Вы находитесь на странице: 1из 279

KARL-OTTO APEL: SELECTED ESSAYS Volume One Towards a Transcendental Semiotics

Edited and Introduced by EDUARDO MENDIETA

HUMANITIES PRESS NEW JERSEY

First published in 1994 by Humanities Press Inte Atlantic Highlands, New Jersey 07716. 1994 by Karl-Otto Apel Library of Congress Cataloging-in-Publication C Apel, Karl-Otto. [Essays. Selections] Karl-Otto Apel : selected essays / edited and introduced by Eduardo Mendieta ; preface by Karl-Otto Apel. p. cm. Includes bibliographical references and index. Contents: v. 1. Towards a transcendental semiotics. ISBN 0-391-03807- 9 (hard) 1. First philosophy. 2. Language and languagesPhilosophy. 3. Semiotics. 4. Hermeneutics. I. Mendieta, Eduardo. II. Title. B3199.A63E5 1993 193dc20 93-12057 CIP A catalog record for this book is available from the British Library. All rights reserved. No part of this publication may be reproduced or transmitted, in any form or by any means, without written permission from the publisher. Printed in the United States of America

SOY INK

PRINTED WITH

CONTENTS

Acknowledgments Preface KARL-OTTO APEL Introduction EDUARDO MENDIETA 1. A N A L Y T I C PHILOSOPHY OF L A N G U A G E A N D T H E GEISTESVOTSSENSC HAPTEN INTENTIONS, CONVENTIONS, A N D REFERENCE T O THINGS: M E A N I N G IN HERMENEUTICS A N D T H E A N A L Y T I C PHILOSOPHY OF L A N G U A G E T H E TRANSCENDENTAL CONCEPTION OF L A N G U A G E COMMUNICATION A N D T H E IDEA OF A FIRST PHILOSOPHY: T O W A R D S A CRITICAL RECONSTRUCTION OF T H E HISTORY OF PHILOSOPHY IN T H E LIGHT OF L A N G U A G E PHILOSOPHY TRANSCENDENTAL SEMIOTICS AS FIRST PHILOSOPHY T H E " P R A G M A T I C T U R N " A N D TRANSCENDENTAL SEMIOTICS: T H E COMPATIBILITY OF T H E "LINGUISTIC T U R N " A N D T H E " P R A G M A T I C T U R N " OF M E A N I N G THEORY WITHIN T H E FRAMEWORK OF A TRANSCENDENTAL SEMIOTICS C . S. PEIRCE A N D POST-TARSKIAN T R U T H TRANSCEN DENTAL SEMIOTICS A N D HYPOTHETICAL METAPHYSICS OF EVOLUTION PRAGMATIC PHILOSOPHY OF L A N G U A G E BASED ON TRANSCENDENTAL SEMIOTICS Index

vii viii

xi

2.

51

3.

83 112

4. 5.

132 175

6. 7.

207

8.

231 255

ACKNOWLEDGMENTS

I would like to acknowledge my deep debt to my long-time mentor, Stephen Eric Bronner, with whom I first studied Critical Theory, who encouraged me on this project and helped make it possible. I would also like to thank Enrique Dussel for his influence on my intellectual development, and Edmund Arens and Greg Klass, who read the manuscript and provided valuable advice. Seyla Benhabib taught me about Communicative Ethics, and Richard J. Bernstein bettered my understanding of American Pragmatism and Wittgenstein. Both thus contributed immensely to my understanding of Professor Apel's philosophy. O f course, this collection would not have been possible without Apel's permission. Through many conversations and a trans-Atlantic correspondence and by making available many unpublished materials he has generously guided me through his work. I would also like to acknowledge the very important help that I received from Jennifer Farquhar, Cornelia Tutuhatunewa, and Kathy Delfosse, whose meticulous editing helped make this collection of essays into a book. Finally, let me thank Keith M . Ashfield, President of Humanities Press, for his enthusiasm and encouragement. EDUARDO MENDIETA THE NEW SCHOOL FOR SOCIAL RESEARCH

vii

PREFACE TO THE FIRST VOLUME

After some struggles with my doubts and scruples, I have accepted Eduardo Mendieta's proposal to collect into two volumes the papers I have published in English, along with some new translations, during the last two decades. Indeed, 1 had primarily thought of elaborating my German works into a more systematic and coherent continuation of what I previously called "transformation of philosophy." But this project will still take some time, and so these two volumes present, in advance, those English papers thatnot quite coherentlymark the two main lines of my transformation project: namely, on the one hand, that of t r a n s c e n d e n t a l semiotics as the novel, post-linguistic turn paradigm of First Philosophy as theoretical philosophy; and, on the other hand, that of c o m m u n i c a t i v e or discourse ethics as the corresponding paradigm of First Philosophy as practical philosophy.
1

The present volume deals primarily with the first of these two dimensions of First Philosophy. It opens with two essays ("Analytic Philosophy of Language and the Geisteswissenschaften" and "Intentions, Conventions, and Reference to Things") that are primarily concerned with building the bridge between language-analytic philosophy and the Continental h e r m e n e u t i c tradition, from which I myself started out. These opening pieces are also closely connected with my long-standing preoccupation with the epistemological and methodological foundations of the social or cultural sciences, that is, with the e x p l a n a t i o n versus u n d e r s t a n d i n g controversy. These opening essays are further supplemented by two essays ("The Hermeneutic Dimension of the Social Sciences and Its Normative Foundations" and "Types of Rationality Today: The Continuum of Reason between Science and Ethics") that have been placed in the second volume because critical h e r m e n e u t i c or reconstructive sciences, being not value-neutral (in contradistinction to the standard natural sciences and the quasi-nomological social sciencesi.e., the behavioral sciences), presuppose a philosophical foundation of ethics and may in turn by their results contribute to an ethical assessment of the human situation as a product of cultural evolution and history.
2

After the hermeneutic inauguration, however, all the subsequent essays of the present volume clearly stand in the service of expounding the idea of a t r a n s c e n d e n t a l semiotics that could take the place of ontological metaphysics and m e n taUstic epistemology by taking over and fulfilling the methodological function of a First Philosophy in our time. This idea is first introduced in the essays "The Transcendental Conception of Language-Communication and the Idea of a First Philosophy" and "Transcendental Semiotics as First Philosophy," which provide viii

PREFACE

iX

a reconstructive outline of the historical sequence of the three main paradigms of First Philosophy. The second of these essays (which served as an introduction to my Ernst C a s s i r e r L e c t u r e s delivered at Yale University in 1977) tries to provide not only a historical reconstruction but also a systematic foundation for the sequence of paradigms, the last of which is transcendental semiotics itself. The last four essays of the present volume, which are more recent, try to show how and in what respects t r a n s c e n d e n t a l semiotics (which in my opinion comprises t r a n s c e n d e n t a l pragmatics of language and t r a n s c e n d e n t a l hermeneudcs) may fulfill its function with regard to the traditional problems of metaphysics and epistemology, that is, regarding the problem of truth. The reader can easily realize, in this context, how much I owe to Charles Sanders Peirce, to whom I devoted much exegetical work before turning to my own conception of a transcendental semiotics.
3

In a sense one may say that, by relying on Peirce's "pragmaticism" rather than on the subjectivist, nominalist, and particularist versions of pragmatism and neopragmatism, I came to take another option than did Richard Rorty in conceiving of a post-metaphysical (or even post-epistemological) conception of philosophy, as is indeed required in our day. Although I can agree with the acceptance of a "de-transcendentalization" with regard to categorical schemes, I would insist that this very argument for de-transcendentalization, through its validity claim, presupposes a transcendental a priori with regard to the necessary presuppositions of argumentative discourseas, for example, the regulative principle and counter-factual anticipation of an ultimate universal consensus to be reached in the long run by the indefinite argumentation community. Rorty himself confirms this structure by the validity claims raised by each one of his own verdicts against all universal validity claims of philosophy. He thus ends upas do the postmodernists following Nietzschewith the novel rhetorical figure of constantly committing a performative self-contradiction.
4

The significance of this crucial point will be further clarified, and also dramatized, in the essays of the second volume, which is concerned with the foundation of a universally valid ethics. KARL-OTTO APEL FRANKFURT AM MAIN

NOTES
1. Karl-Otto Apel, Transformation der Philosophic, 2 vols. (Frankfurt a.M.: Suhrkamp Verlag, 1973). Selective English translation: Towards a Transformation of Philosophy, trans. Glyn Adey and David Frisby (London: Routledge &. Kegan Paul, 1980). 2. Karl-Otto Apel, Die " E r U d r e n N e r s t e h e n " Kontroverse in transiendentalpragmatischer Sicht (Frankfurt a.M.; Suhrkamp Verlag, 1979). English translation: Understanding and Explanation: A Transcendental-Pragmatic Perspective, trans. Georgia Warnke (Cambridge, Mass.: MIT Press, 1984). 3. Charles Sanders Peirce, Schriften, 2 vols., ed. by Karl-Otto Apel (Frankfurt a.M.:

PREFACE

Suhrkamp Verlag, 1967, 1970); Karl-Otto Apel, Der Denkweg von Charles Sanders P e a c e : Eine Einfiihrung in den amerikanischen Prugmatismus (Frankfurt a.M.: Suhrkamp Verlag, 1975); English translation: Charles S. Peirce: From Pragmatism to Pragmatkism (Amherst, Mass.: University of Massachusetts Press, 1981). See also my argument with Rorty and the postmodernists in Diskurs and Veranttwrtung (Frankfurt a.M.: Suhrkamp Verlag, 1988); selective English translation: Discourse and Responsibility, to appear with Columbia University Press).

INTRODUCTION

Karl-Otto Apel is one of the most important philosophers of postwar Germany. Characterized by the dual program of rescue and transformation, Apel's work appropriates much of classical German philosophy as it confronts its reactionary elements. More than that, however, he opens these traditions to a fruitful interchange with other philosophical positions. In Apel's work we encounter both the experiences of a national catastrophe and the crises of Enlightenment thought. Indeed, Apel's essays show a set of deep engagements, confrontations, and juxtapositions with thinkers belonging to what in the 1960s, 1970s, and even still today are considered to be irreconcilable traditions. American pragmatism, Italian humanism, analytic philosophy, German hermeneutics, and semiotics all influence his thought. They serve to establish a process of mutual questioning and correction of the traditions in question. Apel's work, only to be compared to Habermas's, was, thus, extremely important in opening postwar German philosophy to foreign philosophical traditions. Thanks to Apel, for instance, C . S. Peirce has become as much a household name as have Heidegger, Gadamer, and Adorno. However, Apel's work is not limited to illuminating exegesis and panoramic historical reconstructions. It is also deeply marked by a systematic and architectonic drive. In all of his essays one can see the designs of a larger formulation. This became in the late sixties a systematic philosophical program that went by the name of a semiotical transformation of transcendental Kantian philosophy. But as other philosophers made similar moves, Apel complemented his "linguistic turn" with a transcendental semiotics. This transcendental semiotics is a First Philosophy ( p r i m a philosophia), the theoretical part of a system that grounds a practical philosophy that is expressed concretely in a discourse ethics. Along with this more detailed architectonic, Apel has continued to develop his ancillary programs of a transcendental pragmatics, a transcendental hermeneutics, and a theory of rationality. Today, however, what lies at the center of Apel's focus and source of preoccupation is the development of a Discourse Ethic that may act as the grounding for a macro-ethic of planetary coresponsibility, and its possible applications and consequences for our present world situation in which we face accentuated North-South inequities, a precarious global ecological situation, and a deeply suspect "triumph" of Western democracies and their underlying notions of progress.
xi

Xii

INTRODUCTION

II Karl-Otto Apel was born on March 15, 1922, in Diisseldorf. In 1940, as an eighteen-year-old youth, Apel, along with his entire graduating class, volunteered for military service. He fought on the Russian front and became a prisoner of war. Apel's war experiences are of great importance in understanding the nature of his philosophical project. Apel himself has suggested that his war experiences are at the root of his philosophizing. The experience that "everything was false" led him to search for a solid foundation that would allow him to demystify what was dogmatically presented as ethical or true. This way of portraying Apel's motivations for philosophizing borders dangerously on psychologizing if one does not contextualize it within his later university studies, his confrontation with the bankruptcy of the German philosophical tradition that contributed to the failure of moral nerve under Nazism, and the immediate postwar period. Apel describes this in his autobiographical essay "Zuriick zur Normalitat? Oder konnten wir aus der nationalen Katastrophe etwas Besonderes gelernt haben?" (Return to Normality? Or Could We Have Learned Something Special from the National Catastrophe?).
1

In the fall of 1945, Karl-Otto Apel entered the University of Bonn to study philosophy, history, and G e r m a n i s t i k . There he studied with the Rankeans Holtzmann and Braubach, the historian of literature Giinther Miiller, and the neo-Humboldtian language scientist (Sprachwissenschaftler) Leo Weisgerber. He also studied with Oskar Becker, a historian of mathematics who had an existentialist-phenomenological inclination, the neo-Hegelian Theodor Litt, and the important medieval historian and scholar of Romance languages ErnstRobert Curtius. But it was Erich Rothacker who became most important for Apel's philosophical development. Under Rothacker, Apel was influenced by a vision of the G e i s t e s w i s s e n s c h a f t e n that combined a life-philosophy ( L e b e n s p h i l o s o p h i e ) with an anthropological-psychological approach. Studying with Rothacker, who was also Habermas's D o k t o r v a t e r , Apel already evidenced his systematic concerns with delineating a transcendental hermeneutics through an anthropological-epistemological ( e r k e n n t n i s a n t h r o p o h g i s c h e s ) transformation of Kantian philosophy. This program was first articulated in his doctoral dissertation, "Dasein und Erkennen," in which he provided an anthropological reading of Heidegger's categories from Being a n d T i m e . The concern with the a priori conditions for grasping the meaning of existence was also at the center of Apel's later historical research into the different traditions of the idea of language. Most important, however, is that in his student years Apel already saw the need to surmount the gap between the analytic tradition and the hermeneutical tradition of German philosophy. The possibility of bridging the gap between these two traditions of the philosophy of language was seen by Apel as resulting from a transformation of transcendental philosophy. This transformation was to be carried out by questioning both the structures of pre-understanding and the search for criteria that

INTRODUCTION

Xlll

allows for the determination of knowledge claims. After his doctoral dissertation (in 1950) Apel became Rothacker's research assistant and engaged in a project financed by the Mainzer Akademie der Wissenschaften und Literatur. Eventually this group founded the A r c h i v fiir Begriffsgeschkhte, in which Apel's first essays appeared. Here also was published his still-important work on the historical roots of modern philosophies of language: D i e I d e e der S p r a c h e in der T r a d i t i o n des H u m a n i s m u s v o n D a n t e bis V i c o ( T h e I d e a of L a n g u a g e in the T r a d i t i o n of H u m a n i s m f r o m D a n t e to V i c o ) . This historical research led Apel to put aside the systematic elaboration of the project of a transcendental epistemological anthropology laid out in this dissertation. Still, he was able to relate one dimension of his earlier systematic concerns to this new aspect of his work: namely the a priori status of language as the medium through which the world is disclosed to us. Here, in these studies, Apel starts to link Heideggerian philosophy and the neo-Humboldtian language research of Weisgerber with the analytic philosophy of language. In these early formulations, language acts for Apel not only as the precondition of facts and events, but also as the precondition of the possibility and validity of intersubjective knowledge. This would become a distinctive and normative feature of the way Apel discussed and engaged with other philosophers of language.
1

These early years of research, however productive, were still very hard for Apel, who contracted an eye infection that almost left him blind. It required he stop his research for several years. In fact, the book T h e I d e a of L a n g u a g e in the T r a d i t i o n of H u m a n i s m f r o m D a n t e to V i c o was a partial rendering of a much larger project that could not be completed due to his illness. After this hiatus in the early sixties, Apel presented his Hofcilitation, and went on to teach in Kiel. From his Kiel years came his most famous, and now classic, work, Transformation der Philosophie (Transformation of P h i l o s o p h y ) , originally published in 1972. As this tour de force evidences, Apel had made his great discovery, C . S. Peirce, whom he had at first read as a forerunner of Dewey and James, but whom he later came to consider the greatest American philosopher. Peirce would serve as the bridge to link Continental hermeneutical philosophy with the analytic philosophy of language.
3

In the first volume of Transformation der Philosophie, subtitled "Language Analysis, Semiotics, and Hermeneutics," we encounter a confrontation and convergence between an ontological hermeneutics and a transcendental therapeutic critique of language. The key philosophical figures in this confrontation and climactic convergence are Wittgenstein and Heidegger. The point of commonality that allows Apel to bring these two seemingly disparate philosophical giants together is the questioning of Western metaphysics as a theoretical science. Wittgenstein's work was motivated by the question of how to overcome the seduction of the "metaphorical appearance" of the language of philosophy. This demystification is at the center of Wittgenstein's focus on the question of the criteria of the sense and nonsense of propositions. Heidegger, similarly, is concerned with the "oblivion of Being" (Seiravergessenneit) as it

Xiv

INTRODUCTION

takes place in Western metaphysics qua ontology, also a mystification. In both, then, Apel encounters a critique of Western mystifying metaphysics, albeit from the standpoint of an unacknowledged and unthematized transcendentality of language. Both criticize dogmatic metaphysics from a transcendental point that itself remains unelaborated, unjustified, and ungrounded. The question for Apel, then, is how a mutual point of reference and corrective may be located. What emerges from Apel's juxtaposition of Wittgenstein with Heidegger is the "transcendentality" of language. This transcendentality is understood by Apel in the sense of providing the necessary preconditions for perceiving the objects of knowledge and hermeneutically allowing meaning to appear. These insights are further elaborated and substantiated in the first volume of T r a n s f o r m a t i o n , through a series of essays in which language is also discussed as a meta-institution vis-a-vis Arnold Gehlen's theory of institutions. The relationships between language and order and between language and truth are also closely studied. It must be noted perhaps not so parenthetically that Apel and Habermas always need to digest the entire material of a tradition before presenting their own positions. Indeed, when presented within the context of the history of a problematic, their positions seem to emerge dialectically from the inner aporias of that particular philosophical problem. Thus, in each one of his essays, before launching into the particular issue at hand, Apel gives us magnificent overviews of the history and evolution of particular philosophical conceptions: histories of phenomenology and the phases it has undergone from Husserl through Heidegger; histories of hermeneutics and the differences between Schleiermacher, Dilthey, Hegel, Gadamer, and Heidegger. But this encyclopedic knowledge does not restrict itself to Continental philosophy. It also extends to knowledge of the history of analytic philosophy. This is most evident in the essay on Charles Morris. In the uncovering of language as the ground of transcendentality, in the first volume of T r a n s f o r m a t i o n der P h i l o s o p h i c further elucidation of transcendental hermeneutics takes place. Dealing with Gadamer's work, it became clear to Apel that transcendental questions concerning the conditions of the possibility of knowledge must remain primarily a questioning after the conditions of inter-subjective validity claims. These validity claims must not be responded to by recourse to a contingent a priori of the world-pre-understanding, in the sense of a quasi-ontological concept such as "meaning-event" or "truth-event," if one does not want to fall into historicism-relativism. Rather, it must be necessary and possible to appeal to the complementary non-contingent a priori (which cannot be challenged without a self-performative contradiction) of the ideal, universally valid presuppositions of argumentation; that is, -the argumentative discourse of an ideal, unlimited communication community that is always counter-factually presupposed. Gadamer's position, suggests Apel, was still infected with the dogmatism of a metaphysical position that does not submit its own claims to the process of validation implied in every transcendental discus-

INTRODUCTION

XV

sion. Gadamer had skipped over the question of standards and normative rules that allow one to differentiate between meaning as truth and truth as meaning. In this discussion Apel moves towards an overcoming of hermeneutical relativism and historicism by means of his own version of ttanscendental hermeneutics, which in later formulations he would locate within the edifice of a transcendental semiotics. Turning to the second volume of T r a n s f o r m a t i o n der Philosophie, a different "hermeneutical horizon" appears. We encounter Apel's attempt to provide a semiotical transformation of transcendental philosophy and an ultimate foundation ( L e t z t b e g r i i n d u n g ) for ethics. Whereas the first volume was determined by the influence of Heidegger, and the preoccupation with the hermeneutics of language, the second volume seeks normative justification for validity claims and a transcendental grounding in a theory that marries hermeneutics, semiotics, and pragmatics. If in the first volume Heidegger and Wittgenstein were the key figures in the constellation of modern philosophy, Peirce, Royce, and Kant emerge as the key figures in the second volume. This is not to imply that Heidegger and Wittgenstein ate left behind, superseded. Rather, their ideas are subsumed within a new patadigm, a new problematic. Nor does this shift of emphasis between the first and second volume imply a clear break. Rather, thete is a continuity. Indeed, as Apel questions the radical critiques of Wittgenstein and Heidegger, which had elaborated the "oblivion of Being," he asks whether both of them had not engaged in a similar oblivion, but in this case of logos (Logos v e r g e s s e n h e i t ) . Apel's transcendental philosophy discusses the conditions of the possibility of knowledge; still, it focuses neither on ontological questions nor on the epistemological conditions of knowledge. Rathet, departing from the insights already gained in volume one, this new transcendental philosophy begins as a critique and analysis of language. Extending the patallel, if Kant's was a ttanscendental philosophy that tried to establish the conditions of valid knowledge and the conditions of the constitution of objects, contemporary philosophy departs from the establishment of such conditions as conditions that happen in and through language. The so-called linguistic turn is made evident and patently clear by Apel. Thus, from a pure critique of knowledge, knowledge as constituted by and for a monological transcendental subject, we move to a critique of language. This language, however, is not understood by Apel as the logical calculus or lingua philosop/uca of analytic-linguistic philosophy. It appears instead in the . fullness of its triadic dimensions: syntactics, semantics, and pragmatics. Syntactics refers to the relationship between signs. It is from this semiotical dimension of language that the modern logical-mathematical philosophies of science atise. Semantics refers to the relationship between signs and empirical states of affairs. It is from this dimension that empiricist and positivistic projects of a unified science take their departure. Finally, there is pragmatics, which deals with the relationship between signs, the objects they point to, and human beingstheir userswithin contexts of communities of users. It is from here

XVi

INTRODUCTION

that American pragmatists began the attempt to develop a theory of science. In contradistinction to the projects of a philosophy of science, which tries to reduce science to either a syntactics or a semantics, Apel insisted that knowledge as a sign-mediated activity is always an interpretation by humans performed within the context of a community of scientists or a community of interpretation. Here two central ideas are profiled. The communication community turns out to be a transcendental horizon, while theoretical and practical reason converge. This means that every person, in as much as he or she is a member of human society, is always already embedded in a community of communication that" acts as the "institution of institutions." Within this institution of institutions we are always already entwined in a web of normative presuppositions that command the individual to enter into the process of redeeming the validity claims that ale constantly raised in the process of communicative interactions and argumentation. These conditions and presuppositions of discursive argumentation constitute an intranscendable, uncircumventable ( n i c h t h i n t e r g e h b a r ) horizon. Indeed, a person may intentionally pretend to disregard these transcendental presuppositions of the transcendental language game, but he or she can only do so at the risk of falling into a "performative contradiction." A '^performative contradiction" is the disruption and self-annulment that takes place when a person contradicts what they are saying with what they are pragmatically presupposing in order for their claim to make sense. This limit concept becomes for Apel an ultimate foundation ( L e t z t b e g r t i n d u n g ) , which serves as the archimedian point for a grounding of both the theoretical and practical sciences: hence, the convergence of practical and theoretical reason. From T r a n s f o r m a t i o n der P h i l o s o p h y , then, we are able to gather and observe the breadth of Apel's philosophical system. A parallel with Jiirgen Habermas's project of the reconstruction of the social sciences can best illustrate the reach and importance of Apel's philosophical work. The parallel is entirely warranted by their long friendship and mutual intellectual indebtedness, going back as far as the 1950s, when they met in Rothacker's seminars. In fact, the parallel can be extended, and one can speak of a division of labor between Apel and Habermas. If Habermas has formulated his theory of communicative rationality and action in order to extract sociopolitical theory from the cul-de-sac into which it was driven by the conflation of reason with instrumental rationality, Apel has called for and articulated a transformation of philosophy, in terms of a transcendental semiotics that includes both a hermeneutics and a pragmatics, in order to extricate philosophical discourse from the aporias of a reason entrenched in "methodological solipsism" and "abstractive fallacies." In this parallel, the hybrid status of Apel's and Habermas's projects is revealed. Both are engaged in critically preserving and appropriating the best of the German critical and idealist tradition, while at the same time having it interact dialectically with British-American analytic philosophy, pragmatism, and systems theory.
4 5

Among the projects delineated in the T r a n s f o r m a t i o n , then, are the foundations of an anthropological epistemology that analyzes the conditioning charac-

INTRODUCTION

XVJi

ter of constitutive knowledge interests, interests that orient us not only in the tealm of object-subject and object-object relations, but also in those of subjectco-subject. This project is broached once more in a few of the essays collected in this volume, in particular "Transcendental Semiotics and Hypothetical Metaphysics of Evolution" and "Pragmatic Philosophy of Language Based on Transcendental Semiotics." Volume Two of the T r a n s f o r m a t i o n anticipates a transcendental semiotic, and this is the focus of most of the essays here. Also present, as a corollary to a transcendental anthropological epistemology, is the development of a theory of rationality. (This will be taken up once again in the second volume of K a r l - O t t o A p e l : Selected E s s a y s , which will be entitled Ethics a n d the T h e o r y of R a t i o n a l i t y . ) Most important, however, is that in T r a n s f o r m a tion we have the foundations for a formulation of a theory of discourse ethic. This theme has become the focus of Apel's philosophizing and has yielded a very important volume entitled D i s k u r s u n d V e r a n t w o r t u n g : D a s P r o b l e m des U b e r g a n g s z u r postkonventioneUen M o r a l ( D i s c o u r s e a n d Responsibility: T h e P r o b l e m of T r a n s i tion to a Postconventional M o r a l i t y ) . In this volume, Apel deals with the possibility and the necessity of grounding ethics on the uncircumventability of the communication community. After Kiel, Karl-Otto Apel moved to the Goethe University in Frankfurt, where he taught in the philosophy department, along with Habermas, and from which he recently received his emeritus status. From his Frankfurt years there comes a series of in-depth essays confronting the critical rationalists, and in particular Albert's critique of the notion and possibility of a grounding of the sciences and ethics. During his Frankfurt period, Apel also entered into debate with the postmodernists and their challenge to the universality of reason and the possibility of a normative grounding of ethics. In particular, he criticized Richard Rorty's neopragmatism and his relativization of the a priori communication community that acts as the already given background of understanding. Part of Apel's debate with Rorty is over the status of transcendental argumentation; Apel argues that it allows us not to fall back into the kind of extreme contextualisms or cultural relativism that have lead Rorty to abdicate to the postmodernist critiques of universalism ( l o g o c e n t r i s m ) and to retreat behind the hermeneutical horizon of the American political tradition as a point of reference and a warrant for his philosophical claims. Another very important topic of Apel's philosophizing during his Frankfurt tenure was the question of truth and its relationship, on the one hand, to the a priori, historicizing communication community in which truth is claimed and disclosed, and, on the other hand, to the fallibility of knowledge implied by the Peircean notion of truth as the consensus that would be reached in the long run in an unlimited community of researchers and arguers. Some aspects of this discussion are dealt with and articulated in some of the essays of the present volume.
6

XVlii

INTRODUCTION III

The works collected in this first volume of K a r l - O t t o A p e l : Selected Essays span his critical engagement with the Continental and analytic philosophies of language. The first two essays illustrate the concern with this historical convergence of both the analytic and hermeneutical traditions. The first essay, which dates to 1965, is concerned primarily with the history of analytic philosophy and its unfolding towards a point of convergence with the hermeneutical tradition of the philosophy of language. In this essay, Apel delineates three major stages of the evolution of analytic philosophy and the major aporias that led analytic philosophers to move to the next stage. Placing analytic philosophy within the Enlightenment tradition of the attempt to formalize the sciences of humanity along the model of the natural sciences, Apel characterizes the project of analytic philosophy as an attempt to establish the possibility, or as it turned out the impossibility, of a normative and quasi-nomological science of human interactions. In the second essay, Apel once again takes up the project of bridging the gap between analytic and Continental or hermeneutical philosophies of language. This time, however, the project is pursued by means of a discussion of the treatment of meaning in both traditions. Apel provides detailed historical profiles of preceding treatments. In discussing the nature and treatments of the question of meaning, Apel focuses on three different aspects of the description of meaning and what leads to its understanding: intentions, conventions, and reference. These in turn, as Apel shows, have been treated unilaterally or have been over-emphasized by different currents within both traditions. The close and systematic readings of the different ways the question of meaning has been treated under these three headings leads Apel to the formulation of "transcendental semiotics" as the perspective from which the dialectical mediation between these aspects can be achieved without the over-emphasis or neglect of any of them. The status and full description of this transcendental semiotical perspective will unfold methodically through the following and complementary essays. The complementarity between the next two essays is similar to that between the prior two. The first of these two essays is historical in character. In it, Apel presents us with a critical historical reconstruction of the philosophies of language, beginning with Aristotle and ending with the contemporary linguistic turn of philosophy, which takes language as its methodological point of departure. This new vantage point has come to be considered the sine qua non of any contemporary philosophical approach to the perennial questions of philosophy. In both essays, three different paradigms of First Philosophy, p r i m a philosophia, are sketched. The first paradigm is First Philosophy as ontology, and stems from Aristotle and Plato. Here the question is the correspondence between ideas and ontological existents. Under this paradigm language is conceptualized as a means to portray, under the correspondence relationship, things that exist and

INTRODUCTION

XJX

that need to be established befote we can assign names to them. The second paradigm is First Philosophy as epistemology, and stems from Kant; it begins with a discussion of the conditions of the possibility of certainty. The third paradigm of First Philosophy tries to integrate and transform the insights of the first two paradigms. According to Apel's critical reconstruction, this third paradigm was first formulated by Peirce. But its clearest expression is in terms of a transcendental semiotics. This transcendental semiotics is characterized, using Peirce's and Morris's descriptions of semiotics, as the mediation of all claims to knowledge by the interdependence of the three ways in which objects, their signs, and the users of those signs are related. Indeed, in the second of these two essays. "Transcendental Semiotics as First Philosophy," Apel elucidates the two preceding paradigms as moments in the triadic relationship between object, sign, and community of sign users and interpreters. If the first paradigm focused unilaterally on the relationship between things and consciousness, not yet intersubjectively and communicatively conceptualized, the second focused on the internal relationship of sign representation and consciousness, analyzed critically from the standpoint of their making possible knowledge of the world. If the first paradigm abstracted from knowledge of things its mediation by language, the second abstracted from the signs that allow us to cognize the world their belonging to a process of interpretation. Apel demonstrates not only the order of succession of these three paradigms, but also what they either thematize or presuppose. The final essays in this volume deal with very specific philosophical problems and how they are solved by a transcendental semiotical-pragmatic approach. In these last essays Apel deals with the question of truth and its different treatments by Tatski, Searle, Habermas, Putnam, Grice, and other major philosophers. In these essays Apel also broaches the question of the aporias bequeathed to us by Kant as they pertain to the validity of scientific knowledge, natural causality, and teleology. These aporias are viewed from the standpoint of a metaphysics of the evolution of human knowledge that attempts to establish a connection between nature, the pre-history of humanity, and the unfolding of human history as the mutual unfolding of nature and human knowledge. In particular, Apel draws on Peirce's later metaphysical speculations. In a similar fashion to Mead's claims about the metaphysical status of sociability, which as a principle establishes the mutual unfolding of object and social self and therefore the interpenetration of reason and society, Apel establishes that there is a convergence between knowledge of nature and growth in philosophical knowledge. In the last essay Apel proceeds to demonstrate how the pragmatic turn of the philosophy of language, as another moment within the unfolding of the linguistic turn of philosophy, can best be understood and accounted for from within a semiotical transcendental philosophy, qua First Philosophy. The claim is that a pragmatic turn of the philosophy of language is a necessary condition for explicating truth claims and validity claims. But this is not yet a sufficient condition. The sufficiency, so to say, is provided by a fully elaborated

XX

INTRODUCTION

transcendental semiotics, which is what Apel here tries to achieve. These essays constitute a major conttibution to the knowledge of the historical unfolding of one of today's most important philosophical currents, namely the discoutse theory of ethics and its underpinning theory of communicative rationality. These essays also provide a point of departure for furthet study of the problematics formulated by this current. Among some of the most important problems to be studied furthet are the status of transcendental arguments in a postmodern world, the contradictions between the real and ideal communication communities, the tole of a de-utopianized Marxism in criticizing the asymmetries of actual communication communities, the prerequisites for developing a planetary macro-ethic of coresponsibility that would address the widening gap between North and South, and the growing global ecological crises. If nothing else, however, these essays should help English-speaking intellectuals to come into contact with one of the most important Continental philosophets of our time.
7

EDUARDO MENDIETA THE NEW SCHOOL FOR SOCIAL RESEARCH

NOTES
1. Karl-Otto Apel, Diskurs und Verantwortung: Das Problem des Dbergangs zur postkonventioneUen Moral (Frankfurt a . M . : Suhrkamp Verlag, 1988). English translation forthcoming from Columbia University Press. 2. Karl-Otto Apel, Die Idee der Sprache in der Tradition des Humanismus von Dante bis Vico (Bonn: H . Bouvier und C o . Verlag, 1963). 3. This major work of philosophy was, unfortunately, only partially translated into English, most of the essays coming from the second volume of the German original. Karl-Otto Apel, Transformation der Philosophie, 2 vols. (Frankfurt a . M . : Suhrkamp Verlag, 1973). English translation: Towards a Transformation of Philosophy, trans. Glyn Adey and David Frisby (London: Routledge & Kegan Paul, 1980). 4. O n the relationship between Apel and Habermas, see the speech Habermas gave at the colloquium organized on the occasion of Apel's receiving his emeritus status from the Goethe University in Frankfurt. This may be found as an afterword to Walter Reese-Schafer's Karl-Otto Apel zur Einfuhrung (Hamburg: Junius Verlag, 1990), 137-49. 5. See Jiirgen Habermas, Vorstudien und Ergdn^ungen zur Theorie des kommunikativen Handelns (Frankfurt a . M . : Suhrkamp Verlag, 1984), 7. 6. Karl-Otto Apel, "Fallibilismus, Konsenstheorie der Wahrheit und Letztbegriindung," in Philosophie und Begrundung, ed. W . Kuhlmann (Frankfurt a . M . : Suhrkamp Verlag, 1987). 7. For treatments of Karl-Otto Apel in English, see Fred R. Dallmayr, Beyond Dogma and Despair: Toward a Critical Phenomenology of Politics (Indiana: University of Notre Dame Press, 1981); Twilight of Subjectivity: Contributions to a Post-Individualist Theory of Politics (Amherst, Mass.: University of Massachusetts Press, 1981); and Critical Encounters: Between Philosophy and Politics (Indiana: University of Notre Dame Press, 1987). This last book contains an entire chapter devoted to Apel's Transformation.

INTRODUCTION

XXi

See also Josef Bleicher, Contemporary Hermeneutics: Hermeneutics as Method, Philosophy, and Critique (London: Routledge & Kegan Paul, 1980). In German, see Kommunikation und Reflexion: Zur Diskussion der Trons^endentalpragmatzk. Antworten auf KarlOtto Apel, ed. Wolfgang Kuhlmann and Dietrich Bohler (Frankfurt a . M . : Suhrkamp Verlag, 1982); and Walter Reese-Schafer, Karl-Otto Apel zur Einfuhrung (Hamburg: Junius Verlag, 1990).

ONE Analytic Philosophy of Language


and the Geisteswissenschaften

SUMMARY

This study attempts a historical account and critical evaluation of the so-called analytic philosophy of language, considered by the author to be the characteristic and dominating new methodical approach in Anglo-Saxon philosophy in the first part of this century. The perspective from which evaluation and critique is viewed is mainly that of the traditional German G e i s t e s w i s s e n s c h a f t e n , which led in this century to the development of a h e r m e n e u t i c a l philosophy, for which the problem of language is also of paramount importance (from Dilthey to Heidegger to H . - G . Gadamer). The confrontation of the two philosophical currents will lead to a historical reconstruction of analytic philosophy in three phases: logical atomism, logical positivism, analytic philosophy of language. In the evaluation of the third phase a considerable convergence of analytic and h e r m e n e u t i c a l philosophy will become apparent; butand here lies the author's main thesisit will also become evident that it is necessary to go beyond both philosophies and to mediate dialectically between the method of intersubjective understanding of language and the methods of objective explanation of behavior, this mediation being necessary since man is not (yet) able to express completely the actual motives of his behavior in intersubjective communication.

INTRODUCTION: T H E M E T H O D I C A L - M E T H O D O L O G I C A L A M B I V A L E N C E O F ANALYTIC PHILOSOPHY OF L A N G U A G E PERVADING T H E T H R E E PHASES OF ITS HISTORICAL D E V E L O P M E N T

The name "analytic philosophy," as it is used today, contains an ambivalence of meaning, which, as we shall see, is of paramount importance for the problems to be discussed: 1. Analytic philosophy stands first for a school of thought, which recognizes as "scientific" only the methods of the natural sciences in the wider sense of the word, insofar as they objectively explain the phenomena in question by reference to causal laws. This philosophy sees as its main goal the justification of this
1

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

objective knowledge and its separation from any kind of subjective Weltanschauung, that is, theology, metaphysics, or some other normative science. It is clear that the philosophy thus characterized, which in Germany is usually identified with the logical positivism of the so-called Vienna Circle, will have a polemical attitude towards the idea of a Geisteswissenschaft and the philosophical concepts which constitute its systematic foundation and its historical roots. Thus viewed, analytic philosophy, today the most influential school of thought in the Western world (at least in the Anglo-Saxon and the Scandinavian), seems to be the continuation of eighteenth-century Enlightenment, which also held that the only legitimate goal of science dealing with man and his culture was to give explanations in terms of laws of nature, if possible, mathematically formalized. Thus it seems as if a confrontation of analytic philosophy and a philosophy of the Geisteswissenschaften can lead to no more than a renewal of that nineteenth-century discussion, in the course of which J. G . Droysen (in his H i s t o r i k , 1868) and later W . Dilthey (in his E i n l e i t u n g in die Geisteswissenschaft e n , 1883) contrasted the concept of "explanation" (of nature) with "understanding" (of the historical-social world as created by man).
1 2

2. The name "analytic philosophy," however, has still another meaning component which, from the point of view of the historian of philosophy, characterizes the methodical starting point of this philosophy more precisely than the vague explication given above. It was not actually the "analytic" methods of the sciences under study by analytic philosophy which gave this philosophy its name, but rather its own method of analysis, that methodical revolution in philosophy which is dominating the Anglo-Saxon world today. This "analysis," however, which is considered so revolutionary, is not applied to the objective facts of science, but rather to the sentences of science, that is, not to things, but to the language that speaks of these things. "Meaning and Truth," "Meaning and Verification," "Language, Truth, and Logic"these are typical titles to be found in the literature of analytic philosophy; and the distinction between meaningful and meaningless sentences is the characteristic theme of the logical positivist's critique of metaphysics.
3

Proceeding from the dichotomy between "explaining" and "understanding," as established in the German tradition of the philosophy of the Geisteswissenschaften, one might expect that also the analytic philosophers in their discussions of meaning have encountered problems that correspond to the problems involved in the concept of "understanding," forone should thinkthe sentences of the causally explaining sciences, in fact all sentences as vehicles of meaning, must first be understood as expressions of human intentions before one can proceed to deduce them from general laws, and thus explain the facts described by them. From this one might conclude: Although analytic philosophy as a philosophy of science accepts as the goal of science only the objectivistic explanation of facts, nevertheless, the problems involved in the very idea of "language analysis"

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

must lead analytic philosophy "through the back door" into the midst of those problems which the theory of "understanding" of the Geisteswissenschaft was designed to cope with. In the following, we shall, in fact, use the parallels outlined thus far as a heuristic device, in order to facilitate a fruitful confrontation between analytic philosophy and the philosophy that is implied by the idea of a Geisteswissenschaft. In doing this, we shall have to consider and interpret what we may call the "objectivism" of analytic philosophy, long indisputably part of its explicit methodology and which must be well distinguished from its own method of philosophizing. In the following pages we shall therefore endeavor 1. to show the methodical-methodological ambiguity already appearing at the outset of analytic philosophy as a philosophy of language analysis; 2. to discuss the claim of the methodology of logical positivism, to wit, that the realm of the Geisteswissenschaften can also be incorporated into the sciences that "explain"; here the antinomy between the objectivist-physicalist and the language analysis viewpoint will have to be made apparent; 3. to follow the development of the self-understanding of the analytic philosophers up to that aporetic point at which the problems of the "understanding" of the Geisteswissenschaften become televant for the self-reflection of language analysis. There is, in fact, an obvious correspondence between the three-point scheme just given and the actual historical development of analytic philosophy. T o see this, however, we must accept the insistence of British chroniclers like Urmson and Charlesworth that analytic philosophy cannot be identified with logical positivism and its central idea of a "unified science," as is done in Germany and sometimes in the United States. In England, logical positivism is only thought of as one stage in the development of a philosophy, which began with B. Russell's, G . E. Moore's, and especially the young Wittgenstein's ideal of a "logical analysis of language," and which has reached its final stage in "linguistic philosophy," as started by the later Wittgenstein and practiced in Oxford and Cambridge today.
4 5 6

In the following I shall adopt this British viewpoint of analytic philosophy and its historical development, and shall try to establish from the very beginning a connection between the analytic problem of understanding language and the problem of "understanding" as seen by the Geisteswissenschaften.

T H E ORIGIN O F THIS A M B I V A L E N C E IN WITTGENSTEIN'S

TRACTATUS

A suitable starting point is the T r a c t a t u s of the young Wittgenstein: a sketch profound and paradoxicalof a transcendental semantics or logic of language, which cannot justify its own method. The afotementioned ambiguity between method and methodology was already present in this sketch, which determined all subsequent developments of analytic philosophy. With respect to our objectives in this study, the origin of this ambiguity can be shown very well if we turn

PHILOSOPHY OF LANGUAGE & GEISTESWISSENSCHAFTEN

to Wittgenstein's short but influential discussion of the so-called intentional or belief-sentences. Sentences of the form " A believes that p," " A thinks p," " A says p" seem to contradict the main thesis ( T r a c t a t u s 5.54) of Wittgenstein's sentential logic, accotding to which a sentence can only be part of any othet sentence as a "truth condition" of the latter. The models of this logic of ttuth-functions are complex sentences which are put togethet out of elementary sentences and sentential connectives, such as "Today the sun is shining and everyone is happy" or "It is taining outside ot the sun is shining." Wittgenstein applied the logic of truthfunctions to language analysis to teveal the form allowing verification of sentences such as "Not all citizens of the Federal Republic are musical," thereby analyzing it as " A is musical, B is musical, etc., M is not musical, O is not musical, etc." In other words, the application of the logic of truth-functions the so-called thesis of extensionalitywas connected, for Russell and Wittgenstein, with the hope of discovering the true logical structure of all sentences, which was thought to be concealed, indeed misleadingly cloaked in the external form of everyday language.
7

This hope was now seriously jeopardized by the existence of sentences of the form " A believes that p," for example, the sentence "Peter believes that it is raining outside." Fot in this case, the proposition "it is raining outside," which seems to be contained in the intentional proposition "Peter believes that it is raining outside," certainly cannot be considered as a truth condition of the latter compound sentence. The point about these "belief-sentences" (even more obvious in the case of sentences of "inditect speech"), is, aftet all, that the ttuth of that which is believed, meant, or said can remain undecided, while the sentence about the belief can nonetheless be true. Sentences of this kind ate obviously a condition for the possibility of such an entetprise as the Geisteswissenschaften. Therefore, the pertinence of Russell's and Wittgenstein's difficulties with the belief-sentences to the questions concerning us is that this is the first time in the history of analytic philosophy that the language of "unified science" comes in conflict with the language of the Geisteswissenschaft, which consists of intentional sentences; for Wittgenstein's thesis of extensionality is the first radical formulation of a thing-fact-language model, whichaccotding to Wittgenstein holds for all meaningful sentences, that is, explicitly, for all sentences of the "natural sciences" ( T r a c t a t u s 4.11). Insofar as the later "objectivism" and "physicalism" of the neopositivist methodology has to be considered as part of analytic philosophy and not metely a continuation of the old metaphysical naturalism, it remains dependent upon Wittgenstein. Its claim is not that of the older positiviststhat the realm of the mind itself can be reduced to the realm of nature and its lawsbut rathet that any knowledge obtained in the Geisteswissenschaften must be translatable into sentences of the one, intersubjective language of science, that is, into the objective language about things and facts.

PHILOSOPHY O F L A N G U A G E & GE1STESWISSENSCHAFTEN

Wittgenstein's difficulties with belief-sentences thus introduce for the first time the problems which arise if we attempt to incorporate the Geisteswissenschaften into an objectivistic unified science in its modern linguistic form. In this light let us regard Wittgenstein's solution of the problem, which, though short and obscure, nevertheless determined the further development of analytic philosophy. Wittgenstein applies Russell's maxim, according to which the philosopher should look for the true structure of thought hidden behind the misleading form of everyday language, in the case of the belief-sentences as well, and he thetefore postulates for the intentional form of sentences:
8

(5.541) (5.542)

. . considered superficially, it looks as if the proposition p stood in some kind of relation to an object A . It is clear, however, that " A believes that p," " A has the thought p," and " A says p" are of the form "'p' says p . " . . .

This much seems to be obvious immediately: In these passages Wittgenstein claims that the true form of the intentional sentences is that of a sentence about the meaning of a sentence-sign. We can now ask the question: Does this solution speak for or against the possibility of incorporating the Geisteswissenschaften into the unified language of the objective natural sciences? A t first glance Wittgenstein's solution seems to speak against it, since a sentence about the meaning of a sentence-sign, fot example, the sentence ' " i l pleut' means: it is raining," seems to be a characteristic sentence of a Geisteswissenschaft, that is, a sentence which can be true, though its component sentences "il pleut" and "it is raining" are not its truth conditions. Then how could Wittgensteinwe may ask ourselvesthink he had saved the thesis of extensionality (which supposedly determines the form of all meaningful sentences), which is undoubtedly what he intended to do, as the context shows? Wittgenstein's comments to the sentence " 'p' says p" indicate how he thought to have salvaged the thesis of extensionality: (5.542) . . . this does not involve a correlation of a fact with an object, but rather the correlation of facts by means of the correlation of their objects.

In other words, what we are here concerned with is not a fact in the world, which can be depicted by language, but the depicting faculty of language itself, which consists of the correspondence of depicting facts and depicted facts. The thesis of extensionality, accotding to Wittgenstein, is true because language in its capacity to depict the world does not admit of a special kind of facts which would consist of a subject (as an element of that fact) in its relationship to a state of affairs (as the other element of that fact)though this seems to be the case with the intentional sentences if they are interpreted psychologically (cf. Tractatus 5.541). This possibility of interpretation Wittgenstein has excluded by offering the sentence form "'p' says p" as the explication of the sentences in question, thus eliminating the human subject of the proposition. He therefore

PHILOSOPHY O F L A N G U A G E & GEISTESWISSENSCHAFTEN

draws the following conclusion from his semantical teduction of intentional sentences: (5.5421) This shows too that there is no such thing as the soulthe subject, etc. as it is conceived in the superficial psychology of the ptesent day. . . .

Note that the English translation of this deeply ambiguous sentence sounds almost precise compared with the German: Dies zeigt auch, daB die Seeledas Subjekt, etc.wie sie in der heutigen oberflachlichen Psychologie aufgefafk wird, ein Unding ist. . . . The negative interpretation of this sentence which was accepted by the logical

positivists does, however, agree with another of Wittgenstein's sentences: (5.631) There is no such thing as the subject that thinks ot entertains ideas. . . .

And it especially agrees with Wittgenstein's central thesis that only the sentences of the natural sciences are meaningful, that is (as the logical positivists understood that concept), intersubjectively verifiable. The foregoing considerations led the logical positivists to conclude that psychology and sociology, as long as they were using the intentional sentence form, were not genuine sciences. Therefore, if these disciplines together with any possibly scientifically respectable part of the so-called Geisteswissenschaften were to be reduced to the language of science, then their sentences could no longer be about intentional pseudofacts, that is, the relationship between a "soul-subject" and the state of affairs meant by it, but would father have to deal with genuine relationships between objects and states of genuine objects. This is the starting point for a program of "behavioral sciences" as part of an objectivist "unified science" program, to which we shall return later in this essay. The short Wittgenstein interpretation given above will, however, have shown that the reduction of intentional sentences to sentences about behavior is at best only half of what the T r a c t a t u s has to say about this problem. O n the one hand, this behaviorist reduction is the only possibility left by the semantical theory of the T r a c t a t u s to make a science out of the Geisteswissenschaften; on the other hand, this reduction is not in accordance with Wittgenstein's reduction of intentional sentences to semantical sentences. The reason that hints at such interpretation were not followed in the period when the T r a c t a t u s made its initial impression on the philosophical world lay, as we mentioned already, in the paradoxical nature of the language theory of the T r a c t a t u s : Sentences like " ' p ' means p " are not to be understood as linguistic representations of facts, according to this theory; but this implies that the sentence " ' p ' means p"to which the sentence " A says p " had just been reduced by Wittgensteinis still itself of a misleading pseudoform, because it still looks like a sentence (e.g., of the form " a R b " ) , though it is not about a fact

PHILOSOPHY O F L A N G U A G E & GE1STESWISSENSCHAFTEN

in the world but about that special relation between fact and sign-fact, in other words, about what must always be presupposed when we speak about a fact in the world. Wittgenstein infers from this that sentences about sentences, that is, semantical sentences, are impossible: What they try to express "shows itself" in the use of language, but it cannot be "said." As is well known, Wittgenstein drew the consequences of this distinction and therefore declared that his own sentences about language and its depicting relationship to the world were meaningless, and that they only fulfilled the function of a ladder in order to reach the final, mystical knowledge.
9

One might contend in objection that it is very unlikely for such a paradoxical philosophy of language to have seriously influenced the methodology of logical positivism. However, in doing so, one would ignore the perfect consistency of Wittgenstein's philosophy of language as a transcendental semantics of a logically clear language which does not permit any metaphorical usage. If the language to be used in science must be constructed as logical calculus, all semantical talk about the meaning of the signs belongs to metalanguagethat is, not to the logically clear language of science. It is, of course, possible to formalize the metalanguage, the metametalanguage, etc., ad infinitum, as Russell suggested in his introduction to the T r a c t a t u s , but in this way one would never arrive at the final metalanguage, which is actually used in the construction and semantical interpretation of any formalized language. Without this semantical interpretation employing the last metalanguage, the formalized language is not yet a semantically functioning languagethough objectively given as a list of signs as objects in the world. And once it has been interpreted with the help of the last metalanguage, the very fact of such an interpretation has shown that the logical form of language, which enables us to describe facts according to Wittgenstein, can itself not be described or constructed as a fact, but rather must always be presupposed. This is what Wittgenstein means when he says
1 0

(6.13)

The logic [of language] is transcendental.

Compare further the following sentences of the Tractatus: (5.555) . . . A n d anyway, how could it be necessary in logic for me to deal with forms that 1 can invent? What must be necessary is that 1 should deal with that which makes it possible for me to invent them. There cannot be a hierarchy of the forms of elementary propositions. We can foresee only what we ourselves construct.

(5.556)

In other words, we can neither construct nor anticipate the logical form of language which is also the logical form of the world. It always precedes such attempts as the condition for the possibility of any constructions. Of course, all traditional transcendental philosophy has always spoken about the form of language and its relation to the world; and Wittgenstein does the same extensively in the Tractatus, but he also demonstrates that such formulations,

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

judged by the (language) ideal of a logical "object language," must by necessity be metaphorical. A sentence of everyday language like "Words have meanings" sounds like the sentence "Men have beards." If we want to point out in which way that sentence is misleadingly metaphorical by saying "the semantical relation between language and facts is not a relation like that between two given objects in the world," we nevertheless in this sentence have to make use of the lingual picture of a relationship between two objects. This is the very difficulty (for statements) of any transcendental philosophy, which Kant already hit upon when he had to distinguish between the metaphorical schematism of the affectation of our senses by the "thing in itself" as an "analogous schematism" on the one hand and an empirical causal relationship on the other.
11 12

Wittgenstein, like his teacher Russell, held such "metaphorical" or "analogous" use of language, as is unavoidable in philosophy, to be the result of a confusion of types, which comes about because the philosopher utters a selfreflexive sentence whenever he talks about the form of mind or language; therefore, according to Russell's theory of typeswhich, however, cannot be formulated as a philosophical theory by its own standards any philosophical sentence is "nonsensical." Wittgenstein drew all of these consequences.
13

A t this point someone might object with respect to the special topic of this study: If all philosophical sentences about language as a whole, that is, about the class of all sentences, are nonsensical according to the theory of types because they have to be applied to themselves too, then this still does not also imply that empirical semantical sentences have to be nonsensical, that is, according to Wittgenstein's reduction, the sentences of the Geisteswissenschaften (e.g., "Goethe's sentence 'Uber alien Gipfeln ist Ruh' means such and such"). In the case of the latter kind of sentence, no "self-reflexiveness" seems to be involved. Why does Wittgenstein then treat them implicitly as self-reflexive and therefore "nonsensical"? From the point of view of the T r a c t a t u s , this only possible alternative to a behaviorist reduction of the sentences of the Geisteswissenschaften will become more apparent, I think, in the light of the idealistic transcendental philosophy which stands behind the traditional idea of the Geisteswissenschaften. Following this philosophy, one could argue in favor of Wittgenstein thus. If really no self-reflexiveness of "the" language or "the" mind is implied by the sentences of empirical semantics, then we have no reason to object to a radically objectivist Geisteswissenschaft as envisioned by the early "physicalism." This is what naturalistic positivism was always convinced of and, starting with Wittgenstein, what led to the program of a behaviorist reformulation of intentional sentences. If, however, the sentences of the Geisteswissenschaften are supposed to convey a "re-understanding" of any thinkable meaning of sentences (as it is assumed by the transcendental philosophy behind the German Geisteswissenschaften), then we also have to admitfollowing Vico, Hegel, and Dilthey that in any sentence of the empirical Geisteswissenschaften the subject deals with itself in the final analysis and not with another one, foreign to itself. Every

P H I L O S O P H Y O F L A N G U A G E & GE1STESWISSENSCHAFTEN

important achievement of understanding in the Geisteswissenschaften seems to prove the partial truth, at least, of this conception by its effect on the practical shaping of history and thus also on the person who achieved this understanding. (In understanding one of Goethe's sentences we understand ourselves, i.e., especially, the language which we have in common with Goethe and the possibilities of understanding the world which are embedded in that language.) The well-known idea of the "hermeneutic circle"which means that we must have always understood in order to understand and that we nevertheless can correct this "pre-understanding" by methodical attempts to understandalso presupposes for this kind of understanding that, to use a word of Hegel's, the mind in dealing with the other is by itself.
14

Not a dialectical but rather a paradoxical formulation of this insight of transcendental philosophy is Wittgenstein's radical conclusion from Russell's theory of types: In the sentences about the meaning of sentences, that is, about language, the subject of language, according to Wittgenstein as well, deals with itself; and for this reason philosophy and Geisteswissenschaft are impossible, for in the final analysis both of them deal not with facts in the world but with language as the condition for facts to have meaning.
15

From this standpoint the deeper meaning of the following sentence, already quoted above, becomes more apparent: (5.5421) This shows too that there is no such thing as the soulthe subject, etc.as it is conceived in the superficial psychology of the present day. . . .

The logical positivists, in literal agreement with Wittgenstein, drew from this the conclusion: "There is no such thing as the subject that thinks or entertains ideas," but they were right, because the subject, according to Wittgenstein, does not belong to the world but is "a limit of the world" (5.632). Or, as Wittgenstein proceeds to say: (5.641) Thus there really is a sense in which philosophy can talk about the self in a nonpsychological way. What brings the self into philosophy is the fact that "the world is my world". . . .

But, according to Wittgenstein, (5.62) . . . The world is my world: This is manifest in the fact that the limits of language (of that language which alone I understand) mean the limits of m y world.

Compare also 5.6. According to Wittgenstein, behind the apparent form of the sentences of ordinary language and their apparent subjects there is the one universal form of language as a depicting language and the one subject of this language, which is the limit of the world: Only this radical turn, considering the subjective as transcendental, makes it understandable why Wittgenstein identified the logical form of intentional sentences with the logical form of semantical sentences. The

10

P H I L O S O P H Y O F L A N G U A G E & GEISTESW1SSENSCHAFTEN

decision about sense and nonsense is not left to the judgment of the empirical subject, but rather to the judgment of the transcendental subject of (the ideal) language, which "shows itself" in the logical form of the sentences. (This way we can understand how the early Wittgenstein could hold metaphysical sentences to be nonsensical although they were certainly intended to be meaningful by their empirical authors.) Of course, the concrete hermeneutical problem of understanding is carried ad a b s u r d u m by this undialectical transcendental philosophy; for all human subjects patticipating in the transcendental subject's one pure language would this way already formally be in perfect communication. Assuming this one transcendental form of language, "understanding" can only refer to particular information about facts and no more to the intentions of particular individuals as key to the very form of possible understanding the world. For Wittgenstein normal "understanding" is therefore: "Knowing what is the case i f . . . " (i.e., if certain information is true). A n d philosophical "understanding of language" is to show the transcendental form of the depiction of the world which is presupposed in all empirical understanding of information. Therefore, in the work of the early Wittgenstein, the place of a hermeneutics of individual intentions of meanings is taken by a logical analysis of language, which has to show the identity, guaranteed by the transcendental form of language, of "your" and "my," and "their" world and the world whose description is of general validity. This interpretation is supported explicitly by sentence (5.64) Here it can be seen that solipsism, when its implications are followed out strictly, coincides with pure realism. The self of solipsism shrinks to a point without extension, and there remains the reality of coordinated with it.

But this, according to Wittgenstein, "cannot be said, but it shows itself" (5.62). This paradoxical transcendental philosophy, a "critique of pure language" as Stenius calls it rightly, discredited the language of critique and determined the inner discord between methods and methodology in the further development of neopositivist analytic philosophy. Adhering on the one hand to an objectivistphysicalist methodology, as was suggested by Wittgenstein's theory of the one extensional language for all sciences, this philosophy could not on the other hand reflect philosophically on its own method since this would have been nonsensical metaphysics, according to Wittgenstein.
16

In fact, no analytic philosophy of language can with clear conscience reflect upon its own methods and thus upon problems of that area where, according to Wittgenstein's interpretation of intentional sentences, the Geisteswissenschaften should be located. This holds true as long as the concept of "meaningful language" is limited to a model of descriptive language as it was developed in Russell's Principia M a t h e m a t i c a . But this language model remained generally accepted, as we shall see, also during the second period of analytic philosophy when the methodology of science most widely held today was formulated. It
17

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

1 1

was only in the third period of analytic philosophy, which started with Wittgenstein's lectures in Cambridge around 1932, that this language model was abandoned in favor of a description of the infinite variety of actually functioning "language games." But even then, little immediate appreciation was accorded those language games in which other language games are described (i.e., interpreted) ot in which the essence of a language game is determined; recognition of the natute of the Geisteswissenschaften and philosophy itself was insufficient. 1 shall have to come back to the reasons for this later on. We shall now turn to that neopositivist period of analytic philosophy in which the problem of the Geisteswissenschaften was discussed explicitly only in the framework of the objectivist methodology, that is, the so-called unified science program.

