude that scientific progress will become progressively slower because experiments will become increasingly more difficult to conduct for being ever more expensive in a world of finite and decreasing resources. So, if we factor into our understanding of scientific progress the cold and hard economic fact of increasing cost for technology and decreasing resources and productivity, then a logarithmic retardation in scientific progress is inevitable. Thus, as Rescher sees it, while major scientific discoveries will never come to a complete stop (because new questions will always emerge with each discovery), the time between significant discoveries grows larger with the passage of time. In providing an analogy for this, Rescher asks us to consider such progress similar to a person’s entering a Borgesian library but with a card that entitles the bearer to take out one book the first week, one for the second two weeks, one for the following three weeks, and so on, with increasingly long periods of delay unto eternity. The library has an unlimited, or inexhaustible, number of books in it; every time somebody takes out a book somebody adds, we may suppose, other books to it. In this model, our ever-expanding knowledge grows at an ever-decreasing rate. Scientific revolutions will go on forever, but the time gap between them moves toward infinity and carries with it a substantial slowdown in science as an activity.[(SP) 1977a , 230]. In discussing the economics of deceleration in scientific progress, Rescher considers the possibility that, as technology develops, the cost of progress will become less expensive rather than more expensive. He argues that while this prospect is certainly feasible, it is not realized in practice; and hence a non-linear cost increase is required to maintain a constant pace of progress in science, at least into the indefinite future. Further, that there could not be a finite number of empirically answerable questions follows, Rescher argues in the same place, from what he terms the “Kant-

11

RESCHER STUDIES

proliferation effect” (1977a, 245). This is to say, along with Kant, that unending progressiveness follows from the very real phenomenon that in the course of answering old questions we constantly come to pose new ones. Thus, no matter how much science progresses, there will still be questions that need answers. By implication, there will be questions that we will never, in fact rather than in principle, be able to answer, no matter how sophisticated we become in the practice of natural science. Moreover, even in the infinite long run, these questions will remain unanswered because answering them depends upon a greater concurrent commitment of natural resources than will ever be marshaled at any time in a zero growth world: They involve interaction with nature on a scale so vast that the resources needed for their realization remain outside our economic reach in a world of finite resource- availability. [(SP) 1977a, 250)]. Let us turn now to Rescher’s views on Realism and Idealism. 6. SCIENTIFIC REALISM, IDEALISM, AND INSTRUMENTALISM Realism, like idealism and instrumentalism, means different things to different philosophers. So, in the interest of avoiding as much unnecessary confusion as possible, let me briefly characterize these items as they function classically, and then locate Rescher’s position relative to them Classical realists believe that (a) there is an external world, that is to say, a world of objects whose existence and some of whose properties are neither logically nor causally dependent for their existence on any mental activities of human beings. (b) some of our beliefs about that world are, even if somewhat incomplete at any given time, correct descriptions; and (c) we can justifiably determine and say which of those beliefs, including our theoretical beliefs in natural science, are in fact correct descriptions.

12

LIMITS OF SCIENCE, REALISM, AND IDEALISM

Classical scientific realism shares with classical realism acceptance of (a) through (c). But what distinguishes scientific realism from classical realism is that scientific realism simply extends classical realism to include explicitly beliefs asserting the existence of unobservable theoretical entities postulated to exist by empirically adequate theories. The main alternatives to classical realism are non-realism and antirealism. Non-realists are agnostic about the existence of an external world or about our ability to know anything about such a world, were it to exist. They allow that the world of our experience may or may not satisfy conditions (a) through (c), but they invariably insist that we cannot know that all three of these conditions hold. Moreover, scientific non-realists argue that the success of scientific theories does not require acceptance of (a) through ( c) as true particularly of theoretical entities. The only interesting question for scientific non-realists, is whether scientific theories work as effective instruments by allowing us to make reliable predictions and application to phenomenal experience, and for that, they say, we only need standard confirmation theory. Requiring anything more of a scientific theory is philosophically contentious. These agnostics usually see themselves as instrumentalists epistemologically liberated from the need to claim that successful theories are successful for the reason that they accurately describe the external world, its entities, and its properties. They will claim that successful theories in the past were quite successful in offering precise predictions and applicative success in various contexts, but were later rejected as false. Anti realists, whether classical or scientific, are atheists about our ability the show that (a) through (c) holds. Some have urged, along with both the early Richard Rorty and the later Nelson Goodman, that (a) is unacceptable because all properties of the world are linguistic in nature, and they then go on to dismiss (b) and (c) as obviously indefensible [Rorty, 2000. 1-31, and Goodman, (WOW) 1978.] Whereas scientific non-realists willingly concede that our best scientific theories may, for all we know, correctly describe the external world and its theoretical entities, scientific anti-realists reject that concession. Classical anti-realists are typically phenomenalists, restricting reality to the systematized content of our conscious experience. Some scientific anti-realists may allow that observable physical objects exist in addition to our experience of them, but emphatically deny that theoretical entities exist. Some non-realists are professedly agnostic and instrumentalists with regard to the existence of theoretical entities but are common sense realists relative to the existence of observable middle sized physical objects and their observable properties. [Van Fraas-

13

RESCHER STUDIES

sen, (SI) 1980]. We may, if we like, regard anti- realists, both classical and scientific, as rejecting any or all of the conditions (a) through (c). They tend to see themselves as pure instrumentalists regarding rational belief, simply for having empirically well-confirmed beliefs no one of which happens to be that some of the conditions (a) through (c) are true in the ordinary sense of ‘true’. As typical instrumentalists would have it, then, scientific theories provide us with no demonstrable knowledge of a real external world, or of the existence of theoretical entities in it, unless of course we simply define knowledge in terms of well-confirmed beliefs (however acquired) about what we take to be the external world. Given usual instrumentalist arguments, natural science is therefore crucially limited, or fails completely, in its claims to be describing correctly properties and entities in an external world. So where does Rescher’s metaphysics and epistemology fit into all this? As early as 1981 in Empirical Inquiry and in Induction, and in all his later discussions on realism, Rescher consistently defends (a) and (b) and will accept (a) and (b) for theoretical claims in natural science. In the end, for matters falling outside of science but under the rubric of common sense belief, he adopts (a) through (c), sounding very much like a common sense realist. But he hesitates as to accept (c) for propositions asserting the existence of theoretical entities, entities such as atoms, neutrons, psions and mesons, even though he admits that some theoretical claims certainly can be, and are, true in any empirically adequate theory.[(SR)71-73] In matters of common sense we are certain as we can be about some things because we collectively cannot imagine being wrong about those basic common sense beliefs [(CS) ch.1]. But, as the history of science will attest, we can easily imagine being wrong in our claims about theoretical entities posited to exist in theoretical science. In theoretical science we have only approximative estimations of truth, and in that activity there will always be some real likelihood of error no matter how small or contextually irrelevant to our predictive and applicative success. Let me indicate why Rescher does all this, and what it has to do with his idealism, before concluding this little foray into his metaphysics and epistemology. Rescher’s basic response to the instrumentalist claim that natural science is limited in its effort to describe the real world of theoretical entities is that no such limit exists. Rather, natural science provides a correct picture of the world even if its correct descriptions are partial and always in some important sense incomplete and more in the nature of approximative estimations rather than correct descriptions. Very high degrees of eviden-

14

LIMITS OF SCIENCE, REALISM, AND IDEALISM

tial probability will be sufficient for the sort of approximative estimations of truth we need in order to be fully justified in asserting that some theoretical claim is true. We cannot do any better than that, and nothing more is required. Demanding more when more cannot be attained is folly, and implies a sterile Cartesianism More specifically, in arguing for (a) Rescher asserts that the belief that the existence of an external world is not, as pure instrumentalsts will claim, merely a useful thought fiction for explaining phenomenal experience. Rescher has argued that in arguing for (a) we should postulate, or posit, rather than profess to inductively or empirically prove the existence of the external world. (I, 260-66). He often sees such a posit or postulate as a precondition for empirical inquiry, so it functions as a regulative belief allowing for the attainment of knowledge. Moreover, the implication of such a postulate or posit at the level of prediction and applicative success more than vindicates condition (a), as there is no good reason against believing in (a) and a very strong common sense impulse in favor of it, whereas its denial leads to serious troubles, not the least of which is that everything we take our world to be is some basic mental or human construct, and Rescher along with others find this conclusion arguably false and also horrendously counter-intuitive. Because we have no compelling reason to reject belief in physical objects, that belief remains maximally productive. [(SR) 109ff]. The same holds for (b) and (c). Rescher’s basic argument for (b) is that any denial that we have some beliefs about external world that are true or correct descriptions of it flies in the face of common sense, just as does the denial of (c) that we can justifiably determine and say which of those beliefs, including the theoretical ones in science, are correct descriptions. Indeed we have, as a matter of common sense, many readily identifiable and irrefutable general truths about the external world. But Rescher is sometimes keen on advancing in defense of (b) the belief that unless we admit that some of our robustly confirmed beliefs about the external world are in fact correct (true) descriptions of that world, we would have no way of explaining the predictive and applicative success of our scientific theories except by appeal to mystery or miracle.[(SR) 73-79]. So, he seeks to reconcile the view that in natural science all we get are truth estimations and never truth with the fact that success in science requires that at least some of the claims in natural science are true even if we cannot say which assertions in a successful theory are true and which are not. In Empirical Inquiry, for example, he says quite explicitly that the long-term general (or systematic) predictive success of

15

RESCHER STUDIES

scientific theories across the board is evidence of truth implicated somewhere in the theory [(EI) 236-37). Accordingly, as he also says elsewhere, some sentences in successful scientific theories must be true, or have what he calls some ‘kernel’ of truth, even if we cannot pick out those sentences, and even though they are always only partial descriptions. [(LS) 84; (SR) 73] . So even though natural science never gives us the very truth itself in every fully warranted assertion, rather than a fully warranted estimation of the truth, it nonetheless has some true sentences in it. How many, and which ones they are, in any given theory, we cannot say in terms of a reliable decision procedure applied to all our fully warranted beliefs. This last justification helps with the justification of (b) generally and in scientific theories. But it still leaves us in the lurch with regard to (c), if Rescher is to be a classical realist and a classical scientific realist. Do we or do we not have a reliable decision procedure or method for determining which of our beliefs in and out of science are true rather than highly confirmed approximations or estimates of truth that can equally provide us with predictive and applicative success? On the one hand, we have his view that in natural science we never get the truth rather than justifiable estimates or approximations of it. We do not need anything more than that for human knowledge in theoretical science because that is all we can get, and it is enough to provide for predictive and applicative success while admitting that that success cannot obtain unless some of the beliefs in the theory are in fact true even if we cannot decide which beliefs are true rather that confirmed approximations or estimations that are in fact false. On the other hand, we find Rescher affirming that at the level of common sense affirmations outside of theoretical science we have many identifiable common-sense beliefs that are uncontestably true, because they are simply certain. No one can honestly doubt them, and there is no good reason to doubt them. Beliefs such as “I had a biological mother,” “A mature redwood tree is taller than a mature blade of grass,” “ In the year 2008, somebody lived on the planet earth,” “The coast of California is longer than the coast of Delaware” all seem to be paradigmatic of particular physicalistic assertions we certainly know to be true and could not be more justified in asserting to be true rather than approximations or mere estimates of truth. All this seems to confirm that for Rescher, propositions (a) through (c), for all the common sense and pragmatic reasons he gives, hold outside theoretical science; but (a) and (b) and not (c) hold for theoretical claims in natural science. This makes Rescher a classical realist at the level of common sense and empirical knowledge outside of natural science, but not a

16

LIMITS OF SCIENCE, REALISM, AND IDEALISM

classical scientific realist because of the problem of not being able to determine which theoretical claims in natural science are in fact true rather than fully justified estimations or approximations of truth. So, when Rescher characterizes his realism as ‘myopic’, we see him as adopting (a) and (b) but not (c) in natural science. If this construal of Rescher’s realism(s) is basically correct, and if one could argue that common sense beliefs are no different in kind than any other empirical claim about the world in that they too are defeasible even though at any given time we may not be able to imagine being wrong about what they assert, the ‘myopic’ realism he adopts would also turn out to apply to the realm of common sense. “I know that I had a biological mother” may be as certain as certain can be certain; but that still won’t give us the very truth rather than the fully warranted estimation that it is true. One can plausibly argue for the extension of ‘myopic’ realism to the domain of all empirical beliefs outside of, as well as within, natural science, if one can, as I do, regard even common sense beliefs, no matter how true they seem, as empirical assertions that are fundamentally fallible and defeasible. In such a world, one would be more of a classical realist than a classical idealist for adopting (a) and (b) and not (c) although that is still attractive by way of undermining the cultural relativism deriving from a full blown fallibilism that goes too far in seeing that all properties of the world are linguistic in nature. The only downside of ‘myopic’ realism, both in and out of natural science, would be our needing to admit generally that even in the best of our theories, at any given time, we must take seriously that the world may be quite different than we assert it to be in our most robustly confirmed theories. However, that is just to affirm, along with Rescher, that when it comes to saying how the world really is, rather than how we sincerely think it to be, we cannot aspire to anything more than the sort truth estimations or approximations that have allowed us to adapt as well as we have. There is no harm in that as long as our robustly confirmed theories produce those predictive and applicative successes we need for basic adaptation and human happiness. Sure, the world will remain a little more mysterious than if we had been able to adopt (c). But what’s wrong with leaving a little philosophical space for poets, playrights, mystics and an occasional epistemologist? Finally, what has any of this to do with Rescher’s idealism? It depends on how we define ‘idealism’. In Human Knowledge in Idealistic Perspective (1992) Rescher delineates the sort of idealism, conceptual idealism, he endorses while reaping all the benefits of being a realist of the sort he defends as ‘myopic’, and as described above in terms of accepting

17

RESCHER STUDIES

(a),(b) and not ( c). In this venture he avoids adopting any form of idealism affirming that all properties in the world are linguistic in nature and that everything therefore in the world is mental. Rather, he is intent on advancing a form of idealism advancing the idea that the world cannot be understood except through the constructive conceptualizations of the human mind without which we would not even have an idea of the world, and how we conceptualize the world determines crucially what we take it to be. Given this rough and ready characterization of his idealism, Rescher does not see realism and idealism as mutually contradictory stories, and that the ideal strategy in his overall system is to combine the two in terms of what has proven attractive in each.[(HKIP) 125]. Well and good. But a possible problem with the idealism Rescher seeks to advance along side of his ‘myopic’ realism is that it seems not at all controversial because it seems in the end to assert fundamentally the non-controversial point that in order to know anything at all about the external world human minds are necessary. That affirmation seems distinctly removed from the core classical idealisms that deny any real distinction between the mental and the non-mental, and thereby deny that there is an external world as distinct from the world of thought. So, we can characterize Rescher’s realism as it relates to “idealism” in the following manner. If ‘idealism’ refers simply to the belief that human beings are cognitive creatures that have magical mental capacities to create categorial structures with which they can ultimately formulate and classify hypotheses answering certain why questions about their phenomenal experience, and then test those hypotheses in terms of what we should expect of those hypotheses if they are true of our phenomenal experience, then anybody who is a ‘myopic’ realist or even a classical realist in and out of scientific discourse, can also be an idealist. But if ‘idealism’ refers to the belief that we do not know that there is an external world, that all properties of our world are purely linguistic or mental, or that we cannot know that some of our beliefs about an external world are correct descriptions of that world, then one cannot be a ‘myopic’ realist, or a classical realist in any sense and at the same time be an idealist. Further, if ‘idealism’ refers to the belief that we do not know that some of our beliefs are correct descriptions of an external world, then one cannot be both a ‘myopic’ or a classical realist and an idealist. Finally, if ‘idealism’ refers to the belief that (a) and (b) are true but that (c) is not because (c) is essential to any classically defined or acceptable realism, then ‘myopic’ realism’ is in fact idealism. In the end, without caring to defend the claim here, it seems historically accurate to say to say

18

LIMITS OF SCIENCE, REALISM, AND IDEALISM

that classical idealists have consistently denied at least (a) ,(b), and (c). Historically, then, there does not seem to be any room to combine classical idealism and ‘myopic’ realism because ‘myopic’ realism asserts both (a) and (b). A realism that denies (c) and keeps (a) and (b) will certainly be a weak instrumentalism simply for its refusing to read off the nature of the world from the fully justified truth value assignments or estimations we make based on robust confirmation of our beliefs. But, for Rescher’s reasons mentioned above, that is not an instrumentalism that forces us to us to denying either (a) or (b). All in all, given the first definition of ‘idealism’ above Rescher can consistently be the sort of ‘myopic’ realist he has defined, and an idealist. But in that he is not asserting, grace au dieu, that there is an external world all of whose properties are linguistic in nature. That sort of thesis would be incoherent. The only problem is that ‘myopic’ realism seemingly seeks to cure the myopia (and thereby block the possible charge of being a classical idealism at root) by appealing to common sense as a default mechanism for determining which beliefs are true alethically about the world when in fact common sense may be no more privileged than any other source of empirical knowledge. It’s just that common sense beliefs seem more certain and free of any real doubt. That is certainly strong enough grounds for appealing to any common sense belief as a line in any argument we will call sound, but about them we can be as fallible as we are about the existence of theoretical entities. All things swim in the continuum. Still, we know that there is an external world that some of our beliefs correctly describe even if we cannot on any given occasion distinguish those beliefs from any other robustly confirmed belief about the external world.. 7. APPRAISAL By way of brief assessment of Rescher’s above-stated views on the limits of natural science, it is difficult to disagree with his firm rejection of scientism. Nor does this rejection of what has come to be known as ‘the replacement thesis’ imply an abandonment of a rich naturalism, unless one happens to define stipulatively a rich naturalism in terms of the replacement thesis. Science can still be the most privileged of methods for understanding the nature of physical objects and the regularities governing their activities in this world even if there are some legitimate questions about this world that are not scientific questions. Anything else, as Rescher has argued, would be a self-defeating and incoherent form of naturalism.

19

RESCHER STUDIES

Secondly, with regard to Rescher’s distinctive views on the economic limits on scientific progress owing to the ever increasing cost of technology, it is, again, equally difficult to disagree with the impressive contours of the thesis. Admittedly, in a world of finite and decreasing resources where global population tends to increase exponentially, there will inevitably be a time when the technological cost of scientific progress will become a real impediment to robust scientific progress, and that in all likelihood there will be a logarithmic retardation of scientific progress for that reason. On this score, however, there may well be a question about whether global population will continue to increase exponentially into the indefinite future. Some demographers, for example, now predict that the world population will peak soon and then enter into a period of indefinite decline [ See Nicholas Eberstadt “World Population Implosion ? The Public Interest. No.129, ( Fall, 1997) 3-22.]. That the world’s population has in the past tended to increase exponentially may well be the result of certain factors no longer present to the world population. This is an empirical question, of course, and best left to demographers. Even so, it is worth raising the issue because if contemporary demographers are correct in their estimations, then the cost of new technology necessary for robust and continuing progress may not become so high as to induce the logarithmic retardation Rescher predicts. The only question then is whether the cost of technology for doing science would continue to escalate for the simple reason that, owing to increasing scarcity of natural resources, global productivity would drop dramatically enough in the presence of escalating costs for other human needs, even in a world of decreasing population. Seems plausible enough. But it is nonetheless within the realm of possibility that a cost-effective technology could occur for the safe use of atomic or solar energy. If so, then, even though such energy would not be an unlimited resource, it might take a very long time for the logarithmic retardation in scientific progress to emerge. But, presumably, Rescher’s point is that the logic of logarithmic retardation in scientific progress ties tightly to the fact that in this world, natural resources for economic productivity are finite and decreasing and the demand on them is at least constant if not escalating. In such a world the only way to stop the progressive deceleration in scientific progress would be to guarantee that demand for natural resources extend only to those resources that are basically renewable in a way that establishes an equilibrium between the demand and the supply without threatening the capacity for renewal itself to meet existing demand.

20

LIMITS OF SCIENCE, REALISM, AND IDEALISM

Thirdly, on the question of scientific realism, Rescher’s earlier views seem a bit different from his later views. In the earlier works (including LS) his position seems to be one of claiming that scientific theories aspire or intend to succeed in correctly describing the external world but that, as Laudan and others have argued, the history of science justifies the view that science never succeeds, and never will, succeed in correctly describing the external world. That is why, for example, in LS Rescher’s position seems to end up more classically idealistic in endorsing the view that natural science does not in fact ever succeed in correctly describing an external world. But later in Scientific Realism (and in Methodological Pragmatism) we hear him saying, as we noted above, that the success of natural science under the methods of the natural sciences indicates that at some level or other there must be some truth in the system. He says, for example: While action on false belief . . . can on occasion succeed—due to chance or good luck or kindly fate or whatever—it cannot do so systematically, the ways of the world being as they are it is effectively impossible that success should crown the products of systematically error producing cognitive procedures. Inquiry procedures which systematically underwrite success-conducive theses thus deserve to be credited with a significant measure of rational warrant. Put somewhat differently, Rescher ultimately adopts the argument to the effect that the overall and systematic success of scientific theories as predictive devices cannot plausibly be explained in the absence of believing that there has to be some “kernel of truth” to them. As we saw above, this point Rescher emphasizes in Scientific Realism and thus comes to refine his “approximationist realism” as a position asserting that we cannot account for the systematic or overall predictive success of some of our theories without assuming that some of the claims made in the theories are accurate descriptions of the external world even if those descriptions are approximate or incomplete, and even if some of the claims made in the theories are complete failures as successful descriptions. Moreover, in Scientific Realism he argued (as we also saw above) that just as when a theory fails us we are not in a position (being holists of sorts) to say which sentences failed at describing the world, so too when a theory works well, we are equally at a loss to point out which of the beliefs are the successful describers of fact. Naturally, the success of Rescher’s position depends on

21

RESCHER STUDIES

showing that all proposed alternative explanations for the general predictive (or ‘applicative’) success of the theories or hypotheses in question cannot be as plausibly explained in any other way than by asserting that some of the sentences in any successful theory must succeed in correctly describing the world even though we may not be able to pick out the ones that do from the ones that do not. So, scientific theories may all be false and ultimately go the way of the geocentric hypothesis, but even then some of the claims made in them must succeed in correctly describing the external world. Notice, then, that Rescher’s view here is more realistic than the “realism of intent” or the “realism of aspiration” he defended in The Limits of Science and elsewhere. Notice further that the position, as so construed, is consistent with perpetual revolutions in natural science. It is also consistent with the claim that “all theories are born dead” because, as theories, they will be replaced by successor theories. Finally, notice that in this final and refined picture, according to Rescher, we do not have an effective decision procedure for determining which sentences of our theories are doing the successful describing and hence carrying the success of the theory. Of course, not all of the propositions in a successful theory can be true. If they were all true, then scientific theories would never get replaced. And if none of them were true, then we would have no way to explain the predictive success of such theories. Successful theories require true propositions, but eternal revolution in science requires a good deal of false but robustly confirmed beliefs. As so stated, then, Rescher’s scientific realism is, (pending refutation of instrumentalists’ alternative explanations of success) quite defensible as a statement on the theoretical limits of scientific theory in providing for an understanding of an independently given world. It implies that whatever we take as criteria for truth (presumably robustly confirmability or warranted assertibility) suitably explicated in terms of coherence with initially given data), are sometimes satisfied; but it does not imply, even with a high degree of probability, that any particular proposition one might name as being completely justified is in fact satisfying the alethic or platitudinous definition of truth, although we know that some of them must be doing so. In this it differs, of course, from classical scientific realism because the latter typically asserts that we know, and can say, which of our beliefs about the external world are in fact the ones describing the external world. But if the core of classical realism is the belief that some of our robustly confirmed beliefs about the external world are more or less correct descrip-

22

LIMITS OF SCIENCE, REALISM, AND IDEALISM

tions of that world, and thus that science succeeds at any given time in telling us in some important way how the world really is, then Rescher’s position is surely consistent with the core of classical scientific realism. As so construed, with the suggested provisos, Rescher’s position is an attractive, compelling, and pre-eminently defensible answer among all other proposed answers to the question of the extent to which natural science can and does succeed in some important measure in accurately describing the external world. He calls it ‘myopic’ realism and, as we notes above, it is consistent with that form of idealism noted above that is consistent with (a), (b), and not (c). Fourthly, in discussing Rescher’s views about the foundations of empirical knowledge, we saw above that for Rescher there can be no noncircular inductive justification of foundational beliefs that serve as the basic evidence for other beliefs in the structure of science. Rescher’s claim is essentially that when it comes to justifying basic beliefs, we start with certain presuppositions (or posits or postulates) which serve as evidence for other beliefs; and if what is demonstrated either inductively or deductively under these presuppositions satisfies the purpose for which our methodologies are developed, namely, to successfully predict precisely our sensory experience, then that would be pragmatic justification (or perhaps vindication or validation) of the presuppositions as items of knowledge, even though they themselves are not directly inferred by conscious appeal to other known propositions as their evidence. The central claim, then, is that if the presuppositions were not items of knowledge, they could not in the long run, or systematically, lead to pragmatically successful results. Here again, we would have no way of explaining our success in predicting our sensory experience if the presuppositions upon which we rested the whole system were not as warranted as any result producing human knowledge anywhere else in the system. Is this way of justifying or vindicating basic presuppositions as items of knowledge still an instance of vicious circular reasoning because it appeals to inductive reasoning to establish that certain beliefs have been successful? Some philosophers have claimed as much. [See Bonjour in Sosa (PNR)1979. 228, and Rescher’s reply 174 with Bonjour’s reply to Rescher 250] But Rescher’s basic point is that we cannot plausibly justify first premisses about the world a priori, and we simply must avoid both the infinite regress, and directly viciously circular justifications. All this we do by stopping the regress at basic propositions functioning as presuppositions in

23

RESCHER STUDIES

starting the system, and then indirectly and retroactively justifying the presuppositions as items of knowledge as a result of the empirical adequacy of the system taken as a whole over a long period of time. In this way basic propositions are items of knowledge without being conclusions from premisses known and directly supporting them. They are, then, directly known without justification, but only seen retrospectively as items of knowledge because that would be the best explanation of the success of the system which they initiate. Whether we call this ‘vindicating’ or ‘justifying’ first premisses is immaterial to the claim that the activity is not an example of vicious circularity because no particular proposition established by simple induction actually enters into the justification of the claim that the presuppositions must be true. As we noted earlier, Rescher’s procedure seems similar to Goodman’s and Carnap’s mode of vindicating certain answers that cannot be justified because justification is defined as an activity for authorizing moves within the object language whereas the question of the epistemic status of the presuppositions of the method adopted in the language can only be a matter of vindication in terms of the whole system, and its general empirical adequacy. In this way, first premisses can be know to be items of knowledge even though they are not directly inductively justified although, to be sure, the judgment that the system as a whole is adequate as an inference based upon observation. Doubtless, this issue deserves a deeper discussion and a finer distinction on kinds of inductive inference. But if we are willing to draw a distinction between valid inductive inference within the language of science and valid inductive inferences about the language of science, we need not see vicious circularity in Rescher’s way of providing for the epistemological foundations of scientific knowledge. Finally, in spite of all this sweetness and light, one might plausibly register a contrary, and certainly a minority, belief—even if only to raise the issue for further discussion. It has to do with Rescher’s thesis that scientific progress will continue forever and will never be completed, even in the presence of a logarithmic retardation in progress, because the number of non-trivial empirically answerable questions that can in principle be answered in natural science is inexhaustible. As I noted elsewhere,[1990 (LNS)] Rescher’s argument for this thesis is essentially that since there have always been questions generated by past successes in answering questions in science, then there is good inductive reason to believe that as science continues there will continue to be questions emerging from all the

24

LIMITS OF SCIENCE, REALISM, AND IDEALISM

questions answered in the advances of science. Even if there happens to be a finite number of non-trivial questions at any given time, there will always be more questions to answer at any time in the indefinite future. In fact, it looks as if this is a straightforward and uncomplicated inductive generalization from the past history of science. But this inductive argument for the thesis seems problematic. Let me explain briefly. Rescher’s thesis that the number of non-trivial empirically answerable questions is indefinitely many would be undermined if the number of nontrivial questions that science and technology could answer were finitely many in a world existing indefinitely long, and where more and more nontrivial questions get answered under the methods of the natural sciences and advancing technology. Suppose, for example, that Bertrand Russell’s celebrated chicken had strong evidence that his master-farmer had only enough food to feed him for 750 days. In that case, the chicken should not conclude that because the farmer had fed him for 750 days in the past that therefore the farmer would feed him tomorrow, on day 751. Simple induction from the past is a valid indicator of the future only under certain clear proviso clauses to the effect that there is no good reason to suppose that the conditions warranting the inductive inference will change. So, it seems that the basic inductive inference supporting the belief in the eternal progress of science assumes that the total number of non-trivial questions is not finitely many in a world existing indefinitely long, and where progress is defined in terms of answering more and more non-trivial questions in time. If the total number of such questions were finitely many, then the fact that past progress in natural science always generated more questions than existed before the progress would by no means entail that there will always be such questions to answer in the future. In other words, from the fact that there have always been questions generated in the past by answers generated from science and technology, it does not follow that there will always be such questions in the future unless one already knows or assumes that the total number of non-trivial questions in the indefinite future is not finitely many. Similarly, if Russell’s chicken had not assumed that there would be enough food for the farmer to feed him indefinitely long, then he might have been more prepared for his eventual demise. At this point, one might suggest that the belief that the total number of non-trivial questions at any given time is indefinitely large is no better supported by our experience than is its denial. Indeed, what is the inductive evidence that there are indefinitely many non-trivial empirically answerable questions? Again, it

25

RESCHER STUDIES

cannot derive simply from the fact that in the past there have always been questions generated by answers provided. 8. CONCLUSION Suitably qualified in small ways, Rescher’s central views on the knowledge, the limits of natural science, and his general picture about realism and idealism are impressive, quite defensible, and even magisterial. Whether we are talking about the limits placed on natural science by the progressive and inevitable lack of economic resources to conduct science in a vigorous fashion, or whether we are talking about the limits of science to deliver up in any non-controversial ways the propositions that in fact succeed in correctly describing the external world in some fashion, or whether we are talking about the limits of science by way of providing a justification for its own basic beliefs, or whether we are talking about the limits accruing to the fact that there must be some interesting questions about the world that are not scientific questions, or whether we are talking about the limits science will inevitably have in seeking to solve pressing social issues because of the increasing complexity of technology and scientific answers, we have a persuasive panoramic picture presented of the limits of science along with an unbounded enthusiasm for what science has done and continues to do by way of providing for specific predictions of our sensory experience in a way that counters classical idealism by allowing for belief in and knowledge of an external world while granting full force to a fallibilism that may well prevent us from being able to say which of our beliefs is true while showing that some of them are indeed true. Apart from a more fallibilist view about the deliverances of common sense, for what more could one reasonably ask?

REFERENCES Almeder, Robert, Blind Realism (BR) (Lanham, MD.: Rowman and Littlefield, 1992). ———, Praxis and Reason (PR) (Washington, D.C.: University Press of America, 1982).

26

LIMITS OF SCIENCE, REALISM, AND IDEALISM

———, Harmless Naturalism (HN) (Chicago, Ill.: Open Court Publishers 1998). ———, “The Limits of Natural Science,”(LNS) in Martin Carrier et al. (eds.), Science at Century’s End: Philosophical Questions on the Progress and Limits of Natural Science (Pittsburgh, PA: University of Pittsburgh Press, 1990). Bonjour, Laurence, The Structure of Empirical Knowledge.(SEK) (Cambridge, Mass: Harvard University Press, 1985). ———, In Defense of Pure Reason (DPR) (Cambridge, England: Cambridge University Press, 1998). Goodman, Nelson, Ways of Worldmaking (WOW) (Indianapolis, Indiana: Hackett Publishers, 1978). Rescher, Nicholas, The Coherence Theory of Truth (CTT) (Oxford, England: Clarendon Press, 1973). ———,Cognitive Systematization(CS) (Oxford, England: Basil Blackwell, 1979). ———, Empirical Inquiry (EI) (Totowa, N.J.: Rowman and Littlefield, 1981.) ———, The Limits of Science (LS) (Berkeley, California: University of California Press, 1984). ———, Induction (I) (Oxford, England: Basil Blackwell, 1981). ———, Scientific Progress (SP) (Oxford, England: Basil Blackwell, 1977). ———, Methodological Pragmatism (MP) (Oxford: 1977). ———, Skepticism. (S) (Oxford, England: Basil Blackwell, 1978).

27

RESCHER STUDIES

———, Scientific Realism (SR) (Dordrecht, Holland: D.Riedel Publishing Company, 1987). ———, Human Knowledge in Idealistic Perspective (HKIP) (Princeton, N.J.: Princeton University Press, 1992). ———, Realistic Pragmatism (RP) (Albany, NY: SUNY Press, 2000). ———, Cognitive Pragmatism (CP) (Pittsburgh, PA: University of Pittsburgh Press, 2001). ———, Pragmatic Realism (PR) (Pittsburgh, PA: University of Pittsburgh Press, 2003). Rorty, Richard, “Universality and Truth,” in Robert Brandom (ed.), Rorty and His Critics (RAC) (Oxford, England: Oxford Publishers, 2000). Sellars, Wilfrid, Science and Metaphysics (SM) (New York, NY: Humanities Press, 1968). Sosa, Ernest (ed.), The Philosophy of Nicholas Rescher (PNR) (Dordrecht, Holland: Reidel Publishers, 1979). van Fraassen, Bas, The Scientific Image (Oxford, England: Clarendon Press, 1980).

28

On Possibilities and Thought Experiments Diderik Batens 1. AIM OF THIS PAPER

T

his paper concerns two related recent books by Nicholas Rescher, Imagining Irreality on possibilities and What If? on thought experiments—see [23] and [24].1 The books are related because a theory on thought experiments presupposes a theory on possibilities. The effect of this and of the fact that they have a wider range of application than thought experiments will be that possibilities will receive the bulk of the attention. The books are also very different. A view on possibilities is unavoidably highly theoretical and sophisticated. It involves a host of arguments that pertain to ontology, epistemology, and language. A theory of thought experiments, to the contrary, mainly concerns a matter of reasoning. Everyone competent in the domain to which the thought experiment belongs is able to assess its soundness. While the resulting theory may be sophisticated, its subject matter belongs, from a philosopher’s point of view, to a practical realm. On both subjects, Rescher offers a theory that is systematic and embracing. With respect to possibilities, he takes a straight linguistic stand, heavily attacking the reifying views that came into fashion at the end of the twentieth century. The book on thought experiments concentrates on their use in philosophy without avoiding other domains or general characteristics. It presents a comprehensive theory, illustrated by many examples, that deals with an impressive number of aspects. These aspects are handled in a way that does right to sophistication but never gives in on the perspicuity that is so typical for Rescher. A particular strength of both theories is that they are embedded into a larger view that was developed against a broad background in the history of philosophy, present-day systematic philosophy, as well as the contemporary sciences. The reader should not expect an attack on Rescher’s position. Mainly through the influence of Leo Apostel, the nominalistic and conceptualistic critique of platonistic constructions and the constructivist critique of mathematical realism played a pivotal role in my philosophical upbringing. Moreover, from the first books of Rescher’s, [17] and [18], that came into my hands back in the sixties, the clarity and systematicity of the exposition

RESCHER STUDIES

convinced me whenever I had no previously formed opposing views and sometimes even then. Incidentally, this persuaded me to spend 1972-73 at Pitt rather than at UCLA. So the present contribution will, apart from an expository part, consist on the one hand of some proposed elaborations, especially of two technical points, and on the other hand of some questions concerning points where I am in doubt about Rescher’s precise stand and of some suggestions for further research. 2. POSSIBILITIES Rescher is rather tolerant with respect to existence. He allows for existence in different realms: physical, mathematical, sensory, etc. He sees being as always realm-correlative and contextualized; a colour exists as a sensory object, not as a mathematical entity. But this is not the end of it. Existence is governed by a recursive definition. Things may exist, in a realm-correlative and contextualized manner, because they play an active role in the real world. They may also exist because they must be invoked in order to arrive at a satisfactory account of something that is already taken to exist. For all his tolerance, Rescher drastically rejects the idea of granting existence to possible worlds and merely possible objects. In order to make his position clear, it seems best to start from his view on facts. A ‘truth’ is the representation of a fact through some language. Facts are obviously not linguistic entities. Yet, there is a relation. “Anything that is correctly statable in a possible language presents a fact.” ([24, p. 112]). Unlike facts, objects, and the actual world, possible facts, possible objects, and possible worlds are intellectual constructions. They neither exist (pure and simple) nor subsist. The reason for denying them existence is double. First, there is no good reason to grant them existence, whence Ockham’s razor applies. Next, they lead one into trouble of sorts. Unrealized possibilities are projected by minds. To be more specific, they are projected from within the actual world. Our view of the actual world is the starting point. From there possibilities are obtained by assumptions and suppositions, and also by stories or scenarios. Supposing may bring us a long way from reality. By adding suppositions on suppositions one may reach possible situations that are very different from the starting point. This way of proceeding is not without danger, as we shall see in Section 3. One may consider a counterfactual concerning a real person, say Mary. What would have happened if Mary had stayed home yesterday night, in-

30

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

stead of going to the theatre? The object involved in the considered possibility clearly exists. The considered possibility itself is false—it does not become true by our considering it. One may also consider a possibility that does not pertain to a real object, but to a fictional character, to a mythological animal, to a Martian, etc. In such cases, the considered possibility is actually false and, moreover, the object to which it pertains does not exist. It has been claimed that understanding our thinking about possibilities requires possible worlds populated with possible objects and that, as a result, those worlds need to be considered as existing. Rescher disagrees. In view of his recursive approach to existence, he needs to show that those possible worlds are not so required. His alternative is a verbal or linguistic approach to possibilities. Possibilities are obtained by varying on truth as described in our present language. They are obtained by introducing assumptions or suppositions, or by relying on our imagination to create fictions and stories. On this view, possible worlds and merely possible objects are, just like other possibilities, creations of the mind. This status is clearly sufficient to understand their functioning, which concerns our thinking about and reasoning in terms of possibilities. This does not show, of course, that no possible worlds and merely possible objects exist outside of our world. Yet it shows that this existence is not required. As may be expected, Rescher does not leave the matter there. He wants to show that the linguistic approach to possibility is the better one. To do so, he presents several objections to a realistic stand on possibilities. I shall consider the two objections that I take to be most central. The first is that “to be is to function on one’s own, thought independent footing as a constituent of a realm” ([23, p. 26]). Rescher does not consider this argument at great length and justly so. It is obvious enough that the possibilities we are able to contemplate are obtained by varying on our theories about the world, by which I obviously mean ‘our world’. So possible worlds and the objects that populate them miss the required quality. This is easily supplemented with arguments from history. In every period, the described possibilities are variations on people’s views about the world. This holds for myths, scientific thinking, as well as fictions and stories. Note that a considerable number of people have engaged in activities that led to the introduction of possibilities. They were trying really hard. This does not mean that our conceptual system delineates what we may consider as possible, or that an era’s conceptual system delineates what may be considered as possible in that era. Humans are able to vary on the concepts of their culture even if, as history teaches, this particular ability is

31

RESCHER STUDIES

rather limited. Even if unreal possible worlds existed, people would clearly have no access to them. The limits of the possibilities one may consider are determined by the logico-conceptual systems one is able to imagine in view of the logico-conceptual systems one knows, which is one’s own and perhaps some past ones. The second and in a sense more dramatic objection is related to the schematic character of fictional objects. Real objects are complex. They are much more complex than any present description of them. They have more properties than they will ever manifest—this includes dispositional properties. Rescher even argues that they are “endlessly complex”, referring to points of view, to possible functions, and to the complexity that arises at the level of micro-physics. Luckily there are easy means to identify actual objects, for example pointing. The situation for possible objects is very different. As they obviously cannot be pointed to, any attempt to identify them will have to rely on a description. But any description will always be far too incomplete to allow for identification. This holds for things like the possible elephant in that corner. It also holds for such descriptions as “the person just like him (pointing to a black haired person), except that he is red haired”. The latter is schematic because one cannot change the colour of the person’s hair without changing many other features, for example his or her genes, the genes of the person’s parents, etc. Moreover, many different changes will lead to a person that meets the description. Put differently, even if there were possible objects, any description would fit a multitude of them and so would not identify any of them. When one refers to a possible object, one refers to something schematic, not to anything like an object. The same holds even more for worlds. Rescher illustrates his claims by means of objects and worlds that are invoked by philosophers as well as by means of characters from stage plays and novels. For an already convinced reader, like me, it may sometimes sound slightly repetitive. But then the aim of the book is to convince the Platonists. Note that Rescher does not object to statements about ‘possible objects’ and ‘possible worlds’. Such statements do not refer to objects or worlds, but only purport to do so. Their intentional objects are fictional entities, constructs of the mind, not anything like real objects or the real world. The functioning of those entities is different from that of real objects or the real world. For example, there is no point in trying to find out their properties over and above their description. So their possibility is de dicto, not de re. Positing the existence or quasi-existence of possible worlds that differ from

32

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

the real world and of objects that occur only in those possible worlds is useless and inappropriate. It is useless because fictional objects and worlds can be understood without granting them existence. It is inappropriate because fictional objects and worlds are schematic whereas the entities posited to exist are concrete. On the linguistic view, the criterion for (logical) possibility is “internal logico-conceptual coherence” ([24, p. 140]). That something is possible depends on the coherence of the set of propositions characterizing it. As I noted before, this does not mean that the present conceptual system defines what is presently to be considered as possible. Humans are not only able to consider variations of statements, phrased within their conceptual system, that they consider as true. They are also able to consider variations of their conceptual systems—the latter variants are also the result of assumptions, suppositions, fictions, and stories. In this sense, the actual has precedence on the possible. There is a very good reason not to avoid statements about fictional objects and possible situations. We need them to state what is not the case and we need them for planning and projecting, for thought experiments and counterfactuals. Rescher stresses that humans live in two worlds, reality and imagination. Imagination is swarmed with possibilities. The last two points I shall consider concern logic. We have seen that something is possible if its description is logico-conceptually coherent. Again, the description need not be phrased in an actually existing language. It is sufficient that there is a description in a possible language— needless to say “possible language” contains a de dicto modality. Rescher claims that we can make sense of it in terms of a generalized conception or definition of a language. The required coherence refers to logic and to a conceptual system. The conceptual system may obviously vary from one language to the other. What about logic? Although in many passages Rescher is apparently applying classical logic, this is not a matter of principle. He states explicitly that the logic will depend on the assertoric context. He refers to different logics and moreover stresses that “matters of language and logic are to be adjudged not by an absolute and universal standard— God’s logic, so to speak—but rather in a way that is case specifically ad hoc, that is by the local standard that is (ideally) explicit or otherwise (contextually) implicit in the way and manner in which the scenario at issue is presented” [23, pp. 177]. So, here again, Rescher’s position is one of tolerance. This does not come unexpected given his work in logic, which in-

33

RESCHER STUDIES

cludes many valued logic, handling inconsistency, and plausible reasoning—see for example [19] and [25]. For the reasons given above, Rescher rejects the interpretation of worlds semantics in terms of de re possible worlds. He proposes a new semantics for modal logic, viz. one in terms of scenarios. However, as one can hardly object to the worlds semantics itself, a technical device after all, he claims that this semantics can be interpreted in terms of scenarios. The relevant section is terribly short and a precise interpretation is suggested rather than articulated. However, the matter is easily repaired, as I shall show below. 3. THOUGHT EXPERIMENTS Thought experimentation plays a role that is often underestimated. Moreover, it occurs in different contexts and serves different aims. In What If?, Rescher concentrates on thought experimentation in philosophy, but also discusses the topic in general and pays same special attention to scientific thought experiments and historical thought experiments. He discusses a large set of philosophical thought experiments as well as some from outside philosophy, showing the adequacy of his approach. The distinctive feature of philosophical thought experiments is that they concern the realm of human thinking and conception. Rescher offers an interesting categorization of the possible functions of philosophical thought experiments— refutatory, demonstrative, explanatory, and evaluative; see [24, p. 51– 52]—but also stresses (p. 89) that philosophical thought experiments are enormously diversified in their use. Thought experiments rest on suppositions. To be more precise, the supposition should be introduced in order to resolve “a more far-reaching issue”. So the main aim is not the relevant conclusion that is derivable from the supposition, but the lesson can be drawn from the derivability of that conclusion. Put differently, the aim is not to find out whether a certain conditional is true, but to show that the truth or falsehood of the conditional answers a further question. The supposition is often counterfactual, but this is not required: some thought experiments involve suppositions of which we don’t know whether they are true or not. Rescher introduces five elements that occur in every thought experiment: (i) the supposition, (ii) the context of information in which (i) operates, (iii) the conclusion, (iv) the larger question the thought experiment is designed to answer, and (v) the course of reasoning through which the preceding are to be seen as providing grounds for the purposed answer/lesson.

34

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

This little list is extremely important. In my experience, a problem with some of the literature on thought experiments is that it is easy enough to follow the examples and to see that their conclusions are reasonable, but that the exact rules of the game remain obscure. Rescher’s five elements avoid precisely this. Especially the context is central. A condition may lead to conclusions that contradict each other, but the context will settle which is the right conclusion in the present case. That the context may be implicit is not a problem as long as critical analysis enables one to reveal it. Rescher also considers five stages that roughly correspond to the specification of the elements. Malfunction may occur in each of the stages. The precise benefit acquired from thought experiments is widely discussed in the literature. Rescher takes a clear stand here. Thought experiments are unapt to establish matters of contingent fact, but can establish matters of logico-conceptual possibility. The latter concern coherence and consistency and hence are a matter of reasoning. The reasoning involved, the reasoning leading from (i) and (ii) to (iii), never compel the conclusion but renders it plausible. In some passages Rescher refers to the evaluation that is required in order to arrive at the maximally plausible conclusion. Note, however, that Rescher considers the process as an objective matter. The conclusion is the maximally plausible alternative. Needless to say, thought experiments and the involved reasoning require that one considers possibilities. The status and properties of these are as described in the previous section. So the possibilities, and even the whole reasoning, are to be located at the linguistic or communicative level. No possible worlds are involved, except when a thought experiment would be about possible worlds, and no reified possible objects or reified possible worlds enter the picture. The issue is not one of ontology but a semantic one: which statements can or cannot meaningfully be made with respect to some putative existent. If the supposition is counterfactual, there is always a conflict between the supposition and the context. Here Rescher invokes his theory of counterfactuals, which originated in [17] or earlier. The essential idea is that priorities are introduced and that these lead to the selection of a consistent set of (sometimes disjunctions of) salient beliefs. The supposition obviously receives maximal priority. The elements of the context will also require prioritization. Indeed, there will always be a manifold of conflicts if the context is extended with the supposition. The prioritization does not result from a logical analysis but is typically case specific. The general rule is that higher informativeness leads to a higher priority, but this may be

35

RESCHER STUDIES

overruled in specific cases. There is a useful classification of statements according to their informativeness on p. 43 of [24]. A nice example where even more refined priorities are required is the historical counterfactual “If the European component of World War II had been prolonged by six months or more, the U.S. would have used the atom bomb against Germany.” The context contains several factual statements, which normally receive the lowest priorities. In the present case, however, some of them need to receive a higher priority than the others in order to remove all contradictions. The counterfactual assumption makes it clear that the history of that war is involved and not the history of technology. So the factual statement that the U.S. did not use the atom bomb against Germany is rejected (receives the lower priority) whereas the factual statement on the date when the U.S. had its atom bombs ready for use is retained (receives the higher priority). The interesting point here is that the specific supposition has an effect on the context. Incidentally, in What If? Rescher proceeds in terms of rejecting and retaining salient beliefs directly in view of their priorities—rejecting the weakest link in an inconsistent chain is another formulation—rather than making the detour through maximal consistent subsets. There would only be a difference if some but not all salient beliefs of a priority level have to be removed. This would result in several selected maximal consistent subsets. Obviously, the same result is obtained by rejecting all those beliefs while retaining their disjunction. In philosophical thought experiments, the prioritization depends centrally on a philosopher’s fundamental commitments—there is an interesting list of such commitments on p. 107 of [24]. Given this, the prioritization need not be immediate but may require a complex cost-benefit analysis. Rescher explicitly discusses per impossibele suppositions, which introduce the negation of a law of nature or of a logico-conceptual necessity. There is nothing wrong with such suppositions in principle. Quite to the contrary, such reasoning is common for example in mathematics. But obviously, there are limits to this. The central criterion is that the result should not be meaningless, in other words, should not be conceptually incoherent. Suppositions that transgress the limit are unintelligible. Their internal incoherence makes them literally senseless. The criterion does not lead to a sharp boundary. Rescher refers to the relation between unrealism and indefiniteness: the latter is directly proportional to the former. As indefiniteness increases, the resulting range of possibilities is so large that we are at a cognitive loss—all this is illustrated by examples. In line with his

36

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

position, Rescher warns against a philosophical attitude that is disdainful with respect to mere “matters of fact”. Many concepts are experientially based. They are viably integrated units only in view of the factual arrangements of the world. They “miss the abstract integrity of purely theoretical coherence”. The very meaning of such concepts presupposes certain facts. 4. WORLDS SEMANTICS Let us start with a specific (first order) semantics for S5. This is a bit technical and readers willing to believe me may skip the gibberish. L will be a first order language without individual constants and LM the same extended with modal operators. Individual constants may be defined by means of the iota operator. A model is a quintuple M = (W, w0, D, s, v) in which W is a set of worlds, w0 is the actual world, D is a set (the domain), s: W → ℘(D) is a function that assigns to every world a subset of D (the objects that ‘exist’ in w), and v is the assignment function, which in every world assigns an extension to the simple non-logical linguistic entities. Every model M determines a valuation vM, which assigns a truth value to every sentence of LM. The function of w0 is to define a sensible semantic consequence relation: A is true in M iff vM(A, w0) = 1, M is a model of Γ iff all members of Γ are true in M, and Γ╞ A if and only if A is true in all models of Γ. What is the relation between the worlds of such model and the real world? Nearly nothing actually. All the model tells us about any world w ∈ W is, for every sentence A of the language, whether A is true at w or not. All the model tells us about a world is that the members of a subset of LM are true at the world, whereas all other sentences are false. By the definition of vM, the set of the sentences that are true at a world is deductively closed and maximally non-trivial with respect to S5. The so-called real world w0 is no exception. Moreover, w0 need to have nothing in common with the actual world—what is true in w0 may be false in the world. There is nothing concrete about the worlds in W. They are elements of a technical device. They are abstract points to which sets of sentences are associated. So a modal model is basically a set of sets of sentences that are related in a specific way. The objects that ‘exist’ at the worlds of the model have no properties that transcend the given language. The same holds for

37

RESCHER STUDIES

the worlds themselves. Nothing new is to be discovered about these objects and worlds. No study of a worlds model will ever lead to a conceptual change. So an interpretation in terms of concrete possible worlds transcends the model completely. Which lesson is to be drawn from this? On the one hand, one can hardly object to a technical device. It explicates a given logic for a given language and does so correctly in that the soundness and completeness of the logic with respect to the semantics are provable. On the other hand, precisely because the semantics is a technical device which defines a set of related sets of sentences, there is nothing which opposes a ‘verbal’ interpretation of the possibilities represented by the models. Let me now show that an S5-model may be obtained in terms of scenarios, viz. sets of sentences of L. Just pick a (finite or infinite) number of scenarios, Σ0, Σ1, . . . ; let I be the set of the subscripts. Select one of scenarios, say Σ0, as ‘the real scenario’. Every scenario has infinitely many predicative models. Select one predicative model, say µi. = (Di, vi), for each scenario Σi. An S5-model is easily defined from the selected predicative models. Define W = {wii ∈ I}, D = i∈IDi, s(wi) = Di, and, for every non-logical symbol ξ, v(ξ,wi) = vi(ξ). It is easily seen that, for every set of scenarios, there is a S5-model M, and vice versa. The models of other standard modal logics require an accessibility relation R, whence M = (W, w0, D, s, v, R). The simplest way is to proceed blindly: define a relation R of the required sort over the scenarios. By the same approach as before, every set of scenarios on which R is defined results in a worlds model of the required sort by requiring that Rwiwj iff RΣiΣj and every modal model may be obtained from a set of scenarios on which R is defined. For specific applications, however, one may desire to proceed in a planned way. Suppose that one wants to characterize physical necessity. This requires a reflexive and transitive accessibility relation R. So one has to make sure two things. On the one hand, RΣiΣj should only hold if those consequences of Σi that one wants to be necessary at the corresponding world w are also consequences of Σj. On the other hand, if one wants a consequence A of Σi not to be necessary at w, then there should be a scenario Σj for which RΣiΣj and of which A is not a consequence. Any selection of properties of R may be easily turned into requirements on the consequences of scenarios related by R. So any world semantics for one of the standard modal logics is easily obtained from scenarios. Moreover, the way in which the accessibility relation is here introduced shows that it may be

38

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

given a purely linguistic interpretation. Thus the transitivity of the relation in the case of physical necessity derives from the requirements on the scenarios between which the relation holds. In this example, everything that is physically necessary in a scenario Σ should also be physically necessary in every scenario accessible from Σ. What I tried to show up to this point is that it is easily possible to interpret a modal semantics in a way that agrees with Rescher’s linguistic approach to possibility. It is even possible to devise worlds models in terms of scenarios, provided one is willing to proceed in terms of predicative models. Just like these models, the worlds of the semantics will correspond to sets of sentences that are deductively closed and maximally nontrivial—saturated sets to use a shorter description. To consider such sets provides some advantages in that it simplifies the metatheory. For certain other purposes, to consider saturated sets is cumbersome and uselessly complicated. All this, however, is a matter of appropriateness, not of principle. As we have seen, there is nothing concrete about the worlds and objects of a modal semantics. The whole construction stays within the confines of the given language. Rescher seems to prefer a different kind of semantics for scenarios. It is correct that, for most sets of non-saturated scenarios, we obtain an infinity of worlds models. Moreover, each world corresponds to a set of sentences that is deductively closed and maximally non-trivial. This means that all sentences of the language obtain a truth value at every world. The reason is that we pick a predicative model for each scenario. We pick only one, but it corresponds to a saturated set of sentences. So every modal model provides much more information than the scenario idea requires. Moreover, we have to consider all modal models of a set of scenarios in order to obtain the semantic consequence relation. A semantics that is much closer to scenarios is possible. The idea is to devise models that assign values only to linguistic entities occurring in the considered premises and conclusion. So the assignment and the valuation will be partial functions. The resulting semantics somewhat resembles a tableau method.2 Articulating it is not very difficult, but would lead us too far into technicalities. But would such a semantics constitute an advancement in comparison to the usual one? The worlds semantics copies the standard semantic style from predicate logic. There, every model verifies all members of a saturated set of sentences. Such a set corresponds to the complete description of a possible state of the world within the given language. If the premises provide no information on some sentence, it will

39

RESCHER STUDIES

have models in which that sentence is true and others in which it is false. So that gives us the right consequence relation. The standard models have some advantages, technical ones and, in some people’s views, also philosophical ones. So I wonder whether Rescher wants to reform the semantic standard and whether he wants to do so specifically for scenarios or in general. 5. REASONING TOWARDS COUNTERFACTUAL CONCLUSIONS All examples of counterfactuals considered in the books under discussion have a rather simple context. The salient beliefs are enumerated and the conflicts between subsets of them and the counterfactual assumption catch the eye. Depending on the specific situation, a priority ranking of the salient beliefs is introduced. If the salient beliefs are enumerated, the effects of the priority ranking are immediately clear. If this is the case, and the conflicts are known, some salient beliefs are rejected for present purposes and the truth of the counterfactual conclusion is assessed by checking whether it is a consequence of the retained salient beliefs. For some counterfactuals, the situation may be more complicated, as Rescher realizes very well. For one thing, the involved theories may be complex, even from a computational point of view. As predicate logic is not decidable and there is no positive test for consistency (consistency is not even semi-recursive), it may take one a long way to figure out whether the conclusion of a counterfactual is derivable or not. Sometimes one at best arrives at an estimate which relies ‘on present best insights’. Another complication, although not so far reaching, is that not all salient beliefs need be recognized beforehand. When the conclusion of a historical counterfactual is assessed, one historian may bring up certain historic facts that were unknown to the others until then and this may have an effect on the derivability of the conclusion. Where the conclusion of a scientific counterfactual is assessed, one scientist may bring into the picture a theory that was unknown to the others or of which the others did not see the relevance. That new information comes up during the reasoning process will not cause problems for the priority ranking, which depends on properties of the salient beliefs, as was explained in Section 3. As soon as new salient data are adduced and recognized, all parties will agree on their priority ranking. The aim of the present section is to show that, even in such complex cases, it is possible to reason towards the conclusion and to reach either the

40

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

conclusion or its rejection whenever this is possible, that is whenever there is a proof that the conclusion counterfactually follows or whenever there is a demonstration that it does not. Let me state at the outset that there is nothing wrong with Rescher’s definition of the counterfactual consequence relation—I shall strictly stick to it. Moreover, the problem I promise to solve has nothing to do with computational efficiency. The problem strictly concerns articulating a way in which one may proceed to arrive at an answer to the question: Is a considered statement derivable or not under the counterfactual assumption? Allow me to introduce two simplifying assumptions, or idealizations: that none of the salient beliefs is self-contradictory and that A and B are salient beliefs whenever their conjunction is. It is not difficult to overcome these assumptions, but to do so requires some technicalities that are better avoided in the present paper. The recipe I shall propose requires that Rescher’s counterfactual reasoning is first characterized by an adaptive logic. In order to do so (in the specific way I consider here), we need a ‘translation’ from the standard predicative language to the standard modal predicative language. The reasoning will operate on the counterfactual assumption, which is taken to hold indefeasibly, and several levels of prioritized salient beliefs, which belong to the context. The counterfactual assumption is mapped on itself; so it is not translated. If A is a salient belief that has the highest priority, it is mapped to ◊A, if it has the next highest priority it is mapped to ◊◊A, etc. Let us suppose that there are three levels of priorities. The diamond may be read as “it is plausible that”, two diamonds indicating a lower plausibility: “it is plausible that it is plausible that”, etc. An adaptive logic in standard format3 is a triple: a lower limit logic, a set of abnormalities,4 and an adaptive strategy—see below. The lower limit logic is a Tarski logic—a logic of the familiar kind—that has a proof theory and an adequate semantics. For the present application, we shall consider a specific sequence of adaptive logics. The sequence is required to handle the prioritization of the salient beliefs. As we agreed to have three priority levels, the sequence will consist of three adaptive logics which will be called Tm1, Tm2, and Tm3. All of them have the same lower limit logic, a predicative version of the modal logic T, viz. the logic described in Section 4 with a reflexive accessibility relation. The sets of abnormalities of the adaptive logic Tmi is Ωi = {∃(◊i A&¬AA∈Fa} in which i ∈ {1, 2, 3}, abbreviates a sequence of ◊i diamonds, A is the existential closure of A (A

41

RESCHER STUDIES

preceded by an existential quantifier over every variable free in A), and Fa is the set of (open or closed) atomic formulas and their negations.5 Note that an abnormality expresses that a statement which is plausible (to some degree) is false. The strategy is Minimal Abnormality—its effect will become clear below. Where Γ is the premise set, prepared or ‘translated’ as explained above, the sequence of adaptive logics delivers CnTm3(CnTm2(CnTm1(Γ)))—in words: the Tm3-consequences of the Tm2-consequences of the Tm1-consequences of Γ. The standard format defines the proof theory and the semantics of the three adaptive logics. I shall not present these definitions, but rather explain intuitively the way in which the logics work in terms of the semantics. We start from the T-models of the premise set Γ. The ‘Tm1-abnormal part’ of each T model, M, is defined by Ab1(M) = {AA∈Ω1; M╞ A} in which M╞ A abbreviates that M verifies A, in other words that vM (A ,w0) = 1. So Ab1(M) comprises the members of Ω1 that are verified by M . Ab2(M) and Ab3(M) are defined similarly with respect to, respectively, Ω2 and Ω3. A T-model M of Γ is Tm1-minimal abnormal if there is no T-model M′ of Γ for which Ab1(M′) ⊂ Ab1(M). Similarly for the other two adaptive logics. The combined adaptive logic first selects the Tm1-minimal abnormal models from the T-models of Γ. From these models, which provably are the T-models of CnTm1(Γ), the combined adaptive logic selects the Tm2minimal abnormal models, and from these it selects the Tm3-minimal abnormal models. The counterfactual consequences of Γ are the formulas that are verified by all models in the last selection. It is not difficult to see that every model in the last selection verifies a maximal consistent subset of the original, untranslated, premises. This subset contains the counterfactual assumption, contains as many formulas of level 1 priority as are compatible with the counterfactual assumption, contains as many formulas of level 2 priority as are compatible with the previous set, and similarly for the formulas with level 3 priority. In other words, every model in the last selection verifies the members of a prioritized maximal consistent subset. The converse can also be shown: every correctly prioritized subset of the original premises is verified by a model in the last selection. So the combined adaptive logic defines a semantic consequence set that coincides with the consequence set of a counterfactual assumption in view of the prioritized salient beliefs as defined by Rescher.6

42

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

The semantics is useful to clarify the way in which the combined adaptive logic handles the premises, but is does not clarify the way in which one may reason in order to derive the counterfactual consequences. To this end we need the proof theory. I only present an outline, referring the interested reader to the bibliographic items on adaptive logics for details. Adaptive logics are defeasible logics. It follows that they require a dynamic proof theory: formulas derived (locally) at some point may be considered as not derivable at a later point in view of the insights in the premises provided by the continuation of the proof. At a still later point, such a formula may again be considered as derived in view of the meanwhile gained insights in the premises. Adaptive logics handle this dynamics by deriving formulas ‘on a condition’ and by a marking definition. Derivation on a condition is a simple matter. Every line of an annotated proof contains, apart from a line number, a formula, and a justification, also a condition. This is simply a finite set of abnormalities. There is nothing mysterious about this. Where ∆ is a finite set of abnormalities, A is derivable on the condition ∆ from the premise set Γ if and only if the classical disjunction of A and of the members of ∆ is derivable from Γ by the lower limit logic. According to T, p V (◊p&¬p) is derivable from ◊p. So p is Tm1derivable on the condition ◊p&¬p from ◊p. At every stage of the proof, the marking definition defines certain lines as marked, that is OUT (the formula of those lines is not derived), or unmarked, that is IN (the formula is derived). The marking definition proceeds in terms of the disjunctions of abnormalities that are derived unconditionally (on the empty condition) at that stage.7 Let me conclude this section by two comments. First, remember that the combined adaptive logic defines CnTm3(CnTm2(CnTm1(Γ))). As each of the three involved consequence sets are not even semi-recursive8 one might suppose that one never will arrive at deriving any Tm2-consequence. This, however, is a mistake. The proof theory of the combined adaptive logic enables one to conditionally derive Tm2-consequences and even Tm3consequences right from the start of the proof—see [3] and [5] for details. This property of combined adaptive logics is especially attractive. The second comment concerns ‘final derivability’. Intuitively, a formula is finally derived at a line of a proof if and only if the line will be unmarked in any extension of the proof. Interestingly, final derivability may be defined by: if and only if any extension of the proof in which the line is marked may be further extended in such a way that the line is unmarked. Final derivability 1 is computationally complex (in general Π1 for the Minimal Abnormality

43

RESCHER STUDIES

strategy; see [27] but assess this in terms of what is said in the last note). Nevertheless, there are procedures that form criteria for final derivability— see [6] and forthcoming work by Peter Verdée. When applied to a premise set and conclusion, the procedure delivers a finite proof of final derivability if and only if there is one; if it stops and the conclusion is not derived, then the conclusion is not finally derivable. As the consequence set is not semi-recursive, the procedure will never stop for some premise sets and conclusions. This means that, at any point in a proof that has not stopped, one has to make the choice between acting on present best insights or continuing the procedure. The latter decision will in general deliver more unconditionally derived disjunctions of abnormalities and these are crucial for determining final derivability. Nevertheless, for some premise sets and conclusions, every finite set of disjunctions of abnormalities will be unavoidably insufficient as a criterion for final derivability. The course of action to be taken will then depend on the aim we are pursuing. If the aim is intellectual insight, we shall conclude that the answer is, for example, yes by present insights, but that the latter are inconclusive. If the aim is action, we shall sometimes decide to go by present insights, especially if there is no time to postpone the decision. 6. SOME COMMENTS AND QUESTIONS Rescher presents a recursive definition of existence. As a criterion, the definition is somewhat vague, especially the “must be invoked” part. Sometimes several sets of entities provide an equally satisfactory account of something that is already taken to exist. In such a case, none of these sets of entities must be invoked. This seems to be a reason to grant existence to the entities of all those sets, rather than to none of them as a literal application of the criterion requires. Moreover, it will often be nearly impossible to show in an absolute way that something must be invoked in order to arrive at a satisfactory account of something that is already taken to exists. I do not consider this as a hard objection. As we are not near the accomplishment of our knowledge, a vague instruction is preferable over a hard one and tolerance is a virtue. Even if the criterion is somewhat vague, it seems to me that it is more than sufficient itself to reject the existence of possible worlds and merely possible objects—perhaps this is not sufficiently stressed in [23]. Indeed, everyone will have to grant existence to linguistic entities and to their meanings in the sense of entities that occur in people’s minds.9 Possibilities

44

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

play a role in nearly every experience. We see that there is nobody in the room. We often hesitate about the precise nature of the things we are looking at. We sometimes arrive at false conclusions and may later detect ourselves that we made a mistake. So possibilities are intrinsically interwoven in all experience. All this holds even more for beliefs. In short, possibilities that have a purely linguistic basis are within the reach of our mental abilities. But then they are sufficient for thinking about the consequences of their realization, we may rely on them to express goals and desires, we may make up fictions, etc. Given that linguistic entities and their mental counterparts exist, that we understand the way in which they operate, and that they provide a satisfactory account of our handling possibilities, there is no need to moreover invoke possible worlds and merely possible objects. This argument is strengthened by two further considerations. The first is that the proof burden on those that require the existence of possible worlds is enormous. They have to show that ‘negative’ observations, mistaken observations, and hesitancy in observation are impossible if there is only one world. The second consideration concerns the explanatory power of the existence of possible worlds. Suppose for a moment that they would exist, populated with objects and facts. If they would provide an explanation for the fact that we are able to consider certain possibilities, then they would do so only in a circular way. Indeed, the only way in which we are able to obtain any knowledge about their existence—this knowledge is at best schematic, as Rescher convincingly shows—is through our actual linguistic behaviour. Given that our future linguistic behaviour is unpredictable, the introduction of those possible worlds cannot possibly lead to a theory that provides any information that does not also derive from our linguistic behaviour. So this is like explaining gravitation in terms of angels that push objects in the direction of each other. It other words, the explanation is of the vis dormitiva sort. While I agree with Rescher on the ontological status of possibilities, It seems to me that he is overoptimistic in presenting a linguistic approach to facts. For Rescher, a fact is anything that is correctly statable in a possible language—the crucial passage is on pp. 34–35 of [23], footnotes 8 and 9 included. A language should be understood in the standard sense here: the set of statements is at most countably infinite. Rescher is right that it follows from his definition that there may be uncountably many facts and these obviously cannot all be described into a single language. It seems to me, however, that this approach is still too restrictive. We are able to consider the possibility that an object has a property which cannot be de-

45

RESCHER STUDIES

scribed in any countable language, just like some real numbers can only be described by an infinite sequence of symbols. We have no reason to exclude that such possibilities are realized in our world. It would be difficult to justify that they are not called facts. The whole point is that a finite language allows one to describe sets of which some members cannot be described in that very language—the set of functions from natural numbers to natural numbers is a ready example. A possible source of worlds realism is apparently the similarity with mathematical objects. Within arithmetic, there is, for every number, a larger prime number. Within some modal models, p is true in the ‘real world’ w0 but there are worlds, accessible from w0, in which p is false. As some take numbers to exist because they are supposed to exist in the realm of arithmetic, why not say that worlds exist because they are supposed to exist in modal semantics. This brings me to the existence of numbers, a point on which I would like to disagree with Rescher. In order that I be not misunderstood, let me state at once that I have not the slightest desire to grant any form of existence to either Hamlet, Hera, or the possible fat man in the door. My only aim is to question the viability of granting existence to mathematical entities. In most places, Rescher sees the existence of (some) mathematical entities as unproblematic.10 In some passages, for example [23, p. 187], he considers them more closely, pointing out that they are creatures of convention. In consequence, their properties are fully dependent on the stipulations by which they are introduced. Rescher stresses that it is possible to do so because mathematical entities are abstracta. The matter is wholly different for fictional objects and worlds, which are concreta. With all this I agree. Mathematical entities being abstract, a description would be sufficient to individuate them; worlds and objects being concrete, every such description is partial and hence insufficient for individuating them. Let us return for a moment to the two main objections against granting fictional objects a serious existence. The first was that they have no thought independent footing as constituents of a realm. This is correct in the sense that nothing would be left of Sherlock Holmes’s pipe if humans cease to exist. However, the same applies to mathematical entities. If there were no humans, there would still be two oak trees on the island in the river, but the number two would not anymore be around. Of course, numbers and other mathematical entities have a certain kind of independent footing. What I mean is that there is an internal necessity to them. Once we have agreed on the successor function and the addition

46

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

function along the lines of the Peano axioms, 3 + 2 is inescapably identical to the second successor of 3—that we happen to call this number five is of course a conventional matter. So this inherent necessity, deriving from the way numbers are introduced, seems to provide them with an independent footing: you cannot decide things to be otherwise except by changing the rules of the game. But a similar necessity holds for many fictional objects. I may write a new play in which a Hamlet is staged, say a softy who likes video games and fears to act. I may claim in the play that this is the very same Hamlet as Shakespeare’s. Still, in doing so, I am changing the rules of the game. As long as we are talking about Shakespeare’s Hamlet, we are facing certain necessities. Precisely this enables one to study Hamlet’s acts and motives, his tragedy, etc. We can argue about this and convince each other by rational means, post-modernists aside. Of course Shakespeare created Hamlet. So he is not thought independent. But were not mathematical entities created in a rather similar way? They are social creations, but still creations and hence not thought independent either. Incidentally, the gods of the Olympus are also social creations. The upshot is that the thought dependence and thought independence of mathematical objects is very similar to that of other fictional objects. Indeed, there existence within mathematical theories depends on logico-conceptual coherence and this is the characteristic of possibilities. So let us turn to the second objection: schematism. Unlike fictional objects, mathematical entities would be completely determined by their definitions. Moreover, this would be sufficient to individuate them, because they are abstract entities. But is the description of mathematical entities sufficient to individuate them? The limitative theorems show at least that we have no warrant for this. We know from Gödel’s first theorem that even arithmetic cannot be characterized by a recursive set of first-order axioms—we cannot rule out non-standard models. Second-order logic helps us out and so does the standard model of arithmetic. However, both second-order logic and the standard model define arithmetic as a set of statements that is not even semi-recursive. Moreover, if there were a problem with Peano Arithmetic, viz. if it would turn out to be inconsistent and hence trivial,11 then there certainly would be a problem with second-order logic as well as with the standard model. It follows that we have no warrant that mathematical entities may be characterized by reliable means. Worse, reliable means do not allow one to define the set of natural numbers. So we are facing a dilemma. Every first order description is incomplete and every other description is known to be unreliable in that it re-

47

RESCHER STUDIES

quires more than a denumerable language and essentially refers to a set that is not recursively enumerable. But even if numbers were completely defined, this does not show that they exist. We still have to demonstrate that they must be invoked in order to make sense of entities which we take to exist for independent reasons, say physical entities and their properties. But how exactly do we use mathematical entities in order to handle physical objects and processes? We take the values of quantitative properties of physical objects, move into the mathematical fiction with these values in order to perform a number of operations on them, and next return to physics with the outcome of those operations, which is then known to be a quantitative property of the physical objects. Arriving at the same conclusion without the detour through mathematics would certainly be much more complicated. However, this is merely a pragmatic consideration. I do not know of any demonstration that the detour is necessary. I am not advocating that one tries to avoid the detour. I take mathematical theories to be fictions, extremely coherent ones about abstract entities, but still fictions. There is nothing wrong with fictions, with reasoning about them, and with making existential statements within them—I fully agree with Rescher on this. All that is required in order to safeguard clear thinking is that we realize that we are operating within the fiction. Allow me to add to incidental comments to this. The first is that fictional contexts are in one respect very different from ‘reality contexts’. In ‘reality contexts’, possibilities are introduced by (sometimes hidden) assumptions or suppositions in order to obtain a reductio, a counterfactual, an idealization, and the like. In fictional contexts, no assumptions or suppositions are required. A story is told or written or acted. We may derive conclusions from the fiction to reality, moral conclusions, descriptive conclusions, and so on. Nevertheless, reality and the fiction are and should be totally separated ontologically. The fiction is merely about thought objects. The precise circumstances of such derivations, their rules and pitfalls, would be a very interesting object of research. The second incidental remark is that, once we operate within the fiction, we may handle fictional objects in pretty much the same way as we handle real objects. Rescher stresses that a real object may be individuated by pointing at it. In a stage play or movie, one can also point to a character as genuinely as one can point to the cat on the mat. I mean pointing to the character, not to the actor. But obviously there is an actor, the character is fictional, and when we

48

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

point to the character, we point to a schematic object that is represented by the actor. The next point may be minor, but I would like to make it because it concerns logic. We have seen that Rescher takes an extremely tolerant stand with respect to logic. Apparently he does not rule out any logic in advance. Nevertheless, a striking abhorrence of contradictions is implicitly present in his work. This may be seen on the one hand from the fact that, in order to indicate the limits of logical coherence, he often uses ‘contradictory’ rather then non-trivial or incoherent—he does so constantly in the two books under discussion here. A second indication is that, to the best of my knowledge, Rescher’s logical work includes logical systems that allow for inconsistency (both A and its negation being true) whereas he never considers systems in which contradictions (the conjunction of A and its negation) are true—[25] allows for inconsistency but explicitly rules out contradictions.12 I am not complaining about the logical systems Rescher articulated. These are perfectly sensible and well wrought. I also agree with the idea that inconsistencies should be eliminated, although I would add the proviso “whenever possible”.13 I only wonder whether Rescher does indeed have objections against true contradictions and, if so, what is the reason for it. Indeed, relevant logics14 and half of the other paraconsistent logics15 allow for true contradictions. None of them turns inconsistency into triviality, some of them do not allow for triviality, and even those that allow for triviality see it as incoherent.16 So even paraconsistent logics that allow for contradictions define an internal coherence. Does this not mean that, if there are good reasons to apply those paraconsistent logics in certain contexts, then that application is unproblematic? While Rescher might object to these paraconsistent logics even in those circumstances, there can be no objection, given their internal coherence, against introducing them by supposition. Thus one may study the combination of (the first-order version of) Frege’s set theory with a specific paraconsistent logic, as it is done in [15]. That combination is clearly a possibility. There is a question to which I had expected to find an answer in one of the books under discussion. Rescher stresses the importance of conceptual systems and the fact that such systems are devised, historically, as a result of our attempt to form theories about the world.17 As we have seen, one may not only consider a possibility that derives from varying, by introducing an assumption, on actual situations. One may also introduce assumptions that lead to a conceptual system which is different from the present one. Our ability to make such assumptions makes it possible for us to con-

49

RESCHER STUDIES

sider the resulting possibilities. Different conceptual systems seem to lead to different logical spaces, that is sets of possibilities which are exhaustive and mutually exclusive. This evokes the question whether all those logical spaces may be joined into a single one.18 I cannot see that this question could be answered in the positive. Moreover, if the answer is negative, this does not form a problem for Rescher’s position, but it seems to make life sour for the worlds realist. Let us first consider the question itself. It is handy to refer to a mathematical example in this connection because this saves me the trouble form specifying the meaning of the non-logical part of the language. It is shown, for example in [11], that there are non-trivial paraconsistent set theories that validate (a version of) extensionality and full comprehension. However, if classical negation (the Boolean negation of classical logic) is added to the language, these set theories immediately collapse into triviality. Note that the resulting system might be saved from triviality by introducing a set of restrictions on the occurrence of classical negation in the set theoretic axioms. This, however, does not undermine my point because it involves changing the paraconsistent set theory, not only the literal formulation of its axioms but also its concepts. So the situation is as follows. The underlying conceptual system of the paraconsistent set theory is coherent and there are coherent conceptual systems that contain classical negation, but joining one of the latter to the former delivers gibberish. As I noted before, all this is not a problem for Rescher. Possibilities are arrived at by means of assumptions from “a historically developed product” [24, p. 150]. Given this, the conceptual systems mentioned in the previous paragraph may be reached by us, but their combination, resulting from combining assumptions, need not safeguard the possibilities arrived at by the separate assumptions. However, the absence of a unique logical space is a serious problem for the world’s realist. He might try to defend his position by saying that some sets of worlds cannot be reached from each other. The problem with that way out, however, is that the sets should be separate in one sense while being reachable from each other in another sense. Let me try to phrase this briefly. On the one hand, the structure of some worlds is compatible with classical negation whereas the structure of others is not. So here we have two possibilities that we should be able to compare within the same logical space and that we clearly can express within the same language. On the other hand, once we have opted for one of the two possibilities, we obtain a logical space that is incomparable with the one obtained from the other possibility and as a result we must describe

50

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

these possibilities within different languages. The reader may easily replace this example by one in which two geometries are involved, or two views on heat. At this point, I turn to some comments and questions concerning thought experimentation. Rescher stresses that a thought experiment never compels any conclusion but at best renders it plausible. Note that this concerns the conclusion, element (iii) from Section 3, not the lesson drawn from the thought experiment, which is the answer to element (iv). So this means that the conclusion does not follow from the supposition and the salient beliefs by deductive means, but by plausible reasoning. This evokes an interesting question. Let A be the supposition, ∆ the set of statements that make up the context, and B the conclusion. If A is a counterfactual supposition, then it is obvious that B follows from A and ∆ by plausible reasoning only because counterfactual reasoning is defeasible. In some thought experiments, however, we consider a supposition A of which we do not know whether it is true or not. In such cases, there is no conflict between A and the context ∆. Even here the conclusion needs to follow from A together with the context by a form of defeasible reasoning. Moreover, in some cases where A is a counterfactual supposition, it seems possible to transform the thought experiment in such a way that we are agnostic about the transformed supposition A′, that there is no conflict between A′ and the transformed context ∆′, and that the result is still a thought experiment. If this is correct, then that suggests that even thought experiments with a counterfactual supposition involve a form of plausible reasoning apart from the reasoning needed to resolve the conflicts between the hypothesis and the context. In view of all this, it seems to me that the theory on thought experimentation would benefit enormously from the study of the precise forms of defeasible reasoning that occur in thought experiments. The insights gained would lead to a theory that answers questions like the following. Do all forms of defeasible reasoning occur in thought experiments? Are there specific forms of which one needs to occur in every thought experiment? Is counterfactual reasoning itself one of these forms, or need it always be combined with another form? There is a different topic that seems to deserve further research. We have seen that, in the case of a counterfactual supposition, the prioritization of the elements of the context depends on the context as well as on the supposition. All the examples that Rescher gives are certainly convincing. Nevertheless, it seems possible to arrive at certain rules which settle, in

51

RESCHER STUDIES

terms of the content and the form of the supposition and context, which prioritization scheme should be followed. Thus, as Rescher suggests in the atomic bomb example mentioned in Section 3, the supposition sets the topic and salient beliefs that concern this topic should obtain the lower priority. Another rule might be that the lower priority should go to salient beliefs that relate to properties of the subject of the supposition, whereas beliefs about properties of other objects should as much as possible be left apart. These proposals are obviously simplistic. My only aim is to indicate a possible line of research. It should also be checked whether such criteria may be formulated in terms of the form of the English sentences or should refer to the logical form of the ‘translation’ of the statements into a formal language—the latter option would involve some complications because logically equivalent formulas would not necessarily be equally faithful ‘translations’. A question of a very different nature concerns meaningless suppositions in thought experiments. It seems to me that the meaningfulness of a supposition will depend heavily on the context and, in as far as the context is implicit, that the meaningfulness may change as time goes by. A simple example of the last point is provided by the quotation of Quine on p. 155 of [24]. In 1972, Quine uses the phrase “speculation on what we might say in absurd situations of cloning and transplanting”. There is nothing absurd with such situations today, at least if brain transplanting is not involved. As to the role of the context, I would like to refer to two examples from different domains. In [14], Chris Mortensen makes an attempt at developing a geometry that would allow for such objects as are displayed in well-known drawings by M. C. Escher. Suppositions that in isolation seem completely outlandish would obviously become meaningful with such a geometry in the context. My second example concerns a supposition that Rescher [24, p. 57] considers outlandish: “Suppose that bees spoke English.” Again, the matter will depend on the context. The hypothesis might be used with a context in which bees retain the ability to communicate the information they communicate by their present dance language. It is easily seen that their ‘English’ would have to involve some unusual complications: they would need words to describe in a rather precise way distances, the size of angles, and the quality of the ‘source’. The lesson to be drawn might be that the communication which these simple insects manage to perform by means of their analogical language would require a far more complex nervous system if the communication had to be performed by means of a symbolic language. The lesson I would like to draw from this paragraph is

52

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

that the meaningfulness of the supposition of a thought experiment cannot be judged independently of the context. 7. ENVOI It was not sufficiently stressed, in the preceding sections, that the philosophical stand taken by Nicholas Rescher on possibilities and the way in which he handles thought experimentation are intimately connected. Let me illustrate the matter by means of an example. One may consider Rescher’s work on counterfactuals as concerned with criteriology, whereas the work by David Lewis is concerned with the logical structure of the counterfactual implication. This distinction, however, is not essential. Neither is the distinction that derives from explicating or not explicating the matter in terms of a worlds semantics. The essential distinction lies here: for Rescher, a counterfactual implication does not make sense. A counterfactual implication relates the counterfactual supposition to the conclusion. However, this relation is not well-defined as long as the context is not included. The possibilities are not ‘out there’. They are thought objects that come into being, in that capacity, by the introduction of a supposition within a specific context. That I would be the most appropriate person for discussing these two subject matters is questionable. I am glad I accepted to write the essay, even if I had to do so in difficult circumstances, not only because I learned a lot from writing it, but because friendship deserves all consideration one can afford.

REFERENCES [1]

Anderson, Alan Ross and Nuel D. Belnap, Jr., Entailment. The Logic of Relevance and Necessity, volume 1. (Princeton: Princeton University Press, 1975).

[2]

Anderson, Alan Ross, Nuel D. Belnap, Jr., and J. Michael Dunn. Entailment. The Logic of Relevance and Necessity, volume 2. (Princeton: Princeton University Press, 1992).

53

RESCHER STUDIES

[3]

Batens, Diderik, “A general characterization of adaptive logics,” Logique et Analyse, (2001), pp. 173-175 and 45-68.

[4]

———“In defence of a programme for handling inconsistencies,” in Joke Meheus, Inconsistency in Science (Dordrecht: Kluwer, 2002), pp. 129-150.

[5]

———, “The need for adaptive logics in epistemology,” in Dov Gabbay, S. Rahman, J. Symons, and J. P. Van Bendegem, editors, Logic, Epistemology and the Unity of Science (Dordrecht: Kluwer Academic Publishers, 2004), pp. 459-485.

[6]

———, “A procedural criterion for final derivability in inconsistency-adaptive logics,” Journal of Applied Logic, vol. 3 (2005), pp. 221-250.

[7]

———, “A universal logic approach to adaptive logics,” Logica Universalis, vol. 1 (2007), pp. 221-242.

[8]

———, “Adaptive Logics and Dynamic Proofs: A Study in the Dynamics of Reasoning,” 200x. Forthcoming.

[9]

———, “Towards a Dialogic Interpretation of Dynamic Proofs,” To appear in a Festschrift for Shahid Rahman, 200x.

[10] Batens, Diderik, Chris Mortensen, Graham Priest, and Jean Paul Van Bendegem, (eds.), Frontiers of Paraconsistent Logic (Baldock, UK: Research Studies Press, 2000). [11] Brady, Ross, Universal Logic (Stanford, Ca.: CSLI Publications, 2006). [12] Newton C.A. da Costa. “On the Theory of Inconsistent Formal Systems,” Notre Dame Journal of Formal Logic, vol. 15 (1974), pp. 497-510. [13] Meheus, Joke, (ed.), Inconsistency in Science (Dordrecht: Kluwer, 2002).

54

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

[14] Mortensen, Chris, “Peeking at the Impossible,” Notre Dame Journal of Formal Logic, vol. 38 (1997), pp. 527-534. [15] Priest, Graham, In Contradiction: A Study of the Transconsistent (Oxford: Oxford University Press, 2006). Second expanded edition. [16] Priest, Graham, Richard Routley, and Jean Norman, (eds.), Paraconsistent Logic: Essays on the Inconsistent (München: Philosophia Verlag, 1989). [17] Rescher, Nicholas, Hypothetical Reasoning (Amsterdam: NorthHolland, 1964). [18] ———, Topics in Philosophical Logic (Dordrecht: Reidel, 1968). [19] ———, Plausible Reasoning: An Introduction to the Theory and Practice of Plausibilistic Inference (Assen/Amsterdam: Van Gorcum, 1976). [20] ———, The Strife of Systems: An Essay in the Grounds and Implications of Philosophical Diversity (Pittsburgh: University of Pittsburgh Press, 1985). [21] ———, A System of Pragmatic Idealism, Volume 3: Metaphilosophical Inquiries (Princeton, NJ: Princeton University Press, 1994). [22] ———, Satisfying Reason: Studies in the Theory of Knowledge (Dordrecht: Kluwer, 1995). [23] ———, Imagining Irreality (Chicago and La Salle, Illinois: Open Court, 2003). [24] ———, What If? (New Brunswick, New Jersey: Transaction Publishers, 2005). [25] Rescher, Nicholas and Robert Brandom. The Logic of Inconsistency: A Study in Non-Standard Possible- World Semantics and Ontology (Oxford: Blackwell, 1980).

55

RESCHER STUDIES

[26] Routley, Richard, Relevant Logics and their Rivals, volume 1. (Atascadero, Ca.: Ridgeview, 1982). [27] Verdée, Peter, “Adaptive Logics Using the Minimal Abnormality 1 Strategy are Π1-Complex,” Synthese, 200x. Forthcoming. NOTES 1

Research for this paper was supported by subventions from Ghent University and from the Fund for scientific Research — Flanders.

2

Beth devised his two-sided tableaux to obtain a semantics for intuitionism. Later Kripke turned this into a worlds semantics that agrees better with the standard conventions, viz. that a saturated set of sentences (or formulas) is verified by a world.

3

The standard format defines the proof theory as well as the semantics of the logic. Moreover, most metatheoretic properties, including soundness and completeness, were proved in terms of the standard format without relying on the specific properties of the involved logic. For a discussion of application contexts of adaptive logics, see, for example, [5]; for most of the metatheory, including proofs, see [7]. Soon [8] will be the best reference for both.

4

Intuitively, an abnormality is a formula, specified by a logical form, which is taken to be false unless and until it has been proven that the premises require it to be true. Which formulas are considered as abnormalities will obviously depend on the specific form of defeasible reasoning needs to be applied.

5

A formula is atomic if it contains no logical symbol except possibly for identity, “=”.

6

If the counterfactual supposition is p then ◊¬p will be a member of the ‘translated’ premise set. A T-model that verifies ◊¬p also verifies ◊ (¬p V q). It can be shown, however, that no harm originates from there. If this causes q to be verified by some model in the last selection, whereas q is not a counterfactual consequence, then ¬q will be verified by some other model in the last selection.

7

The theory of dynamic proofs is spelled out in [9] and turns out to be astoundingly simple.

8

This is not a property of the adaptive approach, but of the conseqnence relation defined by Rescher. The computational complexity of both is exactly the same.

9

There may be disagreement about the existence of propositions in the sense of Frege. There may also be disagreements about the precise status of natural languages and about the question whether they can be seen as well-defined systems. All this, however, is rather irrelevant for the point I am making in the text.

10

This statement is about the abstract mathematical entities themselves, not about their occurrence in for example measurements. For a very instructive contribution to the meaningful use of numerical values I refer to Chapter 7 of [22].

56

ON POSSIBILITIES AND THOUGHT EXPERIMENTS

NOTES 11

Peano Arithmetic would then be trivial because its underlying logic is classical.

12

Note that negation behaves abnormally, because both A and ¬A may be true, and that conjunction also behaves abnormally, because A&¬A is false even if A and ¬A are true.

13

In [20] Rescher subscribes to this view and documents the way in which one may proceed to eliminate inconsistencies. Here too his formulation suggests that he considers this elimination as an absolute demand. In [4], I argue that there is no warrant that all inconsistencies can be eliminated, but that a sound methodological maxim requires that they are eliminated whenever this is possible.

14

See, for example [1], [2], [11], and [26].

15

Paraconsistent logics are logics in which ex falso quodlibet (to derive B from A and ¬A) is invalid. See, for example [10], [13], and [16]. All relevant logics are paraconsistent.

16

Allowing for inconsistent theories that are non-trivial was the explicit main aim of Newton da Costa’s work, for example [12].

17

I refer to [21] for a very instructive contribution on conceptual systems, their interaction, and their relation to experience.

18

At the level of languages, the corresponding question is whether there is a language in which may be expressed everything that may be expressed in a possible language? Rescher answers that question in the negative [23, p. 35] but offers only the argument that languages are denumerable whereas the set of possible languages is not. Obviously, logical spaces may be non-denumerable.

57

Coherentism and Coherence Truth in the Philosophy of Nicholas Rescher Bryson Brown Natura in reticulum sua genera connexit, non in catenam: homines non possunt nisi catenam sequi, cum non plura simul possint sermone exponere. Albrecht Haller. 1. INTRODUCTION

N

icholas Rescher’s work combines a deep knowledge of the history of philosophy, a broad strategic sense of philosophical ideas and a creative open-mindedness about how to apply them. He has proposed a coherence account of truth, in The Coherence Theory of Truth (CTT, 1973), a pragmatic approach to the justification of cognitive methods in Methodological Pragmatism (MP,1977) and a coherentist epistemology in Cognitive systematization (CS, 1979). Space precludes a detailed examination of all the issues addressed in these three volumes in this review essay. Instead, we will focus on two key problems in epistemology and the theory of truth, and examine the solutions that Rescher proposes. But we begin with an account of the three books and their relations. Rescher’s philosophical strategy in these books promises to make our notions of truth and justification fit together hand in glove. One might worry that the link is a little too direct—but Rescher is well aware of the problem of the criterion—no simple circularity linking methods and truths will pass here. Further, the coherence view of truth adopted in CTT is avowedly criterial, that is, its first concern is not to say what it is for a sentence to be true, but to explain how we decide that a sentence is true. As we will see, Rescher does argue for some constraints on the nature of truth on the basis of his coherence theory, but he adopts a tripartite stand on truth: a correspondence view of the meaning of ‘is true’, a pragmatic account of how we justify a choice of truth-criteria, and a coherentist account of the criterion by which we judge what is true.(CTT, 262) The upshot, as Rescher sees it, is a ‘systematic unity’ joining the three principle accounts

RESCHER STUDIES

of truth together. Thus the three books, each focused on one aspect of this ambitious account, are very closely connected. Rescher’s work in these books illustrates the very coherentism Rescher advocates. It is not a linear argument, aimed at mustering support for each conclusion in turn. Instead, it addresses a range of topics, reaching positions on each that fit together, complementing and reinforcing each other. It is throughout a struggle against the limits our epigram from Haller invokes. 2. THE COHERENCE THEORY OF TRUTH CTT begins with a discussion of two different forms that ‘theories of truth’ can take: definitional or criterial. A definitional theory provides an account of what ‘is true’ means, while a criterial theory provides an account of how we tell what is true. Criterial theories in turn can be guaranteeing or authorizing, where x’s meeting a guaranteeing criterion for being φ entails that x is φ, while x’s meeting an authorizing criterion for being φ merely justifies (provides a ‘rational warrant for’) the claim that x is φ. A theory that offers an authorizing criterion for truth need not provide a logically tight connection between its criterion and a definition of truth, although some sort of connection will have to be made out in the end. The correspondence theory of truth offers the metaphor of correspondence with the world or with the facts as a definition of truth. But Rescher argues that it holds little promise as a starting point from which to seek a criterion of truth. All seems well enough if we stick with observations, where some kind of confrontation with reality determines what sentences we are prepared to assert. But many other sorts of claims, from simple generalizations to modal claims, mathematical claims and counterfactual conditionals, don’t allow for direct confrontation between an epistemic agent and the features of the world that, on a correspondence view, would make such claims true or false. 3. INITIAL POINTS Rescher acknowledges that coherence with other propositions is not a serious candidate as a definition of truth; he proposes it only as a criterion for truth. This leads to an interesting question: how can coherence be a good criterion for deciding which propositions are true absent a definitional connection between truth and coherence? Rescher rejects the ideal-

60

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

ists’ metaphysical insistence that reality itself must be coherent, a view that provides a straightforward answer. Instead, he takes coherence and systematicity to be regulative ideals (CTT 39, CS 127 and elsewhere) governing our description of the world. Of course, if the world were entirely incoherent, a coherent description of it or of anything would be unachievable. Some degree of regularity and systematicity in nature is required for knowers such as ourselves to exist at all. But this starting point still allows the systematicity it demands to be limited and local. So, like the regulative presumption, it provides a starting point for inquiry; a hopeful initial assumption of a causally necessary condition for knowledge, later to be judged by success in application and the coherence of the resulting commitments. The pragmatic theory of truth is also a criterial theory, proposing that we should judge what is true by its pragmatic value, i.e. by whether acting on it serves our ends. Finally, what Rescher calls the intuitionist theory draws a sharp distinction between basic truths established in some immediate way and derivative truths established by an inferential process that takes the basic truths for its premises. This produces a tight connection between truth and its criteria: the basic truths are established by something like an observational confrontation with the way things are, and the derivative truths are inferred by a process that is truth preserving. Nevertheless it is at heart a criterial theory, centered on an account of how we decide what it is true. 4. THE CRITERION AND OTHER WORRIES The problem of the criterion makes the first of many appearances here; Rescher cites Sextus Empiricus’ stratagem for questioning any proposed criterion of truth, forcing its advocates to choose between circular reasoning and infinite regress: . . . (a)nd when the argument thus reduces itself to a form of circular reasoning the discovery of the criterion becomes impracticable, since we do not allow them to adopt a criterion by assumption, while if they offer to judge the criterion by a criterion we force them to a regress ad infinitum.(CTT, 13, from R.G Bury, tr., Outlines of Pyrrhonism, II, 20) Rescher recasts the argument in modern dress: Suppose the criterion C takes the form C: (∀p)[R(p) → T(p)], where T(p) iff p is true.

61

RESCHER STUDIES

Then an argument for the truth of p would run: C, R(p)/″T(p). But unless C and R(p) are settled truths, this argument must be suspected to be unsound. How, then, are we to show that T(C)? We can try the argument C, R(C)/″T(C) But this is circular, since it depends on having already established C. We can appeal instead to an alternative criterion, C1: (∀p)[R1(p) → T(p)]. But to show that an argument for C using this criterion is sound we need to establish C1, and the regress is underway.(CTT, 13) The question is what lesson we should draw from this. Rescher suggests first that logical and conceptual truths may be established without appeal to ‘an external criterion’ (CTT, 15); his strategy at this point is to undermine the problem by rejecting the notion of a single criterion that must rule in all cases, and seeking a source of truths that could provide, in the end, a well-founded criterion for other types of truth. This looks unpromising at first blush—surely it is a case of the second type of response, and vulnerable to the regress objection. But the problem of the criterion arises over and again through these works, and Rescher’s treatment of it needs to be examined more broadly before we evaluate it. Unlike a definitional theory, criterial theories of truth allow the possibility that some propositions that satisfy the criterion may in fact prove to be false: we are not guaranteed that C(p) iff T(p). The familiar distinction between Type I and Type II errors comes up here: a type I error occurs when T(p) holds but not C(p); a type II error occurs when C(p) holds and T(p) fails. Commitment to a general criterion of truth C is commitment to a policy of holding that p (and therefore that T(p)) when and only when C(p). In effect, we treat C(p) as equivalent to T(p)—but only as a faute de mieux approximation: the possibility of cases where the two diverge remains. One important consequence of this is that even when we accept p on the grounds that it meets our criterion of truth, when the risk of being

62

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

wrong outweighs the rewards of being right it is perfectly rational to refuse to act on p. 5. CHARACTERIZING COHERENCE Rescher defends a modest form of coherence theory, claiming merely that coherence is the ‘prime’ criterion of truth.(CTT, 23) He argues further that coherence must play some role as a criterion of truth for all tenable views—that is, in at least some cases, the coherence of a proposition with some others will be the best way to determine whether it is true. Consequently, anyone interested in truth at all should take an interest in coherence and how to evaluate it. But what is coherence? Rescher’s discussion begins with an examination of some historical forms of coherentism. Otto von Neurath’s coherentism is contrasted with Carnap’s intuitionist account of truth, grounded in fixed observation claims and inferences made from them: for Neurath even a protocol sentence may be withdrawn if it fails to fit into an overall system of protocol sentences and other commitments. Notably, Neurath’s description of coherentist methodology insists on consistency as a sine qua non; in his metaphor, the ‘machine’ of knowledge stops when consistency fails, and some protocol sentence must be withdrawn or the inferences made from some protocol sentences altered before it can be set to work again.(von Neurath, 207) Brad Blanshard argued that when coherence is taken as the prime test of truth, we must hold it is ultimately a ‘foolproof guarantee of truth’.(CTT, 30) Since such a guarantee can only hold for a definitional account of truth, Blanshard argued that coherentists have no choice but to take coherence as both the criterion and the meaning of ‘is true.’ But Rescher demurs, rejecting Blanshard’s assumption that the prime test of truth must provide a guarantee of truth: Even if coherence is our only test of truth, it need not be regarded as infallible. The pro-tem fallibility of coherence is obvious, since changes over time often lead us to reject, on coherence grounds, claims that we once accepted on grounds of coherence. Further, there is no reason to suppose that such cases are always ultimately corrected by new considerations. Consequently we can’t rule out the possibility that a claim that in fact always meets the coherence test still fails to accord with the facts.1 Whatever else we may say about it, ‘to cohere’ is a transitive verb; whatever coheres, coheres with something. Since it is claims or proposi-

63

RESCHER STUDIES

tions, or, as I shall say for the most part here, sentences that we are examining for coherence, it seems we must be talking about a relation between sentences. Logical relations between sentences—‘consistency and connectedness,’ as Rescher phrases it, are its main elements. Consistency is described as ‘an obvious minimum’(CTT, 32) condition for coherence, while the notion of connectedness demands further exploration. Some have defended the strong position that a coherent collection of propositions will be inextricably linked, i.e. that each implicitly carries with it as ‘context’ the others it coheres with. In its strongest form, such a view makes each statement equivalent to all the rest, offering only a different way of saying the same thing; coherence is either reduced to maximal redundancy, or embedded mysteriously in the content of each statement. Either way, the position is unilluminating at best. A weaker formulation due to A.C. Ewing holds that any one member of a coherent system can be inferred from the rest, and no proper subsystem of the system is independent of the rest of the system.(CTT, 35) However, Rescher observes that the first feature is trivial if we take the commitments to be closed under any reasonably strong relation of logical consequence: every finite conjunction of elements of the original set (and every disjunction including one of them, and every conditional with one of them as consequent, and so on) appears in the closure. Deleting one sentence inevitably leaves many others from which the omitted sentence follows, by familiar inference rules. Rescher’s second complaint (CTT, 38) is even more telling: on this account any finite set of logically independent propositions can be made coherent simply by adding the conjunction of its members to the set. Surely coherence requires something more than that! 6. AN EXTERNAL CONSTRAINT Rescher follows Neurath in rejecting fixed points such as Carnap’s observations, while retaining the point that observations (along with other sorts of data) still serve as indispensable input for knowledge. Inquiry begins with a ‘body of prima facie truths, i.e., propositions that qualify as potential—perhaps even as promising—candidates for truths.’(CTT, 39). Coherence is how we test these candidates, eliminating those that fail to cohere with the rest and endorsing as true the ones that best cohere with the others, “to ‘make the most’ of the data as a whole”.(CTT, 39) Coherence analysis aims to produce order out of a chaos of conflicting truthcandidates.(CTT, 40) Thus on Rescher’s account, truth and coherence are

64

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

the final results of inquiry, not its beginning. Rescher next turns to tactics for this process of winnowing and refinement. The classic objection, that coherentism offers no way to distinguish the truth from a coherent fiction, finds a straightforward answer here: coherence comes into play only when we already have a set of candidate truths. No fiction that does not result from a coherence-based winnowing of the candidates can meet the test. This also points towards an important aspect of a coherentist response to the problem of the criterion: coherence analysis gets underway only when we already have some sentences and methodological commitments to evaluate. The idea of ‘first knowledge’ is out of place here—if knowledge is a product of coherence analysis the notion that we can know one thing (whether a particular bit of knowledge or a criterion for knowledge) without knowing many others must be rejected. Of course Rescher’s appeal to data needs defense in its own turn. We will return to this issue below. 7. METHODOLOGICAL PRAGMATISM Rescher’s methodological pragmatism adds to this tempered coherentism a test for coherentist method that draws on something other than coherence. It would be circular to apply a pragmatic criterion for methods to evaluate the success of a pragmatic method for determining the truth of individual sentences; Rescher concludes from this that a non-trivial use of the pragmatic criterion to evaluate methods is only possible when the method being tested is not pragmatic.(CTT, 242-3) Further (in an echo of the problem of the criterion) applying the cognitive standard of knowledge of truths to evaluate the success of a method for arriving at truths is similarly circular.(CTT, 242) The only remaining option is a pragmatic justification of a non-pragmatic criterion of truth. MP develops this proposal into a detailed pragmatic approach to the evaluation of methodologies. Rescher argues that the methods of science (allowing for variations in detail from field to field) fall under the rubric of coherence-based reasoning. Consequently, the success of science in terms of predictive power and control provides good pragmatic reasons for accepting coherence as our over-arching criterion of truth. Of course no guarantees of success are possible—Rescher’s stand on the force of this justification is modest: it is a case of “ ‘this or nothing better’ ”(CTT, 259). But the pragmatic standard for methodology does endorse coherence

65

RESCHER STUDIES

as our best choice of method, and will do so for as long as coherence methods continue to succeed pragmatically. Evolutionary considerations contribute to Rescher’s pragmatic case for coherentism: if coherence were not reasonably successful by pragmatic standards, it would not have survived as a method for fixing belief. This appeal to evolution reaches beyond coherence to make an appeal paralleling Quine’s famous remark about the fate of those ‘inveterately wrong in their inductions’(Quine 1969, 126) As Rescher explains in chapter VIII of MP, the evolutionary defense of coherence based on methodological pragmatism resists an objection that is difficult for an evolutionary defense of sentence-by-sentence pragmatism to answer: belief in certain ‘convenient’ falsehoods is quite likely to be compatible with successful action in the world. But when it comes to methods rather than individual claims, we have few to choose from and many opportunities to put them to the test. Consequently, given our natural concern for pragmatic success, methods that survive are very likely to be at least as successful as any alternatives we have been able to turn up. Rescher is confident that it is scientific methods, summed up by the notion of coherence, that best pass this test of time. One point in favor of this claim is that non-scientific methods survive today largely in niche applications where no method seems to work notably well: for example, crystalball gazing, fortune telling, psychics and other ‘paranormal’ methods focus their efforts on claims that we have no established way of settling on presently available evidence. Moreover, they concern topics that people are highly interested in—in such cases, accepting ignorance is harder and clutching at straws more common. But if that is the best we can say for these methods, they are not serious competition for the scientific coherentist. Further, coherence is an open-minded standard: any lower level method whose results regularly meet its requirements will be endorsed as reliable on coherence grounds. The appeal to evolution also plays a role in Rescher’s response to the problem of the criterion. It provides another external source of evaluation or feedback, one that Rescher argues is sufficient to ground the presumption of truth that is needed to get coherence analysis off the ground. This point combines with the earlier appeal to a distinction between “definitional (conventional), logical and conceptual truths” and “factual truths”(CTT, 15). For Rescher the former are not up for grabs under coherence analysis, which aims only to refine data about contingent facts concerning how things stand in the world. If we can show that a justifica-

66

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

tion of coherence is rooted outside of the factual world, then the regress may break down. Again, it’s important to note that the bar is set at a modest height: all we need is a defeasible presumption in favour of coherence. Together these two points undermine the skeptic’s claim that even tentative adoption of coherence methodology is arbitrary and unjustified. 8. SCIENTIFIC PROGRESS Insofar as science leads us to adopt general theses of indeterminate scope, Hume’s critique of induction ensures our conclusions must be tentative. This rules out even objective probabilities as an account of what scientific inquiry provides us. (MP, 172) Avoiding error is no solution either: it will not prevent an unhelpful, random walk from one error to another, and we will never run out of erroneous positions to consider. The upshot of these worries is a generally negative attitude in philosophy of science towards ideas about scientific progress—something that strikes outsiders as very strange indeed, given the burgeoning growth of successful scientific ideas and technology. But if truth is so hard to come by, and progress requires some relation to truth, whether in the form of increasing amounts, or better approximation to some specified truth, then we are stuck: There is no criterion for truth outside of or independent from science that we could use to show that science is really making progress towards the truth in some such sense. But on Rescher’s account this worry does not rule out progress. The methodological turn that is the principle theme of MP comes to bear again here. Science does improve, by providing better methods for testing and answering questions—better not in terms of some independently determined truth, but in terms of ‘thesis substantiation’.(MP, 175). Perhaps the new methods extend the old, applying in new cases or conditions where the old failed. We can also point to cases where new measurement techniques improve both precision and agreement between independent measurements—here we see progress, in the absence of any proof that the results are somehow closer to an independently-defined truth. The key to establishing improvement is dominance: New methods must be an advance on the old in some cases and be at least as good in all. Applying cognitive coherence more broadly, we may also be able to show our beliefs about what is being measured and the principles at work in the measurement imply that the new technique is indeed more reliable and precise. And prag-

67

RESCHER STUDIES

matic efficacy is invoked as well, as a theory-independent measure of success. Rescher advocates a ‘Copernican inversion’ shifting us from an inference of the form ‘because these claims are true, they cohere,’ to one of the form ‘because these claims are coherent, they are true.’ We don’t show that a new method is better by showing its results are ‘truer’; instead, we conclude that the results are truer because the new method is better. This inversion, of course, is part of Rescher’s response to the wheel argument. It also points towards the regulative status of coherence: for Rescher it is a defeasible presumption of inquiry that the truth will cohere and be practically efficacious. Consequently, we measure our approach to the truth by our success at refining and improving the coherence of our system of beliefs as well as our success, when applying it, at prediction and control. The upshot is a two-sided test in which cognitive systematization (explanatory adequacy, coherence) and pragmatic efficacy (problem-solving and control) serve as separate but equal measures of the cognitive success of our methods. A similar double cycle of theoretical and practical evaluation also appears in CS (103). The details of the ‘scientific method’ are left for later consideration—the general focus on coherence and inferential linkages required for successful explanation, prediction and control is sufficient. Rescher also invokes practice to avoid the difficulties about theoretical incommensurability famously invoked by Thomas Kuhn.(MP, 185) Successful application shows that the improvement is not just ‘in the believer’s eye’. Standards of practical success show rich continuities despite the conceptual gaps that separate earlier and later theories. In particular, there is continuity in the technology of measurement, in laboratory practice, in standards for evaluating measurements as more or less precise and reliable, in the production of well-known effects and many other aspects of scientific practice. This ensures a practical link between the standards of success applied to predecessor and successor theories. Rescher cites medical examples in particular (MP, 186); similarly, William Smith once wrote of his biostratigraphic results that, whatever explanation might be offered of them, “Yet this one thing full well we know— how to find them ordered so.”(Smith, 1829, in Albritton, 116) MP also includes a substantial response to skepticism that begins with reflections on the idea of the burden of proof (MP, 201f). The main point of this discussion is that while in the courts and in debates the proposing party is held to carry the initial burden of proof, this burden can be shifted.

68

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

If things were otherwise, that is, if nothing the proposing party could say or present in evidence imposes a burden of refutation on the other side, the proposer could never win. Whatever such an exchange is, it is not a real debate. The lesson about skepticism is simple: when a skeptic simply rejects any and every claim or principle presented by a defender of some knowledge claim, she is cheating. It may be philosophically challenging to decide at what point the burden shifts to the skeptic, but if it cannot ever be shifted, what is going on is not a rational discussion. Finally, chapter 11 of MP addresses the question of logical pluralism, an issue receiving substantial attention again today. Could there be more than one logic? Rescher’s position on this question is a moderate one. While the pragmatic standard allows room for considering different logics, there is some absolutism about basic principles, including the centrality of truth-preservation. 9. COGNITIVE SYSTEMATIZATION In CS Rescher focuses on the role of systematicity as the overarching methodological guide to epistemic reasoning. While the pragmatic argument in MP underpins this commitment to systematization, Rescher links his view here to Hegel, who inverted the standard inference from truth to system, arguing instead from systematicity to truth. (CS, 33-6) Another important step in the argument of CS claims that coherence (as Rescher proposes it) provides a good account of scientific reasoning, which is a paradigmatic example of pragmatic success (at explanation, prediction and control). This clinches the pragmatic case for coherentist methodology first raised in CTT and extended in MP.(CS, 106-11) As in MP, the dual cycle of feedback- testing methods for the coherence of their cognitive outputs, and for their pragmatic vindication in prediction and control, is emphasized. This allows Rescher to accept a certain circularity on the cognitive side, while mitigating its impact by appeal to practical success.(CS, 110) A regulative presumption of the value of coherence is checked and confirmed by success in application: “ ‘Design your cognitive procedures with a view to the pursuit of systematicity!’ is a regulative principle of inquiry whose legitimation ultimately lies in its being pragmatically retrovalidated by its capacity to guide inquiry into successful channels.”(CS, 127) The emphasis is on present results as opposed to ideal ends: This seems to me a crucial insight, with implications for the realism/anti-realism de-

69

RESCHER STUDIES

bate in philosophy of science. The present success of scientific theories, imperfect and presumably ‘untrue’ as they are, is our best evidence for an irenic realism.(CS, 127) It is not that we have or should claim that we have the truth now. But we have reason to claim that our present position is truer than the alternatives, because it is more systematic, coherent and successful. Truth in this sense is not the correspondence that Rescher accepts as the correct definition of truth—truth in the definitional sense is most emphatically not a matter of degree. However, we do have an improving accord between our criterion for judging truth and the features of our best present account of things. Beyond this, ultra posse nemo obligatur.(CS, 141) Chapters IX through XI of CS explore questions about the limits of cognitive systematization. Rescher distinguishes three kinds of limits: incompletability, inconsequence and inconsistency. Chapter IX focuses on the possibility that our system of knowledge is incompletable. While some apparently unanswerable questions may come to be set aside as improper, i.e. as having false presuppositions, others are likely to remain. And practical limits imply that even some answerable questions will not, in fact, be answered. Still, Rescher retains completeness as a regulative ideal: incompleteness implies a high probability of at least some incorrectness, since it limits the use of coherentist methods to cross-check our conclusions.(CS, 152) Inconsequence is the focus of Chapter X: the more isolated our commitments are, the less coherent they are; explanatory completeness requires all of our commitments to be connected in such a way that some other commitments explain them. One form of explanation involves “subsumptive inferences” bringing what is to be explained under some (possibly probabilistic) covering law(s), given some background conditions.(CS, 154-5) This leads to an explanatory regress, since each explanans needs explanation in turn, so completeness is impossible for such explanations. A broader conception of explanation reflects a richer range of connections between our commitments. Commitments accepted on these grounds have an “explanatory rationalization”(CS, 156) within the system of science. This can hold for every member of a system of claims, so completeness is in principle achievable for this sort of explanation, and in fact constitutes part of Rescher’s ideal of cognitive systematization. However, the notion that our explanatory efforts will reach a final conclusion which will not be corrected or improved on is unrealistic given the prospect of more precise and reliable observations of more and more phe-

70

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

nomena and the possibility of conceptual innovations. “This “inexhaustibility” of the potential knowledge of things is implicit in the very conception of a “real thing” as it figures in our conceptual scheme.”(CS, 159) Our inquiry might reach a state that it will never improve on any significant way, but we could never be in a position to judge confidently that it was so.2 In particular, our ability to create and observe extreme conditions of temperature, energy, and other physical parameters is clearly limited; we must always be somewhat unsure of what would happen at more extreme values of these parameters. The autonomy of science generates another sort of completeness— nothing that science can explain can be explained in some other way or as a part of some other enterprise, and there is no source other than science for the correction or improvement of scientific understanding.(CS, 161-2) Chapter XI confronts the third and most drastic type of limit: the failure of consistency. Inconsistency might block the project of systematizing our knowledge, and both incompleteness and ‘inconsequence’ may lead to inconsistency. For example, curve-fitting or gap filling efforts guided by overlapping, conflicting principles could lead to inconsistency, and attempts to compensate for indecisiveness can also produce inconsistency.(CS, 172) The present tension between relativity and quantum mechanics presents a situation where the best theories in distinct but related domains don’t mesh consistently. Such conflicts may be resolved, but there is no guarantee of this. We might wind up instead end up in the situation of Bohr’s motto, contraria sunt complementa, in which incompatible commitments play indispensable roles in our cognitive commitments. The pursuit of systematization itself may lead to inconsistency; the demand that we connect and fill gaps in our commitments makes avoiding mistaken or conflicting commitments at all costs unreasonable. Rescher holds that toleration and coping are possible here. Although the coherence analysis developed in CTT always leads us to a consistent position, the costs of imposing that procedure may be too high: there may be no non-arbitrary way to restore consistency, and the reasons we have for retaining each element in the inconsistent set may outweigh the value of arbitrarily imposed consistency. The fact that consistency is a regulative ideal for inquiry does not show it is achievable; at any point short of an ideal outcome, the costs of eliminating inconsistency may exceed the costs of tolerating it. This is a very serious threat to coherence. Rescher’s treatment of the logic of truth in CTT emphasizes the inferential links between our com-

71

RESCHER STUDIES

mitments. Negation inconsistency leads to disaster given the logic of truth developed in CTT: every sentence follows from such commitments. Further, the cost of giving up closure of our commitments under consequence is extremely high. Reducing this cost requires a new logic that preserves much of the inferential structure of our commitments while preventing the inconsistency from trivializing our commitments altogether. CS concludes with an examination of “cognitive metasystematics,” systematic relations between parts of our cognitive commitments. A natural approach here is to invoke a hierarchical taxonomy of inclusion, with (for example) sub-fields grouped into fields, fields into disciplines, and so on from the most general to particular fields, specialties, and problems. But interconnections make this not nearly as tidy as it is in biological taxonomy. Rescher envisions a network of relations, with increasing specialization balanced by a growing variety of ‘cross-linkages’ between different fields. Increasing complexity is another theme here—both the complexity of nature as we understand it and the richness of opportunities to engage with and explore nature as new instruments and concepts are developed are rapidly growing. 10. LOGICAL MATTERS Coherence analysis and Inconsistent Data. One of the aims of CTT is to explore how consistency, conceived as a sine qua non for coherence, can be imposed on a collection of sentences taken as data, i.e. as ‘putative truths.’ Coherence analysis takes data as input and proceeds to refine them into a system of claims that we can be justifiably committed to. The standard language of propositional logic is taken for granted here, as ways of transforming inconsistent data set into a best (most plausible) estimate of the truth are explored. The method Rescher applies to ‘clean up’ the data begins with maximal consistent subsets (MCSs) of the data—these can be considered our best candidates for truth at the outset: each eliminates the inconsistency of the data while retaining as many as possible of our putative truths. Rescher adds an important further element to his notion of data here: not every MCS is equally plausible. The use of both plausibility rankings and of probabilities for our data are explored. In the case of plausibility rankings, an overall threshold is proposed for separating MCSs that are sufficiently plausible to warrant acceptance from those which are not. If one MCS survives this cut, it becomes our best present take on the truth. Failing that, there may be more than one

72

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

that ranks highly enough to deserve consideration as a possibility. Similar tactics can be applied in the probabilistic case as well. In general, Rescher proposes that we render our commitments consistent by endorsing the shared consequences of the sufficiently plausible MCSs. I take inference to be the addition or elimination of epistemic commitments on the basis of reasoning about logical relations between our commitments. As G. Harman has often pointed out, logical consequence relations do not provide a theory of inference, because we can respond to a valid argument whose premises we have accepted either by accepting its conclusion, or by rejecting one or more of the premises that it follows from. However, logical relations between sentences still structure our inferential options: when we are confronted by a logically valid argument whose premises we accept, we can accept the conclusion, reject at least one premise, or (an option that is often neglected) weaken the logic we apply to our commitments. The last option is the important one here. At the end of the day, Rescher does not always insist on the elimination of inconsistency. The price, in terms of rejecting commitments that are strongly supported by other considerations, is not always worth paying. This calls for a logical investigation of inconsistency, its relation to the idea of consequence as guaranteed truth-preservation and for new ideas about how to avoid logical disaster when reasoning with an inconsistent set of commitments. 11.

TOWARDS A BROADER LOGIC OF TRUTH

The standard definition of inconsistency is due to Post: an inconsistent set has every sentence in the language as a consequence. Commitment to such a set will clearly be incoherent so long as we continue to regard our commitments as closed under the consequence relation—and giving up closure tout court is incompatible with the idea, central to coherentism itself, that our commitments have contents that connect them. This makes inconsistent commitments disastrous. But Rescher regards this threat of trivialization as linked to a very particular “body of logical machinery”(CS, 175), and suggests that with the help of different logical machinery inconsistencies might be confined to a narrow and isolable region of our commitments. Changing the consequence relation transforms the contents of our commitments in a far-reaching way. But if the new logical machinery preserves enough of the inferential links that connect our commitments and make them coherent, this price may be worth paying.

73

RESCHER STUDIES

Before considering examples of such consequence relations, we shall set the entire enterprise in a more general context. The logic of truth developed in CTT, together with the notion that the consequence relation preserves truth, make paraconsistency unattainable. We begin by generalizing on the logic of truth that Rescher develops in CTT (142-7). Rescher’s approach to the logic of truth can be pursued without assuming the standard propositional language Rescher employs. Abstracting from this particular language focuses attention on some key ideas that deserve notice, in particular, the structural rules of consequence. We will assume only a consequence relation defined on the sentences of our language. In general, acceptance (or its dual, denial) of some sentences will commit us to acceptance (or denial) of others. Given the usual structural features of the consequence relation, we can go on to say some interesting things about coherence and its preservation. We begin with a language L; SL is the set of sentences of L and L ∈ P(SL) → SL is L’s consequence relation. We take it that the consequence relation preserves commitment, that is, if someone accepts some sentences s, s', . . . , then if sc is a consequence of s, s', . . . , she is committed to sc as well. On discovering this, she may retain her other commitments and accept sc, she may reject sc together with some sentence(s) in s,s', . . . or she may find fault with L and propose a change to the consequence relation that eliminates this consequence. We say that a set of sentences {s0,s1, . . . } is inconsistent if and only if ∀s∈SL, {s0,s1, . . . } s. Finally, we assume that our consequence relation is reflexive, monotonic and transitive. As Dana Scott notes (Scott 1974), any such consequence relation can be captured by a set of 1/0 valuations on SL: Γ α will hold if and only if every allowed valuation that assigns 1 to every member of Γ also assigns 1 to α. Moreover, every consequence relation that can be captured by such a set of allowed valuations is reflexive, monotonic and transitive. So the standard reading of ‘1’ as truth, ‘0’ as falsehood makes the consequence relation truth-preserving. Thus our assumptions so far are compatible with a theory of truth applying to the sentences of the language and with the standard equation of truth-preservation with consequencehood. We can now introduce some important inferential properties and relations of sentences. A sentence α is absurd iff ∀β, {α} β. A sentence α denies a sentence β iff ∀γ, {α,β} γ. A sentence α non-trivially denies a sentence β iff α denies β and α is not absurd. (Note that denial is symmetrical while non-trivial denial is not.) A language L has denial iff ∀α ∈ L,

74

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

∃β ∈ L: β denies α. Finally, a language L has non-trivial denial iff L has denial and ∃α,β ∈ L: β denies α, α is not absurd and β is not absurd.3 One sentence is said to be a negation of another if and only if it is a weakest contrary. Thus β is a negation of α iff β denies α and ∀γ: γ denies α → {γ} β. A language L has negation iff there is such a weakest denial for every sentence α in L. Interesting results for the logic of negation follow: since every sentence denies4 an absurd sentence, the negation of an absurd sentence must be a consequence of every singleton in P(L). By monotonicity, this negation will follow from every non-empty subset of L.5 We call such sentences trivial, and every trivial sentence is a negation of every absurd sentence: Transitivity implies only an absurd sentence can have a denial of itself as a consequence: if β denies α and {α} β, then by the definition of denial, ∀γ:{α,β} γ, and by transitivity ∀γ:{α} γ. Similar inferential treatments of other connectives can also be given. A sentence β conjoins Γ iff β follows from Γ and everything that follows from Γ follows from β (more formally, β is a conjunction of a set of sentences Γ iff β ∈ L: Γ β and ∀γ: Γ γ ⇒{β} γ). β disjoins Γ iff β follows from each member of Γ and the singleton set of β proves everything that follows from each member of Γ (more formally, β is a disjunction of Γ iff β ∈ L: ∀γ ∈Γ: {γ} β and ∀δ: ∀γ ∈Γ: {γ} δ ⇒{β} δ). L has conjunction iff every finite set of sentences in L has a conjunction, and L has disjunction iff every finite set of sentences in L has a disjunction.6 These inferentially defined types of sentences are related in familiar ways. Suppose L has negation. Then, where ‘neg(β)’ is a negation of β, If {α}

β, then for negations of α and β, {neg(β)} neg(α).

(Proof is by the definition of negation, monotonicity & transitivity.) Further, α

β ⇒ {α, neg(β)} is absurd.

(Proof is by the definition of negation, monotonicity and transitivity.) But the converse implication fails. Finally, If {γ} α, then {γ} neg(neg(α)). By the definition of negation, if {α,β} is absurd, then α neg(β). So α

75

RESCHER STUDIES

neg(neg(α)). Suppose {γ} α. Then by transitivity {γ} neg(neg(α)). We don’t get the classical if and only if form of this theorem because neg(α) is by definition the weakest denial of α but we do not require that α be the weakest denial of neg(α). So α may be stronger than neg(neg(α)). Finally, assume L has negation and conjunction, where ‘or(α,β)’ and ‘and(α,β)’ are, respectively, a disjunction and a conjunction of α, β. Then ∀α,γ {and(α,neg(α))} γ. (Proof is by the definitions of conjunction and negation.) On the other hand, even if L has negation and disjunction we don’t have ∀α, ∅ or(α,neg(α)). By definition this disjunction must follow both from {α} and {neg(α)}. So neg(or(α,neg(α)) neg(α) and neg(or(α,neg(α)) neg(neg(α)), and so implies and(neg(α),neg(neg(α)). This is a conjunction of a sentence and its negation, and therefore absurd; its negation, neg(neg(or(α, neg(α))), is a triviality. But the double negation cannot be removed to obtain the result, since we don’t have DN in the right direction. These results show that even when the usual connectives are not present in a language, they can be conservatively added so long as the language has negation, conjunction and disjunction in the inferential sense, that is, so long as some sentences or other play the required inferential roles. Still, not all languages will have this expressive power.7 This allows a little more breathing room to defenders of logical pluralism, including Rescher himself, as well as to those who advocate alternative logics that don’t meet the conditions Rescher imposes here. This asymmetrical logic fits nicely with Rescher’s discussions of the logic of truth in CTT, but it still makes inconsistency disastrous. However, thinking about how its asymmetry arises leads us to the possibility of another consequence relation: running from right to left, this is the relation under which commitment to denial of a set on the right is closed. This leads to an alternative account of ‘denial’, now defined in terms of right-toleft trivialization: β denies α iff, ∀γ, γ {α,β}. This negation is the strongest such denial, i.e. ∀γ, γ {α,neg(α)} and ∀β: if β denies α, then neg(α) {β}. Similarly ‘right-handed’ versions of conjunction and disjunction can be defined, reversing the asymmetries noted above. The result is a paraconsistent negation dual to the more familiar negation above: in traditional terminology, a left hand negation is a weakest contrary, while a right hand negation is a strongest subcontrary.

76

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

Both types of negation can be accommodated within a set-set consequence relation defined as holding when and only when every assignment that assigns ‘1’ to every sentence in the set on the left assigns ‘1’ to at least one member of the set on the right (or equivalently, when and only when every assignment that assigns 0 to every sentence in the set on the right assigns ‘0’ to at least one member of the set on the left). Classical logic results when we identify the right and left hand operators, that is, when our negation is self-dual and conjunction and disjunction are dual. These two negations raise the possibility of avoiding a problem with Rescher’s response to the liar. In CTT Rescher addresses the threat of paradox in a semantically closed language, raised by (Tarski, 1949)8: Rescher’s response there is that, because the logic of truth as he expounds it is not two-valued, Tarski’s argument does not go through. However, the strengthened liar does not assume bivalence: SL: SL is not true. So the liar paradox remains a serious problem here. But it’s only the right hand sense of negation for which SL is true or SL is not true must hold, and this negation is paraconsistent because it forms a subcontrary, not a contrary. As a result, the strengthened liar argument does not lead to trivialization. This somewhat more abstract logical framework fits nicely with the feedback cycle pictured on p. 276 of MP, in which the interaction between inferential practice and logic is adjusted as the process of inquiry produces commitments whose systematization requires refinements of inferential practice and these subsequently guide revisions of inferential practice and logic. 12.