THE

C O N S E Q U E N C E S O F THIS A M B I V A L E N C E FOR T H E M E T H O D O L O G Y O F S C I E N C E O F L O G I C A L POSITIVISM

E. Husserl says in his C a r t e s i a n i s c h e Meditationen about the way in which other subjects are given to me: 1 experience othets as actually existing and, on the one hand, as world objectsnot as mere physical things belonging to Nature, though indeed as such things in respect of one side of them. They are in fact experienced also as g o v e r n i n g psychically in their respective natural organisms. Thus peculiarly involved with animate organisms, as "psychophysical" Objects, they are " i n " the world. O n the other hand, I experience them at the same time as subjects for this wotld, as expetieneing it (this same world that 1 experience) and, in so doing, experiencing me too, even as I experience the wotld and others in i t .
18

This phenomenological sketch shows, it seems, which possibilities there are in principle for a science of man. It is similar enough to the ideas of the young Wittgenstein, starting as it does from a transcendental subject, that we can use it as comparison to the neopositivists' treatment of the problems of the Geisteswissenschaften. When proceeding from Wittgenstein's treatment of the belief-sentences, we should be inclined to demand that a genuine Geisteswissenschaft be constituted on the basis of that kind of experience mentioned above by Husserl to which 1 and the others, experiencing each other, also experience the same world. Such a Geisteswissenschaft would deal with other human beings, not as objects of meaning and language, not as objects in the world, but rather as those beings who "mean" together with me as partners of communication; in othet words: This Geisteswissenschaft would be constituted on the level of intersubjectivity. Its purpose would be, for example, to reestablish communication between subjects in case this communication had broken down, or to initially establish such communication between different subjects. Seen thus, the interpreter and the translator would be prototypes of a man of

12

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

the Geisteswissenschaften, and Petrarch, the inauguratot of the studia h u m a n i t a as after the Middle Ages, would be the man who set the theme of the Geisteswissenschaften when, in his fictitious letters, he anticipated the possibility of discourse between the august minds of all times and nations, to be realized only later by the Geisteswissenschaften.
19

At this point, however, we must inquire: Can this art of understanding, which does not regard the human being as an object of research but instead assures the intersubjectivity of meaning, rightly be called a W i s s e n s c h a f t a science? (In the Middle Ages, grammar, rhetoric, and dialectics were called the artes s e r m o n i c a l e s , and the first two were the forerunners of present-day humanistic hermeneutics.) Does not science presuppose the intersubjectivity of language as the very condition for the possibility of its own sentences? A n d is not this presupposition to be understood as implying preciseness in the description of possible facts, thus guaranteeing the reproducibility of experienceand not just some more or less well-functioning communication? Do not we have to presuppose for the protocol sentences as a basis of experience for general theories that the linguistic interpretation of the world is undisputed, that is, that intersubjective agreement has been reached about what is to count as experience?
20

The neopositivist program of unified science is based on exactly this presupposition, just as the T r a c t a t u s is; namely, that there is a world of facts which can be described unequivocally; and this presupposition explains why the logical positivists never had any doubts about the inclusion of the sciences of man and his culture, that is, the social and behavioral sciences, in their program of the united science. The last remark is not to be understood as implying that the logical positivists were not interested in the clarification of the conditions of possible intetsubjective communication. Quite the contrarytheir greatest lasting achievements are probably in the field of constructive semantics, that is, in the construction of formal languages which can be interpreted as precise fotmalizations of scientific theoties. But the construction of these "frameworks of language" (Camap), of these quasi-ontological category systems (only within which it is possible to distinguish between logically necessary propositions and factual propositions), is not considered as theoretical science but as a kind of practical work which admits of no further justification.
21

Philosophers, according to Camap, are designers of languages which will or will not stand the test of practical applicability. By turning the theoretical problem with sentences about the meaning of sentences into the practical problem of constructing semantical systems, Carnap avoids the type-theoretical difficulties of philosophical universal sentences about all sentences, that is, about language in general and its relation to the world, which had led Wittgenstein to his paradoxical conclusions. O n the other hand, in doing this Carnap renders impossible his own philosophical reflection on how every semantical system, successfully interpreted, depends on the language of science as it has developed in the course of history and as it is in use now, and with whose help

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

13

the artificially constructed framework could be shown to be a legitimate language of science. T o be more precise, the semantical system only has to comply with the science in question through its "rules of correspondence," "correlating definitions," etc., and by fulfilling the "conditions of adequacy"; Carnap replaces reflection of the above-mentioned dependency by logical reconstruction of parts of the language of science in use; but the presuppositions, implied in the meanings of the fundamental concepts of this science, remain rationally unclarifiable. According to the pragmatic justification of constructive semantics, they can only be accepted or rejected, and their acceptance constitutes a certain categorical "net" for a possible description of the world (Wittgenstein, Popper). This way, however, the complementary relationship between the two endeavors of logical and hermeneutic clarification of meanings is not reflected upon; and it is not recognized, or at least not acknowledged, that this complementary relationship between logical semantics and historical hermeneutics (i.e., the history of philosophy and of science as well as the history of literature, of language, and of social institutions) is an instance of the hermeneutic circle, the latter being fundamental to all Geisteswissenschaften: Man has always tried to unearth the meaning of strange language documents by constructing schemes of interpretation; this is not different from what is done in constructive semantics where everyday language is consciously "estranged" so that it becomes unclear and in need of interpretation. Even the "recoil of the text," that is, the retroactive effect of the object language already in use correcting the scheme of interpretation, can be found in the work of constructive semantics, though in this case the semantical system as a whole is tested and possibly corrected, while in the Geisteswissenschaften the conceptual framework is gradually changed as interpretation proceeds. But some of the empirical Geisteswissenschaften come closer to the former kind of self-correction, because they have laid down more explicitly their methodological conceptual presuppositions (e.g., Max Weber's "ideal types" as hypothetical standards of sociological understanding). Compared with this, Carnap's semantical system can certainly be taken as ideal types for the understanding of language and thus also of the categorical structure of the world. The logical positivists do not see the close connection of their methods and those of the Geisteswissenschaften (both trying to assure the intersubjective understanding between human beings), because they have made their methodological idea of science dependent on the assumption of one objectivist language which has already been constructed. Instead of understanding the function of the Geisteswissenschaften by reflecting on his own methods for the clarification of meanings, the logical positivist prefers to think that the empirical control of his own constructive language analysis should be left to an objectivist description of verbal behavior. In this manner R. Carnap, in his article " O n Belief Sentences," tried in 1954 to solve the problem of an empiricist foundation of the Geisteswissenschaften (after several previous unsuccessful attempts). He claimed that a sentence
22

14

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

like "John believes that the earth is round" is a theoretical construction which can be inferred (though, of course, only with some degree of probability) from a description of John's behavior, for instance, from the sentence "John makes an affirmative response to 'the earth is round' as an English sentence." For the evaluation of this reduction everything will depend on how we interpret the expression "makes an affirmative response." Normally we would understand this as an intentional expression, so that it does not describe a separate objective fact but behavior which is itself part of language and which implies an understanding of the sentence "the earth is round." In this case we would have to assume an understanding of the intentional direction of the affirmative reaction on the part of the person acting as recorder as well; in other words, the objectivist reduction would have failed and this failure could only teach us that human behavior is itself part of language and has to be u n d e r s t o o d as such.
23

Carnap, however, would like the expression "affirmative reaction" to be so understood that it does not imply that John understands English or any other language. If this stipulation were to be taken seriously, we would be justified as Hans Skjervheim rightly objects in admitting the following sentence as an example of a protocol sentence: "John coughs in answer to 'the earth is round' as an English sentence." But in this case it would be impossible to infer anything from the protocol sentence with any degree of probability about John's opinions.
24 25

For the man involved in the actual work of the empirical Geisteswissenschaften, these radical attempts to reformulate his experience in a behaviorist thing-language are usually less interesting than the other attempts to view his methods as a whole in analogy to the objective natural sciences. Especially when he faces the question of the final goal of his knowledge, the suggestive power of the concept of "causal explanation" according to laws is often very strong; a hierarchy of the various disciplines suggests itself according to how close they approach this ideal of science, that is, how "scientific" they are. From this point of view, the philologies are on the lowest level because as auxiliary sciences they simply collect and prepare the material needed by the historian to inform him about singular facts of human behavior; and only the historian is the real empirical researcher who describes the facts of the world of human civilization. But just as the old descriptive natural sciences, once their task had been fulfilled, were slowly replaced by the "explaining sciences," the final purpose of historical research, according to this view, would be its integration into a sociology which deduces and perhaps even predicts particular cases of human behavior from general laws. A theoretical foundation for this project of making the cultural and social sciences more and more scientific was first attempted by the logical positivists in the publications of their journal E r k e n n t n i s (1930-38), which was continued in the United States under the name of ] o u m c d of Unified Science (1939); furthermore in the International E n c y c l o p e d i a of Unified Science (after 1938), and lately

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

15

again in the Minnesota Studies in the Philosophy of Science. In this connection the "theory of scientific explanation" by Carl G . Hempel and Paul Oppenheim is of prime importance. Let us examine this theory more closely.
26 27

A n "explanation" is supposed to answer the question: Why is or was this or that the case? The explaining sciences just take over what is or was the case (i.e., the "explanandum") from the descriptive sciences. Therefore, "explanandum" is to be understood as "the sentence describing the phenomenon to be explained (not that phenomenon itself). " The explanation on the other hand, that is, the "explanans," also consists of sentences and, according to Hempel and Oppenheim, necessarily of two classes of sentences: "One of these contains
2 8

certain sentences C

C , which state specific antecedent conditions;


k 1 ( 2 r 29

the other is a set of sentences L L , . . . , L , which represent general laws." By making this distinction, the authors hope to do justice to the scientific "why-question," which always asks for two different kinds of answers: "According to which general laws and by virtue of what antecedent conditions does the phenomenon occur?" According to the view stated so far, an explanation is supposed to be the logical deduction of a sentence (the explanandum) from other sentences (the explanans). As we shall see in the following, this formulation in terms of sentence analysis is characteristically different from the theory of explanation of the old positivism. This formulation shows for the first time a presupposition of the theory of explanation which allows us to see the connection between this theory and the theoretical problems of understanding in the Geisteswissenschaften. But before we elucidate this connection of actual m e t h o d s , we shall first look at the relation between the neopositivist theory and the theory of understanding of the Geisteswissenschaften, as it appears from the point of view of the explicitly stated methodology of the former philosophy: Hempel and Oppenheim try to show that their theory of explanation also applies in principle to the "nonphysical sciences."
30 31

Therefore the two authors object to the idea, for example, that in the realm of so-called purposive behavior, teleological analysis should replace causal analysis. They claim that this idea is based on an erroneous view of what "motives" are, because motives should not be understood as if a goal, still lying in the future, were determining our present actions; the reason being that this goal might possibly never be reached. Therefore, instead of this as yet unattained goal, the present desire to reach a certain goal should be taken as the motive of the action. This desire, however, as well as the belief, also present before the action, that a certain course will in all probability lead to the desired effect, these "determining motives and beliefs . . . have to be classified among the antecedent conditions of a motivational explanation, and there is no formal difference on this account between motivational and causal explanation. " What strikes somebody accustomed to research in the Geisteswissenschaften is the tacit assumption in this argumentation that the aim of knowledge in the realm of purposive behavior can be nothing but causal explanation just as in the
3 2

16

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

natural sciences. Starting from this assumptionwhich, from the point of view of the Geisteswissenschaften really begs the questionit is not difficult to discover the fallacy in regarding motives as "goals lying in the future." Now 1 certainly do not want to dispute that it is in fact faulty thinking in historical or biological speculation to assume a future goal to be the efficient cause of real events (perhaps even in the sense of a metaphysical hypostasis). In the Geisteswissenschaften, however, which try to understand human actions from their motives, we encounter an altogether different aim of knowledge, and not a residue of metaphysics.
33

To see this difference, we must bear in mind that even in instances of unrealized goals it will still be of interest to us to understand this setting of a goal as the creation of a future possibility for human beings in a particular situation. This specific interest of the Geisteswissenschaftenas opposed to sole interest in the causal explanation of actions which actually occurredmay ultimately be based on the assumption that the goals set up by past generations, insofar as they can be understood, continue to exist as possibilities of human actions; that is, they can be realized by those who understand them or they can be transmitted by them to the following generations as motives for possible future actions. Such a "community of interpretation" in the tradition of meanings is the common ground for all attempts in the Geisteswissenschaften to attain knowledge; and natural scientists as members of a "community of investigators" (C. S. Peirce) are thus, according to J. Royce, also members of that community of interpretation in respect to their motives for research. A n d only when these motives have already been u n d e r s t o o d can they be used as psychological antecedent conditions in a causal explanation.
34

The historical "sciences," which try to understand the purposive behavior of past generations, are really not the most typical kind of Geisteswissenschaft; because in this field the method of objectifying actions as events to be causally explainedthus approaching the methods of natural sciencecan still rather easily take root. However, when man acts he is also part of the community of interpretation and he will therefore try to illuminate his motives of action in literary works. The interpretation of these documents is the theme of the "real"that is, the hermeneuticGeisteswissenschaften: not in order to discover documents for the reconstruction of past events, but rather in order to understand motives and ideas for their own sake, that is, in order to enrich present and future life with these ideas or meanings. These disciplinesfor example, the philologiesare simply ignored in the methodology of neopositivism. The reason for this, however, lies partly in the fact that in the AngloSaxon countries these disciplines are still seen as they were before the age of modern scienceas part of the artes liberates, that is, especially of rhetorics and literary criticismwhile on the other hand the concept of "science" was limited to fit the methodological ideal of the natural sciences.
35 36

The assumption that causal or statistical explanation of objective events by means of general laws is the only conceivable purpose of scientific knowledge was never doubted in the methodology of neopositivism, not even when the

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

17

possibility of an understanding of human behavior was explicitly discussed. In such cases it only appears at first glance as if understanding as a method of knowledge was compared to e x p l a n a t i o n . In fact, instead of comparing the ultimate possibilities and purposes of the two methods, understanding was evaluated right from the start on the basis of how much it could contribute to the objective explanation of facts; and thus it was no wonder that it was found to be only of heuristic value as a "prescientific" method.
37

A teleological explanation tends to make us feel that we really "understand" the phenomenon in question, because it is accounted for in terms of purposes, with which we are familiar from our own experience of purposive behavior. . . . This understanding . . . in terms of one's own psychological functioning may prove a useful heuristic device in the search for general psychological principles which might provide a theoretical explanation: but the existence of empathy on the part of the scientist is neither a necessary nor a sufficient condition for the explanation, or the scientific understanding [sic!], of any human action.
38

"Understanding" is taken here as "empathy"; and this interpretation is actually in accordance with a certain psychologizing tendency also to be found in some of the works of the founders of modern German Geisteswissenschaften, namely from Herder to Schleiermacher to Dilthey. In face of this explicit (methodological) point of viewthat is, that "understanding" is "empathy" and "empathy" is neither a necessary nor sufficient condition for scientific explanation, though it may be a useful heuristic device1 think it appropriate to once again reconsider the implicit methodic relation of the neopositivist theory of explanation to "understanding." This can best be done by reviewing the arguments on which Hempel and Oppenheim base their explicit evaluation of "understanding." According to this view, "understanding" is, first of all, not necessary,

for the behavior of psychotics or of people belonging to a culture very different from that of the scientist may sometimes be explainable and predictable in terms of general principles even though the scientist who establishes ot applies those principles may not be able to understand his subjects empathetically.
39

If we admit, for the sake of argument, the possibility of an objective explanation or even prediction of human behavior in the above-mentioned cases, we still have to ask whether such a possibility speaks for or against the necessity of understanding human behavior. Could not we argue conversely: Only when we are dealing with psychotics or with people of a very strange culture do we get the ideal of doing without an immediate understanding of their motives and try instead to explain their behavior like an event in nature? But in so doing we have abandoned not only any attempt at empathy, but, more important, any attempt to "get into a conversation" with these human beings. In contrast to this quasi-scientific extreme case, the necessity of understanding human motives could well be demonstrated by the case of, say, an English

18

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

psychiatrist trying to understand his French colleague or an American ethnologist, his German colleague. They can do this by going through the writings of these men who may have died long ago; and the purpose of their research will be to understand the viewpoints and the kinds of questions taised by their colleagues. Here it is appropriate to call to mind a point made above, namely, as the logical positivists themselves ate always eager to emphasize, that the logical operations of science are not applied to the phenomena as such, but to sentences which describe these phenomena as something and in a certain way. Befote any explanation is offered science has to "understand" the so-called basic sentences as protocols of facts a n d as human interpretations of these facts. Foi every newly discovered fact must be taken as an interpretation only tacitly agreed upon to be of general validity for the institutionalized "community of interpreters"; this shows that the community of investigators in the natural sciences is also a relevant phenomenon for the Geisteswissenschaften. In this field, understanding is not primarily required as a kind of psychological empathy, but is rather a necessary condition for participation in the intersubjective exchange of ideas. Insofar as such an exchange is necessarywhich has to be admitted, at least in the case of the community of scientistsit cannot be replaced by objective methods for the explanation of behavior. Objective explanation of facts and intersubjective communication about what is to be explained are instead "complementary" aspects of human knowledge in the sense in which N . Bohr used the word. They exclude each other and they presuppose each other. Nobody can just "understand" without presupposing factual knowledge which could be stated explicitly as "explanations." O n the other hand, no natural scientist can explain anything without participating in the intersubjective communication described above. Now it cannot be denied that even in the Geisteswissenschaftenwhich have cultivated the business of intersubjective understanding of different points of view into a methodthe need is felt at times to "estrange" human behavior for a while so that it becomes available as an object of explanation. Thus we may want to explain not only psychotics and people of exotic cultures, but also the classical texts of Western theology and metaphysics by unmasking their true unconscious motives which were no part of the author's self-understanding and therefore cannot be understood. Especially in political history, the method of understanding the motives of actions of the state is often so little satisfying in regard to an undetstanding of the totality of actions involved that the very need of deeper understanding of the true motives will call for a psychological or sociological explanation in terms of causes or statistical laws of human behavior.
40

Now we can appreciate the second argument of Hempel and Oppenheim against "understanding": namely, that understanding is not sufficient, "for a strong feeling of empathy may exist even in cases we completely misjudge a given personality." In such a case of insufficient or even erroneous understanding the question must be raised, however, of what an objective explanation of this behavior can
41

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

19

possibly achieve. For the neopositivists the decisive ctitetion of a scientifically useful explanation is its empirical testability. But how we can test the truth of an explanation of human behaviot which considets motives as causes? How do we determine whethet cettain people have ot had certain motives? Hempel and Oppenheim touch on this problem in a sttangely obscute passage: The ptesence of cettain motivations may be ascertainable only by indirect methods, which may include reference to linguistic uttetances of the subject in question, slips of the pen ot of the tongue, etc., but as long as these methods ate "operationally detetmined" with reasonable clatity and precision, thete is no essential difference in this respect between motivational explanation and causal explanation in Physics.
42

O n a first reading of this passage we might interpret it as saying that also the linguistic utterances of the persons in question are actually playing an important role in ascertaining motives "by indirect methods." If this were indeed what the two authors intended to say, we could only agree with them, as the most immediate possibility to explore someone's motives is to draw him into a conversation (be it the first questioning of patient by doctot or the examination by a judge or a public opinion interview). A n d it should be well noted that even if we want to check our understanding obtained by psychological empathy, the first way of doing this will again be conversation. In this instance, however, we would not have checked our understanding against an empirically tested objective explanation, but we would have simply corrected understanding with better understanding. It would be rather odd to call this understanding via language an "indirect method" of ascertaining motivesunless one wants to assume the existence of a psychological empathy without language as a direct method of understanding.
43

Hempel and Oppenheim, however, talking about "indirect methods" to ascertain motives, cannot mean our normal way of understanding language, because they want to show, after all, that there is no difference in principle between motivational explanation (together with its empirical testing) and causal explanation in physics. "Linguistic uttetances" are supposed to be phenomena like the Freudian "slips of the pen or of the tongue," that is (ftom the point of view of the neopositivist theory), objective events in nature, which can be interpreted as symptoms for hidden motives. This objective method of testing knowledge about human motives, obtained by intetsubjective communication, is in fact what is done in psychoanalysis (andmutatis mutandisin critique of ideology). What seems to be impottant about these methods is that the immediate contact of intersubjective communication must be disrupted first before they are applicable, and the other person has to be placed in the position of an object. (This may happen in the midst of a convetsation, and to some degree it will happen in every conversation.) The question is, however, whether these objective methods to discover

20

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

motives or to test a motivational hypothesis are correctly interpreted philosophically if we immediately take them as parts of a scientific explanation of human behavior, that is, as methods designed to replace "understanding" and to finally enable us to develop a verifiable theory for the prediction of human behavior. For our purposes I think we can state the following condition as criterion for an objective scientific explanation as the logical positivists see it: The theoretical as well as the observational statements of the explanation are independent of the object to be explained insofar as they do not presuppose communication of the "explainer" with the "object" by means of a common language (in case of motivational explanation), which the theory of understanding holds to be essential. Therefore we must ask: Can we actually find such a relationship towatds the object in the case of psychoanalysis or of "critique of ideology"? It cannot be denied that in these cases the contact of linguistic communication has to some degree been "loosened" and that, as a result, the people to be "cured" have become, to the same degree, scientific objects. However, we should not forget that the interpretation of, say, Freudian slips as symptoms for hidden motives is only possible because they can be recognized as "slips" solely in the context of linguistic utterances which have to be understood. T o that extent, therefore, the psychoanalyst has to remain in intersubjective communication with his "object." A n d insofar as he incorporates the linguistic slips as symptoms into the conceptual framework of a theory (e.g., the theory of the Oedipus-complex), it is still not clear whether he is then "explaining" or only "understanding" more deeply. A t any rate, we must remember that the psychoanalyst interprets even such kinds of behavior as meaningful, as expressions of unconscious motives, which would be considered meaningless by a man of the Geisteswissenschaftenfor example, a "tick" or forgetfulness in certain matters, or even physical symptoms, which undoubtedly would be left to physiological explanation by the Geisteswissenschaften. Now it could be said that verification in terms of intersubjective protocol sentences, independent of what the object thinks of himself, is the decisive criterion for an objective scientific explanation. For psychoanalysis this would mean the following. The protocol sentences describing, for example, the disappearance of certain physical symptoms of an illness could be taken in support of the neopositivist thesis as "empirical verification" of the explanation given in the process of psychoanalysis. The following objections, however, can be raised against such an interpretation: The objectively reported success of psychoanalytic treatment can hardly be called a logical consequence (as prognosis) of a scientific explanation, for this result could also be brought aboutdue to the patient's reactive cooperation and contemplation about himselfeven if certain motivational hypotheses of the psychoanalyst were not correct. Therefore it would be much more appropriate to speak of "empirical verification" if the patient not only recovered but also better understood his former behavior in the light of psychoanalytic hypotheses: This "better understanding" would imply that he

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

21

now could see through the illusions of his former goals and could understand how he had missed what was teally impottant fot him in life; in other words, the patient would teachwith the help of psychoanalysisa deepet understanding of his motives as possibilities of his human existence. In this case, however, the patientthe object of analysisfinally meets again with the thetapist in one intersubjective language fot the communication of possibilities of human existence: because the theory is now confirmed by the comments and reports of the "analyzed object," and not by protocol sentences of the psychoanalyst. Therefore the "explanation" of his behaviot was not the first step towards a science of human behavior, but only a methodological device of "estrangement," a quasi-objectification of understandable meaning, which was justified to the degree in which the patient had not yet achieved an undemanding of himself. T o summatize: "Explanation" was setving "undemanding." This would not hold true in case the explaining psychologist wete to allow his theories of motivations to be used for manipulation of human behaviotfot example, for economic or political purposes. In this event, explanation and the prognosis of behavior in fact become autonomous; but then the "understanding of meaning" which, as we saw, is an irreplaceable complementary approach, is set free and becomes autonomous too on the side of the manipulating people; because men whose behavior was completely explainable would not know what to do with these explanations and prognoses.
44

Obviously, these considerations can easily be applied to the relationship between explanatory sociology and "understanding." T o the models of psychoanalysis given here corresponds the model of critique of ideologies. Here, too, human self-understanding is undoubtedly corrected by objective methods. But the result of these latter methodsfor example, the uncovering of economic interestscan always be incorporated in principle into a more profound understanding of oneself. A n d in regard to the above-mentioned criterion of an objective explanation we can say this much: The sociologist, just like the psychoanalyst, cannot completely separate his own descriptive object-language from the language of his "objects," who are his co-subjects. (For this reason the assumption that human attitudes are ideological can never be "total," because in that case it either loses its function as critiqueas it does in fact in the works of K. Mannheimor it becomes applicable to the language of sociology itself, thus invalidating its own truth-claim.) But let us more thoroughly consider the relationship between "understanding" and "explaining" in sociology. For this purpose we shall examine some of the examples given by T . Abel in his article "The Operation called 'Verstehen,'" which is representative of the neopositivist methodology. According to Abel, "understanding" is a kind of internalizing of observed human behavior by identifying it with one's own personal experiences. This way a logical connection between the observed facts is established by means of some sort of emotional syllogism. Abel's concept of "internalizing a situation" embraces all the characteristic fundamental postulates of the theory of under45

22

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

standing since Schleiermacher: sich-einfuhlen (empathy), s i c h - h i n e i n v e r s e t z e n (imagining oneself in the situation of another), n a c h v o l l z i e h e n (reliving of another person's experiences), and so on. Abel admits that this method can lead to an experience of "evidence" on the part of the person trying to understand, that is, to the feeling that the observed behavior becomes understandable. But this experience, according to Abel, is not mote than a feeling (which depends on one's own actual former experiences) of a possible connection between the events and actions in question. Its logical status is one of an ad hoc explanatory hypothesis depending on a general rule of behavior, which was found inductively by observing one's own inner experiences. From this analysis of "The Operation called 'Verstehen'" Abel concludes that "understanding" cannot be accepted as an instrument of scientific analysis, first, because it depends on the subjective capacity for and variety of experiences of the person trying to understand, and, secondly, because it is not a verifiable method: From the point of view of V e r s t e h e n alone any connection that is possible is equally certain.
46

Therefore in any given case the test of the actual probability calls for the application of objective methods of observation; e.g., experiments, comparative studies, statistical operations of mass data, etc.
47

Abel illustrates his analysis with the example of a correlation between a crop failure and the number of weddings in an agricultural area. In this case, by understanding the feats and worries of the farmers, the connection between the bad crop and the behavior reaction can apparently become evident. But actually only a hypothesis has been made, which can now be tested by objective methods of the social sciences; for the "understandable" general rule of conduct (decrease in income makes people reluctant to assume responsibilities) may now be confirmed but also disproven by subsequent investigations. Perhaps the average behavior of the farmers can be explained in terms of a law which does not correspond to any understandable maxim of human behavior. This may leave the scientist feeling vaguely dissatisfied, as Abel admits, but it does not limit the validity of the scientific explanation. Abel's remarks about the scientific explanation of "incomprehensible" behavior cast light on the fundamental difference in goals of knowledge between science and Geisteswissenschafta difference normally not reflected upon, though it will always bias any discussion about these two approaches. We shall return to this fundamental difference between the two theories. Let us first have a look at the interesting critique of Abel's analysis by the Norwegian, Hans Skjervheim, already quoted above, from the point of view of the "school of understanding" ( " V e r s t e h e n s s c h u l e " ) . Citing in his support M . Weber, T . Parsons, and W . J. Thomas, Skjervheim
48

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

23

shows that Abel has cut off some of the problems of undetstanding in the social sciences from the statt by limiting himself to the "why" questionthe same question the "objective explanation" tries to answer. But actually the social scientist, accotding to Skjervheim, needs "understanding" prior to answering the question of what is to be consideted as a "behavior fact." Abel and most behaviotists, howevet, naively considet the facts to be connectedthat is, the stimulating factors and the behaviof-teactionsas "objectively given," while Skjervheim points out that these objective presuppositions of understanding (or explaining) of motives (or causes) themselves can only be determined by an understanding which is "subjective" (in the wider sense of the word): A more careful analysis shows that these facts obtain theit specific meaning only by being facts for a specific person acting in a specific situationot by being facts fot the teseatcher himself. Now it is important to tealize that thete is no difference in principle between the "subjective constitutions of meaning" in teference to the petsons under observation or to the teseatchet himself. (Fot the positivist, howevet, this fundamental difference between "private ways of understanding the wotld" and facts, described in terms of the intetsubjectivist language of science, does, of course, exist.) In fact, a possible communication between the concepts of the persons whose actions ate investigated and the researcher must be ptesupposed. The social scientist could not talk about "suicide rate," "marriage rate," "frequency of divorce," "election tesults," and so on if the people undet consideration could notin principleunderstand their own behaviot in these terms. With regard to the above-mentioned example of a correlation between crop failute and dectease of marriages, Skjervheim asks Abel who has detetmined that the harvest was bad, a botanist, ot was it not tathet the farmets themselves; fot at any rate he will have had to ask the lattetin an intetview and hence again by way of actual undetstandingwhethet they consideted a certain event (which may be botanically definable) a "bad crop." By using M . Weber's concept of "actual understanding," Skjervheim finally manages to dispense with the main argument of the positivists, namely the need for objective methods of verification, or rather he "attacks it from behind." He first asks how the social scientist obtains the data he needs for the objective verification of his explanation (quoting T . Parsons in his own support). The analytically minded scientist speaks about observations, comparisons, and statistical investigations. Skjervheim now points out that all these procedutes to obtain "social facts" do already presuppose the actual undetstanding of meaning: Even statistical investigations depend on interviews or the reading of documents (e.g., of the registrar's offices).
49

If we watch this debate between Skjervheim and the logical positivists from a distance, we notice that the protagonists of "understanding" (i.e., of the Geisteswissenschaften) always attack the supporters of the theory of explanation (i.e., of the objective social or behavioral sciences) from behindand vice versa. The "objective scientists" point out that the results of "understanding" are

24

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

only of prescientific, subjectively heuristic validity, and that they at least must be tested and supplemented by objective analytic methods. The ptotagonists of understanding, on the other hand, insist that the obtaining of any data in the social sciencesand therefore also any objective testing of hypotheses presupposes "actual understanding" (Skjervheim: "observational understanding") of meaning. If we now make use of the considetations developed in our discussion of Hempel and Oppenheim's theory, we can say that both points of view are right, but that they do not do justice to the other side. The weakness of Skjervheim's arguments seems to be that he does not sufficiently acknowledge the importance of the possibility of using objective methods of explanation fot the purpose of checking human undetstanding of oneself and of othets. The impottance of statistical investigations, fot example, may be found in the fact that the results of these investigations can be used as objective models to check and correct understanding of situations and motivations. Should the effect of this double checking be invalidated by the fact that the statistical data themselves presuppose the possibility of understanding? A basic "undetstanding" must even be presupposed for the protocol sentences in physics, as I mentioned above. In most cases, howevet, this fact can be ignored fot ptactical purposes: A specific measurement will usually not for that reason fail to fulfil its function to confirm a theory of physics, for the investigatots may not fully agree upon the meanings of theit basic metrical concepts unless, in the case of Einstein, the problem of understanding these concepts becomes manifest in a so-called crisis of the foundations.
50

The example of such a "crisis"which can even raise the problem of understanding in regard to observational protocols in the exact natural sciencesdemonsttates, however, that the problems involved in obtaining data in the social sciences cannot be totally neglected. For if we claimed, as the neopositivists actually do, that the data of the social sciences must also be facts which can be desctibed unambiguously and objectively, one could rightly say that the social sciences are in a permanent "crisis." Now this difficulty of obtaining objective facts in the social sciences was more adequately analyzed by Skjervheim, who pointed out the necessity of linguistic communications (in Royce's term we could say: "of a community of interpretation") between the subject and the so-called object of the social sciences. However, this emphasis on the hermeneutic problems involved in obtaining facts in the social sciencesas justified as it isshould still not let us forget that "objective methods of verification" do fulfil a controlling function when they are contrasted with the results of understanding: A n d for all practical purposes, the importance of this function of the objective methods is not much diminished by the fact that the individual protocol sentences in turn presuppose understanding. But the same atgumentation from the point of view of the ptacticing scientist can also be said to support the "understanding" of the Geisteswissenschaften.

PHILOSOPHY O F L A N G U A G E & GEISTESWISSENSCHAFTEN

25

Just as the possibilities of objectivizing social phenomena by means of statistical observations should not be forgotten because of the understanding implied in any particular observation, so we should not demand that any understanding of meaning be incorporated into the methods of objective science. This obviously makes the least sense in the case of understanding of written or spoken language. Here also the champion of the objective sciences has to admitas a member of the human community of interpretation, which is complementary, as we said above, to the objectively given factsthat the meaning of the signs cannot be reduced to anything else. (For example, he cannot take the protocol sentences of his colleague as "verbal behavior," because he would need other protocol sentences to desctibe it, and so on ad infinitum.) But the protagonist of the objective sciences must not only admit that "understanding" is an irreducible presupposition for his being a member in the community of interptetation of the sciences; he also cannot seriously deny that this understanding of the world opens up a new dimension of scientific or theoretical problems, which cannot be solved by objective methods. The understanding of a strange language, for example, cannot be replaced by an exact detetmination of the frequency with which the different words are used in this language. Such statistical methods can, however, also be used in the framework of "understanding"as a means of promoting the latter via objective criteria so, for example, if we statistically investigate an unknown language in order to further our understanding of its grammatical structure; but this way we do not replace "understanding." For the knowledge of statistical frequencies with which the words actually appear does not even suffice for us to know that we are dealing with words of a language altogether.
51

Actually, the analytic philosophers had already recognized the "understanding of meaning" as an autonomous theoretical method when they postulated that the analysis of language was to replace the analysis of facts (Carnap's "formal mode of speech"). Thus regarded, it seems paradoxically that the analytic methodology has replaced the "explanation" of the older positivism by "understanding"; because, strictly speaking, it counts only the deduction of sentences from sentences as scientific analysis. It must be admitted, however, that this is only a borderline case of "understanding," which we could call, in the words of Rothacker, the "conceiving" ( " B e g r e i f e n " ) (of timelessly valid relations of meanings), thus distinguishing it from the "understanding" as well as from "explaining." Strictly speaking, the "conceivable" relations of meaning, that is, those relations understandable on purely logical grounds, make up only one dimension of language and, accordingly, of understanding: that dimension which can be projected into the signsyntax of a formalized language. But already in mathematics, and particularly in the factual sciences, the weight of the semantical dimension of language is felt. The logical "conceiving" is, so to speak, embedded in the "understanding" of the content, that is, of the meaning of the concepts and sentences. If it were trueas was assumed by the logical atomism of Russell and the
52 53 54

26

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

young Wittgensteinthat there is only one logical form of language to desciibe the wotld, and accordingly only one transcendental semantics for the language of science, the logical positivists could hope to solve the problem of linguistic undetstanding by consttucting the semantical system for the language of "unified science"completely independent of the language-hermeneutical work in the understanding Geisteswissenschaften (including the hetmeneutical history of science). But actually the logical positivists have long since given up that thought and now just demand that the object-language be intersubjectively verifiable. Carnap's "principle of tolerance," which is now generally accepted among positivists, implies that the interpretation of semantical systems, as logical reconsttuctions of language, is embedded in the established use of language by the vatious sciences as they have developed historically.
55

This change in the concept of language was explicitly acknowledged in the methodology of logical positivism when Carnap accepted C . Morris's idea of the pragmatic aspect of a sign as the third dimension of language beside syntax and semantics. In this connection, Morris, who was strongly influenced by traditional American pragmatism, always emphasized that the formal semantical systems to be interpreted as frameworks of the language of science were, in the last analysis, also founded in the ptagmatic dimension of the signs, while Carnap accepted this only reluctantly. This means in behaviorist terms that the sentences of the language of science do not describe facts "just as they ate," but rathet facts of the behavioral environment of the scientists, who interpret the sentences by using them. In the language of the Geisteswissenschaften we could say that the possibility that linguistic signs have meaning cannot be understood without presupposing a "meaning-intention" which exptesses itself in the signs. In other words, not even the facts of science are facts for the unchanging "subject as such" (of "the language as such"), but they are constituted in a concrete and therefore historically determined human horizon of meanings.
56

Thus changing the concept of language, the logical positivists seem to have given up the very thought which Wittgenstein upheld in the T r a c t a t u s when he reduced the logical form of intentional or belief-sentences to the logical form of semantical sentences. But this seems only logical if we temember that the (neo-Leibnizian) idea of the one logico-ontological universal language, the ttanscendental subject of which matks the "limit of the world," had been abandoned step by step in the second phase of analytic philosophy. If even in the language of science we cannot assume that unambiguous communication between the subjects using this language is guaranteed by the "transcendental subject" of that language (that is, Wittgenstein's "metaphysical subject," cf. T r a c t a t u s 5.641), then the investigation of meanings of linguistic symbols by analytic philosophets must eventually lead to hermeneutical attempts of understanding the meaning-intentions of the users of these symbols.
57

With this conjecture we move the problem of understanding in the Geisteswissenschaften into the horizon of the third phase of analytic philosophy, which abandoned the construction of logical frameworks for the language of science,

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

27

instead considering as its task the descriptive analysis of the use of everyday language.

THE OF

R E S O L U T I O N O F T H E A M B I V A L E N C E IN T H E THIRD P H A S E IN F A V O R A " M E T H O D O L O G I C A L INTERSUBIECTIVISM-AND T H E REMAINING U N D E R S T A N D I N G A N D OBIECTIVE E X P L A N A T I O N

P R O B L E M O F A DIALECTICAL M E D I A T I O N B E T W E E N INTERSUBIECTIVE

The third phase of analytic philosophy, which is considered by its advocates as the real revolution in philosophy, was decisively influenced by the thoughts of the latet Wittgenstein, which were first recorded by students from his lectures in 1933-35 in the so-called Blue a n d B r o w n B o o k s , published in 1958. A short perusal of these notes as well as the Philosophical Investigations, which were published posthumously in 1953, suffices to confirm our expectation that analytic philosophy, having forsaken the idea of a universal language, had to turn to the hermeneutics of meaning-intentions, that is, to the problems of traditional Geisteswissenschaften.
58

In fact, the problems of understanding the meaning of such expressions as "meaning," "believing," and "understanding" itself dominate most of the work of the later Wittgenstein, while he had dealt with them in the T r a c t a t u s in only a few apodictic sentences. Still the difference from the Tractatus is not as great as one might think at first. Language analysis remains Wittgenstein's method: This implies that the radical "antipsychologism" of the Tractatus (that is, the attempt to uncover the illegitimate hypostatizations in the language of "superficial psychology") is upheld in Wittgenstein's later philosophy; in fact he now tries to destroy that ontology of mental or psychic states and actions which G . Ryle, calling it the "paramechanical theory of the mind," attributed to Descattes. This tendency in Wittgenstein's later work makes it even harder than in the case of the Tractatus to see its connection with the traditional philosophy of the Geisteswissenschaften. For the metaphysics of the spirit and of the subject in nineteenth-century idealism, which should be considered the foundations of the Geisteswissenschaften (although the latter certainly put more emphasis on empirical research), are taken by the later Wittgenstein as a "disease" of language together with all other concepts of metaphysics in Western philosophy, whereas in the T r a c t a t u s he had still considered these concepts as "borderline cases" of a transcendental semantics.
59

Still, the point at which Wittgenstein's thinking meets with the modern philosophy of the Geisteswissenschaften can perhaps be shown if we try to understand their common goals, which also means that we understand the dialectical development of analytic philosophy in the course of history as 1 have tried to do in this study. We shall see, however, that the connection with our theme will only become obvious if we consider not only what Wittgenstein explicitly says about philosophy, but rather look at the method of language analysis he is actually using himself.

28

P H I L O S O P H Y OF L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

First of all Wittgenstein agrees on one negative point with the theory of understanding of the southwest German school of neo-Kantianism as well as with E. Husserl's analysis of intentionality: In countless examples he tries to show that concepts like "understanding," "meaning," "believing," and "expecting," do not denote mental states, feelings, experiences, or some other kind of mental processes in time. But now Wittgenstein, as opposed to Husserl, does not instead look for "mental acts" (to which the intentional expressions refer and which aim towards "meaning" as something of timeless validity), because for Wittgenstein this would mean replacing testable by untestable hypostatizations. What is important, according to Wittgenstein, is to realize that words like "understanding," "meaning," and "expecting" do not "denote" anything at all; they are not used like names, which can be used in factual descriptions to name something. The student of mathematics who suddenly exclaims, after some formula has been explained to him, "Now I understand," does not want to give some information to his teacher about the state of his own mind; he does not want to describe anything, he rather wants to indicate "Now I know how to go o n . " If someone says, "I expect that he will come tonight," he does not describe any mental state, as pethaps the person does who says, "I am waiting for him impatiently." A n d we cannot ask (to use an example of G . Ryle, which brings out Wittgenstein's point very well), "How long did you mean last night?," while it is possible to ask, "How long did you discuss the question last night?" Therefore "to mean" obviously does not describe any "activity," either physical or mental.
6 0 61

But what can we say positively about the meaning of the so-called intentional expressions, if we cannot even assume that they denote something? At this point we should recall the basic theme of the philosophy of language analysis as it was originated by Wittgenstein. In the T r a c t a t u s the function of intentional expressions like "to mean" was taken as something which could not itself be "meant," that is, which could not be "denoted," the function of these expressions was considered to be identical with the function of language in general, namely, to mean or denote something. The function of "to mean," according to the early Wittgenstein, "shows itself" in the function of language. In the Philosophical Investigations the solution of this problem is not so very differentas far as we can talk about "solutions of problems" in the philosophy of the later Wittgenstein. The difference, however, is that now the model of logical atomism no longer determines how the functioning of language is to be understoodand this means in general that the traditional model of language since Aristotle has been likewise abandoned: namely, the idea that wordsymbols "denote" objects, the former ones being elements of descriptive sentences, the latter ones of facts. Instead of this modelwhich was prejudiced, in the last analysis, by the concept of "theoria" as a philosophical state of m i n d the later Wittgenstein introduced as a new key concept, the "language game," or rather the "language games." These "language games" are distinguished from the earlier concept of the one

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

29

depictirig language of the "metaphysical" or "transcendental" subject, which marks the limit of the world, in the following respect: They are concrete unities of language usage, of a form of life and of a certain way to see the world, each one different but still related to the others. What a meaning-intention is "shows itself" in describing these language games, according to Wittgenstein, because meaning-intentions cannot be thought of in isolation from the use of language, that is, from a certain habitual behavior which is a form of life only insofar as it is a language game. The questionwhich ideas of the third phase of analytic philosophy could be made use of for a philosophy of the Geisteswissenschaften has to take the following into consideration: the above-mentioned integration of meaning-intentions a n d of the understanding of the observer himself into these language games. At this point I should make a comment on the methodical difficulties one encounters trying to interpret the philosophy of the later Wittgenstein appropriately. The later works do not form a system, but they are rather "a series of landscape sketches," as their author says himself in the introduction to the Philosophical Investigations. It is completely impossible within the scope of our brief historical treatment of analytic philosophy to fully exhaust and interpret in their last consequences all ideas and intentions hinted at in these philosophical remarks. Our aim therefore can only be to generally clarify the implications of the method of analysis of language games for a philosophy of the Geisteswissenschaften. In this position, we are fortunate enough to be able to refer to the work of a British author who took it upon himself a few years ago to carry out exactly that objective, because from the point of view of contemporary German philosophy the book by Peter Winch, T h e I d e a of a Social Science a n d Its Relation to Philosophy (London 1958), can be considered as an attempt to analyze and develop the consequences of Wittgenstein's theory of language games for the philosophical foundations of the Geisteswissenschaften. We must point out, however, that Winch's interpretation of Wittgenstein, which is inspired by Collingwood and Weber, strongly differs from the Wittgensteinian philosophy as pursued in Oxford and Cambridge today. We could call Winch's interpretation a thinking with Wittgenstein against Wittgenstein. Winch is thinking with Wittgenstein insofar as he is the first to call attention to some of the most important implications of the concept of a language game: He shows how, on the one hand, the "constitution" (or "identification") of objects is "interwoven" with rules, and how, on the other hand, rules are "interwoven" with the social forms of life, and finally, how the behavior, which is following a rule, is "interwoven" with the possibility in principle to reflect upon these rules. Winch explains, using Wittgenstein's ideas, how understanding functions in this system of interwoven presuppositions of a language game. But Winch also thinks against Wittgenstein, insofar as he does not draw from the theory of language games the same conclusion as Wittgenstein did: Namely,
62

30

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

that philosophy is only an "idle" language game whichparadoxicallycan at best fulfil the task of uncovering its own "causes," that is, the reasons why people start to philosophize, and that it has to do so when language and thus social communication do not function properly anymore. Winch explicitly contests this purely negative "underlabourer conception of philosophy" of most analytic philosophers as well as of Wittgenstein, who already expressed it in the T r a c t a t u s , thereby thinking a widespread and almost traditional tendency of British philosophy to its paradoxical conclusion. For Winch philosophy becomes instead the science of the a priori forms of understanding of reality. These forms, being constituted by "rules," are also "forms of life" and thus a priori forms of "social interrelation."
63 64

The close connection between "meaning" and human behavior in a given situation, which was claimed by Wittgenstein and before him by the pragmatists since Peirce, was taken by most analytic philosophers to imply a possible reduction of "understanding of meaning" to an objective empirical description of behavior. Winch, however, interprets the close connection between meaning and human behavior the other way round: All human behavior insofar as it has to be conceived of as "following rules" and thus at the same time as "understanding" and "understandable," cannot be a legitimate object of objective empirical methods. Accepting Wittgenstein's claim that the a priori presuppositions of understanding are inseparable from the language games as a "form of life," Winch holds that sociology as the science of the "social forms of life" is not an empirically generalizing science, but is basically identical with philosophy as epistemology, both dealing with the "forms of understanding."
65 66

That strange antagonism between scientific methodology and philosophic method, which we have followed from its beginning in the T r a c t a t u s all through the development of analytic philosophy, was thereby resolved by Winch in favor of the philosophical methods. Thus our conjecture has been confirmed that the problems of meaning would eventually bring analytic philosophy in contact with the problems of the Geisteswissenschaften. In his historical review, Winch discusses J. S. Mill's Logic of the M o r a l Sciences in his system of inductive logic, just as Dilthey had done before him, thus formulating the conception of Geisteswissenschaft polemically as opposed to Mill's views. A n d Winch, in his discussion of the positivist conception of sociology from Mill to Pareto to Durkheim as an explaining science, in fact confirms the thesis of Dilthey and his school that the understanding of human life is essentially different from the explanation of natural events, and that therefore the investigation of the sociohistorical reality cannot adopt the methods of the natural sciences.
67

Winch, however, does not return to the psychologically oriented concept of understanding of Schleiermacher and Dilthey, as is clearly shown by his discussion of Max Weber. Instead, following Wittgenstein's theory of language games, he tries to determine the concept of understanding in such a way that the logician can no longer object to it on the ground that "understanding" can only be explicated either psychologically (as imagining oneself in somebody else's

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

31

situation) or by appealing to some irrational intuition. We shall therefore at this point scrutinize once more the presuppositions of Wittgenstein's method of language analysis in order to see how much it can be used to help solve the fundamental problems of the Geisteswissenschaften. Since its beginning with B. Russell and the young Wittgenstein, a strongly antipsychological attitude has been characteristic for analytic philosophy, which was much more radical than the "antipsychologism" of neo-Kantianism and of Husserl's phenomenology. As a result, the problems of "consciousness" and its "acts of knowledge" as well as the corresponding "mentalistic" theory of meaning or ideas are nowhere treated by analytic philosophy. Even the word Erkenntnistheorie (theory of knowledge) was banned by the young Wittgenstein as being psychological; and in fact this discipline has to a large extent been replaced in the Anglo-Saxon world of today by the "logic of science." Even the behaviorist tendencies of analytic philosophers should not be interpreted as the expression of a naturalistic philosophy, but rather as an attempt to get rid of "mentalism" and psychologism in philosophy. Seen as a whole, analytic philosophy has in a way returned to the pre-Kantian or (even more appropriately) to the preCartesian primacy of formal logic in philosophy.
68 69

A closer examination, however, shows that the emphasis on formal logic and its technical reformation and development does not sufficiently characterize analytic philosophy, though for a time its advocates may have thought this to be their main contribution. They could think so as long as they did not doubt that all a priori true sentences could be analyzed as analytic sentences and that thus the problem of the a priori presuppositions of all knowledge was reducible to the task of analyzing the "logical form of language." But it became more and more apparent that language owed its "form" (compare what Humboldt called the "inner form of language") not only to "logical syntax," and not even to semantics (which was thought of as only presupposing a depiction of facts intersubjectively given by experience), but primarily to "pragmatics," that is, to the use that people made of it in specific situationsto the "language game." Thus the idea of "logical analysis of language," which was introduced into philosophy by the young Wittgenstein as a speculative concept, gradually became clearer and the full implications of this program could now be understood. In fact, already the T r a c t a t u s contains a "transcendental logic," as Kant used the term; because the T r a c t a t u s presupposes as "the form of language" those a priori forms for the possible connection of experiences, which, according to Kant, have to be assumed in order to understand the possibility of objective experience. The idea that language in a way "constitutes" the world was differentiated and at the same time modified by Wittgenstein in his later works insofar as he changed from an "absolutistic" view of language to the idea of various language games relative to various "forms of life."
70 71 72

In analytic philosophy it is therefore language which takes the place of the a priori forms, capacities, and acts of the consciousness (of Kant's "transcendental synthesis of apperception"); and it was again the concept of language which was

32

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

modified during the development of analytic philosophy in a mannet analogous to the modification of Kant's concept of "consciousness" by Dilthey.
73

And now that Wittgenstein's idea of critique of language had developed from a "critique of pure language" (as Stenius calls the T r a c t a t u s ) to a critique of concrete language games as forms of life, Winch trieswith the help of the latter conceptto clarify the problem of "understanding," which Dilthey once tried to solve in his way by transforming the "critique of pure reason" into a "critique of historical teason." However, we must inquire in which way the old problems have become different by "language" having replaced "consciousness" as the key concept for the problem of "understanding." Or, to put it in Winch's terms: Why is "understanding" a more respectable concept if it is viewed in the context of a "language game," rather than meaning "imagining oneself" into "another's inner life as it was structured by his past experiences"?
74

The answer to this question lies in that strange conception of logic which has come to prevail in the third phase of analytic philosophy: Russell's and the young Wittgenstein's identification of the logical with that which corresponds to a rule. The result of this identification was, initially, that the form of language was seen only as the form of logic, and laterthe other way aroundlogic was understood to be dependent on the rules of the language games actually in use. Gilbert Ryle, for example, speaks about the "logical behavior" of words, that is, their behavior according to the rules of the many different language games as "forms of life" (as Wittgenstein used that phrase). Winch considers as the main thesis of his book: that criteria of Logic are not a direct gift of God, but arise out of, and are only intelligible in the context of, ways of living or modes of social life.
75

Thus the understanding of forms of life becomes for him practically identical with the comprehension of different systems of "internal relations." T o put it in other words, the forms of logic as they are actually used are founded in that area treated by "understanding" sociology; for "the whole idea of a logical telation is only possible by virtue of the sort of agreement between men and their actions which is discussed by Wittgenstein in the Philosophical Investigations."
76 7 7

Does this mean that Winch has replaced the "psychologism" of nineteenthcentury theory of understanding by a soft of "sociologism"? To answer this it will be best to forget altogether for the moment the thought of the one formal logic, which supposedly can give us criteria for evaluation of different language games (i.e., which follow different rules) or "forms of life." Instead we should turn to consider what Winch actually means by an understandable "form of life" or "social institution," which is itself a form of understanding. I suppose everyone will admitafter the adventurous attempt of early analytic philosophy to solve all problems of language and of understanding in terms of logicthat hermeneutics deals with something that does not belong in the realm of formal logic; on the other hand, the understanding of social

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

33

institutions as forms of life cannot be taken as an arbitrary kind of empathy because it always follows rules. The question is, therefore, in which respect hermeneutical thinking has gained a new dimension by using the idea of the unity of a language game which can be understood (e.g., by a philosopher) and is at the same time a foundation of understanding. The actual starting point of Winch's I d e a of a Social Science is Wittgenstein's discussionfascinating no matter how obscure in respect to its conclusionsof "what it means to follow a rule." What this discussion can teach us is, according to Winch, that one person alonethat is, in isolation from society cannot follow a rule; in other words, that a "private" language is not possible. According to Winch, the following point is the most convincing argument advanced by Wittgenstein in support of his idea that we have to make the concept of rule as well as of understanding and understandability dependent upon the existence of social customs, of permanent forms of life and institutions. Outside of the social context, on the basis of logic alone, we could not decide whether a person's behavior is following a rule or not, that is, whether his behavior is "meaningful" or not. Winch puts the problem this way.
78 79

What is the difference between someone who is really applying a rule in what he does and someone who is not? A difficulty here is that any series of actions which a man may perform can be brought within the scope of some formula or other if we are prepared to make it sufficiently complicated. Yet, that a man's actions might be interpreted as an application of a given formula, is in itself no guarantee that he is in fact applying that formula. What is the difference between these cases?
80

Winch then finds this criterion for whether a certain behavior follows a rule (i.e., is meaningful) and whether this behavior is correctly interpreted: One has to take account not only of the actions of the person whose behavior is in question as a candidate for the category of rule following, but also the reactions of other people to what he does. More specifically, it is only in a situation in which it makes sense to suppose that somebody else could in principle discover the rule which I am following that I can intelligibly be said to follow a rule at a l l .
81

But according to Wittgenstein, this kind of observation and interpretation of someone's behavior by others is only possible if the rule, which the behavior is following, is part of some social custom.
82

Thus Winch, in discussing a behaviorist interpretation of Wittgenstein as well as Max Weber's concept of understanding, actually comes to a formulation of the foundations of a philosophical hermeneutics. Winch concludes from his critical evaluation of Weber or rather of the traditional German concept of "understanding sociology" and "understanding psychology" that, as a presupposition of understanding, participation in a language game and in the "customs" and "institutions" connected with it must replace all theories of empathy, a concept which traditionally had been

34

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

connected with the ptimacy of the individual's innet expetiences. In accotdance with Wittgenstein and most analytic philosophers in England today, Winch atgues against the methodical solipsism of the older theory of understanding "that the concepts in terms of which we understand out own mental ptocesses and behaviot have to be learned, and must, therefore, be socially established, just as much as the concepts in tetms of which we come to undetstand the behaviot of othet people." Thetefote, accotding to Winch, sttange cultures (as forms of social life and ways of understanding the wotld) cannot be undetstood directly by mere intuition of the historian, but they have to be taken as systems of rules, which are different in principle from the fotms of life of the interpreter.
83 84

Now Winch does not take his own theory as a polemic against "understanding sociology" as such, but tathet as a justification of this conception with the help of Wittgenstein's ideas. This becomes apparent when Winch rejects a behaviotist interpretation of Wittgenstein which assumes "that most human behaviot can be adequately desctibed in tetms of the notion of habit ot custom and that neither the notion of a rule nor that of reflectiveness is essential to i t . " Although he does not carry out this debate against Wittgenstein himself, it should still be pointed out that Winch interprets Wittgenstein's examples in a way that is hatdly in accotdance with theit main tendencybecause Wittgenstein emphasizes time and again that the learning of a language by children (as well as any other behavior which follows rules) is based on "ttaining" ( A b r i c h t u n g ) . He once makes the point very strongly: "When I follow a tule, I do not choose. I follow the tule b l i n d l y . " O n the othet hand, Wittgenstein also makes clear that the question, "How can I follow a rule?," is not asking for the causes but for the justification of my behaviot.
85 8 6

At this point Winch comes in: He admits to Wittgenstein and Ryle that "knowing how to do something" is possible without reflecting upon the rules of such activity, and that, in fact, all behavior following traditions is, in this sense, "without reflexion"; he furthermore emphasizes that the implicit rules of human behavior always contain more than what can be formulated by actual prescriptions: in logic, for example, the activity of inferring (i.e., the insight of what follows from what) is always "ahead" of any procedure for its justification (and any reflection on why this follows from that). O n the othet hand, Winch emphasizes that any behavior which follows a rule can be justified if reflected upon. Fot Winch the important question is not whethet a person can formulate the tule which he follows, but rather whether it is meaningful "to distinguish a tight and a wrong way of doing things in connection with what he does. "
87 8 8

This way Winch has found a new approach to the problems of learning ttaditional forms of behavior, which Wittgenstein always had explicated with the behaviotist concept of "training": "Learning how to do something is not just copying what someone else does; . . . the pupil . . . has to acquire the ability to apply a criterion; he has to learn not merely to do things in the same way as his teacher, but also what counts as the same way."
89

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

35

From this analysis Winch concludes that human behavior which follows a ruleas opposed to the behavior of animals that may be causally explainable insofar as it was induced by trainingcannot be called "blind" because it implies understanding: The fact, for example, that behavior can only be morally relevant if an alternative to it is possible, though it does not imply that the agent is actually aware of this alternative, does imply that he could become aware of this alternative. Therefore, according to Winch, meaningful (i.e., understandable as well as understanding) behavior is characterized as presupposing a principle (a maxim); and on the other hand the very idea of a maxim of behavior presupposes actual behavior, that is, a form of life that embodies this maxim. Following Wittgenstein, Winch formulates this thought this way: "The notion of a principle (or maxim) of conduct and the notion of meaningful action are interwoven in much the same way as Wittgenstein spoke of the notion of a rule and the notion of the same being interwoven."
90 91

I think that Winch here approaches the concepts of "being-able-to-be-in-theworld" and "understanding-oneself-in-the-situation" (which imply a priori forms of understanding) as they were originally explicated by Heidegger in Being a n d T i m e and have now been used as fundamental concepts of a Philosophical H e r m e n e u t i c by H . - G . Gadamer. Common to all attempts at a foundation of hermeneutics in the school of Wittgenstein as well as that of Heidegger, is the rejection of psychological explanations of understanding; that is, both reject the methodical solipsism which is the last consequence of Cartesian subject-object metaphysics. Wittgenstein's and Heidegger's thinking furthermore agree in a positive sense in their new emphasis on language taken as the medium of intersubjective understanding as well as an interpretation of the world. Because of this turning to language, hermeneutics is no longer an auxiliary discipline of epistemology: For Winch as well as for Heidegger every particular knowledge obtained presupposes a certain understanding of the world, which is not rooted in the individual as such, but instead was developed in the process of interaction between many individuals.
92 93

These philosophical reflections on the a priori presuppositions of understanding do, however, pose new problems in respect to a justification of the Geisteswissenschaften: Winch repeatedly states that the problems of understanding as opposed to scientific explanationare not empirical but philosophical ones, because they are about the a priori presuppositions for all possible empirical knowledge. This opinion is in agreement with the fundamental thesis of analytic philosophy that the solution to all questions about the a priori structure of the world can be found by understanding the use of language, the language game (Wittgenstein), the semantical system (Carnap). What analytic philosophers are discussing here corresponds to what Heidegger calls the "understanding of being," which "goes before and along with" all empirical experience, and which is not yet a conceptual ontology, and which he therefore calls "pre-ontological." Analytic philosophers as well as Heidegger are dealing here with models of the world which proved to be of intersubjective validity, as they
94 95

36

P H I L O S O P H Y OF L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

are implied in everyday language or in the language games of the sciences. But is not the task of the Geisteswissenschaftenaccording to the traditional views of their protagonistsan empirical one? They do not want to understand an a priori system of communication but rather individual "objectivations of the mind" (Dilthey), that is, works and actions of historical human beings. It was, after all, this empirical goal that led to the confrontation with the explaining natural sciences, because it provoked the question about the difference between understanding and explaining of motives. How does Winch, from a Wittgensteinian point of view, cope with these problems? Is he able to somehow relate the two goals of understanding, as he sees them: the philosophical one and that task which he considers part of the social sciences? Because Winch orients himself on language as a system of rules, he seems to have difficulties in fully coming to grips with the problems of understanding, which former hermeneutical theories tried to solve with the help of psychological concepts like "empathy," "re-living," "re-experiencing," etc. Winch does not explicate the difference between understanding in the social sciences and explanation in the natural sciences, as the older hermeneutics did, by contrasting "outer" and "inner" experience (the latter being the empathy invested in the act of understanding). Instead he refers to the philosophical "memory" ("recalling") of intersubjective language games, which have to be presupposed for the possibility and validity of outer as well as inner experience. The understanding of someone else is thereforeaccording to Winchnot based on observations and generalizations of these observations, but on reflection upon an understanding of the world, which the researcher, as participant in the language game, has in common with the person to be understood: "any more reflective understanding must necessarily presuppose, if it is to count as genuine understanding at all, the participant's unreflective understanding. A n d this in itself makes it misleading to compare it with the natural scientist's understanding of his scientific data."
96

This argumentation corresponds exactly to the one we used in our discussion of the neopositivists, when we referred to the methodical presuppositions of all empirical observations and explanations implied in the understanding of language. But this argumentation seems to throw hermeneutics back on the paradoxical theory of the T r a c t a t u s , according to which all experience presupposes the understanding of its meaning, so that "understanding of meaning" cannot possibly involve any empirical problems. We therefore still seem to have the following alternative: Either understanding of linguistic meaning consists in "knowing what is the case" if a statement is truein this instance understanding is not hermeneutical, because the problem of intersubjective communications must have been solved or bypassed before this "factual" understanding could be discussed; or understanding refers to the a priori forms of language, to the rules governing the linguistic interpretation of the world, that is, to the presuppositions regarding the possibility of any experiencein which instance understanding is no longer empirically hermeneutical and therefore cannot, it
97

PHILOSOPHY OF LANGUAGE & GEISTESWISSENSCHAFTEN

37

seems, lead to any new experiences about meaning. Now, of course, Wittgenstein himself had replaced the one a priori form of language by an infinite number of language games as forms of life; and Winch in his attempts to formulate the philosophical foundations of the "understanding social sciences" starts out explicitly from Wittgenstein's "quasi-empiricist" statement: What has to be accepted, the given, isso one could s a y f o r m s of l i f e . Winch interprets this sentence thus: Whereas the philosophies of science, of art, of history, etc. will have the task of elucidating the peculiar natures of those forms of life called "science," "art," etc., epistemology will try to elucidate what is involved in the notion of a form of life as s u c h . " This program obviously does not yet go beyond the one of a general philosophy of culture ( K u l t u r p h i b s o p h i e ) .
1 0 0 9 8

The following paragraph illustrates how Winch thinks about the relationship between understanding forms of life and empirical facts: If we are, for example, to understand what happens when two persons "exchange" things, we have to take this behavior as a symbolic act by which the participants bind themselves to follow certain "rules of the game" in the future. The same holds for the case that a single person, having stopped reading a book, puts a slip of paper between the pages. This "private" behavior can also only be understood as "using a bookmark" insofar as it can be taken as an "application" of a "rule" which can, in principle, be checked by other people.
101

This means, in terms of methodology, that understanding of empirical facts of behavior is only possible if the general rules which the behavior is following (thus becoming meaningful behavior), are understood beforehand. It therefore appears as if the individual case to be understood had to be related, by a kind of Platonic "anamnesis," to that which is always understood: rules of social life, which are also the a priori forms for the interpretation of the world. A n d in fact this is what has to be assumed for all prescientific understanding among people who follow the same rules of conduct. But what if we have to understand a behavior very strange to us, whether this strangeness is due to spacial or temporal distance? Here, it seems, we encounter the specific problems the Geisteswissenschaften were designed to cope with; for must we not assume in this instance that the empirical "case" itself suggests its meaning, and that this suggestion does more than evoke our memory of certain rules already known before? Winch does touch on this problem, but, following Wittgenstein, he refers exclusively to the understanding of general rules as a presupposition for the understanding of some specific behavior. So, for example, in this passage: "The behavior of Chaucer's Troilus towards Cressida is intelligible only in the context of the conventions of courtly love. Understanding Troilus presupposes

38

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

understanding those conventions, for it is only ftom them that his acts detive their meaning."
102

Now a protagonist of the Geisteswissenschaften would also admit that Ttoilus' behavior towatds Ctessida can only be understood completely in the context of the medieval conventions of courtly love; but, then, is not our reading of Chaucer's tales and othet similar literary documents the kind of expetience which is necessary in order to acquire an understanding of couttly love? A n d is it not the job of the historico-philological Geisteswissenschaften to enlatge our knowledge of strange forms of life through interpretation of individual works and histotical actions? But considering what has been said befote, how is such enlatgement of our knowledge of sttange tules of behaviot possible? The difficulty seems to lie in the concept of "experience of categotial meaning," that is, "experience of a priori rules of possible experience." What seems to be required is a kind of experience or a kind of empirical observation which furnishes us data that enable us to completely undetstand new horizons, new possibilities of social forms of life and of comprehending the data of expetience. This kind of experience does in fact exist. It is possible because man does not only live in one specific form of life, but beyond that he can telate to othet forms of life as his potential, which, howevet, is not unlimited but is determined by his own actual fotm of life. The "empirical" discovery of new possibilities of living and understanding comes about when we do not subsume the data of expetience as specific cases under a system of tules alteady at hand (as we do in everyday life and as natutal scientists), but when we succeed in letting things, and especially people and theit behavior, "appear" to us as provoking new rules of perception. So, for example, art seems to fulfil the task, especially in our day, of making the phenomena appeal as a fascinating challenge to out sense of perception, instead of seeing them through the cliches of conventional expetience, which are, however, necessary for practical life.
103

Now this kind of experience, which already helped to cteate the wotk of art and latet will be induced by its contemplation, still belongs to a functioning language game and therefore does not lie outside the presuppositions of understanding as postulated by Wittgenstein and Winch. Without them, no understanding of the wotk of art would be possible; but that presupposed understanding need not be the understanding of that form of life, the possibility andpethapsdesirability of which we are only to undetstand in the actual expetience of the work of art. Therefore, a certain discrepancy between the presupposed forms of understanding and the meaning of the wotk of art is likely to atise, which has to be recognized and mediated in the process of interpretation. This confrontation and creative mediation of the different rules, rathet than subsuming facts undet given rules, is the real achievement of the Geisteswissenschaften; and the understanding thus brought about is their contribution to life. Winch comes rather close to this idea when he adapts Wittgenstein's remark that the philosophers' misunderstandings of language could be compared to the behavior of barbarians confronted with a strange culture: The sociologists who

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

39

misunderstand strange cultures can be compared to the philosophers who no longer know their way about in everyday language. Winch thinks, following Wittgenstein, that the situation of being challenged by strange forms of life offers the chance of looking at one's own form of life from a different system of rules. According to this view, the individual case must be undetstood in the framewotk of a complete and finished system of rules if it is to be understood at all. This viewpoint corresponds to the one assumed in the comparative study of cultures; opposed to itot rather, preceding itis that kind of understanding in the process of which the new experience does not only make us aware of our own system of rules, as well as the strange one to be undetstood, but in which the rules for our own form of life are themselves developed and pattly constituted. This truly histotical undetstanding of something unfamiliar ("histotical" insofar as it first creates the possibility of historical thinking) is similar to the understanding of a wotk of ait, because in both of them the understanding of the individual case and of the genetal system of rules presuppose and at the same time correct each other.
104 105

Actually Dilthey has already correctly described the impottant logicomethodological relevance of histotical understanding fot the Geisteswissenschaften by saying: Undetstanding continues to increase our histotical knowledge by making use of histotical documents, by going back into patts of the past hithetto not understood, and finally because of the very progress of history itself, which produces new events and thus enlatges the field of undetstanding itself. This enlatgement demands that we find new genetal truths to penettate this wotld composed of unique events. A n d this widening of the histotical horizon allows us to form more and more general and fruitful concepts. Thus at any point and at any time in the work of the Geisteswissenschaften there is "citculation" between experience, understanding and representation of the wotld of the mind in genetal concepts.
106

H . - G . Gadamer in his Pru'iosopru'sche Hermeneutilc has recently analyzed the ontological status of this understanding as well as the sttucture of dialogue corresponding to this understanding. Instead of Winch's and Wittgenstein's conception of a one-sided dependency of the undetstanding of content on the knowledge of a certain form of life, we find here the idea of the "hetmeneutic circle": The undetstanding of forms (rules) and of content both presuppose each othet, so that the formation of human forms of life in the course of history is effected through acts of understanding.
107

From the point of view of this theory of histotical undetstanding, one could be inclined to consider Wittgenstein's comparison of language games as a mere abstraction, which should be replaced by a "concrete" philosophy in and for the historical situation, which integrates the results of the Geisteswissenschaften. Instead of analyzing and comparing the language games as models for the interpretation of the world, we should rather continue, with hermeneutical methods, the one historical dialogue in which the language gamesin spite of

40

PHILOSOPHY O F L A N G U A G E & GEISTESWISSENSCHAFTEN

their different rules of understanding the worldrelate to one world and to each other, thus forming a unity of human history, however incomplete. In this case philosophy would have to give up its "uncommitted position" (Winch) of merely comparing given forms of language and life, leaving everything "as it i s . " Instead it would join in the historical dialogue by attempting to integrate the results of the Geisteswissenschaften found to be relevant for the creative project of our life.
108 109

The idea of the Geisteswissenschaften seems to demand such reciprocal presupposing of form and content of understanding, as it was, in fact, first envisaged by Hegel; this seems to be the only fitting philosophical foundation of its work. O n the other hand, we should not forget that the Geisteswissenschaften, as they grew out of the "historical school" of the nineteenth century into an empirical discipline, had to contest Hegel's claim that a philosophical mediation between content and form of understanding was possible. They had to protest against the dogmatic way in which Hegel brought about this mediation, though its very idea remained a regulative principle of their own work. What does this mean for the post-Hegelian Geisteswissenschaften of today? I think that the mediation between form and content of understanding can be attempted only as a projection, which admittedly itself belongs to history, that is, as a dogmatic anticipation of the future, which is aware of its dogmatism but keeps it under the control of a highest level of reflection, whichin respect to the contents "leaves everything as it is."
110

We must, therefore, criticize Winch's and Wittgenstein's conception of philosophy as the comparing of the various a priori forms of understanding (an idea which also lies behind all other attempts at a comparative study of cultures), because it abstracts from the reality of historical dialogue. But, on the other hand, we have to admit that we are unable in principle to preserve, and, at the same time, to integrate into our historical understanding, the knowledge gained through that abstraction. In fact, we must concede the importance of comparative studies of forms of understanding for the undetermined future of our understanding of the world.
111

As finite human beings who cannot know sub specie aeternitatis the final result of history, we will have to accept with Wittgenstein the existence of different forms of life as different forms of understanding. But from this it does not follow, in my opinion, that philosophy should give up all attempts to critically evaluate the knowledge attained in various forms of understanding (e.g., in religion, or in myth, science, and philosophy). Instead they all should be related to the common interest in knowledge of mankind, the latter participating in a concrete historical dialogue. Winch, however, due to his Wittgensteinian presupposition that the language games determine the limits of understanding and that any question can be asked meaningfully only within a specific language game, arrives at a kind of "monadology" of different cultural systems. A further consequence is radical relativism in philosophy, or in social science as he sees i t .
112

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

41

According to Winch, the different cultural systems, for instance, the forms of a magical interpretation of the world, can only be understood in terms of their immanent rules, that is, in the context of their own social forms of l i f e . If we think of our often all too hasty explanations of archaic and exotic conceptions of the world in terms of Western, especially of scientific standards of knowledge, we can only approve of Winch's demand as a conditio sine q u a n o n for sociological research of cultures. But, then, the abstractness of the model becomes apparent too. Is not a concrete cultural system like the Western Christian religion really a compound of very different language games (from primitive magical to mystically spiritual to philosophically critical forms of understanding), which were nevertheless united through continuous dialogue with each other? A n d the external critique of religious formsnot admissible according to Winchwas it not always part of the concrete dialogue of Western Christian religion (beginning with the early critique of the sacraments and the dogmas up to the general critique of religion in the Enlightenment)?
113

Wittgenstein argues implicitly in favor of the irreducibility of language games as the last presuppositions for all understanding by pointing out that these units of use of language, practical behavior, and understanding of the world, are really "functioning." This claim may prove its power of persuasion if we only consider his well-chosen or constructed language modelsas opposed to the language of metaphysics. If we, however, like Winch, apply Wittgenstein's principles to the concrete history of the human mind, thus admitting all language games to be understandablealso the metaphysical ones, which really should not have functioned, according to Wittgensteinthen we see that these principles do not fit here. The language games of the concrete history of the human mind are not understandable just by and in themselves; they do not and did not "function" because they "outgrew" themselves in the course of history, producing through permanent dialoguethe reflection and critical questioning of their own principles (as well as the critique and eventually revolutionary corrections of the corresponding forms of life). This leads us to a last critical note on Winch's conception of "understanding science"; it is related to a problem we have already touched upon when we discussed the neopositivist theory of social science: Winch would like to preclude the application of any objectively explaining methods as irrelevant in the "understanding sciences." Insofar as Winch just wants to emphasize the different "goals of knowledge" of understanding and explaining sciences, I should like to agree with h i m . Winch, however, rejects any theory which does not take the human behavior under study in the way it is understood by the people being observed. This is, for example, what Pareto did in his theory of "residues" and "derivations," when he separated certain ideas as "human constants" from the context of their various ideological entanglements, then using these constants as the basis of his critique of ideologies. Against such attempts Winch puts forward his thesis that the "relation between idea and context is an internal one," and that the idea loses its meaning when separated from its context. T o support his
1 1 4

42

PHILOSOPHY O F L A N G U A G E & G E I S T E S W I S S E N S C H A F T E N

argument, Winch even goes so far as to tefet to the example of atithmetic. This methodological approach can actually be legitimate and fruitful from the point of view of the Geisteswissenschaften, but it also shows how much Winch's assumption of the language game, only understandable by and in itself, is teally a mete theoretical absttaction, which is strictly applicable only in mathematics thus beating witness to the origin of Wittgenstein's thinking in logic and mathematics. One could thetefote feel tempted to play off Wittgenstein, the critic of language and metaphysics, against Wittgenstein, the "surveyor" of the language games (who leaves everything "as it is"), by pointing out that most language games relevant fot the history of the human mind, if pethaps not "idling," so still need a supplementary interpretationto be fully understandablewhich goes beyond the internal relations of the concepts used and takes into account the actual behaviot of the language game participants as not consistent with the concepts used in that game fot public interpretation of the behaviot. In other words, I would like to see the language gameas opposed to Winch's conceptionas a dialectical unity of "use of language," "practical form of life," and "understanding of the wotld," which means that these thtee "moments" (as Hegel used the wotd) ot aspects which constitute the language game do not always have to agree with each othet, but tathet can stand in a cettain discrepancy towatds each other and still make up one language game. Winch is right, when he points outagainst the behaviotiststhat the behaviot of a medieval monk cannot be undetstood as being meaningful without knowledge of the rules for his behavior, which stem from his religious conceptions. But does this prove that the actual behaviot of that monk is completely undetstandable in terms of the "institutional fiction" of that religious form of life within which the monk is living?
115 116

If one admits that there is no total congruence between a petson's selfunderstanding in terms of his institutionalized ideology and the motivations of his behaviot, then one also will have to admit that the sociologist interprets a given behavior with the help of concepts which go beyond the conscious hotizon of the epoch or culture to which it belongs. Thetefore, one can even justify that he explains "fragments" of human behaviot (e.g., economical behavior) by means of statistical lawsnot different, in principle, from the explanation of, say, the behaviot of gas molecules. But the possibility of such methods of explanationinsofat as they really belong to the social sciences and do not relate, fot example, to organic processesdoes not in my opinion indicate that the methods of the natutal sciences are beginning to replace the undetstanding of the Geisteswissenschaften. I think, to the contrary, that these objectivations of cettain aspects of human behavior, which cannot (yet) be articulated in the "language of self-understanding," nevertheless are serving to further this selfunderstanding. Statistical methods, for example, are not the beginning of a statistical science of man as a whole, but they are serving objective theories of motivation; the latter are, again, not the beginning of a causally explaining science of human behaviot, but they serveas, for example, in psychoanalysis

philosophy of language

& geisteswissenschaften

43

and the critique of ideologiesthe ever new attempts to understand men better than they understand themselves. These attempts to let "objectification" serve "disobjectification," that is, that condition in which man is freed by knowledge to act responsibly, have to be judged according to whether the "objects" of the theory can become "subjects" who can incorporate that theory into their own language and self-understanding. The possibility for such incorporation does in fact define the purpose of psychoanalysis and critique of ideologies. Now we can also see in which respect the methodical presupposition of analytic philosophy, that all understanding must be explicated as "clarification of language," is correct: not as clarification of either the logical form of a single ideal language, or of monadic and static language games, but as hermeneutical development (widening and deepening) of that dialogue, which we human beingsto quote Holderlin"are."
117

NOTES
1. See W . Stegmuller, Hauptstrdmungen der Gegenwartsphilosophie, 2nd ed. (Stuttgart: Kroner Verlag, 1960), chap. 10. 2. Cf. K . - O . Apel, "Das Verstehen: Eine Problemgeschichte als Begriffsgeschichte," Archiv fur Begriffsgeschichte 1 (1955): 172ff. 3. See A . J. Ayer et al., The Revolution in Philosophy (London: Macmillan, 1957); J. O . Urmson, Philosophical Analysis (Oxford: Clarendon Press, 1956); M . J. Charlesworth, Philosophy and Linguistic Analysis (Pittsburgh: Duquesne University Press, 1961). 4. Cf. Urmson, Philosophical Analysis. 5. Cf. Charlesworth, Philosophy and Linguistic Analysis. 6. See also J. Hartnack, Wittgenstein and M o d e m Philosophy (London, 1965). 7. "Verification" to be understood here as: "To understand a proposition means to know what is the case if it is true" {Tractatus 4.024). Tractatus Logico-Philosophicus, trans. D. F. Pears and B. F. McGuinness (London: Routledge and Kegan Paul, 1922; reprint Atlantic Highlands, N . J . : Humanities Press, 1961). 8. Cf. Tractatus 4.002 and 4.003. 9. Cf. Tractatus 6.54 and 7. 10. Cf. Tractatus, 22-23. The possibility of talking in a metalanguage about an artificial language (i.e., a semantical system) has often been used since Russell's introduction as an argument against Wittgenstein; however, this possibility really illustrates a problem which Wittgenstein, by thinking through all the consequences of Russell's theory of types, had first pointed out: That a formalized language cannot "reflect" on itself. Russell's infinite hierarchy of metalanguages solves Wittgenstein's paradox only by replacing philosophy with "constructive semantics" (a solution which, in fact, Carnap later accepted). If, however, one wants to justify philosophical sentences about the form of language as a whole and its relation to the world (and neither Russell nor Camap have been able to avoid such sentences), then a "deeper" point of departure is needed: One would have to doubt the logical conception of language (as is required by Russell's "ramified theory of types"), according to which language cannot reflect on itself; furthermore, one would have to show how the living languagejust as the thought that it expressesalways stands in a reflexive relationship to itself; and finally, one would have to elucidate upon how a living language, only by virtue of this self-reflexive relationship, can

44

philosophy of

language

&

geisteswissenschaften

interpret the world from a certain perspective. This self-reflexive relationship implicit in the "hermeneutic synthesis" of any linguistic proposition (the truth claim of a factual assertion being merely a borderline case on a scale of possible linguistic utterances) has to be made explicit and put into concepts for ordinary language itself. This, however, means that Hegel's dialectic of the limit also has to be applied to language, in opposition to Wittgenstein's main concern "to draw a limit to thinking, or rathernot to thinking but to the expression of thoughts." (Cf. Theodor Litt's theory of the "Selbstaufstufung der Sprache" in his book Mensch und Welt: Grundlinien einer Philosophie des Geistes [Munich: I. & S. Federman, 1948), chap. 13.) 11. In the sense of a "syntactical metaphor": cf. E. Stenius, Wittgenstein's Tractatus (Oxford: Basil Blackwell, 1960), 211. 12. Cf. Kant, Critique of Pure Reason, A 566, and E. K. Specht, Der Analogiebegriff bei Kant und Hegel (Kantstudien, Erganzungshefte 66) (Cologne: Kolner Universitatsverlag, 1952). 13. Cf. Max Black, Language and Philosophy (Ithaca, N . Y . : Cornell University Press, 1949), 114. 14. Cf. W . Dilthey, Gesammelte Schriften, vol. 7 (Stuttgart-Gottingen: B. G . Teubner, 1957- ), 191. 15. For a positive interpretation of this central idea of Wittgenstein's, see below, pp. 30 and 35, for P. Winch's identification of "understanding social science" and epistemology. 16. Cf. E. Stenius, Wittgenstein's Tractatus, 220. 17. The last remark also appliesas should be pointed out in regard to the following quotationsto many philosophers who today would no longer like to be called "logical positivists." 18. Husserliana, vol. 1 (The Hague: Martinus Nijhoff, 1950), 123; this translation from Edmund Husserl, Cartesian Meditations: A n Introduction to Phenomenology, trans. Dorian Cairns (The Hague: Martinus Nijhoff, 1960), 91. 19. Cf. K. O . Brogsitter, Das hohe Geistergesprdch (Bonn: H . Bouvier, 1958). Also K.-O. Apel, "Die Idee der Sprache in der Tradition des Humanismus," Archiv fur Begriffsgeschichte 8 (1963): 166 and in other places. 20. Cf. on this point K. Popper, The Logic of Scientific Discovery (London: Hutchinson, 1959), 93ff. Popper points out that such an agreement cannot be reached in the way desired by the logical positivists, that is, solely on the basis of facts. The intersubjective agreement in the "community of investigators" (C. S. Peirce) about what is to count as a fact is in the last analysis the product of a pragmatically oriented social process of communicationthe latter being the starting point for the hermeneutical Geisteswissenschaften as well. 21. Cf. R. Carnap, "Empiricism, Semantics, and Ontology," in Semantics and the Philosophy of Language, ed. L. Linsky (Urbana, III.: University of Illinois Press, 1952), 268fF. 22. R. Carnap, " O n Belief Sentences," in Philosophy and Analysis, ed. MacDonald (Oxford: Basil Blackwell, 1954), 129ff. See also Hans Skjervheim's critical study Objectivism and the Study of Man (Oslo: Universitets Forlaget, 1959). 23. Cf. for this Winch's interpretation of the later Wittgenstein (see below, p. 30). 24. Camap, " O n Belief Sentences," 130. 25. Objectivism and the Study of M a n , 24. 26. Cf. Herbert Feigl and May Brodbeck, eds., Readings in the Philosophy of Science (New York: Appleton-Century Crofts, 1953); and H . Feigl and M . Scriven, eds., Minnesota Studies in the Philosophy of Science 1 (1956), 2 (1958). It should be mentioned that the various contributions collected in these volumes can hardly all be called

PHILOSOPHY OF LANGUAGE & GEISTESWISSENSCHAFTEN

45

27. 28. 29. 30. 31. 32. 33. 34.

35. 36. 37.

expressions of neopositivist philosophy. However, because of their common unquestioned assumptions, the different viewpoints (e.g., of Abel) we shall discuss must be attributed to the neopositivist phase of analytic philosophy for our purpose of historical reconstruction. That this classification is justified will become more obvious after our discussion of Peter Winch's Idea of a Social Science, which criticizes the unquestioned presuppositions of the neopositivists' theory of science, thereby using arguments of the older Wittgenstein, that is, of the latest period of analytic philosophy. First published in Philosophy of Science 15 (1948). Quoted here from H . Feigl and M . Brodbeck, eds., Readings in the Philosophy of Science, 319ff. Ibid., 321. Ibid. John Stuart M i l l says: " A n individual fact is said to be explained by pointing out its cause, that is, by stating the law or the laws of causation of which its production is an instance" {Logic, book 3, chap. XII, 12, sec. 1). C . G . Hempel and P. Oppenheim, "Theory of Scientific Explanation," Philosophy of Science 15 (1948), sec. 4. Ibid., p. 327. I shall not here broach the question of whether such erroneous thinking can be found in Aristotle's theory of the causa finalis. Cf. Karl-Theo Humbach, Das Verhdltnis von Einzelperson und Gemeinschaft nach h s i a h Royce (Heidelberg: C . Winter, 1962), 1 lOff. and 137ff. I think the history of ideas has not yet fully appreciated the remarkable fact that in the second part of the nineteenth century "hermeneutic philosophy" developed in the United States as well as in Germany, rooted in Hegel's thinking, but stimulated by positivist and pragmatic logic of science. See below, p. 21. Cf. E. Rothacker, "Sinn und Geschehnis," in Sinn und Sein (Tubingen: M . Niemeyer, 1960), 3. In this context it should be pointed out, however, that for Wittgenstein and Popper the inductive method of the natural sciencesas ars inveniendi in the formulation of hypothesesis also only of psychological interest and hence irrational and prescientific. (Cf. Tractatus 6.3631, 6.36311, and 6.371, and K. Popper, The Logic of Scientific Discovery, chap. 1, 1 and 2.) For the methodology of logical positivism only the logical deduction of sentences from other sentences can be considered "scientific" in the strict sense: In causal explanations it would correspondingly be the logical deduction of observation statements from general (hypothetical) laws and from sentences describing the antecedent conditions for the occurrence of the fact to be explained. In causal explanations in history (which Hempel considers to be no more than "explanatory sketches"), the psychological interest of the scientist, according to Popper, is switched over to the antecedent conditions while the formal structure of "explanation" remains the same. This is so because in history these antecedent conditions are not just initial conditions of an isolated system which develops according to laws of nature, but they themselves are what the historical hypotheses are aboutbeing the causes of the particular events in question. The "general laws" of the explanation, however, are pragmatically presupposed in the way they are known to everybody through daily experience. (Cf. K. Popper, Die offene Gesellschaft und ihre Feinde, vol. 2 [Bern, Francke 1958], 323ff. For the difficulties of this conception, cf. W . Dray, Laws and Explanation in History [Oxford: Oxford University Press, 1964].) About the concept of "explanatory sketch," see C . G . Hempel, "The Function of General Laws in History," in Theories of History, ed. P. Gardiner (Glencoe, III.: Free Press, 1959), 351.

46

PHILOSOPHY OF LANGUAGE & GEISTESWISSENSCHAFTEN

38. Hempel and Oppenheim, "Theory of Scientific Explanation," 330. In 1930 Neurath said: "Empathy, understanding, and the like may help the researcher, but it enters into the system of statements of science as little as a good cup of coffee, which helped the researcher to do his work" (Empirische Soziologie [Vienna, 1931], 56). 39. Hempel and Oppenheim, "Theory of Scientific Explanation," 331. 40. A n example for this is the critical Bible "explanation" of the Enlightenment as started by Spinoza's Tractatus Theologico-Politicus. 41. Hempel and Oppenheim, "Theory of Scientific Explanation," 331. 42. Ibid., 328. 43. See below, p. 33, regarding the objections which can be raised to this assumption, rather widespread among logical positivists, in the light of Wittgenstein's language game theory. 44. Cf. Peter Winch, The Idea of Social Science (London: Routledge and Kegan Paul, 1958), for a similar account of psychoanalysis from the point of view of the later Wittgenstein's philosophy. 45. In Readings in Philosophy of Science, 677-88. 46. Also this thesis gains some plausibility under the assumption that the facts to be connected by understanding exist in isolation, that is, without there being presupposed some understandable contextbut this is exactly what we do not assume in the Geisteswissenschaften. In the context of a more or less understandable complex human situation (e.g., a war, the foreign policy or artistic movement of a certain period), the understanding of possible connections between the individual facts is progressively limited and made more precise by comparing it to possible ways of understanding the period or phenomenon as a wholeand vice versa. Thus the objective control, which Abel requires, is to some extent already achieved by this very process of understanding. 47. 48. 49. 50. 51. 52. 53. 54. Readings in the Philosophy of Science, 685. Skjervheim, Objectivism, 33ff; see also p. 64. T. Parsons, The Social System (1951), 544, fn 4. See above, p. 18. See P. Winch Idea of a Social Science, 115, who, however, does not consider the possibility that understanding can utilize objective methods for its own ends. See below, pp. 37 and 38. Cf. E. Rothacker, Logik und Systematik der Geisteswissenschaften (Bonn, 1948), 119ff. Cf. K. Godel, "Uber formal unentscheidbare Satze der Principia Mathematica und verwandter Systeme I," Monatshefte fiir Mathematik und Physik 38 (1931): H . 1. As the following section will show, this semantical dimension has to be understood as integrated into a pragmatical dimension. The "operative" foundation of logic and mathematics by P. Lorenzen seems to take account of this fact right from the start. From the position of the later Wittgenstein, Lorenzen's theory can be taken as a synthetic a priori construction of a language game. Russell's and Wittgenstein's "logical atomism" as well as the early neopositivism of Carnap and Schlick confusedas Leibniz had done previouslythe "inner form of language" and the "form" of "formal logic," which must be assumed for any consistent use of language. This confusion concerning that key concept of "logical form" (which, though never defined, predetermined all further speculation) had the following results. A t first, invoking "the" formal logic, semantical prescriptions were laid down for the use of language by philosophers (e.g., by Russell in applications of his ramified theory of types for the use of the expressions "is" and "exist," and by Carnapagainst Heideggerfor the use of the expression "noth-

55.

PHILOSOPHY OF LANGUAGE & GEISTESWISSENSCHAFTEN

47

56.

57. 58.

59. 60. 61. 62. 63. 64. 65.

66. 67. 68. 69.

70.

ing"). Later, howeverinvoking the "pluralism" of possible rules of language formal logic as the presupposition for a consistent use of language was unnecessarily put in doubt. Cf. K . - O . Apel, "Sprache und Wahrheit in der gegenwartigen Situation der Philosophic," Philosophische Rundschau, 1959: 161-84; and E. Tugendhat, "Tarskis Semantische Definition der Wahrheit," Philosophische Rundschau, 1960: 131-59. See above, p. 5. Cf. G . A . Paul on Wittgenstein in The Revolution in Philosophy, ed. A . J. Ayer et al. (London: Macmillan, 1957); also cf. J. Hartnack, Wittgenstein and Modem Philosophy (London, 1965). G . Ryle, The Concept of Mind (London: Hutchinson, 1949), chap. 1, "Descartes' Myth." Wittgenstein, Philosophical Investigations 1.154, 179, 180, 321. Ibid., 1.577. Cf. Winch, The Idea of a Social Science, 28, 63; see also Wittgenstein, Philosophical Investigations 1.225. Cf. Winch, The Idea of a Social Science, 3ff. Ibid., 40. Cf. Blue and Brown Books (London, 1958), 69: "The use of the word in practice is its meaning." A n d in the Philosophical Investigations, when Wittgenstein analyzes the language game of construction workers (1.6): "Don't you understand the call 'slab!' if you act upon it in such and such a way?" Cf. also Remarks on the Foundations of Mathematics 3.32: "What interests me is not the immediate realization of a truth, but the phenomenon of immediate realization. Not indeed as a special mental phenomenon, but as one of human action. I am asking: what is the characteristic demeanour of human beings who 'realize' something 'immediately,' whatever the practical result of this realizing is?" For a possible different interpretation, however, see Philosophical Investigations 1.197, 307, 308. Winch, The Idea of a Social Science, 43. Cf. ibid., chaps. 3 and 4. Cf. Tractatus 4.1121. A typical expression of this view in regard to the history of philosophy is found in the historical account Formate Logilc by the neo-Thomist Bochenski (Munich: Verlag F. A . Freiburg, 1956); English translation: Forma/ Logic, tr. Ivo Thomas (Notre Dame, Ind.) See note 55. Typical of this way of thinking is the following remark by Catnap in his article "Die Methoden der logischen Analyse" (Actes du 8eme Congres International de Philosophie a Prague, 1934 [Prague, 1936], 142-45): "We think that there is no third kind of sentences next to empirical and analytic sentences. The idea of a third kind seems to be the result of a confusion of psychological and logical questions. We believe that Phenomenology in the last analysis has not yet overcome the psychologism which it has been fighting so strongly." Meanwhile Carnap in his constructive semantics as well as Wittgenstein in his analysis of language games have, each in his own way, rediscovered the problems of a synthesis a priori of structures of meaning: Because these problems are inherent in the design of a language game, that is, in the "depth grammar" (Wittgenstein) which in a way constitutes its objects, as well as in the construction of an only pragmatically justificable, quasi-ontological framework of language, within which questions about empirical or logical sentences can then be asked. (Cf. R. Catnap's Empiricism, Semantics and Ontology.) About the conventional character of a spontaneous "projection" of a language game, Wittgenstein says in his Remarks on the Foundations of Mathematics 1.74: "But here one would like to retort: There is no

48

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

71. 72. 73.

74. 75. 76. 77. 78. 79.

80. 81. 82. 83. 84-

greater difference than that between a proposition about the depth of the essence and one abouta mere convention. But what if I reply: To the depth that we see in the essence there corresponds the deep need for the convention." Cf. Heidegger's juxtaposition of Weltentwurf and Schiclcung des Seins (= "Zur-Sprache-kommen des Seins"). Cf. E. Stenius, Wittgenstein's Tractatus; and A . Maslow, A Study in Wittgenstein's Tractatus (Berkeley: University of California Press, 1961) xiii f. Cf. E. K. Specht, Die sprachphilosophischen und ontologischen Gnmdlagen im Spatwerk Lwkuig Wittgensteins (Cologne: Kolner Universitatsverlag, L963). Cf. the famous sentence from Dilthey's preface to his Einleitung in die Geisteswissenschaften (Gesammelte Schriften, vol. 1, xiii): "In the veins of the knowing subject, as constructed by Locke, Hume and Kant, no real blood is flowing, but the diluted juice of 'reason' as the mere activity of thinking." We have contrasted here for the sake of clarity Winch's and Dilthey's terminologies and viewpoints; cf. also Winch's discussion of M . Weber, The Idea of a Social Science, 111. Ibid., 100. Ibid., 123. Ibid., 126. Cf. Wittgenstein, Philosophical Investigations 1.197ff. Cf. the following "remarks" by Wittgenstein (ibid. 1.199): "Is what we call 'obeying a rule' something that it would be possible for only one man to do, and to do only once in his life? . . . It is not possible that there should have been only one occasion on which a report was made, an order given or understood; and so on. To obey a rule, to make a report, to give an order, to play a game of chess, are customs (uses, institutions). To understand a sentence means to understand a language. To understand a language means to be master of a technique." Furthermore, 1.243: "Could we . . . imagine a language in which a person could write down or give vocal expression to his inner experienceshis feelings, moods, and the restfor his private use? Well, can't we do so in our ordinary language? But that is not what I mean. The individual words of this language are to refer to what can only be known to the person speaking; to his immediate private sensations. So another person cannot understand the language." Wittgenstein hints at the answer in the following paragraphs, so in 1.256: "Now, what about the language which describes my inner experiences and which only I myself can understand? How do I use words to stand for my sensations?As we ordinarily do? Then are my words for sensations tied up with my natural expressions of sensation? In that case my language is not a 'private' one. Someone else might understand it as well." Cf. also 1.257: "What would it be like if human beings showed no outward signs of pain (did not groan, grimace, etc.)? Then it would be impossible to teach a child the use of the word 'tooth-ache'." Winch, The Idea of a Social Science, 2 9 . Ibid., 30. Cf. Wittgenstein, Philosophical Investigations, sees. 198 and 328. Winch, The Idea of a Social Science, 119. For a very similar way of refuting the methodical solipsism of the theory of understanding (and of modem epistemology in general), cf. Heidegger, Sein und Zeit, sees. 26 and 31. The difference from nineteenth-century hermeneutics becomes apparent when we compare Winch's theory with the following passage of Dilthey in his Einleitung in die Geisteswissenschaften (1883), where he argues against sociology: " M u c h of the goal directed interrelations in a society can be explained by reference to social relations as such. But something like the development of philosophy, for example, is

PHILOSOPHY OF LANGUAGE & GEISTESWISSENSCHAFTEN

49

grounded and has its purpose not only in society but in the individual as such. This 'double foundation' is even more obvious in the case of religion and art. If we were to think of a single human being existing on earth, he would, provided only that he lived long enough, develop these different mental activities all on his own, though living in total isolation" (pp. 422-23). Dilthey's last sentence at least we will no longer be able to accept as expressing a meaningful fictitious model, since Wittgenstein showed the impossibility of a private language. Compare also A . Gehlen's polemic against Dilthey in his "Philosophie der Institutionen" {Urmensch und Spdtkultur [Bonn: Athenaum Verlag, 1956]). Gehlen's "constitution of human consciousness and self-understanding via institutions" comes close to Winch's ideas in many respects. However, the later Dilthey, who, to correct his earlier psychologism, used Hegel's concept of the "objective mind," expressed ideas quite similar to those of Gehlen and Winch. Cf. e.g. Gesammelte Schriften, vol. 7, 146f.: "Every single human expression represents something that is common to many and therefore part of the realm of the objective mind. Every word, or sentence, every gesture or form of politeness, every work of art and every historical deed are only understandable, because the person expressing himself and the one understanding him are connected through something they have in common; the individual always thinks, experiences and acts as well as understands in this 'common sphere'." 85. Winch, The Idea of a Social Science, 57. 86. Cf. Philosophical Investigations 1.219; cf. also 1.198, 206, 217. 87. Cf. Winch ( T h e Idea of a Social Science, 55ff.), who quotes Lewis Carroll on what the tortoise said to Achilles (Complete Works, Nonesuch Press). 88. Winch, The Idea of a Social Science, 58. 89. Ibid., 58f. Winch can refer, in support of this view, to Wittgenstein's own analysis in the Philosophical Investigations 1.143. 90. Winch, The Idea of a Social Science, 62, 65. 91. Ibid., 63; and Wittgenstein, Philosophical Investigations 1.225. 92. Cf. H . - G . Gadamer, Wahrheit und Methode (Tubingen: J. C . B. Mohr, 1960); also of particular interest, Gadamer, "Zur Problematik des Selbstverstandnisses," in Einsichten: Festschrift fur G . Kruger (Frankfurt a . M . : Klostermann, 1962). Also Gadamer, " V o m Zirkel des Verstehens," in Festschrift fur M . Heidegger (Pfullingen, 1959), 24-34. 93. Cf. H . - G . Gadamer on the concept of "language game" in Philosophische Rundschau 11 (1963): 42ff. 94. Cf. Winch, The Idea of a Social Science, 9f., 15-18, 71f., 83-86. 95. Cf. the following sentences from the Philosophical Investigations: "What looks as if it had to exist, is part of the language" (sec. 50); "Grammar tells what kind of object anything is" (sec. 373); "Essence is expressed by grammar" (sec. 371). It would be worthwhile to compare this with what the later Heidegger says about language as "house of being" and "domicile of the human being" (cf. Brief iiher den Humanismus [Bern, 1947]). 96. Winch, The Idea of a Social Science, 89. 97. See above, pp. 9-10. 98. Wittgenstein, Philosophical Investigations, chaps. 2 and 11. 99. Winch, The Idea of a Social Science, 41. 100. This program is very similar to the one developed by E. Cassirer in his Philosophy of Symbolic Forms (New Haven: Yale University Press, 1966), the difference being, however, that Winch emphasizes that the "forms" of human understanding to be investigated have to be taken as rules which are followed " i n the context of interhuman relationships in a sociecy" (The Idea of a Social Science, 40). According to Winch, all former philosophy of language neglected the fact "that those very

50

PHILOSOPHY OF L A N G U A G E & GEISTESWISSENSCHAFTEN

101. 102. 103.

104. 105. 106. 107. 108. 109. 110. 111.

112. 113. 114.

categories of meaning, etc., are logically dependent for their sense on social interaction between men." Winch, The Idea of a Social Science, chap. 2, sec. 2: "Meaningful Behavior." Ibid., 82. This kind of experience has been described as "encounter" (Begegnung) in the existentialist literature of pedagogics and Geisteswissenschaften, which was strongly influenced by M . Buber (left und Du, 1922), but also by the rediscovered Feuerbach (K. Lowith, Das lndividuum in der Rolle des Mitmenschen, Munich, 1928). Cf. also F. Bollnow, Existenzphilosophie und Pddagogik (Stuttgart, 1959), chap. 6. Winch, The Idea of a Social Science, 114. Ibid., 114, 118. W . Dilthey, Gesammelte Schriften, 7: 145. Cf. Gadamer, Wahrheit und Methode; idem, "Zur Problematik des Selbstverstandnisses"; and idem, " V o m Zirkel des Verstehens." Cf. also J. Lohmann, "Die Entfaltung des menschlichen Bewulkseins als Sprache," in Freiburger Dies Universitatis, vol. 11, 1963-64. Winch, The Idea of a Social Science, 102f.; and Wittgenstein, Philosophical Investigations 1.124. For the development of German Geisteswissenschaften out of the "historische Schule," cf. E. Rothacker, Einleitung in die Geisteswissenscha/ten (1920). The last sentence is meant as an objection to Gadamer's critique of the comparative methods in the Geisteswissenschaften, though 1 have followed closely his ideas in the last paragraphs (cf. Gadamer, Wahrheit und Methode, 220, 380). Cf. Winch, T h e Idea of a Social Science, 102: "connected with the realisation that intelligibility takes many and varied forms is the realisation that reality has no key." Cf. ibid., chap. 4, 1 (countering Pareto). In the terminology of the later Wittgenstein we could say that objective "explanation" and "understanding of motives" are two different language games, which correspond to different forms of behavior. Hempel's theory of historical explanation has in fact been criticized from this point of view by analytic philosophers of the Oxford school. B. P. Gardiner ( T h e Nature of Historical Explanation, Oxford, 1952), for example, admits besides causal explanation the "explanation by a motive," which is what a detective does who imagines himself in the position of the probable culprit in order to understand his possible ways of acting in the given situation. Gardiner thus arrives at the concept of "explanation in terms of'intentions' and 'plans'" (pp. 49f.), which is differentiated from Dilthey's and M . Weber's concept of understanding only insofar as the metaphysical background of an idealist philosophy of the mind is missing. For a further critique of the neopositivist theory of "historical explanation," cf. W . H . Walsh, Philosophy of History (London, 1960), and W . Dray, Laws and Explanations in History (Oxford: Oxford University Press, 1957). Dray, besides the "explanation by laws," speaks of an "explanation by the goal," calling the latter "understanding" and stressing that it is not merely a heuristic method serving the explanation by laws. The historian, he says, must "revive, re-enact, rethink, re-experience the hopes, fears, plans, desires, views, intentions, etc. of those he seeks to understand" (P- 119).

115. G . Winch, The Idea of a Social Science, 107, 109. 116. For this concept, see A . Gehlen, Urmensch und Spdkultur. 117. We have quoted the Tractatus from the translation by D. F. Pears and B. F. McGuinness; Philosophical Investigations and Remarks on the Foundations of Mathematics from the G . E. M . Anscombe translation.

TWO Intentions, Conventions, and Reference to Things: Meaning in Hermeneutics and the Analytic Philosophy of Language

H E R M E N E U T I C P H I L O S O P H Y O F U N D E R S T A N D I N G AS A HEURISTIC H O R I Z O N FOR DISPLAYING T H E P R O B L E M - D I M E N S I O N S O F A N A L Y T I C PHILOSOPHY OF M E A N I N G

"Meaning" and "understanding," 1 suppose, are still bound to be associated with different quarters or strands of contemporary philosophy. In Continental Europe at least the term "understanding" is a key term of h e r m e n e u t i c s , from Droysen's and Dilthey's philosophy of history or, respectively, of the Geisteswissenscha/ten through post-Heideggerian h e r m e n e u t i c p h e n o m e n o l o g y . In Anglo-American and Scandinavian circles, on the other hand, at least the term "meaning" is rather bound to be associated with post-Wittgensteinian ormore recently and more correctlypost-Fregian analytic philosophy of l a n g u a g e . In order to deal adequately with my topic, it seems to me useful heuristically to bring together the different philosophical horizons of hermeneutics and analytic philosophy of meaning. In order to make an attempt at doing so, 1 may be allowed to say in advance some words on the historical background and the recent development of both streams or directions of thought. In the Hellenistic age, hermeneutics (or, in Latin, ars mterprefcmd'i) had its place as a somewhat peripherical discipline within the common framework of the Texvat X o y i K a i (or, in Latin, artes s e r m o n i c a l e s ) , that is, together with grammar, logic, rhetoric, and eventually poetics. In the Middle Ages ars interpretandi, as doctrine of the multiple meaning of the Holy Scriptures, still kept its connection with the trivium of the artes liberales, especially with the newly developed doctrine of the proprietates t e r m i n o r u m , that is, significatio and suppositio, although its place became rather that of an auxiliary discipline within the theological and the juristic faculty. From the humanistic Renaissance and the Protestant Reformation, however, hermeneutics took a special career which was particularly determined by Luther's rejection of the principle of tradition and the speculative method of allegorical interpretation of the Bible in favor of
1

51

52

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

the principle of understanding the Bible from its own context ( s c r i p t u r a sui i n t e r p r e s ) . It was also within the frame of seventeenth-century Protestant theology that the Greek name hermeneutic was revived and substituted for the humanistic term ars interpTetandi. The philosophically relevant aspect of the special career of hermeneutics was constituted by the fact that it was a Lutheran theologian and post-Kantian philosopher, namely, F. D. Schleiermacher, who initiated the transition from hermeneutics as an auxiliary discipline to a general epistemological reflection on the conditions of the possibility of understanding and interpretation ( A u s l e g u n g ) of written texts and speech. By this quasi-Kantian radicalization of the pragmatic-methodological questions of interpretation in theology, philology, and jurisprudence, Schleiermacher opened the way for the conception of hermeneutics as basis for the methodology of the philological and historical disciplines (as did A . Boeckh and J. G . Droysen) and, moreover, for all so-called Geisteswissenschaften (thus with W . Dilthey). Finally, even Heidegger's further radicalization of the idea of hermeneutics in the sense of a "fundamentalontological" explication of human "being-in-the-world" as "happening" of the "disclosure" of the "meaning of being" and thereafter Gadamer's "universalityclaim" of "philosophical hermeneutics" may be considered, I suggest, as consequences of the outlined special career of hermeneutics. Now this special theologico-philosophical career of hermeneutics since the time of Luther, in a certain sense (which is not free of exaggeration), may be considered as a special feature of the history of the German mind. This partly explains the fact that until recently there was hardly any communication between hermeneutics, or, for that matter, hermeneutic phenomenology, on the one hand, and, on the other hand, the great movement of analytic philosophy of language and meaning which, so to speak, inaugurated a new paradigm of First Philosophy in our day. In the Anglo-Saxon countries hermeneutics until recently was a topic known only to theologians. Thus, for example, modern semanticists like Quine or Davidson may hardly be aware of the fact that not only the thesis of the "indeterminateness of translation" but even the question of "radical translation" or "radical interpretation" had a certain equivalent in the hermeneutic tradition (e.g., in Schleiermacher's supposition that, even with regard to ordinary conversation, nonunderstanding rather than understanding should be considered as a matter of course). Though little realized in semanticist quarters, from the vantage point of more recent hermeneutic philosophy, even from a Wittgensteinian perspective, the quasi-Cartesian standpoint of "methodical solipsism" which lies at the ground of nineteenth-century psychologists hermeneutics may be rejected by strong a priori arguments. Related areas of possible convergencies, apparently unknown to both sides, are presented, for example, by the question as to the necessary preconditions of actually understanding other people, ranging from general presuppositions of rationality or c o m m u n i c a t i v e competence to their actual application to the situation of communicative understanding by, say, a "principle of charity" (Quine, Davidson)
2 3 4

INTENTIONS, CONVENTIONS, R E F E R E N C E TO THINGS

53

or a "preconception of perfectness" (Vorgriff der V o U k o m m e n h e i t ) (Gadamer) or evenin case of a sttiving for consensus by discourseby an ethically televant "counterfactual anticipation of an ideal speech-situation" (Habermas).
5 6

This unrelatedness of similar problems and even approaches in modern analytic philosophy and philosophical hetmeneutics is of coutse not only due to different national traditions, but also due to different problem-perspectives. Fot analytic philosophy of meaning developed from logical semantics of ideal language as its otiginal point of departure towatds semantics and pragmatics of n a t u r a l language, whereas modern hermeneutics developed out of a methodology of historico-philological interpretation towatds a q u a s i - t r a n s c e n d e n t a l philosophy of c o m m u n i c a t i v e u n d e r s t a n d i n g . Fot a long time this fact of different but convetging perspectives was disfigured or misrepresented by mutual prejudices. This was especially due to the fact that the philosophical foundation of hermeneutics was an achievement of the nineteenth century which, at least fot a while, was heavily burdened with psychologism; whereas analytic philosophy of meaning, starting out from the twentieth century's antipsychologist movement of logical semantics, fot a long time consigned all problems of the subject-relation of meaningand thus the problems of intention and communicative undetstandingto the wastepapet basket of so-called ptagmatics to be dealt with by behaviorism. A nototious culmination-point of this mutual misunderstanding was represented, for a while, by the so-called explanation/undeistanding-conttovetsy; more precisely, by a cettain phase of the conttovetsy. It was based, on the neopositivist side, on the supposition that hermeneutic understanding could be conceived of as "empathy" and hence, at best, as a psychologically interesting means in the setvice of causal explanation. O n the othet sidewhich could be called quasi-hermeneutic or rather pseudo-hermeneuticthe controversy was based on the ambition that undetstanding should figure as a noncausal type of explanation of actions as spatio-tempotal events. That at least this aspect of the explanation/understanding-conttovetsy rested on a mutual misundetstanding was made cleat, I suggest, by the Kuhn/Popper/Lakatos-debate about history of science; fot this debate has shown among other things that hermeneutic understanding of causal explanations within the frame of a reconstruction of the "internal history" of science (I. Lakatos, 1971) is something very different from causal explanation of undetstanding, for example, from external explanation of certain types ot featutes of ptoblem-undetstanding in the history of science by, say, economic or psycho-physiological causes.
7

In what follows, however, I shall not deal with this aspect of hermeneutic understanding, but tathet with the semiotically centtal atea of undetstanding the meaning of linguistic signs which at present, as it seems to me, makes up a broad field or zone of possible meeting and mutual illumination of hetmeneutics and analytic philosophy of meaning.

54

INTENTIONS, CONVENTIONS, R E F E R E N C E TO THINGS

The actual situation here seems to be characterized not only by a far-reaching convergency of problems, but, moreover, by the surprising fact that even a certain exchange or crossing-over of positions appears to have happened. Thus, on the one hand, modern hermeneutics, following Heidegger and Gadamer and drawing on certain affinities with French structuralism, has almost dismissed (dropped) the problematic of understanding subjective intentions, say, of the authors of texts, in favor of explicating the language of texts as an autonomous medium and even origin of meaning. After the later Dilthey already had corrected his former psychologism by the introduction of the "common sphere" of the "objective mind" as basis for "pragmatic understanding"an idea that comes close to Wittgenstein's conception of the "language games" as parts of "forms of life" the later Heidegger arrives at the following declaration:
8

Man speaks only by speaking out of [i.e., by corresponding or coming up to] language. The language is speaking. Its speaking is speaking for us in what we have spoken.
9

H . - G . Gadamer correspondingly declares: it appears to us to make up the distinction and dignity of poetry that in it language is no longer speech, i.e. it possesses in poetry a unity of form [Gestolt] independently from all relationships of speaking and of being addressed or being persuaded.
10

And in another place: In fact scripture (literariness, Scfiri/tlichkeit) is central for the hermeneutic problem in so far as in scripture language has absolved itself from the writer or author as well as from the specified (destined) address of a receiver or reader and has brought itself to an existence of its o w n .
11

O n the other hand, post-Wittgensteinian analytic philosophy of ordinary language, especially speech-act theory since Austin, went beyond the logical semanticism of truth-functional and intensional types of meaning theory towards accounting for the illocutionary and perlocutionary force of utterances within the frame of communication; and while G . E. M . Anscombe and G . H . von Wright reintroduced the concept of "intention" into general-action theory, H . P. Grice has even set up a program of meaning theory that (in the last resort) claims to reduce the "timeless meaning" of an utterance-type, as, for instance, a sentence or a nonlinguistic sign, to the prelinguistic intention of the speaker. Thus it appears that the pragmatic turn of twentieth-century hermeneutics has produced a certain analogy to the semanticist abstraction from understanding subjective intentions within the frame of actual communication.
12 13 14

Yet the situation is still more complicated. For up to now we have not yet considered one feature by which hermeneutic philosophy from the outset has been distinguished from analytic philosophy of meaning. What I mean is the

INTENTIONS, CONVENTIONS. R E F E R E N C E TO THINGS

55

fact that, notwithstanding Schleiermacher's postulate of a hermeneutics of otdinary conversation, the main concern of hermeneutics was of course always understanding the undetstanding of historical texts. Hence even in nineteenthcentury hermeneutics it was not psychological interpretation that constituted the characteristic background for the so-called canons of interpretation, but rather histotical interpretation, which even gave a new twist to the traditional types of, say, gtammatical and genetical interpretation.
15

Thus, even with tegatd to those texts whose message was mote ot less testricted to communication with contemporaries (which, by the way, is not characteristic for the great texts of religion, philosophy, law, and poetry), the primary concern of hermeneutics was not understanding the communicative understanding of the texts by the contemporaries. It was rather understanding how the meaning of the texts can be understood by people of a later epoch by way of their histotical understanding the normal or typical undetstanding of the texts by the contemporaries. In fact, the historical twist of hermeneutic understanding amounts to the demand that even those texts (e.g., of teligion, philosophy, law, and poetry) which in fact wete destined also for future generations have to be interpreted by way of historically exploring the notmal understanding at least of the language of the texts by the contemporaries; notwithstanding the possibility that the peculiar meaning of the texts may still be waiting, so to speak, for an adequate understanding by future generations. (This is the hetmeneutic point of the canon of making oneself contemporary with the contemporaries of the text.) It was by facing these complications of hetmeneutics as historico-philological understanding that the latet Dilthey came to draw a distinction between prehermeneutic or, as he said, "pragmatic understanding," as it normally functions in the context of actual communication within a "common sphere" of (social) life, and, on the other hand, "hermeneutic" or methodical ( k u n s t g e m a f l e s ) understanding. The latter, he admits, becomes necessary only, if there is no longer a "common sphere" of communication, as, for example, in case of an encounter with a foreign civilization or in case our own cultural tradition, as it is documented by the canonical or classic texts, has become foreign (alien) to us.
16

From this it becomes clear already, 1 suggest, that Wittgenstein's approach to the pragmatic understanding of meaning within the context of a "language game" as part of a "form of life" cannot immediately be compared with, and exploited by, the hermeneutic approach to the problem of understanding meaning. This is inditectly illustrated by the (quasi-)Wittgensteinian approach to "understanding a primitive society" by Peter W i n c h . For by preserving the Wittgensteinian premises even in the case of undetstanding a f o r e i g n form of life, Winch, as far as 1 can see, gets entangled in a paradox. For he claims that he himself is able, and we should be able, to understand a foreign form of life (e.g., the institution of witchcraft oracles of the Azande Negroes
17 1 8

56

INTENTIONS, CONVENTIONS, R E F E R E N C E TO THINGS

described by Evans-Pitchard) from within its own context whose paradigms are at the same time supposed to be incommensurable with our own paradigms of understanding meaning.
19

1 do not hereby wish to suggest that the problem of relativism, which in a paradoxical way has grown out of the Wittgensteinian conception of language games, or, respectively, their incommensurable paradigms as belonging to different forms of life, is foreign to the tradition of hermeneutic philosophy. But the distinguishing point is that the problem was explicitly raised and discussed as the problem of historism (not "historicism" in Popper's sense) in the philosophy of Dilthey and his followers. The point of relativism qua historism in Dilthey's view was not, as it is in Wittgensteinianism, that understanding meaning is only possible within the context of a pragmatically functioning language game as form of life. This presupposition would rather have made hermeneutics meaningless or impossible in the eyes of Dilthey. The point of historistic relativism with Dilthey is rather that, according to his insight, which goes back in some sense to Herder, it is not only the special historicity (i.e., the relatedness to a previous epoch) of the texts as interpretanda, but furthermore, the special historicity of the actual interpreter and his possible interpretants (to speak in terms of Peircean semiotics) that has to be reflected in hermeneutics. It is out of this perspective of a reflection on historicity that Gadamer sees the main problem of hermeneutic understanding in the mediation t>etu/een different historical stages of language in a broad sense. A n d he suggests we might overcome the historism of the twentieth century just by dispensing with the naive idea of "objective understanding," that is, the idea of making ourselves contemporary with the original addressees of the text or even reenacting the original intentions of the authors of the texts. Instead, in Gadamer's view, we are to follow Heidegger's "temporal interpretation" of the being of human being-in-the-world and thus of the being of the very process of meaning-tradition and of understanding that tradition through a "melting of horizons." Thus, adequate understanding of meaning according to Gadamer is not the same as understanding the original meaning of the text by restoring the original conditions of its constitution, but rather understanding what the text has to tell us through explicating in our language what the text could mean with respect to the things it deals with and through simultaneously applying what the text could mean to the actual problem-situation.
20

The most radical point in this post-Heideggerian conception of Gadamer's in my opinion lies in the supposition that the meaning of words or sentences is not timeless in a strict sense, but rather subject to the "history of being" as history of the linguistic disclosure of the meaning of being. Since the history of this meaning-disclosure is also a precondition of propositional truth and falsehood, Gadamer, along with Heidegger, even speaks of the historical process (Geschehen) of truth. A n d , since in Gadamer's as in Heidegger's philosophy of being there is no normative orientation-point for the constitution of meaning and truth beyond time, Gadamer is only consequent in stating that the process of inter-

INTENTIONS, CONVENTIONS, R E F E R E N C E TO THINGS

57

pretation (i.e., of "exhaustion of the true meaning" of the gteat texts), is not only infinite, in principle, but cannot even be conceived of as a process of better or deeper understanding of the true meaning, but at best as a process of "understanding differently" accotding to the different setting of the pertinent epoch of the history of the disclosure of being, one might add.
21

1 want to emphasize that 1 am fat from being satisfied by this solution, ot dissolution, of the problem of (historical) understanding. I even think that E . D . Hitsch is tight when he statesin his book Validity of Interpretationthat Gadamet, as well as Heidegger, in a certain conttadiction to theit own intention, represent a radical version of historism and hermeneutic skepticism. For, even if one has to giant that the meaning of wtitten textsas of speech in geneialgoes beyond all conscious intentions of theit authois, one may neveitheless cling (keep) to Schleiermacher's and W . von Humboldt's tenets that language as a geneial system determines the individual thought as well as it in turn takes its oiigin in every act of speech thiough which individual thought realizes itself. Thus notwithstanding the relative autonomy and historicity of language, hermeneutic undetstanding must considet as criteria of adequateness, and of progressive correctness, those constiaints that ate imposed on meaning by the "interpenetiation" (Schleiermacher) of public language and individual (subjective) intentions.
22 23 24

CM the other hand, howevei, I would also like to point out that in the post-Heideggeiian vetsion of hermeneutic philosophy certain problems of the older tiadition of hermeneutics concerning the tempoiality and historicity of understanding meaning have been laised in a radical wayproblems whose adequate treatment is completely missed in analytic philosophy of meaning. This is even true, and precisely so, aftet the piagmatic and intentionalistic turn of analytic philosophy of meaning. Foi example, the question of the historical distance of time is not at all considered in present theory of communication. This fact is of coutse a consequence of the methodical abstiactions implied in the peitinent analytic theories. Thus speech act theory even largely abstracts from the perspective of actual understanding or interpretation of speech acts by the hearers; and modern communication-theory in geneial is ahistoiical in so fai as it does not take into account the hermeneutic quasi-communication of historically changing audiences oi publics with wiitten texts whose messages may even be destined foi a dialogue between the enlightened spiiits of all ages, as the Italian humanists such as Petiatch and Pietio Bembo used to say25

Now theie is no doubt, I think, that the new standaids of pteciseness and controllability in analytic philosophy of meaning, as well as theii close connection to corresponding branches of linguistic semantics and pragmatics (even of linguistic text-theories!), are due to methodical abstiactions. This fact seems to speak in favoi of current methodical abstractions from time and historicity in analytic philosophy of understanding meaning. Yet, on the other hand, it is very doubtful whether the traditional questions of hermeneutics are really different in

58

INTENTIONS, CONVENTIONS, R E F E R E N C E TO THINGS

quality from, and hence irrelevant for, those dealt with in modern theory of communication. Hermeneutics, it is true, might be considered as dealing only with the receptive side (or part) of communication; but modern communication-theory on its patt may be consideted as dealing only with a special case of understanding meaning, that is, with a limit-case of meaningttadition, so to speak. Fot each case of communication may also be conceived of as a phase or stage within a semiotic process of sign-interpretation that constitutes the continuity of a linguistic a n d cultural tradition, or rather the continuity of that unique tradition of history that even comprises our actual discoutse about meaning and understanding as part of itself.

The actual need for taking into consideration the histoticity of understanding meaning becomes even more urgent, if one reflects on the fact that in traditional hermeneutics the question of an adequate understanding of meaning was not only characterized (and complicated) by a tension and a felt need for reconciliation (mediation) between the two dimensions of understanding subjective intentions and understanding linguistic conventions, but moreover by a third dimension of understanding meanings in the light of their implicit reference to things, that is, with regard to their logical implications (in a broad sense) even beyond intentional implicatures.
26

In traditionalor, more precisely, preclassicalhermeneutics there has always been, besides the canons of grammatical and generical interpretation, a canon of interpretation in the light of the subject-matter in question. In the case of theological and juridical hermeneutics the reference to things-canon was of course represented by the dogmatic point of view whose intetest was even safeguatded by certain institutions. In eatly humanistic hermeneutics around the study of classical literature ( s t u d i a h u m a n i t a t i s ) , a certain equivalent for the dogmatic point of view was established and quasi-institutionalized with regard to the normative standards of literary criticism which were based on the paradigmatic function of the classics of the different genera. Later in the Enlightenment the new standatd of scientific truth and philosophical reason (including the norms of natural law) became also a normative standatd of hermeneutics. Thus, for example, in the doctrines of historico-philological interpretation of Hugo Gtotius and, even more definitely, of Spinoza, it was explicitly stated that the ultimate criterion of correct interpretation is not the reconstruction of the intention of the author or of the text, say, of the Bible, but the accordance of the teachings of the text with reason. Following this new canon of reason, J. Martin Chladenius in his Einleitung zur richtigen Auslegung vemunftiger R e d e n u n d Schriften (Leipzig, 1742) nicely elucidates the meaning of "correct interpretation" by stating: "One understands perfectly a discourse or scripture, if one in doing so takes into consideration all thoughts which the words according to reason and the rules of the soul may raise in us" (sec. 155, my trans.). This claim of understanding by judging on the subject-matter of texts and

INTENTIONS, CONVENTIONS, REFERENCE TO THINGS

59

hence on their validity claims in the light of reason was of course rejected togethet with dogmatical hermeneutics of nineteenth-century critical, that is, historico-philological and psychological, hermeneutics. Instead, the scientistic ideal of value-free objectivism led even Dilthey, in spite of his methodological distinction between natutal science and the Geisteswissenschaften, to establish a patallel between scientific and hermeneutic objectivity by suggesting teenactable psychical ptocesses ( E r l e b n i s s e ) , ot latet "exptessive objectivations" of the mind, to be conceived of as objects of understanding capable of being fixed under absttaction from the problem of judging on their validity claims concerning theit subject-mattet. Against this shott-citcuited objectivism, Gadamet, 1 think, has tightly pointed out that undetstanding meaning, in conttadistinction to objective cognition in natutal science, otiginates from (teflexively reestablishing ot improving on) interpersonal communication on (in German, Verstdndigung uber) certain subject-mattet ptoblems (which as successful communication cannot be disentangled, in principle, from normally coming to agteement about the material validity-claims in question. Thus, understanding the meaning of a text ot speech has to be primarily conceived of as answering a question within a dialogue father than as objectifying and hypothetically explaining ot reenacting a psychic process; although it must be added that in otdet to judge on the matetial validity claims of a text ot speech, one has to be able to teflectively distinguish between the questions that ate asked or answered by the act or speech and, on the othet hand, the questions that the authot ot speaket intends to ask ot to answet.
27 28 29

This shows that the hetmeneutic type of cognition, being telated to primary objects of communication by the attempt of sharing and at the same time critically distantiating the truth-claims of one's co-subjects of cognition, has its place on a highet level of leflection, in principle, than has the natutal science type of cognition. In a sense August Boeckh in his E n z y k l o p d d i e u n d M e t h o d o l o g i e der phiologischen W i s s e n s c h a f t e n brought out this point by denning "understanding" as "cognition of the cognized" (Erkennen des E r k a n n t e n ) ; although he, along with Schleiermacher, thought of an identicalreproductionof creative acts of cognition rather than of a critical judging on their matetial validity-claims, which task he would have placed beyond the competence of philological science.
i 0

It may seem plausible that hermeneutic understanding at least within the frame of philological science may methodically abstract from the question as to the justificability of the material validity-claims of speech and textseven if one has to admit that the philological-hermeneutic absttaction would turn into an absttactive fallacy, once the wider context of interpersonal communication on a cettain subject-matter, and hence the matetial validity-claims involved, should be forgotten in favor of equating and thus confusing the act of communicative undetstanding with just a psychologically objectifying cognition of certain mental processes. However, even if one concedes a certain right (justification) to the philological-hermeneutic abstraction, one may nevertheless have strong

60

INTENTIONS, CONVENTIONS, R E F E R E N C E TO THINGS

reasons for doubting its feasibility (practicability) in all cases. For being able to judge upon material validity-claims of speech and texts may be an indispensable heuristic precondition even of merely understanding the meaning of speech or texts. Examples of this are delivered in particular by the history of philosophy and the sciences. Thus Archimedes' hydrostatics could not even be translated into Latin, let alone be interpreted, by the humanists in spite of their achievements in restoring the knowledge of Greek and Latin. It had to wait to be deciphered for the mathematician and technician N . Tartaglia, whose knowledge of Greek was rather deficient but whose knowledge, or rather know-how, in technical hydrostatics could almost match the grade of Archimedes'.
31

From examples like this it becomes intelligible that some philosophers who in hermeneutics were oriented toward history of philosophy or history of science have claimed that reference to things, or rather reference to problems, in the light of a quasi-platonic third world of ideal structural connections rather than grasping individual intentions or even psychical processes should be the crucial dimension of understanding meaning. This is in fact the point of Nicolai Hartmann's tenets in his essay D e r philosophische G e d a n k e u n d seine G e s c h i c h t e (1936), where he plays out understanding the philosophical in Plato's works against understanding the individual Platonic features of them. In a more definite and elaborated way K. R. Popper made the same point in his essay On the T h e o r y of O b j e c t i v e M i n d (1972), where a special quasi-Hegelian twist is given to the notion of the third world by the supposition of a (dialectical) interdependence between its historical growth due to the creative achievements of human subjects of cognition and, on the other hand, the understanding of the problem-situation of these creative achievements in the light of actually available third-world structures of the "objective mind." Drawing inspiration from this latter conception of a historical dynamics, so to speak, between hermeneutic understanding and the progress of our objective knowledge concerning possible reference of speech to things, 1 would like to raise the question whether Gadamer's post-Heideggerian hermeneutics that culminates in the idea of a historical process or, respectively happening of truth and meaning (see above, pp. 57-58) might not be susceptible to (i.e., bound to be supplemented and thus modified by) a normative turn which would substitute the idea of progressively deeper understanding in the place of merely "understanding differently" from time to time. In this case the three dimensions of meaning that have to be considered by all attempts at understanding, namely, linguistic conventions, individual (subjective) intentions, and reference to things (in a broad sense), would have to be historically intertwined in such a way that a "regulative idea" (Kant) of possible progress in hermeneutic truth would become visible. As far as I can see, the quasi-transcendental semiotics of C . S. Peirce, and its application to a hermeneutics of cultural tradition by J. Royce have brought to bear such a regulative idea, that is, the postulate of a final consensus of an indefinite community of interpretation about the "ultimate

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

61

logical interpretants" of sign-interpretation and application to social praxis in the long r u n . It is in the light of this heuristic horizon that 1 shall try to discuss in what follows some aspects of the interconnection of linguistic conventions, individual intentions, and reference of speech to things (in a broad sense) in analytic philosophy of meaning.
32

T H R E E D I M E N S I O N S O F U N D E R S T A N D I N G M E A N I N G IN A N A L Y T I C P H I L O S O P H Y : LINGUISTIC C O N V E N T I O N S , INTENTIONS, A N D R E F E R E N C E TO T H I N G S

As far as 1 can see, there are three main viewpoints from which the discussion about (the understanding of) m e a n i n g in analytic philosophy may be tackled. They may very roughly be centered around the key words c o n v e n t i o n , i n t e n t i o n , and reference to things. Let us consider these three topics in turn.