PARACONSISTENCY

In the languages described above, absurdities as well as any set including a sentence and its left negation are inferentially explosive. But we can modify our initial consequence relation to avoid this disaster. We have assumed so far that guaranteed truth preservation determines what follows from what. So when our premises cannot all be true, anything and every-

77

RESCHER STUDIES

thing follows from them. To avoid this, we must either accept that some inconsistent collections of commitments could all be true and adopt a consequence relation preserving this new and generous view of truth, or we must find some other desirable property of our commitments and adopt a consequence relation that preserves it instead. Important examples of the first approach to paraconsistent logic include Graham Priest’s ‘logic of paradox’ (LP) and first degree entailment (FDE). In these logics no sets are inconsistent by our definition, but we still retain quite a lot of the usual inferential machinery. The standard semantics for these logics produce pretty drastic consequences for our understanding of truth: there is no barrier to contradictions being true, and in FDE there is no barrier to tautologies being false. So contradictions go from being catastrophic, utterly trivializing our commitments, to being no worry at all— the general sense that contradictions and inconsistencies detract from our commitments’ coherence even if they don’t utterly destroy it now needs philosophical support, since the logic alone copes with them easily. This runs against the tone of Rescher’s discussion in CS, where the demand for consistency remains in force as an important regulative ideal. Further, on the practical side of our commitments, supposing that such commitments can all be true, it seems we should be able to use them as premises for practical reasoning. Suppose for example that our commitments are leftnegation inconsistent, i.e. we have a commitment to both p and ]p for some sentence p. Now, when the question whether p is important to a decision, we have some notion of how the assumption that p and the assumption that ]p will bear on the choice. But how ought we to act on the assumption that both p and ]p hold? The second approach to paraconsistency avoids these problems. It also points towards a concrete version of a suggestion from Rescher: “This local-anomaly theory indicates that consistency too is a matter of degree.”(CS, 175) If consistency is a matter of degree, then there should be some way of measuring it. And once we have a measure of degrees of consistency, our new consequence relation can be required to preserve degrees of consistency just as standard consequence relations preserve strict consistency. An illuminating way to define a consequence relation is: Γ

78

α iff ∀Γ' Γ' ⊇ Γ and Γ'  ⊥ ⇒ Γ', α  ⊥

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

where {⊥} has every sentence as consequence. That is, α follows from Γ if and only if α consistently extends every consistent extension of Γ. This is related to the notion that in accepting the consequences of our commitments we beg no questions: no consistent (i.e. acceptable) extension of Γ is ruled out by closing Γ under the consequence relation. A logic that preserves degrees of consistency in this way will ensure that the extension of a premise set by the set’s consequences will not rule out (as making things worse) any extensions of the set that don’t make things worse. The result would be a natural generalization of consequence relations that preserve consistency that does not simply give up when consistency is lost; instead, these logics will preserve whatever degree of consistency our commitments have while begging no questions about how those commitments could be extended while not making things worse. In fact, an alternative interpretation of LP and FDE allows us not only to continue to regard inconsistency as a serious flaw in our commitments but also provides just this sort of measure of how bad that inconsistency is. The measure is based on the idea of an ambiguity set, the set of least sets of atoms whose ambiguous treatment allows us to produce a consistent image of our commitments. The consequence relation of LP preserves a feature of ambiguity sets—the ambiguity set of an acceptable extension of a set of sentences must be a subset of the original set’s ambiguity set; in turn, an LP consequence of the set is a sentence that preserves this feature when it is added to any such acceptable extension. On this approach, inconsistency is managed by a kind of restricted trivialization: in order to make the contradiction ‘p ∧ ] p’ true, we treat ‘p’ as ambiguous, and call any sentence including ‘p’ true if there is some way of distributing classical (1/0) values across the instances of ‘p’ in the sentence that would make the sentence true. The real logical work of determining non-trivial consequences is done by the partial classical assignment to the atoms which are not treated ambiguously. The paraconsistent logic pursued by Rescher elsewhere (Rescher and Brandom, 1980) takes a different approach. Wittgenstein held that it was not a serious problem for a contradiction to be potentially provable in a system, so long as our proofs did not take advantage of this. One way to avoid deriving contradictions from an inconsistent set is to reject aggregation, in which we infer from two or more sentences a sentence that follows from all of them together but not from any one of them individually. Rescher and Brandom’s logic eliminates aggregation altogether. However, we can take a more measured approach by distinguishing how much ag-

79

RESCHER STUDIES

gregation can be safely applied to our commitments. A consequence relation is aggregative if and only if some sets of sentences have consequences that don’t follow from any singleton subsets. One measure of aggregation in a consequence relation is whether the aggregation it creates survives divisions of premise sets amongst some number of cells. Given our definition of aggregation, no aggregation can survive divisions of a premise set of cardinality n amongst n cells. So if we allow all divisions of our premises, we get a fully non-aggregative logic, and the resulting consequence relation is identical with our base logic only if the base logic itself is non-aggregative. On the other hand, if we allow no divisions we retain all the aggregation of our base logic. If this is the only way to retain the consequences of the base logic, the base logic is fully aggregative. In between these extremes things are more interesting. A logic that survives all divisions of any premise set amongst two cells is aggregative but not fully aggregative; the same applies for divisions amongst any finite number of cells. Classical logic, of course, is fully aggregative; conjunction introduction allows us to obtain the closure of any premise set under conjunction. But in general conjunctions don’t follow from any proper subset of their conjuncts. A logic that dispenses with full aggregation but endorses the rule 2/3, allowing us to infer from any triple in a premise set the disjunction of pairwise conjunctions amongst the triple, survives 2-cell divisions of premise sets; in general, 2/n+1 aggregation (forming the disjunction of pairwise conjunctions amongst any n+1 premises) survives all n-celled divisions of our premise sets.9 We can also adapt our principles of aggregation to suit the amount of aggregation that is safe for a given premise set. In such a case, what we preserve in the consequence relation depends on the amount of division necessary to avoid deriving a contradiction. The resulting logics were first proposed by P.K. Schotch and R.E. Jennings—one such logic takes classical logic as its base but restricts aggregation to avoid trivialization, preventing inconsistent sets from exploding except when they include at least one (classically) contradictory sentence. Like other preservationist forms of paraconsistency, these logics underwrite the regulative ideal of consistency while providing means of coping when imposing consistency is too costly. Taken altogether, these developments in logic since the publication of CTT, MP and CS provide additional support for the general thrust of Rescher’s ideas about logical pluralism and inconsistency-tolerance.

80

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

13. THE WHEEL The problem of the criterion, also called the dialellus or wheel argument, plays a central role in all three books. It poses a fundamental challenge to any criterion of truth; responding to it drives much of the dialectic of these books even when it is not explicitly on the table. Rescher’s response is complex, with different elements emerging at different points in the exposition. To put the problem in context, we begin with a brief synopsis of Roderick Chisholm’s response to it. Chisholm distinguishes particularist and methodist responses to the problem. A particularist begins with some particular truth claim(s); if these are able to stand on their own, without regard to any consideration of a criterion, we might then hope to arrive at a defensible account of a criterion that would in turn allow us to identify further truths. A methodist, on the other hand, begins with some criterion which is to stand on its own; once we have such a criterion established, we can apply it to arrive at particular truth claims and begin the process of producing knowledge. At first both responses seem untenable. The particularist begins with truth-claims for which she has no criterion, and the methodist offers up a criterion which has not been tested against any established truths and is not justified by appeal to some other, higher-order criterion. Surely the skeptic will not be impressed by either! Whether we try to begin with particulars or with methods, such first claims seem equally arbitrary (if they did not, the problem wouldn’t get off the ground in the first place). Chisholm responds by framing the issue as a forced choice between particularism, methodism and skepticism; from this perspective, skepticism looks very unattractive even if the skeptic’s objections to the other two are telling. Chisholm buttresses his particularism by appealing to G.E. Moore’s famous rejection of skepticism as less-well grounded than the knowledge claims it is supposed to undermine, and it is fair to say that Moore’s central examples focus on particular instances of knowledge. However, Moore’s approach can also be applied to methods, with equal credibility. For example, I think it’s pretty clear that ‘looking and seeing’ is a good method for arriving at certain kinds of true claims. In fact, like Moore, I may well be more confident of this than I am of the premises of our skeptical argument. Chisholm explicitly denies that his particularism involves any commitments to criteria. But his denial is almost startlingly irenic: “I would say— and many reputable philosophers would disagree with me—that, to find

81

RESCHER STUDIES

out whether you know such a thing as that this is a hand, you don’t have to apply any test or criterion.” (Chisholm, 72) But Chisholm’s allegiance to particular claims in the declared absence of a method for evaluating the standing of particular claims cannot, in the nature of the case, be any less vulnerable to skeptical challenge than the methodist’s commitment to a method absent any means for evaluating methods. As a foundationalist Chisholm maintains that some particular claims are justified in themselves without reference to any other claims or appeal to some method: they are, as he puts it, self-justifying. But to a skeptic pressing the problem of the criterion, this looks just like a method: the fact that we find ourselves making certain claims is itself taken to justify the claims, i.e. our method for determining the truth of these claims is to accept the claims we find ourselves making. This method doesn’t apply to all claims; for Chisholm it is true first and foremost for first-person claims about the character of our experience. But what justifies this automatic endorsement of some claims but not others because we find ourselves making them? The foundationalist needs to distinguish her foundational claims from other claims that require support from some other source, and it seems some sort of criterion must creep in here. 14.

A COHERENTIST RESPONSE

When a question seems insoluble, one important kind of response is to find a framework in which it is soluble or improper. By contrast with both Chisholm’s particularist and methodist, a coherentist rejects the very idea of a first piece of knowledge which can stand on its own. Our groundlevel commitments and our commitments to criteria of truth must fit together, each supporting the other. At the outset neither has more than a presumptive claim to truth, but after coherence analysis, success in application and evolutionary persistence, these commitments have real standing. A coherentist response to the wheel argument need not treat coherence as a proposal for what the criterion is, that is, as the starting point for a process that will lead us to particular knowledge of fact. Instead, we treat coherence as a value that guides our efforts to improve on a starting point that already has credibility, though not sufficient credibility to be regarded as true—and coherence itself has a similarly tentative status at the outset. Rescher begins, in CTT, by proposing an analytical status for coherence. But this analytic status need only be strong enough to shift the burden of proof, that is, to create a presumption that it is an acceptable crite-

82

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

rion to begin with, and will be judged, in the end, by how it works out. Evolutionary considerations (emphasized in MP) and a mutually supportive pair of pragmatic and systemic cycles of evaluation (emphasized in CS) are also brought to bear. No part of the system is asked to stand on its own; it is only as a whole, and in the light of the whole, that we can confidently assert anything. Finally, to return to MP, if there is any interest in the skeptics’ challenge, there must be some way for a non-skeptic to shift the burden of proof onto the skeptic. Thus Rescher’s coherentism leads us to reject the terms of the problem of the criterion. The skeptic here is playing a very strange game whose terms ensure that she will win, though they superficially (and only locally) resemble the rules by which we evaluate and criticize assertions. Her assumption is that everything is entirely up for grabs, that no claim, either particular or about criteria, has any standing that requires a response from her. But to take that stand is to deny, among other things, that one has any idea of what it is for a sentence to be true. It’s hardly surprising that such a starting point leads to skepticism. Most importantly, this stand implies that there is no way for a defender of any claim or criterion to shift the burden of proof to the skeptic. But that is not the game of justification that we learned at mother’s knee—it is something quite different, and quite boring: a game that is trivially won for the skeptic. We would still like to have an answer to a more constructive question: how does (and how could) knowledge get started? The coherentist responds that, if we start from zero, with no idea of what the particular facts are and no idea of what methods we can use to identify truths, then there is no way to recover our knowledge. This seems to end the discussion with an obvious insolubilium: after all, both as individuals and as a species, we started out with no knowledge at all. How could human knowledge get started, if knowledge requires justification, and justification requires appeal to further knowledge? But coherentism points in a more hopeful direction—it suggests a need for holism, and a rejection of the search for ‘first knowledge’. And Rescher does more than simply invoke holism—his response describes a complicated, interwoven path that draws on data, coherence and pragmatic success to provide what he modestly characterizes as a response that is as satisfactory as any could be. Beginning with claims and methods adopted only tentatively, we can apply coherence analysis to transform them into a substantial, well-supported view of the world. It’s only from the later

83

RESCHER STUDIES

standpoint that we can recognize that some individual commitments are well justified.10 Chisholm’s model is, in Rescher’s taxonomy, intuitionist; it requires that we begin with fixed truths, and reason in a way that is at least likely (on the whole) to preserve truth. But Rescher’s model is revisable all the way down: it begins with data together with tentative means for winnowing those data to produce a defensible view of things. Along the way we can refine both the data and our means refining them. On this account truth is a long-term goal, not a piece-by-piece accomplishment gradually extended. Even the logical features of truth and ideal coherence are revisable regulative ideals, not demands set in stone from the outset. Modesty is the watchword here—but it is a hopeful modesty.

REFERENCES Albritton, C. C., The Abyss of Time (San Francisco: Freeman, Cooper and Company, 1980). Brown, B., “A Short History of Preservationism,” in J. Woods and D. Gabbay (eds.), Handbook of the History of Logic, Vol 8: The ManyValued and Non-Monotonic Turn in Logic (Amsterdam: Elsevier, 2007). ———, “LP, FDE and Ambiguity,” in H. Arabnia, ed., IC-AI 2001 Volume II- Proceedings of the 2001 meetings of the International Conference on Artificial Intelligence (CSREA publications, 2001). ———, “Yes, Virginia, There Really are Paraconsistent Logics,” Journal of Philosophical Logic, vol. 28 (1999), pp. 489-500. Brown, B. and Schotch, P. K. “Logic and Aggregation,” Journal of Philosophical Logic, vol. 28 (1999), pp. 265-287. Chisholm, R., “The Problem of the Criterion,” The Foundations of Knowing, (Sussex: Harvester Press, 1982), pp. 61-75.

84

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER

Feigl and Sellars (eds.) Readings in Philosophical Analysis (New York, 1949). Neurath, O., “Protocollsätze,” Erkenntnis, vol. 3 (1932-33), translated by F. Schick as “Protocol Sentences,” in A.J. Ayer (ed.), Logical Positivism (Glencoe, Ill., 1959), pp. 201-14. Payette, G. and P. K. Schotch, “Preserve What?” in Schotch, Brown and Jennings (eds), On Preserving, forthcoming from University of Toronto Press, 2008. Priest, G., Routley, R., and Norman, J. (eds.) Paraconsistent Logic: Essays on the Inconsistent (Munich: Philosophia Verlag, 1989). Quine, W. V., Ontological Relativity and Other Essays (New York: Columbia University Press, 1969). Rescher, N., The Coherence Theory of Truth (London: Oxford University Press, 1973). ———, Methodological Pragmatism (New York: New York University Press, 1977). ———, Cognitive Systematization (Oxford: Basil Blackwell, 1979). Rescher, N. and Brandom, R., The Logic of Inconsistency (Oxford: Basil Blackwell, 1980). Schotch, P.K., Jennings R. E., (1989) “On Detonating,” in G. Priest, Routley, R., and Norman, J. (eds.) Paraconsistent Logic: Essays on the Inconsistent (Munich: Philosophia Verlag, 1989), pp. 306-327. ———, “Modal Logic and the Theory of Modal Aggregation,” Philosophia, vol. 9 (1981), pp. 265-278. ———, “Inference and Necessity,” Journal of Philosophical Logic, vol. 9, (1980), pp. 327-340.

85

RESCHER STUDIES

Schotch, Brown and Jennings (eds), On Preserving, forthcoming from University of Toronto Press, 2008. Sellars, W., Science, Perception and Reality (London: Routledge and Kegan Paul, 1963). ———, “Empiricism and the Philosophy of Mind,” in Science, Perception and Reality (London: Routledge and Kegan Paul, 1963), pp. 127196). Tarski, A. “The Semantic Conception of Truth,” in Feigl and Sellars (eds.) Readings in Philosophical Analysis (New York, 1949), pp. 52-84. NOTES 1

I won’t explore the interesting and difficult question of whether there are good coherence-based reasons for taking such a position. After all, our experience of coherence-based corrections of previously justified conclusions raises questions about what we should expect from an ideal investigation: perhaps it can be argued that such an investigation should lead to coherence support for sentences that accord with the facts in some correspondence sense. Such an argument would reinforce the connection Rescher perceives between a coherence criterion for truth and his (modest) correspondence account of the definition of truth.

2

One exceptional circumstance where this might accepted is a case in which human extinction is imminent—for example, if a massive asteroid or comet, far bigger than the one that struck at the end of the Cretaceous, were about to strike the earth. But of course this is a matter of inquiry reaching its temporal end rather than its goal.

3

These definitions are drawn from Payette, G. and P.K. Schotch, `Preserve What?', in Schotch, Brown and Jennings (eds), On Preserving.

4

And is denied by, as well.

5

For convenience we can declare that such sentences also follow from the null set.

6

See Payette and Schotch, 2008, where these ideas are defined in terms of productivity and its dual, co-productivity.

7

In particular, by our definitions, both Priest’s LP and first degree entailment lack denial.

8

Tarski, ‘The Semantic Conception of Truth,’ 52-84 in Feigl and Sellars (eds.) Readings in Philosophical Analysis (New York, 1949).

86

COHERENTISM AND COHERENCE TRUTH IN THE PHILOSOPHY OF NICHOLAS RESCHER NOTES 9

This is a matter of hypergraph-colourings—any hypergraph with chromatic index ξ is a complete aggregative template for divisions of our premises amongst ξ-1 cells. A proof of this result appears in the appendix of Brown and Schotch, 1999.

10

See W. Sellars “Empiricism and the Philosophy of Mind” for a subtle escape from a very similar regress, on pages 168-9.

87

Philosophical Anthropology James W. Felt

I

t has been said that Leibniz was the last man who knew everything. But the writings of Nicholas Rescher, who professes himself a follower of Leibniz, may cast doubt upon that assertion. At least for a reader used to identifying Rescher’s work principally with philosophy of science it is refreshing to read his insightful and humane observations on the whole range of human experience as he analyzes it in Human Interests: Reflections on Philosophical Anthropology, and in Luck: The Brilliant Randomness of Everyday Life, a book that is a development of one of the chapters of the former. I shall highlight some of Rescher’s interpretations of human nature as found in these books, then reflect on the relation of his philosophical commitments to those interpretations. THE BOOKS In Human Interests Rescher describes philosophical anthropology as a philosophy of the good life, a normative discipline concerning how we ought to live. Such normative judgments essentially concern human values, and deal not with the good of any particular person but with what is the case for people universally. They are validated not a priori but by the agreement of others, particularly those who find real satisfaction in living according to them. Rescher then asks what it is to be a person and identifies seven characteristics which, taken together, define personhood: intelligence, affectivity, agency (the capacity for free, goal-directed action), rationality, self-understanding, self-esteem, and mutual recognizance (fellow-feeling functioning in a context of community). The core “self” of a person is not a static given but a bundle of capacities whose activities are largely observable and thus experientially accessible. This, in Rescher’s view, is a shift from a “substance” to a “process” view of the person. Rescher goes on to consider our human impulse to transcendence—the search for something abiding in this world of change (“The Quest of Sisyphus”). Contemporary society tends to adapt the Stoic program by looking for finality in the findings of science, yet we cannot be sure even of the supposedly permanent laws of nature, only of our own beliefs about them.

RESCHER STUDIES

Given the chanciness of life and our weakness of apprehension, is there any room for optimism in life? Here and elsewhere Rescher argues forcibly that optimism is the best policy, and in doing so he is not simply Dr. Pangloss. He shows that “attitudinal optimism” (of which one exponent was William James) is not only rationally well founded, but much more likely than is pessimism to produce what it envisions. Attitudinal optimism recommends not just patience but confidence; it is an evaluative rather than a predictive posture, and it need not be validated on evidential so much as on pragmatic grounds.1 Through reason the human being has a unique dual citizenship in the realms of facts and of values. This is, in fact, cause for astonishment. As Rescher says: “It is remarkable that nature has managed to evolve a creature who aspires to more than nature can offer, who never totally feels at home in its province, but lives, to some extent, as an alien in a foreign land” (1990, 139). Entertained ideals take us into the realm of imagination beyond the real or even beyond the realizable, but to give up striving after ideals is to give up being fully human. Rationality is a fundamental human value and a source of happiness, for it enables us to pursue appropriate ends. Although reason does not guarantee us affective happiness, it is the best guide we have for reflective happiness and for the goal-directed management of our lives. To read the book Luck may remind you of seeing a dozen circus performers emerge from a Volkswagen bug, for you would not have imagined that the apparently simple, everyday idea of luck is pregnant with the wealth of implications that Rescher unpacks from it. Its subtitle, “The Brilliant Randomness of Everyday Life,” hints that Rescher will in the end argue that we humans, precisely as humans, cannot really give up, nor want to give up, the chanciness that luck affords to our lives. Without it we would live in a world in which our human freedom would be impossible; we would be living in a pre-programmed world, a world without surprises (who relishes reading yesterday’s newspaper?) and we would lose the freedom essential for our human agency. SKETCH OF RESCHER’S PHILOSOPHIC PRINCIPLES In a remarkable book, The Unity of Philosophical Experience, Etienne Gilson wrote: “Philosophers are free to lay down their own sets of principles, but once this is done, they no longer think as they wish—they think as they can” (243). The pressure of these principles, like the massive weight

90

PHILOSOPHICAL ANTHROPOLOGY

of a glacier, forces certain conclusions and precludes others. What then are the philosophic principles laid down by Rescher? He answers this explicitly in a recent book, in the preface of which he writes: “This book sets out the theoretical underpinnings of the idealistic approach to metaphysical realism that its author has developed over several decades” (Rescher 2005, ix). Let us consider—and tentatively evaluate—what seems to be the center and source of those underpinnings, then observe it at work in his understanding of human experience. 1. Rescher’s Epistemological Representationalism, and a First Difficulty In chapter 3 Rescher asks: “1) What entitles us to claim that there is such a thing as mind-independent reality? “2) What can we justifiably say regarding what that reality is like” (Rescher, 2005, 35).” It is enough for our present purpose to focus mainly on the first of these questions. Rescher holds that all perception, even sense perception, is inevitably subjective, personal, always owned by some particular individual. Yet it is supposed to inform us of an objective, extra-mental, bodily world. And, as he says, “this poses the big problem of how we are to get from here to there” (2005, 31). John Locke too had recognized this same problem. “It is evident,” he writes, “[that] the mind knows not things immediately, but only by the intervention of the ideas it has of them. . . . How shall the mind, when it perceives nothing but its own ideas, know that they agree with things themselves?” (Locke 228). It is clear that Rescher, standing in the respectable company not only of Locke but of Galileo, Newton, Hobbes, Descartes, Hume, and any number of subsequent philosophers, holds this representational epistemology. Any such view naturally spawns the same difficulty that troubled Locke, for it assumes that what we are sensibly given is not bodies in a world but rather our own internal sensations or sense-data, and these are presumed somehow to represent that external world. Consequently, Rescher points out, “there is . . . an unavoidable evidential gap between statements regarding the [inner] experience of people, and those that concern the world’s objective arrangements” (2005, 32). He adds, with wistful candor, “This situation is not particularly good news.” No indeed. Now, given his representational perspective Rescher must hold that getting from here to there can hardly be a matter of evidence. Yet we do need to get there somehow, and he thinks the best we can do is get

91

RESCHER STUDIES

there by postulating that an extra-mental world exists. After all, our very concept of reality is committed to a realism that rejects simply identifying reality with our mental conceptions of it. So if we are not given an external world, we must postulate it if experience is to make any sense. The justification of this postulate lies in its utility. In fact, Rescher thinks that this need to postulate an external world forms a kind of transcendental argument for its own justification (2005, 37). He later lists the following particular reasons why we need to make this postulation: 1) to preserve the distinction between truth and falsity regarding factual matters (there must be an external criterion of conformity); 2) to save the distinction between appearance and reality (Rescher is an idealist but no Berkeley); 3) to ground intersubjective communication, with its need for common things to discuss; 4) to provide a basis for shared projects of inquiry, such as science; 5) to account for the fallibility of our knowledge, inasmuch as an objective reality can exceed anything we can know or say about it; and 6) to provide for causal accounts of inquiry, for explaining why things happen (Rescher, 2005, 82-83). These are impressive reasons and Rescher believes they are retrojustified by the very existence of common human enterprises, including science. In the end, however, one must ask whether postulation of an external world, even if it is the best we can do, escapes Locke’s worry about its soundness. If it does not, will not this doubt infect every interpretation that Rescher gives to human behavior? Is this the best sort of realism available to us? I think it is not, but here can only point toward where a better realism seems to lie. It could be called “relational realism.”2 It rejects the assumption—and it is an assumption—that our sense perception has as its object our own sense data rather than objects in an extra-mental world. But our experience presents itself prima facie as encountering a world, not our own sense impressions of a world. Such an encounter involves a world acting on our senses. Aristotle pointed out that the activity of a cause lies in its effect.3 Thus the world as it acts upon us in our sensing is present to us in its acting, and that acting both elicits and gives the specification for the ensuing activity of our senses. Of course we do not find ourselves confronted with a world-in-itself, but a world-as-it-relates-itself-to-us in our sensing— a relational world. We do not have to think that we encounter our sense perceptions, any more than we have to think that we are looking at the lenses of our eyeglasses when we see objects clearly by means of the lenses. If our sense impressions were in fact the objects of our perceiving,

92

PHILOSOPHICAL ANTHROPOLOGY

they would indeed be of little or no real evidential value pointing to extramental reality. But if the world itself, albeit the relational world of perception, is what we encounter in perception, then we are already “there” in the very act of perceiving, so the world itself is already evident to us. The “there” already lies in the “here.” 2. A Second Difficulty Suppose, however, that we were provisionally to accept Rescher’s argument that there is no real evidence for the existence of an external world, so that one is forced to postulate it. We could still reasonably ask whether Rescher is consistent with his argument when he enunciates this view. He asserts that we need to postulate the existence of a world with other people in it in order “to serve as a basis for intersubjective communication” (2005, 82-83). But who are these other people with whom Rescher feels the need to communicate, yet has to postulate rather than simply acknowledge their evident existence? Does he not, by asserting the need for this postulation, thereby contradict his theoretical denial that it is evident that they exist?4 3. Application to Rescher’s Concept of the Person Arguably the richest chapter in Human Interests is the second, “What Is a Person?” It seems therefore the most rewarding and feasible locus for noting the interplay between Rescher’s humanistic insights and his adopted metaphysical perspective. In doing so I consider his conception of the person from the following particular, though somewhat overlapping, perspectives: a) Epistemic and Ontological Realism; b) Free Will; c) Substance and Process; d) Self as Subject. 3a. Epistemic and Ontological Realism It will be recalled that Rescher sets down seven conditions “severally necessary and jointly sufficient” for qualifying a being as a “person.” Here is the seventh of those: “7. Mutual recognition. One must be disposed to acknowledge other duly qualified agents as persons and be prepared to value them as such. With persons there must be not only feeling, but fellow-feeling; persons must function in a context of community” (Rescher, 1990, 7).

93

RESCHER STUDIES

This is a splendid insight into personhood, but as I remarked before, it seems inconsistent with Rescher’s theoretical need to postulate these other persons with whom he is aware of being in community. He postulates them because he thinks he has no real evidence that other persons—or the whole physical world for that matter—actually exist. So he supposes that there is a world with other persons in it, and this supposition turns out to work very well. Or does it? Can it conceivably render intelligible our human experience, for instance, of love? Can lovers seriously persuade themselves that they must constantly make a supposition that their beloved is in their arms although they have no reliable evidence that that is the case. Yet Rescher writes : The presuppositional conception of a mind-independent reality . . . plays a central and indispensable role in our thought about matters [sic; of?] cognition, in the context of the contrast between “the real” and its “merely phenomenal” appearances. Moreover, it is seen as the target of [sic; or?] telos of the truth-estimation process at issue in inquiry, providing for a common focus in communication and communal inquiry. (The “real world” thus constitutes the object of our cognitive endeavors in both senses of this term—the objective at which they are directed and the purpose for which they are exerted.) And, further, reality is also to be seen as the ontological source of cognitive endeavors, affording the existential matrix in which we live and move and have our being—and whose impact upon us is the prime mover for our cognitive efforts (Rescher, 2005, 83-84; his emphasis).

Note that here, as in many passages elsewhere, Rescher takes causality seriously. He writes that the (causal) “impact” of extramental reality is the “ontological source” of our cognitive processes, and this is, after all, essential to interpersonal communication: that persons act upon one another. But if I am to interpret my own activity with others according to Rescher’s theory rather than his insight into human experience, I can only postulate or suppose the existence of these other persons on whom I seem to act and whose action I seem to feel on me. I conduct myself as if I am acting on an extramental object that I choose to treat as if it is a person, and the impact I experientially feel in return must be regarded as if originated by that other postulated person? Is this a credible interpretation of actual human experience? If not, why hold it? We have already seen that Rescher recognizes the puzzle, as did Locke, of how we are to get from our own personal, subjective experiences to an

94

PHILOSOPHICAL ANTHROPOLOGY

objective, extramental world. This is echoed again just above where Rescher contrasts “the real” with its “merely phenomenal” appearances. In short, when we trace the roots of Rescher’s need for postulating an extramental world we find ourselves once again facing the dilemma inevitably posed by representational epistemology. Yet when Rescher treats of the distinction between appearance and reality he defines appearance as “how things are thought to be,” and reality as “how things actually are” (1990, 4). This enables him later on to write: “Reality is not something different from appearance but is just exactly what appearance is whenever it happens to be correct” (1990, 7) In this rather curious move Rescher treats appearance and reality as if they were simply two sets of attributes rather than a perceptual activity on the part of the perceiver contrasted with the existence of an extramental object. The ontological dimension has been omitted altogether. And of course whether the appearance is “correct”could only be determined by establishing the correspondence of the two sets of attributes, so we are right back to Locke’s perplexity about how to get from our perceiver’s “here” to the perceived’s “there.” (This seems rather like noting that a stopped clock is absolutely accurate twice a day, the only problem being to find out when.) Many of these difficulties would vanish if Rescher were to accept as evident that in his perception he is being acted upon by extramental causes, whatever they may prove to be. That he does not do so is somewhat surprising since he is well acquainted with the process philosophy of A. N. Whitehead, one of whose greatest philosophic contributions was to stress the authenticity of the feeling of being acted upon by other things (his “perception in the mode of causal efficacy”). When one couples this aspect of immediate experience with Aristotle’s astute observation, mentioned above, that the activity of an effective cause lies precisely in its effect, then one is in a position to give an ontological justification for holding that the immediate object of perceiving is not one’s own sense impressions or sense data, but rather the extramental object itself as it is affecting one. Thus representational epistemology with all its attendant problems is shown to be bankrupt and useless. There is no need to postulate an extramental reality or extramental persons because they are immediately given—hence evident—in perception. There is no need for a postulational realism when an evidential realism, particularly that of human interaction, lies ready at hand.

95

RESCHER STUDIES

3b. Free Will The third of Rescher’s seven conditions for personhood reads as follows: “3. Agency. One must be capable of goal-oriented action as a free agent who is able not only to pursue goals, but to initiate them, to set them for oneself. As Philosophers have insisted since Aristotle’s day persons are not merely agents, but autonomous agents whose goals proceed from within their own thought processes and who are accordingly responsible for their acts” (1990, 6-7; Rescher’s emphasis). This is very well said, yet Rescher appears later to give it such a narrow interpretation as to subvert the concept of human freedom after all. Here is what I mean. In “The Meaning of Life,” which is the fourteenth chapter of Human Interests, Rescher includes a subsection entitled “Free Will” (1990, 158 - 160) in which he defends his own understanding of human freedom. The freedom that he is interested in is, first of all, freedom of action, in the sense of “being in control of one’s doings” (1990, 159). This is also the position of John Locke who wrote: “So far as a man has power to think or not to think, to move or not to move, according to the preference or direction of his own mind, so far is a man free” (Locke, 228). There are two main points included within the above descriptions: (1) that the concern is with an action that (2) is taken as the result of a different act, the act of one’s own inner volition. Using the word “liberty” to express freedom, Locke goes on to say: “Liberty is not an idea belonging to volition, or preferring; but to the person having the power of doing, or forbearing to do, according as the mind shall choose or direct” (Locke, 317). For Locke (and I think for Rescher) this doing or acting can include taking up or laying aside different thoughts in the mind, but it cannot reasonably include the person’s very act of choosing or deciding to do or not do something, since in that case the choosing and the doing are the selfsame act. Rescher, I think, is content with that point but concentrates on the act’s being the outcome of one’s own inner volition or act of determination, not taken in virtue of any motivations somehow imposed from without, by any other agents. “We thus arrive at the doctrine of compatibilism,” he writes, “—the argument that freedom of the will is compatible with a softened determinism in the order of natural processes . . .” (1990, 159). This appears to coincide with the “soft determinism” that William James called “a quagmire of evasion” (James, 149). For in the first place, it is hardly a philosophical problem whether one can carry out what one chooses to do. A

96

PHILOSOPHICAL ANTHROPOLOGY

locksmith might be needed rather than a metaphysician. The true locus of the problem of personal freedom lies within the psyche or mentality of the acting agent. If all my acts of deciding or choosing are necessitated by motives, proclivities, and my previous decisions, then I literally can’t help here and now deciding what I decide. Rescher clearly implies as much when he writes that a free agent , “‘could have done otherwise’, all right, but only if the agent’s choices and decisions had been different—which itself could have happened only if the agent’s values, goals, and decisions had been different” (1990, 159; my emphasis). But an act of choosing or deciding is free only if, given exactly the same motives and background, the agent could have chosen differently than he did. Rescher is rightly intent to refute the charge that the laws of nature render us “mere automata in that our will is not free, and all our actions merely the result of the play of natural forces beyond our control” (1990, 158). But am I not equally an automaton—albeit a self-created one—if all my choices are themselves the inescapable outcome of my own motives and previous decisions? It does not save freedom to say that my being constrained now is acceptable because it is self-imposed unless my own prior act of deciding is itself free. Yet Rescher, with Locke and countless others, seems to locate freedom only in the carrying out of one’s decisions, rather than in the deciding itself. Freedom in the very act of deciding is the freedom most worth having, and I do not see that it fares well, if at all, in Rescher’s view.5 3c. Substance and Process. In the chapter “What Is a Person?”, under the subheading “Personal Identity”, Rescher asks “What is it that individuates a particular person as such?”6 He goes on to reply: The idea that “the self” is a thing (substance), and that whatever occurs in relation to “my mind” and “my thoughts” is accordingly a matter of the activity of a homunculus-like agency of a certain sort (a “mind” substance) is no more than a rather blatant sort of fiction—a somewhat desperate effort to apply the substance-attribute paradigm to a range of phenomena that it just does not fit. . . . It is, after all, rather repugnant to conceptualize persons as things (substances)—ourselves above all.7

97

RESCHER STUDIES

Rescher thus identifies “substance” with a “thing”—pretty clearly a dead, inert something-or-other—and rightfully rejects that as descriptive of a person. But he has here bought into the common though demonstrably erroneous idea that that is the only meaning of “substance,” particularly when applied to Aristotle’s notion. This error is largely traceable to A. N. Whitehead who, understandably but mistakenly, attributed to Aristotle’s ousía the notion attached to “substance” by Descartes (“that which needs nothing else in order to exist”), Locke (“an I-know-not-what” that stolidly and changelessly underlies its acquisitions or loss of qualities), and Hume (lacking all possibility of being internally related to others). Yet Aristotle in fact conceived ousía (badly translated as “substance” by Boethius) as a dynamic agent engaged over time in a process of activity that includes its own self-development.8 The living, acting, developing human person is in fact the very paradigm of a substance and nothing at all like Rescher’s “thing.” Even so, we can ask whether Rescher’s own alternative concept of the self is viable. To determine just what that concept it is, we can recur to the seven conditions he laid down in Chapter Two of Human Interests, 6-7. He writes: “These conditions are severally necessary and jointly sufficient conditions for qualifying as being ‘a person’ in the standard sense of the term” (7). And he points out that the first four conditions relate to what a person can do (understand, evaluate, initiate acts, guide acts by reason), and the last three to what a person does do (understands and esteems self, and is disposed to value other persons as well). This is an excellent description of what a person can and does do, but it doesn’t tell us what a person is—a person’s ontological structure, if you will. One does find such an expression in the “Personal Identity” subdivision of the second chapter. There Rescher writes: Once we conceptualize the core “self” of a person as a bundle of actual and potential processes—of action and capacities, tendencies, and dispositions to action both physical and psychical—then we have a concept of personhood that renders the self or ego experientially accessible, seeing that experiencing itself simply consists of such processes. On a process-oriented approach, the self or ego (the constituting core of a person as such, that is, as the particular person he or she is) is simply a mega-process—a structured system of processes (1990, 17-18; Rescher’s emphasis).

Now it must be recalled that Rescher also says that a person understands, evaluates, loves, and freely initiates goal-directed activity for which

98

PHILOSOPHICAL ANTHROPOLOGY

the person is morally responsible. In all this, as a being in space and time, the person must maintain an identity over time. Whenever you say, “I did it,” you implicitly claim that the you that said it is the very you that did it. But if we replace “person” in Rescher’s sentences quoted just above with his description of the self as “a bundle of actual and potential processes” or “a structured system of processes,” how plausible is the result? It is worth trying aloud. If however we replace “person” with “a living, bodily being that acts over time through understanding and will”—an authentic “substance” view—are we not better off? I am afraid that the person so accurately identified by Rescher as to what it can and does do, loses its personhood when conceived, in Rescher’s theory, as a bundle of processes that are supposed to be aware of themselves, to esteem themselves, to originate morally responsible activities, and to retain their self-identity over time. 3d. Self as Subject In his seven conditions essential to personhood, Rescher includes selfunderstanding, namely conceiving of oneself as an intelligent free agent, operating in the dimension of belief, action, and evaluation; and also selfesteem, valuing oneself on that very basis. This is very well said, but a further recognition of its implication would have been constructive. For in that same subsection on Personal Identity, Rescher quotes the famous lines of David Hume: “For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never can catch myself at any time without a perception, and never can observe any thing but the perception.”9 To this Rescher remarks, “Here Hume is perfectly right.” Well, yes and no. Rescher goes on to point out, quite correctly, that one can never observe oneself as one would observe an object. But Hume’s intention is broader than that. He is desperately in search of his self and says he never finds it. From that perspective those lines of Hume constitute one of the most flagrant blunders in western philosophy. He writes: “I stumble,” “I never catch myself,” “I never can observe.” But who, please, is doing the searching if not the very self that is sought? While mistakenly searching for himself as an object Hume is all the while implicitly grasping himself as a subject doing the searching. Rescher too must be aware of himself as a subject when he repeatedly refers to his self-awareness and self-esteem. Naturally Hume could not find himself as an object, but is

99

RESCHER STUDIES

Rescher content, with Hume, to leave out recognizing oneself as a subject of experiencing? In considering self as subject, we have implicitly moved beyond Rescher’s first question, “What entitles us to claim that there is any mindindependent reality?” to his second question, “What can we justifiably say regarding what that reality is like?” (2005, 35). In his stress on the fundamentality of interpersonal relations Rescher implicitly accepts—or at least postulates—the existence of other people like himself. He postulates, with considerable insight and eloquence, that he lives in a world of other human beings with whom, by his very human nature, he needs and desires to interact. This postulation rightly overleaps the problem that some pragmatists would presumably have—as would also John Locke—in getting not just to extramental reality but to a person if they think that what their senses deliver to them are only their own sense data, a display of colors, sounds, smells, and so forth—from which they must construct a person. But Lewis Carroll was right about Humpty Dumpty: it can’t be done. Yet though Rescher thinks he must postulate other persons, at least it is persons he postulates. Does he have a right to do so? I think so, because he writes as if it is evident to him (whether in theory he thinks so or not) that he is dealing with other persons. Indeed, no experience seems more concrete to him than that of interpersonal relationships; that is what his seventh condition of personhood, that of Mutual Recognizance, is all about.10 CONCLUSION. Rescher’s philosophic anthropology is rich in insights into an extraordinarily wide range of human experience. He provides good reasons for a refreshing optimism about the possibilities of life. But his thought seems to me in constant tension, bordering sometimes on contradiction, between his expression of insights about human existence and the philosophic principles in terms of which he analyzes them. His phenomenology of the human situation is astute, but his theoretical interpretation of that experience seems unconvincing. An historical analogy comes to mind. Descartes in his sixth meditation thought he had proved, by appealing to God’s veracity, that an extramental, quantified world really exists. He then proceeded serenely to assume that these quantified objects, of whose existence he was now assured, were capable, despite their quantified nature, of exciting within his mind the

100

PHILOSOPHICAL ANTHROPOLOGY

purely mental ideas by which he perceived them. But even given the existence of quantified things, Descartes was no closer to accounting for his perceiving them (as George Berkeley was quick to point out) than he was before he thought he could, through God, prove their existence. Rescher likewise supposes that he has no real evidence for the extramental existence of things, including people, but instead of appealing to God’s veracity he considers pragmatic considerations to be sufficient grounds for supposing their existence. He then proceeds to give an eloquent account of his interactions with these postulated existents. Both philosophers, after accepting in their own ways the existence of an external world, are untroubled in proceeding to describe genuine dimensions of human experience that are nevertheless hard to reconcile with their theoretical principles.

REFERENCES Felt, James W., “Whitehead’s Misconception of ‘Substance’ in Aristotle,” Process Studies (1985), pp. 224 - 236. ————, Coming To Be (Albany: State University of New York Press, 2001). ————, Human Knowing: A Prelude to Metaphysics (Notre Dame, Indiana: University of Notre Dame Press, 2005). Gilson, Etienne, The Unity of Philosophical Experience (San Francisco: Ignatius Press, 1999), Hocking, William Ernest, “Whitehead as I Knew Him,” in George L. Kline (ed.), Alfred North Whitehead: Essays on His Philosophy (Englewood Cliffs, New Jersey: Prentice-Hall, Inc, 1963). Hume, David, A Treatise of Human Nature, ed. L. A. Selby-Bigge (Oxford: Clarendon Press, 1968).

101

RESCHER STUDIES

James, William, “The Dilemma of Determinism,” in The Will To Believe and Other Essays in Popular Philosophy (New York: Dover Publications, 1956). Locke, John, An Essay Concerning Human Understanding, ed. Alexander Campbell Fraser, vol. 2. (New York: Dover Press, 1959). Moore, George Edward, “A Defence of Common Sense,” in J. H. Muirhead (ed.), Contemporary British Philosophy (London, UK. Allen & Unwin, 1925). Rescher, Nicholas, Human Interests: Reflections on Philosophical Anthropology (Stanford, California: Stanford University Press, 1990). ————, Luck: The Brilliant Randomness of Everyday Life (New York: Farrar Straus Giroux, 1995). ————, Reason and Reality: Realism and Idealism in Pragmatic Perspective (Lanham: Rowman and Littlefield Publishers, 2005). ————, “Optimalism and the Rationality of the Real: On the Prospects of Axiological Explanation,” The Review of Metaphysics, vol. 59.3 (March 2006), pp. 503-16. ————, “Causal Necessitation and Free Will,” Process Studies, vol. 35.2 (2007), pp. 195 - 206. NOTES 1

On this topic the reader may wish to read Rescher’s more recent essay (2006) “Optimalism.”

2

I develop this view in Human Knowing: A Prelude to Metaphysics (Notre Dame: University of Notre Dame Press, 2005).

3

Aristotle, Physics III:3, 202a11-22, and VII:2, 243a3-4.

4

A colleague has pointed out to me that G. E. Moore made a similar point in his essay, “A Defence of Common Sense,” especially pp. 203-05.

5

It is worth noticing that in a much later essay (2006: “Causal Necessitation and Free Will”) Rescher provides a different and strong new argument that free acts of choice are compatible with natural physical causality, but does it without retreating from the argument examined above.

102

PHILOSOPHICAL ANTHROPOLOGY

NOTES 6

I am afraid that in his development of this topic Rescher muddles the notion of individuality with that of individuation. I see now that I made the same error in my book Coming To Be, 101 and 121. One must first be individuated, be a separate and autonomous person, before one can develop a particular individuality. For us humans, the root of the former is bodily, quantitative, while the latter is qualitative.

7

1990, 16. Rescher goes on to add: “People instinctively resist being described in thing-classificatory terms” and he cites Sartre as suggesting that a person who has stolen will resist saying “I am a thief.” But has Rescher here misconstrued Sartre’s point? In No Exit, at least, Sartre makes the point that the man who did cowardly acts is in fact a coward, even though he resists saying or even thinking that he is.

8

For detailed substantiation of this claim, see my essay, Felt,1985, 224-236.

9

Hume, 1968, Book 1, Part 4, Section 6, “Of Personal Identity.”

10

The philosophical problem of “intersubjectivity”—of experiencing other persons— is notorious in modern philosophy, especially that of Edmund Husserl who devoted the last and longest of his five Cartesian Meditations to wrestling (not very successfully, as it seems to me) with just that problem. A more refreshing view was expressed by A. N. Whitehead. When asked how we know minds other than our own, he replied “Hang it all! Here we are. We don’t go behind that: we begin with it” (Hocking, 8).

103

Pluralism and Consensus Lenn E. Goodman

A

few years ago, taking a visiting philosopher to lunch, I found myself treated to quite a spate of views that I thought sounded rather cockeyed. Not hearing much in the way of argument, I asked why one should think that way. The answer I got stunned me a bit: ‘If we don’t, we’ll all just be killing each other.’ That, I think, was meant to sound like pragmatism. To me it rang hollow, a sort of updated ad baculum. Does failure to agree, broadly, set us inevitably at each other’s throats? Is conformity a better recipe for peace than recognition of our differences? Can one respect another’s outlook without endorsing it—or reject it without tears? Is agreement in fear of conflict cheap, or bought too dear? Is it condescension, or doublethink? Nicholas Rescher’s Pluralism: Against the Demand for Consensus takes aim at one of today’s fatter sacred cows, the idea of a moral, political, even epistemological imperative for intellectual convergence. “Recognizing that others see some matter differently from ourselves,” Rescher writes, “need not daunt us” (121). “There is nothing rationally mandatory about the quest for consensus (126). What a breath of fresh air in today’s academic hothouse, where rare, self-pollinated specimens exude well rehearsed praises of diversity but can barely countenance an alien thought. Many vocally abhor stereotypes but crowd into a procrustean bed of their own making, mouthing the passwords for admission—and then caricature rival views, or romanticize what they find pleasantly primitive or safely exotic. Squeezing into their partisan mold, they make tintypes of their adversaries, mocking those they’ve already stamped out of their parents or their former selves. So learning becomes not discovery or growth but mimicry—storage and retrieval, cookbook chemistry and McGuffey physics and philosophy. Teaching collapses into a facile relativism, secure in the promise that little on offer will jar against biases duly accredited. Rescher’s chief targets in Pluralism are Jürgen Habermas, Hans Georg Gadamer, and Karl-Otto Apel. But he does not ignore Kuhn or Rorty, Rawls, Mill, Peirce, and James. His case rests on his commitment to what he calls objective rationality. But reason today often gets a bad rap, and that’s not just Hume’s fault. A lot of overweening claims have been made

RESCHER STUDIES

in its name, and false pretenders have so often usurped a mantle that doesn’t fit too well that reason itself can look or sound somewhat pretentious and halloweenish. If its fruits were just the Terror, then Burke’s or Oakeshott’s tradition, or Rorty’s parochial (“local” or “retail”) values begin to look downright cozy, and the idea of a truth “out there” to be grasped and held onto starts to look greedy, a source, or reflex, of an urge to dominate. There’s a variety of responses to the seeming scandal of intellectual or practical diversity. Rescher lists skepticism, relativism, and syncretism among them, each defined in severe but hardly fanciful terms. For no stance seems quite overstated enough in today’s agora to lack its familiar enthusiasts. One need hardly dummy up strawmen. Well documented avatars preen in the literature. Skepticism, here traced back to Xenophanes and Aenesidemus, deems no truth claims tenable. Rescher calls it nihilism and tackles it with an argument garnered and generalized from H. H. Price’s take on testimony: There’s at least as much risk in skirting as in making a commitment— more, in fact. For sitting on the sidelines assures vacuity; but commitment risks only error, in principle open to correction. Rescher’s watchword: Nothing ventured, nothing gained. Skeptics might answer, as Sextus does, that their quarrel is not with ordinary life but with claims to ultimate truth. But Rescher sees forced choices: “We cannot rationally maintain a posture of indifference. . . . One must ‘choose sides” (120-21). For the ultimate questions that skeptics try to shake off are fraught with value issues vital to the conduct of life: “It makes a crucial difference to everybody which position one adopts. . . . We cannot view doctrinal disagreement in the light of a ‘mere divergence of opinion.’ It is a serious conflict—as one would expect when something as important as one’s values comes into play” (122). To opt out is itself to choose. Despite their blanket condemnations, skeptics, in fact, are usually pretty specific in their allergies and agendas: Metaphysics, religion, and (hand picked) value notions top the roster of anxieties and neuralgias from which equipollence promises relief. Philip Hallie sees skepticism as the royal road to tolerance—despite the doctrinaire skeptic’s disavowal of any value that would set tolerance ahead of persecution or oppression, or any other seeming end. Most skeptics reserve a special place in epistemic Hell for superstition. So Sextus singles out astrology for bitter invective and opprobrium, even as his tropes sweep away astronomy at large. Almanacs get a

106

PLURALISM AND CONSENSUS

pass from him, but logic is out. Somehow skeptics hope to hang on to a critical edge, even as they reject reason and the senses wholesale and forswear all objective criteria of truth. Evidently, we humans find it hard to give up theorizing just because some theory (even our own!) pronounces theory impossible. “Homo sapiens,” Rescher writes, “has evolved within nature to fill the ecological niche of an intelligent being” (87). We don’t always fill it very creditably. Syncretism is the quest for a conjunction of all views. Its poster child here is Paul Feyerabend, who shifts the skeptic’s motto “no more” into a new register and comes out singing “Anything goes!” (90, 100). Assiduously uncritical, Rescher’s syncretists hope, delusively, to rise above the fray of outlooks and opinions. But the views they lump together are discordant. Their fusion is not a symphony but “a mess” (102). For Rescher’s syncretists, “The incompatibility of claims does not mean that one or another of them must be denied. . . . Truth is somehow large and complex enough to accommodate contradictions” (91). That appraisal, typifying the extremes, may be somewhat hasty. There are, after all, ways to unite discordant views without simply and mechanically conjoining them at the expense of the excluded middle—if one can see past the limited horizons of rival outlooks. Language and experience canonize all sorts of conflicting views and values, rarely needing to reconcile them since they come into play in disparate contexts. But a good philosopher does not simply play off one of these against another but charitably seeks the motives of each, in hopes of finding a larger context in which apparent disparities are resolved, even those among the sharply antagonistic intellectual forces that have barricaded themselves behind one fragmentary insight or another. Success in devising that kind of creative vision is the hallmark of genuinely great philosophy: Plato overcomes the opposition of Heraclitean flux to Parmenidean invariance by finding the place of change in this world, and seeing a place for ultimate unity beyond the world of becoming, yet shining through it. Aristotle can wed his naturalism to Plato’s idealism by discerning steady, indeed telic patterns in the natural flux. Spinoza discovers liberty in a world of determinacy, when he observes that identity does not mean changelessness, and urges that we too are, in some measure, among the many causes whose agency has effect. Kant can reconcile Hume with Leibniz by sifting the a priori from the innate, separating the warrant for our concepts from the story of their genesis in the mind. Syntheses like these complement the analytic moment in philosophy and build a syncre-

107

RESCHER STUDIES

tism rather more thoughtful than the one that Rescher rightly and firmly excludes when he calls on us to be “open-minded, not empty-headed” (119). The relativism Rescher rejects he calls indifferentist. Like syncretism, he finds, it resembles skepticism in (at least overtly) refusing to discriminate among rival views: It finds no rational grounds for choice, but only taste, or tradition. Rescher fells it with a tu quoque: Indifferentists have foresworn all rational grounds for choosing their own approach. Yet choices must be made, not all of them on a par with ordering coffee instead of tea. Those who take all values to be “irrationalizable” (103) have left themselves no rational leg to stand on. That argument, to be sure, is dialectical, best addressed to those not yet wedded to the nugacity of reason. It won’t gain much purchase on committed anti-rationalists. They can always say that demands for rational argument were just what they were trying to avoid. So Rescher does smoke out his adversaries, but does not exactly handcuff them. They can always say that to ask for reasons for their irrationalism simply begs the question. Still, it’s never useless to expose the pretensions of speciously reasonable dismissals of the use of reason. Does Rescher’s tu quoque slap back at his pluralism? Facing the objection that pluralism too cannot discriminate, he declares it perfectly consistent to respect a variety of views while taking seriously a limited subset and endorsing just one: “The fact that a plurality of positions is available does not mean that we have to see them all as equally correct and adoptable” (96). We need not treat all options equally or avoid all preferences. We have our reasons, and they will reflect our experience. But that’s no call to turn “feckless,” “spineless,” or fainthearted (89, 96). We should not be afraid to “stick our necks out” and “take a position that endorses some answers and rejects others” (96). Indeed, we must. We have to be committed to the views we hold: “We have no choice but to see our truth as the truth” (105). On one level that sounds like a truism. If saying ‘In my opinion...’ is just a kind of ritual forelock pulling, expressive of modesty or an admission of fallibility, it’s probably pretty innocuous. No stack of disclaimers, even if couched in the language of relativism, gets one wholly off the hook: Either we hold a view or we don’t. It’s fine to say we’re not sure and we’re open to conviction. But if conviction means anything, the bottom line is that with whatever caveats and qualifications remain, we’re convinced.

108

PLURALISM AND CONSENSUS

Consensus, as Rescher argues, is a ready substitute if truth itself seems unattainable. And advocates of the various species of “indifferentism”— skepticism, relativism, and syncretism—can always urge that truth itself is a dangerous illusion. They may readily brand commitments to objectivity “absolutist,” muddling into the notion of the name all the bad associations of know-nothing triumphalism, dogma and dictatorship. But holding a view is not the same as imposing it, and it’s the pluralist who recognizes that there are people in the world who live differently or think differently than he. The skeptic, relativist and syncretist, placing all views and ways on a par, pretends that none is any better than the rest, and so reduces all to one, whether in the syncretist’s amazing ball of wax, the relativist’s shallow deference to their diversity, or the skeptic’s wholesale dismissal. Mill’s alternative looks attractive when he promises that truth is what will likelier emerge where many and diverse voices can be heard. What troubles Rescher in that promise is Mill’s further hope that intellectual ferment will yield not just truth in the end but its common acknowledgment, albeit at “a stage of intellectual advancement which at present seems at an incalculable distance” (On Liberty, Part 2, ed. H. B. Acton, p. 105). Peirce strikes a more discordant note when he makes the ultimate consensus of inquirers not just the consummation of inquiry but the criterion of truth. Rescher, for his part, is very ready to admit that reality itself is out of reach. We have no God’s eye view, no secret passage away from and above our own perspectives. But what follows, on Rescher’s account, is no “indifferentist” attraction or repulsion but recoil from the intellectual apathy he finds among others who make and press the same admission: We need to acknowledge the limits of our experience and our methods but get on with our lives. Of course our human limitations and the open-endedness of experience will inevitably lead different persons (and different cultures) in different directions. But the inference Rescher draws is a need to recognize that we are who we are and think as we do: It’s perfectly fine to hold a view. We have to stick to our guns and have every right to do so, so long as we make every endeavor to be rationally conscientious in our thinking. For ‘right’ here just means rational warrant. What recognition of our limitations ought to teach us, Rescher argues, is not that we should hold no view or abide by no practice but only that we should not expect others to see eye-to-eye with us or march with us in lockstep. To be who we are, we can, and should, and must think the thoughts we think, hold the beliefs we hold, cherish the values we prize,

109

RESCHER STUDIES

and live by the practices we admire. We have no obligation to cave in just because others think differently. There is no obligation to ape their ways, and no intrinsic merit in yielding to their outlook. What morality asks of us is simply to give up the expectation or demand that they do or think as we do. And what epistemic humility asks of us is to make every effort to take account of others’ standards in the shaping of our own (110-11). Flexibility is one thing; pliability, quite another. It’s one thing to learn from others and quite something else to embrace a view or practice simply because others seem to expect that of us. One factor that drives an interest in consensus today (quite apart from our penchant as social animals, in many a good and bad sense) is the pressure toward a sort of ideological minimalism. It feels good, of course, to know that others think or feel as we do, that their goals are like ours, that they’re pulling for the same team. But that kind of feeling can goad inquisitorial behavior, jihads, show trials, and crusades, enforcing conformity not in the interest of social harmony or even the salvation of souls but out of an intellectual or moral imperialism that is all too ready to forget about sincerity or depth of commitment, so long as the longed for formula of surrender is heard. But when consensus becomes a goal in its own right (especially among equals), simplicity is at a premium. Clarity and coherence readily give way to a drive for univocity and the reduction of holdouts to (at least outward) conformity. Hence the immolation of consistency in doxastic creeds: The telltale marker of the heavy political hand is the residue of words: shibboleths linger where once, for some, at least, there was a shining, joyous faith. The politics of early Church councils, the secular and mildly skeptical humanism of the Renaissance, the Reformation search for simple, Gospel truths, the new secularity and secularism of the Enlightenment, the quest for some lowest common denominator in public education, or, failing that, a muttering of overlapping sayisms in societies at last ready to acknowledge their diversity, all press toward minimalism: What are the least offending phrases to be uttered, the least obtrusive gestures to be made. The formulae of compromise are battered into shape and into place not to make them ready for real credence but because they least stir the somnolence of consensus with unfamiliar claims, the strong scent of coffee, or of the fire it was brewed on. By a kind of Gresham’s law of ideas, bad thinking drives out good. Laws, of course, do, and rightly, press toward minima, demands that all (or nearly all) can live with. But law, quite naturally, drags morals along

110

PLURALISM AND CONSENSUS

with it, on the vulgar presumption that morals should hardly ask what law does not insist on. Why should we ask more of ourselves, morally, or intellectually, or spiritually, than we expect of one another? We can see the trend pretty clearly in religious institutions, when spirituality is commodified and the tough intellectual work and spiritual discipline of a tradition or personal quest give way to the gestures of commity, vocal lipservice and tacit laxity that passes for tact in the marketplace. Mill says something precious here, and seldom quoted: Though doing no wrong to any one, a person may so act as to compel us to judge him, and feel to him, as a fool, or as a being of an inferior order: and since this judgment and feeling are a fact which he would prefer to avoid, it is doing him a service to warn him of it beforehand, as of any other disagreeable consequence to which he exposes himself. It would be well, indeed, if this good office were much more freely rendered than the common notions of politeness at present permit, and if one person could honestly point out to another that he thinks him in fault, without being considered unmannerly or presuming. (On Liberty Part 4, ed. Acton, p. 134).

What Mill is addressing here is the biblical obligation of reproof (Leviticus 19:17), a sensitive matter, to be sure, but also an obligation constitutive in the larger if more abstract sounding commandment of the next verse, which it helps motivate biblically: Love thy neighbor as thyself. Part of Rescher’s program, intellectually and morally, is to call for the kind of candor that Mill is wishing for and that today, even more than in Mill’s day, is smothered by the demands of false tact—which too readily morphs into the pressure toward conformity and correctness that called forth Mill’s defense of intellectual freedom and diversity of life plans and patterns. One can, Rescher holds, respect another’s outlook without embracing it; and one can recognize another’s lifestyle without adopting it. But recognizing the possibility of respect and recognition of this sort does not commit us to the view that every outlook is worthy of respect or every life course admirable. More than one pathway can be and is rationally pursued, and more than one style or outcome of thinking can be and is adopted. That’s inevitable, Rescher argues, given our finitude and the diversity of human experience. Only think of the variety of grounds on which one theory might be given preference to another. A fortiori should we expect diversity in the larger pathways of thought. But it’s rationality, in whatever form, that wins respect. A viewpoint must be defensible publicly and indeed pub-

111

RESCHER STUDIES

licly defended to be worthy, in Rescher’s eyes not of credence but of deference. So one needn’t take astrology, say, or geomancy, seriously. Indeed, one shouldn’t, and (smart people at least) should be able to say why. But there are alternative rational outlooks, and they need not be cut off from one another. Precisely because rationality was their modus operandi, conversation and debate among them can and should become their modus vivendi. Communication, pace Keith Lehrer and Carl Wagner (137) and Donald Davidson (142) or Apel (148), does not presuppose consensus; and pace Habermas (153), it does not even presuppose a common goal of consensus. Indeed, Rescher argues, conversation presumes a bracketing of our assumptions about whose views are most adequate: “To be sure it is sometimes said that to understand other people’s discourse we must agree with them... But this is simply wrong. . . . we may need to know (or conjecture) what their beliefs are, but we need not agree with them. We can understand what Galen says about the four humours or what Leibniz says about his windowless monads without agreeing with these authors as to the issues involved” (142). We don’t have to agree with Anaximander (or, still less to share his world view!) to understand what he meant when he said that the sun was a hole in the firmament through which we can see the cosmic fire that circles the world (138). Praxis may work somewhat differently here from the more intellectual realms of symbolism or belief. Do we need to agree, if not to understand each other, then to get along? One problem in discussions of consensus and its lack is a slack-key slide from talk about agreement and disagreement over particular (or specific) facts to wider ranging talk about paradigms, or from language games to systems of syntax, from rituals, say, to ways of life. The notion of diverse “forms of life” facilitates the slide. Clearly we can’t have much of a conversation with someone if we don’t have a common language. And many people think there’s always a metaphysic or a cosmology embedded in our language. That’s an inference commonly drawn from Whorf’s work. And many presume that we can’t live together unless we share a world view. Hence the moral pressure to unify our world views. And yet translation, although no easy or mechanical task, is possible. Conversations do take place across cultural divides. And people who share a culture can and do differ profoundly. Poets and philosophers can work creatively at the boundaries of syntax and semantics, and regular Janes and Joes too regularly think and speak and act creatively. They don’t just make fixed moves within established language games. Great painters

112

PLURALISM AND CONSENSUS

don’t just color inside the lines, and neither do the rest of us when we’ve a mind to be ourselves. It’s easy to assume that to live together, if we don’t have common values on the object level, as it were, then we’d better share some values on a meta-level—say, about getting along or keeping the law, or living and letting live. But even that assumption is not perfectly true. Shared values do contribute to the functioning of a society. But value differences are found in any human group, even in a couple or among siblings, or parents and their children. And it’s not always through shared meta-values that such differences are mediated. It can be through common interests, or a sense of shared history or destiny. What unites a community need not be doxastic at all, although I think it will be ideational. Ibn Khaldun calls the commitment that holds a tribe or clan together as,.abiyya, literally, sinew. Pragmatically it means a sense of shared interests, underwritten by the recognition (tested in experience) that another’s gain or loss, pride, grief, or shame is one’s own. Tribally what’s critical is how dependably one can expect help when trouble strikes. It’s easy to reify that prospect, but phenomenologically it’s a matter more of roles than rules, a construal of identity not a body of beliefs. Roles still are values but not in the credal way that the history of European sectarian and secularizing history might have led one to expect; and the counters in play tribally are more often brides or reprisals than, say, oaths of allegiance, trade credits, or scholarships and cultural exchanges. But interests, not ideologies alone are the vitamins that build sinew, not just in tribes but also in nations. One of the merits of Rescher’s Pluralism is to turn our eyes in that direction. It’s pretty clear that people can live together (not just in the same country but even in the same household) when they don’t share the same outlook. And it’s widely observed that for practical purposes people can get along pretty well, keep the laws and uphold the constitutions of their lands, even when they have profoundly different reasons for doing so. Indeed, the liberal dispensation is based on the assumption that one need not (and should not) press one’s fellows as to the ultimate groundings of their allegiance. Outlooks, Rescher reasons, will always be diverse. So the quest for consensus is utopian (181). But what marks and builds the health of a healthy society are shared efforts and allegiances. Individuals and groups of all sorts can find their own good reasons for working together and getting along. “What matters for irenic conflict resolution is not second-order con-

113

RESCHER STUDIES

sensus but second order acquiescence.” And that is more a matter of personal and group concerns than of shared slogans or beliefs: “The rational person’s acquiescence is, after all, based on a cost-benefit calculation. . . . sensible people are distinctly unlikely to acquiesce in arrangements that are oppressive to them”(171-73). What follows, Rescher argues, is that “A political process that is both theoretically valid and practically workable must seek not to abolish divisive self-interest but to co-ordinate it.” The means to that end are “certainly not just the methodology of deliberative reasoning,” but “the forging of a community of consilient interests” (181). For “What is needed for cooperation is not consensus but something quite different—a convergence of interests” (180). Pressure to conform, in the interest of consensus, will prove an impediment to creativity, an invitation to mediocrity, and a disincentive to innovation and productive effort (163). Realism accepts “the unavailability of consensus” and strives instead to create “a communal framework of thought and action where we can come to terms with disagreement and make the best—and most—of it” (193). It’s framing the human political problem in terms of interests rather than agreement that leads Rescher to find Rawls’ posit of unanimity behind the veil of ignorance utopian. But the charge goes further. Rescher wants to speak of real people in the real world. That’s why he dismisses consensus as a chimera. He’s not interested in conjectures about the virtual agreement of ideally rational choosers in a fictive thought experiment. He too, of course, posits sensible people when he speaks of acquiescence. But that, he thinks, is a reasonable expectation. He assumes that people can be counted on (for the most part, at least) to recognize their own interests. That’s the basic premise of any liberal democracy, and it’s why he cites the ability of people to tell the difference between when they’re flourishing and when they’re being oppressed. Like any theory, that one might hold steady only within certain tolerances. It might not work too well with folks who see their interests in terms, say, of reconquering Andalusia, restoring the Caliphate, and toppling every secular state—and who would rather die than acquiesce. Rescher, like Rawls, thinks that people should and would desire to preserve their liberties and then their other interests, which he readily assumes are met with liquid assets. His theory is not well calibrated for societies or movements where liquid assets are so ample for some and so scanty for others as hardly to have weight in the plotting of a course. When conquest matters more than life itself, and dignity is bought by martyrdom, not work

114

PLURALISM AND CONSENSUS

or wisdom, wealth is not a stabilizing factor, and liberty needs to be reborn before it can take wing. Conversation is rarely fruitful with fanatics, and never with armed terrorists. Nor does social mobility unify the societies on which fanatics feed. It can as likely foster that sense of guilt and revulsion with self that (when turned upon the alien other) transforms alienated youths into suicide bombers and ordinary hoodlums and autodidacts into bomb designers and planners of suicide missions. Dialogue is not quite the answer here, and I think Rescher knows that. What he exposes as not just utopian but wholly artificial in Rawls is the notion of “ideally rational people” planning a society in “ideally rational circumstances.” “Knowing how to play a perfect hand in a card game will not help us decide how to play the imperfect hands that fate actually deals to us” (178). Rescher here pinions one of Rawls’ deepest weaknesses: Like the economists who give real world names to the variables in their models, Rawls would like his model to reveal the veritable nature of justice: Knowing what arrangements rational subjects would choose if unconstrained by need or fear or history, unbiased by any thought of the roles they themselves would play in the society they frame, and undistracted even by presumptions about their own strengths or tastes, should show us what is properly called justice. Rational choosers, Rawls informs us, would give lexical priority to the preservation of personal liberty and then use the maximin principle in the distribution of goods and ills. That is, inequalities would be tolerated only insofar as they benefitted even those least advantaged by them. This is not the place to note the many rivals to Rawls’ variant on Pareto equity. Nor should we dwell upon the tensions in the notion of rational choosers who don’t know their own values yet know themselves well enough to know what they would deem a cost and what they would welcome as a benefit and know just how risk averse they’d be in a society they have not yet lived in. (For otherwise, how do we know that they would not risk all, by giving up equality of any sort for a chance at wealth or fame or learning, virtuosity, romance, or power?) What stops Rawls’ car before it gets into motion at all is that it’s not a real car but a schematic. That is, Rawls, no less than Quine, is a mathematician. The results he draws from his analysis of rationality are just the ones that he’s packed into it. In a real society Rawls is impatient with those messy human beings who bring all their cultural and historical baggage along with them, and (in the name of liberalism!) bars political appeal to fundamental principles (“com-

115

RESCHER STUDIES

prehensive doctrines”). But seeing that he is no position to legislate what reasons people might adduce (or have in mind) when they argue about the ground rules under which they live, he confines his legislation to a stipulative definition of political liberalism, withholding the name of liberal from anyone who appeals to comprehensive doctrines in a political context. Rescher is not so Pickwickian. He locates liberty and liberality in free wheeling debate. As a kid getting my feet wet in philosophy, I used to make a certain mark in my notes when I came across the kind of contretemps that I later saw labeled with the rubric: ‘One philosopher’s modus tollens is another’s modus ponens.’ My signal to myself was a bit less wordy, if not quite as cute: just two exclamation points side by side, one upright, the other inverted. (The order of the marks could indicate which view was being rejected and which retained, or even presumed). The irony that bars a reductio from working as a lead pipe cinch is that it can spring an adversary right into the briar patch he was hoping for. It clarifies things to know or show that what one thinker finds absurd another took for granted all along. But the real lesson is that words like ‘tantamount’ will not do much to alter settled convictions. Al-Farabi saw that when he studied Aristotle’s case that assigning determinate truth values to sentences about particular events in the future could license a crippling determinism. That was just the outcome that Aristotle’s (Megarian) adversaries were looking for. Aristotle’s reductio was dialectical, and Farabi needed a stouter stick if he was to head off the Ashcarite predestinarians of his own day. (He found one in the recognition that the necessity by which a true prediction entails the corresponding facts does not infect those facts themselves. Hypothetical necessity does not cash out as categorical.) Does Rescher run that briar patch kind of risk when he makes our lack of instant access to a God’s eye view of things not only a warrant for his pluralism but, simultaneously, a part of his pitch for commitment? How does he avoid landing in the briar patch himself with Rorty, Feyerabend, or even Sextus, all of whom celebrate in different tones the same lack of omniscience? How, does fallibility become an argument for taking a stand? Formalizing his dissolution of the framework-relativity of a claim we actually make, Rescher finds just the claim itself left behind, an insoluble precipitate (105). Does that mean that whatever views we hold we’re stuck with; and the standards we accept we must persist in? Rescher rightly qualifies the impression of fatalism: Of course we want others to see our

116

PLURALISM AND CONSENSUS

viewpoint, and indeed to use our standards. But we are expected (and should expect of ourselves) to take others’ commitments seriously. All of us should make “every reasonable effort” to apply the standards that by our best lights “ought to be everyone’s” (108-09). Flexibility is a virtue. But it’s not an end in itself. Forthrightness is not the same as dogmatism. Rescher grips the tiller tightly here, steering between the Scylla of “megalomania” and the Charybdis of vacuity. He may tack a bit, allowing that “individuals do, in general, have a cognitive orientation and are able to resolve issues by its means” (105), but then urging that “Each thinker, each school, is bound to take a strongly negative stand towards its competitors: belittling their concerns, deploring their standards, downgrading their values, disliking their presuppositions, scorning their contentions, and so on” (108, and again on 120). Rescher’s heading is best plotted as the view that we should respect a variety of viewpoints (but not all) and can learn from any, but need not and should not assign them all “cognitive parity.” We don’t want to be stuck on the corner where a man has passed out, unable to decide between chanting incantations and calling an ambulance. Rescher mounts his commitment to commitment, that is, earnestness and openness about what we actually believe (or practice) on a kind of tripod. His rationalism is one leg, the philosopher’s faith that we should be willing and able to make sense of our commitments to ourselves and others. The second leg is candor, not quailing or paltering about where we stand or why. The third is his free admission that we have no godlike access to the absolute truth about all things (108, 118) or the clear page of the “Recording Angel” (77). The irony not lost on a philosopher of Rescher’s experience, is that this last premise is Rorty’s, and many another postmodernist’s. The critical difference is that Rescher has not relinquished, or even pretended to relinquish, his critical (and constructive) tools. He agrees with the anointed foes of foundationalism that we have no window to climb out of, to escape the envelope of our own views. But our self-styled post-moderns infer from this that truths or worlds are many and indistinguishable by reason. Rescher, by contrast, resolves to make the most of what we have: Truth is one (as it must be if there is to be logic or reasoning of any kind), despite the fact that “paradigms” are many—as they must be, given the divergences among any two slices of experience. But despite the divergences we see, we have good grounds for ignoring some outlooks and privileging others. When we’ve weighed up all the evidence at hand, we still have grounds for drawing our conclusions, and we

117

RESCHER STUDIES

need to draw the conclusions that we draw. We can do so in no expectation of (even ultimate) unanimity. The underlying premise here: Although we don’t own a godseye view of things, we do have ways of checking up on appeals to our credence or allegiance. So Rescher sees no need to play cynical games with inquiry as Feyerabend and his merry cohort do. Nor is he loath to use a moral test or two, alongside the familiar logical tests that philosophers put store in: The American polity is a better model for a pluralistic society than, say, Lebanon or Yugoslavia (167). And dissent can often break through (at least some) cultural trammels: “The Andalusian friar Bartolomeo de las Casas [1484-1566] upheld the human rights of Amerindians against the consensus of Spanish conquistadores and settlers alike that they were inferior beings” (163). Alonso de la Vera Cruz, his younger contemporary, dedicated his life to that cause in the mid-sixteenth century. One hopes it’s not just in a desire to please or to agree that others might say and see that this Bartolomeo and Alonso were right, and the conquistadores and settlers were wrong when they ground the faces of their poor victims. That last is no mere deduction from a general truth. It’s something rather more primary (if hardly universal!)—a moral truth in its own right, and one from upon which general laws might be constructed inductively just as readily as it might be derived from them once they have been framed. Alonso and Bartolomeo did not know that it was wrong to extort and exploit the Indians because they read it in a book. For there were many books, against which they wrote theirs, that taught just the opposite. One of the flaws in the notion of paradigms, world views, or blicks, is the assumption that such entities are airtight and logic-tight, immune to criticism by way of thoughts framed within their own categories or through the critique of those categories. The notion that a theory or construct cannot generate within itself the means of its own critique is one of the deep mistakes that many a twentieth century thinker picked up from the false analogy of thought with language and of natural languages with artificial schemes of logic. That false assumption, which should have been dissolved by Gödel’s discoveries about reflexivity, if not by common sense, fatally infects the notion of paradigms and world systems, treating them as selfcontained and even self-sufficient mechanisms and not as living things capable of biting themselves—or systems of living thinking, capable of being thought through, or seen through. Critiques, in fact, are not exclusively internal. The same sort of crosscultural checks that relativists enlist to undermine claims to objectivity can

118

PLURALISM AND CONSENSUS

also open up the eyes of the morally sensitive or conceptually adept. Moral differences do not disqualify all claims to moral truth, any more than moral affinities flush any such unicorns from forest coverts. If truths must be constructed, as postmoderns so often inform us, the fact remains that some constructs are sounder than others. Not every image is a sham, not every currency is counterfeit, not every relationship is an imposture or an imposition. Categoreal differences, real or imagined, do not discredit taxonomic realism, no matter how much they may give it pause. Category sets, like tools—or metaphors—have their varied uses, and their aptness or ineptitude depends upon their fit. It is a fact, and no mere fantasy that humans are closer kin to chimps than to mushrooms, no matter how many words the Inuit are said to have for snow or the beduins for sand. The uniform deep structure of all (human) languages (if we frame their rules with sufficient generality and abstraction) does not prove that we’re sure to slice the world rightly at the pre-existing joints. And yet, we can in some measure (and for varied purposes) triangulate (as forest rangers do, not candidates for office), to test our presumptions against their logical or cultural alternatives and project alternatives and syntheses that rise above the dichotomies we encounter. So we can open up a rent or two in the fabric of familiar suppositions and let in a little light and air. Granted we don’t have a superhighway to the absolute as dogmatists and the marketers of scientism, positivism, moral or spiritual dogmatism, instant enlightenment or salvation like to claim. But experience can be objectifying. Human it remains, and situated, as every human individual and group must be—and yet, not fixed immovably in place. We can grow a bit through the broadening effects of travel, temporal, cultural, geographical, or conceptual. And it is only human knowledge, after all, that we need; and human virtue, presumably, that we’re after. Our study is not to become angels or jinn, or happy naked ratmoles somewhere in Alpha Centauri. We plant our ladders here on earth, and our effort will never do more (but should never try for less) than to rise somewhat above our starting point, using the tools we are given and those that a culture, increasingly aware of its own limits and diversity and of options that lie somewhat beyond those limits, allows us to make our own. In the end I don’t think Rescher needs to be quite as cautious as he is about building his case for conviction on the fact of our human limitations—although there’s a neat twist in that, since he’s building on his adversaries’ premises. My own take on the situation is that we humans,

119

RESCHER STUDIES

through and not despite, our various traditions, be they scientific or religious, moral, artistic, social or political, do sometimes work our way toward truths, and toward other values that can be crystallized and expressed as truths. That observation seems to me to chime well with Rescher’s pluralist intent—and, for that matter, with the project of the university, properly conceived. Our ability to test one tradition against another, since human ways of thinking and the values that arise in human ways of living are not utterly incommensurable, gives hope to the more critical and selective type of syncretist, who is willing not only to convince or be convinced but also to learn and grow.

120

Meditations, Wagers, and Existential Issues: Nicholas Rescher on Religion and Philosophy John Haldane

N

icholas Rescher is rightly recognised for the breadth of his philosophical interests and for his intellectual industry. Indeed in those respects he resembles a figure from the nineteenth or early twentieth century of European or American philosophy.1 Yet even placed in this comparison class he continues to stand above the assembly in respect of the range of his sustained intellectual enquiries and the scale of his writings. Here, I am concerned with two aspects of his work as a philosopher in which, by contrast to other facets, he has been almost restrained in his output. They are, however, of real importance both for understanding his intellectual dispositions and general philosophical position, and for the insights they offer into some central and pervasive philosophical issues. The aspects in question are his reflections on religion, and on the nature of philosophy itself conceived of not in terms of its internal problems and specific methods and techniques, but as an aspect of a recurrent response to the human condition. Prominent among the connections between these two aspects of his work is an appreciation of the extent to which ideas, values and beliefs may be pragmatically justified. As he puts it in the essay on ‘Religion and Pragmatism’: [A] belief or practice is rationally warranted to the extent that its acceptance and implementation yield results that align with the expectations that its acceptance underwrites. (Rescher 2007, 25).

Before proceeding to discuss Nicholas Rescher’s philosophical position directly, however, it is relevant to contrast it with the general character of contemporary analytical philosophy in respect of the same issues. In order to do that I shall begin with some broadly drawn historical observations, a method which he himself has sometimes adopted. Western philosophy became systematic, dialectical and academic in the period of the high middle ages following the example set by Anselm in his dialogues De Grammatico, and De Veritate, and his ‘meditations’

RESCHER STUDIES

Monologion and Proslogion, which were drawn and distilled from his teaching to monk-pupils. This mode of precise and rigorous argumentation was further developed by Aquinas, Ockham and Scotus leading to such systematic presentations as their various Quaestiones Quodlibetales and Summae Logicae and Theologiae, . Their concerns were typically theological, but to arrive at satisfactory answers to the questions posed by belief they found it necessary to draw a number of metaphysical, conceptual, cognitional and psychological distinctions, and to argue rigorously, and often in detail, against certain errors and fallacies. Of this period, Rescher writes that: It was then, in fact, more than at any other stage of history, that philosophy stood at the center of academic and intellectual culture. And most of the complaints made about the thinkers of the era—their preoccupation with subtle distinctions and logic chopping—fail to do justice at once to the seriousness of their concerns and to the fact that their subtleties generally served a clear and present purpose with regard to the clarification of significant philosophical issues (Rescher 2005, ix)

Thus was born a form of analytical, problematical thought that in its later scholastic stages became increasingly technical and preoccupied with logic and grammar. The parallels between late scholasticism and present day philosophy have been drawn in general terms, but trends in the historical study of the former, and in the development of the latter, provide definite and detailed parallels, particularly in formal ontology and in the philosophy of logic and language. As Rescher notes, however, in an interesting study of one such area of parallelism, namely that of reference to the non-existent: The prime focus of their [medieval philosophers’] attention was theology, and their dealings with non-existence related to the role of such items in the thoughts of God rather than in those of men” (Rescher 2005, 126).

This might seem a point of detail, a reminder that their “clarification of significant philosophical issues” was in the service of theological concerns, a matter of possible interest to intellectual historians but not germane to the philosophical task of understanding the nature of thought and its possible objects. That, however, is a mistake. John Stuart Mill once wrote that “it really is of importance not only what men do but what manner of men they are that do it” (Mill, 1859, Ch. III), and one might add that questions of

122

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

motive, interest, presupposition and ultimate concern also determine the character of one’s intellectual enquiries. To deny this, or to overlook it, would be an error akin to that of the empiricist who thinks that visual (in)discriminability exhausts the grounds of cognitive differentiation. The fact is that in singling out an object, even where that identification begins with or takes in perception, one brings a range of interests and knowledge to bear; and in consequence one may legitimately conclude that what looks the same may yet be different. Similarly, the nature of an activity is given not only by its elements and their empirical or even logical ordering but by its presuppositions, its form and its end. The scholastics would refer to the latter as its “finality”, and it may seem that “end” is just a variant of this; but that easy assumption is indicative of the failure I am concerned with. The notion of finality as it features in scholastic Aristotelianism carries implications of teleology and of an embedding of proximate, distal and ultimate points and purposes which may be related intrinsically (per se) or extrinsically (per accidens). Moreover, even where elements are ordered per accidens, as in the causal series: a causes b; b causes c; c causes d (the sound of my stirring coffee wakens the cat; the movement of the wakening cat knocks over the cup and spills the coffee; the spilling of the coffee stains the floor), there will be additional radiating series in which the same elements are related to others in per se orderings, as in: a causes a′′ in causing a′ (my hand stirs the coffee in turning the grasped spoon; the cat expels carbon dioxide by breathing ; etc). Within this intellectual context it is likely to be further supposed that the existence of per se orderings of causes, and of finalities (which in this view are a kind of cause2) leads to the postulation of a source of causality that is itself a first cause, ('hoc est causa prima') of which Thomas Aquinas observes ‘et hoc dicimus deum’—and this we call God’. It is just such an understanding of causal sequences as ordered per se, and not merely as non-transitive series of extrinsically related events, that differentiates the cosmological and teleological theistic proofs of Aquinas from those of modern day analytical philosophers of religion such as Alvin Plantinga and Richard Swinburne. The general point, to which I return, is that two activities may be empirically indiscriminable and yet be different in kind; and so it is, I believe, with the case of medieval and contemporary metaphysics and philosophy of language; for the assumption of God and of an economy of creation, covenant, vocation, grace, sin and salvation makes quite a difference to the matter and purpose of what it is one is doing. So when Rescher observes

123

RESCHER STUDIES

that the prime focus of the attention of medieval philosophers was theology, and that their dealings with non-existence related to the role of such items in the thoughts of God, he touches on a real and profound difference between the philosophy of intentionality pursued in a religious framework, and theories of mental reference developed in a quasi-scientific one. The last stages of scholasticism were contemporary with the rise of modern philosophy. Thus the lives of the late-classical thomists Francisco Suárez, (1548–1617) and John of St Thomas (1589-1644) more or less coincide with those of Francis Bacon (1561-1626) and René Descartes (15961650), and in view of the latter’s scholastic education by the Jesuits of the Collège Royal de La Flèche, it is unsurprising that there are points of contact between the old and new philosophies. Still, the ‘new philosophy’ really was new in the place it accorded to empirical science. One might say that at this point philosophy was ‘de-theologised’ and made ‘scientific’, but that would miss two points. First, the three great rationalists Descartes, Spinoza and Leibniz all had religious interests, were theologicallyknowledgeable, and were alert to the challenge of expressing themselves in religiously charged times. Likewise, the first two of the three great empiricists, Locke, Berkeley and Hume, were also religious, theologicallyinformed, and engaged with the denominational politics of the period. Notably also, the two great respondents to the sceptical, anti-theistic, theologophobic and secularistic Hume, namely Kant and Thomas Reid, were again religiously minded, the latter being a Presbyterian minister . The second point against the idea that in modernity philosophy was simply and directly de-theologised and made scientific is that it overlooks the impact of philosophy upon science. Here again empiricism is to blame in leading commentators to neglect the extent to which science then, and still today is not simply a matter of highly technical looking, measuring and testing, but is structured by metaphysical concepts and theoretical modes of inference. All of that said, and allowing for pre-modernist revivals (such as neoscholasticism) and the various phases and modes of philosophical idealism, the trend of philosophy, particularly in the English-speaking world, has been towards a form of professionalized academic study modelled on the fundamental sciences, and on mathematics and its applications in economics and computation theory. And although it remains the case that science presupposes a background metaphysics, that framework has been less open to philosophical scrutiny, in part because scientists have been educated without training in philosophical analysis and criticism, and in increasing

124

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

isolation from the humanities; and in part because philosophers themselves have come to adopt an attitude of awe and deference towards science, analogous to that hitherto accorded to theology. To this last point needs to be added the observation that whereas a century ago someone taking up philosophy would tend to have had a literary (often classical) and religious education, for the last few decades the background of philosophers has shifted towards science and mathematics, to the extent that one might almost speak of a ‘dehumanitisation’ of the discipline, certainly in respect of logic, epistemology, and metaphysics. Finally, there is the fact, now almost forgotten, that while philosophy is one of the most ancient intellectual disciplines, its institutionalisation as an academic profession for which people are trained as a principal form of lifelong employment, with associated professional organisations and career structures, is a quite recent phenomenon. Prior to the end of the nineteenth century, English philosophers had either been clergymen, doctors, civil servants or private tutors. The situation in Scotland was somewhat different because of the deep-rooted position of philosophy in its ancient universities, a position re-vitalised in the Scottish Enlightenment; but even there philosophers were typically other things besides, again very often clergymen. Something of that Caledonian tradition was inherited in North America with the result that in liberal arts colleges philosophy was typically yoked to religion and taught by ministers.3 As Dewey remarked of his education at the University of Vermont: “Teachers of philosophy were at the time, almost to a man, clergymen” and he adds “Just how and why Scotch philosophy lent itself so well to the exigencies of religion I cannot say” (Dewey 1930, 1998, 15). • • •

It is against this general background that I wish to place Nicholas Rescher’s contributions to philosophy of religion and meta-philosophy, because doing so serves to highlight how different these are to the common run of analytical work, notwithstanding Rescher’s equal commitment to rigour, precision and the deployment of formal techniques where they may be appropriate. I can put my general points very briefly. First, that whereas analytical philosophy of religion tends to be theoretical and speculative, often a field of applied epistemology and metaphysics, Rescher’s contributions are more in the tradition of practical philosophy and are alive to the fundamental existential question of the meaning of human existence. Second, that while analytical philosophy tends to be unreflective about its own nature

125

RESCHER STUDIES

and purpose, Rescher is alert to the need to understand what philosophy is and how it relates to broader human concerns, particularly ones regarding values. In saying that analytical philosophy does not much reflect on its own character I am not saying, as some, even of its practitioners, might, that it lacks a self-conception or an implicit nature. In fact, as my previous observations suggest, it operates as a kind of science (with ethics being in part analogous to technology). Some philosophers are barely aware of this. Others recall Quine’s naturalisation of epistemology and ontology and the idea that philosophy is the most abstract end of the spectrum of scientific activities; as J.J.C. Smart put it: Metaphysical theories are further removed from control by experiment and observation than are typical scientific theories even though, as I think, this is a difference in degree rather than kind. (Smart 1993, 81).

Moving further in that direction, yet other philosophers celebrate the liberation of philosophy from traditional humanistic and existential concerns, seeing in this the possibility of achieving real and irreversible progress in the resolution of what then appear as theoretical and technical challenges (see Haldane 2001). The point of saying that analytical philosophy does not much reflect on its own character is rather that the question of the nature of philosophy is not much considered as a matter of reflective analysis; and even rarer is the practice of self-criticism. Rescher stands apart in this respect. He has written a number of pieces directly on the nature of philosophical practice, but more importantly he brings a conscious conception of philosophy to the task of investigating issues of a religious and existential nature. One important example is his book on Pascal’s Wager which devotes a good deal of attention to clarifying the context and ambitions of Blaise Pascal’s reflections, and to setting their conclusions within the framework of an account and defence of pragmatic reasoning. This stands in marked contrast with typical analytical treatments of the Wager which generally miss its point and treat it as a kind of failed demonstration. Given my characterisation of the analytical conception of philosophy as theoretical science this is hardly surprising. From that perspective the only question that arises is: is this an adequate existential demonstration? And to settle that one will need to consider whether to the extent that an eliminative induction is implicit or deployed all possibilities have been identified, and whether the antecedent probabilities of these have been determined; and that done what the various entailment relations

126

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

may be. Viewed in this light the state of mind and prior commitments and needs of an enquirer are besides the point, as would be those of a scientific experimenter or theorist. Questions of existential significance or adequacy, or of deliberative agency are quite extrinsic and any reference to them would be at best distracting and at worst distorting. From that perspective Rescher’s reproduction of the prologue to the argument and his attention to it will seem at most irrelevant. Yet this misses the point, and in missing it is liable to miss also the essentially practical nature of Pascal’s and Rescher’s reasoning. Let me quote some short extracts from Pascal’s text in the translation of the pages of the Pensées that Rescher’s presents: Our soul has been cast into the body, where it finds number, time, dimension. . . . The finite is annihilated in presence of the infinite, and becomes pure nothingness. Even thus does our intellect stand to God. . . Who then can blame Christians for not being able to give a logical reason for their belief, professing as they do a religion which they cannot explain by reason? . . . If they proved it, they would give the lie to their own words; it is in lacking proofs that they do not lack sense. Let us examine this point and declare: ‘Either God exists, or He does not.’ To which view shall we incline? Reason cannot decide for us one way or the other: we are separated by an infinite gulf. . . . Which will you choose, then? Since a choice has to be made, let us see . . . Your reason suffers no more violence in choosing one rather than the other, since you must of necessity make a choice. That is one point cleared up. But what about your happiness? Let us weigh the gain and loss involved in wagering that God exists. Let us estimate these two possibilities; if you win, you win all; if you lose, you lose nothing. Wager then, without hesitation, that He does exist.. . . . . . ‘But if I am made so that I cannot believe. What then do you wish me to do?’ . . . That is true. But understand at least that your inability to believe is the result of your passions; for although reason (now) inclines you to believe, you cannot do so. Try therefore to convince yourself, not by piling up proofs of God, but by subduing your passions. . . Learn from those who were once bound and gagged like you, and who now stake all that they possess. . . . Follow the way by which they set out, acting as if they already believed, taking holy water, having masses said, etc. Even this will naturally cause you to be-

127

RESCHER STUDIES

lieve and blunt your cleverness. ‘But that is what I fear’ Why? What have you to lose? (Rescher, 1985, 2, 8-9).

In his analysis and discussion of the Wager passages, Rescher emphasises first, Pascal’s sceptical stance so far as concerns human cognition in general and the possibilities of demonstrative proofs in particular; second, the assumption of an (orthodox) Christian context; third, the practical character of the reasoning, both in the respect of being directed to action rather than theoretical knowledge, and in regard to its essentially prudential character appealing to the hearer’s needs and wants; and fourth, the idea that it is legitimate and possible to bring oneself to belief through practical strategies. It is a revealing indicator, I think, of Rescher’s estimate of how likely analytical readers are to fail to understand these points, or, understanding them, to take them to be reasons for dismissing Pascal’s reasoning, that he repeatedly stresses the sceptical, contextual, practical and apologetic nature of the wager argument: Pascal’s reasoning is only in a position to persuade someone who believes that God may exist. The job of the wager argument is simply to establish that belief is rationally warranted. We [then] step outside the province of the argument as such into the region of effective praxis. The Wager argument’s task is to show that belief is rationally warranted—in the specifically prudential (not evidential!) mode of rational warrant. The argument only maintains that it is prudent (rationally advantageous) to accept (believe, have faith) that God exists. It is prudential and not, strictly speaking, evidential. [I]t is a special-purpose instrument with a limited and special mission—to stiffen the backbone of the slack and worldly Christian. It is theology alone that moves Pascal toward [the prospect of reward rather than through the threat of punishment]. Just how God is conceptualized thus becomes the crucial issue. It is addressed only to people who have a very definite sort of view of what God (should he exist) would be like—to the person who is committed to the Christian idea of God but hesitates about believing in him. (Rescher, 1985, 16, 17, 19, 21, 27, 92 and 98).

128

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

One response to this interpretation of the Wager might be to say that so understood it is devoid of philosophical interest. If it is not about establishing knowledge of the existence of God on the basis of some or all of empirical observation, conceptual analysis, and deductive inference; but is only a strategy for achieving personal satisfaction based on dogmatic doctrine and ritual, then whatever its interest it cannot be that of a philosophical argument. Quite apart, however, from the fact that Rescher extracts ideas about the nature of practical and probabilistic reasoning that are evidently of philosophical interest, the rejoinder I envisaged is an illustration of the question-begging conception of philosophy as speculative science, a conception I have challenged and with which I have contrasted Rescher’s own approach to the discipline. What is at issue in the case of Pascal’s Wager is not simply the character and merit of that piece of reasoning; but the nature and aims of philosophy itself. • • •

Here it may be helpful, and is relevant given Rescher’s knowledge and admiration of the that tradition, to draw upon the scholastic way of thinking about action and activity. First, then, recall the Aristotelian principle, at once methodological and metaphysical, that acts are specified by their objects, powers by their acts, and substances by their powers. Adapted to present purposes this tells us that what philosophy is, is determined by its proper object, insofar as it is an activity shaped by and completed in the attainment of that end. On one account the end of philosophy is truth, but that is not very informative: first, because truth may be the end of other activities also; and second, because truth itself may be an analogical notion applicable to a range of different though intelligibly related ends. Part of what it would mean to uphold the analogicality of ‘true’, is that truth is distributed between speculative, practical and contemplative dimensions, with these themselves subdividing into further regions. Suppose however that we distinguish philosophy from other truthseeking activities by saying that it is essentially reflexive and involves judgments of second intention; or equivalently that it takes as its subject matter first-order activities and their constitutive principles, such as concepts and rules. Then we may link this to the idea of dimensions or truth and derive the result that philosophy is itself an analogical notion completed in relation to different forms of truth. The correct conclusion of this reasoning, I believe, is that there is no primary form of philosophy but that philosophical activity gathers around a number of dominant modes. In try-

129

RESCHER STUDIES

ing to identify these it is useful to consider the practice of philosophy both present and past, and I suggest that as one does so a four-fold division emerges. First there is philosophy ordered or directed towards the understanding and realisation of beauty. This is to be distinguished from, but may include, aesthetics as that is usually understood; for an interest in ‘beauty’ might not be in the attainment of it as a good or, as it would be more apt to say, a form of ‘truth’. Philosophy for the sake of beauty draws in the aesthetic more deeply as a per se interest, and is itself made answerable to beauty as a governing principle of that very activity. Thus it is that philosophy in this sense is itself a kind of art-making. Second, there is philosophy as science whose end is the understanding and attainment of natural truth. By the latter I mean both the truth about nature, but also truth that is itself part of nature, and is attained by natural processes. This way of putting things obviously begs a question as to the character and extent of nature. In antiquity ‘physis’ or ‘natura’ were wideranging notions that could easily encompass immaterial realities, and in the middle ages nature is often treated as extensive with created reality. It is clear, however, that within contemporary English-language philosophy this question is commonly answered, explicitly or implicitly, in a broadly materialist fashion, and therein lies the appeal of the idea of naturalising philosophy itself. Third, there is philosophy as morality and politics, whose telos is the good itself further distinguishable in terms of virtue, welfare, and justice. Here again the end specifies the nature of the activity but also characterises it. So we find that serious moral philosophy seeks not only to understand the good but to attain it, and to do so in ways that are themselves virtuous and justice.4 Fourth and last, there is philosophy as religion. Here again I need to distinguish between the way in which religion features in contemporary analytical philosophy, if it features at all, and the manner of engagement and determination that I am concerned to identify. Religious phenomena may be made the objects of non-philosophical enquires, and also the objects of philosophy as art, as science, and as morality and politics. But besides these there is a more intimate formation effected by having as the end of philosophical activity the understanding and realisation of religious states and conditions, and by having this end determine the form of one’s activities. So just as the philosopher as artist works in aesthetically sensitive ways, e.g. through a certain literary style; and the philosopher as scientist

130

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

works in accord with scientific modes and methods, favouring datagathering, theoretical formulations, and so on; and the philosopher as moral or political agent is concerned to think fairly and justly; so the philosopher as religious agent adopts attitudes of conscious reverence and respect (understood in a non-sentimental manner). Just as I suppose that philosophy is a range of related activities, and do not presume that any one of these is primary, so I do not exclude the possibility of combining aspects of them in a process of intellectual enquiry (though there will be obvious possibilities of tension depending upon how the objects are conceived). Thus, I am not saying either that there is only one correct way to do philosophy in general, or for any given individual. What I do suggest, however, is when one looks at particular philosophers, or philosophical movements, or traditions, and when one survey the history of philosophy both in the West and more broadly, one finds philosophical thought conforming to and grouping around these four forms. Further to that I suggest that most philosophers are drawn to one or other orientation. Obviously this claim is impossible to establish by any empirical survey, in part because the terms of it are not themselves empirical ones but draw on conceptions by which activities will be described, interpreted an evaluated. Still, it can be illuminating to apply the scheme, and insofar as one may do so with regard to so productive and complex a figure as Nicholas Rescher, I suggest that he may usefully be thought of as engaging in philosophy as religion. Since this is a somewhat arresting claim let me immediately qualify it by saying that I am not offering this as a fully general characterisation of his thought and writings, and nor am I restricting the idea of religion to dogmatic or revealed religion. Rather I have it in mind that his specific writings in philosophy of religion, such as Pascal’s Wager, and the essays contained in Issues in the Philosophy of Religion, reveal an attitude and character to his intellectual activities that is not concerned to understand religious things in a speculative, scientific mode, but to understand and attain religious ends. For example, in “On Faith and Belief” he draws a very interesting distinction between doxastic and axiological attitudes towards the existence of God; that is to say between an interest in grounds for positing the existence of God, and desires or evaluative reasons for believing in God. As he rightly observes, these two sorts of attitudes admit of positive and negative orientations. Thus arises the following possibility: [S]omeone could be a doxastic theist but an axiological atheist. Indeed, axiological atheism is a far more drastic position than doxastic atheism, seeing

131

RESCHER STUDIES

that the axiological atheist is someone who, irrespective of whether or not he believes in God, would prefer to have him non-existent, and would do away with him if he could. (Rescher 2007, 2).

Lest it may seem an extravagance, born of the logic of his taxonomy, for Rescher to envisage such a possibility as the axiological atheist, consider the following from Thomas Nagel: I want atheism to be true and am made uneasy by the fact that some of the most intelligent and well-informed people I know are religious believers. It isn’t just that I don’t believe in God. It is that I hope there is no God! I don’t want there to be a God, I don’t want the universe to be like that. (Nagel, 1997, 130).

Rescher has interesting things to say about the possible psychological motivations towards axiological atheism which provide a telling counter to the more familiar antitheological psychologising. He observes, for example, that: Many people feel threatened by a belief that God might exist, because they feel that, were it so, God would not approve of them. For them, atheism is a security shield that protects them against an ego-damaging disapproval by somebody who ‘knows all, sees all’. Atheists are not infrequently people on whose inmost nature the vice of self-contempt has its strongest hold. (Rescher, 2007, p. 7).

The point of drawing the axiological/doxastic distinction is not, however, to classify holders of attitudes, or to speculate about the motives for belief and unbelief, but to set the scene for an argument to the effect that it is rational to hope for that which one conceives to be good, and that the best means for validating a hope takes the form of “trying and seeing and finding it suitably rewarding”. That is a pragmatic test, and Rescher adds the arresting thought that if the set of hopes and desires that constitute a religious aspiration pass the test of lived legitimacy thereby showing themselves to have value, it is reasonable, assuming no defeating doxastic argument, to proceed from axiological to doxastic theism. Let me quote: Accordingly, the failure to make a transition from axiological to doxastic theism, though certainly not manifesting any deficiency in logic, does betoken a somewhat regrettable failure of moral nerve. In the absence of a conviction that its realizability is infeasible, the rational person seeks to implement his

132

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

desires in action—and one of the ways in which we can implement a desire is by letting it guide our beliefs, by letting the wish be father to the thought insofar as the existing circumstances of the situation permit. (Rescher, 2007, 11).

Evidently we are back to the thinking of Pascal’s Wager as Rescher interprets it, but I wish to suggest three qualifying points. First, that the kind of reasoning to a religious commitment being recommended is broader than the specific example offered by the Wager. Second, that analogous pragmatic justification is available with regard to the acceptance of moral and other values as determinants of practice. Third, that there is a broader sense in which Rescher is a religious philosopher; one that combines ‘theological’, aretetic and eudaimonistic concerns. Put another way, Nicholas Rescher is a philosopher preoccupied with existential questions and with the meaning of life considered not primarily as objects of theoretical speculation but as matters about which one must arrive at an answer which, since the reasoning is ultimately practical, is also a decision about how to live. In short, he is an existential pragmatist. • • •

I began by observing that Rescher is a philosopher of incomparable range. But as G. K. Chesterton once observed, a scheme may be extensive though small in conception and simply endlessly repetitive. Such, I fear, is true of some philosophers; but it is not at all the case with the writings of Nicholas Rescher. Notwithstanding the pervasive orientation I have just proposed, in reading him one does not have the sense of a general formula being applied now here, now there. Rather, as one studies his work one finds ideas being discovered and explored in the very process of his writing. In that respect he is something of an imaginative author, and a philosopher as artist: setting a scene, introducing themes and characters and then charting the working out of an intelligible sequence. This aspect comes to the fore in his writings on value and meaning as these are intimated in human experience. In acknowledgement of that aspect of his philosophy, and somewhat in the same spirit, I should like to conclude by reflecting on the general issue of seeking meaning in human life and making sense of what we find there. Deep down in human kind and never far from the surface of serious human activities lies a desire for fulfillment. That desire has several aspects including cognitive, emotional and practical ones. In one or another way we seek meaning in the world around us while also trying to make sense of our

133

RESCHER STUDIES

own lives and of the human condition more generally. Sometimes the particularities and the details of immediate concerns obscure the fact that what are at issue are larger questions concerning the content and status of morality, the character of community, the purposes of society, the role of the arts, sciences and philosophy, the possibilities of transcendence, and similarly fundamental matters. Evidently, however, one may seek something without finding it; but to speak of 'making sense' suggests an achievement, and in the context of the search for meaning it further suggests a successful quest. There is an ambiguity in the expression ‘making sense’ which conveniently presents us with two contrasting attitudes to the whole question of whether there is any such thing as meaning in the various departments of human thought and practice, and if there is, what it may amount to. For we may think of ‘making sense’ as a discovery of what is there already, some intrinsic order of intelligibility, purpose or value, waiting to be made sense of; or we may think of the process as being one of construction, of making sense as one might make a picture or make (up) a story. It is a familiar trope of ‘postmodernist’ critiques that traditional notions of meaning and value provided by religious narratives or by philosophical theories of human nature are no longer tenable, since they rest on false beliefs in objective metaphysical and natural realities. For these critics searching for meaning in life is like hunting for unicorns—both are pointless activities based on empty myths. There are, I think, three broad styles of response to such subversive scepticism or nihilism: first, romantic reaffirmation, second, self-conscious irony, and third, reform and renewal. I favour the last of these, and if I have him aright so too does Nicholas Rescher. This involves rearticulating and where necessary amending older conceptions of human nature and human values so as to show their coherence, plausibility and contemporary relevance. One way of vindicating traditional ideas is by showing that they have a part in our best, and lived, understanding of human experience at a very general level—and not a mere walk-on contribution, or even a small speaking part, but a central and indispensable role in the human story. That is something Rescher has done extensively in his reflections on morality and value more generally (Rescher 1987, 1989, 1990, 1992, 2004). The sceptic or nihilist may respond that it is one thing to find that human beings recurrently resort to certain styles of thought and evaluation, and quite another for those to be coherent, let alone justified. After all, it is a familiar theme of subversive criticism that what proclaims itself as ra-

134

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

tional is in fact a pathological search for security, and what asserts itself as good is in truth a strategy for exercising domination. Arresting as these suggestions may appear, such criticisms like those of religious belief as depending upon psychological insecurities are themselves, as Rescher pointed out, open to just the same speculative analysis. Might not the postmodernist be asserting the absence of truth and of value out of a fear that if they exist then he or she will be under the burden of recognising and respecting them? Might not the denial of a religious reality be an expression of fear lest there be a standpoint of judgment beyond human prejudice, and a prospect of eternal gain or loss? These are serious matters to contemplate, and it would not be a surprise if rather than face them inventive people devised the coping strategy of saying that, after all, they are only fearful fictions. Moreover, there is a second-order consideration which suggests that where a style of thinking and its associated presuppositions, are recurrent and even constitutive of areas of thought and practice that are otherwise found to be valid and productive, then the indispensability of those presuppositions may be the best argument there is for believing in them. Consider, for example the practice of arbitrating competing claims, or that of imposing sanctions for misconduct. In the course of these, assertions may be made about what is fair or fitting, and these may themselves be contested by appeal to the idea of justice, either that dispensed by a higher authority, or ultimately that of justice itself. The subversive critic may then suggest that talk of fairness and of fittingness is simply a veiled invocation of power or interest. Taking that prospect seriously, however, does not eliminate the possibility that there is right and wrong in the matter. Someone may assert out of a motive of greed or fear, that justice requires such and such, and institutions may perhaps be created for the same reason; but it seems impossible not to think that these very facts require us to ask whether the verdicts and sanctions were then themselves unjust. Indeed, the natural, if not the logical implication of a claim to ulterior motive is that the action in question was wrong. Of course it may not be. A mathematical calculation done out of a wish to impress rather than out a concern for truth, may nonetheless be correct; and a verdict handed down and a sanction imposed out of a wish to dominate may nonetheless be in accord with abstract principles of impartial justice. Not only that, but the latter promise answers to a question that will not go away: was what was done, in itself right? It is imaginable that we might be so innocent as not to wonder about possible ulterior motives in the dis-

135

RESCHER STUDIES

pensation of ‘justice’. It is inconceivable that we could have arbitration and the administration of sanctions without the question arising: is this fair? This is simply part of the human form of life. ‘Well it may be’ concedes the critic, ‘but that does not show that there is any such thing as justice per se, or equivalently, that there are truths about what justice (the principle not some particular institution) requires’. Now, however, we are entitled to ask where lies the standpoint that the critic occupies and which is outside the human form of life. This is, of course, an ironical question intended to make the point that our judgments about what is the case are and can only be our judgments, informed by what we know, or believe we know—from small detail to general structure. And what we believe we know, as evidenced by what we say and do, and by how we evaluate what others say and do, is that there is truth and falsity, there is virtue and vice, there is justice and injustice, there is beauty and ugliness, there is nobility and depravity, and there is reverence and blasphemy, and there is the hope of a good God and the fear of existential meaninglessness. This being so, there is also the question, now an unambiguously philosophical one, of what all of this implies about the nature of reality. The unsubtle critic having been given something to think about, there is a more nuanced questioner to be addressed. This person agrees that human experience is structured by meanings and values, and further agrees that this fact cannot be set aside by positing alternative explanations of why we think in these terms, and therefore accepts that we are, and are set to remain meaning and value governed beings. Yet he nonetheless withholds assent to meaning and value as objective realities. This subtle doubter acknowledges that we are meaning-seeking and sense-making beings, recognises that sense-making is not the same as asserting power or attempting to resist death, but yet thinks that making sense is not a matter of discovery but one of construction or invention. On this account we are indeed answerable to standards of value and virtue, and are aesthetic and even spiritual beings, but we are all of these to the extent that, and because, we are imaginative and creative animals that construct an intermediate surface between ourselves and the purely material world, an intermediary lining on which we draw and colour our compositions. This is a concession to objectivity, allowing that there is something we seem to see which calls for our attention and earns our regard, but it is also and fundamentally a form of subjectivism inasmuch as it regards the source of that ‘something’ as lying within us. By way of illustration of

136

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

what I have in mind consider the following from Roger Scruton, a brilliant practitioner of the art of illuminating the structure of human meanings: Birth, copulation and death are the moments when time stands still, when we look on the world from a point at its edge, when we experience our dependence and contingency, and when we are apt to be filled with an entirely reasonable awe. It is from such moments, replete with emotional knowledge, that religion begins, and the rational person is not the one who scoffs at all religions, but the one who tries to discover which of them, if any, can make sense of those things, and, while doing so, draw the poison of resentment. (Scruton, 2007).

Admittedly this is a kind of psychologising, but not one that seeks to subvert the dimension of human experience that it describes. It is rather an exercise in reflexive ‘natural history’, noting that this is a part of the human form of life, and explaining what is going on within it in terms borrowed from the relevant sensibility and practices themselves. Be it respectful, however, it remains a challenge to the interpretation in which human practices are contrasted with transcendent realities. There seem, then, to be three positions in contest: the subversive sceptic-cum-nihilist who says it is all a meaningless projection; the hermeneutic constructivist who recognises the dimension of meaning as having its own structure and hierarchy of values; and the value realist who urges that what is rejected tout court by the former, and is recognised by the latter, however partially, is something real and unconstructed, though it may be contributed to by the forms of our sensibility. How to resolve the dispute between the second and third parties? In the year following the end of World War I, the British political theorist and economist Harold Laski wrote as follows: . . . Our business if we desire to live a life not utterly devoid of meaning and significance, is to accept nothing which contradicts our basic experience merely because it comes to us from tradition or convention or authority. It may well be that we shall be wrong; but our self-expression is thwarted at the root unless the certainties we are asked to accept coincide with the certainties we experience. That is why the condition of freedom in any state is always a widespread and consistent skepticism of the canons upon which power insists. (Laski, 1919).

Laski was writing about a political situation, in a time of post-war trauma, in the wake of the Russian Revolution, and from the perspective of a so-

137

RESCHER STUDIES

cialist intent upon creating a more just society. His socialism, however, owed more to nineteenth century British moral thinkers, in particular John Stuart Mill, than to Marx and Engels. Here Laski adopts a familiar revolutionary theme, namely the attack upon tradition, convention or authority; but the real target of his concern seems to be inauthenticity. The reason for questioning established ideas is not that they may be wrong, or even that they are promulgated from on high, but that if we are to be self-expressing creatures then whatever we give life to must originate within, or at least be confirmed from that quarter. Scepticism here is not directed against the possibility of objective truth handed down to us but suggests that our own avowals may be inauthentic, not issuing from our own certainties but from those of others which we then unquestioningly accept. This theme was once very familiar. It is, though, flawed in two respects. First the elevation of individual personal experience over other sources of knowledge is a residue of empiricism, i.e. of the belief that all true knowledge has to be traceable to the impact of the world upon one’s own senses. The reality, however, is that before we are in a position to validate propositions against experience we must first be able to formulate them, and that calls for conceptual resources that were acquired in the process of being taught a language, and which were deployed in experience long before we were able to adopt a critical perspective upon it. In short our ability to question authority itself depends upon authoritative instruction that, logically, we could not have questioned. Second, Laski’s highlighting of the themes of scepticism and certainty suggests the application of a misapplied standard to external sources, which, they having failed it, are then substituted for by internal conviction. In other words, Laski agrees with what he takes to be the proper demand of certainty, but because of his empiricism he relocates this internally, within the individual subject. Later thinkers, having learned the lesson of the many attacks on the very idea of objective certainty, then take the further step of concluding that the only way in which something could enjoy authority is if it were beyond questioning by the subject; and the only way in which this possibility could be provided for is if the subject is his own authority in virtue of having created the thing in question. We can only find meaning and make sense if we are ourselves the authors of meaning and the constructors of sense. Given this diagnosis a pragmatist Rescherean remedy suggests itself. If you agree that seeking meaning and making sense are part of the human form of life, and if your reason for not allowing that these activities could

138

MEDITATIONS, WAGERS, AND EXISTENTIAL ISSUES: NICHOLAS RESCHER ON RELIGION AND PHILOSOPHY

have objective counterparts is that we could never be certain of having identified them unless they were our own creations, then let go of the demand of certainty. Making sense is a fallible business, and just like any other human endeavour that seeks to align itself with objective standards there is no reason to think that a first-person perspective is more likely to be right than an external guide. Seeking meaning is something that human beings do, as is making sense. It is hard even to describe these activities without presupposing objective standards for them; and an acceptance of human fallibility and of our dependence on the authority of others releases us from the pressure to protect the idea of making sense by treating it as a matter of construction (about which we cannot be accused of error), rather than one of discovery (with the implication that we may fail). Interestingly, therefore, the most intellectually modest position is the pragmatic realist one. It postulates a destination without assuming that we are on the right road towards it, or that having set out we must arrive. The corollary of this modesty, however, is the prospect that if we reach journey’s end in the search for meaning then we will indeed be able to make sense of things by having discovered real truths about human life and its fulfillment. In this context to speak as I have done of ‘realism’ and the ‘real’ is not to stand in opposition to the philosophical pragmatism I have drawn from Rescher, since the source and test of the real is experience and practice. To repeat an earlier quotation from his essay on “Religion and Pragmatism”: [A] belief or practice is rationally warranted to the extent that its acceptance and implementation yield results that align with the expectations that its acceptance underwrites” (Rescher 2007, 25).

REFERENCES Dewey, J., “From Absolutism to Experimentalism,” in Hickiman, L. A. and. Alexander, T. M (eds.) The Essential Dewey (Bloomington, IN.: Indiana University Press, 1930, reprinted 1998), Volume 1. Haldane, J., “Has Philosophy made a Difference and could it be Expected to?” in A. O’Hear (ed.), Philosophy at the Millennium (Cambridge: Cambridge University Press, 2001). ———, “American Philosophy: ‘Scotch’ or ‘Teutonic’,” Philosophy, vol. 77 (2002), pp. 311-29.

139

RESCHER STUDIES

———, “Gravitas, moral efficacy and social causes,” Analysis, vol. 68, (2008), pp. 34-9. Laski, H., “The Dangers of Obedience,” Harper's Monthly Magazine, vol. 159 (1919). MacIntyre, A., Ethics and Politics: Selected Essays, Volume 2, (Cambridge: Cambridge University Press, 2006). Mill, J. S., On Liberty (1859). Nagel, T., The Last Word (Oxford: Oxford University Press, 1997). Rescher, N., Pascal’s Wager: A Study of Practical Reasoning in Philosophical Theology (Notre Dame, IN.: University of Notre Dame Press, 1985). ———, Ethical Idealism: A Study of the Import of Ideals (Berkeley, CA.: University of California Press, 1987). ———, Moral Absolutes: An Essay on the Nature and the Rationale of Morality (New York, NY.: Peter Lang, 1989). ———, Human Interests: Reflections on Philosophical Anthropology (Stanford, CA.: Stanford University Press, 1990). ———, The Validity of Values: Human Values in Pragmatic Perspective (Princeton, NJ.: Princeton University Press, 1992). ———, Value Matters: Studies in Axiology (Frankfurt: Ontos Verlag, 2004). ———, Scholastic Meditations (Washington, D.C.: Catholic University of America Press, 2005). ———, Issues in the Philosophy of Religion (Heusenstamm: Ontosverlag, 2007). Scruton, R., “The Sacred and the Human,” Prospect, vol. 137 (2007). Smart, J. J. C., “Why Philosophers Disagree,” Canadian Journal of Philosophy, Supplementary 19, (1993). NOTES 1

My main title draws from parts of the titles of three books which represent Nicholas Rescher’s principal contributions to philosophy of religion and religious philosophy, namely: Rescher (1985) Pascal’s Wager; Rescher (2005) Scholastic Meditations; and Rescher (2007) Issues in the Philosophy of Religion.

2

On the variety of causes, including but going beyond the four of Aristotle: material, formal, efficient and final. See Haldane 2008.

3

For more on this subject and its implications see Haldane 2002.

4

On this conception of moral philosophy see MacIntyre 2006, especially chapters 1 and 2.

140

Rescher’s Metaphysics William Jaworski

R

escher’s metaphysics is wide-ranging and innovative. The chapters that follow display some of its major themes. These include a commitment to a metaphysics of experience similar to Kant’s transcendental idealism—a conceptual idealism that is nevertheless compatible with a causal materialism. He also endorses a functionalistic pragmatism which he uses to defend a robust form of metaphysical realism, and which provides him with the resources to articulate a metaphysics of value. The latter provides a basis in turn for a metaphysics of moral value with a biological orientation similar in its outlines to Aristotle’s ethics. It also provides resources for answering what we might call the fundamental question of existence: Why is there something instead of nothing? Rescher endorses an optimalist answer along Leibnizian lines; although unlike Leibniz, his account is not committed to theism: There is something instead of nothing, he says, because that is for the best, and the best can be reckoned in purely naturalistic terms. In what follows I will try to provide some orientation to Rescher’s rich and detailed discussion of these points by sketching its main contours. Rescher’s metaphysical realism comprises two ideas: the idea that there is an objective, mind-independent world, and the idea that we can have knowledge of that world. To support his realism Rescher employs a broadly Kantian strategy. Where Kant appeals to a notion of experience, however, Rescher appeals to a notion of practice. Metaphysical realism, he argues, is a precondition for the existence of cognitive practices such as inquiry, communication, and evaluation. There can be such a thing as empirical inquiry, for instance, only if there is a mind-independent reality that cooperates in certain ways with our inductive practices. We can make sense of our scientific practice as a process that yields knowledge of the world, for instance, only if we presuppose that there is a world, and that certain “ampliative” modes of inference are legitimate: that subjective phenomena give us some (albeit defeasible) information about objective realities, that particular cases give us some (albeit incomplete) information about universal conditions, etc. Scientific practice is made possible by a commitment to metaphysical realism. The same is true, Rescher argues, about intersubjective communication, evaluation, and all the other cogni-

RESCHER STUDIES

tive practices that depend on the distinction between truth and falsity, the notion of truth as agreement with reality, the distinction between reality and appearance, and the possibility of error. All of these are made possible by a tacit endorsement of metaphysical realism. According to Rescher, when we examine the ways metaphysical realism is implicated in our cognitive practices, it becomes apparent that it is not something we discover, but something we presuppose and that we must presuppose if we are to engage in the cognitive and communicative practices we do. Moreover, Rescher argues, we have no choice but to engage in such practices. Creatures like us have no other modus operandi and can live no life that holds more promise than a life that seeks good reasons for thinking and acting in certain ways, and that means a life that involves cognitive practices. Engaging in rational inquiry, communication, and other cognitive practices is something that is for us unavoidable, and for that reason in the absence of something better we have no choice but to embrace metaphysical realism. Rescher’s approach to metaphysical realism stands opposed to a familiar type of skepticism. Very roughly the skeptic assumes that belief in an objective world can be warranted only by appeal to rational inquiry, and then argues on broadly Humean grounds that rational inquiry cannot warrant this belief because it already presupposes it. Rescher is free to agree that rational inquiry cannot warrant its own starting points in the way it warrants, say, the acceptance of empirical results by means of induction. But he rejects the idea that the methods of rational inquiry are as limited in scope and variety as the skeptic conceives them to be. When it comes to metaphysical realism, something that is, to use Kant’s expression, a condition for the very possibility of engaging in the cognitive practices we do (including, importantly, the practice of formulating skeptical arguments), our endorsement of it is ultimately warranted not by some process internal to the inquiry it enables—at least not by scientific induction or similar process of the sort the skeptic has in mind. Our endorsement of it is instead warranted by the success of the practices it makes possible. According to Rescher, metaphysical realism not only enables us to engage in cognitive practices, it is ultimately vindicated by their results. The results in question are both practical and theoretical. On the practical side we develop a picture of the world that enables us to operate effectively in it: we might think here of the technological advances that have accompanied the growth of scientific knowledge. On the theoretical side the picture we develop enables us to predict and control natural occurrences, and also

142

RESCHER’S METAPHYSICS

enables us to explain how creatures like us with our particular cognitive resources and modus operandi would be able to develop a picture of the world and our place in it that was more or less accurate. Metaphysical realism is thus warranted in a twofold way: first, it enables rational inquiry; second, it is justified by its success. Rescher refers to this second point as retrojustification. The success of an inquiry based on certain assumptions enables us to assert with hindsight that those assumptions were the right ones to make. Hence, the success of our cognitive endeavors—of science, of communication, and so forth—vindicates the realist presupposition. Rescher’s approach to metaphysical realism is an expression of his more general commitment to pragmatic criteria of justification and what he calls ‘functionalistic pragmatism’. According to Rescher, we are purposive animals. We do things in order to achieve certain ends: We have laws to establish and endorse rules of conduct; we construct ethical systems to guide behavior in ways that are conducive to living well; we engage in inquiry to remove doubt, and so on. Theorizing is ultimately something we do, a process that aims among other things at answering various questions we have. As a result, we can evaluate theorizing according to the same sort of criteria we use to evaluate other goal-directed processes: the criteria of efficacy and efficiency: whether something achieves its goals and whether (all things being equal) it does so in the most efficient and economical way. According to Rescher these criteria can be brought to bear on the evaluation on an enormous variety of procedures and processes: theoretical, legal, and moral ones included. Functionalistic pragmatism fits hand-in-glove with a theory of rationality since rational beings will prefer (all things being equal) those methods and procedures which are most effective at achieving their ends and which achieve those ends in the most efficient ways. This connection with rationality might suggest a stereotype of pragmatism as a crass utilitarian doctrine that sees reason as a mere tool for calculating means to ends but that pays no heed to the evaluation of ends themselves. Hume endorsed a theory of rationality of this sort. Reason, he said, was the slave of the passions. The latter furnish us with ends, and to those ends reason merely reckons the means. Consequently, Hume claimed, “It is not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.”1But Rescher argues against this conception of rationality and articulates an alternative. Reason, he says, has two aspects: an instrumental aspect and an axiological one. The former aspect concerns the calculation of means to ends, as Hume thought, but the latter concerns the evaluation

143

RESCHER STUDIES

of ends themselves. Clearly if we adopt inappropriate ends; if, for instance, we pursue an end in a way that outstrips its true worth, we cannot lay claim to being fully rational. In addition, Rescher reminds us, ends often concern values, and values are not tastes. Tastes do not admit of rational evaluation; they are unreasoned preferences. Values on the other hand have an instrumental aspect; they serve to guide and direct behavior. As a result they admit of functionalistic evaluation. Values that direct us to pursue ends that are clearly not in our best interests or that do not satisfy our real needs are not values it is rational for us to adopt. This axiological component of reason provides Rescher with a basis for articulating a metaphysics for moral evaluation. Like Kant’s metaphysics of morals, Rescher’s is based on the requirements of rationality. But unlike Kant’s metaphysics of morals and like Aristotle’s, Rescher’s account has a biological orientation that grounds evaluation in an account of human flourishing. Rescher talks about the notion of what is in our best interests in a twofold way: in terms of the conditions that are necessary for our being, and in terms of the conditions that are necessary for our well-being. Human life involves a manifold of processes, he says. Some of these are chosen; they mark the kinds of behaviors we call ‘wants’, ‘wishes’, ‘desires’, and so on. But some are not chosen; they mark genuine needs, ones determined in large part by our biological heritage. Because we are animals with a particular evolutionary provenance there are certain things we need in order to live such as air, food, and shelter, and there are also certain things we need in order to live well: self-respect, companionship, a sense of belonging, and the like. These needs provide an important part of the basis for assessing what is or is not in our best interests. Like Aristotle, then, and unlike Kant, Rescher’s metaphysics of moral evaluation has a biological orientation that centers on the notion of specifically human agents as opposed to rational agents in general. And again, like Aristotle, Rescher looks to ground evaluation in the requirements enddirected behavior must satisfy if it is to achieve the goal of human flourishing. Rescher’s metaphysics of experience, on the other hand, is very similar to Kant’s. He endorses a type of idealism—what he calls conceptual idealism in contrast to the ontological idealism associated with philosophers like Berkeley. Conceptual idealism is not the claim that everything is mental, or that the mind generates or constitutes everything that exists. It is not a causal claim. It is instead the claim that our experience of reality is minddependent. It claims that the basic categories in terms of which we describe

144

RESCHER’S METAPHYSICS

and explain the physical world involve an implicit reference to mental operations. More precisely, it claims that the basic predicates and terms we use to describe the world express relations, and each of these relations has some aspect of the mind as one of its terms. Consider an analogy. Artifacts cannot be defined apart from the purposive activities in which they are involved, activities which themselves involve mental operations. Books, for instance, cannot be defined apart from activities such as reading and writing—activities involving mental operations. According to Rescher all the concepts we use to understand the world are analogous to the concept of a book; all of them advert in some way to the operations of mind. Consider, for instance, basic metaphysical notions such as the notion of particularity. That notion adverts to the activity of identifying, a mental operation. Consider likewise basic physical notions such as the notion of spacetime. It involves the operation of locating, another mental operation. Mental operations of this sort are built into the very concepts we use to understand the world; they constitute our very idea of physical reality. Conceptual idealism claims, then, not that the mind is a causal source of the world, but that the mind provides part of the basic interpretative mechanisms whereby we experience and understand the world. Conceptual idealism does not deny that reality is mind-independent in the sense that the basic furniture of the universe might remain unchanged if beings with minds had never existed. The world might be mindindependent in a causal sense, Rescher argues, but it is not mindindependent in a conceptual sense. Consider: what exactly would remain unchanged in a world without minds. Would, for instance, roses exist? Consider the various features we attribute to roses, especially the ones that categorize something as a rose—its smell, color, size and shape. The concepts of all of these features advert in some way to mental operations such as identifying, classifying, and measuring. Take away those features, and what is left? Nothing that could be classified as a rose surely, for being a rose depends on having features of the aforementioned sorts. Without those features, ones defined in part by reference to mental operations, there is no longer any basis for categorizing something as a rose. Rescher doesn’t deny that there might be other ways of conceiving the world that are not mind-involving, but he does argue that there is no way of articulating such a conceptual scheme given that our own is mindinvolving. Any act of articulation requires the use of conceptual tools already at our disposal, and the only ones currently at our disposal are mindinvolving. There is no way of articulating a conceptual scheme that tran-

145

RESCHER STUDIES

scends the limits of the scheme we work with in fact, and that means, then, there is no way of articulating a conceptual scheme that is not mindinvolving. Importantly, conceptual idealism is compatible with causal materialism. The idea that our concepts of physical processes depend on mental operations is perfectly compatible with the idea that our mental operations depend causally on physical processes. Conceptual idealism is not an explanatory theory regarding the causal source of the mind, says Rescher; it is instead an analytic or hermeneutic theory regarding the basic categories in terms of which we understand the world. Conceptual idealism is thus officially neutral about physicalism, substance dualism, neutral monism, and other theories constituting philosophy of mind’s stock in trade—theories that are causal in the relevant sense. Rescher also sketches a strategy for answering what I called earlier ‘the fundamental question of existence’: Why is there something instead of nothing? In line with his functionalistic pragmatism, he defends an axiological approach along Leibnizian lines: There is something instead of nothing, he argues, because that is for the best. An answer along these lines appeals to what Rescher calls the Law of Optimality: whatever possibility is for the best is the possibility that is actualized—precisely because it is for the best. Rescher refers to an approach based on this principle as a type of optimalism. Optimalism claims that it is in the nature of things that the best possible alternative is realized. One advantage of optimalism is that the Law of Optimality is selfsubstantiating, a desirable feature in a first principle. A principle that could not be explained except by appeal to some further principle would make a poor candidate for a first principle since it could not lay claim to being foundational in its domain. If asked, however, why the Law of Optimality obtains, the optimalist can appeal to the Law itself: it obtains because that is for the best. This makes the Law of Optimality as good a candidate for being a first principle as any; it ends up being part of the very world order whose existence it explains. Applying the Law of Optimality to the prevailing world order requires something of a promissory note which Rescher describes as a “scientifically reputable” set of value parameters that are in some way physically relevant. With such a set of value parameters in hand, the optimalist can explain the existence of the world using the following line of reasoning: (i) whatever possibility is for the best is the possibility that is realized; (ii) the

146

RESCHER’S METAPHYSICS

prevailing world order is the best that can be realized; therefore, (iii) the prevailing world order exists. Rescher defends optimalism against three competing approaches to the fundamental question of existence. First, he defends it against an approach that rejects the question on the grounds that it is ill-posed. Rescher shows that the arguments for this approach fail. Second, he defends optimalism against an approach that insists on a non-axiological answer to the question, one that appeals, for instance, to a notion of efficient causation. According to Rescher the question needn’t be taken to ask for a productive force or other causal factor. Causal explanation is one type of explanation, but it is not the only type. It is possible to explain the existence of something without identifying a cause that produces it. This is what the optimalist answer does. According to the optimalist, value operates on the range of logical possibilities not as an agent that produces the best alternative, but as a factor that constrains the range of possibilities so that only the best are available to be realized, and only the uniquely best is realized in fact. A constraint of this sort doesn’t require the postulation of a causal agent. By analogy, says Rescher, consider the operation of certain rules for English word formation. English allows double letters but not triple ones. This rule constrains the range of possible English words. It thereby explains why, for instance, ‘pussy’ is an English word but not ‘pusssy’, but it surely doesn’t cause this to be the case in the sense of having produced the double letter in ‘pussy’. To insist, as non-axiologists do, that an answer to the fundamental question of existence invoke a cause is to insist on a model of explanation the optimalist is not obliged to accept. Finally, Rescher defends the optimalist answer against what I will call an axiological supernaturalism which argues that there can be no value without a valuer—that an axiological answer to the fundamental question of existence must appeal to a supernatural source of value and/or purpose. Supernaturalism poses a serious challenge to Rescher’s optimalism because if it is right, the Law of Optimality is not self-substantiating. If value requires a valuer, then the Law of Optimality would have to be explained by appeal to some further principle positing a valuer such as God. Rescher argues, however, that the major premise of the supernaturalist argument is false. It assumes that values are analogous to purposes. A purpose must be someone’s purpose, and hence purposive explanations must appeal to the purposes of an agent. But values can be completely impersonal. They are less like purposes in this sense and more like facts. Something’s being a fact doesn’t depend on someone endorsing it; in the same way, says Re-

147

RESCHER STUDIES

scher, something’s being valuable doesn’t depend on someone actually valuing it. To be valuable is to deserve to be prized, but that doesn’t mean anyone actually prizes it. Hence, according to Rescher, the supernaturalist argument fails: a commitment to axiology does not imply commitment to theology. Rescher’s optimalism does, however, imply a commitment to teleology since it implies that things tend toward optimality. But teleology in general does not require agents with purposes. In a way that is once again reminiscent of Aristotle, Rescher argues that there are natural end-directed processes that do not require conscious agency: homeostasis and heliotropism are just two examples.2 Further, Rescher argues that a naturalistic axiology is superior to the supernaturalist alternative for several reasons. It is, for instance, more economical. The supernaturalist adds a theological component to an account that is capable of standing on its own axiological feet. In addition the supernaturalist view is regressive from the standpoint of the history of science—a history in which God has been invoked with decreasing frequency to do increasingly little explanatory work. Moreover, it seems that supernaturalism inverts the natural order of explanation: it tries to explain what is better understood—the natural world—by appeal to what is not understand at all: God. Rescher’s axiology is not committed to theism, but it is not committed to atheism either. If axiology needn’t be grounded in theology, it remains compatible with theism, and indeed Rescher’s axiology would appear to be perfectly in harmony with theism since it is reasonable to suppose a divine creator would proceed optimalistically. As with the other claims he advances, however, Rescher thinks the ultimate warrant for an axiological metaphysics along the lines he endorses is to be found its practical application—in the way it enables us to systematize our knowledge, in particular. This, he thinks, is something that will depend in the long run on the future progress of science. Rescher’s ideas have any lucidity and candor that is refreshing to read. Methodologically he is engaged in what Strawson calls descriptive metaphysics.3 He is most profound in articulating ideas that are integral to our intuitive pre-philosophical understanding of the world, but that are for various reasons difficult to articulate systematically. Rescher’s work thus strikes one as being highly insightful without being shocking or counterintuitive. It also has a broad, systematic vision. In a profession whose practitioners increasingly operate with a tight-angle shutter and zoom lens, a phi-

148

RESCHER’S METAPHYSICS

losopher with a panoramic view of a broad constellation of issues marks a refreshing deviation from the norm.

NOTES 1

Treatise on Human Nature Book II, Part iii, Section 3.

2

Aristotle argues that not all teleology involves deliberation and choice. His examples include web-building in spiders, nest-building in birds, and leaf-growing in plants (Physics ii.8: 199b20-33).

3

P. F. Strawson, Individuals, An Essay in Descriptive Metaphysics (London: Methuen, 1959), pp. 10-11.

149

The Limits of Science Reconsidered Ulrich Majer

C

ontemplating the extraordinaire success of modern science three ideas come to mind regarding its future development:(i) the unity of science (ii) the completeness of science and (iii) possible limits of science. We feel immediately that these ideas hang closely together, but if we are asked, what are the exact logical relations between the ideas, we will have some difficulties, because the logical dependencies of the three ideas are not immediately visible. For example does the unity of science imply its completeness or inversely, does the completeness of science presuppose its unity? These and similar difficulties have their origin not, at least not primarily, in an imprecise meaning of the three core-concepts “unity”, “completeness”, and “limit”—it should be noticed that all three concepts posses an exact mathematical definition—but in a nonconformity or an incongruence of these concepts to science. We simply do not know for example, what the concept of ‘completeness’ means with respect to science, because we do not know how to apply it science. We have no intelligent idea, despite or just because of its precise mathematical definition, how it should be applied to science. The same is true for the other two concepts ‘unity’ and ‘limit’. Hence, the question arises what shall we do in such a situation? There is, however, no reason for despair! We do, what reasonable persons always do in such a situation: We study different cases, make conceptual and contextual distinctions, and investigate the sources where the difficulties may come from. Finally, we reflect on the adequacy of our traditional concepts and explanations and introduce, if necessary, completely new concepts and theoretical approaches. This is roughly what Nicholas Rescher has done in his skillful book ‘The Limits of Science’, in which he considered the old and intriguing question: Does science have limits—internal or external ones—and what does it mean, if the answer is Yes and No at the same time? Does it mean a contradiction or only a contraposition, a Yes in one and a No in another direction or aspect? My aim in this paper is a twofold yet very moderate one: First I want to spell out what are from my point of view the essential points in Rescher’s discussion of the aims and limits of science and the ideas of an ultimate theory and a perfected science. Second, I will deepen these considerations

RESCHER STUDIES

by including some aspects of modern science that are—or at least seem to me—lacking in Rescher’s discussion of these intricate questions. It was by no means an accident, when I wrote that the answer to the question “Does science have limits?” may be Yes and No at the same time, because that seems to me the overall structure of Rescher’s answer in his effort to come to terms with the main question. This, of course, does not mean a contradiction but only a contraposition, i.e. an affirmation in one and a negation in another direction of the multi-dimensional domain of science. This is a rather cryptic remark and I will explain it in a moment. But first let me point out that there is an interesting parallel between Rescher’s answer and Kant’s approach to the question “does science have limits”. Kant makes in the Prolegomena a distinction between limits [Grenzen] and boundaries [Schranken], which is very helpful in understanding Rescher’s position. He says: “So lange die Erkenntnis der Vernunft gleichartig ist, lassen sich von ihr keine bestimmten Grenzen denken. In der Mathematik und Naturwissenschaft erkennt die menschliche Vernunft zwar Schranken, aber keine Grenzen, d. i. zwar, dass etwas außer ihr liege, wohin sie niemals gelangen kann, aber nicht, dass sie selbst in ihrem innern Fortgange irgendwo vollendet sein werde.“ (A 168; my italics)

In other words, mathematics and natural science have boundaries but no limits, which means two things: First, from an external point of view we recognize that there exist “spheres of knowledge”, beside mathematics and natural science, which both disciplines cannot enter or penetrate into. However, and this is the second point, from within both disciplines we cannot recognize any limits, which means that both disciplines proceed in their research ad infinitum and will never be finished or completed. Now, this reading of Kant may be criticized either as trivial or as unintelligible. Sure, at first glance it seems so and I will respond to this objection in a moment. But before let me point out that my interpretation is at least correct from a ‘literary point of view’. This becomes clear immediately if I quote the next 3 sentences in the same passage: Die Erweiterung der Einsichten in der Mathematik, und die Möglichkeit immer neuer Erfindungen geht ins Unendliche; ebenso die Entdeckung neuer Natureigenschaften, neuer Kräfte und Gesetze, durch fortgesetzte Erfahrung und Vereinigung derselben durch die Vernunft. Aber Schranken sind hier gleichwohl nicht zu verkennen, denn die Mathematik geht nur auf Er-

152

THE LIMITS OF SCIENCE RECONSIDERED

scheinungen, und was nicht ein Gegenstand der sinnlichen Anschauung sein kann, als die Begriffe der Metaphysik und Moral, das liegt ganz außerhalb ihrer Sphäre, und dahin kann sie niemals führen; . . . . Naturwissenschaft wird uns niemals das Innere der Dinge, d. i. dasjenige, was nicht Erscheinung ist, aber doch zum obersten Erklärungsgrunde der Erscheinungen dienen kann, entdecken; aber sie braucht dieses auch nicht zu ihren physischen Erklärungen. (Prolegomena, A 168; my italics)

Evidently Kant maintains two assertions at once: (i) The expansion of the domains of mathematical and scientific knowledge goes to infinity; consequently, the domains of the two disciplines cannot have limits; (ii) nonetheless they have boundaries, because there exist “spheres” of cognition (i.e. metaphysics and morality), which lie outside the domain of sensible intuition and experience, to which mathematics and science are restricted. Taken in itself each assertion seems trivial, i.e. true with a high probability, but taken together they are highly problematic, if not incompatible. That mathematics and natural science have no internal limits is one of the oldest experiences of our occidental culture; and that there are spheres of cognition outside the two disciplines is also hardly deniable. But that this state of the art will remain so forever, i.e. that mathematics and natural science will expand their domains of knowledge unlimitedly and yet will not enter or incorporate in the long run other spheres of cognition is hardly believable and vehemently denied by all those scientists and philosophers, who dream of a universal science comprehending all types of knowledge. In other words, it seems not intelligible to maintain on the one hand an unlimited expansion of mathematics and science and at the same time the existence of certain boundaries that mathematics and natural science cannot overcome. This looks like a ‘contradictio in adjecto’ and Kant’s subtle distinction between limits [Grenzen] and boundaries [Schranken] cannot really conceal the contradiction. So far so good, but the whole objection rests on an unproven presupposition, which is patently wrong, namely the assumption that ‘unlimited’ means the same as ‘infinite’ and that consequently ‘unlimited’ and ‘finite’ denote contradictory properties. But this is not the case as every model of non Archimidean Geometry shows, in which one can adjunct a line segment CD unlimitedly (starting from point A on a given straight line AB and moving always in the same direction B) without exceeding point B on the straight line in question. This means that the notions ‘unlimited’ and ‘finite’ do not denote contradictory properties. And therefore we can maintain without a whiff of a contradiction that our physical space is in fact

153

RESCHER STUDIES

unlimited and finite at the same time, as the theory of general relativity demands it for certain models. Unfortunately, Kant himself was convinced that the last assertion entailed a contradiction, as it indeed does in Euclidean geometry because in this geometry the axiom of Archimedes is valid. However, because Kant was equally convinced that the development of mathematics and science embodied a certain non-Archimedian structure representing both properties in tandem he tried to find ‘a way out’ of the apparent contradiction. His solution was the sublime distinction between ‘Grenze’ and ‘Schranke’: The expansion of cognition, the discovery of new facts, their unification by reason, as well as the deepening of understanding, have no internal limits, neither in mathematics nor in science, but both disciplines are bound to certain areas of cognition, namely logic and intuition—pure and empirical -, which both cannot transcend without becoming entangled in contradictions and hence losing all meaning. After the elimination of the suspicion that Kant’s claim regarding the double character of science could entail a contradiction I will return to Rescher’s book and explain why also Rescher—like Kant before him—has a kind of non-Archimedean structure in mind, when he discusses the question of the limits of science. Let me begin with the very last chapter 15, which bears the title: “The Limited Province of Natural Science”. Unfortunately, Rescher makes no terminological distinction between limits and boundaries, but that he has a kind of restriction of natural science to certain ‘provinces’ of cognition in mind cannot be mistaken, because this is roughly that what the title of the last chapter announces. Hence, we have to consider only two questions in order to see, whether Rescher shares Kant’s non-Archimedean view of science: First, what kind of ‘limitation’ of science has he in mind and, second, in which way is science unlimited? The answer to the first question is not easy, because it presupposes an analysis of the meaning of the expression province of natural science. This analysis is not trivial, because it entails a twofold problem, namely first the demarcation of science from other forms of cognition, and second the task to do this in a non-circular way, which avoids the danger of a ‘petitio principii’ or ‘question begging’ regarding the limitation of natural science to a certain province that is fundamentally different from other domains of cognition. Rescher endeavors to solve this tricky problem in one ingenious stroke. Though, I agree basically with his procedure and the overall solution, I feel a certain difficulty in some of the details of his proposal. Let me

154

THE LIMITS OF SCIENCE RECONSIDERED

first quickly explain what his general strategy is, before I identify the particular difficulty. The leading idea in Rescher’s attempt to demarcate the province of science from other domains of cognition is the proposal first to identify province of science intensionally by means of a distinction between the subjective and the objective moments of scientific cognition, and then to use this distinction in a second step for the extensional demarcation of the proper domain of natural science from other spheres of cognition not belonging to the natural science. Consequently, Rescher’s first aim is the clarification of the distinction between subjective and objective. After some preliminary differentiations he comes to the point: Natural science is a mission-oriented endeavor, [. . .]. It inquires into what sort of things there are in the world and how they work at the level of lawgoverned generality, [. . .]. Given this mission, the concern of science is, and must be, with the public face of things—with their objective facets. It strives for reproducible results, and its focus is on those objective features of things that anybody can discern, [. . .]. Science deliberately sets aside the observer-relative dimension of experience. (p. 243; italics are Rescher’s; spacing is mine)

Having made the distinction between the objective and subjective features of things Rescher sets out to demarcate the proper domain of science from other areas of cognition. He does this in several small steps, beginning with the exclusion of the “person-linked dimension of human cognition” from the domain of natural science: The “facts” to which science addresses itself are [. . .] those that arise from intersubjectively available observation rather than personal sensibility. [. . .] Thus, science ignores the individualized, affective, and person-linked dimension of human cognition: sympathy, empathy, feeling, insight, and personal reaction. (ibid. p. 244)

From here he strives—taking some intermediary steps like the exclusion of “the value dimension of experience” from the proper domain of science— to the final conclusion: Accordingly, science omits from its register of facts [that are worth taking at face value] the whole affective, emotional, feeling-oriented side of our cogni-

155

RESCHER STUDIES

tive life. [. . .] The sort of observation-transcending experience that lies at the basis of norms and values remains outside its range. (ibid. p. 244)

Now, having reached this point, it is easy to see that the province of science is limited in the sense of Kant’s concept of “boundary “ (Schranke) that science cannot enter certain spheres of human cognition, because these lie outside its concern with indubitable objective facts; this is what Rescher says too, although he does not use Kant’s term: Thus, no matter how far we push science forward along the physical, chemical, biological, and psychological fronts, there are issues about humanity and its works that will remain intractable by scientific means—not because science is impotent within its [own] range but because these issues lie outside it.

Of course, Rescher does not stop here but draws further conclusions regarding the inability of science to explain the affective, evaluative and emotive dimension of human cognition: “Thanks to its exclusion of normative and evaluative issues, science approaches people as objects of study, as things, and not as persons. [. . .] cognitive modes as sympathy and empathy are put aside.” (ibid. p. 245) Although these considerations are very interesting, I will bypass them and instead return to the main argument regarding the non-Archimedean structure of science. We have seen that Rescher agrees to Kant’s view that science has boundaries; we still lack a demonstration that he also accepts the other part of Kant’s claim that science has no (internal) limits. But before entering this task let me first explain why I see a certain difficulty in Rescher’s general procedure to identify the boundaries of science by means of a distinction between the subjective and the objective parts of science. Let me first stress that his general strategy is basically quite convincing, but “the devil sits in the detail” as we say in German; this means in the concrete case: The characterization of the “objective features of things” as those that “anybody can discern” in distinction to the “subjective” aspects as the “observer-relative dimension of experience” is in general quite right but, I suspect, too imprecise if we come to a concrete situation and ask what does the distinction mean in this case. Lets take a stellar observation, for example the distance of the Venus to the sun at a certain time. Every earthly observer will observe approximately the same distance, although his observation is obviously observer-relative, because it depends on his position on the earth instead on the moon.

156

THE LIMITS OF SCIENCE RECONSIDERED

This very simple example already shows that inter-subjectivity and observer-relativity are not mutually exclusive but occur together in most practical cases. Physicists have known this for a long time—at least since Galileo—and have formulated their theories accordingly, which means two things: First, they represent all processes and events in a “frame of reference”. Second, they formulate a principle of invariance stating which frames are physically equivalent. (In classical mechanics these are the inertial-systems, which are invariant under Galileo-transformations.) Returning to Rescher’s distinction one should notice that in most cases the frame of reference is also the frame of the observer in an objective sense. This means that observer-relativity and objectivity (in the sense of inter-subjectivity) can go hand in hand. Consequently the distinction between subjective and objective features must be drawn differently. But the question is: How? Rescher’s answer I can imagine. He will reply that he had something different in mind than just the opposition between inter-subjectivity and observer-relativity—something more impersonal on the objective, and something more private or idiosyncratic on the subjective side. But the question remains: Which distinction? What kind of distinction has he in mind? But notice, Rescher cannot simply reply: “the whole affective, emotional, feeling-oriented side of our cognitive life” belongs to the subjective side. Such an enumeration would be question begging, because it is an open question, whether these psychological phenomena can be integrated eventually into the natural sciences or not. He must come up with a genuine distinction that allows to determine the extension of science by a characterization of its objectivity in distinction to mere anthropomorphic aspects and effects. But this is extremely difficult, because these two aspects are very closely intertwined in modern theories like general relativity and quantum mechanics. I’ll come back to this point later. For the moment it suffices to remind the reader that the question whether the lengthcontraction is a real (objective) or merely an apparent effect of the observer moving in another frame of reference than the rigid body, whose length shall be determined, is still not settled—not to speak of other questions in statistical mechanics and quantum theory, which are much more intricate. Coming back to Rescher’s presumed ‘non-Archimedean’ view of science we have still to show that he shares Kant’s view that science has no internal ‘limits’ in the sense that it can be expanded infinitely without touching any limit (Grenzen). At first glance this task seems easy, because he

157

RESCHER STUDIES

summarizes his point of view at the very end of his book in four “significant conclusions”, of which the first affirms explicitly: Natural science has no limits (in the strict sense of that term). There is no reason to think, on the basis of general principles, that any issues within the domain of natural science lie beyond its capabilities (chapter 6 and 7).

Consequently, we have only to ask what does the qualification “in the strict sense of that term” mean? A first hint that things are not so easy as one perhaps thinks is given by the subsequent remark regarding the nonexistence of “any issues within the domain of natural science [that] lie beyond its capabilities”. I suppose this is not intended as an analytical assertion (as one could read it too) but as a substantial remark pointing to an essential characteristic of natural science. Hence, we have to ask which characteristic has he in mind. The answer, as I understand it, has two parts: a negative and a positive one. Let’s take the negative first. If one supposes for a moment that it is true that natural science (like mathematics) has no inner limits in Kant’s sense (Grenzen) then it is quite natural and reasonable to ask: Why is this the case? What is the reason for this remarkable meta-scientific fact? And this question seems to have a plain and simple answer: Because nature itself is infinite, or more precisely, because the domain of natural science itself is intrinsically infinite. But Rescher rejects this answer right away, because it is neither necessary nor helpful. On the contrary, it disguises the true reason for the limitlessness of science, as we will see in a moment. But first let me quote his rejection of the simple answer: A “working hypothesis” of the structural infinitude of nature is not needed to assure nonterminating progress in science. [some pages later he summarizes even more explicitly] . . . an assumption of the quantitative infinity of the physical extent of the natural universe or of the qualitative infinity of its structural complexity is simply not required to provide for the prospects of ongoing scientific progress. (pp. 81 and 84; italics in the original; spacing is mine)

To this I can only agree. It is a silly misunderstanding of the nonArchimedean structure of science to think the limitlessness of natural science rests in an intrinsic infinity of its domain. Finiteness of space in the large as in the small is completely compatible with its un-limitedness, as David Hilbert has shown almost a century ago. And the same holds true

158

THE LIMITS OF SCIENCE RECONSIDERED

not only for physical space but also for the human recognition of nature. Our scientific knowledge of nature is certainly finite—in any reasonable respect I can think of—but nonetheless unlimited in the sense of a potential infinite expandability. But why is that the case? What justifies this optimism? This brings me to the positive part of Rescher’s answer. This, however, is more difficult to ‘pin down’, and I am not quite sure whether I got to the right point. The core of the positive answer seems to me to comprise two complementary, equally important points: First, the reason for the limitlessness of science does not rest on the side of nature, at least not in the first line, but on the side of the explorer! And second, the limitlessness of science is closely tied up with the intrinsic mathematical character of natural science. It is a deep-rooted fact about man and “man made” science that we can explore nature in a rational way (satisfying the demands of reason) only by applying mathematics to the “phenomena” and their symbolic representation. This has been so since the very beginning of science in Greece (take for example Plato’s Cosmology in the Timaeus) and continues to the present day in which the mastership of mathematics is an indispensable tool for a deeper exploration of nature and the further development of science. Before I explain why both points belong inseparably together like the two sides of one coin let me first quote some remarks that show that both points are in fact Rescher’s central arguments for the limitlessness of natural science. Having pointed out that the “infinity-of-nature principle is strictly one-sided, placing the burden of responsibility for the endlessness of science solely on the shoulders of nature itself” (“on the side of its objects”, so to speak) Rescher develops his own positive answer in opposition to this mistaken view: Science, the cognitive exploration of [the ways of] the world, is a matter of the interaction of the mind with nature—of the mind’s cognitive exploitation of the data to which it gains access in order to penetrate the secrets of nature. The crucial fact is that scientific progress hinges not just on the structure of nature itself but also on the structure of the information acquiring processes by which we investigate it. (p. 81; italics in the original; spacing is mine)

The idea that the limitlessness of science is in the first line a matter of the interaction of mind with nature is further developed in the consideration “that the complexity of endless hierarchy levels need not enter in through the structural makeup of nature itself but merely in the conceptual apparatus being deployed to study it at greater depths of sophistication.” (p.

159

RESCHER STUDIES

82; spacing is mine). Hence, the circumstance that we explore nature by means of a conceptual apparatus is the decisive reason for the endlessness of science. In this context Rescher points to the similar situation in geometry: The case is similar to that of the geometer, whose hierarchy of definitions [. . .] does not reside in his materials as such but in the conceptual taxonomy that he chooses, for cognitive reasons, to impose on these materials. Nature herself has no “depth”. For depth is an inherently relative matter, generated through the operation of a cognitive perspective. The endless levels at issue will not be physical levels but levels of consideration inherent in the activities of inquiring beings. Complexity, after all, lies less in the objects than in the eyes of their beholder. . . . Thus responsibility for the open-endedness of science need not lie on the side of nature at all but can rest one-sidedly with us, its explorers. (p.82/83; italics original; spacing mine)

The mere similarity between science and geometry regarding the cognitive reasons for the limitlessness is turned into a necessary relation a few lines later when mathematics is identified as an indispensable part of science whose prospects cannot be limited to a finite domain: Who is prepared to say that the realm of pure mathematics . . . . represents a finite domain? If mathematics is incomplete, so is a natural science that relies in its exploitation for the description of nature. If new mathematics makes new laws of nature accessible to our cognitive grasp, and if the prospect of new mathematics is literally endless, then so, clearly, is the prospect of new science. (p. 83)

Although nobody can say [myself included] why nature follows or obeys mathematical laws, I take the quotation as a reduction of the limitlessness of natural science to that of mathematics so that either none or both are limitless in the sense of Kant. And this seems to me the core of the matter: Rescher shares Kant’s view of the non-Archimedean character of mathematics and natural science in the meta-theoretical sense that both can be expanded unlimitedly and yet have boundaries that they cannot overstep. That my interpretation of Rescher’s viewpoint is basically correct can be inferred from a detail that seems rather unimportant at first view but, in the end, turns out to be quite essential for his position. What I have in mind is this: If we consider the history of science, particularly that of modern physics, one can get the impression that the progress of science becomes less

160

THE LIMITS OF SCIENCE RECONSIDERED

and less significant, because the changes of the observational consequences from one theory to next become minor and minor. Take for example the sequence of astronomical theories from Copernicus to Kepler, then to Newton and last not least to Einstein. Copernicus initiates the heliocentric worldview taking circles as models for the motions of the planets; Kepler improves this by taking ellipses instead of circles; Newton again corrects Kepler’s model taking the interference into account that each planet exercises on the motions of the others; Einstein finally tops this by introducing a relativistic correction for the observed ‘anomalous perihelion’ of Mercury. Now, everybody [acquainted with the history of modern astronomy] knows that these changes become smaller and smaller, measured quantitatively in meters and seconds,1 but it would be a heavy mistake to infer from this circumstance that the corresponding changes in the theories become less and less significant. Almost the inverse is the case because the changes in the subsequent theories have ontological consequences in other unexpected respects that go far beyond the sheer quantitative aspects that initiated the changes. Take for example the first step from Copernicus to Kepler: The substitution of circles by ellipses was connected with the idea of forces that explain the motions of the planets and thus with an assimilation of the supra- with the sub-lunar sphere of the universe, or, to express it differently, with an assimilation of mathematical astronomy with terrestrial physics—thus paving the way for Galileo and, of course, Newton.2 The incongruence between the quantitative improvements on the one hand and the objective significance of the theoretical innovations on the other is even more obvious in the next steps from Kepler to Newton and, of course, from Newton to Einstein. Coming back to Rescher’s view, the interesting point in face of the historical example is now that he puts forward a very similar argument— although less concrete and more abstract—in favor of the incongruence between the quantitative improvements and the objective significance of the theoretical innovations regarding the progress in science. Supposing a natural system with a parameter p, whose quantitative determination needs different amounts of time in different historical situations, Rescher argues as follows: [Even if the system is very simple] endless cognitive progress is nevertheless available. For as our capacity to make p-determinations down to smaller and smaller time intervals increases from minutes to seconds to milliseconds to microsecond, and so on, we will obtain an increasingly comprehensive in-

161

RESCHER STUDIES

sight into the modus operandi of the system and can obtain even fuller information about it. There is no reason to think that the discoveries being made as one moves further down the line are of lesser significance than those achieved early on. The discoveries made at each iteration are alike major innovations. […] even a simple system can afford the prospect of an ongoing discovery of facts whose overall significance does not decrease with the succession of stages. (p.81/82; italics original; spacing mine)

I understand this somewhat artificial example as an argument for the logical independence of the significance of a discovery (and its subsequent theoretical representation) from the quantitative scale of the relevant parameters p, q, . . . . r, at which the discovery was made: Discoveries at smaller scales can have, and often do have the same or even greater significance than those at larger scales. In short, the objective significance of a discovery (and its theoretical representation) is scale-independent! This is not only in agreement with the outlined example of astronomy, but also in concordance with the non-Archimedean character of natural science because in non-Archimedean-geometry no rigid measure, no rigid yardstick for the distance (of points) exists. Consequently, in non-Archimedean geometry no universal scale is available to measure the distance between points. The analogous is true in science with respect to the significance of a discovery or a theoretical innovation. The quantitative improvement of a prediction is no measure for the significance of the corresponding theoretical innovation. In other words Rescher’s insistence on the logical independence of the significance of a discovery or theoretical innovation from any quantitative scale of improvement is in complete harmony with his non-Archimedean conception of natural science as an unlimited—yet bounded—process of experimental discoveries and theoretical innovations. Both, the non-Archimedean character and the ‘scale-independence’ of the significance of theoretical innovations are the ‘two sides of one and the same coin’! This interpretation, I think, will be vindicated by the following sequence of quotations: (I have concentrated on the “Kernsätze” and omitted all intermediary considerations.) Why should the successive “layers” of physical theory be seen as increasingly less important, less interesting, or less valuable? [ ] The fact is that progress in natural science carries us through a series of transformations and revolutions whose overall significance is essentially uniform. [ ] Neither theoretical issues of general principle nor the actualities of historical experience suggest that scientific progress need ever come to a stop. [ ]

162

THE LIMITS OF SCIENCE RECONSIDERED

A very small-scale effect, . . . can force a far-reaching revision and have profound impact by way of major theoretical revisions. (Think of special relativity in relation to ether-drift experimentation, or general relativity in relation to the perihelion of Mercury.) (pp. 84—86; italics original; spacing mine) It seems obvious from the quotation that Rescher’s view of the natural science bears a certain similarity to Kuhn’s conception of science and in particular to his well-known conception of scientific revolution. But there is to my mind—beside the similarity—an important difference between the two positions that is worth to spell out; let me begin with the similarity: Both agree that on the quantitative level an increasing precision of observation and measurement can be recorded during the recent history, particularly during the 19-th and 20-th century. And both agree too that the increasing precision can be taken as a “sign of scientific progress” on the practical level of prediction and experimentation. And last but not least both agree that the fact of a growing precision of observation and measurement cannot be extrapolated to the theoretical level and be regarded as a yardstick of theoretical progress. The reason for this impossibility is, however, very different for both authors: Whereas Kuhn argues that theories—separated by a scientific revolution—are incommensurable because they denote different domains of objects and, consequently, cannot be compared logically with respect to a notion of objective progress in the sense that the later theory entails the former, Rescher does not make such a claim. He only insists that the growing precision is no yardstick for the significance of a theoretical change or revolution, neither in the negative nor in the positive direction. Whether there exist “other means” to judge the significance of a theoretical development (and, hence, its meaning for the objective progress in science) he leaves more or less open. He only stresses that there is no limit for further progress of science on the theoretical level. If I am asked, which of the two positions seems more reasonable, my answer is somewhat swaying. I was always convinced that Kuhn launched an important logical point with his thesis of incommensurability. The point is to my mind the following: Theories separated by a scientific revolution usually denote domains of objects with disjunctive elements; the one domain is not a subset of the other and vice verse. To compare such theories is difficult because we have not the appropriate ‘meta-logical’ (or modeltheoretic) means to analyze the logical relations between such theories. On the other hand, if we had such means Kuhn’s incommensurability thesis

163

RESCHER STUDIES

would be pointless. The case is different with Rescher: His objection against precision as a quantitative measure for the significance of a theoretical development and the progress of science does not rest on a particular logical relation among revolutionary separated theories but on the nonArchimedean-structure of science in general. And this seems to me the saver and more sustaining, hence, the more reasonable position. I will not close my analysis of Rescher’s Limits of Science without a discussion of the two main objections against his Kantian view of science. Both objections try to undermine respectively one of the two pillars, on which the non-Archimedean character of science is founded. The first objection is directed against the existence of boundaries (Schranken) that science cannot overcome. The second objection supposes—contrary to the first—the existence of certain limits that science cannot surmount. To be sure, part of the objections is created by terminological confusion, calling ‘limits’ boundaries and vice verse; these can, of course, be disregarded. There are, however, also certain objections that don’t rest on terminological confusion, but instead either maintain the existence of limits (in the right sense) or deny the existence of genuine boundaries. Let me take the second type of objections first, because these can be easily refuted as shallow or empty. The principal form of these objections is always the same: They deny the existence of boundaries (in the sense that there are spheres of cognition and knowledge that science cannot enter) by supposing a kind of universal reductionism that can be applied ‘in principle’ to every sphere of cognition no matter what the logical and epistemological status of that sphere is: whether it belongs to the domain of sense and intuition, or feelings and emotions, not to speak of the domains of ethics and aesthetics. ‘In principle’ here means that the reduction of a certain sphere of phenomena to science, especially to physics is not really carried through but merely promised for the future. What this means let me explain by a historical example taken from mathematics. Although I share David Hilbert’s optimism repudiating Du-BoisReymond’s “Ignorabimus” I also know (like Hilbert himself) that the reduction of one theory to another is an extremely difficult and tough business, whose final result cannot be predicted in advance—before it has been established! Hilbert’s program “to prove the consistency of Peanoarithmetic within PA itself” is the best example for what has just been said. He hoped in the twentieth to be able to reduce analysis to arithmetic by means of consistency proofs, and yet had to learn—despite remarkable

164

THE LIMITS OF SCIENCE RECONSIDERED

success in the beginning—that this is impossible—at least that it cannot be carried through in the way he had hoped. Now, the moral of the example is this: What can happen in pure mathematics is even more probable in the domain of physics itself, let alone in cases of theories belonging to totally different domains. In other words, the burden of proof that a reduction is indeed possible is on the side of those who maintain it; mere reductionism ‘in principle’ is only a case of wishthinking without any warranty regarding its outcome. Therefore reductionism in principle—without a demonstration that it is indeed possible to reduce a certain sphere of cognition as ‘seeing of colors’ or ‘feeling of emotions’ to physics—cannot be taken as a serious objection against the existence of boundaries and will not be discussed further. Consequently, the only objection that remains is the assertion that genuine limits of scientific cognition exist within science itself! This objection seems to me better justified than the former—at least intermittently although I don not agree with it in the very end! If we review the development of physics during the last hundred years or so we recognize a very remarkable and unique type of theory-change that never occurred before. What I have in mind is, of course, the transition from classical mechanics to the theory of relativity and the establishment of the modern theory of quantum-phenomena. What makes this development, this twofold change in the foundations of physics so unique and unprecedented? The answer is closely connected with the question of the limits of science: For the first time in the history of physics (natural philosophy) the “observer” became an integrated part in a theory of a domain of physical objects submitted to the same laws as the other objects in the “world”. This happened first in the special theory of relativity and then, twenty years later, in the modern theory of quantum phenomena. The expressions ‘observer’ and ‘world’ must here be taken in a peculiar way: “Observer” does not mean necessarily a human individual but a measuring device that obeys the same laws as the other objects in the world, and “world” means a domain of special objects that obey the laws of the theory in question (to which the observers belong). But what is so special in relativity that we call it ‘special theory of relativity’? Has the relativity of rest and motion not always been part of natural philosophy—at least since Galileo? The answer is, of course, YES. But the fact that there exists no motion with a velocity faster than the velocity of light was no constituent part of classical mechanics! Einstein first incorporated this fact into mechanics (although the fact itself was known for a long time). And

165

RESCHER STUDIES

this made fundamental revisions in our conceptions of space and time necessary as Minkowski has pointed out: The Galilean theorem of the addition of velocities had to be substituted by a new law for the “addition” of velocities reflecting the fact that the velocity of light is the upper limit of the possible velocity of a moving object—including any physical observer independent whether it is a human being or a measuring device. Now we see why the incorporation of a “universal constant” into a fundamental theory of nature—like the velocity of light in the theory of special relativity—leads to certain limits of cognition—limits that cannot be violated except by sake of contradiction. The most important aspect of the limitation of velocity in the theory of special relativity is the circumstance that we have no means at all to detect whether an object is at rest or in uniform-linear motion. But is this not the same situation as in classical mechanics? Yes, but with an important difference: In classical mechanics, where the velocity of objects is not axiomatically limited, we could detect by means of experiments with objects “faster than light” (like the instantaneous transmission of gravitation) whether an object is at rest or in uniform-linear motion. But this is physically impossible since the theory of relativity has been accepted as empirically appropriate, because it denies the existence of objects “faster than light” and, hence, the possibility of such experiments. This was one of the fundamental insights: it made the search for such experiments and observers futile. The epistemological situation in modern-quantum theory is rather similar with respect of the “integration of the observer” into the theory— although much more intricate and a lot more subtle! Here it is Planck’s constant in form of Heisenberg’s uncertainty relation h ≤ ∆q • ∆p that determines the limits of precision in a combined measurement of the place and impulse of a “quantum-object”: the two quantities together cannot be measured more precisely than h/2 m; h being Planck’s constant or ‘Wirkungsquantum’ as the German word names the universal constant introduced by Planck more appropriately. This seems easy enough but in fact it is a great mystery that has to be revealed in the future. The most prominent (although somehow unsatisfactory) ‘solution’ of the riddle is the notorious Copenhagen-interpretation. It supposes that the observer has in the act observation—or better in the measurement-process—an irreducible (disturbing) effect on the quantum-object that leads to the uncertainty with respect to the place and impulse of the quantum-object after the measurement has been finished—respectively the observer has been “cut off”.

166

THE LIMITS OF SCIENCE RECONSIDERED

I will here not discuss the Copenhagen-interpretation nor defend it against other, more radical proposals like the theories of ‘hiddenvariables’. For the moderate aims of this paper it suffices to note that the limits of precision as stated in the uncertainty relation cannot be superseded by an omniscient observer nor outwitted by an ingenious experiment, the fantasies of some philosophers notwithstanding. Quantum theory is the best-corroborated theory we ever had; it withstood until today all experiments to falsify it—even the ingenious tests on the basis of Einstein’s proposal and Bell’s inequalities. Consequently we are confronted with the question whether the uncertainty-relation, as well as the velocity of light, establishes ‘limits of science’ in the required sense, or are just contingent ‘limits of cognition’ that may be eliminated by a further unanticipated scientific revolution and, therefore, do not stand in conflict with the non-Archimedean character of science? Before I answer this question one point should be clear: Neither myself nor anybody else can look far enough into the future to exclude the possibility of such a scientific revolution; that is by the way the reason why scientific revolutions cannot be anticipated. In this way Rescher can defend his position (that science has no internal limits) ad infinitum. But this seems to me no honest answer because it trivializes the concept of limit in way that separates it from the very idea of limit as it is used and presupposed in non-Archimedean-structures.3 I hope I can make a better proposal. It is my conviction that the ‘limits of cognition’—as they became an essential element of our modern worldview in form of the theory of relativity and the uncertainty relation—will not be erased but, on the contrary, somehow conserved by the future development of science because any future theory—whether revolutionary or not—must be in accordance with those experiences that have given rise to the new worldview in the first place. But if this is indeed the case, if these limits will not disappear, then Kant’s non-Archimedean-view of science seems to be refuted. But this ‘conclusion’ seems to me too rash. I agree there must be something wrong with Kant’s view, insofar as some limits in science exist that cannot be overcome because they rest on natural constants, but this does not imply that Kant is completely wrong. Natural science in totality could still possess a non-Archimedean character despite the fact that in some respects limits exist. Science has many dimensions that are not internally limited by natural constants. Take the energy of a particle or the structural complexity of a system of molecules—to give only two examples. And for this reason, I think, Rescher is on the ‘right track’ if he insists that natural science has in

167

RESCHER STUDIES

principle a non-Archimedean character, although we know what Kant could not anticipate that it has in two respects well defined limits.

NOTES 1

Indeed the quantitative differences between the observed ellipses and the Copernican circles becomes so small, that one cannot draw them with pencil on paper— without grotesque exaggerations—because the lines drawn by the pencil are in the reduced scale much too broad to keep circles and ellipses apart.

2

See Alexandre Koyré’s From the closed World to the infinite Universe, John Hopkins Press 1957 and his Études galiléennes, Paris 1939 and in particular his wonderful essay “Attitudes esthétique et pensée scientifique” Critique, Sept.- Okt. 1955

3

The operation of extension (i.e. the adjunction of a line segment) must be uniform—not arbitrary—in order to speak of a non-Archimedean structure with respect to a certain operation or extension.

168

Conceivability Diego Marconi 1. RESCHER AND POSSIBILITY

A

t least since his book on Leibniz’s philosophy (1967), and all the way to the very recent Conditionals (2007) Nicholas Rescher has been deeply interested in the notion of possibility and the philosophical issues it originates. He has written extensively on both the logic and the metaphysics of possibility, the semantics of counterfactuals, imagination and its artifacts such as non-existent entities. On such issues, Rescher’s views have often been at odds with widespread philosophical opinions and fashions. He has strongly criticized naïve uses of the notion of a possible world, showing in great detail that there is no possible world that is “just like the actual world, except that ~p” (whereas, in the actual world, p is the case); for a world that differed from the actual world with respect to p would be bound to differ in indeterminately many other respects (C, Ch.7; ImI, p.169). More radically, he has attacked the very notion of a possible world as a basis for the semantics of modal discourse, arguing that possible world semantics “is built upon a foundation of sand” (ImI, p.189). In the actual world’s case, he notices, it is in order to distinguish between objectively existing things and states of affairs and our descriptions or thoughts of such things and states of affairs (TP, pp.196-197). When we say that our knowledge, or our description of the world is incomplete (and bound to remain such), we are presupposing that distinction. But with alleged possible worlds, no analogous distinction is viable: (merely) possible worlds have no ontological autonomy. Possible worlds are not there for us to explore or investigate: they are hypothesized by way of linguistic description or imagination. However, whatever we can imagine or fix by linguistic description “will never be something that deserves to be called a world” (ImI, p.173), for “whatever has an authentically world-like constitution . . . has a complexity that cannot be completely described” (ImI, p.174; see TP, p.208). A genuine world—such as the actual world—exceeds our descriptive capacities: but with possible worlds, there is nothing there but our descriptions. It is not that possible worlds are inaccessible— this presupposes that they are, somehow, there, though epistemically secluded from us. Rescher’s point is, rather, that whenever—as so often—we claim to be imag-

RESCHER STUDIES

ining or describing a possible world, it is not a possible world we are imagining or describing. Rescher’s point here is that a world is determined to the utmost detail, which complete determination no description of ours can ever approximate. But—it could be retorted—couldn’t we imagine a world as determinate, even though we cannot determinately imagine it? I.e., couldn’t we imagine that it possesses determinate features for each of its determinables, even though we do not (and could not) specify in each case what the determinate feature is?1 Rescher’s response, I suppose, would be that as long as the details have not been fixed (but only imagined to have been fixed) we have failed to describe or imagine one world rather than another: at most, we could be said to have imagined a class of worlds. Even this could be misleading, though, for it could be taken to mean that we managed to imagine not one world but many; whereas, according to Rescher, we haven’t imagined or described any single world. Worlds cannot be imagined.2 Coherently, and no less outrageously, Rescher has been deeply skeptical of standard, purely logico-semantical accounts of counterfactual conditionals. In his view, counterfactuals do not have truth conditions (C p.137) and counterfactual reasoning is not governed by logic alone (ib.). The reason is that no counterfactual is inherently acceptable or unacceptable: a counterfactual’s acceptability depends on epistemic factors, such as which issue is being addressed and (consequently) which beliefs about the relevant situation are regarded as salient (C Ch.8; P p.147). In general, to evaluate a counterfactual we add its antecedent to our beliefs, adjusting them so as to preserve consistency (this step, Rescher says, is “simply a special case of apory resolution”, C p.135);3 we then determine whether the counterfactual’s consequent would be acceptable given the modified beliefs (we “derive” the consequent from the modified beliefs). As is well known, however, there is more than one way in which our beliefs can be adjusted (C p.86); so, principles must be found that would govern the required belief revision while “saving the intuitions”, i.e. validating all and only counterfactual conditionals that look intuitively acceptable. However, intuitions of acceptability vary, depending on the issue which is being raised. Thus, the evalutation of a counterfactual depends on both (a) the issue and the ensuing saliences, and (b) general principles of belief revision. The standard literature on counterfactuals only marginally noted the first factor,4 and reduced the second to just one principle, minimal difference: a counterfactual is true if its consequent holds in all antecendentverifying worlds that are “closest” to the actual world. In Rescher’s terms,

170

CONCEIVABILITY

this would correspond to the consequent being derivable in all belief systems that differ minimally from our previous belief system (besides including the antecedent). It is to Rescher’s credit that he tried to both flesh out and improve upon the vague notion of minimal difference. As he shows with great abundance of examples, it seems to turn out that our intuitions about the acceptability of counterfactuals can be accounted for by assuming that belief revision respects a hierarchy of “systemic fundamentality”. Beliefs are sacrificed in the following order (C pp.109-110; see P pp.142150): 1. Item-specific beliefs (“That creature is an elephant”), 2. Nomic beliefs (“Elephants are grey”), 3. General existential beliefs (“Elephants exist”), 4. Definitional/taxonomic beliefs (“Elephants are vertebrates”). More fundamental beliefs are less easily given up than less fundamental beliefs. If the hierarchy is borne out by our actual intuitions of counterfactual acceptability (Rescher himself points out some complications, C pp.116, 125-129), then we could say that the notion of “closeness” or “similarity” among belief systems (or worlds) has been given a relatively precise meaning.5 While objecting to many received views in the area of modality, Rescher has been careful never to throw away the baby with the bathwater. We saw that, in his theory of counterfactuals, he preserved one important intuition that has been basic to much standard work in the field, i.e. the idea that counterfactual evaluation depends on selecting antecedentverifying situations that are both consistent and minimally different from the situation that is regarded as obtaining. Similarly, while rejecting the metaphysics of possible words Rescher wants to countenance both modal discourse and its semantics, which he has proposed to re-write in terms of scenarios (a now popular term), i.e. “oversimplified conceptual thought artifacts answering to thoroughly incomplete world descriptions” (ImI, p.190). Rescher seems to think that such a “linguistic” revision of the model structures of modal discourse would not involve the rejection of de re modalities (see ImI, §8.5), a point that would deserve scrutiny. One also wonders whether the proposed revision in terms of scenarios could account

171

RESCHER STUDIES

for the variety of modal systems, and correlative notions of possibility and necessity, that have sprung from possible world semantics, or whether the revised system would be confined to just one concept of modality (essentially, S5-modality). Finally, one wonders whether the proposed revision, once spelled out in full technical detail, would share the relative simplicity and undeniable beauty of Kripkean semantics. 2. RESCHER ON CONNECTION

THE

CONCEIVABILITY-POSSIBILITY

None of these issues will be taken up in what follows. I will limit myself to a single, basic problem within Rescher’s vast philosophical work on modality: namely, how is possibility itself to be conceived, particularly in relation with such (distinct?) notions as conceivability, describability, imaginability, and more. There are two reasons for this choice: first, the relation of possibility to conceivability has turned out to be an extremely controversial issue, of wide-ranging implications for many important areas of philosophy, including the philosophy of language and the philosophy of mind in addition to metaphysics. Secondly, it seems to me that Rescher’s discussion of the issue provides a useful background for a re-examination of the conceivability/possibility debate, for it essentially restates the “classical” view while dealing with crucial objections that have been raised against it. As we saw, Rescher’s basic intuition about modality is that there is no separate, ontologically autonomous realm of possibility alongside the realm of reality: there is only one world, the actual world. “The sphere of altogether mind independent reality includes only the actual” (ImI, p.183). Consequently, the possible must be built with materials provided by reality. Rescher’s suggestion is that we see the possible as stemming from a particular subdomain of the actual, namely minds. “The foundation of the possible lies in the actual capabilities of real minds” (ImI p.181, TP p.216). This is to say that “to be possible is to be conceivable by an actual mind” (ImI p.172); “Possibility is tantamount to conceivability” (TP 216). Now, what kind of identity is this? Rescher warns us that it should not be taken as a definition of possibility, but as an explanation (ImI, p.172); perhaps he regards possibility as stricto sensu undefinable. Anyway, there is little doubt that it (at least) implies extensional equivalence: whatever is possible is conceivable, and conversely. So this is indeed a strong thesis concerning the relation of conceivability to possibility. As we shall see,

172

CONCEIVABILITY

many philosophers have seen reason to challenge both directions of the equivalence, most insistently the right-to-left direction, i.e. the entailment from conceivability to possibility. It is also a classical thesis, clearly stated by David Hume (1973, p.32) and reiterated by Ludwig Wittgenstein in the Tractatus: “What is thinkable is possible too”.6 Let us try better to understand Rescher’s thesis. First of all, what is conceivability? Rescher appears to regard ‘conceiving’ as synonymous with ‘imagining’: “Imagination is our only accessway to the realm of the unreal” (ImI, p.170); “Possibilities are products of our imagination and are thereby geared to the resources that the emplacement of our intellects in this world’s scheme of things puts at its disposal”(ib.). However, the synonymy in itself does not bring much light. In contexts of this kind, ‘imagination’ may have two senses, one relatively clear but of little use in this context, the other rather obscure. In the (relatively) clear sense, ‘imagining’ means imaging: imagination is the capacity to form mental images or pictures. Suppose we grant that mental images exist and that we have a clear notion of them. Still, conceiving could hardly be identified with imagining in this sense. There are just too many things and states of affairs that we intuitively take to be conceivable though we have (and can have) no mental pictures of them, from Descartes’ chiliagon to sounds that no human ear could hear, from the existence of God to gold’s having atomic number 79.7 Conceiving, in both its philosophical and its everyday use, does not reduce to imaging. Rescher is of course aware of all this, so he hastens to make it clear that the conceivable must be construed so that it “includes imaging and imagining” (TP, p.213), and that imagining must be “purged of psychological connotations” (ImI, p.182, TP, p.213). Though this is fine as far as it goes, it leaves us with the question of what might imagining be, as distinct from imaging and once de-psychologized (this is the word’s second sense, the one that stands in need of clarification). Though this question will be answered eventually, for now it must wait. In Rescher’s view, conceivability is said to pertain to “actual minds”; or at any rate, we are equating possibility to conceivability by actual minds. Indeed, this is for Rescher the indispensable link to actuality that makes possibility a respectable notion. Now, clearly, ‘actual minds’ cannot mean ‘present minds’: we would hesitate to limit the conceivable to what we can here and now imagine on the basis of our present mental resources. For suppose we apply the limitation retrospectively: as (presumably) Julius Caesar could not imagine radiotelescopes, we would have to say that they

173

RESCHER STUDIES

were inconceivable for him. It would follow that either the connection between conceivability and possibility is severed (though radiotelescopes were inconceivable for Caesar, nevertheless they were possible) or possibility is turned into an epoch-relative notion (and so why not culturerelative, or individual-mind-relative, as well?). Granted, we do say such things as ‘Genuine European unity only became possible after the fall of the Berlin wall’. However, it is generally acknowledged that such uses of ‘possible’ should not be taken too seriously: what is meant here is that certain obstacles to European unity were only recently removed, not that an impossibility was turned into a possibility.8 Again, Rescher is fully aware of these difficulties (the Caesar example is his: ImI, p.178, TP p.215). His response is twofold. On the one hand, he allows for relativized notions of both conceivability and possibility: we can speak of what is possible for us, based on what we can conceive. For example: “The only possibilities for us are those which can be projected in terms of our conceptual scheme” (ImI, p.178). On the other hand, he also allows for the corresponding absolute notions. When we say that something is (unqualifiedly) conceivable, “we mean ‘conceivable per se’, not in terms of the currently available conceptual mechanisms” (ImI, p.179; TP, p.215). In saying that something is conceivable “we are concerned not with the activities of this or that possibility-projecting mind, but with the possibility-projecting capabilities of minds in general” (TP, p.213). In other words, ‘conceivable’ is an implicitly two-place predicate (“x is conceivable by y”); we can existentially quantify over the second variable, and this is just what we do when we say that something is conceivable, period. A further problem may look threatening. Suppose we grant Rescher that when we say that something is conceivable we mean that it can be conceived by some actual mind, not excluding future minds. How much do we know about what future human minds will be able to excogitate? If conceivability is anchored to the mental capacities of human minds in general, the anchoring may be weak. We simply have no idea, right now, of what may turn out to be conceivable. A deep revision of scientific knowledge might imply the conceivability of perpetual motion. But then, we just don’t know—as of now—what is really conceivable and what isn’t (with the possible exception of explicit contradictions). Consequently, we don’t know what is possible: our modal knowledge is nil. But perhaps the threat can be dissolved. For possibility to be effectively explicated in terms of conceivability, it is not necessary that the extension of conceivability is specified. We do not disqualify Tarski’s definition of

174

CONCEIVABILITY

truth because it does not yield the extension of the concept of truth. As in every other such case, as far as extension is concerned we cannot go beyond our present beliefs, hypotheses, guesses. Rescher’s explication of possibility instructs us to go by our present views of what is conceivable and what isn’t to determine what is possible and what isn’t, in full awareness that such views are defeasible: what we regard as possible need not coincide with what is possible. This entails that we lack modal certainties, not that we have no modal knowledge: for some of our views about what is conceivable (and therefore possible) may well be correct. Modal knowledge, in this respect, is in the same boat as ordinary knowledge. Finally, let us see how Rescher deals with a more subtle objection, one that was repeatedly voiced by Wittgenstein in the “later” stage of his reflection. There is no denying that, in speaking of conceivability by actual human minds, we at least seem to be hinting at certain mental faculties. Now, faculties properly so called—such as the ability to add small numbers or to understand spoken language—may be the possession of some, many, or almost all human beings; they may be stronger in certain individuals and weaker in others; they may be lost and regained; in short, they are entirely contingent features of certain organisms and their brains. Is this compatible with what we have in mind when we say that something is conceivable? For example, when we say that “we can’t imagine that p”, do we mean that our imagination faculty is not strong enough—not at the present moment, anyway—to generate the imagination or conception that p (though at another moment, it might be, as it might be for a better endowed fellow human)? Is “I can’t imagine that p” to be compared with “I can’t mentally add 356 and 915 (maybe you can)”? Wittgenstein didn’t think so. He thought that “I can’t imagine that p” was a logical, not an empirical remark. Roughly, it was not about anything we could or could not do; it was about p’s making or not making sense. “I can imagine the colour transition” isn’t an assertion here about a particular power of my own imagination, in the way that “I can lift this stone” is about the power of my own muscles [. . .] it is to be understood as a proposition of grammar. It looks as if we could say: Word-language allows of senseless combination of words, but the language of imagining does not allow us to imagine anything senseless. (Wittgenstein 1974, pp.128-129; cf. 1953, §251)

To say that we cannot imagine that ~p is just to say that ‘~p’ doesn’t make sense (which, for Wittgenstein, crucially entails that ‘p’ doesn’t make sense either: 1974, p.129; 1953, §251).

175

RESCHER STUDIES

Again, Rescher is well aware of the risks that are involved in defining conceivability in terms of faculties of “actual human minds”, though he is mainly concerned with not having conceivability (and, consequently, possibility) depend upon the idiosyncrasies and accidentalities of “the mental life of people” (TP p.213). Anyway, his remedy is essentially the same as Wittgenstein’s: conceivability is made “entirely public and interpersonal through its objectification in language” (TP, p.213); thanks to the use of “the structured and structuring . . . linguistic machinery”, the mind is extricated from its idiosyncratic vagaries and induced to proceed in “restricted and canalized ways” (ImI, p.183). Thus the conceivable amounts to “possible descriptions, language enshrined concepts” (ImI, p.180); consequently, “To be a possibility is to answer . . . to a certain description” (ImI p.185 = TP p.210). Note that this is, at the same time, an answer to the question of what might imagination be as distinct from imaging: imagination is just linguistic description; imaginability is describability in language. There is, however, a standard objection against the linguistic interpretation of conceivability, which has led many philosophers to reject it. There appear to be intelligible linguistic descriptions of impossible states of affairs: necessary falsehoods such as ‘2+2=5’ are intelligible (Tidman 1994, p.302, Fiocco 2007, p.365), and so are explicit contradictions of the form ‘p & ~p’. If what is expressed by such linguistic formulations is regarded as conceivable via the identification of conceivability with linguistic describability, then conceivability cannot seriously be regarded as an explanation of possibility, for contradictions are impossible if anything is. A radical reply to this objection was first put forth by Wittgenstein in Tractatus logico-philosophicus: it consists (as so often in the Tractatus) in biting the bullet and simply denying that contradictions, and necessary falsehoods in general are genuinely intelligible.9 In the Tractatus, this is part of a coherent (though famously controversial) view of the relations between possibility and conceivability, based on the notion of making a picture of a state of affairs. It is doubtful that Rescher—like most contemporary thinkers—would endorse Wittgenstein’s full view. Anyway, his reply to the anti-linguistic objection is prima facie different. After granting that “the impossible may well be described or stated” (TP, p.211), as the antilinguistic objection points out, he goes on to say that, nevertheless, the impossible “cannot correspond to a ‘meaningful thought’” (ib.). In other words, the conceivable is not merely what can be described in language (even contradictions can): rather, it is what can be meaningfully thought.

176

CONCEIVABILITY

Though language, with its “structured and structuring” machinery, may be helpful—perhaps even indispensable—in “objectifying” and disciplining thought, it is not sufficient to guarantee conceivability. Now, at this point in the argument, the notion of a meaningful thought is apparently not to be explicated in terms of human mental faculties: for if it were, we could hardly avoid the impression of being back to square one. How, then, is it to be explicated? Though Rescher is not explicit about this, occurrence of the word ‘meaningful’ authorizes the suspicion that there may be something semantic about a thought’s being meaningful: as if a meaningful thought were a thought that can be expressed by a meaningful linguistic description. If this were the correct explication, then Rescher’s response to the anti-linguistic objection would not, after all, be essentially different from Wittgenstein’s: for, like Wittgenstein, he would be relying on a distinction between what can be stated in language (which includes contradictions and necessary falsehoods in general) and what can be meaningfully stated in language. Some such distinction would turn out to be essential to an understanding of conceivability. This, however, is a narrow and thorny path. Not everybody is happy with the notion of a genuine linguistic expression—not a word jumble— that would be, nonetheless, meaningless and unintelligible; particularly, many have found the view that contradictions are meaningless and unintelligible deeply counterintuitive. However, Rescher is not explicitly offering an alternative, nor is it clear that there is an alternative. Except, of course, taking the notion of a meaningful thought as primitive. But then, one would wonder whether explaining possibility in terms of an unexplicated notion of “meaningfully thinkable” is a substantial theoretical gain with respect to not explaining it at all, as in the views of those philosophers that hold that possibility judgments are based on “direct modal intuition”.10 Let me summarize Rescher’s views on possibility and conceivability. Possibility amounts to conceivability. Conceivability involves actual minds, though not necessarily present minds: to be conceivable is to be imaginable by some human mind. The imaginative operations of human minds are carried out in language: hence to be conceivable is to be linguistically describable. To the objection that even inconsistencies, though surely inconceivable, may be said to be linguistically describable Rescher replies that the linguistic description of an inconsistency does not “correspond to a meaningful thought”. So, finally, to be conceivable is to be describable in language in such a way that the description corresponds to (= expresses?) a meaningful thought. If “expressing a meaningful thought” is

177

RESCHER STUDIES

taken as equivalent to “being meaningful”, then Rescher’s view of the relation of possibility to conceivability essentially coincides with Wittgenstein’s view in the Tractatus. If not, then Rescher owes us an alternative explication of what it is to correspond to, or express a meaningful thought. Such difficulties with the notion of conceivability are by no means characteristic of Rescher’s analysis. Indeed, the notion has proved extremely resistant to satisfactory clarification. But, on the other hand, the notion is not easily renounced either. For if we give up conceivability we seem to lose our epistemic grip on the notion of possibility: how are possibility claims to be assessed, if not by testing the conceivability of their contents? We seem to be left with the prospect of basing modal knowledge on some sort of direct modal insight. However, the problem with this solution is that it is not clear that it provides a genuine alternative to basing modal knowledge on conceivability tests. It is generally agreed that possible worlds are not there for us to inspect: in Rescher’s words, we have no “intuitive insight into the modus operandi of nonexistent possible worlds” (C p.135). But then, what is exactly the difference between “seeing” that soand-so is possible and imagining it to be so-and-so? It appears that what we do in both cases is examining certain thoughts of ours, usually expressed in language. If modal intuition is about examining thoughts, it is not clear that it is not just imagining or conceiving; if on the other hand it is about examining states of affairs, i.e. possible facts, doesn’t this suppose that we have epistemic access to possible worlds? To be sure, there is plenty of room for explanation concerning both “thoughts” and “examining”; however, it doesn’t seem that appealing to direct intuition removes the need for such explanations. At most, it expresses our unwillingness to provide them. This is why, in spite of difficulties, we have reasons to stick to the connection between possibility and conceivability. 3. POSSIBILITY AND CONCEIVABILITY AFTER KRIPKE AND PUTNAM There is a less controversial side to the equivalence of possibility and conceivability: the inference from “possible” to “conceivable” has been widely regarded as safe (Rescher goes as far as stating that “possible though not conceivable” is not “a viable locution”, ImI p.181). Recently, even this side of the equivalence has been challenged. Stephen Yablo (1993) claimed that Goldbach’s conjecture (that every even number is the sum of two primes) is neither conceivable nor inconceivable: “No thought

178

CONCEIVABILITY

experiment that I . . . can perform gives me the representational appearance of the conjecture as possible or as impossible, or the slightest temptation to believe anything about its modal character” (p.11). However, if the conjecture is true then it is surely possible, so that—if the inference holds—it should be conceivable. But we don’t take the inconceivability of Goldbach’s conjecture to be proof of its falsity. This particular argument clearly depends on a surely legitimate but special sense of ‘conceivable’. Anyway, I am not going to take up this side of the issue. I will focus on the inference from “conceivable” to “possible”, which has proved far more controversial. Long ago, Thomas Reid noticed that the inference was not as straightforward as Hume had taken it to be: many mathematical propositions had been regarded as conceivable that were later proved not just false but impossible.11 However, the issue came to be dramatized after Putnam’s and Kripke’s “discovery” of the necessary a posteriori. The propositions expressed by ‘Hesperus is Phosphorus’, ‘Salt is NaCl’, and other such sentences were then recognized to be necessary, if true. On the other hand, many—including Putnam and Kripke themselves—conceded that it was at least prima facie conceivable that Hesperus was not identical with Phosphorus, or that salt was not NaCl. For such truths are a posteriori: they have been revealed through experience. Different experiences—so the intuition went—might have revealed different facts, and there was nothing a priori to rule out the possibility of such different experiences. Given the appropriate experiences, we would have come to the conclusion that salt is KCl, not NaCl, or that Hesperus, the evening star, is not the same celestial body as Phosphorus, the morning star. As it is conceivable that such experiences might have occurred, it is likewise conceivable—imaginable, not a priori ruled out—that salt is not NaCl, or that Hesperus is not the same as Phosphorus. We thus seem to have examples of situations that are both conceivable and impossible. So perhaps conceivability does not entail possibility. Not everybody drew this conclusion, however; not Kripke himself, for one. He pointed out that perhaps we were mistaken in thinking that the envisaged states of affairs—that Hesperus is not Phosphorus, that salt is KCl—were genuinely conceivable. Perhaps they only looked conceivable. Something may be conceivable that is easily mistaken for the state of affairs that (say) Hesperus is not Phosphorus, though it is not that state of affairs. For example, we may think that we have exactly the evidence that we do have—we are in the same epistemic situation as we in fact are—except that such evidence is caused by two different celestial bodies, only one of

179

RESCHER STUDIES

which (or none of which) is the planet Venus. That, however, would not be a situation in which we would be epistemically accessing the state of affairs that Hesperus is not Phosphorus, i.e. that the planet Venus is not identical with itself.12 It would be a situation in which, in Crispin Wright’s words (2002, p.410), we are accessing a symptomatic imitation of the (impossible) state of affairs that Hesperus is not Phosphorus. By this move, the inference from conceivability to possibility is salvaged: though conceivability does entail possibility, it is not really conceivable that Hesperus is not Phosphorus. But, on the other hand, the problem arises of how can we ever tell whether what we are conceiving is really p or a symptomatic imitation of p. Many other moves have been made in connection with the conceivability → possibility inference and the alleged counterexample of necessary a posteriori propositions. To scrutinize such moves, let us characterize the issue in some detail. We have a thesis, deriving from Putnam’s and Kripke’s work on reference and rigid designation: that certain theoretical identities are both necessary and a posteriori. Let us take “Salt is NaCl” as an example. It is maintained that the proposition that salt = NaCl is necessary: (Th)

Nec ()13

We then have a Simple Argument purporting to refute (Th). Suppose it is conceivable that salt is not NaCl, and further suppose that whatever is conceivable is possible. It follows that it is possible that salt is not NaCl, against (Th): 1. Conc () 2. (p) (Conc (p) → Poss (p)) 3. Poss () Rescher would say that the set {(1), (2), (Th)} constitutes an aporetic cluster:14 granted that Poss (~p) → ~Nec (p), (1), (2), and (Th) cannot all be maintained. In principle, there seem to be three ways out of the aporetic cluster:

180

CONCEIVABILITY

(A) Reject (Th): accept the argument and claim that it is simply not the case that salt is necessarily NaCl. This view—in a way, the most obvious option—is not often defended in the literature; it can be found e.g. in Ayer (1982, p.270). I will not consider it in what follows, not because I regard it as a dead option but because I want to focus on the current debate on the issue. (B) Reject (1): claim that is not really conceivable. As we shall see, this was Kripke’s original move. Crispin Wright (2002) suggests another way to make the same move. (C) Reject (2): claim that conceivability does not entail possibility. This was Putnam’s initial, off-the-cuff choice in (1975, p.233). He argued for it in (1990). The same view was later espoused by Tidman (1994) and Fiocco (2007).15 However, this is not all. There is a fourth possible move, which does not consist in rejecting any of (1), (2), or (Th) but in rejecting the Simple Argument purporting to refute (Th) on the basis of (1) and (2). The Simple Argument is rejected as based on equivocation: ‘conceivable’ and ‘possible’ are not consistently used in the same senses throughout the argument. A typical proponent of this view is David Chalmers (e.g. 2002): there are senses of ‘conceivable’ and ‘possible’ such that both (1) and (2) are true, but, taking ‘possible’ in that sense, (3) is not the negation of (Th). Yablo (2002) can be read as working along similar lines: he wants to define a notion of possibility such that (2) is plausible, but—understanding ‘possible’ in that sense—(3) does not negate (Th). In all these cases, the way out is (D) Reject the Simple Argument, {(1), (2)} |—~(Th), as based on equivocation. People who make this move are often aiming at preserving the inference from conceivability to possibility in some sense of ‘conceivable’ and ‘possible’. Now, our present theoretical interest is in the validity of the conceivability → possibility inference. Hence, we are addressing the resolution of the aporetic cluster with two main questions in mind: (i) Are there good direct arguments against the inference’s validity? By a direct argument I mean an argument that does not simply consist in assuming (1) and (Th)

181

RESCHER STUDIES

and deriving the inference’s invalidity by reductio. (ii) Can we vindicate the conceivability → possibility inference without simply denying (Th) — the easy, but most unpopular way out? As we shall see, (B), (C), and (D) provide different answers to both questions. To put it briefly, proponents of (B), such as Kripke, think that the inference can be preserved even in the face of (Th), since it can be shown that (1) is false: we cannot really conceive that salt ≠ NaCl. If they are right, the inference, be it valid or invalid, is anyway not threatened by (Th). Proponents of (C), such as Putnam, Tidman, and Fiocco, claim that the inference is generally invalid, independently of whether we can or cannot conceive that salt ≠ NaCl; indeed, invalidity of the inference is their reason for not seeing the Simple Argument as a threat to (Th). Finally, proponents of (D) such as David Chalmers believe that there are senses of both ‘conceivable’ and ‘possible’ such that the inference from conceivability to possibility is valid (as well as senses in which it is not). However, there are no senses of ‘conceivable’ and ‘possible’ that are such that premises (1) and (2) in the Simple Argument both hold, and they jointly entail the negation of (Th). Thus for Chalmers the inference can indeed be preserved (in more than one way) without having to give up (Th). Alternative (D) agrees with (B) in holding that—in one sense in which the conceivability → possibility inference is valid—we cannot really conceive that salt ≠ NaCl; and it agrees with (C) in maintaining that there are senses of ‘conceivable’ and ‘possible’ in which the inference is invalid. Let us start with (D). I will present Chalmers’ views rather sketchily, as they are by now well known and have been thoroughly discussed.16 There are two main notions of possibility (and, correlatively, of necessity). One, called ‘secondary possibility’, is just the familiar notion of metaphysical possibility: p is secondarily possible iff there is a counterfactual world such that p In this sense, given that salt is NaCl, it is impossible that salt ≠ NaCl. The other notion, primary possibility, is based on the idea of the actual world being different from what we—a posteriori—know it to be. For all we know a priori, our world—the actual world—could be such that salt is not NaCl. ‘Salt’ might—for all we know a priori—designate a substance that is not NaCl. In this sense, it is possible (primarily possible) that salt ≠ NaCl.17 Conceivability is a more intricate matter. Chalmers introduces three distinctions, which (in principle) originate eight different notions of conceivability. The first distinction is between negative conceivability (= compatibility with what we know a priori) and positive conceivability (= imagin-

182

CONCEIVABILITY

ability, though not just image-ability). The second is between prima facie conceivability and ideal conceivability, i.e. conceivability upon ideal rational reflection (Thomas Reid’s mathematical impossibilities may have been prima facie conceivable, but they are not ideally conceivable). The third, and crucial distinction is between primary and secondary conceivability. p is primarily conceivable iff it is (negatively or positively, prima facie or ideally) conceivable that p is actually the case. It is secondarily conceivable iff it is (negatively or positively, etc.) conceivable that it might have been the case. With these definitions in place, and disregarding many interesting details, Chalmers’ main theses are the following: (D.1) Primary conceivability entails primary possibility, (D.2) Secondary conceivability entails secondary possibility, (D.3) Primary conceivability does not entail secondary possibility. Supposing that Chalmers can convincingly argue for (D.1)-(D.3), then (1) he would have given a qualified, though essentially affirmative answer to the question whether conceivability entails possibility, and (2) he would have shown that we can preserve the inference without giving up (Th), provided ‘Nec’ in (Th) stands for secondary necessity (as it is usually taken to be the case). In fact, that salt ≠ NaCl is both primarily conceivable and secondarily inconceivable. As it is secondarily inconceivable, validity of (D.2) does not threaten (Th); as it is primarily conceivable, validity of (D.1) only implies that it is primarily possible (as it arguably is), which, again, does not contradict (Th) in the assumed (“secondary”) interpretation. So (Th) is neutralized as a counterexample to the conceivability → possibility inference, once both ‘conceivable’ and ‘possible’ are properly interpreted. However, one wonders whether much has been accomplished in terms of clarifying the inference. For example, have we been provided with clear notions of conceivability? Primary and secondary conceivability appear to be tailored upon the corresponding notions of possibility: p is primarily conceivable iff it is conceivable that it is primarily possible, and similarly with secondary conceivability. Hence, in order to understand, say, primary conceivability we have to fall back onto other notions of conceivability, particularly, negative and positive conceivability (though it surely makes sense to note that something might be only superficially (prima facie) con-

183

RESCHER STUDIES

ceivable while further reflection would persuade one that it is really inconceivable, it falls short of explicating conceivability itself). Now, negative conceivability is a relatively well defined notion, but Chalmers himself has doubts whether it can be said to entail possibility in any sense (2002, p.161). Positive conceivability, on the other hand, remains rather obscure. It is clearly to be distinguished from “perceptual imaginability” or imageability; moreover, positively conceiving that p does not coincide with entertaining the proposition that p, for there are propositions (e.g. “complex unknown mathematical propositions”) that are not easy to imagine (p.151). Eventually, we are told that the notion “resists straightforward definition” (ib.): we are only provided with a synonym, ‘modal imagina[bility]’, and with examples such as modally imagining Germany winning the last World War by imagining “a world in which certain German armies win certain battles and go on to overwhelm Allied forces within Europe” (ib.). (Note, however, that it is easy to suppose all this taking place without Germany winning the Second World War). All in all, a rather lame notion. I, for one, cannot figure out what it would be to modally imagine, or positively conceive water not being H2O. If the explication of the notion(s) of conceivability is less than satisfactory, results (D.1)-(D.3) are themselves less than completely convincing. Even taking Chalmers’ two notions of possibility for granted, one feels that much remains to be done to articulate notions of conceivability that would make the inference from conceivability to possibility go through, in some sense of ‘conceivable’ and ‘possible’. Chalmers’ rescue of the inference is thus partly based on a promissory note. Also, one doesn’t feel that much progress has been made with respect to the Wittgensteinian (and, perhaps, Rescherian) notion of conceivability as meaningful linguistic describability. Let us now look at the critics of the conceivability → possibility inference, i.e. let us consider alternative (C). As I mentioned, the first to argue for this option was Hilary Putnam. To show that conceivability does not entail (metaphysical, or, as he chose to say, logical) possibility, Putnam relied on Kripke’s notion of an epistemic counterpart, but he put it to a very different use. In Kripke’s view, when we say that it is conceivable (say) that salt is not NaCl we are misdescribing something else: namely that it is conceivable that there is an epistemic counterpart of salt which is not NaCl. This is indeed conceivable, according to Kripke; but then, what we are asserting to be conceivable is not that salt ≠ NaCl. Putnam, on the other hand, thinks that ‘it is conceivable that salt is not NaCl’ means that some

184

CONCEIVABILITY

epistemic equivalent of salt may not be NaCl: “What is conceivable is a matter of epistemology” (1990, p.62). Kripke used the notion of an epistemic counterpart to explain how it may (wrongly) seem to us that it is conceivable that salt ≠ NaCl; Putnam uses it to explicate ‘it is conceivable that salt ≠ NaCl’. Having assumed that this is what ‘conceivable’ means, it is not hard to show that conceivability does not entail possibility. It may well be that some epistemic counterpart of salt is not NaCl, so that—in Putnam’s preferred sense—it is conceivable that salt is not NaCl. On the other hand, that salt ≠ NaCl is not (metaphysically) possible. So conceivability does not entail possibility. Something like Putnam’s argument was later revived by Chalmers to show that primary conceivability does not entail secondary possibility. Is there any reason to attribute wider import to Putnam’s argument? It stands to reason that, if what we mean by saying that it is conceivable that salt is not NaCl is that something, not necessarily salt but something epistemically indistinguishable from salt, might not be NaCl, then this does not entail that salt itself might not be NaCl. The question is whether this is all we can mean by ‘conceivable’, as Putnam seems to be assuming. We saw that, for partisans of alternative (D), other senses of ‘conceivable’ are available; we also saw that this is neither Wittgenstein’s nor Rescher’s sense of ‘conceivable’. It thus seems that Putnam only showed that there is a sense of ‘conceivable’ in which ‘conceivable’ does not entail ‘metaphysically possible’. Paul Tidman, a forerunner of the present debate on the relation of conceivability to possibility, put forth the following argument. Grant that it is prima facie plausible that we can imagine impossibilities: for example, that this particular cat is a robot. As this cat couldn’t be a robot, the inference to possibility does not go through, at least prima facie. To maintain that in fact it does, we have to show that, appearances to the contrary, we cannot imagine impossibilities. But how could we? “If you cannot, and therefore do not, really believe that Hesperus is not identical with Phosphorus, it must be that you cannot tell that you lack this belief” (1994, p.303). For— the argument seems to be—if you could tell that you lack the belief you wouldn’t be holding precisely that belief, as we are assuming you are. This is much too quick. First of all, it may be that we cannot conceive of impossibilities even though we are under the unavoidable delusion that we can. To deny this—that we may, in some cases, inevitably misrepresent to ourselves the content of our own beliefs—is to assume a form of belief transparency: a substantial philosophical thesis. But, secondly, why

185

RESCHER STUDIES

couldn’t I, upon reflection, come to the conclusion that what I was imagining is not best described by “this cat is a robot” (but, say, by “this cat-like object is a robot”)? Tidman appears to be assuming that I couldn’t have a reason to choose the latter description over the former. This, however, amounts to assuming that we cannot rationally reflect upon the content of our conceivings, i.e. that we can’t go beyond prima facie conceiving. Here, Chalmers’ motivations for his distinction between prima facie and ideal conceivability are in point. Very recently, a new argument against the conceivability → possibility inference was proposed by M. O. Fiocco (2007). After surveying several notions of conceivability, Fiocco comes to the conclusion that to imagine that p is to suppose that some world verifies p. However, this is not established by examining possible worlds; we stipulate that a world verifies p. Not that we can stipulate that a given world is possible (p.374): modal facts, according to Fiocco, are independent of our stipulations. However, we are free to stipulate which world we are considering (be it possible or not). What we are imagining is up to our stipulations: “In order to be assured of introducing a world that verifies the proposition that transparent iron exists, one need only stipulate consideration of a possible world in which transparent iron exists” (pp.375-376). But then, “any world whatsoever [be it possible or not] is modally imaginable” (p.375). So conceivability (= modal imaginability, = stipulability) does not entail possibility. There is some confusion in the way the argument is phrased. Though Fiocco insists that any world whatever, possible or not, can be stipulated, he also says that “one stipulates which possible world one is introducing” (p.375, my italics), and that “. . . one need only stipulate consideration of a possible world in which transparent iron exists”. The latter formulations seem to imply that stipulation selects a world, or a class of worlds, among possible worlds. If this were Fiocco’s view, then conceivability, i.e. stipulability, would indeed entail possibility, against his explicit claim. Hence, we must take the latter formulations as slips and stick to the claim that conceiving that p consists in stipulating that we are considering a world such that p (be it possible or not). But then, the same point we made against Tidman is relevant here: why should we take our prima facie stipulations as the end of the matter? We are usually interested in whether what looks prima facie conceivable to us is really conceivable, i.e.—in Fiocco’s terms—in whether a p-world is a possible world. We do no go about stipulating consideration of a world such that p &~p, because we can’t conceive of such a world. Though our prima facie stipulations may lead us to con-

186

CONCEIVABILITY

sider impossible worlds, we usually do not stop there and there is no reason why we should. So, either conceivability does not reduce to stipulability, or our stipulations are subject to rational reflection. To be sure, there is no guarantee that such reflection would succeed in pruning out all hidden impossibilities; however, Fiocco doesn’t show that our attempts are bound to fail. The mere fact that our prima facie stipulations may be untenable does not show that conceivability does not entail possibility. Direct arguments against the conceivability → possibility inference only succeed in establishing that there are senses of ‘conceivable’ such that conceivability does not entail metaphysical possibility (as we saw, Chalmers came to the same result as a side-effect of a strategy aimed at vindicating the conceivability → possibility inference). Hence, these criticisms are not the last word on the inference. However, the Simple Argument still stands: supposing the inference to be valid, how do we reconcile (Th) with the alleged conceivability of ? Supporters of alternative (B) like Kripke answer that we do not: we just have to give up the mistaken hypothesis that we can conceive of such impossibilities as Hesperus being other than Phosphorus or salt not being NaCl. To defend this view, they are under the obligation of explaining how it comes to seem to us that such (alleged) impossibilities are conceivable. We saw how this obligation was met by Kripke: in his view, we are deluded into believing that we can conceive that salt is not NaCl because we actually can conceive that something else—a “symptomatic imitation” of salt—is not NaCl. In his (2002), Crispin Wright suggests another possible explanation. We believe that Fermat’s Last Theorem has been proved, so that it is true and necessarily so. Suppose someone claims she can imagine that the theorem is false: e.g. she imagines finding mistakes in Andrew Wiles’ proof, finding counterexamples to the theorem, etc. We could then say she is not really imagining the theorem being false; rather, she is imagining states of affairs that would be the case if, per impossibile, the theorem were false. Other examples of the same kind are easily made up: for example (exploiting a suggestion of Tidman’s, 1994, p.299), someone could claim he can conceive that gold does not have atomic number 79, as he can picture scientists labeling certain rocks with the numeral 78. Thus, there is a whole “category of conceivings that fall short of being genuine counterconceptions to a given [necessary] proposition. . .because it is insensitive to another distinction: that between genuine conceivings of a scenario in which p fails to obtain and conceiving, rather, what it would be like if , per impossibile, p were (found to be) false” (Wright 2002, p.437).

187

RESCHER STUDIES

Hence, there are two ways in which step (2) in the Simple Argument— the conceivability → possibility inference—can be preserved without having to give up (Th). One is Kripke’s way: showing that what is being conceived is not ~p but a symptomatic imitation of ~p. The other consists in showing that what is being conceived is, again, not ~p but “what it would be like if” ~p (per impossibile) were the case. However, Wright sees a crucial difference between the two ways. With Kripke’s way, cases like the Hesperus-Phosphorus case can be contrasted with cases in which we cannot distinguish between conceiving that ~p and conceiving of a symptomatic imitation of ~p Such is, for example, the case of pain: as there is no symptomatic imitation of pain, we cannot distinguish between genuinely conceiving that pain is not the stimulation of C-fibers and conceiving that some epistemic counterpart of pain is not the stimulation of C-fibers. Not so with per impossible pseudo-conceivings: here, if p is true then every conceiving that ~p, “no matter how intricate and coherent”, is open to the charge of being a per impossibile pseudo-conceiving, for it is always possible to draw the distinction between conceiving that ~p and conceiving of what would be the case if—per impossibile—~p. Wright’s conclusion is devastating, not for the tenability of the conceivability → possibility inference but for its “modal significance”: as premise (1) in the Simple Argument can always be challenged, we can never rely on the inference to draw any conclusion concerning the modal status of p The inference from conceivability to possibility is preserved at the very high price of not knowing for sure if it can be applied to derive a modal conclusion from a conceivability premise, though “it can continue as a defeasible operational constraint on the ascription of necessity in cases where contingency is an epistemic possibility” (Wright p.438, his italics). Perhaps, however, the conceivability → possibility inference has been too quickly downgraded.18 Let us go back to Wright’s discussion of Kripke’s strategy. Wright says that, within that strategy, the inference “is allowed to retain some teeth” for there are cases, such as pain, in which we cannot disqualify a conceiving from being a conceiving-that-p by appealing to the possibility of a symptomatic imitation of p One gathers that otherwise the inference would be epistemically emptied (as with the second strategy, according to Wright). Why so? Suppose we could find no case such as the pain case: every time, we are as a matter of fact able to show that one who believes he is conceiving that p may be plausibly regarded as conceiving that q, where q is an epistemic counterpart of p. Notice that we are not supposing that it is assumed that this can be done in every case; we

188

CONCEIVABILITY

are supposing that we have been able, de facto, to do it in every forthcoming case. I.e. we are supposing that it is plausible to surmise that one who says she can conceive that Hesperus is not Phosphorus is actually thinking of two different planets having the same observational effects as Venus has, and that it is plausible to surmise that one who says he can conceive that water is not H2O is actually thinking of a substance such as Putnam’s XYZ, and so forth. Nothing seems to be wrong so far: all seems to depend on whether our conjectures are, indeed, plausible. If they are (which may be doubted, of course) then we have succeeded in explaining why one might wrongly believe to be conceiving that Hesperus is not Phosphorus, etc. We have succeeded independently of whether there are cases in which we can advance no plausible conjecture to a comparable effect. Things would be different if we were just assuming that it could be done in every possible case, without submitting to the burden of actually doing it. Then, indeed, we would be epistemically trivializing the conceivability → possibility inference, for we would be making it in principle impossible for anybody to argue that he is genuinely conceiving that p and not of something else. But that’s not what we are supposing we are doing: we are just showing, in each forthcoming case, that we can plausibly think of an epistemic counterpart of p. The possibility remains open that we might encounter a case in which this cannot be done (and, if Kripke is right, such a possibility will be actualized once we come to think of pain). Now, why is it any different with the second strategy? The difference ought to consist in something like the following: with the second strategy, we can prove that (if p is an alleged necessary a posteriori truth) every conceiving that is claimed to be a conceiving-that-~p can be plausibly argued to be, not a conceiving-that-~p but a conceiving of something that would be the case if (per impossibile) ~p. This would amount to proving that any redescription of the fact that ~p, “no matter how intricate and coherent”, might have different truth conditions from ~p itself. Discounting problems with negation, this, in turn, would be like proving that no two distinct sentences, p and q, must have the same truth conditions. Which seems widlly implausible. As we saw, there are many cases in which it is quite natural to say that what has been presented as a conceiving-that-~p is really a conceiving of something that would be the case (or even, might be the case) if ~p. But proving that this could be so in every case, for every candidate redescription of ~p, is another matter. Thus it is not clear that the conceivability → possibility inference can be vindicated only by making it useless as a step in arguments from con-

189

RESCHER STUDIES

ceivability to possibility. Clearly, however, alternative (B) succeeds in showing that we can countenance both a posteriori necessity and the conceivability → possibility inference only to the extent that we can plausibly show that the negation of a given a posteriori necessity only looks conceivable, though it really isn’t. When this cannot be done, as in the pain case,19 we have a choice between (1) allowing for exceptions to the inference, (2) acknowledging that the relevant proposition is not necessary after all, and (3) maintaining that it is not even true (Kripke’s choice in the case of ). Even if we choose the first option, however, conceivability remains a prima facie good, though defeasible guide to possibility. It may have been noticed that, throughout the discussion of (B), the notion of conceivability was left unanalyzed. In terms of Chalmers’ taxonomy, Kripke’s conceivability must be secondary, not primary, as negations of necessary a posteriori truths are primarily conceivable. On the other hand, it seems that nothing in the discussion commits Kripke to either positive or negative conceivability. As for Wittgenstein conceivability, i.e. meaningful describability in language, it seems to be ruled out by the consideration that Kripke would regard, e.g., ‘water is not H2O’ as fully meaningful (hence conceivable in this sense). 4. RESCHER’S WAY OUT Finally, what would be Rescher’s way out of the aporetic cluster? It appears that it couldn’t be (C), for he is a staunch defender of the inference from conceivability to possibility (as one side of their equivalence). Moreover, it seems that it couldn’t be (D), for Rescher appears to be persuaded that ‘conceivable’ is not equivocal at all. Could it then be (B)? If we load Rescher with Wittgenstein’s notion of conceivability (as I am inclined to do), then regarding as prima facie conceivable though ultimately inconceivable, as per (B), amounts to taking ‘salt ≠ NaCl’ to be prima facie meaningful but ultimately meaningless.20 Would Rescher accept to bite this particular bullet and maintain that ‘salt ≠ NaCl’ is (ultimately) meaningless? At first sight, it is hard to believe that he would. Thus, Rescher’s way out seems to be (A), i.e. denying that is a necessary proposition; which, once more, would put him at odds with widespread philosophical opinion. However, I am not so sure that this would be his final word. First of all, Rescher made room for a notion of nomic modality (TP p.146), on which whatever is inconsistent with

190

CONCEIVABILITY

the laws of nature is (nomically) impossible. In this qualified sense, Rescher could say (and probably would say) that it is impossible that salt is not NaCl, and would have no objection to saying that it is (nomically) inconceivable that it is. He would thereby be joining the supporters of alternative (B), though not as far as the “full”, or unqualified senses of ‘conceivable’ and ‘possible’ are concerned. But there is more. We saw that Rescher understands conceivability in terms of meaningful describability. Now, at least once, while stating the connection between possibility and describability, Rescher said that “To be a possibility is to answer . . . to a certain description—one that is framed by means of the language-grounded taxonomic and nomic machinery designed to cope with the ‘real world’“ (TP p.210, italics added). If ‘describable’ means ‘describable on the basis of (our) taxonomic and nomic machinery’, then perhaps we should say that ‘salt ≠ NaCl’ is not an admissible description, or in other words that it is inconceivable that salt is not NaCl. Could it be that, for once, Rescher and Kripke are on the same side?

REFERENCES Ayer, A. J., Philosophy in the Twentieth Century (Weidenfeld: London, 1982). Chalmers, D., “Does Conceivability Entail Possibility?” in T.Szabó Gendler, J.Hawthorne (eds.) 2002, pp. 145-200. Fiocco, M. O., “Conceivability, Imagination and Modal Knowledge,” Philosophy and Phenomenological Research, vol. 76 (2007), pp. 364380. Garcia-Carpintero, M., and J. Macia (eds.), Two-Dimensional Semantics (Oxford: University Press, 2006). Hume, D., A Treatise of Human Nature (1739-40) (Oxford: Oxford University Press, 1973).

191

RESCHER STUDIES

Lewis, D., Counterfactuals (Cambridge, Mass: Harvard University Press, 1973). Marconi, D., “Two-Dimensional Semantics and the Articulation Problem,” Synthese, vol. 143 (2005), pp. 321-349. Putnam, H., “The Meaning of ‘Meaning’,” in his “Mind, Language an Reality,” Philosophical Papers, vol. 2, (Cambridge: Cambridge University Press, 1975), pp. 215-271. ———, “Is water necessarily H2O?” in his Realism with a Human Face (Cambridge: Harvard University Press, 1990), pp. 54-79. Reid, T., Essays on the Intellectual Powers of Man (1785), (Cambridge, Mass: MIT Press, 1969). Rescher, N., A Theory of Possibility [=TP] (Oxford: Blackwell, 1975). ———, The Strife of Systems, (Pittsburgh: University of Pittsburgh Press, 1985). ———, Imagining Irreality [=ImI], (Chicago and La Salle: Open Court, 2003). ———, Presumption and the Practices of Tentative Cognition [=P], (Cambridge: Cambridge University Press, 2006). ———, Conditionals [=C], (Cambridge Mass: MIT Press, 2007). Szabó Gendler, T. and J. Hawthorne (eds.), Conceivability and Possibility, (Oxford: Clarendon Press, 2002). Tidman, P., “Conceivability as a Test for Possibility,” American Philosophical Quarterly, vol. 31 (1994), pp. 297-309. Van Cleve, J., “Conceivability and the Cartesian Argument for Dualism,” Pacific Philosophical Quarterly, vol. 64 (1983), pp. 35-45.

192

CONCEIVABILITY

Yablo, S., “Is Conceivability a Guide to Possibility?” Philosophy and Phenomenological Research vol. 53 (1993), pp. 1-42. ———, “Coulda, Woulda, Shoulda”, in T.Szabó Gendler and J.Hawthorne (eds.) 2002, pp. 441-492. Wittgenstein, L., Tractatus logico-philosophicus (London: Routledge and Kegan Paul, 1922; Engl. tr. by D. Pears and B. McGuinness, 1961). ———, Philosophische Grammatik, ed. by R. Rhees (Oxford: Blackwell, 1969; English tr. by A. Kenny, Oxford: Blackwell, 1974). ———, Philosophische Untersuchungen, ed. by G. E. M. Anscombe and R. Rhees (Oxford: Blackwell, 1953). Wright, C., “The Conceivability of Naturalism,” in T. Szabó Gendler, J. Hawthorne (eds.) 2002, pp. 401-439. NOTES 1

See Yablo 1993, p.28.

2

In TP Rescher had adopted a more orthodox, though still considerably original view of the metaphysics of possible worlds.

3

Given an aporetic cluster, i.e. a set of mutually incompatible beliefs that are all prima facie well motivated, apory resolution consists in restoring consistency by forsaking at least one of the incompatible beliefs. See Rescher 1985, Ch.1., 2-3.

4

See, however, Lewis 1973, pp.65-77.

5

Rescher himself regards his approach to counterfactual evaluation as alternative to “closeness”—based approaches: see C, p. 171 ff.

6

Wittgenstein 1922, 3.02. Concerning the other side of the equivalence (possible thinkable) Wittgenstein is not equally explicit. It seems, however, that there is nothing to keep us from picturing a state of affairs (i.e. a possibility) except perhaps lacking names for its objects. It also seems, however, that we can help ourselves to all the names we may need (5.535).

7

Descartes’ chiliagon example is in Meditationes de prima philosophia, VI, AT 72. For the other examples see Tidman 1994, p. 299, Chalmers 2002, p.151, Fiocco 2007, p.367.

8

Notice another consequence of the limitation of conceivability to present conceivability, given the conceivability/possibility connection: as Caesar could not have imagined that salt = NaCl, so it was not possible for him that salt = NaCl. How-

193

RESCHER STUDIES

NOTES

ever, that such was the nature of salt was not just possible but actual for Caesar, as for everybody else. 9

In fact, for Tractatus all necessary falsehoods must turn out to be contradictions, once analyzed. —To be intelligible, a proposition must have sense, but contradictions lack sense (1922, 4.461). Or again, to understand a proposition means to know what is the case if it is true (4.024), but contradictions, being necessarily false, cannot possibly be true, as they do not depict possible facts.

10

Tidman 1994, p.308, Fiocco 2007, pp. 379-380. They both refer to Van Cleve 1983.

11

“It is easy to conceive, that in the infinite series of numbers, and intermediate fractions, some one number. . .may bear the same ratio to another, as the side of a square bears to its diagonal; yet, however conceivable this may be, it may be demonstrated to be impossible” (Reid 1969, p.434; see also p.435).

12

See Kripke 1980, pp.103-104.

13

I am adopting the convention on which

= the proposition that p.

14

See fn.3 above.

15

In his (1993), Yablo took it for granted that the inference from conceivable to possible failed in some cases, such as the case of necessary a posteriori propositions. His problem was whether conceivability was at all a reliable guide to possibility, or whether it did not provide even defeasible evidence of possibility, as others later claimed (e.g. Fiocco 2007, p.379, and (more guardedly) Tidman 1994, p.308).

16

See Garcia-Carpintero and Macia (eds.) 2006, Marconi 2005.

17

Yablo (2002) criticizes this notion and puts forth a different proposal: he says p is conceptually possible iff it could have turned out that p. For example, if it had turned out that ‘Hesperus’ referred to Mars, then it would have turned out that Hesperus Phosphorus (whereas in no circumstances could it have turned out that Hesperus Hesperus). I believe there are problems with Yablo’s suggestion; however, it is just for reasons of space that I do not discuss it here.

18

Wright actually discusses what he calls ‘the Counter-Conceivability principle’, i.e. the thesis that “a posteriori necessities, no less than a priori ones, are defeasible by lucid counterconception” (p.434). This is “Conc (~p) ~Nec (p)”, which is equivalent to the conceivability possibility inference.

19

Wright appears to believe, but doesn’t actually show that even in this case the second strategy is viable.

20

For the Tractatus, ‘salt NaCl’ is meaningless, both prima and secunda facie, as it is the negation of an identity statement trying to say what can only be shown. On

194

CONCEIVABILITY

NOTES

the other hand, what the statement tries to convey would not be regarded as necessarily false, for in the Tractatus every composition is inherently contingent.

195

Weird Worlds Robert K Meyer INTRODUCTION

F

or over 3 decades, I have been enamored of possible worlds semantics. I was inspired by Gottfried Wilhelm Leibniz (and Saul Kripke). But under the influence of my colleagues Richard (Routley) Sylvan and Val (Routley) Plumwood, I was emboldened in papers from [11] on to take seriously weird “possible worlds” in which contradictions are true but laws of logic are false. Still, the scare quotes around ‘possible worlds’ are there for a reason. Even a (classical) relevantist like me fears harsh judgment from his peers. Nor is it any fun being thought a nut by other logicians (although, as the self-appointed Maximum Leader of the Logicians Liberation League, I do not recoil completely from departures from sobriety). From time to time, I sobered up from the fantasy that there were yet more worlds to conquer. I would go on to explain that what we really meant were pieces of information (Urquhart [15]), setups (Routley & Routley [12]), theories (Fine [3], though we suggested them first in [11]) or states (your favorite cognitive or information or computer scientist). Thanks to my old Pitt teacher Nicholas Rescher (co-author of [9] with Pittsburgh’s Robert Brandom), I announce here and now my abandonment of such cravenness. For [9] introduces “schematic” possible worlds (the negation incomplete ones) and “superposed” possible worlds (the negation inconsistent ones). What had kept me craven, it is my shame to have to confess, was the “possible” in “possible world”. I had hitherto lacked the chutzpah of Mortensen [6], with its conviction that “anything is possible.” Making p true, and simultaneously making its negation ~p true in the same respect, had unfortunately unnerved me. Danke, Nick. Merci, Bob. 1. NEITHER HERE NOR THERE Before we get into the exciting business of making contradictions true (sort of), let us try the more familiar task of making excluded middles false (again, sort of). Recall Aristotle and tomorrow’s sea battle (not to mention

RESCHER STUDIES

Brouwer and his bevy of mathematical intuitionists, rejecting papers invoking the tertium non datur). Anyway, Aristotle was famously uncomfortable with either asserting or denying “There will be a sea battle tomorrow.” (And he was the guy who thought up the law of excluded middle!) The ground of his discomfort would seem then to have been simply that tomorrow’s sea battle was a future contingent, whence taking it as already definitely true or definitely false was jumping the gun. Like Aristotle’s, here is a case where we can both have and eat our cake. So [9] holds, and I agree. If neither p nor ~p has happened yet, the future may (so far as we know) be open regarding p. In that case, it would be churlish to already pronounce p either definitely true or definitely false. (Granted, some deterministic science such as physics {or astrology} may induce us to be so churlish. We ignore that point here.) But it is certainly not churlish to pronounce p ∨ ~p nonetheless true. We might, e. g., appeal to van Fraassen’s method of supervaluations to justify that choice. Or, more lazily, we can just go along with [9]. 2. SUPERPOSING WORLDS As in discussion of the Heisenberg uncertainty principle in quantum physics, we may think of “superposed” worlds as contradictory possible states of a physical system. Only measurement, one has heard, settles the actual state of the system, by “reducing the wave packet”. Candidate “attributes” exist as complementary pairs according to the Heisenberg relations, such as the position and momentum of a particular elementary particle. According to the popular Copenhagen interpretation of quantum physics due to Bohr, Heisenberg and others, exactness in the measurement of one of these attributes must be paid for in a corresponding imprecision in fixing the value of its complement. Against this, [9] recalls the Everett-Wheeler theory, which says that all possibilities are realized. Put otherwise, let w1 and w2 be possible worlds as usually conceived. Let p be true in w1 and ~p be true in w2. [9] enjoins us to take seriously also the superposed world w1 .∪. w2, in which each of p, ~p count as true. But wait! Does not logic then collapse? Doesn’t absolutely everything (or at least enough things to be sadly depressing) then count as true, in view of the lamentable theorem (p ∧ ~p) → q of Logic?

198

WEIRD WORLDS

2.1 DOUBLE EXPOSURE Well, No! By the lights of [9], the lamentable theorem does not in this case apply. True, we may (at least for the moment) be stuck with each of p, ~p, for some particular choice of p. That’s as far ([9] tells us) as its authors are willing to go! We certainly don’t want to be stuck with every q, which the lamentable theorem would force upon us. Indeed a large supply of q ∧ ~q would prove distressing. Even the simple conjunction p ∧ ~p (for our chosen p) is going too far. For Rescher and Brandom wish to preserve one premiss valid classical arguments, among which the argument from p∧~p to irrelevant q certainly figures. So what do they permit, among the sentences made true in superposed worlds? Think of it this way. Suppose that, out of your enthusiasm for p, you have photographed world w1. But then, sensing the appeal of ~p, you have on the same plate photographed world w2. The result is a double exposure. And, if you print it out, you’ll have a superposition. Here is some material about how messed up things can get. 2.2 MODES OF INCONSISTENCY How do I contradict myself? Let me count the ways ([9], p. 24). Below, the true-in-w predicate is in all cases tw. Here are some options. (Weak Inconsistency) To accept the prospect that for some genuinely possible world w: tw(P) and tw(~P), for some P. (Strong Inconsistency) To accept the prospect that for some genuinely possible world w: tw(P & ~P), for some P. (Hyperinconsistency) inely possible world w:

To accept the prospect that for some genu-

tw(P & ~P), for all P. (Logical Chaos) To accept the prospect that for some genuinely possible world w:

199

RESCHER STUDIES

tw(P), for all P (and accordingly tw(P) and tw(~P), for all P). [9] deplores (B)-(D), while being willing to put up with (and even to champion) (A). But is not this a rather strange divide? (A) and (B) seem but across a creek from each other; (B) and (C), by contrast, across an ocean. Granted, the (lamentable) ex contradictione quodlibet theorem p∧~p → q which was drilled into most of us when we were (logically speaking) impressionable youths has much to answer for. “If you’ve got one contradiction, you’ve got them all,” is (among other things) its lamentable motto. It is that lamentable theorem which makes (B) and (C) seem neighbors (indeed, it makes (B) and (D) seem neighbors). But (B) and (C) are not neighbors. For all the good reasons that Rescher and Brandom consider in [9] there are grounds to take contradictory pairs seriously. Indeed, there are even more grounds, as we shall see below. 3. UNDER THE HOOD We next peek under the hood (or bonnet, as we say Down Under) of the Rescher-Brandom Formula One Speedster (Possible Worlds model). While we welcomed above the ontological courage of the authors of [9], how does their bravery manifest itself via collections of formulae? When logicians get serious there tends to be something syntactic which underlies their metaphysical play. More specifically, as Bertrand Russell observed in [13], “almost all thinking that purports to be philosophical or logical consists in attributing to the world the properties of language.” We turn immediately to a consideration of theories, as collections of pieces of language that stand in for pieces of worlds. 3.1 THEORIES A collection S of (formal) sentences that hangs together logically is a theory. Yes, but what is it to “hang together”? The answer to which I have inclined from [11] on has been deeply indebted to my old mates Belnap and especially Dunn. Let ≤ be the binary relation of logical entailment on sentences, and let ∧ be ordinary conjunction. Definition 1. A set S is a theory if and only if it satisfies, for all sentences A and B, the following two conditions:

200

WEIRD WORLDS

≤E If A ≤ B then if A ∈ S then B ∈ S (Entailment closure) ∧I If A ∈ S and B ∈ S then A∧B ∈ S (Conjunction closure) A theory, in short, is what a lattice theorist would call a filter. Would Rescher agree that ≤E and ∧I characterize “hanging together logically”? At least where his contribution to [9] is concerned, he would probably not agree.. True, he should have no objection to ≤E. In fact, he could be a lot more specific, since [9] is content to take A ≤ B as classically indicated by the theoremhood of the material conditional A ⊃ B in classical 2-valued logic, henceforth 2. (We relevantists, by contrast, might prefer the theoremhood of A→B in the One True Logic as our criterion. Ah, yes, but what is the One True Logic? We shall return to this point in discussing Rescher’s [8] below.) Meanwhile let us consider the ∧I condition in Definition 1. Rescher and Brandom, I think, would be in two minds about that one. It is certainly satisfied by the sets of sentences true in what they call standard worlds. But [9] expunges ∧I with great haste when it comes to superposed worlds. Indeed, we saw in section 3.2 the distinction that [9] attempts between weak and strong inconsistency (these are (A) and (B) respectively above). That distinction is nothing but the rejection of ∧I for superposed formula sets! 3.2 SEMI-THEORIES So what would [9] accept, as a minimal condition that a bunch of sentences hang together logically? Definition 2. A set S is a semi-theory if and only if it satisfies, for all sentences A and B, the following condition: ≤E If A ≤ B then if A ∈ S then B ∈ S (Entailment closure) Of course! Even a superposed world w, as [9] sees it, should respect onepremiss (classical) entailments. 3.3 LEIBNIZ THEORIES But should we not, after all, prefer standard worlds, and the theories that correspond to them? No, not really, answers [9]. Granted, we can cook up our superposed and schematic worlds out of standard worlds. First, let w

201

RESCHER STUDIES

be the schematic world w1 .∩. w2, where w1 and w2 are standard. As [9] sees it this should make true in w exactly the sentences that are true in both of (the standard worlds) w1, w2. There is an obvious semi-theory residue of this thought. The truths at w are the intersection of the truths at the standard world w1 and the truths at its standard pal w2. As things now stand, those respective sets of truths are not merely semi-theories but theories. And since (as everybody knows) the class of theories is closed under arbitrary intersections, the truths at w constitute a theory. Let’s be a little more careful in characterizing the truths at a standard possible world. They will constitute, at the sentential level, what we’ll call an ultratheory U (corresponding to an ultrafilter in a corresponding Boolean algebra). Necessary and sufficient conditions for membership in U are the following (truth-like) conditions: T~.

~A ∈ U iff A ∉ U

T∧.

A∧B ∈ U iff (A∈U) ∧ (B∈U)

T∨.

A∨B ∈ U iff (A∈U) ∨ (B∈U)

T⊃.

A⊃B ∈ U iff (A∈U) ⊃ (B∈U)

T≡.

A≡B ∈ U iff (A∈U) ≡ (B∈U)

We have gone out of our way to indicate in each case that an analogous condition is imposed in (what [2] would call) the U-language on the right as the connective on the left suggests. (We could have made this clearer by expressing T~ as ~A ∈ U iff ~(A ∈ U). ) Can we do something similar at first order? Well, not really. Nothing daunted, let us do it anyway. Where A(x) is a formula and A(t) is the result of proper substitution of t for x therein, the ultratheory U will be moreover a Leibniz theory provided that it also satisfies, for all sentences, T∀.

∀xA(x) ∈ U iff A(t) ∈ U for all terms t

T∃.

∃xA(x) ∈ U iff A(t) ∈ U for some term t

202

WEIRD WORLDS

Caution: What we have said here isn’t quite what we mean. The quantifier clauses presuppose that what occurs on both the left and the right are sentences—i. e., formulae in which no variables occur free. Even then they only make sense for a substitution interpretation of the quantifiers—one for which every object in the intended domain of quantification has a name (which may, given the tricks that logicians play, be the object itself). We note that [9], which lets some really queer objects into its superposed and schematic worlds, has nasty words for substitutional quantification therein. Anyway, Definition 3. A set S of sentences is a Leibniz theory if and only if it is a theory, and satisfies moreover all of the above conditions T~, T∧, T∨, T⊃, T≡, T∀, T∃. So a Leibniz theory is a set of classical first order sentences which, when its syntactic primitives are interpreted, may be taken to describe a standard possible world. And now let us cook up (the syntactic residue of) all the Rescher-Brandom worlds out of Leibniz theories. 3.4 PUTTING IT ALL TOGETHER So here is the linguistic residue of the Rescher-Brandom story, as I see it. We have semi-theories, theories, and Leibniz theories. If we wish, we may begin with the Leibniz theories, which are (in the nice phrase of [1]) truth-like. As we have seen, we get schematic theories (corresponding to schematic worlds) by intersecting Leibniz theories. And we can now complete the story by taking unions of these, which will produce superposed theories (akin to superposed worlds). Now let us note how these maneuvers decrease our level of truthfunctional involvement. Either union or intersection of Leibniz theories will (in general) spoil T~. In the same vocabulary, distinct Leibniz theories T1 and T2 must differ with respect to some sentence A. Without loss of generality, A may be chosen to be an atomic sentence P, with P ∈ T1 and ~P ∈ T2. So the schematic theory T1 ∩ T2 will lose negation completeness; while the superposed semi-theory T1 ∪ T2 will sacrifice negation consistency. The reason for these outcomes, so to speak, is that intersection and union of Leibniz theories mess around also with the conditions T∨ and T∧ respectively. Evidently both T1 and T2 above contain the instance P∨~P of the Law of the Excluded Middle (along with all other instances of classical first-order theorems, as [9] points out). But intersections of theories are not

203

RESCHER STUDIES

necessarily ∨-prime. In particular, our schematic theory T1 ∩ T2 contains neither P nor ~P, violating T∨. Exactly dual remarks apply to the superposed semi-theory T1 ∪ T2 and the condition T∧. For this does contain each of P, ~P. But although the union of semi-theories definitely preserves the relation of entailment (and any other binary relations that hold in both), we have seen and commented above on its failure to preserve ∧-closure (which is, after all, half of T∧). 4. IS THE GAME WORTH THE CANDLE? [9] makes an enormous number of extremely trenchant points (whose surface we have only just begun to scratch here). Having made many of the same points myself, over the years, I can only applaud them. One does not hear quite so often, these days, about how wonderful a thing it is that contradictions entail everything—giving us, or so it used to be intimated, more reason to avoid them. That sounded too much like “Capital punishment for shoplifters!” That penalty, more or less in effect in late 18th century England, no doubt did discourage petty theft. But what more often happened was that petty thieves had their sentences commuted to transportation as convicts to Australia, whence those who did not fit into what Napoleon called a “nation of shopkeepers” came here instead to build a nation of shoplifters! Actually, the historical truth is that Australians became sensitive to “the convict stain.” Not having been put to death for truly bad reasons, we built instead a society that was exceptionally law-abiding (at least for a while). So indeed let it be with inconsistency. I also applaud [9] for its insistence that, ceteris paribus, inconsistency is a bad thing. But Right Reason is no less to be prized when it is under strain. Rather then we need it all the more! 4.1 BUT THE DEVIL IS IN THE DETAILS! Like Routley’s [11], taking note of a “Meinong revival” is part of the business of [9]. So is a respect for Wittgenstein’s willingness to put up with contradictions in [12]. Still, one needs to assess the specifics of the Rescher-Brandom plan for the Logic of Inconsistency. We are led down the garden path, with the thought that we can save classical logic, even in inconsistent situations.

204

WEIRD WORLDS

Can we? Well, maybe. As a classical relevantist, I incline that way myself, at least on Mondays, Wednesdays and Fridays. But today is Sunday, and it has 90 minutes to go. So today, I would like to see two classical negations in the Logic of Inconsistency. I am more than content to go along with [9] in using ‘~’ for the DeMorgan negation that goes with outright denial. But as in [13] I prefer that, when the authentic semantical classical negation stands up, it should be Boolean ¬. 4.2 RAINY DAYS AND MONDAYS Oops! Monday has struck, and its coming has left me feeling particularly classical. So I am feeling good about all of (p∧~p) ⊃ q, (p∧¬p) ⊃ q and even (p∧¬p) → q. But I am enough of a relevantist to remain aghast at (p∧~p) → q. A reason for this, no doubt, is that I remain attached to the adjunction rule ∧I above. Despite [9], I don’t think it suffices to be classical merely to accept the classical tautologies and the one-premiss entailments. Since Aristotle’s syllogistic, Logic has been (to the contrary) about combining premises for the sake of argument. So, having commended Rescher and Brandon for their ontic courage in taking superposed worlds as authentically possible, I am distressed at their unwillingness to proceed from taking each of p, ~p severally to their conjunction p∧~p. The ground of this unwillingness is not far to seek. [9] is classical enough to have preserved the inference from p∧~p to q. So, to avert explosion, [9] blocks the move from p, ~p severally to their conjunction p∧~p. That’s lamentable. 5. MANY-VALUED LOGIC? Professor Almeder has also asked me to comment on Rescher’s book [8] about many-valued logic. This too is a fine and useful volume. But, having devoted almost this entire essay to the inconsistent, I shall have to leave my comments about [8] mainly incomplete. 5.1 TRUTH VALUES? The many-valued logics of Vasil’ev, Łukasiewicz and their fellow innovators do raise interesting questions. Since it is possible to add others to the simple truth-values {true, false}, what are we to make of these extra values?

205

RESCHER STUDIES

The many-valued pioneers, [8] tells us, assimilated their explorations in non-Aristotelian logic to the groundbreaking explorations of 19th century mathematicians in non-Euclidean geometry. But the analogy only holds in part. Before the geometric radicals Gauss, Lobachevski, Bolyai and Riemann, one has heard that no alternative to Euclidean geometry was envisioned. On the other hand, Aristotle himself seemed to have had qualms about the principle that every sentence was exactly one of true, false. We have already asked about future contingents, such as “There will be a sea battle tomorrow”. It is at least arguable (and was so argued by Łukasiewicz—see [8]) that not even Aristotle was obliged to assign one of the standard truth-values to such statements. Anyway, talk of more truth values is merely scratching the surface. Who says that the only semantic values relevant to sentences are their truth or falsity? It was a daring logicist hypothesis (and no self-evident truth) that the extension of a sentence was its truth-value; and that, moreover, all that truly counts in sentential logic can be worked out from truth-functional composition of such truth-values. True, we have been teaching this daring hypothesis for (at least) the past century. It has caught on. Insofar as it has given rise to the many-valued logic surveyed in [8], the hypothesis has provided more—much more—of the same. When additional “truth-values” were introduced, they also were subjected to functional composition. But this does raise a problem. We can form pretty firm ideas about how classical true and false compose truth-functionally. The negation of a truth is false; of a falsehood, true. The conjunction of two truths is true; otherwise, conjunctions are false. Granted, the story gets more ad hoc when we get to disjunctions, while it involves downright special pleading to accommodate conditionals. But suppose, like Łukasiewicz [14], we desire just one more truthvalue. We’ll call it N, and think of it as intermediate between classical T and F. Let’s presume, as Łukasiewicz did, that the truth-functional behavior of T and F will be the same in the 3-valued logic as in 2. What are we to make of combinations in which N occurs, such as ~N and T ∧ N? Łukasiewicz had a definite plan on which he answered these questions. Speaking personally out of my own experience with some elementary logic classes, it seems to me that Łukasiewicz had the right plan. Or at least, on the 3 or 4 occasions when I asked my beginning logic students how they would fill in the 3-valued truth tables, what they produced were the Łukasiewicz truth tables. (I hope that my questions were not like those of

206

WEIRD WORLDS

Socrates to the slave boy in Plato’s Meno, excerpted in [15], eliciting the answers that were thought correct. At the very least, I had not consciously loaded the dice. For my own favorite 3-valued logic RM3 was produced by another Pole, Sobocinski [16].) Anyway, Rescher does not shrink when it comes to truth-functionality. For he holds that Probability Logic PL, as ordinarily conceived, is manyvalued logic but not truth-functional. [8] goes on to say (p. 188) that once ‘truth-functionality is dropped as an essential feature of a “logic” (as it certainly ought to be), there is no longer any basis for failing to recognize the probabilistic logic as a system of many-valued logic.’ Nonetheless, I am (like Rescher) much impressed with the fecundity of classical 2. But, having promised to return to it, there is still the question whether the One True Logic exists. 5.2 THE ONE TRUE LOGIC? We come in conclusion to the question whether there is, or is not, a preferred logic. Rescher finds ample ground to argue on both sides of this question in Chapter 3 of [8]. He notes immediately that the plurality of logical systems is a fact. So there is at least a prima facie case that there are many logics. Is, then, the question settled immediately on the side of pluralism, relativism and perhaps even irrationalism? Rescher responds with an emphatic “No.” He wishes rather (p. 215) to furnish a clear picture regarding the exact sense in which there are “different, competing systems of logic.” Rescher wishes to inquire whether, and to what degree, some systems of Logic are to be given preference over others. He wants to know specifically whether the grounds for such preference are empirical, pragmatic or something else entirely. Rescher divides the perspectives from which Logic may be viewed into the psychologistic, Platonistic and instrumentalistic. He finds the first two seriously wanting. But Logic, as [8] sees it, is above all an instrument (or, as Aristotle said, an organon). Rescher sees “doctrinal correctness as regards logic … to hinge upon that of instrumental superiority.” (p. 222) And so [8] finds the correct position “to lie somewhere in the middle—with the functionalistic or pragmatic version of the instrumental approach.” With this I agree. There is no One True Logic. But there is Our Logic— which is as firm a guide to right reason as a programmed lift-off is to take us to the moon. We do not make up for ourselves the laws of logic—any

207

RESCHER STUDIES

more than we can make up for ourselves the laws of physics. The world intrudes upon us, with its facts and its challenges. We respond with our theories and our formal systems, to explain our successful rational behavior and to guide us away from dumb mistakes. None of these theories—none of these systems—is uniquely privileged. Nor is any particular squiggle the unique formal counterpart of some intuitive notion. That’s not how it works. Said John Robinson, “I am very confident the Lord hath yet more light and truth to break forth from his word.” I retain that confidence.

REFERENCES Anderson, Alan R., Nuel Belnap and J. Michael Dunn, Entailment, vol. 2 (Princeton: Princeton University Press, 1992). Curry, Haskell B., Foundations of Mathematical Logic (New York: McGraw-Hill, 1963). Fine, Kit, “Models for Entailment,” Journal of Philosophical Logic, vol. 3 (1974), pp. 347-372; substantially reprinted in [1], 208-231. Łukasiewicz, Jan, “O logice trojwartosciowej,” (On 3-valued logic), Ruch Filozoficzny, vol. 5 (1920), pp. 169-171. (Tr. in S. McCall (ed.), Polish Logic: 1920-1939 (Oxford, 1967), pp. 16-18) Meyer, Robert K, “A Boolean-Valued Semantics for R,” Research Paper No. 4 (Department of Philosophy, RSSS, Australian National University, 1979). Mortensen, “Anything is possible,” Erkenntnis, vol. 30 (1989), pp. 319-37. Plato, “Meno,” http://www.cut-the-knot.org/proofs/half_sq.shtml Rescher, Nicholas, Many-Valued Logic (New York: McGraw-Hill, 1969). Rescher, Nicholas, and Robert Brandom, The Logic of Inconsistency, (Oxford: Basil Blackwell, 1980).

208

WEIRD WORLDS

Routley, Richard, “Exploring Meinong’s Jungle and Beyond,” Departmental Monograph No. 3 (Department of Philosophy, RSSS, Australian National University, 1980). Routley, R., and R. K. Meyer, “The Semantics of Entailment (I),”in H. Leblanc (ed.), Truth, Syntax & Modality (Amsterdam: N. Holland, 1973). Routley, R., and V. Routley, “Semantics of First-Degree Entailments,” Nous, vol. 6 (1972), pp. 335-359. Russell, Bertrand, “Vagueness,” Paper read before the Jowett Society of Oxford University in November, 1922. http://www.cscs.umich.edu/~crshalizi/Russell/vagueness/ Sobocinski, B., “Axiomatization of a Partial System of Three-Valued Calculus of Propositions,” The Journal of Computing Systems, vol. 1 (1952), pp. 23-55. Urquhart, Alasdair, “Semantics for Relevant Logics,” The Journal of Symbolic Logic, vol. 37 (1972), pp. 159-169; substantially reprinted in [1], section 47. Wittgenstein, Ludwig, Foundations of Mathematics (Oxford, 1956).

209

Nicholas Rescher on Greek Philosophy and the Syllogism Jürgen Mittelstrass Peter Schroeder-Heister 1. THE DISCOVERY OF RATIONALITY

G

reek philosophy has long fascinated philosophers and historians of philosophy alike. Here we will consider not a particular phase of the development of philosophy (or science) but its very foundation, at least in its European sense. Closely related to this is the idea that not only did a particular form of philosophy (and science) come into being (like, say, European philosophy), but that philosophy itself emerged according to its concept. Hence, if we identify the beginning of Greek thought with the beginning of reason, it becomes immediately clear that, in the broadest sense, terminological and methodological insights play a decisive role.1 Among these insights is the discovery of the possibility of theoretical propositions and proofs (Thales). In fact, it is both the geometrical (theoretical) propositions of Thales, which are copied by the general propositions of pre-Socratic philosophy, as well as the discovery of proof, later extended with the help of logical instruments (within the framework of axiomatic conceptions), which provide the first example of philosophical, i.e., rational thought and earned Greek thinking in general the validity that enabled it to step out of the shadow of myth once and for all. Plato, for example, based his theoretical philosophy on the insights of geometry and introduces logic as a philosophically fundamental discipline in the framework of his epistemological insights in the Theaetetus, his linguistic investigations in the Cratylus, and his theory of truth and falsehood in the Sophist. Aristotle outlines his theory of a demonstrative science, and thereby his model of exact science, with respect to the axiomatic structure of geometry already under development, and with his syllogistics he establishes logic as a theory of justification in the narrower sense. However, in contrast to Plato’s narrower definition of philosophy, which in terms of its subjects remains oriented on the geometric idea of ideality, Aristotle resumes the physical considerations of the pre-Socratic, especially Ionic, tra-

RESCHER STUDIES

ditions. It was only through Aristotle that this tradition became part of the prehistory of philosophy, and cosmology the roots of philosophy in the Greek and modern sense. Moreover, we have become used to understanding what we think about philosophical (and scientific) knowledge, by thinking in the language of Greek philosophy. Our concept of philosophical (and scientific) thought, as well as our concept of philosophical (and scientific) rationality both have a distinctly Greek history. Greek thought catches up with us not only historically—as an integral part of our cultural and historical consciousness—but also systematically: the form of our knowledge and our questioning is determined by rationality, the beginning of which we describe with Greek thought. A consequence of this is that we cannot think of ourselves outside (the form) of Greek thought. And this also means that we cannot reflect the origin of this thought from the outside, as distant and disinterested observers, but only from the inside, as a part of a cause-effect relation. Methodologically (and epistemologically) formulated: Our reconstructions of the Greek origin of philosophical (and scientific) thought and of the rationality which realizes itself within this thought, are unavoidably determined by those (conceptual) orientations that we borrow from what we are reconstructing. This need not be circular nor, methodologically speaking, ‘historicist’. In essence, the point is to understand that in (the development of) Greek philosophy and science, an essential part of our philosophical and scientific self-understanding is also enclosed within it. For if it is true that Greek thought catches up with us both historically and systematically within the framework of the questions posed here, and that our form of thinking is determined by its Greek form, then we are investigating ourselves when we inquire into Greek thought. This is not only relevant insofar as we do philosophy (and science), but also in that we generally identify efforts of rational orientations with the essential idea of man. In this sense, as members of a rational culture, we are still Greek, and in this sense there exists no alternative to Greek thought and the concepts of reason and rationality, which came into the world through Greek thought. To use the language of Greek philosophy: The beginning is the essence—the Greek origin of philosophical and scientific thought, and its concepts of reason and rationality, are its essence. It is, therefore, unsurprising that no philosopher can afford not to reflect on Greek philosophy, and that even the major philosophers continue to return to it again and again when posing questions, offering explanations, re-

212

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

constructing, and participating in other forms of philosophical debate. Nicholas Rescher is no exception. 2. COSMOS AND LOGOS The importance and unique status of Nicholas Rescher in modern philosophy is plain to see in the number of his voluminous books, such as A System of Pragmatic Idealism in three volumes (1992-1994) and Philosophical Reasoning (2001). In comparison, his numerous shorter works are often overlooked, but it seems that these greatly complement the picture of his great contribution to philosophy, which is nonetheless impressive in and of itself. These shorter works often seem to be just preliminary studies for a more comprehensive work. But, this impression is false, and does not do justice to these shorter works, which include books such as The Riddle of Existence (1984) and Epistemetrics (2006), and also other books such as Galen and the Syllogism (1966) and the collection Cosmos and Logos: Studies in Greek Philosophy (2005). Galen and the Syllogism offers a discussion of Galen and the fourth figure of the syllogism in the light of new data from Arabic sources, while Cosmos and Logos deals with various aspects of mainly pre-Socratic philosophy. In this collection, Rescher summarizes his evaluation of Greek philosophy. This will here be shown by a perusal through the contributions summarized in this volume and in the analysis of his important earlier contribution to Galen and the history of logic. In the studies found in Cosmos and Logos, we find masterstrokes of an interpretation which is argumentative, and which draws simultaneously on well-chosen source texts. This results in a high level of understanding that intentionally avoids much of the common literature analysis that normally suffocates an original approach. Here, Rescher follows two principles. The first of these principles, which is explicitly stated, is that of an interpretative restoration of a theory that directs historical analyses, particularly in case the textual basis and the tradition is poor.2 It corresponds to a concept of reconstruction, which can be defined as follows: A reconstruction is given or can be considered as successful or adequate, if a construction K′ not only reproduces K correctly in all its essential parts but also, at the same time, fulfils the intentions pursued by K better (at least not worse) than K3—and this in Rescher’s sense of supplementing information that makes the reconstructed argument more explicit.

213

RESCHER STUDIES

The second of these principles, which is implicitly expressed, says that historical clarity (obtained from historical reconstruction) presupposes systematic clarity. That is, where there is no systematic clarity, nothing is seen, not even when reading texts. This can be rephrased in the following way using a well-known formula of Kant4: historical ideas (Anschauungen) without systematical knowledge (Begriffe) are blind—thereby taking into consideration the other side of the formula that concepts without contents are empty, which means here that a systematical interest needs a substantial basis, in this case philosophical thought in its historical development. Genesis and validity are tied together—also in Rescher’s thought. Central to the studies in Cosmos and Logos we accordingly find not the author and his work, but instead terminological constellations and methodological conceptions, which—as the following analysis will show— constitute the entire thought style of a philosophical era: of Greek philosophy in its essential cosmological and epistemological components, and simultaneously, for the most part, determine the further (philosophical and scientific) development. Examples are the concepts of archē (as primary matter and as principle), of opposition, of thought experimentation, of scepticism, of rational choice, and of logical reasoning. 3. FROM EVOLUTIONARY THOUGHT TO LOGICAL ANALYSIS The first study is dedicated to cosmic evolution in Anaximander, and appeared before Charles H. Kahn’s important book on the Greek philosopher5 (the latter confirmed the former rather than rendering it outdated). The analysis is based on the assumption that Anaximander had an evolutionary concept of the development of the cosmos. The starting point here is the concept of apeiron, which is described and explained by means of Greek sources (Simplicius and others), with the result that the apeiron “lacks any quantitative as well as any qualitative definition”6. The correct conclusion is that “Anaximander’s apeiron is limitless or boundless, but not literally infinite.”7 This, again, could serve as an example of the Aristotelian concept of hypokeimenon, which also Simplicius employs in his description of Anaximander’s position.8 A spatial interpretation of the apeiron, which existed early on and which seems also to be shared by Rescher in his presentation of the cosmogonic process9, leads, however, to more difficulties than solutions insofar as the ancient ideas of a material primary substance are again thrown into relief (how can something, which is itself not spatial, limit something

214

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

that is?). An analysis of principles in the Aristotelian sense, which, in a way, is already present in Anaximander, again becomes an analysis of matter. However, it is still fascinating just how well Rescher succeeds in describing the evolution of Anaximander’s cosmos (cosmogony) and its structures (cosmology) in his recourse to material and spatial ideas. More importantly, though, as far as the systematic interest of philosophy in its own history (as well as the history of science) is concerned, is that the concept of scientific explanation10 gains high profile with Anaximander, since here, for the first time, the explanation of natural phenomena, i.e., astronomical and meterological phenomena, succeeds, and succeeds in the form of a strictly naturalistic account (“the founder of scientific cosmogony”11). It should be added—diverging from Rescher’s account—that this might also be valid in the sense of an analysis of principles when one interprets the concept of the apeiron not spatially, but in the sense of the Aristotelian concept of principle (archē)—whereas Aristotle curiously ignores Anaximander’s conception in his interpretation of the pre-Socratics12 which is oriented on the concept of the archē. The analysis of the concept of contrastive opposition, on the other hand, is treated much differently. This concept can be held as a central element not only of pre-Socratic philosophy, but also of Greek philosophy as a whole, including Plato and Aristotle: the opposite determination of objects (through the determination of opposite qualities) and their balance. In a precise manner Rescher presents various conceptions from Pythagoras through Aristotle, e.g., Heraclitus’ theory of the unity of opposites, together with the claim that the development of a theory of opposites demonstrates a natural unfolding of a key idea “through substantively interconnected stages”13. This, however, would be too bold of a claim. For one, these conceptions are too different, and they also refer to a wide range of physical, arithmetic, ontological and logical objects. The idea seems plausible that processes can be explained through the existence of opposites and their balance (a central motif in Greek thought) within theoretical as well as practical philosophy. This idea was supported by both empirical and theoretical evidence, whereas the logical meaning of the concept of the opposite used here, namely as the opposition between statements, when one and the same object is assigned two contradictory determinations, was not clarified until much later, probably not until Plato, namely within the framework of his theory of truth in the Sophist. The investigation into thought experimentation in pre-Socratic philosophy leads directly into methodology. Thought experimentation is seen by

215

RESCHER STUDIES

Rescher to be a cognitive procedure which can be defined as follows: “A ‘thought experiment’ is an attempt to draw instruction from a process of hypothetical reasoning that proceeds by eliciting the consequences of an hypothesis which, for aught that one actually knows to the contrary, may well be false. It consists in reasoning from a supposition that is not (or not yet) accepted as true—and perhaps is even known to be false—but is assumed provisionally in the interests of making a point or establishing a conclusion.”14 With this very clear definition, Rescher, using examples from pre-Socratic argumentation, shows that an important instrument of generating ideas is at hand, and that it consists at the same time of “different styles of argumentation and reasoning”15, which from the very beginning have determined the development of philosophical and scientific rationality. One after the other, Rescher proves, in concise analyses, that in pre-Socratic philosophy six various forms of thought experiments can be identified and reconstructed: explanatory conjectures (Thales), negatively demonstrative reasoning (Anaximander), reductio ad absurdum (the Pythagoreans), sceptical thought experimentation (Xenophanes), and analogical thought experimentation (Heraclitus). This last model, value dominance argumentation (Heraclitus), however, reinforces the fact that thought experimentation is a “cognitive instrument of substantial value”16. This procedure is, at the same time, typical of Rescher’s philosophical approach: clarity in the systematic approach leads to clarity in the philosophical interpretation. This is something that the so called hermeneutic philosophy still has to learn. On a much different level, Rescher deals with sophistic and sceptical streams of Greek philosophy. He attempts the rehabilitation of the sophists against Plato’s verdict that—with very few exceptions, such as Hegel— greatly influenced how the sophists had been viewed within the history of philosophy. The chapter opens with the sceptical tropes of Aenesidemus and begins anew with a brief definition: “They were apories, that is, paradoxical considerations designed to manifest the infeasibility of obtaining secure knowledge about the world. For such knowledge is supposed to be objective and impersonal—uniformly true and valid for all people— whereas the tropes indicate that such solidity is unrealizable in matters relating to this world of ours. It thus emerges that knowledge of reality is unattainable.”17 This definition allows us to describe a sceptical process which follows, on the one hand, Protagoras, i.e., his sceptically interpreted maxim (“man is the measure of all things”18), while following, on the other hand, the principles of perspectivity (isostheneia, “the equivalency of dif-

216

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

ferent views”), of the plausibility or similarity with truth (ta eikota, “the likely-seeming”) and of common sense (epochē, “the suspension of judgement and belieflessness”). At the same time, scepticism and sophistry are clearly differentiated both systematically and historically: “what the Sceptics urged was a mental disengagement (ataraxia), an epochē-geared detachment from judgmental thought in favor of living by ‘appearances.’ By contrast, what the Sophists proposed was: a commitment through interests and the requisites of effective praxis as mediated through our personal sense of desirability and value.”19 With this in mind, the way towards a more sympathetic understanding of sophist thought is open for Rescher. While the Sophists anticipate the anti-cognitivism of the Greek Sceptics, they at the same time avoid the cognitive nihilism of the Sceptics. Short commentaries on the nomos-physis distinction and its linguistic analysis in Plato’s Cratylus make this clear, whilst at the same time the epistemological basis is layed deeper. Here, Rescher is again led by a systematic interest, namely to preserve systematic thought from its own inborn naïvité, as expressed in epistemological dreams of omnipotence. The last two studies in this volume are dedicated again to logical and methodological themes. The chapter entitled “Anaximander, Aristotle and ‘Buridan’s Ass’” takes up the issue of the problem of choice without preference, as well as the suggestion that this problem has its roots in very different historical contexts: in the problem of physical symmetries (Anaximander, Plato, Aristotle), in the problem of how to explain God’s choices to reason and to human rationalization (Ghazālī, Averroës), and in the problem of man’s freedom of the will (Thomas Aquinas and others). Again, Rescher is not concerned about historical information as such, but instead about the systematic issue of a reasoned choice among equivalent alternatives, as relates to the difference between reasons and inclining motifs. The historical examples are well-chosen—these reveal the greater context in which this problem occurs in the history of philosophy—and simultaneously contribute significantly to the theory of rational choice, including the idea of randomness. The final study (“Aristotle on Ecthesis and Apodeictic Syllogisms”) rounds off the volume with a small masterpiece of logic. The focus here is on the method of ecthesis, which in Aristotle’s logic is presented as a third possibility, in addition to conversion and reductio ad absurdum, for reducing syllogisms back to those of the first figure. The question arises as to how far the principle of modal ecthesis, required in the proof of the modi baroco and bocardo with necessary premisses and necessary conclusion, is

217

RESCHER STUDIES

based on strictly logical considerations. This question is answered with reference to the fact that they are primarily metaphysical considerations, which here determine the argumentation and Aristotelian modal logic as a whole. Nonetheless, it is demonstrated that modal logic suffices for the same systematic demands as formal logic without modalities. This will be later on demonstrated in more detail (Section 6). A comparable task confronts Rescher in his short book on Galen and the syllogism. 4. GALEN AND THE FOURTH FIGURE In his Prior Analytics, Aristotle distinguished only three figures of categorical syllogisms. Based on this tripartition, the syllogistic moods of the fourth figure have often been treated as ‘indirect’ moods of the first figure with the terms in the conclusion interchanged. The introduction of the fourth figure as an inference schema in its own right was credited to Galen for many centuries, and was sometimes even called the Galenian figure.20 This attribution was called into question by Łukasiewicz21, and many historians of logic followed him in that respect. Rescher, in his book Galen and the Syllogism22, challenges this claim and argues that the attribution of the fourth figure to Galen is correct after all. To support this claim, he presents new evidence by providing an edition of a treatise by the Baghdad logician, physician and scientist Ibn al-Sarī, also known as Ibn al-Salāh (ca. 1090-1153)23 on the fourth figure of the syllogism. In this treatise, al-Salāh not only defends the fourth figure, but also refers to (today no longer extant) sources available to him, which credit it to Galen. The (Arabic) edition and annotated translation of an Istanbul manuscript of this treatise— copied in 1229 from a valid source and in the modern Western scientific community known to exist since at least the 1930s24—is a masterpiece of philological and historical scholarship.25 The ability to read and edit Arabic manuscripts is not normally something one would expect of a ‘universal’ philosopher but only of a specialist.26 The problem with ascribing to Galen the authorship of the fourth figure is that this claim rests only on indirect evidence. The extant writings of Galen tell us nothing directly about this issue. On the contrary, in his Eisagogē dialektikē (the only original Greek logic text known to have survived), he speaks of only three figures in the usual Aristotelian way.27 Therefore, in order not to blame him for inconsistency, Rescher has to explain why Galen can both speak of three and of four figures arising from categorical syllogisms with two premisses. In fact, this should have (at

218

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

least implicitly) been undertaken within the tradition that credited the fourth figure to Galen. This makes it necessary for Rescher to discuss systematically the very concept of a syllogism. In doing so, he ends up with a book that, in addition to its historical and philological parts, contains a thorough examination of the foundations of syllogistics. It deals with three issues: (1) the edition and translation of al-Salāh’s treatise, (2) the concept of a syllogism and the significance of the fourth figure, (3) the problem of crediting Galen with the authorship of the fourth figure. Issues (1) and (2) can be studied independently of one another and of (3), whereas (3) relies on both (1) and (2). Rescher’s explanation as to why Galen may have considered both a three-figure and a four-figure treatment of syllogistics, is based on a distinction he draws between two conceptions of syllogisms, which lead to different views of syllogistic figures. Rescher distinguishes between (i) the Aristotelian two-premiss and (ii) the later two-premiss-cum-conclusion views of the syllogism, which may be described as follows. (i) Given two premisses with one term shared between them (the middle term, M), we may pose the syllogistic question, asking what follows from them with respect to the two terms which are not shared by the premisses (the extreme terms, S and P). We would then distinguish between: I.