Most characteristic for ordinary language philosophy appears to be the c o n v e n tionalist approach, for example, the later Wittgenstein's understanding of meaning in terms of the use of language, that is, in terms of linguistic conventions or rather conventions of different language games that as "parts of forms of life" rest on an interweaving of the use of language with ways of experience and ways of behavior, especially social interaction. A more precise, but also more restricted version of this conventionalist approach is presented, 1 suggest, by J. R. Searle's combination of linguistic semantics with Austin's pragmatics of illocutionary force in his theory of speech acts. What I mean at present is the claim Searle has formulated as follows: There are . . . not two irreducibly distinct semantic studies, one a study of the meaning of sentences and one a study of the performances of speech acts. For just as it is part of our notion of the meaning of a sentence that a literal utterance of that sentence with that meaning in a certain context would be the performance of a particular speech act, so it is part of our notion of a speech act that there is a possible sentence (or sentences) the utterance of which in a certain context would in virtue of its (or their) meaning constitute a performance of that speech act.
33

I think, Searle's claim is very plausible under the presupposition of his principle of expressibility. (which reads: "for any meaning X and any speaker S whenever S means . . . X then it is possible [in principle] that there is some expression E such that E is an exact expression of or formulation of X . " I would even consider this principle itself acceptable on the assumption that the meaning of an utterance that in a strict sense can and has to be understood does not include all kinds of "perlocutionary effects" (as, e.g., emotions, beliefs, or responses by actions) but only what Searle, along with Austin, calls the
3 4

62

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

"illocutionary effect," that is to say, the heatet's pute undetstanding of what is meant by a singula! uttetance of a sentence.
35

Even on this assumption, howevet, Searle's conventionalist approach seems to me to be unsatisfactory for the following teasons: Although I believe that the principle of expressibility is valid " i n principle," I would claim at the same time that it has the status of a "tegulative idea" to which "nothing empitical can precisely correspond," to speak in Kantian terms. This would mean that, undet empitical conditions, we always have to reckon with what 1 would call a "pragmatic difference" between the literal meaning of the sentence uttered and the meaning of the individual (singulat) utterance in its pragmatic context (ot mote precisely, in its linguistic "context" and/ot in its situational-pragmatic "co-text"). This latter meaning must not be confused with "perlocutionary effects," I would think. Fot the question is not as to the intended response to be reached by a speech act as pure means of a purposive rational action, but father as to the "utteter's occasion-meaning" (dice's term) as it is to be undetstood, and that means: be shared by speaker and hearer, independent of all possible responses, as being the occasion-meaning of a linguistic uttetance. Thus it is the intended "illocutionary effect," or, in othet words, the public meaning to be shared as an effect of verbal communication that may be different, according to the "ptagmatic difference" from the conventional meaning of the linguistic expressions used. From this it follows, I think, that an utterer's occasion-meaning is different from unintended logical implications of the meaning of the actual utterance, but nevertheless may be identified with what Grice called "conversational implicatures" in so fat as these can be supposed to be meant, ot communicatively intended, by the utterer. A general argument in favor of the "pragmatic difference" in this sense, and that is against Seatle's claim that "an adequate study of speech acts is (i.e., only) a study of langue," may be derived, I think, from Searle's own distinction between universal rules of speech and their c o n v e n t i o n a l realizations in different languages, as, for example, English, French, or German. For as long as this difference exists, we have reasons to suppose that, in order to satisfy (fulfill) the requirements of the universal pragmatic rules of speech in a particular situation, speakers have to make use of the compensative powet of a general "communicative competence" that ttanscends the "linguistic competence" of performing speech-acts in accordance with the syntactic and semantic rules (i.e., conventions) of certain languageseven if, and precisely if, these languages are conceived of as institutions of normal communication, as it is suggested by Searle.
36 37

Thus we are in fact directed toward the assumption that, in order to understand the concrete meaning of utterances, we must not only look for conventions of the use of language, but rather for intentions which may determine the meaning of illocutionary acts even beyond their normal expressions by sentences according to linguistic conventions. This, as 1 mentioned already, corresponds in a sense to a turn speech-act theory has taken due to the approach of Paul Grice. A n d the universal-pragmatic rules of communication, whose fulfillment,

INTENTIONS, C O N V E N T I O N S . R E F E R E N C E TO THINGS

63

as I suggested, sometimes makes us deviate from the linguistic conventions of speech, may appear to be identified, at least partly, as being different from mere linguistic conventions by Grice's "conversational postulates." Let us then consider Gtice's reintroduction of an intentionalist theory of meaning into analytic philosophy.
38 39

*
40

A precise resume of the intentionalist program may be found in G r i c e where the following "four main forms of meaning-specification" are presented: 1. " X (utterance-type) means '. . . " ' [Specification of timeless meaning for an utterance-type which is either (la) complete or (lb) incomplete]: 2. " X (utterance-type) meant here '. . . " ' [Specification of applied timeless m e a n i n g for an utterance-type which is either (2a) complete or (2b) incomplete]: 3. " U meant by X (utterance-type) '. . . " ' [Specification of u t t e r a n c e type occasion-meaning]: 4- " U meant by uttering X that . . . " [Specification of u t t e r e r ' s o c c a s i o n meaning]. It is with regard to this sequence that Grice argues "that timeless meaning and applied timeless meaning can be explicated in terms of the notion of utterer's occasion-meaning (together with other notions), and so ultimately in terms of the notion of intention." A n d , in order to grasp the point of this reductionist thesis, it must be added that Grice uses the term "utterance" in such a way that the value of X in "uttering X " is not taken to be fixed by linguistic or any other convention. He only restricts the meaning of X in "utterer's occasion-meaning" to being "nonnatural" (nn) in the sense ofbeing not the meaning of a symptom, but the expression of an intention. Thus he suggests the following "initial definition" of "utterer's occasion-meaning":
41 42

Li meant something by uttering X " is true iff, for some audience A , U uttered X intending (1) A to produce a particular response r (2) A to think (recognize) that D intends (1) (3) A to fulfill (1) on the basis of his fulfillment of (2). Trying to grasp the gist of this intentionalist approach to meaning, I would suggest that there are two essential features to which it may be traced back: First, the claim that the structure of meaning in speech may be reduced to the structure of meaning in prelinguistic communication. Second, the claim that the structure of communicative acts in general may be reduced to the structure of purposive-rational actions (to use the terminology of Max Weber) that use cettain actions (i.e., prelinguistic and linguistic utterances) as means or instruments in order to produce some effect (a "tesponse") in other people. Now, having turned from Searle to Grice because of the "pragmatic difference" (between the universal rules of verbal communication, suggested by Grice's

64

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

"conversational postulates," and Searle's "conventional realizations" of the rules of speech), I am now inclined to argue against both of Grice's reductionist claims partly with the aid of Searle and partly in the light of what I would call universal or transcendental pragmatics.
43

Before going to examples, let me first outline my general outlook and, so to speak, my prejudices in the matter at stake: 1. Although I agree with Grice as against Searle in so far as I think that "communicative competence" goes beyond "linguistic competence" by being able to use verbal utterances in a nonconventional way and even use extralinguistic or paraverbal signs as vehicles of meaning-communication; I do not think, on the other hand, that the structure of meaning in verbal and paraverbal communication between humans can be reduced to the sttucture of meaning as it may be conceived of as possible in an extralinguistic or prelinguistic type of communication (i.e., in a type of communication where a verbal fixation of general meaning according to linguistic conventions cannot be presupposed). In other words, I think that the possibility of our sharing fixed common or public meanings, as it is made possible by linguistic conventions, is the basis of the institutional fiction of timeless meaning; and, furthermore, that it is only in the light of this public timeless meaning that we may conceive of meaning at all and thus far also of meaning in prelinguistic communication. From this I conclude that, notwithstanding the g e n e t i c priority of prelinguistic communication, we cannot, in principle, reduce the concept of meaning to preor extralinguistic meaning. The final reason for this impossibility, it seems to me, lies in the fact that our question for the nature of "meaning" is not a quaestio facti, but a quaestio j u r i s (Kant), namely, a question concerning the conditions of the constitution of the validity of the "meaning n n " of human utterances. Now such a question, I suggest, cannot be answered by going back to prelinguistic stages of intentionalist behavior (which we can at best imagine by a sort of privative analogy to our intentions), but only by "transcendental reflection" to the conditions of the possibility of our own meaning-intention in our very argument about this subject. The pursuit of this nonpsychological self-reflection must lead us to the result, I think, that our meaning-intentions are in fact made possible by the public meanings of languageeven if, by virtue of "communicative competence," we transcend the conventional linguistic meanings of words and sentences in accordance with our particular intentions. The same point has always been expressed by the insight that unliteral meaning presupposes the literal meaning of speech. Thus far, I suggest, my preconception of the subject-matter is determined by a transcendental-pragrnatic argument. It is "transcendental" in so far as it follows the Kantian method of the q u a e s t i o j u r i s and of "ttanscendental self-reflection," that is, reflection on the conditions of the validity of one's own thought. In reflecting on the fact that the meaning of my thought qua argument is made possible not only by synthetic functions of my consciousness, but by public

INTENTIONS. C O N V E N T I O N S . R E F E R E N C E TO THINGS

65

meanings of language, however, my argument becomes transcendentalpragmatic. For, deviating thus far from classical Kantianism as well as from modern semanticism, it presupposes that the subjective formation of i n t e r s u b j e c lively valid thought is made possible, at least in part, by the public meanings of language, w h i c h , nonetheless, o n their part are dependent o n interpretation, and possibly creative reinterpretation, by human subjects of meaning-intentions. I should think that this type of argument amounts to an overcoming of what Husserl in his Cartesian Meditations called "methodical solipsism" and considered as a necessary implication of transcendental reflection i n the sense of the " p h e n o m e n o l o g i c a l e p o c h e . " A n intrinsic motive of my opposition to G r i c e ' s approach is indeed constituted by the impression that he and his followers are i n fact going to reestablish "methodical solipsism" (if not "psychologism" and, as a consequence, all kinds of reductionist n a t u r a l i s m ) by falling back b e h i n d the meaning a priori of public language and thus behind what I would consider the new paradigm of a transcendental-semiotic type of First Philosophy that was initiated by, among others, Frege, Peirce, and Wittgenstein.
44 45 46

From these transcendental-pragmatic preconceptions, w h i c h b y the n o t i o n of a nonpsychological self-reflectionin fact predetermine my methodical approach w i t h regard to the way of asking for the nature of meaning, I am led to a certain universal-pragmatic preconception w i t h regard to different types of rule-following actions and hence of h u m a n r a t i o n a l i t y . T h i s leads me to my second general argument against Grice's approach.
47

2. A l t h o u g h I agree w i t h G r i c e (and w i t h traditional hermeneutics as against certain post-Heideggerian and structuralist hermeneutics) that the intention of a speaker (or, for that matter, of the author of a text) is i n fact a crucial criterion for the understanding of meaning i n any k i n d of communication, I do not think, on the other hand, that the structure of communicative actions may be reduced to the structure of purposive-rational actions, that is, to the structure of actions that use extralinguistic and linguistic signs as pure means or instruments in the service of producing some "perlocutionary effect" (i.e., a response) i n a communication-partner. It may be very difficult to make a distinction between purely instrumental and communicative actions o n a prelinguistic level, say, i n animal behavior. A n y way, in order to adequately deal w i t h human communication by speech, I think we must make a distinction between instrumental (i.e., purposive-rational) actions and communicative actions for the following reasons: A first argument, w h i c h may be derived immediately from the foregoing transcendental-pragmatic argument, amounts to saying that a single person could not understand the intentions of his purposive-rational actions (or even the r u l e s of means-ends rationality) without presupposing already the intersubjective, that is, c o m m o n and, as it were, timeless meaning that is fixed by the sign-types of a language, even by those of indexical expressions. T h i s presupposition of language implies, I suggest, that there must be special communicative actions, that is, actions of sharing linguistic meaning.-These actions must
48

66

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

be different fiom purposive-rational actions in that their conditions of success (ot felicity) are not satisfied by somebody's fulfilling the rules of means-ends rationality solitarily, so to speak, that is, independently from fulfilling rules of agreement with othet persons. This point may be eventually clarified by the following line of atgument. It seems to me that, on the human level of possible intetaction by speech, it is possible to make the following distinction between types of actions in general (and, for that matter, of rule-following competence or rationality). With regard to man's relation to inotganic nature, thete is the possibility of a pure instrumental rationality of purposive-rational actions, say, of btinging about some effect by performing some intetvention in natute. This type of technical rationality even seems to be the basis fot the notion of causal necessity and hence fot expetimental physics. A t the same time it obviously excludes the possibility of a communicative relationship. With tegatd to man's relation to human beings, on the othet hand, it seems to be necessary to make a distinction between, at least, three different types of tational actions accotding to different pie-attitudes and corresponding cognitive interests, so to speak:
49

i. One may to a cettain extent succeed iri treating human behavioi like that of nature, accotding to a scientific and hence virtually technological subjectobject relation. In this case, howevet, fellow human beings ate not ptopetly treated as pattnets of communicative actions, and hence of communicative undetstanding of meaning, but lathet tendentially precluded from relevant (pettinent) communication because it would distutb the effectiveness of predictions, based on causal ot statistical explanations, by the nototious effects of self-denying ot self-fulfilling prophecy.
50

In our present context, we are not concerned with this type of relationship between social engineets and theii human objects. But, since this relationship in fact provides the paradigm for the notion of insttumental action with legatd to human beings, it may already appeal doubtful, in view of this relationship, whether communication can be conceived of according to the model of people's treating each othet as objects of purposive-rational action. ii. There is, however, a special type of purposive-rational action with regard to other people which does not tendentially reduce them to natutal objects of instrumental actions, but tathet acknowledges (ot even tespects) them in a sense as partners of intetaction. I shall call this type of purposive-rational action, and the corresponding type of rule-following competence or rationality, the strategical one. What I mean may be provisorily illustrated by the intetaction of adversaries (opponents) in a war or rather of partnets in business negotiations. If the pertinent rules are to be studied in a mote restricted and controllable form, a model of this type of action may be provided by the interaction of the partnets of a game as it is dealt with in modern game theory, as it was founded by von Neumann and Morgenstern. It is important to note with regard to this model of interaction that the relationship between the actors of the game are not
51

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

67

necessarily antagonistic, but may be cooperative as well, depending only on the agent's assessment of the relationship between his own interests and those of his opponent ot pattner. This makes it understandable that the game-theoretical model of social interaction has even served as a basic model for understanding the constitution and function of linguistic conventions and thus of mutual understanding through language.
52

iii. I think however that c o m m u n i c a t i v e action in the proper sense constitutes a further type of action and rationality that may not be reduced to that of strategical action or, respectively, rationality. The reason for this assumption may be suggested by the following considerations. The notion of pute sttategical rationality of interaction between opponents in a game indeed implies reciprocity of rule-following actions and thus far implies equal status of the partners; but it does not imply, but pragmatically ptesupposes and thus fat excludes, the notion of (the rules of) coming to agteement about the rules of the game, that is, of agteements about possible purposes, means, and conditions of relevant actions within the game. Now this is the same, I suppose, as the claim that the notion of strategical action excludes and presupposes the notion of coming to agreement about, and thus shating, the meanings also the occasion-meanings!of linguistic (and patalinguistic!) utterances by communication. Now let me try to corroborate this last claim by dealing more closely with the tenets and examples of H . Paul Grice. It seems obvious to me that at least the original version of Grice's attempt at reducing the notion of m e a n i n g to that of intention could be considered as an attempt at reducing c o m m u n i c a t i v e actions to strategical actions. I would also think this to be the main point of Seatle's example of the "American soldier" who in the Second Wotld War was captured by the Italians and intended to make the Italians believe him to be a German officer and therefore let him free by citing the verse of Goethe "Kennst du das Land, wo die Zitronen bliihen?," this line being the only piece of German speech available to h i m .
53

To be sure, I am committed to claim that Searle's proposal of a revision of Grice's intentionalist analysis goes too fat if he substitutes for "S means nn something by x" the formula "S utters sentence T and means it (i.e., means literally what he says)." For, as I pointed out earlier, it must be possible fot us to mean something by using types of linguistics uttetances in a way that deviates from the pertinent conventions. This, 1 said, is a consequence of the essential difference between communicative competence and mere linguistic competence, or, in other words, of the "pragmatic difference" that constitutes "utterer's occasion-meaning." But the crucial question, it seems to me, is whether Grice can himself provide an amended version of his analysis of m e a n i n g n n that evades the counterargument provided by Searle's example and at the same time preserves his claim of reducing the notion of conventional meaning to that of intentional meaning or, in other words, communicative actions to purposive-rational actions.

68

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

Grice has tried to cope with this problem in his paper "Utterer's Meaning and Intentions" (1969). There he claims that the example of the American soldier could be understood in systematically different ways: In the first case the American soldier would only intend the Italians to believe him to be a German officer without intending at the same time the Italians to believe this on the basis of their recognition of his intention. In this case the American soldier, also on Grice's account, could not be described to mean that he is a German officer, but only to intend, by means of his utterance, that the Italians should come to that c o n c l u s i o n . In the second case, however, the American soldier would intend the Italians to reach a belief that he was a German officer via a belief that the words which he utteted were the German for "I am a German officer." In this case, Grice claims, it would eventually be possible to say that the American did m e a n : "1 am a German officer"; although he would not mean that by the words which he uttered (i.e., "Kennst du das Land, . . ."), but rather "by saying the line in a particular way."
54

In a similar way, Grice suggests, a shopkeeper in Port Said who sees a British visitor could mean that the visitor should come in by saying to him with an alluring smile, the Arabic for "You pig of an Englishman."
55

I understand part of the point of this claim to be that it must be possible to communicate with other people, and hence to mean something, by using utterances in a way deviating from the pertinent conventions. Thus far I would agree, as I pointed out earlier. But 1 would nevertheless reject the alleged examples for m e a n i n g n n s o m e t h i n g by X for the reason that they may be understood as examples of strategic behavior which, as I already suggested, must not be confused with communicative behavior. Focusing on the second example (concerning the Port Said shopkeeper), one can on no account agree that the speaker did succeed in performing an act of verbal communication, for he simply did not fulfill the condition of sharing the public meaning of his verbal utterance with his communication-partner. Note well that he not only failed to share with him the conventional meaning of the Arabic sentence (for "You pig of an Englishman"), but also the nonconventional occasion-meaning which his verbal utterance would have had, also on my account, if the speaker had succeeded in sharing it with his communication-partner. For, in order to share the occasion-meaning of the merchant's verbal utterance, the British tourist had to understand both the literal meaning of the Arabic sentence and its intended occasion-meaning. But in this case he would not have understood it in the sense the shopkeeper intended him to. He would rather have seen through the feigned speech act of the merchant, and if he nevertheless had followed the invitation to come in, he would have done so not on the basis of genuine communication but rather on the basis of feigned communication whichon both sides!stood in the service of strategical actions. As against this analysis one might argue that by saying the Arabic for "You pig of an Englishman" in a particular way (namely, "with an alluring smile") the merchant after all might have succeeded in communicating the invitation that

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

69

the tourist should come in without the latter's having understood the literal meaning of his Arabic sentence, that is, just by sharing with the tourist the meaning of his particular utterance, the meaning of an ad hoc chosen sign, so to speak. Now, one may of course grant (concede) that by using ad hoc signs we may succeed in performing extralinguistic acts of communication which are not to be confused with strategic actions. However, this semiotic possibility does not save the Gricean case for two reasons. First, it does not concern the problem of the utterer's occasion-meaning of a verbal utterance; at best it does show that, within the broader frame of semiotics, there is in fact the possibility of communication by sharing the meaning of extralinguistic sign-tokens besides communication by sharing the (conventional or nonconventional) meaning of verbal utterances. This possibility, by the way, does not provide an argument in favor of a semiotic reduction of the second type of meaning-sharing to the first type, for in my opinion at least all differentiated types of extralinguistic communications of sharable meanings may be considered to be parasitic upon the linguistic types. Second, the appeal (reference) to extralinguistic communication furthermore misses the point of Grice's example. For there it is not by sharing with the tourist the meaning of a harmless extralinguistic sign-token that the shopkeeper is to perform a successful act of communication but by meaning "come in" by saying (instead) the Arabic for "You pig of an Englishman." In short, the shopkeeper does not at all share the (thus far public) meaning of his (verbal or nonverbal) utterance with his communication-partner, but rather merely makes him believe in his sharing such a meaning, in order to, by means of this suggested belief, reach his strategic aim or purpose. A n d precisely this method of a feigned or pseudocommunication (which does not share the meaning-understanding of the communication-partner, but uses it "only as a means" within the frame of a purposive-rational action) does also constitute the structural twist of Grice's interpretation of the other example, according to which the American soldier captured by the Italians would be able to m e a n "1 am a German officer" by saying in a certain authoritarian way "Kennst du das Land, wo die Zitronen bliihen." In other words, in both examples the speaker elicits a response as a perlocutionary effect of his utterance without sincerely mediating it by the illocutionary effect (i.e., the sharable meaning-understanding) of his communicative act. This method that does not intentionally give the hearer a chance of first understanding and then responding points to the peculiarity of bad rhetorics, I suggest. A n d of course, by only suggesting the possibility of basing one's response on genuine understanding of a sharable meaning, it shows its dependence, in principle, on the structure of genuine communication as being mediated by the sharing of meaning.
56

Thus Grice's examples (or respectively his interpretation of Searle's counterexample) will at best count as deficient cases of genuine communication in the service of strategical interaction, since they do not fulfill at least three

70

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

essential conditions of genuine (linguistic or extralinguistic) communication: 1. T h e c o n d i t i o n that an uttetance must be u n d e r s t a n d a b l e in the sense that its meaning (also its occasion-meaning!) may be shared (at least) by the speaker and the hearer 2. T h e c o n d i t i o n of the truthfulness of an uttetance as an act of meaning something 3. T h e c o n d i t i o n of symmetry w i t h regard to the speaker's and the heater's supposed ability, i n principle, of using all kinds of signs or utterances i n performing all kinds of communicative acts. Thus fat three of the ethically relevant universal or transcendental-ptagmatic validity claims that as normative preconditions are necessarily involved in a l l h u m a n acts of c o m m u n i c a t i o n are not fulfilled, but suggested as being fulfilled, by Grice's examples of communicative acts. A n d if one should concede that a fourth validity-claim, namely, the truth-claim, is not only directly expressed by statements but is at least inditectly implied by the informative content of every speech-act (and of every extralinguistic utterance that w i t h regard to its meaning is parasitic upon a speech-act), then it becomes immediately cleat that also this fourth normative c o n d i t i o n of the possibility of genuine acts of communication cannot be fulfilled by Gtice's examples, since fulfilling the c l a i m to ttuthfulness or veracity is a precondition for fulfilling a truth-claim (even if the inverse should not be the case). Indeed, not only can the A m e r i c a n soldier not fulfill the claim of conveying a true information by his G e r m a n verse, but also the Pott Said shopkeeper cannot do so w i t h regard to the full informative content of his cynical utterance, whose meaning at least for the speaker himself is not restricted to that of a k i n d invitation.
57

G r i c e himself seems to feel that to be the problem; for he tries to exclude from his conception the possibility of the speaker's deceiving the hearer and nevertheless preserve the point of reducing communicative action to purposive-rational action. In order to make clear this point, he introduces the following example from his o w n experience. H e had observed, by listening to a French lesson, that the little daughter of a friend thought that a certain sentence i n French means " H e l p yourself to a piece of c a k e , " though in fact it means something quite different. N o w he made use of this knowledge at the coffee table by addressing the French sentence to the girl. A n d the girl i n fact helped herself to a piece of cake and thus showed the response he had intended.
58

G r i c e obviously thinks this to be a case of unconventional but successful communication fot the following reasons: In contradistinction to the " A m e r i c a n soldier" example w h i c h could be understood i n different ways, the " l i t t l e g i r l " example is thought to have made clear that the following c o n d i t i o n (let us call it condition C ) is fulfilled: T h e utterer U "intends a hearer to recognize (and to think himself intended to recognize) some crucial feature F, and to think of F (and to t h i n k himself intended to think of F) as correlated i n a certain way w i t h some response w h i c h the utterer intends the hearer to p r o d u c e . " If this
59

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

71

c o n d i t i o n is fulfilled, it does not matter, G r i c e suggests, "whether F is thought by U to be really correlated i n that way w i t h the response or not; though of course i n the normal case U w i l l t h i n k F to be c o r r e l a t e d . " In both cases G r i c e thinks it to be possible, i n principle, to say that the utterer U m e a n s by (uttering) X the correlated response he intends the audience to produce.
60

N o w I t h i n k that the "little g i r l " example is still ambiguous and that Grice's explication of the example by the c o n d i t i o n C , that is, by the crucial feature F of the utterance (which is thought by the audience to be correlated, and to be intended by the speaker to be thought to be correlated, w i t h the intended response) does not dissolve the ambiguity. For, I suggest, that, o n Grice's account, there are at least three possibilities of interpreting the example: 1. In one sense, one could understand the example as meaning that the utterer would make use of a certain utterance X only as a means (of feigned speech) i n order to reach immediately a certain perlocutionary effect. T h i s interpretation of the example is suggested at least by Grice's c o n d i t i o n C (although it is not necessarily suggested by the example itself!). For the c o n d i t i o n C only demands the crucial feature F of the utterance to be thought of by the hearer as correlated (and to be intended to be thought to be correlated by the speaker) w i t h a certain response, w h i c h the speaker intends the hearer to produce. It does not demand that feature F of the utterance be thought of as correlated (and intended to be correlated) primarily w i t h a certain legitimate understanding of its public meaning, that is, w i t h an "illocutionary effect" i n the sense of A u s t i n and Searle. H e n c e the condition C would not warrant, I suppose, that the utterance of the French sentence could be used and understood independently from situations where its utterance would be immediately followed by the k i n d of response it is thought to be correlated w i t h . Thus, if the utterance of the French sentence is only thought to be correlated with actually taking a piece of cake, it is not clear but rather precluded that it may also be used (in an appropriate context) i n order to merely m e n t i o n the possibility of taking a piece of cake. Precisely this possibility, however, constitutes the distinguishing character of meaning o n the level of h u m a n speech, I suggest. A n d it is made possible by the difference between the illocutionary and the p e r l o c u t i o n a r y effect of speech-acts as demanded by A u s t i n and Searle. It seems clear to me that Grice's neglect (disregard) of this difference as a condition of the possibility of h u m a n verbal communication (in contradistinction to mere c o n d i t i o n i n g behavior) is connected w i t h Grice's neglect (disregard) of the c o n d i t i o n of sharing the meaning of an utterance by the utterer and his audience. For i n case, say, of a dog's "understanding" of his master's "utterances" one may indeed dispense w i t h both conditions, at least to a certain extent. 2. A second interpretation of the "little g i r l " example may neglect Grice's neglect of the distinction between the illocutionary and the perlocutionary effect of a communicative act i n view of the fact that his example rather suggests that the little girl thinks that the public meaning of the English sentence " H e l p yourself to a piece of cake" is correlated by c o n v e n t i o n w i t h the

72

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

French sentence used by the speaker O n this account, the undetstanding of this meaning is in fact suggested to het independently from the tesponse which she is intended to produce. A n d , if the girl is also intended to think that the speaket intends that illocutionary effect, one may not say, either, that she is deceived with tegard to the speaket's intention. Nevertheless, I think, she would still be object of a sttategic manipulation by feigned speech as long as the speaket himself does not shate the public meaning of the French sentence which he suggests her to undetstand. 3. Now thete is in fact a thitd interpretation of the "little girl" example that may even suppose this last condition to be fulfilled, such that the example might indeed pass as an illustration of an unconventional but successful act of communication, or even, as a ttiumph, so to speak, of communicative ovet mete linguistic competence. In this case, we have not to suppose that the speaket is using the uttetance of the French sentence only as a means (of feigned speech), in otdet to teach his aim, that is, to produce a perlocutionary or even an apparently illocutionary effect in the little girl; we now have rather to suppose that the speaker, notwithstanding the fact of linguistic convention that the Ftench sentence does not teally mean "Help yourself to a piece of cake," is really prepared to share for a while the little girl's peculiar understanding of the French sentence, in order, on the basis of this agteement on its public meaning to teach a genuine illocutionary effect and eventually, furthermore, a perlocutionary effect that is mediated by the illocutionary effect. In this case, howevet, the meaning of the unconventional but successful act of communication would not be based on the intention of the uttetet alone, as Grice would have it, but rather on his successfully supposing and exploiting a momentary quasi-convention between him and the little girl. This quasi-convention concerning the meaning of a sentencewhich in fact is patasitic upon the conventional meaning of an English sentence!would be the crucial condition of the possibility of his meaning something by utterance X, that is, of his intention's sharing a public and, as it wete, timeless meaning with his audience. This interpenetration of individual intention and public meaning, 1 would like to call to mind, was the point in the hetmeneutics of Schleietmachet and W . von Humboldt.
61

In fact, successful quasi-conventions like that which is supposed in the "little girl" example may sometimes even lead to an enlargement of the meaningconventions of a natural language by certain sttange (alien) phrasesas in case of the Getman phrase "Machen Sie keine Fisimatenten," which, according to one account, originally meant "Do not say 'visitez mes tentes'" (the meaning of the French phrase being delivered seems to have been a tathet empty politenessceremony of the French officers of the Napoleonic occupation forces). In any case, successful communications by conventionally incorrect uses of uttetance-types must be considered, I think, as presupposing, in principle, and hence as parasitic upon, genuine linguistic conventions. Hence they cannot be claimed to show that (the meanings of) communicative actions may be reduced

INTENTIONS. C O N V E N T I O N S . R E F E R E N C E TO THINGS

73

to (the intentions of) purposive-rational actions. T h e same seems to me to be true w i t h regard to examples of extralinguistic c o m m u n i c a t i o n like using a whistle, i n order to inform somebody, i n accordance w i t h a previous arrangement (agreement), about the occurrence of an expected event. A different case would be the slamming of a door w h i c h might have only the prelinguistic meaning of an expressive symptom and thus would not at all count as a case of "meaning n n " i n the sense of Grice's intentionalist a p p r o a c h a l t h o u g h , by the way, its meaning would still have to be explicated in the light of the public timeless meaning of signs based o n linguistic conventions. But, I think, there is still another reason, besides the need for linguistic conventions, why public meanings cannot be reduced to intentions. A s we have seen, o n Grice's program for a meaning-theory not only "Utterer's meaning by X " is to be reduced to (his) intention, but finally even the so-called timeless meaning of utterance-types. T h i s contention, I t h i n k , is only consequently from his point of view. But it seems to me to be just as consistent from my point of view to claim that the m a i n reason why even utterer's meaning cannot be reduced to utterer's intention, is indicated by the fact that utterer's meaning is made possible by his sharing the so-called public timeless meaning of types of linguistic utterances. N o w this latter aspect of meaning is itself not only dependent o n linguistic conventions, I suggest, but also o n what we have called reference to things i n a broad sense. (This is at least true for the timeless meaning of propositional sentences and their parts, although it seems to be inadequate to speak of reference to things even in the broadest sense w i t h regard to the conventional public meaning of mode-indicatots and of performative sentences and their self-referential parts. For these seem to be complementary to meaning qua objective reference, if they are considered w i t h i n the frame of the propositional-performative double-structure of speech-acts and their explicit expression by sentences of natural language. T h i s complementarity structure of objective reference and subjective self-reference must be postulated, I suppose, as a ttanscendental semantic and pragmatic precondition of the h u m a n capability of not only representing objective states of affairs but moreover expressing the truth-claim that is connected with propositions w i t h i n the frame of statements, and besides that the veracity-and-tightness c l a i m that is connected w i t h any type of speech-act.) A t this point the third key word and great topic of meaning theory comes into play.
6Z

In order to introduce a vantage point from w h i c h it seems easy to grasp the prima facie difference between the problem of understanding particular meaningintentions and the problem of understanding so-called timeless meaning as dependent o n reference to things, let us teflect o n the fact that, i n our mutual constitution of, let us say, philosophical texts, we are ourselves engaged i n understanding the meaning of meaning, or i n understanding understanding. T h i s type of understanding seems to be very different from understanding the

74

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

intentions of certain speakers or authors of texts. It even seems to be different from understanding the meaning-conventions of a certain language, for we are not properly concerned with understanding the meaning of "meaning" and "understanding" in English, or the meaning of "signification" or " c o m p r e n d r e " in French, or the meaning of "Bedeutung" and " v e r s t e h e n " in German. We are rather supposing to be engaged in understanding what should be understood as the unitary meaning of these words of the three languages. How is this problem associated with the problems of understanding linguistic conventions and subjective or individual intentions? For a traditional Platonist it seems to be easy to solve our problems at this point; for he could direct us to the timeless ideas as the ultimate paradigms of the timeless meaning that is shared by the wotds "meaning," "signification," and " B e d e u t u n g " or, respectively, by the words "understanding," "comprendre," and " v e r s t e h e n . " A n d the Platonist would of course claim that the individual intentions of the different speakers in the present discussion as well as the possibly different linguistic conventions of the English, French, and German words for "meaning" and "understanding" had to be understood in the light of the timeless ideas in which they participate. Now, even if one does not wish to follow the Platonist suggestion towards a hypostatization of extralinguistic meanings (or of linguistic meanings as being extralinguistic entities, as some people would say), one nevertheless cannot deny, 1 suggest, that there is a good point in postulating a unitary normative dimension to be presupposed in our understanding of m e a n i n g and understanding; although we must take our point of departure from the conventional meanings of particular languages. If we would seriously deny that the supposition of a unitary meaning of m e a n i n g and u n d e r s t a n d i n g may function as a "regulative idea" of our common inquiry in spite of possibly different subjective intentions and linguistic conventions, we would deny implicitly that our present attempt at coming to agreement about these problems by discussion could make any sense. This seems to me to be a transcendental argument against the very idea of reducing the meaning of m e a n i n g , that is to say, the timeless m e a n i n g of special words of special languages that may be used by individual persons in special contexts, to either special linguistic conventions or to individual intentions. In fact, the (very) idea of performing such a reduction by intersubjectively valid arguments is selfcontradictory and hence absurd, 1 would suggest. In this context, it seems to me interesting that Paul Grice, in displaying his reduction-program, does not even attempt to make his m e t h o d consistent with his p r o g r a m . For, in order to do this, he would have to show that his very method of reducing timeless meaning to intentions could presuppose, or rely on, the truth of his reducibility thesis; that is to say, he could exclude from his method any recourse to a normative standard besides that of correctly understanding a particular intention. Now Grice does not try to explicate the m e a n i n g of "X means, in a timeless
63

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

75

way, that P" by reducing this m e a n i n g of m e a n i n g to, say, " I , H . P. Grice, m e a n by X that P," which then would be reduced ultimately to an explication in terms of Grice's intention to bring about some effect in his audience. Contrary to this, he rather follows a method of intentionalist reduction that implies a kind of Davidsonian theory of m e a n i n g . For his scheme of reduction reads like this: "Utterer U means something by uttering X " is true if, a n d only if, for some audience A , U utters X intending . . . " (my emphasis).
6 4

This modern way of understanding the meaning of certain linguistic sentences in terms of their truth-conditions is of course different from a Platonist account, but, by its recourse to the idea of truth, it combines tendentially the normative dimension of Plato's way of ideas as paradigms with an empirist and languagerelated way of grounding the meaning of propositions on their reference to things. This tendency, though, toward a synthesis of a normativist and an empirist way of meaning-explication, 1 think, is not sufficiently developed and brought to bear by the Tarskian convention T that makes up the core of Grice's scheme of meaning-explication. The deficiency seems to me to rest on the fact that Tarski's logico-semantical scheme expresses only a necessary condition for a meaning-theory as well as for a truth-theory, presupposing already the possibility, in principle, of a transcendental-pragmatic explication of the possible criteria of redeeming truth-claims as well as meaning-claims.
65

Davidson, in his application of Tarski's scheme in the service of a semantic theory of linguistic meaning, provides an interesting pragmatic supplementation of the logical-semantical scheme by substituting for the right branch of the biconditional sentence-form "s is true (in the object language) if and only if p , " a phrase that requires empirical evidence for the fact that p is held to be true by a native speaker of the object-language. Thus Davidson gets empirically testable sentences (so called T-sentences) of a truth-theory for natural languages such as '"Es regnet' is true in German when spoken by x at time t if and only if it is raining near x at t." A n d he uses these T-sentences as a clue to an empirically testable theory of "radical interpretation," that is, of explicating linguistic meanings without presupposing a meaning-understanding.
66

Now, although this approach is very plausible as an empirist and languagerelated way of correlating the explication of meanings with reference to things, it obviously does not sufficiently take into regard the quasi-Platonist normative dimension of truth and (its relation to) meaning. For it takes for granted that the competent speakers of a natural language are already "sharing a picture of the world that must, in its large features, be true" and hence may authoritatively bear witness of meaning in as far as it is based on reference to things. Such a confidence (trust) in a commonsense background of shared truth is in fact a necessary presupposition of any attempt at "radical interpretation," as is expressed by Quine's and Davidson's principle of charity (which corresponds to Gadamer's hermeneutic principle of perfectness); but, being thus a necessary condition of any attempt at communicative understanding, it nevertheless cannot serve as a sufficient condition for philosophically grounding our understanding of linguistic
67

76

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

meanings on reference to things, since we have to reckon with a possible deepening of out meaning-explications based on a possible ptogtess of ttue knowledge about reference of signs to things beyond the stage of common sense. This problem-situation, it seems to me, is the teason why the Socratic-Platonic question as to what is (the essential meaning of), say, j u s t i c e cannot be sufficiently answered by empitico-semantical (i.e., linguistic) methods of meaninganalysis. A n d the same is true, it seems to me, with regatd to out present question as to the m e a n i n g of m e a n i n g or the m e a n i n g of u n d e r s t a n d i n g meaning. The notmative dimension of the ideal unity of meaning in as fat as it is based on the normative dimension of a possible true knowledge about the reference of signs to things can nevet be sufficiently represented by empirically ascertainable linguistic conventions, although there is no other medium of definitely comprehending and expressing intetsubjectively valid truth and meaning besides the public meaning of language. In view of this ptoblem-situation it appeats to me that only C . S. Peitce's way of "making out ideas cleat" by thought-expetiments concerning the possible ptactical consequences of a correct interpretation of signs in the long run (i.e., the method of "ultimate logical interpretants") provides a method of explicating meaning by truth-conditions (and thus by reference to things) that goes beyond the linguistically testified stage of common sense and can take into account the possible progress of out knowledge about the reference of signs to things. It was a triumph of the Peitcean method of meaning-explication, when Einstein, in his special theory of relativity, succeeded in clarifying the meaning of propositions about the simuhaneousness of events in terms ofI should saythe ttuth-conditions for propositions about possible measurings of the simultaneousness.
68

The point of Peirce's normative conception of an opetationalist explication of meaning (which, by the way, is different from Btidgman's quasi-behaviotist "operationalism") was not abolished, 1 suggest, but only specified or made precise when later, partly as a consequence of Einstein's general theory of relativity, it was suggested that powerful scientific theories have a depth-dimension of semantic reference that cannot be completely explicated in terms of experimental experiences. 1 would indeed consider this turn of meaning-explication within recent philosophy of science to be a testimony (or at least a symptom ot index) fot the interpenetration of the empirist and the Platonist dimension in the idea (notion) of meaning as reference to things (in a broad sense). If, howevet, this latter assumption holds good, then the whole ttadition of philosophical and, in recent times, scientific explication of meaning by reference to things must present an important atgument in favor of our conception of the so-called timeless meaning of linguistic signs (i.e., wotds and complete sentences). For then it amounts to adopting a bad type of Platonism, ot rather pseudo-PIatonism, if one supposes that the conventions which are constitutive for special languages should not only make possible the f a c t u a l p u b l i c n e s s of meanings but moreover the timelessness that we associate with the idea (notion)

INTENTIONS. C O N V E N T I O N S , R E F E R E N C E TO THINGS

77

of linguistic meanings. I would suggest that this timelessness is n o t h i n g less and nothing more than the institutional fiction that is constitutive for the function of human language. T h i s institutional fiction, although it is never completely realized (or redeemed) by human language or, respectively, speech, seems to make possible argumentation, and hence science, philosophy, and especially logic. It may be noted at this place that the construction of ideal languages does not essentially change this situation. For the precise meanings of their expressions must i n principle be explicated w i t h the aid of natural languages as ultimate pragmatic metalanguages, so to speakjust as the definitions of classical p h i losophy had to tely o n the explicatory power of undefined words of natural language. T h i s means, however, I suggest, that the inctease of preciseness and definiteness of meanings made possible by constructed languages is always relative to certain special disclosures of the life-world according to special pragmatic contexts. From this it seems to follow that the publicness of meaning, as it is constituted by linguistic conventions, must always be open for being filled up or refilled, so to speak, by new experience i n a broad sense, including scientific experience, experience of philosophical thought and, not to forget, poetic expetience about the life-world and h u m a n n e e d s . But, if this is true, then the public meaning of words and sentences, notwithstanding their teference to things and their being grounded by conventions, must also be open for its refreshment, so to speak, by individual intentions of creative speakers and authors of texts. T h u s , i n rejecting the program of reducing the n o t i o n of meaning to that of intentions, 1 am far from denying the importance of the idea of individual intentions for our understanding the meaning of utterances and texts and hence even the n o t i o n of public meaning.
69

[ would even go further and accept the postulate w h i c h , 1 understand, underlies Husserl's transcendental intentionalism: namely, the postulate that the h u m a n subject of understanding by its "transcendental r e d u c t i o n " t h a t is, by its identifying itself w i t h the ttanscendental subject of understanding w h i c h is unavoidable i n a sense for any philosophermust claim to conceive of any possible timeless meaning as a virtual correlate of (to) his intentional act. T h i s postulate is, so to speak, the institutional fiction that constitutes t r a n s c e n d e n t a l phenomenology o] meaning. It makes up the result of Husserl's Auffiebung of classical Platonism, that is, of the eidetical theory of meaning, into the postK a n t i a n paradigm of transcendental philosophy of consciousness. But this A u f h e b u n g of the first paradigm into the second patadigm of First Philosophy was still before the thteshold of the t h i r d paradigm of First P h i l o s o p h y . For Husserl, biased by his transcendental version of methodical solipsism, had not yet taken into account the following consideration of a transcenaental semiotics: A s legitimate and unavoidable as it is for the philosopher to claim any possible meaning to be a virtual correlate of his intentional act, it is no less legitimate and necessary for h i m to reflect o n the fact that even the meanings of those concepts that make his actual self-understanding as a subject
70

78

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

of intentional acts possible are always already mediated by the public meanings of language, that is, by a medium of collective experience, so to speak. This fact, however, implies that our understanding public meanings can be deepened and thus modified by an indefinite argumentative discourse on future experience and hence by communicative intentions that may be articulated in all possible languages. Hence (therefore) the philosopher, who in arguing about the meaning of meaning must anticipatorily identify himself with the transcendental subject of true (or essential, or real) meanings of words or sentences, cannot definitively rely either on his eidetic-noematic intentions or on empirical semantic explorations of the meaning-conventions of special languages; he has rather to envisage and try to anticipate again and again the possible consensus about the "ultimate logical interpretants" of signs (Peirce) by an indefinite community of interpreters; although he can and must at the same time rely (along with Wittgenstein and Davidson) on the (transcendental-pragmatic) fact that already "in sharing a language . . . we share a picture of the world that must, in its large features, be true."

Thus, from the perspective of a transcendental semiotics, which includes transcendental pragmatics a n d /lermeneutics, one may only come to the conclusion that the key notions of subjective intention, linguistic convention and reference to things (in a broad sense) are of equal importance for the understanding of meaning. They complement each other and restrict each other as regulative principles of inquiry within the context of the so-called hermeneutic circle of meaningdisclosure. For, in order to initiate the hermeneutic enterprise of understanding and explicating the meaning of utterances or of written texts, it is possible, in principle, to start out from each angle, so to speak, if a certain preliminary understanding of the language can be taken for granted. But it seems equally necessary in this enterprise to oscillate between the three viewpoints and regulative principles, in order to correct and deepen one's meaning-conjectures. This thesis concerning the hermeneutic circle between the three regulative principles of understanding meaning is not contradicted, I think, by the fact that there is of course a (crucial) methodological difference between, say, a purely linguistic, a purely philological, and a normative-critical understanding of a text. In the first case, one may concede that linguistic conventions must be the focus of the cognitive interest, while understanding of individual intentions and considerations about reference to things or to normative paradigm can only but must indeed have heuristic functions. Similarly, in the case of philological text-interpretation the expressed intention of the author must be a crucial gage or restrictive regulator of understanding; notwithstanding the fact that the verbally expressed meaning of the text may go far beyond the conscious intentions of the author and may have significant implications that can only be grasped by an expert of the matter in question. Finally, in a normative-critical understanding
71

INTENTIONS, C O N V E N T I O N S , R E F E R E N C E TO THINGS

79

of meaning, say, i n the reconstruction of the internal history of science or of philosophy, already the selection of the relevant materials (subjects) must be determined by an anticipatory preconception of the ideal norm of possible true knowledge of the subject-matter i n question; and the elaboration of this conception i n a quasi-communicative discourse, so to speak, w i t h the classics of the discipline must remain the focus of understanding. Nevertheless, the very fact of a hermeneutic mediation of our direct study of the things i n question through a critical understanding of the intentions of the classics i n the field shows that, i n this case also, all three dimensions of meaning and of understanding meaning are i n play.

NOTES
1. Cf. H . G . Gadamer, "Hermeneutik," in Historisches Wbrterbuch der Phibsophie, ed. J. Ritter (Basel, Stuttgart: Schwabe, 1974), 3:1061-73. 2. C f . , e.g., F. D. E. Schleiermacher, "Hermeneutik," in Hermeneutik und Kriuk, ed. M . Frank (Frankfurt a . M . : Suhrkamp, 1977), sees. 15, 16. 3. For a comparison between the later Wittgenstein's argument against the possibility of a "private language," Peirce's and Royce's philosophy of the "interpretationcommunity" and Heideggerian and post-Heideggerian hermeneutics, cf. K . - O . Apel, Transformation der Phibsophie, 2 vols. (Frankfurt a . M . : Suhrkamp, 1973); English selective translation: Towards a Transformation of Philosophy (London: Routledge and Kegan Paul, 1980). 4. Cf., e.g., D. Davidson, "Truth and Meaning," Synthese 17 (1967): 304-23; and idem, "Radical Interpretation," Dialectica 27 (1973): 313-27 (also in D. Davidson, Inquiries into Truth and Interpretation [Oxford: Clarendon Press, 1984]; all page references are to the original place of publication). 5. See H . G . Gadamer, Wahrheit und Methode (Tubingen: Mohr, 1960), 277f. 6. See J. Habermas, "Towards a Theory of Communicative Competence," Inquiry 13 (1970): 360-75. 7. For an overview, see K . - O . Apel, "The 'Erklaren/Verstehen' Controversy in the Philosophy of the Natural and Human Sciences," in Chronicles (of the International Institute for Philosophy), ed. G . Fl0istad, vol. 2, pp. 19-50 (The Hague: Martinus Nijhoff, 1982). For an attempt at a solution or rather dissolution of the problem that underlies the controversy, see idem, Die Erklaren/Verstehen Kontroverse in transzendental-pragmatischer Sicht (Frankfurt a . M . : Suhrkamp, 1979). 8. For a comparison of the later Dilthey and Wittgenstein, see K . - O . Apel, "Wittgenstein und das Problem des hermeneutischen Verstehens," in Apel, Trans/ormation der Phibsophie, 1:335-77. 9. Heidegger writes: "Der Mensch spricht nur, indem er der Sprache entspricht. Die Sprache spricht. Ihr Sprechen spricht fur uns im Gesprochenen." M . Heidegger, Unterwegs zur Sprache (Pfullingen: Neske, 1959), 44. 10. Gadamer, Wahrheit und Methode, 177. 11. Ibid., 369. 12. See especially J. L. Austin, How to Do Things with Words (Oxford: Oxford University Press, 1962); a n d ] . Searle, Speech Acts (Cambridge: Cambridge University Press, 1969). 13. See G . E. M . Anscombe, Intention (Oxford: Basil Blackwell, 1957); and G . H . von Wright, Explanation and U n d e r s t a n d i n g (Ithaca, N . Y . : Cornell University Press, 1971).

80

INTENTIONS. C O N V E N T I O N S , R E F E R E N C E TO THINGS

14. See H . P. Grice, "Meaning," The Philosophical Review 66 (1957): 377-88, and idem, "Utterer's Meaning and Intentions," The Philosophical Review 78 (1969): 147-77. 15. See, e.g., A . Boeckh, Enzyklopadie und Methodologie der philologischen Wissenschaft, ed. Bratuschek (Leipzig, Neudruck, Darmstadt: Wissenschaftliche Buchgemeinschaft, 1877 [1966]). 16. See W . Dilthey, Der Aufbau der geschichdichen Welt in den Geisteswissenschaften (reprinted in Gesammelte Schriften, 7, Stuttgart, 1910 [1973]), 206ff. 17. Cf. K . - O . Apel, Transformation der Philosophie. 18. See P. Winch, The Idea of Social Science and Its Relation to Philosophy (London: Routledge and Kegan Paul, 1958), and especially idem, "Understanding a Primitive Society," American Philosophical Quarterly 1 (1964). 19. Cf. K . - O . Apel, "Die Kommunikationsgemeinschaft als transzendentale Voraussetzung der Sozialwissenschaften," in Apel, Transformation der Philosophie, 2:220-63. 20. Gadamer, Wahrheit und Methode, 245, 281. 21. Ibid., 280ff. 22. See E. D . Hirsch, Validity and Interpretation (New Haven, Conn.: Yale University Press, 1967), appendix 2. 23. See Schleiermacher, "Hermeneutik," 32. For a confrontation of this hermeneuticdialectic conception with the formal-abstract distinction between the general "form" of language and the individual "content" of thought in Wittgenstein's Tractatus and in Moritz Schlick, cf. "Sprache und Ordnung," in Apel, Transformation der Philosophie. 24. Apel, Transformation der Philosophie, 79. 25. Cf. K . - O . Apel, Die Idee der Sprache in der Tradition des Humanismus von Dante bis Vico (Bonn.: Bouvier, 1963 [3rded., 1980]), 166f. 26. For the difference between implications and implicatures, see H . P. Grice, "Logic and Conversation," in Pragmatics, ed. P. Cole and ]. L . Morgan, vol. 9 of Syntax and Semantics (New York: Academic Press, 1978), 41-55. 27. Cf. Wittgenstein's remarks in The Blue and Broivn Books (Oxford: Basil Blackwell, 1958), 1.242: "If language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgements. This seems to abolish logic but does not do so. . . . It is one thing to describe methods of measurement, and another to obtain and state results of measurement. But what we call 'measuring' is partly determined by a certain constancy in results of measurement." Furthermore, cf. Davidson's use of Quine's principle of charity as a presupposition in "radical interpretation," especially in "The Method of Truth in Metaphysics" {Midwest Studies in Philosophy 2 [1977]: 224-54), where he states: "In sharing a language, in whatever sense this is required for communication, we share a picture of the world that must, in its large features, be true. . . . Agreement does not make for truth, but much of what is agreed must be true, if some of what is agreed is false" (244f.). 28. Besides Gadamer, Wahrheit und Methode, cf. W . Kuhlmann, Reflexion und kommunikative Erfahrung (Frankfurt a . M . : Suhrkamp, 1975). 29. See Gadamer's argument with Collingwood's "logic of question and answer." Gadamer, Wahreit und Methode, 352ff. 30. Boeckh, Enryklopadie und Methodologie, 10. 31. See L. Olschki, Geschichte der neusprachlichen wissenschafthchen Literatur, 3 vols. (Heidelberg: Halle, 1919-27), 2:200ff.; also 3:71ff. 32. Cf. K . - O . Apel, "Szientismus oder transzendentale Hermeneutik?" in Hermeneutik und Dialektik, ed. R. Bubner et al. (Tubingen: Mohr, 1970), 1:105-55, sec. 3. 33. J. Searle, Speech Acts (Cambridge: Cambridge University Press, 1969), 17. In fact, Searle claims to "capture both the intentional and the conventional aspects" of

INTENTIONS. C O N V E N T I O N S . R E F E R E N C E TO THINGS

81

34. 35. 36. 37.

38. 39. 40. 41. 42. 43.

44. 45. 46.

47.

48.

49. 50. 51. 52.

illocutionary acts "and especially the relationship between them." But he also claims that in the normal case of performing speech acts, i.e., if the speaker is using words literally, he "intends" the recognition of his intentions "to be achieved in virtue of the fact that the rules for using the expression he utters associate the expression with the production of that effect" (Speech Acts, 45). Thus he comes to claim that "an adequate study of speech acts is a study of langue" (ibid., 17). Ibid., 18. Cf. note 6. Searle, Speech Acts, 39f. For the universal-pragmatic notion of "communicative competence," cf. Habermas, "Towards a Theory of Communicative Competence"; cf. also K . - O . Apel, "Sprechakttheorie und transzendentale Sprachpragmatik zur Frage ethischer Normen," in Sprachpragmatik und Philosophie, ed. K . - O . Apel (Frankfurt a . M . : Suhrkamp, 1976), 3.1 and 2. See Grice, "Logic and Conversation." See Grice, "Utterer's Meaning and Intentions," The Philosophical Review 78 (1969): 147-77. Ibid., 149. Ibid., 150. Ibid., 151. See J. Habermas, "Was heifk Universalpragmatik?," in Apel, ed., Sprachpragmatik und Philosophie, (also in Jiirgen Habermas, Vorstudten und Erganzungen zur Theorie des kommunikativen Handelns [Frankfurt a . M . : Suhrkamp, 1984]; all page references are to the volume edited by Apel); and Apel, "Sprechakttheorie und transzendentale Sprachpragmatik." See Apel, "Sprechakttheorie und transzendentale Sprachpragmatik." Cf. Grice, "Logic and Conversation." This is illustrated, I suggest, by Schiffer, whereas I consider Parret's insisting on the publicness of meaning to be an argument against any reestablishment of psychologism. See Stefan Schiffer, "Truth and the Theory of Content," in Meaning and Understanding, ed. Herman Parret and Jacques Bouveresse (Berlin, New York: Walter de Gruyter, 1981), 204-22; and Herman Parret, "Perspectival Understanding," in ibid., 249-79. See Habermas, "Aspects of Rationality of A c t i o n , " in Rationality Toddy, ed. T. Geraets (Ottawa: Ottawa University Press, 1979), 185-204; and K . - O . Apel, "Types of Rationality Today," in Rationality Today, ed. T. Geraets (Ottawa: Ottawa University Press, 1979), 307-39. Although it must be pointed outalong with Peirce against Hegel's treatment in the Phenomenology of Mindthat the use of indexical expressions is different from that of words as symbols for general concepts, it must also be emphasizedalong with Hegel and Peircethat even the situation-bound meaning of "indexical expressions" cannot be understood by us as public meaning without presupposing the general meaning of symbols like "thisness," "nowness," "I-ness," etc. We need these public general meanings as transcendental-hermeneutic presuppositions in the light of which we may understand experiences, although we may experiencesay in psycholinguistics or even in empirical hermeneuticsonly more or less similar meaning-intentions. Cf. Apel, Die ErfclarenA/erstehen Kontroverse. See ibid. See J. von Neumann and O . Morgenstern, Theory of Games and Economic Behavior (Princeton: Princeton University Press, 1944). See, e.g., D. Lewis, Convention (Cambridge, Mass.: Harvard University Press, 1969).

82

INTENTIONS. C O N V E N T I O N S . R E F E R E N C E TO THINGS

53. 54. 55. 56.

57. 58. 59. 60. 61. 62. 63.

64. 65.

66. 67-

68. 69. 70.

71.

Searle, Speech Acts, 44f. Grice, "Utterer's Meaning and Intentions," 169. Ibid. I am intentionally alluding here to Kant's verdict against using humanity in a person "only as a means," being convinced of the internal relationship between (i.e., ultimate unity of) communicative and ethical rationality. Cf. Apel, "Sprechakttheorie und transzendentale Sprachpragmatik," and idem, "Types of Rationality Today." Cf. Habermas, "Towards a Theory of Communicative Competence," and idem, "Was heiBt Universalpragmatik?" Grice, "Utterer's Meaning and Intentions," 162f. Ibid., 163. Ibid. Cf. Schleiermacher, "Hermeneutik," 32; and the chapter "Sprache und Ordnung," in Apel, Transformation der Philosophie. Cf. above, pp. 63 - 64. Notice, this is not an appeal to a petitio principii within the frame of logical deduction, but a demand for a reflective self-application of one's principle which is legitimate with respect to any philosophical program of reductionism; it is indeed destructive (in the sense of a pragmatic self-contradiction) with respect to any such program that disavows its own universal validity-claim (i.e., timeless meaning-claim and truth-claim). Grice, "Utterer's Meaning and Intention," 15Iff. Cf. K . - O . Apel, " C . S. Peirce and Post-Tarskian Truth," in this volume (originally published as " C . S. Peirce and the Post-Tarskian Problem of an Adequate Explication of the Meaning of Truth," The Monist 63, no. 3 [1980]). See, e.g., Davidson, "Radical Interpretation," 321ff. See D. Davidson, "The Method of Truth in Metaphysics," in Studies in the Philosophy of Language, ed. P. A . French, T . E. Uehling, Jr., and H . K. Wettstein. Volume 2 in the Midwest Studies in Philosophy (Morris: University of Minnesota, 1977), 245 (also in Davidson, Inquiries into Truth and Interpretation; page number is to original publication). See K . - O . Apel, Der Denkuieg von C . S. Peirce (Frankfurt a . M . : Suhrkamp, 1975). Cf. F. Koppe, Sprache und Bedilrfnis (Stuttgart, Bad Canstatt: Frommann-Holzboog, 1977). Cf. K . - O . Apel, "Transcendental Semiotics as First Philosophy," in this volume (originally published as "Transcendental Semiotics and the Paradigms of First Philosophy" in Philosophical Exchange 2, no. 4 [1978]: 324). Cf. above, p. 57.

THREE The Transcendental Conception of Language-Communication and the Idea of a First Philosophy: Towards a Critical Reconstruction of the History of Philosophy in the Light of Language Philosophy

EXPOSITION

O n e might perhaps, among philosophers of our time, come to agreement about the fact that analysis of language i n a broad sense makes up the methodological vantage point of modern philosophy, in a similar way as analysis of consciousness did after Descartes, and as analysis of the essence of things had done still earlier from the time of A r i s t o t l e . In other words, one c o u l d a n d , as I think, one s h o u l d w o n d e r whether i n our day philosophy of language has i n fact taken over the role of a First Philosophy w h i c h was ascribed (attributed) to ontology by Aristotle and later claimed for epistemology or transcendental philosophy in the sense of K a n t . T h i s question, it seems to me, is almost forced upon us by the recent history of philosophy; for nearly all of the really modern directions of philosophy i n this century have made claims to the effect that language is the crucial c o n d i t i o n for the possibility and validity of our knowledge of the structure of the world. I think, i n this context, of Wittgenstein's claim that philosophy is "critique of language," that "the limits of my language mean the limits of my w o r l d , " or, i n his later work, that the essence of things lies i n grammar and that the existence of a language game is the c o n d i t i o n for the possibility of somebody's following a rule and identifying phenomena. I further think of the V i e n n a C i r c l e and its struggle for a meaning criterion for a language of science and especially of Carnap's reduction of ontological and/or transcendental questions of traditional philosophy to questions of constructing syntactico-semantical frameworks for formalized languages. But at the same time I might also think of a turn i n continental neo-Kantianism and phenomenology, starting, respectively, w i t h Ernst Cassirer's "philosophy of symbolic forms" and with Heidegger's consideration 83

84

T H E IDEA O F A FIRST P H I L O S O P H Y

of language as "house of being" and going on to H . G . Gadamer's universality claim for a philosophical hermeneutic of language. A n d for illustration's sake I might also mention, in this context, my own claim of Nichthintergehbarkeit of natural language as a medium of transcendental reflection.
1

In connection with the transcendental claims of language philosophy one should also mention that modern linguistics, which for a long time kept away from philosophical problems by restricting itself to phonological and morphological subjects, has more recently intervened in philosophy. 1 think of N . Chomsky's renewal of the old program of u n i v e r s a l g r a m m a r and of the semantic neo-Humboldtism of so-called metalinguistics (Sapir, Whorf) and the inhaltsbezogene Sprachwissenschaft (L. Weisgerber and others). In reflecting upon the transcendental function of language, one should, however, not restrict one's attention to its role as a condition for the possibility and validity of description as if language were only a medium for anyone's solitarily objectifying the world including his fellow human beings. The later Wittgenstein's attack on the idea of a private language should open our eyes to a consideration of the intersubjective dimension of speech as communication and social interaction. Also in this dimension the philosophy of our day brought about, it seems to me, convergent trends toward a fundamental change in the level of First Philosophy. In this context, I think of the American movement of Jjragmatic semiotics with its central idea of a community of sign interpretation and of universal discourse, which leads from C . S. Peirce via J. Royce to G . H . Mead's explication of the very idea of mind or self-consciousness in terms of the internalized reciprocity of symbolic interaction in a society qua communication community. After an interlude of behavioristic reductionism, most recently this idea seems to be reconstructed in Austin's, Grice's, and Searle's philosophy of speech acts and in terms of a universal pragmatics which supplements Chomsky's idea of "linguistic competence" by the idea of "communicative competence."
2 3

This philosophy of communication and interaction may lead to a thorough transformation of the idea of "ego cogito" as the transcendental foundation of theoretical philosophy as well as of practical philosophy; and thus, in my opinion, the transcendental conception of l a n g u a g e ' C o m m u n i c a t i o n should be considered as the most important aspect of the supposed revolution of the idea of First Philosophy in terms of language philosophy. It is for this reason that in the title of this paper I juxtapose the transcendental conception of languagecommunication not simply the idea of language philosophywith the traditional idea of a First Philosophy.
4

As the subtitle of my paper indicates, I will not try to explicate directly the transcendental conception of language-communication by a systematical approach, rather 1 will try to show indirectly its meaning and significance by a critical reconstruction of the traditional idea of First Philosophy and its relation to philosophy of language. The transcendental conception of languagecommunication will in this enterprise successively display itself as a critical

T H E IDEA O F A FIRST P H I L O S O P H Y

85

standard in the light of which the traditional conceptions of First Philosophy show their internal aporetic. If one takes up this point of view as a heuristic guide one must, I think, pose the following questions with regard to the history of philosophy: What is the significance, in the light of language philosophy, of the fact that, in the beginning, "First Philosophy" was founded by the Greeks as an ontology of the essential structure of things; that later on, in the so-called "new age," it was transformed into, or replaced by, a critical epistemology or philosophy of consciousness; and that finally, in the twentieth century, both ontology and epistemology wete questioned of transformed by language-analysis in a broad sense of this term? I will deal with the different parts of this question in turn, and, after doing so, I hope to be able to suggest an answer to the question, whethet 01 in what respect the philosophy of language-communication may justifiably take over the role of First Philosophy. Needless to say, I can carry out the task I have set myself only by a tough sketch, in which there will be found provoking theses rathet than a thotough argument.

L A N G U A G E - C O M M U N I C A T I O N A N D T H E G R E E K IDEA O F FIRST P H I L O S O P H Y

Let us begin thenaccording to our programwith the question: What, in Greek philosophy, is the telation between First Philosophy and teflection on language? Whyto put the question directlydid ontology become the Fitst Philosophy at that time rather than philosophy of language? From our historical distance and level of reflecrion one could consider the last question somewhat cutious, and one might answer it by calling attention to the fact that language as a condition of knowledge is much more difficult to gtasp and to analyze than the realm of things given by the senses. A t firstone might sayattention focuses on what can be shown in unreflective experience, in the so-called intentio recta or prima; later one comes to reflectwithin the so-called intentio obliaua or secundaon cognition itself as a function of consciousness, and, finally, one may reflect on the function of language as a condition of the possibility and intetsubjective validity of knowledge. Certainly, this answer is not false; we will even accept it as a guideline fot understanding the sequence of petiods in the history of philosophy. However, it must be stressed that Greek philosophy itself went through this cycle of stages in a way. In the age of Soctates and the Sophists it already turns away from ontological questions about the nature ((piiatc,) and origin ( a p y f \ ) of things, and taises questions as to the correctness of names (dpGoTTjg 6vo|idTu)v), the function of speech (Xoyoc,), and the meaning of words as concepts or definitions ( o p o i , o p i o u x n ) - Plato, through whom we know about these discussions, already achieves the insight that the truth is not to be sought in the quality of

86

T H E IDEA O F A FIRST P H I L O S O P H Y

single names but that it is a function of their connection into a statement (X.OY05), and Aristotle, especially in his De I n t e r p r e t a t i o n e , laid the foundations
5

of a philosophy of grammar, which was further elaborated by the Stoics and thus decisively influenced the grammar of the schools in the Western world up to the present day. But why did not Plato already, as Wittgenstein suggests, look for the rule of
6

the use of words in order to find an answer to the famous questions of Socrates as to what courage or justice is? A n d why did he not see in his own definition of thinking as a voiceless dialogue of the soul with itself a clue to the fact that thinking is to be considered as a function of communication by language? A n d Aristotle, who so often opens his questions about the essence ( c u o i a ) of being (6v) by an inquiry into the use of the wordswhy did he not consider the possibility that his ontological categories are relative to the Greek language? The answer to these questions, in my opinion, has to be a twofold one: O n the one hand Plato and Aristotle would have had good reasons for being dissatisfied with doctrines which claim to "reduce" their question as to the essence of things to a mere question about the use of w o r d s . (We shall come back to this point at the end of this chapter.) O n the other hand, however, we must not overlook the fact that Plato and Aristotle did not have a concept of language adequate to enable them to see that their very questions, not to speak of the answer, were dependent on the learned use of a certain language. The classical philosophy of the Greeks had at its disposal essentially four concepts for comprehending the essence of human speech or communication: ovouxx (name), oufiPoXov, OT][ieiov (symbol or sign), opoc; (concept), and X.OY05 (speech, oratio, ratio, statement, etc.). (It is worth mentioning that it had no concept of a special language. Only the Romans had the word "lingua
7

latina".)