M occurring as the subject term in one premiss and as the predicate term in the other,

II.

M occurring as the predicate term in both premisses,

III. M occurring as the subject term in both premisses. Using x, y and z as variables for the syllogistic judgement forms a, i, e and o, this leads to the three syllogistic figures:

219

RESCHER STUDIES

I. MxP SyM Z

II. PxM SyM Z

III. MxP MyS Z

The task of syllogistic theory would then be to establish which judgements Z of the form S z P or P z S are valid consequences of the respective premisses, depending on the judgement forms a, e, i or o assigned to x and y. Whether Z takes the form S z P or P z S is nothing that affects the distinction between the different forms. In fact, in figures II and III we can, without loss of generality, assume that Z takes the form S z P, as all conclusions of the form P z S are obtained from the same figure by exchanging the order of the premisses, which is a logically insignificant operation. As the distinction between consequences of the form S z P and P z S is a subordinate matter even for the first figure, Aristotle is right to treat figure-I inferences with the conclusion P z S as special (secondary, indirect) modes of the first figure. They are secondary, as the figure-I modes with S z P as their conclusion are self-evident28 and superior to all other forms of reasoning. (ii) Given two premisses with a middle term M and extreme terms S and P, we may instead ask whether a conclusion of a particular form follows from them. In other words, we take the form of the conclusion to be part of the question. Therefore, in the case of the first figure, we do not just ask: “What follows from M x P and S y M ?” but rather, for a given judgement Z: “Does Z follow from M x P and S y M ?” As the form of Z is now relevant, this question must be divided from the very beginning into the two questions: “Does S z P follow from M x P and S y M ?” and “Does P z S follow from M x P and S y M ?”. This leads to splitting up the first figure into two figures, the first (in the new sense) and the fourth: I. MxP SyM SzP

IV. PxM MyS SzP

There is no difference with respect to the second and the third figure. Whereas figures I and IV are duals of each other with respect to permut-

220

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

ing premisses and converting the conclusion, figures II and III are selfdual. This view of the syllogistic figures has remained the standard. From a modern point of view, the difference between (i) and (ii) is that between a consequence-set view (“What are the consequences of two given premisses?”) and a consequence-relation view (“Which consequence relations between two given premisses and a given conclusion are valid?”). Aristotle’s consequence-set view leads to a coarser classification than the medieval (and modern) consequence-relation view. Rescher not only draws a conceptual distinction between these views, but also finds a terminological distinction which reflects this: A syllogistic figure, according to the three-figure view, is called a schēma by Aristotle (syzygia by Alexander of Aphrodisias), whereas according to the fourfigure view, a figure is called systasis (symplokē by Alexander) and the corresponding syllogisms technai29, with corresponding distinctions in Arabic. While only the first terminology is present in the extant texts of Galen (where he speaks of three figures), the second terminology occurs (in Greek) in his listings of his own works, and its Arabic counterpart is used in the Arabic texts which associate him with the fourth figure. Thus, Rescher makes the ascription of the fourth figure to Galen plausible by attributing to him the explicit awareness of these two views of the categorical syllogism.30 Not only those who, like Rescher, credit Galen with the fourth figure, have to explain how a three-figure and a four-figure view may coexist. Those who, like Łukasiewicz, deny Galen’s authorship, must do so as well, in order to comply with the sources who claim his authorship (such as, e.g., Averroës), if they do not wish to simply disregard them as not trustworthy. Their way of arguing is that Galen, when speaking of a fourth figure, does not mean the fourth figure in the standard sense, but something different, e.g., syllogisms with more than two premises.31 In any case, the ascription of the fourth figure to Galen is based on indirect sources and on the weight given to them, as none of the relevant papers by Galen are known to have survived, and even the text edited by Rescher does not unequivocally show this, as is demonstrated by Sabra’s independently published commentary on this text.

221

RESCHER STUDIES

4.1 AL-SALĀH’S TREATISE: RESCHER VS. SABRA The very same Istanbul manuscript, which is the topic of Rescher’s book, was published independently as a photographic reprint with a detailed commentary by the renowned historian of Arabic science A. I. Sabra.32 It appeared a year earlier than Rescher’s book and roughly at the same time as Rescher’s preliminary report33, but not before Rescher’s book was completed.34 In contradistinction to Rescher, Sabra follows Łukasiewicz in that Galen did not invent the fourth figure and interprets the references to Galen and the fourth figure in the introduction to al-Salāh’s treatise (which is the basis of Rescher’s historical argument) accordingly. In the introduction to this treatise35, al-Salāh makes the following four points which Rescher uses as arguments for Galen’s authorship of the fourth figure: (1) There is a commentary by Abū ‘l-Faraj ibn al-Tayyib on the Prior Analytics, which criticizes Galen’s introduction of the fourth figure. (2) A treatise by Ahmad ibn al-Tayyib al-Sarakhsī refers to a report mentioned to his master al-Kindī about the existence of a Syriac treatise of Galen dealing with the syllogistic figures. Al-Kindī explicitly rejects the fourth figure. (3) There is said to be a treatise by al-Fārābī criticizing the fourth figure. (4) There is a treatise by someone named Dinhā the Priest, entitled The Fourth Figure of Galen, which is full of errors. Sabra deals with these points in the following way: (ad 1) This is acknowledged, but not given decisive evidence.36 (ad 2) This interpretation suffers from the vagueness of an Arabic expression. It is not clear from the text whether al-Kindī’s rejection

222

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

of the fourth figure presupposes a claim by Galen against which it is directed.37 (ad 3) There is no surviving manuscript by al-Fārābī which supports this claim.38 Actually, Rescher was hoping that the edition of further manuscripts of al-Fārābī might bring to light some reference to Galen and the fourth figure. This has not materialized so far.39 (ad 4) The identity of Dinhā is not clear. Sabra does not further comment on this source. He would probably call into question its reliability, as apparently Dinhā is not a logician. It may be regretted that neither Rescher nor Sabra returned to the subject later. The merit of Sabra’s paper is his extended commentary on the full text. Rescher essentially discusses its introduction, as this is the passage relevant to Galen, and considers the rest “largely a routine”40. However, the main text does contain some interesting observations on the fourth figure that may have fit into Rescher’s general discussion on the merits of this figure (see below). In general, it is somewhat surprising how little attention Rescher’s and Sabra’s discussions of al-Salāh’s treatise received by the experts in the field. In the philosophical literature, there is no reference to Sabra’s paper whatsoever, and Rescher’s is, for the most part, quoted vaguely in the style of “for a discussion of Galen and the origin of the fourth figure see Rescher”. Even in the Arabistic literature there are references to Rescher and Sabra only here and then.41 Given that there was so little general interest in the topic, the coincidence of these two projects which made al-Salāh’s treatise accessible to the public in 1965 is even more remarkable. 5. RESCHER AND THE FOURTH FIGURE Though the consequence-relation-view of syllogistics, which gives the fourth figure its proper place, has dominated since medieval times, the fourth figure has nevertheless often been considered inferior to the other figures. Rescher discusses six arguments given in the literature to support this claim. He discards all of them and concludes with an argument of his own to establish it. Unfortunately, he does not deal with arguments which explicitly support the fourth figure (as compared to, say the second and

223

RESCHER STUDIES

third), although that would have fit well into his book, as al-Salāh’s text is one that does exactly this. In the following, we discuss Rescher’s reply to three traditional objections against the fourth figure (objections 2, 3 and 4 in his enumeration), as they are the most interesting from the systematic point of view. His replies to the three other objections, though historically highly relevant, are fairly obvious. Objections 2, 3 and 4 all deal with the deductive power of certain moods with respect to certain inference rules. (1) Objection 2 (stated in many traditional textbooks): The fourth figure is the first figure with an inverted conclusion. Rescher correctly replies that this is not true. One has to invert both the premisses and the conclusion to reach figure IV from figure I. Rescher correctly remarks that the true underlying reason for this objection is that only conversion is needed to deduce figure IV from figure I (which is traditionally considered the perfect figure), whereas without reductio, figures II and III cannot be obtained from I. So according to Rescher, figure IV may at best be considered deductively ‘less independent’ of figure I than figures II or III. However, even that depends on the fact that conversion is considered a weaker rule than reductio, as reductio alone suffices to generate figures II and III from I, so with respect to the number of rules required, figure IV is as near to figure I as are figures II and III. That conversion is weaker than reductio might actually be claimed as conversion can be justified using reductio, if a limiting case such as S a S is admitted as a premiss.42 (2) Objection 3 (due to Leibniz): Figure IV is farther away from figure I than figures II and III, since in addition to reductio, conversion must be used to derive it. Rescher replies that Leibniz’s assumption that conversion is less natural than reductio, is not warranted at all, and refers to the derivability of conversion from reductio (see above). (3) Objection 4 (credited by Rescher to Kneale & Kneale43): The fourth figure is inferior to the other figures, as reductio does not lead out of it. Rescher replies that this is no more than a fact. He does not mention that it is due to a distinguishing characteristic of the fourth figure, namely that the two occurrences of the three terms involved always occur in opposite positions—one on the left

224

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

and one on the right. This feature may actually give the fourth figure a superior quality (see Section 5.2). Overall, according to Rescher, the objections to the fourth figure describe, if they are based on correct observations, mere features of this figure, without rendering it inferior. Nevertheless, Rescher shares the opinion that the fourth figure is indeed inferior to the others, but on the basis of an argument which seems to him “to be the only one that posseses genuine force”44. It is based on the insufficient deductive power of the fourth figure, given the basic Aristotelian inferences to derive syllogistic moods from others (and in the end to reduce all valid moods to the moods of the first figure): conversion and reductio . Using these two inference principles, the valid moods of each of the figures I, II and III suffice to derive the valid moods of all four figures, whereas this is not true for the valid moods of figure IV. The moods barbara-I, baroco-II and bocardo-III cannot be reduced to the fourth figure by means of conversion and reductio alone. “It is yet another of the ironies of the history of logic that this (to our mind solitarily appropriate) ground for maintaining the inferiority of the fourth syllogistic figure is of our own devising and is nowhere to be found in the ramified literature of the subject, despite the innumerable discussions aimed at showing the inferiority of the fourth figure.”45 Rescher’s systematic conclusion, then, is that the fourth figure, though indispensable, is “somewhat less central or fundamental”46 than the others. Rescher attempts to put the traditional view of the inferiority of the fourth figure on a firm base by comparing its deductive power with that of the other figures. 5.1 AL-SALĀH’S ARGUMENTS IN FAVOUR OF FIGURE IV As Rescher wants to dissociate his own objection to figure IV—which is not an argument against the figure as such, but against its deductive power—from other objections found in the literature, he does not discuss arguments in favour of figure IV. However, as they are the topic of the work edited, it is interesting to discuss al-Salāh’s arguments. Al-Salāh makes at least three points: The first one is in favour of figure IV as a genuine syllogistic figure, the second and third are arguments for the superiority of figure IV over figures II and III. He even uses a new numbering system, calling figure IV the second, and the (traditional) figures II and III the third and fourth figures, respectively.47 The status of figure I as the fig-

225

RESCHER STUDIES

ure of the self-evident or self-justifying perfect syllogisms is not questioned. (1) Al-Salāh’s central point in favour of figure IV is that it covers forms of reasoning that actually occur and as such must be addressed. Its reducibility to figure I is no argument, as figures II and III are reducible to figure I as well. Reducibility does not mean that the figure need not be considered. Al-Salāh argues that once we make a distinction between the two premisses of a syllogism, we are automatically committed to distinguishing between the different positions of the middle term separating figures I and IV. To support this view, he gives a nice analogy: When we classify bodies into heavy and celestial, and correspondingly motion into straight and circular, and then associate heavy bodies with straight motion towards the center, then with qualifying straight motion into the two options towards the center and its opposite from the center, we are automatically committed to a corresponding division of the noncelestial bodies into heavy and light ones. This is a strong argument in favour of the consequence-relation view of syllogisms and thus for considering figure IV as a genuine form of reasoning. (2) Figure IV is superior to figures II and III, as it allows for three judgement forms e, i and o in its conclusion, whereas figure II only allows for e and o, and figure III only for i and o. Figure I allows for all four forms a, e, i, and o. Thus the syllogistic figures are ranked according to the number of judgement forms possible in their conclusion. This is an argument concerning the expressive power, but not the deductive power, of figure IV. It is certainly a distinctive feature of this figure, but not necessarily one that makes this figure superior. (3) As only conversion is required to pass from figure I to figure IV, figure IV is nearer to figure I than are figures II and III. This is closely related to the objection discussed at the beginning of Section 5. It relies on the assumption that conversion is a more natural form of reasoning than reductio. However, even if the distance of figure IV from figure I is smaller than that of figures II and III from I, we do not have a linear order. Figure IV does not lie in between figures I and II/III in strength.

226

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

5.2

Merrill’s Ignored Characteristic of Figure IV

At the end of chapter II, which contains the systematic discussion of figure IV, Rescher refers to a paper by Merrill (1965) that establishes the superiority of figure IV over the other figures and thus calls into question Rescher’s own stance as elaborated upon in this very chapter. This footnote was obviously added after the book was completed, perhaps even added after the book was in print. It says: “It is, moreover, undoubtedly the case that there are other systematic considerations—of perhaps a kind that seems artificial from the traditional approach to syllogistic logic—which militate favourably on behalf of the fourth figure. See, for example, D. D. Merrill [Reference to Merrill (1965)].” The rather moderate and indirect way of expressing himself: “perhaps a kind that seems artificial” shows that Rescher was aware that Merrill’s paper challenges his basic systematic claim, namely that the fourth figure is inferior due to its lack of deductive power. Reading Merrill’s paper, it turns out that it puts forward an argument which supports a position opposite Rescher’s (and at odds with the whole tradition of syllogistic theory). Quoting the paper and pointing to its significance demonstrates Rescher’s intellectual honesty, as no one could or would have complained about not quoting a paper which appeared after the book was in print.48 From today’s point of view, this is even more remarkable. Merrill’s paper, though highly relevant, short, clear, and to the point, has not been taken notice of at all, although Mind is one of the leading philosophical journals. As of now (December 2007), the WebSci and Philosopher’s Index databases list no reference to this paper in any journal. Not even Thom in his extensive book on the syllogism49 mentions it. We could not examine all of the books ever written about syllogistics, but Rescher’s quotation is the only one we could find, although more then forty years have passed. It is unbelievable, how a most significant systematic contribution to syllogistics could remain unnoticed.50 Merrill’s approach rests on the use of obversion rules, which were not introduced into the formal apparatus of syllogistics until the 19th century.51 In this sense, Rescher is right to say that an approach based on obversion is “artificial from the traditional approach to syllogistic logic”. On the other hand, as the idea of concept-negation which underlies obversion, is already present in Aristotle’s work52 (though not within his syllogistic formalism),

227

RESCHER STUDIES

Rescher is also right in not repudiating such an approach outright. This is even more appropriate, as Rescher’s own of method of assessing the significance of a syllogistic figure in terms of its strength to axiomatize syllogistics is not the most traditional way of treating syllogistic figures, either. A proper traditional approach would always put the allegedly selfjustifying ‘perfect’ moods of figure I first, which Rescher does not. As Merrill’s paper is highly relevant, we sketch its content here and relate it to Rescher’s approach. Suppose we have obversion rules at our disposal, which allow us to pass from a judgement to its contrary by changing a term to its complementary term, such as “All men are mortal” to “All men are not immortal” (i.e., “No man is immortal”), or “Some animals are mammals” to “Some animals are not non-mammals”. Denoting the term complementary to P by P’, and assuming that P’’ yields P, obversion justifies the following equivalences: S a P ⇔ S e P’

S e P ⇔ S a P’

S i P ⇔ S o P’

S o P ⇔ S i P’

Call a semi-syllogism an inference figure, which looks like a syllogism, with the only difference being that any occurrence of a term S, M or P may be replaced by its complementary term S’, M’ or P’, respectively. Note that we speak of occurrences of terms. This means that, in a semi-syllogism, a term may occur in one position in its plain form and in another in its complementary form. For example, P i M S’ a M’ S i P would be a semi-syllogism of the second figure (which is, of course, invalid). Using obversion, any syllogism whatsoever can be converted into a semi-syllogism, in which only the judgement forms e and i occur. As eand i-judgements are convertible, this semi-syllogism can be transformed into a semi-syllogism of any figure by conversion (where premisses must be permuted if the conclusion is converted). Now only the fourth and no other figure has the property that by means of obversion, any semisyllogism of a given figure can be transformed into a syllogism of the same figure. This is due to the fact that only in the fourth figure do the two occurrences of a term always occur in opposite positions (something on

228

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

which Leibniz’s objection to the fourth figure is based, i.e. something which was considered a negative feature of figure IV, see (2) at the beginning of Section 5). This means, in particular, that one of the occurrences must occur in the predicate position and therefore obversion is permitted. This holds equally for valid and for invalid moods. Consider, for example, barbara-I: MaP SaM SaP Obversion of all three judgements gives M e P’ S e M’ S e P’ from which by conversion of both premisses we obtain the following semisyllogism of the fourth figure P’ e M M’ e S S e P’ From this, by obversion of the first premiss, we obtain P’ a M’ M’ e S S e P’ which is an instance of calemes-IV. If, in this example, we replace a and e with o and i, respectively, we obtain a reduction of the invalid mood ooo-I of the first figure to the invalid mood oii-IV of the fourth figure. In this way, Merrill shows that the valid moods of the fourth figure axiomatize the valid moods of all other figures by means of conversion and obversion only. Moreover, the invalid moods of the fourth figure axiomatize the invalid moods of all other figures. Axiomatization of the invalid moods goes beyond the classical picture of syllogistics, although it is not

229

RESCHER STUDIES

too far from it, since most reduction procedures studied in classical syllogistics are symmetrical and independent of the validity of the moods considered.53 If obversion is considered an admissible way of reasoning, this shows that the fourth figure is not inferior, but indeed superior to figures I, II and III. Actually, for the reduction of the valid moods of any figure to those of IV, conversion restricted to the premisses (conv) and obversion (obv) is needed. For the reduction of the valid moods of any figure to those of one of the figures I, II or III, we need reductio (red) plus unrestricted conversion (convx).54 Rescher’s claim that the fourth figure is inferior can only be upheld if red+convx is considered less elementary or less natural than obv+conv. This essentially amounts to the question of whether obversion is considered an elementary mode of inference. Merrill’s paper demonstrates again that all claims based on the deductive power of syllogistic moods or figures strongly depend on the transformation rules assumed, and that the appropriate choice of such rules is not a trivial matter at all. 6. ECTHESIS, MODAL SYLLOGISTICS, AND THE METAPHYSICAL BASIS OF LOGIC In the final paper of Cosmos and Logos55, Rescher extends the principle of ecthesis introduced by Aristotle in his assertoric syllogistics, to the modal realm, something which by Aristotle is only sketched in a few lines of text, but not developed in detail.56 Using this extension, he is able to explain certain features of this system which have been a stumbling block to many prominent authors dealing with the subject. Ecthesis, in the standard non-modal sense, turns a particular judgement into two universal judgements, namely SiP SoP

into into

N a S and N a P N a S and N e P

claiming that a term N can always be found. Actually, in the first case, for N we can choose the term (S∩P), whose extension is the intersection of the extensions of S and P, whereas in the second case, we can choose the term (S\P), whose extension is the difference of the extensions of S and P.57 In the modal case, ecthesis turns an apodictic particular judgement into two apodictic universal judgements, namely □(S i P)

230

into

□(N a S) and □(N a P)

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

□(S o P)

into

□(N a S) and □(N e P)

As in the assertoric case, this can be validated by assuming N to be (S∩P) and (S\P), respectively. However, Rescher wants to characterize N not just extensionally, but by means of a choice operator which from S and P selects a natural kind, for which the universal apodictic assertions hold true. This is then metaphysically interpreted as expressing Aristotle’s tenet that “science, since it deals with the necessary, cannot but deal with the universal as well”58, i.e., a necessary particular judgement like □(S i P) or □(S o P) becomes a scientific judgement by turning it into necessary universal judgements in the way indicated. This is, of course, a highly metaphysical claim linking logic with metaphysics. It cannot rest on logic alone, as the assumption that there is a natural kind N with the mentioned properties is not a purely logical postulate. In order to be able to “construct”59 such an N, Rescher introduces a bracket notation [SP] for terms S and P such that, if all S are P, then all S are necessarily [SP]. Thus Rescher uses as an axiom the principle

(1)

SaP □(S a [SP])

His second axiom is

(2)

□(S a P) □([QS] a P) 60

From (1) and (2) together with corresponding principles for i- and ejudgements SiP (1’) □(S i [SP])

□(S e P) (2’) □([QS] e P)

he is able to derive the valid moods of the first figure of Aristotle’s modal syllogistics and show that the invalid moods are not derivable.61 This includes the validity of the (infamous) barbara LXL □(M a P) SaM □(S a P)

231

RESCHER STUDIES

and the underivability of barbara XLL MaP □(S a M) □(S a P) Modal ecthesis is then axiomatized using the axiom

(3)

□(S i P) (∃Q)(□([SQ] a S) ∧ □([SQ] a P))

and analogously for o-judgements.62 So the intuitive idea is that there is some predicate Q constituting a natural kind [SQ], such that from the particular modal judgement we obtain two universal modal judgements with the subject term [SQ]. By adding the standard reduction rules, this constitutes an axiomatization of the full system of Aristotle’s modal syllogistics. Though as a formal system this is impressive, its intuitive interpretation is not entirely clear. The ecthesis axiom (3) suggests that in [SQ] the Q denotes a property essential to all S’s, i.e., a differentia specifica singling out in a natural way those S’s that are also P’s. This view also underlies the reading of [QS] in (2): If in some possible world a [QS] did not have the property S, then we could not infer from □(S a P) that it has the property P. On the other hand, the term [SP] in axiom (1) is interpreted differently, as here [SP] denotes the natural kind N within P which is specified in a way such as to make □(S a N) deducible from the non-modal statement S a P. The reading of [SP] with P being essential to S would naturally validate the judgement □([SP] a P) which, together with (1), makes it possible to derive □(S a P) from S a P, thus trivializing modal logic. Though the system is formally consistent, Rescher’s bracket terms are semantically ambiguous.63 This is not surprising, as in Aristotle’s modal syllogistics the seemingly trivial judgement □(S a S) cannot be valid. This is due to the fact that using barbara LXL, □(S a S) trivializes modal logic.64 In Rescher’s axiomatization, this is obtained by applying (2) to the premiss □(S a S), yielding the trivializing judgement □([SP] a P) mentioned above.65 From this point of view, Rescher’s axiomatization of modal syllogistics using bracket terms is only formally, but not semantically, significant. However, there is a way out of this in keeping with the essentialist interpretation that guides Rescher’s formalization, though it means sacrificing

232

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

certain parts of Aristotelian modal syllogistics. We interpret “every S is necessarily P” not as the de dicto statement “it is necessary that every S is P” (□(S a P), in modern terminology □∀x(Sx → Px)), but as the genuinely essentialist de re statement “every S has P as a necessary (= essential) property” (S a □P, in modern terminology ∀x(Sx → □Px)). Then it is obvious that S a □S is not universally valid, as not every S that we pick in the actual world need to be an S in every other possible world. The bracket term [SP] now denotes a rigid concept66 which in the actual world has the extension of (S∩P). Then every object which, in the actual world, has the property [SP] (and thus the properties S and P), has the very same property [SP] (but not always the properties S or Q) in every other possible world. Aristotle’s basic metaphysical claim expressed in axiom (1) would then be that whenever S a P holds in the actual world, there is some natural property [SP], which every S (picked in the actual world) has in the actual as well as in every other world. Axiom (2) is validated likewise: If every S (picked in the actual world) has the property P in all possible worlds (i.e., S a □P holds true), then every [QS] (picked in the actual world), as being an S in the actual world, has the property P in all possible worlds.67 Correspondingly, the validity of barbara LXL and the invalidity of barbara XLL result naturally. Obviously, this view invalidates all parts of Aristotelian modal syllogistics that rest on conversion laws, as S i □P does no longer imply P i □S. However, these parts of modal syllogistics can more easily be discarded68 than Aristotle’s general metaphysical views whose intimate relation to logic is pointed out by Rescher. Converting the title of a book by Michael Dummett69, we may speak of the metaphysical basis of logic which is evident in Aristotle’s modal syllogistics. We would like to thank Perdita Rösch and Thomas Piecha for bibliographical assistance.

REFERENCES Becker, A., Die aristotelische Theorie der Möglichkeitsschlüsse: Eine logisch-philologische Untersuchung der Kapitel 13–22 von Aristoteles’ Analytica Priora I (Berlin: Junker und Duennhaupt, 1933; reprint Darmstadt: Wissenschaftliche Buchgesellschaft, 1968).

233

RESCHER STUDIES

Dummett, M., The Logical Basis of Metaphysics (London: Duckworth and Cambridge Mass.: Harvard University Press, 1991). Geach, P. T., “Review of Rescher (1966),” Mind, vol. 77 (1968), pp. 144145. Jager, M., “Review of Rescher (1966),” Foundations of Language, vol. 6 (1970), pp. 104-105. Kahn, Ch. H., Anaximander and the Origins of Greek Cosmology (New York: Columbia University Press, 1960). Kneale, W. & Kneale, M., The Development of Logic (Oxford: Clarendon Press, 1962). Lagerlund, H., Modal Syllogistics in the Middle Ages (Leiden and Boston Mass. and Cologne: Brill, 2000). Lameer, J., Al-Fārābī and Aristotelian Syllogistics: Greek Theory and Islamic Practice (Leiden, New York and Cologne: Brill, 1994). Lameer, J., “The Organon of Aristotle in the Medieval Oriental and Occidental Traditions,” Journal of the American Oriental Society, vol. 116 (1996), pp. 90-98. (Review article of Ch. Burnett (ed.), Glosses and Commentaries on Aristotelian Logical Texts: The Syriac, Arabic and Medieval Latin Traditions, London: The Warburg Institute, 1993). Łukasiewicz, J., Aristotle’s Syllogistic from the Standpoint of Modern Formal Logic, 2nd edition (Oxford: Clarendon Press, 1957). McCall, S., Aristotle’s Modal Syllogisms (Amsterdam: North-Holland Publishing Company, 1963). Menne, A., “Obversion,” Historisches Woerterbuch der Philosophie, vol. VI (Basel: Schwabe, 1984), pp. 1089-1090. Merrill, D., “Reduction to the Fourth Figure,” Mind, vol. 74 (1965), pp. 6670.

234

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

Mittelstrass, J., “Scientific Rationality and Its Reconstruction,” in: N. Rescher (ed.), Reason and Rationality in Natural Science: A Group of Essays (Lanham Md. and New York and London: University Press of America, 1985), pp. 83-102. Mittelstrass, J., “Griechische Anfaenge des wissenschaftlichen Denkens,” in: G. Damschen, R. Enskat, and A. G. Vigo (eds.), Platon and Aristoteles—sub ratione veritatis (Goettingen: Vandenhoeck & Ruprecht, 2003), pp. 134-157. Nortmann, U., Modale Syllogismen, mögliche Welten, Essentialismus: Eine Analyse der aristotelischen Modallogik (Berlin and New York: de Gruyter, 1996). Parks, R. Z., “On Formalizing Aristotle’s Theory of Modal Syllogisms,” Notre Dame Journal of Formal Logic, vol. 13 (1972), pp. 385-386. Pingree, D., “Review of Rescher (1966),” Speculum vol. 43 (1968), pp. 188-189. Rescher, N., Al-Fārābī’s Short Commentary on Aristotle’s Prior Analytics: Translated from the Original Arabic with Introduction and Notes (Pittsburgh: University of Pittsburgh Press, 1963). Rescher, N., “New Light from Arabic Sources on Galen and the Fourth Figure of the Syllogism,” Journal of the History of Philosophy, vol. 3 (1965), pp. 27-41. Rescher, N., Galen and the Syllogism: An Examination of the Thesis that Galen Originated the Fourth Figure of the Syllogism in the Light of New Data from Arabic Sources including An Arabic Text Edition and Annotated Translation of Ibn al-Salāh’s Treatise “On the Fourth Figure of the Categorical Syllogism” (Pittsburgh: University of Pittsburgh Press, 1966). Rescher, N. and Parks, Z., “A New Approach to Aristotle’s Apodeictic Syllogisms,” Review of Metaphysics vol. 24 (1971), pp. 678-689.

235

RESCHER STUDIES

Rescher, N., Cosmos and Logos: Studies in Greek Philosophy (Frankfurt: Ontos Verlag, 2005). Sabra, A. T., “Review of Rescher (1963),” Journal of the American Oriental Society, vol. 85 (1965), pp. 241-243. Sabra, A. T., “A Twelfth-Century Defence of the Fourth Figure of the Syllogism,” Journal of the Warburg and Courtauld Institutes, vol. 28 (1965), pp. 14-28. Street, T., “On Studying Medieval Arabic Logic,” Journal of the American Oriental Society vol. 117 (1997), pp. 536-541. (Review article of Lameer 1994.) Thom, P., The Syllogism (Munich: Philosophia Verlag, 1981). Thomas, I., “Review of Rescher (1966),” Philosophy of Science, vol. 34 (1967), pp. 198-200. VS = Diels, H., Die Fragmente der Vorsokratiker: Griechisch und Deutsch (Berlin: Weidmann, 1903, vols. I-III, 6th edition, ed. W. Kranz. Berlin: Weidmann, 1903). NOTES 1

See Mittelstrass (2003), 134-157.

2

Rescher (2005), 1.

3

Mittelstrass (1985), 92.

4

Critique of Pure Reason B 75.

5

Kahn (1960).

6

Rescher (2005), 3.

7

Ibid.

8

In Phys. 24, 13f. (VS 12 A 9).

9

Rescher (2005), 9.

10

Rescher (2005), 24.

236

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

NOTES 11

Rescher (2005), 26.

12

Met. A3.983b6ff.

13

Rescher (2005), 44.

14

Rescher (2005), 47.

15

Rescher (2005), 60.

16

Ibid.

17

Rescher (2005), 63.

18

Plato, Theaet. 152a; Aristotle, Met. K6.1062b13-14.

19

Rescher (2005), 76.

20

This view was propagated by Averroës (see Sabra 1965b, 14).

21

Łukasiewicz (1957), § 14 (38-42).

22

Rescher (1966). For reviews of this book see Thomas (1967), Geach (1968), Pingree (1968), Jager (1970).

23

Al-Salāh died in Damascus in A.H. 548 = A.D. 1153-54. See Sabra (1965b), 15.

24

See Sabra (1965b), 14-16.

25

It is a striking coincidence that the same manuscript was published independently at the same time by A. I. Sabra (1965b). See the next section.

26

Actually, this book is not Rescher's first publication of this kind—see Rescher (1963).

27

Rescher (1966), 3-4.

28

“Perfected through themselves” in Aristotle's terminology (An. pr. A5.26b26-35, 29b6-8).

29

Rescher (1966), 17-19.

30

Jager (1970) argues that the ancient Greek tradition after Galen should be taken into account in order to make this “admittedly indirect and circumstantial evidence” (Rescher 1966, 17) more conclusive.

31

Rescher (1966), 2. This was first claimed by Łukasiewicz (1957, 39), who refers to an anonymous scholium on Ammonius' commentary on the Prior Analytics.

237

RESCHER STUDIES

NOTES 32

Sabra (1965b). Sabra did his doctorate with K. R. Popper in London and later became Professor of the History of Arabic Science at Harvard University. Actually, he wrote a (not so favorable) Review of Rescher (1963), see Sabra (1965a).

33

Rescher (1965).

34

The preface of Rescher (1966) dates from February 1965. We owe the Sabra (1965b) reference to Thomas Piecha.

35

Rescher (1966), 52-54; Sabra (1965b), 16-18.

36

Sabra (1965b), 18.

37

Ibid.

38

Sabra (1965b), 18f.

39

Lameer (1994, 1996); Street (1997).

40

Rescher (1966), 50.

41

For example, in Lameer (1994).

42

Use datisi-III and cesare-II, which themselves can be obtained from figure I using reductio.

43

Kneale & Kneale (1962), 101.

44

Rescher (1966), 47.

45

Ibid.

46

Rescher (1966), 48.

47

In the following, to avoid any confusion, we use the traditional numbering throughout.

48

The preface dates from February 1965, and the first 1965 issue of Mind was available in the U.S. probably considerably later than this date.

49

Thom (1981).

50

In an e-mail, Professor Merrill confirmed to us that he knows of no other quotation beyond that of Rescher's.

51

Menne (1984).

52

De Int. 10.20a19-26.

238

NICHOLAS RESCHER ON GREEK PHILOSOPHY AND THE SYLLOGISM

NOTES 53

Exceptions are ecthesis and rules which involve subalternation, such as conversio per accidens.

54

Here, the “x” stands for the permutation of premisses, which is needed when the conclusion is converted, as, for example, in the reduction of bocardo-III to darii-I.

55

“Aristotle on Ecthesis and Apodeictic Syllogisms”. This paper is a revised version of a paper co-authored by Z. Parks (Rescher & Parks 1971) entitled “A new approach to Aristotle's apodeictic syllogisms”. Some of the revisions are substantial, so the serious student of Aristotelian modal syllogistics should consult the 1971 paper as well. Unfortunately, the 2005 edition also contains several serious misprints.

56

An. pr. A9.30a6-14.

57

We disregard the problem of existential import, which becomes relevant when reading these inference steps from right to left.

58

Rescher (2005), 126.

59

“Construction” here is understood in the metaphysical sense, not in any “constructivist” way of thinking.

60

Rescher (2005), 117. In Rescher & Parks (1971, 680), this axiom is just presented as such, whereas in Rescher (2005) it is reduced to its nonmodal variant, assuming the (Aristotelian) principle that every assertoric inference yields a corresponding apodictic inference (which, in modern terminology, comes down to assuming the principles of necessitation and distribution).

61

Rescher (2005), 117-120.

62

Rescher (2005), 122f.

63

This is reflected in the terminology used by Rescher and the different ways in which he describes the meaning of [SP]. He speaks of [SP] “to represent the Pspecies of S” (1971, 678; 2005, 115, text in angle brackets only in 2005), of [SP]'s as “specifically those S's that are by nature P's”, as “those S's which must be P's in virtue of their being S's (i.e. by conditional or relative necessitation)” (116, “must” emphasized in 2005), of [SP] “such as to validate the inference (1)” (text in angle brackets added 2005), as “the P-oriented subspecies of S (consisting of S's that are P's by virtue of being S's)” (only 2005), of [SP]'s as “P's with a relative necessity, subject to the condition of their being S's” (only 1971), as “S's-that-in-fact-are-P's” (only 1971, footnote 4). Concerning □([SP] a P), Rescher just writes in 2005 that this is “unattainable in the circumstances” (Rescher 2005, footnote 3), without giving any further reason. (It is clear that it must be unattainable.) Rescher gives the impression of an author still struggling to make explicit what he has in mind. He vacillates between claiming that [SP] is a subspecies of S with P being a specifying term, and considering it a

239

RESCHER STUDIES

NOTES

species that is specified in some other way, though it contains the same objects. Our proposal at the end of this section attempts to give [SP] a precise meaning. 64

As has often been stated, Aristotle is silent about that result, although he may have been aware of it. See McCall (1963), 50.

65

As a potential way out, the 1971 paper (Rescher & Parks 1971, footnote 2) contemplates an observation put forward by its coauthor in Parks (1972), where he argues that an identification of variables as in □(S a S) is never considered by Aristotle. However, in the end, Rescher & Parks (1971, footnote 7) prefer accepting the invalidity of □(S a S). This is also correct from the perspective of our reconstruction below. Changing the substitution rule appears ad hoc without systematic warrant.

66

Rigid in the sense that its extension is the same in all possible worlds. For simplicity, we assume the domain of objects to be constant throughout possible worlds.

67

This is now a direct semantic justification of axiom (2). Rescher's justification in the 2005 paper, which relies on the validity of the assertoric counterpart of (2), is no longer valid under the essentialist reading of universal apodictic judgements.— The idea of interpreting apodictic judgements as de re statements is not new. Its modern history goes back to Becker (1933, see also Nortmann 1996). The problems which apodictic judgements pose to the conversion laws (see below) were a central point in medieval discussions of modal syllogistics (Lagerlund 2000).

68

Namely by attributing them to a lack of distinction between de re and de dicto modalities.

69

Dummett (1991).

240

Rescher’s Evolutionary Epistemology Jesús Mosterín

P

hilosophy of science can be applied to the analysis of the theory of evolution, but the theory of evolution itself can also be applied to the analysis of philosophy of science (and indeed of culture in general). This second task (the application of the theory of evolution to philosophy of science) can be understood and carried out in two different ways: (1) Making use of the theory of evolution to account for the structure and function of our cognitive apparatus, and (2) making use of the theory of evolution to account for the evolution of the scientific theories and methods, and of science itself. Konrad Lorenz provided an explanation of our cognitive capacities based on the Darwinian theory of evolution. By so doing, he retrieved well-known Kantian thesis, although turning them outside down and transforming the Kantian individual a priori in the biological a posteriori of the species. Rupert Riedl, Gerhard Vollmer and others have followed this path. Karl Popper developed an evolutionary conception of our cognitive apparatus, identified knowledge with expectation and adaptation, and maintained that most of knowledge is congenital.1 Donald Campbell, Popper, David Hull and others have used the theory of evolution to explain the evolution of scientific theories. These authors have emphasized the remarkable structural analogies between biological evolution and the evolution of scientific knowledge. Popper underlined the method of trial and error, and the subsequent elimination of error, as the way to articulate the principles of Darwinian evolution in their application to the evolution of science. Campbell emphasized blind variation and selective retention. It is in this context that Popper developed the theory of world 3, devoted to the objective contents of thought processes, theory plagued by an unresolved tension between Darwinism and Platonism. 1. SYSTEMIC PHILOSOPHY Modern science requires for its advance an ever increasing specialization of their practitioners. The special researcher tends to know more and more about less and less, until he (or she) knows almost everything about almost nothing. Science is supposed to offer some sort of mirror of reality,

RESCHER STUDIES

but, as a consequence of extreme specialization, the mirror is broken down into a thousand disparate pieces. One important task of philosophers and scientists is to try to reconstruct the whole mirror from the broken pieces, and to be able to offer a unitary, global and coherent world view. Not in vain Plato characterized the philosopher as “the one with the overview” (synoptikòs).2 Classical philosophers, such as Plato, Aristotle or Descartes, tried to articulate a global world view in the form of a philosophical system. This system was supposed to be encyclopedic in scope, internally coherent, externally compatible with the results of current science and satisfying the methodological criteria of the particular thinker. The majority of contemporary philosophers have abdicated the ambition to construct such a system. Only a few, like Bertrand Russell, Mario Bunge and Nicholas Rescher, have dealt with many diverse fields of philosophy and beyond, and have kept alive the torch of universal curiosity and classical, systematic philosophy. Bunge has always felt special esteem for Rescher as a philosopher. At a meeting of the Institut International de Philosophie held in Moscow in 1993, challenged to mention a philosopher better than Hintikka and several other proposed candidates, Bunge mentioned Rescher. In the World Congress of Philosophy held in Boston in 1998, the analytical philosopher Georg Von Wright declared that enough juice had already been squeezed from the analysis of language, and that the time was ripe for shifting interest to the new philosophical challenges coming out of science. Few ideas have challenged, enlightened and ignited the contemporary philosophical scene as much as the Darwinian theory of evolution. It suffices to witness the commercial success of authors at the frontier between evolutionary biology and philosophy, like Richard Dawkins or Daniel Dennett, and the heated discussions in the press and even in American politics on these issues. In particular, a whole direction of philosophical inquiry has developed under the names of evolutionary epistemology and evolutionary psychology. ‘Analytical philosophy’ in the broad sense implies little more than a striving for clarity and precision of language and for logical cogency of argumentation. Obviously, Rescher can be considered an analytical philosopher in this broad sense. But he is not a mere analytical philosopher; he is also a classical, systematic philosopher, attuned to the new developments in science and aware of the powerful explanatory role that evolution can play in epistemology.

242

RESCHER’S EVOLUTIONARY EPISTEMOLOGY

2. EVOLUTIONARY EPISTEMOLOGY Rescher published Methodological Pragmatism: A Systems-Theoretic Approach to the Theory of Knowledge in 1977. Especially, chapters VIII (on “Evolutionary Epistemology”) and IX (“A Critique of Thesis Darwinism”) were directly devoted to the analysis and partial criticism of evolutionary explanations in epistemology. In 1990 Rescher published A Useful Inheritance: Evolutionary Aspects of the Theory of Knowledge, whose chapter 2 retakes the arguments of chapter IX of the previous work, and where other chapters also contain material from previous books. As Rescher himself has written, “one of the inevitable drawbacks of producing a systematic work in philosophy [. . .] is that one cannot avoid letting each book overlap to some extent with some others”. [1977, p. XIV] Taken together, these works develop an important contribution to evolutionary epistemology. Rescher got seriously involved with evolutionary epistemology, but was not much concerned with the application of the Darwinian theory to the evolution of culture in general, a subject developed by Richard Dawkins, Dan Sperber, Luca Cavalli-Sforza, Marc Feldman, Robert Boyd, Peter Richerson and others. The first of these authors coined the term ‘meme’ for the units or elementary chunks of cultural information. The term does not seem to appear in the works of Rescher I have read. (I have not read all of them; probably no one else has, with the exception of the prolific author himself). Science is of course a part of culture, and evolutionary epistemology cannot be kept separated from the evolutionary theory of culture. Repeatedly, Rescher makes a clear distinction between the evolutionary explanation of the structure and function of our cognitive apparatus (what he calls “cognitive Darwinism”) and the different evolutionary explanation of the development of our cognitive contents (notions, thesis, theories, methods and procedures). What he has to say about the first (“knowledge as a tool for survival”) is rather standard and straightforward. What he writes about the second is somehow more contentious. I will limit my comments to some of these contentious issues. 3. FORCES AS EXPLANATORY FACTORS In mechanics there is a clear distinction between kinematics (the description of the movements of bodies) and dynamics (the explanation of those movements). In order to explain movement, we have to introduce

243

RESCHER STUDIES

forces, conceived as explanatory factors, and we have to offer theories of those forces. One important force is gravitation, other one is electromagnetism, and another is the strong nuclear force. General relativity is the theory of gravitation; quantum electrodynamics is the theory of electromagnetism, and quantum chromodynamics is the theory of the strong nuclear force. Shifting from physics to biology, the analogue of kinematics is paleontology, which is a descriptive endeavor devoid of explanatory functions. Paleontology describes which live forms existed at different past moments (as revealed by fossils found in old strata), and consequently how live forms change in time, i.e., evolve. Evolution in this sense is a plain fact, and no explanation or theory is involved. In order to provide explanations for evolutionary changes, we have to introduce forces (again conceived as explanatory factors), like natural selection and drift. Darwinian evolutionary theory is basically the theory of natural selection. Kimura’s neutral theory is the theory of drift; it asserts that the majority of evolutionary changes at the molecular level are caused not by Darwinian natural selection but by random drift of selectively random mutants. 3 Initially controversial, neutral theory has gained widespread acceptance in genetics and evolutionary studies, as more detailed molecular knowledge about genomes has become available. There is no contradiction, but complementarity, between the Darwinian theory that explains adaptive evolution and the neutral theory that explains nonadaptive molecular evolution. 4. CULTURAL EVOLUTION In the case of culture, there is also a clear distinction between the descriptive history of culture, which is the analogue of kinematics and paleontology, and the attempts at theoretical explanation, i.e., the attempts to construct a cultural dynamics. If any explanations are to be provided, forces of cultural evolution have to be introduced, forces like random drift, rational selection and coercion, among others. Rescher maintains that, while genetic or biological variation is produced by random mutations, cultural variation arises through purposeful, directed activity. He also holds that biological natural selection is blind while cultural selection is rational: “Biological variation is undoubtedly Darwinian, with teleological blind natural selection operating with respect to teleologically blind random mutations. Cultural evolution, on the other hand, is generally Teilhardian, governed by a rationally-guided selection among purposefully devised mutational variations”4. This is doubtful. Think of

244

RESCHER’S EVOLUTIONARY EPISTEMOLOGY

linguistic evolution –not the biological evolution of the linguistic capacity, but the cultural evolution of particular languages, like the processes that slowly transformed Indo-European into Latin and Latin into French and Spanish, and that go on slowly changing languages all the time. The pronunciation of words is transmitted by learning and imitation with frequent errors in the transmission. These errors are unconscious and not purposeful. Nevertheless in the course of time they accumulate and transform languages. The same can be said of grammar and lexicon. Or, to take a quite different case, think of the cultural history of religious distribution. Drastic changes in the adoption of religion are often the result of wars of conquest, the victorious army imposing its own religion by shear coercion. The north of Africa was Christian until the Arab armies conquered it. Now it is Muslim. Catholicism arrived in Mexico with the Spanish Conquistadores. The line dividing Catholics and Orthodox Christians in the ex-Yugoslavia is the result of the battles between the Austrian and Russian (or Turk) armies. When people have changed religion in the past, it has seldom been a matter of rational decision; it has more often been a question of submission to an invading army. Of course some individuals convert voluntarily to a new religion, like Richard Gere to Buddhism, but that is rather the exception than the rule. Summarizing, there are several forces driving cultural evolution, and rational selection is just one of them. 5. SCIENTIFIC EVOLUTION Rescher could be tempted to concede the role of non rational forces in culture at large, but still insist in its prevalence in the particular parcel of culture that is science. Indeed rational selection is more important in science than, say, in fashion. But even focusing our attention on science, it is doubtful that a “Teilhardian” evolution takes place, “governed by a rationally-guided selection among purposefully devised mutational variations”. In his brilliant criticism of Karl Popper’s evolutionary epistemology, Rescher repeatedly attacks Popper’s idea that hypothesis are somehow random and his exclusive reliance on hypothesis elimination to account for the progress of science. I have discussed this issue with Popper, and he did not insist on randomness. Popper just thought that the character of the hypothesis (random or purposeful) did not matter: a random hypothesis could come out true, and a purposeful one could well be false. So, the best we could do would be to concentrate our efforts on the testing. In the jargon of the contexts of discovery and justification, he wanted to regulate provi-

245

RESCHER STUDIES

sional justification, but leave quite open and free the context of discovery. August Kekulé said that he had come to his famous discovery of the ring shape of the benzene molecule in a day-dream of a snake seizing its own tail. Popper would not care about Kekulés’s dream, as long as his hypothesis stood up to severe and repeated tests. I agree with Rescher’s point that we possess an inborn heuristic capacity for induction; this point is an adequate corrective to Popper’s excessive polemics against induction. While Rescher is right is pointing to the heuristic value of induction, Popper is also right in denying it any value in the validation of hypothesis. Whatever relative and provisional validation there is, it must come from repeated tests, experiments and observations. Science (I mean, empirical science) is not an enterprise grounded on reason, as mathematics is and as 17th century rationalism imagined. Science is based on mistrust of reason and reliance on experience. Of course, the whole of science is a rational endeavor that makes constant use of reason. But mere plausibility, clarity or evidence of ideas is not enough to accept a thesis in the standard model of a science. Let’s consider the development of cosmology. Every philosopher, physicist and astronomer thought that the universe was something static and repetitive like a clock. No one (no philosopher or scientist, no prophet or poet) had imagined that the universe, far from being a clock, was an explosion. The unsettling discovery was made by an observational astronomer, Edwin Hubble, in 1929. He did not like his discovery, and kept looking for alternative explanations of the data, but the explosion or expansion of the universe was here to stay. Many people disliked the incipient big bang model and greeted with relief the steady state universe model of the early fifties. Again, the unexpected and serendipitous discovery of the cosmic background radiation by Arno Penzias and Robert Wilson in 1965 caused the ruin of the cherished steady state model, to the consternation of many people and of Penzias and Wilson themselves. At the end of the 20th century there was an almost complete consensus in the scientific community that the expansion rate of the universe was decreasing, due to the pull of gravity. In 1998, Saul Perlmutter and Brian Schmidt published the results of their precise measurements of the distances to certain far away supernovae of type 1a. Unexpectedly, these measurements implied that the expansion of the universe is accelerating, not decelerating, as the consensus of the astronomers assumed. We could go on. Sometimes cosmologists have purposefully devised hypothesis that later happened to be right, but many other times new observations forced the adoption of hypothesis no one had purposefully thought of and

246

RESCHER’S EVOLUTIONARY EPISTEMOLOGY

no one liked. When nature cannot be directly asked, scientists are at a lost. Superstring theory –notoriously unconnected to any empirical checking— allows now for 10500 different solutions, which is a fantastic number, especially for a theory looking for a unique solution. In the course of his criticism of Popper for being unable to account for scientific progress, Rescher repeats that there are many more hypothesis than biological forms (or mutations): “But genetic mutations are of limited variety—unlike the possibilities for hypothesis formation.”5 This again is doubtful. All possible hypotheses are expressed in some language, for example, in English. The set of all possible hypotheses expressed in English is a subset of the set of all possible combinations of the letters and other typographical signs of the enlarged Latin alphabet. That entails that the cardinality of the set of all hypothesis is at most denumerable. There are at most as many possible hypothesis as natural numbers. But the set of all possible mutations and of all possible genetic forms is also a subset of the set of all combinations of the 4-letter alphabet of DNA, and consequently, there are at most as many different mutations as natural numbers. The fact of having 4 letters, or 26, or 80, is irrelevant, as long as the number of letters is finite. All languages based on a finite alphabet can be reformulated equivalently in a binary language, based on just two letters (0 and 1, say). Of course, linguistic formulations can be extremely complex and long, but so can be genomes. There is no a priori reason to think that the number of possible hypothesis is greater than the number of possible genetic forms. 6. EXTRATERRESTRIAL LIFE AND INTELLIGENCE One of the most appealing aspects of Rescher’s thought is his deep interest in the tantalizing possibility that alien creatures on far away planets might have developed a kind of science quite different from ours. I share his fascination for this issue. I have been a paying member of the Planetary Society and have visited some of the main venues of the search for extraterrestrial intelligence, like the radio telescope in Arecibo (Puerto Rico), which is conducting this research on the background all the time. In 1984 Rescher published The Limits of Science, whose chapter 11 is devoted to “extraterrestrial science”. Similar preoccupations, themes and arguments appear again in chapter 5 of A Useful Inheritance (1990). I agree with Rescher that an extraterrestrial form of science, if it exists at all, is bound to be quite different and probably even incomparable with ours.

247

RESCHER STUDIES

The existence of life elsewhere in the universe is a necessary condition for the existence of extraterrestrial intelligence. But the notion of life in general is quite unclear, and will remain so as long as we do not detect, encounter or contact alien forms of “life”. All life on Earth is very similar. (Yes, jellyfish and humans are quite similar below the surface; actually, we descend from them and still carry some of their genes). This similarity rests on descent from a common ancestor, something we would certainly not share with alien life forms on other planetary systems; they could be utterly different. The laws of physics and chemistry are obeyed everywhere in the universe (as far as we can tell), but they are compatible with an infinity of different courses of events. Many intricate features shared by all living organisms on Earth are not the result of any physical laws or any chemical or structural constraints. The only explanation of their ubiquitous presence in terrestrial organisms is that these features are inherited from a common ancestor. All living creatures on our planet share many characteristics, from the basic chemistry of organic molecules dissolved in liquid water to the code for translating nucleic acid information into protein structure. We do not know which of those features result from universal regularities and which ones are just contingent accidents developed by chance and inherited. Is life a cosmic joke or a cosmic imperative? Is life an extremely rare exception in an abiotic universe, or is life an extremely probable occurrence in a universe teeming with life? We do not know, and something more than armchair thinking is necessary to settle this question. Life on Earth is a chemical form of life, and, as chemistry is completely based on the electromagnetic interactions between atoms, life on Earth is based on the electromagnetic force. Alien life forms could be based on gravitation or the strong nuclear force. More than 95 % (by weight) of living matter on Earth is composed of the four elements hydrogen, oxygen, carbon and nitrogen. That could be otherwise. Life as we know it on Earth is based on carbon compounds. Some scientists think that any life will be based on carbon, but the element silicon has been suggested as an alternative, as silicon can also combine with four hydrogen atoms, like carbon. Life as we know it is based on water as a solvent. Liquid water provides a stable medium in which organic molecules can dissolve and interact. Life reactions take place in water, their products and wastes are carried by water, and living beings (bacteria or humans) are 70 % water by weight. Solvents other than water are conceivable in alien life processes. A solvent must of course be able to dissolve other chemical compounds. Among the

248

RESCHER’S EVOLUTIONARY EPISTEMOLOGY

plausible alternative candidates liquid ammonia (NH3), methyl alcohol (CH4O) and liquid hydrogen fluoride (HF) have been suggested. To be useful as a solvent for life on other planets, any solvent must remain liquid within a large range of temperatures typical of planet surfaces, so that the temperature variations on those planets do not make the solvent freeze or boil. Most of our life molecules are made up of a small number of types of rather simple organic molecules, monomers, joined together in complex molecules called polymers. Among the polymers are the proteins (made of amino acids) and the nucleic acids (made of nucleotides). Proteins consist of a backbone of repetitive elements, the same for all amino acids, and a series of different side chains (peculiar to each amino acid). There is an immense variety of possible amino acids, as many as organic side chains. So it is surprising that just 20 kinds of amino acids are used by life on Earth to make proteins. Many other kinds of amino acids exist. The amino acids are asymmetric around their central carbon atom, so they have chirality or hand. Each amino acid can occur in two versions: right-handed and left-handed. All the amino-acids found in life on Earth (with the exception of glycine, which has no hand) are left-handed, never right-handed. The amino acids obtained in the Miller-Urey type experiments, and the amino acids found in meteorites, are of mixed hand, about 50 % left-handed and 50 % right-handed. The left-handedness of biogenic amino acids arose by chance and was preserved by descent. In alien life proteins could be made from different types of amino acids than on Earth, and the number of amino acid types to choose from need not have been 20, it could be 15 or 63 or 210. Only twenty amino acids occur commonly in life as we know it. Still, an average protein molecule, containing something like 100 amino acids, could be made in 20100 (= 10130) different ways, producing a correspondingly huge variety of proteins (more than the number of elementary particles in the observable universe). Most organisms make and use fewer than 100,000 types of protein molecules. So life on Earth is very selective and uses only a small fraction of the molecules it could use. It is not to be expected that any other life form, even if protein-based, would make exactly the same narrow selection. All this, assuming the alien life form would be build out of proteins; of course, it could be otherwise. Reproduction and inheritance of structure require a system of information storage and transmission, which (in Earth's organisms) is provided by the double helix strand of DNA. DNA is a nucleic acid, a polymer made up

249

RESCHER STUDIES

of nucleotides. Another nucleic acid is RNA. These nucleic acids consist of a backbone of repetitive elements (a sugar and a phosphate), the same for all nucleotides, and a series of different nitrogenous bases (peculiar to each nucleotide). There are only four types of nucleotides in DNA. Many other different bases could possibly be used to build nucleotides, but life has chosen just those four types. So a four-letter code was fixed by the last common ancestor and subsequently inherited by all living organisms. An extraterrestrial life form is unlikely to have the same biochemistry, the same genetic code for encoding information in a DNA-like double strand of nucleotides and the same proteins made up from the 20 standard amino acid types as on Earth, all of them left-handed. All energy produced by living things on Earth by the processes of fermentation, respiration and photosynthesis is immediately saved in the form of ATP (adenosine triphosphate), a nucleotide with three phosphates attached. The cell invests the energy obtained in attaching the phosphates. When these bonds are broken by hydrolysis, the free energy is released and put to useful work wherever needed. ATP is the universal energy currency of life on Earth. All living beings, from bacteria to plants to people, invest their newly gained energy into ATP, and use the ATP to obtain the free energy they need for all work done by their cells. If there is life somewhere else than on Earth, it can hardly be expected to rely on ATP the way we terrestrial organisms do. In the evolution of life on Earth, the most difficult step seems to have been the formation of the eukaryotic cell. At least it took the longest time (about 2 billion years). Once eukaryotic cells were available, the pace of evolution accelerated. The production of mammals and men did not present especial difficulties (at least it did not take such a long time). Measured in terms of time needed, to create human beings from chimpanzeelike creatures was 400 times easier than to create amoebas from bacteria (5 million years against 2 billion). Surely chance and random encounters played an important role in the origin of the new cells. The particular symbioses of bacteria and archaea that led to the modern animal, vegetal and other eukaryotic cells could have been quite different. So, if there is life elsewhere, it is bound to be rather different from what we know on Earth. Just how different, we do not know, and can only learn from future observations. I am digressing from my commentaries on Rescher’s evolutionary epistemology, but I think there is a very close parallelism between the possibility of there being quite different forms of life in the universe and of there

250

RESCHER’S EVOLUTIONARY EPISTEMOLOGY

being quite different types of science. Nick is renowned as an incisive and cheerful conversation partner, combining learning and kindness in equal proportions. I have allowed myself to be carried out by our imagined conversation on these fascinating questions. Rescher is one of the very few philosophers with whom I can imagine having such a conversation.

NOTES 1

Karl Popper, “Towards an Evolutionary Theory of Knowledge”, in A World of Propensities (1990).

2

Plato, Republic, VII, 537.

3

Motto Kimura, The Neutral Theory of Molecular Evolution (Cambridge University Press, 1983).

4

Nicholas Rescher, A Useful Inheritance (Rowman and Littlefield Publishers, 1990), p. 8.

5

Nicholas Rescher, A Useful Inheritance (Rowman and Littlefield Publishers, 1990), p. 21.

251

Common Sense Joseph C. Pitt or present purposes the salient fact is that common sense constitutes a mode of endeavor with a characteristic aim of its own, namely, to achieve a rationally grounded commonality (uniformity) of opinion and evaluation in a commonality of agents dependent on cooperative and collaborative action in meeting human needs. (Nicholas Rescher, Common Sense, A New Look at an Old Philosophical Tradition, pp. 41-42.) Here in the quiet hills of Appalachia we describe someone who knows how to get things done, usually expeditiously, as being possessed of “walkin’ around sense”. That is usually in opposition to those described as having “book learning: or “perfessers”. While it is generally assumed that someone who has walkin around sense might also have book learning, it is rarely the base that someone primarily described as having book learning is also assumed to have walking around sense. Translation: perfessers don’t have walkin around sense, or rarely do, which is something we all know, which makes it a bit of common sense according to Nicholas Rescher or is it? In the short history of humankind, there is a long tradition of seeking out wise men and women for advice in living. These wise persons usually don’t do anything—they just dispense wisdom.1 This is inherently paradoxical for how can someone who doesn’t do anything be in a position to tell me how to live my life? But the tradition of getting such advice from a shaman, elder, priest, psychologist, etc. is persistent. Seeking out an elder makes some sense, it is supposed that they have lived their life and learned in the process and are therefore in a good position to dispense well earned wisdom. Priests, psychologists and shaman are another story and explaining that will take us too far afield. This wisdom of the elders seems close to something like common sense, but if it were truly in Rescher’s sense of common sense, common sense, then we wouldn’t need to seek out the elder for we ought to already know it. Since, as he puts it,

F

Common-sense belief is generally seen as neither a matter of sophisticated reasoning nor elaborate inquiry but merely a matter of good judgment (gesunder Menschenverstand). In general, such things “go without saying”

RESCHER STUDIES

since they would not tell interlocutors something that they did not already know . . . Accordingly, common sense is a matter of judgment—both as to what is the case and as to what is to be done. (Rescher 2005, pp 52-53)

And, appropriately, he seeks it’s rationale, it’s probative force. In short, why should we follow common sense? Put that way, the question appears to be a rather silly one—after all, not to do so is tantamount to admitting to irrationality, no? Well, yes and no. It really depends on how we unpack the concept. So far we have been told that common sense is a matter of judgment. In what immediately follows I am going to list some of the other salient points Rescher makes and see what kind of a picture we get. • The credibility of common sense is bound up with the fact that its validity prevails only in a very limited domain, based as it is, on the ordinary course of the everyday life of people-in-general. (p. 55) • Common sense is a matter of what is available to all on the basis of shared experience. (p. 56) • . . . common sense contentions are not categorical certitudes but powerful presumptions. (p.57) • In the end, matters of custom and usage are only able to exist by persisting, by propelling themselves down the corridor of time through its becoming manifest to people in the course of experience that they can serve to meet human needs efficiently and effectively. • [common-sense beliefs] . . .reflect the thought-stance of people-ingeneral, yet not on the basis of bio-physical considerations, but rather on the basis of culturo-cognitive considerations. They are the fruits of experience—the collective experience of people on a large scale and over a long time, having prevailed in the struggle for cultural survival through providing information that meets the needs of the group. Accordingly, they issue from principles rooted in human culture by means of rational selection. (pp 61-62) The sum of these quotes comes to this: common-sense beliefs are the ideas about everyday living it would be foolish to reject because they are the

254

COMMON SENSE

product of cultural evolution having been rationally selected for their survival benefits. Most of this I agree with. In many respects Rescher is also in large agreement with William James, who says My thesis now is this, that our fundamental ways of thinking about things are discoveries of exceedingly remote ancestors, which have been able to preserve themselves throughout the experience of all subsequent time. They form one great stage of equilibrium in the human mind’s development, the stage of common-sense. (James 1907, p. 114)

Both Rescher and James agree that common-sense beliefs are the product of some form of cultural/intellectual evolution and that they have survived over time because they have been useful for everyday survival. The problem here is that of justification. James sees common-sense in terms of categories of thought that have been useful in dividing up the world for us and that they are pretty much the same everywhere because we, all of us, encounter things, sameness, difference, kinds, minds, space, time, etc. (James p.114)2 Rescher, on the other hand, says the beliefs have survival value and are therefore rationally selected. James goes on to admit the defeasibility of these categories; Rescher too acknowledges that commonsense beliefs are not certain and eternal. But he does insist they are rationally selected and that is where we part company. Whatever “rational selection” means, it surely means using generally agreed upon principles of reasoning that represent the standard for best reasoning in these or those circumstances. Further, if one, as Rescher and James both think, concurs that the fundamentals of common sense are universal, then it would have to follow for Rescher that there are fundamental principles of good reasoning agreed upon by all peoples and cultures. But that, prior to our most recent past, there have been such principles is manifestly false.3 Learning how the other thinks has been one of the great obstacles to world order and peace. For example, coming from a democratic environment has certainly not prepared the George W. Bush government for the ways of thinking that have evolved from the politics of centuries old religious animosities and tribal allegiances. It therefore seems that we need a different account of how the principles, if not the beliefs of common sense came to be and to have the probity they do. Rescher goes to some pains to separate common-sense beliefs from common sense principles. Common sense beliefs are on the order of “try to stay dry when it is raining.” Common sense principles are broader in scope, they are in effect general rules of procedure specifying modes of ac-

255

RESCHER STUDIES

tion designed to meet universal human needs (for food, health, safety, information, etc.) For example: do not take pointless and needless risks. (p. 41) Other examples would include “trust your senses except in those cases where you know you shouldn’t,” trust your mother,” etc. But it is very hard to believe that these principles have the probity they do because of some rational decision to adopt them. As noted above, it is not clear there are universal rules of rational thinking. Second, even if there were, where did they come from? I will argue that the adoption of such principles, to the extent that they are principles at all, are the product of other features of our constitution—whether it is biophysical or social is hard to say, but it is not through some rational process. The argument here is speculative to the point where it asks about the development of processes whose origins are beyond our ability to verify. As an argument it will also take the form of a story, much on the order of Sellars’ story of our ancestors in his “Empiricism and the Philosophy of Mind.” It goes like this: Trying to answer the question “what does it mean to be human?” is clearly beyond my capacity—but it at least needs to be addressed for it is tightly connected to what we mean by common sense, for humans have common sense, whereas other species don’t unless they share some of the same characteristics we do as humans. When I work this issue with my undergraduates, I suggest that our chief individuating characteristic is a sense of humor. There are good reasons for this suggestion, but going into them will take us off topic. There is a certain sense in which no one characteristic marks us out as human, but, rather, several taken together. So, in looking for those other characteristics that make us human, several suggest themselves. First, we are curious. But other creatures are also curious; consider, for example, dolphins, monkeys, dogs, and the list goes on. Second, we have a sense of death, mortality, the temporary nature of our being. Third, we are capable of transmitting our experiences and what we have learned from them to others of our kind. Fourth, we tell stories, some of which in more developed cultures also become histories. I am sure there are others, but these seem to me to be central. And of these the most central is the ability to learn from our experience. My guess is that is a function of evolution, not social/cultural evolution as much as good old survival-of-the-fittest-sort-of-evolution. It almost goes with saying that the ability to learn from experience has great survival value, especially in a race of rather puny beings. That we can learn from experience means, among other things that if we survive whatever dangerous situation we stumbled into, we can figure out what to do the next time

256

COMMON SENSE

the situation presents itself. That it almost goes without saying that this is the case means it is as good an example of common sense as you will find on grounds shared by Rescher, Reid, and James. It is such a good example that elsewhere I have referred to it as The Common Sense Principle of Rationality (CPR) Learn From Experience. (I like the pun because it really does have cognitive value) (See Pitt 2000) The ability to learn from experience is built into us. It is what makes us work. To learn from experience requires that we be able to figure out what we had wrong when we got into trouble. Was our background knowledge defective? Did our values lead us astray? Did we have inappropriate goals? Background knowledge, values and goals and the ability to assess each of them in light of failure are what make us able to succeed in this very hostile world. It is what gives the weight to claims like “whoever fails to learn from history, is doomed to repeat it.”

Notice, however, CPR is not the sort of thing that one of our ancestors, let us call her Smith, decided to adopt as a principle of rationality. That we function using CPR on a daily basis and could not even begin to question its probilty is what makes it the principle of common sense. All the others Rescher alludes to are parasitic on this idea, that we should learn from experience. Yes, CPR is the mother of all laws of prudence, and according to Rescher common sense principles are principles of prudence. But from whence prudence? Prudence stems from knowing about the uncertainty of our knowledge of how to make our way around in the world. The acquisition of prudence requires, in a transcendental manner, CPR. Prudence is both normative and epistemological. Normatively speaking, that we ought to learn from experience, that it is prudential to do so, seems unassailable. But what does it means for us epistemologically? Here I want to take issue with one of Rescher’s main claims about common sense. Rescher claims that common sense principles have the justification they do because they are restricted in their use to the narrow domain of everyday issues. We don’t appeal to common sense when we are dealing with highly technical and specialized matters. In those domains, it is alleged, other principles of rationality come into play. Now it is just this kind of thinking that has fueled the long debate over the relation between science and common sense. Are they separate domains or are they continuous? Sellars, in “Philosophy and the Scientific Image of Man,” (see Sellars 1963) sees them as separate, although he thinks the scientific image of man is an outgrowth of what he calls the manifest image, the everyday image of the world we have, what I would call the world of common sense. But, for Sellars, the two images will be

257

RESCHER STUDIES

sufficiently different that some means must be devised in order for us to come to have a full understanding o the world and our place in it. He thinks this is to be found in a combining of the two images into a stereoscopic image. I, on the other hand, wish to argue that in its basics, scientific thinking is much the same as common sense thinking. Science proceeds by way of trial and error. Trial and error is a direct function of CPR. Rescher’s attempt to restrict common sense to the domain of everyday life is a result of his failure to ground common sense thinking in something like CPR. The principles of commonsense cannot be rationally selected for the reasons we have discussed above. But if there is no other justification for them, then their success in the world of everyday life is, for Rescher, their own justification. But, as I have suggested, Smith did not choose to operate using the principles of common sense, she realized that proceeding by way of learning from her mistakes and failures and adjusting her actions accordingly gave her a competitive advantage. And obtaining a competitive advantage is not something our ancestors decided to seek, it was mandatory for survival. Smith did not think this through and decide to adopt CPR, but rather as a creature fighting for survival in a very dangerous world, she found that this way of operating was successful and she taught it to her children. In its most basic form this might even be possible without language. We have evidence that monkey can teach their young by example. Surely that is how we got started as well. And if it is possible to learn by example without language, then since language is crucial to rational thinking, it seems possible for our humanoid ancestors to have developed CPR without rational thought, nay it was necessary. Smith succeeded not because she thought about adopting a mode of action—she succeeded because she acted first and then thought about it—in short, she had to have something to think about. So what does this all mean for Rescher’s account of common sense. If he is serious about having the principles of common sense be rationally selected, then he is not talking about common sense. This emphasis on rational selection suggests more of the rationalist than a pragmatist commitment to common sense. This is too bad since the heart of pragmatist epistemology is the idea of a feedback loop involving reassessment and reevaluation and Rescher surely wants to be a pragmatist. But the pragmatist always has the same problem as anyone else—the how-to-get started problem. I am suggesting that it can be solved by not thinking about it.

258

COMMON SENSE

That is, first comes action, then if you live to think about it, then think about it. A bit primordial perhaps, but Smith would agree.

REFERENCES James, William, “Pragmatism and four essays,” in The Meaning of Truth, (Cleveland: Meridan Books, 1907). Peirce, C. S., Philosophical Writings of Peirce, edited by Justus Buchler (New York: Dover Publications, 1955). Pitt, J. C., Thinking About Technology, (New York: Seven Bridges Press, 2000). Also available at http://www.phil.vt.edu/HTML/people/pittjoseph.htm Resher, Nicholas, Common-Sense: A New Look at an Old Philosophical Problem (Marquette: Marquette University Press, 2005). Sellars, Wilfrid, “Empiricism and the Philosophy of Mind,” in Science, Perception and Reality (London: Routledge and Kegan Paul, 1963). Sellars, Wilfrid, “Philosophy and the Scientific Image of Man,” in Science, Perception, and Reality (London: Routledge and Kegan Paul, 1963). NOTES 1

Ample evidence of this practice and that it is a well-established fact of our lives is to be found in cartoons in virtually in every issue of The New Yorker.

2

This is one of the few places where James reverts to one of the pragmatist’s main moves, abandoning the individual-centered epistemology of the traditional western stance in favor of an appeal to the community. Despite the title of the work from which this view of common sense is drawn, James is a strange pragmatist, making most issues turn on the needs, desires, etc, of individuals. C. S. Peirce’s great contribution to philosophy was to show the way out of the solipsistic bind rationalists and empiricists alike have put us in by appealing to the community as the final arbiter of truth, reality, standards, etc. See Peirce 1955, Chapters 2 &3.

3

The development of the internet and the phenomenon of globalization are indeed creating a situation in which we might not agree that this way of thinking is the best, but that we must all think this way if we are going to do business.

259

Systematic Philosophy and Theoretical Framework on Nicholas Rescher’s A System of Pragmatic Idealism Lorenz B. Puntel

T

he expressions ‘system of philosophy’ (or ‘philosophical system’) and ‘systematic philosophy’ can be B and often in fact are B taken as synonymous, thus being only two syntactic variants for designating one and the same concept. But they have significantly different connotations, most especially in a historical perspective. ‘System of philosophy’ (or ‘philosophical system’) evokes memories of the so-called great philosophical systems of the past, especially those developed by the German idealists Fichte, Schelling, and Hegel, and thus of megalomaniacal philosophical projects that history has shown to be unavailing. Against this background, the association of ‘system of philosophy’ or ‘philosophical system’ with the presentation of a grandiose, absolutely complete, closed, and definitive philosophical conception is in most cases almost inevitable. For this reason the author of this paper avoids using ‘system of philosophy’ or ‘philosophical system’, preferring instead the more modest and less problematic expression ‘systematic philosophy’.1 Between 1992 and 1994, Nicholas Rescher published his monumental three-volume work A System of Pragmatic Idealism.2 The title amounts to A Pragmatic-Idealistic System of Philosophy. But Rescher’s usage of ‘system’ it far away from having the historical connections just alluded to. Indeed, Rescher’s System of Pragmatic Idealism has nothing to do with the megalomaniac “systems” of the past. For the sake of clarity, however, in this paper the author will use the expression ‘pragmatic-idealistic systematic philosophy’ when dealing with Rescher’s work. This paper does not examine the entire work; that would be too huge a task to be tackled in the space available. It addresses only one aspect of the work, although to be sure one that is central or fundamental: the question of the relation between systematic philosophy and what will be called theoretical framework. When addressed to Rescher’s work this general question must be specified by distinguishing several subquestions. For this reason, the paper is divided into three sections. Section 1 explains the con-

RESCHER STUDIES

cept of theoretical framework. Section 2 investigates the theoretical framework underlying Rescher’s pragmatic-idealistic systematic philosophy. Section 3 presents a partial critical appraisal. 1. THE CONCEPT OF THEORETICAL FRAMEWORK 1950 R. Carnap introduced the concept of linguistic framework.3 His starting point was the question about the status of abstract entities. He characterizes his understanding of “linguistic framework” indirectly by saying: If someone wishes to speak in his language about a new kind of entities, he has to introduce a system of new ways of speaking, subject to new rules; we shall call this procedure the construction of a linguistic framework for the new entities in question.4

Carnap then introduces a “fundamental distinction between two kinds of questions concerning the existence or reality of entities” (ibid.). The first kind he calls internal questions, i.e., questions formulated with the help of the theoretical resources provided by the accepted linguistic framework. The second kind of questions he dubs external questions; these he characterizes as questions “concerning the existence or reality of the system of entities as a whole” (ibid.). It is not clear how he understands “the system of entities as a whole,” but a certain clarification can be obtained from a linguistic framework he provides as an example: the linguistic framework of “thing language”, which expresses “the world of things”. On the basis of this linguistic framework one must, according to Carnap, say the following: To recognize something as a real thing or event means to succeed in incorporating it into the system of things at a particular space-time position so that it fits together with the other things recognized as real, according to the rules of the framework.5

From the internal question posed within the “thing language” Carnap distinguishes “the external question of the reality of the thing world itself” (ibid.; emphasis added). Carnap rejects such questions and interprets them as being questions treated by realisitic as well as idealistic metaphysics. Here again, Carnap’s exact thesis is not clear. On the one hand, he says that an external question “cannot be solved because it is framed in a wrong

262

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

way” (ibid.). This seems to mean that questions asked (and treated) outside a linguistic framework are meaningless questions, devoid of any sense. On the other hand, he characterizes external questions as questions relating to “the world itself”. But surprisingly, he fails to point out that questions referring to “the world itself” can be B and in fact are B questions based on a linguistic framework, namely the framework containing expressions that refer precisely to “the world itself”. One must conclude, that the rationale behind Carnap’s distinction is the thesis that nothing can be established in a meaningful and theoretical way without recognizing (i.e. making explicit or at least presupposing) an adequate linguistic framework. Carnap’s concept of linguistic framework is too narrow to provide an adequate clarification of philosophical theorizing. It articulates only a partial aspect of a much more extensive concept: the concept of theoretical framework. Language is only one element of the entire apparatus every theory must presuppose and rely upon. Besides language and its semantics there are at least two other indispensable components of every theoretical framework: the formal dimension (logic and mathematics) and the dimension of ontological structures (or categories). For the sake of brevity and clarity one can simply speak of three kinds of structures that constitute every theoretical framework: formal (logical/mathematical) structures, semantic structures, and ontological structures.6 Theoretical frameworks for philosophical theories must include other components as well, but which ones are required depend upon the specifics of the theories in question. In many cases, what can be called the epistemological dimension, the dimension of knowledge, is among the most important of those components. Indeed, most philosophical conceptions or theories understand themselves to be special forms or cases of knowledge. Nevertheless, according to the author’s conception the epistemological dimension cannot be attributed a central B let alone an all-determining B place within systematic philosophy. Many, perhaps most philosophers would probably say that the justformulated claim is unreasonable and even false, especially because it is allegedly manifestly counterintuitive. They would argue that unquestionably, theories are necessarily connected to the dimension of knowledge, because theories articulate what we know or believe that we know. But this widely accepted argument must be contested, as can be shown by working out what can be called the structure of theoretical sentences. Purely theoretical sentences are indicative/declarative sentences that bear no explicit reference to subjects, speakers, knowers, or the like. Genuinely theoretical sen-

263

RESCHER STUDIES

tences do not have the (explicit or implicit) form, “Subject S believes/knows that p”; they have instead the form, “It is the case that p”. Indeed, one can say that a sentence φ is indicative (hence theoretical) if and only if a syntactically correct sentence results from the application to φ of the operator “It-is-the-case-that”, an application that can be indicated by introducing the notation ‘ φ’. With respect to any theory as such, the attitudes of subjects towards it is in a fundamental perspective completely insignificant: whether subjects believe or doubt or accept the theory has no impact on the theory as such. The theory is so to speak immune to the vicissitudes of the attitudes subjects might entertain towards it. A theory has its value solely on the basis of what it articulates and of the force of the arguments that sustain it. This is not to deny that in other respects the attitudes of subjects towards theories are an important matter. In principle, every single philosophical theory is based on an explicitly formulated or at least implicitly presupposed theoretical framework. But this fundamental feature of philosophical theories is a fortiori of utmost importance for systematically oriented philosophies; indeed, what characterizes systematic philosophies in the first place is the interconnectedness they articulate among the many specific philosophical theories that address special topics. But interconnectedness is unconceivable without a welldefined theoretical framework. A further point that must be emphasized is the thesis that there is a plurality of (explicitly formulated or implicitly presupposed) theoretical frameworks—that there are in fact many different theoretical frameworks and that even more can be devised or are possible. This state of affairs is a fundamental constitutive part of every sound systematic view of philosophy and science, as is shown below. 2. THE THEORETICAL FRAMEWORK UNDERLYING RESCHER’S PRAGMATIC-IDEALISTIC SYSTEMATIC PHILOSOPHY [1] Indisputably, Rescher is a systematically oriented philosopher. He expresses his understanding of philosophy in very clear terms: [A] tenable philosophy must be a systematically dovetailed whole. For in the end, the range of our philosophical concern is a network where everything is interconnected with everything else. [. . .] Only systemic, holistically attuned positions can yield truly satisfactory solutions to philosophical problems B solutions that do not impose undue externalities even in the wake of the nearterm advantages they yield within the immediate problem-area of their local

264

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

doctrinal neighborhood. For better or worse, viable philosophizing has to be a matter of systematization. (III 35).

And Rescher describes his own monumental work as follows: In numerous and varied publications appearing over the last thirty years, I have developed a system of philosophy B a complex but (I hope) coherent position embodying a unified approach to a wide variety of interrelated philosophical issues. [. . .] The larger project here at issue seeks to present my philosophical ideas in a sufficiently comprehensive and coordinated form that their systematic interrelatedness becomes clearly manifest. (I, Preface).

The central point concerns the systematic interrelatedness of the issues treated. But this general feature presupposes that some specific questions are clarified in advance, most especially the following two questions: what kind of treatment is applied, and what kinds of issues or questions are treated? The first question directly addresses the theoretical framework presupposed by the elaboration of the systematic philosophy. The second question concerns the manner in which the theoretical framework is applied in the treatment of substantial philosophical issues. This paper focuses on both questions, considering them together. Clarifying these questions in the case of Rescher’s pragmatic-idealistic systematic philosophy requires a brief indication of how his systematic work is composed. The subtitles of the three volumes identify the contents presented in each of them: Human Knowledge in Idealistic Perspective (Volume I); The Validity of Values (Volume II), and Metaphilosophical Inquiries (Volume III). The work does not include a complete, separate exposition of the pragmatic-idealistic theoretical framework as such, but there can be no question that Rescher avails himself of such a theoretical framework. He relies upon it, but his explicit presentation of it is incomplete. This procedure is perfectly legitimate and coherent given a central methological stance underlying Rescher’s entire conception. This is the strong rejection of every kind of foundationalism, i.e., of the doctrine that holds that every kind of discursive claim to truth requires truths as inputs. More exactly, the foundationalist assumes at the beginning of his theorizing a very small initial collection of absolutely certain truths and from these proceeds by procedures of supplementation to reach a wider domain of truth. In sharp contrast with this foundationalist methodology, Rescher develops a coherentist methodology that inverts the foundationalist ap-

265

RESCHER STUDIES

proach by thoroughly dispensing with any appeal to basic, foundational truths, and by extracting (presumptive) truths discursively from initial data understood as consisting merely of truth candidates and not of certain or fixed truths. Whereas the foundationalist methodology yields a basically axiomatic and, therefore, linear systematization of reasoning (as in mathematics), the result of coherence methodology is a cyclical structure, where one can repeatedly return to and reappraise any item already introduced. In short, coherentist methodology produces a theoretical network, not a theoretical linearity. From the coherentist methodology a highly significant consequence can be derived as regards the question and the shape of the exposition or the presentation of the systematic philosophy B and Rescher, indeed, explicitly draws this consequence by introducing a fundamental distinction between the order of exposition and the order of theoretical or rational connection: the first must be linear, the second need not be so. The idea of a network of substantive interconnections is, after all, the very heart and core of the coherence criteriology of justification that has traditionally played an important role in idealist philosophy. Thus while mere expository convenience can be allowed to prevail in the order of presentation, this need not (and generally does not) reflect the substantive order of fact. The systemic coherentist can quite appropriately reject the challenge to justify one’s mere expository proceedings in advance. Conforming practice to theory, it suffices to let the justification of one’s beginning lie retrospectively with the wisdom of hindsight B in the capacity of the discussion as a whole to accommodate comfortably the overall system it is designed to present. (I xiv, Introduction) This explains (in the sense of: clarifies and legitimates) Rescher’s dealing with his own theoretical framework: initially he partially explains it, but on the whole his framework is simply used in his work; additionally, as the treatment of substantial issues develops, further aspects of the theoretical framework are made explicit. [2] In his explicit characterization of his theoretical framework Rescher concentrates almost exclusively on one aspect or element: the epistemological component. The reason for this concentration is that the kernel of Rescher’s theoretical framework is human knowledge, which he undertands and explains in pragmatic-idealistic terms. All of volume I of his trilogy is devoted to this topic. Rescher holds a “middle-of-the-road-idealism that makes significant concessions to realism” (I xiv). He defends a special kind of interaction be-

266

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

tweeen mind and reality, according to which the two sides are inextricably intermingled. As is shown in the next section, in Rescher’s handling of the interaction the main weight lies on the side of the mind: Idealism centers on the conception that reality as we understand it reflects the workings of mind. And it construes this as meaning that the inquiring mind itself makes a formative contribution not merely to our understanding of the nature of the real but even to the resulting character we attribute to it. Through its generality and indefiniteness, this sort of view is not so much a philosophical doctrine as a general strategy for developing one. (I xiii)

Rescher links this view of idealism with a pragmatically oriented rationality and epistemological stance. He holds that “the ultimate arbiter of rationality is represented by a very basic concept of knowledge-wed-topractice, and the ultimate validation of our beliefs lies in the combination of theoretical and practical success, with ‘practice’ construed in its pragmatic and affective sense” (I 225). To refute the objection that this stance amounts to an instrumentalistic position that abandons truth for utility, Rescher notes the following: Only at a remove, at the methodological level, do pragmatic considerations enter in. Successful praxis is seen as adequacy indicative with respect to our generalized criteria of acceptance and verification. The link between truth and praxis is not seen as direct, but as mediated by the methodology of inquiry. Our justificatory rule is, Truth claims via methods, and methods via praxis. (I 226)

To be sure, this is to be understood in the sense that “the practical sector is ultimately in the controlling position as regards cognition B that praxis affords our quality-control monitor over theoria, with applied science constituting the ultimately commanding factor with respect to theoretical science.” (Ibid.) With the exception of Volume II, which treats the topic of the validity of values, Rescher’s entire systematic work is an extensive exposition of the pragmatic-idealistic component of his theoretical framework. One of the most striking features of the work is the almost completely absence of a philosophy of language, especially of a semantics, and of an explicit ontology. Rescher’s systematic philosophy is essentially epistemologically oriented, and pays little or no attention to semantical and ontological problems. Section 3 presents some critical considerations of this orientation.

267

RESCHER STUDIES

3. A PARTIAL APPRAISAL: LIMITATIONS AND DESIDERATA Rescher’s great work is a major achievement in contemporary philosophy. It shows that and how systematic philosophy can and must be done. Probably no other contemporary author has so convincingly succeeded in presenting a systematic view that emerges from the best to be found in the great tradition of philosophy, from the achievements of the analytic philosophy of recent times, from the rich and sophisticated instruments of modern logic, and from taking seriously the magnificent progress of modern science. Instead of praising the immensely positive aspects of the work, however, this section identifies some deficiencies and desiderata, addressing two points: first, shortcomings of the pragmatic-idealistic theoretical framework; second, brief final remarks on some absent or insufficiently treated central philosophical areas 3.1 SHORTCOMINGS OF THE THEORETICAL FRAMEWORK

PRAGMATIC-IDEALISTIC

[1] Rescher’s entire systematic conception is centered on subjects. One of the most frequently occurring expressions in Rescher’s exposition is “we”. (Rescher never uses the singular “I”.) That subjects should be accorded absolute centrality might at first glance appear to be beyond dispute. After all B so it is said or presupposed B philosophy is our enterprise, our work; theories are our creations; the theoretical tools are our inventions, etc. In Rescher’s words: all this is the result of the “workings of (our) mind” (I xiii). Moreover, it is said or presupposed, we have to decide whether we accept or reject a theory; we engage in discussions, etc. Taking seriously the centrality of subjects seems to be a philosophical stance of indisputable evidence. If the philosophical enterprise starts with or from subjects and in fact from the outset establishes subjects as the absolutely central point of reference for the whole of philosophical theorizing, then of course the matter is settled once and for all. But a closer analysis shows that this allegedly indisputable evidence is only superficial, only a surface phenomenon. In contrast, a factor that is not superficial but instead is truly indisputable is the following: one of the essentials of every theory is the language in which the theory is expressed or articulated. The main components of theoretical languages are indicative (declarative) sentences that deserve to be called

268

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

theoretical sentences. From the structure of indicative/theoretical sentences some far-reaching consequences can be drawn. Their structure has been correctly articulated by Wittgenstein, who writes, in the Tractatus, “The general form of the [indicative, theoretical] sentence is: It is the case that such-and-such (es verhält sich so und so)” (Tractatus 4.5). This form is extraordinarily simple and at the same time extraordinarily profound and important. As is shown above, this reveals that the theoretical standpoint is articulated by the operator “It is the case that. . .” B the theoretical operator B that is an explicit or implicit component of every theoretical sentence.7 Philosophy, as a theoretical science, presents only theoretical sentences. In the present context the important point is the fact that the structure of theoretical sentences does not make any reference to anything like subjects or speakers or knowers, or the like. It says simply: something is articulated absolutely. What does this mean and what implications does it have? First of all, this structure makes fully clear that no opposition or dichotomy between subjects (or speakers or knowers) and something like the world (the domain of what is bespoken or known) is either explicitly articulated or implicitly presupposed. The structure is the expression of a presupposed unity between the entire theoretical dimension and the dimension it is about, call it “object(s)”, “the world”, “the universe”, or “being”. (For the sake of brevity the expression ‘the world’ is used in what follows.) Every indicative/theoretical sentence articulates one “point” or “aspect” or “segment” of “the world”. But a more precise characterization of “the world” must be given. This expression can be taken as designating only the objective pole of the relation between mind and world; in this sense, “the world” is taken in the objective sense: “the worldO”. But “the world” can be understood as the dimension that encompasses both mind and worldO: “the world” in the comprehensive sense—“the worldC”. The just formulated thesis must now be stated in more precise terms. Every indicative/theoretical sentence articulates one “point” or “aspect” or “segment” of “the worldC”. From this a further consequence follows. The entire theoretical dimension B what Rescher according to his pragmatic stance calls “the workings of mind” (I xiii) B is not something separate from the worldC, not something that could or should be introduced into this comprehensive dimension, so to speak from outside this dimension; instead, the theoretical dimension is a dimension of the worldC itself in the most literal sense. It becomes clear that it is inadequate to single out the side of mind by characterizing the relationship of mind and worldC as being exclusively de-

269

RESCHER STUDIES

termined by the mind. This is what Rescher does, when he repeatedly says that in philosophy and science only “our truth” (not “truth itself”), “ou reality” (not “reality in itself”) is attained or articulated.8 On the contrary, from what has been shown it follows that the worldC itself is being manifested or articulated: theorization is in a non-metaphorical sense the automanifestation or auto-articulation of the worldC itself. Against this view one can raise the objection that it fails to do justice to the undeniable fact that theories, concepts etc. are products of our own making. This is certainly true, but it is only a partial truth, an inadequately captured and articulated truth. It is partial because it ignores the fact that our workings are a dimension of or occur within the worldC itself. Our workings B however they may be described from our side B are fundamentally and in the first place workings of the worldC itself. The objection is the result of overlooking the entire multi-level structure of the mindBworld relationship. In some passages of his oeuvre Rescher fully recognizes this fundamental point. An example: Our mathematics is destined to be attuned to nature because it itself is a natural product as a thought instrument of ours: it fits nature because it reflects the way we ourselves are developmentally emplaced within nature as integral constituents thereof. Our intellectual mechanisms B mathematics included B harmonize with nature because they are themselves a product of nature’s operations as mediated through the cognitive processes of an intelligent creature that uses its intelligence to guide its interaction with a nature into which it is itself fitted in a particular sort of way. (I 99)

This passage is to a certain extent a partial description of what is called above “the worldC”, although some wordings raise questions. But the idea that mathematics B and, as can coherently be added, the entire theoretical dimension B “fits or is fitted into nature” (or “world/reality/being”) is a powerful idea that has far-reaching consequences. Unfortunately, Rescher does not draw those consequences. To be sure, these formulations are highly abstract and general. Still, they articulate decisive guidelines for the philosophical treatment of this extremely complex and difficult topic. One fact they reveal is that an exclusively or even primarily epistemological approach is unavailing, because it cannot adequately articulate what may be called the real theoretical situation. One factor that makes it difficult to specify Rescher’s theoretical framework with adequate clarity is that Rescher introduces considerations

270

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

whose status—as epistemological, or ontological, or perhaps as of some other sort—is not entirely clear. Nevertheless, these are important and worth taking into account. They are to be found most especially in Part IV of Volume I, which is titled “Realism and Idealism.” [2] Rescher defends an overall position with respect to the idealismrealism debate. He rejects some of the stronger versions of idealism, but endorses some of the less ambitious. He calls his own brand of idealism “pragmatic/cognitive/conceptual idealism” and he thinks this position is perfectly compatible with a “realism of sorts” he dubs “situational realism” (I 122-125). As regards the long-standing quarrel between realism and idealism, he comes to the conclusion that “the most sensible stance is that of a mixed, halfway-house position that gives both realism and idealism their respective due” (I 125). He devotes an extraordinary effort, in lengthy expositions, to explaining and justifying this position. An appraisal of his immense effort that would do justice to his arguments and reflections would require far more space than is available here. The following account presents only a brief exposition of his main thesis and his main argument, and then some critical remarks. Rescher summarizes his position as follows: Our position is a realism, since it acknowledges a realm of ontologically mind-independent existence. But in accepting a cognizability-in-principle standard of real/true/actual, it is also a cognitive idealism that holds that existence is in principle knowable B although known by us in terms of reference of a characteristically human orientation. What seems right about idealism is inherent in the fact that in investigating the real, we are clearly constrained to use our own concepts to address our own issues B that we can learn about the real only in our own terms of reference. But what seems right about realism is that the answers to the questions we put to the real are provided by reality itself B whatever the answers may be, they are substantially what they are because it is reality itself that determines them to be that way. Mind proposes, but reality disposes. But of course, insofar as one can learn about this reality, it has to be done in terms accessible to minds: our only access to information about what the real is [is?, LBP] through the mediation of mind. (I 323-4)

Appraising Rescher’s position requires that two points be initially spelled out.

271

RESCHER STUDIES

[i] In the context of his treatment of a rather marginal topic B the problem of extraterrestrial science B Rescher introduces a distinction that is highly illuminating for understanding and appraising his systematic position as a whole. It is the distinction between the process and the content (or product) of science: Is the requisite for a civilization’s ‘having science’ primarily a matter of the substantive content of their doctrine (their belief structures and theory complexes), or is it primarily a matter of the aims and purposes with a view to which their doctrines are formed (control, prediction, explanation, and the like)? (I 112-3)

The issue of substantive content enters in this context with the question of how similar the scientific beliefs of the members of other civilizations are (would be? could be?) to ours. Significantly, Rescher says that this question of content “is clearly something in which we would be ill advised to put much stock”. And he adds: “. . . in adjudging scientific status, we had best leave the matter of doctrinal content aside and give prime emphasis to matters of process and purpose rather than products and results” (I 113; italics supplied). This philosophical stance decisively explains the insufficient character of Rescher’s “overall position” (I 323): it is manifestly a one-sided position that has very serious and negative consequences. It does not take B at least not sufficiently B into account the fundamental fact that the most important “point”, the very undisputable core of any theory is its doctrinal content: what it says about the world, what the theory is about. All other aspects, including those mentioned by Rescher (control, prediction, processes etc.), are certainly important in different respects, but in the final analysis they are entirely subordinated to the central factor “(doctrinal) content”. [ii] Rescher’s argumentative strategy for substantiating his “mixed position” consists essentially in introducing some fundamental and far-reaching distinctions as the ultimate basis on which it relies. Some examples of the most important distinctions: B

our science vs. science(s) (I 110)

B

our science vs. ideal science (I 297)

B

our side (with respect to science) vs. nature’s side (with respect to science) (I 99-104)

272

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

B

intent of science vs. achievement of science (I 58)

B

what science asserts vs. how science makes its assertions (I 285)

B

“. . . one cannot appropriately maintain that our science depicts reality” vs. “all we can say is that it affords an estimate of it” (I 297)

B

“The thesis that ‘science describes the real world’ must be looked upon as a matter of intent rather than as an accomplished fact” (I 299)

B

the truth/the real truth vs. our truth (I 52-53)

B

the truth vs. our truth = our best estimate of the truth (I 52)

B

our best estimate of truth vs. getting nearer to the truth (approximation to the truth) (I 59)

B

reality as such/itself vs. reality that is available to us (I 271)

B

reality itself vs. reality as we can manage to conceive of it (I 32-33)

B

“‘Reality as such’ is no doubt independent of our beliefs and desires, but what can alone concern us is reality as we view it” (I 271).

B

“The world that we describe in science is one thing, the world as we describe it in science is another, and they would coincide only if our descriptions were totally correct B something that we are certainly not in a position to claim” (I 279)

B

“the world” vs. the world as we can manage to conceive of it (I 319)

B

realism of intent vs. realism of achievement (I 284)

All these distinctions are based on the assumption that knowing or theorizing subjects are the all-determining point of reference. For “knowing or theorizing subjects” Rescher significantly says simply “we”. At first glance

273

RESCHER STUDIES

his distinctions appear to be highly illuminating and even supported by compelling evidence. But what does a closer analysis of the “real meaning” of such distinctions yield? The following considerations attempt to shed light on this central topic. [3] Interestingly, Rescher himself formulates what can be taken to be the most serious objection against his own position: On conceptual idealism’s telling, there is something indelibly anthropomorphic (human patterned) or, more accurately, noomorphic (mind-patterned) in the way in which we think about the world. ‘But’, someone may say, ‘this does not show anything about the world, but rather about us’. (I 319)

Rescher dismisses this objection by simply reiterating one of his central distinctions (the numerical indices/subscripts appended to the term ‘world’ are not in the following quotation; they are added here only to clarify the references to Rescher’s text in the considerations that follow): However, this objection loses the point of conceptual idealism, which is exactly that there is no clear separation between the world1 and the domain of thought, because our only possible route to cognitive contact with the world2 is through mediation of our conceptions about it, so that for us, ‘the world’3 is inevitably ‘the world4 as we can manage to conceive of it’. In consequence, how we conceive of the world5 (namely, in noomorphic terms) has to be seen as a fact not just about us but about ‘the world’6 as well. (Ibid.)

This response reveals a profound ambiguity in Rescher’s position, one that becomes manifest when his usage of the term ‘world’ is submitted to an accurate analysis. There are six occurrences of the term, whereby two of them have the term in quotation marks. Presumably, in those two occurrences (3 and 6) the term has the same meaning. But the text has a point only if one presupposes that there is a difference between the occurrences of the term in quotation marks and the occurrences without quotations marks. What then is the difference? In occurrences 1 and 5 the term seems to be used in the same sense—but what sense? It is manifestly not identical to the sense of the term in occurrence 4, because in this occurrence the term is specified as the-world-as-we-can-manage-to-conceive-of-it. So we have: the world (occurrences 1 and 5), ‘the world’ (occurrences 3 and 6), and the-world-as-we-can-manage-to-conceive-of-it (occurrence 4). How are they related to each other? The most plausible interpretation seems to be the following: in occurrences 1 and 5 the term the world (without quota-

274

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

tion marks) is used in a very general and undifferentiated or unspecified sense. But a strong difference seems to be admitted by Rescher between ‘the world’ (occurrences 3 and 6) and the-world-as-we-can-manage-toconceive-of-it (occurrence 4). ‘The world’ seems to be identical to what Rescher calls ‘the world itself’ or ‘world as such’ (see the quotations above). But what is the difference between ‘the world itself/as such’ and the-world-as-we-can-manage-to-conceive-of-it? The problem is that “we”, in speaking of “the world itself”, really speak of “the world itself”, that is, “we” already use a concept and, according to Rescher, a concept of “ours”, of course. But then the alleged difference between “the world itself” and “the-world-as-we-can-manage-to-conceive-ofit” simply evaporates. To think and to articulate “the world itself” along the lines of inquiring and reasoning pursued by Rescher involves a plain incoherence. The only way out seems to be to distinguish two different levels of thinking or articulating: the first would be the level in which “we”, the subjects, are the determining point of reference, being, so to speak, the subjective pole (or the subjective relatum) of a relation whose objective pole (or objective relatum) is “the world”. The other level is a higher level: it is the dimension that embraces or encompasses both “us” and “the world” (understood as the objective pole of the relation having “we/us” as the other relatum; “world” in this sense is characterized above as “world in the objective sense and indexed as: worldO). The higher, encompassing dimension is characterized above as the world in the comprehensive sense: the worldC. Plainly, no incoherence emerges if the difference between “the world itself B in the sense of: the world-as-notconceived-of-by-us” and “the-world-as-conceived-of-by-us” is established from the standpoint of the worldC. This still purely negative result of the above considerations is in need of further elaboration. [4] This subsection sketches an alternative conception, but not one that should be understood as an all-round rejection of or diametrically opposed alternative to Rescher’s position. Instead, at least with respect to many points, the author’s proposal understands itself as being another interpretation of some fundamentally valuable intuitions and insights that Rescher has tried to articulate within his theoretical framework. Four factors must be addressed in this context. [i] The central point is changing the focus of the (philosophical/scientific) perspective from subject(ivity) (from “we”) to pure or genuine theoretical language (the perspective from the world/being). This point

275

RESCHER STUDIES

is initially exposed and explained above (see 3.1 [1]), but here important additional clarifications must be added. This shift of perspective has profound and far-reaching consequences. One of the most important is that what matters in the first place is not that and how we address ourselves to the world, but rather the fact that the world (the world itself!) is articulated, that it manifests itself in a certain manner and to a certain extent. This is not to deny that our concepts and other tools are at work; it is instead to emphasize that our concepts et alia are in the first place modes of the self-manifestation of nature or the world itself. Interestingly, in some passages of his oeuvre Rescher himself explicitly touches on this fundamental factor, although to be sure without drawing from it the crucial theoretical consequences. For example: [W]hat seems right about realism is that the answers to the questions we put to the real are provided by reality itself B whatever the answers may be, they are substantially what they are because it is reality itself that determines them to be that way. Mind proposes, but reality disposes. (I 323-4) We need the notion of reality to operate the conception of truth. A factual statement on the order of ‘There are pi-mesons’ is true if and only if the world is such that pi-mesons exist within it. By virtue of their very nature as truths, true statements must state facts: they state what really is so, which is exactly what it is to characterize reality. The conception of truth and of reality come together in the notion of adaequatio ad rem B the venerable principle that to speak truly is to say how matters stand in reality, in that things actually are as we have said them to be. (I 260)

In both passages Rescher switches the perspective: from the weperspective to the purely theoretical perspective (“it is the case that. . . ”), and then back. The second passage is especially important in this respect: “we” occurs in the first sentence and in the last. But why articulate a determinate state of affairs from our perspective? Why not say, for example, “The notion of truth is a presupposition for the operability of the conception of truth” (first sentence)? For highly similar reasons, the last subordinate clause in the last sentence is also highly significant: “. . . in that things actually are as we said them to be”. After having formulated purely theoretical sentences in accordance (in effect) with the structure “ φ”, Rescher reintroduces the we-perspective, saying “as we said things to be. . .”. Why we? Compare Tarski’s famous so-called informal characterization of the

276

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

(intuitive) concept of truth that underlies Rescher’s characterization of truth in the quoted passage: “A true sentence is one that says that the state of affairs is so and so, and the state of affairs indeed is so and so.”9

Tarski formulates: “. . . a true sentence says that. . .”; he does not say: “. . .we say that. . .”. This may appear to many as a triviality or as an unimportant or even nonsensical subtlety, but it simply is not. The we-perspective introduces a one-sided standpoint, a restricted standpoint, whereas the purely theoretical sentence articulates the only absolutely universal, unrestricted perspective. This universal, unrestricted perspective is the perspective from the standpoint of that dimension that has been called above the world in the comprehensive sense: the-worldC. In every theoretical sentence whose structure is “it is the case that (for example) φ,” the particle “it” can be interpreted as being the abbreviation for “the-worldC”: in those sentences the-worldC itself articulates or manifests or expresses itself in such and such a way (in the example: in the way “φ”). [ii] Rescher defends the thesis that there is a mind-independent world or reality. But he speaks of a “postulation” (I 260-266) and raises the question of what legitimates this postulation. He answers the question from a pragmatic perspective: “in terms of its ultimate utility” (I 260). This pragmatic view of the mind-independence of the world/reality does not do justice to the status of theoretical discourse. Theoretical discourse is simply a fundamental fact of human life; it requires no justification and none can be provided for it, simply because a justification would presuppose the validity of theoretical discourse. But theoretical discourse is always in some way or another about the world or reality B and this world or reality is different from and independent of the theoretical mind (except in those cases when the theoretical mind articulates itself). Since human theoretical minds are contingent beings, there is a mind-independent world/reality, “mind” here meaning “human mind”. This is not the result of a “postulation”, because the concept of “postulation” presupposes that there is a meaningful question to be clarified. Of course, there is such a question if one contemplates the history of ideas (of philosophy). Especially in modern times the human mind has been conceived of as a subject enclosed in itself, separate from “the real world”. If this stance is accepted as a meaningful and serious position, then there is indeed a question of the “mind-independence” of the world. And one of the responses that has been given to that question has been to postulate such a mind-independent

277

RESCHER STUDIES

world. Rescher explains his concept of postulation thus: “. . . the justification for this fundamental presupposition of objectivity [the mindindependence of world/reality] is not evidential at all; postulates are not based on evidence. Rather, it is functional. We need this postulate to operate our conceptual scheme.” (I 264-5) But this stance is simply the result of overlooking and consequently ignoring the very status of the theoretical dimension: this dimension is simply what it is, namely the presuppositionsless basis for every theoretical sentence, question, statement. . . [iii] Rescher admits the intelligibility of nature: mind is attuned to nature and nature is attuned to mind (I 96-109). It appears that he understands intelligibility primarily, if not exclusively, from an evolutionary perspective: “A world in which intelligent creatures emerge through the operation of evolutionary processes must be an intelligible world.” (I 103) This is a convincing argument B as far as it goes. But it falls short of what a systematic philosophy can and should show. A more comprehensive way of understanding and characterizing the intelligibility of the world/reality emerges from a different approach. A closer analysis of the initial or basic theoretical situation brings to the fore what can be dubbed universal expressibility, i.e., the expressibility of the whole of the world/reality/being.10 To articulate this feature in another manner: the whole of the world/reality/being is the unrestricted universe of (theoretical) discourse. It would be self-contradictory to admit of something beyond the unrestricted universe of (theoretical) discourse, because to admit it is to refer to it, and this amounts to considering it as an element within the unrestricted universe of (theoretical) discourse. The term ‘expressibility’ is introduced here as a general term, as an abbreviation for many different concepts including “intelligibility,” “understandability,” “articulability,” “explainability,” “graspability, “explicability,” “conceivability,” etc. In this context a few remarks on Rescher’s understanding and usage of the central concept of systematicity are in place. In A System of Pragmatic Idealism he equates systematicity with cognitive systematicity; in this sense systematicity is a hallmark of the scientific. But in another book11 he explicitly treats the question of the status of systematicity within a systematic philosophy. He distinguishes between cognitive and ontological systematicity and tries to clarify the relationship between the two. To cognitive systematicity he attributes a regulative/methodological character. Moreover, he considers cognitive systematicity to be an indicator of ontological systematicity and ontological systematicity to be a causal precondi-

278

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

tion of scientific inquiry.12 In A System of Pragmatic Idealism the question whether the world is a system is left open: In the final analysis, philosophy is a system because it is concerned to indicate or at least to estimate the truth about things, and ‘the truth about reality’ is a system. Its various sectors and components are bound to dovetail smoothly with one another. Even if reality as such is not systematic, an adequate account of it must be. [. . .] It [philosophy] must accommodate and coordinate all aspects of the real in a unified and coherent whole, with the result that any viable philosophical doctrine will and can be no more than a particular component piece fitting smoothly into the wider puzzle. (III 21)

There can be no doubt that this stance is a “natural” result B but perhaps not a necessary inferential consequence B of the privileged status Rescher attributes to the epistemic dimension. In any case, it seems hardly possible to understand the rationale of a statement like “Even if reality as such is not systematic, an adequate account of it must be.” This establishes B at least as a possibility B a profound gap or separation between cognitive systematicity and reality. But then the question arises: what exactly would be an adequate account that has the fundamental feature of being systematic? An adequate account of what? This seems not at all intelligible. The most fundamental reason for Rescher’s strange thesis seems to be his implicit assumption that admitting of an ontological systematicity would lend support to the “The One-World, One-Science Argument” (I 118), which he emphatically rejects. In Cognitive Systematization he asserts that the equation “‘The world itself is systematic’ is tantamount to ‘An adequate account of the world is systematic’” does not obtain. And he presents a kind of argument by appealing to a comparison: “A poor intellectual workman can present information regarding even a simple and regular (etc.) configuration of objects in complex, disorganized form; and a clever workman may be able to describe a disorganized chaos in relatively simple and systematic terms”.13 But this is simply not the case. Rescher’s comparison is ill-explained: it does not at all support his thesis. A closer analysis of the comparison demonstrates the contrary. Concerning the first part of the comparison: what would “information in complex, disorganized form” be? Would it be real, intelligible information? Would it be real and intelligible information “regarding a simple and regular configuration of objects”? Not at all. There would be no correspondence relation between the alledged information and the simple and regular configuration; briefly stated, the alleged information

279

RESCHER STUDIES

would not be “information about. . . something”. As for the second part of the comparison. What would a “disorganized chaos” be? The usage of the surprising conjunction of the words ‘disorganized’ and ‘chaos’ is revealing. In general the expression ‘chaos’ is used in a wholly indeterminate way. Close analysis can always show that the word refers to something viewed only from a very particular and inadequate point of view, such that in that perspective this something does not appear as structured or organized in the expected manner. From this it does not follow at all that the something is, simpliciter a “disorganized chaos”. When a “clever workman” describes this allegedly “disorganized chaos” “in relatively simple and systematic terms”, he is simply removing the indeterminacy of the indeterminate conjunction “disorganized chaos”; he is fixing and clarifying the real reference of this strange conjunction “disorganized chaos” to something that is “simple and systematic”, as his description says it is. Rescher’s example/comparison is mistaken. [iv] The last of Rescher’s intuitions and insights that can be reinterpreted another way is of special importance. Rescher has repeatedly written about the limits of science, about the discontinuities in the history of the scientific process, about scientific progress, about scientific relativism, about the cognitive inexhaustibility of nature, etc. His conception of this cluster of topics is summarized in the following passage: Significant scientific progress is generally a matter not of adding further facts (on the order [of?] filling in of a crossword puzzle) but of changing the framework itself; substantial headway is made preeminently by conceptual and theoretical innovation. (I 49)

This far-reaching thesis fits very well in with one of the central theses propounded by the author in his Structure and Being: the thesis that there is a plurality of theoretical frameworks.14 Furthermore, the thesis that there is B at least for us humans B no highest or absolute theoretical framework is accepted by both Rescher and the author. But the manner in which Rescher understands the shift from one theoretical framework to another seems to be highly problematic with respect to the ontological consequences of the shift. Rescher is a staunch critic of every sort of convergentism—every view according to which “the progress of science proceeds by way of cumulative accretion (like the growth of a coral reef)” (I 49). And he explains:

280

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

If the idea that the progress of science proceeds by way of cumulative accretion [. . .] collapses, then so will the position [i.e., the convergentism] as a whole. And collapse it does. For science progresses not just additively but also largely subtractively. As Thomas Kuhn and others have persuasively argued, today’s most significant discoveries generally represent a revolutionary overthrow of yesterday’s, the big findings of science taking a form that contradicts its earlier big findings, and involves not just supplementation but even outright replacement on the basis of conceptual and theoretical innovation. (Ibid.) Given that science in the main develops not by way of culmination but by way of replacement, it makes no sense to think of the development of science in terms of convergence or approximation. [. . .] [W]e can claim neither that in inquiry we attain the truth, nor that we steadily draw nearer to this goal. (I 50).

It appears difficult to see the coherence in Rescher’s statements. First of all, it makes no sense to speak of “progress of science” (as a fact) and at the same time to characterize this “progress” as a process in the sense of a sequence of “outright replacements”. If only replacements take place, there is no progress. Nevertheless, Rescher is committed to the view that later theories are better theories (see I 48)—so there is progress. Firmly sticking to its almost exclusively epistemic stance, Rescher interprets this scientific progress by saying that it correlates with warranted confidence: “With the transition from a corpus of scientific knowledge to its successor, we neither get more of the truth nor draw nearer to it; rather, we get a better-based estimate of the truth B one that gives us a firmer warrant for our claims.” (Ibid.) But this is not satisfactorily intelligible. Indeed, why should we have a better warranted confidence in the later theory, a better-based estimate of its truth, if all matters of content B and this means: all aspects concerning the ontological import of the theory B are simply overlooked, ignored, brushed aside? What is the basis of the “better-based estimate” of the truth of the “better” theory? Rescher provides no answer to this question. If the history of science is not to be considered an immense irrational and, therefore, senseless process, it must be acknowledged that the sequence of theories is not devoid of any truth, i.e., of any ontological import. This is not to say that a (every) theory attains “the truth”. It means instead that any theory and thus every theory is the articulation of at least some aspects of the self-manifestation of the world/nature. To be sure, it is

281

RESCHER STUDIES

extremely difficult to explain this state of affairs. In his Structure and Being the author defends the thesis that every theory can be qualified as true only in a relative sense, namely, as being relative to its theoretical framework (more exactly, to the theoretical framework underlying the theory). This thesis does not imply an unlimited relativism with respect to truth, it yields only a moderate, non-selfcontradictory relativism.15 The term ‘convergentism’ is not suitable for characterizing this thesis, because it has the connotation of the approach to an end of the process, of something that will be reached or given ultimately “in the limit”. For the same reason, talk of “approximation” (to the truth), and the like, is also highly problematic and should be avoided; the process in question, with respect to its ontological dimension, must instead be considered to be an open-ended one. Rescher rightly criticizes the idea that science develops by way of cumulative accretion or of adding further truths or facts (comparable to the filling in a crossword puzzle); instead, science’s process and progress, as it relates to ontology, must be conceived of as an extremely complex non-linear process. One way of describing this process is to have recourse to the concepts of coarse-grained and fine-grained structures of the world/reality. How this view should be spelled out in detail cannot be elaborated in this paper.16 3.2 BRIEF FINAL REMARKS ON CENTRAL PHILOSOPHICAL TOPICS INSUFFICIENTLY TREATED IN RESCHER’S SYSTEMATIC PHILOSOPHY The aim of the present paper is a limited one, namely, the investigation of the theoretical framework that is either implicitly presupposed by or explicitly articulated in Rescher’s pragmatic-idealistic systematic philosophy. One aspect of this investigation addresses the question how and to what extent Rescher’s pragmatic-idealistic theoretical framework is applied. A systematic philosophy is supposed to have a comprehensive character in the sense that—at least in principle—it can encompass all the central topics that must be conceived of as philosophical topics. For this reason, it is remarkable that some of the most important areas of philosophy are either not addressed at all or are insufficiently treated in Rescher’s systematic work. [1] A first important topic that is absent from Rescher’s philosophy is the relationship between philosophy and science (or the sciences). It should be noted that Rescher, as can be seen from what is said above, al-

282

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

ways speaks of science, but never raises—let alone clarifies—the question how philosophy relates to science. What is the status of his—allegedly philosophical—considerations of science? Does he present a “theory of science” (Wissenschaftstheorie)? The clarification of this question is certainly an urgent desideratum for Rescher’s pragmatic-idealistic systematic philosophy. [2] A second central topic, this one largely absent, is language—and most especially semantics. The famous linguistic turn has never taken place in Rescher’s entire philosophical oeuvre; rather, his philosophy is characterized by what is now called the cognitive turn. Due to the cognitive character of his philosophy, the absence of semantics will not be considered a lack or fault by Rescher. On this matter the author of this essay defends a diametrically opposed view.17 Taking language—and, consequently, semantics—into explicit consideration appears to be indispensable for clarifying the status of philosophical questions and solutions. The main reason is this: if a philosophical conception is not presented, it does not enter into the actually occurring philosophical dialogue. But any conception that is presented is always—inevitably—presented by means of language. For this reason, language cannot be a purely arbitrary “tool” that simply expresses thoughts; on the contrary, the linguistic structure of sentences must have an intimate connection to the contents language expresses in order to be suitable to perform this task. The analysis of linguistic structures brings the structures of the conceptual contents and ultimately of the ontological dimension to the fore To be fair, it must be noted that after the appearance of A System of Pragmatic Idealism Rescher published a new book entitled Process Metaphysics—An Introduction to Process Philosophy,18 in which he presents in an Appendix a highly interesting sketch of a process semantics (175-182). [3] A central philosophical area that is insufficiently treated in Rescher’s systematic philosophy is the immense area of ontology. This is—at least to a certain extent—understandable, because it is difficult to see how this area could be addressed explicitly and systematically within a pragmatic-idealistic theoretical framework. This appraisal is underpinned by the critical considerations presented above in section 3.1. But Volume III of A System of Pragmatic Idealism contains a programmatic chapter titled Philosophy in Progress (III Chapter Eleven, 168-181). In the wake of Alfred North Whitehead, Rescher gives valuable indications concerning the application of the idea of process to different philosophical areas. He briefly characterizes the main fields of its application: Process Ontology;

283

RESCHER STUDIES

Process and “The Problem of Universals”; Process Philosophy of Nature; Process Psychology: Difficulties of the Self; and Process and Metaphilosophy. Though he devotes only little more than two pages (171-173) to process ontology, he clearly criticizes the substance-bias of Western philosophy and defends the view that processes are basic and things derivative: process ontology “sees things not just as the products of processes (as one cannot avoid doing) but also as the manifestations of processes—as complex bundles of coordinated processes” (III 173). This is highly promising. In his above-mentioned book Process Philosophy Rescher expounds this view to a certain extent. But because of the central significance of this philosophical area a more extensive treatment remains a very important desideratum to be addressed by Rescher’s systematic philosophy.19 [4] According to Rescher philosophy has a holistic nature. He is strongly “against dismissing synoptic questions”, i.e., against the rejectionist project that simply dismisses as “illegitimate the entire project of seeking for a reason for the existence of things” (III 8). And he has developed interesting reflections on, for example, the famous “big question,” why is there anything at all, rather than nothing? (See, for example, III 8-11) Rescher does not shy from the so-called “big questions” that have been raised and treated in the history of metaphysics. Although he does not explicitly address the question of God in A System of Pragmatic Idealism, he addresses it in other writings, most especially in the paper “God’s Place in Philosophy. Non in Philosophia Recurrere est ad Deum.”20 He defends what he calls “theistic minimalism in philosophy” based on the “principle of disciplinary autonomy”—“When we ask explanatory questions that lie within a domain we generally expect that the answer to lie within that domain”21—and abiding by the following maxim: “In transacting your philosophical affairs make do without God insofar as possible, and only bring him into it as a last resort, when there is no viable alternative.”22 This is a well-balanced philosophical stance, so far as it goes—but it does not go far enough. What about questions that do not lie within any specific domain, but that rather address the totality of specific domains? This would be the adequate philosophical way of treating the so-called question of God. Seeing the question only as a an isolated question, namely as a “religious question”, almost inevitably leads to isolated answers, i.e., answers that cannot be considered adequate from the philosophical perspective, at least as this perspective, in this respect, has been established by the mainstream of the great metaphysical tradition. This is the perspective of a comprehensive conception of being as such and as a whole.23 Such a

284

SYSTEMATIC PHILOSOPHY AND THEORETICAL FRAMEWORK

conception is lacking in Rescher’s otherwise impressive systematic philosophy. The author of this paper identifies the development of such a conception as the task most requisite for the further development of Rescher’s pragmatic-idealistic systematic philosophy.

NOTES 1

See the author’s book Structure and Being. A Theoretical Framework for a Systematic Philosophy. Translated by and in collaboration with Alan White (University Park: Penn State University Press, 2008), especially Chapter 1. (This book is the English translation of the German original edition of the book: Struktur und Sein. Ein Theorierahmen für eine Systematische Philosophie (Tübingen: Mohr Siebeck Verlag, 2006).

2

Vol. I: Human Knowledge in Idealistic Perspective (1992). Vol. II: The Validity of Values (1993). Vol. III: Metaphilosophical Inquiries (1994) (Princeton: Princeton University Press). Page references to this book are indicated in the text by the abbreviations: I page(s). . ., II page(s). . ., III page(s). . .

3

See his paper “Empiricism, Semantics, and Ontology”, first published in Revue Internationale de Philosophie vol. 4 (1950), pp. 20-40, reprinted in: R. Carnap, Meaning and Necessity: A Study in Semantics and Modal Logic (Chicago-London: The University of Chicago Press, enlarged edition 1956). Page references are to this last edition.

4

Ibid. 206.

5

Ibid. 207.

6

Structure and Being (see footnote 1), Chapter 3, section 3.2.

7

A possible misunderstanding must be avoided from the outset. The operator A Itis-the-case-that@ could be misread as applicable only to purely empirical sentences: A It is (implicitly understood: empirically) the case that. . .@ If this operator were used in this sense, however, one could not understand (for example) necessary, a priori, etc., theoretical sentences as having the form qualifiable by means of the operator. But this understanding or way of reading the operator is neither compelling nor here intended. A Empirically,@ Aa priori,@ A a posteriori,@ A necessarily,@ etc., can indeed qualify the operator. That the A such-and-such=s@ of sentences of the form A(It is the case that) such-and-such,@ without further qualification, are often understood as empirical sentences is the result of a thoughtless habit. But this is a matter of a specific linguistic custom, not a factor given with the sentence-form itself.

8

Cp. for instance I 283-286.

285

RESCHER STUDIES

NOTES 9

Alfred Tarski, “The Concept of Truth in Formalized Languages (1933)”, in A. Tarski, Logic, Semantics, Metamathematics: Papers from 1923 to 1938. (Oxford: Clarendon), pp. 152-278; see 155.

10

For a comprehensive treatment of this feature see Structure and Being, especially Chapter 5.

11

Cognitive Systematization. A Systems-Theoretic Approach to a Coherentist Theory of Knowledge (Oxford: Blackwell, 1979).

12

Ibid. Chapter VIII.

13

Ibid. 118.

14

See Structure and Being Introduction [3], sections 3.3.4.3, 5.1.5.2.2.2, 6.4.

15

See Structure and Being, section 3.3.4.3.

16

See Structure and Being, section 5.1.5.2.2.3.

17

See ibid., especially Chapter 2, section 2.2, Chapter 3, section 3.2.2, Chapter 5, section 5.1.4.

18

Albany, N.Y.: State University of New York Press, 1996.

19

In Structure and Being the author has shown that a substance ontology cannot be supported. He develops an innovative ontology fundamentally in relation to its semantics, relying on the following axiom: the semantics and the ontology of a philosophical language are fundamentally two sides of one and the same coin. See Introduction [2], Chapter 2, section 2.2.3.2, Chapter 3, sections 3.2.2, 3.2.3.

20

First published in Philosophy and Theology, vol. 12 (2000), pp. 95-115. Reprinted as Chapter 5 in volume: Nicholas Rescher, Issues in the Philosophy of Religion (Frankfurt, Lancaster, New Brunswick: Ontos Verlag, 2207), pp. 43-53. Page references are to the reprinted text in this volume.

21

Ibid. 45.

22

Ibid.l 49.

23

See Structure and Being, Chapter 5, sections 5.2 and 5.3.

286

On Rescher’s View of Idealism (and Pragmatism) Tom Rockmore

T

he Pittsburgh polymath, Nicholas Rescher, has written on many themes, but none closer to the heart of his philosophical position than idealism. This is at unusual in a period still dominated by Anglo-American analytic philosophy, one of whose founding acts more than a century ago was to reject idealism in all its forms. This ban has never been rescinded or weakened. To the best of my knowledge, no analytic thinker has ever examined the concept of idealism in detail, which they have continued to reject. Analytic figures are now returning to such German idealists as Kant and Hegel. Yet significantly, this mainly leads to interpretation and appropriation of ideas and insights from their positions that are treated as if they were unrelated to idealism or at least, as if their idealist aspects could simply be disregarded. Yet as soon as we consider “idealism,” it becomes clear that the meaning of the term is difficult to pin down. Different observers use it in different ways. Others, for instance Plato, practice forms of idealism without appealing to the term. And it is even questionable that there is any single doctrinal commitment covering all the ways that “idealism” is understood. Accordingly, the aim of this paper will be identify what Rescher means by “idealism,” to determine how that relates to kinds of idealism that are often dismissed out of hand, to measure the achievement of his position regarded as idealism, and to raise a few critical points with respect to possible difficulties. 1. RESCHER’S CONCEPTION OF IDEALISM

Rescher, who has very broad interests, has written often on idealism over a period of many years. His publications on idealism include many articles, too numerous to mention here, as well as a number of books, such as Conceptual Idealism,1 and more recently, A System of Pragmatic Idealism.2 Rather the tracing the development of Rescher’s view of idealism, the present paper will concentrate on its mature form in his recent trilogy on pragmatic idealism. We can begin with two remarks on the general title.