By means of these four concepts it was impossible to grasp that

meaning is essentially a function of a language. For these four concepts form two clusters between which the problem of linguistic meaning slips through: X-oyog (ratio) and opoc; (concept) were a priori directed to something universal which was thought to be independent of the use of language; 6vou,a (name) and rruu^oXov or OT|[ieiOV (sign), on the other hand, did in fact mean something which differs according to the use of different languages, but for Aristotle, at least, it had nothing to do with the meaning of thoughts; it was only a conventional means of designating, in the setvice of the "logos." (Perhaps it was precisely this progressive step of no longer asking for the correctness of single names but rather for the truth of statements that caused the Greek philosophers to overlook the cognitive function which languages have by virtue of the determinate meanings of their words and phrases. )
8

I will illustrate and support these assertions by two texts, which are representative of a position which has both guided and misguided all philosophical thinking up to the present day. The first passage is to be found in Aristotle's De Interpretatione (16a3ff.):

THE

IDEA O F A FIRST P H I L O S O P H Y

87

The things of the voice ate symbols of the things of the mind, and the things of writing ate symbols of the things of the voice. A n d as the lettets are not the same for everybody, neither are voices. The states of m i n d , however, to which those signs primarily refer are the same for e v e r y o n e , as are the reflections of things which ate the same fot everyone.
9

In this passage, which is as concise as it is fundamental, Aristotle established, neatly as a commonsense dictum, the intersubjective identity of possible m e a n i n g s as correlated to the ontological s t r u c t u r e of things and theteby independent, in principle, of the use of wotds. A n d here we may show more exactly what is the characteristic point of the insufficiency of the Gteek foundation of philosophy, consideted in the light of philosophy of language. It is true, as still temains to be shown, that it is not possible to "reduce" the question of the essential structure of things to the question of the actual use of wotds; but from this it does not follow that the interpretation of the structute of the wotld is independent of the use of language. The assumption of an intersubjective identity of meanings as reflections of the things is rathet an expression of the deeply tooted interest of Aristotle (as alteady of Plato) in the unity of a Fitst Philosophy as universal ontology founded in the unity of logic. Aristotle had to postulate the intetsubjective identity of meanings of the different signs of the voice as a guarantee of the (applicability of the) logical principle of i d e n t i t y .
1 0

(By the way, J. Locke's thesis of the radical variety of meanings as private ideas should not be taken to be mote than a psychologistic antithesis to the ontological conception of Aristotle. The function of language as a condition of the possibility and intetsubjective validity of undetstanding a sense datum as something is passed ovet in both cases because meanings ate, as a mattet of principle, treated as something outside of language. One may, at most, concede that Lockelike Epicutus in ancient timescalled attention to the genetical interdependence between the vatiety of languages and the variety of impressions and ideas of experience. Thereby J. Locke indirectly questioned ot, so to speak, undermined the standatd conception initiated by Aristotle, accotding to which the variety of languages is nothing elsethat is, nothing motethan a variety of sounds used as signs fot the same ideas.)
11

As is well known, even professional linguistics, as founded in the nineteenth century, for a long time preserved in practice the Aristotelian separation between signs and meanings by systematically abstracting from the problems of semantics and even syntax of language. So, at the beginning of the twentieth century the following commonsense sedimention of an ancient conception of the relation between things, thoughts, and language was still predominant: First we have to knoweach for himself the elementary data of the world; then, on the ground of these data and by abstraction with the help of universal logic, we have to grasp the (ontological) structute of the wotld; then we may designate the elementary data by signs and represent their structural connections by special signs or rather by connections

88

T H E IDEA O F A FIRST P H I L O S O P H Y

of signs; finally we inform other people, with the help of the possible signconnections, about the state of affairs we have apprehendedparticular features of this conception were questioned more than once, but in its kernel it was not overcome. Indeed the "analytical philosophy of language" of B. Russell and of the young Wittgenstein started with a rigorous renewal of this conception in so-called logical atomism.
12 13

However, the rigorous renewal of basically Aristotelian conceptions in a logically refined form brought to light still another difficulty of the ontological foundation of the First Philosophy of the Greeks: If we presuppose that language is, in principle, only an instrument for communicating to other people what everybody had previously apprehended and designated for himself, the question arises: How are we to conceive the possibility of an intersubjective convention about the meaning of signs, or if you like, about the use of signs? If the determination and designation of things and of their relations is always already presupposed for communicating and n o t alsothe other way around, then we need not and cannot in any way communicate about the relation between language and the world, and that means we cannot communicate about the meaning and reference of signs. That, in fact, is demonstrated by a paradoxical argumentation in the T r a c t a t u s L o g i c o Philosophicus of the young Wittgenstein. Only the following alternative seems to be open under these preconditions: Either there are only private languages and no communication is possible at all; or everybody has to be provided a priori with the same ideal language as onto-semantical framework, so that "solipsism . . . coincides with pure realism" and everybody must, by metaphysical guarantee, speak about the same world. To put it in other words, the language which is presupposed as possible in the ontology as First Philosophy should have been established by God and communicated to every single man by a mystical metalanguage. It is not difficult to illustrate this idea of a lingua a d a m i c a sive universalis sive naturalis sive philosophica by the history of philosophy. But what about thinking the other way around: May not communication by language, which under human conditions of life has to replace the instinctive signal-functions of animal behavior, be an intersubjective function which isat least in principleboth a convention about the use of signs and a foundation of a specific interpretation of the world?
14 15 16

It is just this way of thinking which has been prevented and suppressed by the ontological foundation of philosophy by the Greeks. This thesis may be illustrated with the following text of the commentator of Aristotle, Ammonius, who traces the point of the text back to Theophrastus:
17

Since speech ( X o y o q ) has a twofold relation . . . one to the hearers for whom it has some meaning, the other to the things, about which the speaker wishes to suggest some belief to the hearers, so with regard (to the relation) to the hearers poetics and rhetoric arise . . . with regard to the relation of the speech to the things, however, the philosopher will in the first place take care to refute the false and to demonstrate the true.

T H E IDEA O F A FIRST P H I L O S O P H Y

89

This extremely instructive text may be seen in a way as a result of the long struggle for emancipation of philosophical logic and ontology from the truthclaims of the poets and rhetoticians. By distributing the telations of speech as he does, Theophtastus rigorously separates what is nowadays called the pragmatical from the semantical dimension of speech. The pragmatical dimension is completely handed ovet to the arts of rhetoric and poetics which, fot theit patt, ate so to speak, required to give up, in favot of philosophy, all claims tegatding the semantical relation of speech as the dimension of truth. (As the full text shows, thetotic and poetics have as theit subject only those features of speech intended, "to give enjoyment, to fill with enthusiasm and to overwhelm the heaters in the way of petsuasion.") It goes without saying that this text (which is attf ibuted to Theophtastus) is in full agteement with the Aristotelian presupposition of the intetsubjective identity of the meanings of signs as correlated to identical things. Were it not so, that is, were the meanings of signs to depend on the use of some language, one would have to make sure, by communication with the hearets, what the meaning and even what the reference of the used signs ate before being able to decide about the truth of a given statement. Even Aristotle is well awate of these problems, as his practice and the last patt of the O r g a n o n show; but it is outside the limits of his ontological approach to teflect upon the pragmatical dimension of the problem of ttuththe dimension constituted by intetsubjective communication about a possible interpretation of the world by means of the signs of a language. Now it is again important to temark that Theophtastus' distribution of the dimensions of speech, as well as Aristotle's distribution of signs and meanings, has temained influential up to the ptesent day as a commonsense sedimentation. Admittedly, the privilege of philosophical logic in matters of truth was challenged in the name of thetotic and poetics. This movement can be traced from Isocrates through Cicero and the Italian humanists up to the background of modern humanities, but the ontological ptesuppositions of Theophrastus' distribution wete not overcome. A s in the case of Aristotle's model of the signfunction, they had to be renewed in a rigorous form in modern philosophical analysis of language, namely, in the distinction, already mentioned, of C . Morris and Carnap between the semantical and the pragmatical dimension of the sign-function.
18

Here at last the shortcomings of Theophtastus' separation came to light. Fot, under the conditions of a constructive semantic in the style of Camap or Tarski, it was no longet possible to leave unexamined the pragmatical implications of the verification ot falsification of propositions, which Theophrastus could overlook by tacitly conceiving the so-called things as interpreted in the light of the Greek language. Now it turned out that a strictly semantical c o n c e p t i o n of t r u t h , defined by abstraction of the sign-re/erence from the pragmatics of meaning as the possible sign-interpretation, is void of any content, as is shown by Tatski's famous scheme of definition: "The proposition 'the things are so and so' is true if and only if things are so and so." If we illustrate this scheme by saying, for instance,

90

T H E IDEA O F A FIRST PHILOSOPHY

"The proposition 'it is raining', is true precisely if it is raining," it becomes immediately clear that the decision whether it is raining implies interpretation of the meaning of the word "raining." That is, of course, no problem in ordinary language. But what about "radioactive raining"? In short, Carnap had to concede that verification is a pragmatical problem which contradicts, strictly speaking, the distribution made by Theophrastus.
19

So here again the rigorous reconstruction of the traditional approach in the framework of modern language analysis suggests a totally new approach in First Philosophy as still has to be shown.

LANGUAGE-COMMUNICATION A N D T H E MODERN ERA O F EPISTEMOLOGY AS FIRST PHILOSOPHY

Let us now pass to our second heuristic question: What may be saidin the light of the philosophy of languageabout the ttansformation of ontology as First Philosophy into an epistemological analysis of consciousness .
1

The key for understanding this turn which becomes obvious with Descartes and Locke and which finally leads to Kant's t r a n s c e n d e n t a l philosophy as a new foundation of First Philosophy, the key for understanding this turn in the light of philosophy of language lies, as 1 see it, in the nominalistic movement of the late Middle Ages, or, to be more precise, in the emancipation of European philosophy and science from the verbal authority of the Scholastic ttadition of the ancient worldview. In order to simplify this complicated process in an appropriate way it is perhaps necessary to start out with a point made by A . Toynbee: that the culture of the European Middle Ages was an offshoot, indebted to the ancient authors and to canonical texts for the whole of their religious and secular knowledge about the world. Hence the painstaking elaboration of a subtle logic of language ( D e proprietatibus terminorum and De modis significandi) as a methodological foundation of scholastic ontology. In a sense it may be true that the Greeks unconsciously borrowed their ontological categories from their native language; it is, howevet, literally true that the Scholastic doctors derived their ontological categories of world-interpretation quite consciously and methodically from the canonical texts laid down in Latin as a universal l a n g u a g e . They proceeded in a much more literal sense than Aristotle from understanding a language to understanding the world; but they have also far more rigorously called in question the truth or reality of all universal concepts of language. The later Scholastics in a way were compelled to examine a verbally fixed world-picture by its application to experience; and the final result of this comprehensive test was the doctrine of the immediate intuition of the individual, as a content of consciousness preceding all use of languagea doctrine which has determined modern philosophy from Ockham through Descartes and the Btitish empiricists up to Kant. By this doctrine the turn from ontology to epistemological analysis of consciousness was made.

THE

IDEA O F A FIRST P H I L O S O P H Y

91

Now this fundamental turn to a new Fitst Philosophy proves to be highly ambivalent if we try to judge it in the light of the philosophy of language. O n the one hand nominalism was a progtessive and indispensable step in the histotical process of breaking through the naive language-realism of the eatly thinking of mankind. Seen in this perspective, nominalism has opened the way fot a new interpretation of the wotld by the mathematical language of modern natural science, and at the same time it made it possible to view the variety of languages in a detached and objectifying way as a vatiety of sign-systems. (This may be illustrated by the wotk of the Btitish empiricists as well as by the wotk of Leibniz, who in a way anticipated the modem logistic analysis of language.) But, on the othet hand, nominalism also set fotth epistemological docttines which added to the shortcomings of ancient ontology with regard to the function of language: Here, in the first place, the conception of cognition as an intuitive function is to be named which, in principle, is independent of language and only subsequently needs the use of signs fot fixing memory and for communication. Connected with this conception, from the beginning up to now, was the conception of methodical solipisism, that is, the belief that man could attain an undetstanding of the data of his consciousness including an undetstanding of himself as an "I" withoutin ptincipleptesupposing himself to be alteady socialized in a communication-community. The philosophy of language implied by nominalism, and especially by methodical solipsism as an aspect of nominalism, became explicit in the following passage of John Locke's Essay C o n c e r n i n g H u m a n U n d e r s t a n d i n g (11.2.2): "Wotds, in theit primary or immediate signification, stand fot nothing but the ideas in the mind of him that uses them, howevet imperfectly soevet ot catelessly these ideas ate collected from the things which they ate supposed to teptesent." In this text, as mentioned above, a psychologistic antithesis to Aristotle's ontological conception of meaning is offetedan alternative which Leibniz and the eatly Wittgenstein ttied to avoid by postulating an ideal ontological ftamework of language by which everybody is a priori confronted with the same world. Locke's epistemological approach, strictly speaking, accounts only fot ptivate languages; and no consistent connection seems to be possible between this apptoach and the insight, fotmulated by the same Locke in an othet place of his Essay (III. 2.8), that "common use, by a tacit consent, appropriates cettain sounds to certain ideas in all languages, which so far limits the signification of that sound that, unless a man applies it to the same idea, he does not speak ptopetly." Methodical solipsism, regardless now of whethet it is conscious ot not, is also found in the work of Descartes, who, in his methodical doubt, does not teflect on the fact that he, even in questioning all tradition, is still making use of a language game of atgumentation which is, in principle, a public one. So, for instance, he does not teflect on the fact that, in a consistent language game, it is not permitted to use the phrase "merely my dream" in such a universal way that it would make sense also in the hypothesis "perhaps the whole world is merely my dream." _

92

T H E IDEA OF A FIRST PHILOSOPHY

T h e same k i n d of an appeal to the semantical consistency of a language game also provides, i n my o p i n i o n , a telling argument against the radical assumption of a nominalistic pragmatismthe view that the function of the universal concepts of language might be reduced to that of a conventional arrangement i n the service of practical life without any truth-claim. A n o t h e r matter is a critical nominalism i n the sense of fallibilism w h i c h a s C . S. Peirce s h o w s m a y well be brought to agreement w i t h a critical r e a l i s m . A f t e r a l l t o sum up these argumentsone cannot, i n principle, dispute the truth-claim of language universals by argumentation w i t h the help of language.
20

N o w , insights such as these reveal, 1 believe, the characteristic illusion or self-misunderstanding of modern, nominalistic epistemology. T h e y show that this approach to a First Philosophy from a standpoint of a methodical solipsism did not reflect upon language as a condition of the possibility and of the intersubjective validity of all knowledge, including even the critique of knowledge. T h i s last remark and its formulation leads us to see also the deep ambiguity of Kant, who explicitly replaced the traditional ontology as First Philosophy by transcendental philosophy as critique of reason. For, o n the one h a n d , K a n t asked for the first time precisely the question as to the c o n d i t i o n of the possibility and intersubjective validity of knowledge; o n the other hand, however, he does not take account of language as such a c o n d i t i o n . H e does not say a wotd about language i n the C r i t i q u e of P u r e Reason and he mentions it only i n his " A n t h r o p o l o g y from a pragmatic point of v i e w , " whete it is called an important medium of understanding the wotld and o n e s e l f w h i c h is, of course, very true. Since, however, K a n t does not reflect upon language as a c o n d i t i o n of the possibility of knowledge and of the critique of knowledge his undertaking of a critical First Philosophy ultimately falls a v i c t i m to the basic fallacy of modern nominalism. For K a n t believes that he could at the same time dispute, in principle, the possibility of k n o w i n g something about things-inthemselves and speak about things-in-themselves as causes of the world of experience. If K a n t had recognized that the synthesis of apperception is always a function of interpretation mediated by signs, he could not have believed that his distinction between cognition of a thing and mere thinking of a thing could justify his speaking about the function of unknowable things-in-themselves as causes of sense-data.
2 1 22

Moreover, i n another respect, K a n t also would have h a d to question the apodictic c l a i m of deducing the categories of possible experience from his so-called table of logical forms of judgment; he would have had to take into account the syntactical and semantical categories of language as forms of possible experience, as was suggested later o n by W . von H u m b o l d t . A s it stands, K a n t has, in his conception of transcendental philosophy as First Philosophy, practically reconfirmed the classical position of Aristotle, according to w h i c h the variety of language is only a variety of sounds used as signs. Nevertheless, by his way of asking the question for the conditions of the possibility and intersubjective validity of cognition, K a n t d i d , in fact, also open the way for a

THE

IDEA O F A FIRST P H I L O S O P H Y

93

questioning of his own conception of a First Philosophy by a philosophy of language. This leads us to our third heuristic questionas to the reasons for the ttansformation of critical epistemology into critical analysis of language as the latest fundamental turn in the history of philosophy. The program of a critical questioning of epistemology by philosophy of language is alteady suggested by the conception of a Metafcritik of the ctitique of pute teason to be found in Kant's contemporary critics ]. G . Hamann and ]. G . Hetdet. Thus Hamann writes: No deduction is necessary to show that language is the otiginal ancestot in the genealogy of the seven sacted functions of logical premises and conclusions. Not only is the entire possibility of thinking founded in language . . . but language is also at the centet of teason's misunderstanding with itself, in patt due to the ftequent coincidence of major and minor terms, to its vacuity and abundance of ideal phtases, in patt because of the infinite number of verbal figures in respect to syllogistic ones.
23

"Without language," Hamann temarks, "we would not have teason," he goes so far to say: "Reason is language, namely Xoyoc;."
25

24

and

Hetdet, in his "Metakfitik det teinen Vernunft," followed the line of argument of his mentot and friend Hamann, especially that atgument of Hamann's by which he suggests that the problem of the transcendental unity of sensation and undetstanding as toots of cognitiona problem exposed by Kantis always alteady solved by the unity of wotld-interpretation attested to by language. Howevet, it has to be tematked that Hetder understood Kant's ttanscendental question even less than did Hamann and, thetefote, was unable to fulfill the tole of a "Metakfitik" of Kant's critique. (Latef on C . S. Peirce showed what this means: to transpose the Kantian problem of the unity and difference of sensation, intuition, and undefstanding into the framewofk of language analysis. He illusttates this by minute semiotical analysis of the intetaction of the functions of "symbols," "indices," and "icons" in a synthetic logic of hypothesis and confifmation.)
26 27

The merits of Herder in founding a philosophy of language lie in anothet ditection: As an anthtopologist and philosopher of history he was the first since G . B. Vico to have shown in detail what Hamann called the "chronological a ptiori of language"that is, he showed the history of languages as accumulations of different world-interpretations into a condition of the possibility of actual expetience and intetsubjective understanding. Thereby Herder, as Vico in his S c i e n z a N u o v a had done, outlined the program of humanistics as a new btanch of science, which in fact may be televant fot the foundation of philosophy itself, if the a priori of language is a transcendental ptesupposition of philosophy itself.
28

The next stage in the ttansformation of the epistemological philosophy of consciousness into the language-analytical ot langiwge-hermeneuticoT philosophy of our time is the wotk of W . von Humboldt. He is at the same time the successor

94

T H E IDEA O F A FIRST P H I L O S O P H Y

of Herder and the forerunner of a philosophically relevant foundation of linguistics. In conscious opposition to the whole traditional conception of language since Aristotle and by anticipating in a sense the conceptions of N . Chomsky he defines the essence of language as follows: Language must be considered as immediately laid down in man; for as a product of his understanding in the clearness of consciousness it is completely unexplainable . . . language could not be invented if its type did not already exist in the human understanding. In order to enable man to understand only one single word . . . as an articulated sound which designates a concept, the whole of language in its structure must already lie in him. There is nothing single in the language, everyone of its elements announces itself only as part of a whole. . . . The language necessarily comes forth from man, and certainly only gradually, but so that its organism determines the functions of the faculty of thinking, not, of course, as a dead mass lying in the dark of the soul but as a law. . . . If therefore this phenomenon which has no equivalent in the whole realm of the thinkable, is to be compared with anything else, one may well call to mind the natural instinct of animals, and one may call language an intellectual instinct of reason.
29

Language is no work (ergon), but an activity ( e n e r g e i a ) . Its "true definition, therefore, can only be a genetic one. It is the continually repeating activity of the mind to enable articulated sounds to express thoughts."
30

While, on the one hand, these definitions of the human faculty of language, not surpassedI would sayup to now, may be considered as a hint pointing toward the generative g r a m m a r , on the other hand, Humboldt also anticipates a program of comparative linguistics which would be relevant to epistemology. (One may think today of metaiinguisrics in the sense of Sapir and B. L. Whorf or of "inhaltsbezogene Sprachwissenschaft" in the sense of L. Weisgerber). So Humboldt writes: By the interdependency of thoughts and words it becomes quite clear that languages are to be considered not so much as means for representing the already known truth, but rather as means for discovering the previously unknown truth. Their variety is not one of sounds and signs, but a variety of world views.
31

The sum total of the recognizable as the field to be operated upon by the human mind lies between all languages and independent of them in the middle of them; man can approach this purely objective realm only according to his way of cognition and sensation, that is, in a subjective way. . . . But the subjectivity of all mankind in itself becomes again something objective. So the original agreement between man and the world, on which the possibility of all knowledge rests, is regained piecemeal and progressively.
32

These statements of Humboldt's obviously allude to Kant's critique of knowledge, and they, in fact, give a hint as to how Kant's problem of the unknowable things-in-themselves could be solved by a transformation of his critique in the

THE

IDEA O F A FIRST P H I L O S O P H Y

95

light of the philosophy of language. This transformationas I interpret Humboldt's remarkswould integrate the functions of Kant's finite consciousness into the virtually infinite faculty of interpretation and criticism of interpretation which is the faculty of all mankind seen as a language community. O f coutse, it has to be admitted that Humboldt's conception is too restticted to the comparison of languages as collective wotldviews to provide the basis fot a ditect challenge to Kant's ctitique of knowledge. In this respect, however, Humboldt's approach may be supplemented by the semiotical transformation of Kant's critique by C . S. Peirce.
33

Peirce in a sense understood himself as a Kantian; but, fat mote thoroughly than even the neo-Kantian E. Cassitet in his Philosophy of Symbolic F o r m s , Peirce drew all the consequences from the insight that the synthetic function of cognition is mediated by signs. Accotding to Peirce, it is a function of an hypothesis (that is, an abductive infetence) which in the dimension of intersubjectivity has to be supplemented by the interpretation of signs. (The complementarity between the cognition of things and the intetsubjective communication by interpretation of signs was later on further clarified by ]. Royce.) So Peirce finally reached two important results by his semiotic transformation of Kant's critique of knowledge:
34

1. Kant's "highest point," which is the ttanscendental unity of the consciousness of objects and of self-consciousness, has to be replaced by the postulate of a final unity of sign-interpretation which would be the consensus about truth, teached in an indefinite community of investigatots and interpretets. This community is, according to Peiice, the subject of the process of cognition, a subject whichin contradistinction to an individual consciousnesscould keep up in principle with the indefinite Reality as it is in itself. 2. By this last remark also the second tesult of Peirce's semiotical transformation of Kant's critique is indicated. Kant's talking about a principally unknowable thing-in-itself shows itself as conttadiction if cognition is conceived as hypothesis mediated by signs; and so, instead of Kant's distinction between unknowable things-in-themselves and knowable phenomena, Peiice introduces anothei fundamental distinction: that between the real as the indefinite know* able and what can actually be k n o w n . So a finite subject of cognition, also according to Peirce, cannot know the leal in itself. Moreover, Peirce thinks, in contradistinction to Kant, that scientific knowledge is, in principle, also fallible. Nevertheless, he need not talk about unknowable things in themselves; for, under the presupposition of a superindividual, communicative function of cognition, there exists a subjective equivalent of the infinite reality as object of cognition, namely, the indefinite power of cognition of the indefinite community. Later on ]. Royce tried to work out the idea of an indefinite c o m m u n i t y of interpretation as a semiotic foundation for the social sciences and the humanities as well. And, in a sense, G . H . Mead's symbolic interactionism, combined with the idea of universal discourse, could have been considered as the culmination point
35

96

T H E IDEA OF A FIRST P H I L O S O P H Y

of semiotic pragmatism as a transcendental pragmatics of c o m m u n i c a t i o n had it not been misunderstood, partly by M e a d himself and especially by his follower and editor C . W . Morris, as a special k i n d of behaviorism, namely, social behaviorism. In fact, a closer examination of Mead's posthumous work M i n d , Self a n d Society shows, i n my o p i n i o n , that M e a d is not reducing the phenomena of language-communication to some collective aspect of observable behavior but rather the other way around: H e i n fact shows that human behavior as symbolic interaction can only be adequately recognized and accounted for if the involved reciprocity of expectations is understood, w h i c h in turn can only be achieved by reflective participation i n interpersonal communication. By showing the simultaneous origin of understanding other people by placing oneself into their situations and understanding oneself as 1 by being oneself w i t h the eyes of others, M e a d indeed surpassed the o l d idea of introspectionism as a function of methodical solipsism; but he did so not by substituting external observation of behavior for internal observation or introspection, but by transcending this whole alternative of Cartesian epistemology, that is, by showing that selfreflection and understanding other people are two sides of the same phenomenon. In this respect, his position is very close to Dilthey's conception of h e r m e n e u t i c s as a reciprocal deepening of h u m a n self-understanding and understanding of the geschichtlich-gesellschaftliche W e l t and to Heidegger's conception of existential h e r m e n e u t i c s o n the basis of a simultaneous origin of Mitsein and S e l b s t v e r s t d n d n i s . In fact, G . H . Mead's analysis surpasses the C o n t i n e n t a l tradition of Hermeneutics w i t h regard to the social concretization and biological prefiguration of symbolic i n t e r a c t i o n f o r example, in the process of socialization and especially by showing that the "significant symbols" of language, so to speak, function as an institutionalization of the reciprocity of human interaction, and thus of understanding and self-reflection.
36

W i t h this suggestion of a convergence of A m e r i c a n semiotic pragmatism and the G e r m a n tradition of hermeneutics, we have crossed the threshold of the twentieth century where analysis of language i n the broadest sense was, i n fact, to take over the role of First Philosophy. However, as we already have mentioned, so called analytic philosophy i n its first stages (namely, in the conceptions of logical atomism and logical positivism) was to renew i n a more rigorous form the basic A r i s t o t e l i a n conception of the relation between language and things. Ludwig Wittgenstein, i n his later lectures, was the first to see that in logical a t o m i s m w h i c h implicitly provided the metaphysical foundation also for logical p o s i t i v i s m a n archaic idea of language, namely the o n t o s e m a n t i c a l idea of words as names and languages as combinations of names corresponding to combinations of things, was at the same time presupposed and reduced ad a b s u r d u m if taken as a model of natural language. But let us examine more closely the idea of language-analysis as First Philosophy.
37 38

THE

IDEA O F A FIRST P H I L O S O P H Y

97

L A N G U A G E - C O M M U N I C A T I O N A N D T H E IDEA O F L A N G U A G E - A N A L Y S I S AS FIRST P H I L O S O P H Y

The conception of language presupposed in thefirsttwo stages of analytic philosophy may be considered, in my opinion, as a combination of the language-conception of British empiricism with that of mathesis universalis stemmingfromLeibniz. The idea of language connected with the idea of mathesis universalis and thus with the rise of symbolic logic does not, as Locke's idea of language, start out with the conception of words as signs of "private ideas" but with the conception of words as Rechen?etteln (Leibniz) as functions of a language of calculation which is a priorithat is, by its logical formintersubjective. Hence, elimination of uncleamess and misunderstandings in science and philosophy is not expected to be reached by intuitive ascertainment (by subjective-solipsistic introspection) of the evidence of signification, but by the syntactico-semantic consistency of the language-system which as a ready-made "calculus ratiocinator" would allow its users to reduce all rational discussions to a "calculemus," that is, to "blind" or "symbolic" thinking, as Leibniz says, which need not reassure itself intuitively of its semantic contents. Now, the "aporia" of this approach became apparent, as I see it, in the T r a c t a t u s logico-philosophicus of the early Wittgenstein. It may be interpreted as polar opposite to the aporia of empirism as solipsism.
39

If one presupposes, as the early Wittgenstein does, that ordinary language disguises the logical form of a universal languageso that the logical form of deep grammar would provide an intersubjectively valid representation of all elementary facts by elementary sentences and a logical reduction of meaningful sentences to elementary sentences-if one adopts this supposition, then, to be sure, the difficulty raised by Locke, the problem of how private ideas as meanings of words can be communicated and thus private experience may be intersubjectively valid, no longer arises. But the elimination of the problem is provided by the fact that personal experience and communication by language are now completely separated from the constitution of the meanings of words. Meanings are presupposed in the onto-semantical system of language as objective "substance" of the world which has only to be named by the users of language as elements of possible "states of affairs." Now, since the "form" of "states of affairs" (the logical form of language and of the world) is also a priori the same for all users of the language, the problem of solipsism is dissolved by the hypothesis that, without presupposing implicit agreements about the meaning of words and thus about the significance of things brought about by the very process of language-communication, every user of the (universal) language is a priori confronted with the same world.
40

As Wittgenstein puts it himself: Here we see that solipsism strictly carried out coincides with pure realism. The I in solipsism shrinks to an extensionless point and there remains the reality coordinated with it. ( T r a c t a t u s 5.64)

98

T H E IDEA O F A FIRST P H I L O S O P H Y

If one does not wishas it might be suggested by the Tractatustogether with the problem of communication about meaning of words, to eliminate the whole problem of subjective meaning-experience and intentionality, then only the following interpretation of these phenomena, as far as 1 can see, might be connected with the early Wittgenstein's solution of the problem of intersubjectivity: Communication has to be conceived of as a process of transmission of messages between persons who for that purpose have to encode their private thoughts by means of a common sign-code or language-system which by its structure guarantees intersubjectivity of meaning. This implies, however, as Moritz Schlick has shown, that only the a priori intersubjective f o r m or structure of states of affairs, which is represented by the form or structure of sentences, can be communicated, whereas the content of the message can only be won by private interpretation which is not influenced by the structure of language and does not, on its part, influence that structure.
41

Of course, the difficulties of this model of language-communication, which provides a transition from mathesis universalis to technological theory of information, become most obvious if one takes into regard the well-known phenomena of historical interference between the development of language-systems, social communication, and experience within the process of sociocultural evolution. But, even if these historical experiences with natural language could be disregarded, the idea of language as a universal o n t o - s e m a n t i c system (or deep s t r u c t u r e of all systems), as it is suggested by the early Wittgenstein, seems to be, in principle, incompatible with the structure of human communication or speech. The main point of this incompatibility has been made visible by Wittgenstein himself, that is, by the famous paradox inherent in the structure of his T r a c t a t u s . To speak or communicate about (the structure of) a language which a priori by its structure guarantees intersubjectivity of information about the world seems to be neither necessary nor possible. Since every participant in a communication must already presuppose for himself the structure of language in order to say something about something, he need not and cannot say something about that structure: It only "shows itself," as Wittgenstein puts it. 1 would like to stress here that the idea of a separation between object-language and (an endless hierarchy of) metalanguages which was introduced by B. Russell and elaborated upon by Tarski and Carnap does not dissolve the paradox of the Tractatus, as it was hoped by B. Russell. For this fundamental presupposition of logical semantics is, from the beginning, restricted to artificial frameworks which themselves need interpretation by natural languages which, so to speak, must fulfill the role of a last metalanguage not foreseen in the hierarchy of metalanguages of logical semantics. Now, with regard to this actually last metalanguage of interpretation Wittgenstein's problem recurs, and the paradox of the Tractatus would in fact reappear, if in fact natural language could, or ought to, be considered as a ready-made instrument of intersubjective worldrepresentation which need not be mediated a priori by communication about world-representation and thus by self-reflection.

THE

IDEA O F A FIRST P H I L O S O P H Y

99

Now, from an anthropological point of view one could postulate that the telationship specific for human beings between language as a system, communication by speech-acts, expetience as wotld-interpretation mediated by signs, and practice of life regulated by signs should be possible only if natutal language as a system does not exclude self-reflection but is, at least partly, established and reestablished as an institution by implicit self-reflection, inherent in communication. For, in conttadistinction to the function of so-called "languages of animals" or of "signal-codes," language-communication between men may neithet be sufficiently conceived of as pure ttansmission of messages (information) about states of affairs, which does not touch upon the subjective modes of interpreting the significance of experience, nor may it be thought of as ptivate actualization of a system-structure which would, in the long run, leave untouched the syntactico-semantical structure of the language-system. A n d these two statements are obviously concerned with two aspects of one and the same structure of language-communications; fot the possibility and necessity of interpersonal agreements about the significance of the objects of experience and the possibility and necessity of interpersonal agreements about the meaning of language-signsalready on the level of wotdsmay be considered as an expression of one and the same reflectivity of human intelligence. This reflective intelligence ot Besonnenheit, as J. G . Herder called it, may be undetstood, according to Hetder and A . Gehlen, as a kind of compensation for the lack of instincts characteristic of human beings. Not compelled in an unequivocal way to react on signals as behavioral stimuli, but, rather, exposed to the openness of significant situations, man is, from the beginning, engaged in interpreting the wotld with the help of language and, at the same time, in reconstructing the semantical system of language with the help of his interpretation of the wotld as a significant situation for action. Of course, successful communication, especially information about a virtually infinite set of new states of affairs by means of a finite set of signs, would also be inexplainable if no relative stability of an objective world-interpretation on the level of wotds, and thus of the semantical structure of language as a system, could be presupposed. A n d , certainly, the genesis of language as a system cannot be undetstood with the help of the theory of convention or invention, which in fact is an idealization of the conception of language as institutionalized communication. It is in this place where W . von Humboldt's idea of a human instinct of language, which was elaborated in our time by N . Chomsky and Lenneberg, could provide a solution, if it can be mediated with the idea of reconsttucting the language system as an institution of communication. But such a solution would presuppose a foundation which had to ttanscend, in principle, the alternative of empiricism, in the sense of Locke, and rationalism, in the sense of mathesis universalis. For, as little as a solipsistic conception of designating one's intramental ideas may explain the intetsubjective validity of meanings and thus of objective experience and rules of action, even that may be achieved by a conception of a language-system which is supposed as intersubjec-

100

T H E IDEA OF A FIRST PHILOSOPHY

tively valid without being mediated by its actual use i n c o m m u n i c a t i o n and i n organizing the meaning-intentions of the single members of the language c o m munity. T h e patadox of the neo-Leibnizian construction of an ideal languagesystem, according to my o p i n i o n , lies i n the fact that it allows us to t h i n k of speech or communication only as a private and automatic actualization of a preestablished system-sttuctute; now, since such a type of use by definition cannot influence the structure of the system, it cannot explain just the c o n s t r u c tion of an ideal language-system w h i c h is nothing else than a l i m i t i n g case of reflective and reconstructive use of language. ( A n d so long as linguistic structuralism conceives of "parole" or performance only as actualization of a ready-made language-system it shows its internal dependence o n neo-Leibnizianism and, consequently, participates i n its aporetics.) T h e previous confrontation of the aporetics of the empiristic-solipsistic model of language w i t h the logistic model of language shows, as I t h i n k , that (also) a simple combination of both models, as it was intended i n logical positivism, is not enough to cope with the problem of relationship between system-structure and actual communication i n natural language. O n e of the reasons for the deficiency of the neopositivist combination is shown by the conception of Moritz S c h l i c k already alluded to: From his combination of the logistic concept i o n of form or structure w i t h private interpretation of c o n t e n t it may be learned, I would suggest, that the methodical solipsism inherent i n the nominalist tradition of empiricism is by no means surmounted by the Leibnizian idea of a language-system w h i c h would be a priori intersubjective by its syntactical and onto-semantical structure. Indeed, it was a progressive s t e p w h i c h , as was mentioned already, connects Leibnizianism w i t h modern linguistic structuralismto replace the A r i s t o t e l i a n idea of extralinguistic meaning and Locke's psychologistic transformation of this idea by the conception of meaning as function of a language-system; but so long as the conception of the system is only based o n logic and ontology (or "onto-semantics")that is, so long as it cannot be thought of, at the same time, as institutionalization of human communication (and thereby as objectification and alienation of thought, w h i c h is n o t h i n g else than internalized language-communication) it must leave the problem of meaning-intentions outside the realm of language, and thus must deliver it once again to the prelinguistic conception of methodical solipsism underlying contemporary epistemology. Speaking with W . v o n H u m b o l d t , one could say that a languagesystem in the sense of neo-Leibnizianism is lacking energeia which has its pendant in the fact that creative thought, w h i c h has to provide the e n e r g e i a of language, is conceived of without presupposing "language as its forming o r g a n . " Thus, language-communication, not being mediated dialectically w i t h the languagesystem o n the one hand and w i t h personal t h i n k i n g o n the other hand, cannot be conceived of as a ttanscendental c o n d i t i o n of possibility of thought and knowledge but (rather is thought of) only as a process of transmission of private thoughts encoded i n the public medium of a language-system. T h i s technologi-

T H E IDEA OF A FIRST P H I L O S O P H Y

101

cal reduction misses the point of language-communication i n principally the same way as Aristotle does, if he finds the paradigm of language-variety i n the variety of artificial letters; fot, i n the case of writing as in the case of encoding, it is presupposed that language as "forming organ" of thought functions alteady as an unreflected condition fot the translation between ptivate thought and the public code. T h e situation so fat chatactetized (as combination of the idea of an a ptioti intersubjective language-system with the idea of thought i n the sense of methodical solipsism) was not essentially changed w h e n i n a second phase of logical p o s i t i v i s m t h e problem of actualization of the language-system was no longer to be solved by the assumption of a private interpretation (by introspective experience) but rather by the postulate of an objective description of the use of language as overt behavior. For, if i n a strict sense of behaviorism, the understanding of intentions i n people's speech-acts is replaced by external observation and description of the data of observation, the observer must give up his prescientific relation of interpersonal communication w i t h those people he has to observe. T h i s reduction of co-subjects of communication to mere natural objects, by the way, makes it impossible, i n principle, to decide whether speech is the object of observation or n o t n o t to speak of the question of what rules of speech are f o l l o w e d . T o this extent, linguistics or the social sciences are impossible on a strictly behavioristic basis, i n my o p i n i o n . Besides this, however, it may be shown that a strictly behavioristic science of the use of language would not, as is usually assumed, overcome the solipsistic implications of traditional empiricism but, quite to the contrary, would reconfirm them. For, i n order to thoroughly avoid the problem of understanding so-called private meaning-intentions and meaning-interpretations, the behaviorist must universalize his attitude toward men, as for instance Skinner i n fact does. T h i s means that he must also replace communication w i t h his colleagues by, instead, observing and describing their verbal behavior. N o w , i n this case, w h i c h is i n fact the point of scientistic reductionism, the observer becomes a methodical (if not a metaphysical) solipsist; for he would need a language of science w h i c h he, in principle, must be able to use privatim, that is, without presupposing a communication-community. For this "lonesome" behaviorist solipsism and realism must, indeed, " c o i n c i d e , " as the early Wittgenstein had postulated i n the T r a c t a t u s .
42

T h i s consideration shows, i n my o p i n i o n , that behaviorism and introspectionism as combined w i t h the logistic idea of an ideal language-system are two c o m plementary aspects of methodical solipsism w h i c h underlie modern scientism. W h a t is overlooked i n both cases is the a priori of language-communication as a transcendental c o n d i t i o n for understanding one's o w n intentions and experiences i n the light of other people's and vice versa. But this shows that at the ground of scientism and methodical solipsism lies the commonsense idea of language and communication as mere instruments i n the service of a prelinguistic faculty of t h o u g h t a n idea w h i c h we have traced back to its ancient origins.

102

T H E IDEA O F A FIRST PHILOSOPHY

It is this conception of language and communication w h i c h was reconstructed, on the highest level of precision, in the first two phases of modern analytic philosophy of language; and by this very reconstruction it was reduced ad a b s u r d u m , as I tried to show by displaying the aporetics of logical atomism and logical positivism. T h e later Wittgenstein was the first, it seems to me, to draw the consequences of the problem-situation so far characterized. H e d i d so, roughly speaking, by replacing the n o m e n - n o m i n a t u m model of logical a t o m i s m a n d thereby the onto-semantical language conception of the Greeks and the solipsistic meaningconception of contemporary epistemologyby the idea of "language games," and especially by the thesis that a private language would be impossible since nobody could "privately follow a r u l e . " In order to reconstruct the philosophical result of this approach in our present context it seems to me necessary, however, to think w i t h Wittgenstein against Wittgenstein and beyond Wittgenstein. Thus, fot instance, it does not suffice to replace the designation-model of languageand the corresponding objectified conception of m e a n i n g b y the postulate of describing the vatious functions of usealthough this postulate of Wittgenstein is fully justified as a pragmatic alternative, widening the heuristic horizon, to the onto-semantical and epistemological orientation of traditional philosophy of language. But, if one wete to understand the demand for description of use strictly i n the sense of e m p i r i c i s m o r even of behaviorism, as Wittgenstein sometimes seems to suggestthen it becomes impossible to justify the most important implication of the model of the language game: the thesis, already mentioned, that a "private language" would be impossible. If our previous criticism of behaviorism as a variety of methodical solipsism was tight, then from Wittgenstein's insight that (meaningful talk about) following a rule is, i n principle, dependent o n the context of a public language game it follows that the describer of a language game must participate, i n a certain sense, i n the language game to be described. Otherwise the describer could never be sure that the rules w h i c h he supposes as followed i n his description are the same as those followed by the participants of the language game. Peter W i n c h was tight, I think, to derive from this insight the possibility and necessity of a new, q u a s i - h e r m e n e u t i c , foundation of the social sciences. T h o u g h he confuses, it seems to me, the phenomenon of "interwovenness" of language-use, bodily expressions, social activities, and world-interpretation w h i c h is, according to Wittgenstein, the c o n d i t i o n for a child's learning a language and being socialized into a "form of l i f e , " w i t h the ideal postulate of internal relations (of reciprocal interpretation) between these moments as they ought to be realized in an ideal language game of an ideal communication community. T h i s confusion misleads h i m to hypostatize, o n the line of an idealism w h i c h is also relativism, the factual language games and forms of life thematized i n the social sciences as incommensurable quasi-transcendental frameworks of possible world-interpretations. Thus, i n his foundation of social science, he can neither explain the possibility of h e r m e n e u t i c understanding as a faculty of transcending and

THE

IDEA O F A FIRST P H I L O S O P H Y

103

mediating different language games and forms of life, nor can he justify critique of ideology, which, in fact, mediates dialectically between hermeneutic understanding of the social forms of life and a quasi-naturalistic explanation of those inconsistencies in the factual language games which can only be teduced to external (i.e., causal) telations between use of language, social practice, and actual wotldinterpretation.
43

Now, in the context of out question fot the possible tole of languagephilosophy as Fitst Philosophy, the most important conclusion from our (previous) considerations about the latet Wittgenstein concerns the philosophical language game. It follows ftom what we have said so fat that the philosophet himself, fot his so-called descriptions of language games, needs a language game which stands in a specific relation to all othet language games that may be conceived of; this relation must, at the same time, be a communicative one and a ciitical-ieflective one. Otherwise Wittgenstein's piogiam of a critique of language, manifesting itself in his talk of "nonsense" ot "idle facing of the machine of language," would be unintelligible. Hence a language game must be postulated by which, in principle, communication with all language games and forms of life is possible without getting dependent on the different, and eventually incommensurable, patadigms of the different forms of life; rathet the postulated language game must provide itself a patadigm ot ideal norm for judging all othet language games.
44

(This postulate seems to contradict Wittgenstein's thesis that nothing else is common to the different language games than a cettain "family-resemblance," that is, no essential featute that is a priori valid fot all of them. Indeed, the common essential featute of all language games lies, as it seems to me, in the fact that by learning one languagethat is, by learning one's native language in the context of socializationone also learns something like the deep structure of a universal language g a m e or human form of life. For, togethet with his linguistic competence, every native speaker acquires the furthet competence of transcending his own language game and form of life by reflection and by communicating with other language games. This argument is converging with anothet one which is not reflected upon, at least not explicitly, in Wittgenstein's discussion of the "private language" problem. Although it has to be admitted that following a rule is, in principle, a public affair and that it at least needs to be connected with an existing language game, it must nonetheless be possible, in principle, to inttoduce new rules which eventually cannot be controlled as rules by the communication-community on the basis of the patadigms of the existing language games. This is the case, I think, with all inventots and innovatots not [yet] understood by society, especially of visionaties [in a good sense] of new moral and social forms of life. Now, since in these cases we also are not allowed to suppose "private language games," we must postulate an ideal language g a m e of an ideal [indefinite] communication community as instance of judging or controlling the meaning of the rules followed by the revolutionaries of human form of life.)

104

T H E IDEA O F A FIRST P H I L O S O P H Y

It is this ideal language g a m e , not yet realized in the factual language games, that, in my opinion, is presupposed, although counterfactually, as the condition of the possibility and validity for u n d e r s t a n d i n g human forms of life: It is, at least implicitly, anticipated in all human actions claiming to be meaningful and it is explicitly anticipated in philosophical arguments claiming to be valid. I would like to call this ideal language game which can justify Wittgenstein's thesis of the impossibility of a private language the transcendental language g a m e , which would correspond to the idea of a communicative competence of man in the sense of a universal or transcendental pragmatics of language-communication.
45

THE

T R A N S C E N D E N T A L C O N C E P T I O N OF L A N G U A G E - C O M M U N I C A T I O N

The transcendental language g a m e of the indefinite ideal communication community which we have postulated with Wittgenstein against Wittgenstein defines, in my opinion, that basic conception which may be accepted, on the one hand, as the last presupposition of language analysis as critique of metaphysics and, on the other hand, as the basis for a transformation of the traditional idea of First Philosophy in the light of language-philosophy. As a last presupposition of a critique of metaphysics by language analysis, the normative idea of the transcendental language game may, for instance, lead to an adequate reconstruction of Wittgenstein's critique of Plato's ontological hypostatization of (the ideal unity of) meanings of words, which, it seems to me, cannot convince by recourse to the actual use of WOTCIS only. For, although it may be methodologically useful to replace the seducing ontological questions of the Platonic Socrates as to "what is courage" or "what is justice" or "what is truth" by (heuristic) questions regarding the use of these words (these questions providing at least a suitable method for destroying the metaphorical illusion that all words should have referents like proper names), it would nevertheless be unsatisfactory, if we were to accept a description of an actual use of words (which may eventually be based on sociometrical methods) as a definite answer to the question what we should adequately understand as meaning of the concepts "courage," "justice," or "truth." A solution of this problemand thereby a dissolution of the old syndrome of problems characterized by the philosophical terms "essence," "definition," "idea," "concept," and "meaning" might however be reached if one would not expect to get an immediate answer to the Platonic questions about the "essence" of things by description of the actual use of the corresponding words, but rather by postulating a consensus about an adequate use of the words which would be reached by all participants of a language game if they could discuss the problems long enough under the conditions of an ideal communication-community. In other words, although a philosophically relevant definition, in order to be intelligible, must always be connected with a given use of words (as also a philosophical argument must start out from accepted premises) it must nevertheless take account of the newest state of human experience and argumentation, thus anticipating in a given
46

THE

I D E A O F A FIRST P H I L O S O P H Y

105

language game the structure of an ideal language game that could be valid for all rational beings. However, such a normative interpretation of the thesis that the "essence" of things lies i n the use of words has to face the quasi-transcendental problem w h i c h is raisedaccording to W i t t g e n s t e i n b y the pluralism of language games (e.g., of those belonging to science and those interwoven w i t h pre- and extra-scientific forms of life), and besides t h i s a c c o r d i n g to W . v o n H u m b o l d t b y the "variety of the structure of human language." For, properly speaking, the "essence," according to Wittgenstein also, may be found not so much i n the actual use of words as i n the deep g r a m m a r regulating that use; and according to Humboldt (and B. L . W h o r f ) our understanding of the essential structure of the world is always already prejudiced by the different "worldviews" suggested by the different types of language-sttucture. H o w might this pluralism of possible systems of "internal f o r m " (Humboldt) or "deep grammar" ( W i t t g e n stein) be reconciled w i t h out postulate of a final consensus about the use of concepts i n an ideal language game? A r e not the different syntactico-semantical systems or types of deep grammar different ways of a possible formation of a consensus about rules of the use of words, so that it a priori makes no sense to expect or postulate a universal consensus about questions of meaning-rules and thus about questions of "essence"? T h e relativistic tendency of these objections is enforced, so it seems, by the consideration that the attempts, so far undertaken, to construct an ideal language of science d i d not lead to a lingua universalis sive pfiiiosophica, as it was postulated by Leibniz, but rather have reconfirmed the assumption of an a priori existing pluralism of possible "semantical frameworks." T h i s result seems to be i n harmony w i t h the conventionalism and pluralism of theoties or paradigms as it has been tecently defended as ultima ratio i n the philosophy of s c i e n c e .
47

W h i l e I admit that these questions may be considered as the most difficult problems to be faced by our transcendental conception of language-communication as basis fot First Philosophy, I w i l l try to cope w i t h t h e m by starting out from a historic-anthropological consideration: W h i l e it is true that, today as thousands of years ago, there are incommensurable differences of "internal f o r m " or "syntactico-semantical structure" between the systems or types of language, the corresponding differencesalways stressed by experts for primitive culturesbetween language games or sociocultural forms of life have not been preserved i n modern planetarian civilization. T h e variety of language games as parts or forms of life has not disappeared; it has been superformed or played over, so to speak, by the language game of science and technology w h i c h , i n spite of its o w n complexity, has brought about something like the unity of a form of life. It is remarkable, i n this context, that the semantic component of the human languages, notwithstanding the petsistent diversity of the system-structures, apparently was able to adapt itself to the process of unification o n the level of the scientific-technological language game. For the languages of the Far East (as C h i n e s e and Japanese) o n the one h a n d and

106

T H E IDEA O F A FIRST P H I L O S O P H Y

the European languages on the other hand, in spite of the diversity, or even incommensurability, of their structure as systems, seem to be able today to formulate by practically equivalent expressions nearly all essential ideas of scientific-technological civilization. Moreover, it is probable that even the intimate spheres of the different sociocultural forms of life whose verbal expressions are often considered as untranslatable may become indirectly interpretable, at least for purposes of a practical understanding between different civilizations, by h e r m e n e u t i c knowledge about the diversities of sociocultural structures including the inherent worldviews of the language-systems. What kind of interpretation may be coordinated with these remarks by a philosophy of language? It seems to me that the most important conclusion suggested by the history of understanding between human civilizations aims at a simultaneous distinction and dialectical meditation between syntactico-semantical language-systems and semantico-pragmatic language-games. While it may be possible to think of language-systemsespecially if they are idealized according to the paradigm of artificial frameworksas incommensurable conditions (frameworks, perspectives) of possible concept-formation, this view is obviously misleading with regard to language-gamesif these are understood as pragmatical units of communication or social interaction. In other words, while it apparently makes no sense to expect at any time something like a synthesis of the different methods of g e n e r a t i o n on the level of linguistic (grammatical ot, eventually, semantical) competence (in the sense of N . Chomsky) to come about, it is by no means unreasonable to expect something like coming to a better understanding between members of different language communities on the level of communicative competence (which, as every translation shows, is not only dependent on its preformation by the linguistic competence but, besides this, is dependent on pragmatic universals). A n d the historical progress of communication in the pragmatic dimension may even influence the semantical component of languages notwithstanding its dependence on different system-structures.
48

The reason for this possibility may be found, in my opinion, in the fact thatcontrary to the opinion of M . Schlickthe structure of natural languages is not independent of its pragmatic interpretation, in contradistinction to the form or structure of artificial language-systems. Thus, with regard to natural languages, it is possible, ot even necessary, to think, on the one hand, that by the coining power of their "internal form" (e.g., by the so-called semantical field-structures) they have in fact preformed through millenaries the use of language and thereby social interaction and world-interpretation, while, on the other hand, it may also be imagined that by the effects of successful communication in the pragmatic dimension of language-use the semantical system-structures of language may be modified in the long run. This reciprocity of influence may be interpretedfollowing Hegel and W . von Humboldtas a historic-dialectic interaction between the "subjective mind" and its alienation in the "objective mind"; and it may well be imagined that in the present era of history the communicative competence of man is coming home, so to speak, from its

THE

IDEA O F A FIRST P H I L O S O P H Y

107

objectifications and alienations in the institutional power of the languagesystems. Thus, it would be understandable that by the development of communicative competenceespecially by the development of the inherent competence of reflection upon the structure of language-systems and of pragmatical universalsthe so-called incommensurability of system-structures is approximately surmounted. In this context (of a historic-philosophical consideration) special attention should be paid to the fact that the transition from use of words to the use of concepts (mediated by definitions), going along with the rise of philosophy and the sciences, was a crucial innovation in the semantico-pragmatic level of languagecommunication. By the invention of concepts the meaning of language-signs was for the first time explicitly made a function not only of experience and of a certain language-system but also of logical operations valid for all languages. Hence it is not surprising, after all, that the Greek founders of philosophy and science tended to overlook the fact that the meanings of concepts mediated by definitionsnot to speak of the meanings of wordsalso remain dependent on the syntactico-semantical rules inherent in special language-systems and, besides this, on the semantico-pragmatical rules of certain language games. Thus the Greek founders of First Philosophy solved the whole problem of meanings by abstracting from the a priori of language-communication, that is, by an ontosemantic reduction of meanings to extralinguistic entities, as "ideas" (Plato) or intrapsychic impressions (jta0r|u,aTa) of things (Aristotle). First Philosophy thus far seemed only to be a matter of the "logos" and its onto-logical relations to the "essence" of things. Contemporary transformation of First Philosophy into epistemology insisted on the subjective evidence of the "ideas" in the soul qua consciousness and at the same times radicalized the illegitimate abstraction of the a priori of language-communication concealed in the h u m a n logos to the point of methodical solipsism. Finally, the latest transformation of the idea of First Philosophy in the name of language-analysis has reached a point, it seems to me, where it is able to cope with the old problem of essence and meaning by integrating the a priori of language-communication into the idea of the human logos. The transcendental conception of language-communication on the level of pragmatics or hermeneutics may show that, notwithstanding the indispensable mediation of meaningsand hence of all personal intentionsby the use of language, the ancient postulate of intersubjectively valid concepts of the essence of things may be fulfilled in the long run by the process of communication in the indefinite communication-community of rational beings, which was intended and also brought along in all civilized language-communities by the invention of discussion of concepts.

NOTES
1. Cf. K.-O. Apel, Die Idee der Sprache in der T r a d i t i o n des Hurnanismus v o n Dante bis Vico (Bonn: Bouvier, 1963); and also "Sprache als Thema und Medium der

108

T H E IDEA O F A FIRST P H I L O S O P H Y

transzendentalen Reflexion," Man and World 3 (1970): 323-37. 2. Cf. K . - O . Apel, "Szientismus oder transzendentale Hermeneutik," in Hermeneutik und Diakktik, ed. R. Bubner et al., vol. 1 (Tubingen: Mohr, 1970), 105-44, reprinted in K . - O . Apel, Transformation der Philosophie, vol. 2: Das Apriori der Kommunikationsgemeinschaft (Frankfurt a . M . : Suhrkamp, 1973). 3. Cf. J. Habermas, "Toward a Theory of Communicative Competence," in Recent Sociology, no. 2, ed. H . P. Dreitzel (New York: Macmillan, 1970), 115-48; reprinted in Inquiry 13 (1970): 205-18, 360-75. Cf. further K . - O . Apel, "Noam Chomskys Sprachtheorie und die Philosophie der Gegenwart," in K . - O . Apel, Transformation der Philosophie, vol. 2 (Frankfurt a . M . : Suhrkamp, 1973). 4. Cf. the essays in ibid. 5. Cf. Plato, Soph. 261c-262c. 6. Cf. Soph. 263d. 7. See ]. Lohmann, "Uber den paradigmatischen Charakter der griechischen Kultur," in Hermeneutik und Dialektik, Festschrift fur H . G . Gadamer (Tubingen: Mohr, 1970), 171-89; see further]. Lohmann's papers in Lexis 1 (1948): 49-106; Lexis 3, no. 1: 5-49; Lexis 3, no. 2: 169-217; and in Festschrift fiir L . Weisgerber (Dusseldorf, 1958). 8. So it is not quite surprising that the Neoplatonist tradition which interpreted Plato's Cratylus as defending the theory of the correctness of names had some beneficial influence by preserving the notion that words are not simply sounds arbitrarily used as signs. Finally, the strongest argument of the Seoet-theory of names was answered in the Neoplatonist tradition by the fruitful idea that the variety of words standing for the same things is not necessarily explained by different conventions but could also be explained by a variety of experienced aspects of things. This view may be traced in, for instance, Nicolaus Cusanus, Leibniz, and even in W . von Humboldt. Cf. K . - O . Apel, "Die Idee der Sprache bei Nicolaus von Cues," in Archiv fiir Begriffsgeschichte, vol. 1 (Bonn: Bouvier, 1955), 200-21. 9. The Greek text reads: " E c r u |J,EV o v v x a EV rrj wvfJT(I>v fev xfj ipuxf) ita0T}u.dTU)v at>|j,pcAa, K a i x a ypa(p6|j,eva xcbv EV if) (ptovfj. K a i w o i i E p o v b z y p a u . u . a x a Jtdot xa auxa ovbe tpwvaL <ov |J,EVXOI xaijxa ormeia Jtpwxov, x a i j x a Jta0fju.axa xfj;
HwXTJS- K a i (bv x a i j x a o u o i w u a x a u p d y u t r t a rjSr| x a r i x a . "

10. Cf. Tullio De Mauro, Ludwig Wittgenstein: His Place in the Development of Semantics (Dordrecht, Holland: D. Reidel, 1967), 8f. 11. Cf. below, p. 91. 12. In order to avoid Bradley's paradox of relation (according to which the idea of a real relation a R b presupposes further relations between R and the relata a and b, and so forth ad infinitum), L. Wittgenstein in his Tractatus (3.1432) postulates a language which does not represent relations by names (as if they were things) but by relations of names: "We must not say, 'The complex sign "a R b" says "a stands in a relation R to b"'; but we must say, 'That " a " stands in a certain relation to " b " says that a R b."' 13. Tullio De Mauro, Ludwig Wittgenstein; cf. also K . - O . Apel: "Wittgenstein und Heidegger," in Philosophisches Jahrbuch 75 (1967): 56-94, reprinted in Uber Heidegger, ed. O . Poggeler (Cologne: Kiepenheuer and Witsch, 1969). 14. Cf. L. Wittgenstein, Tractatus logico-philosophicus 5.6415. This problem is illustrated by the hypothesis of a first superhuman founder of names in Plato's Cratylus as well as by the eighteenth-century controversy about human or godly invention or institution of language (especially in Sussmilch, Herder, and Hamann). 16. See K . - O . Apel, Die Idee der Sprache. 17. Ammonius, In Aristotelis " D e Interpretatione" Commentarius (ed. A . Busse, 1887, p. 65), 31-66, 19.

THE

IDEA O F A FIRST P H I L O S O P H Y

109

18. Cf. K . - O . Apel, Die Idee der Sprache, esp. chap. 5. 19. Cf. R. Camap, Introduction to Semantics (Cambridge, Mass.: Harvard University Press, 1942), para. 38. 20. Cf. K . - O . Apel, "Introduction to C . S. Peirce," Schriften, vol. 1 (Frankfurt a . M . : Suhrkamp, 1967) and vol. 2 (Frankfurt a . M . : Suhrkamp, 1970). 21. I. Kant, "Anthropologic in pragmatischer Hinsicht," i n Samthche Werke, ed. K . Vorlander, vol. 4 (Leipzig, 1920), 101. 22. See below, pp. 95-96, about C . S. Peirce's semiotic transformation of Kant's critique. 23. J. G . Hamann, "Metakritik iiber den Purismus der Vemunft," in Samthche Werke, ed. ]. Nadler, vol. 3 (Vienna, 1949-57), 286. 24. Ibid., 231. 25. ]. G . Hamann, Letter to Herder of 10 August 1784, in Hamann's Schriften, ed. F. Roth, vol. 7 (Berlin, 1821-25), 151f. 26. See ]. G . Hamann, "Metakritik," 286; ]. G . Herder, Samthche Werke, ed. B. Suphan, vol. 21, 314ff. 27. Cf. K . - O . Apel, "From Kant to Peirce: The Semiotic Transformation of Transcendental Logic," in: Proceedings of the Third International Kant Congress, 1970 (Dordrecht, Holland: Reidel, 1972), 90-104. Cf. also my introduction to Peirce, Schriften, vol. 2. 28. Cf. K . - O . Apel, Die Idee der Sprache, Introduction. 29. W . von Humboldt, "Uber das vergleichende Sprachstudium," in Gesammelte Schriften, ed. Berliner Akademie der Wissenschaften (Berlin, 1903-36), vol. 4, 14ff. 30. W . von Humboldt, "Uber die Verschiedenheit des menschlichen Sprachbaues," 45. 31. Ibid., 27. 32. Ibid., 27f. 33. See K . - O . Apel, "Introduction to C . S. Peirce," and "From Kant to Peirce." 34. See ]. Royce, The Problem of Christianity (New York, 1913), vol. 2, 146ff. Cf. K . - O . Apel, "Szientismus oder transzendentale Hermeneutik?" in Hermeneuti/c und Dialektik, Festschrift for H . G . Gadamer (Tubingen: Mohr, 1970) vol. 1, 105 - 4 4 . 35. See ibid.; see also my paper "Communication and the Foundation of the Humanities," in Acta Sociologica 15, no. 1 (1972): 7-26; expanded version in Man and World 5 (1972): 3-37. 36. In the context of the German tradition of philosophical anthropology, A . Gehlen reconfirmed many of the ideas of G . H . Mead in his "philosophy of institutions." But Gehlen, stressing the necessity of stabilization of human behavior by social institutions, arrives at an anti-intellectualistic polarization between institutionalization and critical self-reflection. What helike many pluralistic Wittgensteinians identifying "language games" and "institutions"does not see is the fact that even human self-reflection and critique of institutions is always already institutionalized and in a way stabilized by language communications, for example, by argumentation as a philosophical institution. Thus language-communication is not only an institution among other institutions stabilizing interaction, but at the same time it is a metainstitution with regard to all special institutions, including language games. A s the meta-institution of all institutions, language-communication is not restricted to definite social groups and their languages, and so must not be considered as a function of a finite social system; but as an expression of the communicative competence of all human beings which is in itself a meta-communicative competence, it should rather be a priori related to an indefinite community of interpretation or of universal discourse, in accordance with Peirce, Royce, and Mead, and therefore it may be attributed a transcendental function on the level of theoretical and practical reason.