RESCHER STUDIES

Not all but at least some forms of idealism, for instance Kant’s critical philosophy, are strongly systematic. Yet the system-building impulse rapidly expired in the wake of German idealism. Few later philosophers are systematic in anything approaching the conception of system as it was earlier defended by Kant, under the influence of Descartes, namely rigorous organization in the form of a science constructed on the basis of a single principle or leading idea. In the twentieth century, perhaps only Whitehead even approaches such a standard. American pragmatism, both in the writings of the first generation (Peirce, James and Dewey) as well as in more recent neo-analytic pragmatisn since Quine, has sometimes been rigorous, more so in Putnam than in Rorty, but, if compared to views of system illustrated by German idealism, non-systematic. The conjunction in the title of “pragmatism” and “idealism” is also uncommon. Though there is little clarity about the term “idealism,” which is often believed to have run its course earlier in the nineteenth century, after Hegel’s death, hence before pragmatism emerged in the 1870s in Peirce’s early writings. Peirce, James and Dewey, the American pragmatists of the first generation, are very different from each other. Peirce is a highly original thinker, whose importance has arguably still not been fully acknowledged. Rorty renewed interest in pragmatism after Dewey passed from the scene when it was languishing. That was useful. But he also offered a distorted version of pragmatism, which was not useful. Peirce’s work centers broadly on philosophy of science as it was understood when he was writing, hence on knowledge. Rorty, who was an epistemological skeptic, turned attention away from Peirce and toward Dewey in the process of deflecting our gaze from probably the single most important American thinker. James, who, as Peirce’s self-appointed publicist, called attention to his position, was a philosophically weaker figure. His interpretation of Peirce’s position led the latter to dissociate it from what James said about it. Dewey, who was genuinely important, though perhaps not up to the level of Peirce, did not know much about his distinguished predecessor and struck out in a different direction. Though they are not often considered in the same breath, the link between the American pragmatists and their German idealist predecessors is very strong. James, of course, knew next to nothing about the idealists, whom he typically distrusted. Yet Peirce claimed to know the Critique of Pure Reason nearly by heart. His initial reservations with respect to Hegel were gradually overcome. He later claimed that if Hegel had had what Peirce called secondness, he would have been able to regard himself as a

288

ON RESCHER’S VIEW OF IDEALISM

Hegelian. Dewey, who began in the camp of the St. Louis Hegelians, was very critical of Kant but remained close to Hegelian themes, which he restated in a Darwinian, naturalistic idiom. There is, hence, a deep, but not often noticed continuity between German idealism, particularly Hegel, and early American pragmatism. Royce, who was there at the beginning, and who knew German idealism very well, thought that idealism and pragmatism meant about the same thing. He correctly observed early in the twentieth century that at the time pragmatism was what observers were calling recent forms of idealism.3 In this way, he points to an underlying continuity, which is not often perceived, but which is offers a key to Rescher’s position. In choosing to call his position pragmatic idealism, Rescher, who perceives the continuity between idealism and pragmatism, draws on the resources of both tendencies instead of restricting himself to merely one of them. Among the idealists, he is especially interested in Kant and among the pragmatists he is especially concerned with Peirce. In exploiting a link that is not often perceived much less developed. Rescher concentrates broadly on theory of knowledge, especially with respect to the sciences. I will be focusing on the first volume of his System of Pragmatic Idealism, which is entitled Theory of Knowledge in Idealistic Perspective. 2. CONCEPTUAL IDEALISM “Idealism,” which is certainly far from self-explanatory, is also very different from popular views of it. Idealism is more often criticized and rejected but less often studied, and not often understood. Well-known claims about idealism are frequently demonstrably in error. A well-known instance is G. E. Moore’s assertion that idealism is the view that denies the existence of the external world. The paper in which he formulated this claim is often cited, but little read, yet accorded near canonical status.4 Moore’s influential claim rests on a misreading of Kant, who objects to deficient views of idealism, which he attributes to Descartes and Berkeley, but defends good idealism, which he himself professes. Moore, who misunderstands Kant’s objection, which is directed to some forms of idealism, illegitimately extends it to idealism in general. The result is a complaint that has the virtue of being clear, but, since no one denies it, since it is in fact a straw man, applies to no single idealist position. There are, of course, many different views of “idealism,” whose relation is often not clear. The philosophical term “idealist” seems to have

289

RESCHER STUDIES

been invented by Leibniz. In responding to Bayle, he objects to “those who, like Epicurus and Hobbes, believe that the soul is material” and adds that in his own position “whatever of good there is in the hypotheses of Epicurus and Plato, of the great materialists and the great idealists, is combined here.”5 Leibniz’ usage of the terms “idealism” and “materialism” implies that they differ, but, since they are not antithetical, they can be combined in a single position. He suggests, as Fichte suggests, a simultaneous commitment to idealism and materialism (or realism).6 Such later tendencies as Marxism and Anglo-American analytic philosophy understand themselves in terms of the supposed incompatibility of idealism and realism (materialism). Leibniz, who does not regard them as incompatible, implies they are compatible. In this, of course, he is close to many idealist figures, for instance Kant, who was not only not opposed to realism as such, but described his own position as transcendental idealism and empirical realism. Since Leibniz invented the term “idealism,” the doctrine has been described in many different ways. Forms of idealism differ widely. It is at least plausible that there is no single overarching doctrine, idea, insight, or commitment to which all thinkers who sail under the idealist banner subscribe. In our time, “idealism” functions more often as a designation for theories one rejects than for a descriptive term linking together a series of thinkers under a single banner. Rescher’s conceptual idealism draws on many figures and doctrines, which he welds together in his position. His argument in favor of idealism is formulated in terms of a series of distinctions, to begin with between ontological and epistemic types. He further subdivides ontological idealism into causal and supervenient forms, and epistemic idealism into so-called fact, cognitive, strong substantival, weak substantival, explanatory and conceptual subforms. Rescher, who concentrates on epistemic idealism, is centrally concerned with the problem of knowledge. He regards idealism as a cognitive approach to getting knowledge and not, for instance, as a theory of what is. Moore’s attack on the supposed idealist denial of the existence of the external world, which is unrelated to the theme Rescher is pursuing, does not count against his conceptual idealism. Rescher’s epistemic idealism is mainly geared to his view of epistemic practice, particularly in the realm of scientific inquiry. Theory of knowledge, hence idealism, is based on the assumption of the existence of a mind-independent external world. There is no way to grasp, seize, or oth-

290

ON RESCHER’S VIEW OF IDEALISM

erwise know the world directly. Rescher denies any version of the conviction, which can be traced back at least to Plato, that under appropriate conditions at least some of the time some observers can directly grasp what is as it is. He rather agrees with those, such as Kant and Hegel, who believe that our views of what is are never immediate, but always mediated by some sort of conceptual framework, or interface, which is situated as it were between the observer and nature. In virtue of the need to approach nature through a conceptual framework, any and all knowledge claims invariably depend on the human mind. For instance, any cognitive claim is finally expressed in, or mediated by, language. All of these points come together in Rescher’s definition of “conceptual idealism”: “Any fully adequate descriptive characterization of the nature of physical (“material”) reality must make reference to mental operations; some recourse to verbal characteristics or operations is required within the substantive content of an adequate account of what it is to be real.”7 Rescher’s argument in favor of conceptual idealism includes criticism of other epistemic strategies as well as other forms of idealism he does not favor. He begins by suggesting, almost like Hegel, that whatever is real is knowable through conceptualization, what Hegel might calls concepts (Begriffe). Rescher’s view turns on the distinction between the real as such and as we know it. Knowledge claims refer to, or as he also says intend the mind-independent real world, which we cannot claim to know. In saying that we cannot claim to know the way the real is, Rescher is not, like Berkeley, saying that the world is merely mental. He is also not saying that the real is dependent on mind. In that sense, Rescher is certainly not endorsing the kind of view that is often attributed to Berkeley. Rescher is rather insisting that there is a basic distinction between the world and our view of it, and the latter is in some sense to be specified always in part mental. This view is close to a standard way of reading Kant as proposing what in modern times has often been called a causal view of perception. Hegel understands the problem of knowledge to consist in going from false appearances (Schein) to true appearances (Erscheinung). Kant, on the contrary, concentrates on the true appearance of the real, understood as the mind-independent world. He holds that the world appears to us in the form of representations that we work up into the objects of experience. From this perspective, the critical philosophy asserts the existence of a mindindependent external world. Kant, who thought that this point was denied by “bad” idealism, asserted it strongly in the “Refutation of idealism”

291

RESCHER STUDIES

added in the second edition of the Critique of Pure Reason. For the critical philosophy we do not know this world as it is in itself, but our claims for knowledge are directed toward, intend, or refer to it. Kant notoriously contends that in arriving at experience and knowledge of objects, we work up the sensory information by bringing them under an invariant group of twelve categories, or rules for synthesis. A main difference between Rescher and Kant lies in the fact that he declines the idea that there is a single fixed set of ways in which we process information in arriving at our view of the world. Another important difference concerns the concept of the subject, which for Kant is a mere epistemological principle. In this respect, Rescher follows post-Kantian idealists, who, beginning with Fichte, understand the cognitive subject to be one or more finite human beings, who are influenced in a multitude of ways by the social world into which they are born. The result is to close up a glaring gap in the critical philosophy between what, according to Kant’s epistemological commitments, is attributed to the epistemological subject and whatever finite human beings are in fact known to be capable of doing. Yet this leaves unresolved the difficulty for which Kant introduced this distinction between the claims of knowledge, in his case as based on transcendental logic, and a mere psychological account, what Husserl, under Frege’s prodding, later called psychologism. It is not clear, since Rescher does not discuss this problem, how he would handle it. One way to do so would be to sacrifice the ideal claim of perfect knowledge, which inspires opponents of psychologism like Kant, Frege and Husserl, for a more realistic view, a view “tailored” to the nature of real human subjects, which emerged in post-Kantian German idealism. This effort to “naturalize” the process of knowledge after Kant is continued in Rescher’s own “naturalistic” approach to epistemology. 3. CONCEPTUAL IDEALISM AND CAUSALITY German idealism is continuous with various idealist conceptions of science under the heading of philosophy of nature (Naturlehre). Fichte was concerned with scientific philosophy but not with science. With this prominent exception, all the great German idealists were interested in modern science, even if what Schelling makes of it is certainly problematic. When German idealism declined after Hegel’s death, the link between idealism and science was strained, even broken. The later neo-Kantian return to Kant succeeded in calling attention to the relation of Kant and sci-

292

ON RESCHER’S VIEW OF IDEALISM

ence. The work of Hermann Cohen, which was crucial, turned on uncovering the a priori elements that give rise to different types of scientific experience. Attention to the link between Kant and science was further extended in the work of Ernst Cassirer, who, other than Rescher, is arguably the only other important idealist in the twentieth century to centrally feature science. Cassirer, whose work is still under appreciated, was certainly the idealist thinker best informed about the sciences since Kant. Yet in this respect, Cassirer is an anomaly. The British idealists were mainly uninterested in modern science. This is also the case for Brand Blanshard, who continued their views under American auspices. In that sense, Rescher, in whose long career the nature and limits of science has never been less than a central theme, makes a qualified return to the earlier German idealist concern with science. Causality has long been a central idealist concern. In the Phaedo Socrates advances the notorious theory of forms to counter the explanatory deficit of modern science. Kant was famously concerned to respond to Hume’s refutation of causality that he believed undermined modern science. Causality is further a theme in his moral philosophy in the compatibilist claim that, as he put it, reasons can be causes. After Kant, a form of the Socratic objection to ancient Greek science is repeated two millennia later in Hegel’s complaint that modern science fails to explain the particular. Rescher believes that mind and causality are not only inseparable but also coordinated. He interprets this to mean that causality points to causal laws or principles, which, unlike mere generalizations, have hypothetical force, hence concern possibility. The obvious question is what possibility means with respect to nature. Rescher distinguishes between real and mere possibilities, or possibilities de re and de dicto. He suggests the former concern nature but the latter concern mind. To put the point in other language, real processes can be and are realized in nature, for instance, in the Aristotelian example of an acorn that becomes an oak. But mere possibilities, which, as he notes, are imagined, supposed, conjectured or hypothesized, possibilities that, one is tempted to say, are contrary to fact, or subjunctive, depend on mind. These mere possibilities are, since they are not real, what he calls mere fictions. But Rescher is not thinking of the wholly fictitious, but rather of what, under appropriate conditions, is not, but conceivably could, become a real possibility. One thinks here of Plato’s carpenter who imagines and then constructs a bed. To the critic, who complains that laws simply characterize what happens in the world, with regard to possibility, Rescher replies that natural

293

RESCHER STUDIES

law, which is more than a simple generalization, describes what must happen. On his view, what he calls lawfulness is about the nomic necessity and hypothetical force attributed to empirical generalizations.8 I take his claim to mean that law is not a mere description for two reasons. First, since what is described is not merely contingent, it has law-governed status. Second, what distinguishes law-governed status from merely contingent events is its if-then hypothetical structure, or its hypothetical force. Rescher goes on to note that nomic necessity and hypothetical force cannot result from mere experimental evidence. It follows that a merely empirical stance is unavailing. Though it is correct that descriptive statements about the real appertain to the real world, statements about what is necessary apply, Rescher points out, to all possible worlds. The surprising moral, which Rescher carefully draws, is that an analysis of the lawfulness of physical nature leads beyond traditional scientific empiricism to the traditional idealist view that, as Rescher puts it, “causality is mind dependent.”9 In this context, Rescher alludes to Kant in claiming that a network of possibility points to the category of possibility. Perhaps. Certainly, possibility is one of the modalities Kant considers in his list of categories. There is perhaps an important non-Kantian element as well. Rescher stresses, in a way similar to the deductive-nomological model of scientific explanation popularized by Hempel and Oppenheim, that if the antecedent is present the consequent must follow.10 He seems, then, to believe that the nature is not only independent of us but that its laws are also mind-independent as well. Kant, however, notoriously claims that modern natural science entered on the secure path of science as we now understand when it was realized that we only know what we in some sense construct. A corollary of this principle, which lies at the epicenter of the so-called Copernican turn, to which I will come back below, is that in Kant’s view we literally prescribe laws to nature. If this is true, we do not, as Newton claimed, derive laws of nature directly from the empirical world, but rather impose them so to speak. 4. CONCEPTUAL IDEALISM AND CONSTRUCTIVISM The point concerning Kant turns on the concept of constructivism. Kant can be interpreted as committed simultaneously to two different approaches to knowledge: representationalism, or the view that knowledge depends on correctly representing the mind-independent external world, and constructivism, or the view that knowledge depends on constructing

294

ON RESCHER’S VIEW OF IDEALISM

what we know. The first approach is related to the well-known causal theory of perception, with respect to which Rescher responds by calling attention to the difference between what we know and the way the mind-independent external world is. Very much like Kant, Rescher contends there is an unbridgeable difference between what, on the basis of our theories, we take to be the real and the mind-independent real that we must presuppose but cannot claim to know. The second approach concerns the sense in which for Kant the subject in some sense constructs what it knows as a condition of knowing it. This leads to two very different approaches of knowledge cohabiting as it were in the critical philosophy. One approach is the idea that we find, discover or uncover the cognitive object. The other very different approach is that we construct, make, or construct what we know. The former view relies on consciousness of the object, which is fully constituted in independence of the cognitive subject. The latter strategy depends on the idea that in knowing an object, we are conscious of ourselves, or self-conscious. This is a form of the famous philosophy of identity (Identitätsphilosophie) that developed after Kant. The philosophy of identity is continuous with Kant’s Copernican turn, or the view that we know only what we in some sense construct. The point is that for Kant the world we do know is constructed as it were through the interaction between the subject and the mind-independent world. Rescher implicitly responds to this Kantian claim in further elaborating conceptual idealism. Constructivism goes all the way back to ancient Greek mathematics. Geometry relies on the construction of plane figures. Constructivism comes into modern theory of knowledge in Hobbes and Vico. Kant, who comes to constructivism independently, applies his view of mathematics as based on the construction of concepts to the solution of the problem of knowledge. The difficulty in any form of constructivism is how to make sense of the idea that knowledge if and only if the cognitive subject constructs its cognitive object. Kant, who is unable to clarify this point, compares construction, in a famous passage, to “a hidden art.”11 Rescher clarifies this point in a further elaboration of his position in distinguishing conceptual idealism from spiritualism and panpsychism. The mind, he believes, “contributes essentially to the constitution—as well as the constituting—of our knowledge of reality by characterizing things in its own terms, in using inherently mentalistic conceptions and paradigms.”12 Rescher here goes beyond

295

RESCHER STUDIES

his view that an adequate description of nature crucially relies on mind in indicating its specific role. I interpret this statement to mean, not that mind literally constitutes nature or cognitive objects. This is consistent with his vision of conceptual idealism as epistemic and not ontological. The role of mind rather lies in constituting knowledge of reality, or, in a word, reality as we know it as opposed to the way it is. The way reality is for us depends on conceptual models, or concepts and paradigms that we come up with. I take it that what this amounts to is the familiar point that there are things, which we characterize in various ways by adopting different models, concepts, and ways of depicting them. Yet if this is indeed a fair exposition of Rescher’s view, then it requires further clarification. If there is a difference in kind between mindindependent objects that we take to exist in nature and the objects of nature as we claim to know them in our theories about the world, we cannot claim there are cognitive objects, which we then characterize. We do not experience mind-independent nature, and we can only make cognitive claims about objects that we do experience. Yet there is tension between the very useful suggestion that nature as we know it depends on the mind of the observer, and the further idea that we characterize things in various ways. On Rescher’s own view, we do more than characterize things by describing them through “mentalistic” conceptions and paradigms. In using “mentalistic” conceptions and paradigms we in effect not only bring these objects into cognitive focus. We do not only constitute our knowledge of them, since we cannot know objects as they are and only know them as they are given to us in cognitive experience. We further constitute the objects simultaneously with our knowledge of them. The point then is that what we claim we know, more generally our understanding of, hence knowledge of, nature as experienced is not independent of, but rather dependent on, the conceptual frameworks, explanatory models, paradigms and other ways we routinely employ to cognize nature. The corollary of this view, which I think is part of what Rescher is saying, can be stated in two related points. First, if the object depends on the theory about it, then different theories lead to different objects; and, second, a change in theory leads to a change in objects, hence, if the objects comprise nature, literally to a change in nature. In abstract form, Hegel expounds this general doctrine. He affirms that, when a theory is tried out in practice by comparing it to experience, there are in fact only two possible outcomes. Either in practice experience corresponds to our theory about it, or, on the contrary, the theory fails that test. In the latter case, the theory

296

ON RESCHER’S VIEW OF IDEALISM

needs to be reformulated to bring it into line with observation. But in its very reformulation the cognitive object or objects to which the former theory applied is transformed.13 Hegel’s claim is a version of the basic idealist principle, which is already stated before him in the German tradition by Kant, Fichte and Schelling, that we construct what we know. Though Hegel claims to be presenting a phenomenology of conscious experience, his statement is certainly obscure. It is given a more popular formulation by Kuhn minus the historical dimension on which Hegel strongly insists. In a famous passage, Kuhn remarks on the difference between the views of Priestley, who relied on phlogiston, and Lavoisier, who discovered oxygen. According to Kuhn, it is not that they saw nature differently but rather that in a sense they lived in different worlds.14 The important point is that there is no invariant world to which we refer and that, as Putnam thinks, we know from different angles of vision.15 For as Collingwood points out, what we mean by “world” is a historical variable.16 Just as the idea of world is a historical construct, which in turn depends on the adoption of one or another perspective in a given time and place, so also the cognitive object literally depends on the theory for which it is an object. If a change in the theory brings about a change in the object, for instance in the difference between the ordinary view of water and the chemical view of H2O, then it follows that we do not apply our theories to stable objects. We rather apply our theories to objects that, in turn, are as it were “constructed” by the theory about them. Thus if the ongoing research at CERN in Switzerland is successful with regard to determining the existence of the Higgs boson, it will contribute toward extending the so-called standard view of matter that is currently in vogue. But what we mean by this sub-atomic particle is inseparable from, hence dependent on, our ability to detect it. In other words, the particle itself literally depends on the subjacent categorial framework, in short the current view of physical theory, in terms of which this idea is meaningful. And if at a later date this underlying theory is abandoned then it will be necessary to give up the idea of the Higgs boson that will, when this occurs, vanish as well. 5. REALISM AND IDEALISM Rescher devotes careful attention to the relation of idealism and realism. The alleged incompatibility of idealism and realism is one of the main rea-

297

RESCHER STUDIES

sons why idealism is often rejected. The argument against idealism, which takes many forms, can be reconstructed as a four-step critique. The first step lies in the acceptance of the idea that there is a mind-independent world. The second step consists in claiming that knowledge consists in knowledge of the mind-independent world as it is. The third step is the claim that idealism, which denies the reality of the mind-independent world, hence realism understood as the claim for the existence of the real, is incompatible with ordinary experience and claims to knowledge. The fourth step lies in claiming that common sense tells us that the world, whose existence is denied by idealism, in fact exists and is known to exist, so that idealism conflicts with common sense. This argument presents philosophical arguments about realism as the scholarly reformulation of ordinary beliefs. Rescher does not consider this argument. Yet his account of the relation of realism and idealism contradicts at least three, perhaps even all four of its steps. Though he says nothing about common sense, he also does not support the idea that it is, as someone like Moore believes, the court of last appeal. Common sense suggests that ordinary individuals, those who are not philosophically sophisticated, in fact know, even directly know the way the world is. Moore, for instance, thought he knew with certainty that he had two hands, though it is not clear that anyone, certainly any idealist, denies this claim. Moore and many other critics of idealism depend on common sense to refute idealism. Now common sense relies on the idea that there is a mind-independent external world and that under appropriate circumstances at least some of the time we can reliably claim to know it as it is. But Rescher denies common sense as the road to knowledge on the grounds that there is a distinction between nature as it may exist and nature as we know it either through science, common sense, or in some other way. He further fails to endorse the idea that we know the mind-independent world exists. This is something we could only know if in some way we could cognize it. He believes, on the contrary, that the existence of the mind-independent external world cannot be known, but must be assumed as that which science is about. To put the point in other language, science is about a world whose existence is presupposed as that which science studies but which simply cannot be demonstrated. In essence, he finds that the existence of the world is an indispensable scientific hypothesis, which is the condition of science as it were, but which cannot be proven through science, common sense, or in any other way.

298

ON RESCHER’S VIEW OF IDEALISM

This view is clearly comparable to Kant’s idea, which underlies his view in the famous Herz letter (21 February 1772) that the problem of knowledge consists in analysis of the relation of representations to objects. Kant, who understands causality along the lines of Leibniz’s principle of sufficient reason, thinks that objects of experience are representations, which can be considered without contradiction to be effects caused by a mind-independent external world. The difference is that Kant clearly believes, but acknowledges we cannot know, that there is such a world which impacts on us. In this respect, he is caught in contradiction on several levels. On the one hand, he needs to, but cannot, show that the mindindependent world is the cause of knowledge. On the other hand, his causal analysis of perception runs against his own theory that at least “officially” restricts causality, which for Kant has categorial status, to the perceptual level. Rescher’s analysis is preferable here since in interpreting reality as a mere hypothesis he avoids difficulties arising out of any form of an ontological commitment, which, Kant notwithstanding, simply cannot be made out. The price, however, is that although he does not deny the existence of the external world, he also cannot affirm it. Kant, who refuted those who, in his view, were not committed to the existence of the external world, would have found the latter point unacceptable. The problem is what, if anything, we can know about the way the world is on Kant’s or a similar theory. Kant seems to be of two minds on this crucial question. He both claims that, since all knowledge necessarily begins in experience and there can be none of the noumenal world, we can know nothing about it. Yet he also claims, under the heading of the double aspect theory, according to which appearance and reality are two sides of the same coin as it were, that appearances are appearances of mindindependent objects.17 Kant’s twin claims are clearly inconsistent. Since he has no way to show that the double aspect theory is true, he cannot invoke it to claim we do know things in themselves. A claim of this kind would be inconsistent not only with his understanding of noumena but further with the entire thrust of his effort in the critical philosophy to limit cognitive claims to the bounds of experience. Rescher avoids this difficulty in the critical philosophy in affirming against those committed to the view that science discovers the world as it really is that there is no reason to think that we ever can close the gap between the world as it is and as we know it. In fact, as he points out in remarks on what he calls “the cognitive opacity of real things,” which are simply inexhaustible, there is no reason to think

299

RESCHER STUDIES

we can ever approach anything like cognitive finality.18 Despite repeated dreams of a final theory, which have never been stronger than at present, there is no reason to think we are closer to that day, nor even that any particular “constative” claim is beyond possible future revision. On this version of idealism, there is no epistemological closure since the future is wholly open. Rescher then agrees with such idealists as Hegel, Bradley and Royce that we do not know and probably cannot ever reach absolute cognitive claims. Though frequently mentioned, there is nothing in the literature approaching agreement about the meaning of “realism.” This term is often (mis)used to refer to what is understood as metaphysical realism. Michael Dummett, who raises questions about Platonic realism, a close synonym of metaphysical realism, is often understood as an anti-realist. There are many kinds of realism. Rescher devotes individual chapters to metaphysical realism and scientific realism. He defines “metaphysical realism” in what is currently standard manner as the doctrine that the world as it is, as it exists, is simply independent of thought, of whatever we may think about it. This is not an inductive inference, not a result, but a regulative presupposition of scientific inquiry. Science needs to postulate an objective order of mind-independent reality for at least six reasons.19 The assumption of realism is validated retrospectively since it leads to practical success in achieving theoretical understanding. If we did not assume a mind-independent real world, we could not study it. Yet study of the world is useful in generating results about it validates this assumption. That leads to the conclusion, which I think parenthetically only the most blinkered observer would deny, that values and purposes play a role in science.20 Rescher’s view of metaphysical realism is certainly innovative. There are still some philosophers as well as some scientists who think that we reach knowledge by simply unveiling nature. Yet it is becoming increasingly infrequent for observers to claim to know the mind-independent world as it is. Rescher’s sensible view is useful in countering indirect efforts to make what can be regarded as a similar point, for instance Davidson’s view of triangulation, which, among other things, enables observers to fix the object as it were.21 Rescher’s helpful counterclaim is that we may or may arrive at a view of the world we can accept, but we can never argue that this is the same as the world. Unlike scientific realism, or the version of things as science sees them, which bases itself on the substantive knowledge of the sciences, meta-

300

ON RESCHER’S VIEW OF IDEALISM

physical realism is justified by the shortcomings, or imperfections, of scientific knowledge. As concerns scientific realism, Rescher’s effort is geared to seeking out a middle point between untenable forms of scientific realism predicated on the idea that science describes physical reality and an instrumentalism subscribing to the thesis that natural science merely coordinates phenomena. Rescher sensibly holds that our current body of scientific beliefs consists largely of false beliefs. This is the obvious reply to the suspicion that science is close to reaching an end. Since there is no way to know how theories relate to the world whose existence we presuppose, there is also no way to know whether we are in fact getting closer to knowing it. In pointing to the limits of science, he is especially concerned to distinguish between the world we describe in science and the world as described in science, or again the realism of intent and the realism of achievement, which should not be conflated. In this connection, Rescher examines the ancient view of man as the measure to which he objects that what we mean by reality must be linked to causality.22 He is certainly correct that we ought not to “relativize” our claims to know to anyone at all. Yet, once we accept that the real cognitive subject is one or more finite human beings, who not only formulate, but further evaluate and accept or reject cognitive claims, we also cannot get away from some form of the doctrine of homo mensura. In other words, what we mean by causality and a successful causal inference, and what we in practice accept as meeting these standards, is not independent of, but clearly relies on a version of the view of man as the cognitive measure. Rescher ends his remarks on the relation of realism by introducing what he calls schoolboy realism. This is a form of realism that cannot claim to characterize reality correctly, but that also cannot be dismissed as wholly in error. The result is to argue, like all the great German idealists, that we must “invest” in realism in the science we have in view of ideal science, which, unlike present science or any that could occur, is not and may never be complete. 6. IDEALISM AND PRAGMATISM So far I have been concentrating on Rescher’s view of idealism. I would like now to say a word about his view of pragmatism. Like Peirce, one of his strong influences, and with whose ideas he is in constant dialogue in working out his own view of idealism, Rescher’s views of idealism and

301

RESCHER STUDIES

pragmatism are strongly related. It is not well known that Peirce had deep links to idealism, especially to the German idealist tradition. He knew Berkeley, about whom, he wrote very well. He had a very thorough grasp of Kant. Though he was early very critical of Hegel, as he grew older, he came to believe that the differences between him and his German predecessor were mainly verbal. Almost as soon as he began to think about the problem of knowledge, Peirce came up with the description of the cognitive process as one that navigates between doubt and belief. The relevant question is what is at stake over time in reaching belief, that is, how belief relates to the real, or again what it is that we can say that we know when in the fullness of time the investigative process converges on a belief. In attacking this problem Peirce makes two important points as concerns what we mean by an object, and how that meaning relates to the world. The second point is important to Rescher’s view of idealism. Peirce spells out his view of an object in a passage, which is repeatedly cited and often seen as expressing a central insight of pragmatism, he writes: “Consider what effects, which might conceivably have practical bearings, we conceive the object of our conception to have. Then, our conception of these effects is the whole of our conception of the object.”23 Peirce immediately connects this view of an object as understood in terms of its practical effects to the idea of reality. Reality can be understood in two ways: as what is independent of you or me—he says what you or I think about it24--and what, on the assumption that the correct application of scientific method will produce no more than a single result, will eventually be accepted by all concerned at the end of the inquiry, in a word in the long run. The question is what is finally the upshot of the scientific process over time. In an early paper, Peirce controversially contends that “the opinion which is fated to be ultimately agreed to by all who investigate, is what we mean by truth, and the object represented in this opinion is the real.”25 One can interpret Peirce’s claim in various ways. One way is to regard it as a mere joke since strictly speaking the process of science never comes to an end in time, so no opinion is ever finally agreed on by everyone. If this is the case, then Peirce is not addressing, but rather only circumventing the question. A second way is to regard Peirce as claiming that at the end of the day all investigators will converge on a single view, a result that in fact discloses the real. Rescher, who discusses this reading of Peirce’s theory several times, properly objects that, since we cannot overcome the dif-

302

ON RESCHER’S VIEW OF IDEALISM

ference between our view of the real and the real, Peirce’s claim for the results of scientific inquiry goes too far. We may indeed agree on what we take the real to be as the outcome of the process of inquiry, but it does not follow we have agreed on the real. In other words, even though we may agree, it does not follow, as Rescher correctly points, that we agree about reality. For even if we agree, we may still be deluding ourselves about what is correct.26 In other words, in scientific practice epistemic fallibilism remains a permanent possibility that simply cannot be vanquished through some particularly perspicuous scientific theory. 7. CONCEPTUAL IDEALISM AND PRAGMATIC IDEALISM This paper has focused on identifying the main features of Rescher’s conceptual pragmatism. Rescher’s work centers roughly on philosophy of science and related issues in suggesting an idealist alternative for a debate where, with the prominent exception of Cassirer, idealism of any form has been largely absent over the last hundred years. Pragmatic idealism is intended to combine ideas from related sources in idealism, particularly German idealism, above all Kant, as well as American pragmatism, especially Peirce. Rescher’s minimal claim is that mind is ingredient in the construction of our cognitive claims. The significance of this claim can be seen against the background of the contemporary turn away from subjectivity toward objectivity. Analytic thinkers like Thomas Nagel27 believe that we can disclose objectivity as it really is since we can somehow subtract subjectivity from it. And historians of philosophy like Frederick Beiser28 contend that post-Kantian German idealism represents the decline and fall of objectivity through increasing subjectivity. These and other contemporary thinker share the illusion that we can and must somehow center our efforts on objectivity that consists in revealing the mind-independent world as it is and in opposing views that deny this goal. This approach is linked to a kind of neo-Cartesianism that consists in a direct inference from ideas about the world to the world. But there is another lesson in Descartes, which can be paraphrased as the message that the road to objectivity necessarily runs through subjectivity. Rescher’s work is centered on the effort, which develops in Kant and expands radically in the post-Kantian German idealists to take the subject seriously while working out a conception of objective cognition. This effort is prolonged in a different way in American pragmatism, which, like post-Kantian German idealism, features the primacy of practice over the-

303

RESCHER STUDIES

ory, and as concerns science the primacy of a practical, a posteriori approach to knowledge over a more theoretical, a priori approach. Conceptual idealism belongs to this effort. It is another step in the ongoing process of developing a theory that relates idealism and realism, the contribution of the cognitive subject to constituting what it knows, and the role of the real in the process of arriving at knowledge. Depending on how we interpret Kant, Rescher’s solution remains close to the critical philosophy. In suggesting that the real is not and cannot be disclosed in the cognitive process through which it is represented, Rescher provides an updated version, two centuries later, of an analysis of the relation of representations to objects, which can never be known. As for Cassirer, for Rescher as for Kant theories symbolize but do not cognize the real. It has not been possible to mention more than selected aspects of Rescher’s view that he has worked out over many years in a large number of writings. Yet, even on the basis of this limited account, several points suggest themselves as concerns the relation of pragmatism and idealism, idealism and scientific inquiry, and how to do philosophy of science. Above I have mentioned Royce’s view of the continuity between idealism and pragmatism. In calling attention to the continuity between pragmatism and idealism, thus confirming Royce’s view, Rescher strengthens the understanding of idealism as itself pragmatic since it turns, as Kant already knew, on the relation of theory and practice. Marx did not discover this problem, to which he gave a name, and which runs throughout the writings of all the major figures of this period. In applying idealism to epistemology, Rescher, shows the resources of this approach for the contemporary discussion of knowledge. Though this is not his intention, in this respect his work is complementary to recent efforts to call attention to the epistemological interest of such major idealist figures as Hegel. One should not forget that more than thirty years ago in one of his first major works on idealism Rescher suggested that importance of what he called a Hegelian inversion in working out a viable concept of system. At a time when philosophy has largely turned away from system, hence forgetting an important theme in Kant, in describing a generally Hegelian approach Rescher identifies the importance of turning from a conception of system as build up out of previously secured items of knowledge toward a view of system in which what counts as knowledge depends on the relation of items to an overall, or holistic framework. A second theme concerns the application of idealism to science for the first time in many years, and in that way its role in renewing the great ide-

304

ON RESCHER’S VIEW OF IDEALISM

alist tradition of scientific inquiry. In a time when idealism is largely absent from the debate, hence absent from contemporary philosophy of science, Rescher renews with the great idealist tradition in focusing once again on science from an idealistic perspective. Plato, Aristotle, Leibniz, Kant, Hegel, and Cassirer are some among the names of the great idealists who in the past have contributed to working out an idealist theory of science. With the exception of Plato, each of them in some way makes knowledge of reality depend on ideas that are not found, discovered or uncovered, but rather constructed, constituted, made or shaped by the mind. It might seem odd to include Aristotle since we tend to forget his pioneer claim that, though it seems absurd, in a certain sense the objects that we perceive are in the mind, a claim that, hence, suggests a distinction between what we know and what is.29 The exception is perhaps Plato, who, though he relies on the distinction between appearance and reality, simply identifies reality and forms or ideas. Unlike modern types of idealism, which, with the possible exception of Berkeley, tend to emphasis the mind’s role in constituting cognitive object, for Plato’s position, if indeed there is a position, ideas seem to be always already constituted in independence of the human mind that can at most only know them. His theory is hence very different from and incompatible with the later constructivist move worked out independently in modern philosophy by Hobbes and Vico, then Leibniz, followed by Kant, and a long list of postKantian constructivists, including at least Fichte, Schelling, Hegel, Marx, Cassirer, perhaps Peirce, who sometimes seems committed idealism,30 Dewey and now Rescher. Ever since Plato, there has been a strong link between many forms of idealism and science. The introduction of the theory of ideas as a replacement theory for current views of science, whose advantage lies in providing a better account of the causal origin of things, was mentioned above. Many, though not all modern idealists have been interested in and contributed to our understanding of science. Rescher’s conceptual idealism is specifically interesting for renewing attention to the relation of idealism to modern science. This point can be advantageously stated in reference to Kant and the post-Kantian German idealist reaction. Above I pointed to Kant’s view that cognition as based on constructing the cognitive object, which can never be simply identified with mind-independent reality. Later German idealists turn away from Kant’s conception of the thing in itself, hence from any effort to know the world as it is in restricting attention to experience. Their

305

RESCHER STUDIES

theories, like those of later idealists, tend to be variations on the general constructivist theme. At the conceptual heart of Kant’s Copernican turn is the insight that we construct what we know. Rescher’s conceptual idealism draws on and carries further this insight in applying it to current views of science. This point, which is not only of antiquarian interest, concerns the very nature of scientific knowledge. To those concerned to know the mindindependent world as it is, idealist constructivists oppose two points already made by Kant. On the one hand, if the object is already fully constituted in independence of the cognitive subject, then there is in fact no way reliably to claim to know it. Yet we can reliably claim to know what we in some sense construct, above all as a byproduct as it were in working out our theories of experience and knowledge. In this respect, the merit of Rescher’s conceptual idealism is two-fold. Current philosophy of science tends to run along representational lines in seeking to show that through the formulation of scientific theories science successfully grasps the real. This identification of our knowledge of the world with the world takes many forms. They include insistence on forms of metaphysical realism, scientism understood as the idea that science and only science as the source of knowledge, reductionism of various types such as extensionalism and physicalism, and so on. The alternative, which Rescher illustrates, lies in differentiating what we know at any given moment from what is, hence thought from being, which, despite claims to the contrary, are never finally brought together. He shows on the one hand that the now standard view of science from the perspective of metaphysical realism is fraught with difficulties. And, on the other hand, he further shows the considerable interest of the idealist alternative, which consists in pointing out that through the mind we constitute what we claim to know. In this respect, Rescher’s careful work on aspects of modern science in many volumes over decades is literally unrivaled. I conclude that the considerable merit of Rescher’s conceptual idealism lies in renewing the idealist, constructivist approach to knowledge in the domain of modern science.

NOTES 1

See Nicholas Rescher, Conceptual Idealism (Oxford: Blackwell, 1973).

2

See Nicholas Rescher, A System of Pragmatic Idealism (Princeton: Princeton Uni-

306

ON RESCHER’S VIEW OF IDEALISM

NOTES

versity Press, 1992-3), including Volume I: Human Knowledge in Idealistic Perspective (Princeton: Princeton University Press,1991); Volume II: The Validity of Values: Human Values in Pragmatic Perspective (Princeton University Press, 1992); and Volume III: Metaphilosophical Inquiries (Princeton University Press, 1994). 3

See Josiah Royce, Lectures on Modern Idealism (New Haven: Yale University Press, 1964), p. 85.

4

See “The Refutation of Idealism,” in G. E. Moore, Philosophical Studies (London: Routledge and Kegan Paul, 1922).

5

G.W. Leibniz, Philosophische Schriften, Vol. IV, edited by C. I. Gerhardt (Berlin: Weidmann, 1883), p. 238.

6

See “First Introduction to the Science of Knowledge,” in J. G. Fichte, The Science of Knowledge, edited and translated by Peter Heath and John Lachs (Cambridge: Cambridge University Press, 1982), pp. 138-39

7

Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, p. 305.

8

Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, p. 315.

9

Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, p. 317.

10

See Carl G. Hempel and Paul Oppenheim, “Studies in the Logic of Explanation,” Philosophy of Science, number 15, (1948),pp. 135-75; reproduced in Carl G. Hempel, Aspects of Scientific Explanation (New York: Free Press, 1965).

11

See Immanuel Kant, Critique of Pure Reason, translated by Paul Guyer and Allen Wood (New York: Cambridge University Press, 1998), B 181, p. 273.

12

Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, p. 317.

13

See the Introduction to G. W. F. Hegel, Phenomenology of Spirit, translated by A. V. Miller (New York: Oxford University Press, 1977).

14

Thomas Kuhn, Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962), p. 118.

15

See Hilary Putnam, Knowledge, Truth and History (New York: Cambridge University Press, 1981).

16

R. G. Collingwood, The Idea of Nature (New York: Oxford University Press, 1965).

307

RESCHER STUDIES

NOTES 17

See Kant, Critique of Pure Reason, B xviii-xix, p. 111.

18

See Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, pp. 248-251.

19

See Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, pp. 263-264.

20

See Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, p. 271.

21

See Donald Davidson, “Three Varieties of Knowledge,” in A. Phillips Griffiths , editor, A. J. Ayer Memorial Essays: Royal Institute of Philosophy Supplement 30 (Cambridge: Cambridge University Press, reprinted in Donald Davidson, Subjective, Intersubjective, Objective, Oxford: Clarendon Press, 2001).

22

See Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, p. 271.

23

“How To Make Our Ideas Clear,” in The Essential Peirce (Bloomington: Indiana University Press, 1998), I, p. 132.

24

See “How To Make Our Ideas Clear,” in The Essential Peirce, I, pp. 138, 139.

25

“How To Make Our Ideas Clear,” in The Essential Peirce, I, p. 139.

26

See, e. g., Rescher, A System of Pragmatic Idealism, Volume I: Human Knowledge in Idealistic Perspective, pp. 272-274

27

See Thomas Nagel, The Last Word (New York: Oxford University Press, 1997).

28

See Frederick C. Beiser, German Idealism: The Struggle Against Subjectivism, 1781-1801 (Cambridge: Harvard University Press, 2002).

29

See Aristotle, De Anima, III, 2, 317 and 331.

30

It is not well known that Peirce on occasion claimed his pragmatism was idealism in a Hegelian sense. He makes this point in different ways in the following two passages: “I will so far follow Hegel as to call this science phenomenology although I will not restrict it to the observation and analysis of experience but extend it to describing all the features that are common to whatever is experienced or might conceivably be experienced or become an object of study in any way direct or indirect.” C. S. Peirce, Collected Papers of Charles Sanders Peirce, 5.37. see also his passage at 5.39: “The science of Phenomenology is in my view the most primal of all the positive sciences. That is, it is not based, as to its principles, upon any other positive science. By a positive science I mean an inquiry which seeks for positive knowledge, that is, for such knowledge as may conveniently be expressed in a categorical proposition.”

308

Rescher on Arabic Logic Tony Street

T

he study of Arabic philosophy has come on by leaps and bounds in recent years, as is clear from a bibliography of the field compiled in 1999 by Hans Daiber.1 Of course, the work going on at the moment depends in large measure on the great pioneering studies of years ago. Those studies are dutifully and compendiously listed by Daiber, and among them are thirty-five publications by Nicholas Rescher, mainly on Arabic logic.2 I do not intend in this paper to go through more than a few of these contributions; indeed, I limit myself to Rescher’s overview history of Arabic logic, the Development of Arabic Logic, and his series of studies on Arabic modal syllogistic as laid out by Nağmaddīn al-Qazwīnī al-Kātibī (d. 1276; henceforth Kātibī) and Muhammad ibn Faydallāh aš-Širwānī (d. 1707).3 In what follows, I direct a number of critical remarks to aspects of these contributions, and it will be helpful if I make clear at the outset the nature of the debt that students of Arabic logic owe to Rescher. None of the current work on Arabic syllogistic would be possible without Rescher’s studies. And even today, no one dares write about Arabic logicians, particularly those after Avicenna, without citing Development. With Rescher’s work on Arabic logic, in short, we are confronted by one of those rare cases in scholarship where a scholar has single-handedly opened up a field. 1. THE DEVELOPMENT OF ARABIC LOGIC Development came out in 1964. It consists of two parts, the first, a survey of Arabic logic (pp. 15-82), the second, a register of Arabic logicians (pp. 83-255). While the first part resembles other modern histories of logic, tracing pedagogical affiliation and attempting to isolate important disputed points, the second looks very like a work from the grand Arabic tradition of tabaqāt, or bio-bibliographical literature. The register offers a biography of each logician, a list of his works, and lists of relevant translations and studies. A review at the time of its publication helps us appreciate, fortythree years later, how welcome the book was, and why. The author’s interest in substantive logical questions is evident, especially in Part I. Nevertheless, his main contribution consists in bringing together and

RESCHER STUDIES

making available to future students of Arabic logic the vast amount of literature that has been written in Western languages by Arabists and historians of logic alike. It is thus an indispensable tool for anyone who plans to do further work in the field.4

It was indispensable in 1965 and so it remains in 2008. Obviously, the sixty-seven pages of the survey in Part I cover many periods, scholars and translation-circles about which a great deal more is known now than in the early 60s. I have no intention of listing points that the editor of a revised edition of Development would need to pick up, a project which would make worthy but excruciating reading. I prefer to comment on, first, the method Rescher followed in writing his survey and, secondly, a substantial problem in the historical narrative that Development proposed. 1.1 APPROACHES TO WRITING THE HISTORY OF ARABIC LOGIC Development came out exactly thirty years after the first edition of Madkour’s overview of Arabic logic, L’Organon d’Aristote dans le monde arabe, and five years before L’Organon’s second edition. Rescher and Madkour follow two extremely different approaches to writing logical history, and reflection on the major difference between them is still helpful for anyone working on the history of Arabic philosophy. Madkour assumed that logic is somehow separate from other philosophical doctrines, not fitted out to serve given metaphysical purposes. He wrote of the logicians writing in Arabic that Il serait fastidieux de suivre ces logiciens dans leurs divers exposés; d’ailleurs, il n’y aurait pas grand intérêt à mettre un tel projet à exécution; car si les philosophes musulmans diffèrent entre eux en ce qui concerne certains problèmes physiques ou métaphysiques, ils sont tous d’accord sur les grandes questions logiques.5

Madkour took this to mean that it only remained to find a paradigm author to give a systematic outline of Arabic logical doctrine. Madkour would have preferred if that author could have been Alfarabi, but given the fragmentary nature of his surviving writings, it has to be Avicenna . . . qui représente à juste titre l’école arabe et offre une doctrine complète sur laquelle on peut se prononcer aisément. Ses écrits, que nous avons en main,

310

RESCHER ON ARABIC LOGIC

présentent les différentes manières dont les philosophes musulmans ont traité la logique aristotélicienne; ils en contiennent des abrégés très précis et des commentaires assez étendus. Ibn Sīnā est surtout le philosophe de langue arabe, et sa logique est encore aujourd’hui enseignée dans les écoles musulmanes. . . 6

Let me dwell briefly on why Madkour’s approach cannot produce a helpful history of Arabic logic. There is no “Arab school”, unless by “school” one means a range of dramatically different systems that either talk past each other or squabble about fundamental and foundational issues. One need only consider the differences between Avicenna and Averroes on modal syllogisms, on hypothetical syllogisms, on the importance of Aristotle’s writings. . . In section two I’ll be returning to one of many disputed points within what might loosely be called the Avicennan school of logic, typical of a number of major questions on which even pedagogically affiliated logicians could not agree. Madkour’s approach makes Rescher’s alternative approach seem all the more clear-sighted. Rescher assumed that logical differences map precisely onto the differences among philosophical schools; to write a logical history of the realms of Islam, one need only write their more general philosophical history. Thus we find Rescher in Development tracing the filiations of the philosophical schools, and placing logicians and their writings onto the genealogy produced. Logicians inherit as proponents or opponents of their logic the same philosophers who promote or oppose their metaphysics. This approach commends itself for at least two reasons. In stark contrast to Madkour, Rescher provides an account that finds space to accommodate the vast differences mentioned above that we find among the various Arabic logicians. Further, it assumes that logicians will be working on their logics with some larger philosophical purpose in mind; that is, the logic will have at least some relation to other work that each logician has in hand. For all its merits, one needs to be wary of a few of Rescher’s methodological assumptions. First, the notion of “school” as a way of collecting groups of logicians is fine, so long as it is recognized that these “schools”, particularly later on, may have had no fixed point of convention, no set curriculum, and no doctrinal unanimity. They tend to be united only by pedagogical lineage, itself often tenuous. The Baghdad school is closest to being a school in our usual sense of the word, but it is unclear that its teaching practice was common later on and further east. Secondly—and this is not so much a criticism as advice to future historians of logic— mainly due to his laudable conviction that divisions among logicians fol-

311

RESCHER STUDIES

low divisions among philosophical schools more broadly conceived, Rescher tended to write his history from bio-bibliographical accounts of Islamic philosophical history. But post-Aristotelian logicians generally speaking are given to referring to other logicians, and the logicians writing in Arabic are no exception. Such references make us modify aspects of Rescher’s narrative, and will ultimately provide us with preferable sources for writing the next history of Arabic logic. I doubt that we’ll end up with such rigid isomorphies between metaphysical and logical schools. Finally, the periodization of Rescher’s study (800 to 900, 900 to 1000, and so on) does not even have the doubtful merit of following the temporal boundaries of ruling dynasties. Abbasid imperial policy began the earliest sustained efforts to translate logical works,7 but by the end of the second Abbasid century (950), it was debate about that translated logical problematic conducted among logicians writing in Baghdad and elsewhere that determined the further fortunes of logic in the Islamic world. It is relatively easy to find moments of particularly intense logical activity in the realms of Islam to use as points in periodization of the subject. 1.2 THE WESTERN SCHOOL OF ARABIC LOGIC At one point in its narrative Development goes seriously wrong, and the mistake is such that it clouds our ability to appreciate what is happening in what has been called the Golden Age of Arabic philosophy.8 The problem to which I refer is covered in chapter 5 (“The Clash of the Schools”) and chapter 6 (“The Period of ‘Reconciliation’ ”). According to Rescher, from the mid-eleventh century Avicenna’s writings determined the course of philosophical doctrines and discussions in the realms east of Baghdad. But in the early twelfth century, a major philosopher, Abū l-Barakāt alBaghdādī (d. 1165; referred to by Rescher as Ibn Malkā), began systematically to challenge these doctrines. The spirit in which he did so is characterized by Rescher as Farabian. Further, his impact was such in Baghdad and further east that one may speak of a school of logicians once again writing in the spirit of Alfarabi or, at least, challenging Avicenna’s logic in ways congenial with Alfarabi’s writings. This school of logic was called the Western school. One of the most important scholars to be influenced by Baghdādī was Fakhraddīn ar-Rāzī (d. 1210), a prolific Western logician. So powerful was the Western refutation of Avicennan logic that it was not until Nasīraddīn at-Tūsī (d. 1274) that a convincing set of counterarguments in support of Avicennan logic was produced. In constructing

312

RESCHER ON ARABIC LOGIC

these counter-arguments, Tūsī was working within what Rescher calls the Eastern tradition, badly debilitated but holding on to solutions put forward by, among others, `Umar ibn-Sahlān as-Sāwī (d. 1145). Such was Tūsī’s achievement that by the early fourteenth century there were two great schools, the Eastern and the Western. The reconciliation of these two schools in the fourteenth century by (among others) Qutbaddīn ar-Rāzī atTahtānī (d. 1365) was the most important logical event of the period.9 This account not only lies behind the narrative in Development, it also determines the table of affiliations given on pages 68 and 69. But this story is wrong. One should be aware in reading the tables that (Table VIII) there is no discernible evidence of influence by Alfarabi on Baghdādī, and none of influence by Baghdādī on Rāzī. Table IX is more misleading: Abharī may well have studied briefly with Tūsī under Kamāladdīn ibn Yūnus, but his allegiances were recognized as being to Rāzī;10 and Kātibī was Abharī’s student, not Tūsī’s. Does it matter? Yes. First, it makes Alfarabi loom large in a period of logical activity in which his books were in fact left unread. Secondly, it makes a marginal though interesting logician (Baghdādī) central to Rāzī’s logical development, whereas in fact they agree on few things in logic. Thirdly, it clouds the nature of Tūsī’s project. And finally, it makes us expect to find that the most widely read books of Arabic logic of all time, those by Abharī, Kātibī and Tahtānī, are either presenting Tūsī’s doctrine, or reconciling Tūsī’s doctrine to Rāzī’s; they are not. Arguments from silence are notoriously slippery, but in this case I think decisive. The evidence we have for Alfarabi influencing logicians in Syria and eastwards after the 1150s is fairly thin. In any event, the evidence we have for him influencing Fakhraddīn ar-Rāzī is almost non-existent, given that at the two points (or really, one point examined twice) that Rāzī mentions Alfarabi, Avicenna does too. It’s perfectly possible Rāzī never saw a logical work by Alfarabi, and merely took his references to him from Avicenna. Certainly Rāzī follows Avicenna on every point at which Avicenna declares himself to be disagreeing with Alfarabi.11 Further, Rāzī doesn’t look like he’s followed Baghdādī on any logical point. Baghdādī disagrees with Avicenna on truth conditions for the absolute propositions; he argues for weaker conclusions to modal mixes than Avicenna has; he doubts the utility of examining the syllogistic with conditional premises; he coins new technical terms for aspects of the logic. On all these points, Rāzī follows Avicenna.12

313

RESCHER STUDIES

Turning from Rāzī to Tūsī, we may ask what Tūsī does, in the middle of the thirteenth century, in the defence of Avicenna’s logic. To defend Avicenna’s logic, it is true that Tūsī feels compelled to criticize Rāzī. More precisely, Tūsī seeks to show that Avicenna’s exposition is exactly right to get across the point he was most concerned to convey.13 But his defence of Avicenna’s substantive logical doctrines is carried out in polemic with Abharī and Kātibī. Neither studied under Tūsī, but rather under students who had studied under Rāzī. Together, they provoke the most penetrating works on logic Tūsī produced.14 Finally, there never was a reconciliation between two schools. Tahtānī’s most influential logical contribution is his commentary on Kātibī’s Šamsiyya,15 and there Tahtānī clarifies and seems largely to adopt exactly the doctrine to which Tūsī objects. Tūsī was always read by advanced logicians, but any wider influence he had was in Šī`ī seminaries, through his non-polemical introduction to logic with commentary by Hillī.16 Rescher is right to ask us to look to the larger projects which logicians have in hand to work out what is going on in their logical writings. But the main source of dynamism in the Golden Age is not a return to Alfarabi and his textual Peripateticism; it’s Aš`arite theology. I hope to provide some speculative argument for this claim at the end of this paper. 2. RESCHER ON ARABIC MODAL LOGIC Rescher’s single major contribution to Arabic philosophy is his series of works on the modal logic; it is the most wrongly neglected set of studies in the field of Arabic philosophy today.17 In the following subsections, I’ll look at the context in which Rescher did his work, and the scholarly impact it has had. 2.1 THE CONTEXT FOR RESCHER’S STUDIES For a variety of reasons, mainly accidents of history, the best research from a philosophical point of view in Arabic logic has focused on the syllogistic, and it is interesting to rehearse why that is. Aloys Sprenger, a missionary in India in the mid-nineteenth century, realized that those Muslims who had studied logic in the madrasa had all read at the very least Kātibī’s Šamsiyya, so he and the Reverend Kaye set about preparing an English translation with which they also presented the Arabic text. They did it for

314

RESCHER ON ARABIC LOGIC

want of what they thought of as the real thing; as Sprenger said in his introduction, The right way of studying Arabic Logic would be to take the Arabic translations of Aristotle’s organon and by comparing them with the Greek text to ascertain the original import of the terms, and then to observe how their meaning has changed in the progress of time.18

This is a philological approach to the study of logic which seems to me doomed to failure. It will only work if the same technical terms are used in the later logic as are used in the translations of Aristotle. But they’re not. It’s instructive to contrast this approach with the one Rescher will follow. In any event, Sprenger and Kaye set about their translation, though their resolve grew faint at the point at which the Šamsiyya turned to modal logic. The sad result was that their translation lacked a full treatment of what was Kātibī’s primary concern. Still, since the Šamsiyya was replete, from beginning to end, with references to the modal propositions, even the shortened translation gestured towards a rich and interesting system. This was to prompt Rescher over a century later to begin his long battle with Kātibī’s modal system. He bravely translated what Sprenger and Kaye had omitted, and set out symbolically the syntax of the system (which is to say, he set out the way the various dummy letters and operators were arranged, and the rules by which they could be rearranged). He did this a number of times (at least four that I know of), beginning in 1967 and ending in 1975.19 The early efforts were dogged by a mistranslation of a diabolically difficult paragraph in the Šamsiyya, a passage anyone could have mistranslated. Rescher in consequence misdescribed the syntax of the text’s system, and wisely did not attempt to provide a semantics for it. But his uncle, the famous Orientalist Oskar Rescher, had made him aware of an extraordinary little text written by Muhammad ibn Faydallāh aš-Širwānī (d. 1707), which was handed over to the British Museum in the late 50s (now in the British Library as codex Or. 12405). By great good luck, Or. 12405 presented the same system we find in the Šamsiyya, with a few extensions, and with much fuller explanations. It even concludes with a set of tables laying out the conclusions of various premise-pairs. Or. 12405 enabled Rescher to right the syntax, on which basis he could then provide a semantics in modalized predicate calculus. He concluded by investigating in a cursory fashion some interesting mathematical aspects of

315

RESCHER STUDIES

the system, and by appending some fairly doubtful historical observations.20 We are now in a position to say that the system identified in the Šamsiyya is a sub-system of the Avicennan system.21 But I want to dwell on how Rescher went about his task to produce such valuable results, because I think the way he did what he did is even more edifying than what he did. Rescher had always intended first to provide a syntactic description of the whole system, that is, a non-invasive outline of the set of inferences that were being defended. He only then tried to provide a semantics for the system, that is, to stipulate truth-conditions for the propositions such that the set of inferences could be validly derived. I think the semantics Rescher provided has its short-comings (specifically, the way it handles existence of the subject, the way it represents the copula, the way it leaves underdetermined the metaphysical engagement of the system), but it gave us a formal grip on the elements of the propositions that we had never had before, and it gave us a system we could treat formally which was inferentially equivalent to a system we found frankly baffling. In my opinion, that’s the way to do it: Describe the syntax of a system—the whole system—with one distinct symbol for each operator, quantifier, term-variable and propositional reading, preferably describing the assumptions and proofs used by the logician for each move. Then, and only then, attempt to give a semantics for the system under investigation. Any other way ends up with fragments of the system interpreted in ways that won’t fit in with other inferential commitments. Never begin an interpretation until the description is complete. 2.2 SUBSEQUENT RESEARCH ON ARABIC MODAL LOGIC I find it very hard to criticize any aspect of Rescher’s work on modal syllogistic, because it opened up a vast world of Arabic philosophical writings. Hard, but not impossible. Rescher could have laid out the Arabic terms he was representing in his system,22 and he should have used Tahtānī’s published text rather than the fascinating but virtually inaccessible manuscript by Širwānī.23 Further, Rescher’s interpretation did not follow the Šamsiyya point by point, and I think that’s a weakness; several distinctions Kātibī dwells on don’t figure in Rescher’s model. But what Rescher realized—something, bizarrely, no one up to that time had realized—was that even an apparently unmodalized (mutlaqa) proposition has a hidden temporal modal operator modifying its predicate: “at one

316

RESCHER ON ARABIC LOGIC

time”. Its dual (“always”) rarely appears in Avicenna, though is a fully explored modality in Kātibī’s exposition of modal syllogisms. The other pair of modalities are “possibly” and “necessarily”. (Both “at one time” and “possibly” also have two-sided counterparts which I won’t consider in this paper.) Rescher came up with a way of reading kullu jīm bā (“every J is B”) as a mutlaqa dhātiyya (the most basic proposition in these systems) that made it come out: X: There is something that is at one time J, and whatever is at one time J is at one time B, or, in his symbolization (note that ∃ before a predicate letter is read “at one time”, and that I make explicit the existence of the subject that he says he will assume at the outset of his renditions): (∃x) ∃Jx ^ (∀x)( ∃Jx ⊃ ∃Bx) The other modalities are L: There is something that is at one time J, and whatever is at one time J is a necessary B M: There is something that is at one time J, and whatever is at one time J is a possible B, with broadly similar renditions. With these propositions, Rescher can give a plausible account of why Kātibī takes the following first-figure mixes to be perfect: Barbara XXX (every J is B, every B is A, therefore every J is A), LLL (every J is necessarily B, every B is necessarily A, therefore every J is necessarily A), LXL (every J is B, every B is necessarily A, therefore every J is necessarily A), XLX (every J is necessarily B, every B is A, therefore every J is A); and why he takes all premise pairs with M-minor premises (e.g. every J is possibly B, every B is A) as sterile or—to quote Kātibī more precisely—as “unknowable as to productivity.” When I first realized what Rescher had done, and how he’d done it, I decided to compile what I intended to be a completely neutral description of Avicenna’s modal syllogisms.24 Because I had relied so much on Rescher’s account of Kātibī, however, I couldn’t stop myself thinking about

317

RESCHER STUDIES

Avicenna in terms of a similar simple de re model. This generated problems in having all but only the syllogisms Avicenna wanted; specifically, allowing syllogisms with M-minor premises (e.g. every J is possibly B, every B is A, therefore every J is possibly A) while rejecting certain stronger conclusions (e.g. every J is possibly B, every B is A, therefore every J is A). I gave up, and decided Avicenna’s account of the modal syllogisms flattened out the distinction between the temporal and alethic modalities and, embarrassingly, made a snyde comment in print to the effect that Avicenna’s syllogisms were historically rather than philosophically interesting.25 In what is now the second paper a student of Arabic logic must read,26 Paul Thom sets out an alternative to the simple de re model, a mixed de dicto/de re model, which gives Avicenna’s mutlaqa dhātiyya as: X: Necessarily, whatever is J is at one time B, or, to use a symbolism like Rescher’s: …[(∀x)(Jx ⊃ ∃Bx)] The other modalities will be L: Necessarily, whatever is J is a necessary B M: Necessarily, whatever is J is a possible B. There are good reasons to adopt these truth-conditions simply from the examples Avicenna gives for the various propositions in Pointers, but what interests me is that—assuming S5 modal notions—this is a model that provides a system inferentially equivalent with Avicenna’s. It’s worth pausing at this point to note that, with Thom’s article, we have at last achieved some understanding of Avicenna’s modal logic (a subject to which Avicenna devoted reams of work), and this allowed Thom to speculate plausibly as to its metaphysical applications. This in turn depended on an overview of Avicenna’s logic only made possible by Rescher’s methods and results in describing and analysing Kātibī’s system. We are able to use these models for Avicenna and Kātibī to come to some understanding of what differentiates the two logicians in the modal syllogistic. Here are a couple of passages from the Šamsiyya to which Re-

318

RESCHER ON ARABIC LOGIC

scher never adverted, texts that give Kātibī’s way of distinguishing different understandings of the modals. Our statement every J is B is used occasionally according to the essence (hasab al-haqīqa), and its meaning is that everything which, were it to exist, would be a J among possible individuals would be, in so far as it were to exist, a B; that is, everything that is an implicand of J is an implicand of B. And occasionally [it is used] according to actual existence (hasab al-khāriğ), and its meaning is that every J actually (fī l-khāriğ), whether at the time of the judgment or before it or after it, is B actually.27 [. . . ] The distinction between the two considerations is clear. Were there no squares actually it would be true to say a square is a figure under the first consideration and not the second; and were there no figures actually other than squares, it would be correct to say every figure is a square under the second consideration but not the first.28

Without adverting to this distinction, Rescher rightly rendered Kātibī’s modals with an actualist reading, and Thom has rightly rendered Avicenna’s modals with an essentialist reading. I speculate, in my concluding remarks, about ways to tie this difference together with some of the things that were issue between Tūsī and the logicians he opposed. 3. CONCLUDING REMARKS Kātibī belongs to the Aš`arite school of theology, a school united over the centuries mainly by its commitments to acknowledging a completely untrammeled divine omnipotence. In consequence, only limited claims to knowledge may be made about an ultimately inscrutable world. Other components of Aš`arism were either revised or thrown out in the process of its clash with Avicennan philosophy. In other words, controlling notions for investigating God and man stayed the same, but methods for conducting that investigation had changed. The use of modal syllogistic was the most obvious change in method. It was introduced into Aš`arism in the twelfth century C. E., and scholars must have realized in the course of that century that it carried with it—in its Avicennan form—certain metaphysi-

319

RESCHER STUDIES

cal baggage. Kātibī contributes to the Aš`arite attempt to throw out that baggage. Kātibī recognizes that the disputed inferences (that is, premise pairs with an M-minor) are valid with an essentialist reading of the premises, but he only traces inferences with the actualist reading of the premises. Both Avicenna and Kātibī want to be able not only to trace valid inferences, but also to use the system they produce for extra-logical purposes. They want arguments that are not only valid, but also sound, that is, arguments that are not only formally perfect, but that have true premises. To use the essentialist reading to say every cow is necessarily four-stomached, as Avicenna would, is to claim necessarily, every cow is necessarily four-stomached; this is in one important respect much stronger than Kātibī’s there are actually cows, and everything that’s actually a cow is necessarily fourstomached. I think Kātibī is backing away from the optimistic insights into reality Avicenna is claiming to have. And Tūsī is fighting a battle already lost to the Aš`arites when he writes in defence of Avicenna’s system. This is speculation, of course, and only the editing, reading and analysis of many more manuscripts will allow us to reach firmer conclusions. But the fact that we can speculate about this at all is thanks to Rescher’s work. He has given us a tentative sketch of Arabic logical history about which we may enter into dispute. He has pushed us towards the heuristic expectation that the development of any given logician’s work will be linked to his broader philosophical commitments. And he has shown us how to read the telegraphic texts that the Arabic logicians produced on the modal syllogistic. The work of Nicholas Rescher in Arabic logic shows us, in short, that a scholar not primarily trained in the field of Arabic philosophy can make invaluable contributions to its methods and to its historical and analytical models.

REFERENCES Baghdādī, al-Kitāb al-Mu`tabar fī l-hikma, edited by S. Yaltakaya (Hyderabad, 1357 A. H). Daiber, H., Bibliography of Islamic Philosophy, 2 vols. (Leiden: Brill, 1999).

320

RESCHER ON ARABIC LOGIC

Gutas, D., Greek Thought, Arabic Culture (London and New York: Routledge, 1998). ———, “The Heritage of Avicenna: The Golden Age of Arabic Philosophy, 1000 ca. 1350,” in J. Janssens and D. De Smet, (eds.), Avicenna and His Heritage (Leuven: Peeters, 2002), pp. 81-97. Hillī, al-Ğawhar an-nadīd fī šarh mantiq at-tağrīd (Qum, 1410 A. H.). Kātibī, “ar-Risāla aš-Šamsiyya” in A. Sprenger (ed.), Bibliotheca Indica; a collection of Oriental works, No. 88 (Calcutta, 1854). First appendix to the Dictionary of the Technical Terms used in the Sciences of the Mussulmans, containing the Logic of the Arabians. ———, ar-Risāla aš-Šamsiyya (Cairo, 1948). With Tahtānī’s commentary. Madkour, I., L’Organon d’Aristote dans le monde arabe (Paris: Vrin, 19692). Mahdi, M., “Review of Rescher’s Development of Arabic Logic,” Islamic Studies, vol. 4 (1965), pp. 449-450. Mohaghegh, M., and T. Izutsu (eds.), Collected Texts and Papers on Logic and Language (Tehran, 1974). Perler, D. and U. Rudolph (eds.), Logik und Theologie. Das Organon im arabischen und im lateinischen Mittelalter (Leiden: Brill, 2005). Pococke, E., Historia Compendiosa Dynastiarum authore Gregorio AbulPharajio (Oxford, 1663). Rāzī, Mantiq al-Mulakhkhas (Tehran, 1381 A. H.). Rescher, N., The Development of Arabic Logic (Pittsburgh: University of Pittsburgh, 1964). ———, “Arabic Logic,” The Encyclopedia of Philosophy, vol. 4 (1967). pp. 525-527.

321

RESCHER STUDIES

———, Temporal Modalities in Arabic Logic (Dordrecht: Kluwer, 1967). ———, “Temporal Modalities in Arabic Logic,” in his Studies in Arabic Philosophy, (Pittsburgh: University of Pittsburgh, 1968), pp. 81-110. Rescher, N. and A. vander Nat, “The Theory of Modal Syllogistic in Medieval Arabic Philosophy” in N. Rescher, R. Manor, et al., editors, Studies in Modality, (Oxford: Blackwell, 1974), pp. 17-56. Rescher, N. and A. vander Nat, “New Light on the Arabic Theory of Temporal Modal Syllogistic,” in G. Hourani (ed.), Essays in Islamic Philosophy and Science (Albany: State University of new York, 1975). Street, T., “Toward a History of Syllogistic after Avicenna: Notes on Rescher’s Studies on Arabic Modal Logic,” Journal of Islamic Studies, vol. 11.2 (2000a), pp. 209-228. ———, “Avicenna and Tūsī on the contradiction and conversion of the absolute,” Journal of the History and Philosophy of Logic, vol. 21 (2000b), pp. 45-56. ———, “‘The Eminent Later Scholar’,” in Avicenna’s Book of the Syllogism,” Arabic Sciences and Philosophy, vol. 11 (2001), pp. 205-218. ———, “An Outline of Avicenna’s Syllogistic,” Archiv für Geschichte der Philosophie, vol. 84 (2002), pp. 129-160. ———, “Fahraddīn ar-Rāzī’s Critique of Avicennan Logic,” in Perler and Rudolph 2005, pp. 99-116. Thom, P., Medieval Modal Systems (Aldershot: Ashgate, 2003). ———, “Logic and Metaphysics in Avicenna’s Modal Syllogistic,” in S. Rahman, T. Street & H. Tahiri (eds.), The Unity of Science in the Arabic Tradition: Science, Logic, Epistemology and their Interactions (Dordrecht: Kluwer, 2008).

322

RESCHER ON ARABIC LOGIC

———, “Avicenna’s Modal Logic: de re interpretations,” in J. Marenbon and M. Cameron (eds.), The Prior Analytics in Two Traditions (New York: Fordham, 2009). Tūsī, “Ta`dīl al-mi`yār fī naqd al-asrār,” in Mohaghegh and Izutsu 1974, pp. 137-248. ———, “Mutārahāt mantiqiyya bayna l-Kātibī wat-Tūsī,” in Mohaghegh and Izutsu 1974, pp. 279-286. NOTES 1

Daiber [1999].

2

Daiber [1999] 1: 776-771.

3

Khaled El-Rouayheb has shown, in a study yet to be published, that the logical text in British Library codex Or. 12405 is in fact written by an Ottoman scholar Muhammad Sādiq ibn Faydallāh ibn Muhammad Amin ibn Sadraddīn aš-Šīrwānī, 1630-1707, a number of centuries after the time Rescher thought he lived. This means that the text analysed by Rescher in his later studies of Arabic modal logic was written at pretty much the same time as the Port-Royal Logic.

4

Mahdi [1965] 449.

5

Madkour [19692] 9.

6

Madkour [19692] 9-10.

7

Gutas [1998].

8

Gutas [2002].

9

Rescher [1964] 64-72, 73-75. Rescher [1967a] gives an abbreviated variant of this history, dating the origin of the Western school to Fakhraddīn ar-Rāzī.

10

So Bar Hebraeus in Pococke [1663] 485.7-13 (Arabic); Bar Hebraeus’s account allows us to understand the texts, their differences and the points they share.

11

Avicenna’s references to Alfarabi are gathered together in Street [2001]; the profoundly Avicennan base to Rāzī’s work is established in Street [2005]. To evaluate Rāzī’s debts to Alfarabi, see Rāzī [1381 A. H.] 142.5 (on ampliating the subject term to the possible) and 287.10 (on the effects such ampliation has on mixed modals).

12

See Baghdādī [1357 A.H.] respectively 120-121, 148-152, 155; cf. Rāzī [1381 A. H.] 88, 118, where Baghdādī is referred to on purely verbal disputes.

13

Something I try to show in Street [2000b]; cf. Street [2005].

323

RESCHER STUDIES

NOTES 14

Tūsī [1974a], Tūsī [1974b].

15

Main text in Kātibī [1948]; Kātibī’s text is given as lemmata.

16

Hillī [1410 A.H.].

17

Neglected at least in terms of citation, but honoured by plagiarism. I’ve recently advised journals to reject submissions of Rescher and vander Nat [1974].

18

Kātibī [1854], at page ii.

19

Rescher [1967b], Rescher [1968], Rescher and vander Nat [1974], Rescher and vander Nat [1975].

20

Rescher and vander Nat [1974] is better than Rescher and vander Nat [1975]. The Arabic terms for the elements of its symbolic rendition are given in Street [2000a]; to guard against at least one of the mistakes in my article, avoid using the term “composite” for wasfī .

21

Thom [2009].

22

For the Arabic of those terms, see Street [2000], though now I would tend to stick with Rescher’s choices for translating the terms.

23

Kātibī [1948].

24

Street [2002]. See Thom [2003] chapter 4 for an exploration of the system under a simple de re interpretation. See also Thom’s chapter 9, because Buridan’s is the western system most like Avicenna’s.

25

Street [2005].

26

Thom [2008]

27

Kātibī [1948] 91.1-4.

28

Kātibī [1948] 96.12-14.

324

Philosophy in the Future Avrum Stroll

T

here has recently been a spate of publications about the future of philosophy in the 21st Century. Perhaps the most interesting of such endeavors is Nicholas Rescher’s “Philosophy at the Turn of the Century—A Return to Systems?” that appeared in his Studies in 20th Century Philosophy, published in 2005.1Rescher is avowedly more cautious than most authors about predicting the nature of what may happen, but he is willing to speculate on how whatever happens will happen.2 As he writes: What is the philosophy of the (new) century going to be like? A famous jazz musician was once asked, “Where is jazz going?” He replied, “If I knew that we’d be there already.” But while it is indeed effectively impossible to say what those philosophers of the future will produce, it is possible to make a plausible conjecture about how they will produce it. They are likely for some time to proceed in much the same way as at present—by the same sort of disaggregated collaboration that we are currently witnessing3.

Although Rescher is sensibly reluctant to predict what the content of future philosophy will be, he not only believes that it will happen by disaggregated collaboration but also that it will bring about a resurgence of systematic (or “holistic”) philosophy. Let us look at these two “predictions” if one may label them in that way. In order to understand what Rescher means by “disaggregated collaboration,” one should begin with a contrast he wishes to draw between an older philosophical tradition and philosophy as practiced today. Because his discussion of the historical situation is complex, I will quote him in full. Once upon a time the philosophical stage was dominated by a small handful of greats and the philosophy of the day was what they produced. Consider German philosophy in the nineteenth century, for example. Here the philosophical scene, like the country itself, was an aggregate of principalities— presided over by such ruling figures as Kant, Fichte, Hegel, Schelling, Schopenhauer, and a core of other philosophical princelings. But today this “heroic age” of philosophy is a thing of the past.

RESCHER STUDIES

The extent to which significant, important and influential work is currently produced by academics outside the high-visibility limelight has not been sufficiently recognized. For better or worse, in the late twentiethcentury we have entered a new philosophical era where what counts is not just a dominant elite but a vast host of lesser mortals. Great kingdoms are thus notable by their absence, and the scene is more like that of medieval Europe—a collection of small territories ruled by counts palatine and prince bishops. Scattered here and there in separated castles, a prominent individual philosophical knight gains a local following of loyal friends or enemies. But no one among the academic philosophers of today manages to impose his or her agenda on more than a minimal fraction of the larger, internally diversified community. Given that some ten thousand academic philosophers are at work in North America alone, even the most influential of contemporary American philosophers is yet another—somewhat larger—fish in a very populous sea. If 2 or 3 percent of professional colleagues pay attention to a philosopher’s work, this individual is fortunate indeed. As regards those “big names,” the fact is that those bigger fish do not typify what the sea as a whole has to offer. Matters of philosophical history aside, some of the salient themes and issues with which American philosophers are grappling at the present time include: ethical issues in the professions, the epistemology of information processing; the social implications of medical technology (abortion, euthanasia, right to life medical research issues, informed consent), feminist issues, distributive justice, human rights, truth and meaning in mathematics and formalized languages, the merits and demerits of relativism regarding knowledge and morality, the nature of personhood and the rights and obligations of persons, and many more. None of these topics was put on the problem-agenda of present concern by any one particular philosopher. They blossomed forth like the leaves of a tree in springtime, appearing in many places at once under the formative Zeitgeist of societal concern. Accordingly, philosophical innovation in America today is generally not the response to the preponderant effort of pace-setting individuals, but a genuinely collective effort that is best characterized in statistical terms4.

In an earlier book, in describing the present situation in philosophy, Rescher explains what he means by “disaggregated collaboration:” (Philosophy) is no longer an intellectual enterprise of the “great thinker, great system’ type familiar from the classical tradition. Systems are nowadays constructed like ant-hills rather than like pyramids that are the product of centralized direction. Unprogrammed and disaggregated collaboration among many workers distributively addressing large and complex projects

326

PHILOSOPHY IN THE FUTURE

has become the order of the day. And in every area of philosophy a literature of vast scope and complexity has emerged whose mastery is beyond the capacity of single individuals. Systematization is at work but rather at the collective level than at that of individual contributions.5

Although Rescher’s account begins with German philosophy and a “small handful of greats,” these references are essentially used as examples in support of a main thesis. This is that these “greats,” the equivalents of individual kings or emperors, were mostly concerned with system building, i.e., with the construction of broad philosophical conceptions, whose internal details were generally used in defense of the system. The historical materialism of Marx and Comte’s “law of three stages” are illustrations of what he has in mind. As he sees the present scene (as of 2006) there has been a return to systematic philosophy, after some interludes in which the idea of synoptic philosophy was rejected in terms of piecemeal and detailed analysis. Rescher’s suggestion is that systematic philosophy has been revivified and he predicts that we can expect to see more of it in the future. He also suggests that it will no longer be dominated by a handful of eminent persons but will involve many independent and often widely dispersed contributors working on themes of common interest. This is what he means by “disaggregated collaboration.” Rescher is not the only distinguished scholar who sees the future as revitalizing an older tradition. John Searle ends a 2003 essay in which he also predicts the return of systematic philosophy. As he says: What does philosophy look like in a post-epistemic, post-skeptical era? It seems to me that it is now possible to do systematic theoretical philosophy in a way that was generally regarded as out of the question half a century ago.6

By contrasting the present scene with philosophy of about a half century ago, Searle is obviously referring to such practitioners as G. E. Moore and J. L. Austin who thought that philosophy should deal with piecemeal, detailed, issues. G. J. Warnock described Austin’s approach as follows: He believed (like Moore) that, if progress was to be made, many questions would have to be raised, many facts surveyed, many arguments deployed, step by step and narrowly criticized; questions ought to be distinguished and considered strictly one at a time, and no effort spared to make it wholly clear what question was being asked and exactly what answer was proposed to it. The effect, in discussion in the thirties, of this dogged resistance to haste has

327

RESCHER STUDIES

been described as ‘powerfully negative’, and no doubt it was (if one remembers how philosophers are prone to go on); but conspicuously it was not dull, and above all, not negligible.7

Searle argues that the era of piecemeal philosophy is now over. He believes that the growth of scientific knowledge has transformed philosophy in ways that were not possible around the middle of the last century, and that it is the existence of such knowledge that has made it possible today to do systematic philosophy. His argument in support of this position depends on an historical thesis. Searle begins with the 17th C., and his view is, roughly, that the skeptical paradoxes of Descartes engendered the idea that epistemology was the central discipline of philosophy. Its main task, for more than three hundred years, was to show in the face of skeptical challenges that knowledge exists. Against the centrality of epistemology, Searle claims: “At present it is psychologically impossible to take Descartes’ project seriously in the way he took it: We know too much.” In opposing Descartes, Searle states that there is now a large body of knowledge that is theoretical, comprehensive, systematic, and universal in subject matter. We know with certainty, for example, that the heart pumps blood, that the earth is a satellite of the sun, and that water is composed of molecules of hydrogen and oxygen. His view is that the evidence that supports these claims is overwhelming, and furthermore, that they are embedded in theories that are so well established it would be irrational to doubt them. Although I agree with most of what Searle says I have some substantial disagreements with him as well, and these lead me to a somewhat different picture of the future than that which he portrays. As we shall see, Searle’s reasons for thinking that philosophy will take a systematic course differ from those of Rescher, and it is Rescher’s that I find the more compelling. Rescher also carries the subject further than Searle. He not only predicts the resurgence of systematic philosophy but also justifies it. Altogether his is a powerful position that is supported by much of the classical philosophical tradition. By way of contrasting these two views, let me begin with Searle’s account. I accept Searle’s theses that since the 17th C. epistemology has been the central discipline of philosophy and that many practitioners, since that time, have directed their efforts to refuting it. But we differ over what it has threatened. According to Searle it is the existence of knowledge; but in my view this is an incorrect interpretation of the history of the subject. I think it is not knowledge that the sceptic has traditionally challenged but

328

PHILOSOPHY IN THE FUTURE

certainty. Such 20th C. writers as Ryle, Moore and Wittgenstein have thought this to be so. Wittgenstein’s last notebook, for example, is entitled On Certainty and in Philosophical Papers, Moore published an entire essay “Certainty,” which first saw the light of day as the Howison Lecture at UC Berkeley in 1941. If we go back to Descartes and the Meditations of 1641 we find further justification for my construal of the situation. In Meditation II, Descartes writes: Yesterday’s meditation has thrown me into such doubts that I can no longer ignore them, yet I fail to see how they are to be resolved. It is as if I had suddenly fallen into a deep whirlpool; I am so tossed about that I can neither touch bottom with my foot, nor swim up to the top. Nevertheless, I will work my way up and will once again attempt the same path I entered upon yesterday. I will accomplish this by putting aside everything that admits of the least doubt, as if I had discovered it to be completely false. I will stay on this course until I know something certain, or, if nothing else, until I at least know for certain that nothing is certain. Archimedes sought but one firm and immovable point in order to move the entire earth from one place to another. Just so, great things are also to be hoped for if I succeed in finding just one thing, however slight, that is certain and unshaken.

It will be noted that in this passage, Descartes does not speak about the quest for knowledge. It is true that he twice uses the word “know” but the passage makes it clear that what he wishes to know is whether anything is certain. He says he is searching for “just one thing, however slight, that is certain and unshaken.” So the emphasis is clearly on certainty and not on knowledge. I think it is fair to say that this emphasis runs through all six Meditations. We can find similar quotations from many major philosophers, past and present. My main disagreement with Searle, however, is over the impact and importance of science for philosophy. As he sees it science will and should transform philosophy. He writes: This growth of knowledge is quietly transforming philosophy, making it possible to do a new kind of philosophy. With the abandonment of the epistemic bias in the subject, such a philosophy can go far beyond anything imagined by the philosophy of a half century ago. It begins, not with skepticism, but with what we know about the real world. It begins with such facts as those stated by the atomic theory of matter and the evolutionary theory of biology. . . . Such a philosophy is theoretical, comprehensive, systematic and universal in subject matter.8

329

RESCHER STUDIES

Searle thinks that most major philosophical problems are resolvable, in principle, by science. He says this about the traditional mind-body problem, for example. It seems to me the neurosciences have now progressed to the point that we can address this as a straight neurobiological problem, and indeed several neurobiologists are doing precisely that. In its simplest form, the question is how exactly do neurobiological processes in the brain cause conscious states and processes, and how exactly are those conscious states and processes realized in the brain. . . . It looks similar to such problems as: “How exactly do biochemical processes at the level of cells cause cancer? and, “How exactly does the genetic structure of a zygote produce the phenotypical traits of a mature organism?”9

I disagree with this form of scientism. I hold instead that philosophy is a conceptual activity, and accordingly, that the sorts of factual problems science deals with are not at all comparable to the kinds of issues that philosophy has traditionally faced or will face. In emphasizing the conceptual nature of philosophy, I concur with Rescher who, though impressed by science and its success, has throughout his eminent career been an advocate of conceptual analysis. In my view, philosophy is and always has been an autonomous activity: it is not history, psychology, linguistics, sociology or any sort of natural science, and the methods of science will not resolve its problems. I think Searle has confused causal questions with conceptual ones. Without trying to explain what I regard as his basic confusion, let me provide an illustration of what I mean by a conceptual problem. This is a problem about the justification of abortion, a topic that, as Rescher stresses, is at the forefront of philosophical concern today. As we shall see, it leads eventually to the mind-body problem and raises puzzles that go beyond anything science can deal with. Let us consider the issue in some detail. The extreme right-to-life position advances the following considerations in support of its opposition to abortion. First, it argues that from the moment of conception, what has been produced is a human being, and that all human beings are persons. Second, it states that the unborn are innocent of any crime. Third, it contends that it is necessary to find a coherent set of principles—that is a defensible philosophy—that would justify killing the unborn. Fourth, it affirms that since similar cases must be treated in similar ways, such principles would justify the abortion of an innocent prenatal child only if they

330

PHILOSOPHY IN THE FUTURE

would also justify the killing of an innocent postnatal infant. Fifth, it holds that no considerations can be found that would justify the latter course of action, and, accordingly it concludes that no principles can be found that would justify the former. Therefore, abortion is never justified. The pro-choice position attacks each of these premises and disagrees with the conclusion. But to bring out the special nature of a conceptual problem, I will simply concentrate on the first premise of the extreme right-to-life argument. It holds, as indicated, that from the moment of conception the unborn entity is both a human being and a person. Some advocates of the pro-choice position hold that not even the fetus is a person, although it has the potentiality of becoming one10. This contention is supported by an analogy. Suppose—as many would agree—that cutting down an oak tree is, in a specific case, a bad thing to do. Yet those same persons might well agree that an acorn is not an oak tree, so destroying an acorn that eventually might become a tree is not identical with destroying a tree. According to this comparison, the zygote or even the fetus is like an acorn. It is not yet a person, with aspirations or a will; it is incapable of motion on its own, and lacks thought or intention. The prenatal entity is simply a mass of tissue, analogous to a benign growth. It can thus be excised and abortion can be thought of as analogous to a surgical procedure that eliminates an unwanted cyst or tumor. This counter argument to the right-to-life thesis rests on an analogy, on the idea that the unborn is like an acorn. This analogy occurs in an early phase of the argument between these opposing approaches, but it already raises a conceptual question: How good is the comparison? It will be noted that the appropriateness of the analogy is not a factual or scientific question. It cannot be decided by a description of the scientific facts connected with conception. Both sides, I believe, would agree on the facts. These include such pieces of information as the following. The female germ cell, or ovum, is fertilized by a male germ cell, the spermatozoan. When this occurs, the cell possesses a full complement of twenty-three pairs of chromosomes, one in each pair from each partner. When fertilization first occurs, the ovum is called a “single-cell zygote.” Within twentyfour hours, the single cell begins to divide. It acquires sixteen cells by the third day, and continues to grow as it moves through the fallopian tube into the uterus. During the first week, it implants itself in the uterine wall, and then is called a “conceptus.” By the end of the second week, it is fully embedded in the uterine wall, and from this point until the eighth week, it is called an “embryo.” Some human features appear by the fourth week—

331

RESCHER STUDIES

the embryo acquires a face and incipient limbs—and by the eighth week brain waves can be detected. From this point until birth it is called a “fetus.” These facts are not disputed by either side. The philosophical question is rather: “When does the fertilized egg become a person?” But as one examines the facts they do not speak to that issue. Is a single-cell zygote a person? Is the fetus a person? Is neither a person? The scientific facts are silent with respect to these questions. The justification of abortion is thus not decidable by an appeal to the facts. Whether an abortion is justified or not clearly depends on the context and also on the point of view one takes.11. They may arise from religious or nonreligious perspectives. In many religions personhood occurs only when an entity develops a soul. But when is that? And what is the soul? Such questions are likewise not decidable by scientific means. The issue of whether the unborn entity is a person or not must thus be resolved, if it is at all possible to resolve it, in some other way. This is the typical situation in dealing with philosophical issues. Such issues typically turn on crucial premises that are conceptual, not factual, in character. We can quickly deepen the issue. In the case of abortion, there is a perplexity about the relationship between bits of tissue and personhood. When do pieces of flesh become persons? If a person is not merely flesh, what is it? Is it some kind of nonmaterial entity, and if so, how can we identify it? Such questions give rise to one of the deepest and most intractable problems in the philosophical lexicon, the so-called “Mind-Body Problem.” Philosophers from the time of the Greeks to the present, theologians of all stripes, and ordinary persons have been puzzled about the relationship between personhood and the material constituents that constitute living objects. When the Gospel According to St. John states “The word was made flesh”12 we have a theological version of the problem. In my view, all philosophical problems are like the mind-body problem in being essentially conceptual in nature. I thus disagree with Searle. But my disagreement with him goes well beyond the Mind-Body problem. It is that no serious philosophical problem can be solved by science. And my reason for saying this is that I regard all such issues as primarily conceptual and not as factual. But a defense of this statement would require another and different essay, so I will say no more about it here. I agree with Rescher’s contention that at present, in 2006, there are no more greats on the philosophical scene. By “greats” I mean such figures as Frege, Wittgenstein, Russell and Quine. We do have many distinguished

332

PHILOSOPHY IN THE FUTURE

figures still practicing. But as Rescher points out, they are analogous to counts palatine and prince bishops. I also concur that philosophers have expanded their concerns to include such practical problems as abortion, euthanasia, the nature of personhood and the rights and obligations of persons. The focus on a narrow set of abstract problems, such as the relationship between language and the world, is no longer of central concern. His contention that disaggregated collaboration will lead to systematic philosophy seems to me to rest on a deep understanding and appreciation of the lengthy history of the discipline. Many of its major figures have been systematic thinkers. The list includes Parmenides, Aristotle, Augustine, Aquinas, Bonaventure, Descartes, Spinoza, Kant, Hegel, Carnap, and Quine, inter alios. Rescher’s idea seems to be that piecemeal philosophy as practiced by Moore and Austin was a temporary interlude, though of considerable interest, and that the discipline will now revert to its traditional, holistic mission. As I mentioned above, Rescher wishes to do more than predict such a renewal; he also wishes to justify it. In a large number of books and essays, he has powerfully argued his case. I will concentrate on one such study Philosophical Reasoning, originally published in 1975. In Chapter 12, entitled “Why Philosophizing Must Be Systematic: The Holistic Nature of Philosophy,” Rescher lays out in a careful and articulate manner his justification of systematic philosophy. He gives five reasons for this judgment, each supported by a panoply of examples. Rather than expounding all five—a lengthy task indeed—I will confine my comments to the first and to some of its supporting examples. The five are (not in his own words but according to my interpretation): 1) A philosophical doctrine that is compelling in a particular area is likely to have unwanted repercussions in a different area of the subject. Only systematization will reduce or eliminate this possibility. 2) Philosophical problems arise from conceptual predicaments, are organically connected, and reach across taxonomic frontiers. 3) We must therefore proceed holistically in philosophy, abandoning piecemeal endeavors. 4) From inconsistencies everything and nothing follows. Therefore any philosophical theory must be consistent 5). But to be a theory at all, any philosophical account must be systematic. Rescher supports 1) by a large number of examples. A farmer’s fertilizers may contaminate a neighbor’s drinking water. This is analogous to what may happen in philosophy. In solving a problem within some particular domain we may create difficulties for the solution of problems else-

333

RESCHER STUDIES

where, even in areas seemingly far removed from the original issue. As William James pointed out: “If I refuse to stop a murder because I am in some doubt about whether it is justifiable homicide, I am virtually abetting the crime.” As Rescher indicates, radical skepticism in epistemology may well be an active ally of immorality. A doctrine that makes perfect sense in a particular sub-branch of philosophy can have unwelcome consequences in another. Only systematization can prevent such untoward effects. In his advocacy of systematization in philosophy, Rescher is presupposing a logical principle that is indisputably correct. This is the principle that any body of discourse that is inconsistent says nothing. One who asserts p and then later asserts not p is ultimately saying nothing. The entire philosophical tradition from Aristotle to the present would agree with this remark. But Rescher also assumes that it follows from this axiom that anything other than a fully formalized and consistent system may contain such contradictions. So his thought is that systematization runs counter to the sort of piecemeal philosophy advocated and practiced by Moore and Austin. In holding to these two theses, the need for consistency and the possibility that a non-formal approach may result in contradictions, Rescher is indeed supported by most of the Western tradition. His justification of systematicity in philosophy thus rests on these two maxims. I agree with Rescher that inconsistent systems say nothing, and I also agree that there are obvious dangers in doing philosophy in a way that is non-systematic. Still, one must acknowledge that Moore and Austin were representatives of a different approach that also has ancient antecedents. It seems to me that Rescher is making a solid case for one of these ways. Let me describe another which I shall call “informalist philosophy.” This is a different option from that which he espouses. We can begin its depiction with the ancient Greeks and in particular with Socrates: Socrates is a paradigm of the philosopher who was engaged in conceptual analysis and whose views did not issue in a holistic theory. I say this on the supposition, now widely accepted, that Plato’s early dialogues, Charmides, Lysis, Laches, and Euthyphro are primarily reports by the young Plato of the activities of a man whom he admired, even worshipped. These short works show that Socrates was engaged in conceptual analysis, asking his interlocutors what they mean by certain concepts: friendship and piety, for example. He describes Socrates as a kind of gad fly, asking ordinary Athenians to define these concepts which are in common use, and finding that they are invariably unable to do so. The first two books of the

334

PHILOSOPHY IN THE FUTURE

Republic, a more mature work, finds Cephalus, Polemarchus, Thrasymachus, Adeimantus and Glaucon unable to provide a satisfactory characterization of the notion of justice. In these early works, Socrates is engaging in the analysis of certain fundamental notions, justice among them, trying to discover their real meaning and without success. No system emerges from such activities. It is a piecemeal process that is being described. But it has been argued that as he matured Plato developed a metaphysical system of the sort that Marx and Hegel were later to create. This view has also been contradicted by outstanding scholars, one of them being Raphael Demos. Demos denied that Plato’s philosophy is a system but further he denied that the thought of any great philosopher could be encapsulated into a simple formula. Here is what he said about Plato, for example: . . . . Plato’s philosophy takes the form of wonder, of continual questioning, of continual testing of hypotheses; it is the narrative of a search, an account of the process itself of thinking, rather than of conclusions arrived at. So Plato’s attitude is a mixture of certainty with doubt. His dialogues—and particularly the early ones—set forth problems which often they do not solve; in the dialogues each different character stands for some different point of view; thus, Plato’s works are dramas of intellectual conflict, whose vividness and value to the reader lie frequently in their evocation of the actual history of a mind, candidly facing the confusions of the problems, and struggling toward some solution. The reader who is baffled with Plato’s frequent digressions, with the casual way in which he drops a topic in which he had seemed so intensely interested only a short time before, with his leisurely meanderings among all kinds of subjects, may well ask whether the author is driving at any definite point at all. A possible answer is that Plato is averse to providing the reader with answers to questions; since such generosity on his part would remove from the reader that very sense of wonder which is the stimulus to speculation . . . . So Plato, in his dialogues, is not giving formed and finished ideas to the reader; he is assisting the reader to bear his own intellectual children. Plato’s philosophy is not a system; and it evades any formulation in a deductive pattern. His mind does not seem to stop at any definite idea and rest there; it is in movement.

Demos ends his essay with these words:

335

RESCHER STUDIES

Plato’s writings are rich in insights on specific matters of fact. Of equal importance with these insights is his general doctrine. But of still greater importance than both his insights and his doctrine, is his method. And his method is this: that the mind must approach its task of explaining nature, armed with a variety of categories. The world is complex; it cannot be comprehended under any single notion. Explain the universe by any one specific doctrine, and you will find that you have left something out. Thus, hypotheses are tentative; the mind can never rest in any formulation. Clarity is provisional; and within certainty there must lurk the doubt and the wonder which are the animating forces of the human mind.13

The informalist method that Socrates used in trying to capture the ordinary meaning of such concepts as justice, friendship and piety is not much different from Moore’s practice in Principia Ethica with respect to the notion of “good.” There is clearly a prominent strain of informalism in David Hume’s Treatise, especially in Book II (Of the Passions). In the 20th C, J. L. Austin’s explanation of the notion of responsibility in such essays as “A Plea for Excuses,” and “Three Ways of Spilling Ink” is a good example of how an ordinary language philosopher deals with a concept that was originally surfaced by Aristotle. But Austin is only one of many who have engaged in this activity. It is thus plausible to infer that an informalist tradition has always been part and parcel of philosophy. I agree with Rescher that no greats are writing at present in philosophy, and I also agree that the topics being discussed today vary widely from those under scrutiny in earlier times, but I don’t think it follows from either of those statements that the future will be confined to systematization. I believe that we can expect both approaches to be exemplified but perhaps in different proportions. I therefore predict that we shall see two forms of philosophical practice in the future. One of them is the tradition of system building so eloquently described by Rescher. The other will be some form of informalism. As I see the situation, they will exist happily side by side.

NOTES 1

This essay is dedicated to Nicholas on the occasion of his 80th birthday. I wish him well and hope, indeed expect, that he will continue his brilliant contributions to the philosophical literature.

336

PHILOSOPHY IN THE FUTURE

NOTES 2

While recognizing the risks, John Searle is among the authors who are willing to speculate about the future of the subject. Searle, for example, writes: “General ruminations on the state and future of philosophy often produce superficiality and intellectual self-indulgence. Furthermore, an arbitrary blip on the calendar, the beginning of a new century, would not seem sufficient, by itself, to override a general presumption against engaging in such ruminations. However, I am going to take the risk of saying some things about the current and future state of philosophy, even though I think it is a serious risk. A number of important changes have occurred in my lifetime and I want to discuss their significance and the possibilities they raise for the future of the subject.” “Philosophy in a New Century,” in Philosophy in America at the Turn of the Century, Philosophy Documentation Center, 2003. p.3. I am among those who, like Searle, have ventured onto this slippery terrain. See my Twentieth-Century Analytic Philosophy (Columbia U. Press, 2000), esp. ch.9.

3

Studies, p.107

4

ibid, p.17

5

Profitable Speculations: Essays on Current Philosophical Themes (New York, Rowman and Littlefield, 1997), p.24

6

op.cit.., p.8

7

Symposium on J.L. Austin ed. by K.T.Fann, 1969, p.7

8

Op.cit., p.3.

9

Op.cit., p.9

10

Judith Jarvis Thomson, for example, has defended a variant of this position. See “Symposium on Abortion,” Philosophy and Public Affairs, 1, no.1 (1971): pp.4756.

11

In Roe vs. Wade, (1973), the US Supreme Court stated that fetuses are not fully legal persons, and hence that a woman has a right to an abortion at any stage of pregnancy. The decision also mentioned that the right to an abortion is not absolute, and is limited by the state’s right to regulate medical practice to protect maternal health. Justice Harry Blackburn wrote for the majority that “a state may properly assert important interests in safeguarding health, in maintaining medical standards, and in protecting potential life.”

12

John, 1:14

13

Introduction to The Dialogues of Plato, ed. by B. Jowett, Random House, 1937, pp.vii-xii.

337

Rescher on Explanation and Prediction Bas C. van Fraassen O. PRELIMINARIES1

T

he middle years of the sixties, when I was studying with Nicholas Rescher and his colleagues in Pittsburgh (Adolf Grünbaum supervised my dissertation on the theory of time, Nuel Belnap was my initial mentor in logic, Wilfrid Sellars my guide into the thickets of scientific realism) were full of philosophical excitement. The times they were a-changing … we felt ourselves in the vanguard of new revolutions in philosophy of science and philosophical logic. In my second year, Rescher took me on as a research assistant for his Temporal Modalities in Arabic Logic, and gave me a directed reading course in philosophical logic. The topic was quantificational tense logic—as part of my paper for the course I presented my first completeness proof for free logic.2 But here I will concentrate on other issues about which I learned from Rescher, at that time, which also led me into new paths: explanation and prediction. 1. RESCHER’S WORK ON EXPLANATION When I was thus coming of philosophical age, the philosophical disputes over scientific explanation and prediction, and how they are related to each other, were raging. Nicholas Rescher was one of the main participants, and we graduate students in Pittsburgh had a ring-side seat. I will write here mainly about the papers which led up to his Scientific Explanation, which arrived just a few years after I had finished my studies. The background to this work is, as Rescher documents, the trend setting 1948 paper “Studies in the Logic of Explanation,” by Carl Hempel and Paul Oppenheim. It will suffice here to recall Hempel’s summary of the account of explanation, listing as main criterion for explanation: Explanatory relevance: “the explanatory information adduced affords good grounds for believing that the phenomenon did, or does, indeed occur”.3

The ‘good grounds’ can take two forms:

RESCHER STUDIES

—conclusive reasons (the information adduced implies that the phenomenon occurs); —reasons that bestow a high probability on the phenomenon. In the first case Hempel speaks of deductive-nomological explanation, and in the second of inductive-statistical explanation. 1.1

A LIMITED ASYMMETRY BETWEEN EXPLANATION AND PREDICTION

Shortly after I arrived as a new graduate student in Pittsburgh, Nicholas Rescher’s paper on discrete state systems and Markov chains appeared in Philosophy of Science, establishing surprising limits to the possibility of prediction and retrodiction, even under conditions in which certain kinds of explanation are readily possible. By studying the subject in the precisely demarcated domain of deterministic and indeterministic discrete state systems, Rescher thus supported his previous (1958) more general argument that Hempel and Oppenheim’s contention of a symmetry between explanation and prediction (except for purely pragmatic factors) was mistaken.4 But at that time Rescher still saw strong connections between explanation and prediction, due in part to the orientation that focused on those concepts abstracted from pragmatic factors. This orientation was inherent in the discussion initiated by Hempel and Oppenheim, and it remained dominant for a good two decades. What Rescher did establish, in effect, was that the famous statement the difference between the two [i.e. prediction and explanation] is of a pragmatic character: If E [the conclusion of the explanatory schema] is given, i.e. if we know that the phenomenon described by E has occurred . . . we speak of an explanation of the phenomenon in question. If . . . E is derived prior to the occurrence of the phenomenon it describes, we speak of a prediction5 is mistaken even at that level of abstraction, where explanation, prediction, and retrodiction are all explicated in terms of concepts relating to the ‘theory of confirmation’ developed by Hempel, Carnap, and their colleagues. We can see what still remained in place of this approach, in Rescher’s

340

RESCHER ON EXPLANATION AND PREDICTION

thinking, in the 1963 paper. After a careful analysis of explanation, prediction and retrodiction in terms acceptable and in use by Hempel himself, Rescher summarizes: From the foregoing definitions it follows that whenever a prediction or retrodiction (of whatever type) is given, so a fortiori is an explanation (of the corresponding type). For our defining conditions for prediction and retrodiction in effect add to the conditions for explanation (in which the element of time does not function) certain added restrictions of a temporal character. (p. 329)

But the symmetry thesis was under attack from other directions as well, and some of what Rescher still shared with Hempel was also vulnerable. Rescher’s colleague Adolf Grünbaum gave the example of how we may predict the coming of the storm from the falling of the barometer, but not explain it that way.6 The correlation between the two has a common cause, the falling pressure in the atmosphere, which could be cited to explain both. But this has an extra bite. It goes beyond the asymmetry that Rescher had shown here. Grünbaum example shows that the concept of explanation that Rescher was working with at the time was definitely too minimalist, since the example goes against the above “whenever a prediction or retrodiction (of whatever type) is given, so a fortiori is an explanation (of the corresponding type).” The prediction of a storm could have been made on the basis of a falling barometer, even if—quite contrary to the actual history—the use of barometers had developed without insight into what they were measuring or how storms develop. 1.2

CRITERIA FOR ASSESSING EXPLANATIONS

A clearly salient question at the time was, of course, to what extent the resources relating solely to the ‘theory of confirmation’ could be mobilized to counter or at least analyze those criteria. Rescher addressed this in a joint article with Brian Skyrms a few years later (Rescher and Skyrms 1968). Using only conditional probability to define the relevant notions, they arrived at two criteria for assessment of offered or competing explanations (while noting that there are other relevant criteria as well); (i) Its comparative strength as determined by the extent to which it renders the occurrence to be explained more likely than competing al341

RESCHER STUDIES

ternative explanations manage to do, and (ii) Its explanatory power as determined by the extent to which it renders the occurrence to be explained more likely than possible alternative occurrences.7 The main result then demonstrated is that the two criteria are in tension, leading sometimes to opposite evaluations, though in practice, “an outright conflict between the explanatory power and the comparative strength of explanations can only occur in cases of explanations with low explanatory power”. (ibid.) 1.3

A FIRST SALLY AGAINST SCRIVEN’S PARESIS EXAMPLE

But these criteria can be used to illuminate Michael Scriven’s wellknown counter-example to Hempel’s requirement that to be explanatory, a hypothesis must bestow high probability on the facts to be explained: We can explain but not predict whenever we have a proposition of the form “The only cause of X is A” (I)-for example, “The only cause of paresis is syphilis.” Notice that this is perfectly compatible with the statement that A is often not followed by X-in fact, very few syphilitics develop paresis (11). Hence, when A is observed, we can predict that X is more likely to occur than without A, but still extremely unlikely. So, we must on the evidence, still predict that it will not occur. But if it does, we can appeal to (I) to provide and guarantee our explanation. . . . Hence an event which cannot be predicted from a certain set of well-confirmed propositions can, if it occurs, be explained by appeal to them.8

In their terms, Rescher and Skyrms could then put Scriven’s point as follows: Scriven has focused on a class of explanations which, although very weak in explanatory power, are yet the strongest available explanations simply by virtue of being the only available explanations. Explanations with very little explanatory power may nevertheless qualify as good explanations by virtue of their high comparative strength. (ibid.)9

1.4

342

MORE TROUBLES FOR HEMPEL

RESCHER ON EXPLANATION AND PREDICTION

But the scene was changing more and more on the explanation stage. Sylvain Bromberger’s lecture “Why-Questions” in the Pittsburgh Philosophy of Science Lecture Series had introduced context-sensitivity.10 There we heard his famous flagpole example: the height of a given flagpole (with fixed position of the sun) and the length of its shadow are each calculable given the other. But, Bromberger claims, only one direction of the argument serves as an explanation, while the other only justifies a claim to knowledge: the height of the flagpole explains the length of its shadow.11 Meanwhile Wesley Salmon, in yet another Pittsburgh Philosophy of Science Lecture, offered many examples of statistical explanation that did not fit the Hempelian framework.12 To begin Salmon suggested that explanatory relevance might just be statistical relevance, but he soon found the need to add in notions of causality, though with a sustained attempt to keep that on an empirical basis. 2. AN EROTETIC TURN As a student I could only take in all these new developments, that were simultaneously broadening the scope and enlarging the budget of problems and paradoxes for accounts of explanation. When I turned to the subject myself a decade later, it was to propose a different take on those problems and paradoxes. My approach was at odds with the orientation that I mentioned above, that focused on those concepts abstracted from pragmatic factors. Although Nicholas Rescher had shared that orientation, it was Nicholas Rescher who published my “The Pragmatics of Explanation” in the American Philosophical Quarterly. While this approach handles the main puzzles differently, its criteria by which explanations are assessed follow the lines of Rescher’s work. My take on explanation started from Bromberger’s “Why-questions”, so Salmon rightly called it “erotetic”. But my emphasis was on the pragmatics: the sensitivity to contextual factors that relate to the questioner’s interests, concerns, and values. In this respect I gladly admit to working in Rescher’s shadow, given how much he has emphasized and advocated a pragmatic over-all orientation in philosophy, while always insisting on support in the form of technical studies. 2.1

TASKS FOR AN ACCOUNT OF EXPLANATION

An adequate theory of explanation must (a) clarify the exact relationship

343

RESCHER STUDIES

between explanation and description, (b) make clear how there can be asymmetries and legitimate rejections of explanation, and (c) show how explanations are to be evaluated as better or worse explanations. I submit first of all that: (1) what is or counts as explanation depends on the context. The explanation must be a statement, but one that answers a request of a certain sort (a why-question) whose features are contextually determined. (2) The main contextual factors are the relevance relation and the contrast class. These factors determine exactly what is requested, and account for the possibility of rejections and asymmetries that we have encountered. Thirdly, (3) while any given explanation is in itself simply a description (factual information of some sort), its evaluation as explanation depends on how good it is as an answer to the why-question which it attempts to answer, in the context in which that question was posed. 2.2

THE LOGICAL STRUCTURE OF QUESTIONS

According to this account, being an explanation means being an answer to a certain sort of question. A question is a request for a certain sort of information—in other words, a description. But it does not follow that explanation is the same as description! Here is an analogy: Every daughter is a woman and vice versa, necessarily. Every man is a son and every son is a man. Yet being a daughter does not mean being a woman; and man is not the same as son. Daughter and Son are relational concepts. Similarly, explanation is a relational concept: it classifies the given description (which can be any bit of factual information) as related to something else (a question, an explanationrequest) in a certain way. No particular sort of information is definitive of explanation. What is definitive is: being the sort of information that happens to be requested (through a particular why-question). But what exactly is a why-question and

344

RESCHER ON EXPLANATION AND PREDICTION

what does it ask? To determine this we must first describe questions in general.13

2.2a QUESTIONS IN GENERAL A question poses a set of alternatives (more or less precisely specified) and requests that some of these alternatives be either asserted or eliminated. Here are three question forms: • The simplest is a yes-no question: “Did Adam eat the apple?” Here there are just two alternatives posed, and the request is that one be eliminated, the other asserted. • “Which fruit grew on the Tree of Knowledge of Good and Evil?” also requests that all but one alternative be eliminated, but the set of alternatives is not precisely specified. The answer “Apples” can be taken to assert one alternative and eliminate all the others, whatever those might be. Thirdly, the question • “What are some examples of prime numbers between 4 and 25?” can be answered by giving a list of one or more numbers, asserted to be prime; here there need not be any elimination. Next we must note a clear distinction here between direct answers and other responses. Consider the following question and responses: “Which mountain is the largest active volcano in Europe?” (1) Etna (2) Etna is a mountain and is the largest active volcano in Europe (3) Mt. Blanc is a mountain and is the largest active volcano in Europe

345

RESCHER STUDIES

(4) Etna is a mountain and is the largest active volcano in Europe, while Vesuvius is the second largest (5) Etna is a large active volcano in Europe (6) There are no active volcanoes in Europe (7) Ask me something else! These responses exemplify types of answers. (1) we can take as short for (2), a common sort of abbreviation for a fully explicit answer. (2) is a direct answer to the question. So is (3), though most of us know that it is false. (4) is not a direct answer, because it says more than was requested. (5) is not a direct answer because it says too little, yet it might help—it is a partial answer. (7) should perhaps be called merely a response, and not an answer at all, for it does not purport to provide any of the requested information. Answer (6) is especially interesting, because it rejects the question. It bases this rejection on the claim, in effect, that this question has no true direct answers. Let us call this a corrective answer: the question had a presupposition which is here denied. This gives us a simple, general way to characterize presuppositions of questions, at least in part: If proposition A is implied by every possible direct answer to a question, then the question presupposes A.

A question can then be corrected or rejected by denying such a presupposition. 2.2b WHY-QUESTIONS With this in hand, let us see what sorts of requests are made by whyquestions. A why-question has a (tacit or explicit) contrast class which poses the alternatives. At least in the typical case, the questioner presupposes that the topic of his question is a true proposition while the other members of the contrast class are false. For example, “Why did Adam eat the apple?” has topic: Adam ate the apple

346

RESCHER ON EXPLANATION AND PREDICTION

contrast class: Adam ate the apple, Adam gave the apple back without eating it, Adam threw the apple to the ground, . . . etc. The question “Why did Adam eat the apple?” has the same topic, but a different contrast class (“Eve offered it to Adam”, “Apples are Adam’s favorite food”, etc.) The questioner presupposes that Adam ate the apple, and did not do the contrary things contemplated by way of contrast. So if a direct answer to the first question were completely spelled out, it would begin like this: Adam ate the apple, rather than giving it back, throwing it to the ground, . . . etc., because ... What follows the “because” is also contextually (usually more or less tacitly) limited. The request comes with a specific relevance relation: the questioner wants information relating to one specific sort of factor in the situation he or she is asking about. In this example, what counts as a relevant sort of answer for the questioner may well be whatever relates to Adam’s psychology—his thinking and feeling about Eve’s offer (he wanted to please her)—or alternatively with his physiology (he was hungry). From this it follows that the questioner also presupposes that at least one of the factors which is relevant in that particular way is present. We conclude therefore that the complete form of a why-question and its direct answers is this: Question: Why A rather than B, C, D, . . . ? [please give me some information which is relevant in sense R] Direct Answer: A, and not B, C, D, . . . because Q [which is relevant in sense R to A] and the question can be corrected by mentioning anything which rules out that there can be a direct answer of this form. 2.2c CONTEXTUAL VARIANCE Scientific explanation is explanation in response to any sort of why-question, provided it draws on science for crucial parts of the supplied

347

RESCHER STUDIES

information. The contrast classes and relevance relation may be still differ from case to case. In many cases, the request may well be for the sort of thing characterized roughly as causal factors, factors contributing to the production of the event. But if, for example, certain aspects of atomic structure are explained by the Exclusion Principle, or even more fundamentally by a symmetry principle, the explanation is scientific though not causal. We can illustrate the contextual variance of contrast-class and relevance with “ordinary” examples. Alan Garfinkel gave the following example: A prisoner was asked by the prison chaplain why he robbed banks. The prisoner answered: “because that is where the money is”.

What makes the example funny is that we are sure the chaplain had a different contrast-class in mind. In this context the reply is inappropriate It might well have been appropriate as answer to the same question if voiced by another prisoner. Aristotle already gave an example of how the relevance, as opposed to the contrast, can vary contextually (the lantern example of Posterior Analytics II, 11). In modern dress the example would run as follows. Suppose a father asks his teenage son, ‘Why is the porch light on?’ and the son replies ‘The porch switch is closed and the electricity is reaching the bulb through that switch.’

At this point you are most likely to feel that the son is being impudent. This is because you are most likely to think that the sort of answer the father needed was something like: ‘Because we are expecting company.’ But it is easy to imagine a less likely question context: the father and son are rewiring the house and the father, unexpectedly seeing the porch light on, fears that he has caused a short circuit that bypasses the porch light switch. In this second scenario, he is not interested in the human expectations or desires that led to the depressing of the switch.

2.3

HOW ARE ASYMMETRIES AND REJECTIONS HANDLED?

We have seen how a why-question can be rejected: by a denial of its main presupposition, that is, by the assertion that all its possible direct

348

RESCHER ON EXPLANATION AND PREDICTION

answers are false. But what about the asymmetries? Think now of Bromberger’s example of the flagpole and its shadow. If someone asks “Why is that pole so high?” we must try to determine the contrast class and relevance relation, which together give content to his question. Most likely the contrast class is simply a set of possible heights for flagpoles. But what sort of information is being sought? If (as is also likely) the questioner seeks information about the origin of this pole (how it was made, what sort of flag it is made to support, etc.) then the response “Because it has a shadow 70 ft. long” will not do! For though true, it is not relevant. But remembering the porch light example, we can imagine that the question “Why is the pole so high?” is instead a request for information about what the pole is for, for what purpose it was made that height. And in that case, one can imagine a scenario in which the salient relevant feature is precisely the length of the shadow that the pole casts at a certain time. In general, suppose that relative to a given theory (say, total accepted science) A and B are equivalent, but A describes a factor of one sort, and B describes a factor of another sort. Suppose moreover that in a single context, the question “Why . . . ?” is understood to request factors of the type described in A. (That is, the intention underlying the question makes that sort of factor relevant.) In that case we will at once have an asymmetry. The answer “Because A” to the question “Why B?” satisfies the relevance condition, but in that same context the reply “Because B” to “Why A?” does not. From the preceding examples we conclude also, however, that asymmetries are not cast in stone. Both the prisoner-chaplain dialogue and the one about the porch light show that the relevance relation can vary drastically from context to context. 2.4

EVALUATION OF EXPLANATIONS

There may not be something wrong, in itself, with a request for statistically relevant astrological factors. However, science may tell us that astrological factors just are not very good indicators of anything. We need not reject the question as logically defective, but we can simply deny, as a mere matter of empirical fact, that it has any good answers. This brings us to the last item needed in any adequate theory of explanation, and we can make contact with Rescher and Skyrms 1968. We need a scheme for evaluating explanatory answers in general as good, better, and best, but as they point out, this evaluation can take several, relatively

349

RESCHER STUDIES

independent forms. So far we have only dealt with a delineation of the set of answers, i.e. possible explanations, good, bad and indifferent. Evaluation becomes pertinent only when a given response qualifies as a genuine, relevant answer to the explanation request, and only then, only in the context of that request. Some of the difficulties of previous theories of explanation may have been due to a failure to separate classification from evaluation. Suppose a specific why-question is asked. The respondent draws on theories and information which he himself believes or considers adequate. The questioner may not accept all the same theories, and some of the information will be new to him. A third person who witnesses the dialogue may share only some theories with the questioner and respondent, and may have his own ideas about how reliable the new information is. The basis for evaluation will therefore not be the same for these three evaluators. So the first point is that evaluation will be relative to a given basis. We will call the evaluation scientific if the basis adopted is entirely part of accepted scientific theory and opinion. Let us suppose that the answer given really is a direct answer in the sense described above, and that the basis for evaluation has been chosen. Then the first evaluation question is: • how likely is it that this answer is true? If there are rival direct answers, we can evaluate them comparatively in the first instance by asking how probable they are (relative to the chosen basis). The second evaluative question to ask is: • how telling is the answer, i.e. how much does it favor the topic against the rest of the contrast class? If the question is why an automobile accident happened, we may for instance say that the bad condition of the brakes is a much better answer than the rainslippery road condition. The reason for this evaluation would be that the brake defect made an accident much more likely, while the rain made an accident only a little more likely than not. In that sense, then, the former explanation adduces a much more telling bit of information than the latter. We can easily see here that to be telling, the answer does not need to make the topic highly likely. It only needs to shift the probability further toward the topic, so to speak. This parallels the Rescher-Skyrms diagnosis of Scriven’s paresis example. We could imagine for example, that perhaps

350

RESCHER ON EXPLANATION AND PREDICTION

the probability of an accident on this road was normally 1%, and rain increased this to 4%, while the brake defect increased it to 20%. Then the latter would provide a more telling answer than the former. • If an answer is both likely to be true and telling, does that suffice? No, not in general—there is a further criterion. This brings us back to Grünbaum’s example of the barometer and the storm. To vary the example a bit: if we ask why a certain person died, we may first hear that he had a very high blood cholesterol level, next that he had a heart attack, and finally that he had accidentally ingested a large dose of cyanide. The first answer was the most likely to be true, and it is telling: it does make death more probable. But the cholesterol, which certainly makes death by heart attack more likely, becomes irrelevant when the second answer arrives. The third answer makes both of the preceding irrelevant. Ingesting a large dose of cyanide is sufficient to produce death (by stopping the heart), while heart attacks in general are not always fatal. We say that the later answers provide factors which screen off the earlier mentioned factors. This screening off relation is very important to the evaluation of answers, because simple statistical information may include accidental correlations. Conversely, there may be no general statistical correlation and yet the answer may turn out to be very telling! This can happen once we look at a finer classification of cases, a point that was effectively emphasized by Nancy Cartwright. Suppose that in a certain small population there is no correlation between heart disease and smoking. Should we at once deny that smoking is a telling explanatory factor in cases of heart disease? Not necessarily. For this absence of overall correlation is compatible with the following: in the case of people who exercise and are physically active, there is a higher incidence of heart disease among smokers, and similarly among those people who do not live a physically active life, the incidence of heart disease is also higher among the smokers among them.

The averages over the population as a whole may not reveal this. (This possible reversal of correlation when we look at a finer classification of cases is often called Simpson’s paradox, after a statistician who studied it.)14 So we have three initial criteria to apply when evaluating explanations. We must ask:

351

RESCHER STUDIES

• how likely we consider the answer to be true, • how much it favors the topic as against the rest of the contrast class, and • whether or not it is screened off by other relevant factors. These are very general criteria, that apply to explanations of all sorts. It will be readily apparent how these criteria are formulated entirely in terms of probability, and in that respect conceived along the same lines as those discussed by Skyrms and Rescher in 1968. Of course, as they also emphasized, neither set of criteria purports to be exhaustive for the evaluation of offered explanations. In fact, the two sets of criteria, while formulated in the same terms, are complementary. To single out scientific explanations we have the further criterion: does the answer draw on science? This is a criterion to apply to the basis, and removes some of the relativity of the evaluation. We might grant that something is a good explanation as evaluated relative to one basis and not good relative to another basis. But then we can also go on to point out that these bases themselves can be evaluated. The empirical sciences are likely to do better in this regard than astrology, for example. 3. THE LIMITS OF PREDICTION Already in the 50s and 60s we can see the beginning of Rescher’s long fascination with the limitations to prediction. The fruits of this fascination, both reflections and technical results, were eventually elaborated systematically in his Predicting the Future: An Introduction To The Theory Of Forecasting. What I would like to do here is just comment on one strand in this complex subject, and connect the theme to Rescher’s lecture “Luck”, his 1989 Presidential Address to the Eastern Division of the American Philosophical Association. 3.1

LIMITS TO PREDICTION IN DISCRETE STATE SYSTEMS

Rescher’s 1963 article discussed above, on discrete state (DS) systems, introduced serious limitations to predictability. Two types of systems were considered:

352

RESCHER ON EXPLANATION AND PREDICTION

• deterministic, subject to laws of the type: If st(t) = Si,then st(t + 1) = Sj. •

indeterministic, subject to probabilistic laws of the type: If st(t) = Si, then st(t + 1) will be one of the states Sjl, Sj2, ..., Sjn,with probabilities p1, ..., p2 , …, pn respectively. (These must sum to 1)

The concept of prediction was precisely defined for such systems, as follows: In the context of DS systems, a predictive argument is one that concludes that : st(t) = Si on the basis of an identification of a group of states, st(tl), st(t2,), ..., st(tn), where all tj < t. The argument may be either deductive (Dprediction), or probabilistic, either in the strong sense (Ps-prediction: the state is (conditionally)more likely than not) or in the weak sense (Pwprediction: the state is (conditionally) more likely than any of its competitors, taken singly).

But then come the no-go theorems. For example, Rescher’s Theorem T6.1 and T6.2: in an indeterministic DS system, both strong and weak prediction may be either uniformly or selectively impossible. Another result just mentioned along the way, but very telling, is that an indeterministic DSsystem may be such that its near-term future is predictable (even Dpredictable), but its distant future is not even Pw-predictable.15 3.2 FORECASTING SCHEMES Early on in his career, Nicholas Rescher cooperated with Olaf Helmer on the development of the Delphi method. The theoretical basis of the Delphi methodology was first set out in a 1958 RAND publication (paper P-1513) later reprinted as Helmer and Rescher1959; the history of its development and subsequent application is richly detailed in Rescher’s Predicting the Future. The basic idea of the Delphi method is to tap into the accumulated experience of a profession, but in an interactive process that is meant to elicit 353

RESCHER STUDIES

conclusions that are ratifiable by that professional community itself. A series of successive questionnaires is used to elicit responses from a panel of experts to arrive at an aggregate prediction about future developments. The experts are interrogated individually (without face-to-face interaction) about their expectations for hypothetical future events. The responses are assembled and statistically analyzed. Then the participants are provided with some aggregated information regarding the initial response, and possibly (anonymous) comments about or arguments for or against various positions. The respondents are asked to submit revised estimates, together with reasons for agreeing or disagreeing with the initial consensus. In later rounds the procedure is repeated; normally the responses will converge, in the sense that the range of estimates is narrowed. While the Delphi methodology enjoyed popularity only for a limited time, we can see it as part of the development of expert systems and of consensus-deliberative methods (such as are more informally exercised in the many ‘consensus conferences’ between experts and laity in e.g. cancer research and environmental impact studies). What requires empirical study of course is whether the outcome of such a consensus reaching process yields better calibrated expectation values. Psychological and sociological factors may be operative despite the precautions that are integral to this methodology. 3.3

RELEVANCE TO EPISTEMOLOGY

It is part of Bayesian epistemology to argue that expertise is largely unformulated, hardly ever capturable in a set of explicit propositions, but manifest in the expert’s expectation values. What is needed for the expertise to be drawn on as practical guidance clearly requires listening to all the relevant experts in the field—but first of all they are not generally in perfect agreement, and secondly, the expression of opinion can on specific opinion easily be swayed by contextual factors. The Delphi method attempts to combine the different opinions in a way that relies again on the same implicit expertise, and tends to eliminate the random disturbances that affect the expression of that expertise. Assumed from the beginning in such an approach is precisely what sets Bayesian epistemology apart from traditional epistemology, where it seems generally simply assumed that a person’s opinion can be captured in principle in a set of propositions expressible in his/her own language. But then, traditional epistemology is also largely immersed in the presumption 354

RESCHER ON EXPLANATION AND PREDICTION

that there is an objective weighing of the extent to which opinion is proportionate to the evidence—with the evidence also conceived of as captured in such a set of propositions. This is of course the ideal of Induction, according to which there is a rule or recipe that is applicable, and reliably leads to a unique extrapolation of given evidence. Blatantly stated in this way, the idea is of course vulnerable to wellknown traditional objections. Recognizing these, Hans Reichenbach introduced a more modest claim. First of all he diagnosed the British empiricists’ debacle: they had accepted a criterion of adequacy imposed by their rationalist opponents, who asked for certainty. By such a criterion it would follow at the very least that, to be justified, an inductive method would have to be shown to lead to reliable conclusions with certainty, whenever applicable, and the conditions of applicability directly accessible to us. But the world might be such as to confute any method at all for predictions—without that being something we could rule out beforehand. So he proposed a weaker criterion, that would if satisfied be the best we could have and good enough. It was this: that the method should lead demonstrably to the truth, if any method could— that is, under conditions in which there is any method at all that could succeed

Reichenbach advanced the claim that there is such a method, and indeed, offered the notorious straight rule, the very numerical induction rule that historically everyone had despised, as the basis of all induction. And his justification was this: if in a series of instances there is a limiting relative frequency for a certain property, then the straight rule provides a series of estimates for prediction of what will come next that converge to that frequency—and if there is no such limiting relative frequency then no method at all can concoct a reliable scheme for prediction.16 So wasn’t this success? And wasn’t it the very success that was wanted? Not at all. If a weather forecaster gave us his predictions for rain the next day, following Reichenbach’s rule, he would after a thousand days or so become extremely boring and uninformative. For the probability for rain he would announce for tomorrow would differ from today’s by at most 1/1000, even if meanwhile the once blue heavens had become dark with threatening clouds. We would want his predictions to be sensitive to conditions that single out sub-sequences in the long series of days. To make this challenge realistic, these conditions must be ones that we and he could check on before 355

RESCHER STUDIES

the announcement. But no rule could demonstrably meet that standard, not even if we add “if any rule could”. For any scenario at all there is a rule that does well in that way, but there is demonstrably no learnable rule that does well under all scenarios. It was his own student Hilary Putnam who first introduced the argument to prove this, though it was later made precise by statisticians.17 So precisely what Reichenbach wanted to have met as criterion, when refined by what interests us in practice, cannot be had. 3.5

THE THEME OF CHANCE, LUCK, AND FORTUNE

Nicholas Rescher avoided the impasse of traditional epistemology from the beginning. His theory of evidence (1958) already proceeds entirely in terms of probability, and concentrates on how we can assess evidence offered, rather than entering upon the illusory quest for a guarantee or demonstration that by proceeding in one way rather than another, our acting on what we take as evidence will reliably lead to true conclusions. When Rescher later addresses forecasting in general, it is with an eye throughout to the limits of prediction. While early on Rescher does not deny the possibility of a unique objectively rational likelihood measure, even his 1958 proceeds carefully by way of general conditions that any such measure must satisfy. But this brings us of course to the question: are we to think that, if we manage our opinion rationally, we are likely to have our expectations met? This brings me to the theme of Rescher’s 1989 APA Presidential Address, on the role of luck in our lives. There is no rule that will, with an a priori guarantee, lead us to reliably fulfilled expectations: no matter how conscientiously we “play by the rules” in matters of factual inquiry, there is no categorical assurance that we will answer our questions correctly.18

Nevertheless, there are opinions, rationally extrapolated from our evidence, that do issue reliably fulfilled expectations—we are happy to record this fact, since it is the very basis of our survival. But we need to record the fact in humble gratitude, for it depends not just on our prudent and rational management but on nature’s going along with it. The dinosaurs were long the masters of the earth; their inductive habits were reliable for almost all of that time. There is no reason to think that they were epistemically unadapted to their environment during that time. But their physical environ-

356

RESCHER ON EXPLANATION AND PREDICTION

ment changed, their expectations that had been regularly fulfilled for so long, were increasingly and in the end fatally disappointed. The dinosaurs were epistemically (as well as in other ways!) unlucky in precisely the sense that Rescher explicates: The core of the concept of luck is the idea of things going well or ill for us due to conditions and circumstances that lie wholly beyond our cognitive or manipulative control.19

The dinosaur example is not an argument against prudent and rational opinion management. On the contrary, we know very well that that is the best we can do, and will be the best we can do under any circumstances. But it is in the end our luck that needs to hold out, the world has to remain hospitable to what we have developed, if our opinion is to remain reliable. We have to be lucky.

REFERENCES Baumrin, B. (ed.), Philosophy of Science, The Delaware Seminar, Vol. I. (New York: Interscience Publishers, 1963). Bromberger, Sylvain, “Why-Questions,” in Colodny 1966, pp. 86-108. Bromberger, Sylvain, “An Approach to Explanation,” in Butler 1965, pp. 72-105. Butler, R. (ed), Analytical Philosophy, Second Series (Oxford: Blackwell, 1965). Colodny, R. G., Mind and Cosmos (Pittsburgh: University of Pittsburgh Press, 1966). Colodny, R. G., The Nature and Function of Scientific Theories (Pittsburgh: University of Pittsburgh Press, 1970). Dawid, A. P., “Calibration based empirical probability,” Annals of Statistics, vol. 12 (1985), pp. 1251-1273.