110

T H E I D E A O F A FIRST

PHILOSOPHY

Cf. my review "Arnold Gehlens Philosophie der Institutionen," Philosophische Rundschau 10 (1962): 1-21, reprinted in K . - O . Apel, Transformation der Philosophie, vol. 1. As a representative displaycontrary to Gehlenof Mead's idea of symbolic interaction on the level of social philosophy the work of J. Habermas may be considered; see especially his "Arbeit und Interaktion," in Technik und Wissenschaft als "Ideologic" (Frankfurt a . M . : Suhrkamp, 1968), 9-47, and his contributions to J. Habermas and W . Luhmann, Theorie der GeseUschaft oder Sozjaltechnologie (Frankfurt a.M.: Suhrkamp, 1971). 37. See especially Philosophical Investigations 1.46, where Wittgenstein traces the idea of logical atomism down to Plato's Cratylus. 38. In fact, the onto-semantical and isomorphic idea of language may be considered as a model for axiomatized and formalized languages of science serving as instruments for objectifying aspects of the world according to the Kantian device that the conditions of the possibility of experience (i.e., in this context, of description) are at the same time conditions of the possibility of the objects of experience. However, if Carnap's "Empirism, Semantics and Ontology" is interpreted in this way it becomes also clear that the onto-semantical idea of language always already presupposes a transcendental-pragmatical idea of natural language as a language of communication, for instance, as a language of conventions about construction and interpretation of special onto-semantical frameworks. 39. C f . , for example, L. Couturat, ed., Opuscules et Fragments inedits de Leibniz ( H i l desheim: G . Olms, 1966), 153ff. 40. Cf. L. Wittgenstein, Tractatus logico-philosophicus 2.02Iff. 41. Cf. M . Schlick, Gesammelte Aufsatze, 151-250. 42. Cf. N . Chomsky's " A Review of B. F. Skinner's Verbal Behavior," Language 35, no. 1 (1959): 26-58, reprinted in The Structure of Language, ed. J. A . Fodor and]. J. Katz (EnglewoodCliffs/N.Y.: Prentice Hall, 1964). 43. See: K.-O. Apel, "Communication and the Foundations of the Humanities"; cf. also my review of P. Winch's position in the last section of my "Analytic Philosophy of Language and the Geisteswissenschaften," in this volume (originally published by Dordrecht, Holland: Reidel, 1967); and my more thorough criticism of this position in K . - O . Apel, "Die Kommunikationsgemeinschaft als transzendentale Voraussetzung der Sozialwissenschaften," N e u e Hefte fur Philosophie 2/3 (1972): 1-40, reprinted in Apel, Trans/ormation der Philosophie, vol. 2. 44- A growing acknowledgment of this fact by Wittgensteinians and Ordinary Language Philosophers confronted with the claims of empirical linguistics is documented in C . Lyas, ed., Philosophy and Linguistics (London and Basingstoke: Macmillan, 1971). 45. See K.-O. Apel, "Die Kommunikationsgemeinschaft als transzendentale Voraussetzung." Cf. also J. Habermas, Zur Logik der Sozialwissenschaften (Frankfurt a . M . : Suhrkamp Verlag, 1970), 251ff. and his recent contributions to a theory of communicative competence (]. Habermas, Vorstudien und Ergdn^ungen zur Theorie des kommunikativen Handelns [Frankfurt a . M . : Suhrkamp, 1984]). 46. Wittgenstein himself in a way seems to confirm our critical view by the following remarkable consideration in his Bemerkungen uber die Grundlagen der Mathematik, 1, para. 74, where he questions his own reduction of Wesen (essence) to Ubereinkunft (convention): " U n d da mochte man doch entgegnen: es gibt merits Verschiedeneres als ein Satz uber die Tiefe des Wesens und eineruber bloBe Ubereinkunft. Wie aber, wenn ich antworte: Der Tiefe des Wesens entspricht das tiefe Bediirfnis nach Ubereinkunft." Should not the postulate of an indefinite community of communication correspond just to the deep need of convention in the traditional questions as to the essence of things? 47. As a representative documentation of the invasion of relativistic Wittgensteinianism

THE

I D E A O F A FIRST P H I L O S O P H Y

HI

and Whorfianism into the philosophy of science, one may consider the debate between T . Kuhn (and P. Feyerabend) and the Popperians in Criticism and the Growth of Knowledge, ed. J. Lakatos and A . Musgrave (Cambridge: Cambridge University Press, 1970). A n d it may be remarked in this place, that the prevailing logic of science, which considers the pragmatical dimension of argumentation in the community of researchers as a topic of empirical psychology, is hardly competent to deal with problems of scientific language games and their paradigms, which are apparently introduced as transcendental conditions of the possibility and validity of knowledge. Cf. to this, K . - O . Apel, "Zur Idee einer transzendentalen Sprachpragmatik," in Aspekte und Probhme der Sprachphilosophie, ed. ]. Simon (Freiburg: Alber, 1974). 48. Cf. J. Habermas, "Toward a Theory of Communicative Competence"; and K . - O . Apel, "Noam Chomskys Sprachtheorie und die Philosophie der Gegenwart."

FOUR Transcendental Semiotics as First Philosophy

EXPOSITION

In order to expose the main purpose of my lectute, 1 must point out that I do not intend to give a histotical account of First Philosophy. I intend tathet to display the systematic conception of transcendental semiotics as First Philosophy. 1 shall try to realize that aim through an ideal-typical reconstruction and ctitique of some historically given positions of philosophy. In taking the license of a rather simplified reconsttuction of those histotical positions of philosophy, I w o u l d like to account for the transcendental-semiotical presuppositions of m o d e m (language-)analytic philosophy by claiming that transcendental semiotics may in fact be conceived as a new patadigm of Fitst Philosophy, that is, as the completion of language-analytic philosophy as the third historical paradigm of First Philosophy. A s such, it may in fact supersede, or rathet "suspend" and preserve i n a Hegelian sense, the two preceding paradigms of First Philosophy, namely, ontology ot metaphysics i n the Aristotelian sense and critique of knowledge or philosophy of consciousness i n the sense of K a n t (ot even i n the sense of modem philosophy from Descartes through Hussetl). In otdet to show this, 1 shall try to derive a series of possible abstractive ot reductive fallacies from the well-understood conception of the sign-function or semiosis and thereby ideal-typically teconstruct and account fot some of the m a i n shortcomings that are involved in the three paradigms of First Philosophy and in some of their subparadigms.

T H E TRIADIC S I G N - R E L A T I O N A N D !TS IMPLICATION IN T H E L I G H T O F T R A N S C E N D E N T A L SEMIOTICS

1 start out by taking up the Peircean thesis that the sign-function or semiosis is an illustration of an irreducible triadic or three-place relation, as it is expressed, for example, in the following definition of Peirce's: " A sign, or representation, is something w h i c h stands to somebody for something i n some respect or capacity." T h e three elements or relatives of the triadic relation i n this definition are the
1

112

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

113

sign itself, its object w h i c h is denotated i n some tespect, and the addressee or interpretet of the sign. It is this explication of the ttiadic sign-telation that was taken as point of departure by C . W. M o r r i s i n his " F o u n d a t i o n of the T h e o r y of Signs," where it was made the basis of a three-dimensional semiotics through supplementing the semantic fetation between the sign and its object and the pragmatic felation between the sign and its interpretef by the syntactic telation between signs w i t h i n the frame of a sign-system ot language. (See figure 4.1.)
2

semantic

real object (denotatum)

r
language= system ^ (theory)

pragmatic

sign

real subject of sign-interpretation

III

N5
sign II FIGURE 4.1

M y motive fot taking this Peirce/Morris-schema of respectively, as statting and vantage point fot exposing dental semiotics is provided by two interconnected tal reflection upon the conditions of the possibility of thought.

semiosis or semiotics, the idea of a ttanscenfacts of transcendenintersubjectively valid

O n the one hand, the triadic sign-relation makes up a m i n i m a l basis element of any intetsubjectively valid knowledge, since its object-meaning has to be mediated through the intetsubjectively valid meaning of language-signs. O n the other hand, any argument concerning the validity of knowledge or any other human validity-claim is itself a triadic sign of some sort, not withstanding the fact that i n dealing w i t h aiguments i n terms of formal logic one usually abstracts from the pragmatic dimension of these signs, as, for example, from the fact that propositions as parts of arguments have to be communicated by somebody through speech-acts. N o w arguments at the same time make up that c o n d i t i o n

1 14

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST

PHILOSOPHY

of the possibility of intersubjectively valid philosophical thought behind which self-reflective thought cannot go back, since even solitary thought, being the "voiceless dialogue of the soul with itself" accotding to Plato, must participate in the structure of public argumentation. Hence it seems suggestive, ot even necessary, to take the semiotic structure of cognition and atguments as starting and vantage point for ttanscendental semiotics to be an actual possibility of fundamental grounding in philosophy. Now from this program it follows that from the outset I have to assign an interpretation to the Peifce/Morris-schema of semiosis that, being ttanscendental-semiotic, deviates in some tespects at least from Morris's interpretation. My ttanscendental reinterpretation will refer to all three dimensions of the triadic sign-relation which, on Morris's account, ate to be considered as many topics of eithet empirical-behavioristic or formal-constructive disciplines; that is, syntactics dealing with the telation between signs that makes up the grammatical structure of a language as a sign-system, semantics dealing with the telation between the sign and its real denotatum, and ptagmatics, dealing with the telation between the sign and its user or interpreter, be it a speaker or a hearer. First 1 have to make clear that talking about the semantical dimension of the sign-function as that of standing fot, ot denotating, teal objects can only be justified by the fact that in what follows we are only dealing with the so-called representative or, in Morris's terminology, designative "mode of signifying," in other words, with the function of speech as the mediatot of cognition, as it is also tegatded in the case of Carnap's use of the term semantic within the ftame of his (semiotically founded) logic of science. So we can, in the ptesent context, absttact from many othet "modes of signifying," as, for example, "appraisive," "prescriptive," and "formative" ones, which also have a semantical dimension, according to Morris's latet elabotation of his semiotics, not to speak of the later Wittgenstein's questioning the ttaditional absolutization of the designative ot naming function of words in his conception of the diversity of language games.
3

Second, 1 have to point out that, even under the abstractive presupposition of the representative function of signs in the setvice of cognition, we ate not entitled to identify the semantical dimension with the sign's denotation of teal objects as long as we absttact from the pragmatical dimension of sign-use, as it is done, for example, in empiric-linguistic semantics of language-systems, especially in F. de Saussure's "linguistique de la langue," and in formal constructive semantics, especially in Tarski's semantical explication of ttuth. For as long as we abstract from the ptagmatic dimension of people's communication ot speech (parole) in the context of situations we cannot deal with the possibility of identifying designated objects as real objects in space and time and hence are not entitled, in principle, to speak of denotata as objects of the semantical dimension, but only of designata, that is, of objects of sign-reference within the frame of abstract semantical systems hence, eventually, of scientific theories. The designata as objects of the abstract semantic dimension of language-systems may eventually represent empty classes without a corresponding extension of identi-

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

11 5

fiable denotata, or they may represent classes with a corresponding extension of identifiable denotata, or they may represent classes with a corresponding extension of fictive denotata, as, say, witches or angels; or, finally, they may represent classes of theoretical entities, that is, of denotata that may not be directly identified in space and time but whose assumption may onlyat bestbe indirectly legitimized, so to speak, in connection with a corroboration of the whole theory by way of its identifiable experiential consequences. This point has been recognized in modern logic of science since Camap's acknowledgment that the problem of verification cannot be settled within the frame of semantics but only within the frame of pragmatics. Furthermore, it is impottant to take notice of Peirce's semiotical insight that the function of identifying real objects in space and time cannot be fulfilled by purely conceptual signs or "symbols," in Peircean terminology, but only by "indexical" expressions, as, for example, "this there," whose s e m a n t i c a l reference cannot be conceived of independently of the pragmatic context of situationbound sign-use. So, when in what follows I am talking about the real denotatum as an object of the semantical dimension of the sign-function, as it is suggested also by the diagram (see figure 4.1), 1 shall always presuppose not the abstract semantical dimension of a language-system, but the pragmatically integrated semantical dimension. Thus far, my interpretation of the three-dimensional schema is in accordance, I think, with the general spirit of Morris's conception, as it was accepted by Carnap. But now 1 wish to introduce the main point of the idea of a transcendental semiotics which can no longer be covered by the Morris/Camap conception of semiotics.
4

I start from the consideration that the sign-function through which our cognition of real objects is mediatedor, for that matter, the representative function of the language-systems or semantical frameworks through which the object reference of theories is made possiblecannot itself be philosophically thematized as just a semantical object of the sign-function but must be considered as a condition of the possibility of describing and hence interpreting something as an object of intersubjectively valid theoretical knowledge. This view is reconfirmed by Carnap's conception of the function of "semantical frameworks" in his essay "Empiricism, Semantics, and Ontology." From the viewpoint of constructive semantics in Camap's and Tarski's sense, one may argue that the sign-function or the function of language as a semantical framework of interpretative worlddescription may itself be made the object of semantical reference in a metalanguage, and so on ad infinitum. However, this argument reconfirms rather than refutes my view: it suggests that the actual function of language as a mediation of interpretation and intersubjectively valid representation of objects cannot, in ptinciple, be objectified. Hence it turns out to be a transcendental condition of the possibility, rather than a possible object, of sign-mediated knowledge.
5

Also this view is reconfirmed rather than rejected, 1 think, by Carnap and his followers; for they themselves suggest that the function of philosophy is to

1 16

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

be distinguished from that of empirical science by the citcumstance that philosophets have to consttuct ideal languages or semantical systems as possible ftamewotks of the language of science. Thus constructive semantics, i n a sense, has to ptactice the " C o p e r n i c a n t u r n , " postulated by K a n t , thtough projecting those quasi-transcendental frames that prescribe the form of possible appearance to the world i n as far as it can be made the object of intersubjectively valid desctiption. Howevet, if my transcendental interpretation of language-systems, that is, of the syntactico-semantical patt of semiosis, is justified, then the same interpretation with respect to the (integrative) ptagmatic dimension of the sign-function must hold good; fot, without it, the semantical dimension of reference cannot imply concrete denotation of real objects of knowledge in space and time, as we alteady have pointed out. Just as we cannot considet the semantical conditions of world-description as possible objects of that very description, but must tathet asctibe a ttanscendental function to them, i n the same sense we must assign a ttanscendental function to outselves in so fat as we ate constructing and successfully interpreting language-systems as semantical frameworks of possible world-description. T h i s seems indeed to be as plausible, in principle, as attributing a transcendental function to outselves i n so fat as we ate btinging forward validity-claims of propositions or atguments to be confirmed or contested by any virtual member of an unlimited community of argumentation. Fot in both cases we are not primarily playing a role that is empirically relevant i n a psychological or sociological sense; we are rathet taking ovet a tole that obliges us to be i n chatge, so to speak, of the ttanscendental subject of knowledge ot thought. T h i s may be shown most convincingly through an examination of the current countetatguments to the possibility of a (transcendental-philosophical) abstraction from the empirical properties of ourselves as subjects of cognition or argumentation. For these counterarguments have at least implicitly to take the fotm of assettions like "I heteby assert that i n making and judging statements, we cannot absttact ftom the fact that we ate individuals w i t h specific biographies and sociohistotical backgrounds, etc." N o w this very argument against the possibility of the idea of a ttanscendental subject of cognition ot thought shows that, by the first " w e , " in contradistinction to the second " w e , " it must appeal to ourselves in as far as we are i n charge of the transcendental subject of thought. T h e fact that we can make out validity-claims explicit by performative expressions like "I hereby state that . . . " gives also a hint as to how it is possible for us to know about the actual sign-function as c o n d i t i o n of the possibility of thought w h i c h , as 1 have alteady pointed out, cannot be objectified, o n principle. In fact, out ttanscendental-semiotic knowledge about the actual function of signs ot, tespectively, language as c o n d i t i o n of the possibility of intetsubjectively valid thought and hence knowledge may be conceived of as reflective radicalizat i o n , so to speak, of that reflective knowledge that is first brought to vetbal expression through the self-referential performative parts of constative speech-

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

117

acts. O n the level of transcendental semiotics that reflective knowledge takes the form of propositions that are self-referential by their universal truth-claim. Should it be objected that self-referential speech must lead to semantical antinomies, i n the present lecture 1 can only answer that o n my account this talk tests o n a confusion between formalized semantical systems, w h i c h are to be immunized a ptiori against any possible antinomies and hence must not c o n t a i n self-teferential propositions. T h e language of philosophy must take the risk and the advantage of the pragmatic self-referentialness of natural language, i n order to talk about formalized semantical systems i n general. T h i s view is at least strongly supported by the fact that neither B. Russell i n his semantical theory of types nor A . Tarski i n his semantical (meta-logical) theory of metalanguages or language-states could avoid using a special (philosophical) metalanguage that cannot itself be placed into the infinite hierarchical order of types or, respectively, metalanguages w h i c h it is talking about. In fact, 1 am prepared to deal w i t h the current renunciation (disavowal) of pragmatically self-referential speech and hence self-reflection of thought as being itself one of the abstractive and reductive fallacies to be uncovered and analyzed i n the light of a transcendentalsemiotic account of the triadic sign-relation.
6

From this excursion into the problematic of philosophical self-reflection of thought we may come back to the c l a i m that we must assign a transcendental interpretation not only to the syntactico-semantical patt of the sign-function but also to its actual pragmatic dimension, that is, to the dimension of our actual language-use, and in that context to the dimension of our constructing and successfully interpreting language-systems as semantical frameworks of possible world-description. N o w this means that all philosophical talk about the problems of language-construction and interpretation, even if it is atguing for a purely conventionalist chatacter of all semantical frameworks of possible worlddescription, fulfills the reflective function of a transcendental pragmatics w i t h i n the frame of a transcendental semiotics. Fot, in so far as semantical frameworks are in fact conditions of the possibility of world-interpretation, philosophical talk about out construction and interpretation of semantical frameworks fulfills the function of a transcendental reflection o n the subjective-intersubjective conditions of the possibility of world-interpretation and, in addition, o n the subjective-intersubjective conditions of the possibility of conventions lying at the ground of semantical frameworks. In fact, the question as to the subjectiveintetsubjective conditions of the possibility of conventions lying at the ground of semantical frameworks appears to me as the most radical question of a transcendental pragmatics. For it eventually brings to bear transcendental reflection even in the back, so to speak, of conventionalism w h i c h is usually considered as a radical alternative to transcendentalism.
7

N o w , at this point, my transcendental interpretation of the Morris/Carnap scheme of three-dimensional semiosis or semiotics, respectively, definitively departs from the usual interpretation accepted in analytical logi<s.of language and science. For Morris and Carnap and their many followers i n analytic philosophy

1 18

T R A N S C E N D E N T A L SEMIOTICS AS FIRST P H I L O S O P H Y

take it as a matter of course that the pragmatic dimension of semiosis, that is, interpretative use of signs or whole language-systems by human subjects, may be objectified as a topic of empirical-behavioristic pragmatics in principally the same way as the sign-mediated behaviot of animals and human beings, that is, in otganisms, as it is investigated, say, by psycholinguistics and sociolinguistics.
8

O f coutse, as I alteady indicated, accotding to Morris and Camap, empirical pragmatics, in the same way as empirical-syntactics and semantics, may have a formal constructivist pendant that provides it with a conceptual framework. But formal pragmatics, as it is conceived in analytic philosophy, has nothing to do with a self-reflective thematization of the conditions of the possibility of language-consttuction and interpretation but tathet has been developed in the meantime as a semantization-sttategy with tespect to primarily subjective and self-reflective aspects of the actual ptagmatic dimension (e.g., as in performatives and indexical expressions). These aspects of self-reflective speech-acts, which, in my opinion, may be consideted as conditions of the possibility of human self-reflection and hence of human teason, ate hopefully made the objects of semantical reference by the metalanguage of pure pragmatics, eventually, in connection with a semantics of possible worlds which deals with propositional attitudes not as a topic of ttanscendental-ptagmatic self-reflection of human speech and thought, but tathet as a topic of epistemic and modal logic that accounts fot the meaning of the different self-teflective attitudes in terms of the possible truth of the cootdinated propositional contents in those possible wotlds whete facts wete corresponding to the propositional contents.
9

A t best, I could claim a correspondence between my conception of a ttanscendental-ptagmatic dimension of semiosis and the fact that Camap conceives of the choice of semantical frameworks, and hence of the pragmatical conventions concerning theit construction and interpretation, as a mattet of ptaxis tathet than of theory as it is teptesented for him by the empitical sciences which answet the so-called internal questions that ate made possible by the semantical frameworks. However, through his distinction between consttuctive praxis and empirical theory, Camap does not tenew the transcendental diffetence in the sense of Kant but rathet pleads fot a noncognitivist conception of philosophy. In the present context I shall not yet atgue for the ptiotity of my transcendental interpretation of semiosis and semiotics ovet the usual one. Instead I shall now turn to the application of the transcendentally interpreted scheme of semiosis for the purpose of providing a distinction between and an ideal typical teconstruction of the thtee paradigms of First Philosophy.

POSSIBLE T Y P E S O F FIRST P H I L O S O P H Y IN T H E LIGHT O F T R A N S C E N D E N T A L SEMIOTICS

If one takes the triadic sign-function or semiosis, as we have thus far, as the necessary mediation of world-interpretation and hence as a condition of the

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

1 19

possibility of any cognition of reality, then one may in the first place introduce a rather clear-cut distinction between three possible types of First Philosophy, according to whether the foundation of First Philosophy takes into consideration only the fitst, or the first and the third, or all three places of the triadic sign-relation, in order to account fot the primary topic of philosophy. (The presupposed sequence of the places of the sign-relation has been matked by roman numerals in our diagram; see figure 4 . 1 ) . The three possible types of First Philosophy, distinguished according to the proposed semiotical order of succession, may be approximately identified with the thtee paradigms of First Philosophy as they actually followed one another in the history of philosophy. For it may be said that general metaphysics or ontology in the Aristotelian sense considers real being (I) as it may designate and denotate by naming signs (proper and general names) as the primary topic of philosophy. Furthermore, transcendental philosophy or critique of pute reason in the Kantian sense may be said to consider being (1) only insofar as it is a possible cognitive object of the ttanscendental subject or consciousness (III) as a topic of philosophy, without consideting language or the sign-function as a transcendental condition of the possibility of intetsubjectively valid worldinterpretation and hence object-constitution. Finally, transcendental semiotics, as it has been outlined in the pteceding, may be considered as a third paradigm of First Philosophy that considers being as possible object (denotatum and designatum) of sign-mediated world-interpretation and hence considets the (entite) tfiadic sign-relation as the primary topic of philosophy. This first introduction of the three paradigms of First Philosophy stands of course in need of a great deal of further clarification which I shall try to give in what follows. First 1 should pethaps make clear that from our schematic derivation of the paradigms of First Philosophy it does not follow that, for example, the fitst paradigm (ontology) would completely exclude epistemology and/or philosophy of language, or that the second paradigm (transcendental philosophy) that considers being as possible object of knowledge would still completely exclude philosophy of language. Such an interpretation of the two first paradigms would not only contradict the historical facts but would also miss the point of the idea of patadigms of First Philosophy. This idea is illustrated by the fact that under the domination of the ontological paradigm questions of epistemology paradigm, questions of language-philosophy can be dealt with only as possible questions tegarding certain special objects of knowledge (cognition). Furthermore, it may strike you that transcendental semiotics itself figures, on the one hand, as a paradigm of First Philosophy and, on the other hand, serves as the basic idea from which the possibility of all three paradigms of First Philosophy is derived. Is there perhaps some sort of question-begging or at least some dogmatic prejudicing present here in the play? My answer to this question is that the double function of- transcendental semiotics in fact expresses the claim that transcendental semiotics is not simply a

120

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST

PHILOSOPHY

third type i n a (random) enumeration of possible types of First Philosophy. Rather, it takes its place accotding to a sequential otdet w h i c h it is able to justify such that it may figure itself as a possible and necessary synthesis of the two preceding patadigms of First Philosophy. Thus my conception of a revolutionary succession of different patadigms of thought differs from that of T . K u h n i n that it implies some sort of a Hegelian idea of possible progress i n the history of human thought. Ptopetly speaking, it does not imply any c l a i m of a causally explicable and hence ptedictable necessity of ptogtess. Rathet, it implies the claim that the three paradigms of First Philosophy make up a hierarchical order of levels of critical reflection and also make up an otder of necessary succession in the teleological sense without providing any guarantee of its being realized i n advance of the facts of histoty.
1 0

N o w , after taking a closer look at our schema of the triadic relation of semiosis, one might come to question our suggested detivation of the thtee patadigms of First Philosophy by the aigument that thete ate still fout othet possible detivations apparently atbitrarily neglected i n our approach thus fat. In fact, why should we not be allowed to detive the following fout possible types of First Philosophy from our schema? O n e that considers only real being (I) and signs or language (II) as primary topic of philosophy; one that assigns this dignity only to signs or language (II) and for the subject of sign-interpretation (III); and, finally, two (different) types that resttict the topic of First Philosophy to eithet signs (II) ot to the h u m a n consciousness and its contents (III). N o w , I think that it is neithet necessary to fotbid these derivations nor difficult to coordinate conceivable types ot paradigms of First Philosophy to them. Thus, one may relate the last mentioned possibility to subject idealism w h i c h was hypothetically supposed by Descartes before (previous to) his attempt to prove the existence of an external world and was adopted latet by Betkeley and eventually refuted by K a n t i n his argument for the transcendental presupposition of the structure of external objects of expetience.
11

T h e alternative possibility of distinguishing signs as the only primary topic of philosophy could be called semioticism and may be associated historically w i t h some suggestions of the early Peirce w h i c h seem to dissolve the whole world (including nature and the human subject of cognition) into ptocesses of semiosis. In a sense, Camap's constructive semanticism, w h i c h does not recognize eithet ontological or ttanscendental-epistemological problems that could not be teduced to empirical ones or be translated into the "formal m o d e " of dealing w i t h syntactico-semantical frameworks, could also be associated w i t h this type of First Philosophy.
12

T h e possibility, then, of a type of First Philosophy that only take signs and subjective sign-interpretation into regard could be called semiotical idealism. It has perhaps only been hypothetically or approximately adopted i n the history of philosophy, thus, for example, by the later Berkeley who conceived of the world as a context of natural signs through w h i c h G o d is speaking to u s . Finally, the most interesting of the four newly suggested types is the first one
13

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

12 1

w h i c h only takes real being and signs or language into regard. For it must, I t h i n k , be considered as more or less realized by the thusfat developed patadigm of (language-)analytic philosophy. For, the early Wittgenstein's immersion of the transcendental subject of sign-interpretation into the transcendental-logical form of language that delimits the world of possible facts for u s , ttanscendental self-teflection of the subjective-intersubjective conditions of sign-mediated wotld-interpretation has not been renewed in language-analytic philosophy. Instead, some sort of onto-semantics has been developed. Thus, for example, in the turn from Carnap's semanticism Quine's conception of ontological r e l a t i v i t y or, i n a very different way that sometimes comes close to our conception of transcendental pragmatics or semiotics, in Strawson's conception of the "bounds of sense."
14 15 16 17

H a v i n g detived and tentatively illustrated the total of seven possible types of First Philosophy i n light of our semiotical scheme, we must face the question whether this fertility i n our vantage point must not ovetthrow ot discredit our conception of the teleologically intelligible succession of just three paradigms of First Philosophy that would make up the backbone of the internal history of philosophy, according to a cettain developmental logic. N o w , the fact that all seven types could be detived ftom the transcendentally interpreted scheme of semiosis i n my o p i n i o n should, o n the conttary, suggest that it might be possible to adapt the detivation of the last four types to that of the fitst three types. For example, by showing that the last four types may be undetstood as subtypes and hence subpatadigms of the three m a i n paradigms of First Philosophy. In order to show this, we should reflect upon how it actually was possible for us to detive the seven types or paradigms of First Philosophy from the transcendental-semiotical presuppositions invested, so to speak, i n our triadic or three-dimensional scheme of semiosis. W h a t is the internal relationship of the seven detived paradigms to the presupposed ttanscendental scheme of semiosis? It seems clear from the outset that this question cannot be answered by any Pythagorean mathematical play w i t h the possibilities of combination or isolation of three places of the ttiadic sign-relation, unless some deeper interpretation is connected w i t h the schematical heuristics. Nonetheless, let us facilitate the business of schematical heuristics by symbolizing the seven possible types of First Philosophy w i t h the aid of the seven possible combinations of the three roman numerals. T h u s , we get the following seven items: (I), (II), (III), (I/II), (I/III), ( M i l ) , (I/II/III). A s an interpretative answer to the question of the relationship between the seven paradigms and the triadic scheme of semiosis, I then would suggest first that besides the paradigm of transcendental semiotics w h i c h reflectively takes into regard all three places of the sign-relation as conditions of the possibility of a meaningful world, all the other six types are constituted to by some abstraction. T h e abstraction by w h i c h each of these six types is constituted may easily be read from the pertaining symbolizations. N o w , abstraction in-First Philosophy may be methodologically justified so long as it is under control of transcendental

122

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

reflection as abstraction from a dimension of semiosis; otherwise, I would suggest, an abstractive fallacy must come about. From this point of view one may consider all six possible types of First Philosophy besides transcendental semiotics as based on some type of abstractive fallacy. But this formal-systematical consideration does not yet account for the privileged status of those three types I have distinguished as main paradigms of First Philosophynamely, (I), (I/III), and (I/II/III) in contradistinction to the other four types which were considered as mere subparadigmsnamely, (II), (III), (I/II), (II/III). Why, for example, should it not be plausible to regard ontology (I), semioticism or semanticism (II), and subjective idealism (III) as the fundamental paradigms of First Philosophy? Could they not have historically followed and replaced one another in this very order of sequence? This last question compels us to connect the viewpoint of historical ex post facto understanding with the formal-systematical distinctions. Thereby, I think, it becomes clear, that to begin with, the sequence (I), (II), (III), even if we equate it for the moment with the sequence (I), (I/II/III), (I/III) and hence take the positions (II) and (III) as sufficient representations of transcendental semiotics and transcendental epistemology, does not make up a plausible order of historical succession for the paradigms of First Philosophy. The reason for its implausibility is that it suggests the following dialectics of critical self-reflection in the history of philosophy; the first step after the aporetic or skeptical discussion of the problems of being and nothing, would not have been the question whether or how it is possible at all to know the objective truth about being; but, rather, whether or how it is possible at all to communicate one's cognition through intersubjectively meaningful signs and thereby make possible intersubjectively valid knowledge. Now, I think, not only the factual succession of the three paradigms of First Philosophy, but also the developmental logic of a possible radicalization of critical self-reflection in philosophy speaks against this order of succession. For, notwithstanding the fact that semiotical or language-analytical problems had to be dealt with from the beginnings of a propositional logic, it nevertheless seems to be only after the Kantian question as to the subjective conditions of the possibility of objectively valid knowledge that the question as to the semiotical conditions of the possibility of intersubjectively valid (i.e., primarily intersubjectively meaningful mediation of intersubjectively valid knowledge can be raised in its full critical significance). This consideration becomes still more persuasive if we compare the antimetaphysical suspicion of Kantian critique of knowledge and Wittgensteinian or Carnapian critique of language, respectively. For then it seems rather plausible to consider the idea that some or even all questions of metaphysics should be meaningless (ot nonsensical) due to semantical confusions as a subsequent radicalization of the Kantian thesis that metaphysical questions are necessary but unsolvable for a finite reason (intellect) than the other way around. In other words, the ideal that our questions may be meaningless for semiotical reasons

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

123

seems to presuppose the consideration that answers cannot be given to certain questions for epistemological reasons. To resume this point of developmental order, I think that already the Greek Sophist Gorgias, in his famous three theses, which he set up in reaction to the Eleatic dialectic of being and nothing, not incidentally set the stage for the subsequent succession of three paradigms of First Philosophy. For Platonic and Aristotelian ontology, Kantian or, for that matter, modern epistemology, and twentieth-century language-analytic philosophy may be understood as responses to the challenge of Gorgias' three theses: First, there is nothing; second, even if there were something it could not be known; and even if it could be known, it could not be communicated. But what about the justification of our claim that the developmental order of paradigms of First Philosophy is primarily represented by the type-sequence (I), (I/III), (I/II/III), that is, ontology, transcendental epistemology, transcendental semiotics, rather than by ontology (I) being followed by other possible types that also have an affinity with epistemology and semiotics respectively, such as subjective idealism (III) and semioticism/semanticism (II), or semiotic idealism (II/III) and onto-semantics (I/II)? This question becomes even more intriguing and urgent in view of the circumstance that within the factual course of the history of European philosophy, Berkeley's subjective or semiotic idealism in fact preceded Kant's transcendental idealism, which included empirical realism, and, last but not least, in view of the circumstance that the present stage of (language-) analytic philosophy is characterized by semanticism (Carnap, Tarski) and onto-semantics (Quine, or for that matter "semantics of possible worlds") rather than by the general adoption of something like transcendental semiotics, including the transcendental pragmatics of language. In order to answer this question I should first remind you of my former thesis that the very fact that transcendental semiotics provides a basis for the derivation of the other six possible types of First Philosophy indicates its systematical priority as an all-embracing paradigm of First Philosophy. Thus far we have elucidated this viewpoint only with respect to the claim of transcendental semiotics to suspend and preserve the function of ontology and transcendental epistemology as preceding main paradigms of First Philosophy. It remains, then, to show, from the same point of view, that transcendental semiotics should also be capable of proving its priority as a full-fledged paradigm of First Philosophy with respect to its semiotical rivals, namely, semanticism and onto-semantics. The preceding full-fledged paradigm of First Philosophy, transcendental epistemology, in a similar sense should be capable of proving its priority with respect to ontology as preceding the main paradigm and with respect to its epistemological rivals, subject idealism (or, for that matter, Humean positivism) and semiotic idealism. Now, by these theses, I have set up (designated) the main tasks of a critical reconstruction of the informal history of First Philosophy in the light of

1 24

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST

PHILOSOPHY

transcendental semiotics. A t ptesent I wish only to sketch roughly the sttategy of my envisaged answets, because there is still a furthet objection to be dealt w i t h against the sufficiency of the whole approach. Regatding the ptiotity of ttanscendental semiotics ovet its semanticist and onto-semanticist tivals with tespect to the status and function of the full-fledged third patadigm of First Philosophy, I can only say this m u c h at this point: Semanticism and onto-semanticism along the lines of the eatly Wittgenstein, B. Russell, Tatski, Carnap, Popper, and Q u i n e and even the language-ptagmatical philosophy of the latet Wittgenstein (in as fat as it sticks to the plutalism of factual language games and does not answet the question as to the conditions of his o w n capability of dealing w i t h all of them) are positions of (language-) analytic philosophy that may be shown to be only provisory stages of the t h i t d patadigm of Fitst Philosophy. Fot, they ate based o n absttactive fallacies w i t h tespect to a ttanscendental reflection of the integtative ptagmatical dimension of semiosis. Regarding the ptiotity of transcendental epistemology as the full-fledged second patadigm of Fitst Philosophy i n compatison w i t h its tivals, subjective idealism and positivism, I must say a bit mote at this point. I start out from the thesis that, i n a sense, Kantianism is able to supetsede and save the functions of ontology by speaking of beings as possible objects of knowledge and especially by showing, against subjective idealism and positivism, that thete is a crucial difference to be accounted fot between the objective otdet of appeatances i n the sense of empirical realism and the subjective succession of ideas i n a private consciousness. Moreover, the lattet presupposes the former, rathet than the other way around, as is supposed by subjective idealism since its hypothetical assumption by Descartes. A s a supplement to Kant's "refutation of idealism," one may also show that D . Hume's ot even E. Mach's positivism must, i n otdet to give meaning to its conception of sense-data ot neuttal elements of expetience, necessatily presuppose some version of that very unity of consciousness w h i c h it denies i n older to surmounr the metaphysics of subjective idealism. Thereby it also may be shown to be only a deficient subparadigm of the modern philosophy of consciousness w h i c h has its paradigmatic form i n a transcendental idealism that includes an empirical realism of objective beings. Yet i n regard to the K a n t i a n claim w h i c h , in a sense, is equivalent to the claim of preserving or saving the indispensable functions of ontology by transcendental epistemology, a reservation must be expressed at this point. By his distinction between objects of experience as mere appearances and unknowable things i n themselves, Kant has shown, against his intention, I t h i n k , that transcendental epistemology (by its combination of transcendental idealism and empirical realism) is not definitely capable of completely doing justice to and saving the functions of ontology as a patadigm of First Philosophy. For, i n order to enter into his system of philosophy, K a n t has to presuppose unknowable things i n themselves as causally affecting our senses. Thereby he is i n fact

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

125

presupposing a piece of pre-Kantian ontology that he can neither satisfy nor supersede by his own conception of transcendental idealism conjoined with empirical realism. For the causally effective function of things-in-themselves, accotding to Kant's presuppositions, cannot be conceived of as a function of the category of causality with respect to the formal constitution of the objects of experience but must rather be presupposed as a metaphysical-ontological cause of the material constitution of experience in general. I think that this central aporia of the Kantian system, again and again recognized since the day of Jacobi, cannot be overcome, in principle, within the frame of ttanscendental epistemology as paradigm of First Philosophy. But it may be dissolved within the frame of transcendental semiotics, which has to be shown.
18

Within the present context I must, however, take into consideration still one other possible objection against the whole approach of a ttanscendental semiotics that stands in close connection to the residual problematics of metaphysical ontology, as it has been identified as implied in Kant's doctrine of unknowable things-in-themselves. This residual problematics may be brought to bear on ttanscendental epistemology not only with respect to the being of things-in-themselves behind the possible objects of experience but also with respect to the being of (the same?) things-in-themselves behind the transcendental subject of knowing. In fact, Kant himself makes use of such a metaphysical presupposition first in his solution of the antinomy of freedom of the will versus causal determination and again in his F o u n d a t i o n of the M e t a p h y s i c s o/Morals. But still in anothet and more radical sense, one may claim that any type of ttanscendental epistemology or philosophy of consciousness must tacitly presuppose some answer to the question of the being of the transcendental subject or consciousness and its functions. Now this very question may also be raised with respect to the ontological conditions of the possibilities of sign-interpretation and hence may also be considered a serious objection against the priority of the (third) paradigm of transcendental semiotics over the (first) patadigm of ontology. For it seems to be an argument against any ptiority-claim of the transcendental philosophy of the conditions of the constitution of being over ontology as the philosophy of being itself. This objection may be connected with, and further illusttated by, anothet objection that concerns the fact that until now I have not teally shown that all historical paradigms of First Philosophy (i.e., all great philosophies that have represented that claim) may in fact be derived as possible paradigms or subparadigms from the transcendentally interpreted scheme of semiosis. This was not shown with respect to the modern philosophies of being represented by Hegel, Marx, and Heidegger. These modern philosophies of being, notwithstanding their differences, may be characterized precisely by the fact that they ask and try to answer the question of what being is with respect to the subject or the united subjects of knowing and of communication. Must not their approach necessarily transcend that of a transcendental semiotics as well as that of anranscendental epistemology ?

126

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST

PHILOSOPHY

In conclusion, I can only point, by some ptogrammatical ternaries, to the envisaged strategy of answering these last two interconnected questions as to the ontological status of semiosis and to the possibility of doing justice also to the modern philosophies of being from the vantage point of transcendental semiotics. First, I would like to point out that the modernity of the philosophies of Hegel, M a t x , and Heidegget as philosophies of Being lies i n the fact that they ate dealing w i t h history and the historicity of being including the being ot substance of the human m i n d ot consciousness. N o w , w i t h tespect to this problematics, I first of all wish to state my systematical c l a i m that it can be shown to be impossible a priori to deal philosophically w i t h history without doing justice to the transcendental status of the theoretical truth-claims and notmative ptactical validity-claims of the human subjects of history and of cognition of history. For history cannot be adequately conceived of as simply a natutal process of change to be completely objectified and eventually nomologically explained as a reality (that is) independent of its being continued and continuously altered by human theory and subjective praxis. Rather it must be defined from the outset as an (irreversible) process of continuous dialectical mediation of theory and praxis. T h a t is, as a process that comprises, i n principle, the human attempts of recognizing and tightly continuing this very process. Hence it can only be adequately dealt w i t h by those philosophies that ate able to cope w i t h theit o w n validity-claim as a possible (if not necessary) result of history. Therefore, it may be furthet postulated that an adequate philosophy of the history of being cannot conttadict the type of transcendental philosophy that does not ptevent us from considering the human subjects of knowledge and action (notwithstanding theit transcendental status as carriers of truth-claims and normative validity-claims) as the teal beings belonging to teal history. N o w it is precisely this possibility that is opened up by ttanscendental semiotics, insofat as it poses the ttanscendental language game, w h i c h petvades a l l factual language games due to the communicative and virtual atgumentative competence of m a n , i n the place of an extramundane pure ttanscendental consciousness and theteby surmounts "ttanscendental idealism" in the K a n t i a n sense. T h i s geneial thesis may be provisionally elucidated by some remarks about the metits and limitation of Hegel's, Matx's, and Heidegget's philosophies from the petspective of our approach. Hegel, i n a sense, can be considered as the very fathet ot inauguratot of that philosophy of history that, as I claimed, must account fot its o w n ttanscendental status as the possible (ot even necessary) result of history. Thus far his conception of the history of being or substance as possible subject of its o w n cognition does not at a l l overthrow transcendental philosophy as such. Rather, for the fitst time it claims to show the possibility, and moreover even the actuality, of a complete mutual mediation of ttanscendental idealism and quasi-Aristotelian ontology of substance. W i t h respect to ttanscendental reflection upon the conditions of the possibility of cognition (knowledge), H e g e l goes even beyond K a n t by taking into tegaid the knowledge-claim of critical-reflective philosophy

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

127

itself up to the point of knowledge of knowledge. A l t h o u g h he simultaneously criticizes so-called "philosophy of reflection" including K a n t i a n i s m , he does so by understanding his o w n reflective science of the appearances of consciousness (in the P h e n o m e n o l o g y of Spirit) as self-movement of the idea i n its development from being i n itself through "being for itself" to "being i n itself and for itself," I t h i n k that a true core of this approach of a dialectical mediation between transcendental idealism and ontology of history may be preserved by transcendental semiotics. T h e weakness, though, of Hegel's approach, w h i c h made it appear to the Young Hegelians as the conclusion of theoretical metaphysics that had to be surpassed by something beyond philosophy, lay i n Hegel's c l a i m , or suggestion of a total mediation of ontology and transcendental idealism of consciousness, and hence of the matetial of objective expetience and the form of conceptual understanding. A n d a total mediation of reflective theory and subjective praxis, from the vantage point of an a priori conceptual understanding that has already attained the point of ultimate and hence absolute knowledge. By this suggestion of an absolute theoretical metaphysics of history that urged the Y o u n g Hegelians, and among them Kierkegaard and the young M a t x , to project philosophies of existential or social praxis that sometimes wete considered as ttansgressing the very idea of p h i l o s o p h y . A t the same time, it became clear that Hegel's speculative anticipation of the theoretical truth of the whole of historical being could not do justice to the difference of transcendental philosophical reflection and empirical cognition by sensual expetience as it was supposed by K a n t and was to become the bteaking-off point of the empirical sciences from Hegelianism in the name of positivism.
19

Regarding these shortcomings of Hegel's system of total mediation w i t h respect to sensuous experience and the need for practical engagement toward the future, I consider it possible to show that transcendental semiotics may provide a sort of preserving correction of Hegel's approach i n three respects. First, w i t h the aid of Peirce's analysis of different epistemological (cognitive) signfunctions, it can be shown that the truth about indexical expressions testifying to sensuous certainty mediates between Hegel's position i n his famous opening chapter of the P h e n o m e n o l o g y of Spirit and Feuerbach's critique of it (and thus between idealism and matetialism or, respectively, positivism). Second, w i t h the aid of some ideas of C . S. Peirce and J. Royce about the indefinite community of interpretation, to be postulated, in a sense, as the ttanscendental subject of v a l i d cognition of the level of transcendental semiotics, it can be shown that Hegel's speculative A u f h e b u n g of the truth of the whole history of being into the reflective monologue of one finite philosopher may i n fact be corrected i n favor of the dialogue of a l l rational b e i n g s , that is to be mediated through the progress of empirical science. But this correction of Hegel can even preserve and reconfirm Hegel's idea of a necessary anticipation of the absolute truth by transforming it into a transcendental-semiotical postulate of the possibility, i n principle, of c o m i n g to consensus about meaning and truth w i t h i n the
2 0 21

128

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST

PHILOSOPHY

frame of the infinite dialogue of the indefinite community of interpretation. Finally, it can be shown that the introduction by transcendental semiotics of an ideal community instead of a self-sufficient I as transcendental subject brings about a radical new perspective w i t h tespect to the problems of practical teason, ot of the mediation between theory and ptaxis. But this last point may be provisionally illusttated by some tematks about the post-Hegelian philosophies of the mediation, or separation, between scientific theory and subjectiveintersubjective praxis. T h e dissolution of the Hegelian system of absolute (speculative-theotetical) mediation of theory and praxis i n the nineteenth century finally led to a k i n d of complementarity system of positivism ot scientism without ptactical reason and irrational existentialism of private praxis w h i c h was to become the basic ideology of the bourgeoisie i n the Western world of the twentieth century. O n the other side, Marx's ingenious conception of a critical philosophy of praxis (or of dialectical mediation between critical theory and subjective ptaxis) partly i n his o w n hands and definitely i n the hands of Engels and L e n i n , turned into a dogmatical ontology of historical being that is supposed to be capable of totally mediating between scientific theory and the subjective praxis of m a n k i n d even w i t h regard to the u n k n o w n futute. Seen as a theoretical philosophy from the point of view of our account of the intelligible succession of three paradigms of First Philosophy, M a r x i s m by its objectivist-scientistic transformation into ontological " D i a m a t " underwent a tegtession into a p i e - K a n t i a n type of First Philosophy that is to be considered as anachronistic o n our account. The same is true with respect to those sociologist-naturalist and, for the mattei, sttuctuialist accounts of history that undetstand themselves as ovettaking ot making obsolete First Philosophy from an empirical-theoretical petspective outside of philosophy. T h i s has to be shown in more detail through a systematical detivation of naturalistic fallacies as abstractive and reductive fallacies in the light of out scheme of semiosis as precondition of object-constitution. A s a philosophy of historical being even w i t h tespect to future being, M a i x i s m has to be valuated also as practical philosophy that is to be confronted to the ptactical aspect of the Western complementarity system. In this respect, it may be shown that, by claiming to solve ethical problems of human praxis by a theory of the necessary course of history into the future, marxism became a dogmatic counterpart of the Western ideological complementarity system of positivism/scientism and existentialism by settling down as the Eastern integration-system of " D i a m a t . " N o w , it may be shown that this ideological East-West constellation in the twentieth century amounts to a philosophical dilemma w i t h respect to the problem of ethics i n the age of science; for the alternative of the Western complementarity system and the Eastern integration system turns out to be an alternative between personal freedom and free science without intersubjectively binding ethical norms, values, or aims, o n the one hand, and the dogmatic, fixed, and institutionalized mediation of theory and praxis without a change of
22

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

1 29

mediation through the free decisions of conscience of single human persons as citizens o n the other hand. In view of this dilemma, I think it possible to show that o n the basis of transcendental semiotics, or, for that matter, of transcendental pragmatics, an ultimate foundation of intersubjectively binding ethical norms may be provided through surmounting the methodical solipsism of ttaditional philosophy of consciousness i n favor of the transcendental a priori of the communication-community. Furthermore, it can be shown, I think, that a dialectical philosophy of the undogmatic reconstruction of social history and an ethical strategy for the radiation of theory and praxis w i t h respect to the future may be outlined by starting from the teleological postulate that the ideal communication-community has to be realized w i t h i n the real ope under the historical conditions of the survival of m a n k i n d as a real community. A n d w i t h i n the frame of this program a critical-emancipatory theory of social reality of neo-Marxist inspiration may find its place. From this programmatical point of view, I may finally say some words about the relationship between ttanscendental semiotics ( w h i c h includes transcendental pragmatics and transcendental hermeneutics), and the last great attempt at a foundation of First Philosophy as a philosophy of the history of being, as it has been displayed by M a t t i n Heidegger. H a v i n g myself started out in philosophy from interpreting Heidegger's transformation of the Husserlian phenomenology as a hermeneutics of speech and of language, respectively, I would like to point out that I still see the great achievement of this philosophy, i n its account of the historical events, not to be disposed of scientifically, but rather achieved by the prescientific workings of mythos, poetry, and the arts, of the disclosure of meaning as the precondition of possible truth and falsehood of p r o p o s i t i o n s . I even think that these Heideggerian perspectives may be liable to be further displayed in the future, along w i t h Cassier's philosophy of "symbolic forms" and together w i t h a reconstruction of topics along the lines of G . B . V i c o and the modern conceptions of a historical-philosophical t o p i c s , w i t h i n the hermeneutical part of semiotics.
23 24 25 26

However, I also have come to the c o n v i c t i o n that the normative problems of the conditions of the possibility of intersubjective validity w i t h i n the realm of theoretical and practical reason cannot be thematized, let alone solved, along the lines of a Heideggerian philosophy of the "fate of b e i n g " (Seinsgeschick) w h i c h claims to supersede the Western philosophy of subjectify (Subjektitat) including the autonomy-claim of human reason through an "attentive" ( a n d a c h tiges) t h i n k i n g that justifies its validity-claims by its belonging to (Zugehdrigkeit or even Horigkeit)the fate of being. I am afraid that this type of thought, along w i t h certain stripes of M a r x i s m , of structuralism, and of functionalistic system-theory, may succeed i n rendering obsolete (together w i t h the bourgeois illusion of a theoretical and practical autonomy of the solipsistic subject) the very idea of a h u m a n subject of theoretical and practical validity-claims and of solidary responsibility. In view of this situation, I have come to t h i n k that, instead of trying to overcome { v e r w i n d e n ) the philosophy of the "transcendental
1

] 30

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

subject" togethet w i t h pre-Kantian metaphysics, (ot, fot that mattet, trying to dissolve human subjectivity into anonymous structures), we should rathet try to think of the ttanscendental subject of h u m a n cognition and of human ptaxis as something supetindividual but interpersonal. T h a t is, as an instance of intrinsic solidarity that we must counterfactually presuppose but still have to realize as ideal communication-community. Thus I would indeed claim that ttanscendental semiotics as ttanscendental ptagmatics and transcendental hermeneutics may also cope w i t h the challenge of Heidegger's philosophy of being (and of structutalism, fot that matter) from the vantage point of a transfotmed idea of ttanscendental subjectivity.

NOTES
[Due acknowledgment is made for the editorial assistance of Professor George Stack, Department of Philosophy, S U N Y at Brockport.] 1. C. S. Peirce, Collected Papers, ed. C . Hartshome and P. Weiss (Cambridge, Mass.: Harvard University Press, 1931-35, 2nd ed. 1960), 2.228. 2. C . W . Morris, Foundation of the Theory of Signs, International Encyclopedia of Unified Science, vol. 1, no. 2 (Chicago: University of Chicago Press, 1938). 3. Cf. C . W . Morris, Signs, Language, and Behavior (New York: G . Braziller, 1946). 4. Strictly speaking, this assessment of concordance is already doubtful. For it seems clear to me that the fact that the real object of cognition can only be considered to be the object of the pragmatically integrated semantic dimension of the sign-relation implies already that the idea of a satisfactory explication to "truth" to be provided by pure semantics must amount to an abstractive fallacy. This does not mean to equate the problem of truth-explication with that of verification but, rather, to postulate a truth-explication on the basis of (a transcendental semiotics of) the triadic signrelation. 5. R. Camap, "Empiricism, Semantics, and Ontology," Revue Internationale de Philosophie 4 (1950): 20-40; repr. in R. Carnap, Meaning and Necessity (Chicago: University of Chicago Press, 2nd ed. 1956). 6. Cf. my essay "Types of Rationality Today: The Continuum of Reason between Science and Ethics," Proceedings of the International Symposium Rationality Today, October 1977, ed. T. F. Geraets (Ottawa: Ottawa University Press, 1979). 7. Cf. my essay "Sprechakttheorie und transzendentale Sprachpragmatik zur Frage ethischer Normen," Sprachpragmatik und Philosophie (Frankfurt a . M . : Suhrkamp, 1976), 10-173. 8. For a critical examination of this approach, see my introduction to C . W . Morris, Zeichen, Sprache undVerhalten (Diisseldorf: Schwann, 1973), 9-66. 9. Even if the program of dealing with the epistemologically relevant problems of pragmatics within the frame of a semantics of possible worlds should be successful as an extension, so to speak, of that projection of actual thought ("thirdness" in Peirce's terminology) into the realm of structural possibilities ("firstness" of "thirdness") that always was the business of mathematics and formal logiceven then the question of the transcendental (subjective-intersubjective) conditions of the possibility of the constitution of the notion of "possible worlds" remains to be answered. Cf. on this question Z. Vendler, " O n the Possibility of Possible Worlds," Canadian Journal of Philosophy 5, no. 1 (1975): 57-71.

T R A N S C E N D E N T A L S E M I O T I C S A S FIRST P H I L O S O P H Y

131

10. Cf. T. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962). 11. Cf. I. Kant, Critique of Pure Reason, 2nd ed. (1962). 274-79. Cf. also P. F. Strawson, The Bounds of Sense (London: Methuen, 1966), 125ff. 12. C f . , e.g., C . S. Peirce, Collected Papers, 5.285 and 313f. 13. C f . , e.g., G . Berkeley, N e w Theory of Vision, sees. 51, 159; idem, The Principles of Human Knowledge, sec. 44. 14. Cf. L. Wittgenstein, Tractatus logico-philosophicus 5.62, 5.632, 5.64, 5.641. 15. Cf. G . Janoska, Die sprachfichen Grundlagen der Philosophie (Graz: Akademische Druck- und Verlagsanstalt, 1962); and E. K. Specht, Sprache und Sein: Untersuchungen zur sprachanalytischen Grundlegung der Ontologie (Berlin: de Gruyter, 1967). 16. W . V . Quine, Ontological Relativity and Other Essays (New York: Columbia University Press, 1969). 17. Strawson, The Bounds of Sense. 18. Cf. F. H . Jacobi, Gesammelte Werke (Leipzig, 1812-25), vol. 2, 304. Cf. also Strawson, The Bounds of Sense, especially part 4. 19. C f . , for an overview, K . Lowith, Von Hegel zu Nietzsche (Zurich: Europa Verlag, 1941), especially parts 1,2, and 3. 20. Cf. H . - G . Gadamer, Wahrheit und Methode (Tubingen: Mohr, 1960), 351; cf. also K . - O . Apel, "Szientismus oder transzendentale Hermeneutik," in Hermeneuti/c und Diakktik, Festschrift far H . - G . Gadamer, ed. R. Bubner et al. (Tubingen: Mohr, 1970), vol. 1, 105 - 4 4 . 21. Cf. C . S. Peirce, Collected Papers, 8.13. 22. Cf. my essays "Das Apriori der Kommunikationsgemeinschaft und die Grundlagen der Ethik," in Apel, Tfans/ormation der Philosophie (Frankfurt a . M . : Suhrkamp, 1973), vol. 2, 358-436; and "Conflicts of Our Time and the Problem of Political Ethics," in From Contract to Community: Political Theory at the Crossroads, ed. W. Dallmayr (New York: Marcel Dekker, 1978). 23. Cf. my essay "Szientismus oder transzendentale Hermeneutik." 24. Cf. the "Einleitung" of my book Die Idee der Sprache in der Tradition des Humanismus von Dante bis Vico (Bonn: Bouvier, 1963, 2nd ed. 1975). 25. Cf. E. Tugendhat, Der Wahrheitsbegriff bei Husserl und Heidegger (Berlin: de Gruyter, 1967); and Heidegger's subsequent self-correction in Zur Sprache des Denkens (Tubingen: Niemeyer, 1969), 76f. 26. Cf. my book Die Idee der Sprache in der Tradition des Humanismus.

FIVE The "Pragmatic Turn" and Transcendental Semiotics.The Compatibility of the "Linguistic Turn" and the "Pragmatic Turn" of Meaning Theory within the Framework of a Transcendental Semiotics

EXPOSITION

From the somewhat complicated subtitle of my papet it becomes obvious that thete ate some presuppositions concerning facts as well as concerning tetmsas for example the tetms "linguistic t u r n , " "pragmatic t u r n , " and especially the somewhat esoteric term "transcendental semiotics." T h e best way fot me to clarify, at least initially and approximately, those presuppositions, it seems to me, is to tell a story about the prehistory of the present stage of the philosophy of language and semiotics from my subjective perspective. T h i s might reveal and expose to some extent my o w n preconceptions and prejudices i n the matter; and thereby, I think, bettet pteconditions fot mutual undetstanding ate provided, at least at the beginning, than would be by abstract definitions and apparently self-sufficient (unconditional) arguments. I shall tty to develop my atgument w i t h i n the context of my story w h i c h w i l l gradually become a systematic account of the problems. T h e frame of my background-story may be marked by the following scheme of three stages: 1. T h e linguistic turn i n philosophy and the abstractive fallacy of transcendental semanticism 2. T h e pragmatic turn as the ovetcoming of the abstractive fallacy of transcendental semanticism and hence as the completion of the linguistic turn as transcendental-pragmatic turn w i t h i n the framework of a transcendental semiotics 3. T h e overturning of the pragmatic turn by reducing meaning to prelinguistic intentions and the possibility of its refutation. 132

PRAGMATIC T U R N " & T R A N S C E N D E N T A L SEMIOTICS

133

Let

me then start out with a very very rough and short teconstruction of the

fitst stage.

THE

LINGUISTIC T U R N IN P H I L O S O P H Y A N D T H E ABSTRACTIVE F A L L A C Y O F TRANSCENDENTAL SEMANTICISM

The first part or stage of my story may be represented by the T r a c t a t u s logicophilosophicus of the eatly Wittgenstein and also, in a slightly different sense, by the consttuctive logical semantics of Carnap and Tarski. In Wittgenstein's T r a c t a tus the linguistic turn of First Philosophy becomes obvious as a ttanscendentalsemantic turn, so to speak. As such, it is matked first by the rigorous and thoroughgoing substitution of linguistic concepts in the place of the mentalistic concepts of modern philosophy from Descartes to Husserl, such as consciousness, judgment, thought ot intention(ality). A good example of this substitution is the sentence " D e i Gedanke ist dei sinnvolle Satz" ("The thought is the meaningful sentence"). This fitst maik of the linguistic turn is supplemented by a second one which definitively detetmines the ttanscendental-semantic chaiactet of Wittgenstein's linguistic turn. What 1 mean is the replacement, so to speak, of Kant's "supreme principle of synthetic judgments" (namely, that "the conditions of the possibility of experience are the conditions of the possibility of the objects of experience") by an equivalent postulate with tegatd to the ttanscendental-logical conditions of pute language as being the conditions of the possibility of facts as elements of a desciibable wotld. This ttanscendentalist interpretation of the T r a c t a t u s , which was thoroughly carried through by Eric Stenius, is sometimes called into question, but, in my opinion, it is sufficiently confirmed by Wittgenstein himself, fot example, in the following dictum from Vermischte Bemerkungen:
1

Die Grenze der Sprache zeigt sich in der Unmoglichkeit, die Tatsache zu beschteiben, die einem Satz entspricht . . . , ohne eben den Satz zu wiedetholen. (Wit haben es hiet mit der Kantischen Losung des Problems der Philosophie zu tun.) [The limit of language shows itself by the impossibility of desctibing the fact that corresponds to a sentence . . . without tepeating ptecisely that sentence. (We are dealing here with the Kantian solution to the problem of philosophy.) (Trans, mine)].
2

I would like to nail down the point of this statement by calling it the principle of the onto-semantic autonomy and methodological nontranscendability of language; and I want to make cleat from the beginning that, in a cettain sense still to be clarified, I consider this ptinciple an irreversible standatd of this century's philosophy and a criterion by which at least part of the meaning of the so-called linguistic t u r n is defined.
3

However, according to Wittgenstein's account in the T r a c t a t u s , the meaning of linguistic sentences about facts is not only intranscendable^at least in an intersubjectively valid formby a prelinguistic intuition of facts in themselves

1 34

"PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

but even more so by a metalinguistic reflection (i.e., reflection upon the representative function of language by language); for according to Wittgenstein there is no self-referential use of language. Hence Wittgenstein's own metalinguistic sentences about the relationship between the structure of language and the structure of the world are denounced and renounced as "nonsensical" at the end of the T r a c t a t u s , as is well known (cf. 6.54 and 7). This paradoxical overstatement of the transcendental character and function of the structure of sentences implies among other things that there is no I or subject of meaning-intentions or interpretations that could provide a ttanscendental status not only to the logical sttucture of language but also to the reflective use of language (or to reflective conventions about the use of language). Instead, there is, according to Wittgenstein in the T r a c t a t u s , only a transcendental I thatlike an "extensionless point"is absorbed, so to speak, by the logical form of language, such that the "limits of language are the limits of my world" and "solipsism coincides with pure realism" (cf. 5.64ff. and 5.62ff.). In my opinion, this version of linguistic transcendentalism has a degree of plausibility at best on the following presuppositions: 1. There should be only one pure language, or at least only one logical deep structure of all possible languages, that prescribes the ontological sttucture of the describable world once for all, so to speak, so that there would not be any need for a reflective communication about language or for conventions about the logical grammar of language. 2. The whole language should consist of propositional sentences which have only one function, namely that of a representation of states of affairs, so that there are no self-expressive and communicative functions of language that could serve as a basis for a reflective communication about language and thus even fot reflective conventions about its use. 3. The reference of signs to real objects (within the context of the representation of existing states of affairs by propositions) should be guaranteed somehow by the structure of language (say, by the depicting-function of elementary sentences as protocols of facts). The first part of these presuppositionsnamely, the idea of one pure language or at least one logical deep structute of all possible languageswas given up already by R. Carnap; it was replaced by a pluralism of possible syntacticosemantical frameworks which took over, so to speak, the quasi-transcendental function of prescribing the structure of scientific descriptions and explanations of the wotld of experience.
4

Now, even if Carnap's "principle of tolerance" with regard to the logical form of language might not be completely justifiable, in any event it became clear in this context that Wittgenstein's principle of the intranscendability of the syntactico-semantical form of language has to be transcended in a sense by the principle of metalinguistic reflection upon language, and that means eventually: reflective communication by language not only about (conventions about) the logical form of artificial languages that are to be constructed, but also about the

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

1 35

use of the nonformalizable language that is ptesupposed by the construction as the ultimate pragmatic metalanguage, so to speak. I think this methodological problem-situation already demonsttates that the quasi-transcendental conception of syntactico-semantical frameworks, w h i c h leads from Wittgenstein's T r a c t a t u s to the logical semantics of Carnap and Tarski, suffers from an absttactive fallacy; in othet wotds, that it has to be supplemented by the conception of a (ttanscendental) ptagmatics that could deal w i t h the subjective-intetsubjective conditions of the possibility of reflective communication by language about language. Carnap however, as is well k n o w n , did not recognize these problems as those of a theoretical philosophy but precluded theit teflection by declaring all so-called "external questions" about the constitution of "semantical frameworks" to be purely "practical questions."
5

In close connection to this aspect of logical semantics' absttactive fallacy stands anothet one (that was also inhetited from Wittgenstein's Tractatus): that of tefetence to the teal, or i n Carnap's terminology, that of verification. Fot it quickly became clear that this problem ttanscended the scope of Tatski's explication of the concept of ttuth fot abstract constructed semantical systems. Should thete not also be a problem of a ttanscendental ptagmatic explication of a criteriologically relevant concept of truth that could serve as a basis for the methodology of verification? Here also, however, C a m a p sees, not a further problem of theoretical philosophy, but only a problem of empirical pragmatics, that is, of empitical (behavioristic) philosophy of scientific behaviot of verification.
6

A t this point, I suggest, Charles W . Morris intervened in the development with his Foundations of the T h e o r y of Signs, where he proposed a three-dimensional semiotics, consisting of "syntactics," "semantics," and "pragmatics." Theteby Morris, in a sense, brought to bear o n the issue Peirce's conception of a semiotics that is based o n the ttiadic ot three-places relation of the sign-function ot semiosis: namely, the telation between the sign itself, its teal referent and its user or interpretet. But, o n a closet look into the mattet, this claim of Morris's turns out to be thoroughly misleading. For the ptagmatic dimension, w i t h w h i c h Morris actually supplements Carnap's logical syntactics and semantics, is not allowed to take o n the same quasi-ttanscendental status as the syntacticosemantic framework has in Carnap's account. T h a t is to say, the intentional ot interpretative use of the ftamework is not reflectively integtated, so to speak, into the quasi-ttanscendental function of the syntactico-semantical ftamewotk, but is conceived only as a possible object of empirical science, that is, as a possible object of the semantic sign reference w h i c h already presupposes the syntactico-semantical framework and its actual interpretation (see figure 5.1).
7

T h e reason for this depotentiation of the ptagmatic dimension of semiosis or, respectively, semiotics is to be found, I think, i n the dogma of post-Russellian or, respectively, post-Tatskian logical semantics that there could not be such a thing as a linguistic self-reference of the sign-function ot semiosis, for example, of the meaning-intentions ot the meaning-interpretations of the human subjects of semiosis as subjects. A s a consequence, Morris and C a m a p (who eventually

136

"PRAGMATIC T U R N " & T R A N S C E N D E N T A L SEMIOTICS

accepted Mofris's proposals) could only conceive of pragmatics as a matter of empirical behavioristics; or, beyond that, at best as a matter of a constructiveformal pragmatics, that is, of a metalinguistic constructing of a syntacticosemantical framework for an empirical description of the use of language.
8

But it seems clear that such a program of'pragmatics cannot deal reflectively w i t h the pragmatic dimension of the constitutionsay by explicit conv e n t i o n s o f syntactico-semantical frameworks for empirical descriptions of the world. Hence it cannot deal w i t h the quasi-transcendental functions of the pragmatical dimension of semiosis. Therefore it cannot complete the linguistic turn of First Philosophy by showing that the whole triadic f u n c t i o n of semiosis i n Peirce's sense is the transcendental-semiotic c o n d i t i o n of the possibility of our having intersubjectively valid knowledge about the world. O n the contrary, Morris's and Carnap's introduction of the pragmatic dimension of semiosis reconfirmed the abstractive fallacy of the semanticist stage of the linguistic turn (see figure 5 . 1 ) .
9

In what follows, I w i l l try to illustrate i n a n affirmative way the difference between the Morris/Carnap conception of pragmatics and my o w n idea of a transcendental pragmatics w i t h i n the framework of a transcendental semiotics. I thereby come to deal w i t h the second part of my story.