357

RESCHER STUDIES

Dawid, A. P., “Comment [on Oakes (1985)],” Journal of the American Statistical Association, vol. 80 (1985), pp. 340-41. Gaifman, H. and M. Snir, “Probabilities Over Rich Languages, Testing, and Randomness,” Journal of Symbolic Logic, vol. 47 (1982), pp. 495548. Grünbaum, Adolf, “Temporally Asymmetric Principles, Parity between Explanation and Prediction, Mechanism versus Teleology,” in Baumrin 1963, pp. 57–96. Helmer, Olaf and Nicholas Rescher, “On the Epistemology of the Inexact Sciences,” Management Sciences, vol. 6 (1959), pp. 25-52. Hempel, Carl G., Philosophy of Natural Science (Englewood Cliffs, New Jersey: Prentice Hall, 1966). Hempel, Carl G. and Paul Oppenheim, “Studies in the Logic of Explanation,” Philosophy of Science, vol. 15 (1948), pp. 135-175. Oakes, D., “Self-calibrating priors do not exist,” Journal of the American Statistical Association, vol. 80 (1985), p. 339. Putnam, Hilary, “Reflexive Reflections,” Erkenntnis, vol. 22 (1985), pp. 143-154. Putnam, Hilary, “Degree of Confirmation and Inductive Logic,” in Schilpp 1963, pp.761-783. Rescher, Nicholas, Predicting the Future: An Introduction To The Theory Of Forecasting (Albany: State University of New York Press, 1998). Rescher, Nicholas, Scientific Explanation (New York: Free Press, 1970). Rescher, Nicholas, Temporal Modalities in Arabic Logic (Dordrecht: Reidel, 1966). Rescher, Nicholas, “A Note on ‘About’,” Mind, vol. 72 (1963), pp. 268-

358

RESCHER ON EXPLANATION AND PREDICTION

270. Rescher, Nicholas, “A Theory of Evidence,” Philosophy of Science, vol. 25 (1958), pp. 83-94. Rescher, Nicholas “Discrete State Systems, Markov Chains, and Problems in the Theory of Scientific Explanation and Prediction,” Philosophy of Science, vol. 30 (1963), pp. 325-345. Rescher, Nicholas, “H2O: Hempel-Helmer-Oppenheim, an Episode in the History of Scientific Philosophy in the 20th Century,” Philosophy of Science, vol. 64 (1997), pp. 334-360. Rescher, Nicholas, “Luck,” Proceedings and Addresses of the American Philosophical Association, vol. 64 (1990), pp. 5-19 Rescher, Nicholas, “On Prediction and Explanation,” The British Journal for the Philosophy of Science, vol. 8, (1958), pp. 281-290 Rescher, Nicholas and Brian Skyrms, “A Methodological Problem in the Evaluation of Explanations,” Noûs vol. 2 (1968), pp. 121-129. Salmon, Wesley, “Statistical Explanation,” in Colodny 1970, pp. 173-231. Salmon, Wesley, Statistical Explanation and Statistical Relevance (Pittsburgh: University of Pittsburgh Press, 1971). Schervish, M. J., “Comment [on Oakes (1985)],” Journal of the American Statistical Association, vol. 80 (1985), pp. 341-342. Schilpp, P. A. (ed.), The Philosophy of Rudolf Carnap (La Salle, Ill.: Open Court, 1963). Scriven, Michael, “Explanation and Prediction in Evolutionary Theory,” Science, vol. 130 (1959), pp. 477-482. van Fraassen, Bas C., “The Pragmatics of Explanation,” American Philosophical Quarterly, vol. 14 (1977), pp. 143-150.

359

RESCHER STUDIES

NOTES 1

I very happily dedicate this to Nicholas Rescher, on his 80th birthday, in gratitude for his guidance and help during my graduate studies. My intellectual debts to his work are not at all exhausted by what I mention below, but extend to many other subjects, including tense logic, the logic of commands, and the importance of pragmatism for the history of epistemology.

2

Free logic is, as the name indicates, free of certain existence commitments, and seemed needed in this context: the present tense “Charles V is not alive” is true although the subject does not exist at present. This was the beginning, for me, of a good deal of work along this line.

3

Hempel 1966, p. 48.

4

The history and pre-history of this episode is presented at length, from a participant’s point of view, in Rescher 1997.

5

Hempel and Oppenheim 1948, p. 138.

6

Grünbaum 1963, p. 309.

7

Rescher and Skyrms 1968, page 128

8

Scriven 1959, p. 480; cited Rescher and Skyrms, p. 128-129.

9

Rescher and Skyrms 1968, p. 129.

10

Bromberger 1966.

11

Bromberger1966, p. 92; see also Bromberger 1965.

12

I am referring here to Salmon 1970; more detailed studies appeared later at his hand, beginning with Salmon 1971.

13

In this subject I was indebted to Rescher’s colleague Nuel Belnap; the first encounter with logic I had in Pittsburgh was Belnap’s seminar on the logic of questions.

14

To see how this is possible, imagine that in the case of active people the probability of getting heart disease is low, while among non-active people it is pretty high. On the average it could be, let us say, 20%. Among smokers, in each case, the probability may be raised by a small amount. But this is still low among the active smokers and high among the non-active smokers. The average may be again somewhere between the high and the low, depending on the proportion of active people in the population. That new average can be just 20% again,

15

Rescher 1963, p. 338.

16

Just to elaborate on the first part: the sequence of predictions will be well calibrated if the announced probabilities match the actual frequencies. Since the

360

RESCHER ON EXPLANATION AND PREDICTION

NOTES

straight rule’s announced probability for the next item is always the actual frequency so far, that will match what has been found up to that point—and these probabilities will converge to the relative frequency in the entire series even if it is infinite. It is therefore easy to see how poorly this criterion, taken by itself, evaluates what we would really want from a predictor. 17

In Putnam’s 1963 contribution to the Schilpp volume on Carnap’s philosophy he gave a diagonal argument to refute the hope for a ‘universal’ inductive method. See further Putnam (1985), Gaifman and Snir (1982), and Oakes (1985) with the discussion between Oakes, Dawid, and Schervish accompanying that article on the next three pages. The argument is related to the simple proof that there is no ‘universal’ computable function. For suppose the computable functions are {F(i)} and U(i,n) = F(i)(n) for all i, n. Then U+1 is computable if U is, and differs from each F(i) at argument i; hence is not computable. Of course this shows that the list {F(i)} is not effectively specifiable. The same is true of the computable forecast systems, as defined by Dawid et al., and applies mutatis mutandis to any significant subclass we might want to consider (such as those basing their calculations on ‘acceptable’ scientific theories plus data, etc.).

18

Rescher 1990, p. 6.

19

ibid. p. 7.

361

Nicholas Rescher on Scientific Progress: Science in the Face of Limited Cognitive and Technological Resources Theodor Leiber Roland Wagner-Döbler We are like sailors who have to rebuild their ship on the open sea, without ever being able to dismantle it in drydock and reconstruct it from the best components. Otto Neurath1 It is difficult, indeed impossible, to predict the future of natural science, for we cannot forecast in detail even what the questions of future science will be—let alone the answers. […] [T]he future of science is an enigma for us. Nicholas Rescher2 1. INTRODUCTION: VARIETIES OF PROGRESS

F

irst of all, ‘progress’ is a basic concept in the philosophy of history of modern times. Mostly, under ‘progress’ is understood a teleological, or intentional change brought forward by human action insofar the change is assessed as improvement. Therefore, in general progresses differ from developments by the criterion of increasingly better realized aims. Within the philosophy of science of modern times ‘progress’ usually means the growth of theoretical knowledge and its technical utilitization which seems to take place within ever shorter time intervals. In contrast to theories of cultural decline (starting with an hypothesis of a ‘golden age’) and theories of circulation of history, progress in the modern sense presupposes the discovery of ‘historical consciousness’. Thereby, as an idea of unlimited progress in modern times (in the occidens) it is sometimes seen as a secular reincarnation of the Christian theology of history. Whether in the form of a secularized Christian-Jewish expectation of salvation, or as a consequence of the development of the natural sciences and the advance-

RESCHER STUDIES

ment of control of nature caused by it, ‘progress’ is the concept by which the expectation of the future is formulated in modern times, sometimes connected with theories of lawful successions of certain phases of (progressive) development. The conception of progress in European enlightenment is based on the assumption that a free development of the intellect and an enhancement especially of knowledge from the natural sciences will lead by itself, automatically, to a humanization of society. Insofar, ‘progress’ is taken to be synonymous to enlightenment as a historical process. In the course of the 19th century, the concept of progress gradually looses this practical philosophical characterization which connects it with the aims and methods of enlightenment. An advancement of mankind is, then, increasingly less seen in its political and moral abilities set free be enlightenment, but rather in the (pretended) ‘unlimited’ technical possibilities of control of nature. The rise of the standard of living gained through the power of scientific knowledge and technical know-how is taken as the measure of progress; while enlightenment’s ‘progress of mankind’ is more or less reduced to the forceful production of industrial culture. With respect to theory formation the concept of progress has found increasing interest in philosophy and history of science in recent times. In general, there is almost no disagreement in the scientific as well as in the majority of the philosophical community that science, as a whole, is a progressive enterprise. Nevertheless, there has been, and continues to be, a lot of dispute how ‘scientific progress’ can be adequately characterized. Some spectators and philosophers of science, however, deny any kind of scientific progress. What is commonly assessed as scientific progress, in their opinion erroneously, has to be characterized as an irrational sequence of different discourse fashions. With this opinion, however, only one of those fashionable discourses might be played. Undoubtedly, ‘progress’ is a term intimately connected with normative issues, i.e., it contains value judgements (Niiniluoto 1980). Generally speaking, a ‘progressive’ sequence is constituted by stages of increasing superiority.3 The two extreme forms of such a progressive sequence can be characterized by (1) a gradual and incremental, or cumulative increase of superiority, or by (2) a sharply discontinuous or interruptive one. To the advocates of pattern (1) the idea of gradual incorporations of earlier achievements into present scientific theories seems to be preferable; while the acceptance of pattern (2) goes with the strong emphasis of so-called revolutionary phases during which theories are overthrown and—at a first glance irrationally—replaced by superior competitors. Among scientists

364

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

and philosophers of science there is still no consensus about which of these patterns is more appropriate. There is also no broad agreement about the content of the superiority, or quality criteria for scientific theories, models, explanations, observations, experience, and experiments (in the course of time): Which features are to be valued as ‘scientifically progressive’? Candidates proposed are (Losee 2004), e.g., predictive success (logical empiricism), convergence upon truth (Charles S. Peirce, Karl R. Popper), consilience (William Whewell), reduction (Ernest Nagel), incorporation-with-corroborated-excess-content (Imre Lakatos), asymptotic agreement of calculations (Niels Bohr), and problem-solving effectiveness (Larry Laudan, Philip Kitcher, Nicholas Rescher). Further quality criteria of scientific theories can be identified (Leiber 2007, 209–211): power of unification or force of explanation—guaranteed by causal or functional mechanisms or empirical invariances, respectively, and inferential correctness or systematicity, and simplicity; history of success, i.e., being a successful explainer of (known) facts; fecundity, or progressivity with respect to epistemology, methodology, technical application or heuristics; mathematization4; continuity of mistake and correction5; observation inclusion.6 From a general point of view, it is important to distinguish the (more) descriptive and the (more) explanatory aims of the scientific enterprise. According to the first perspective, science aims at describing the properties and relations of various substances, while in the second perspective scientists concentrate on establishing explanations for as many substantial properties and relations as possible. As far as the descriptive and explanatory efforts of natural science are concerned in general, there has certainly been a growth of knowledge: we know more, and we know more details about the constituents of matterenergy with increasing precision; and we have gained more knowledge of lawlike regularities (i.e., system-law hypotheses) than our predecessors had. In these respects, we cannot but state that there has been an increase of scientific knowledge over time which surely is to be counted as “descriptive progress”.7 Accordingly, almost nobody denies that science exhibits and produces technical progress in the sense of increasing power and expanded options. It is, however, questionable, whether this ‘accretion of knowledge’ thesis is also true of the successive theories, i.e., can we identify a corresponding ‘theoretical progress’? Or: Is the development of theories and their

365

RESCHER STUDIES

theoretical concepts as cumulative as descriptive progress? Should we identify scientific progress with an increase of descriptive knowledge and be agnostic about theoretical progress? Scientific progress, then, would be instrumentalistically characterized solely by the increasing range and success of science’s forecasts, i.e., it would be taken as irrelevant whether the explanatory power of successive theories increases or not. Then, it simply would be meaningless to speak of theories as being ‘more adequate’, or (approximatively) ‘true’. However, those who emphasize the relevance of theoretical progress in science have basically followed two different approaches. The first one is (gradually more) normative and consists in identifying the type of such progress, as, e.g., incorporation or revolutionary overthrow, and then setting forth the necessary and/or sufficient conditions of scientific progress. A second approach is (gradually more) descriptive and consists in developing a theory of science which identifies the content of ‘goodness’, ‘quality’, or ‘superiority’ that should increase when progress is achieved.8 Perhaps, the two most influential recommendations in this sense are: scientific progress is characterized by (i) the convergent approximation to truth achieved by successive theories, or by (ii) the increasing effectiveness of problem-solving achieved by the application of successive theories. Within a third approach, which also establishes a theory of science, one seeks to explain why science develops as it does. Here, a favourite and most widely promulgated conception is to explicate the mechanisms of scientific progress by analogy to (the mechanisms and theories of) biological evolution. 2. SCIENTIFIC PROGRESS AS COGNITIVE INCORPORATION The most prominent approaches to ‘scientific progress as cognitive incorporation’ are William Whewell’s tributary-river image, scientific progress through reduction as advocated by logical empiricists, and Imre Lakatos’ version of scientific progress as incorporation. Whewell’s Tributary-River View of Scientific Progress According to William Whewell (1847; 1857) the historical development of mechanical physics, and especially of astronomy up to the midst of the 19th century “resembles the confluence of tributaries to form a river” (Losee 2004, p. 11). Accordingly, in his philosophy of science Whewell

366

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

emphasized the process of incorporating past theories and results into present theories. In other words, his tributary-river view conceives progress in science as a continual and homogeneous process of unification. Thereby, according to Whewell even discredited theories may ‘contribute tributaries to a river’. In summary, it is clear from Whewell’s writings that his reconstruction of the historical development of his favourite disciplines (e.g., classical mechanics, astronomy) emphasizes continuity rather than revolutionary overthrow. Whewell tried to historico-inductively establish a philosophy of science, i.e., to develop it bottom-up starting from the inspection of the history of science. Therefore he identified—by induction, or abduction?—a principle of evaluation from the history of science: he called it ‘consilience of induction’. ‘Consilience’ is characterized by the following necessary conditions of the acceptability of progressive theory-replacement: (i) the successor theory must be consistent; (ii) it must be more inclusive than its predecessor, i.e., factual statements must be subsumable under the successor theory; and (iii) this increase in subsumptive, or unificational power must be accompanied by an increase in simplicity (Losee 2004, p. 13). It is not easy, however, to apply Whewell’s criteria of scientific progress because he delivered no general criterion of comparative simplicity.9 Fortunately, he specified a sufficient condition of consilience: the achievement of ‘undesigned scope’ which is characterized by psychological as well as logical novelty. Unfortunately, however, the conceptions of novelty are notoriously vague. In the case of Whewell’s suggestions this means: he thinks of a theory as ‘psychologically novel’ if it is both successfully applicable and unexpected; and he calls a theory ‘logically novel’ if it is successfully applicable to factual statements which are “different in kind from those taken into account when the theory was formulated” (Losee 2004, p. 15). But what does ‘different in kind’ mean objectively? Clearly, the epistemological crux with Whewell’s criterion of undesigned scope is the vagueness of the concept of unexpectedness. In other words, if it is applied in an adequately strict sense, cases of ‘undesigned scope’ show up rather rarely in the history of science, i.e., the criterion singles out very few theories as being progressive. Therefore in the vein of Whewell’s approach, in most cases one must go back from the vague requirement of undesigned scope to the (similarly?) vague requirements of inclusiveness and simplicity (Losee 2004, p. 16).

367

RESCHER STUDIES

3. SCIENTIFIC PROGRESS COMMENSURABILITY

THROUGH

REDUCTION

AND

Reduction (or explanation) as a specific type of cognitive incorporation was especially prominent as a feature of scientific progress in logical empiricism. Its program, which was dominant in the philosophy of science during a period after World War II, is constituted by the following assumptions: 1.

The theories and models (i.e., definitions, axioms, law hypotheses) as well as the experimental findings of empirical science (empirical facts) can be precisely represented by declarative sentences.

2.

Relevant systems of declarative sentences can be reconstructed in the symbolism of formal (statement and predicate) logic.

3.

Philosophers of science have to specify and formulate criteria for the evaluation of laws and theories, i.e., they concentrate on the context of justification.10 These criteria are to be applied, for instance, to assess the correctness or truth of proposed explanations, to determine the degree of evidential support of a law, or to gauge the rationality of theory choice.

4.

The objective evaluation of empirical theories of the natural sciences is possible because there exist elementary empirical sentences (or protocol sentences), i.e., basic sentences the meaning of which is given independently of the theories compared.

5.

The development of science is rational, i.e., theories which were dominant (or accepted) within a domain of science at different times are commensurable (at least to a relevant extent), so that they can be compared systematically.

On this epistemological and methodological basis logical empiricists tried to develop criteria to assess the relational superiority of competing theories. One prominent approach is Ernest Nagel’s theory of reduction which formulates necessary formal and empirical criteria for the replacement of one theory by another, more inclusive one. Appealing to deductive logical

368

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

structure, in logical empiricism reduction (or reducibility) is a quite natural and most important criterion of scientific progress.11 More specifically, Ernest Nagel identified four conditions of successful reduction (Nagel 1961): (1) Relevant theoretical terms of the reduced theory must be linked to the theoretical terms of the reducing theory (connectability; “heterogeneous reduction”); (2) the laws of the reduced theory must be derivable from the interpreted axioms of the reducing theory (derivability); (3) the empirical support of the reducing theory must be better than that of the reduced theory; (4) the reducing theory must prove theoretically fertile, i.e., its theoretical premises must contain potential for developing further the original theory (Losee 2004, p. 29). Several authors, like Paul Feyerabend and Hilary Putnam, have critizised Nagel’s rather narrow, or impractical early conception of scientific progress via reduction, or (strict) deduction, respectively. Later on, these authors agreed that in homogeneous reductions it suffices that the reduced laws are either (strictly) derivable from the reducing premises, or are acceptable approximations to the laws derivable from those premises (Nagel 1974, p. 101). In other words, certain types of approximative reduction are also to be counted as a criterion of scientific progress. On that stage of the discussion Feyerabend insisted that heterogeneous reductions are of much greater importance. According to him, at least for successful high-level theories the predecessor-incorporation model is (completely) inadequate. Actually, Paul Feyerabend recommended the following incommensurability thesis: “what happens when transition is made from a restricted theory T* to a wider theory T (which is capable of covering all the phenomena which have been covered by T*) is something more radical than incorporation of the unchanged theory T* into the wider context of T. What happens is rather a complete replacement of the ontology of T* by the ontology of T, and a corresponding change in the meanings of all descriptive terms of T* (provided these terms are still employed)” (Feyerabend 1962, p. 59).12 It is clear from these statements that an important question in the context of scientific progress for Feyerabend is: How can theories be compared? For low-level theories he conceded that they may be evaluated by comparing their ‘basic’ statements and consequences directly with observational statements. This is possible because for low-level theories there exist background theories which provide a common interpretation for those observational statements and consequences. In turn, this is so because the status of such background theories does not depend on the fate of individ-

369

RESCHER STUDIES

ual low-level theories. These conditions are not fulfilled, however, for high-level theories, i.e., a (relatively) theory-neutral observation language is not available for high-level theories. Consequently, theory comparison becomes an extremely complex and difficult task because high-level theories specify their own observation language (Losee 2004, p. 34). This incommensurabilty approach not withstanding, Feyerabend still maintained that even conflicting high-level theories can be evaluated by rational procedures. Actually, he suggested three of them (Losee 2004, p. 34–35): (1) One could formulate a still more general theory that contains statements which can be used to test the competing theories. (2) One could compare the internal structures of the high-level theory-competitors: the derivational length from basic premises to observation statements may differ, as well as the number and severity of approximations made in the course of the approximative derivations. It was the suggestion of Feyerabend that, ceteris paribus, a shorter derivation and a smaller number of approximations must be estimated higher.13 (3) One could compare the observational consequences of theories with ‘human experience as an actually existing process’. In the end, Feyerabend’s approach to evaluating competing high-level theories offers no definite objective answer (practical incommensurability), while for low-level theories an appeal to (elementary) observation sentences may decide the issue. In other words, his conception of scientific progress follows the accretional incorporation model as far as low-level theories are concerned, while it is revolutionary overthrow which is the typical relation between competing high-level theories. Later on Paul Feyerabend maintained that it is aesthetic considerations—“matters of taste”— that may decide the issue between incommensurable theories: the evaluation of high-level scientific theories can be compared with the evaluation of poems—and that process is not beyond the reach of argument. Lakatos’ Version of the ‘Progress Is Cognitive Incorporation’ Thesis According to Imre Lakatos it is often rational not to accept apparently negative evidence against a hitherto successful theory. This is the case as long as by modifications of the protective belt of the corresponding research programme we can generate theories that ‘contain’ their predecessors and achieve additional success. The research programme is then called ‘progressive’. Lakatos explicitly declared that:

370

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

“within a research programme a theory can only be eliminated by a better theory, that is, by one which has excess empirical content over its predecessors, some of which is subsequently confirmed. And for this replacement of one theory by a better one, the first theory does not even have to be ‘falsified’ in Popper’s sense of the term. Thus the progress is marked by instances verifying excess content rather than by falsifying instances; empirical ‘falsification’ and actual ‘rejection’ become independent” (Lakatos 1971, 100–101).

More explicitly, theory Tn-1 can be replaced by theory Tn within a research program provided that (i) Tn accounts for the previous successes of Tn-1; (ii) Tn has greater empirical content than Tn-1; and (iii) some of the excess content of Tn has been corroborated. Under such conditions the research programme is ‘progressive’. A research programme that fails to advance in this way is ‘degenerating’. Is Lakatos’ criterion of ‘incorporation-with-corroborated-excesscontent’ a necessary and/or sufficient condition of scientific progress? Neither nor. Since on the one hand, it is common opinion that progress is already achieved within a research programme if there is an improvement of the adequacy of theory and observations—even if there is no increase of the range of covered phenomena at all. And sometimes progressive theoryreplacement can even involve a reduction of scope (Losee 2004, p. 40). In contrast to Lakatos’ opinion, the incorporation criterion is also not a sufficient condition of progress within a scientific research programme. The systematical, or logical argument is that a purely extensive addition of a well-confirmed theory X, which is consistent with the central core of a research programme, i.e., the transition from T1 to T2 = (T1 & X), is to be qualified as ‘progressive’ on Lakatos’ incorporation criterion; it is common consensus, however, that such ad hoc conjunctions can hardly be counted as contributions to progress in science. It goes without saying that the appraisal of a scientific research programme normally will change in diachronic perspective: progressive programmes may become degenerate, and vice versa. Also, it is clear that, according to Lakatos, competing research programmes should be compared be referring to their relative rates of progress, and to the relative importance of scientific successes. Lakatos also applied his progress-as-incorporation criterion to the evaluation of competing methodologies which runs as follows (Lakatos 1971, pp. 108–122): Having selected a set of competing methodologies one elaborates the type of rational reconstruction of scientific progress specific

371

RESCHER STUDIES

for each methodology. Then, those rational reconstructions have to be compared against the history of science. If methodology Mn+1 reproduces all the historical episodes reproduced by Mn, and additional episodes besides, the Mn+1 is superior. According to Lakatos, his methodological approach—“incorporationwith-corroborated-excess-content”—is superior to inductivism as well as to methodological falsificationism: In difference to empiricist inductivism, the methodology of scientific research programmes allows for theoretical concepts of unobservables like, e.g., ‘point masses’, or ‘electromagnetic waves’. And contrary to Popper’s methodological falsificationism, Lakatos’ methodology of scientific research programmes accounts for the fact that scientists continue to accept research programmes even if they have encountered dramatic falsifications. In addition, it seems to cover scientific progress in formal sciences like mathematics. 4. SCIENTIFIC PROGRESS AS REVOLUTIONARY OVERTHROW The principal alternative to ‘scientific progress by cognitive incorporation’ is ‘progress through revolutionary overflow’ which involves discontinuity. It is a standard assumption that after such a revolution science is practiced at a higher level represented by the victorious theory. According to Thomas Kuhn, the paradigm case of a successful scientific revolution is the replacement of Aristotelian physics by Newtonian physics. There are a few minimal requirements for a ‘revolutionary’ replacement to be in hand: (i) it must involve activities of human beings that result in (ii) an important change involving (iii) the overthrow of a current interpretation—theory, model, methodological rule, evaluative or experimental practice—such that (iv) scientific communities utilize (for a time) that replacement in their work. The main vagueness of this requirement lies in the notion of ‘important change’. I. Bernard Cohen has maintained that scientific revolutions can be identified despite the vagueness of the general (minimal) characterization of “scientific revolution”. He suggested four tests of revolutionary status (Cohen 1985, pp. 41–47): (1) the testimony of contemporary witnesses; (2) references to the achievement in subsequent documents in the field; (3) the judgments of competent historians of science; and (4) the opinions of scientists working in that field today.

372

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

Kuhn’s Three-Beat Pattern Thomas Kuhn had demanded holistic change, i.e., he maintained that in the situation of a “scientific revolution” one must either accept an incoherent (or contradictory) theory, or even world-view, or else revise so many fundamentals of that theory or world-view together that it is correct to speak of a scientific revolution or the merging into a new world. There is no coherent account of nature between the old and the new world-view (Kuhn 1987, p. 19). Here, a central problem is to decide when a taxonomic change is to be evaluated as being of ‘holistic impact’, i.e., revolutionary. Or else, we might gradualize the notion of ‘important scientific revolution’. What Kuhn had called a scientific revolution, Stephen Toulmin subsumed under the concept “replacement of an ideal of natural order” (Toulmin 1961). One objection to Toulmin’s approach is that he seems not to be able to establish a clear-cut demarcation of revolutionary episodes. For characterizing a scientific revolution, the proposed criteria of “holistic taxonomic change” (Thomas Kuhn) or “replacement of an ideal of natural order” (Stephen Toulmin) are unsatisfactory. Moreover, it seems as if no condition can be identified that is necessary or sufficient for the occurrence of a scientific revolution. However, Kuhn suggested at least a unique pattern of development characteristic of revolutionary episodes, a threebeat sequence of stages: a period of “normal science” followed by a period of conflict followed by a third stage of a new period of normal science under the victorious second paradigm. Clearly, the usefulness of Kuhn’s three-beat pattern of scientific revolutions depends upon what counts as a ‘paradigm’. Authorized by Kuhn himself are the concepts ‘exemplar’ and ‘disciplinary matrix’ from the second edition of The Structure of Scientific Revolutions; they stand for a narrow and a broad sense of ‘paradigm’, respectively (Kuhn 1970, p. 174–175). However, Kuhn’s three-beat pattern fails to fit historical episodes in several respects: (1) There may be revolutionary changes without prior accumulation of anomalies (e.g., the change from Ptolemaic normal science to the Copernican paradigm); (2) contrary to Kuhn’s conviction (Kuhn 1970, p. 77) which has been adopted by Imre Lakatos (Lakatos 1970, p. 119), a higher-level theory may be rejected although there is no competing theory available (e.g., the principle of conservation of mass was abandoned in the face of evidence about radioactive decay processes while a novel theory—later called the theory of special relativity—was not yet available).

373

RESCHER STUDIES

Popper on Progress Through Overthrow-with-Incorporation For Karl Popper, scientific progress also involves growth by cognitive incorporation as well as by revolutionary overthrow. However, while according to Kuhn some developments are progressive because of incorporation and others because of revolutionary overthrow, Popper assessed ‘progress’ when one and the same developmental episode displays both incorporation and overthrow: “[F]irst, in order that a new theory should constitute a discovery or a step forward it should conflict with its predecessors; that is to say, it should lead to at least some conflicting results. But this means, from a logical point of view, that it should contradict its predecessor: it should overthrow it. In this sense, progress in science, although revolutionary rather than merely cumulative, is in a certain sense always conservative: a new theory, however revolutionary, must always be able to explain fully the success of its predecessor. In all these cases in which its predecessor was successful, it must yield results at least as good as those of its predecessor and, if possible, better results. Thus in those cases the predecessor theory must appear as a good approximation to the new theory; while there should be, preferably, other cases where the new theory yields different and better results than the old theory” (Popper 1981, pp. 93–94).

Thus, according to Popper, for progress to be achieved in science the following requirements must be met: (1) There is some evidence statement e implied by successor theory T2, but not by its predecessor T1, (2) e falsifies T1 but not T2, and (3) T2 reproduces all the explanatory successes of T1, and (4) T2 achieves some explanatory success not achieved by T1. 5. SCIENTIFIC PROGRESS AS INCREASING PROBLEM-SOLVING EFFECTIVENESS According to Larry Laudan: “no one has been able to say what it would mean to be ‘closer to truth’, let alone to offer criteria for determining how we could assess such proximity” (Laudan 1977, p. 11).

Accordingly, he concluded that the key criterion for scientific progress is not ‘approximation to truth’, ‘truthlikeness’, ‘verisimilitude’, or the like,

374

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

but it should be the problem-solving ability of scientific theories. Thereby, a scientific ‘solution’ to a problem is accepted even if it is merely approximative14: “Any theory, T, can be regarded as having solved an empirical problem, if T functions (significantly) in any schema of inference whose conclusion is a statement of the problem” (Laudan 1977, p. 25).

For Laudan, the problem-solving model: “transcends the particularities of the past by insisting that for all times and all cultures, provided those cultures have a tradition of critical discussion (without which no culture can lay claim to rationality), rationality consists in accepting those research traditions which are most effective problem solvers” (Laudan 1977, p. 130).

In his view, it is epistemologically unacceptable that ‘Kuhnian holism’ introduces an evaluative relativism during the process of a scientific revolution. His own model—the “reticulational model of scientific change” –, however, seems to fit scientific practice rather well, and avoids such relativistic attitudes. Laudan himself claimed the epistemological superiority of his approach in comparison to Kuhnian holism in which theories, methodological rules and cognitive aims are replaced simultaneously in a socalled revolution. According to Laudan, the reticulational model of scientific progress has normative-prescriptive status, i.e., it: “is intended to be more than descriptive of existing practice. It purports to give a normatively viable characterization of how discussions about the nature of cognitive values should be conducted” (Laudan 1984, p. 104).

A descriptive conception of scientific progress rather similar to that of Larry Laudan has been developed by Philip Kitcher (1993, esp. pp. 127– 177) who concentrates on ‘consensus practices’ within communities of scientists. In difference to Laudan, Kitcher “introduced three types of problem-solution—practical, conceptual, and explanatory—that contribute to scientific progress” (Losee 2004, p. 126). He did not propose necessary or sufficient conditions for scientific progress.

375

RESCHER STUDIES

Some thinkers, e.g., Charles S. Peirce, Nicholas Rescher, Larry Laudan, Philipp Kitcher, David Hull, Stephen Toulmin, and Karl Popper, conceived the mechanisms instantiating progressive processes—in one or the other way—in analogy to theories of organic evolution. Here, increasing fitness in solving problems posed by nature is seen as an analogy of increasing scientific problem solving effectiveness. Generally, if the comparative fitness of theories is ‘measured’, or assessed by standards of evidential support, problem-solving effectiveness, inclusiveness, completeness, and the like, then the evolutionary analogy may certainly have explanatory applications. A number of disanalogies between biological evolution and theorychoice remain, however. As Michael Ruse commented, e.g., “Popper has been no more successful than others in making traditional evolutionary epistemology plausible. The growth of science is not genuinely Darwinian” (Ruse 1986, p. 65). Nicholas Rescher also advocates a differentiated view on the (restricted) analogy between scientific progress and biological evolution. According to him it would be just absurd to contend that the development of, e.g., mathematics or theoretical physics “is somehow an evolutionary requisite” (Rescher 2000, p. 138). At the same time, however, it cannot be denied: “that the capacities and abilities that make such enterprises possible are evolutionarily advantageous—that evolution equips us with a reserve capacity that makes them possible as a side-benefit” (Rescher 2000, p. 138).

The project of natural science certainly is not a necessary, or strictly deterministic, or analytic consequence of (our) biological evolution. But phylegenetic evolution ‘has endowed us’ with a body of skills which are the constitutive basis for our abilities necessary for occupying our evolutionary niche as well as for developing science (Rescher 2000, p. 139). 6. RESCHER’S METHODOLOGICAL PRAGMATISM The Convergence View of Scientific Progress and its Flaws For Nicholas Rescher, it is indisputable that any conception of scientific progress as a convergent process will be obsolete as soon as conceptual innovation becomes unavoidable to grasp the new phenomena which arise from ongoing technical escalation. That is the case because our questions as well as our answers are constituted and at the same time limited by the

376

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

cognitive and technological state of the art. Obviously, such conceptual constructionist innovations necessarily lead to “an ongoing wholesale revision of ‘established fact’” (Rescher 1996a, p. 59–60). In this sense, scientific progress cannot take place (solely) within Kuhn’s pattern of ‘normal science’, i.e., it must be more than only working out and corroborating established theories to higher degrees of accuracy (i.e., to a few more decimal places) by providing ever more precise empirical data. On the contrary: “[s]ignificant scientific progress is genuinely revolutionary in involving a fundamental change of mind about how things happen in the world. Progress of this caliber is generally a matter not of adding further facts […] It is, rather, a matter of changing the very framework itself. And this fact blocks the theory of convergence” (Rescher 1996a, p. 60).

Moreover, according to Nicholas Rescher, scientific progress is also not a divergent process: “Later is not lesser [ - and not more]; it is just different. Theoretical progress in [natural] science is essentially a matter of substantiation” (Rescher 1996a, p. 62).

In other words, scientific enquiry and especially its technological capacities allow for different conceptions of the objects and law hypotheses of science so that we arrive at “substantially different overall” stories in the course of time (Rescher 1996a, p. 62). Since every major successive stage in the development of science brings forth innovations which are— roughly—of similar interest and importance, the convergence view has to be rejected, i.e., it is not corroborated by factual history of science (Rescher 1996a, p. 62). Of course, while convergentism would imply that ‘later is lesser’, only a non-convergence view is compatible with the problematic fact and serious economic limitation of the “ongoing cost escalation of scientific progress” in natural science (Rescher 1996a, pp. 62, 63). For Rescher it is not possible to demonstrate that the succession of scientific theories converges upon truth.15 Charles Sanders Peirce’s idea of convergent approximation (upon truth), which is based on the assumption that successive changes of our theories are of ever-diminishing significance, expresses a purely theoretical possibility, “but neither historical ex-

377

RESCHER STUDIES

perience nor considerations of general principle provide reason to think that it is a real possibility” (Rescher 1996a, pp. 57–58). Nevertheless, he holds to the opinion, which has been prominently advocated by Peirce, that (natural) science is characterized by a process of self-correction. By the thesis of self-correctedness of science it is meant that in the course of history (i.e., in the long run, or at ‘infinity’) false propositions will be identified and subsequently replaced.16 While Nicholas Rescher concedes the difficulty, if not impossibility, of convincingly proving the self-correctedness of hypothetico-deductive theories, he nevertheless intended to develop a plausible version of Peirce’s self-correction thesis. According to Rescher, Peirce tried to demonstrate that the hypotheticodeductive method of science including its self-correctedness is based on quantitative induction (Rescher 1978b, pp. 1–17). Thereby, however, Rescher maintained that it is not theories that are subject to self-correction, but standards and procedures according to which they are evaluated. In this sense scientific progress is marked by: “increased refinement of the procedures and methods used in the acceptability-screening processes for truth-candidates” (Rescher 1977, p. 175).

In other words, it is Rescher’s conviction that scientific progress (in the natural, or empirical sciences) is achieved, if novel evaluative standards and procedures are developed and implemented that (i) account for the performance of the old ones, and (ii) add new methodological virtues (Rescher 1977, p. 175). In order to gauge the ‘strength’ of evaluative standards and procedures it is indispensible to consider “internal factors” as well as “external factors”. Relevant ‘internals’ are, according to Rescher, coherence, selfsustainability and correctability of the method by feedback from prior variants. Additionally, methodically developed theories must exhibit properties of systematicity, comprehensiveness and simplicity. The important ‘external’ condition, which is rather specific for Rescher’s philosophy of science, is that the method guarantees theories, models and laws which increase problem-solving effectiveness and control: “We thus envisage a two-fold criteriology of cognitive adequacy (and ergo of scientific progress) namely the internal factor of systematisation and the external factor of pragmatic efficacy” (Rescher 1977, p. 177).

378

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

In the (wide) sense of a “Copernican inversion” this conception of scientific progress: “proposes that we do not judge a method of inquiry by the truth of its results, but rather judge the claims to truth of the results in terms of the merit of the method that produces them (assessing this merit by both internal [coherentist] and external [pragmatist] standards)” (Rescher 1977, p. 179).

Of course, according to this view ‘(absolute) truth’ is not acheivable, or available in (natural, empirical) science. However, there is a relevant concept of comparative, gradualized truthfulness which amounts to the thesis that “the application of increasingly more meritious evaluative standards and procedures” guarantees “the presumption that later theories are ‘truer’ than earlier ones” (Losee 2004, p. 106). In this sense, greater truthfulness is “a matter of increasingly more firmly rationally warranted presumption” (Rescher 1977, p. 183): “[W]e thus envisage a complex criterion of merit which stresses the technological/productive side of inquiry in natural science” (Rescher 1977, p. 177).

In other words, Rescher here factually replaces the metaphysical concept(s) of (closeness to) (absolute) truth by the empirio-pragmatic concept of worthness of acceptance, or reliability which basically is ‘measured’ through the properties of increasing capacities of problem-solution and of control of the scientific enterprise as well as of nature. 7. THE ACCRETIONAL MODEL OF SCIENTIFIC PROGRESS AND ITS FLAWS Nicholas Rescher identifies as the core characteristic of the accretional view of the progress of (natural) science “that each successive accretion inevitably makes a relatively smaller contribution to what has already come to hand.” Accordingly, scientific progress comes along by “driving questions down to a lesser and lesser magnitude, continually decreasing their inherent significance.” As Rescher points out, this is—more or less— synonymous to Charles S. Peirce’s “vision of ultimate convergence in scientific inquiry”, and to Thomas Kuhn’s pattern of ‘normal science’ (Rescher 1978a, p. 47).

379

RESCHER STUDIES

Rescher has listed a number of flaws of this ‘accretional model of convergent normality’ which is based on: “two gravely mistaken ideas: (1) that the progress of science proceeds by way of cumulative accretion (like the growth of a coral reef), and (2) that the magnitude of these additions is steadily decreasing” (Rescher 1978a, p. 48).

It is typical for significant steps of scientific progress that new findings are not coherent with all well-established ‘facts’, and that they inevitably challenge revisions of the conceptual and theoretical framework itself— without which there would be no “[s]ubstantial headway”. Of course: “[i]t will not serve to take the preservationist stance that the old views were acceptable as far as they went and merely need supplementation— significant scientific progress is genuinely revolutionary in that there is a fundamental change of mind as to how things happen in the world” (Rescher 1978a, p. 48).

So, science is moving “through an endless sequence of fundamental revisions” (Rescher 1978a, p. 50). 8. SCIENTIFIC RATIONALITY AS COGNITIVE ECONOMY Expectedly, Rescher does not hold a naïve, or scientistic realism, or a certistic empirism, but votes for a pragmatist model character of (natural) science including the hypotheticity of ‘laws of nature’. In other words, fallible is our data-base as well as the adequacy of our theoretical approximation to it (Rescher 1996a, p. 66; cf. Rescher 1978a, 246). Consequently, concerning scientific matters the declared fallibilist Rescher refrains from the systematic use of the conception of a “final and definitive truth”, and declares it “an unrealizable dream” (Rescher 1996a, p. 68–69). He holds it that “a gradual convergence toward a definitively true account of nature at the level of scientific theorizing” “is very implausible” (Rescher 1996a, p. 51). General epistemological fallibilism has several further implications, of course: (i) the demarcation between science and non-science is dynamical and hardly predictable, and in that sense vague; (ii) the science of tomorrow may resolve issues that today’s science qualifies as intractable; (iii)

380

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

tomorrow’s science may endorse principles that are rejected by contemporary scientists. With the words of Nicholas Rescher: “[I]t is infinitely risky to speak of this or that explanatory resource (action at a distance, stochastic processes, mesmerism, etc.) as inherently unscientific. Even if X lies outside the range of science as we nowadays construe it, it by no means follows that X lies outside science as such. We must recognize the commonplace phenomenon that the science of the day almost always manages to do what the science of an earlier day deemed infeasible to the point of absurdity (split the atom, abolish parity, etc.). With natural science, the substance of the future inevitably lies beyond our present grasp” (Rescher 1996a, p. 77).

In the light of his epistemological pragmatism and fallibilism, Rescher finally conceives ‘truth’ as “the best estimate that is actually obtainable here and now” which is the only realizable thing for finite beings without access to “certain knowledge” arising from “complete and perfected information” (Rescher 2000, p. 12). Quite obviously, within this approach: “rationality has an ineliminable economic dimension since the effective use of limited resources is, after all, a crucial aspect of rationality” (Rescher 2000, p. 12).

The same is true for “[c]ost-effectiveness—the proper co-ordination of costs and benefits in the pursuit of our ends” (Rescher 2000, p. 13). In other words, according to Rescher rationality and epistemic economy are “inextricably interconnected” in the sense that “[r]ational inquiry is a matter of epistemic optimization, of achieving the best overall balance of cognitive benefits relative to cognitive costs” (Rescher 2000, p. 14). Within that pragmatist approach the sub-properties of cognitive systematicity— Rescher here mentions: simplicity, regularity, coherence, and conformity etc.—represent “principles of economy of operation” (Rescher 2000, 16). For Rescher, theorizing in natural science is generally based on inductive generalization from the observed data.17 Induction is a “rational process of enquiry” that “constructs the simplest, most economical cognitive structures”, “the simplest pattern of regularity”, “the least complex theory structure”, “that can adequately accommodate our data” (Rescher 1996a, pp. 52–53). In other words:

381

RESCHER STUDIES

“Induction is no more than a search for cognitive order in the resolution of our questions, and our processes of inductive inquiry into nature are geared to reveal orderliness if it is there” (Rescher 2000, p. 15).

Rescher also puts forward a list of “established principles of inductive systematization: uniformity, simplicity, harmony, and such principles that implement the general idea of cognitive economy”, such as “consilience [Whewell], mutual interconnection, and systemic enmeshment” (Rescher 1996a, p. 53). Thereby, Rescher explicitly declares »simplicity« as the “key principle” (which many scholars qualify as vague and imprecise) while the key injunction is that of “cognitive economy” (Rescher 1996a, p. 53). In coherence with his pragmatist epistemology, Rescher also advocates a conception if instrumentalistic simplicity, i.e., there is no ontologically relevant principle of simplicity in nature, and accordingly, at first glance the world may always appear as complex. In that situation, a principle of complexity management reads: Keep your conceptions, models and theories simple as far as possible. “Only introduce as much complexity as you really need for your scientific purposes of description, explanation, prediction, and control” (Rescher 2000, p. 11). Of course, it goes without saying that “we favor uniformity, analogy, simplicity, and the like exactly because they ease our cognitive labour” (Rescher 2000, p. 11). At the same time Rescher reminds us that “[i]t would be naïve—and quite wrong—to think that the course of scientific progress provides a world-picture of increasing simplicity” (Rescher 2000, p. 10). Rather, cognitive evolution is characterized by ongoing exfoliation of diversity and complexity because of recursive nonlinear dynamic processes (‘similar’ to organic evolution. For Rescher, the claim of “the ontological simplicity of the real is somewhere between hyperbolic and absurd. Instead, our inductive practice insists upon the simplest issue-resolving answer—the simplest resolution that meets the conditions of the problem. And we take this line not because we know a priori that this simplest resolution will prove to be correct. […] Rather, we adopt this answer—provisionally at least—just exactly because it is the least cumbersome and most economical way of providing a question resolution that does justice to the facts and demands of the situation” (Rescher 2000, p. 10).

382

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

9. PRELIMINARY CONCLUSIONS Undoubtedly, in the history of (natural) science we find examples of scientific progress by accretion, or incorporation as well as by revolution, or overthrow. Obviously, there is a type of scientific progress during Kuhn’s “normal science” achieved by accretion and incorporation, while another type of progress is instantiated by Kuhn’s “scientific revolutions” (or something similar to them). Kuhn also held correctly that there is no ‘algorithm’ available for determining and predicting whether a certain theory-replacement is progressive, or not. In other words, we do not know the necessary and sufficient conditions to be fulfilled by such a replacement process. Thus, it seems fair to state that the numerous attempts to establish necessary or sufficient conditions of ‘progressive incorporation’ or ‘progressive overthrow’ “have been largely unsuccessful” (Losee 2004, p. 156), i.e., we do not have an exhaustive criteriological characterization of the concept of scientific progress.18 This can be clearly recognized since for almost every proposed condition for scientific progress countercases can be identified from the history of science. In conclusion, we must admit that the corresponding condition is not necessary, or not sufficient, or neither necessary nor sufficient. This is right of consilience, undesigned scope, incorporation-with-corroborated-excess-content, asymptotic agreement of calculations, the testimony of ‘crucial experiments’, and the resolution of anomalies (see also Losee 2004). Despite the negative result concerning most of the proposed criteria for scientific progress (1) scientists do practically base judgments on such criteria, and (2) such criteria are pragmatically certainly helpful. In other words, questions like the following ones are not irrelevant: 1. Has consilience been achieved within a domain? 2. Are the anomalies that beset T1 overcome by T2? 3. Is the replacement of T1 by T2 accompanied by a display of undesigned scope? 4. Do calculations derived from T2 agree asymptotically with those of its (partially) successful predecessor T1?

383

RESCHER STUDIES

5. Are the theories under discussion consistent, coherent, systematically structured, simple, mathematized? 6. Do they exhibit novel predictions? 7. Do they contribute to pragmatic efficacy and control? As soon as we have decided which developments in science are progressive we may ask (1) what kind of property is it that is increased when progress is at stake, and (2) by which mechanism is this progressive development generated? Any list of quality criteria for convincing scientific explanations (presumably) is neither exhaustive nor can it be brought into a unique hierarchy. Generally, not all criteria just mentioned above have to be or can be applied (rigorously) in any concrete research situation. Especially, these criteria are interpretable, normic, and fallible. Therefore, often complicated situations of competition arise between different criteria, which in most cases are only decidable with difficulties, or not uniquely. In any case, such situations are not explicable (monotonically) deductively, or algorithmically. Nevertheless, the above criteria mark important (normic) standards of scientific rationality, i.e., elements of a general scientific methodology, and criteria of evaluation for the (quasi-abductive) conclusion to the most reliable explanation—in short: for dealing with (certain and most important aspects of) science’s (and nature’s) constitutional, organizational, hierarchical, operational, and nomic complexity. Finally, it goes without saying that in scientific research these types of complexity are generally also intimately interwoven with problems of descriptive, generative, computational, and taxonomical complexity (Rescher 1998, p. 9; cf. Leiber 2007, p. 197). Given the set of obviously tensional quality-criteria for scientific theories, what seems to be needed is a weighting of their relative importance. However, there has been no general agreement among methodologists in the philosophy of science on such weighting. 10. NON-INTELLECTUAL LIMITS OF SCIENTIFIC PROGRESS: THE LAW OF LOGARITHMIC RETURNS OF RESEARCH Much energy in theory of science was devoted to questions of scientific reasoning. But however philosophers interpreted the scientific modus operandi intellectus, and however controversially and inconclusively they interpreted it, scientists successfully paced their way to new explanations, detected new laws of nature or refuted flawful assumptions. In contrast to that, much less attention by theorists of science was paid to another crucial

384

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

factor of scientific progress, i.e., technological power of squeezing out nature. And investment in that power is inseparably connected with an economic aspect, a cognitive profit-and-loss consideration. Rescher turned out to be one of the rare philosophers of science who examined that aspect, among other aspects, of a ‘cognitive economy’. On the History and the Future of Scientific Progress Nicholas Rescher’s “main aim in writing” Scientific Progress was “to vindicate [...] the Leibnizian vision of an open and hopeful future for scientific inquiry.” He wanted to argue against the pessimistic end-to-progress speculations with its “tragic implications for the human condition” (Rescher 1978a, p. xi). To that end, he voted for the conception of “scientific deceleration” while rejecting the idea of “scientific stoppage”, because in Rescher’s view: “the truth of the matter appears to lie in this middle ground. A just view of the future of science will accordingly be neither over-sanguine nor apocalyptic, but realistic in its conception of the long, hard haul that lies ahead” (Rescher 1978a, p. xii).

Rescher’s Scientific Progress is “substantially oriented towards the economic dimension of natural science—a scrutiny of the productive aspect of inquiry in this sphere in terms of inputs and outputs” (ibid.). The two most important inspirers of that approach are Gottfried Wilhelm Leibniz and Charles Sanders Peirce. According to Rescher, Leibniz was the first philosopher who understood science as a productive, in principal businesslike enterprise. Leibniz’s projects of academy, of encyclopedism, and of a scientia universalis give rich evidence of his efforts to organize and systematize scientific work. Obviously, Leibniz fully grasped the prospects and implications of Francis Bacon’s prophetic visions in methodology as well as institutional organisation of science and its (future) progress (ibid.). Beyond Bacon’s views, Leibniz also had an understanding of the importance of public interest and governmental sponsorship for future science (Rescher 1978a, p. 6). The philosopher who first concerned himself with the economic aspects of scientific research was Peirce. He suggested the founding of a new ‘economy of research’ which adopted a central place in his epistemology (Rescher 1976). Peirce holds that scientific inquiry must also be viewed in

385

RESCHER STUDIES

the light of “the concepts and categories of technical economics” (Rescher 1978a, p. xiii). Contrary to many earlier writers on the subject of (the limits of) scientific progress, Rescher stresses that the most important limits of scientific inquiry reside in the—physical or economic—limitations “imposed upon us by the nature of the world”, and do not (so much) result from “human failings and deficiencies (in intellect, learning-power, memory, imagination, will-power, etc.)” (Rescher 1978a, p. 2). Rescher’s approach in Scientific Progress proposed five major alternatives for conceiving the long-range future of scientific discovery (see Table 1), which seem to fulfil certain significant merits: they are coherent with the available empirical information, and thus at least plausible; (ii) they seem to be sufficiently precise in order to be testable. ____________________________________________________________ Table 1 Name of Conception

Horizons of Scientific Innovations

I. Nature Exhaustion

limited by the inherent finitude of nature and ultimately attainable II. Nature Saturation limited by the inherent fini(Asymptotic Completion) tude of nature but only approachable asymptotically III. Capacity Exhaustion limited by the inherent finitude of man and ultimately attainable IV. Capacity Saturation limited by the inherent fini(Asymptotic Incapacitation) tude of man but only approachable asymptotically V. Unlimited Horizons unlimited (Potentially Unending Pro- (the traditional view supgress) ported by many of the greatest natural scientists) Table 1: Conceptions of the Future of Science (Rescher 1978a, p. 7) ___________________________________________________________

386

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

From Rescher’s point of view, the Potential Limitlessness of Science does it not make necessary “to maintain the structural infinitude of nature in order to underwrite the prospect of unending scientific progress” (Rescher 1978a, p. 40). This is so because “the prime task of science lies in discovering the laws of nature. And even a mechanism of finitely complex make-up can have endlessly complex laws of operation. The workings of structurally finite and indeed simple system can yet exhibit a finite intricacy in functional complexity, exhibiting this limitless complexity at the operational or functional—rather than at the spatio-structural or compositional—level. And this sort of operational complexity would be quite enough to underwrite the limitlessness of prospective scientific discovery regarding the behavioral laws of even a system whose physical make-up is extremely simple. (For a useful analogy, think of the theory of chess.) And so structural infinitude becomes irrelevant” (Rescher 1978a, p. 40–41): “After all, the “completeness” of science is a function of our interests—a largely homocentric issue. It is a matter of what sort of things we want to have explained and how our curiosity is oriented” (ibid.). And: “There is no adequate basis for the view that the search for greater “depth” in our understanding need ever terminate at a logical rock-bottom. Nor need it ever abort for lack of value or interest” (ibid.).

According to Rescher, theoretically “the prospect of […] ongoing ‘scientific revolutions’ is potentially unending, and there is no warrant for a theory of convergence that sees the innovations of theorizing science as being of constantly diminishing scope; with discoveries in natural science, later does not mean lesser” (Rescher 1996a, p. 51). “The thesis of endless scientific progress is thus perfectly compatible with the view that every question that can be asked at every stage is going to be answered at some future stage: it does not commit one to the idea that there are any unanswerable questions placed altogether beyond the limits of possible resolution. All that is needed for unending progressiveness is the very real phenomenon that in the course of answering old questions we constantly come to pose new ones” (Rescher 1978a, p. 245). Rescher calls this phenomenon the “Kant proliferation effect”, or the “conservation-law for scientific problems” (ibid., pp. 245, 246).

387

RESCHER STUDIES

The Historic Acceleration of Science: Adam’s Law of Exponential Growth and the Logarithmic Retardation Henry Brooks Adams (1838–1918) was one of the first who made explicit in detail the increasing tempo of the advancement of science. Accordingly, Adams conceptualized a “law of acceleration” to which he gave a mathematical formulation, which says that the production of scientific work(s) is adequately approximated by an exponential growth-rate (Rescher 1978a, p. 54). In other words, “the percentage-rate of the per annum increase in the levels of scientific effort and quantitative productivity remains constant with the passage of time” (Rescher 1978a, p. 55). In the sixties and seventies of the past century, Derek deSolla Price in his widespread “Little Science, Big Science” (1963) illuminated the consequences of that exponential growth for the structure and organization of science. Following Price, numerous case studies exhibited different theoretical and empirical aspects (and the existence) of the exponential growth of science. From these investigations a “picture of science as an enormous industry” arose. This was a decisive factor because of which one must view science “in essentially economic terms, as a production process with an input and an output—as a productive enterprise with costs (in terms primarily of manpower and resources) and with returns (in terms of an output in the first instance of ‘literature’, but more significantly of ‘findings’). This ‘science-industry’ has a far-flung network of training centers (schools, colleges, universities), and of production centers (laboratories and research-institutes)” (Rescher 1978a, p. 65–66).

Data at least up to the early 1970ies, according to Rescher and others, indicate exponential growth of input (human and material) as well as output (e.g., publications). Not in the same way as this kind of input and output expands, however, the amount of first-rate outcomes which grow, according to Rescher, only in a linear manner. Of crucial importance here is the "Law of Logarithmic Returns", i.e., the cumulative output of first-rate findings is proportional to the logarithm of cumulative investment. First-rate findings grow in a linear manner as long as investment in science grows exponentially. With the deceleration of investment in science, which began, according to Rescher (and other observers) in the later seventies of the past century, a "logarithmic retardation" of first-rate findings is only a

388

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

logical consequence. It is the first-rate findings which Rescher identifies with "scientific progress", not the inflating mass of standard contributions. Similar to deSolla Price, Rescher posed a question of much interest for future prospects of science: “Is the exponential growth of science itself a ‘law of nature’, as it were, or is it descriptive of the transitory constitution of one particular era?” (Rescher 1978a, p. 69). Obviously, and also in agreement with writings of deSolla Price, Rescher holds it that the assumption would be “grotesque” that “the human and material resources committed to scientific research” or “the literature that is its output” would follow an “unendingly continued exponential growth” (Rescher 1978a, p. 70). Rescher gives no cognitive criteria for the definition of first-rate findings, and insofar the critique of I. Niiniluoto is legitimate (1980, p. 451). Rescher only assumes that there in fact are different levels of significance, but concedes that there will be different approaches to indicate or characterize those different levels. According to Rescher the scientific situation is the opposite of what we know as mass production in industry: Here the principle is not valid; a certain amount of investment into production leads to additional outcomes in a proportional way or even in a more than proportional way. Rescher “does not deny that further “revolutions” (=discoveries of the very first importance) lie ahead—quite to the contrary”, he “maintains that they do, and indeed do so in endless numbers” (Rescher 1978a, p. 122). However, “such revolutions will become less and less frequent […] purely and simply because revolutions are becoming increasingly expensive to mount” (Rescher 1978a, p. 122). The crucial role of technology and applied science “Progress in natural science is a matter of dialogue or debate between theoreticians and experimentalists” (Rescher 1996a, p. 54); and “in natural science, imperfect physical control is bound to mean imperfect cognitive control” (Rescher 1996a, p. 58). “What we find in investigating nature always in some degree reflects the character of our technology of observation; what we can detect, or find, in nature is always something that depends on the mechanisms by which we search” (Rescher 1996a, p. 55–56). Science of our days is more technology-hungry than ever. “To all appearances, the technological demands of modern science render it emphatically capital-intensive. A fact to which any theory of scientific progress will also have to give close scrutiny” (Rescher 1978a, p. 69).

389

RESCHER STUDIES

“Since Bacon’s day, this idea of control over nature has been recognized as the keynote of scientific progress” (Rescher 1978a, p. 191) - and this control includes, according to Rescher, what we can call observational control. “In particular, physical limits are always amplified by technologiceconomic limitations […]” (Rescher 1978a, p. 237): “The economic limits to scientific progress always inhere in a physical basis: increasing costs of inquiry are invariably associated with physical resistance to large-scale interactions” (Rescher 1978a, p. 237).

Concerning the main stream of philosophy of scientific progress, Rescher critically writes: “The traditional theories of scientific progress join in stressing the capacity of the ‘improved’ theories to accommodate new facts. Agreeing with this emphasis on ‘new facts’, we must, however, recognize two distinct routes to this destination: the predictive (via theory) and the productive (via technology). And it would appear proper to allow both of these routes, the predictive and the productive, to count. To correct the overly literary bias of the purely theoretical stress of traditional philosophy of science requires a more ample recognition of the role of technology-cum-production. To say this is not, of course, to deny for one moment that the two (theory and technology) stand in a symbiotic and mutually supportive relationship in scientific inquiry. But the crucial fact is that the effectiveness of the technological instrumentalities of praxis can clearly be assessed in the absence of any invocation of the cognitive content of the body of theory brought to bear in their devising” (Rescher 1978a, p. 190).

Progress of applied science and technology, is, according to Rescher, comparatively easy to assess—in contrast to assessing theoretical progress. “When the ‘external’ element of control over nature is given its due prominence, the substantiation of imputations of scientific progress becomes a more manageable project than it could ever possibly be on an ‘internal’, content-oriented basis. The progressiveness of science appears most strikingly and decisively on its technological side, marked by an ever widening ability to operate at further and further removes in parametric remoteness or complexity of detail, and by an ever expanding predictive and physical control over nature. […] The old Hobbesian conception of scientia propter potentiam provides a perfectly workable basis for taking

390

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

the expanding horizons of technological capacity as an index of scientific progress” (Rescher 1978a, p. 192). Eventually, as a consequence of the logarithmic retardation, Rescher prognosticates: “In the future, work on analytic problems is bound to become increasingly prominent in the setting of a zero-growth world. An ultimate shift in the nature of natural science must be anticipated—away from a predominantly power-intensive, synthetic phase in which ever new regions of ‘parametric space’ are opened up (the sort of scientific progress that has in fact predominated since the industrial revolution), and towards a predominantly complexity-intensive, analytical phase. The central issue of this latter-day phase will be the increasingly intricate relationship obtaining in already accessible regions, but only brought to light through increasing sophistication, relying upon increasingly refined mathematical methods in data-utilization and theory-formation” (Rescher 1978a, p. 240–241).

Implications of Logarithmic Retardation According to Rescher’s analysis, scientific progress will not terminate, i.e., there will be no literal end to the history of scientific enquiry. There will, however, appear an era of slow-down and almost zero-growth of science instead of of exponential growth (and linear progress) (Rescher 1978a, p. 255). Nevertheless, our future prospects in matters of new achievements in (natural) science seem to be in principle without limits. Consequently, Rescher interprets the logarithmic retardation view of (future) scientific progress as a clear reason for optimism—because it combines the realistic prospect of “unending” scientific progress with its factual cognitive manageability by us limited beings (Rescher 1978a, p. 259). Rescher does not see his theory of logarithmic retardation of scientific progress carry over from the natural sciences to the social, or to the formal sciences (i.e., mathematics and logic), or to the humanities. Basically, the reasons are the following: the subject of the social sciences is the organizational structure of human activity, and the “cognitive exploration” of these offer the “prospect of unending innovation”; on the other hand, the purely formal (intellectual) sciences are free from the problems of “technological escalation”; finally, the humanities’s methodology is hermeneutic, or interpretative, and therefore there is “no end to innovation here as long as there is no end to the formation of intellectual perspectives that afford the com-

391

RESCHER STUDIES

binations and the emphases through which new humanistic interpretations always become possible” (Rescher 1978a, p. 261). Of course, it can be expected that a slowing-down of scientific progress will cause a corresponding deceleration of the rate of technological progress (Rescher 1978a, p. 262). However, also technological progress in itself might be governed by a Law of Logarithmic Returns—completely independent of scientific progress. Here an analysis has to be considered which exhibits Rescher-analogous lines of thought concerning returns of applied sciences: the “Law of Decreasing Marginal Returns of Technology” (Giarini & Loubergé 1978). Theoretical and Empirical Evidence A Law of Logarithmic Returns has to be empirically valid. It can fit to reality or reality can overthrow its empirical implications. Beyond Rescher's own empirical examples (presented in Scientific Progress), appropriate investigations are rare.19 Of course, the examination and comparative evaluation of large bodies of scientific work is cumbersome, and needs comprehensive competence, not to mention many methodological traps. However, a huge mass of related empirical work was done in certain areas of scientometrics. One of the most stable and ubiquitously confirmed scientometric regularities is called Lotka's Law. This regularity describes a quite asymmetric productivity distribution with a small nucleus of extremely prolific individuals in a scientific discipline and a “long tail” of far less productive contributors as characteristic (a phenomenon related to the well-known Pareto distribution). The uneven pattern of contributorship has been revealed in any scientific discipline or speciality examined. It is a direct implication of Lotka’s Law that the different levels of importance and quality grow at different rates (see numerical examples at Rescher 1978a, pp. 104–108). Empirical evidence is available, e.g., for the literature of formal logic, being evaluated by the outstanding logician Alonzo Church. The growth rates of different quality—or significance—levels obviously differ in agreement with Lotka’s Law and in agreement with a Law of Logarithmic Retardation of first-rate innovations (Wagner-Döbler 2001, p. 430f.). Further examples and empirical case studies confirm that pattern (Wagner-Döbler 2001). Lotka’s Law turned out to be valid, however, not only in natural sciences, but also in social sciences and humanities, and in mathematics. This undermines the argument that a logarithmic retardation goes back to de-

392

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

creasing marginal returns of costly technology used by science, because in most humanities or in mathematics capital-intensive technology seems to play no role at all. It supports, however, a general Law of Logarithmic Returns. 11. PHILOSOPHY OF SCIENCE AND SCIENCE OF SCIENCE: A CONCLUDING REMARK The relationship between philosophy of science, on the one hand, and science of science, on the other hand, even of the last decades resembles in some respects the relationship between philosophy of nature, on the one hand, and physics and other natural sciences of 18th and 19th century, on the other hand. A clear demarcation line was not visible, and from a modern point of view philosophy went quite too far into areas which today are definitely in the hands of natural scientists. White areas of factual knowledge led philosophers into the temptation of speculation. Natural science blossomed through interaction between new experiments, empirical knowledge, and mathematical thinking, with certain distance to philosophical speculation. Correspondingly, philosophy of science grew in certain distance to philosophical self-interpretations of natural scientists. Scientific methodology, as a cognitive undertaking, can be analyzed from a purely empirical and naturalistic angle. On a micro level of such an approach, this is done, e.g., in cognitive psychology or evolutionary epistemology. In addition, Rescher illuminated another non-normative, factual aspect of scientific progress, the aspect of cognitive returns and investment ignored by most philosophers of science. Rescher studied these phenomena on a micro-level (e.g., in his Priceless Knowledge), but also on a macro-level (e.g., in his Scientific Progress), paving the way to a fruitful interaction between ‘naturalistic’ and theoretical philosophy of science.

REFERENCES Cohen, I. Bernard, Revolution in Science (Cambridge (MA): Harvard University Press, 1985).

393

RESCHER STUDIES

Cohen, L. Jonathan, “Is the progress of science evolutionary?” British Journal for the Philosophy of Science, vol. 24 (1973), pp. 41-61 Dilworth, Craig, Scientific Progress: A Study Concerning the Nature of the Relation Between Successive Theories (Dordrecht: Kluwer, 1994). Donovan, Arthur, Larry Laudan & Rachel Laudan, Scrutinizing Science. (Dordrecht: Kluwer, 1988). Doppelt, Gerald, “Relativism and the reticulational model of scientific rationality,” Synthese, vol. 69 (1986), pp. 225-252. Feyerabend, Paul K., “Explanation, reduction and empiricism,” in: Herbert Feigl & G. Maxwell (eds.) Minnesota Studies in the Philosophy of Science. Vol. III. (Minneapolis: University of Minnesota Press, 1962), pp. 28-97. Giarini, O. & H. Loubergé, The Diminishing Returns of Technology (Oxford: Pergamon Press, 1978). Kitcher, Philipp, The Advancement of Science. Science without Legend, Objectivity without Illusions (Oxford: Oxford University Press, 1993). Krohs, Ulrich & Georg Toepfer (Hg.), Philosophie der Biologie. Eine Einführung (Frankfurt am Main: Suhrkamp, 2005). Kuhn, Thomas S., The Structure of Scientific Revolution, 2nd ed. (Chicago: University of Chicago Press, 1970). Kuhn, Thomas S., “What are scientific revolutions?” in Lorenz Krüger, Lorraine J. Daston & Michael Heidelberger (eds.), The Probabilistic Revolution. Vol. 1: Ideas in History. Cambridge (MA): MIT Press (1987), pp. 7-22. Lakatos, Imre, “Falsification and the methodology of scientific research programmes,” in Imre Lakatos & Alan Musgrave (eds.), Criticism and the Growth of Knowledge (Cambridge: Cambridge University Press, 1970), pp. 91-195.

394

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

Lakatos, Imre, “History of science and its rational reconstructions,” in R. Buck & R. S. Cohen (eds.), Boston Studies in the Philosophy of Science, Vol. VIII. (Dordrecht: Reidel, 1971),pp. 91-108. Laudan, Larry, Progress and its Problems. Toward a Theory of Scientific Growth (London: Routledge & Kegan Paul, 1977). Laudan, Larry, Science and Values (Berkeley: University of California Press, 1984). Laudan, Larry, “Progress or rationality? The prospects for a normative naturalism,” American Philosophical Quarterly, vol. 24 (1987), pp. 1931. Laudan, Larry, Science and Relativism. Some Key Controversies in the Philosophy of Science (Chicago: University of Chicago Press, 1990). Leiber, Theodor, Vom mechanistischen Weltbild zur Selbstorganisation des Lebens. Helmholtz’ und Boltzmanns Forschungsprogramme und ihre Bedeutung für Physik, Chemie, Biologie und Philosophie (Freiburg: Alber, 2000). Leiber, Theodor, “Structuring nature’s and science’s complexity: system laws and explanations,” in Theodor Leiber (ed.), Dynamisches Denken und Handeln. Philosophie und Wissenschaft in einer komplexen Welt. Festschrift für Klaus Mainzer zum 60. Geburtstag (Stuttgart: Hirzel, 2007), pp. 193-212. Losee, John, Theories of Scientific Progress. An Introduction (New York: Routledge, 2004). Nagel, Ernest, The Structure of Science (New York: Harcourt, Brace & World, 1961). Nagel, Ernest, “Issues in the logic of reductive explanations,” in his Teleology Revisited (New York: Columbia University Press, 1974).

395

RESCHER STUDIES

Neurath, Otto, “Protocol statements,” in: R.S. Cohen & M. Neurath (eds.), Otto Neurath: Philosophical Papers (Dordrecht: Reidel, 1983), pp. 9199. Niiniluoto, Ilkka,” Scientific progress,” Synthese, vol. 45 (1980), pp. 426462. Perrin, Jean, Atoms (London: Constable1913 [1922]). Popper, Karl, Objective Knowledge (Oxford: Clarendon Press, 1972). Popper, Karl, “The rationality of scientific revolutions, in: Ian Hacking (ed.), Scientific Revolutions (Oxford: Oxford University Press, (1981), pp. 80–106. Price, Derek deSolla, Little science, big science (New York: Columbia University Press, 1963). Puntel, Lorenz Bruno, „Einführung in Nicholas Reschers pragmatische Systemphilosophie,“ in Nicholas Rescher, Die Grenzen der Wissenschaft (Stuttgart: Reclam, (1985), pp. 7-47 Rapp, Friedrich, Fortschritt. Entwicklung und Sinngehalt einer philosophischen Idee (Darmstadt: Wissenschaftliche. Buchgesellschaft, 1992). Rescher, Nicholas, “Peirce and the economy of research,” Philosophy of Science, vol. 43 (1976), pp. 71-98. Rescher, Nicholas, Methodological Pragmatism (Oxford: Oxford University Press, 1977). Rescher, Nicholas, Scientific Progress. A Philosophical Essay on the Economics of Research in Natural Science (Oxford: Basil Blackwell, 1978). [German translation: Wissenschaftlicher Fortschritt. Eine Studie über die Ökonomie der Forschung (Berlin: de Gruyter, 1982).] Rescher, Nicholas, Peirce’s Philosophy of Science (Notre Dame: University of Notre Dame Press, 1978).

396

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

Rescher, Nicholas, Die Grenzen der Wissenschaft. Stuttgart: Reclam, 1985), [orig.: The Limits of Science (Berkeley: University of California Press, 1984)]. Rescher, Nicholas, Priceless Knowledge? Natural Science in Economic Perspective (Lanham: Rowman & Littlefield, 1996). Rescher, Nicholas, “Theoretische Grenzen der Naturerkenntnis,“ in his Studien zur naturwissenschaftlichen Erkenntnislehre ed. by Axel Wüstehube (Würzburg: Königshausen & Neumann, 1996), pp. 143– 163. Rescher, Nicholas, Complexity. A Philosophical Overview (New Brunswick: Transaction Publishers, 1998). Rescher, Nicholas, Nature and Understanding. The Metaphysics and Method of Science (Oxford: Clarendon Press, 2000). Ruse, Michael, Taking Darwin Seriously (Oxford: Blackwell, 1986). Ruse, Michael, Evolutionary Naturalism (London: Routledge, 1995). Schurz, Gerhard, „Das Problem der Induktion,“ in: Herbert Keuth (Hg.) Karl Popper. Logik der Forschung (Berlin: Akademie Verlag, 1998), pp. 25-40. Toulmin, Stephen, Foresight and Understanding (New York: Harper Torchbooks, 1961). Wagner-Döbler, Roland, “Rescher’s principle of decreasing marginal returns of scientific research,” Scientometrics vol. 50 (2001), pp. 419-436 Whewell, William, History of the Inductive Sciences (London: Cass, 1857, 3rd edn. 1967 (orig. 1837) Whewell, William, Philosophy of the Inductive Sciences (London: Cass, 1847, 2nd edn. 1967). Zilsel, Edgar, “The genesis of the concept of scientific progress,” The Journal of the History of Ideas, vol. 6 (1945), pp. 325–349.

397

RESCHER STUDIES

NOTES 1

Neurath [1932] 1983, p. 92

2

Rescher 1996a, pp. 65, 70.

3

For some of the following considerations see also (Losee 2004, pp. 1–3).

4

Clearly, mathematized models and theories make possible quantitative forecasts and explanations which constitute high-level criteria of scientific quality.

5

The property of continuity of mistake and correction allows, e.g., for a controllable estimation of error as well as degrees of approximation.

6

A (novel) theory normally should have at least the same observable consequences as its predecessor.

7

This term is borrowed from (Losee 2004, p. 2).

8

Such an approach may or may not contain normative postulates as to how scientific investigation ought to be conducted. Within more descriptive approaches to ‘scientific progress’ it is intended to discover and display those repeated patterns deemed progressive by the practitioners of science (in historical or contemporary perspective). ‘Purely descriptive’ and ‘purely normative’ approaches are extremely idealized ideas, to say the least.

9

As usual, ‘simplicity’ is not given a clear definition. According to Whewell, a theory might be called ‘simple(r)’ if its explanatory power increases without introducing novel assumptions. But how can one theory differ from another one without introducing some novel assumption? In contrast to the third, the first two progressiveness-criteria of Whewell are accepted by most philosophers of science.

10

This implies that the history of science and the psychology of scientific discovery are almost irrelevant, or unimportant for the philosophy of science. Such a position was advocated, e.g., by Hans Reichenbach, the logical empiricists, and Karl Popper.

11

On the other hand, Nagel’s theory of reduction is applicable only to theories which can be formalized as deductive systems.

12

Thomas Nagel, Dudley Shapere, and Peter Achinstein, amongst others, maintained that Feyerabend had overstated the case for meaning variance, or incommensurability, respectively.

13

Seemingly, Feyerabend did not give further reasons for that conviction.

14

Of course, in the strict sense any (scientific) solution to a problem seems to be approximative, even in the ‘hardest’ empirical discipline—there are no nonapproximative solutions.

398

NICHOLAS RESCHER ON SCIENTIFIC PROGRESS: SCIENCE IN THE FACE OF LIMITED COGNITIVE AND TECHNOLOGICAL RESOURCES

NOTES 15

Peirce’s convergence-upon-truth thesis has been criticized by many scholars, among them Peter F. Strawson and Larry Laudan (see Losee 2004, pp. 100–101), and Gerhard Schurz (1998). In summary, the thesis rests upon the unprovable assumption that, in the long run—probably ‘at infinity’—“science achieves all and only true statements about the universe” (Losee 2004, 102). This assumption is, for example, far from obvious in the light of the Duhem-Quine thesis as well as in the light of structural (non-naïve) realism.

16

According to Peirce, it is this self-correcting aspect of the practice of science which sets it apart from other forms of inquiry.

17

Rescher sometimes uses ‘induction’ even synonymous to “the scientific method” (cf. Rescher 2000, p. 11).

18

Remember that we practically use quite a lot of concepts for which that statement is true.

19

For an overview and some own empirical examinations see Wagner-Döbler (2001).

399

Rescher on Dialog Systems, Argumentation, and Burden of Proof Douglas Walton David M. Godden

N

icholas Rescher has succeeded in developing a metaphilosophy (2006) based on a theory of formal disputation that is dialectical in nature (Rescher, 1977). Formal dialog theories of this sort have now become central to the methodology of the rapidly growing interdisciplinary field of argumentation study. Argumentation can be abstractly defined as the interaction of different arguments for and against some conclusion. To evaluate a claim, on the argumentation approach, it is necessary to examine both the plausible arguments for the claim and the plausible arguments against it. Hence the right framework is that of a dialog, where two parties take turns making claims and criticizing them. The advantages of the dialog approach have recently become apparent in the field of computer science, most notably in artificial intelligence. Over the last few years, argumentation has been gaining increasing importance in multi-agent systems, mainly as a vehicle for facilitating rational interaction (interaction that involves the giving and receiving of reasons). Argumentation has provided methods that have led to the development of automated systems for designing, implementing and analyzing sophisticated forms of interaction among rational agents, like defeasible arguments based on expert opinion. Argumentation has already made solid contributions to the practice of multiagent dialogues, including application in domains like legal disputes, business negotiations, labor disputes, scientific inquiry, deliberative democracy, risk analysis, scheduling, and logistics. In this paper, we consider how Rescher dialog systems might be applied to the central problem of how burden of proof shifts back and forth in argumentation. We first consider some problems arising out of attempts to apply formal dialog systems as normative models to philosophical argumentation of the kind studied by Rescher (2006). Carrying this quest for applicability to realistic argumentation even further, we next consider how such a formal dialog system could be applied to everyday conversational argumentation and to legal argumentation. As we will show, the medium between the formal dialog system, and a real case of argumentation where a decision

RESCHER STUDIES

needs to be made about whether a given argument is reasonable or not, is the argumentation scheme. The argumentation scheme represents the form of the argument, but in this paper we consider not only the familiar deductive argument schemes, like modus ponens, but also defeasible argument schemes, like argument from expert opinion. Instances of such defeasible schemes need to be evaluated on a burden of proof basis in a given case (Reed and Walton, 2005). This is done by using a set of appropriate critical questions that matches the scheme that applies to the argument in that case. One of the main problems so far unsolved in argumentation theory concerns how the dialog is managed when a proponent puts an argument forward that fits a scheme and the respondent poses a critical question matching that scheme (Walton, 2007). The asking of a critical question is often said to shift a burden of proof to the proponent’s side, so that if the proponent does not answer the question, her argument fails. On the other hand, asking some kinds of critical questions do not defeat the original argument unless evidence is given to back up the question. So how should the burden shift from one side of a dialog to the other when a critical question appropriate for an argumentation scheme is asked? How should the dialog proceed when an argument that otherwise seems reasonable is challenged by the asking of a critical question? Our study will center around this problem. 1. RESCHER’S METAPHILOSOPHY Rescher’s metaphilosophy (2006) is built around the observation that any philosophical view, once formed, seems plausible in the beginning, but then is subject to qualification, which in turn leads to the drawing of distinctions. The process of building a philosophy is dialectical because philosophy begins with a theory that is formed as an attempt to answer questions that arise out of puzzling inconsistencies of the kind Rescher calls apories. Critical questions are always asked, however, because general rules that are formed as part of the philosophy are constantly confronted with exceptions. Moreover, as a philosophy becomes more comprehensive, inevitably it leads to inconsistencies that need to be resolved. Rescher (chapter 6) offers many historical examples of actual philosophical theories that fit his metaphilosophy very convincingly. He offers a highly persuasive argument that the development of a philosophy fits the model of his theory of dialectic. Philosophical argumentation, on his view, is built around a conflict of opinion formed from two conflicting theses about

402

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

some topic, each of which by itself seems plausible. To respond to the objections of the opposing side, each side introduces a conflict-resolving distinction in its own position, while at the same time raising questions that suggest that the position of the opposed side is implausible in some respect. Through this process, each side produces a more sophisticated formulation of its initial position. The philosophical apories Rescher (2006, chapter 2) describes are collections of plausible philosophical propositions that lead to an inconsistency. One example he gives (p. 20) consists of the following four propositions. (1) A (cognitively) meaningful statement must be verifiable in principle. (2) Claims regarding what obtains in all times and places are not verifiable in principle. (3) Laws of nature characterize processes that obtain in all times and places. (4) Statements that formulate laws of nature are cognitively meaningful. From propositions (2) and (3), it follows that laws of nature are not verifiable in principle. From statements (1) and (4) it follows that statements that formulate laws of nature must be verifiable in principle. Thus collectively, the four propositions composing the apory are inconsistent. This inconsistency is a problem for logical positivists, who wanted to claim that all meaningful statements must be verifiable in principle (statement (1)). Positivists would of course want to include laws of nature in this claim. However, the problem, expressed by statement (2), is that claims regarding what obtains in all times and places are not verifiable principle, along with statement (3), stating that laws of nature characterize processes that obtain in all times and places. This apory represents what was an issue in a real dispute that took place in philosophy. Positivists wanted to maintain the verification principle, as a way of expressing their general viewpoint that only scientific propositions are meaningful because they are verifiable in principle, whereas metaphysical statements are not meaningful because they are not verifiable in principle. The apory above, however, attacks this philosophical viewpoint

403

RESCHER STUDIES

by putting forth the basis of a plausible argument to the effect that laws of nature are not verifiable in principle. This attack raises questions about the tenability of the verifiability principle and raises doubts about whether this principle can be applied in the way the positivists intended. There are many questions that can be raised about how this process of development of a philosophy stemming from an initial apory works. Are there rules governing what sorts of questions can be asked. What sort of answers to them are appropriate? What forms of argument should be used in supporting one’s own philosophical view and attacking the philosophical view of an opposed party? At what point, if any, should such a process of increased sophistication of one’s philosophical theory be terminated? Rescher’s system of dialectics provides a framework for giving some answers to these questions. 2. RESCHER’S SYSTEM OF DIALECTIC Rescher’s basic system of dialectic has several distinctive features that make it different from other formal dialog systems. In the Rescher system, each party attempts to lead the other to violate some rule of the game. This adversarial aspect fits the pattern of the medieval obligation games, where there were two participants who took turns making moves of various sorts, and where the goal of one (called the proponent) was to trap the other (called the respondent) into conceding a contradiction. This kind of dialogue is clearly meant to be asymmetrical, in the sense that the roles of the two participants are different. The proponent has the more active goal of trying to win the dialog by moving the respondent through a series of moves to a move where he concedes a contradiction. The respondent, on the other hand, wins the dialog if he manages to survive for a designated number of moves without becoming trapped into conceding a contradiction. In the Rescher system (1977, p. 4), the proponent opens the disputation, formulates a special proposition called her thesis T, and then produces arguments that support T. The opponent uses challenges and counterassertions to try to rebut the proponent’s moves. In this system, there are three fundamental moves (1977, p. 6). Making the move of categorical assertion ‘!P’ means that the participants who asserted P maintains that P. We’re not told what it means for a participant to maintain some proposition, but what we take it to mean is that the person has not only asserted that proposition but has done so in such a way that he is committed to it.

404

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

Making the move of cautious assertion ‘†P’ means that the person who cautiously asserted P is maintaining that P is the case for all the other party in the dialog has shown. Rescher tells us that the requirement for cautious assertion is that P’s being the case is compatible with everything the other party in the dialogue has said. We take this requirement to state that P must be compatible with all the previous commitments of the other party in the dialog. Making the move of provisoed assertion P/Q means that P obtains, other things being equal, in ordinary circumstances in which Q obtains. Provisoed assertion can be described as a defeasible conditional relationship stating that if Q holds then defeasibly it can be inferred that P holds. This interpretation is confirmed when Rescher tells us (1971, p. 7) that the rule of detachment holds for provisoed assertion: P/Q, Q; therefore P. In Rescher’s system there are two kinds of primary objections that may be made in a dialog against the assertion of a claim C made by one party (Freeman, 2007, p. 26). One is the reply †~C, meaning “~C is the case, for all that you have shown”. The other is the reply ~C/P and †P, meaning “Ceteris paribus, when P is the case, ~C is the case, and P is the case for all you have shown”. The first kind of reply is called cautious denial and the second is called provisoed denial. The provisoed denial is an interesting weak kind of objection against a claim made by the other party. It’s not really a rebuttal, in the sense that it is giving an argument that shows that C is false. Nor is it really a refutation of the claim C. It could be described perhaps as an attack on the claim C, but a sort of hypothetical attack that claims that the proponent has not disproved an argument that could be made against the claim C. It’s certainly a subtle and interesting form of attack, and one wonders how it should affect burden of proof in the dialog between the claimant and the attacker. For example, to adapt a case cited by Freeman (2007, p. 29), suppose the proponent claimed that a witness saw the defendant enter the house, and the opponent replied, “Ceteris paribus, when the witness is lying when he makes a claim C, ~C is the case, and the witness is lying for all you have shown”. This would be a pretty weak objection, if made in a trial. The opponent would likely retort, “You show that the witness is lying. The burden of proof is on you, because the witness took an oath to tell the truth when he agreed to testify”. Such examples make us wonder how Rescher’s two kinds of objections fit into a more standard kind of language of objections and rebuttals, and exactly what the rules should be for deciding allocation of burden of proof with respect to moves like cautious denial and provisoed denial. We suspect that although there are now many formal dialog systems that have been devel-

405

RESCHER STUDIES

oped, to model legal argumentation as well as everyday conversational argumentation, the basic notions of attack, rebuttal, objection and refutation are still not very well defined or understood. One feature of the Rescher system of disputation that is especially important is the asymmetery between the two parties. He characterizes it by saying (1977, p. 20) that the “ultimate burden of proof” is on the proponent as a dialog proceeds. Only the proponent is allowed to make moves that imply a definite commitment, and the proponent’s commitments change during the course of the disputation. The opponent does not take on commitments during the course of the dialog, or does so only in lesser way when he concedes a committed thesis when he subjects it to a distinction (p. 71). The opponent concedes a committed thesis when he is given an opportunity to attack it but fails to take up the opportunity of attacking it. This form of concession, however, appears to be different from the kind of commitment that the proponent takes on when he makes an assertion. Important features to note are the adversarial aspect of the system and the asymmetry between the two participants. The Rescher dialog is set up in such a way that there is meant to be a winner and loser, and what each party has to do in order to win is precisely defined so that any sequence of moves can be judged as winning or losing. Another important feature (p. 4) is that the outcome of a dialog is judged by a determiner, a third party who presides as a referee, and judges the conduct of the dispute. What Rescher (1971, p. 1) sees as the clearest and historically most prominent instance of the dialectical process is formal disputation of the kind that was well known as a common procedure in universities in the middle ages. According to Rescher (1971, p. 1), this procedure of disputation before a master and an audience served as a “major training and examining device” in the middle ages in the universities of the time. He also sees the procedure as “closely akin to a legal trial”, presided over by a judge who determines the outcome (p. 1). Rescher observes that such a university disputation was set up in the fashion of a courtroom. It was presided over by a supervising master called the determiner, who summarized the result of the disputation and made a ruling on its outcome (p.1). This feature of having a third party determiner is especially significant in characterizing Rescher’s system, because other dialog systems have for the most part been of a simpler kind where there are only two parties, a proponent and a respondent, whose moves are governed by rules, but there is no third party to enforce the rules or to judge the outcome (Walton, 2007). There is no conflict between the two party dialectical system and a

406

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

three party dialectical system. The latter can simply be seen as a special case of the former where a third party is added on as a method of judging the outcome and seeing that the moves of the two other parties are made in accord with the rules. However, the fact that Rescher has opted for a three party system as his basic dialectical system is significant. It shows the strong historical influence of the model of formal disputation that was developed as a procedure in the universities in the middle ages. We find this characteristic of Rescher’s dialog system a distinctive feature of it, and will comment further on it in connections with Rescher’s metaphilosophy in the next section, and also discuss its significance for the study of formal dialog systems generally in the concluding section (section 8). 3. APPLYING DIALECTIC TO METAPHILOSOPHY As we return to considering Rescher’s metaphilosophy, one rather large question naturally arises. Who is the third party determiner who sees that proper dialectical procedures are followed in philosophical disputation, and who decides the outcome on whether an argument was properly defended or not? In the case of a forensic debate, there is such a judge or determiner who decides who won and who lost the debate. In the case of a philosophical discussion in a university classroom, the professor or instructor guides the discussion, tries to keep it on track, tries to see that questions are answered and objections addressed, and perhaps makes some attempt to judge whether the discussion was successful or not in light of the pedagogical aims of the class. But in many of the kinds of philosophical discussions in ordinary conversation, and also those that take place in philosophical journals or other written forms of communication, there really is no such third party judge or determiner who oversees the procedure and judges the outcome. Perhaps the referees who decide whether in article submitted to a journal, or a book manuscript submitted for publication should be acceptable or not, can be cast in this third party role. But it looks like they do not really fit into this category. One of their jobs is to criticize weak points in the argument, and they do have to make a quality judgment of whether this admission is publishable or not, a kind of success criterion. But this does not seem to be the kind of case that Rescher is thinking about when he gives examples of how philosophical theories develop out of apories. If we look at such cases of how philosophical theories develop through criticism and through increased sophistication by dealing with inconsisten-

407

RESCHER STUDIES

cies and exceptions to general rules, we see that the process does definitely have the form of a dialog in which there are two opposed sides. It appears to be the type of dialog that has been called the persuasion dialog or critical discussion that begins with a conflict of opinions and then proceeds through an argumentation stage in which probing questions are asked and each side not only supports its own viewpoint but criticizes and attacks the viewpoint of the other side. But in typical instances of such a process of the kind cited by Rescher, each side responds to weaknesses or inconsistencies of commitments pointed out by the other side. Since such a weakness would make his view appear implausible, a philosophical arguer is compelled to make his theory more sophisticated by dealing with such weaknesses and problems found in it. But there does not seem to be any third party judge or determiner who says things like, “Here is an inconsistency in your position as developed so far, so how can you deal with it?” The best we can say here perhaps it is that there is an audience or readership of philosophical literature, and the aim of a philosophical dispute is to produce a theory or position that is plausible to this audience. Is such an audience well cast as a single determiner or judge of the discussion? We do not think so. We think that anyone can judge philosophical argumentation in such a context by applying the rules of a critical discussion and then judge for him or herself whether that position developed by the arguer is consistent, plausible, or has responded in a persuasive way to objections made against it. In our opinion, the way to judge the argumentation used in a philosophical theory or in disputation, from a critical point of view is to identify, analyze, and evaluate specific arguments in it by using tools developed in argumentation theory. However, there are problems here too in cases where a third party referee or determiner or cannot be counted upon to make rulings. 4. THE PROBLEM OF BINDING COMMITMENT IN DIALOG SYSTEMS Certain common forms of argument have been formalized as argument schemes (Hastings, 1963; van Eemeren and Grootendorst, 1984; Kienpointner, 1992; Walton, 1996; Grennan, 1997; Verheij, 2003; Reed and Walton, 2005). Especially interesting are the defeasible schemes, where the premises do not imply the conclusion either by deductive or inductive inference. A main problem is how defeasible argumentation schemes are ra-

408

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

tionally binding. Properties of defeasible argumentation schemes are elucidated in this study through an important example of a type of argumentation related to fallacies, namely the appeal to expert opinion. This form of argument is often reasonable, but has been traditionally studied in logic under the heading of the argumentum ad verecundiam fallacy. The problem posed is to grasp how an argumentation scheme that is neither deductively valid nor inductively strong can be binding. If the person to whom the argument was directed accepts the premises, and the argument has the form of an argumentation scheme, how does the argument compel him rationally to accept the conclusion? This problem is framed in the paper as one of commitment in dialogue, using the device of the commitment store invented by Charles Hamblin (1970) as a tool for analysis of informal fallacies. Leaving defeasible forms of argument aside for further consideration below, the problem can be expressed in terms of a deductively valid form of argument like modus ponens. Suppose a respondent accepts the premises of an argument put forward to him by its proponent, and the argument has the form of modus ponens. Logically speaking, the respondent, assuming he is a rational arguer who is taking part in some sort of discussion where rationality is important, should also accept the conclusion of the argument. Of course, other responses should be allowed as well. He might decide to question one of the premises, or retract his previous commitment to it. He might ask critical questions suggesting that the modus ponens form does not apply to the argument in this particular case in the way the proponent assumed. Or he might bring forward a counterargument that has as its conclusion the opposite of the proposition that was the conclusion of the original argument. But barring one of these legitimate responses, should he not be held to accept the conclusion of the proponent’s argument, given that he is participating in a discussion in which both sides are claiming to be rational arguers? The tool called the profile of dialogue (Krabbe, 1999), in which a relatively short sequence of connected moves with the proponent’s moves are paired with those of the respondent, can be used to frame an example. In the example, the proponent opened at her first move by asking the respondent for reasons why she should accept the proposition A. He replied by putting forward a sequence of argumentation through seven moves offering a compelling reason why she should commit to A. As the dialogue proceeded, the respondent continued trying to use arguments to get the proponent to commit to A. He used a modus ponens form of argumentation

409

RESCHER STUDIES

____________________________________________________________ Table 1: Example Dialog of Commitment Problem Proponent 1. Why should I accept A? 2. Yes. 3. Yes. 4. Yes.

Respondent Do you accept B? Do you accept ‘If B then C?’ Do you accept ‘If C then A?’ Do you accept modus ponens as valid? 5. Yes. Well then you must accept A. 6. Why? You must accept C, by moves 1 and 2. 7. I still don’t accept A. You must accept A, by moves 3 and 6. ____________________________________________________________ by first of all getting her to commit to two conditional propositions, and then by getting her to accept the antecedent B of the first conditional. C then follows by modus ponens, so she must now accept C. But then, by the second condition she accepted at move 3, combined with C, she must accept A. Once again the result follows deductively by modus ponens. At her reply at her move 4, she reaffirms that she accepts modus ponens as a valid form of argument. But she won’t accept A, even though this refusal of commitment would not follow the requirement of rational argumentation that if you are committed to the premises, and committed to the form of argument, you should be committed to the conclusion (unless you make another appropriate responses). In a Rescher dialog, the problem appears easy to solve. The respondent can simply point out that the proponent is committed to an inconsistency, and that is the end of the dialog. The proponent loses. But this avoids the most central and substantive problem of retraction of commitment. Both participants need to be encouraged to take on commitments, to make the dialog move forward to resolving the ultimate conflict of opinions and reveal the deeper commitments on the issue. The problem is how to present some sort of incentive in the system for the proponent to either make such an appropriate response or to maintain consistency in her commitment set by accepting the conclusion of the argument put forward by the respondent. At the same time, another problem is to encourage and guide the

410

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

party who has had an inconsistency exposed in her commitment set to come to reasoned judgment of what to give up. In this instance, the proponent can still fail to commit to A, but there needs to be some penalty. According to the solution of retraction of commitment advocated in (Walton and Krabbe, 1995), she must either retract commitment to B or C, or to one of the conditionals ‘If B then C’ and ‘If C then A’. Various solutions to these problems of commitment are put forward in the dialog systems of (Walton, 1984). But since that time, many defeasible argumentation schemes of the kind that are common in everyday conversational argumentation have been identified. The even more urgent problem is to try to grasp how these schemes are rationally binding on the commitments of the respondent when put forward by a proponent. Hastings (1963) presented the first systematic identification and analysis of many of the most common defeasible argument schemes. He described each scheme as having a set of critical questions matching it. More recent work that has identified and analyzed many such defeasible argument schemes has followed this basic format (Kienpointner, 1992; Walton, 1996). To illustrate how defeasible schemes work, the example of appeal to expert opinion is often used. This form of argument, called “appeal to authority” had long been taken to be a fallacy. But argument from expert opinion is a very common form of argument, especially in legal argumentation, where expert testimony is one of the most widely used forms of evidence. The following argumentation scheme for appeal to expert opinion was presented in (Walton, 1997, p. 210). Argument Scheme for Argument from Expert Opinion Major Premise: Source E is an expert in subject domain S containing proposition A. Minor Premise: E asserts that proposition A in domain S is true (false). Conclusion: A may plausibly be taken to be true (false). One can easily see from the argument scheme for argument from expert opinion that it represents a defeasible form of argument. Although there is a natural tendency to defer to experts, experts are often wrong, and appeal to expert opinion should not be taken as a substitute for objective or ex-

411

RESCHER STUDIES

perimental evidence when it is available. For these reasons, argument from expert opinion is best evaluated as a defeasible form of argument that should be subject to critical questioning. The following six basic critical questions are given as appropriate for the scheme in (Walton, 1997, p. 223). Basic Critical Questions for Appeal to Expert Opinion Expertise Question: How credible is E as an expert source? Field Question: Is E an expert in the field that A is in? Opinion Question: What did E assert that implies A? Trustworthiness Question: Is E personally reliable as a source? Consistency Question: Is A consistent with what other experts assert? Backup Evidence Question: Is E’s assertion based on evidence? The expertise question concerns the knowledge of the expert source. The trustworthiness question concerns the reliability of the source. The field question is important from a point of view of fallacies because the fallacy of appeal to expert opinion often occurs where the source is an expert, but in a field different from that in which the appeal falls. The opinion question is important from a point of view of fallacies, and especially in legal argumentation as well (Godden and Walton, 2006), because the exact wording of what an expert says may need to be carefully quoted and examined. The backup evidence question is important because the user will often be able to use expert advice better if he understands something about the evidence the expert’s opinion was based on. 5. HOW COMMITMENTS ARE BINDING IN TWO PARTY AND THREE PARTY SYSTEMS There is one especially important fundamental difference between ordinary conversational argumentation of the kind typical of a philosophical discussion and legal argumentation of the kind found in a trial. In ordinary argumentation each party directs his argument to the other. For example,

412

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

the aim of the proponent is to persuade the respondent to come to accept the proponent’s thesis. If any fallacies are to be found in the arguments of one party, it is the other party who identifies them, and makes judgments about how strong or weak these arguments are. The dialog is selfmonitored. If one party breaks a rule, for example by preventing the other party from having a proper chance to speak, it is the other party who complains. There can, in many instances, be a third party involved, like an audience or referee, but in many cases there is not, and the third party may have different roles. For example, it could be a passive audience or an active third party who argues against the other two, or who plays a role like that of a referee. The very simplest case of a dialog needs two parties. They don’t have to be actual persons, but there must be two sides, the pro and the contra. The theory of dialog requires two participants whose views or interests are opposed in some way, and who take turns supporting or questioning these views or interests by making moves. In the literature on fallacies and argumentation, it has been this two party model of dialog that has predominated so far, even though with some fallacies, like the ad hominem, a third party audience needs to be part of the dialog setup. In the ad hominem argument, one party attacks the argument of the other by attacking the other’s ethical character (for example, for honesty), and the uses that attack to try to denigrate the plausibility of the other party’s argument to make it seem less plausible in the eyes of an audience. For example, in negative campaign tactics of the kind so often used in political argumentation, the audience is the public, or perhaps more specifically, the voters. The kind of legal argumentation used in trial, for example, seems fundamentally different. In a trial, each side directs its arguments to a third party, the so-called trier of fact, whether it be a judge or jury. This fundamental distinction makes a big difference in how rational argumentation needs to be treated in these two different contexts, and how commitment is binding in different ways in the two contexts. In the two party dialog system, when the proponent puts forward an argument that fits an argument scheme that is binding on the respondent, and the premises are commitments of the respondent, the respondent has to accept the conclusion of the argument. Suppose for example the proponent puts forward an argument that fits the argument scheme for defeasible modus ponens.1 Suppose, in other words that the argument is an instance of Rescher’s rule of detachment for provisoed assertion. In a two party dialog system such an argument once put forward has

413

RESCHER STUDIES

to be binding on the respondent in the following way. He either has to accept the conclusion or he has to put forward some appropriate critical questions matching that argument scheme. Or alternatively, he has to put forward a counter-argument, or otherwise respond to the argument by raising critical objections to it, unless he is willing to accept it. This requirement means that arguments that fit argument schemes are rationally binding on a respondent when put forward by a proponent in a two party dialog system. This particular feature is vitally important to understanding how argumentation in a dialog system representing ordinary conversational argumentation is binding on the commitment of a respondent to whom an argument is addressed. Suppose however we compare the case of a legal argument used in a trial by the one side or the other. It is clear in such a case that the respondent does not have to raise an objection immediately or concede the conclusion of the argument. In a trial, when the lawyer for one side puts forward an argument, the lawyer for the other side has the option of (a) responding to that argument immediately by making an objection or producing a counterargument, or (b) saying nothing at all. For strategic reasons, the lawyer may decide not to respond to the argument at all. This distinction between how commitments are rationally binding in a two party dialogue system and how they are rationally binding in a three party dialogue system is fundamentally important to grasping how these two different kinds of systems model rationality. In the two party system, there is no third party judge or determiner to enforce rules or to see to it that either of the other two parties pays attention to each other’s arguments or modifies their commitments in accord with these arguments when they are put forward. The fundamental problem in such systems (Walton, 1994) is that a respondent can simply avoid commitment by saying in response to any move of the other party, no matter how reasonable or powerful, “No commitment”. The basic problem in such systems is there needs to be some way of encouraging or even forcing commitment to the conclusion of a strong argument that has met all the requirements needed to change the respondent’s commitments. For example, the strategy of the proponent in a persuasion dialog system is to use arguments that make the respondent become committed to propositions that can later be used as premises to prove the proponent’s thesis in the dialog. But if the respondent continues to avoid commitment, even when confronted by a strong argument based on premises that are prior commitments of his, the proponent can never be successful in persuading the respondent.

414

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

This principle of requiring the modification of commitment by a respondent who is confronted with an argument that should be rationally persuasive was held in (Walton and Krabbe, 1995) to be the central problem of commitment in formal dialog systems designed to represent everyday conversational argumentation. How otherwise can formal dialogue systems have normative bite when applying to real cases of everyday argumentation? In such cases, participants may simply ignore rational arguments, and decide they’re not going to conform to valid argument forms, or to withhold commitment even when it is clear that such a refusal is irrational. To make formal dialogue systems useful as normative models of real cases of everyday argumentation therefore, some rules need to be put in assuring that commitment is binding on a respondent under the right conditions. The system is very different in a three party system of dialog in a legal trial. The trial is highly adversarial in the way it works to uphold legal goals of fairness and finding the truth of a disputed matter. Advocates on both sides are free to put forward whatever arguments they want, and respond in any way they think strategically advantageous for opposing the arguments put forward by the other side. Arguments put forward by one side are not really directed to the other side. They are directed to the third party trier of fact who will decide the outcome. An advocate will try to put forward the strongest possible arguments supporting the ultimate probandum of her own side, while making the strongest possible objections and counter-arguments to raise doubts about the arguments put forward by the other side. But the decision on who won the case will be made later as the arguments of both sides are compared, put together, and evaluated as they are recalled by the trier. If it is right that how commitments are rationally binding is a matter that needs to be treated differently in two party systems versus three party systems of dialog, some questions need to be raised on what the best applications of Rescher’s system are. Is his system more usefully applicable to cases of argumentation in everyday conversation, or are they more usefully applicable to legal cases of argumentation of the kind found in a trial? It would seem that it might be most usefully applicable to cases of legal argumentation. But this remains to be seen. What is interesting to comment on is that the fundamental problem of how commitments are binding in formal systems of dialogue needs to be considered in judging how useful such systems are as normative models that can be applied to real instances of argumentation.

415

RESCHER STUDIES

6. IS THERE A BURDEN OF QUESTIONING? A suggestion considered so far is that if formal models of dialogue are to be useful to study fallacies and other kinds of potentially illegitimate moves in everyday conversational argumentation, reasonable arguments should be viewed in such systems as normatively binding on the respondent. This normative binding property can appear in two forms. The first one has been described above. It requires the respondent to accept the conclusion of an argument having the form of one of the schemes provided he has already accepted the premises of the argument. Of course, such an acceptance is defeasible. In lieu of accepting the conclusion he should have the option of questioning or attacking one of the premises, or questioning whether the scheme really applies in the case at issue. The second form of the normative binding property relates to the asking of critical questions. If the respondent asks an appropriate critical question of the kind matching the scheme fitting the proponent’s argument, the making of such a move would seem to imply that the argument would default. It would either be defeated, at least temporarily, or would be put into a state of suspension, once the asking of the critical question has raised doubts about it. Because argument schemes do have this normatively binding quality, once put forward in a dialog, and questioned, they impose a burden on the proponent to satisfactorily answer the critical question posed. There is also a normative obligation on the part of the respondent to either accept the argument or to raise such questions. When the issue of critical questions was first discussed in the literature, the prevailing view was that no burden of proof attached to asking critical questions (Godden and Walton, 2007). It is commonly accepted that parties making assertions incur a burden of proof to successfully defend their assertions with acceptable reasons, and that they bear an obligation to retract those assertions that they cannot successfully defend. It was thought that “to ask an appropriate critical question in a dialogue shifts the burden of proof back onto the side of the proponent of the original argument to reply to this question successfully” (Walton, 1996, p. 15). Recent developments have challenged these views. It has recently been argued (Walton, 2003) that there may be a burden of questioning in certain types of dialogs, or in certain argumentative circumstances in a dialog. Also, it has been suggested that, when trying to specify how critical questions can be represented in models diagramming the structure of argument schemes, some critical questions might best be seen as having a burden of proof attached to them (Walton and Reed, 2003;

416

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

Bex, Prakken, Reed and Walton, 2003). Walton and Reed (2003) argued that sometimes there is a burden of proof attached to raising a critical question while in other cases there is not. Most critical questions, once asked in a dialog, shift the burden of proof back to the original proponent of the argument. But some critical questions rest on some implicit claim which acts as an objection to the argument. These kinds of critical questions have a positive burden of proof attached. In order to make the original argument default they have to be backed up with additional evidence needed to support the question. These observations have led to two theories about requirements for burden of proof for refuting shifting when critical questions matching the argument from expert opinion are asked (Gordon, Prakken and Walton, 2007). According to one theory, in a case where the respondent asks any one of these critical questions, a burden of proof automatically shifts back to the proponent’s side to provide an answer, and if she fails to do so, the argument defaults. On this theory, only if the proponent provides an appropriate answer is the plausibility of the original argument from expert opinion restored. This could be called the shifting burden theory, or SB theory. According to the other theory, asking a critical question should not be enough to make the original argument default. The question, if merely deflected back to the other side, needs to be backed up with some evidence before it can shift any burden that would defeat the argument (Gordon and Walton, 2006). This could be called the backup evidence theory, or BE theory. Only if the respondent specifically cites evidence of the sort that would back up the question does the proponent’s argument default. Consider the opinion question matching the scheme for the argument from expert opinion. It asks what E asserted that implies A. If the proponent supplies some specific proposition allegedly implied by what the expert then said, that this should be enough to satisfy the questioner, and the burden of proof should shift back to the respondent once the proponent answers the question satisfactorily. What should happen, however, if the proponent fails to supply such a proposition? The argument should default, because unless the proponent can quote, or otherwise represent accurately the proposition that was supposedly asserted by the expert, the respondents or the audience are in no position to judge whether what the experts said was represented accurately and fairly. Now consider the trustworthiness question matching the scheme for argument from expert opinion. It refers to the reliability of the source as one that can be trusted. It would be a powerful attack on an argument from ex-

417

RESCHER STUDIES

pert opinion if the expert was shown to be biased or a liar. But unless there is some evidence of this, the proponent should be able to simply answer “Yes, the expert is trustworthy.” Such a reply should be enough to answer the question appropriately. The questioner surely needs to produce some evidence of bias or dishonesty to make the charge stick. Otherwise the proponent can reply, “There is no evidence of that at all”, shifting the burden to the respondent’s side to back up his question with some evidence this sort. So far we have only considered problems of what happens in ordinary conversational argumentation when one party puts forward an argument from expert opinion and the other party raises appropriate critical questions about the argument that matches the argument scheme. The basic problem is how the argument, once put forward, should be binding on the commitment of the respondent. But this problem manifests itself in several distinct questions concerning the right sequence of dialogue moves that need to follow the putting forward of such an argument. Does the respondent have to accept the argument if he is committed to the premises? Can the respondent ask critical questions instead of accepting the argument immediately? The answer to this question is surely ‘yes’, since the argument from expert opinion is inherently defeasible as a form of reasoning. But now some even larger questions need to be posed. What happens when the respondent asks one of these appropriate critical questions? Does that simply defeat the argument, and that is the end of the matter, or can the proponent restore the plausibility of the argument by answering the question? It might seem straightforward that the answer to this question is that the proponent can restore the plausibility of the argument by answering the question. But as we’ve seen above, the situation is not so simple because of the dual nature of critical questions. In some cases, the proponent doesn’t need to answer the question but can simply shift the burden of proof back onto the questioner by asking for evidence of the kind needed to backup the question and give it enough force to defeat the argument. This problem cannot be avoided by any system of formal dialectics, because it brings into question whether any such formal system can function usefully as a normative model of argumentation that can be applied to real cases. This latter problem of how to differentiate between the effect of the two different kinds of critical questions in dialogs has led to the development of the Carneades system (Gordon, Walton and Prakken, 2007), based on a distinction between three kinds of premises in argumentation scheme: ordinary premises, assumptions, and exceptions. Ordinary premises are the

418

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

ones explicitly stated in the argumentation scheme. Assumptions are additional premises that are assumed to be true, and that have to be proved in the same way as ordinary premises for the argument to hold. Exceptions are premises that are assumed to be false, and not to hold, unless evidence is provided by the questioner that such a premise does hold. We now come to consider how such problems about normative conditions regarding burdens of proof in questioning arise in to arguments from expert opinion in legal argumentation. We’re thinking of cases where expert testimony is used as a form of evidence in a trial. An expert is brought in to testify for one side and then cross-examined by the other side. The difference between this kind of case and that of ordinary conversational argumentation is that there are many procedural rules, especially rules of evidence, that are brought to bear in a trial regarding the admissibility and evaluation of expert testimony as a kind of evidence. In a trial, there is a third party, namely the judge, who sees to it that these rules are followed. The judge may also have the role of argument evaluator who decides at the latter stage of the trial which side met its burden of proof. In other cases, this rule is performed by the jury. Such legal uses of argumentation, especially in the context of the trial, definitely fit Rescher’s three party formal model of disputation better than they fit the simpler two party model. Thus it could well be a strong advantage of Rescher’s system of dialectic that it is much better suited to legal argumentation than the traditional two party model that has dominated the bulk of literature on formal dialogue systems for argumentation so far. 7. EXPERT TESTIMONY IN LAW The advantage of the Rescher-style three party dialog system is that it makes room for a third party adjudicator who can rule on these problems of burden of proof when they arise in a trial where expert testimony is put forward as evidence. If the two sides get into a dispute about the asking of critical questions of the kind considered above, the judge can intervene and make a ruling. If there is a dispute about whether expert testimony should be admissible in a given case or not, once again the judge can make a ruling. In fact, this is exactly what tends to happen in legal argumentation in a trial. There is a meta-dialogue interval (Krabbe, 2003) during which the judge considers the moves that have taken place in the dialog between the two sides during the trial, and makes a ruling concerning these moves in a

419

RESCHER STUDIES

case where there is some procedural controversy about their admissibility, or some controversy about matters relating to burden of proof. We have little space here for examining cases in detail, but before getting into discussion of theoretical problems arising from a shift in a burden of proof, it might be good to consider an outline of a case that is highly typical of how rulings on bias in expert opinion evidence cases actually work. The following passage has been excerpted from Yoho v. Thompson, 336 S.C. 23, 518 S.E.2d 286 (Ct. App. 1999). Petitioner Dorothy Yoho sued respondent Marguerite Thompson to recover damages for injuries she sustained when Thompson’s car struck Yoho's car from behind. Prior to trial, Thompson's insurer paid Yoho the policy limits of $50,000 and Yoho’s underinsured motorist carrier, Nationwide Insurance Company (Nationwide), then assumed Thompson’s defense. At trial, Thompson admitted liability, leaving damages as the only issue for the jury. Prior to trial, Thompson indicated she would call Dr. William Brannon as a witness. Dr. Brannon had reviewed Yoho's medical records and would give an opinion as to the extent of her injuries. In motions made before trial and prior to Dr. Brannon's testimony, Yoho asked the trial judge to allow her to cross-examine Dr. Brannon regarding his relationship with Nationwide to establish possible bias. Yoho presented Dr. Brannon's deposition testimony from another case that he did “a fair amount of consulting work with Nationwide” and had given lectures to Nationwide agents and adjusters. Yoho also presented information that ten to twenty percent of Dr. Brannon's practice consisted of reviewing records for insurance companies, and that his yearly salary was based on the amount of money his practice earned, which included his consulting work. The trial court denied Yoho’s motion on the basis that the probative value of the content of the cross-examination would be outweighed by the prejudicial effect of injecting the issue of insurance into the proceedings. The court informed Yoho that she could discuss Dr. Brannon’s bias by using generic terms such as “defense”, “defendants”, and “defense lawyer”, but that she could not discuss his possible bias by using the word “insurance”.

This case reveals several features of the practical reality on how bias objections to arguments from expert opinion are routinely handled as evidence by the courts. The first thing to notice is that the jury would probably not be much influenced by the mere asking of the critical question of whether Dr. Brannon might be biased. In order to make much impact on them, Yoho would surely have to offer some evidence of this alleged bias. The real problem for the judge posed by the case is whether the evidence Yoho

420

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

collected is admissible. The problem posed by this kind of case comes under the trustworthiness critical question treated in section 6. Asking whether the expert has a bias is a subquestion of the trustworthiness critical question. Just asking whether the expert is biased is not enough by itself to shift the burden of proof, because the other side can simply ask, “Can you give any evidence of bias?” Unless some evidence is given to support the question, the bias question lacks enough force to shift the burden of proof. This case suggests that the real value of having critical questions matching an argument scheme like argument from expert opinion is that the questions enable an arguer to seek out avenues that identify points of challenge in an argument. These points of possible attack can then be followed up or not, depending on what is known or available as evidence in the case. Thus the value of the set of critical questions is that they provide entry points for a further search into how to attack an argument, or to locate missing premises in it that can be questioned or at least recognized. Yet the problem remains of determining how critical questions, rebuttals, and objections should be replied to in systems of dialog that can accommodate the complex moves that are made in trials, and how burden of proof rules should apply to such moves. In the reality of a courtroom, the argumentation is complex, outranging the parameters of a simple dialog model. Although the respondent is questioning the proponent who originally put forward the argument from expert opinion, it is the jury who is the real target of persuasion. No argument is worth putting forward unless it will have a significant impact on the jury. The respondent is probing into the commitments of the proponent in an examination dialogue that is embedded into a persuasion dialogue in which the jury is the target of the persuasion attempt. Despite this rather complex situation however, the general expectation of theory, to the effect that the bias critical question needs to be backed up, does appear to hold. It suggests that the burden of proof needs to be on the questioner to not just ask whether the expert is based, but to offer some reason to think so. 8. CONCLUDING REMARKS We begin this section with some remarks about Rescher’s metaphilosophy, based on his notion of an apory, and then conclude with some general remarks about how we see Rescher’s system as fitting in with the recent literature on dialog theory.

421

RESCHER STUDIES

The way we see Rescher apories is that they represent attacks on a philosophical position of an important kind that has often been maintained and been found to be plausible and defensible by philosophers. Thus to analyze an apory you really need to begin with some position that can be defined as a set of commitments that have been held to be plausible and that represent a philosophical viewpoint of some interest or importance. Then you need to show, by isolating a set of propositions, that there is an inconsistency that can be generated by adding a new proposition to the set of the existing arguer’s commitments. This new proposition needs to be highly plausible, or at least it needs to appear to be a proposition that could be defended by giving plausible reasons to support it. The defender of the original philosophical viewpoint must then deal with the apory on pain of being committed to an inconsistency. What is called for is an adjustment of commitments by retracting one or more of the propositions in the original set. Finally, we take up what we consider to be the fundamental question concerning the relationship Rescher’s formal system of dialectics to argumentation of the kind that we are so often interested in analyzing and evaluating. Should the three party system or the two party system better be used to model such argumentation? An apory presents an inconsistency that provokes a philosophical discussion falling into two sides or opposed positions. Confronted with such an inconsistency, people are free to develop their own views and to defend the one side or the other. In this respect, we think that philosophical argumentation of the most common kind is similar to everyday argumentation, and that it is up to the two parties themselves to follow the conversational rules for public discussion. In modeling instances of this kind of argumentation, the other systems in the literature (Walton, 2007) do not see the value of having a third party arbiter or judge. They see the basic kind of philosophical argumentation in everyday conversational argumentation of the kind most often taken as the target in the literature as best modeled by a two party dialectical system. In this respect, there is a fundamental difference between ordinary philosophical argumentation and legal argumentation of the kind encountered in a trial. In the latter kind of argumentation, a three party model appears to be the better one to apply. However, one might say, what’s the difference anyway? One can always use the two party model as the basic system, and then add on a third party in more complex systems as this special feature is needed. One could reply, however, that the question of how commitments should be held to be binding has proved to hard solve, precisely because

422

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

there is no third party to make clear and decisive rulings. In a three party dialog the judge can always step in and see to it that a question is answered, that an argument is relevant, that a party does not speak out of turn, that a party makes an attempt to deal with an inconsistency in his or her position, and so forth. However, in the two party model, it may seem that there is really no assurance that argumentation schemes will be binding on a respondent. The respondent may accept the premises of an argument presented by the proponent, and the argument may fit an argumentation scheme, but the respondent may simply decline to accept commitment to the conclusion. The example dialog shown in table 1 is precisely such a case in point. This is a problem for a formal system of dialectic of the two party type, because one might well question whether such a formal system of dialogue represents a normative model of argumentation at all. In the three party system, this third party can always ensure that such requirements are reasonably met. So far, the three party model seems not only necessary, but intrinsically better than the two party one. But does the three party model offer hope of solving these problems only by introducing other problems? It can be argued that there is a dilemma for the three party model, a kind of apory in fact that needs to be resolved. The function of the rules of the dialog is to enable the two parties who are contesting with each other to each be sure that both parties are following these rules. The ability of each party to appeal to the rules to protest in an instance where the other party appears to have violated one of the rules is supposed to be an important feature of the existing dialog systems. We could argue that this feature of having such rules that limit needs the need for a third party arbitrator. There is also a danger in three party system that the third party, the rational judge or mediator, may make mistakes in applying the rules, or even favor one side over the other in a biased way. In effect the dialog may become an authoritarian system of the kind that dialog models of rational argumentation are specifically designed to combat. So there is a dilemma here. If the rational judge just always follows the rules, then there is really no need for her. The two parties themselves can follow the rules, and each party can insist that the other party follows the rules if he thinks that she has violated one of them. On the other hand, if a rational judge has the freedom to apply the rules in the way that she thinks fit, what we have is an authoritarian system. Indeed, what is often taken to be the central part of the notion of the critical discussion type of dialogue is the rule that a participant is committed to retraction if the other party shows that one of his arguments is faulty.

423

RESCHER STUDIES

If a participant is not committed to the vulnerability of his view, he’s not really gaining from the critical discussion anyway. Thus it seems that an important part of a dialog in which rational argumentation is taking place is that the dialog is self-policing. The most important part of it, arguably, is that participants need to spot the fallacies and procedural faults in the moves and arguments of the other party, and also recognize and correct the faults in their own arguments. These remarks offer a rationale to help us understand why the literature is dominated by two party systems which eliminate the need for third party arbitrator. Another reason may be that formal systems of dialogue tend to become very complex very rapidly. Many such complex systems already exist in the argumentation literature and in the literature on artificial intelligence as well (Walton, 2007). Given this fact, it has begun to appear recently at the best approach is to start with simple systems of a basic sort to which features can be added for specific applications, rather than beginning with complex systems that have too many bells and whistles to provide a basic structure applicable to the many and varied needs of computing and other applications. Despite these reservations, we still find reasons for concluding that a case can be made for using a three party model to represent legal argumentation of the kind found in a trial. If the rules are clear and precise, and it is clear how they apply to a given case, then a two party model might do the job. However, cases that go to trial tend to be those where a good deal of judgment is required in applying the rules to the facts. The facts of the case may be uncertain, and many of them may simply be unknown because of lack of evidence. Legal rules are defeasible, and often conflict with each other on the question of how they apply to a particular case. These are precisely the kinds of hard cases where we needed a third party trier, a judge or jury. We conjecture that it is precisely to this type of dialog that the Rescher three party model is best applicable.

REFERENCES Bex, Floris, Henry Prakken, Chris Reed and Douglas Walton, “Towards a Formal Account of Reasoning about Evidence, Argument Schemes and Generalizations,” Artificial Intelligence & Law, vol. 11 (2003), pp. 125165.

424

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

Freeman, James B., “Progress without Regress on the Dialectical Tier,” in Hans V. Hansen and Robert C. Pinto (eds.), Reason Reclaimed, (Newport News, Virginia: Vale Press, 2007). Godden, David M. and Douglas Walton, “Argument from Expert Opinion as Legal Evidence: Critical Questions and Admissibility Criteria of Expert Testimony in the American Legal System,” Ratio Juris, vol. 19 (2006), pp. 261-286. Godden, David M. and Douglas Walton, “Advances in the Theory of Argumentation Schemes and Critical Questions,” Informal Logic, vol. 27 (2007), pp. 267-292. Gordon, Thomas F. and Douglas Walton, “The Carneades Argumentation Framework: Using Presumptions and Exceptions to Model Critical Questions,” in P. E. Dunne and T. J. M. Bench-Capon (eds.), Computational Models of Argument: Proceedings of COMMA 2006 (Amsterdam: IOS Press, 2006), pp. 195-207. Gordon, Thomas F., Henry Prakken and Douglas Walton, “The Carneades Model of Argument and Burden of Proof,” Artificial Intelligence, vol. 171 (2007), pp. 875-896. Grennan, Wayne, Informal Logic (Kingston: McGill-Queen’s University Press, 1997). Hamblin, Charles L., Fallacies (London: Methuen, 1970). Hastings, Arthur C., A Reformulation of the Modes of Reasoning in Argumentation (Evanston, Illinois: Ph.D. Dissertation, 1963). Kienpointner, Manfred, Alltagslogic: Struktur und Funktion von Argumentationsmustern (Stuttgart: Fromman-Holzboog, 1992). Krabbe, Erik C. W., “Profiles of Dialogue,” in J. Gerbrandy, M. Marx, M. de Rijke and and Y. Venema (eds.), JFAK: Essays Dedicated to Johan van Benthem on the Occasion of his 50th Birthday (Amsterdam: Amsterdam University Press, 1999), pp. 25-36.

425

RESCHER STUDIES

———, “Metadialogues,” in F.H. van Eemeren, J.A. Blair, C.A. Willard and A.F. Snoek Henkemans (eds.), Anyone Who Has a View: Theoretical Contributions to the Study of Argumentation (Kluwer: Dordrecht, 2003), pp. 83-90. Reed, Chris and Douglas Walton, “Towards a Formal and Implemented Model of Argumentation Schemes in Agent Communication,” Autonomous Agents and Multi-Agent Systems, vol. 11 (2005), pp. 173-188. Rescher, Nicholas, Dialectics: A Controversy-Oriented Approach to the Theory of Knowledge (Albany: State University of New York Press, 1977). ———, Philosophical Dialectics: An Essay on Metaphilosophy (Albany: State University of New York Press, 2006). ———, Presumption and the Practices of Tentative Cognition (Cambridge: Cambridge University Press, 2006). Verheij, Bart, “Dialectical Argumentation with Argumentation Schemes: An Approach to Legal Logic,” Artificial Intelligence and Law, vol. 11 (2003), pp. 167-195. Walton, Douglas, Argumentation Schemes for Presumptive Reasoning, (Mahwah: Erlbaum, 1996.) ———, Appeal to Expert Opinion (University Park: Penn State Press, 1997). ———, “Are Some Modus Ponens Arguments Deductively Invalid?” Informal Logic, vol. 22 (2002), pp. 19-46. ———, Dialog Theory for Critical Argumentation (Amsterdam: John Benjamins Publications, 2007). Walton, Douglas and David M. Godden, “The Nature and Status of Critical Questions in Argumentation Schemes,” in David Hitchcock and Daniel Farr (eds.), The Uses of Argument: Proceedings of a Conference at

426

RESCHER ON DIALOG SYSTEMS, ARGUMENTATION, AND BURDEN OF PROOF

McMaster University (Hamilton, Ontario: Ontario Society for the Study of Argumentation, 2005), pp. 476-484. Walton, Douglas and Thomas F. Gordon, “Critical Questions in Computational Models of Legal Argument,” in Paul E. Dunne and Trevor Bench-Capon, Nijmegen (eds.), Argumentation in Artificial Intelligence and Law, (IAAIL Workshop Series: Wolf Legal Publishers, 2005), pp. 103-111. Walton, Douglas and Erik C. W. Krabbe, Commitment in Dialogue (Albany: State University of New York Press, 1995). Walton, Douglas and Chris Reed, “Diagramming, Argumentation Schemes and Critical Questions,” in Frans H. van Eemeren, J. Anthony Blair, Charles A. Willard and A. Francisca Snoeck Henkemans (eds.), Anyone Who Has a View: Theoretical Contributions to the Study of Argumentation (Dordrecht: Kluwer, 2003), pp. 195-211. NOTES 1

According to (Walton, 2002), defeasible modus ponens is the following form of argument: if A then defeasibly B; A; therefore (defeasibly) B.

427

Rescher on Process Thought Michel Weber

M

ost, if not each, of the topics that the editor has aptly chosen to allocate among the present book’s separate chapters could claim to embody both the marrow of Nicholas Rescher’s work and the very core of all possible philosophical analysis (or even synthesis). In each case, one can easily foresee the argument that could be put forward, but it would be far more problematic to frame a set of meta-criteria to comparatively assess the pretensions of each of these fields. In the case of “process thought”, however, a double historical evidence combined with a systematical one could help the author wishing to foster such a bold thesis. On the one hand, the crowning of Rescher’s philosophical works by five of his most recent monographs, all process-oriented, together with the fact that, according to his own testimony, the impact of the process mode of thought is as old as it is deep, is truly remarkable. PM’s and PP’s Prefaces underline indeed Rescher’s enduring interest and specify that it spurred from Walter Terrence Stace’s (1886–1967) 1949 seminar in Princeton. On the other hand, the process approach in general and its Rescherian version in particular have most obviously renewed all the traditional areas organizing the philosophical work: from logic and mereo-topology to natural theology . . . Last but not least, process philosophy is not only endowed with a thorough selfconsistency and a high-grade applicability: it also provides a coherent meta-philosophical framework immediately accounting for the endless development of philosophy itself.1 Such are not, however, our primary aims in this short study. Since the publication of After Whitehead: Rescher on Process Metaphysics (2004), there already exist a companion-type volume to accompany the reader in his or her Rescherian process inquiries. Neither do we plan to debate in details Rescher’s pragmatist methodology and his well-known “conceptual idealism”—although, in the course of our argument we will point at the strong Kantian ring his systematic distinctions often bear and thereby rediscover its importance for the type of process thought that our author fosters. In order to sail close to the Rescherian wind and to remain as straightforward as possible in these historically and speculatively intricate matters, only three main conceptual threads are exposed in this paper, all three be-

RESCHER STUDIES

ing recurrently correlated in Rescher’s own process inquiries: first, the distinction between the philosophical doctrine or position on the one hand and the philosophical approach or tendency on the other; second, the definition of process as a sequentially structured succession of stages or phases; third, the relativization of the Whiteheadian legacy of process thought. In conclusion, we briefly contextualize Rescher’s processism with his methodological pragmatic idealism. 1. THE PROCESS APPROACH OR TENDENCY Of the three main threads that recur in Rescher’s process inquiries, the insistence on the perennial dimension of the process approach obviously stands out. We postpone to the next section his definition of process itself to focus here on the useful transhistorical taxonomies that our author provides in order both to unite and to discriminate the most remarkable historical vicissitudes of the concept. Rescher claims that one should distinguish between the philosophical doctrine or position on the one hand and the philosophical approach or broad programmatic tendency on the other: “The one is a specific and definite substantive position, the other a generic and potentially diffuse doctrinal tendency. And it is, of course, mistaken to look for doctrinal unanimity within any broad philosophical tendency.” (PP 44-45, cf., e.g., PM 27 sq. and PPD 3, 10, 33, 47, 99) More precisely, process thinking involves a clear emphasis on activity through the concepts of motion, change, passage, transition, contingency, eventfulness, becoming, creative synthesis or even creation. What basically matter are “interactive relatedness, wholeness (totality), activity (selfdevelopment), innovation/novelty, unity of law (functional typology), productive energy, drive, etc., fluidity and evanescence, activity (agency)” (PM 35; cf. also, e.g., PM 31). As such process metaphysics has thus to be contrasted with substantialism, that considers precisely that all contingent features of experience should be at best ignored and at worst revoked from our systematic attempts. Plato and Aristotle stand out as the faithful wardens of this tradition and even if a more processual interpretation of their work is possible, it remains by far totally marginal.2 Process thinking targets nothing less than the sumbebekos, i.e., to repeat, the very evidences that have been willingly downgraded—if not simply ignored—by mainstream philosophers . . . Anyway, Rescher’s first move in his analysis of the process philosophi-

430

RESCHER ON PROCESS THOUGHT

cal approach is to distinguish processual primacy and processual priority (PM 2-3, 27-34, 42-46, 90, 115); his second move is to contrast ontological processism and conceptual processism (PM 57-60; cf. 27 sq., 33, 46, 48, 56 sq., 115). The entire philosophical landscape is thus partitioned in the following way: on the one hand, the primacy of process exemplified by Heraclitus, as against the Parmenidian primacy of substance, and the priority of process of Empedocles, as against the Democritean priority of substance; on the other, the process ontological commitment of a thinker such as the late Whitehead versus the more careful process epistemological commitment of Rescher himself. But let us not anticipate. In sum, here are the four basic types of process philosophy that issue from these 2 x 2 set of criteria:3 Approach Priority of processes (weak) Primacy of processes (strong)

Epistemological (weak) Our understanding depends on process ideas. Our understanding is constituted by process ideas.

Ontological (strong) Things depend on processes. Things are constituted by processes.

The boldest thesis is of course the strong ontological approach: “Esse sequitur operari: things are constituted out of the flow of process, and substantiality is subordinate to activity” (PM 44; cf. PP 7, PPD 5 and NU 31). The ontological primacy of process over substance is understood by way of reducibility (to be contrasted with the “Process Reducibility Thesis” of PM 42 sq. and 63; cf. PP 6) and leads, for obvious reasons of consistency, to the thesis of the epistemological primacy of processes. The tamest thesis is the weak epistemological one that refuses to make any ontological commitment and only argues that we cannot understand the world without some recourse to the idea of process. Within that general matrix of possibilities, heuristic tools are needed to further discriminate the historical doctrines that have variously exploited the basic contrasts already mentioned (see supra our quote of PM 35) in order to do justice to the specific intuition of their respective authors. How are exactly activity, process, change and novelty articulated with substance, product, persistence and continuity? In the same way that Aristotle’s concept of substance does not amount either to Descartes’ or to

431

RESCHER STUDIES

Spinoza’s, Heraclitus’ ever-living fire is neither James’ stream of experience nor Whitehead’s creative advance of nature. Hence the introduction of two further pivotal distinctions: first, between product-productive processes and state-transformative processes (PM 41; cf. 31-40 and PP 28) and second, between owned and unowned processes (PM 42-46, 49, 63; cf. PP 28-29). Productive processes produce substantial things while transformative processes transform states of affairs. To exemplify: manufacturing processes and natural processes such as seed germination differ from windstorms and earthquakes. Owned processes represent the activity of actual agents (“the chirping of birds, the flowering of a bush, the rotting of a fallen tree”—PM 43) while unowned processes do not: they are “free-floating” (“the cooling of the temperature, the change in climate, the flashing of lightning, the fluctuation of a magnetic field”—ib.). According to Rescher, only state-transformative and unowned processes are fully relevant to “process metaphysics” because of the intrinsic substantial ring of their respective alternatives (PM 41-44). In conclusion, a trans-historical and cross-cultural heuristic grid is applicable, ultimately because, as we will see, “process” is nothing less (or nothing more . . . ) than a transcendental category. 2. THE SEQUENTIAL DEFINITION OF PROCESS AND EVENTS The trans-historical thesis being established, let us peruse Rescher’s definition of process in PM,4 which involves three main quasi-Leibnizian ideas5: sequential structure, eventuation and overlapping hierarchy. Needless to say that the stakes are high: the process philosopher has not only to do justice to process, flux, becoming and the like, but also to stability, being and permanence—and thereby to account for everyday commonsensical experience, according to which both being and becoming are to be taken at face value (cf. PM 31 and the discussion around the priority/primacy of process over substance). As Whitehead wittingly remarked: “you may polish up commonsense, you may contradict in detail, you may surprise it. But ultimately your whole task is to satisfy it”.6 Process should be construed first as “a coordinated group of changes in the complexion of reality, an organized family of occurrences that are systematically linked to one another either causally or functionally [ . . . ], an integrated series of connected developments unfolding in conjoint coordination in line with a definite program.” (PM 38, cf. PP 26) Elsewhere, Rescher even speaks of a process “recipe” (PP 19 and 25-26). This sequential

432

RESCHER ON PROCESS THOUGHT

definition brings thus to the fore a linear process series of the type: A—B—C— . . . Needless to say that processes are durational, i.e., that their sequential unfoldment necessarily develops over time. (This being almost a tautology.) Second, these “dispositionally structured modes of development” (PM 46; cf. PP 7-9 and 131) are to be correlated with occurrences or events: “processes always involve various events, and events only exist in and through processes” (PM 38).7 A serial qualification is thus in order: A !—B !—C !— . . . By which we mean that each sequential step is eventful, unique, that it constitutes an hapax symbolized with “!”. Third, “processes will always involve a variety of subordinate processes and events” (PM 38; cf. PP 23, NU 67): process, subprocess (that could be called prehension) and superprocess or megaprocess (or nexus-like togetherness) do not constitute an absolute context-independent fact (cf. PP 15). To further exemplify: “the ongoing self-identity of mind consists in a continuity of psychic process—even if it lies beneath the threshold of awareness.” (PM 110) More precisely, there is an eventful Aufhebung of sorts: There is an endless “sequence of levels of integrative complexity of phenomenal order” (NU 60 and 70). Hence two immediate consequences: first, past processes impacts present processes and will impact future processes—a fundamental pattern that makes temporalization explicit—; second, process is “mereologically homogeneous” (PP 23): a part of a process is itself a process.

In sum, there are three definitional steps: sequence, eventuation sequence, nested hierarchy of eventuation sequence. As a result, we rediscover activity and the other key-concepts announced above (cf. again our quote of PM 35): sequence qua interactive relatedness, eventuation qua innovation, and

433

RESCHER STUDIES

nested hierarchy qua wholeness. In other words: process qua “sequential pattern of action” (PM 40-41; cf. PP 23, 27 and PPD 2, 27) is, by hypothesis, “a fixed sort of eventuation-sequence” (PM 40) that constitutes a “hierarchic assemblage of micro into macro units” (PM 54; cf. NU 38, 42-44) and “inherently exhibits a structure of spatiotemporal continuity” (PM 39; cf. PP 23). Rescher further specifies that “two quite different sorts of complexity hierarchies” (PM 79; cf. PM 53 and PP 20) should be evoked: physical hierarchies of compositional structure and law hierarchies of physical order (that does not presuppose the former). The former are related by physical inclusion (e.g., molecules, cells, organs, organisms, colonies, etc.), the latter by nomic inclusion, from base-level laws governing phenomena to higher-level laws coordinating laws, and so forth and so on. Since the sequential dimension of process is firmly rooted in Rescher’s definition as well as in common-sense, solely its two specifications seem demanding of further specification at this stage of our argument. Two questions then immediately come to mind. First, what is exact status of novelty? Does it leak or does it jump into the world:8 does it smoothly unfold or does it abruptly break in? Second, how does Rescher’s optimism (e.g., PM 99, PPD 176 sq.) impacts on this definition? The issue of progress will be introduced cosmologically below; for the time being, let us focus on the first puzzle. The overall picture disclosed by this threefold originative definition is continuist: no analysis of the continuing unfoldment of processes over time and space will ever reach a more fundamental discrete process: it is “processes” all the way down (cf., e.g., PP 129 and especially all of Rescher’s discussions of Whitehead’s epochal ontology, i.e., the theory of concrescence, that is interpreted qua atomistic). The origination, flourishing and fading away of processes is akin to a bird's life made of an “alternation of flights and perchings”.9 To use another metaphor: stability is understood within a manifold of continual change in the same way “stability-waves of continuous process provide for any sort of continuity of existence” (PM 89 and PP 12). Even in epistemology the accent falls on continuity: erotetic discovery is the “phenomenon of the ever-continuing ‘birth’ of new questions” (PP 63). Processes construct a seamless tapestry, flowing as waves. Hence the conundrum: how exactly can we understand emergence, novelty, innovation, and creativity—i.e., unpredictability—in such a continuist system? Zeno’s paradoxes, Plato’s exaiphnes, Leibniz’s monads, James’ budlike experiences, Weiss’ gulps and Peirce’s synechism boldly argue for

434

RESCHER ON PROCESS THOUGHT

some process form of ontological discontinuism or intrusiveness (cf. Pepper quoted by PM 75; see also AW 57-59)—but this ontological commitment is precisely what Rescher does not deem wise because of the radical cognitive opacity limiting our ongoing interaction with nature (PP 91). In PM 74 sq., he argues with the help of a distinction between ontological innovation (new kinds: things, processes, products, states of affairs), phenomenical innovation (new types of events or courses of events) and epistemic innovation (new sorts of knowledge, ideas, information, problems, questions). In the context of our present sketch, three broader complementary levels of argumentation seem however more to the point: epistemological, moral, and cosmological. “Perhaps no form of novelty is more intractable from the predictive point of view than conceptual novelty—innovations in ideas and conceptions.” (PM 76) First, Rescher’s epistemology, broadly understood as fostering both his theory of knowledge and philosophy of science, is articulated around a double focus imaginarius. On the one hand, the pragmatically validated Kantian regulative postulate that qualifies the commitment to an objective world of real things as nothing more than a fundamental convention (PP 92), a common-sensical construal of the world10: according to the “essential Kantian insight”, intelligibility is not a constitutive feature of nature, but a regulative one.11 On the other hand, the (quasi) Platonic ingressive postulate that understands conceptual innovation as a process of temporalization of pre-existing potentialities. Take for example Rescher’s quote of Plato’s Symposium: “what is ‘in reality new [is also continuous with the old] according to that law of succession by which all mortal things are preserved, not absolutely the same, but by substitution, the old wornout mortality leaving another new and similar existence behind.’ Along these lines, processists stress that while we humans live on ‘borrowed time,’ we nevertheless have a foothold in eternity.” (PM 122)12 Or his idea of the exploration of parametric space introduced to provide a basic model for understanding the mechanism of technoscientific innovation in mature natural science (PP 74 sq.): its “steady stream of new data” is not new in kind “but in their functional interrelationships” (PP 86-87). Of the two postulates, the former rules it all and tames this Platonic trend into a conjunctivism: the medieval metaphysical dispute between nominalism, conceptualism, and Platonism needs indeed to be resolved conjunctively (PP 115; also PPD 16 and 53); “no belief system, no value framework, and no item of moral norms is an extra-mundane absolute, mandated in unchangeable perfection by circumstances beyond our control. The best we can do is to

435

RESCHER STUDIES

figure things out on our own.” (PPD 43 and 78; cf. PM 69 sq. and PP 9 sq.) There are however occurrences suggesting the necessary correlation between the innovative emergence of ever new unforeseeable existences and the breaking of pre-discernible patterns. When we speak of real conceptual novelty, it is not just a matter of degree and the spatio-temporal continuity of processes does not prevent the actuality of a conceptual irruption.13 Second, Rescher’s moral philosophy, that reinforces his pragmatic interpretation of Kant. Analysing the old problem of the apparent antagonism between causation and liberty, Rescher introduces the term “rent”14 to point at the well-known difficulty: the singularity that constitutes the act of free will has to be both causality-suspensive and inscribed in the causal matrix. Precisely, this has been put forward very boldly by the Critique of Pure Reason’s “third conflict of the transcendental ideas”, where Kant links together the freedom of the will and the general mechanism of nature, the autonomy involving noumenal a-causality and phenomenal causality. Third, Rescher’s cosmological optimism directly follows from his Kantian ethics: progress in the ethical sphere must be granted if one wishes to provide the conditions of possibility of moral action; this is the closest Rescher gets to an involvement with the ontological an sich. In sum, at all three levels, Rescher hints, just like the protopragmatist Kant, at the necessity to give some status to creative discontinuity. 3. THE POST-WHITEHEADIAN LEGACY Interestingly enough, the threefoldness of sequential structure, eventuation and overlapping hierarchy exploited in the previous section offers an indirect bridge with the process philosophy of science Whitehead has developed mainly during his so-called “London epoch” (1910–1924)15 and a direct one with his later ontology, as it was embodied, during his “Harvard epoch” (1924–1947), in his proto-idea of creative advance.16 First, how does the threefoldness somewhat matches Whitehead’s London epoch? The Method of Extensive Abstraction axiomatized in 1919 aims at extracting from sense-experience the primitive notions of science and common-sense alike (mainly space and time). It constitutes a skilled generalization of the instinctive procedure of habitual experience in the light of the Fregean definition of cardinal numbers with equinumerical classes. As such, the Method mainly exploits overlapping hierarchies of sequentially structured events that feature objects. In sum, whereas Rescher mereologically articulates processes and events, Whitehead

436

RESCHER ON PROCESS THOUGHT

mereologically articulates events and objects, the former being the relata of the fundamental relation of “extension” and the latter embodying the permanences allowing recognition and hence measurement (and hence technoscience). Second, Whitehead’s proto-idea of creative advance can be approximated with the threefold of sequential structure, eventuation and overlapping hierarchy. Before suggesting such an approximation, let us introduce to the proto-idea itself. Three functors are involved in the creative advance: the gift of creativity, the power of efficacy and the bliss of vision. Each functor specifies a particular speculative locus, but only the whole speaks. Creativity basically means the irruption of the unheard, the beginning of a new causal chain. In common philosophical parlance, it refers to becoming. It necessitates the concepts of epochality (the “epochal theory of time” that amounts to James’ bud-like experience) and liberty (that, at the universal scale, amounts more to spontaneity than to free-choice). Difference necessarily involves discontinuity. Here dwells actuality, i.e., the present per se that is worthy of the concept of duration. Efficacy basically means the reproduction of patterns. In common philosophical parlance, it refers to being. It necessitates the concepts of continuity and determinism. Repetition involves continuity. It belongs to potentiality, i.e., to the pervasive past. Vision basically means an eschatological horizon, a melioristic trend. In common philosophical parlance, it refers to God luring all processes. It necessitates the concepts of (hierarchies of) eternal objects and of (primordial nature of) God. In a purely metaphorical way, it can be attached to the future. These functors should be regrouped in two waves, according to Whitehead’s own understanding of creativity qua actuality, efficacy qua potentiality, and vision qua eschaton. First, the togetherness of creativity and efficacy is embodied in the concept of growth. Second, the overall togetherness of growth and vision embodies the creative advance. In other words, two things matter: on the one hand, the interplay of creativity and efficacy secure a genuine growth of the mundane tissue; on the other, that growth is for the better. To resume our argument: Whitehead’s creative advance and Rescher’s threefold definition offer the striking parallelisms—sequential structure/efficacy, eventuation/creativity and overlapping hierarchy/vision— that make even more obvious the basic bone of contention between them: at the end of the day, only the question of the conditions of possibility of novelty, emergence, innovation and creativity (PM 31-32, 35, 40, 74-78,

437

RESCHER STUDIES

80-82, 90, 93, 96-97, 98-103, 121, 133-134, 156, 159) make their respective approach idiosyncratic. Rescher basically claims that James, Bergson and Whitehead are wrong to argue for the bud-like theory of process (PM 14 sq.): “Once this atomistic doctrine is abandoned, the matter becomes simplicity itself. Nothing is more natural than that mini-processes should join and combine into macro-processes, and a process metaphysic that does not commit itself to a Whiteheadian atomism needs no special machinery to accommodate this fact because it allows reality to be seen as processual ‘all the way down.’ A more satisfactory approach is thus reflected in the doctrine of ‘synechism’ introduced under this name by C. S. Peirce, who defined it as ‘that tendency of philosophical thought which insists upon the idea of continuity as of prime importance in philosophy,’ with particular stress on the idea that ‘a true continuum is something whose possibilities of determination no multitude of individuals can exhaust.’ In the end, individuality can be conceived of in terms of unity of process itself.” (PP 3132)17 According to Rescher, they are furthermore wrong to dare to expand human’s experience to all possible experiences (cf. Whitehead’s “reformed subjectivist principle” in the light of Rescher’s discussion of “owned processes”) and especially to find an iota of liberty in all experiences, whatever its grade: liberty is confined to rational human agents (PPD 73-74). Besides the fact that there are traces of discontinuism in Rescher and that he misunderstands epochality for classical atomism, Whitehead would himself reply with a Kantian argument of sorts: first, becoming cannot be understood without disruption; second, there is nothing wrong with anthropomorphism (that needs to be departed from anthropocentrism) as long as it is a critical one . . . 4. CONCLUSION: METHODOLOGICAL PRAGMATIC IDEALISM AND PROCESS Ultimately, one should peruse Rescher methodological standpoint in order to understand the special blend of process thinking he has developed in his last works. Rescher’s advocation of processism cannot indeed be separated from his strictly methodological realistic pragmatism as it is correlated to his conceptual idealism: for instance, when he argues that “being is ultimately a matter of activity” (PPD 2) or that that infinity is no more than a matter of cognitive complexity (PPD 107), these are basically epistemic claims. Process is for him a transcendental idea that applies to the world

438

RESCHER ON PROCESS THOUGHT

qua known, not necessarily to reality an sich. His “juxtaposition” of pragmatism and processism is of course rooted in Peirce, James and Dewey (PM 14 sq. and PP 46-47) but Rescher furthermore creates a symbiosis with the epistemic analyses fostered in the Kantian idealist tradition in order to alleviate the anticognitivism of neo-pragmatism and “the knownothing prosaicism of the later Wittgenstein” (PM 165, cf. PP 96 sq.). Process thinking actually establishes a double cognitive inexhaustibility or opacity: ontological (linked with “being”) and existential (linked with “becoming”). On the one hand, there is no way we can exhaust the “being” of past events; on the other, there is an intrinsic opacity to all “evergrowing manifold of process[es]” (PM 132). Here lies the ultimate difference between the various forms of actuality and possibility, between concreteness and abstractedness. In a way, the Kantian noumenalization is thus reinforced and the need for a uniformity structure is more dramatic in such a chaosmos (cf. AW 53 sq.). Hence two complementary possible responses: Rescher’s development of a renewed coherence theory of truth that is geared with the utilitarist criteriology and Whitehead’s epochal theory of eventuation. Maxwellian electromagnetism (1873) and the emergence of psychodynamics18 put aside, the influence of Herbert Spencer (1820–1903) on process utilitarianism stands out. More precisely, Spencer’s Principles of Psychology (published in 1855, four years before Darwin’s Origin of Species) put into motion the classical “biological theory of knowledge”19 with an amazing claim: the structure of human intellect is “a priori for an individual” but “a posteriori for the whole species.” His core argument is fairly simple: the original function of knowledge is purely utilitarian because our mental apparatus is the product of the struggle for life, i.e., of our continual adjustment to the sector of reality important for survival purposes (the mesocosm). Since our cognitive functions are of empirical origin, they can have only limited applicability. Now it goes without saying that utilitarianism means here limited but real applicability: some form of necessity obviously seals the relativity of our categories. But Whitehead’s epochal trail still seems an option in this speculative atmosphere saturated with Kantian insights: in order to settle the controversy around noumenal causality, Rescher carefully distinguishes the genuinely experientiable (because constructed) causality and the noumenal ground demanded by the epistemological use of the Principle of Sufficient reason.20 Similarly, Whitehead articulates two types of processes, each exerting its own causation, as it were: transition describes the phenomenal,

439

RESCHER STUDIES

conditioned (hetero-nomic) flux of past events—being—whereas concrescence describes the unconditioned (auto-nomic) genesis of contemporary events themselves—becoming. More to the point: in both thinkers one finds the perennial idea that the reason of the world lies hidden in “something extramundane”—better: in some borderline, edge-like process that Whitehead calls concrescence. Furthermore, principles of uniformisation are for both theoretically postulated and knowledge-productive rather than discovered and knowledge-produced but practically meaningful if not downright experientiable in the case of the British. On the other hand, an interesting difference between Kant and Whitehead lies in the status given to space-time, that Whitehead understands as the surface-effect of a more fundamental, objective and plastic (i.e., neither subjective nor rigid),21 structure of uniformity and solidarity (the extensive continuum). J.-Cl. Dumoncel, J. Seibt, and G. W. Shields have, among others, aptly claimed that with Rescher’s process books, and subsidiarily with the collective AW, a major conceptual regrouping is taking place: the so-called continental and analytical trends are again reinforcing each other.22 From that perspective, Rescher’s works are most definitively at unison with powerful systematic philosophers such as Leibniz, Kant, Grassmann, Brentano, Peirce and Whitehead . . . 23

REFERENCES PRIMARY BIBLIOGRAPHY ID:

Inquiry Dynamics (New Brunswick, Transaction Publishers, 2000). NU: Nature and Understanding (Oxford, Clarendon Press, 2000). PM: Process Metaphysics (Albany, SUNY Press, 1996). PP: Process Philosophy (Pittsburgh, University of Pittsburgh Press, 2000). PPD: Process Philosophical Deliberations (Frankfurt, Ontos Verlag, 2006). SECONDARY BIBLIOGRAPHY AW:

440

Michel Weber (ed.), After Whitehead: Rescher on Process Metaphysics (Frankfurt, Ontos Verlag, 2004).

RESCHER ON PROCESS THOUGHT

NOTES 1

Cf. especially PM’s chapter X: “Process philosophy [ . . . ] has the virtue of selfsubstantiation” (PM 168).

2

One could mention by the same token the processualization of Thomas Aquinas advocated by Felt, Fetz and Clarke who claim that the Aristotelian and Thomistic concepts of substance—not Descartes’— should be understood relationally and are thus immune to process-like criticisms. Cf., e.g., James W. Felt s. j., “Whitehead's Misconception of ‘Substance’ in Aristotle”, Process Studies, Vol. 14, No. 4, (1985), pp. 224-236; Reto Luzius Fetz, Whitehead. Prozeßdenken und Substanzmetaphysik (Freiburg und München, Verlag Karl Alber, 1981); William Norris Clarke, s. j., The One and the Many. A Contemporary Thomistic Metaphysics (Notre Dame, Indiana, University of Notre Dame Press, 2001).

3

This matrix is adapted from A. Weekes’ sharp analysis in AW 223 sq.

4

PP 22’s definition would lead us in a slightly different direction, it says: “A process is an actual or possible occurrence that consists in an integrated series of connected developments unfolding in programmatic coordination: an orchestrated series of occurrences that are systematically linked to one another either causally or functionally.”

5

Compare with what PM 12-13 claims of Leibniz’ processism: “The whole world is one vast systemic complex of such active processual units. They are programmed agents—‘incorporeal automata’— developing in coordinated unison as individual centers of activity operating at different levels of sophistication within an allcomprising unified cosmic whole. [ . . . ] Leibniz developed a complex theory of nature as an integrated assemblage of harmoniously coordinated processes so that processes, rather than substantial objects furnishing the basic materials of his ontology.” See also J. Seibt’s Leibnizian remarks in her “Rescher” entry written for Michel Weber and William Desmond, Jr. (eds.), Handbook of Whiteheadian Process Thought, (Frankfurt / Lancaster: Ontos Verlag, forthcoming).

6

Alfred North Whitehead, The Aims of Education and Other Essays (New York and London, The MacMillan Company and Williams and Norgate, 1929), p. 107.

7

“A process is a sequentially structured sequence of successive stages or phases which themselves are types of events or occurrences (in the case of an abstract process) or definite realization of such types (in the case of a concrete process).” (PP 26)

8

Cf. PM 77’s quote of James’ Pluralistic Universe.

9

William James, The Principles of Psychology (New York, Henry Holt & Co., 1890), Vol. I, p. 243, quoted, e.g., in PM 38.

10

“Communication requires not only common concepts but common topics— shared items of discussion, a common world of selfsubsistently real “an sich” objects basic to shared experience. The factor of objectivity reflects our basic commitment of a shared world as the common property of communicators. Such a commitment in-

441

RESCHER STUDIES

NOTES

volves more than merely de facto intersubjective agreement. For such agreement is a matter of a posteriori discovery, while our view of the nature of things puts “the real world” on a necessary and a priori basis. This stance roots in the fundamental convention of a socially shared insistence on communicating—the commitment to an objective world of real things affording the crucially requisite common focus needed for any genuine communication.” (PP 92, cf. 98 sq.) 11

Cf., e.g., Nicholas Rescher, The Primacy of Practice. Essays Towards a Pragmatically Kantian Theory of Empirical Knowledge (Oxford, Basil Blackwell, 1973), pp. 91-92.

12

Quoting Plato, Symposium, 208A-B, Trans. by Benjamin Jowett.

13

“And new and different concepts is just that—new and different: there just is no matter of degree here.” (PP 88) “The key issues are ones of significance, centrality, priority, and emphasis. And in this regard the position of process metaphysics is that the interests of a just appreciation of the world's realities call for prioritizing activity over substance, process over product, change over persistence, novelty over continuity. Seen in this perspective, process philosophy does not—or need not— deny the reality and validity of the second members of these pairs, but rather maintain that in the order of significance they must be subordinated to the first.” (PM 31) “Realms are megaprocesses, as it were. And such mega-processes need not necessarily be continuous in space and time; it makes perfect sense to see all processes of the same sort (all pencil sharpenings, for example) as constituting not merely a processual thing-kind but also as thereby constituting a spatio-temporally distributed mega-process.” (PM 72)

14

In PPD ch. V, originally published as “Causal Necessitation and Free Will”, Process Studies, 35/2, 2006, pp. 193-206.

15

I.e., with works such as “La Théorie relationniste de L’espace” (1916), An Enquiry Concerning the Principles of Natural Knowledge (1919), The Concept of Nature (1920) and The Principle of Relativity, with application to Physical Science (1922).

16

Whitehead has gently shifted from the concept of extension, to the concept of extensive abstraction and finally to the creative relation of extensive connection. The works of his first epoch—Cambridge, U.K. (1880–1909)— share one common concern: to question the foundations of geometry and, thereby (roughly speaking, of course), to provide an account for the relatedness of all possible worlds. Formalism is a tool (an organon) to come to terms with reality understood from the standpoint of a relationist theory of space (cf. his 1905 “Theory of Interpoints”). The later Whitehead transcends this formal ontological standpoint with a proper existential ontological standpoint. Process and Reality’s purpose is to display the gearing of actuality per se or existence (which is subjective and qualitative) and of the various layers of potentiality or being (basically objective and quantitative) through a bipolar relation (the relation of extensive connection) operating on regions. Now, the point of inflexion between these two seemingly foreign perspectives is to be found in his London epoch, and more precisely in Principles of Natural Knowledge’s Part III, where the bipolar relation of extension operating on events is intro-

442

RESCHER ON PROCESS THOUGHT

NOTES

duced. Here geometry is explicitly studied qua physical science: the “concreteness” it addresses is no longer active as the goal of a priori principles but it is not yet the locus of the synthesis of the knower and the known. In London, Whitehead extends (no pun intended) his focus to space / time / matter with the help of the mereology axiomatized in the Principles of Natural Knowledge on the basis of a fundamental binary relation (irreflexive, assymmetrical, transitive) whose relata are events: the relation of extension. In doing so, he adopts a non-Euclidean—in so far as it is pointless—geometry. Of special significance is the fact that actuality is conceived as continuous and hence is understandable with a part-whole relationship. The application of the logic of relations to the perception of space and time allows him to bridge the gulf between the world of sense-perception and the world of science. In Harvard, Whitehead’s interest shifts to ontological uniformity, a question that he believes require a mereo-topology of sorts. Two facts explain this reform: on the one hand, Theodore de Laguna’s criticisms of the Principles of Natural Knowledge’s mereology; on the other hand, Whitehead's decision to “throw a match into the powder magazine”, i.e., to cross through the gates of metaphysics and hence to adopt an “epochal theory”. There is no need to specify here the exact status of this ontological atomism; suffice it to say that actuality per se is now discontinuous. The binary relation of extension whose relata are events is replaced by the binary relation of extensive connection (irreflexive, symmetrical, non transitive) whose relata are regions. Inclusion (irreflexive, assymetrical, transitive) is rebuilt from there. 17

Quoting Peirce’s Collected Papers; cf. PM 55-56 and PM 29: “On closer inspection, the idea of discrete ‘events’ dissolves into a manifold of processes which themselves dissolve into further processes.”

18

With the work of thinkers such as Herbart (1824), Weber (1829), Helmoltz (1859), Fechner (1860), Wundt (1878), Lotze (1884), Ward (1886), Münsterberg (1889) and, most importantly, Myers (1889–1895): it is by no means restricted to Freud’s Traumdeutung (1900).

19

On the development of “evolutionary epistemology”, see especially Milic Capek’s New Aspects of Time. Its Continuity and Novelties. Selected Papers in the Philosophy of Science (Dordrecht-Boston-London, Kluwer Academic Publishers, 1991).

20

Cf. Nicholas Rescher, The Primacy of Practice, op. cit., pp. 70 sq. and his Kant and the Reach of Reason: Studies in Kant's Theory of Rational Systematization (Cambridge, Cambridge University Press, 2000), pp. 23 sq.

21

Cf. also James: “Plasticity, then, in the wide sense of the word, means the possession of a structure weak enough to yield to an influence, but strong enough not to yield all at once. Each relatively stable phase of equilibrium in such a structure is marked by what we may call a new set of habits. Organic matter, especially nervous tissue, seems endowed with a very extraordinary degree of plasticity of this sort; so that we may without hesitation lay down as our first proposition the following, that the phenomena of habit in living beings are due to the plasticity of the organic materials of which their bodies are composed.” (The Principles of Psychology [1890], op. cit., I, p. 105).

443

RESCHER STUDIES

NOTES 22

Jean-Claude Dumoncel, “Critical review of After Whitehead”, in Michel Weber et Pierfrancesco Basile (sous la direction de), Chromatikon II. Annuaire de la philosophie en procès—Yearbook of Philosophy in Process (Louvain-la-Neuve: Presses universitaires de Louvain, 2006), pp. 261-279 (Dumoncel’s 1987 doctoral dissertation precisely dealt with Le système de Whitehead et la philosophie analytique); Johanna Seibt, op. cit. Cf also G.W. Shields, Process and Analysis. Whitehead, Hartshorne, and the Analytic Tradition (New York: State University of New York Press, SUNY Series in Philosophy, 2002).

23

Frederic Brenton Fitch (1908–1987) deserves a mention as well: cf. his “Combinatory Logic and Whitehead's Theory of Prehension”, Philosophy of Science, vol. 24 (1957), pp. 331-335. In his Yale 1934 dissertation, Fitch acknowledged J.B. Rosser, Alonzo Church and F. S. C. Northrop . . . Cf. also John W. Lango “Fitch” entry for the forthcoming Handbook of Whiteheadian Process Thought, op. cit.

444

How is Scientific Knowledge Economically Possible?: Nicholas Rescher’s Contributions to an Economic Understanding of Science James R. Wible 1. INTRODUCTION

M

ore than a century ago philosopher Charles Sanders Peirce asked a question which contributed to the development of American pragmatism. In his essays on pragmatism, Peirce rephrased Immanuel Kant’s famous question about whether knowledge is possible into a similar but more realistic question. Peirce asked whether there were any types of reliable statements that have informative value for human beings. In the wake of both Humean scepticism and the advent of Darwinian conceptions of evolution, Peirce restated Kant’s question in the following way asking whether any kind of synthetic statement is possible: Late in the last century, Immanuel Kant asked the question, “How are synthetical judgments a priori possible? . . . But before asking that question, he ought to have asked the more general one, “How are any synthetical judgments at all possible?” (C. S. Peirce, “Probability of Induction,” 1878, pp. 303-304).

But Peirce went beyond the bounds of the discipline of philosophy in his day and brought new perspectives to this question. Among the other disciplines which Peirce explored and which affected his conception of pragmatism was economics. This economic aspect of Peirce’s thought raises the possibility of another version of Peirce’s and Kant’s questions. Taking into account that the methods of the sciences were a significant part of the essays on pragmatism and also that Peirce was keenly interested in economic aspects of science and inquiry, an economic version of Peirce’s question can be formulated. His question—“How are any synthetical judgments at all possible?”—if reformulated in economic terms would be something like the main title above—“How is any scientific knowledge at all economically possible?” While philosophers mostly have not raised an economic version of

RESCHER STUDIES

Kant’s question about the nature of science, it could be the case that it is both a philosophically and economically significant question. In our time, the philosopher who has most considered the economic problems of science is Nicholas Rescher. For at least three decades, Rescher has authored articles and books on economic aspects of science. From his latest book in this area, Epistemetrics (2006) back to one of his first, Scientific Progress (1978b), Rescher has argued that economic aspects of science and inquiry are a significant part of a more comprehensive understanding of science. It is clear in reading these works, that Rescher has taken into account much of what Peirce wrote about economic aspects of science and human inquiry. Furthermore, going beyond Peirce, Rescher has made additional contributions to understanding the economic dimensions of science. It is also clear that Rescher gives an answer to the economic version of Peirce’s question in the main title of this paper. His answer is that scientific progress depends on scientific observation and it in turn depends both on the availability and types of resources required for scientific observation. Before further exploring Rescher’s contributions on economic aspects of scientific inquiry, it is useful to make some observations about an economic approach to science. In itself, it is quite remarkable that economic dimensions of human knowledge have been largely overlooked by most philosophers and scientists. It certainly suggests that many of them may think that economic dimensions of science are not a significant intellectual matter. Certainly based on the number of books he has written on economic aspects of science, Rescher would disagree with that point of view. However, there may be economic reasons for ignoring economic aspects of science. Conceptions of science and human inquiry without economic dimensions might not be as naive as one might initially imagine even from an economic perspective. For example, one might offer the interpretation that conceptions of science which ignore economic factors simply assume that science is costless. One justification for ignoring the cost of doing science would be an assumption that a great deal of science could be done at a negligible cost. If the search for knowledge is nothing more than appraising the truth status of a proposition in the mind of an inquirer in light of knowledge immediately available to that inquirer, then we may have nothing more than pencil and paper operations at negligible expense of resources.1 Another variation of the cost question would be an assumption of constant cost. For instance, a few decades ago economists typically assumed that cost curves would not rise as more of a product was produced. This is the idea of a horizontal cost curve. Now there is a general realiza-

446

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

tion that even if the cost curve of the private producer may not rise, there is certainly a realization that social costs may rise if negative external effects such as pollution are taken into consideration. So perhaps those who mostly pay no attention to economic aspects of science are simply like economists before environmental limits were recognized. They essentially assume that cost factors are constant as science expands the scope of its investigations with more scientists and more experimental research projects. A constant cost argument might be a reason for economic factors receding into the background of an understanding of science. Rather than zero, negligible, or constant cost, an entirely different cost scenario could be at work. Extremely high costs may be another reason for ignoring economic factors. If it is the case that some imagined science projects could be prohibitively costly, then it would be instinctive for scientists to dismiss them without further experimental consideration. Clearly it is economically rational to waste no time thinking about experiments that may be infinitely costly in terms of the contemporaneously available observational and computational tools of an existing scientific discipline. However, this possibility and the preceding considerations do not explain the nearly complete absence of economic questions about scientific inquiry. The science that actually gets done lies somewhere in between the special cases of zero and infinite cost. Good science is both scarce and costly so one would imagine that it would be informative to pay attention to the economic aspects of science. This is essentially what Rescher has done. 2. NICHOLAS RESCHER’S ECONOMICS OF SCIENCE Other than Peirce, the philosopher who has paid most attention to economic aspects of human knowledge is Nicholas Rescher. In 1978, Rescher published two books on the economics of science. One was Peirce’s Philosophy of Science and the other Scientific Progress: A Philosophical Essay on the Economics of Research in Natural Science. Those two books were followed by Cognitive Economy in 1989, Priceless Knowledge in 1996, and Eptistemetrics in 2006. The main aim of what follows is to present the most significant features of Rescher’s economic philosophy of science.

447

RESCHER STUDIES

A. Peirce’s Philosophy of Science While Peirce’s theories of science, scientific research, and the economic dimensions of science are given due reference in each of the books mentioned above, the one which most clearly shows Rescher’s indebtedness to Peirce is the first one, Peirce’s Philosophy of Science. This book is relatively short and consists of four article-length chapters. Taking the chapters in reverse order is useful for considering the economic themes of this essay. The last chapter concerns Peirce’s Coast Survey paper on the economy of research and is titled, “The Economy of Research.” In his paper, Peirce (1879) created an economic model of allocating funds to increase the precision of scientific research projects.2 The idea in Peirce’s model was that the researcher should allocate funds so that equal additional scientific value would be received from each additional dollar allocated to each research project. Departing somewhat from a narrow interpretation of Peirce’s economic model as one of utility maximization, Rescher reinterpreted it as a cost-benefit model for appraising possible research projects: Peirce proposed to construe the economic process at issue as the sort of balance of assets and liabilities that we today would call cost-benefit analysis. On the side of benefits, he was prepared to consider a wide variety of factors: closeness of fit to data, explanatory value, novelty, simplicity, accuracy of detail, precision, parsimony, concordance with accepted theories, even antecedent likelihood and intuitive appeal. But in the liability column, there sit those hard-faced factors of “the dismal science”: time, effort, energy, and last but not least, crass old money (Rescher 1978a, p. 69).

Near the middle of this chapter, after he had finished his interpretation of Peirce’s essay on the economy of research, Rescher rather emphatically portrayed how much Peirce’s economic project has been neglected and what value it could have for understanding science: To this idea of the economy of research—of cost-benefit analysis in inductive inquiry and reasoning—Peirce gave as central a place in his methodology of science as words can manage to assign. Yet no other part of this great man’s philosophizing has fallen on stonier ground. The Peirce bibliographies now run to over eight hundred entries and currently add over forty items annually. Yet in all of this flood of writing over the century since Peirce flourished, there is not one single item that devotes substantial attention to this aspect of his theory of science (Rescher 1978a, pp. 72-73).

448

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

Rescher continues commenting on the dearth of attention given to Peirce’s project of the economy of science: This is a great misfortune, for there is virtually no part of Peirce’s philosophy that is currently more relevant and is capable of rendering greater service to the solution of current disputes. Everything considered, it is no exaggeration to say that Peirce’s project of the economy of research is an instrument that can cut through many recent disputes about inductive reasoning like a hot knife through butter (Rescher 1978a, p. 73).

This chapter ends with illustrations of several philosophical problems such as Carnap’s total evidence requirement, Hemple’s paradox of the ravens, Goodman’s grue paradox, falsificationism, and others. In each case, Rescher maintains that economic considerations show how to make epistemic choices that would otherwise be more difficult. Other chapters in Peirce’s Philosophy of Science deal with the self-correctiveness of science, its compleatability, and the role of abduction in making science more efficient. These are themes that would be addressed more broadly in subsequent books by Rescher. B. Scientific Progress The next book to consider which also appeared in 1978 is Scientific Progress. From an economic perspective, this is Rescher’s longest and most detailed book. I also believe that it is his best book on the economics of science. In the beginning of the work, Rescher (1978b) acknowledges the Peircean origins of this research project. Obviously less attention is given to Peirce in this book than the previous one just discussed. The theme of the title is also very important. When Scientific Progress appeared, Thomas Kuhn had published his relatively recent book on scientific revolutions which led many to question whether scientific progress was an intellectually valid notion. An important aspect of Kuhn’s work is a conception of scientific revolutions. Scientific revolutions give the appearance of a theory of progress but they also lead to a thesis of incommensurability. This is the idea that the science done before and after a revolution has so little in common that one cannot meaningfully draw the conclusion that scientific progress has taken place. Rescher’s Scientific Progress can be taken as an argument against the extreme interpretations of Kuhn and later deconstructionists who would take the position that any conception of scientific progress must be abandoned. From an economic

449

RESCHER STUDIES

perspective, one wonders why society would devote such a high level of resources to the enterprise of science if its prospects for progress were so dismal. Rescher seems to sense that extremely negative views of scientific progress would imply the wasting of resources on a grand scale. A wellinformed democratic society would never support the misuse of resources on the scale implied by the strong incommensurability arguments. An economic point of view could provide for an intellectual middle ground between those who say science makes no progress and those who portray science as an efficient market place of ideas. The economic truth is somewhere in between. Science is neither as efficient as its most ardent proponents might argue nor as inefficient as its most extreme critics implicitly suggest. Rescher clearly sees that an economic perspective can make a contribution in understanding science from a philosophical point of view. His position is that science advances and that its advancement crucially depends on resources available for observation. However, observation escalates in complexity and cost. Consequently, cost escalation implies that the growth of scientific knowledge will slow down. This is a thesis of logarithmic retardation. This is especially true for the natural sciences where observation seems to go through stages of technological innovation that open up new domains of more complex phenomena. Scientific Progress is divided into five sections. In each section, one finds a description of broad patterns of scientific activity that an economist would describe as features of the economic landscape of science. Many of the patterns constitute stylized facts that perhaps no one else has called to our attention.3 Often they are presented in a manner that portrays the growth of science over time. Some of them are described with functional relationships or with examples illustrating an exponential pattern of change. These examples clearly evoke an economist’s conception of a production function and the theories of growth that have been developed from those functions.4 One gets the sense that the economic features of an important segment of the economy are being described here for the first time. The main claim of Scientific Progress is stated at the beginning of the book. Near the end of the first chapter, Rescher informs us: It will be the main thesis of this book that the latter-day prophets of doom who espouse a model of saturation or exhaustion are wrong—that the demise of science is grossly exaggerated, and the situation we actually confront is not one of termination of science but merely one of its deceleration (Rescher 1978b, p. 15).

450

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

In this introduction Rescher has already reviewed a number of different perspectives regarding the future of science. Over the past several centuries there have been a number of theories that maintained that scientific progress would end. However, some scientists like Kelvin and Newton have held that science will make advances without any limit except for the extinction of human life. The main factors leading to a slowing down of the growth of scientific knowledge are economic constraints: Now over and above the limits of man’s intellect and of time for humanity at large there looms the issue which concerns us here, that of the physicoeconomic limits on knowledg -yielding interactions with nature . . . In the final analysis, it is harmless to disregard intellectual or biological limits in a context where—as we hope to show—it is the physico-economic limits and limitations that play the decisive role (Rescher 1978b, p. 15).

In other comments, Rescher likens his thesis to that of a famous classical economist on the limits of population growth. He maintains that the thesis of Scientific Progress is “Malthusian in its structure” (p. 2). Furthermore the overall method of the book is conducted somewhat in the style of economic analysis: The discussion accordingly proceeds in the highly aggregate manner familiar from economics. In predicting the general movement of an economy—or even the more specific movements of its constituent industrial sectors—the economist need not say anything about the activities of a particular factory, the fate of a particular product, or the dealings of a particular company. His deliberations can, nay must, proceed on a far more generalized and aggregated level (Rescher 1978b, p. 4).

After introducing the main themes of the work and various views about scientific progress, the first part of the book ends with discussion of one of the stylized facts of science. Rescher maintains that science in some of its features has grown at an exponential rate doubling over time. Apparently the first person to note the exponential growth of science was the historian Henry Adams. For this reason this historical pattern of exponential scientific growth is termed “Adam’s Law.” Here Rescher provides descriptive evidence that the manpower or labor involved in science has grown at an annual rate of 6%, that scientific literature has been growing about 5% over two centuries, and that research and development expenditures related to science have grown as high as 10% per year after WW II. Near the end

451

RESCHER STUDIES

of this discussion of Adam’s Law, Rescher even gives us a simple equation which divides the resources used in science into two categories, labor and capital. This formulation clearly seems in harmony with the production function framework of growth studies in economics. Exponential growth implies the doubling of the activity over some time horizon. Even modest single-digit growth rates imply a doubling over 20 to 25 years. Near the end of the discussion of Adam’s Law, Rescher considers whether growth must decelerate and what that could mean for scientific progress. He even considers the possibility that resources may reach a steady state of zero growth. Here he clearly recognizes that this means that science will have a constant annual flow of new resources and that scientific activity will continue. However, it seems to imply that scientific innovation will also decelerate. In the second major section of Scientific Progress additional stylized facts about the deceleration of the growth of science are identified. For example, the cost of doing scientific research apparently escalates dramatically as a given field of research expands. Again, Rescher frames this phenomenon in economic terms: [T]here remains the economists’ uncompromising question regarding the actual structure of the relationship between resource investment and product output. In particular, the problem arises of whether, as science progresses, a fixed amount of effort continues to yield uniformly significant results, or whether a process of declining yields is operative in this respect . . . A great deal of impressionistic and anecdotal evidence certainly points towards the increasing costs of high-level science. Scientists frequently complain that “all the easy researches have been done.” The need for increasing specialization and the division of labor is but one indication of this (Rescher 1978b, p. 79).5

If the cost of doing science escalates as a field of research expands, over time ever more significant demands are placed on the shoulders of researchers. This was apparently noticed by the physicist Max Planck and is termed “Planck’s Principle of Increasing Effort.” An implication of escalating cost and effort in science means that the rate of achieving significant findings in science may slow down. Rescher calls this the “Law of Logarithmic Returns.” Here Rescher goes so far as to provide an implicit log function relating a measure of scientific findings, F, to total resources available for scientific research, R. He likens it to the Weber-Fechner law from the psychology of sensation. Although he does not graph this rela-

452

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

tionship, anyone who is basically familiar with log functions knows that findings will increase but at a slower rate than resources. This is a phenomenon economists call diminishing returns. It usually occurs when one resource cannot be increased in proportion to other resources. For example, total output functions for labor exhibit diminishing returns when the capital available to the worker is assumed to be fixed. Workers produce more but the rate of increase is dampened if non-labor resources cannot be increased in tandem with the additional efforts of the worker. In a subsequent chapter, the economic arguments that led to Planck’s Principle are applied to the question of significant findings. If science has grown at an exponential rate over most of the past two centuries, then concepts and publications of those concepts have grown as well. Here Rescher aims to focus on the broad pattern of truly important findings. He hypothesizes that important findings are related to the total number with an exponential equation. This relationship is named for Rousseau and termed Rousseau’s Law. Rousseau’s Law is the assertion that the number of important findings is approximately equal to the square root of total findings. Furthermore, the patterns of aspects of scientific inquiry described by Rousseau’s Law are consistent with the thesis of logarithmic returns as science grows in human and non-human resources. Moving to the next chapter, Rescher again restates the main theme of the book. Since this is his thesis, it would be logical to follow his example and name this stylized aspect of science after him, as “Rescher’s Thesis of Logarithmic Retardation” in science. Restating this thesis, Rescher maintains: The rising cost of scientific discovery at issue in Planck’s Principle combines the hypothesized onset of zero-growth condition in the resources available for scientific work to yield the deceleration of scientific progress as an immediate consequence. When one’s budget remains fixed while the price of commodities increases, one cannot but scale down one’s purchase (Rescher 1978b, p. 112).

At a later point, he gives another version of the Thesis of Logarithmic Retardation: Our deceleration-thesis rests neither on the claim that the well of undiscovered significant findings is running dry, nor on the claim that the limits of man’s capacities for inquiry are being exhausted. The slowing it anticipates is not theoretically inevitable for any such absolutistic reason, but rather is practically unavoidable for the merely economic reason that it is getting

453

RESCHER STUDIES

more and more expensive to run the increasingly complex machinery of scientific innovation (Rescher 1978b, p. 121).

The remaining three sections of Scientific Progress develop and support the main argument of deceleration in the growth of science. In the third section, Rescher argues that advances in the natural sciences require new data and new data processing capabilities. These in turn are dependent of the development of new technologies of observation and information technology. He formulates this as a “Technological Dependency Thesis” (p. 142). In the next chapter, he argues for domains of phenomena which are accessible with observational technologies geared to specific ranges and types of phenomena. As domains and ranges of phenomena become more difficult to observe, new and higher cost instruments of observation need to be created. New instruments and methods of observation lead to rapid and then to diminishing progress in science. Rescher describes the issue as a complex interaction of the mind of the scientist with his surrounding environment: What is at issue is thus not a one-sided thesis about the structure of the physical world itself, but a deeper and two-sided thesis about mind/nature interaction and the structure of our cognition in relation to nature. The pivotal point is that man as an inquirer is enormously efficient and effective. We are able to derive less and less cognitive benefit from the later steps in the explanatory process of moving further outward in parameter space precisely because we have been so efficient in the cognitive exploration of the phenomena that came to hand early (Rescher 1978b, p. 164).

The idea of successive levels of parameter space that become increasingly difficult to explore scientifically suggests another stylized fact about science. “Gore’s Law,” named for a late 19th century figure, states that the number of important findings in an area of research is proportional to the power available to the researcher to explore a domain or parameter space (p. 167). When new technology opens up new domains for observation, its power of observation is greater. New findings escalate at first and then eventually diminish. When science has been growing exponentially, this pattern is not as easy to see. But when resources available for observation eventually diminish, then scientific progress will inevitably slow down. In part four of Scientific Progress, Rescher likens the development of science to an arms race against nature. Nature presents itself as stratified

454

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

levels of phenomena that become increasingly difficult to observe. As the scientist moves to new and previously unstudied levels of phenomena, cost and effort escalate and advances become more difficult. At the micro level of individual research areas, the pattern of science may be cyclical as new domains are uncovered and lead to higher rates of important findings. Periods of rapid success are then followed by times of greater scientific difficulty as the pace of successful scientific research slow down. The cycles of progress and retardation could even grow longer as time passes. In part five, Rescher presents his most general conclusions about the prospects for the future of science. Also, he wants to address possible critiques of his thesis regarding the deceleration of science. His conception of one important critique of deceleration in the growth of science can be framed by making an analogy between science and the economy: Let us begin with an economic analogy, viewing science as a whole as an economy, its branches as industries, its problem-areas as firms. Now it is one thing to deal with the microeconomics of firms and quite another thing to deal with the economy as a whole, and it may well prove to be a complex and difficult issue to aggregate the partial situations into a synoptic whole. Discrepancies can easily arise—many firms or even industries might flourish while the economy as a whole stagnates, or conversely (Rescher 1978b, p. 222).

In the discussion which follows, Rescher clearly recognizes that dynamic new areas of science might be continually created dominating older declining areas of science so that his “Logarithmic Retardation Thesis” might not hold for the entire scientific enterprise. Here Rescher creates a taxonomic argument regarding new branches of science and concludes that it would be unrealistic to expect enough new areas of science to offset overall deceleration. In the last chapter, another possible critique of the deceleration thesis is considered. There Rescher maintains the “Logarithmic Retardation Thesis” against a “Kant Proliferation Effect.” This effect would result in accelerating or constant scientific progress if new scientific concepts and questions proliferate. As new unfathomable conceptual and theoretical frameworks are conceived in the minds and discourse of future generations of creative scientists, science could again grow at a higher rate. Near the end of the work, Rescher offers reasons for discounting this Kant Proliferation Effect and he again offers an economic rationale for his vision of science growing at a decreasing rate:6

455

RESCHER STUDIES

And the burden of the present course of argument is that the most pressing and immediate limitations to scientific progress are those operative on the side of physico-economic limits to the acquisitions and processing of the data themselves, rather than those relating to their strictly intellectual exploitation (Rescher 1978b, p. 236).

C. Cognitive Economy Eleven years after Scientific Progress appeared, Rescher’s third book on the economics of knowledge appeared. In 1989, Cognitive Economy was published. As with the earlier books, Rescher (1989, p. ix) begins by expressing his indebtedness to the “American polymath Charles Sanders Peirce” who “was the first to stress the role of economic considerations in the theory of knowledge.” Cognitive dimensions of science are the major focus of this book which is much shorter than Scientific Progress. Issues of cognition were developed previously, but not to the extent presented in Cognitive Economy. In the very first paragraph, Rescher asserts the crucial nature of economic factors in understanding human knowledge: The theme of this book is the importance of economic considerations in the conduct of our cognitive affairs. Knowledge has a significant economic dimension because of its substantial involvement with costs and benefits. Many aspects of the way we acquire, maintain, and use our knowledge can be properly understood and explained only from an economic point of view . . . Any theory of knowledge that ignores this economic aspect does so at the risk of its own inadequacy (Rescher 1989, p. 4).

Cognitive Economy is a relatively small book in seven chapters. The first chapter elaborates more fully the benefit-cost interpretation of Peirce’s economy of research. Rescher’s interpretation here is a much more detailed extension of themes from the last chapter of Peirce’s Philosophy of Science. In Cognitive Economy, Rescher maintains that the benefit of knowledge is that it gives us a better awareness of our environment and the human condition. He also asserts that cost effectiveness is an aspect of human rationality. Human rationality is the exercise of reason in making the best available choices including considerations of cost. One of the most interesting sections of the first chapter is an economic discussion of risk-taking. Here Rescher develops an argument which implies that scepticism is an economically irrational attitude toward risk-taking. The argument makes use of conceptions of error similar to those in the discipline of

456

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

statistics. One type of error is to reject concepts which are informative which should have been accepted but were not. The other type of error is to accept concepts which turn out to be wrong and should have been rejected but were not. To help us imagine how these two risks work together, he creates a graph which visually appears quite similar to the one that Peirce created in the “Note on the Economy of Research.” There are significant differences between the graphs, but each one tells a story of balancing benefits per unit of a costly alternative. Rescher’s graph is reproduced as Figure 2 and Peirce’s as Figure 1. ____________________________________________________________ Figure 1 Peirce’s Bi-directional Utility Graph for Additional Resources Allocated to Two Research Projects from the “Note on the Theory of the Economy of Research,” 1879

____________________________________________________________ In Peirce’s graph, two research projects are represented. The first project is read from the left hand side of the graph and the second from the right hand side. The additional value of greater precision in a research project diminishes for each extra dollar of expense as more funds are allocated to the project. The optimal allocation of funds occurs where the two curves intersect. In his graph, Rescher imagines an inquirer conceiving of all situations of risk. Suppose all risk situations for creating knowledge that are known to the scientist are represented on the horizontal axis. As we move from left to right, the inquirer initially encounters the situations of the first type

457

RESCHER STUDIES

of risk. These are the situations with the least likelihood of erroneous outcomes and the most significant potential for informative value. Moving further to the right are found the situations of the second type with the greatest risk of being false and the least expected informational value. If the first type of error is graphed, it greatest on the left side of the graph and diminishes as we move to the right. For the second type of error, the converse holds true, it increases as we move from left to right on the graph. From an economic perspective, maximizing what humans know would entail minimizing risk of both types of error. This would be the situation illustrated where the two curves of error intersect. Rescher (1989, p. 18) ____________________________________________________________ Figure 2 Rescher’s Analysis of the Two Risks of Rejecting a Valid Hypothesis and Accepting an Invalid Hypothesis from Cognitive Economy, 1989, p. 19

____________________________________________________________ concludes “the reasonable thing to do is to adopt a policy that minimizes misfortunes overall.” After analyzing the graph, he continues applying the implications of this model to two very different cognitive attitudes towards risk. Those whose attitudes are represented towards the left hand side of Rescher’s graph are identified as sceptics. Those whose attitudes are represented towards the right hand side of the graph are identified as syncretists.7 Sceptics would waste cognitive resources by failing to pursue the most fruitful opportunities for gaining knowledge due to being too risk

458

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

averse. Syncretists would waste cognitive resources on far too many situations where there is little expected gain. They would be “risk lovers” as economists use that term. Sceptics would under-spend on scientific investigation and syncretists would over-spend. Both patterns would have an adverse impact on the growth of scientific knowledge. Near the end of the discussion of risk and scepticism, Rescher (p. 29) presents the options in the form of a 2X2 pay-off matrix like the game theorist would use. It appears to be a game of a single player against nature. In the two columns of the matrix, he has payoffs for acting and for not acting on a belief. In the two rows he has outcomes for guessing correctly and incorrectly. His point is that the pay-offs appearing in one column, the one for not acting, are zero regardless of whether one’s guess is correct or not. However, the pay-offs in the other column are both positive. There is value to making a guess, even if it turns out to be wrong, as long as the excesses of the syncretist are avoided. Negative information narrows future choices even if the pay-off is less than guessing correctly. The second and third chapters of Cognitive Economy explore more subtle and complex aspects of science than Rescher has taken up in the previous two books. Chapter two takes up the theme that for science to be successful in economic terms it needs to foster attitudes and patterns of trust and cooperation. One reason for this is that information must be shared for science to be successful. Also, the creations of the researcher are intellectual property. The scientists needs to work within institutional arrangements that give appropriate credit for innovation and also allows other researchers to be quickly informed of the content of those new creations. Beyond the proper recognition of intellectual innovation, scientists need to be trustful of each other. Here Rescher uses an analogy with banking: [W]e proceed in cognitive matters in much the same way that banks proceed in financial matters. We extend credit to others, doing so at first to only a relatively modest extent. When and if they comport themselves in a manner that shows that this credit was well deserved and warranted, we proceed to give them more credit and extend their credit limit, as it were. By responding to trust in a responsible way, one improves one’s credit rating in cognitive contexts much as in financial contexts (Rescher 1989, p. 38).

Shortly after these comments, Rescher notes that similar aspects of trust operate in most professions such as medicine and law. In economic terms, the reason why so much trust and cooperation is required is that complex social processes typically are characterized by asymmetric information.

459

RESCHER STUDIES

Asymmetric information occurs when the buyers and sellers or the two or more parties in a strategic game have significantly different information sets about shared economic circumstances. An awareness of asymmetric information now characterizes many of the sub-disciplines of economics. It is a dominant aspect of research in finance and banking and in understanding specialized knowledge professions like law, medicine, and accounting. The knowledge structures of science and the professions are all complex, specialized, and create informational difficulties for non-specialists. Asymmetric information also leads to the theory of implicit contracts. In the context of a system of common law and complex economic processes with asymmetric information, unique arrangements often evolve that allow mutually beneficial and efficient transactions to take place. In terms of economic processes, arrangements for complex informational and knowledge situations may start out as tacit or oral arrangements. These are implicit contracts. Implicit contracts are recognizable patterns of economic processes that get established over time. These avenues of cooperation often start out spontaneously and may gradually gain the force of law. As time passes, eventually some of the most standardizable aspects of such arrangements take the form of explicit legal arrangements. Since the transactions patterns of science are as complex as the most complicated ones in the economy, one would expect implicit contracts to emerge in science as well. Although the term itself cannot be found in Cognitive Economy, in all other respects Rescher makes the case that implicit contracts are operative in science:8 Only through cooperation based on mutual trust can we address issues whose effective resolution makes demands that are too great for any one of us alone. In the development and management of information, people are constantly impelled toward a system of collaborative social practices—an operational code of incentives and sanctions that consolidates and supports collective solidarity and mutual support. In this division of labor, trust results from what is, to all intents and purposes, a custom consolidated compact to conduct their affairs in friendly collaboration. If its cognitive needs and wants are strong enough, any group of mutually communicating, rational, and dedicated inquirers is fated in the end to become a community of sorts, bound together by a shared practice of trust and cooperation, simply under the pressure of its economic advantage in the quest for knowledge . . . It can emerge for reasons of prudential self-interest alone (Rescher 1989, p. 43).

460

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

The next chapter continues with these themes of the efficiency of trust and cooperation within science. The sense of implicit contracts is extended to the communications processes of science. Accuracy and honesty tend to prevail in science because deceit is so harmful and wasteful of the resources and time of the scientific community. The economics of trust and communication is also at work in creating many of the unique institutional arrangements of science as compared to other sectors of the economy: The same economies of scale, resulting from the efficiencies of mutual access, that draw people together in cities draw scholars and scientists together in universities, research centers, institutes, academies, and professional associations. Had we not inherited such collectivities, we would not have to reinvent them. And much the same holds good for professional conferences, journals, preprint exchange networks, and the like (Rescher 1989, p. 58).

In the remaining four chapters of Cognitive Economy there are a number of linkages of aspects of science to concepts from economic analysis. In one chapter, Rescher claims that the idea of scientific importance is intertwined with an economic conception of rationality. With regard to choices that scientists need to make, he remarks that “the intelligent allocation of effort, and of resources in general, is a critical factor for rationality at large” (p. 81). In another chapter, he describes the economic efficiencies that result from the systemizing of scientific knowledge. Harking back to Peirce’s idea of abduction he also remarks that:9 Any inductive process is inherently chancy. Induction is always a matter of guesswork, and its results are always at risk to further or better data. But of course what is involved is responsible, rather than wild, guesswork—rational conjecture rather than fanciful speculations (Rescher 1989, p. 85).

In the last chapter of Cognitive Economy Rescher asserts again that economic aspects of empirical research will lead to problems for scientific progress. He again emphasizes the escalation of the technology for and costs associated with making ever more complex observations. Moreover, there is less use of terms such as “deceleration” and “logarithmic retardation” of the growth of science as compared to Scientific Progress. Also, the focus on growth rates of various aspects of science is absent. However, the overall theme of the importance of the economic constraints facing scientific inquiry remains:

461

RESCHER STUDIES

The technological/economic requirements of scientific progress mean that we will never be able to advance the project to our total satisfaction . . . As best we can tell, the limits of science are economic, we reach them not because we have exhausted the novelties of nature but because we have come to the end of our economic tether (Rescher 1989, p. 149).

D. Priceless Knowledge Less than ten years after Cognitive Economy, another book appeared in what clearly amounts to a series of books on the economics of scientific inquiry. In 1996, Priceless Knowledge: Natural Science in Economic Perspective was published. Priceless Knowledge combines and expands the themes of Scientific Progress and Cognitive Economy. It’s rhetorical style is more like that of Cognitive Economy but it takes up more of the economic issues like Scientific Progress. Mostly omitted from Priceless Knowledge are the details of the various stylized facts portraying the broad economic patterns of scientific inquiry as found in Scientific Progress. Instead there are more qualitative summaries of the patterns of economic activities in science and sharper philosophical arguments. My guess is that a philosopher might prefer Priceless Knowledge and a social scientist Scientific Progress. In the preface to Priceless Knowledge, Rescher tells us that his goal is “to give a unified expression to an economically realistic view of the future prospects of natural science” (p. vii). In the introduction, he tells us that “natural science is by nature an incompletable project” (p. ix). Themes from Scientific Progress and Cognitive Economy appear again in revised form. Rational scientific inquiry requires scientists to make economic choices. Science will continue without an end as long as humans survive as an intelligent species. Technology will escalate in complexity and cost. And scientific discovery will decelerate. Of particular note is a restatement of the law of logarithmic returns for significant findings in science. In Scientific Progress the relationship was asserted quite specifically in terms of a square root between significant and total findings. In Priceless Knowledge, the pace of scientific progress is imagined simply as deceleration rather than as logarithmic retardation. There is also a discussion of whether scientific realism is adequate as an understanding of what scientists do. There are two chapters of Priceless Knowledge which share no overlap with Rescher’s three previous books on the economics of science. In one chapter, Rescher argues that we cannot predict the future of science. This

462

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

subject would soon be more fully treated in an entirely separate book.10 In the other chapter, he raises the question of whether a scientific enterprise comparable to ours will ever appear somewhere else in the universe. Here Rescher goes through extensive stylized multiplication factors and offsetting fractions so that there are enough details for an interesting but speculative conclusion. He first summarizes the prospects for other planets like ours in the universe. He estimates that there should be 1022 planets in the universe possibly in some respects like earth. This is a huge number and it is obvious that many of them may not be good candidates for intelligent life and the emergence of science. At this point, Rescher takes the reasons for decreasing the number of planets where intelligent life and science might emerge and formulates them as fractions. These fractions relate the proportion of planets similar to earth to those where life itself might appear, where a temperate climate might emerge, where intelligent species might evolve, where culture and technology would develop, and finally planets where intelligent life rises to the creation of science and epistemology in something like their present forms. All of the fractions when multiplied together amount to 10-22. Obviously if the multiplication factors, 1022, and the fractions, 10-22, are multiplied together their product is 1. If everything is taken as presented, this amalgamation of stylized facts about the universe present an argument that humans are alone in the universe.11 Furthermore, it implies that science and epistemology as practiced here on earth may happen no where else in the universe. The overall sense of this highly speculative calculation ends with the implication that humans should not hope that another intelligent species somewhere else in the universe may have resources to do better science than we do. A second implication is that science creating species in the universe may be quite scarce. If science is going to be done anywhere in the universe, most likely it will be done by humans on earth. Since our resources and our numbers are limited, this means that all of the possible science that will be done in the universe will be done on earth. If this is the case then the economic limits of science being done on earth are also the economic limits of science being done anywhere else in the universe. 3. EPISTEMETRICS AND PHILOSOPHICAL ARGUMENTS FOR THE LIMITS OF SCIENCE Just a few years ago Rescher’s most recent book on the limits of science appeared—Epistemetrics was published in 2006. If you run eyes over the

463

RESCHER STUDIES

contents, it is obvious that some if not many of the themes of the previous four books are present. However, economic themes nearly disappear receding into the background. There is almost a total absence of the crisply phrased claims that economic limits are slowing the growth of scientific knowledge. Given that Rescher has authored four books with clear economic themes relating to cognition and the development of scientific knowledge, it seems odd to find a fifth book with economic factors deemphasized. Perhaps the reason for this is that specialists in one discipline do not like to have their ideas developed in terms of another discipline. In itself this is an economic dimension stemming from an intellectual division of labor. If knowledge is specialized in each discipline, there is so much invested in the ideas, intellectual constructs, vocabulary, specialized terms, and results of one discipline that there may be significant disciplinary incentives against considering the insights of another discipline. For economic reasons, economic speculations about science could be significantly diminished in Epistemetrics. Instead of statements favoring an economic understanding of our cognitive endeavors, in Epistemetrics one finds the case being made for a new discipline for studying science. This discipline would study the quantifiable patterns of the scientific community and its processes of inquiry. Most likely it would be a sub-discipline of philosophy. Obviously the title of the book would be the name of this new discipline. Rescher imagines the birth of epistemetrics as a new discipline in the following way: The discipline represented by the domain of inquiry to which the present book is addressed does not yet exist. Epistemetrics is not yet a scholarly specialty. To be sure, as regards scientific information in specific there is a discipline of scientometrics, represented by a journal of that name. But while this book too will keep scientific knowledge in the foreground, various of its key principles . . . hold every bit as much for our knowledge of every-day life matters as they do for the natural and human sciences . . . And so, seeing that epistemetrics is not as yet an established field of inquiry, the present discussion can do no more than offer a preliminary glimpse into the nature of such a discipline (Rescher 2006, pp. ix-x).

Rather than framing the study of science with economic ideas, in Epistemetrics, Rescher gives priority to the ideas from philosophy rather than economics. Aspects of conceptions of knowledge from the ideas of Kant and Leibniz are given priority over his own and Peirce’s on the economics of science. Although he begins with a discussion of the trade-off between

464

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

the security of our knowledge and its precision, which evokes some of the themes of the previous books, it does not take long to take up some of Kant’s ideas regarding human knowledge. In chapter two, Rescher takes up Kant’s thesis that human knowledge is rational to the degree that it is systematized. Apparently Kant held that knowledge must be systematized in an organic way as parts of an entity are to related to the whole. Cognitively significant knowledge must be part of a system of knowledge. Pieces of information need to be related to each other and prioritized systematically before one has knowledge. Such systematized knowledge is often organized hierarchically as well. Knowledge consists of nested aggregations of concepts and information that are located within other broader systems of knowledge. One can go from the study of specific scientific problems to sub-disciplines, to specialties within a discipline, and to groups of related disciplines. The last chapter of Epistemetrics also takes up ideas attributed to Kant. This chapter takes up ideas of limits and finitude and applies them to the discovery of knowledge. Here Rescher argues that human minds and the subject matter that all of the sciences aim to explain are both finite. However, that does not mean that finite creatures will one day complete the study of a finitely limited subject matter. Rescher following Kant argues that the capacities for inquiry by humans is so limited relative to the complexity of the phenomena of the universe, that humans will never end their quest for knowledge. As Rescher puts it, human knowledge is limited but unbounded: “Reality in all its blooming buzzing complexity is too rich for faithful representation by the recursive and enumerable resources of our language” (pp. 95-96). In the next to last chapter, Rescher presents an argument that goes back to Leibniz regarding what can be expressed with symbols in language and mathematics. Although symbolic representation is in principle infinite, as a realistic matter humans are limited in how many sentences and equations they can write, read, and comprehend. The question is whether there is a limit to how much humans can know. Arguing from the idea of an upper limit for humans, Rescher maintains that an upper bound exists on the quantity of symbolically represented ideas that can be accessed by human beings: “The moment ones sets a realistic limit to the length of practicably meaningful sentences one has to realize that the volume of the sayable is finite—vast though it will be” (p. 77)). In later paragraphs, Rescher holds that it was Leibniz who maintained that human understanding cannot keep up with the detail and richness of the universe in which we live.

465

RESCHER STUDIES

Other chapters of Epistemetrics provide accounts of some of the broad stylized facts of human inquiry. Often they appeared in one of his previous books, but here they are always presented in a new way. Kant’s principle of concept proliferation, logarithmic retardation of important scientific findings, Planck’s principle of escalating effort, Adam’s thesis of exponential growth, the deceleration of qualitative results, and the growing complexity of cognitive taxonomy are here considered again and offered as illustrations of what could be gained by a new discipline of epistemetrics. 4. PREDICTION, COMPLEXITY, AND PLURALISM: IMPLICATIONS OF RESCHERS’S PHILOSOPHY OF SCIENCE FOR ECONOMICS Certainly it is clear that Rescher has devoted a significant fraction of his professional career in thinking about the economic limits of science and the implication of such limits for philosophy of science. Depending on how one interprets Epistemetrics, five books have been written explicitly addressing the limits of science from both economic and philosophical perspectives. However, one should not get the impression that the cross fertilization of thought is one directional. There is much that economists could learn from Rescher’s approach to philosophy and from his economic understanding of the stylized patterns of scientific research. One area where economists could learn from Rescher concerns an emerging new, but small field in economic methodology known as the economics of science. Economic methodology is the field of economics which historically has been concerned with relating conceptions of economic science to philosophy of science. Within economic methodology only in the last decade or so has a new economics of science emerged as a sub area relating economic ideas about science to those in the philosophy, history, and sociology of science.12 One main difficulty that emerges in an economics of science is the problem of the discipline of economics. An economics of science in many respects would be an economic theory of economics. This would pose special conceptual problems for an economics of science. This problem is an issue of reflexivity or applying economic ideas to the discipline itself.13 Another difficulty is that economics as a discipline has been more inclined toward behavioral theories. This implies that cognitive interpretations of economics have been difficult to make consistently. Rescher’s writings on cognitive dimensions of knowledge as they appear in the five books of his economic approach to philoso-

466

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

phy of science should be of interest to economists and methodologists. Also, his cost based interpretation of the technological limits of scientific observation and its implications represent a unique contribution to the economics of science that no economist has made. Another way economists could learn from Rescher is to read other more broadly focused books bridging his economics of science and his broader approach to philosophy. Here there are three books which economists could benefit from reading. The books I have in mind are Predicting the Future: An Introduction to the Theory of Forecasting (1998a), Complexity: A Philosophical Overview (1998b), and Pluralism: Against the Demand for Consensus (1993).14 The titles of all three books have relevance to economics as it is presently practiced. Economists have devoted considerable resources to forecasting many and various patterns of detail at the micro level and the overall pace and pattern of the business cycle and economic growth at the macro level. Economic phenomena clearly are complex phenomena as that term is now being used. Also, economists have never been able to agree on the nature of the science or its most important policy implications. The three books taken together are in many respects interrelated and quite complementary. They cover many of the issues that would be needed in fashioning a fairly comprehensive methodology of a social science like economics. In Predicting the Future, Rescher aims to create a general philosophical theory of prediction. Prediction is an issue of great interest to economists. Issues surrounding prediction have been a central part of economics since the rise of econometrics in the 1930s, the advent of macro forecasting in the 1950s, and in econometric methodology. In the book, Rescher aims to develop a theory of empirical inquiry that takes a middle road between the excessive optimism of positivism and operationalism and the pessimism of extreme post-modern views of science. This middle road is essential because the world we live in is pervaded by chaos and chance. In such a world, prediction is not an easy matter and forecasting models may be quite narrow in their implications for the future. There are obstacles to successful prediction. Uncertainty is taken as an epistemological barrier and chance, chaos, and complexity are taken as ontological obstacles to predictability. The overall result is that we live in a world which is only partly predictable. At one point, Rescher even criticizes Milton Friedman’s essay on positive economics and his claim that economic models need to be judged only on their predictive accuracy even if they are quite unrealistic in their assumptions. With regard to the prospects for highly accurate pre-

467

RESCHER STUDIES

diction in economics, Rescher provides this judgment: For what makes prediction in economics . . . .a particularly difficult and skittish matter is the fact that the behavior of such complexly interactive systems as human free agents operating in social settings just does not exhibit the uniformity and regularity of behavior possessed by, say, carbon atoms. The actual behavior of a real-life economy is faced with contingencies of innumerable sorts—political, sociological, meteorological—whose aggregate effect is that even the best-made economic models cannot engender accurate forecasts. Choice, chance, and contingency undermine the stability of economic processes, so that the prospect of precise and secure prediction is severely compromised (Rescher 1998a, p. 198).

Another book which could be read to advantage by economists is Rescher’s Complexity (1998b). Aspects of complexity have already been introduced above and have been suggested as a reason for some of the difficulties that economists face with prediction. In Complexity Rescher sets out to give a philosophical account of nearly every aspect of complexity. Rescher recognizes that Peirce was an original intellect who theorized that the world moves towards more complicated patterns of diversity and complexity.15 Complexity is an aspect of the real world. There are three modes of complexity—ontological, epistemic, and functional. With regard to ontological complexity, things, entities, and processes in the real world seemingly exhibit structural and organizational patterns of complexity. The world seems to change as new patterns of order and pattern evolve, new patterns of individuals emerge, and whole new levels of phenomena appear. With regard to epistemological complexity, human theories of the world about us also grow and develop in a hierarchical fashion mirroring some of the complexity of the world being studied. With regard to functional complexity, the common sense knowledge of life outside of science is also dominated by organizational arrangements, structures, and computational processes which exhibit an increase in complexity. The overall implication of complexity is that it places limitations on what can be known. This means that scientists and others need to make economic choices that meet the requirements of what ever problem is being encountered. The third book in this interrelated trilogy is Pluralism. Rescher begins by considering various conceptions of philosophy and science which hold that a credible intellectual position should be the product of consensus. In different ways Habermas, Rawls, Rorty, and Kuhn offer such positions. However, Rescher maintains that science is pluralistic rather consensual.

468

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

A major reason for scientific pluralism is that human beings have considerable experiential diversity. In the domain of science this diversity results in scientific diversity. Although there is one world and one universe behind our experiences, our capacity for differing experiences and conceptualizations of that experience imply an unlimited variety of descriptions and ideas about our world. It is because of evolutionary complexity and the limitations it implies for what can be known, that pluralism is a consequential result. Next Rescher turns to the two philosophical positions previously encountered with Figure 2 from Cognitive Economy. Those positions are portrayed as ways of dealing with the limitations of what can be known— scepticism and syncretism. Scepticism is the thesis that none of the ideas about our world should be accepted while syncretism is an extreme form of relativism that every alternative idea is equally valid. For Rescher pluralism is based on “perspectival rationalism” (p. 80). In this view it is rational for one to believe that a particular theory is better than its rivals but one must realize that others may arrive at support for an alternative theory in a similar way. Thus it is rational to take one’s own position as best among rivals and realize that others, because of complexity and the limitations of empirical evidence, will go through a similar process and rank another theory as the best among available alternatives. Rescher asserts: Pluralism holds that it is rationally intelligible and acceptable that others can hold positions at variance with one’s own. But it does not maintain that a given individual need endorse a plurality of positions . . . That others agree with us is no proof of correctness; that they disagree, no sign of error. A pluralistic diversity of conflicting positions certainly is no basis for scepticism” (Rescher 1993, p. 89).

Clearly anyone familiar with the discipline of economics knows that economists are typically divided into two or more schools of thought in nearly every important subdivision of the discipline. If one believes that social phenomena are complex phenomena and lend themselves to alternative theories, theoretical vocabularies, and models, then we should expect economics to be a scientific discipline with a high degree of pluralism. The historic patterns of disagreement and pluralism of economics need not be interpreted as lessening the scientific character of economics. Instead, the pluralism exhibited in economics in light of Rescher’s contributions needs to be viewed as a consequence of the inherent limits all inquirers face. Those limits are ontological, epistemological, functional, and economic—where the economic limits of inquiry have been

469

RESCHER STUDIES

highlighted in so many of Rescher’s books on the economics of science that have been the focus of this essay. 5. THE SIGNIFICANCE OF RESCHER’S ECONOMIC PHILOSOPHY OF SCIENCE Nicholas Rescher has written five books that relate to an economic version of Peirce’s and Kant’s questions about the nature of human knowledge. To the question which forms the title of this essay—How is Scientific Knowledge Economically Possible?—Rescher provides an extended answer. The world in which we live is a partially stable array of hierarchical and nested patterns of natural and social phenomena. These patterns can be observed often with complexity, with great difficulty, and with limited resources. However, instruments of observation and computation are required to document the most important properties of those patterns. The instruments of observation are often resource intensive, expensive, and usually quite specific to a narrow domain or range of phenomena. This means that a fundamental aspect of science faces economic limits. But economic limits are not the only significant limits that science faces. Human concepts and our intellectual capacities are also limited. Our cognition is limited, our capacity to create theoretical and observational vocabularies is limited, and our ability to agree on which theories are best is also limited due to the limitations of evidence in allowing us to winnow theories at any point in time. Two other aspects of our world and science imply that science will be inherently pluralistic. The minds of professional scientists will proliferate alternative scientific theories faster than relevant evidence can be produced. Also, the properties of the natural and social worlds are so vast relative to our scientific abilities that science may be unbounded and continue indefinitely into the future. In several of his books, Rescher concludes that the economic and cognitive limits of science suggest that scientific progress will continue but decelerate. Mostly this is in opposition to the thesis that someday science will end—that science has discovered all of the major phenomenal patterns of the natural and social worlds that have already come into existence. Rescher rejects this view maintaining that scientific progress will continue indefinitely into the future. Rescher also portrays many quantifiable patterns of scientific activity which an economist would recognize as important stylized facts of science. Few others have presented so rich an informational picture of the patterns

470

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

of the scientific enterprise. These stylized facts give us an idea as to how synthetic knowledge about the world in which we live is developed. In his last book, Epistemetrics, Rescher mentions very little about the high costs of technology and instruments of observation limiting the pace of scientific progress. In the years since his first book on economics of science, Scientific Progress, was written three decades ago, science seems to have undergone a revolution in its observational technologies. There are so many new ways to make images and visualize data in the various sciences. There are new images of the vast astronomical patterns of deep space, neurologists can now watch our brains function with new brain imaging equipment, the human body has been visualized in path-breaking ways, and advances in computer algorithms now allow searches for patterns in large data sets. Virtually every discipline in the natural and social sciences has acquired significant new capacities for observation and/or computation over the past three decades. In econometrics, more data is available and more models of the data can be tested than was humanly possible just a few years ago. In mathematical economics, computer programs now solve systems of equations that might not have been solved a generation ago. At this point, we do not know if the enhanced observational capacities we are experiencing are temporary or whether they will last indefinitely into the future. Isn’t it interesting that Rescher was writing about the deceleration of science in the late 1970s just before it apparently accelerated again? But will this observational revolution and acceleration in the growth of science continue indefinitely? Rescher’s theory of science suggests that this recent increase in the growth rate of science may slow down in the future. He is clear that the observational capabilities of the sciences go in cycles. Rescher’s ideas about the limits of science predict that science may decelerate and grow at a much slower pace some day in the future. However, if there is evidence of an acceleration of the observational capacities of science over the past few decades, that would make it more difficult to hold the thesis that scientific progress will end at some point in the future. Of what significance are Rescher’s books that have been the focus of consideration in this essay? The five books of Rescher’s with economic themes certainly represent one of the most important contributions to a small, but emerging field known as the economics of science. Until Rescher and a few economists began thinking about economic aspects of science, economic dimensions of scientific inquiry mostly were mostly ignored over the past century since Peirce wrote his short essay on the econ-

471

RESCHER STUDIES

omy of research. Without considering the economic aspects of science, philosophers, economists, and others who studied science were and are like those who ignored the environmental implications of economic activity. Both the environmental and the economic implications have proven to be too significant and too costly to be absent from an intellectually comprehensive approach in understanding science. This is true from either an economic or philosophical point of view. A further area of significance for Rescher’s work concerns the last three books considered above, the ones with the themes of prediction, complexity, and pluralism. Those three books taken together would make an excellent starting point for considering the contemporary methodological problems of the discipline of economics. The subject matter of the various subdisciplines of economics are obviously complex phenomena. The complex phenomena of economic processes are difficult to predict. The weakly predictable complex phenomena of economic processes tend to generate competing theoretical frameworks in the many sub-disciplines of economics. And the competing theoretical frameworks in turn create the conditions for pluralism. If the empirical results of the various theories are very weak in comparatively determining which theoretical approach is better than its rivals, there should be a great deal of pluralism across the discipline of economics. These disciplinary patterns just described are all implications of Rescher’s understandings of the roles of prediction, complexity, and pluralism in scientific investigation including economics. Finally, as Rescher certainly recommends, individual economists would do well to adopt an attitude of perspectival rationalism, that each individual economist holds a theoretical position which she or he considers to be the one best supported by available evidence, even if others disagree with that appraisal of the evidence.

REFERENCES Hands, D. Wade, Reflection without Rules: Economic Methodology and Contemporary Science Theory (Cambridge, U. K.: Cambridge University Press, 2001). Jevons, W. S., The Theory of Political Economy, 5th ed. (London: Macmillan, 1957 [1871]).

472

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

Okun, Arthur M., Prices and Quantities: A Macroeconomic Analysis. Washington, D. C.: The Brookings Institution, 1981). Peirce, C. S., “The Probability of Induction,” Writings of Charles S. Peirce: A Chronological Edition, 1872-1878, vol. 3, C. J. W. Kloesel, ed. Bloomington: Indiana University Press, (1986[1878]), pp. 290-305. ———, “Note on the Theory of the Economy of Research,” United States Coast Survey for the fiscal year ending June 1876, U.S. Government Printing Office, 1879, reprinted in, Writings of Charles S. Peirce: A Chronological Edition, 1879-1884, vol. 4. (Max Fisch ed. Bloomington: Indiana University Press, 1986[1879]), pp. 72-78. Rescher, Nicholas, “Peirce and the Economy of Research,” Philosophy of Science, vol. 43, (1976), pp. 71-98. ———, Peirce’s Philosophy of Science (Notre Dame: University of Notre Dame Press, 1978a). ———, Scientific Progress: A Philosophical Essay on the Economics of the Natural Science (Oxford: Blackwell, 1978b). ———, Cognitive Economy: The Economic Dimension of the Theory of Knowledge (Pittsburgh: University of Pittsburgh Press, 1989). ———, Pluralism: Against the Demand for Consensus (Oxford: Clarendon Press, 1993). ———, Priceless Knowledge? Natural Science in Economic Perspective (New York: Rowman and Littlefield, 1996). ———, Predicting the Future: An Introduction to the Theory of Forecasting (Albany: State University of New York Press, 1998a). ———, Complexity: A Philosophical Overview (New Brunswick: Transaction Publishers, 1998b) ———, Epistemetrics (Cambridge University Press: New York, 2006).

473

RESCHER STUDIES

Romer, David, Advanced Macroeconomics (Boston: McGraw-Hill Irwin, 2006) Sagan, Carl, The Varieties of Scientific Experience: A Personal View of the Search for God, 1985 Gifford Lectures, ed. Ann Druyan (New York: Penguin Press, 2006). Wible, James R., “Implicit Contracts, Rational Expectations, and Theories of Knowledge,” Review of the History of Economic Thought and Methodology, vol. 7 (1990), pp. 141-70. ———, “Rescher’s Economic Philosophy of Science: A Review of Nicholas Rescher’s Cognitive Economy, Scientific Progress, and Peirce’s Philosophy of Science,” Journal of Economic Methodology, vol. 1, (1994a), pp. 314-29. ———, “[A Review of] Cognitive Economy: The Economic Dimension of the Theory of Knowledge by Nicholas Rescher,” History of Political Economy, vol. 26, (1994b), pp. 177-180. ————, The Economics of Science: Methodology and Epistemology as if Economics Really Mattered (London: Routledge, 1998). ———, “Prediction, Complexity, and Pluralism: More on Rescher’s Post Post-Modern Economic Philosophy of Science,” Journal of Economic Methodology, vol. 7, (2000a), pp. 294-301. ———, “What is Complexity?” in David Colander (ed.), Complexity in the History of Economic Thought, Perspectives in the History of Economic Thought (London: Routledge, 2000b), pp. 15-30. ———, “Complexity in Peirce’s Economics and Philosophy: An Exploration of His Critique of Simon Newcomb,” in David Colander (ed.), Complexity in the History of Economic Thought, Perspectives in the History of Economic Thought, (London: Routledge, 2000c), pp. 74-103. NOTES 1

Another justification of an assumption of costless science might be something like an infant industry argument—that noneconomic factors are more important in understanding science especially in an early phase of scientific development. This

474

HOW IS SCIENTIFIC KNOWLEDGE ECONOMICALLY POSSIBLE?

NOTES

implies that the attention of the most innovative scientist turns toward the core logic of successful experimental design rather than toward economic aspects of an experiment. 2

Peirce used a reformulated model from one of the first mathematical models of utility maximization. His model draws from Jevons’s (1871) path-breaking work on consumer optimization.

3

Stylized facts are a short list of the most important facts that a theory should explain. Existing theories typically explain most but not all of the stylized facts.

4

One of the better introductions to theories of economic growth is Romer (2006).

5

The phrase in quotes is attributed by Rescher to William George in The Scientist in Action, published in London in 1936.

6

In macroeconomics, there is an important question about the growth of the whole economy which is similar to those raised by Rescher about science. The question is what determines the pace and pattern of economic growth. Solow’s growth theory implies that physical capital and saving for investment purposes should be the most important determinants of growth. However, that has not been supported by recent research. Instead economists now study endogenous growth models where the growth of technical knowledge has an important role to play. A good summary of this material can be found in Romer (2006) chapters 1 to 3. Romer (2006, pp. 38-42) also provides an overview of research on whether the fixed nature of land works as a drag on economic growth. Estimating models of growth with and without the amount of land being fixed, it has been found that the drag on economic growth is quite small. Apparently, human beings are very creative in innovating around physical constraints and limitations.

7

Rescher (1989, p. 17) specifically has in mind radical Popperians such as P. K. Feyerabend.

8

One of the best economics books on implicit contracts and trust is Okun (1981). The role of implicit contracts in macroeconomics and science is explored in Wible (1990). Cognitive Economy is also reviewed in Wible (1994a, 1994b).

9

Abduction was the term Peirce created for making a guess that took the form of an hypothesis. An abduction goes way beyond existing facts and needs to be supported with new evidence

10

Rescher’s Predicting the Future appeared in 1998 and is discussed in a subsequent part of this paper.

11

A similar calculation appears in Carl Sagan’s (2006, pp. 108ff) posthumously published 1985 lectures titled, The Varieties of Scientifics Experience.

475

RESCHER STUDIES

NOTES 12

In his recent book on economic methodology, Wade Hands (2001) has called this the “economic turn” in science studies. He provides an excellent overview of the writings of several economists on the economics of science.

13

The problem of reflexivity in economics is also explored in Wible (1998).

14

Since I have reviewed these books previously, only the most important points will be considered here. See Wible (2000a).

15

For a related discussion of ideas of complexity as they relate to Peirce’s ideas see Wible (2000b, 2000c)

476

Possibility, Plenitude, and the Optimal World: Rescher on Leibniz’s Cosmology Catherine Wilson

N

icholas Rescher’s contributions to Leibniz-studies constitute a large and important part of his oeuvre. Few translations and critical studies, especially in English, were available when Rescher began his dissertation, completed in 1951, on Leibniz’s cosmology, his metaphysical physics, at Princeton,1 and indeed 1951 might be taken as the start of the Leibniz Renaissance in the English-speaking world.2 Interest in Leibniz has accelerated in the years since, with new theses, translations, articles, and monographs appearing at an increasing rate. Rescher’s role in stimulating this revival was substantial. Not only did he contribute a number of papers, beginning with “Contingence in the Philosophy of Leibniz,” published in the Philosophical Review for 1952, he produced an introduction to Leibniz, a general survey, a translation with commentary of the Monadology and three collections of essays. He was a founding member of the International Leibniz Society, an early editor with Studia Leibnitiana, and he encouraged younger academics, including the present writer, at early stages of their careers. While Rescher notes that he is not “a Leibniz scholar in that mode of total commitment that typifies the European philosophy historians of the old school,” this comment reflects the range of his interests, not the depth of his contributions. Rescher’s work on Leibniz falls into two main categories: historical studies and philosophical analysis. In the first group are papers falling under the heading of contextual history, situating Leibniz within the social, intellectual and political climate of the late 17th century. An early essay in this group addresses Leibniz’s opinion of William Penn and the Quakers, the most unorthodox of the respectable Protestant sects of the 17th century.3 Others papers examined Leibniz’s relations with John Frederick, the Duke of Hanover in the 1670s, with the Emperor Charles VI, with Prince Eugene of Savoy, and with the Electress Sophie Charlotte.4 These articles and chapters bring to life a philosopher who, for all his authorship of seemingly fantastic metaphysical doctrines, was a somewhat burdened but remarkably competent historian, engineer, and librarian with broad interests and excellent social skills. It is a different set of papers, however, papers that are

RESCHER STUDIES

addressed to what the Germans term “systematic” philosophy, on which I will concentrate in the remainder of this chapter, and indeed a subset of them. Rescher’s most sustained and important investigations in this area were the papers collected in his Studies on Leibniz’s Cosmology,5 on the general notion of a Leibniz-world and the puzzles and problems related to its origins and constitution, and its moral and aesthetic value. Rescher has been uniquely attentive to the formal and aesthetic principles of theory construction in Leibniz, to the philosopher’s search for reduplicative patterns and echoes to be found at many levels. This, in my view, is his enduring legacy in Leibniz studies. Few philosophers in the first half of the 20th century, a century of war, atrocity and fear, were intrepid enough to venture on any serious academic discussion of Leibniz’s doctrine of the perfection of the world. His view that we inhabit “the best of all possible worlds” had been developed to extremes never intended by Leibniz by his academic follower Christian Wolff and wittily satirized by Voltaire in his Candide. Leibniz’s reputation had suffered accordingly. Bertrand Russell treated the best of all possible worlds doctrine as a piece of frivolity, suggesting that temporary or apparent evil was related to overall good in just the way that “a drink of cold water when you are very thirsty on a hot day,” can be so satisfying as to make one’s earlier thirst seem to have been worthwhile. So absurd did the doctrine of perfection appear to Russell that he suggested that Leibniz’s metaphysical views belonged to his “exoteric” philosophy, while his logical doctrines were the basis of an “esoteric” doctrine of the self-creation of the universe and the universal reign of determinism. Rescher brought out the esoteric aspects of Leibniz’s doctrine of the best, He invited, and indeed compelled readers to consider seriously the arguments and considerations Leibniz advanced in favour of what he knew to be a priori an implausible thesis. Rescher’s work not only brought out the sophistication of Leibnizian optimism, it also generated a set of questions –some textual, some metaphysical—on the notions of plentitude, perfection, and choice. These are questions of continuing interest, both philosophical and exegetical. By way of making a personal contribution to the present volume, I would like to illustrate what I and others have been prompted to think about, and what we have learned, from reading Rescher on Leibniz’s cosmology.

478

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

1. LEIBNIZ-WORLDS AND THE DOCTRINE OF PERFECTION Leibniz was the first philosopher since medieval times to consider the logical features of possible worlds and their inhabitants, and he also considered their qualitative and quantitative features. The Discourse on Metaphysics was his first attempt to articulate systematic views that would put him into relation with the “moderns”—Hobbes, Descartes, and Spinoza. These philosophers differed in various ways one from another, but they held in common the view that good and evil named the subjective evaluations humans attached to events, and that God was either the same thing as the corporeal world, or a vast arbitrary power. Leibniz led off his Discourse with a resounding declaration of the motivation and competence of a god who cared: To show that an architect could have done better is to find fault with his work….It is enough, then, to have confidence in God that he does everything for the best, and that nothing can harm those who love him…[H]e who acts perfectly is like an excellent geometrician who knows how to find the best constructions of a problem; or a good architect who makes the most advantageous use of the space and capital intended for a building, leaving nothing which offends or which lacks the beauty of which it is capable; or a good family head…; or a skilled machinist who produces his work by the easiest process that can be chosen; or a learned author who includes the greatest number of subjects in the smallest possible volume.6

“God,” he declared: has chosen to create the world which is the most perfect, which is to say that which is at the same time the simplest in its hypotheses and the richest in its phenomena7

Leibniz also seems to have held that the actual world—and probably any world—contains a infinite number of substances, or at least that there is no end to the number of new creatures, who are substances and who contain multitudes of substances within themselves, that an investigator could go on discovering, were she able to travel to remote regions of the cosmos, or to explore the microworld with ever-improving instruments for visualization. He went on to develop the notion of a Leibniz-world, an L-world, that has the following features:

479

RESCHER STUDIES

1.

An L-world is a set of substances.

2.

There is one actual L-world and an infinity of possible L-worlds.

3.

A substance is an indestructible unit that experiences an infinite series of representations, emerging spontaneously and sequentially from its depths because of its appetites and in accord with its divinely predefined “concept.”

4.

A given substance exists in only one L-world.

5.

All substances in an L-world represent all others in their world.

6.

The actual world is the most perfect of all the L-worlds.

7.

The existence of the actual world is morally necessary.

Despite his adherence to Theses 3 and 4, Leibniz maintained that what happened in the world was not necessary, and his thinking about necessity was, as Rescher showed, bound up with his combinatorial approach to the question of the assembly of worlds, an approach that both built on and revised the Cartesian precedent. Descartes, who considered the physical world apart from human souls to be a collection of material particles in motion and at rest, maintained that the visible world of animals, plants, celestial objects and geographical features took on all forms that were possible, given the laws of motion laid down by God at the creation. Every possible collection of entities and every possible sequence of events will be realized sometime in the history of the universe.8 We simply happen, on Descartes’s view, to find ourselves in a tolerable state of the universe, in which it is stocked with identifiable objects, large and small, and a mixture of moral and physical good and evil. But universes with only one or two humanly recognizable entities, or universes of (on our view) malevolent humans living in a landscape of cinders and floods and mistreating one another, have already been or will be realized some day if the principles of matter and motion permit it. And why should they not? All such entities and state of affairs, it seems, can be produced by matter in motion in accord with the laws of nature we currently recognize. Further, moral properties depend on the will of God; they are laws laid down by him that could have been otherwise. If, counterfac-

480

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

tually, God had willed that creatures should hate him, Descartes maintained, this would have been a good thing to do.9 Leibniz’s instincts rebelled against these conclusions. They implied that God was simply the amoral demiurge, neither benevolent nor malevolent, who inscrutably, for no evident reason, had brought a material world of no real moral qualities into being, decided what its rules of conduct were to be, and left it to run, with the understanding that he would punish infractions of his arbitrary rules with eternal torment.: “[A] God like Descartes's allows us no consolation other than that of patience through strength. Descartes tells us… that matter passes successively through all possible forms, that is, that his God created everything that can be made, and passes successively through all possible combinations, following a necessary and fated order…. God is merely this necessity or this principle of necessity acting as it can in matter. Therefore, it is impossible to believe that this God cares for intelligent creatures any more than he does for the others; each creature will be happy or unhappy depending upon how it finds itself engulfed in these great currents or vortices. Descartes has good reason to recommend, instead of felicity, patience without hope.10

Some constraint, above and beyond the laws of motion and mechanical principles governing particles had to exist, on Leibniz’s view, that would limit the possible visible forms the universe could take. Descartes’s claim that everything possible happens implied that if something never happens, it is not possible. So by asserting that some things that are possible nevertheless will never happen, that there are possible collections of things and sequences of events that have never been and will never be realized, Leibniz had the basic plank of an alternative theoretical construction in hand. His theory of possible and actual individuals can be approached via his theory of creation. 2. HOW DOES THE VISIBLE WORLD COME TO BE? What is the process by which God created the world? Leibniz had no interest, no more than Descartes did, in the biblical story of the hexameral creation of fish, quadrupeds, humans, etc. He maintained that “all possible things, or things expressing an essence or possible reality, tend toward existence with equal right in proportion to the quantity of essence or reality, or to the degree of perfection which they involve.”11

481

RESCHER STUDIES

This claim might rouse the mental picture of a vast horde of possible substances each struggling to race to the finish line to be crowned with actuality, or racing to get across a border separating the realm of possibility from the realm of actuality. Interpreted in this way, the theory of the striving possibles is puzzling, as Rescher points out, for it suggests that God’s goodness and his unique ability to realize his preferences fully are not the cause of the existence of the actual world. Rather, it seems, the world has brought itself into existence on account of the sheer force of the substances it now includes, substances whose tendency to exist was simply stronger than that of their competitors, with God a passive witness, handing out blue ribbons at the end. Rescher decisively rejects this interpretation.12 God’s choice of it must be the cause of the existence of the unique actual universe. Moreover, one might observe, the race-model cannot represent Leibniz’s real intention with respect to the striving possibles, for, in such races, even if a certain degree of speed is required to get across, the competitors that cross the line do so on account of their individual excellence, not because of their relations with one another. But in an L-world, only substances that fit into the same world can exist in it. The world is just a collection of compossible substances.13 The claim on existence of an individual substance can only be a claim-to-co-exist-with-these- particular-others, and if compossibility is an all or nothing relation, there cannot be degrees of individual perfection. Because a given substance can exist only in one world, it cannot enter into a less perfect confederation with a set of other substances. So the relative “perfection” of a substance cannot make a difference to whether it belongs to the unique existing world or to any other. Judas contributes to the perfection of this world –he is perfect for it—as Rescher observes, and, accordingly, Judas has more right to exist than some possible person rather like Judas but who did not betray Christ.14 As a result, we have to say that there are no deserving substances, held back by their compossibles and doomed to remain merely possible because of bad companions. There are no bad companions, strictly speaking; only denizens of overall inferior worlds. Belonging to an inferior world limits the entitlement of a subject to exist, or its ability to strive to exist. The upshot, however, is that we have to conceive worlds as conceptually prior to individuals. There is no model in which free-floating substances can, by their individual striving, aggregate into many possible worlds, but only one actual world.15 Rather, collections of individuals face God’s existence tribunal all together, like teams competing for selection, and God’s prefer-

482

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

ence for them reflects their excellence as a team. At the same time, their energy is proportional to their excellence, as Aristotle seems to have thought in his analysis of the moral relations of victor and vanquished.16 3. HOW CAN WHAT HAPPENS IN THE WORLD BE ANYTHING BUT NECESSARY? According to Theses 3, 4, and 5, the experiences and actions of a substance cannot be other than what they are. To say “Caesar might not have crossed the Rubicon” or “Eve might not have eaten the apple” is to say something literally false, just as most17 statements invoking the activities of ghosts and demons are deemed false, even though we can seemingly form mental representations of the conditions that would make them true. There are “other Caesars” and “other Eves,” in possible worlds, but they bear only a relation of similitude to our substances, and the sequences of events in which they occur prominently—their narratives—are different. If we were to consider them as determinate individuals, we would have to assign these counterparts different names.18 So it seems that all actions of all creatures in our world are necessary, and that we are under a sort of cognitive illusion in supposing that we have the capacity mentally to represent what it is like for counterfactual situations to be realized. But it is difficult to reconcile that claim with the more intuitive view that our true (and false) counterfactual thoughts that are genuinely about ourselves and the substances around us. 19 Further, while it might have been “mechanically necessary” for Eve to eat the apple, given the laws of nature of her world, it seems strange to insist that it is “logically necessary.” But if it is not the case that, in some possible world, Eve does not eat the apple, it would seem to be logically necessary that Eve does eat the apple, and if Eve is logically compelled to do so, what becomes of free-will, sin, divine justice, and so on, as Leibniz’s critic Arnauld asked, sending him back as it were to the drawing board? Why, one might wonder, did Leibniz adhere to Thesis 4, rather than endorsing a version of counterpart theory on which an individual can exist in more than one world? This would have enabled him to analyze counterfactual statements more intuitively, and would have supported better his claim to be defending theology and morals from the Cartesian threat. Thesis 4 is, however, essential to the concept of a Leibniz-world because of its relations with Theses 3 and 5. If a substance represents every other substance in its world, and if all the representations of a substance are given in ad-

483

RESCHER STUDIES

vance, then the existence of one substance in a world effectively determines the rest. So Leibniz’s semantic and moral-theological problems, given his conception of substances and worlds, were inescapable. One might insist that when I say that Caesar need not have crossed the Rubicon, I have a true thought whose truth-conditions reside simply in the existence of a similar-to-ours possible world in which a Caesar counterpart does not cross the Rubicon. Some counterfactual thoughts about N are made true by the possible states and doings of some other entity N’ and some features of N’s world; they are nevertheless thoughts about N in our world, just because it is something about N and our world that is generating the thought. But Leibniz did not go down this thought pathway. Rather, as Rescher shows, he argued that, for Eve’s, Caesar’s, and, for that matter, my actions to be logically contingent, it is not essential that some counterfactual statement about what we might have done instead be true. The contingency of an action by an agent does not require that that very action by that very agent fail to occur in a possible world. All that it requires, Leibniz maintained, is that that agent, who is predestined to perform that action, would not have existed had there not been the God there is, the God who chooses to realize the best of all possible worlds. (Leibniz assuredly does not infer that if an action A occurs in all possible worlds in which A exists, it is necessary.) But how is it possible that any actual substance could have failed to exist if God exists necessarily and if he cannot fail to choose the best ensemble of substances? As Rescher points out, although it is certainly not the case that there are possible Gods who act otherwise, it is not metaphysically necessary for God to create the actual world he does, the most perfect possible. To describe God’s actions as necessary would be to say, nonsensically, that God’s actions are the same in all possible worlds, or to put God under some form of necessity, to posit a source of that necessity—another God who ordains how things are to be or who has selected this God who ordains things to be a particular way—and this too is absurd. God cannot do otherwise than act for the best, but this is because he has a divine character, perfect intelligence, and a perfectly efficacious will. His is an intentional choice of the best entities, those that together will result in the best world. Rescher intends this interpretation I think when he says that “God is in a certain sense self-constituted . . . in certain respects God freely decides various aspects of (not his nature, but) his actions and modes of acting. And this includes, in specific, “being good” by way of creating this particular world of ours.” 20

484

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

Thus the team of individuals preferred and chosen by God is the unique combination with which “entity space” (not to be confused with imaginary physical space and time) can be “filled,” and the sequence of events they experience is the unique sequence that can “fill” it, but the reality of this team is still contingent on the goodness and power of God. Accordingly, 1.

God is not determined by anything outside himself, in addition to his knowledge, his preferences and his capabilities, to create the best. In this sense, he is free.

2.

It is not possible for God to act so as to create a world that is not the best possible. His freedom does not encompass this possibility.

Analogously, an artist might be observed to enjoy a vast range of choices over materials, procedures and techniques, including dyes, pigments, and forms to employ in creating a work of art, but, in virtue of having a perfectionist personality, she might not be able to complete and put her name to anything substandard. Her temperament precludes this. Rescher invites us to see God as just such an artist.21 The source of the goodness of the actual world and its contingency are the same: the preferences of God and God’s ability to realize them. But if God’s preference for it makes the choice of the actual world inevitable, the human will is not it seems perfectly free insofar as Eve, Caesar, etc. could not have done otherwise. A substance cannot have experiences other than the experiences it has, in the order it does, for these flow from its concept and harmonize with the experiences of all the others in its world. Leibniz’s version of compatibilism, often ridiculed under the heading of the preestablished harmony, was genuinely new, and his treatment of a substance’s determinations is parallel to his compatibilist treatment of God’s will: 1. No creature, whether actual or possible, is determined to act by anything outside itself, but only as its perceptions, inclinations and powers dictate. 2. It is not possible for an actual creature to act other than as the moral constitution of the universe requires. Its free will does not extend to this possibility.

485

RESCHER STUDIES

If a substance were free in the sense that, at a given time t, it could, as a result of what it willed, experience either x or y, God would not have preferred and realized a genuine L-world. 4. WHY IS THERE ONLY ONE ACTUAL L-WORLD? Why, one might wonder next, does God not create many worlds, even if some are less good than our world? Would not a multitude of distinct Lworlds better satisfy God’s desire to enjoy as much essence as possible? If the right to exist of a substance is the contribution it makes to the richness and economy of the world in which it is situated with all its compossibles, why isn’t the right to exist of a world the contribution it makes to the richness and economy of the whole set of worlds? If there is a “best” world, why are there not better-than-average and worse-than-average worlds, with greater or lesser claims on existence? Perhaps a certain number of betterthan-average worlds could have been realized? At some points, Leibniz suggests that each substance is like a world, but it is also true that a world is like a substance. For the substances of any world, possible or actual, all contain in one sense the same information, insofar as they represent all the experiences of the others, though from their own perspectives. All the information about the world in which it is located is contained in any arbitrarily chosen substance. To have one substance, from one world, is to have them all. To have different nonreduplicative content, it seems, one must have different worlds, not merely different substances. So why should not multiple, non-interacting worlds contribute to the excellence of some larger totality in which they exist? Leibniz even seemed to entertain this suggestion at one point early in his career: . . . [I]t does not follow from this [the eternity and infinity of our world] that there is not another world, or other minds which cohere among themselves in a way which is different from that which holds in our case …. God is equally present both to this and to that world, for there could be a different law of nature in that world. 22

But he soon rejected this position: There is no need for the multitude of things to be increased by a plurality of worlds; for there is no number of things which is not in this one world, and indeed in any part of it.

486

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

To introduce another genus of existing things, and as it were another world which is also infinite, is to abuse the name of existence; for it cannot be said whether those things exist now or not.23

In the Theodicy, Leibniz defines “the world” as “the whole succession and the whole agglomeration of all existent things, lest it be said that several worlds could have existed in different times and places.”24 God’s realization of all possible worlds, merely for the sake of plenitude, would raise the Cartesian threat: everything possible happens. Realizing this multiplicity would express his love of essence, but God does not desire to maximize essence at the expense of moral worth. The best world must exemplify its bestness by being unique in existing, and, as Rescher explains, God is not maximizing any particular quality in choosing the best of all the worlds for realization. 5. WHAT MAKES ONE L-WORLD THE BEST AND SO ACTUAL? Rescher showed how Leibniz understood the best of all possible worlds doctrine by reference to problems in engineering involving the optimization of initial settings. Optics, especially, provides of range of examples in which there is a trade off between values when two parameters are in competition. In a camera, for instance, a wider aperture admits more light, allowing the image to be registered on a cloudy day, but depth of field is lost. The best choice of settings is a delicate problem, and not every picture, taken even with a well-adjusted camera, can be in perfect focus everywhere. Maximizing perfection is really optimizing, when two variables both represent desiderata but cannot be simultaneously maximized. The two values Leibniz thought God specially cared about were simplicity of underlying principles—these to be held to a minimum—and richness of effects—these to be drawn out to a maximum. God, according to Leibniz, establishes the settings for the universe only once, producing as much richness and variety as he can, but at the same time keeping the laws of nature as simple and uniform as he can. This world cannot be improved upon; for any change to its underlying settings would produce a world that was either impoverished with regard to what it contained, or else more chaotic and unpredictable: the phenomenal world must exhibit a scientific basis that is “simple, elegant, discernible, calculable.” Rescher points to a number of examples of variety-within-constraint: baroque music (innovative tunes vs.

487

RESCHER STUDIES

structural unity); statecraft (freedom vs. law and order); science (wide explanatory power vs. simple concepts).25 He shows how Leibniz was moved by parallels and analogies that go beyond the parallelism of mind and body, or the analogy between appetite and physical force. One of the papers in Studies in Leibniz’s Cosmology brings out Leibniz’s irrepressible habit of thinking analogically: As the traditional atoms are the ultimate physical building blocks of material objects, so the monads are the ultimate metaphysical building blocks of substance . . . As an embryonic state of an organism is to its later more developed stages, so is the present state of a substance to its later more developed stages . . . God is to the world that is his creation as the inventor is to his machine . . . ” 26

The process of world-choice can be elucidated by thinking of machines that generate outputs. Both the machine-design and its output can be evaluated separately according to distinct criteria: Imagine a set of weaving machines that are being inspected and ranked. These machines can be evaluated by engineers who have no aesthetic sense for tapestries purely on the basis of their design-properties. Some have few moving parts, operate quietly and smoothly, and are not prone to breakdown; others are Rube Goldberg contraptions. (Leibniz’s rejection of the Newtonian world-machine was based on Newton’s own admission that the Newtonian cosmos was prone to breakdown.). Quite separately, a group of connoisseurs with no sense for machinery could evaluate the output of the machine separately from the qualities of the machine that produces it. How many different kinds of design and texture—stripes, brocades, velvets, tapestries, satins, checks, tweeds, can this machine produce? Machines composed of many parts that can weave only a few patterns are not eligible for realization, but how does God choose between a machine that is finicky and has many complicated and prone to breakdown parts, but that weaves more and more beautiful patterns and textures than its competitors, and a machine that is sleeker, more elegantly designed, and more reliable, but that cannot weave as many fine things? One can imagine a human chooser stymied by such a dilemma, but Leibniz does not envision God as having to make this choice, or to order his values. Rather the machine God constructs is such that there is no simpler machine that could produce such a variety and beauty of tapestries, and there is no tapestry too lovely for the machine God brings into existence to produce.

488

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

The result of God’s fixing of the settings is a phenomenal world that inevitably appears aesthetically and morally variable to ordinary observers. In the actual world, there is a variety of living forms, of colours, pieces of music, chemical combinations, geographical formations, faces, hands, and leaves, each of which is a perfect individual like no other. Note that Leibniz does not demand that the machine be able to produce all possible forms of cloth, as the Cartesian machine does, if left to run indefinitely, but even rough weaves and garish plaids may have their place in the range of textiles. So in the world, there are ugly objects and horrific episodes, but new varieties of injustice and maltreatment, if there really are new varieties, do not add to the richness, in the sense in which God desires it, of the world; they were not what God had in mind when setting things up. That we feel, intuitively, that the balance between the experiential complexity of our world and its intellectual simplicity is a good one is not the same as knowing that it is the optimal balance. Many of Leibniz’s readers, like Russell, can attach no credence to the claim that the ultimate origins of the world depend on the preferences and actions of a supramundane person, and we believe that the perfect individuality and uniqueness of everything in it is only highly probable and not metaphysically assured. Nevertheless, the sense of living in a well-ordered world, into which we fit as we would fit into no other world is, as Kant often pointed out, important to cultivate, and Rescher’s investigations contribute no little to that end, as well as to a correct understanding of what Leibniz really meant to say.

NOTES 1

Where primary sources in English were concerned, students of Leibniz had little to turn to except Langley’s translation of the New Essays, (1896), Robert Latta’s edition of the Monadology, (1898); and the extracts made available by Bertrand Russell in his The Philosophy of Leibniz of 1900. John Dewey’s study of Leibniz was published in 1888, followed by Russell’s, Carr’s (1929) and Joseph’s (1949).

2

The translations and collections of Loemker, Weiner, Alexander, Huggard, Lucas and Grint, and Morris all appeared between 1951 and 1956. Over the last twentyfive years, there have appeared Anglophone monographs by Adams, Hartz, Jolley, Mates, Mercer, Nachtomy, Phemister, Ross, Rutherford, Sleigh, and Wilson.

3

Rescher, “Leibniz and the Quakers,” Bulletin of the Friends’ Historical Association, 44 (1955) pp. 100-07.

489

RESCHER STUDIES

NOTES 4

See especially “Leibniz finds a Niche,” pp. 162-198; “Leibniz visits Vienna,” pp. 199-231 of On Leibniz, Pittsburgh, University of Pittsburgh Press, 2003.

5

Rescher, Studies in Leibniz’s Cosmology, Collected Papers, Vol XIII, Frankfurt, Ontos, 2006.

6

Leibniz, Discourse on Metaphysics, Sects. 3-5, Die Philosophische Schriften von Leibniz, ed C.I. Gerhardt, 7 vols., Hildesheim Olms, 1965, IV: 430, tr. Loemker, Philosophical Papers, Dordrecht, Reidel, 1969, p. 305-6.

7

Ibid., Sect 6. G IV: 431, tr. Loemker, p. 307.

8

Descartes, Principles of Philosophy, Pt. III, Sect. 47; in Oeuvres 11 vols., ed. C. Adam and P. Tannery, Paris, J. Vrin, 1964-74, IX-2: 103.

9

Descartes, “Conversation with Burman,” AT Vi:160, tr. CSM III:343.

10

Leibniz, Letter to Molanus, in Philosophical Essays, tr. and ed. Roger Ariew and Daniel Garber, Indianapolis, Hackett, 1989, p. 242; Philosophische Schriften, ed. Gerhardt, IV:299-300.

11

Leibniz, “The Radical Origination of Things,” in Gerhardt, ed. Philosophische Schriften, VII: 303; tr. Loemker, p. 487.

12

Rescher, “Leibniz on Ontological Perfection,” in Studies, p. 55.

13

There is an analogy here with recent approaches to natural selection; genes face the tribunal of selection as confederations, not as individuals, according to Richard Dawkins, The Extended Phenotype, 2nd ed. Oxford, Oxford University Press, 1999, esp. Ch. 4.

14

Rescher, “Leibniz on Ontological Perfection,” p. 58.

15

Or so I argue in "Plenitude and Compossibility," Leibniz Review 10 (2000), pp. 120.

16

Aristotle, Politics, Bk. I, Ch. 6 1255a: “[I]n some sense, excellence, when furnished with means, has actually the greatest power of exercising force…power seems to imply excellence.” In Jonathan Barnes, ed., Complete Works of Aristotle, 2 vols. Princeton, Princeton University Press, 1984, II:1991.

17

I say “most” to exclude statements like “Ghosts do not walk on ramparts” which are true.

18

On “possible Adams,” see Leibniz, Letter to Arnauld July 14, 1686, G II:54f.

19

See Saul Kripke’s criticisms of Lewis’s counterpart-theory in Naming and Necessity, Princeton, Princeton University press, 1980, p. 45.

490

POSSIBILITY, PLENITUDE, AND OPTIMAL WORLD

NOTES 20

Rescher, “Leibniz on the World’s Contingency,” Studies, p. 31

21

Rescher, “Leibniz on Ontological Perfection, Studies, p. 64, et passim.

22

Leibniz, De Summa Rerum, tr. and ed. G.H.R. Parkinson, New Haven, Yale University Press, 1992, p. 67.

23

Leibniz, “My Principle is…” in Parkinson, op. cit., p. 103.

24

Leibniz, Theodicy, Sect. 8, G VI: 107, tr. in Theodicy: Essays on the Goodness of God, the Freedom of Man, and the Origins of Evil, ed. August Farrer, tr. E.M. Huggard, La Salle, Open Court, 1985, p. 128.

25

Rescher, “Leibniz on Ontological Perfection,” Studies, p. 69

26

Rescher, “Analogy and Philosophical Method in Leibniz,” in Studies, pp. 161-167

491

Rescher on Aporetics and Consistency John Woods

I

t is by now a well-worn legend that Isaiah Berlin dined out often and to advantage on the strength of the observation by the poet Archilocus (680- 645 BC) that the fox knows many things but the hedgehog knows one mighty thing. Given his breathtaking versatility, no one could deny to Nicholas Rescher the appellation “fox”. But it would be quite wrong to overlook Rescher’s appreciation of big ideas, and their potential for unification and systematicity. One of Rescher’s big ideas is that of limitations, whether on actions and understanding, on reference and ascription, on the acquisition of knowledge, on access to information, on the capacity to process it. These for the most part are not discoveries of Rescher’s making, but they stand out in his writing, and they do so in noticeably Rescherian ways. An important example of these Rescherian limitations are those that involve wayward quantification and anomalous conjunction. Indeed a second theme prominently on display here is that of problematic totalities. Chapter 1 deals with quantifications that place themselves beyond the range of identifying specification. For example, there are ideas that have never occurred to anyone, but precisely because the quantified sentence is true no one will ever come upon a sentence knowing it to be an instantiation of it. When this happens, Rescher says that the quantifications embed “vagrant predicates”. A vagrant predicate “P” has this property: The dots in the sentence “There are things that have P, e.g., …” are by the meaning of “P” impossible to fill in. Since vagrant predicates must have nonspecifiable application, we have it that individuation doesn’t require identification. And since we can know the unknowability of something without knowing what it is, Rescher takes this as compromising the prospects of substitutional quantification and intuitionistic logic. In chapter 2, we have it similarly that there are facts that can never be expressed; no one will ever come upon a fact knowing it to be an instantiation of that limitation. Each kind of case lands us in an instantial blackout. Instantial blackouts make for underdetermination. Chapter 4 deals with aporetic clusters, after the Latinized Greek aporos for “impassable”. Aporetic clusters are the stock and trade of philosophy. They are families of claims of which Rescher writes:

RESCHER STUDIES

(1) as far as the known facts go there is good reason for accepting them all; the available evidence speaks well for each and everyone of them, (2) taken together, they are mutually incompatible; the entire family is inconsistent (chapter 4, p. 7). While predicate-vagrancy lands us in the bind of underdetermination, aporetic clusters pull in the opposite direction. They are overdetermining. They are structurally linked to arguments of the following type: 1. Every natural event except possibly those in the micro-order is caused. 2. Every human action is a natural event of this very sort. 3. No caused action can be free. 4. So, human action can’t be free. The argument proceeds from premisses not antecedently in doubt, yet they jointly ground a truth-preserving conclusion which everyone would find, to say the least, unwelcome. Assuming its deductive validity, the argument is open to two conflicting interpretations. On one of them, the argument is a reductio of one or more of its premisses by virtue of the absurdity of its conclusion. It bears on this that a principal virtue of the very enterprise of proof is that it not merely rubber-stamp what is already known; it sometimes overturns wholly confident convictions to the contrary. This being so, we have a second interpretation: our present argument might be one of those, that is, a sound demonstration of something surprising (or gobsmacking). This gives rise to what with some justice might be called Philosophy’s Most Difficult Problem1. Save for a particular class of exceptions, there seems to be no suitably general and grounded policy for determining when a valid argument is a reductio of something heretofore taken as true, and when it is a sound demonstration of something previously thought to be absurd. The exceptions, as is widely supposed, are those valid arguments whose conclusions are contradictions. Here, it is said, there is no question of there being a sound demonstration of such a conclusion, for we would have it

494

RESCHER ON APORETICS AND CONSISTENCY

otherwise that some contradictions are actually true. Still, this is precisely what some philosophers do in fact think, although Rescher is decidedly not in their number. These are the dialetheists, subscribers to the proposition that some (very few) contradictions are indeed true. This occasions difficulties, one for the non-dialetheist and the other for the dialetheist. For the non-dialetheist, there is the problem of showing in a non-question-begging way that the proof of, say, Russell’s paradox is not a sound demonstration of a surprising—indeed, a shocking—truth. For the dialetheist, the problem is that of producing a policy for determining when a valid argument whose conclusion is a contradiction is a reductio of something in the premiss-set or is a sound demonstration of one of those few true contradictions. Consider, for example, the Curry Paradox, whose conclusion is in the trivialist claim that every sentence is true. Dialetheists abhor trivialism as much as the rest of us abhor dialetheism. The rest of us, Rescher included, have a reason to reject the Curry proof of it. The reason is that the proof’s conclusion entails contradictions. But for dialetheists this is not a generally available manoeuvre. In each case, then—for us and for them—Philosophy’s Most Difficult Problem recurs. Aporetic clusters face us with the task of “epistemic damage control”. In chapter 4, Rescher relativizes weakest-link identification to the nature of the enquiry in which the aporetic cluster arises and the kind of the aporetic trigger that produces the problem. There are five cases to consider. (1) In the area of scientific inquiry aporized by new discoveries, the weakest link is “the most weakly evidenced proposition”. (2) If in the same field, aporia arises from scientific conjecture, the weakest link is “the systematically least well entrenched proposition”. (3) In philosophy, under press of a speculative supposition, the weakest link is the thesis least consonant with the fundamental commitments of one’s overall position”. (4) In mathematics under a consistency-distributing hypothetical assumption, the weakest link is by reductio the assumption itself. (5) In the area of pure speculation under a wholly speculative supposition, the weakest link is the systematically least fundamental proposition.” (4, p. 6; my emphases.) Chapter 5 is about paradoxes. Their resemblance to aporetic clusters speaks for itself, as does the element of unspecifiability that they introduce. The set of all sets includes itself as a member, but no one will ever come upon something knowing it to be that set. Like aporetic clusters—and contrary to my suggestion two paragraphs ago—paradoxes offer promise of resolution at their weakest link. Rescher here conceives of the weakest link as the least plausible of the propositions involved in the paradox. If it is not

495

RESCHER STUDIES

obvious where the greatest implausibility lies, Rescher offers an algorithm that identifies that proposition whose abandonment entails least cost. The algorithm operates on the priority ranking of propositions, where rankings are reflections of the reliability of a proposition’s source or ground. The theme of inconsistent totalities is picked up again in chapter 6, and so too the idea of experiential inaccessibility. Consider illegitimate totalities such as the set of all sets or the truth that conjoins all truths. A good many philosophers side with Russell in determining that no such totalities exist. Russell judged that they offend against the Vicious Circle Principle. The chapter is important for the wedge that it drives between two kinds of illegitimacy, the ontic and the epistemic. Rescher thinks that Russell’s Vicious Circle Principle goes too far precisely because it is an existencenegating rationale for rejecting these totalities. By Rescher’s lights, the trouble caused by these totalities is epistemic (“experiential”), not ontological. The Vicious Circle Principle is over-kill. What’s wrong with the set of all sets is: a problem of identification [of knowing that this is it] rather than one of being or existence—an epistemic rather than ontological failing.

This is a distinction that also bears on instantially non-specifiable generalizations such as “There are ideas that have never occurred to anyone” in which “ideas that have never occurred to anyone” is a vagrant predicate. Here, too, it is easily seen that the problem is not that there aren’t such ideas but rather that they can’t be knowingly specified. Accordingly, Rescher replaces the Vicious Circle Principle with the Successful Introduction Principle (SIP): The identifying condition for an item must not involve a reference—explicit or tacit—to that item itself.

If we grant that SIP is correct, then it offers promise of repair for the entire class of self-referential paradoxes. But it is not a wholly general paradoxsnuffer. It leaves non-referential variants—for example, Steven Yablo’s non-referential form of the Liar—beyond its reach.2 Chapter 3 is an excerpt from The Logic of Inconsistency (1980), coauthored with Robert Brandom. The Logic of Inconsistency is an essay on paraconsistency. A logic is paraconsistent precisely when it blocks the derivation of ex falso quodlibet, the theorem that provides for the equivalence of negation inconsistency and absolute inconsistency. A system is

496

RESCHER ON APORETICS AND CONSISTENCY

negation-inconsistent if and only if for some wff A, both A and ∼A are theorems. A system is absolutely inconsistent if and only if every sentence is a theorem—in other words, is trivial. Paraconsistent logics divide rather neatly into two camps. In the one, ex falso is blocked by imposition of a relevance constraint on deducibility and entailment. In the other, ex falso is blocked in some other way, for example, by suppressing the rule of adjunction, without the need to impose a relevance condition. At the time of the book’s publication, most of the paraconsistent work done by American logicians was of this first type. Indeed some of the leading relevant logicians were Rescher’s colleagues at the University of Pittsburg. Rescher himself, and Brandom too, are in the second camp, in the slipstream of paraconsistent developments in Brazil and Poland. (The paraconsistent terrain in Australia tended to subdivide into dialetheic logic on the one hand and relevant logic on the other. The paraconsistent turn in Canada is substantially influenced by adaptations of Rescher and Brandom, got by weakening some of the latter’s inferential constraints)3. The Rescher and Brandom essay is a response to two different impulses. One is the impulse to free set theory from the Russell contradiction, and to do so with least possible disturbance to the theory’s original intuitions. The other is to have a logic that is duly sensitive to the particularities of what for present purposes I will call “assertion”. Although these are clearly inequivalent objectives, they embed a common strategy. It entails, in the manner of Jaśkowski, suppression of the adjunction rule, the rule that permits unrestricted premiss-aggregation.4 It is a fateful rejection. In enables Rescher and Brandom to drive a wedge between the classical equivalence between 1.

The implication of X by {A, B, C, … }

2.

The implication of X by A ∧ B ∧ C ∧ …

The pages dealing with the first issue—that of attaining a consistent version of intuitive set theory—are not reprinted here. Even so, it is useful to have some idea of how Rescher and Brandom proceed with this task. For one thing, they see it as instantiating Rescher’s econometric criterion for paradox resolution, discussed in chapter 6. As he there points out, paradox resolution is a matter of converting an inconsistent set of sentences into a consistent set. Conversion is achieved by deletion.

497

RESCHER STUDIES

And the object is to achieve this at minimal cost—with the least possible sacrifice among the theses towards which we were, in the first instance, favorably inclined. (Chapter 5, pp. 1-2).

In the matter of Russell’s Paradox, Rescher is one of those philosophers who think that getting rid of the old axioms and finding some new (and presumably consistent) axioms for sets is more trouble that it’s worth. It is doubtless true that that replacement systems such as ZF or ZFC have attained a sure footing in modern mathematics, but Rescher thinks that this success has come at a price. The new set theories are complicated counterintuitive and ad hoc. It is doubtful that, when writing in 1980, Rescher and Brandom were seriously proposing that ZF and the like be given up on. But they did want to make the point—if only as a matter of principle—that there was a better way in which to have proceeded with the rehabilitation of set theory. Toward the end of The Logic of Inconsistency we find the idea of a “minimally consistent complete” or “intelligible” formal system. Something is an intelligible system if and only if it meets the following three conditions. (1) Its theorems include all classical logical truths. (2) No selfcontradiction is a theorem. (3) Any classical consequence of a theorem is a theorem. Condition (3) does not apply to sets of theorems, however. An intelligible formal system not only need not be closed under conjunction, it may contain pairs of incompatible propositions (but not their conjunctions). We see, then, that an intelligible system is a set of propositions generated by non-empty (though not necessarily finite) intersections and unions of complete consistent theories. Rescher and Brandom’s logic is an intelligible logic in this sense. It is a non-classical system built up from classical components. Could we apply to Russell’s proof the method recommended in chapter 4 for aporetic clusters in mathematics? If so, we would have to find a hypothetical assumption in the Russell proof for which the proof itself would serve the purpose of reductio ad absurdum. Everyone agrees that the comprehension axiom is the source of the contradiction, and Rescher appears to believe that the paradox downgrades a would-be axiom to a “hypothetical assumption”, which in turn is brought down by Russell’s contradiction-proving argument. We might agree that Rescher’s test succeeds in finding the proof’s weakest link. But it leaves undealt with the question of what now? This, in large part, is the business of the RescherBrandom monograph.