THE THE

PRAGMATIC T U R N O F M E A N I N G - T H E O R Y A S T H E O V E R C O M I N G O F ABSTRACTIVE F A L L A C Y O F T R A N S C E N D E N T A L S E M A N T I C I S M A N D AS THE COMPLETION O F T H E LINGUISTIC T U R N AS

T R A N S C E N D E N T A L - P R A G M A T I C T U R N WITHIN T H E F R A M E W O R K O F T R A N S C E N D E N T A L SEMIOTICS

O n the c o n d i t i o n of the triadic or three-places relation structure of the signfunction or semiosis, there are two aspects or directions for a necessary overcoming of the abstractive fallacy of semanticism: one w i t h regard to the so-called s e m a n t i c dimension of sign reference and one w i t h regard to the so-called pragmatic dimension of the sign-use. T h e reason for this double aspect to the problem lies i n the fact that even the semantic dimension of reference poses a problem of pragmatics as has to be shown. Let us then first consider the so-called semantic dimension of sign-reference (figure 5.1, left side of the scheme).
THE INTEGRATION OF SEMANTICS AND (TRANSCENDENTAL) PRAGMATICS WITH REGARD TO T H E SEMANTIC DIMENSION OF SIGN-REFERENCE
1 0

Contrary to Morris's suggestion i n the F o u n d a t i o n s , it has to be pointed out that the problem of identifying the real denotatum, w h i c h may or may not correspond to a designatum i n the world of space and time, is not only a matter of the semantic dimension of a syntactico-semantical system. T h i s is easy to recognize if one considers the fact that an abstract semantical system of language may contain designata that cannot at a l l be identified as denotata, as for example, unicorns or h e l l or Santa Claus. H e n c e identifying a denotatum is a

PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

137

fundamental complementarity-situation interpreter (subject of meaning-intention) as o b j e c t o f behavior description

cognitive reference

III(I) interpersonal communication (reciprocal understanding)

denotatum (real object o r referent)

interpreter (subject of meaning-intention)

red uction

interpreter (co-subject o f meaning-intention)

_ -

r e d u c t i o n of t h e P e i r c e a n scheme of s e m i o s i s w i t h i n t h e f r a m e of l o g i c a l empiricism a n d behaviorism (Carnap a n d Morris) t r a n s c e n d e n t a l - p r a g m a t i c s u p p l e m e n t a t i o n ( i n t e g r a t i o n ) of t h e scheme semiosis w i t h i n the f r a m e of t r a n s c e n d e n t a l semiotic of

F I G U R E 5.1

138

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

matter not only of the semantic but also of the ptagmatic dimension of semiosis. Thete has to be a sign-interpretet (see figure 5.1, right side of the ttiadic scheme of semiosis) i n otdet to identify the teal denotatum of a sign (figure 5.1, left side of the ttiadic scheme of semiosis). But even Morris's pragmatics, that is, his account of "sign-mediated beh a v i o t , " cannot show what identifying the referent of a proper name (or identifying an individual referent of a geneial name) really i s . Fot he can at best desciibe an external causal c h a i n of behavioi that is supposed to connect the o i i g i n a l object of denotation w i t h the " o i g a n i s m " that uses the name (see figure 5.1, left side: "empirical semiotics"). But, as J. Seaile has recently shown once moie i n a brilliant discussion of the "causal theory of reference," an external description of the causal c h a i n of b e h a v i o i can nevet ensure that the meaning-intention of the usei of a name correctly refers to its denotatum. In case of a so-called histoiical causal c h a i n between the o i i g i n a l baptism of a t h i n g and the actual use of the p i o p e i name, the intentional content of the name may very well have changed at a ceitain stage of the c h a i n , notwithstanding the continuous causal connection of veibal b e h a v i o i . T h i s happened w i t h many p i o p e i namesas w i t h the name " M a d a g a s c a i , " w h i c h before M a i c o Polo referred to some part of continental A f r i c a a n d , I guess, w i t h many geneial names whose etymological meaning is different from its actual meaning.
1 1 12

Thus it becomes cleai, it seems to me, that a piagmatics of identifying denotata must be an account of causal a n d intentional chains simultaneously. It must be able, i n principle, to tiace the use of a name down to a situation where the meaning of the n a m e b e it a p i o p e i name o i a g e n e i a l c a n be ostensively intioduced, o i teintioduced, i n such a way that the usei can intentionally and reflectively identify an existing and hence also causally effective referent. T h i s is piecisely the way Peiice has explicated the function of identifying leal denotata w i t h the aid of indexical s i g n s . O n e should t h i n k that identification of leal referents as denotata of signs is eventually a case of complete integiation of a l l three dimensions of the tiiadic sign-function o i semiosis. Fot, i n the situation of identifying a denotatum as an instance of a general name, the meaningintention of the sign-usei must not only coincide w i t h the perception of the existing and causally effective referent but also fit i n w i t h the syntacticosemantical framework of a language. T h u s identifying a denotatum is a case of an encounter w i t h the leal wotld that is at the same time mediated by language and language-constitutive. Therefore it may seive as a patadigm-case f o i undetstanding, what I would call, the piagmatic completion of the linguistic turn i n philosophy w i t h i n the framework of a transcendental semiotics.
13

T h u s fai my semiotic account of the problems of sign-reference has been very roughly sketched. It passed over many intricate problems and, i n order to suggest the possibility of a transcendental-semiotic integration of piagmatics and semantics, it neglected intentionally the differences between different special cases of identifying denotata: thus for example, the difference between a subsumptive identifying of an exemplar of a w e l l - k n o w n species and, o n the other hand,

'PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

139

introducing a new general name, in order to be able to subsume a hitherto unknown phenomenon under this head; or the difference between these procedures and identifying the referent of a proper name. In what follows, I will discuss these problems a bit more thoroughly from the point of view of the following question: Is it really true that in all these cases a complete integration of the semantic and the pragmatic dimension of language comes about? That is to say, may one still claim, according to the linguistic turn, that mediation by language is a nontranscendable condition of the possibility of our identifying of something as something and thus of our intersubjectively valid knowledge of the world? One could conceive of objections to this claim from two perspectives: O n the one hand, one could think that identifying the real denotatum is a matter of the mental intentionality of a subject of perception that is independent from, or goes beyond, the meaning of the signs of a language-system. Thus the problem of identifying would lead us back to a mentalistic philosophy of intentionality, say in the sense of Brentano's philosophical psychology or of Husserlian phenomenology. This seems indeed to be the opinion that underlies certain tendencies of a new philosophy of intentionality. O n the other hand, one could rather argue that intentionality and linguistic meaning in the sense of intentionality belong together, whereas reference to the real denotatum would lead us beyond both of these notionssay toward grasping the real essence or nature of things which causally determines the meaning of signs qua extensionality of names. This latter contention seems to make up the point of the post-Wittgensteinian realistic semantics of Saul Kripke and Hilary Putnam. A n d this conception could also be considered incompatible with the linguistic turn of philosophy, it appears.
14 15

In the face of these diverging trends of the present theory of meaning qua reference, I shall try to show two things: First, there are indeed different empirical affinities, and hence priorities, of the three angles of understanding meaning, namely extensionality, intensionality, and intentionality, with regard to the problem of reference. But, notwithstanding these differences there is also an internal relationship between the three concepts of meaning, such that it is not allowed, in the last resort, to separate them from one another. In other words, although the theory of reference leads beyond semanticism (in the Carnapian sense and in a narrow sense of linguistic meaning or usage) it does not lead beyond the transcendental-semiotic postulate of an integration of semantics and pragmatics of language. I shall try to show this in a triangle-discussion, so to speak, with regard to the relations between the three concepts of meaning. To begin with, I want to argue that it is not possible to separate the notion of referential intentionality, say in perceptual identifying of something as something, from the notion of public meaning qua intensionality of (proper or general) names. This holds in my opinion, although it is "often possible and necessary to ascertain a difference between the factual intentionality of

140

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

perceptual reference and the conventional meaning qua intensionality of those wotds in the light of w h i c h the identification of the tefetent is possible. In this context, the factual intentionality may both fall short of the public meaning qua intension ot even transcend it i n grasping the teal essence of things. W h a t does this mean? Ftom a logician's point of view, I t h i n k , Ftege was fight to make an abstractive distinction between Sinn and Bedeutungor, for that matter, intensionality and extensionalityof names, on the one hand, and the psychologically relevant concept of meaning-intentions, f o i example, "ideas" i n the sense of John Locke, on the othet hand. H e theteby was able to separate methodically between ideal, "timeless," and "intetsubjectively v a l i d " meanings, to be dealt w i t h i n the fotmalizable context of logical ttuth transfer, and particulat subjective mental states, to be dealt w i t h in empitical psychology. If one were to atgue that, aftet a l l , different people can be in the "same" psychological statesay by grasping the same intetsubjectively valid meaningFtege could answet, I suggest, that this is just a postulate ftom the point of view of the logical semantics of Sinn. Indeed, a psycholinguist would nevet find empirically the s a m e state of m i n d in two different p e o p l e o t even i n the same human being at different t i m e s b u t only, at best, sufficiently similar states of intending the same public meaning; the lattet being, so to speak, the regulative principle of theit thought and the necessary institutional fiction of theit possible c o m m u n i c a t i o n by language. T h a t is to say, public meanings of sign-types of language, to be explicated as intensions, are indeed the ideal standatds of patadigms of meaningidentity and hence of intetsubjective (sometimes called "objective") validity of meaning. Platonism, always recurrent among mathematicians and logicians, is just an ontological hypostatization of the absttactive n o t i o n of public meaning ot intensionality, I suggest. T h i s m u c h ftom a logician's point of view. Epistemologically, howevet, the absttactive sepatation between p u b l i c meaning-intensions and subjective meaning-intentions must be suspended; the logical semantics of a Ftegean ot C a r n a p i a n type would lose its meaningful function if we did not suppose that the ideal, intetsubjectively valid meanings of language-signs can be grasped, i n principle, by every human being qua subject of meaning-intentions. Otherwise we could not undetstand that different people can agtee, i n principle, o n the intetsubjective validity of an atgument. Futthermore, it must be possible fot every subject of meaning-intention to contribute by his ot het meaning-intentions to the public meaning-intensions. Otherwise we could not understand that human expetience can constitute the content of public meaning-intensions of language-signs: only the human subject of meaning-intentions may realize the reference-dimension of public meanings by identifying real denotata. In short, what we have postulated before, an integration of abstract semantics by pragmatics, implies suspending the abstractive separation between public meaning-intensions and subjective meaning-intentions. However, after what we have said about the difference, i n principle, between ideal timeless meaning-intensions and empirical concepts of mental states, it must also be clear that the subject of grasping ideal meaning-intensions by

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

141

meaning-intentions cannot be a possible object of empirical psychology (or another discipline of empirical pragmatics). In other words, i n order to cope w i t h traditional third world Platonism of intersubjectively valid m e a n i n g intensions by a philosophy of m e a n i n g - i n t e n t i o n s , one needs to reintroduce the K a n t i a n idea of a "transcendental subject." (Every h u m a n being must be able i n principle, so to speak, to take over the role of the transcendental subject of cognition and meaning-intentionality.) T h i s is what Hussetl cleatly undetstood, I suggest, i n conttadistinction to the present revivers of so-called philosophical psychology (who sometimes even dteam of reducing intentions to something in the btain). But Husserl, i n accordance w i t h the ttadition of mentalistic transcendentalism prior to the linguistic turn, went so far as to suppose that a transcendental ego-consciousness that would survive the "bracketing" of the world (including language and communication w i t h othet subjects) could constitute, in principle, all timeless meanings by its intentionality. In contradistinction to this methodological solipsism, it has to be pointed out, i n my o p i n i o n , that the single subject of intentionality even fot his ot het self-understanding as " I " has to presuppose public, intensional meanings that must be always alteady carried by languagesigns. T h e subject of intentionality can at best c l a i m to share public meanings by sign-interpretations or to contribute to the publicly shatable meanings of language by his meaning-intentions. T h e transcendental subject-function of the single subject of intentionality lies precisely i n his or her proposing (carrying forward) those meaning-claimsfor example, i n the context of atgumentative discoursewhose intetsubjective validity may be confirmed by an indefinite community of sign-interpretation. Therefore the indefinite community of signinterpretation, i n accotdance, it seems to me, w i t h cettain suggestions of Chatles Peirce and Josiah Royce, is the definite ttanscendental subject of meaningintentions (and hence virtually of intersubjectively valid knowledge).
16

T h e point of this semiotic tiansfotmation of transcendentalism lies in its accounting for the empirical facts, I suggest. T h u s it is i n good accotdance w i t h the fact that i n modern societies the meaning-intentions of the single subjects of sign-interpretation, or of cognition i n the light of sign-interpretation, w i l l usuallythat is, i n case of the l a y m e n f a l l short of the public meanings (intensions and extensions) that ate valid i n the language-community, whereas s o m e t i m e s i n case of the creative expertsthey w i l l even transcend them in addressing, so to speak, the indefinite community of sign-interpretation. B o t h facts are possible, I suggest, since (because) intentionality, i n conttadistinction to public meaning qua intensionality, is needed fot realizing pragmatically the semantic dimension of sign-reference by identifying real denotata. In older to illustrate this point further, let us now consider the relationship between intensionality and extensionality. T h e classical doctrine w i t h respect to this problem says that the (ideal) intension of a name determines its possible e x t e n s i o n , and this should h o l d quite independent of the factual intentions of human beings. A g a i n , this tenet appears

142

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

to me to be correct as a logician's argument against any version of psychologism. A n d i n this sense one could be inclined to say that "meanings," that is, intensions and e x t e n s i o n s of terms, "are not in the h e a d , " to use Putnam's phrase. Putnam, however, does not use this slogan in order to defend the classical tenet of logical semantics against psychologism but rather, I suggest, i n order to question the classical position as a psychologistic one from an epistemological and ontological point of view. For h i m e x t e n s i o n s of terms may be different from intensions and intentions (both of them being taken as mental states and thus far as "something i n the head"); and i n this case extensions have priority over both intensions and intentions, since the latter two express only subjective states of knowledge, so to speak, whereas the extensions virtually represent the real essence of t h i n g s .
17 18

N o w , i n the fact of this challenge, a first thought might be that, nonetheless, there must be after all some internal relationship between the extensions and the intensions and intentions, respectively. For how should it otherwise be possible to speak meaningfully about extensions as meanings that are "not in the h e a d . " If the extensions were not determined at least by possible intensions, they could not be conceived as extensions of terms or names; and if they were completely different from our possible meaning-intentions, w h i c h may function i n our perceptual identifications of real denotata, then we could not k n o w anything about them and could not even m e a n them. I t h i n k indeed that at least o n the level of possible intentions, intensions, and extensions an internal relationship between the three concepts of meaning must be realizable. But we have already seen that this type of relationship does not dispense us from considering, and accounting for, important differences between the factual realizations of the three dimensions of meaning. N o w , w i t h regard to Kripke's and Putnam's point, this problem may be illustrated by a case in w h i c h we do not succeed i n subsuming a given m a t e r i a l l e t us say, some extraterrestrial stuffunder the head of a general concept. H o w should we identify i n this case the strange stuff as subject of a sentence by w h i c h we could state the very impossibility of its determination w i t h the aid of a possible intension of a term? If it were true that given things, that is, individuals and supposed exemplars of natural kinds, could only be identified w i t h the aid of possible intensions of termsas it indeed was the assumption of Leibniz and i n a sense of H e g e l t h e n we would need to know everything about the strange t h i n g i t s complete intensional d e f i n i t i o n i n order only to say that we do not know anything about it. For i n order to say this, we would need to identify the thing w i t h the aid of our conceptual knowledge about it. It was this paradoxical situation, w h i c h was exposed by New-Hegelians like Bradley and Royce, that provoked Charles Peirce to his explication of the semiotical and epistemological function of indexical s i g n s l i k e " t h i s , " "there," " h e r e , " " n o w , " " t h e n , " " I , " " y o u , " e t c . i n the context of identifying real objects of perceptions as subjects of any further determination w i t h the aid of general c o n c e p t s . A n d it was one of the m a i n points i n Kripke's and Putnam's
19

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

143

realistic semantics, I suggest, to connect this indexical function of identifying with the function of names (i.e., proper names and general names of natural kinds. T h a t is to say, K r i p k e and Putnam tried to detach the determination of the e x t e n s i o n of names from the i n t e n s i o n s or conceptual contents of names by instead connecting it w i t h the indexical function that a name takes o n by the "original baptism" of an i n d i v i d u a l s a y " A r i s t o t l e " o r an (exemplar of a) natural k i n d s a y " g o l d " or "water" or " f i s h . " In the case of identifying i n d i viduals by proper names, the determination of the extension by baptism would comprise everything that is identical w i t h this (indexically identified) individual in every possible world. A n d this would detach the extension of the proper name from any definite description that could constitute its intension o n the basis of the fates (destinies) of the baptized individual i n some possible world. In the case of a natural k i n d , the determination of extension by baptism would comprise evetything that is identical i n its real essence w i t h this (indexically identified) stuff i n every possible world. A n d this determination would detach the extension of the general namesay of "water" or " g o l d " f r o m any conceptual content that could constitute the intension of the term. In both cases, the name or term whose extension is indexically defined would be a "rigid designator," since its extension would be independent from the different and changing conceptual contents that constitute the intensions of (proper or general) names, that is, identical i n every possible world. N o w , against these proposals of indexical determinations or definitions of the extensions of names or terms, the following objections could be directed i n the name of the postulate of the internal relationship between all three concepts of meaning. First, it may be said that only w i t h the aid of i n d e x i c a l words, that is, without any support by intensional determination, it is not possible, o n Aristotle's and Peirce's account, to identify an individual or an exemplar of a natural k i n d . Indeed, an individual entity that is baptized i n such a way that its name could only be defined indexically by the word "this (there)" (Aristotle's T65EJI) or, for that matter, even a supposed natural k i n d that is baptized i n such a way that the extension of its name could only be defined by the phrase "everything that is identical i n its essence w i t h this": S u c h a thing would be a good example for an unknowable D i n g a n s i c h . N o w , this interpretation is obviously not i n accordance w i t h Kripke's or Putnam's intention to rehabilitate an epistemological and ontological r e a l i s m and even essentialism w i t h the aid of a new semantics. Indeed, this method would also not fit i n w i t h Peirce's idea of e n s u r i n g b y indexical d e f i n i t i o n t h a t something encountered as real, but u n k n o w n thus far, may be progressively determined conceptually as subject of a sentence. W h a t then would be the right method i n order to make use of a semantics of indexical definition of names? I t h i n k the first requirement of the method would be to make sure that the indexical definition that is connected w i t h the "original baptism" can be remembered and transferred by communication. For as we have pointed out before i n accordance w i t h Searle, the " o r i g i n a l baptism" and the communicative

144

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

transfer of the name must not be confused w i t h a causal reaction and its transfer, w h i c h could be the object of an external description of a "causal c h a i n " of behavioi. (Thus far the talk of a "causal theory of m e a n i n g " is at least misleading.) N o w , i n order to provide for the fact that the causal transfer goes along w i t h its control by an intentional ttansfei of meaning, we need a formulation of the indexical definition connected w i t h the o i i g i n a l baptism that goes beyond the above report about the encounter w i t h the Ding a n s i c h . H o w could it read? In order, for example, to refer by baptism to an individual i n such a way that the indexical definition can be remembered and transferred by communication, it would at least be necessary to detetmine the extension of the attached p i o p e i name by a phiase somehow like this: "the c h i l d that was baptized at the time t at place p by peison p . " N o w this is a definite description that provides a Ftegean Sinn that at least initiates the intensional detetmination of the extension of the p i o p e i name by f u i t h e i definite desctiptions. In the case of the " o i i g i n a l baptism" of some (supposed) exemplat of a natutal k i n d that cannot yet be subsumed undet the head of a geneial conceptsay by the name "baboo" it would not even suffice to provide a definite desctiption like: "the matetial that was baptized at time t and place p by peison p , " fot this would not provide a sufficient detetmination of the extension of a (supposed) natutal k i n d by i n dexical identification of its teal essence. In otdet to achieve this, one eithet has to complement the indexical definition of "baboo" by some picture (say by a photogiaph) o i to provide some desctiption of the phenomenal qualities (and relations of qualities) that point to the sttuctute of the given matetial, such that one could arrive at a definition somewhat like: "baboo" means everything that is identical i n its structure w i t h that matetial that was baptized at time t and place p by person p; and the structute of "baboo" may be desctibed by the following list of phenomenal qualities: . . . But if this proposal should be relevant, then it would appeal as if the intended detachment of the indexical definition of the extension of the name from its intension has already been proved impossible. T h i s verdict, however, would be precipitous i n a sense. For what has been shown thus far is only (again) that there must be an internal relationship between the indexical definition of the possible extension and the definition of the possible intension of the name. But this does not prevent us from providing a definition of the extension of a name that by its indexical dimension is partly independent ftom out piesent concepts (i.e., from the factual intensions of the wotds of out language) and thus open to a f u i t h e i intensional detetmination, say, by the progress of science. In this sense it may indeed be claimed that pragmatically the extension of all p i o p e i names (especially as they are used i n history) is not defined definitively by definite desctiptions (not even by a clustei of them!) but at least also by the open dimension of the indexical definition of proper names as "rigid designators." A n d it seems to be clear that only this type of (pragmatical and semantical) definition ensures the possibility of progress i n historical researchsay about

PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

145

the teal biogtaphy of Plato ot A r i s t o t l e . (This interptetation would not exclude, 1 suggest, that out definition of the meaning of " P l a t o " and " A r i s t o t l e " by a clustet of definite desctiption could ptovide a necessary c o n d i t i o n fot out use of these proper names. T h e proper names as rigid designatots would lie i n the fact that histotical petsons have attracted our cognitive intetest not just by their being born and baptized, but by theit having continued theit teal existence not in a possible world, but i n our factual world of history.) Similarly, it may be claimed that pragmatically the extension of all general names of natutal kinds (especially as they ate used i n natural science) is not defined definitively by the factual intensions (i.e., the conceptual content) of out terms but at least also by the open dimension of the indexical definition of the general names ot terms as "rigid designators." A n d again only this type of a (pragmatical and semantical) definition ensures the possibility of scientific progress. In order to illustrate this lattet point, I w i l l briefly discuss Putnam's theory of a "linguistic division of l a b o r , " w h i c h , it seems to me, is a very important empirical illustration for the possibility and necessity of an ovetcoming of the methodological solipsism of the ttaditional epistemology. T h e point of Putnam's theory may be summatized i n the following way: A t least i n m o d e m societies w h i c h ate characterized by a differentiated social division of labot there must prevail a corresponding linguistic division of labot. T h i s means that fot almost every patt of the first-hand knowledge of things by experience, and hence for the corresponding knowledge about the extensions of general names, there are experts who by far outtun the laymen. These experts therefore really detetmine the public meaning of wotds.
20

N o w this theory is rather ambiguous w i t h regard to our problem of the factual relationship between e x t e n s i o n s or intensions as determinations of linguistic meanings. Fot, i n one sense, the theory is i n good accordance w i t h the classical assumption that the extensions of terms are detetmined by theit intensions. T h e expetts, it may be claimed, ate just marking out the public standatds about the intensions of w h i c h the extensions ate logically dependent; and thereby they may even determine, i n a sense, the use of words by laymen, although these have no first-hand knowledge about the extensions of most terms. (This interpretation is indeed supported linguistically by the sttuctutalist theory of "fields" of meanings ot "wotd-contents." For this theory may explain to a cettain extent how laymen as competent native speakets may i n a certain sense correctly use words about w h i c h they have neither an adequate knowledge concerning the intensions nor the extensions. T h i s holds good, e.g., w i t h tegatd to my o w n use of the wotd " g o l d " or " h e m l o c k , " and even of words like "elms" and "beeches.") In brief, the c o m m o n use of a language just ptovides, so to speak, a ptagmatic proliferation of the collective knowledge about intensions of words and thereby enables laymen to dispose i n a sense of the knowledge of the experts. S t i l l there is another interpretation of the theory possible i n light of the realistic semantics of Kripke and Putnam. T h e experts, it may^be said, are not only competent for the factual standard determination of the intensions of terms

146

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

and thereby for the public use of language, but they are also i n charge of those (partly) indexical definitions of the extensions of terms as rigid designators that keep the dimension of intensional meaning of terms open to further determinat i o n ; and they w i l l usually even be a little bit ahead of the collective standard determination of intensions by their "tacit knowledge" by experience w i t h respect to those extensions of terms that ate not yet covered by the collective standard intensions. Thus far it may indeed be claimed that, insofar as scientific progress is determining the use of modern languages, there is a factual priority of the exploration of the extensions of words over the fixation of their intensions. A g a i n this statement does not of course refute our general claim that there is an internal relationship between possible ideal extensions and intensions of terms and that o n the level of this relationship the logical postulate of the determination of extensions by intensions must h o l d . But this claim must not be confused w i t h the claim (suggested by abstract logical semantics of constructed language-systems) that the extension of the words of a l i v i n g language may be determined once and for all by the well-defined intensions of the terms of a semantical system. For a constructed semantical system simply cannot keep up w i t h the pragmatic difference between the factual intensions and the factual extensions of the words of a l i v i n g language, w h i c h has to keep up w i t h the growth of h u m a n experience and especially w i t h the progress of science. T h i s shows only that the pragmatic difference between intensions and extensions requires a pragmatic completion of the linguistic turn of philosophy i n order to overcome the abstractive fallacy of logical semantics i n the C a r n a p i a n sense, as we have suggested previously. N o w the pragmatic difference between extensions and intensions of the words of a living language must obviously be realized by acts of identifying real denotata of words, that is, by referential intentionality. T h i s fact leads us to the last problem of our triangle-discussion of the relationship between the three paradigm concepts of meaning: W h a t about the internal and factual relationship between Kripke's and Putnam's indexical definition of the extensions of names and the concept of referential intentionality? After what we have said before about the role of intentionality i n the (semantic and pragmatic) context of indexical identifying, it seems to be clear that not only the possible but even the factual intentionality of our indexical references must be able to keep up w i t h the determination of meaning as extension. For it is just by the intentional content of our indexical definitions that we can open up a dimension of extensional meaning that transcends the factual intensions of our terms. A t first sight, this point appears to be a twist i n Searle's recent argument against Putnam's c o n t e n t i o n that "meanings are not i n the h e a d . " A n d I would indeed agree that the intentional content of an indexical definition must reach precisely as far as the extension of a term that is defined by the indexical definition; otherwise it could not determine the empirical conditions of its satisfaction. Thus far it may be claimed that meaning is " i n
2 1

PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

147

the head." (But this meaning of the metaphorical phrase can only be justified, in my opinion, if we at this point presuppose a transcendental-epistemological, and not an empirical-psychological, concept of intentional-meaning. The metaphysical paradigm or, respectively, prefiguration of what I called a transcendental conception would be Aristotle's or, respectively, Thomas Aquinas's dictum "anima quodammmodo omnia." It seems clear that Putnam did not have this conception in mind when he thought of "meanings in the head." He rather thought of contents of mind in an empirical-psychological sense; and thus far he is right in claiming that "meanings are not in the head.") But Searle not only claims to be right with regard to the meaning-equation of "intentional content" and "indexical definition" but also with regard to the meaning-equation of "intentional content" and (factual) intension and thus thinks it possible to defend the traditional thesis that the intension determines the extension on this presupposition. Thus far his argument against Putnam may be summarized as follows: P r e m i s e (Searle along with Putnam): Two people, say Jones on Earth and his Doppelgd'nger twin Jones on twin Earth have type-identical mental states in saying: "I have a headache," but the extension of "I" in their verbalized thoughts is different because it is determined by indexical definition. C o n c l u s i o n (Searle against Putnam): In both cases the intension determines the extension since the intentional contents (i.e., the self-referential concepts Jones and twin Jones have of themselves) are also different. It seems clear to me that Searle would be right in arguing this way against Putnam, ifand only ifit could be allowed to equate the different situationbound indexical meanings of " I " in the two different utterances of, respectively, Jones and twin Jones with two different general concepts, and hencein the case of our examplewith two different intensions of the same sign-type. Searle, it seems to me, is seduced into making these assumptions by considering the indexical, self-referential self-conceptions of Jones and twin Joneswhich are indeed different according to the different intentional contents of their verbalized thoughtsas different concepts and hence intensions of the term " I . " But this seems to me to be semiotically untenable since it in fact obscures the fruitful and indispensable distinction between general concepts or intensions of terms, whose meaning must not be situation-bound, and indexical definitions. A n d it thereby obscures the good point in Putnam's distinction between the indexical definition of the extension of a term from a situation-bound point of view and the conceptual contents or intensions of terms. But things are indeed different with regard to the intentional content of the utterance of an indexical definition. This content, which must only be partly indexical, in order not to be cognitively blind, as we have argued previously (see above, p. 146), may also be called intensional. Thus far Searle may be entitled to argue, first, that if by "intension" we mean intentional content then the i n t e n s i o n of a n u t t e r a n c e of a n indexical e x p r e s s i o n [my emphasis] precisely does determine

148

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

extension; and, second, that i n perceptual cases two people can be in typeidentical mental states . . . and theit intentional contents can still be diffetent; they can have different conditions of satisfaction.
22

But this point is, of coutse, i n accotd w i t h the c l a i m of tealistic semantics that it is possible to ttanscend the intensional scope of out concepts (i.e., of the factual intensions of out terms or, respectively, wotds) by indexical definition of the extensions of terms as rigid designatots. T h i s becomes clear from Searle's interpretation of Putnam's example concerning the meaning of "water" o n Earth and o n t w i n Eatth w h i c h accotding to Putnam is supposed to be intensionally identical, but extensionally different. It teads: T h i s indexical definition given by Jones o n eatth of "water" is defined indexically as whatevet is identical i n structure w i t h the stuff causing this visual expetience, whatevet that structute is. A n d the analysis fot twin Jones on t w i n eatth is: " W a t e t " is denned indexically as whatevet is identical i n structute w i t h the stuff causing this visual experience, whatever that structute is. Thus, i n each case we have type-identical expetiences but i n fact i n each case something different is meant. T h a t is, in each case the conditions of satisfaction established by the mental content ( i n the head) is different because of the self-tefetentiality of perceptual expetiences.
23

S t i l l this example is patadoxical precisely from the point of view of Seatle's concept of "intentional c o n t e n t . " For it obviously supposes that Jones and t w i n Jones may actually mean different things although they know nothing about this difference. If we univetsalize this supposition, we arrive at the conclusion, already suggested pteviously, that the indexical definition of the extension of the real (i.e., its definition i n the vein of Kripke's and Putnam's causal and tealistic theory of meaning) amounts to a definition of the unknowable Ding an s i c h . Fot we would attive at a situation where people could not know anything about that w h i c h they must define purely indexically, namely, the extension of the geneial name of the real w h i c h could read: "whatevet is identical in essence w i t h that w h i c h is causing (Kant says "affecting") this (i.e., my ptesent) expetience (e.g., of tesistance), whatever that structute i s . " A c c o r d i n g to Chatles Peiice, such a puiely indexical definition is nonsensical because it cannot show, i n principle, how the meaning of "identical in essence w i t h . . . this . . . " could be conceptually interpreted. Thus it teduces the meaning of the leal to the limit case of a bumping (of the w i l l of the I) against something i n the night (the tesistance of the non-I). N o w this is obviously not supposed to be the meaning of Putnam's ot, tespectively, Seatle's example. But it has to be noticed that Seatle's insertion of the phtase " i n structure" (instead of " i n essence") i n the indexical definition and his emphasis o n the "self-referentiality" of "perceptual experiences" does not change the paradoxical situation. O n the contrary, the paradoxicality of the example rests precisely on the suggestion that Jones and twin Jones ate considered to have the s a m e experiences, as far as their knowledge reaches, but different indexical (even self-referential) intentions.

'PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

149

T h i s shows in my o p i n i o n that the given example, taken abstractively, demolishes the good sense of both Kripke's and Putnam's realistic semantics and Searle's theory of intentionality. In order to avoid an abstractive fallacy and make visible the good point of both theories, it is necessary, I suggest, to considet the example not as a paradigm case but as a limit case w i t h i n the historical context of the development of human knowledge and of the meaningcontent of human languages. T h e given example must not be universalized as such but considered from the point of view of those people who already know (say, by a better theory of water than that of Jones and t w i n Jones) that Jones and twin Jones i n fact referred to different natural kinds (say, different species of watet and something else). But, in order to conceive of the possibility of such a development of knowledge and of language, the example of the l i m i t case is misleading since it cannot show, i n principle, how the fact that Jones and t w i n Jones were referring (on Searle's account, even somehow intentionally) to different matetials can be discovered and thus how the supposition of the semantic theories can be verified. T h i s becomes clear if we imagine the case where Jones detectssay, as a member of a scientific e x p e d i t i o n t h a t the "water" o n t w i n Earth is something quite different from "water" o n Earth. T h i s would possibly be an occasion for an "original baptism" i n Kripke's sense. A s I suggested previously, i n such a case the indexical definition of the extension of the name must not only contain phrases like " i d e n t i c a l i n structure" and "causing this visual experience, whatevet that structure i s , " but it must be supplemented by a picture or a description of the structure of the visual (or, for that matter, nonvisual but sensual) experience, say, by a list of the qualities (and relations of qualities) that appear to make up the structure of the causally effective entity that is pointed to by " t h i s . " It has to be noticed that such a supplementation of the indexical definition of the extension of the name would not mean that the discoverer could already provide a conceptual definition by w h i c h he could subsume the phenomenon under the head of some class. But he could indeed make his indexical definition of the extension epistemologically relevant. For he would provide a meaning to the intentional content of his definition of the name that is neithet purely indexical (and hence cognitively blind) nor completely conceptual (and hence not open for the still u n k n o w n real extension of the name qua rigid designator). But if this characterization of the role of the representation of the phenomenal (qualitatively given) structure is relevant, then the whole dichotomy of indexical definition and conceptual definition (in the sense of a possible subsumption) does not suffice i n order to deal adequately w i t h the problem of the intension of the intentional content of a baptism-protocol. It does not sufficiently explain how we can introduce into the scientific discourse a newly discovered object of experience as subject to further determination. T h e theories thus far discussed do not yet suffice, it appears, to explain semiotically and epistemologically the pragmatic difference and the possible c o n t i n u u m of meaning between extensions, intensions, and intentions.

150

PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

In what follows I shall thetefote try to review the whole triangle-discussion in terms of a ttanscendental semiotics that dtaws heavily on some basic conceptions of Chatles Peiice. A n d , before entering into this conclusive discussion of the problem of teference, I want to recall out leading question (see above, p. 135) as to whethet it may be shown that the identification of real referents as denotata of our signs amount to showing that the linguistic turn of philosophy may be completed by a pragmatic turn, that is, by a pragmatic integtation of the semantic dimension of sign-reference (see figure 5.1, left side!). In order to introduce the tools of a Peitcean semiotics into the discussion of out problem, let us return to our example of the "oiiginal baptism" of a piece of extraterrestrial material by the name "baboo." As I pointed out previously, the baptism cannot be done by just pronouncing the name "baboo" as a physical reaction to the causal affection by the stiange stuff, but it has to provide something like a baptism-protocol; that is, a little story about the procedure of identifying the referent of "baboo" which makes it possible to remember and to communicate the new name in such a way that its referent can be intentionally teidentified as that which was intentionally identified as cause of his expeiience by the fiist discoverer. In a Wittgensteinian vein one could say: The "oiiginal baptism" has alteady to be performed along the lines of a public lule of identifying that can be followed in a sense by the fiist discoveiet as well as by all potential reidentifiers. Thus far I would claim that the procedure of identifying within the context of the "original baptism" is already a step beyond the piivacy of a prelinguistic procedure of ascertaining evidence foi just one consciousness in the sense of methodological solipsism (from Descartes and Locke through Husserl). It is already the entrance into the range and realm of a public language game (Wittgenstein) and of the indefinite community of sign-interpretation (Peirce). But it has to be noticed that the tule of the language game of "oiiginal baptism" in the sense of Kiipke precludes the possibility that the meaning of the name "baboo" might be explicated in terms of some factual use of language. Since the meaning of "baboo" is that of a ligid designatot whose extension is the same in any possible woild, it must tianscend, in a sense, the range of any language game that is centered around the rules of a factual use of language. But this, it seems to me, is an essential feature of every explication of meaning on the level of the scientific (ot philosophic) language game, a feature that is only made visible in a special way by the case of introducing a new and partly open meaning by the procedure of "oiiginal baptism." In principle, no explication of the meaning of a scientific termsay of "heavy" or "simultaneous" on the level of physics, or of "justice" or, for that matter, "meaning" on the level of philosophycan be reached by just describing a given use of language. Rathet we need an experiment of thought by which we could imagine how the term that is to be explicated could or had to be interpreted and hence used on the giound of all conceivable experiences or practical consequences. This point of Peitce's "pragmatic maxim," in my opinion, constitutes the difference and the superior-

PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

151

ity of a normative ttanscendental-pragmatic approach to meaning-explication over a Wittgensteinian pragmatics of language use, not to speak of Morris's behavioristical pragmatics.
24

Let us then try to give a Peircean account of the procedure of identifying that goes along with the discovery and original baptism of what is called "baboo." The discoverer, I suggest, might say to his companion (or, if this is not possible, to himself) something like this: This thing over thereunder the big ttee in the foregroundlooks so and so (to be specified by describing the perceptible qualities and relations of qualities that make up the phenomenal structure). O n the basis of these perceptible qualities, 1 don't know how to determine what it is (i.e., under which general concept it could be subsumed). That is, I find it impossible at present to provide a conceptual subsumption of the given stuff on the basis of an abductive inference of the form: "This there is so and so; what is so and so might be an exemplar of ' A ' . " Hence, in order to provide for the possibility of a later determination on the basis of a reidentification, I will give it a name: I hereby baptize the given stuff by the name "baboo." I thereby define the extension of "baboo" as "being everything that is the same as this over there which now causes my present experience by presenting the following phenomenal structure (to be specified by a picture or by a list of phenomenal qualities)." The semiotically interesting feature of this account of an original baptism, and that which goes beyond our former accounts, lies in its recourse to two classes of nonconceptual signs (in Peirce's terminology, "indices" and "icons") whose function is connected in our baptism-protocol in some way with the function of signs of general concepts (i.e., "symbols" in Peirce's terminology). Thus all three classes of signs together will constitute, so to speak, the intension (in Searle's sense) of the "intentional content" of out protocol. Let us illustrate that in more detail: The use of linguistic indices (or rather quasi-indices, as has still to be shown) like "this" is needed, in order to testify to the existence of the given stuff and its causally affecting and, at the same time, being intentionally identified by the sign-user. The use of further linguistic indices like "over there" and "now" or "at present" concretizes the situationreference of the protocol by indicating the space-time relation of the phenomena to the sign-user and vice versa. Finally linguistic indices like "I" and "hereby" (and possibly "you" for addressing a companion) indicate the performance of the speech-act of baptizing as part of an explicitly or implicitly communicative situation that may serve as the point of departure for a further process of communication about the object of the original baptism. Thus far the recourse to indices reconfirms our prior analysis of the "original baptism" as utterance of an "indexical definition." But, in our example, the function of the linguistic indices is supported and supplemented by the function of icons (e.g., photographs) or what could be called linguistic quasi-icons in the sense of a Peircean semiotics. This function obviously makes up a new feature of

1 52

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

our Peircean account w h i c h is not easy to explicate. Linguistic icons ate used i n the desctiptive patts of out ptotocol; and they ate functioning, so to speak, inside the conceptual ptedicatots by w h i c h the discovetet is desctibing the given qualities (and relations of qualities) that make up the phenomenal structure of the matetial that is to be baptized. T h i s function of linguistic icons belongs to the indexical definition of "baboo" insofar as the qualities that make up the phenomenal structure are not desctibed as pure phenomena of relation-free "suchness" (i.e., of "firstness" i n the sense of Peirce's theory of categories). T h a t is to say, the phenomena ate not desctibed as merely possible qualities, but as qualities given fot the consciousness of the discoverer whose attention is drawn to them by the function of the " i n d i c e s . " (The function of the indices corresponds to Peirce's category "secondness" w h i c h expresses a dyadic or two-place telation, e.g., the encountet between the 1 and the non-1.) But the function of the linguistic icons belongs also togethet w i t h that of the conceptual ptedicatots, inside w h i c h they ate functioning as quasi-icons in describing the qualities of the given phenomena. By this quasi-iconic function the conceptual ptedicatots of the desctiption ate, so to speak, recharged w i t h a meaning-evidence that they can only acquite i n the situation of perception but lose as abstract conceptual ptedicatots. T h i s means that the quasi-iconic funct i o n of ptedicators i n the context of desctiption (e.g., perceptual judgments) does not so m u c h setve to subsume a given phenomenon undet the head of a general concept ot class but tathet, i n advance of that absttactive logical opetation, to gtasp and ptesent the given quality (ot perceptible sttuctute) of the referent. Therefore linguistic icons cannot function i n the context of absttact ttue ot falsepropositions about facts but only inside of perceptual judgments about what is actually given w i t h "phenomenological" ot "phaneroscopic" evidenceas i n our example of the carefully described qualities of the sttange stuff that cannot be subsumed under the head of some concept ot class as yet. It is this function of the linguistic icons that obviously makes it possible to suppott and supplement the indexical definition of the extension of "baboo" by a desctiption that is not yet a conceptual subsumption but that makes the indexical definition cognitively (and hence epistemologically) relevant by preparing, so to speak, for a later conceptual subsumption. Thus fat this element of out semiotic account goes beyond the purely indexical account of the definition that belongs to the oiiginal baptism, but it is obviously i n accord w i t h Kripke's and Putnam's approach as well as w i t h Searle's conception of the " i n t e n s i o n " of the "intentional content" of the "uttetance of the indexical d e f i n i t i o n . " It covets indeed both the causal and the intentional aspect of referential identifying, and it especially supports the K r i p k e a n c l a i m that by baptism the real essence of individuals and of natural kinds is somehow grasped by and integtated into the names qua rigid designatorsin conttaposition to the nominalistic claim that all description w i t h the aid of geneial concepts amounts merely to a linguistic arrangement of terms i n the service of pragmatic purposes.

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

1 53

T h i s point of convetgence of our semiotic approach w i t h that of essentialistic realism is certainly i n agreement w i t h the general spirit of Peirce's emphatic antinominalism. But i n the face of this fact the crucial question of our present apptoach atises i n a new and acute version: the question whethet this semiotic approach must not rathet represent a contraposition (with regard) to the linguistic turn of contemporary philosophy. T h i s has i n fact been claimed for Kripke's and Putnam's realistic theory of reference; and if this apptoach may be integrated into and, so to speak, "aufgehoben" i n a Peircean semiotics, then the question arises, how the latter may be interpreted as a pragmatic completion of the linguistic turn, even i n the sense of a ttanscendental semiotics. M y answet to this question would be: A transcendental semiotics based o n the Peircean transformation of Kant's "transcendental l o g i c " provides a t h i r d way beyond and apart from the traditional alternative of nominalism and essentialistic realism or, for that mattet, of commonsense realism and transcendental criticism. It does so especially by a conception of semiotics that is capable of going beyond and mediating between the semanticist version of the linguistic turn, w h i c h is based only o n the function of conceptual symbols (if not o n the notion of abstract syntactico-semantical frameworks), and, o n the other hand, the position of prelinguistic ontology and transcendental philosophy (including Kantianism and Husserlian phenomenology). 1 can suggest this point i n the present context only w i t h regard to the problem of reference or, respectively, of the triangle-discussion of the three concepts of meaning.
25

Thus fat 1 have wanted to suggestalong w i t h P e i r c e t h a t the function of linguistic indices and icons, o n the one hand, transcends any possible function of conceptual signs by fixing or restticting the language to causally effective real objects of perception and to their structural qualities. T h i s function of the nonconceptual sign-types of language, w h i c h is bound up w i t h the context of the perceptual situation, may indeed account semiotically for the indispensable evidence-basis of human cognition: T h a t is, i n Peircean tetms, the dyadic relation ("secondness") of the l's clash w i t h the causal affections of the non-1 and the monadic (relation-free) suchness of the given phenomena ("firstness"). Thus far Peircean semiotics refutes the precipitate contention (of logical semantics and semanticist philosophy of science, including even Poppetianism) that evidence may be reduced to just a psychological feeling and hence means as much as n o t h i n g fot epistemology i n the face of the fact that a l l intersubjectively valid results of perception are impregnated by linguistic interpretation or, respectively, by theories. Peircean semiotics indeed saves from modern semanticism the truth-core of a Husserlian phenomenology of evidence and especially of the A t i s t o t e l i a n c l a i m that perceptual judgments are incorrigible i n a sense, that is, w i t h regard to given qualities i n the sense of secondness and firstness. (Peirce offers also an evolutionist explanation of this property of perceptual judgments: They function as boundaries (borders) and vehicles of transition between, o n the one hand, natural processes of sign-information, not to be influenced by man, and, o n the other hand, processes of sign-interpretation that

154

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

make up the subject ot topic of the notmative semiotic logic of s c i e n c e . ) O n the othet hand, however, Peircean semiotics also provides crucial arguments fot the transcendental-ptagmatic completion of the linguistic turn. For, in conttadiction to Hussetl's ptelinguistic vetsion of phenomenology, Peircean semiotics holds that evidence of given phenomena i n the sense of "firstness" and "secondness" is not (yet) the same as (intetsubjectively valid) knowledge, as long as it is absttactively sepatated from the symbolic interpretation w h i c h has to complete out cognition in the sense of the category thitdness, that is, of the conceptual mediation of the intuitively given phenomena w i t h undetstanding ot, respectively, teason. (Here Peitcean semiotics amounts to a ctitical reconstruction of the epistemological development from Kant to Hegel.) Correspondingly, Peircean semiotics cannot agree either w i t h a ptelinguistic version of real-essence ontology or naturalistic causal theory of reference. For it must insist on the fact that even indexical and iconical signs, insofar as they function w i t h i n the context of the "original baptism" (i.e., of indexical definition and, beyond that, of a structural description of the phenomenon), are linguistic signs, after a l l . T h i s means that their function is interwoven w i t h that of conceptual signs (i.e., symbols) i n a twofold way.
26

First, it has to be noticed that the whole text of our baptism-protocol, that is, the text of our indexical definition and moreover the text of the accompanying structural desctiption, must combine the function of all three classes of signs, namely, i n d i c e s , i c o n s , and symbols i n Peircean terminology, i n order to constitute the intension of the intentional content of that protocol. Second, it is time now to make clear what it means to say that the linguistic indices and, respectively, icons are only quasi-indices and quasi-icons. In a sense one could say that i n both cases they function only inside conceptual signs (i.e., symbols). Thus, linguistic indices like " t h i s , " " t h a t , " " h e r e , " " t h e r e , " " n o w , " " t h e n , " " 1 , " " y o u , " etc. do not function like indices outside of language, say smoke as an index of fire or the pulse as an index of blood-pressure. T h e y rather include an element of conceptual thitdness, w h i c h even determines the type of their situation-bound secondness. Thus, the meaning of " t h i s , " i n contradistinct i o n to " t h a t , " is somehow determined by the conceptual distinction between "thisness" and "thatness" w h i c h is not situation-bound. T h e same holds w i t h regard to "here," i n contradistinction to "there" etc. Nonetheless, Hegel was of course semiotically wrong when he acknowledged only this conceptual part, and thereby denied the situation-bound indexical part, of the meaning of these linguistic indices i n his chaptet o n " S i n n l i c h e Gewissheit" ("sensuous certainty") i n his P h a e n o m e n o l o g i e des G e i s t e s . E v e n his antagonist Ludwig Feuerbach, who insisred o n the evidence of intuition, o v e r l o o k e d l i k e H e g e l t h e fact that there are linguistic signs, different from conceptual symbols, whose function is precisely to inttoduce and integtate the situation-bound evidence of intuition into the conceptual meaning of language. S t i l l , Hegel's talk of the "indetermined immediate," w h i c h , on his account, is denoted by the indexical signs, would be correct if it were not the case that the

'PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

1 55

indexical signs are capable, within the context of the actual situation, to direct our attention (and intention) to given qualities (i.e., "firstness"), possibly to qualities of hitherto unknown phenomena. This leads us to the other systematicsemiotical connection of sign-functions that was recognized by Peirce: that of a quasi-iconical grasping and presenting of qualities (in the sense of firstness) inside the conceptual meaning of predicators (i.e., thirdness). As has already been suggested previously, it is the internal connection between all three types of linguistic signs that is the basis of the openness of the living language to an enrichment of meaning by human experience, especially by the progress of the sciences. In this context, it has to be emphasized that the integration of iconical meaning into the conceptual intensions of predicators within the context of the intensions of baptism-protocols remains semantically effective beyond that stage. It operates as a normative steering-function towards reaching the ideal "logical interpretant" throughout the indefinite process of sign-interpretation that has to be postulated on Peirce's account of meaningexplication. Thus, one might conceive of the entire history of the word "heavy" from its introduction by an ostensive definition of the underlying quality through its further determination by Newton's and finally Einstein's theory of gravitation. A n d also the later stages of such a process are not merely determined by conceptual interpretation but also by new situational confrontations (e.g., of the experiential consequences of theories) with the phenomena and hence by new occasions for an integration of all three sign-functions. O n considering this semiotical theory of an integtation of the conceptual and nonconceptual sign-functions, it becomes clear, I think, that this approach, which leads beyond abstract semanticism and saves phenomenological evidence, does not abandon, however, the linguistic turn in philosophy but rather completes it. It does so by means of a pragmatic integration of the semantic reference of linguistic signs that may be called transcendental-pragmatic for the following reasons: First, the Wittgensteinian point of a transcendental semantics, which 1 have introduced at the beginning (see p. 132) as a criterion for the linguistic turn, has not been abandoned but onlyindeedwidened by the paradigm of a baptismprotocol. This becomes clear only if we consider the intentional content of the original baptism, which may be expressed by a ceremonial speech-act of the first discoverer that sounds like this: I hereby baptize the material over there by the name "baboo," defining the extension of this name as comprising everything that is the same as this over there which now causes my present experiences by presenting the following phenomenal structure . . . We are here no longeras in Wittgenstein's dictumonly concerned with a prepositional sentence that describes a fact, but rather' with an expanded performative sentence that reflectively describes and thereby also performs the speech-act of identifying, baptizing, and defining the given name. Thus, not

1 56

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

only an objectively given state of affaits (a Wittgensteinian "fact") is desctibed, but rathet a histotical situation and a human response to it. Nevertheless, the transcendental point of the Wittgensteinian statement is preserved. For it is not possible, in principle, fot the performer of the o i i g i n a l baptism, who theteby has to provide the initial stage of an interpretation process, to go somehow beyond the limits of his language game. In order to achieve the same identification, baptism, and definition, he has to repeat the same performative sentence. Second, as we have already intimated previously, the tule and the tange of the language game of the oiiginal baptism ate somehow continued throughout the furthei interpretation-process, notwithstanding the possible progress in meaning-determination, i n otdei to ensure the possibility of an intentional teidentification. N o w , if we univeisalize this point w i t h tegatd to identifying and interpreting what is called the real, we must tecognize that there is no possibility, at any stage of the process, to go beyond the limit of the pettinent language game. Just by pragmatically integtating the phenomenological dimension of sign-tefeience, the transcendental-pragmatic point of the linguistic turn has been strengthened and shown to be nontianscendable. T h i s may be clarified further if we ask f o i the ttanscendental subject of the entite achievement of identification and interpretation. It obviously is no longet the ttanscendental consciousness of the 1 i n the sense of the ttaditional mentalism and methodological solipsism; although the ttanscendental consciousness of the 1 still retains the function of ascertaining evidence together w i t h the function of intentionality and also of the synthesis of apperception in the sense of K a n t . However, the transcendental subject i n the sense of transcendental semiotics must not be defined w i t h respect to the evidence of experience and meaning-intentionality but tathet w i t h tespect to the possible intetsubjective validity of meaning-interpretation and hence of possible knowledge. It has to be capable, in principle, of functioning as subject of an ultimate consensus about sign-interpretation and thus about all conceivable truth-criteriaas evidence of correspondence between intentions and given phenomena, coherence of sentences or theoties, piagmatical fruitfulness of assumptions or strategies, e t c . In brief, it can only be the indefinite, ideal community of sign-interpretation, w h i c h o n the level of the atgumentative discoutse could arrive at the "final o p i n i o n " about the real, as we must postulate by a "regulative idea" whatever the facts about the future might be. Consideted as referent w i t h tegatd to this subject of sign-interpretation, the teal must no longer be defined as the unknowable Ding an s i c h , in the sense of Kant, but r a t h e t w i t h P e i i c e a s the indefinitely knowable that can nevet be factually k n o w n definitely. T h i s definition i n my o p i n i o n provides the point of an "Aufhebung" of commonsense realism into transcendental semiotics; fot the point of the f o r m e r t h a t the real is independent from anybody's thought about i t i s quite compatible w i t h the point of the latter: That the real as identifiable and conceivable must be the object of sign-interpretation. T h i s position may also be marked by the label of a meaning-critical realism,
27 28

PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

1 57

THE INTEGRATION OF SEMANTICS AND TRANSCENDENTAL PRAGMATICS WITH REGARD TO THE PRAGMATIC DIMENSION OF SIGN-USE U p to now we have considered primarily the problem of overcoming the absttactive fallacy of semanticism w i t h regard to Morris's semantic dimension of sign-reference, that is, of identifying real referents as denotata of signs (see left side of figure 5.1). Since, however, this problem turned out to be one of a (transcendental-)pragmatic integtation of semantics, we became already c o n cerned w i t h Morris's pragmatic dimension of semiosis, that is, w i t h the use of signs by an interpreter (as a member of a community of interpretation) and in this context w i t h self-referential intentionality (e.g., of the ceremonial speechact of "original baptism"). In what follows, we shall consider primarily the problem of overcoming the absttactive fallacy of semanticism w i t h regard to Morris's pragmatic dimension of semiosis, that is, w i t h primary regard to intentionality and its expression by speech-acts as communicative acts (see tight side of figure 5.1). W i t h regard to the dimension of sign-teference, the abstractive fallacy of semanticism came about by especially overlooking the problem of a pragmatical integration of semantics, fot example, of dealing w i t h the problems of intentional identifying of real referents and w i t h the pragmatic difference between (factual) e x t e n s i o n and (factual) intensions of linguistic terms. W i t h regard to the dimension of sign-use, the abstractive fallacy of semanticism comes about rathet by overlooking the possible semantical integration of pragmatics, that is, the linguistic aspect of the meaning-intentionality of speech-acts. In order to undetstand this point, we have to remember that the original distinction of the semiotic dimensions in Morris's F o u n d a t i o n s of the T h e o r y of Signs was conceived primarily as a supplementation of Camap's conception of abstract syntactico-semantical frameworks of constructed languages. N o w , o n these conditions, there is no need for a linguistic semantics of the pragmatic use of language, since a C a m a p i a n semantical system allows only for propositional sentences whose function is metely that of a representation of states of affairs. But things are quite different w i t h natural languages, w h i c h serve not only the purpose of representation but also that of expressing the pragmatic by nonpropositional sentences. A good absttactive distinction between the different functions of a natural language was offered by K a r l Biihler i n his S p r a c h t h e o r i e . H e distinguished between the tepresentation-function of propositions and, o n the other hand, the functions of self-expression and of communicative appeal. But even Biihler was not prepared to account for the latter two functions as those of linguistic "symbols" but only as functions of "symptoms" and "signals." A c c o r d i n g l y , he and K a r l Popper, who follows h i m , consider the two pragmatic functions as the "lower functions" of language w h i c h man has i n c o m m o n w i t h the animals, in contradistinction to the representation-functions of propositions as carriers of possible t r u t h . T h u s far Biihler's and Popper's account is still i n accord w i t h Camap's abstract semanticism, since it conceives of the function of
2 9 3 0

158

"PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

self-expression and appeal not as possible intentional and symbolically expressed dimensions of linguistic meaningto be taken into account, e.g., in a theory of argumentationbut as merely pragmatical functions to be dealt with only in empitical psychology and social science. A decisive bteakthrough towatds reflecting the nonptopositional but nonetheless symbolical, and hence pragmatical a n d semantical, functions of natural language was achieved, I suggest, by J. L. Austin's discovery of the "performative" phtases as linguistic expressions of "illocutionary acts." This holds especially with regard to his late insight that the linguistic expression of the "illocutionary force" of speech-acts by performatives is not only a matter of institutionalized, ceremonial formulaslike those of baptism ot marriagebut is a possibility, in principle, of making explicit the potential ptagmatic force of every sentence whose utterance would constitute a speech-acteven of a constative, or assertive, act. Why is this detection of the function of the performatives a decisive breakthrough that goes beyond Buhlet's and Popper's understanding of the nonrepresentative functions of language?
31

The point I have in mind was brought out partly by Sttawson, partly by Searle, and partly by Habetmas. Let me explicate each contribution in turn and allow me to go sometimes a little beyond the presumable self-understanding of the authors mentioned. First, Peter Strawson brought together Austin's theory of "speech-acts" and Paul Grice's theory of "intentions." In doing so he tried to understand Austin's distinction between "illocutionary" and "perlocutionary" acts as indicating different kinds of intentions, so that the former are merely expressions of genuine meaning-intentions by public language-signs, whereas the latter are a kind of covertly strategical acts whose intentions cannot be made public by conventional signs of language in order to fulfill their purposes.
32

Already in Strawson's conceiving of illocutionary acts as expressions of meaning-intentions, it becomes clear, I think, that the nonrepresentational functions of speechthat is, self-expressions and appeal in terms of Biihler's theoryare not merely functions of "symptoms" and "signals," as is the case with animal behavior (or, with humans, in the case of nonintentional expressions of psychical states or interpersonal relations that are studied in psychology and especially in psychopathology). In contradistinction to these subconscious functions, the nonrepresentational functions of speech as, for example, of performatives, obviously may symbolically express self-referential meaningintentions and thereby become the medium of a public expression of selfreflection and responsibility with regard to interests and social claimseven with regard to the truth-claim that in statements is connected with the representational function of propositions. For this truth-claim may be explicitly expressed by the performative sentence "I hereby affirm . . . . " Furthermore, with Strawson's distinction between illocutionary acts and (covertly strategical) perlocutionary acts, it becomes clear that the genuine meaning-intentions, which may be expressed by performatives (i.e., by nonprop-

PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

159

ositional types of sentences), are different from those intentions that cannot at all be expressed by the public medium of language, namely those of strategical intentions, say, of suggesting a certain image of the speaker. B o t h of these Strawsonian points, I believe, were pointed out once more by J o h n Searle i n his book Speech Acts. For Searle made it clear i n his book that genuine meaning-intentions, w h i c h are to be understood as direct intentions of illocutionary acts must also be capable of being linguistically expressed by explicitly performative sentences. Searle made this point w i t h his "principle of expressibility," w h i c h states that "whatever can be meant can be s a i d , " w i t h the following qualifications: . . . even i n cases where it is i n fact impossible to say exactly what I mean it is i n principle possible to come to be able to say exactly what I mean. I can i n principle if not i n fact increase my knowledge of the language, or more radically, if the existing language or existing languages are not adequate to the task, if they simply lack the resources for saying what I mean, I can i n principle at least enrich the language by introducing new terms or other devices into i t .
3 3

T h i s qualification is supplemented by another, that the principle of expressibility does not imply that it is always possible to find or invent a form of expression that w i l l produce all the effects i n the hearer that one means to produce; for example, literary or poetic effects, emotions, beliefs, and so o n .
3 4

I would understand these qualifications as stating something like this: Corresponding to what I have previously called the pragmatic difference between the fixed intension of terms and their possible extension as names of real referents, there is also a pragmatic difference between the conventional linguistic rules for expressing illocutionary acts and the meaning-intentions one may wish to express by speech-acts. But, as i n the case of the triangle-discussion, we must also i n this case take into account something like an internal relation between possible meaning-intentions, and possible conventional linguistic meanings. T h i s means that, i n both cases, one may suppose that a living language is open, i n principle, to a public expression by conventional sign-types of those meaningintentions that cannot be actually expressed i n such a way as yet. In my o p i n i o n , these qualifications of the principle of expressibility are still to be supplemented by the following clause ( w h i c h allows for a compensation, so to speak, of the pragmatic difference by the communicative competence of the speaker): W i t h i n the situational context, it is always possible to express genuine meaning-intentions by nonconventional signs, for example, by suggesting n o n literal, occasional meanings of conventional linguistic signs or by the use of paralinguistic or extralinguistic ad hoc signs. But also i n these cases the possibility of sharing a public meaning, at least by the speaker and the hearer, is dependent o n the presupposition of the conventional meaning of sign-types of language by the occasional meanings of the ad hoc signs; thus, for example, o n

160

"PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

the presupposition of the literal meaning of words by their metaphorical meaning. A n d beyond that it is dependent, I think, on the fact that shatable meanings of ad hoc signs ate based on successful ad hoc conventions about the use of the ad hoc signs, so to speak. (I need not speak in this context of nonlinguistic sign-typesas, for example, those of signal-languages whose meanings are parasitic on those of conventional linguistic sign-typesot of the use of signs by animals whose valid understanding as meaningful for us humans is always dependent on a reconstruction in the light of our understanding of conventional linguistic meanings.) In brief, in as far as the occasional meanings of ad hoc signs may be understood as public ot intersubjectively valid meanings, they ate in fact both dependent on conventional linguistic meanings and virtually supplementing those meanings by ad hoc conventions about public meanings.
35

O n the condition of these qualifications, I think, Seatle's "principle of expressibility" is valid; for it provides indeed the only critetion fot a demarcation between genuine meaning-intentions, which may be expressed, in principle, by illocutionary acts, and sttategical pseudo-meaning-intentions, that is, ditect intentions of perlocutionary effects which cannot be expressed by illocutionary acts and hence cannot become publicly shatable (and thus intersubjectively valid) meanings at all. O n the condition of the "principle of expressibility," ditect intentions of perlocutionary effects, and, in this sense, attempts at "perlocutionary acts," can indeed only be understood as intentions of purposive rational actions that incidentally use language-signs as means or instruments for suggesting certain conclusions to the heatetconclusions that cannot be expressed publicly, in ptinciple. In accordance with the preceding, I think that Searle's "principle of expressibility" provides a ctucial atgument fot the pragmatic completion of the linguistic turn in philosophy in the sense of my contention. Fot it shows, as Seatle puts it himself, that there ate . . . not two irreducibly distinct semantic studies, one a study of the meaning of sentences and one a study of the performances of speech acts. Fot just as it is part of our notion of the meaning of a sentence that a literal utterance of that sentence with that meaning in a certain context would be the performance of a particular speech act, so it is part of our notion of a speech act that there is a possible [my emphasisl sentence (or sentences) the uttetance of which in a certain context would in virtue of its (or their) meaning constitute a performance of that speech act.
36

In my opinion, it is an implication of this position that genuine meaningintentions must be capable, in ptinciple, of being articulated publicly (i.e., as meaning-claims whose intersubjective validity can be redeemed in principle by the indefinite community of interpretation). A n d this would not be possible if there were no constitutive rules of "the" language, realized conventionally by some special language (i.e., "language"), by which the meaning-intentions can be articulated publicly. Every special language, so to speak, is an institution for articulating meaning-intentions publicly, and since every special language is

PRAGMATIC T U R N " & T R A N S C E N D E N T A L SEMIOTICS

161

open to articulating any possible meaning-intention, it is, so to speak, a quasi-transcendental institution providingin a still contingent, histotical formthe conditions of the possibility of intersubjectively valid meanings. This is of course a restatement of the point of the linguistic turn in philosophy; and from this restatement it follows already, I suggest, that (intersubjectively valid) meanings cannot be reduced to prelinguistic intentions, fot the simple reason that they ate the necessary public articulations of genuine meaning-intentions, which of course are, from an empirical-genetical point of view, the origins of meaning. This point has to be defended, I suggest, against those attempts at reducing meaning to prelinguistic intentions that have become fashionable again in recent years in a revived "philosophical psychology" or "philosophy of mind." A n d in this context it has even to be defended, as has yet to be shown, against the psychologistic turn of John Searle himself in his last work, Intentionality. This brings me to the third part of my story.