498

RESCHER ON APORETICS AND CONSISTENCY

How, then, do these authors deal with the paradox? In the spirit of Rescher’s ontological openness to vicious circularity, they would allow the intuitive comprehension axiom to stand, as well as the contradictions it generates, but without having also to allow for the explosiveness of those contradictions. The set theoretic contradictions are tolerable precisely because they can be isolated in ways that not only avert trivializing explosion but also avoid the considerable ac-hocness that attaches to the replacement of the intuitive axioms with new ones. Here the basic idea is to consider the naïve axioms as describing an impossible world. But since impossible worlds are superpositions of possible worlds, it ought to be possible to split the comprehension axiom into two consistent axioms, each of which operating distributively rather than aggreggatively would deliver the theory of sets without further ado. Clearly an interesting idea, it turns out that splitting is met with technical difficulties. Since that material is not printed here, it suffices for me simply to remark that Rescher and Brandom want axiom-splitting to do two things for set theory. One is to keep the contradiction out. The other is to let the rest of set theory in. It is a fairly straightforward matter to achieve the first goal, but uncommonly difficult to achieve the second. 5 The result is that Rescherian set theory is threatened with a complexity and ad hocness that rivals that of ZF. This is a telling lesson. It helps us see that in theoretical contexts, small changes have a way of turning out to be big ones. Given its paraconsistent credentials, Rescher and Brandom’s logic has a stake in averting ex falso. There is a well-known and much-approved way of doing this. It turns on a distinction between the consequences a sentence has and the consequences of a sentence it would be appropriate to draw. On this view, it is one thing to assert that B is a consequence of A. It is another thing to assert B on grounds that it is a consequence of A. Having, we might say, is up to A. Drawing, we might also say, is up to us. With the distinction at hand, it is left open that there are consequences that a sentence has that would not be appropriate to draw. And with this possibility in mind, one might be inclined to suppose that although everything whatever is a consequence of a contradiction, nothing even close to everything is appropriately drawn from a contradiction. We might even go so far as to say that nothing whatever is appropriately drawn from a contradiction. The distinction between having and drawing is as old as logic itself, and is implicit in the difference between Aristotle’s logic of immediate inference and his logic of syllogisms. In more recent times, the contrast is

499

RESCHER STUDIES

clearly present in Gilbert Harman’s contention that conditions on entailment are one thing and rules of inference are another, and, this being so, that A ∧ ∼A might well entail every sentence without licensing the inference of every sentence therefrom.6 Rescher and Brandom approach ex falso is in the spirit of the distinction between having and drawing, between entailing and inferring. But the details are significantly different. Harman says that the rules of logic aren’t the rules of correct inference. The rules of inference—of consequencedrawing—are non-classical. Rescher and Brandom say something similar. They say that there is nothing wrong with the rules of classical logic. Rather, it is the semantics of his system that can’t be classical. It is well to note there that Rescher is drawn to a distinction between logic in the strict sense, in which the classical approach does just fine, and logic in a more generic sentence that includes Rescher’s own particular handling of semantics. The distinction will become clearer as we proceed. Harman’s view is that it never was the business of logic to give expression to the rules of right reasoning. Rescher says something weaker. In consistent worlds, the laws of logic and the laws of inference are equivalent. This equivalence is lost in impossible worlds. There, and only there, does inference lift away from logic and become non-standard. Rescher is a non-classical minimalist. In their approach to the paradoxes, he and Brandom want to be as classical as is consistent with their resolution of them. In The Logic of Inconsistency, they take pains to make this clear. The rule of adjunction, they say, is valid both as a principle of deduction and as a metatheoretical principle that preserves theoremhood (that is, it is a valid principle of logic in the strict sense). It fails only as what they call a semantic rule (logic in a more generic sense), that is, a rule of the form tw(P), tw(Q) ⇒ tw(P ∧ Q) in which ‘t’ is a truth predicate and ‘w’ indexes to worlds. Below I will say something further about Rescher’s notion of semantic rule and of the role played by ‘t’. But for now I want to say something about the classical latitude offered by both the Harmanian and Rescherian handling of ex falso. To do this, we might ask ourselves what is the job of a mathematical theory, say a theory of sets? The answer, certainly, is that its job is to attain a knowledge of sets. This is done by constructing a theory. A theory is a

500

RESCHER ON APORETICS AND CONSISTENCY

partition of a class of sentences about sets into those to which hold (rather than not hold) and those which don’t hold (rather than hold). The two sets are strictly disjoint, although sometimes the partition that begat them may not be complete. Suppose now that we had a theory in which every sentence held; in which, that is, there is no sentence that holds whose negation doesn’t also hold. This is the so-called Catastrophe Thesis. If ex falso were true, then inconsistent theories would simply blow up, explode, detonate, or what have you. It is easy to see that no such theory could afford us a knowledge of sets—none, not a jot. If ex falso were true, then there is nothing whatever to be learned about sets from intuitive set theory. Frege published the bad news of the paradox in an appendix of volume 2 of the Grundgegetze. If ex falso is true, the preceding pages of the Grundgegetze teaches us not one blessed thing about sets. The same is true of the calculus before the discovery of limits and, later, of the hyperreals. The calculus harboured a known inconsistency for a very long time. Old quantum theory was also known to be inconsistent. If ex falso is true, the calculus told us nothing at all about differentiation and integration, and old quantum theory told us nothing at all about the interior of the atom. As the history of mathematics and science attest, no one has had the slightest inclination to accept these implications. If those histories are anything to go on, inconsistent theories of X routinely give us lots of principled information about X. The Catastrophe Thesis has no legs. In the circumstances of the case before us, if the classical laws of entailment did hold true (and entailment were also theoremhood-preserving), then for no sentence of a negation-inconsistent theory would we have a principled reason to pick it over its negation. The theory would speak equally well for the negation of any sentence it spoke well of. But, again, as the history of actual mathematical practice tells us, this is nonsense. Ex falso can’t be true of entailment. For ease of reference, let us call this the Contra-Catastrophe Thesis (CCT). CCT bears weightily against a certain picture of inconsistencymanagement. The picture is roughly as follows. Suppose that T is an inconsistent theory. If the inconsistency is known to the theorist and if he is unable to block its emergence in a principled way, he is able to work out the findings of his theory by taking pains to keep the inconsistency out of harm’s way. If the theorist is unaware of the inconsistency, he proceeds in the same way, applying tacit check to the inconsistency. So while it may be true that every sentence is derivable in T, the competent theorist restricts his T-findings to a consistent subset of T’s sentences.

501

RESCHER STUDIES

If CCT is true, there is something fatally wrong with this picture. It presupposes that one can discern which particular sentences hold in T in a way that gives you a knowledge of T’s subject-matter, notwithstanding that they all hold in T. It imputes to the theorist capacities for discernment which exceed those exercised by the working mathematician. For in no case would picking a particular winner make a loser of its negation, picked or not. Picking is not in the theorist’s gift. Picking is the prerogative of the theory. One picks what one’s theory separates out as a winner. But in an inconsistent theory everything is a winner and nothing is. I want to turn now to the particular focus of Rescher’s nonadjunctivism. Besides being of interest in its own right, this will help us to assess CCT. Rescher and Brandom are careful to say that adjunction is all right for logic in the strict sense, and that its suppression is intended only as a semantic rule. They emphasize that it is their semantics that is “unorthodox” and non-Tarskian, not their logic (in the strict sense). What is the nature of this semantic heterodoxy? In chapter 6, “Reification fallacies and inappropriate totalities”, Rescher characterizes his semantic rules as principles of communication. He writes that “the salient lesson of these deliberations is that the effective ground-rules of effective communicative practice suffice to side-line a significant range of epistemic puzzles and paradoxes.” We have it then that whatever else they may be, Rescher’s semantic rules are pragmatic rules of communication. They are rules of utterance. Of assertion. Rescher began the pragmaticization of logic as early as 1959.7 A further development was Hintikka in Knowledge and Belief (1962).8 One of the job descriptions of a formal semantics is to define the system’s logical truths. Hintikka includes in this category sentences whose negations would be self-defeating from someone to utter, some of which fail the Tarskian test of having a model in every interpretation. Rescher’s treatment of the truth-predicate, as with expressions in the form t(P) occurring in his semantical rules is also instructive. Truth is not a model theoretic property. Tarski’s semantic maps truth values to sentences via the relations borne by them and their parts to a family of set theoretic structures, which Tarski clearly thinks of as a simulation of the world. The truth and falsity of sentences is a matter of how they stand to the word; it is also a matter of how the world is. In its occurrences in Rescher and Brandom’s semantical rules, the sentence operator ‘t’ is not functioning as Tarski’s truth predicate. Their semantical rules, ‘t’ functions as an assertion operator and assertion is allowed to be world-relative. In world w1 A might be asserted with good

502

RESCHER ON APORETICS AND CONSISTENCY

authority, and in world w2 ∼A might be asserted with good authority. But it cannot be asserted on good authority that A and ∼A. Assertion is Moorean; it is intrinsically truth-committed. I cannot assert “A, but A is not true”. If we now interpret the operator ‘t’ as an assertability operator, this is tantamount to making it a truth predicate of kinds. For an assertible claim is a claim whose assertion honours all its commitments. Assertion is truth from the inside, so to speak. Assertion and assertibility have not stirred much interest in the mainstream of logic, but Rescher has been thinking about it for a long time. His “Assertion logic” appeared in 1968 in Topics in Philosophical Logic.9 Rescherian truth is such that both the inaccessible and misbehaving bits of the world can’t be thought. And what can’t be thought can’t be reasoned from. As Rescher and Brandom write in chapter 3, “The logic of inconsistency”, “[a]ssertions is, of course, only a meaningful prospect in contexts where not everything can be asserted—where there is some exclusion. For assertion must be determinative, and omniis determinatio est negatio …, as Spinoza’s precept has it”. It bears on our reading of Rescher that Jáskowski’s founding investigations were directed to the logic of discussion. Jáskowski called his systems “discussive” logics (and, in a variation, “discursive”). They are a branch of dialogue logic. Discursive systems are non-classical in certain predictable respects. In a conversation, one party might forward the claim that A, and another might assert a proposition incompatible with it. It might even be the case that there is a positive degree of reasoned support for each of A and ∼A. But there are two things Jáskowski won’t allow. He won’t allow that the combination of the reasons for A and the reasons for ∼A constitutes a reason for A ∧ ∼A. Neither will he allow that A ∧ ∼A has even occurred in this discussion (so the question of what are the reasons for it doesn’t actually arise). N-party discussions are non-adjunctive. The rule of adjunction is not a universally valid rule there. Imagine a conversation between two speakers S and S′ over some disputed matter. It is natural to suppose that there will be many propositions on which S and S′ disagree and that, for the most part, they will not shrink from giving them voice. If each is performing self-consistently, we could speak of the worlds constituted by their respective assertions in this dispute. World w1 would be made up of S’s assertions and w2 would be made up of S′’s assertions. Could we now make a further world—not a world “thrust upon” us “by an ineluctable force” of its own (chapter 3), but a world “of our devising” whose “descriptive make up” would be “as we please”? Rescher’s answer is Yes. In Imagining Irreality (2003), Rescher

503

RESCHER STUDIES

writes: “These [= possible worlds], of course, are not possible worlds as such but conceptual constructs”.10 We “construct” the superposition of w1 and w2 to make the impossible world w3. But since the proposed construction would be along discussive lines, it can only be expected that adjunction would not be valid in w3. And whatever the details (the devil is in the details), Rescher and Brandom are entirely right in thinking that in the old calculus, old quantum theory and old set theory, what those theories taught us was produced paraconsistently. Perhaps we now have the basis on which to mount a rejoinder to CCT. Recall that if CCT is true, so is ex falso; and any readiness, whether Harman’s or Rescher and Brandom’s or anyone else’s, to preserve its truth for strict logic is doomed. CCT presumes that a mathematical theory is a class of sentences closed under logical or logico-mathematical consequence. If T is an axiomatic theory, then what T tells us about its subject-matter is conveyed by those axioms and their closure under consequence (whatever the details). As CTT points out, if T’s closure-mechanisms fail to partition T’s sentences in the requisite way, then whatever we might already know about T’s subject-matter or whatever we might come to know about it, this could not be a knowledge afforded by T. Rescher’s insistence, in chapter 3 and elsewhere, that his semantic rules are, in effect, assertibility rules suggests a different conception of what a theory is. Contrary to what is assumed by CCT, a theory is not a set of sentences closed under sentential relations. Rather a theory is a set of assertions, closed under certain “semantic” rules, that is, rules of an assertibility logic. An assertibility logic can guard against catastrophe in two ways. It can forbid the assertoric aggregation of any assertion and its negation. It can also suppress ex falso for assertion. Even if someone did assert something in the form A ∧ ∼A, the assertion of everything would be flatly disallowed (recall the Spinozistic constraint). What do we now make of the theories which provoked CCT— old calculus, old set theory and old quantum theory? How, we asked, could these inconsistent theories give us any instruction if deduction were closed under classical consequence? The answer is that this is all beside the point. Perhaps the old comprehension axiom does indeed entail everything whatever. But Rescherian theories aren’t closed under strict logic. They are closed by the rules of assertion. Could such theories give us instruction about integration and differentiation about sets and about the interior of the atom? Yes, they could. I am not aware of any place in Rescher’s writings, including those represented here in which he either identifies the CCT or marshals the present rebuttal of it. But, if I am not mistaken, all the ele-

504

RESCHER ON APORETICS AND CONSISTENCY

ments required for the rebuttal are prefigured in Rescher’s notion of a semantical rule. I mentioned earlier Rescher’s approach to aporetic resolution. This is done by spotting the weakest link in an aporetic cluster. The procedures of chapter 5 has its origin in Rescher’s theory of plausibility screening, developed in his ground-breaking Plausible Reasoning (1976).11 Consider a simplified example. X, Y and Z are expert witnesses in a criminal trial. Although they agree on some matters, they disagree on others. Thus the totality of their evidence—call it Σ—is inconsistent. We might note in passing that although every sentence of Σ has been attested to by an expert, no one has attested to Σ. The relation of being attested to is not closed under aggregation. Suppose now that there were some reasonable basis on which to rank the reliability of these experts. If we used the numerical scale 1-10, with 10 highest and 1 lowest, we could also rank the reliability of the claims given in evidence; a proposition’s reliability is the same as the reliability-ranking of its source. When confronted with Σ, it is a juror’s job to pick a maximal consistent subset of it. One might think that the best way to do this is to form that set whose propositions have the highest reliability ranking; that is, to accept everything the most highly ranked expert says and reject any contrary claim of the lesser-ranked experts. It is to Rescher’s considerable credit to have taken note of an important empirical fact about resolution of testimonial conflict. It is that often a resolver’s subset will contain the negation of a more highly ranked proposition. Such a proposition might be picked because, for example, it is backed by two experts of less than top ranking. It is true that in figuring out what evidence to believe, Rescher’s method gives overall preference to propositions of high rank. Not in the direct way just noted, but rather by application of the following rule: Pick that maximal consistent subset of Σ that excludes Σ-propositions of least reliability. In the case of a tie, pick the common components of the tied sets.

In Plausible Reasoning, Rescher generalizes the judicial example in an interesting way. Any proposition that one might assert or consider asserting has a source. The source might be sense-perception or memory or mathematical proof or the sayso of a pundit or interlocutor, and so on. Sources in turn are more or less reliable. Sources can also be construed as witnesses to the propositions they ground. Some sources speak their truths literally, others in a more metaphorical way, suggestive of the Book of Nature.

505

RESCHER STUDIES

Sources, then, are sources of information. When the information that comes one’s way is inconsistent, one has the task of conflict resolution. One way of proceeding is to handle all these cases in the way in which testimonial conflicts in court are handled, that is, by applying the method of plausibility screening. Considered as a general method of conflict resolution, plausibility screening embodies some fairly unrealistic idealizations. Leading the list is the assumption that our epistemic sources are subject to credible and predictable reliability-indexing, even on so gross a scale as 1 to 10. Another possible difficulty is Rescher’s equation of plausibility with a degree of source-reliability. In fact, one might say that the greater the reliability of its source, the greater latitude a proposition has on the score of implausibility. Perhaps a more interesting question is whether plausibility screening (or anything like it) actually works for aporetic clusters. We should ask the same question for the closely related issue of paradoxes. I lack the time to give these questions the attention they deserve. But before moving on, a few brief observations might be in order. First, in the selections printed here, especially in chapter 4, Rescher also entertains what can only be consider a rival of plausibility screening as a method of conflict-resolution. It is the econometric method of costs and benefits crudely formulable as follows: Pick that maximal consistent subset of Σ that excludes an Σ-proposition at least over-all cost to one’s epistemic priorities.

The dominant conception is now of a proposition’s priority-ranking, which while a generalization of plausibility-ranking, is one in which some of the trouble with plausibility evaporates. This, I think, is a considerable improvement over the provisions of 1976. There may well be classes of cases in which the two methods trigger the same exclusions, but they are far from inequivalent. A second point is that a defining characteristic of aporetic clusters is not only that the propositions they contain are highly plausible, but it is not always possible to discern reliability-variations in the sources of these conflicting plausibilities. But, even assuming tenability of the equation of the most plausible and the most reliably-sourced, consider again the deterministic cluster. If we opted for plausibility screening, we would order the conflicting propositions according to source-reliability. But what are the sources of these propositions? Perhaps they are rooted in “our” intuitions or in “common knowledge”. If so, the prospect of ordering would seem to

506

RESCHER ON APORETICS AND CONSISTENCY

evaporate. But if we opted for the econometric method in which priorities were ordered by costs and benefits, there would be some clear advantages. One is that we would not need to worry about plausibilistic/reliability indiscernibilities. Another is that weakest-link identification need not be an affair of the moment. The econometric method leaves plenty of time to sort out what the costs of the various exclusions actually turn out to be. A case in point is the handling of ex falso by the early proponents of relevant logic. By a large majority, relevant logicians elected to crimp the reach of the law of disjunctive syllogism. It was not realized at the time that the cost of this rejection is the thorough-going intensionalization of the resulting logic. As Quine observed, when we meddle with a given connective, we meddle with them all. (Recall difficulties of the same sort attending the Rescher-Brandom repair of intuitive set theory.) The determinism cluster is important in an obvious way. It can serve as a test case for Rescherian weakest-link identification. If we follow the procedures of chapter 4, we are to do two things. We must first identify something as “a speculative supposition” in virtue of which the cluster is engendered. We must then reject whatever proposition that is “least consonant with the fundamental commitments of [our] overall position”. This may strike the reader as problematic on both fronts. Which is the speculative supposition that does all the damage here? Perhaps Rescher would say that they all are. Which, then least comports with the fundamentals of our overall position. Consider the options: 1.

Human beings are in part outside the natural order.

2.

The law of causality breaks down in the mezzo-world.

3.

Human beings aren’t free after all.

Hobson’s Choice, some would say, and historically none of these is the preferred solution anyway. The preferred solution is to disambiguate, and to do so in ways that all the propositions in the cluster turn out to be both true and welcome. This is precisely what compatibilism seeks to deliver. True, it is in its own right a contested solution, but it’s one whose heart is in the right place. It seeks for something for these conflicting intuitions to be true of. For one sense of “cause”, everything in the natural order is caused. For another sense of “cause”, anything caused is incompatible with

507

RESCHER STUDIES

freedom, yet lots and lots of human actions aren’t caused. Solutions of this sort deserve a name. Let us say that they are a kind of reconciliation. A third observation also concerns costs and benefits. There is a wellestablished mathematical theory of costs and benefits. In speaking of the econometric method of conflict resolution, one might suppose that the conflict-resolver applies the theorems—tacitly perhaps—of this mathematical model. There is not much in the way of empirical evidence to suggest that this is actually the case. We need only reflect on the history of the determinism-free will conflict to see that precisely what is missing here are the calculations mandated by the mathematical model. This suggests not only that as used here the cost-benefit idiom is a metaphor. It also suggests that its cash value is dialectical rather than mathematical. Seen this way, the value one attaches to a claim is reflected in what one is prepared to give up to old it consistently. Embedded in this conception is a third notion of conflict resolution. One finds it championed in Henry Johnstone’s thesis12 that philosophical argument and counter-argument has the ad hominem character of arguments discussed (without disapproval) by Locke in the Inquiry Concerning Human Understanding (1690). It is a form of argument in which one party presses the other with consequences of the latter’s “own principles and concessions”, and in that way seeks his surrender. The dialectical conception of cost-benefit conflict resolution is an easy generalization of this notion. One of the virtues of Johnstone’s approach is that since different people often value things differently, philosophical pluralism is an entirely intelligible state of affairs. And, of course, pluralism is an issue about which Rescher himself has written valuably and with insight.13 In bringing this introduction to a close, I want to say a few words about chapter 7, which is a discussion of an early paper on plural quantification, originally published in 1964, and reprinted here as an appendix. Part of what makes the Rescher-quantifier interesting and important is the kind of philosophical issue to which it is a response. Also of interest is its timeliness. What is striking about Rescher’s writings is, among their other very obvious virtues, is how ahead of the curve they tend to be. Rescher is not the originator of all that he writes about, but he has an admirable ability to spot trend-lines very early. Plural quantification is a case in point. Not much was going on with plural quantifiers in 1964, but today they are all the rage.14 Especially important is what motivated Rescher’s interest in them. One way in which to be drawn to the subject is by noticing the plain linguistic fact that natural languages possess quantifiers other than those of mainstream logic. They are right “there” in the language, so wouldn’t it be

508

RESCHER ON APORETICS AND CONSISTENCY

interesting to expose their logical structure? This was not Rescher’s motivation. Rescher was struck by the frequency with which the lawlike generalizations of successful science failed the standard of strict universality. The universal quantifier overstates the insights of much of science. Rescher was not alone in noticing this. In 1959 Michael Scriven published an important paper on “normic” claims in science, statements that hold for the most part.15 (And before Scriven, Harold Langford was writing about generic claims in a paper he published in 1949).16 Scriven and Rescher appear to mark two different approaches to nonuniversal generalization, Scriven anticipating developments in the logic of generic statements and Rescher taking the different route of plural quantification. For certain classes of cases, there are significant differences between the two, as may be seen by considering the four-leggedness of ocelots or the malaria-carriage of the Aropheles mosquito. We might note not only that “Ocelots are four-legged” is not a Rescher quantification, but that in any event not every generic statement even implies a corresponding Rescher-quantification. “The Aropheles mosquito carries malaria” is a true generic claim, but fewer than 5% of these creatures are carriers.17 The semantically distinctive feature of generic truths is that they can remain true in the case of true negative instances. The same is also true of “Most ocelots are four-legged”, but with a large difference: a three-legged ocelot is an exception to the generic claim but not to the Rescher-quantification claim. Non-universal generalizations exhibit even more heterogeneity than one finds in the contrast between Rescher-quantification and the genericity of the four-leggedness of ocelots. Aside from the obvious variations on Rescher’s “most” (“many”, “several”, etc.), we also have it that genericity comes in varying kinds, covering among other things claims about what is typical, usual, characteristic and normal. Even “Ocelots are four-legged” can claim membership in this set, but with a peculiarity: “Ocelots are fourlegged” is always true except where something has gone wrong, such as congenital defect or injury. Accordingly, notwithstanding the intuitions noted just above, “Ocelots are four-legged” is true just when “All nondefective ocelots are four-legged” is also true. It might be supposed that we get a similar paraphrase with typicality-generics (“Zs typically W”) and characteristicness-generics “(Xs characteristically Y”), as follows: i. It is characteristic of Xs that all of them Y.

509

RESCHER STUDIES

ii. It is typical of Zs that all of them W. But this is wrong. It is characteristic of Xs that most of them Y, and it is typical of Zs that most of them W. So some generics—quite a range of them in fact—do indeed incorporate the Rescher-quantifier. Here again, as early as 1964, Nicholas Rescher was on to something of major importance.

NOTES 1

John Woods, Paradox and Paraconsistency: Conflict Resolution in the Abstract Sciences. Cambridge: Cambridge University Press, 2003; pp. 14-16 and 76-78.

2

Steven Yablo, “Paradox without self-reference”, Analysis, 53, 251-252, 1993.

3

For the tie between preservationist logics and Rescher and Brandom’s brand of non-adjunctive logic, see Bryson Brown, “Rational inconsistency and reasoning”, Informal Logic, 14 (1992), 5-10.

4

S. Jaśkowski, “On the rules of suppositions in formal logic”, reprinted in Storrs McCall, editor, Polish Logic: 1920-1939, Oxford: Oxford University Press, 1967. First published in Studia Logica, 1 (1934).

5

See, for example, David Makinson’s review in the Journal of Symbolic Logic; 47 (1982) 233-236 and A J. Dale’s “The illogic of inconsistency”, Philosophical Studies, 46 (1984), 417- 425.

6

Gilbert Harman, “Induction”. In Marshall Swain, editor, Induction, Acceptance and Rational Belief.. Dordrecht: Reidel 1970.

7

“The Logic of existence and denotation”, Philosophical Review, 1959. See also “Belief Contravening Suppositions”, Philosophical Review, 1961 and “Plausible Implication”, 1961.

8

Jaakko Hintikka, Knowledge and Belief, Ithaca, NY: Cornell University Press 1962. Reprinted under the same title by College Publications, London, in 2005.

9

Dordrecht and Boston: Reidel.

10

Chicago and LaSalle: Open Court; p. 144, n. 9.

11

Assen and Amsterdam: van Gorcum

12

Henry W. Johnstone, Jr., “Philosophy and the Argumentum ad Hominem”, Journal of Philosophy, 49 (1952) 489-498.

510

RESCHER ON APORETICS AND CONSISTENCY

NOTES 13

Pluralism: Against the Demand for Consensus. Oxford: Clarendon Press 1995.

14

See, for example, Gila Sher, The Bounds of Logic, Cambridge, MA: MIT Press 1991.

15

Michael Scriven, “Truisms as the grounds for historical explanations”. In Patrick Gardiner, editor Theories of History, pp. 443-75. Glencoe IL: Free Press 1959.

16

C.H. Langford, “The institutional use of ‘the’”, Philosophy and Phenomenological Research, 10, (1949), 115-120.

17

See here Manfred Krifka, Francis Jeffry Pelletier, Gregory N. Carlson, Alice ter Meulen, Gennaro Chierchia and Godehard Link, “Genericity: An Introduction”. In Gregory N. Carlson and Francis Jeffry Pelletier, editors, The Generic Book, pages 1-124. Chicago: University of Chicago Press 1995; p. 44.

511

Responses to the Contributed Essays

B

efore embarking on specific comments on the particular contributions to this volume, I would like to express my appreciation to those who have contributed. They have diverted time and effort from their busy academic lives to perform this act of supererogation, and I am deeply grateful to them for this. Beyond this, I would like to seize this opportunity to make a brief observation regarding the general tenor of reactions to my position regarding the fundamental issues of ontology. Here my readers sometimes reveal a certain queasiness about the postulational realism that I defend on the basis of pragmatic considerations. They sometimes incline to think that an endorsement or realism on pragmatic grounds leaves the idea of mindindependent reality stranded in the shallows of convenient fiction. I myself certainly do not see the matter in this light. Rather, my approach takes the more robust line of the following three-fold challenges: • If there is indeed mind-independent reality then what sorts of grounds for its acceptance could possibly be available to us? • When duly identified, just which of these can one reasonably complain of as being absent? • If we indeed do have all the substantiation that there possibly could be, then what more could we reasonably ask for? On this basis, so it seems to me, a robust realism can be fitted out with a justificating rationale of plausible cogency. We are emphatically not being asked to be realists on grounds that it is somehow convenient to do so. Rather, the position is enjoined upon us because—as I see it—the stern realities of the situation are such that we do actually have in substantiation of ontological realism everything that could, in the circumstances, reasonably be asked for.

RESCHER STUDIES

RESPONSE TO ROBERT ALMEDER It has long been a source of pleasure and benefit to me to exchange philosophical ideas and arguments with my old friend Bob Almeder. His lively and penetrating mind has helped me to clarify and refine my ideas on various key topics in the philosophy of knowledge. The main point I want to pursue on this occasion relates to the doubt that Almeder casts on my contention: “that scientific progress can continue forever and will never be completed.” I grant that this view seems controversial, but I believe that there is effectively conclusive reason for it, reason that lies in what I characterize as the technological dependency of natural science. The leading idea here is that the progress of natural science hinges on technologically mediated observational and experimental interactions with nature. The result is one not of geographical exploration but rather of the physical exploration—and subsequent theoretical systematization—of phenomena distributed over the parametric space of the physical forces spreading out all about us. This approach in terms of exploration provides a conception of scientific research as a prospecting search for the new phenomena demanded by significant new scientific findings. As the range of telescopes, the energy of particle accelerators, the effectiveness of lowtemperature instrumentation, the potency of pressurization equipment, the power of vacuum-creating contrivances, and the accuracy of measurement apparatus increases—that is, as our capacity to move about in the parametric space of the physical world is enhanced—new phenomena come into view. After the major findings accessible via the data of a given level of technological sophistication have been achieved, further major findings become realizable only when one ascends to a higher level of sophistication in data-relevant technology. Thus the key to the great progress of contemporary physics lies in the enormous strides which an ever more sophisticated scientific technology has made possible through enlarging the observational and experimental basis of our theoretical knowledge of natural processes.1 The crux, however, is that technological progress confronts a natural resistance barrier. As we increase the power of those armamentaria of scientific investigation, the prospect of further improvements, while never altogether blocked, becomes increasingly difficult and demanding. Every additional step gets to be as costly in its resource requirements as the sum total

514

REPLIES

of all of those that have preceded. The prospect of improvement is always there, although its realization becomes increasingly impracticable. And yet with every step forward there are new phenomena, new problems, new questions, new so-far unresolved issues. The prevailing situation never provides for the as-yet-unrealizable data and these data provide for new issues and open up new horizons. And it is this technological escalation which, as I see it, provides the fundamental and inescapable ground for the in-principle incompletability of the scientific project. The nature of its technological dependency means that natural science is embarked on a journey that can never actually be completed. Nature’s resistance barriers predate the perfectability of natural science. RESPONSE TO DIDERIK BATENS Diderik Batens offers an interesting and insightful critique of my ideas regarding thought experiments, hypothetical possibility, and imaginary worlds. I am grateful for his suggestive discussion, though here I can only deal briefly with a few points. Logical Tolerance. Batens is entirely right that I am tolerant with regard to logic. And this holds both as regards our logica utens, the logic we use in our reasonings, and the logica docens, the logic we employ in formatizing, explaining, and expositing this aforementioned logic. (Ideally these two should become convergently one and the same.) However, once such a logic is settled on, the room for further tolerance is (or should) diminish to the point of virtual nothingness. Once we have a logic in place, the issue of consequence and consistency relative to this accepted logic should almost universally become a fait accompli. The ongoing development of logical theory diminishes the room for tolerance. Mathematical objects. These differ from purely hypothetical (or “imaginary”) individuals being abstract rather than (putatively) concrete. Hence mathematical objects admit of identifying descriptions that are complete in a way that merely possible objects do not. For language alone is insufficient to identify them. For no merely finite formula can specify in individuating detail the features of a nonexistent item whose nature is (supposedly) concrete. Here the distinction between concreta and abstracta is, as I see it, absolutely fundamental. The detail of the concrete precludes purely linguistic identification. Possibilities and Language. The ontological reach of language is limited. The denumerability of the linguistic formula of arithmetic means that

515

RESCHER STUDIES

there are more real numbers than can ever be named and identified. And given the recursive nature of language there will be more problematic than can ever be identified and considered. The possible worlds that theorists so glibly invoke can almost never be specified: it takes more than hints to individual worlds. A vast gulf separates language and ontology where possible worlds and possible individuals are concerned. And so realism and modesty alike speak for a semantics of scenarios as against one of possible individuals and worlds, because the former—in effect “stories”—are linguistic objects we can get a grip on. Here we can achieve by honest toil what is otherwise not even actually accessible to theft. While Batens acknowledges that the reach of language is far less than that of possibility, he insists that those unidentifiable possibilities are nonetheless perfectly good, real, acceptable possibilities. But are they? I myself see them as only second-order possibilities. They could possibly be identified by possible identifiers. But they are not first-order possibilities at all. Consider an analogy. The ocean contains a lot of water. And it contains zillions of little splugets of water. But until you take one out and form a drop with it, there just is no drop there. Until that drop is actually formed and individuated there just is no drop there. It is not literally correct to say that the ocean consists of countless drops of water. Counterfactual reasoning. Here I fully acknowledge the utility of Batens’ suggested resort to adaptive logic. However, my own way of handling the issues in terms of plausibility and presumption based on finite comparisons has (I think) the comparative advantage of intuitive accessibility and practicable economy. RESPONSE TO BRYSON BROWN I much appreciate my former student Bryson Brown’s clear and constructive exposition and development of the coherentist position I advocate. Brown’s discussion is accurate in diagnosing the guiding ideas of the coherentist’s program as I have endeavored to cultivate it. For while any sensible epistemic program must encompass the idea of what truth is, it can dispense with the foundationalist idea that we must already from the very outset possess much knowledge of what is true. In rational inquiry the demand for truths need not be met at the outset (as would have to be the case if deduction were our only inferential resource) but can be postponed until later stages of the game.

516

REPLIES

And in this regard, Brown’s deliberations about the logic of truth-claims are illuminatingly constructive, and gets good mileage from semantics of classically binary (1-true, 0-false) situations. However, much could also be done with a multi-valent semantics whose inferential rules follow a weakest-link principle with valid references as those which “never make matters worse.” At any rate the coherentist’s insistence on consistency (i.e., the non-derivability of contradictions) is going to prove something of a loose end, with the prospect of variation with respect to derivation. But there is no reason for the coherentist to feel embarrassed about this because everybody else has to come to terms with it as well. I found Brown’s comments on “The Wheel” particularly interesting, because he rightly stresses that despite my appeal to this trope I reject both the bottom-up epistemic approach from certified truths to validating standard and the reverse move from truth-guaranteeing principles of validation to thereby certified contentions. The process I envision is more convoluted and complex. Most critics do not have the patience to see how the pieces fit together here. Brown to his credit ably does so. The obvious difficulty with a Chisholmian resort to “the evident” in a viable program of cognitive validation is that this category lumps the sheep together with the goats—exactly as the skeptics of classical antiquity already emphasized. And the merit of coherence, duly construed, is that it affords the best process of separation that is operatively available to us. Still the devil is in the details here, and Brown is admirably clear-eyed in spotting them. However, one aspect of his discussion that leaves me uneasy is that it does not indicate the extent to which the concept of plausibility (that is, prima facie or “initial” credibility as it is sometimes called) is pivotal for my approach to the theory of knowledge and inquiry. That this is so perhaps lurks between the lines of the penultimate paragraph of Brown’s exposition, but I see it as something that should be right up there, front and center. Interestingly enough, the principle at issue figures prominently in Brown’s exposition of his own epistemic allegiance. Consider his statement It’s pretty clear that “looking and seeing” is a good method for arriving at certain kinds of truth claims. In fact, like Moore; I may well be more confident of this than I am of the premises of sceptical arguments.

517

RESCHER STUDIES

In this instructive avowal Brown in one breath both acknowledges the plausibility (prima facie credibility) of certain claims, and endorses the principle of not arguing for a problematic thesis from premises that are even more so. Coherentism is not just a theory: it is a living process of rational procedure. Finally, I cannot conclude without acknowledging Brown’s splendid epigraph from Albrecht von Haller (1708-77). In an era of ever-fading Latinity, it deserves translation: “Nature is connected in a network of its components, not in a chain-like sequencing. Humans, however, can only proceed in a chain-like succession, since discourse cannot present several items at once.” RESPONSE TO JAMES FELT In his interesting and instructive discussion of my contention that the epistemic status of our commitment to mind-independent reality is that of a postulate, James Felt “challenges Rescher’s starting point by arguing that . . . extra-mental reality cannot be argued to because it is immediately evident . . . [so that] to claim that one must postulate it is to adopt only a second-best kind of realism.” Given this formulation of the divergence of our views, the following points seem to me in order. Felt and I are in complete agreement with respect to the existence of a realm of mind-independent reality that mind-equipped beings find but do not make. However, while both of us agree to the fact of realism we differ as to the epidemic status of an acceptance of this fact. I represent it as a rationally warranted postulate, Felt represents it as immediately evident. And as I see it, Felt’s formulation of his objection overlooks the rather important consideration that in the first instance postulates are not argued to but argued from. Felt wants to insist that extra-mental reality is not inferred— is epistemically not the product of an inference. He fails to acknowledge that my insistence on postulation puts me entirely on his side at any rate in this regard. The point I want to stress is that we must distinguish clearly between the epistemic route leading to acceptance of a contention and the contention that is accepted. And the nature of this probative route does not affect the substance of the conclusion. Thus a long-winded argument to a conclusion is not an argument to a long-minded conclusion, nor is a clever substantiation of a thesis necessarily the substantiation of a clever thesis. And

518

REPLIES

in this light one must recognize that the epistemic basis for accepting a contention does not affect the substance of the contention itself. Accordingly, the mind-projected (or postulated) acceptance of a realm of reality is not the acceptance of a mind-projected reality. Felt himself favors Etienne Gilson’s appeal to the “données immédiates de la conscience”. But Gilson’s appeal to such givens—and Whitehead’s “immediate awareness” which Felt brings into battle as a reinforcement— seem to me to come up against one and the same difficulty: the pervasive gap between seeming and being. Consider on the other hand the sort of reasoning that Felt propounds later on in his discussion: Let us then tentatively suppose that our feeling of being impacted by a surrounding [external] world may just be authentic, revalating of, because instantiating, things that are actually “out there.” That immediate impression . . . is an experience in need of being explained . . . [And] we have good metaphysical grounds for its explanation.

But what is this movement for tentative supposition to accepted explanatory fact other than the sort of “inference to the best explanation” that any appeal to immediate evidence is designed to avert. I do not think that this is really the sort of thing that Felt really wants, and I confidently believe that it cannot be made to work. We cannot extract from experience any sort of externality that we do not tacitly smuggle into it. Externality as I see it is not given but taken. In the end, the problem with those so-called immediately self-evident facts is that there is bound to be a gap between what is seen as (regarded as, taken to be) self-evident and what actually is mind authentically selfrevelating and self-substantiating. The all-too-pervasive gap between what seems and what is—between subjectivity and objectivity—enters also upon the scene with respect to self-evidentiation, On the other hand, with postulation we ourselves are in charge. And here we can resort to the ex post facto validation of seeing with the wisdom of eventual hindsight that this postulation was in fact well advised. So we can have our substantiation both ways: up front and after the fact. The role of experience is crucial here. But the term has two senses, the one episodic (Erlebnis: “I had a strange experience yesterday evening”), and the other systemic (Erfahrung: “Experience teaches that _ _ _.”) The trouble with the “experience” at issue with immediate-evidentiation is its episodic nature. For this spells its potential fallibility: that what it puta-

519

RESCHER STUDIES

tively shows cannot appropriately be accepted as self-authenticatingly definitive without further ado. On the other hand, “experience” in the systemic sense is something else again able to do effective probative work. But this comes only with the wisdom of hindsight and what is needed is foresight. The advantage of postulation is that it enables us to have it both ways: to take a leap forward and then retrovalidate doing so in the light of systemic experience. As the immediate-experience validation contemplated by Felt sees it, experience hands Reality to us on a silver platter—no work required. As I see it more of an effort on our own front is demanded here. But what we are agreed on—and this is important because it flies in the face of much empiricist theorizing—is that what needs to be done is not a matter of inference. For postulation after all is not a matter of inferring something to be so, but of taking it as such. However, this taking is also something quite different from anything of the order of an experiential giving. And as I see it, this idea of “the given” reflects the unrealistic hope of obtaining by gift (if not by theft) something that is available only through honest toil. I do not quarrel with Felt’s contention that “our experience presents itself prima facie as encountering a world, not our sense impressions of a world.” (How could it be otherwise?!) But this of course is as true of mirages as of actual sightings. Immediate experience alone cannot certify its own authenticity. The step from appearance to reality is neither dispensable nor achievable by direct evidentiation alone—there has to be a prior determination of what can count as evidence. And so there has to be an initial presumption before evidence can get a purchase hold on the issue. Only after that past-postulational stage can evidence retrovalidate (i.e., retrospectively evidentiate) that postulation. Thus here some actual toil is, as I see is, unavoidable. RESPONSE TO LENN GOODMAN Notwithstanding my skepticism towards the demand for consensus, I must avow a preference for agreement over disagreement. And in this regard I find it heartening that so sensible, sensitive, and astute a humanist as Lenn Goodman finds much to approve of in my deliberations about consensus. The key point of my Pluralism book is one whose importance I still cannot overemphasize, namely that the salient task of contemporary public policy is the judicious use of sticks and carrots to create conditions in which people who disagree with one another—even on fundamentals—can

520

REPLIES

nevertheless coexist peacefully and even to some extent collaborate productively. And in this regard I see Goodman’s emphasis on shared interests rather than shared opinions as being right on target. The Philosopher’s stone of recent days does not aim to turn base metal into gold but to extract normative value from descriptive fact. J. S. Mill toyed with the idea of a shift from being desired to being desirable. Consensus theorists envision a transit from agreement to agreementworthiness. Social anthropologists want to reduce morals (behavioral propriety) to mores (behavioral custom). None of these moves will work. The principle ex nihilo nihil also applies in rational deliberation: if you seek normativity out you must put normativity in; if you want normative conclusions you must have normative premisses. So if you are unwilling to undertake normative commitments on your own account and responsibility, you had best give up on the whole normative enterprise. This is something you can certainly do. But it will not earn you the Good Housekeeping approval of common sense, let alone rationality. But now if normative engagement is called for whence can it come? Is it simply a matter of arbitrariness—of taking a stance—never mind why and how? Is one set of norms as good as any other? Surely not! Yet if normativity cannot be based entirely on fact, whence can it come? In the final analysis it comes not from investigation but from affective deliberation, not from intellection but from judgment. Yet can such judgment not go amiss? Alas it can. Granted we humans cannot achieve a God’s eye point of view freed from all that is potentially parochial in cultural and personal relativity. But we can certainly enlarge and broaden our perspective of consideration. The search for consensus is not our only resort here. The motto Tout comprendue c’est tout accepter would not get it right—understanding how others act is not necessarily to accept its appropriateness even for them, let alone as a model for ourselves. But it is (by hypothesis) to understand it—to render it humanly comprehensible. And this is a critical factor. For frequently when we consider what someone has done we say that “We just cannot understand how he could have done that.” Despite our best efforts we cannot integrate those acts into the normal repertoire of human agency and cannot but send them into a classifactory exile from the human sphere. Any rational social polity will encounter its limits, and even the best of states will need its police and prisons. There will always be sociopaths, madmen, fanatics—people whose pleasure lies in the infliction of pain and anguish. The best hope for effective containment here lies in the widely

521

RESCHER STUDIES

shared interest of the preponderant majority in achieving satisfactory conditions of life. And yet Goodman is entirely right here to raise the problem of fallibilism—to human imperfection in regard to information and evolution, to intellection and judgment. While these are issues that have preoccupied me greatly in recent years, Goodman inclines to think I am too cautious here. Winston Churchill said of Clement Atlee that he was a modest man who had a great deal to be modest about. But I believe that just this is the case with us all. It has always struck me that the very fact that we have to live our lives amidst conditions of imperfect information about a world we inadequately comprehend itself carries important lessons. Prominent among them is that those grand ideological systems have their vulnerabilities in relation to the everyday specificities of concrete situations. The siren call of grand theories—however appealing—should never lead us to sacrifice the concrete common sense of everyday life—let alone to sacrifice those who live it. RESPONSE TO JOHN HALDANE I am grateful to John Haldane for his knowledgeable and thoughtful deliberations on issues in the philosophy of religion. In particular, I found much of interest in his comments on the relation between philosophy of religion and analytic philosophy. In ruminating on this theme the following line of thought sprung to mind. It seems to me that in the course of its development, analytic philosophy underwent a sea change as the twentieth century moved on. Initially launched with a view particularly to matters of logic and natural science, it subsequently enlarged its focus of attention to embrace increasingly matters of humanistic concern in ethics, history, and rational deliberation. In this transit, its thematic range increasingly encompassed the Geisteswissenschaften in addition to the Naturwissenschaften and Formalwissenschaften of its earlier emphasis. Given this phenomenon, the diffusion of the analytical ethos amongst Anglophone philosophers made it inevitable that the philosophy of religion would eventually be drawn into its orbit. Unquestionably some institutions (Notre Dame, for example) and some societies (The Society for Christian Philosophers) were prominent in this trend, but they were as much part of its effect as of its cause. The spirit of the times (Zeitgeist) was astir and those of us who moved in this direction were as much its servants as its masters.

522

REPLIES

On to another theme. With respect to the lawfulness of nature (and so not its particular laws) Immanuel Kant wrote that: We ourselves introduce the order and regularity in the phenomena; what we enable nature, . . . However exaggerated and absurd it may sound, the understanding is itself the source of the laws of nature. (CPuR, A126-27.) What Kant here says regarding the lawful order of nature is echoed in the idea—critiqued by Haldane—that “We can only find meaning and make sense if we are ourselves the authors of meaning and of sense.” But just here we must, I think, pay close attention to that little word “we.” Neither Kant nor I nor—ultimately, I would think—Haldane want this to be construed distributively, personalistically, and psychologically. Rather, it has to be construed as meant collectively, impersonally, and geared to humannature-in-general. Be it in matters of cognition, valuation, or praxis, making sense of things should ultimately be a communal enterprise. And here the praxis of the community is our readiest guide in travelling beyond the eccentricities of solipsism. Tradition matters not because it is infallible— since the expulsion form Eden we humans simply have no infallible guide—but because it affords a prospect whose alternatives are less promising yet. Whenever we place our bets in matters of moment, we act, in the final analysis, on our own responsibility. (I do not offload responsibility by following your advice—the decision to do so is unavoidably mine.) But in bearing this responsibility I act foolishly when I refuse to avail myself of the benefit afforded by other and wiser heads. RESPONSE TO WILLIAM JAWORSKI I am grateful to William Jaworski for his perceptive discussion of some of my writings on issue of metaphysics. He acutely apprehends that there is something ironic to the way I propose to defend a basically realistic commitment to a reality that transcends mind by argumentation of a decidedly idealistic cost. For my defense of realism is based on the idea that it is exactly by taking this position that we can best realize the characteristic mission of mind in enabling us to achieve success in explaining predicting and controlling the world we live in.

523

RESCHER STUDIES

Then too Jaworski is keenly aware of yet another non-standard feature to the realism I endorse, one that resists mapping the distribution of fact and value into a separation of the realm of mind from that of nature. For I see no separation here but insist on an acknowledgement of evaluative facts and an embedding of value in the realm of nature’s realities. A commitment to ideals and values is as natural to us as beings of the sort we are as is the need for air or food. Accordingly, norms and values are, as I see it, not artifacts of thought but characteristic features of various sectors of nature’s modus operandi. And in this context I much appreciate the open-minded and generously accepting spirit displayed by Jaworski as regards the Leibniz imposed optimalism that stands at the core of my metaphysical position. In recent days, philosophers have attenuated the line of separation between fact and value. But the closer these two factors come, the easier the transit between them. And as the barrier of separation grows thinner, this transit is bound to be a two-way street. Here however theorists have tended to envision only one direction in seeking to naturalize value. But the road goes two ways and the reverse direction of valuizing nature is no less available. And to my mind what is at issue here is not a matter of wishful thinking but a foray into the realm of “inference to the best systematization” made in the interests of providing a viable account of the nature of reality at the level of grand theory. The one caveat I have about Jaworski’s helpful discussion relates to the sharpness of his contrast between a metaphysic of morals that is Kantianly geared to rational beings at large (little green men included) and an Aristotelian one geared specifically to humans as a biological species. In this regards Jaworski places me on the Aristotelian side. But as I myself envision it, the situation is a bit more complex. For as I view the matter, it lies in our nature as the sort of beings we (biologically) are to take the line that we can, do, and should regard ourselves as rational beings with all the privileges and obligations theoretically appertaining. Accordingly, as I see it, a full-fledgedly Kantian perspective rides piggy-back as it were, on the Aristotelian perspective which Jaworski rightly ascribes to me. RESPONSE TO ULRICH MAJER Ulrich Majer’s stimulating discussion brings to the fore the intriguing question of the limits of science which has long figured in my writing. However, Majer’s discussion provides a vivid reminder of the pitfalls in

524

REPLIES

writing about the work of someone who has himself published extensively over many years. When critiquing the work of a philosopher, one is well advised to have at them for what they say, not for what they omit. With errors of commission, focus on a single sentence will often serve. But with errors of omission, one comes up against the entire opus. Majer writes: [Kant draws] the subline distinction between ‘Grenzen’ [boundary] and ‘Schranken’ [limit] . . . Unfortunately, Rescher makes no terminological distinction between limits and boundaries.

Now this may be true with regard to the particular passage that lies immediate in Majer’s field of vision. However, when one looks further afield to my 2005 book on Epistemic Logic one reads: Even though the thought of finite beings is destined ever to be finite, it nevertheless has no fixed and determinable limits. The line of thought operative in these deliberations was already mooted by Kant: [I]n natural philosophy, human reason admits of limits “excluding limits,” Schranken) but not of boundaries (“terminating limits,” Grenzen), namely, it admits that something indeed lies without it, at which it can never arrive, but not that it will at any point find completion in its internal progress . . . [T]he possibility of new discoveries is infinite: and the same is the case with the discovery of new properties of nature, of new powers and laws by continued experience and its rational combination . . .2 And here Kant was right—even on Leibnizian principles. The cognitive range of finite beings is indeed limited. But it is also boundless because it is not limited in a way that blocks the prospect of cognitive access to ever new and continually different facts thereby affording an ever ampler and ever more accurate account of reality.

Rather than overlooking this issue it figures quite prominently in my deliberations. Majer presses me hard, and rightly so, on the issue of demarcation between the questions that are proper objects for scientific resolution and those that are not, between the inside and the outside of scientific concern. Two comments come to mind here: (1) Irrespective of how one draws the boundary, and indeed even irrespective of whether a precise boundary can

525

RESCHER STUDIES

be drawn at all, the fact remains that some issues are clearly and patently inside and others clearly and patently outside. (2) That as a first approximation issues of what is to be said (correctly and with due care for truth and falsity) belong to science, and issues of what is to be done (appropriately and with due care from what is right and proper) are science-external. On the one hand this has the consequence that issues of efficient means to given ends belong to science, to the determination of how things work in the world. And on the other hand it means that issues of appropriate ends for the intelligent design of means belong to extra-science. But as things go, these two considerations of ends and means are often so intricately intertwined that no easy separation between the two is readily practicable. But then no-one has ever guaranteed to us that philosophizing should be simple. RESPONSE TO DIEGO MARCONI I am grateful to my former student Diego Marconi for his astute diagnosis and insightful analysis regarding problems of possibility/conceivability coordination. His subtle suggestions strike me as constructive and fertile, although my own practice has been to address these issues by more traditionalistic and perhaps thereby also less technical means. Marconi is certainly right to stress that conceivability talk has its difficulties. How do we know what future humans, let alone non-human aliens can accomplish in the way of thought? The answer, of course, is that we don’t. But what we can, I think, maintain is that in saying that somethingor-other is possible we stand committed to the claim that in principle some (finite) mind can get a coherent grip on it, so that possibility is effectively tantamount to conceivability (i.e., coherent thinkability). And our own incapacity does not matter here. It is not being claimed that possibility is thinkability-by-us-here-and-now. For the issue is not one of what you or I or Wittgenstein can conceive of—or at least is so only in one direction: what these individuals can conceive of is ipso facto conceivable, whereas the reverse does not hold. What I have proposed doing is not to “reduce” possibility to something different, but rather to coordinate possibility-at-large with one particular sector of itself, namely possible conceivability. And this is inherently something ideal, not something actualistically geared to you or me or Wittgenstein or homo sapiens. Here we once more reach one of those junctures where idealization is called for.

526

REPLIES

Again, I join with Marconi in rejecting a linguistic approach to conceivability/possibility. For it is clear that language, being developed recursively from a finite basis, can reach no further than the denumerable, while the domain of the real (let alone that of the possible) is presumably of transcendental complexity in its detail. As regards Hesperus/Phosphorus, Salt/NaCl, and the issue of a posteriori necessity, I think we do well to re-deploy the medieval distinction between an absolute and conditional necessity. Given certain actualistic facts about how things stand in the world, certain necessities obtain: given the chemico-minerological realities we can unfailingly identify salt with NaCl. But the necessities at issue, such as they are, are conditional necessities. And those “givens” on which they are predicated are mere takens that could in theory be mistaken. So all in all, I would recategorize those a posteriori necessities as conditional ones. The crux here is that we conceptualize by means of the concepts we have in place, which may well be based on erroneous presuppositions, and could, in theory, even entrap us in impossibilities. With conceivability too the classical distinction between appearance and reality cannot be averted. With possibility/conceivability—as with so much else—there is the unavoidable fact that we can be mistaken, and to be so in ways somewhere between difficult and impossible for us to see. But this merely reemphasizes the element of idealization at issue in the proposed equating of possibilities with coherent conceivability. So while Putnam may be right in holding the “What is conceivable is a matter of epistemology,” nevertheless what is coherently conceivable will have to go beyond this, as the concept of possibility surely requires. In gearing possibility to conceivability I am reluctant to abandon the normative move from putative and de facto to coherent and ideal—and thereby adopt normatively laden—conceivability. For I do not think we can do justice to the complex conceptual phenomenology at issue while yet remaining wholly outside the normative/ideal domain here. And the primary/secondary distinction to which various theories resort seems to me to be an acknowledgement rather than avoidance of this fundamental point. And so in acknowledging the gulf between the real and actual on the one hand and the putative and apparent on the other, the approach I favor has little difficulty in accommodating that Fermat’s Last Theorem example of Crispin Right’s. For with specific judgments of possibility, a conceivability geared to a mistaken view of things is not only possible but likely.

527

RESCHER STUDIES

In all, then, I am Grateful for Marconi’s insightful diagnosis of the way in which my idealistic position of the matter of possibility can be fortified in some of its details. RESPONSE TO ROBERT MEYER I much enjoyed my former student Bob Meyer’s sprightly discussion of Weird Worlds, which seeks to give the nonstandard possible world theory of Rescher-Brandom 1980 a run for its money. In reaching to the interesting ideas canvassed by Meyer I would like to make three points: I. WEIRD WORLDS A good deal of contemporary possible world talk seems to me to be predicated on a confusion. For those who engage in it fail to distinguish between the epistemologically geared idea of • a possible description of a world and the ontologically geared idea of • a description of a possible world After all, not everything that answers to the former succeeds in answering to the latter, no more than is the case with the thesis: A verbose (or confusing or rambling) description of a world is a description of a verbose (or confusing or rambling) description of a world.

The features of a description, in sum, need not carry over to the objects described. And what is true if a world need not be true in it. (It may be true of a world that it is contemplated by many but this need not be true in it; and it may be true of a world that it is discussed by me but that may not be true in it.) On this basis it would appear that a weird world-description need not necessarily be a weird world—nor conversely. And even a theorist who wants to deny that there are any such things as weird (or indeed even

528

REPLIES

merely possible) worlds at all can be prepared to come to terms with weird (and also even merely possible) world descriptions. Given these considerations, there emerges, so seems to me, the significant point that one can afford to indulge in weird-world talk as a useful thought-device even if one lacks—as I do—the speculative recklessness required for accepting weird worlds as live prospect at the level of serious ontology.3 De dicto weirdness is one thing, and de re weirdness quite another. II. A LAMENTABLE BLOCKAGE? Meyer does not approve of the Rescher-Brandom strategy of blocking the conjunctive inference from p and not-p to p-and-not-p in its unrestricted generality. But even Aristotle was prepared to block it absent the proviso that the p at issue is on both sides construed in just exactly the same respect (in point of time, place, and other modes of categoricity). Yet over and above this issue of respectivity there is also the prospect of the course of contextuality and of modality (fact, supposition, possibility-speculation). In various settings we should indeed want to proscribe the conjunction at issue, and only in certain rather special conditions would (or should) we feel comfortable about admitting it. There is really nothing uncompromisingly lamentable about blocking it as a universally and unqualifiedly appropriate modus operandi. III. ONE TRUE LOGIC I would readily agree with Meyer that there will not be One True Logic. But we are nevertheless not free to make up our logic in any-old-way that we may please. Within a coherent organon of rational inference there are certain things that a human must not do. It may in certain cases (e.g., when p is self-contradictory) be practicable to allow not-p to be inferred from p, but one clearly cannot permit this universally and in general. And again one must in some cases refuse to let p be derived from q. When we travel such a negativia there will be a great deal that a viable logic must exclude. No doubt these exclusions will leave many possibilities open and will seriously underdetermine the detailed lineaments of acceptable logics. But they betoken the fact that the situation cannot be one of “anything goes.” And for sure, several important positive features should be there with anything worthy of designation as a “logic,” namely that an adequate sys-

529

RESCHER STUDIES

tem of logic should be self-enabling in providing at the meta-logical level an instrument adequate to provide for its own formalized development.4 And further, to revert to point II, it must certainly be the case that p (and also q) follows from p-and-q. RESPONSE TO JUERGEN MITTELSTRASS AND PETER SCHROEDER-HEISTER My old friend Juergen Mittelstrass and his collaborator Peter SchroederHeister have produced an informatively scholarly account of ancient syllogistic, a topic to which I too devoted attention many years ago. I much appreciate their careful analysis of some key developmental and theoretical issues involved in the histography of the syllogism. When working on Arabic logic, I entertained the hope that the example of an amateur shedding light into some corners of this much-neglected field might inspire (or shame!) professional Arabists to provide more grist to the mill of their philosophical colleagues. But for the most part, they found other things to do, and even now, some two generations later, there is yet a substantial mass of unpublished manuscript material to be worked on. To be sure, there are honorable exceptions here particularly in regard to modal syllogistics, a good deal of it is at the hands of Tony Street, whose discussion in the present volume should certainly be consulted by interested readers. For me, one of the main lessons has been the extent to which fashion can dominate over inherent significance in shaping the flow of scholarly production. Greek logic has been such a prominent topic among students of classical philosophy that one would expect this to have substantial repercussions among those who study the penetration of Greek thought into the cultural orbit of the Islamic world. But in great measure this yet remains as something for the future, and in the meantime we must be grateful for the work, modest in scope but high in quality, that a small cluster of dedicated and conscientious scholars are putting at our disposal. This said, there is yet another phenomenon that merits comment. Perhaps because the field is so small, some of its practitioners have been inclined to treat it as a private preserve that must be guarded against potentially intrusive outsiders. Instead of encouraging participation in the work, they act as though it has to be protected against the contamination by somehow unworthy influences. In this regard I incline to think that stu-

530

REPLIES

dents of Greco-Arabic philosophy have much to learn from their GrecoLatinist classicalist colleagues. Such a circling of the wagons is especially regrettable in this particular field. For just as the natural scientist must establish an equilibrium between observation and theory, so the historian of thought and ideas must establish an equilibrium between his sources and their interpretation. What is critical here is the basis—the experimental observation on the scientific side and the textual on the side of scholarship. Specifically, with regard to GrecoArabic philosophy there is an urgent need for the publication and analysis of texts. And just here scholarship needs all the help that it can get, a circumstance that renders discouragement to active participation particularly unfortunate. And there is more work to be done. For as I see it, that elaborate model syllogistic of the medieval Arabic logicians is a thing of mystery. Neither do we adequately understand where it comes from, nor do we yet comprehend where people were intending to go with it—for what use it was intended or what theoretical need it was seen as meeting. Like the Antikythera mechanism of Greek antiquity, it looks to be a rather surprising anomaly in the history of thought. RESPONSE TO JESUS MOSTERIN Jesus Mosterin gives a clear and instructive commentary on the motivation and elaboration of the approach to evolutionary epistemology that I have sought to articulate in various publications. Mosterin quite justly seeks for clarity about the prospects of cultural evolution through deliberate variation and rational selection. Clearly this is a process whose range of effective operation is going to be decidedly limited. He points out that it is unlikely to work in the context of linguistic communication in general, and there are surely many other examples— matters of diet and cuisine included. But as the modus operandi of the matter grows more sophisticated and the prospect of rational selection grows more hopeful. And here we have not only such rather crude examples as fighting with spears or hunting with a bow and arrow but such decidedly more sophisticated endeavors as determining the ballast needed for ships or forecasting the weather. Granted, culture evolution driven by applicative efficacy is not an unrestrictedly ubiquitous process, but there are many important areas of cognitive endeavor where it does indeed do useful work.

531

RESCHER STUDIES

However, one topic where Mosterin and I see matters in a rather different light is in regard to the role of reason. Modern science, he tells us, “is based on mistrust of reason and relies on experience,” and as such looks askance at “mere plausibility, clarity, or evidence of ideas.” Now true enough, familiarity, intuitive appeal, and surface plausibility will not inform us reliably as to how things work in the world. But that surely is not what reason is about, and failing to place our cognitive bets on these factors is not to distrust it. Reason is certainly not a matter of heeding our everyday intuitions. On the contrary, it is reason itself as applied to the course of actual experience that instructs us as to when, why, how, and how far reliance can be put on considerations of this sort. As I see it, the phrase “experience teaches” involves an abbreviated elision. What teaches us is reason, in interpreting and utilizing the informative materials provided by experience. And experience does not teach us to mistrust reason but rather to acknowledge its decisive arbitrament. The great lesson of the struggle between the empiricist aficionados of experience and the rationalist aficionados of reason is, in my view, aptly put in Kantian language: in matters of scientific knowledge experience without reason is blind and reason without experience empty. Only through the symbiotic cooperation of these two great potencies can we erect the edifice of reliable knowledge. RESPONSE TO JOSEPH PITT The instructive discussion by my former student Joseph Pitt puts front and center the questions of the relationship between science and common sense. And this question is both interesting and complex. Theorists often see an antagonism here. But actually there is—or should be—symbiosis. It is—or should—nowadays be a matter of common sense that the scientific method affords the right and proper means for addressing the questions belonging to the domain of science. And on the other hand the appropriateness of common sense ideas and procedures should— insofar as they indeed are appropriate—be something that scientific inquiry can bring to light. On the one hand it is only common sense to follow the advice of one’s physician. On the other hand, even one’s physician may advise one to use common sense in matters of diet and regimen. It is this complexity that makes it problematic to look at common sense in the light of Sir Arthur Eddington’s dualism of the physicists’ table and the table of ordinary experience as replicated in the distinction drawn by

532

REPLIES

my late colleague Wilfred Sellars between the manifest image of ordinary life and the scientific image of technical expertise. This sort of distinction—important though it be—is not really the most appropriate basis for invoking the contrast of science with common sense. Something else is at work here, a matter not of different objects but of different issues. That the table will (just) fit into our narrow hallway is a matter of science—of measurement. That it is not a good idea to put it there is, however, not a matter of science but of common sense. That it is a bad idea to poke a stick into a nest buzzing with wasps is not really a matter of science. The point is that these two important cognitive resources are designed to address different topics of concern and answer different questions. And one of the most fundamental laws of epistemology is that the proper answer depends on just exactly what the question is. In the crucial issues of everyday life, the relation of science to common sense is somewhat analogous to the relation of statistics to policy in public affairs. To discover how matters actually stand we must consult the one, but to decide what should be done about it we must consult the other. Both are needed because neither does the job of the other. To map the world’s highways and byways we indispensably need science. But in deciding how to travel them we had best use common sense. RESPONSE TO LORENZ PUNTEL It gives me much pleasure that my old friend Lorenz Puntel, himself a master of systemic thinking, has contributed a discussion of my own effort at systematization. His astute and scholarly examination of what he calls the “theoretical framework” of my System of Pragmatic Idealism raises some important and far-reaching issues. In particular he diagnoses what he sees as some significant shortcomings of the project. For as Puntel sees it, my philosophical system has certain lacunae that undermine its systematic adequacy. To meet the challenge of Puntel’s probing discussion would require not just a brief response, but a book, or indeed several books, some of which, I incline to think, I have already written. Let me pass the principal problems he envisions in brief review. [1] The situation of philosophy in relation to the sciences of man and of nature. As I see it, philosophy’s definitive task is to answer “the big questions” about the situation of us humans and our place in the

533

RESCHER STUDIES

world’s scheme of things—and to do this on the basis of the best available information that can provide evidence in this regard. Unquestionably, this information can and does come in substantial measure from the sciences. Granted, this is a vast topic. But I do have discussions out there that make a stab at it.5 [2] Language and especially the semantics of truth and inference. In taking what Puntel justly calls “the cognitive turn” I have endeared to integrate truth conditions (semantics) with use conditions (pragmatics).6 So I don’t think that I have grossly neglected semantics, but rather have sought to deal with its issues in my own nonstandard (pragmatic-epistemic) way. [3] Ontology. Perhaps Puntel is right in complaining that my treatment of ontology is underdeveloped. But it is certainly not omitted in my work, and ontological topics figure prominently in various contexts.7 Nevertheless, what philosophical readers may possibly not welcome in my work is the extent to which I am prepared to leave substantive issues of ontology to the practitioners of science. [4] Philosophical Holism. What Puntel here deplores is not so much omission as insufficiency; for him my discussion “does not go far enough.” It is difficult to know how to respond here,8 for corpus has become so extensive and multi-faceted that its systemic unity is too easily lost sight of. Perhaps the problem is not that my holism is inadequate, but rather that it views the whole at issue as so complex that its own treatment thereof becomes difficult to survey (unuebersichtlich). And this brings us to an interesting point raised by Puntel, viz., that my philosophy looks constantly to us, to the issues “we” face and to the nature of “our” cognitive situation. And one may rightly ask: just who is at issue here, who is this “we”? For sure, it is not the I, the self of the Cartesian ego. Rather, the “we” I have in view is clearly social and communal—the confraternity of intelligent beings who want to think those larger issues through. And this catchment group includes not just our contemporaries and predecessors, but must also be taken to encompass our posterity as well. The community at issue accordingly stretches along the corridor of time, so that we must look with some degree of hopefulness to our collabo-

534

REPLIES

rative successors and proceed not just under the guidance of regulative principles but of regulative aspirations as well. There emerges, once again, an impetus to idealization. RESPONSE TO THOMAS ROCKMORE I much appreciate my old friend Tom Rockmore’s thoughtful analysis of my position on realism/idealism, developed with a view both to its sources of inspiration and its doctrinal objectives. Idealism, of course, acquired a bad name among philosophers in the “age of analysis” and even those who sympathize with idealistic positions formulate their view in other terms. The fact of it is that anyone willing to call himself an idealist carries on his back a great burden of potential misunderstanding—above all the relation to the Berkeleyan contention that minds and their thoughts are all that there is. This strictly nomistic ontological idealism is of course something that few—if any—present-day philosophers would readily espouse. Certainly what I myself have defended is that vastly less problematic idea that the features we ascribe to reality invariably involve an implicit reference to a relationship to mind—that our descriptions of the natural world proceed in terms of reference that envision some sort of concrete relationship to minds and their operations: in sum, that we characterize the real in mindreferential terms. Regrettably, the ill-report of its exaggerated versions have made it difficult to gain an open-minded hearing for such a more modest and plausible version of the doctrine. Critics have tended to trivialize the issue along the lines of Santayanao’s complaint against Schopenhauer that “he proclaimed that the world was his idea, but meant only (what is underavailable) that his idea of the world was his idea.” It is not this triviality that lies at the core of my “conceptual idealism” (or Begriffsidealismus as Germans would call it). Rather, the core of its position is that in our thought about Reality and its nature we invariably impute to it features that it could not possess but for its relationship to mind. The concept of causality, for example, is such that a causal contention not only claims what does actually happen but also incorporates a claim as to what-would-happen-if, a hypothetical situation which reality as such cannot manifest but which rests on the sort of actuality-transcending process in which only minds with their capacity of fact-transcending supposition and assumption can engage. This pivotally central aspect of this position seems so antipathetic to some readers as to be invisible to them and I appreciate Rockmore’s clear and unflinching vision of it.

535

RESCHER STUDIES

Rockmore rightly stresses that one of the central issues of my idealism relates to the question of the status of the laws of nature: are they mental constructs or mind-independent realities? As I see it, what we have here is not an either-or but a range of distinctions. On the one hand, all the relationships that we claim to be natural laws are matters of human conceptual artifices, in that come into being through human cognitive effort on conceptual artifice and are vulnerable to demolition on that same basis. Such laws are nothing apart from their formulation and this is always the fruit of our symbolic inventiveness. On the other hand, we do not frame these laws arbitrarily. Their reason for being is application, and we constantly use them to predict and to intervene in nature. However, reality gives its verdict independently of our wants and wishes—we do not control the outcomes of our applications. The formulation and implementation of those laws is ours, but the resultant eventuations are up to nature. So the unavoidable of idealizing construction must be tempered by a realistic recognition that man proposes but nature/reality disposes. Those laws are human constructs all right, but constraints which being both possibilized and shaped by reality’s modus operandi, could not be what they are but for the fact that mind-independent reality is as it is. How we rub those sticks together is up to us, but it is Nature that decides whether or not a fire starts. Another key issue in the dialectic between reality and our knowledge of it lies in the fact that our knowledge is always and inevitably putative knowledge, subject to clarification and correction. And while we recognize full well that our knowledge of the real truth of things is imperfect, we cannot possibly say how it is so. The injunction “Tell me what is so, distinct and apart from whatever you merely think to be so,” is an in-principle unrealizable request. The gap between “knowledge” and “psychologism” can neither be avoided or crossed. The most fundamental reality with which realism must come to terms is the cognate inaccessibility of the real save via the mediation of idealization. If this be idealism we have little alternative but to make the most of it. And I am grateful to Rockmore for helping to clarify how my writings point the way along this—to my mind inevitable—route.

536

REPLIES

RESPONSE TO TONY STREET I am very grateful for Tony Street’s informative and generous account of my work in Arabic logic and I welcome it as affording me the opportunity to offer some autobiographical clarifications. I came to Arabic scholarship through a family connection, via my grandfather’s cousin Oskar Rescher (Osman Reşer), a German-Turkish orientalist of wide knowledge and energetic productivity who worked for half a century amidst the manuscript collections of Istanbul. When my father died in 1953, I inherited the correspondence with him, and this contact motivated me to spend some time studying Arabic. (I worked largely on my own, but during 1959-60 I also audited some classes of S. D. Goitein at the University of Pennsylvania.) As I entered further into the subject I learned that there was a substantial tradition of logical studies in medieval Islam and, since very little scholarship on the subject existed at the time, I set out to give some account of it. In The Development of Arabic Logic (1964) I mapped out the general outlines of the terrain, and in several subsequent publications I sought to explain what struck me as clearly the most substantial innovation contributed to the subject by the Arabic logicians—the doctrine of temporal modalities. What I found here was a decidedly complex and subtle theory, and I still have difficulty in comprehending how those who developed it were able to do so without any machinery of formalization. In the overall setting of a long academic career of active research and teaching—extending from my doctorate in 1951 to the present day (2008)—my work on Arabic logic was but a brief interlude roughly covering the 1960s. Why did I not persist? Basically for two reasons. Seeing that I was not a professional Arabist, and considering the limitation of my philological training, I felt by the end of this period that I had just about accomplished everything readily accessible to me. Moreover, I had reached a point of division in the road ahead. The question now before me was whether to convert myself into an Arabist or to foster my development as a philosopher? Since the latter was a course on which I had been enthusiastically embarked ever since my late teens, the choice was easy. Then too there was the decidedly unencouraging response to the work that I had been doing. As best I could then see, Arabists had little if any interest in logic and logicians little if any interest in the contributions to their subject by medieval Mediterranean civilization to their subject.

537

RESCHER STUDIES

For virtually a generation after I laid this work aside not a mouse squeaked with regard to Arabic modal syllogistics, and no-one was more impressed than I when several scholars—Tony Street first and foremost among them—picked up the subject where I had left it and proceeded to grow a fruitful orchard where I had planted some seeds. My investigations in Arabic logic thus represent what is by now a longclosed chapter of my work. It is, however, one in whose absence an overall account of my professional development would be seriously incomplete. And I am grateful to Tony Street not only for his benevolent account of my work, but also—far more importantly—for carrying the subject onwards to larger fruition. RESPONSE TO AVRUM STROLL Avrum Stoll’s stimulating critique of my ruminations on the future of philosophy brings to light some instructive discrepancies of opinion. Stroll and I are in complete agreement that prediction is a risky business—and never more so than in relation to matters of speculative inquiry. My own inclination here is to the old familiar model of the pendulum with its back-and-forth swing. The 19th century was the era of great systems and the 20th century the era of small-scale detail-work that shunned big-picture thinking in favor of perceived specialization and division of labor. Against this background it seems only plausible to expect the 21st century to witness a sway of the pendulum back to the systemic sympathies of the 19th. But of course what recurs is never the same, unaffected by intervening developments. And so one could (or might) expect the latter-day systems to embody a mosaic of detail. I certainly have no quarrel with Stroll’s insistence that “conceptual analysis has always been a major feature of philosophy at least from the time of the Greeks.” But of course stone-masonry too has been with us since the time of the ancient Greeks—or for that matter the Egyptians. But this does not mean that these processes, while in many respects the same, were then used to create the sort of grandiose structures—Gothic cathedrals, for example—that later workers erected by use of substantially kindred processes. Analytical methods can serve the systemic aspirations of the schoolmen as well as the detail-work favored by their 20th century successors. So while I entirely agree with Stroll that the methods of analytic philosophizing are here to stay, I would submit that the modest aspiration-

538

REPLIES

targets with respect to which they were deployed by the “analytical” school of 20th century philosophizing are likely to yield to more ambitious, synoptically systemic objectives as we move past the crises of philosophical confidence that intimidated that era’s philosophers in the wake of the scientific revolutions of the early 20th century. I find Stroll’s suggestion intriguing that “it is not knowledge that the sceptic has traditionally challenged, but certainty.” I grant that much 20th century epistemology has been directed against “the quest for Cartesian certainty.” But what has always troubled me throughout these discussions is their very tendency to detach the issue of certainty from that of information and informativeness. For what then gets overlooked is that there is in fact a see-saw tradeoff between certainty (assurance) and informativeness (cognitive detail). You ask me about what it is we saw flying off from yonder tree. I am pretty sure it was a sparrow; I am confident that it was a bird; but I am certain that it was something. Certainty is a matter of informative detail. One can always enhance certainty at the cost of imprecision. Had Descartes been concerned for certainty alone, he would have been better advised to curtail his thesis “I am a thing that thinks” after those first two words. For hard though it is to deny his contention, that abbreviated version, being vaguer, is more secure yet. There is one important issue on which Stroll and I are in wholehearted agreement, namely that the progress of neuropsychology will not supplant the course of conceptual analysis. Even supposing a rigid, lock-step coordination between brain activity and thought, there yet remains the question of who is in charge? Which is the dependent and which the independent variable? Who commands and who follows? And the tighter the coordination the more pressing this questions becomes. This critical point is almost universally overlooked. For coordination is a two-way street where things can go either way in the psychological state carry changes in cerebral physiology in their wake: when the mind frets, the brain buzzes. And conversely, changes in brain states carry changes in mind-states in their wake. One can think of mental activity as a matter of the mind’s awareness of what the brain is doing. Sometimes the balance runs one way and sometimes the other. And so regardless of how tight the correlation of mind and matter may be, there is no ground for construing this circumstance as precluding the efficacy of conceptualizing thought in effecting change, and no reason to refrain from maintaining that it is sometimes mind rather than matter that affords the independent variable that takes intuition in the inauguration of

539

RESCHER STUDIES

change. The tighter the interrelatedness of brain and matter, the ampler the prospects of transactions where mind has the initiative. It is not functional coordination as such that is the pivotal consideration but the difference in the direction of the dependency at issue. If mind were “nothing but” the machinations of matter, if brain psychology were all there is to it, then mind would never be able to do this characteristic work of providing a bridge for the domain of physical processes to the domain of ideas. We would never get from here (processual physicality) to there (conceptual thought): all possibility of achieving meaning, significance, information, would be lost. The contrast between the causality of physical process and the psychic causality of thought processes of intelligent beings is crucial here. The issue is once again that of where the initiative lies, whether with mind or matter, processes of brain or of thought. And nothing about mind-matter correlation—no matter how tightly-woven—precludes the prospect of a two-way street here. RESPONSE TO BAS VAN FRAASSEN I am grateful to Bas van Fraassen for his interesting and insightful contribution to this volume, and am proud as a peacock at being able to number him among my students. Van Fraassen’s account of the explanation-controversies of the 1950s and 60s in their interrelationship with matters of prediction is an instructive supplement to the fine book by my late colleague Wesley Salmon on Four Decades of Scientific Explanation.9 Van Fraassen’s instructive take on the role of these issues in his own philosophical development has led me to contemplate their bearing upon mine. As a student of Hempel’s, a collaborator of Oppenheim’s, a colleague of Helmer’s, and an occasional visitor with Carnap—all during the 1946-56 decade—my interests were much drawn to issues of explanation, prediction, and confirmation.10 But in the course of time I came to take a more synoptic view on the cluster of issues to which these thinkers had devoted so much fruitful specialized attention. It seemed to me that the shortcomings of their efforts as noted by various friendly critics—myself included— were due to one fundamental common cause, viz. that the phenomena of the relevant range were so diverse, many-sided and complex that important aspects of them always eluded the comparatively straightforward theoretical models that were being propounded. Now while most investigators in the field reacted by abandoning theoretical generalities in favor of empiri-

540

REPLIES

cally grounded case-studies, I myself took the route of seeking for a more flexible and versatile basis for theorizing that not only permitted but invited contextualization. As my thinking about these issues evolved, I came to believe that two seemingly discordant remedial steps were needed—one by way of an enlargement of the theory perspective and the other by way of its contraction. Along the former lines it seemed to me fruitful to combine and conjoin that entire spectrum of clusters of relevant epistemic issues— explanation, prediction, confirmation, probability, erotetics, etc.—all needed to be coordinated within the synoptic perspective of one synoptically unified systemic approach. This line of thought led me to envision the integration of all of these diverse issues under the aegis of a coherence theory of truth-estimation. This concern for systemic integration and coherentist coordination developed under the influence of my philosophical hero, Leibniz, with his vision of a systemic integration of science under the motto sympnoia panta (“Everything conspires together”). Thus where those Germanic 20th century masters preached the unity of science, I envisioned unification at the level of meta-scientific epistemology. The second—and to some extent countervailing—perspective developed under the pragmatist inspiration of C. S. Peirce. For it seemed to me only plausible that the methods of cognitive procedure must also be attuned to the particular problems and questions at issue, and that the paramount test of adequacy must thereby pivot on the efficacy and adequacy of those proceedings in meeting the objectives for whose sake they are being put into operation. Pragmatic efficacy with respect to the particular applications at issue should accordingly be seen as the appropriate standard of quality control for respect to our cognitive methods and procedures. And so for the ethereally abstractive theoretical stance of the founding fathers of analytical philosophy of science in Berlin and Vienna, I proposed to substantiate the systemic approach of a Leibnizian even if not Hegelian perspective with a Peirce-inspired pragmatism of characteristically American provenience. This revision of perspective impelled me into the coherentism already envisioned through a glass, darkly, by Otto Neurath, but for me inspired by the intellectual globalism of Leibniz—combined with a considerable admixture of Peirce. Insofar as one’s own case is not lost in an historical blind-spot this is pretty much how I see my own development in the postpositivist.

541

RESCHER STUDIES

RESPONSE TO WAGNER-DÖBLER AND LEIBER Theodore Leiber and Roland Wagner-Döbler present an interesting analysis of recent thought about scientific progress, and they deserve my thanks for their efforts to situate my own work within the larger setting of 20th century thought in this important problem area. And their clear and cogent discussion facilitates access to my ideas in various constructive ways. As I see it, my position regarding scientific progress has two characteristic aspects. Most theorists look to cognitive progress with a view to its informative superiority—its being more truth-like (verisimilar), more detailed, more testable, or some such. My view, by contrast, is to look at the matter not from the angle of information but rather from that of evidentiation. That later information is simply better evidentiated and for this reason more deserving of rational credence. This shift of perspective has the important consequence of putting the main burden for progressiveness not on the side of theory but on the side of praxis. Superior claims to acceptance will then rest not on any textual merit but on the support they get from being able to function effectively in the technologically more sophisticated context in which our later (and putatively superior) thesis must function. And so, the progress of science will, as I see it, ride piggy-back on a correlative technology whose operative efficacy matters of observation and experimentation those theories must both enable and accommodate. (Compare my preceding commentary on Almeder.) I think (or at any rate hope) that this position is something that is to be found on—and not just between—the lines of my work. And for this reason I would feel ambivalent about the criticism that “Rescher gives no cognitive criteria for the definition of first-rate findings”. For as I see it, there are exactly those that open up new identical pathways to application in the development and utilization of innovative technologies in theoretical and practical contexts alike. As a pragmatist I am inclined to judge the merits of theoria in terms of praxis—none resorting to the prospect of assessing theoretical devices in terms of their capacity to facilitate further theoretical work—with theorizing itself thus oriented as a mode of praxis. And for this reason too I am prepared to assess the quality of scientific work on the basis of citation statistics—perhaps sophisticating the process by looking not just to brute citations as such, but rather to citations in most-cited work. Here too, a pragmatism that sees utility as the arbiter of theoretical merit comes into play.

542

REPLIES

RESPONSE TO WALTON AND GODDEN Douglas Walton’s and David Godden’s instructive discussion offers some illuminating insight in combining issue of rational dialectics with those regarding metaphilosophy. There are two basic alternatives for a model of dialectic. One is the bilateral dialogue between two more or less equivalently situated discussants. The other is the trilateral inter-action of three functionally tasked parties proceeding in the manner of a medieval disputation with a proponent, an opponent, and an adjudicator. As Walton and Godden rightly stress, it has been this second model upon which my own work on dialectics has been based. The reasons for this lay in the sorts of applications that were envisioned. In particular, the analogy of such a triadic situation in regard to metaphilosophy would be that of a heretofore non-committed inquirer (as deciding arbiter) studying the pro’s and con’s—with a view to making up his own mind regarding a philosophical issue as articulated by its proponents and opponents. Very little by way of imaginative projection is now required for a particular philosophical writer to situate himself in such a context. Thus a medieval theorist such as the Thomas Aquinas of the Summa Theologica took exactly this route, canvassing the pro’s and cons of a controverted question with a view to seeing where the appropriate answer lies. And even apart from this, a philosophical writer will generally assume a place in such a schema by taking on the role either of advocate of a position or of critic of it—setting out the substantiating considerations and arguments one way and rebutting the opposition’s responses in order to defuse and refute them before the reader and envisioned adjudicator. A great deal of philosophical exposition can thus accommodate itself easily and naturally to this dialectical model, even with those who do not join with Plato in viewing dialectic as the very essence of philosophical procedure. In thought, the shadow of metaphilosophy certainly looms large in the dialectical landscape that I envision. For philosophical doctrines are inherently controversial. With any one of them there is much to be said for and against. And each participant in the discussion strives for the approbation of the as-yet uncommitted, realizing that in the nature of the situation the aspiration of convincing one’s opponent in the context of a head-to-head, two-sided, winner-take-all confrontation is pretty much of a forlorn hope. However, as Walton and Godden duly emphasize, the problem with a triadic model of dialectic lies with that third adjudicating arbiter party. If it

543

RESCHER STUDIES

is a totally impersonal agency—“the established rules and conventions of controversy”—then the contenders are likely to spin-control matters to their own advantage. (And just here we enter into those difficulties of the rule-bound systems in law and elsewhere.) But when an actual person is at issue, there is always the prospect that this will prove to be not an impartial arbiter and neutral bystander, but rather with yet another committed party with an agenda of its own. The best hope here lies with the prospect of a sincere and open-minded inquirer, someone who as yet has no committed position of his own, but is genuinely concerned for the rational settlement of an issue, letting the chips fall where they may. Then too, when we envision that arbiter as an amorphous multitude of generally sensible but as yet committed readers, the situation offers a more hopeful prospect—one that also fits well to the triadic model of dialectic to which Walton, Godden and I share sympathetic favor. RESPONSE TO MICHEL WEBER Michael Weber’s informative examination of process philosophical issues raises two problems that seem to me to be of particular interest: 1. Why think of nature as being processual all the way through, rather than consisting of irreducibly atomic processual units or globules? 2. Why have an idealistic theory of process that brings thought into it, rather than an altogether mind-independent theory of process? Though it may seem strange on first view, my own position sees these issues as closely interrelated—as parts of one single holistic picture. As I see it, nature is a seething cauldron of process with respect to which identification and individuation—and thereby separation, distinctness, and the like—are always matters of mind-imposed artifice. Separative individuation with respect to any thing-kind whatsoever—individual units of processes included—is the work of mind. A Leibnizian Principle of Continuity obtains: nature conjoins and compounds—only mind separates. The ancient Greeks distinguished between those realities made by nature (phusis) and those constructed by human artifice and convention (nomos). There is, however, another, different and (as I see it) more fitting perspective, one which sees everything mundane as natural, the mind and

544

REPLIES

its works included, and only then distinguishes between those things that are natural directly and immediately—humans and their minds included— and those which are so indirectly and obliquely through the intervention of the mind and its operations. The at-large processuality of nature is directly natural, but its disaggregation into individual process is only indirectly so in that it awaits the emergence of minds and the intelligent beings that are their owner-operators. Identity and individuation are to be seen as matter of interaction. After all, where do such items begin and end? The spatial and temporal boundaries at issue are as little fixed independently of mindimposed determinations as are the boundaries between infancy and childhood. And with this perspective in view, one can then espouse an ontology which on the one hand is naturalistically processual through and through but still makes room for discreteness and unicity at the level of conceptualizing artifice, analogizing the nature’s churning processual sea with its multiple waves and wavelets to nature but as product of mind. But why should one take this individuation as mind-dependent line in relation to these fundamental issues of process thought posed in Michel Weber’s deliberations. I believe basically for two reasons. (1) That it provides for a more thorough-going and complete natural philosophy of process, and (2) because it looks to be a more realistic and faithful account of the way things actually are, because differentiation roots in distinction. Granted, mental process roots causally in physical process even as the idea of physical process roots conceptually in the operations of mind. But something new, distinctive, and different arises on nature’s stage in the wake of the developmental emergence of mind. Seeing mind as a product of nature allows nature to enjoy priority in the causal/productive order. But any characterization of nature—any sort of model or account of it—must be the work of mind, so that mind enjoys priority on the conceptual order. Nature’s processuality is the productive source of mind but the mind’s processuality is the productive source of whatever characterization or account of nature there can be. (My comments on James Felt’s discussion should help to clarify how this view of things negotiates the difficult connection between realism and idealism.) Consider an analogy: moving from point A to point B along a zigzag path as follows:

A

B

545

RESCHER STUDIES

But now suppose that each step along the path is itself a similar zigzag. And then each step in this zigzag is similarly zigzagged, and that this process goes on and on. Then as we proceed to move from lesser to greater detail the path elongates at every step and there just is no definite, well defined distance unless and until we select a particular level of scale for our analysis. With this done, however, the realities of the situation (i.e., nature) will determine a definite outcome, but without that essential (“mind supplied’) level-of-sorts deliberation there is no definite answer. That “distance” is a function of a mind-supplied perspective of consideration. It is, of course, nature that determines the answer, but only on the basis of a mind-projected perspective of consideration. It is this sort of perspective that underlines the “idealistic” view of processuality that I envision. RESPONSE TO JAMES WIBLE I much appreciate the careful and conscientious labor that James R. Wible has devoted to my work on the economics of cognition and science. His exposition of the ideas is elaborate and instructive and will certainly help interested readers in orienting themselves amongst the issues at stake. I would like to take this opportunity to elucidate the philosophical background somewhat further. Throughout modern times philosophers have divided their field into two main domains, the theoretical and the practical, the one oriented to knowledge and its acquisition, the other to decision and action.11 The school of thought that most decidedly rejected this dualistic approach is that of Pragmatism. From Peirce to Dewey and beyond, pragmatists have argued that no Chinese Wall can be erected between our theoretical and practical concerns. They insist that theorizing and the development of knowledge itself should be seen as a praxis, at the same time insisting that we must cultivate this praxis under the guidance of theory. Put in Kantian terms, their position is that praxis is blind without theory and theory empty without praxis. Just this symbiotic view of the matter mandates an economic perspective. Any human activity—rational inquiry and theorizing included— demands the deployment of resources (time, energy, effort, ingenuity). And no view of reality is available to us—no model of the world’s operation accessible—without the expenditure of effort and the risk of error. Any adequately grounded theory regarding the nature and ramification of knowledge cannot but see information as a product that must be produced,

546

REPLIES

systematized, disseminated and utilized in ways that are inherent to economic processes in general, with costs and benefits prominently in play. What specifically interests philosophers and economists in their shared concern for the field at large will doubtless differ. But one fundamental fact links their deliberations in systemic unity: any theory of knowledge that does not acknowledge and exploit the subject’s inherent economic realities is in significant measure destined to be an exercise in futility. For once one looks to the evolution of research technology in natural science (as described in my response to Robert Almeder), some of the salient aspects of the economics of natural science leap to the fore. For here we come up against the resource-cost escalation inherent in the phenomenon of resistance barriers (as described there). Every added step forward along the route of progress leaves more difficult—and more expensive in its resource-expenditure demands. The theoretical prospect of unending cognitive progress comes up against the practical realities of economic limitation. What we are led to is certainly not scepticism but rather a decidedly chastening economic realism. RESPONSE TO CATHERINE WILSON I am very grateful to Catherine Wilson for her knowledgeable but generous appreciation of my work on Leibniz. The difficulty with Leibniz has always related to that aspect of his philosophy which I have found particularly fascinating—his metaphysics. Just here, Leibniz aficionados have always found the greatest difficulty. Leibniz the mathematician, Leibniz the logician, Leibniz the physicist—the geologist, the computational pioneer, the historian, the politician, the panEuropean—all the multitudes that are Leibniz have found a respectful and appreciative reception. But Leibniz the monadologist, the optimalist, the theodicist has generally been met with derision at worst and benign neglect at best. Yet I myself have always felt that there is much to be said on his behalf and I have gone ahead to say a good deal of it. What I particularly admire about Leibniz—his sheer brain-power apart—is the range of his interests, the breadth of his sympathies, and the constructiveness of his aspirations. When we encounter unfamiliar ideas many of us—perhaps most—of us look to what can be said against them. But not Leibniz. Ever on the look-out for the useful and beneficial, his open-minded and receptive spirit is a lesson that can and should draw benefit.

547

RESCHER STUDIES

However, the gulf of modernity separates our own era from that of Leibniz. No 17th century philosopher—not Descartes, not Malebrache or Spinoza, not Locke or Berkeley, and certainly not Leibniz—could get by without calling upon God to accomplish needed philosophical work. But any such approach goes decidedly against the grain of the philosophical sensibilities of the present era. Then too, there is the issue of sufficient reason. Leibniz himself seems to reason as follows: (1) The real is rational: every fact has its explanation (sufficient reason) for being as is. [This is the PSR.] (2) The explanation (sufficient reason) for the existence of something can either lie within the nature and conceptualization of that item itself, or lie in something different from and (conceptually) external to it. (3) If its raison d’etre lies within the thing itself, then this thing is a necessary existent. But only God can possibly qualify here. (4) Thus if there is to be any explanation (sufficient reason) for natural reality (the world, the universe), it must be external to this realm itself and thereby be extra- or supra-natural. THEREFORE: The existence of the universe admits of and indeed requires an extra- or supra-natural explanation. If modernity wishes to avoid this conclusion, it must reject one of its premisses. Since Premiss 2 is more or less a matter of logic that is not eligible for rejection it must give up on: Premiss (1) The pan-explicability maintained by a principle of sufficient reason. That is, one must accept inexplicable mysteries. Premiss (3) The contingency of the world. That is, one must adopt the Spinozism of a (self-)necessitated universe. Premiss (4) The impracticability of circular, self-certifying explanation. That is, one must adopt the Spinozism of a causa sui universe.

548

REPLIES

One or another of these aporetic options is now unavoidable, and none of them look particularly attractive—even from the perspective of a theologyaversive modernity, seeing that all invoke a fairly complex metaphysics. Kant sought to de-theologize the doctrines of Leibniz by relocating their theistically grounded necessities in the modus operandi of the human mind. But in the wake of the Darwinian revolution this business emerged as the contingent product of evolutionary processes and thereby became unable to realize its hoped-for explanatory ends. For a time it looked as though philosophers would have to make do with contingency all across the board and all the way down. Only as the conceptual turn succeeded the linguistic turn late in the 20th century did philosophers once more become comfortable with a necessity of the sort mooted by those whom Kant denigrated as “dogmatic” philosophers, thereby enmeshing them in those uncomfortable a posteriori necessities discussed in my reply to Marconi. There is, I think, much in the metaphysics of Leibniz that can be of good use even in present-day philosophy. RESPONSE TO JOHN WOODS John Woods’ very instructive examination of those of my views on logic centering around inconsistency and aporetics rightly stresses its perspectival shift from final semantics to pragmatics. Its approach could be described in caricature as based on the principle that “Form follows function.” For just here Rudolf Carnap came (rather belatedly) to the correct insight. In the development of logic—as in so much else—we face the prospect of options and alternatives. And the decision of which way to go and which fork in the road to follow is one that need not (and indeed should not) be made once and for all, but rather admits of contextual variation in the light of the variable aims and purposes of different areas of inquiry and deliberation. In sum, the pragmatic issue of aims and purposes must come to the fore here. In particular, this perspective allows epistemological considerations to invade our deliberations about matters of logic. As the 19th century was giving way to the 20th the theorists who developed modern symbolic logic, proceeded largely in the interest of mathematics and were accordingly bent on de-psychologizing logic. But in the course of pursuing this (substantially right-minded) venture, they also left matters of functional and purpose context by the wayside. The recovery of the functions at issue here

549

RESCHER STUDIES

became the task of a later generation and—as Woods so clearly observed—I became an enthusiastic recruit to its ranks. As I see it, there is a natural progression in philosophy from eroletics to aporetics to dialectics to systematics. We begin with questions, and in particular the big questions regarding reality and man’s place within it. (Hence eroletics.) And so we proceed to balance considerations against each other, and then become enmeshed in auto-consideration that need to be resolved in time. (Hence dialectics.) And now we come to what Woods astutely labels as “Philosophy’s Most Difficult Problem”—the fact that one man’s modus ponens can be another’s modus tollens. For to handle this we must now proceed systematically and look not just to individual arguments but to entire clusters of them. This superior (for the sake of oversimplification and schematic example). Consider a schematic example. You are deliberating about from theses P, Q, R, S and your initial approach of their possibility is as follows: High: P-or-Q, R Medium: R Low: P, Q, S But suppose further that an analysis of the argumentation reveals that: (1) P & R → S (2) Q & R → S Looking at (1) alone you might reason: “I want to reject the (very) implausible S, so I’ll jettison R.” And looking at (2) alike you will say “I want to reject the implausible S, so I’ll jettison the implausible Q.” And you include: “By adding the symbol P not the implausible Q I manage both to get rid of S and retain (if plausible) the plausibilistically premiss R, so I must jettison P and Q.” But in the circumstances this is not a good option. So systematic consideration of plausibility speak in R’s favor after all.

550

REPLIES

NOTES 1

I have always thought a homely fishing analogy of Sir Arthur Eddington’s to be instructive here. He saw the experimentalists as akin to a fisherman who trawls nature with the net of his equipment for detection and observation. Now suppose (says Eddington) that a fisherman trawls the seas using a fishnet of two-inch mesh. Then fish of a smaller size will simply go uncaught, and those who analyze the catch will have an incomplete and distorted view of aquatic life. The situation in science is the same. Only by improving our observational means of trawling nature can such imperfections be mitigated. (See A. S. Eddington, The Nature of the Physical World [New York: The Macmillan Company, 1928]).

2

Nicholas Rescher, Epistemic Logic (Pittsburgh: University of Pittsburgh Press, 2005), pp. 123.

3

See, for example, N. Rescher Imagining Irreality (Chicago and La Salle: Open Court, 2003).

4

On this issue see the discussion of autodesciptivity in N. Rescher, Many-Valued Logic (New York: Macmillan, 1969), pp. 83-90 and 141-142.

5

See, for example, Philosophical Reasoning (Oxford: Blackwell, 2001).

6

See, for example, Communicative pragmatism (Lanham: Rowman & Littlefield, 1998). Sand also Conditionals (Cambridge, Mass: MIT Press, 2007).

7

See especially Metaphysics (Amherst, NY: Prometheus, 2006).

8

I would argue that books like Cognitive Systematization (Oxford: Blackwells, 1979) and Methodological Pragmatism (Oxford: Blackwells, 1977), do go a fair bit of the way here.

9

Minneapolis: University of Minnesota Press, 1989.

10

On the historical background here see my papers on . . .

11

The ancients favored a triparitative division discourse: Logic and Language, Theoretical Philosophy, and Natural Philosophy. With the last setting up shop in its own as Natural Science, the former two morphed into the dichotomy of the theoretical and the practical, a duality in the attachment of university chairs that so continues in Scandinavia to the present day.

551

Work by and about Nicholas Rescher ____________________________________________________________ CONTENTS I.

Chronological List of Books by Nicholas Rescher

553

II. Translations of Nicholas Rescher’s Books

566

III. Books and Publications about Nicholas Rescher’s Philosophy

569

IV. Ideas and Innovations Associated Eponymously with Nicholas Rescher’s Name 571 ____________________________________________________________ I. CHRONOLOGICAL LIST OF BOOKS BY NICHOLAS RESCHER 1960 On the Epistemology of the Inexact Sciences. Santa Monica, CA (RAND Corporation) 1960; RAND Research Monograph R353. Coauthored with Olaf Helmer. 1962 Al-Farabi: An Annotated Bibliography. Pittsburgh Press), 1962.

Pittsburgh (University of

1963 Al-Farabi’s Short Commentary on Aristotle’s “Prior Analytics.” Translated from the Arabic, with Introduction and Notes. Pittsburgh (University of Pittsburgh Press), 1963. Studies in the History of Arabic Logic. Pittsburgh (University of Pittsburgh Press), 1963.

RESCHER STUDIES

1964 Al-Kindi: An Annotated Bibliography. Pittsburgh (University of Pittsburgh Press), 1964. The Development of Arabic Logic. Pittsburgh (University of Pittsburgh Press), 1964. Hypothetical Reasoning. Amsterdam (North-Holland Publishing Co.; “Studies in Logic” series edited by L.E.J. Brouwer, E.W. Beth and A. Heyting.), 1964. An Introduction to Logic. New York (St. Martin’s Press), 1964. 1965 The Refutation by Alexander of Aphrodisias of Galen’s Treatise on The First Mover. Karachi (Publications of the Central Institute of Islamic Research), 1965. Co-authored with Michael E. Marmura. 1966 Distributive Justice. New York (Bobbs Merrill Company), 1966. Reissued in 1982 by the University Press of America (Washington, D.C.). Galen and the Syllogism: An Examination of the Claim that Galen Originated the Fourt Figure of the Syllogism. Pittsburgh University of Pittsburgh Press), 1966. The Logic of Commands. London (Routledge & Kegan Paul), 1966. 1967 Temporal Modalities in Arabic Logic. Dordrecht (Reidel), 1967; Supplementary Series of Foundations of Language. The Philosophy of Leibniz. Englewood Cliffs (Prentice Hall), 1967.

554

WORK BY AND ABOUT NICHOLAS RESCHER

Studies in Arabic Philosophy. Press), 1967.

Pittsburgh (University of Pittsburgh

1968 Topics in Philosophical Logic. Dordrecht (Reidel), 1968; Synthese Library. 1969 Essays in Philosophical Analysis: Historical and Systematic. Pittsburgh (University of Pittsburgh Press), 1969. Reissued in 1982 by the University Press of America (Washington, D.C.) Introduction to Value Theory. Englewood Cliffs (Prentice Hall), 1969. Reissued in 1982 by the University Press of America (Washington, D.C.). Many-Valued Logic. New York (McGraw-Hill), 1969. Reprinted: Aldershot (Gregg Revivals), 1993. 1970 Scientific Explanation. New York (The Free Press), 1970. 1971 Temporal Logic. New York and Vienna (Springer-Verlag), 1971. Coauthored with Alastair Urquhart. 1972 Welfare: The Social Issues in Philosophical Perspective. Pittsburgh (University of Pittsburgh Press), 1972.

555

RESCHER STUDIES

1973 The Coherence Theory of Truth. Oxford (The Clarendon Press/Oxford University Press), 1973. Reissued in 1982 by the University Press of America (Washington, D.C.). Conceptual Idealism. Oxford (Basil Blackwell), 1973. Reissued in 1982 by the University Press of America (Washington, D.C.). The Primacy of Practice. Oxford (Basil Blackwell), 1973. Translated into Spanish as La Primacia de la práctica, Madrid (Editorial Tecnos), 1980. 1974 Studies in Modality. Oxford (Basil Blackwell), 1974. (American Philosophical Quarterly, Monograph Series.) 1975 A Theory of Possibility. Oxford (Basil Blackwell), 1975. Co-published in the USA by the University of Pittsburgh Press. Unselfishness: The Role of the Vicarious Affects in Moral Philosophy and Social Theory. Pittsburgh (University of Pittsburgh Press), 1975. 1976 Plausible Reasoning. Amsterdam (Van Gorcum), 1976. 1977 Dialectics: A Controversy-Oriented Approach to the Theory of Knowledge. Albany (State University of New York Press), 1977. Translated into Japanese as Taiwa No Roni Tokyo: Kinokuniya Press, 1981.

556

WORK BY AND ABOUT NICHOLAS RESCHER

Methodological Pragmatism. Oxford (Basil Blackwell), 1977. Copublished in the USA by the New York University Press. 1978 Peirce’s Philosophy of Science. Dame Press), 1978.

Notre Dame (University of Notre

Scientific Progress: A Philosophical Essay on the Economics of Research in Natural Science. Oxford (Basil Blackwell), 1978. Copublished in the USA by the University of Pittsburgh Press. Translated into German as Wissenschaftlicher Fortschritt (Berlin: De Gruyter Verlag, 1982). Translated into French as Le Progrès Scientifique (Paris: Presses Universitaries de France, 1993). 1979 Cognitive Systematization. Oxford (Basil Blackwell), 1979. Copublished in the USA by Rowman & Littlefield. Translated into Spanish as Sistematización cognoscitiva (Mexico City: Siglo Veintiuno Editores, 1981). Leibniz: An Introduction to His Philosophy. Oxford (Basil Blackwell), 1979; APQ Library of Philosophy. Co-published in the USA by Rowman & Littlefield. Reprinted in 1986 by the University Press of America (Lanham, MD). Reprinted in 1993 by Gregg Revivals (Aldershot, UK). The Logic of Inconsistency: A Study in Nonstandard Possible-World Semantics and Ontology. Oxford (Basil Blackwell), 1979; APQ Library of Philosophy. Co-authored with Robert Brandom. Published in the USA by Rowman & Littlefield (Totowa, NJ; 1979). 1980 Induction. Oxford (Basil Blackwell), 1980. Co-published in the USA by the University of Pittsburgh Press. Translated into German as Induktion (Munich: Philosophia Verlag, 1986).

557

RESCHER STUDIES

Scepticism. Oxford (Basil Blackwell), 1980. Co-published in the USA by Rowman & Littlefield. Unpopular Essays on Technological Progress. Pittsburgh (University of Pittsburgh Press), 1980. 1981 Leibniz’s Metaphysics of Nature: A Group of Essays. Dordrecht and Boston (Reidel), 1981. 1982 Empirical Inquiry. Totowa, N.J. (Rowan & Littlefield), 1982. Copublished in Great Britain by Athlone Press (London, 1982). 1983 Kant’s Theory of Knowledge and Reality: A Group of Essays. Washington, D.C. (University Press of America), 1983. Mid-Journey: An Unfinished Autobiography. Lanham MD (University Press of America) 1983. Risk: A Philosophical Introduction to the Theory of Risk Evaluation and Management. Washington, D.C. (University Press of America), 1983. 1984 The Limits of Science. Berkeley and Los Angeles (University of California Press), 1984. Translated into German as Grenzen der Wissenschaft. Dietzingen: Reclam Verlag, 1985. Translated into Spanish as Las Limites de la Sciencia (Madrid: Editorial Tecnos, 1994). Translated into Italian as I Limita della Sciencia (Rome: Armando Editore, 1990). Second (revised and enlarged) edition (Pittsburgh: Unviersity of Pittsburgh Press, 1999).

558

WORK BY AND ABOUT NICHOLAS RESCHER

The Riddle of Existence: An Essay in Idealistic Metaphysics. Washington, D.C. (University Press of America), 1984. 1985 Pascal’s Wager: An Essay on Practical Reasoning in Philosophical Theology. Notre Dame (University of Notre Dame Press), 1985. The Strife of Systems: An Essay on the Grounds and Implications of Philosophical Diversity. Pittsburgh (University of Pittsburgh Press), 1985. Translated into Italian as La Lotti dei Sistemi (Genoa: Marietti, 1993); into German as Der Streit der Systeme (Würzburg: Königshausen und Neumann, 1997); into Spanish (in progress). 1986 Ongoing Journey: An Autobiographical Essay. Lanham MD (University Press of America) 1986. 1987 Ethical Idealism: A Study of the Import of Ideals. Berkeley, Los Angeles and London (University of California Press), 1987. Forbidden Knowledge and Other Essays on the Philosophy of Cognition. Dordrecht (Reidel Publishing Co.), 1987. (Episteme Series, No.13). Scientific Realism: A Critical Reappraisal. Dordrecht (Reidel Publishing Co.), 1987. 1988 Rationality. Oxford (Clarendon Press), 1988. Translated into German as Rationalität (Wuerzburg: Koenigshausen & Neumann, 1993); into Spanish as La Racionalidad (Madrid: Editorial Tecnos, 1993); into Italian as Razionalità (Rome: Armando Editore, 1999).

559

RESCHER STUDIES

1989 Cognitive Economy: Economic Perspectives in the Theory of Knowledge. Pittsburgh (University of Pittsburgh Press), 1989. Moral Absolutes: An Essay on the Nature and the Rationale of Morality. New York (Peter Lang Publishing Co.), l989. A Useful Inheritance: Evolutionary Epistemology in Philosophical Perspective. Savage, MD (Rowman and Littlefield), 1989. Translated into German as Warum sind wir nicht klüger (Stuttgart: Hirzel Verlag, 1994). 1990 Human Interests: Reflections on Philosophical Anthropology. Stanford (Stanford University Press), 1990. 1991 Baffling Phenomena and Other Studies in the Philosophy of Knowledge and Valuation. Savage, MD (Rowman and Littlefield), 1991. Frank Plumpton Ramsey: On Truth, ed. by Nicholas Rescher and Ulrich Majer (Dordrecht: Kluwer, 1991). Human Knowledge in Idealistic Perspective. Princeton (Princeton University Press), 1991. Leibniz’s Monadology: An Edition for Students. Pittsburgh (University of Pittsburgh Press), 1991. Co-published in the United Kingdom by Routledge (London). 1992 The Validity of Values: Human Values in Pragmatic Perspective. Princeton (Princeton University Press), 1992. 1993

560

WORK BY AND ABOUT NICHOLAS RESCHER

Pluralism: Against the Demand for Consensus. Oxford (Clarendon Press), 1993. Standardism: An Empirical Approach to Philosophical Methodology. Pittsburgh (University of Pittsburgh Press), 1993. Reissued in paperback, 2000. 1994 American Philosophy Today, and Other Philosophical Studies. Savage, MD (Rowman & Littlefield), 1994. Animal Conversations: A Collection of Fables. Verona PA (NAP Publications), 1994. Metaphilosophical Inquiries. Princeton (Princeton University Press), 1994. 1995 Essays in the History of Philosophy. Aldershot, UK (Avebury), 1995. Luck. New York (Farrar, Straus & Giroux), 1995. Translated into German as Glück (Berlin: Berlin Verlag, 1996); into Spanish by Carlos Gardini as Suerte, azard destino: Aventuras y desaventuras de la vida cotidiana (Santiago de Chile: Editorial Andrés Bello, 1997). Also translated into Japanese and Korean. Process Metaphysics. Albany (State University of New York Press), 1995. Satisfying Reason: Studies in the Theory of Knowledge. Dordrecht (Kluwer), 1995. 1996 Instructive Journey: An Autobiographical Essay. Lanham MD (University Press of America), 1996.

561

RESCHER STUDIES

Priceless Knowledge? An Essay to Economic Limits to Scientific Progress. Savage, MD (Rowman and Littlefield), 1996. Public Concerns: Philosophical Studies of Social Issues. Lanham, MD (Rowman & Littlefield), 1996. Studien zur naturwissenschaftlichen Erkenntnislehre. (Königshausen & Neumann), 1996.

Würzburg

1997 Objectivity: The Obligations of Impersonal Reason. Notre Dame (University of Notre Dame Press), 1997. Predicting the Future. Albany NY (State University of New York Press), 1997. Profitable Speculations: Essays on Current Philosophical Themes. Lanham MD (Rowman & Littlefield), 1997. 1998 Communicative Pragmatism: And Other Philosophical Essays on Language. Lanham MD (Rowman & Littlefield), 1998. Complexity: A Philosophical Overview. New Brunswick NJ. (Transaction Publishers), 1998. 1999 Kant and the Reach of Reason. Cambridge (Cambridge University Press), 1999. Razón y valores en la era cientifico-tecnológica. Barcelona (Editorial Paidos), 1999. Realistic Pragmatism: An Introduction to Pragmatic Philosophy. Albany (State University of New York Press), 1999.

562

WORK BY AND ABOUT NICHOLAS RESCHER

2000 Inquiry Dynamics. New Brunswick, NJ (Transaction), 2000. Nature and Understanding: A Study of the Metaphysics of Science Oxford (Clarendon Press), 2000. 2001 Cognitive Pragmatism. 2001.

Pittsburgh (University of Pittsburgh Press),

Minding Matter and Other Essays in Philosophical Inquiry. Lanham, MD (Rowman & Littlefield), 2001. Paradoxes. Chicago, Ill. (Open Court Publishing Co.), 2001. Philosophical Reasoning. Oxford (Blackwell), 2001. Process Philosophy: A Survey of Basic Issues. Pittsburgh (University of Pittsburgh Press), 2001. 2002 Enlightenting Journey: An Autobiographical Essay. Lanham MD (Lexington Books, 2002). Fairness. New Brunswick, NJ (Transaction Publishers), 2002. Rationalität, Wissenschaft, und Praxis. Würzburg (Konigshausen & Neumann), 2002. 2003 Cognitive Idealization: On the Nature and Utility of Cognitive Ideals. Uxbridge, UK: (Cambridge Scholars Press) 2003.

563

RESCHER STUDIES

Niagara-on-the-Lake as a Confederate Refuge. Niagara-on-the-Lake, Ontario, Canada (Niagara Historical Society Museum), 2003. On Leibniz. Pittsburgh (University of Pittsburgh Press) 2003. Sensible Decisions On the Ways and Means of Rational Decision. Totowa, N.J. (Rowman & Littlefield) 2003. Rationality in Pragmatic Perspective. Lewiston, N.Y. (Mellen Press) 2003. Epistemology: On the Scope and Limits of Knowledge. Albany NY (SUNY Press) 2003. Imagining Irreality: A Study of Unrealized Possibility. Chicago, Ill. (Open Court Publishing Co.) 2003. 2004 Value Matters: Studies in Axiology. Frankfurt (Ontos Verlag), 2004. 2005 Cognitive Harmony. Pittsburgh (University of Pittsburgh Press), 2005. Common Sense. Milwaukee, WI (Marquette University Press) 2005. [Aquinas Lecture] Cosmos and Logos: Studies in Greek Philosophy. Frankfurt (Ontos Verlag), 2005). Epistemic Logic. Pittsburgh (University of Pittsburgh Press) 2005. Realism and Pragmatic Epistemology. Pittsburgh (University of Pittsburgh Press), 2005. Reason and Reality: Realism and Idealism in Pragmatic Perspective. Lanham, MD. (Rowman & Littlefield), 2005.

564

WORK BY AND ABOUT NICHOLAS RESCHER

Scholastic Meditations. Washington, DC (Catholic University of America Press), 2005. What If? Thought Experimentation in Philosophy. New Brunswick, NJ. (Transaction Books), 2005.

2005-2006 Nicholas Rescher: Collected Papers, 10 vol’s; Frankfurt (Ontos Verlag), 2005: Studies in 20th Century Philosophy, Studies in Pragmatism, Studies in Idealism, Studies in Philosophical Inquiry. 2006: Studies in Cognitive Finitude, Studies in Social Philosophy, Studies in Philosophical Anthropology, Studies in Value Theory, Studies in Metaphilosophy, Studies in the History of Logic. 2006 Metaphysics: The Key Issues from a Realistic Perspective. Amherst, N.Y. (Prometheus Books), 2006 Essuis fur les fondements d’ontologie du procès. Edited by Michel Weber. Frankfurt (Ontos Verlag), 2006. Process Philosophical Deliberations. Frankfurt (Ontos Verlag), 2006. Presumption and Tentative Cognition. Cambridge (Cambridge University Press), 2006. Philosophical Dialectic: An Essay in Metaphilosophy. Albany, NY. (SUNY Press), 2006. Epistemetrics. Cambridge (Cambridge University Press), 2006. Error. Pittsburgh (University of Pittsburgh Press), 2006. 2007 Issues in Philosophy of Religion. Frankfurt (Ontos Verlag), 2007.

565

RESCHER STUDIES

Interpreting Philosophy: The Elements of Philosophical Hermeneutics. Frankfurt (Ontos Verlag), 2007. Is Philosophy Dispensable? And Other Philosophical Studies . Frankfurt (Ontos Verlag), 2007. Conditionals. Boston (MIT Press), 2007. Dialectics: A Classical Approach to Inquiry. Frankfurt (Ontos Verlag), 2007. 2008 Aporetics: Rational Deliberation in the Face of Inconsistency. Pittsburgh (University of Pittsburgh Press), 2008. Ignorance: On the Wider Implications of Deficient Knowledge. Pittsburgh (University of Pittsburgh Press), 2008. Free Will: A Philosophical Reappraisal. New Brunswick, NJ (Transaction Books) 2008. Being and Value: And Other Philosophical Essays. Frankfurt (Ontos Verlag), 2008. II. TRANSLATIONS OF NR’S BOOKS STUDIES IN THE HISTORY OF ARABIC LOGIC (Arabic Translation). Cairo (Der al Ma’arif), 1982. THE PRIMACY OF PRACTICE: La primacia de la práctica (Spanish tr. by Ana Sanchez Torres). Madrid (Editorial Tecnos), 1980. DIALECTICS: Taiwa No Roni (Japanese tr. by Taneomi Uchida). Tokyo (Kinokuniya Press), 1981.

566

WORK BY AND ABOUT NICHOLAS RESCHER

SCIENTIFIC PROGRESS: (1)

Wissenschaftlicher Fortschritt (German tr. by Gerhard Schaeffner). Berlin (de Gruyter Verlag), 1982.

(2)

Le progrès scientifique (French tr. by Michel Rosier). Paris (Presses Universitaires de France) 1993.

COGNITIVE SYSTEMATIZATION: Sistematización cognoscitiva (Spanish tr. by Carlos Rafael Luis). Mexico City (Siglo Vientiuno Editores), 1981. INDUCTION: Induktion: Zur Rechtfertigung des Induktiven Schliessens (German tr. by Gerhard Schaeffner). Munich (Philosophia Verlag), 1987; “Introductiones” Series. THE LIMITS OF SCIENCE: (1)

Grenzen der Wissenschaft (German tr. by Kai Puntel). Dietzingen (Reclam Verlag), 1985. [Contains an introduction to NR’s thought, “Einführung in Nicholas Rescher’s pragmatische System philosophie,” pp. 7-47.]

(2)

I limiti della scienza (Italian tr. by Rosalba Camedda). Rome (Armando Editore), 1990.

(3)

Los límites de la Ciencia (Spanish tr. by Leonardo Rodríguez Duplá (Madrid: Editorial Tecnos) 1994.

THE STRIFE OF SYSTEMS: (1)

La lucha de los sistemas (Spanish tr. by Adolfo GarcEia de la Sienre. Mexico City (UNAM = Universidad Nacional Autónoma de México), 1995.

(2) La lotti dei sistemi (Italian tr. arranged by Andrea Bottani.) Genoa (Marietti) 1993.

567

RESCHER STUDIES

(3)

Der Streit der Systeme (German tr. by Birger Brinkmeier). Würzburg (Königshausen & Neumann), 1997.

RATIONALITY: (1)

Rationalität (German tr. by Axel Wüstehube). Königshausen und Neumann, 1993.

Würzburg:

(2)

Razionalità (Italian tr. by Francesca Febbraro). Rome Armando Editorial, 1999).

(3)

La racionalidad. (Spanish tr. by S. Nuccetelli). Madrid. Editorial Tecnos, 1993.

(4)

Chinese translation by Cau Guan Fa in progress.

COGNITIVE ECONOMY. (Chinese translation). Beijing (Jiang Xi Education Publishers), 1999. A USEFUL INHERITANCE. Warum sind Wir nicht klüger? German tr. by Astrid and Helmut Pape. Stuttgart (Hirzel Verlag, Edition Universitas), 1994. PLURALISM: Bulgarian translation by T. Petkov and K. Gebrahova, Sophia (Critique and Humanism Publishing House), 2001. LUCK:

568

(1)

Glück: Die Chancen des Zufalls. German tr. by Jens Hagesledt. Berlin (Berlin Verlag), 1995.

(2)

La suerte: Aventuras y desaventuras de la vida cotidiana. Spanish tr. by Carlos Gardini. Barcelona (Editorial Andrés Bello), 1997.

(3)

Japanese translation Tokyo (PHP Publishing House), 1999.

(4)

Korean translation Seoul (Good Guy Publishing Co.), 2000.

WORK BY AND ABOUT NICHOLAS RESCHER

COMPLEXITY: (Chinese Translation) Shanghai (Scientific & Technological Eduction Publishing House), 2007. PROCESS METAPHYSCIS: Fondements de l’ontologie du procès. French translation by Michel Weber. Frankfurt (Ontos Verlag), 2005. III.

BOOKS AND PUBLICATIONS ABOUT NR’S PHILOSOPHY

1. Ernest Sosa (ed.), The Philosophy of Nicholas Rescher (Dordrecht: D. Reidel, 1979). [A collection of critical essays with brief replies by N.R. [The contributors include: Annette Baier, Stephen Barker, Nuel D. Belnap, Jr., Laurence BonJour, Robert E. Butts, Roderick M. Chisholm, L. Jonathan Cohen, Jude J. Dougherty, Brian Ellis, R.M. Hare, Hide Ishiguro, George Von Wright, and John W. Yolton.] 2. Heinrich Coomann, Die Kohaerenztheorie der Wahrheit: Eine kritische Darstellung der Theorie Reschers von Ihrem historischen Hintergrund (Frankfurt am Main: Verlag Peter Lang, 1983). 3. Robert Almeder (ed.), Praxis and Reason: Studies in the Philosophy of Nicholas Rescher (Washington, D.C.: University Press of America, 1982.) [A collection of critical and expanding essa ys with brief replies by NR. The contributors include: Timo Airaksinen, Robert Almeder, Antonio Cua, John E. Hare, Risto Hilpinen, John Kekes, Gerald J. Massey, Jack W. Meiland, Mark Pastin, Friedrick Rapp, James Sterba, and Dennis Temple.] 4. Lorenz Puntel, Einführung in Nicholas Rescher’s pragmatische Systemphilosophie, in Nicholas Rescher Die Grenzen der Wissenschaft (Ditzingen: Redam, 1985), German translation of The Limits of Science. 5. Andrea Bottani, Verità e Coerenza: Suggio su’ll epistemologia coerentista di Nicholas Rescher, (Milano: Franco Angeli Liberi, 1989). [A systematic study of NR’s coherence theory of truth.] 6. “On Nicholas Rescher’s Work on Argumentation.” A special issue of Informal Logic, vol. 14 (1992) No. 1, pp. 1-58. [Critical discussions of

569

RESCHER STUDIES

NR’s work by Bryson Brown, David Goodman, Harvey Siegel, and Douglas Walton, with responses by NR.] 7. “Book Symposium” on Nicholas Rescher’s trilogy, A System of Pragmatic Idealism. Philosophy and Phenomenological Research, vol. 54 (1994), June issue. [Critical discussions of NR’s “Pragmatic Idealism” triology by Cornelius Delaney, Jack Meiland, Timothy Sprigge, John Kekes, Terrance McConnell, Joseph Margolis, and Johanna Seibt.] 8. “Symposium on Nicholas Rescher’s Pluralism,” Deutsche Zeitschrift für Philosophie, vol. 42 (1994), pp. 291-334. [Critical discussion of NR’s Pluralism by Hans-Peter Krüger, Lothar Schäfer, Logi Gunnasson, and Axel Wüstehube.] 9. Timo Airaksinen, “Methodological Pragmatism: The Pragmatic Epistemology of Nicholas Rescher” in Pragmatik: Handbuch Pragmatischen Denkens ed. by Herbert Stachowiak, Vol. V (Hamburg: Felix Meiner, 1995), pp. 179-95. 10. Michele Marsonet, The Primacy of Practical Reason: An Essay on Nicholas Rescher’s Philosophy (Lanham, MD: University Press of America, 1995). 11. Axel Wüstehube and Michael Quante (eds.), Pragmatic Idealism: Critical Essays on Nicholas Rescher’s System of Pragmatic Idealism (Amsterdam: Rodopi, 1998). [Critical essays on NR’s “Pragmatic Idealism” trilogy by eighteen contemporary philosophers in Europe and the USA.] 12. Martin Carrier et. al (eds.), Science at the Century’s End: Philosophical Questions on the Progress and Limits of Science (Pittsburgh and Konstanz: University of Pittsburgh Press and University of Konstanz Press, 2000). [Pp. 40-134 contains a symposium devoted to NR’s work on the Limits of Science with contributions by Robert Almeder, Laura Ruetsche, Juergen Mittelstrass, and Martin Carrier.] 13. Wenceslao J. González, “Racionalidad científica y actividad humana Ciencia y valores en la Filosofía de N. Rescher,” a long introductory

570

WORK BY AND ABOUT NICHOLAS RESCHER

discussion essay to NR’s Razón y valores en la Era científicotechnológica (Bacelona: Editorial Paidós, 1999), pp. 11-44. 14. Paul Damian Murray, Reason Truth and Theology in Pragmatist Perspective, R. Rorty, N. Rescher, and D. MacKinnin (Leuven: Peeters, 2004). [A theological study dealing substantially with NR’s ideas.] 15. Lotfallah Nabavi, Avicennan Logic Based on Nicholas Rescher’s Point of View (Tehran: Scientific and Cultural Publication Co., 2003). 16. Symposium: “Nicholas Rescher on Pragmatic Philosophy,” Contemporary Pragmatism, December 2005 issue. 17. Michael Weber (ed.), After Whitehead: Rescher and Process Philosophy (Frankfurt: ONTOS Verlag, 2004). 18. Nicholas J. Moutafakis, Rescher on Rationality, Values, and Moral Responsibility (Frankfurt: ONTOS Verlag, 2007). 19. Robert Almeder (ed.), Rescher Studies: Essays on the Philosophy of Nicholas Rescher (Frankfurt: ONTOS Verlag, 2008). IV. IDEAS AND INNOVATIONS ASSOCIATED EPONYMOUSLY WITH NR’S NAME • The Rescher Quantifier in symbolic logic. See David Kaplan, “Rescher’s Plurality Quantification,” The Journal of Symbolic Logic, vol. 31 (1966), pp. 153-55. See also NR’s “Plurality Quantification Revisited,” Philosophical Inquirer, vol. 26 (2004), pp. 1-6. • Rescher's Law of Logarithmic Returns [also Rescher’s Principle of Decreasing Marginal Returns] in scientometrics. See James R. Wible, The Economics of Science (London & New York: Routledge, 1998), p. 151 and the Index). See also Roland Wagner-Döbler, “Rescher’s Principle of Diminishing Marginal Returns in Scientific Research,” Scientometrics, vol. 50 (2001), pp. 419-36. • Rescher’s Effective Average Measure in the theory of distributive justice. Se L. H. Powers, “A More Effective Average,” Philosophi-

571

RESCHER STUDIES

cal Studies, vol. 24 (1970), pp. 74-78); and Geoffrey Ross, “Utilities for distributive Justine,” Theory and Decision, vol. 4 (1974), pp. 239-58 (especially pp. 250-57). • The Dienes-Rescher Inference Engine (also Rescher-Dienes Implication) in nonstandard logic. See Olaf Wolkenhauser, Fuzzy Inference Engines (Manchester GB: University of Manchester in Stockport, 2000), pp. 21-22. • The Rescher-Manor Consequence Relation in nonmonotonic logic. See Diderik Batens, “A Survey of Inconsistency-Adaptive Logics and the Foundtion of Non-Monotonic Logics,” Logique et Analyse, vol. 37 (1994), pp. 57-94; see also J. Merhuis (ed.), Inconsistency in Science (Dordrecht: Klewer, 2002), pp. 144-46. • The Rescher-Brandom Semantics for paraconsistent logic. See Edward V. Zalta, Notre Dame Journal of Formal Logic, vol. 36 (1997), pp. 640-60. • Rescher’s Model [or Theory] of Formal Disputation in dialectics. See Gerhard Brewka, “Dynamic Argument Systems,” Journal of Logic and Computation, vol. 11 (2001), pp. 257-82. • The Rescher Operator in temporal logic. See Lennart Aqvist, “Combinations of Tense and Deontic Modality: On the Rt approach to Temporal Logic with Historical Necessity and Conditional Obligation,” Journal of Applied Logic, vol. 3 (2005), pp. 421-60. ((For detailed information about these items see the references available via Google .))

572

Contributors Robert Almeder Georgia State University Diderik Batens University of Ghent Bryson Brown University of Lethbridge James Felt Santa Clara University Lenn Goodman Vanderbilt University John Haldane University of St. Andrews William Jaworski Fordham University Ulrich Majer University of Göttingen Diego Marconi University of Turin Robert Meyer The Australian National University Jürgen Mittlestrass and Peter Schroeder-Heister University of Konstanz and University of Tübingen Jesús Mosterin University of Barcelona

RESCHER STUDIES

Joseph Pitt Virginia Tech University Lorenz Puntel University of Munich Thomas Rockmore Duquesne University Tony Street Cambridge University Avrum Stroll University California, San Diego Bas van Fraassen Princeton University Roland Wagner-Doebler and Theodor Leiber University of Augsburg Douglas Walton and David Godden University of Winnipeg Michael Weber Center for Philosophical Practice, Brussels James Wible University of New Hampshire Catharine Wilson City University of New York John Woods University of British Columbia

574

Ontos

NicholasRescher

Nicholas Rescher

Collected Paper. 14 Volumes Nicholas Rescher is University Professor of Philosophy at the University of Pittsburgh where he also served for many years as Director of the Center for Philosophy of Science. He is a former president of the Eastern Division of the American Philosophical Association, and has also served as President of the American Catholic Philosophical Association, the American Metaphysical Society, the American G. W. Leibniz Society, and the C. S. Peirce Society. An honorary member of Corpus Christi College, Oxford, he has been elected to membership in the European Academy of Arts and Sciences (Academia Europaea), the Institut International de Philosophie, and several other learned academies. Having held visiting lectureships at Oxford, Constance, Salamanca, Munich, and Marburg, Professor Rescher has received seven honorary degrees from universities on three continents (2006 at the University of Helsinki). Author of some hundred books ranging over many areas of philosophy, over a dozen of them translated into other languages, he was awarded the Alexander von Humboldt Prize for Humanistic Scholarship in 1984. ontos verlag has published a series of collected papers of Nicholas Rescher in three parts with altogether fourteen volumes, each of which will contain roughly ten chapters/essays (some new and some previously published in scholarly journals). The fourteen volumes would cover the following range of topics: Volumes I - XIV STUDIES IN 20TH CENTURY PHILOSOPHY ISBN 3-937202-78-1 · 215 pp. Hardcover, EUR 75,00

STUDIES IN VALUE THEORY ISBN 3-938793-03-1 . 176 pp. Hardcover, EUR 79,00

STUDIES IN PRAGMATISM ISBN 3-937202-79-X · 178 pp. Hardcover, EUR 69,00

STUDIES IN METAPHILOSOPHY ISBN 3-938793-04-X . 221 pp. Hardcover, EUR 79,00

STUDIES IN IDEALISM ISBN 3-937202-80-3 · 191 pp. Hardcover, EUR 69,00

STUDIES IN THE HISTORY OF LOGIC ISBN 3-938793-19-8 . 178 pp. Hardcover, EUR 69,00

STUDIES IN PHILOSOPHICAL INQUIRY ISBN 3-937202-81-1 · 206 pp. Hardcover, EUR 79,00

STUDIES IN THE PHILOSOPHY OF SCIENCE ISBN 3-938793-20-1 . 273 pp. Hardcover, EUR 79,00

STUDIES IN COGNITIVE FINITUDE ISBN 3-938793-00-7 . 118 pp. Hardcover, EUR 69,00

STUDIES IN METAPHYSICAL OPTIMALISM ISBN 3-938793-21-X . 96 pp. Hardcover, EUR 49,00

STUDIES IN SOCIAL PHILOSOPHY ISBN 3-938793-01-5 . 195 pp. Hardcover, EUR 79,00

STUDIES IN LEIBNIZ'S COSMOLOGY ISBN 3-938793-22-8 . 229 pp. Hardcover, EUR 69,00

STUDIES IN PHILOSOPHICAL ANTHROPOLOGY ISBN 3-938793-02-3 . 165 pp. Hardcover, EUR 79,00

STUDIES IN EPISTEMOLOGY ISBN 3-938793-23-6 . 180 pp. Hardcover, EUR 69,00

ontos verlag

Frankfurt • Paris • Lancaster • New Brunswick 2006. 14 Volumes, Approx. 2630 pages. Format 14,8 x 21 cm Hardcover EUR 798,00 ISBN 10: 3-938793-25-2 Due October 2006 Please order free review copy from the publisher Order form on the next page

P.O. Box 1541 • D-63133 Heusenstamm bei Frankfurt www.ontosverlag.com • [email protected] Tel. ++49-6104-66 57 33 • Fax ++49-6104-66 57 34