THE

O V E R T U R N I N G O F T H E P R A G M A T I C T U R N BY R E D U C I N G M E A N I N G TO

PRELINGUISTIC INTENTIONS A N D T H E POSSIBILITY O F ITS R E F U T A T I O N

REMARKS ON GRICEAN "INTENTIONALIST SEMANTICS" One strand of the new psychologistic turn is represented, as fat as I can see, by Paul Grice's "intentionalist" theory of meaning and the long series of attempted improvements and elaborations of this approach which, via S. R. Schiffer and J. Bennett, leads to an integration of D. Lewis's theory of "conventions" and thus to a strategical game theory of petlocutionary interaction.
37 38

It is this strand of communication-theory that, in my opinion, may in the first place be characterized by calling it an "overturning of the p r a g m a t i c turn of meaning-theory." For the leading strategy here is, from the beginning, to play down, so to speak, the linguistic institution of publicly sharable "timeless" (Grice) meanings by reducing meanings to the intentions of purposive-rational actions of single actors. These actors, it is supposed, try to realizejust by the instrument of languagesome perlocutionary effect, that is, some response in or by the heater that lies beyond the illocutionary effect of just understanding the publicly sharable meaning of a speech-act (say, some overt reaction or only some mental dispositiona belief or an emotionof the heater). The conventions of interaction and communicationand in this context even the meaning-conventions and hence finally the intetsubjectively valid "timeless" meaningsare to come about by the strategical game, that is, by the rules of strategical reciprocity, of the intertwined purposive-rational actions of the single actors. I think that this strategy of an intentionalist meaning-theory can be refuted as being .incompatible even with the genuine meaning-intentiqns of human speakers. In order to show this, one need not deny that in almost every human communication also covertly strategical actionssuch as suggestions of certain conclusions and emotional attitudes by rhetorical or other means and especially

162

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

such a t h i n g as " i m a g e - c u l t i v a t i o n " a r e in the play. But, as Strawson has already s h o w n , the intentions of those directly perlocutionary intentions may be sharply distinguished ftom the nonsttategical meaning-intentions that undetlie illocutionary acts. These lattet intentions may be called "verstdndigungsor~ ientiert" (approximately: "oriented towatd undetstandable and acceptable validity-claims") i n conttadistinction to those that ate simply "erfolgsorientiert" (approximately: "oriented towatd achieving a success fot one's petlocutionary intentions"), i n Habermas's t e t m i n o l o g y . A n d it is important to notice, i n this context, that those speech acts that ate erfolgsorientiert and hence use language only as an instrumentas, for example, acts of suggesting a cettain "image" of the speaketmust nonetheless appeal to be veTstandigungsorientiert at the surface, i n otdet to achieve theit covettly sttategical purposes. In short, direct petlocutionary acts must be parasitic upon the possibility and pretended actuality of illocutionary acts. Howevet, i n otdet not to misundetstand this point, I think two comments are needed.
3 9 40

First, it has to be gtanted that also out illocutionary acts ate usually erfolgsorientiert, i n a sense, and thus are means fot teaching a petlocutionary effect i n the heatet. Thus, fot example, informative statements wish to bting about some belief in the heatet; orders wish to be obeyed and hence to bting about some teaction, and even atguments ate to convince the opponent and hence to bting about a c o n v i c t i o n i n his m i n d . But these petlocutionary intentions are not those of direct perlocutionary (i.e., covertly sttategical) acts i n the sense of Sttawson. Fot they make theit intended petlocutionary effects depend upon being first u n d e t s t o o d o n the level of a publicly shatable m e a n i n g - c l a i m a n d then, on the basis of this understanding, being judged and eventually accepted by the hearer w i t h regard to their further validity-claims. T h u s an informative statement is to be accepted as ttue before it is allowed to produce a belief; an otdet is to be accepted as legitimate befote being obeyed; a condolence is to be accepted as sincere prior to its btinging about some feeling of consolation; and an atgument is to be accepted as valid ot sound before a c o n v i c t i o n is acquired. (This obviously was Socrates' and Plato's point against the persuasive rhetorics.) I am here following Habermas's claim concerning the fout univetsal "validityclaims" ("meaning-claim," "truth-claim," "sincerity-claim," normative "tightness-claim") that are bound up w i t h illocutionary acts and make up the " b i n d i n g force" of human speech by w h i c h a "communicative c o o t d i n a t i o n " of h u m a n a c t i o n s i n conttadistinction to a sttategical c o o t d i n a t i o n m a y be teached. But I do not considet it useful to c h a n g e a l o n g w i t h H a b e r m a s t h e usual terminology and have the terms "illocutionary act" and "illocutionary effect" covet not only the "uptake" (Austin) or understanding of the speech-act but also its being accepted on the ground of its " b i n d i n g f o r c e . " W e simply cannot say "I heteby convince y o u , " therefore the act of c o n v i n c i n g cannot be an illocutionary act, although it surely cannot be a direct petlocutionary act (i.e., a covertly sttategical act) i n Sttawson's sense either. W e should rathet allow for a terminological distinction between three things:
41

'PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

163

1. Perlocutionary effects that are brought about accidently 2. Perlocutionary effects that are brought about by a covertly strategical use of language 3. Perlocutionary effects that are brought about by the binding force (of the validity-claims) of our speech. A second comment may be needed with regard to the relationship between Strawson's distinction of acts or, respectively, intentions and the G r i c e a n type of "intentionalist semantics." A s a matter of fact, S t r a w s o n even tried to improve on Grice's approach, i n order to make it compatible w i t h his distinct i o n . A n d this tendency has been continued by offering further improvements of the so-called G r i c e a n mechanism. T h i s means that the speaker-hearer reciprocity of intentionality and knowledge was to be reflected i n advance by the self-referential meaning-intention of the speaker to such an extent that a deception of the hearer by a covertly strategical perlocutionary act would be excluded. But I t h i n k it can be shown that this is impossible, i n principle, as long as there is an asymmetry between the perlocutionary intention of the speaker and the possible knowledge of the hearer about the speaker's i n t e n t i o n . T h i s asymmetry, however, is a necessary implication of the G r i c e a n approach. It can only be prevented if (and when) the speaker gives up using language only as an instrument, i n order to reach some perlocutionary effect be it even the effect of becoming totally transparent in his intentions for the hearer.
42 43

In order to realize a genuine meaning-intention, that is, to express a meaning that is symmetrically intelligible for the speaker and the hearer, the speaker has to restrict his intention to that of an illocutionary act whose effect is publicly sharable meaning. A n d i n order to achieve this, he has to make his o w n and the hearer's understanding of his meaning-intention depend on the conventional meanings of language. For only conventional meanings of language provide the conditions of the possibility of intersubjectively valid meanings, as we have already pointed out. Since this situation holds for us humans t o d a y a situation that is a transcendental c o n d i t i o n of our argumentsit seems to me to be completely irrelevant, in this context, to take recourse to the psychological, historical, or even evolutionist priority of prelinguistic meaning-intentions w h i c h a s may perhaps be c l a i m e d a r e not yet underlying the distinction between illocutionary and direct prelocutionary (i.e., covertly strategical) acts. T h o u g h language may have come about genetically, for us it is the c o n d i t i o n of the possibility of sharable meanings. Let us now consider the psychologistic turn of meaning-theory as it is represented by Searle's recent work o n Intentionality.
REMARKS O N SEARLE'S "INTENTIONALIST SEMANTICS"

T h e strategy of argument i n Searle's latest books seems to be rather different from that of the Griceans; although it appears that for Searle also the difference between genuine meaning-intentions to be expressed by illocutionary acts and

164

PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

the intentions of direct petlocutionary acts has now lost its significance. This must be the case, it seems to me, since Seatle now has abandoned altogethet his former account of meaning-intentions in terms of communicative intentions. This means, as I undetstand it, that he no longet holds his concept of meaning accotding to the performative-ptopositional double structute of illocutionary acts or, respectively, explicir sentences that express those acts. Instead, Seatle now turns back to a precommunicative and ptelinguistic patadigm of meaning in the sense of representation. He summarized his present position as follows:
44 45 46 47

Communicating is a matter of producing certain effects on one's heaters [here obviously illocutionary a n d perlocutionary effects are taken together!], but one can intend to represent something without eating at all about the effects on one's heaters. One can make a statement without intending to produce conviction or belief in one's hearers or without intending to get them to believe that the speaket believes what he says or indeed without even intending to get them to understand it at a l l . [My emphasis.]
48

In the last sentence Searle's abandonment of his former ("illocutionary") concept of meaning is clearly formulated; and this means that his abandonment of what I have called in this chaptet the pragmatic completion of the linguistic turn by the theory of speech-acts. The deeper reason for this turn is that he now thinks that "intentional states of the mind," as "beliefs," "desites," and "intentions" in the narrower sense of "intentionality," are more fundamental with regard to meaning than the linguistic meaning-conventions that rule the possible expressions of the meaningintentions. What arguments could be proposed in favor of these assumptions? Among the reasons for Searle's attempt at basing the philosophy of language on a philosophy of the m i n d , there are especially the following, as fat as I can see:
49

One teason is the quasi-Husserlian point that linguistic expressions as physical entities would not carry meaning at all if they were not animated, so to speak, by our prelinguistic intentions and thus made the exptessions of meaning. Another important atgument says that the "conditions of satisfaction" that are connected with illocutionary actsas, say, with statements, otders, or promisesare primarily "conditions of satisfaction" with tegard to the "intentional states" of the mind.
50

Thus, fot example, my statement will be ttue if and only if the exptessed belief is correcr, my order will be obeyed if and only if the expressed wish or desire is fulfilled, and my promise will be kept if and only if my expressed intention is carried out.
51

Now, in trying to deal with these atguments, I will of course not deny that intentionality is a necessary precondition or ingredient of speech, that is, of the use of language; or, in other words, that a satisfactory (adequate) philosophy of language must at the same time be an adequate philosophy of mind. For this contention belongs to the very point of my thesis that the abstractive fallacy of

PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

165

the semanticist stage of the linguistic turn has to be overcome by the pragmatic completion of the linguistic turn. I think that a particular merit of the speech-act theory lies precisely in the fact that the self-referentiality of the m i n d , w h i c h was made the foundation of critical philosophy in modern times but was tabooed by logical semanticism since G . Frege and B. Russell, has been rediscovered and rehabilitated in a sense by Austin's detection of the performatives. T h i s is even further confirmed by Searle's analysis of the "self-referentiality of indexical propositions," where he shows that and how the utterance of an indexical expression can have a "completing Fregean sense" because it expresses an " i n t e n t i o n a l content" that indicates relations that the object he is referring to has to the utterance itself. This pragmatic completion of "sense" was overlooked by Frege and his semanticist followers, I would t h i n k , because they had a purely symbolicalconceptionalist idea of "sense," as Peirce would say.
52

W h a t then should I have to say in defending the linguistic turn against the psychologistic turn of the latest Searle? In general I would argue that Searle is now sinning against the spirit of his own "principle of expressibility," because he only considers the dependence of illocutionary acts, and hence sentences, o n intentional states of the m i n d (and the intentions to express them), but does not consider the inverse dependence of the meaning of intentional states o n that of their possible expressions by illocutionary acts, and hence o n explicit linguistic sentences. Let me try to show this i n more detail with regard to his two atguments quoted before: In the first quasi-Husserlian argument, Searle is overlooking or d i s r e g a r d i n g along w i t h the whole prelinguistic-turn p h i l o s o p h y t h a t one may not conceive of physical entities as linguistic entities without presupposing already a structurefor example, of relevant oppositions of soundsthat corresponds to a linguistic meaning-structure. T h i s reflects the complementary prejudice that the intentional states of the m i n d could have the status of articulated meaningintentions without already presupposing a structure of meaning-differences that is conventionally and thus publicly fixed (restricted) to the linguistically relevant structures of physical entities, as, for instance, sounds as phonemes. So it looks as if a solitary person, or m i n d , could animate the purely physical entities of language w i t h publicly valid meaning-intentions. Searle illustrates this conception of the relationship between prelinguistic meaning-intentions and l i n g u i s t i c o r other possibleexpressions of them by the following example: It sometimes happens to me i n a foreign country . . . that I attempt to communicate w i t h people who share no c o m m o n language w i t h me . . . i n such a situation I mean something by my gestures, whereas i n another situation, by making the very same gestures, I might n o t , m e a n a n y t h i n g .
53

Searle seems to completely overlook in this story that those of his gestures which are animated by meaning-intentions are by no means a functional equivalent of

166

"PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

linguistic entities, since their meanings are parasitical upon those linguistic meanings which Searle and his communication-pattnets presupposed, even when they did not share a common language with each other but only with their compatriots, in forming theif meaning-intentions. Instead Searle offers the following one-sided account of the relationship between intentionality and linguistic meaning: Since linguistic meaning is a form of derived Intentionality, its possibilities and limitations are set by the possibilities and limitations of Intentionality. The main function which language detives ftom Intentionality is, obviously, its capacity to represent. Entities which are not inttinsically Intentional can be made Intentional by, so to speak, intentionally decreeing them to be so . . . new language games ate expressions of pre-existent forms of Intentionality.
54

This time-honored commonsense idea of the relation of mind and language has been refuted most thoroughly, 1 think, by F. de Saussute. Fot he has shown at least this much: That language is the realm of m u t u a l differentiation of signifiants and signifies, that is, of physically based sign-vehicles and mentally based meanings, such that there are no prelinguistic entities that could already claim the status of intersubjectively valid meanings. De Saussure made this point clear in passages such as the following:
55

Thought taken for itself is like a cloud of mist within which nothing is delimited. There are no pre-fixed ideas (representations), and nothing is determined before the appearance of language. Now, compared to this nebulous region, the sounds by themselves wouldn't present firmly outlined objects either. The sound mass is a little something definitely delimited and clearly determined; it is not a hollow form into which the thought fits itself but rather a plastic material that in its turn is divided in separate parts, in order to deliver designations that are needed by the thought. . . . Hence neither a materialization of thoughts nor a spiritualization of sounds takes place but there is the somewhat mysterious fact that the "sound-thought" brings about divisions and language works out units by shaping itself between two shapeless masses. (My translation from the German.)
56

Moreover, even Searle himself, in his Speech A c t s gave an important hint to the indispensability of the conventional rules of language for the constitution of publicly valid meanings. What I mean is his distinction between "constitutive rules," which are the conditions of the possibility of all "institutional facts" of human culture, including special languages, and "regulative rules," which, as, for example, rules of technical skill, may refer immediately to natural activities like eating or running. 1 think, if the conventional rules of language could be dispensed with in constituting ("institutional facts" of) meanings, then we could dispense with constitutive rules in favor of resorting to the regulative rules for purposive-rational actions. This is precisely the strategy of those representatives

PRAGMATIC T U R N " & T R A N S C E N D E N T A L SEMIOTICS

167

of an "intentionalist" meaning-theory who try to explain the public validity of "timeless meanings" of language by tecoutse to prelinguistic purposes of human actots within the frame of sttategical intetaction. But, as 1 alteady pointed out, these followers of the eatly Grice and of the economic theory of sttategical games cannot explain the difference between illocutionary acts that constitute public meanings and those direct petlocutionary acts that ate patasitic on the public meanings conveyed by illocutionary acts. The question of the constitution of the public validity of meanings btings me to answeting the second atgument of the latest Seatle: A t first sight, it seems to be a very strong argument to say that the condition of satisfaction of illocutionary acts, as statements, otders, or promises, are primarily those of intentional states of the mind, as beliefs, wishes, ot purposive intentions. But, on a closet look, one sees that also this atgument is one-sided or, respectively, ambiguous. The ambiguity of the argument, as far as 1 see, rests on the concept of "conditions of satisfaction." This concept suggests a parallelism and a one-sided dependence telation between illocutionary acts and intentional states, whereas, 1 think, there is a parallelism and a difference and a mutual dependence between them. To suggest the point provisionally with regard to one of Seatle's examples, 1 would say that it is of course true that "my statement will be true if and only if the expressed belief is c o r r e c t [my emphasis!]," but I would like to supplement this truism with the following claim: If the correctness of my belief is to mean as much as intersubjective validity which is to be examined and reconfirmed publicly, then the following proposition also holds: My belief may be correct if and only if my corresponding statement may be proven to be true, that is, intersubjectively valid. Fot, otherwise, proving the correctness of my belief may mean no more than providing representational evidence myself of the fact that a certain uninterpreted and possibly uninterpretable phenomenon that 1 suppose to exist is in fact given. This should in a sense provide a condition of satisfaction for my belief but, nevertheless, it is not yet (the same as) satisfying the truth-claim of a statement (cf. above, p. 144, concerning the difference between a Husserlian and a ttanscendental-semiotic theory of truth). In order to elucidate this somewhat complicated contention, I will introduce at this point some tenets of J. Habermas's "universal" and "formal pragmatics." Habermas, I would say in this context, received and elaborated precisely those points of Searle's early speech-act theory that have been disregarded or dropped, as it appears to me, by the later Searle. Thus from the "ptinciple of expressibility," Habermas detived the conclusion that speech-acts as well as explicit sentences of human language have a partly performative and partly propositional "double structure." From this he further concluded that, by virtue of this double structure, sentences can make explicit the universal validity-claims, that is, claims of public or intersubjective validity, that constitute the "binding force" of, and hence are necessarily connected with, communicative acts of speech. Those validity-claims are:
57

168

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

1. T h e t r u t h - c l a i m , w h i c h refers to the propositional part of a statement (or possibly to its existential presuppositions) 2. T h e veracityor sincerityclaim, w h i c h refers to the intentional state of the m i n d that is expressed by a speech-act 3. T h e n o r m a t i v e Tightness-claim, w h i c h refers to a communicative act as part of a social action 4. T h e m e a n i n g - c l a i m , w h i c h refers to the intended effect of an illocutionary act (in contradistinction, e.g., to the strategically intended purpose of direct perlocutionary acts), and w h i c h , 1 would emphasize, is the precondition of the other three validity-claims. 1 am emphasizing this last point because it is only the sharing of public meaning-claimed by illocutionary acts that makes it possible that the other three validity-claims of speech becomes necessary. A t first sight, one could t h i n k that this conception of satisfying or redeeming validity-claims should aim at something very similar to the later Searle's concept i o n of satisfying the "conditions of satisfaction" of illocutionary acts and, more fundamentally, of " i n t e n t i o n a l states of the m i n d . " However, precisely at this point, an important difference becomes visible: a difference, 1 suggest, between a position that follows the strategy of a pragmatic completion of the linguistic turn and a position that falls back behind the linguistic turn of philosophy i n favor of a version of psychologism and methodological solipsism. Let us consider this more closely: A s 1 already intimated, redeeming the truth-claim of a statement may be something quite different from proving the correctness of a belief if the latter is not made dependent o n the redeemability of the truth-claim of a statement. T h e reason for the difference is the following: A correctness-claim of a belief that could be satisfied independently from understanding the truth-claim of a statement and thus be considered "more fundamental" than the latter, can only, indeed, rest o n the representational evidence of a perception, say, the percept i o n that there is a state of affairs like this (see figure 5.2). N o w , o n Peirce's semiotic account, w h i c h 1 would follow, such evidence is that of an iconical representation of phaneroscopic firstness. A n d this is not even a sufficient c o n d i t i o n for grounding the truth-claim of a perceptual judgment. It is indeed only a necessary c o n d i t i o n for an "abductive inference" that must include a language-mediated the form: T h i s there looks so and so; what looks so and so should be a case of p; hence this there should be a case of p. T h e result of such an interpretation may then be formulated by the statement " T h e cat is o n the m a t . " interpretation. The abductive inference would have acts may be satisfied or redeemed, if it

'PRAGMATIC T U R N " & TRANSCENDENTAL SEMIOTICS

169

But this result of the interpretation is by no means a matter of course, philosophically, as Searle himself makes clear in his chapter on the "background" of understanding meaning which is at least also determined by our linguistic competence. There may indeed be other interpretations of our picturefor example, those that are based on a different apprehension (conception) of the animal or of the base it rests on, or even of the situation that is referred to by the " o n . " (This situation, for instance, might be interpreted with the aid of a Newtonian or Einsteinian gravitation theory.) In any case, it is only the result of the interpretation, fotmulated by a statement, that can be made the object of a truth-claim and of its satisfaction. O f course, in real life a belief will usually be already the result of a linguistic interpretation, but to this extent it is dependent on the conditions of satisfaction of the truth-claim of a statement, and not the other way around, q u o d erat d e m o n s t r a n d u m .
58

The difference and the mutual dependence between the conditions of satisfaction of intentional states and those of public validity-claims of speech-acts becomes even more clear in the case of orders and the underlying wishes and desires. For in this case, the specific validity-claim of an order and its difference from the meaning of a private wish or desire does not even become visible on Searle's account. In fact, it cannot be satisfied or redeemed just by satisfying the underlying wish ot desire, as Searle suggests; for, as a normative rightness-claim, the validity-claim of an order has to be acknowledged and, if necessary, legitimated, whereas the wish or desire as an intentional state of the mind may only be fulfilledpotentially, perhaps, as consequence of acknowledging the validity-claim of an order. Thus it becomes clear that the public meaning of a speech-act, in contradistinction from the intentional state A f the mind that is aid to determine the satisfaction conditions of a speech-act, must imply a reason

1 70

"PRAGMATIC TURN" & TRANSCENDENTAL

SEMIOTICS

for the hearer's understanding and accepting the message. In the case of a statement the immediate reason is provided by its truth-claim, in the case of an older by its rightness-claim but in the case of a simple imperative, the reason for obeying, which is suggested by the speech-act, need not be the legitimacy of a validity-claim. It may rather be provided in this case by the power that is behind the imperative or, respectively, by the sanctions to be expected in case of not obeying.
59

Thus far it might even appear, as if there was only a difference, and no mutual dependence at all, between the satisfaction conditions of wishes and otdets, and hence of "intentional states," on the one hand, and those of speech-acts, on the other. But, I suggest, one should rather say-with Habermas that recognizing the conditions of the satisfaction of intentional states in Searle's sense is just one component of the conditions of the acceptability of speech-acts, since the latter must convey a binding force that is based on reasons beyond the satisfaction-conditions of intentional states.
60

Thus far one has to grant that the illocutionary act of an order must, indeed, presuppose some sort of a wish or desire, which has to be fulfilled, even if, for example, a military order may sometimes be not the exptession of the private wish or desire of the speaker. O n the other hand, however, one must also realize that a wish or desire, in order to be fulfilled, in many cases has to be expressed by a speech-act that has some "binding force," say, by an order, a demand, or a question. Beyond this, 1 would claim that even the meaning of the most intimate wishes or desires, which may be fulfilled by good fortune, must be interpreted in light of language in order to be the meaning of some specific intentional state. This brings us back to the public meaning-claim of illocutionary acts, which is indeed the most fundamental thing from the point of view of the linguistic turn in philosophy. For it is simply not possible to go behind this claim if we want to have a common basis for arguing. A n d arguing is the intranscendable basis for a philosophy after the linguistic turn if this turn is conceived as a transcendentalsemiotic turn.
61

If my preceding critique of Searle's main arguments for this intentionalistic turn is justified, then it obviously cannot be true either that one can make a statement without intending to get some hearer to understand i t . For, if one makes a statement, one must commit oneself to those universal validity-claims that are connected with speech, that is, with its public meaning-claims. The case of a lonesome judgmentsay a perceptual judgment on some given state of affairsmay easily be explained as a case of the "voiceless dialogue of the soul with itself" (Plato). A n d it seems clear thatafter the linguistic turnthis "voiceless dialogue" can only be understood as an internalization of the normal public dialogue by speech. Otherwise we could not understand that even a lonesome perceptual judgment participates in a public world-interpretation. To summarize my critique of the intentionalist move of the later Searle, I am inclined to say that by resorting to prelinguistic and precommunicative represen62

'PRAGMATIC T U R N " & T R A N S C E N D E N T A L SEMIOTICS

171

tational intentional states, Seatle seems to fall back to what I have called the "methodological solipsism" of the philosophy of mind ot consciousness from Descartes to Husserl. It is intetesting to compare this result of Seatle's intentionalism with the result of the other vetsion of intentionalism which starts out from P. Grice and, via Schiffer, Lewis, and J. Bennett, leads to a type of communication-theory that integtates a theory of conventions by a strategical game theory of perlocutionary intetaction. In a sense, this movement is a contraposition to Seatle's mentalism, since it does not reduce meaning to a precommunicative state of mind but rather to a transcommunicative purpose of insttumental actions.
63

With tegard to the linguistic turn of philosophy, it eventually has the same effect. For in both cases the ptagmatic turn toward the meaning-intentions of the human use of signs or language has been overturned in such a way that it becomes incompatible with the philosophical achievements of the linguistic turn. This means that in both cases the function of the transcendental institution of language, that is, the function of constituting intetsubjectively valid meanings by illocutionary acts, has been overlooked. (To say this with regard to Seatle is a bit odd, since without his early work on "speech-acts" I would not have reached my present view of the pragmatic completion of the linguistic turn.) The most characteristic symptom of a complete loss of the transcendental point of the linguistic turn is the tendency in the latet Searle as in other representatives of "philosophical psychology" to go even beyond "intentional states of the mind" and ground the theory of meaning on a theory of the "brain." There is of course nothing to say against a theory of the telations between language, the mind, and the btain. Therefore I do not wish to atgue against Seatle's special account of this topic. But the question is how to understand the talk about something being "mote basic (or fundamental) than something else." There is no doubt that language and meaning "comes very late" in the biological-evolutionist "order of priority." A n d there is no doubt eithet that the development of the brain is a biological condition of the possibility of the mind and to that extent of language. But the point of the linguistic turn was a transcendental one, as I tried to show from the beginning by recourse to Wittgenstein (see above, p. 133). That is to say, it was a new answetdifferent from that of Kantto the question of the conditions of the possibility of the intetsubjective validity of what we can mean. Now it seems clear to me that from this point of view a theory of the brain cannot compete setiously with the transcendental claim of modern philosophy of language (just as before it could not compete seriously with the transcendental claim of a philosophy of the mind ot consciousness). For a theory of the btain cannot, of course, answer the question concerning the conditions of the possibility for validity-claims in talking about the brain.
64 65

My contention in this paper was to show that the transcendental function of First Philosophy cannot be fulfilled by a purely semanticist version of the linguistic turn either. What I wanted to suggest is that it can be fulfilled by a
66

172

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

transcendental-pragmatic completion of the linguistic turn w i t h i n the framework of a transcendental semiotics.

NOTES
1. Eric Stenius, Wittgenstein's "Tractatus" (Oxford: Basil Blackwell, 1960; Ithaca, N . Y . : Cornell University Press, 1964). 2. Ludwig Wittgenstein, Vermischte Bemerkungen (Frankfurt a . M . : Suhrkamp, 1977), 27. 3. Cf. Joachim Leilich, Die Autonomic der Sprache (Munich: Profil-Verlag, 1983). 4. Cf. especially Rudolf Carnap, "Empiricism, Semantics, and Ontology," in Semantics and the Philosophy of Language, ed. L. Linsky (Urbana: University of Illinois Press, 1952). 5. Cf. ibid. 6. Cf. Karl-Otto Apel, " C . S. Peirce and Post-Tarskian T r u t h , " i n this volume (also published as " C . S. Peirce and the Post-Tarskian Problem of an Adequate Explication of the Meaning of Truth," in The Relevance of Charles Peirce, ed. E. Freeman [La Salle, 111.: The Hegeler Institute, 1983]; all page references are to this present volume). 7. Charles W . Morris, Foundations of the Theory of Signs, International Encyclopedia of Unified Science (Chicago: University of Chicago Press, 1938). 8. Cf. Rudolf Camap, " O n Some Concepts of Pragmatics," Phibsophical Studies 6:8591; Richard Martin, Toward a Systematic Pragmatics (Amsterdam: North Holland, 1959). 9. I am here intentionally preferring the epistemologically and ontologically relevant explication of the triadic structure of semiosis to the immanent semiotic one which is based on the relation of the sign to its "immediate object" and its "interpretant," leaving aside the relations of the sign-function to the real object and the real subject of sign-mediated cognition. Cf. Karl-Otto Apel, "Transcendental Semiotics as First Philosophy," in this volume (originally published as "Transcendental Semiotics and the Paradigms of First Philosophy," Philosophic Exchange 2, no. 4 [1978]: 3-22); and Apel, " C . S. Peirce and Post-Tarskian Truth." 10. Morns, Foundations of the Theory of Signs, 81. 11. Morris, Foundations of the Theory of Signs; and especially Charles W . Morris, Signs, Language, and Behavior (New York, Braziller, 1946). Morris's chief semiotical work Signs, Language, and Behavior entrusted the function of a pragmatic integration of semiotics to a behavioristic approach which cannot cope with the communicative and self-reflective understanding of intentions. Cf. Karl-Otto Apel, "Charles W . Morris und das Programm einer pragmatische integrierten Semiotik." Introduction to Charles W . Morris, Zeichen, Sprache und Verhalten (Diisseldorf: Schwann, 1973), 9-66. 12. John R. Searle, Intentionality (Cambridge: Cambridge University Press, 1983), chap. 9. 13. Cf. John J. Fitzgerald, Peirce's Theory of Signs as Foundations for Pragmatism (The Hague, Paris: Mouton, 1966), chapter 3; Karl-Otto Apel, Towards a Transformation of Philosophy (London: Routledge and Kegan Paul, 1980), chapter 3; idem, Charles S. Peirce: From Pragmatism to Pragmaticism (Amherst: University of Massachusetts Press, 1981); and idem, " C . S. Peirce and Post-Tarskian Truth." 14. Saul Kripke, Naming and Necessity (Oxford: Basil Blackwell, 1980). 15. Hilary Putnam, Mind, Language, and Reality, Philosophical Papers, vol. 2 (Cambridge: Cambridge University Press, 1975), chap. 12.

PRAGMATIC T U R N " & T R A N S C E N D E N T A L SEMIOTICS

173

16. See Apel, Towards a Transformation of Philosophy, chapters 4 and 5; and idem, " C . S. Peirce and Post-Tarskian Truth." 17. Putnam, Mind, Language, and Reality, 223ff. 18. Ibid., chap. 12. 19. Apel, Charles S. Peirce, 137ff. 20. Putnam, Mind, Language, and Reality, 227ff. Cf. also Leilich, Die Autonomie der Sprache, Iff., 162ff. 21. Searle, Intentionality, especially 205f. 22. Ibid., 207. 23. Ibid., 208. 24- Cf. Karl-Otto Apel, "Charles W . Morris und das Programm einer pragmatisch integrierten Semiotik." Einfiihrung zu: C h . W . Morris, Zeichen, Sprache und Verhalten (Diisseldorf: Schwann, 1973), 9-66; idem, Towards a Transformation of Philosophy, chap. 4; and idem, Charles S. Peirce. 25. See Apel, Towards a Transformation of Philosophy, chapter 3; idem, Charles S. Peirce; and idem, " C . S. Peirce and Post-Tarskian Truth." 26. Cf. Apel, Charles S. Peirce, 166. 27. See Apel, " C . S. Peirce and Post-Tarskian Truth." 28. Cf. Apel, Towards a Transformation of Philosophy, chap. 3; and idem, Charles S. Peirce, 25ff. 29. Karl Buhler, Sprachtheorie (Jena: G . Fischer, 1934), sec. 2. 30. Cf. Karl Popper, Logic of Scientific Discovery (London: Hutchinson, 1972); idem, Objective Knowledge: A n Evolutionary Approach (Oxford: Clarendon Press, 1972); and idem, Conjecture and Refutations: The Growth of Scientific Knowledge (London: Routledge & K e g a n Paul, 1972); cf. also Karl-Otto Apel, "Zwei paradigmatische Anrworten auf die Ftage nach der Logos-Auszeichnung der menschlichen Sprache," in Kidturwissenscha/ten, ed. H . Liitzeler (Bonn: Bouvier, 1980), 13-68. 31. Cf. J. L. Austin, How to Do Things with Words (Oxford: Oxford University Press, 1962). 32. Peter F. Strawson, "Intention and Convention in Speech Acts," The Philosophical Review 73 (1964): 439ff.; cf. Paul Grice, "Meaning," The Philosophical Review 66 (1957): 377-88; and idem, "Utterer's Meaning and Intentions," The Philosophical Review 78 (1969): 147-77. 33. John R. Searle, Speech Acts (Cambridge: Cambridge University Press, 1969), 19f. 34. Ibid., 20. 35. Cf. Karl-Otto Apel, "Intentions, Conventions, and Reference to Things: Meaning in Hermeneutics and the Analytic Philosophy of Language," in this volume, 51-82 (originally published as "Intentions, Conventions, and Reference to Things: Dimensions of Understanding Meaning in Hermeneutics and in Analytic Philosophy of Language," in Meaning and Understanding, ed. H . Parret and J. Bouveresse [Berlin: de Gruyter, 1981]; page references are to this volume). 36. Searle, Speech Acts, 17. 37. Cf. Grice, "Meaning" and "Utterer's Meaning and Intentions." 38. S. R. SchifFer, Meaning (Oxford: Oxford University Press, 1972); Jonathan Bennett, Linguistic Behavior (Cambridge: Cambridge University Press, 1976); David Lewis, Convention (Cambridge: Cambridge University Press, 1969). 39. Strawson, "Intention and Convention in Speech A c t s . " 40. Jiirgen Habermas, Theorie des kommunikativen Handelns, 2 vols. (Frankfurt a . M . : Suhrkamp, 1981), vol. 1, chap. 3. 41. Cf. ibid., vol. 1, 391 ff.; and Karl-Otto Apel, "Lark sich erFiische Vernunft von strategischer Zweckrationalitat unterscheiden? Zum Problem der Rationalitat sozialer Kommunikation und Interaktion," Archivio di Filosofia, LI, N . 1-3 (1983): 411fF.

174

"PRAGMATIC TURN" & TRANSCENDENTAL SEMIOTICS

42. Strawson, "Intention and Convention in Speech A c t s . " 43. Cf. Apel, "Intentions, Conventions, and Reference to Things," 63-73; and especially idem, "LaBt sich ethische Vernunft," 385ff. 44. Cf. Searle, Speech Acts, 43ff., to idem, Intentionality, 161. 45. Cf. Searle, Intentionality, 166. 46. This was Habermas's understanding of Searle's meaning-theory in Speech Acts. Cf. Jiirgen Habermas, "Vorbereitende Bemerkungen zu einer Theorie der kommunikativen Kompetenz," tn Theorie der Gesellschaft oder Sozialtechnologie, ed. J. Habermas and N . Luhmann (Frankfurt a . M . : Suhrkamp, 1971); idem, "Was heiBt Universalpragmatik?" in Sprachpragmatik und Philosophie, ed. K . - O . Apel (Frankfurt a . M . : Suhrkamp, 1976), 244ff. (and also i n ] . Habermas, Vorstudien und Ergdnzungen zur Theorie des kommunikativen Handelns [Frankfurt a . M . : Suhrkamp, 1984]; all page references are to the volume edited by Apel); and idem, Theorie des kommunikativen Handelns, vol. 1, chap. 3. Cf. also Apel, "Zwei paradigmatische Antworten." 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. Searle, Intentionality, 164fF. Ibid., 165. Ibid., 160ff. Ibid., 161f., 176ff. Ibid., 11. Ibid., 220ff. Ibid., 162. Ibid., 175. Cf. Apel, "Charles W . Morris und das Programm." Ferdinand de Saussure, Cours de Imguistique generak, trans, into German by H . Lommel (Berlin, Leipzig: de Gruyter, 1931), 132ff.; my translation from the German. Cf. Habermas, "Was heiBt Universalpragmatik?"; and idem, Theorie des Kommunikativen Handelns, vol. 1, chap. 3. Cf. Searle, Intentionality, chap. 5. Cf. Habermas, Theorie des kommunikativen Handelns, vol. 1, 403fF. Ibid. Cf. Karl-Otto Apel, "The Problem of Philosophical Fundamental Grounding in Light of a Transcendental Pragmatic of Language," Man and World, 8, N o . 3 (1975): 239-75. Cf. Searle, Intentionality, 165. Cf. Apel, Towards a Transformation of Philosophy, passim. Cf. Searle, Intentionality, 160ff. Cf. ibid., chap. 10. Cf. Apel, "Transcendental Semiotics as First Philosophy."

62. 63. 6465. 66.

SIX C. S. Peirce and Post-Tarskian Truth

As the title of my paper indicates, I wish to establish a relationship between the problem of an adequate explication of the truth-conception that underlies modern empirical science and the philosophy of C . S. Peiice, who is often called the foundei of American pragmatism. In speaking of the truth-conception of modem empirical science, I am thinking of a conception of truth that is necessarily presupposed fot an adequate epistemological and methodological undetstanding of experimental and theoretical natutal science and, indeed, fot such types of quasi-nomological social science as can be practiced accotding to the paradigm of natutal science. This means that I do not propose directly to thematize the truth-problematics of the so-called "hetmeneutic Geisteswissenschaften" ot "critical-reconstructive social sciences" which, in my opinion, ttanscend the truth-conception as well as the very concept of (natural) science. But why must we inquire into the relationship between the truth-conception of modem empirical science and the philosophy of C . S. Peirce in view of the fact that modem analytic philosophy has invested so much enetgy in the logic of modem empitical science and, in that context, in the semantic explication of ttuth?
1

I
In otdet to account for my interest in Peirce's conception of ttuth and, furthermore (moreover) to introduce my general approach to Peirce's philosophy, I propose to begin by placing my approach in a philosophico-histotical context. The context is best explicated in terms of semiotics, especially in tetms of the three-dimensional semiotics of the sign-relation ot "semiosis," comprehending syntactics, s e m a n t i c s , and p r a g m a t i c s , which via C . W . Morris goes back to Peircean ideas and via Morris and R. Carnap has also provided the semiotical background-conception fot "logical semantics." With regard to this semiotical setting-story I want to give at least some vague elucidations in the present context.
2 3

II

For semiotic reasons I do not believe that the concept of truth needed in the philosophy of empirical science can be adequately explicated merely on the basis
175

176

C S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

of logical semantics (i.e., in abstraction from pragmatics) (as little, by the way, as I think that in this way an adequate explication of the concept of explanation can be provided).
4

This means for example that 1 do not consider Tarski's semantical explication of the meaning of truth for formalized languages as a sufficient basis for a realistic correspondence-theory of truth, as it was considered by Karl Popper and others, but only as an explication of a necessary condition for any philosophical theory of truth that is to account for the possibility of logical implication, that is, of truth-transfer, as it is isolated in formalized semantical systems. Seen in itself, this conception of truth is neutral with regard to different ontological or epistemological positions of philosophy, as Tarski himself emphasized; and this means, I suggest, that it is also neutral with regard to the different classical philosophical conceptions of truth, as, for example, metaphysical correspondencetheory, evidence-theory, coherence-theory, and pragmaticist theory.
5 6

My chief reasons for considering Tarski's semantic conception a formalist minimal conception of truth are semiotic reasons in so far as I reflect upon the fact that, being restricted to formalized languages, this conception must abstract from the pragmatic dimension of actual use of interpretation of language. Thus, it must for example abstract from the situation-bound use of objective indexical expressions like "this" and "there" and of subjective indexical expressions in the context of performatives such as "I hereby state that . . . ." Now, as C . S. Peirce first made clear in his semiotic logic of inquiry, without the situation-bound function of i n d e x i c a l expressions the language of science cannot really grasp (get hold of) reality, and without the later so-called performatives (e.g., verbalized acts of "assertion") the human subject of science cannot reflect upon his or her truth-claim in such a way as to take over responsibility for it. This means the following with regard to the problem of a philosophical explication of the meaning of truth:
7 8

O n the one hand, the equivalence-relation expressed in Tarski's convention T (i.e., "x is true if, and only if, p") cannot pass as an explication of that relation between thought or language and reality that is meant in a realistic correspondence theory of truth. For it only deals with that relation insofar as it tacitly presupposes that the meaning of p can be interpretedhowever indirectly with the aid of a natural language which, as a pragmatically ultimate metalanguage of any hierarchy of formalized languages, can, with the aid of indexical expressions, provide a real denotatum for p as a mere designatum of an abstract semantical system. (That the possibility of such an interpretation is not unproblematic may be illustrated with regard to so-called theoretical concepts of deep theories.) O n the other hand, the abstractive deficiency of the logico-semantical explication of truth is even more radical with regard to its failure to reflect on the truth-claims of the human subjects of knowledge as they are expressed and reflected upon in performative phrases like "I hereby state that . . . . " For it is only by reflection on those subjective truth-claims that we can understand the

C. S

PEIRCE A N D POST-TARSKIAN TRUTH

177

meaning of the strange predicate "is true," whose bearer is sentences or rather propositions as stated through statements or assertions. For the fact, reflected upon by the so-called redundance theory of truth, that the predicate "is true" is implied i n the very statement of a proposition simply means that truth is not a strange property of some entities in the world we c o u l d get special informations about but rather a c l a i m that human subjects of knowledge connect w i t h propositions by asserting them and that they can make explicit through performative phrases like "I hereby state that . . . . " T h u s "is true" is, i n fact, redundant as long as our truth-claim is simply implied i n our communicative statements, but it is no longer redundant when our implicit truth-claim is called into question and hence has to be made explicit o n the level of argumentative discourse.
9

Against these semiotic arguments for the need of reflection o n the pragmatic (i.e., objective denotative as well as the subjective performative) presuppositions of the semantic explication of truth as predicate of sentences (of formalized languages), there have been directed the following standard semanticist arguments: W i t h regard to the question of securing real denotata (e.g., identifying real objects to w h i c h the ptedicates of propositions may apply), it has been said that it belongs to the problem of verification and thus has to be solved as a problem of empirical pragmatics w h i c h , as a meaningful question, presupposes already the logico-semantical solution of the problem of an explication or definition of the meaning of truth. Similarly w i t h regard to the question of reflecting o n out subjective truthclaims, it has been said that it too constitutes a problem of empirical pragmatics (say, of psycholinguistics)a problem moreover that should be thematized as a topic of the semantical reference of a metalanguage (say of a behaviotistic type of psycholinguistics), i n order to avoid the semantical antinomies that are implied in the self-referential language of inttospection. In order to show that this problematic has nothing to do w i t h the question of an adequate explication of the meaning of truth, it has been argued that propositions may be ttue or false quite independently of their being asserted as well as of theit being verified. T h e plausibility of these semanticist arguments, as far as 1 can see, rests o n a semiotic a x i o m (or prejudice) that was introduced by M o r r i s and Carnap under the impact of Russell's semantic "theory of types" and Tarski's verdict against self-referential use of language. T h e axiom says that the pragmatic dimension of the triadic sign-relation ("semiosis") that is, the dimension of the actual use of interpretation of signs by the human subjects of c o m m u n i c a t i o n c a n only be thematized as a topic of empirical pragmatics (possibly of a behavioristic type), or as a topic of a formal-constructive pragmatics that should provide the theoretical metalanguage for (the semantical thematization of the pragmatical dimension by) empirical pragmatics. From the point of view of this semiotic axiom we cannot conceive of a thematization of the whole actual triadic relation of semiosis (or e.g., sign-mediated cognition) by philosophic reflection o n its actual pragmatic dimension. Hence there cannot, it seems, be a semiotical
10

178

C . S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

equivalent to transcendental epistemology of a Kantian or Husserlian type. (And this preconception has in fact predetermined the development not only of the Carnapian or Hempelian type of a "logic of science" but also, I think, the development of Popper's "logic of scientific discovery" up to the later conception of an "epistemology without a knowing subject.") I think, howevet, that this semiotical axiomlet us call it the axiom of the impossibility of a ttanscendental pragmaticsis mistaken and, especially, is incompatible with Peirce's foundation of semiotics and semiotical logic of inquiry on the basis of the triadic structure of semiosis, which was considered by Peirce as a semiotic equivalent of Kant's "synthesis of apperception" and, beyond that, of Hegel's concept of "mediation" as a structure of the spirit. 1 shall latet come back to this thesis. (See figure 6.1.)
11 12

For the moment let us see in what respect the semanticist arguments in favor of the insufficiency of Tarski's explication of the meaning of truth stand and fall together with (in dependence of) the Morris/Carnap-axiom of semiotics that there cannot be such a thing as a transcendental-reflective pragmatics of the act of sign-interpretation or, for that matter, of sign-mediated cognition. First, it seems plausible that the question of how to settle the problem of verification, or rather confirmation, of scientific hypotheses in one sense might be conceived of as a problem for empirical scientists and in another sense as a problem of an empirical "science of science" (e.g., history of science) and thus far might be subsumed under the topics of empirical pragmaticsalthough not of a behavioristic type but rather of a hermeneutic-reconstructive type which should be capable of understanding and evaluating the good or bad reasons of the empirical scientists in the light of some normative ideas of what verification or confirmation should be. But then the question arises as to what kind of explication of the meaning of truth is presupposed by such a type of "empirical" pragmatics: Should it be that of logical semantics that cannot even cover the general ideas of verification (or confirmation of falsification) and hence is criteriologically irrelevant? Or should it be a type of explication that would be criteriologically relevant by involving some general idea of verification in the broadest sense, that is, some idea of possible criteria of truth and possibly even of a highest or ultimate regulative principle, so that in its light the very explication of the meaning of truth could be summed up in terms of a possible integration of possible truth-criteria? It seems clear to me that under the presupposition of the Morris/Carnapaxiom of semiotics, it must be postulated that the criteriologically irrelevant type of semantical explication of the meaning of truth is not only a necessary but a sufficient presupposition for an empirical pragmatic thematization of the problematic of verification. This postulate is simply enforced by the presupposition that there is no possibility envisaged of conceiving of a reflective philosophical thematization of the whole triadic relation of semiosis or, for that matter, sign-mediated cognition of the real by a human subject. O n the other hand, in terms of the presupposition that there can be a reflective philosophical

C . S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

179

III (I)

real subject -^(interpreter)

sign

I&
E o c
v

real co-subject

r e d u c t i o n of t h e P e i r c e a n scheme of l o g i c a l e m p i r i c i s m

of s e m i o s i s w i t h i n t h e f r a m e of

t r a n s c e n d e n t a l - p r a g m a t i c s u p p l e m e n t a t i o n ( i n t e g r a t i o n ) of the scheme s e m i o s i s w i t h i n the f r a m e of t r a n s c e n d e n t a l s e m i o t i c s

FIGURE 6.1

'i

180

C . S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

knowledge a priori of the triadic structure of actual semiosis and of sign-mediated cognition, it seems also clear that the logico-semantical explication of the meaning of truth must tacitly presuppose its possible supplementation by, and possible integration into, a nonempirical pragmatical explication of the very meaning of truth, such that this criteriologically relevant type of meaningexplication must be presupposed by the empirical pragmatics of verification. T h e standard semanticist argument for a dichotomy of logico-semantical meaningexplication of truth and empirical pragmatics of verification would then amount to an abstractive fallacy, w h i c h is based o n the nonobservance of Peirce's insight into the triadic structure of semiosis. 1 shall indeed i n the following defend this latter thesis and to the extent heuristically start out from the idea of a transcendental semiotics including a transcendental pragmatics.
13

W i t h regard to the question whether an adequate explication of the meaning of truth needs to take into account the truth-claims as they may be made explicit by performative phrases, the difference between an abstractive semanticist and a transcendental-pragmatic view of the situation is even more striking and important than w i t h respect to the problem of verification. T h e difference at stake hete may be understood as one between the usual (semanticist) supposition that truth is a matter of propositions and not of speech-acts and the (transcendentalpragmatic) thesis that only by recourse to our subjective truth-claims and hence to the meaning of assertions can we understand the (nonredundant) meaning of the predicate "is true." For, as the transcendental-pragmatic thesis w i l l claim, we have to understand "is true" as a predicate that w i t h i n the frame of semantics cannot belong to the level of object-language, and hence of the sentences that express the true or false propositions, but must belong to the level of a metalanguage through w h i c h our subjective reflection upon the language of propositions is expressed and objectified, so to speak. N o w this means that without the possibility of an actual self-reflection of our truth-claims, as it is made possible by performatively expressed assertions, we would be unable philosophically to understand the very difference between object-language and metalanguage w h i c h is presupposed by the so-called semantical explication of truth. T h e standard-argument for the semanticist position, namely, the argument that propositions, and thus semantically interpreted sentences, can be true or false quite independently of their being asserted or not asserted, obviously does not really contradict the transcendental-pragmatic thesis; for it only states correctly, of coursethat the truth or falsehood of propositions, and hence of sentences, is independent of their being factually asserted. If it were instead to claim that one could conceive of the truth or falsehood of propositions quite independently of the conception of human truth-claims and assertions, as it was in fact claimed by Thomas A q u i n a s and Bolzano, then transcendental pragmatical reflection upon the conditions of the possibility of the meaningclaim of phrases like "proposition x is true" must maintain that an abstractive fallacy of a semanticist type has been committed. T h i s is indeed a reproach that

C. S . P E I R C E A N D P O S T - T A R S K I A N T R U T H

181

applies to Popper's conception of an "epistemology without a k n o w i n g subject," I suggest. (It does not apply to Thomas A q u i n a s and to Bolzano only i n so fat as both of these theologically inspited thinkets could presuppose the divine intellect as a transcendental subject of ttue knowledge i n theit radical abstract i o n ftom human t r u t h - c l a i m s . )
14 15

But the consequences of the difference between the semanticist and the ttanscendental pragmatic position have not yet been exhausted by what has been pointed out thus far. For if the arguments advanced thus far are sound, and the meaning of ttuth cannot be sufficiently explicated merely i n tetms of a semantical account of propositions but only by recourse to propositions and to performatively expressed assertions as speech-acts through w h i c h our propositions are proposed and thus connected w i t h out truth-claims, then there are consequences to be drawn both w i t h tegatd to the semiotical conception of language that is presupposed by an adequate explication of the meaning of truth and w i t h regard to the subjective-intersubjective dimension of the idea of truth-claims as implied i n the meaning of the idea of truth. In the first respect, we have to take into consideration that it is not the concept of a semantical system of propositional sentences i n the sense of Tarski and C a r n a p that has to be made the basis of framework for an adequate semiotical explication of the concept of ttuth (and hence of epistemology), but rather a concept of natural language that, from the outset, must account for the twofold (i.e., pattly propositional, partly performative) structure of those sentences that can express speech-acts like assettions and thus fat c a n reflectively (i.e., i n a pragmatical sense of self-reference) express human ttuth-claims. Thereby a ctitical spotlight is shed o n Tarski's conception of natutal language as a semantically "closed system" (of propositional sentences), w h i c h thetefore could not be made the basis or framework for an adequate philosophical explication of the meaning of truth. Looked upon from a transcendental-pragmatical point of view, natural language is not a "semantical system" (of propositional sentences) at all, but rathet a (quasi-institutional) system of twofold (propositional-performative) sentence-patterns as possible virtual expressions of human speech-acts, w h i c h for theit part have to integrate the semantical dimension of propositions (abstractively isolated by Tatski and Carnap) into the pragmatical dimension of speech. Due to its twofold sentence-patterns, natural language is indeed pragmatically self-tefetential; but far from thereby necessarily " i m p l y i n g " semantical antinomies w h i c h disqualify natutal language as the medium of philosophy, it thereby first of all makes possible h u m a n cognitive self-reflection and hence even that type of philosophical (metahgical) self-reflection of thought that has to be presupposed by Tarski's postulate of an infinite hierarchy of metalanguages.
16 17

(In fact the pragmatically self-referential structure of performative phrases of natural language provides the cognitive basis for those general philosophical insights that must be formulated w i t h the aid of implicitly self-referential universal propositions. But it cannot thereby have become the cause [i.e.,

182

C. S

PEIRCE A N D POST-TARSKIAN

TRUTH

sufficient conditionl of semantical antinomies; for all philosophical positionsentences, including those of the Sceptics, must take the form of implicitly self-teferential univetsal propositions, and among them also those sentences through w h i c h Russell's semantic "theory of types" and Tarski's theory of the necessary splitting of the language [of scientific philosophy] into object-language and metalanguage can be made the object of argumentative discourse. There are in fact good reasons for the view that Russell's and Tatski's vetdicts against self-referential language must not be considered as philosophical solutions of the problem of semantical antinomies, but rathet as technical devices of avoiding the rise of the very problem by establishing formalized language-systems as "otgans" of logico-mathematical ["apodeictic"] reasoning w h i c h precisely for this reason cannot be substituted fot natural language as a medium of reflective philosophical r a t i o n a l i t y . A s far as a substantial solution, fot example, of the antinomy of the Liat is required, it might rather be reached along the lines of Hegel's verdict against contradicting one's necessary truth-claim. For this vetdict seems to represent the most fundamental principle of a transcendental pragmatics of h u m a n argumentation).
18

In fact, if there must be a theory of truth that may be applied to itself (as, I should t h i n k , it must be postulated fot an adequate philosophical theory of truth), then the only chance of avoiding an infinite regress i n foundation is provided by a ttanscendental-ptagmatic account, w h i c h should explicate the meaning of ttuth i n terms of the necessity and possibility of redeeming one's o w n truth-claims. T h e talk of redeeming truth-claims, however, besides the general problematic verification, points to a further dimension of the meaning of truth w h i c h may be referred to provisionally by the term "intersubjective v a l i d i t y . " A n d here again our transcendental-semiotical approach opens up a problematic of a possible explication of the meaning of truth that, as fat as I know, was seriously taken into account for the fitst time by C . S. Peitce. W h a t I mean may be systematically reconstructed as follows: If the human subject of truth-claims is conceived of not only i n a K a n t i a n way as subject of the prelinguistic "synthesis of apperception," but, moteover, in a transcendental-semiotical way, as subject of sign-interpretation, then his truthclaims cannot, i n principle, be sufficiently redeemed by the evidence of his o w n synthesis of apperception w i t h tegard to data; they have rather to be redeemed by an intersubjectively valid synthesis ot unity of sign-interpretation. But this means that the methodical solipsism of modern epistemology from Descartes to Husserl has to be surmounted in favor not only of an a priori warranted intersubjectivity, as it is presupposed i n Kant's idea of a "transcendental consciousness," but i n favor of the postulate of a n indefinite community of signinterpretation w h i c h could reach discursive agreement about the confirmation or falsification of propositions, for example, of scientific hypotheses and theories, by evidential criteria. A n d this implies that the old idea of the "consensus o m n i u m , " w h i c h from Aristotle and the Stoics to K a n t was c o n -

C . S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

183

sidered rather a commonsense index of a merely "subjective criterion of t r u t h , " has to be given the status of a "regulative p r i n c i p l e " of redeeming truth-claims that as such is "constitutive" for the meaning of truth qua intetsubjective validity.
1 9

A t this point let us attempt a ptovisional resume of the tesults of our introduction of the ttanscendental-semiotic or, respectively, transcendentalpragmatic approach: A f t e t teducing Tatski's semantical conception of t r u t h i n accotdance, 1 suggest, w i t h its o w n n e u t r a l i t y - c l a i m t o the status of an abstractive m i n i m a l conception w h i c h only reclaims a logically indispensable necessary c o n d i t i o n of any philosophical explication of the meaning of truth, we have charted i n tough outlines the semiotic hotizon of a critetiologically relevant explication of the meaning of truth. W e have postulated that it should encompass the geneial idea of possible verification ( i n a broad sense, including confirmation or corroboration and hence falsification) as well as that of intetsubjective validity to be testified by universal consensus. N o w the question may be laised as to the relationship between out semiotically postulated hotizon of a ciiteriologically relevant explication of the meaning of truth and the ttaditional truth-theoties, as, fot example, the correspondence-, evidence-, coherence-, and the piagmatist satisfaction-theory of truth. A n d , i n accotdance w i t h out suggestion, we may even supplement this enumeration by the supposition of a consensus-theory of truth. In this context, it is interesting to observe that i n the last decades, pattly from the vantage point of Tatski's semantical theory, partly from the vantage points that transcend the absttactive semantical, but hatdly the semiotical approach, almost all the ttaditional conceptions have been renewed. (Thus, e.g., the semantical conception has been widened by integrating ptagmatical ptesuppositions by D . D a v i d s o n and i n a m u c h mote tadical sense by E. T u g e n d h a t . Ttanscending the semantical conception i n the ditection of an ontological ot tathet onto-semantical conception by recourse to W i t t g e n stein's T r a c t a t u s , W . Sellats has developed a theory of "similarity ot correspondence or isomorphy between two systems of objects w h i c h belong to the natutal o r d e r . " T h e coherence-theory of ttuth w h i c h has a neo-Hegelian and a neopositivist otigin was recently revived and elabotated by N . R e s c h e t and L . B. P u n t e l . Finally, different types of a consensus-theory of ttuth wete proposed by the Erlangen school of dialogical constructivism [W. K a m l a h and P. L o t e n z e n and K . L o t e n z ] and by J. H a b e t m a s i n the name of a univetsal pragmatics.)
20 21 22 23 2 4 25 26 27

N o w I think that the hotizon of a ttanscendental semiotics as Fitst Philosophy is suited also to settling the question as to the systematical otdet of, and hence interrelationship between the different approaches of ttuth-theoties thus far proposed as possible conttibutions to a full-fledged explication of the meaning of truth. A n d since I conceive of the philosophy of Peirce as a first project of a transcendental semiotics, including a ttanscendental piagmatics, I heuristically assume i n this essay that it can provide important viewpoints for settling the

184

C. S

PEIRCE A N D POST-TARSKIAN TRUTH

question of a systematical integtation of the different conceptions of truth. Thus, I shall i n the next section try first to outline from a quasi-Peircian point of view the critical (sttategical) restrictions that have to be imposed upon the possible configurations of truth-conceptions as ingredients of a ttanscendentalsemiotic explication of the meaning of truth. By this method I hope finally to approach that type of ttuth-theory that comes closest to Peirce's conception and may both count as a critical reconstruction of it and be itself most fruitfully reconstructed in the light of Peirce's contributions to a transcendental-semiotic or, respectively, tianscendental-pragmatic theory of truth.

Ill

1. First, I suggest that an ontological-metaphysical conception of truth that views the correspondence between things and ideas as a relation between entities w i t h i n the world is ruled out as anachronistic by the critical standards of present philosophy. These standards at least since K a n t have been prejudiced by the idea of the uniqueness of the epistemological subject-object r e l a t i o n , which cannot i n principle be reduced to an object-object r e l a t i o n . Hence any criteriologically relevant attempt to think of an ontological correspondence between thought and reality as though we could conceive of such a relation from a standpoint outside of the epistemological subject-object relation is confronted w i t h a standard-argument of the post-Kantian (epistemological) paradigm of First Philosophy. T h e correspondence, the argument runs, can only be verified w i t h respect to the a p p e a r a n c e s of reality by judgments that w i t h respect to their correspondence to the things i n themselves have to be verified o n their part, w h i c h can only be done by judgments about the appearances of reality and so o n ad infinitum. ( A cross-check of the validity of this argument seems to be provided on the one hand by those modern versions of the correspondence-theory that avoid the regressus ad infinitum by simply talking of a correspondence between propositions and facts that are conceived as existing if and only if the corresponding propositions are true. These theories, i n order to avoid the regressus ad infinitum, have to take a circulus vitiosus into the bargain; and, of course, they must be criteriologically irrelevant. A n o t h e r , even more important cross-check of the transcendental-epistemological argument against the metaphysical correspondence-theory is provided, it seems to me, by those biological-evolutionist theories that conceive of correspondence as a relation of real adaption between the cognitive equipments or apparatuses of organisms and their e n v i r o n m e n t s . These empirical-scientific theories are, i n my o p i n i o n , the historically legitimate followers and substitutes for the ontological-metaphysical correspondence-theories of truth; but they testify to their scientific status by loosing direct philosophical relevance, for example, as possible justifications of their o w n truth-claims.)
28

2. N o w the aporia of the infinite regress of any attempts to check the ontological-metaphysical correspondence between thought and reality has often

C . S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

185

been considered as an argument for the thesis that the truth of ptopositions can in fact only be checked by ptopositions, and ftom this it has been concluded that the correspondence-conception of truth has to be abandoned in every respect and to be replaced by a coherence-theory of truth. Aftet O . Neurath and B. Blanshatd, L. B. Puntel has recently taken this turn; and he has combined it with anothet impottant atgument of the coherence-theory since Hegel: It by definition must be ptedestined to account for the systematical connection between all conceivable criteria (in a nondefinitive sense) of ttuth and hence, being the ultimate standard or regulative principle of a possible integtation of truth-criteria, provides the definitive explzcans ot definiens for a philosophical explication of the meaning of truth.
29 30

In dealing with this claim from a transcendental-semiotical point of view, I would first concede that the impottance of the cohetence-conception lies in fact in the citcumstance that it provides a critetion fot the connection of tiuthcriteria which must be used by us in the process of redeeming our ttuth-claims. From this it follows, among othet things, that the coherence-theory of truth cannot be refuted alone by the argument that thete may be different, that is, tival consistent systems that may even then be submitted to the criterion of theit coherence with othet truth-candidates, if they all happen to be coherent in themselves. But the pity is that the criterion or regulative ptinciple of coheience, if it is not supplemented by othet independent critetia of ttuth, applies to too much. It provides no criterion for a distinction between our thoughts about possible worlds and our knowledge about the real wotld. The teason fot this idealistic deficiency of a pure coherence-theory of truth may be exposed, I think, in the light of a transcendental semiotics of Peircean provenience. First, it may be shown that the coherence-theoretician goes too fat when, from the fact that out expetiential data are always language-impregnated ot theory-impregnated data, he concludes that when testing hypotheses we can in the last analysis only compare propositions with propositions. The ctucial semiotic atgument against this thesis is provided by Peirce's distinction between three types of signsicons, indices, and symbolsand his proof that human language, as it is used in "perceptual judgments" and hence in the context of confirmation and innovative discoveties, cannot work alone on the basis of the geneial conceptual meaning of symbols which are determined merely by the linguistic conventions, as Hegel supposed in his chapter on "sense-certainty," (sinnliche G e w i s s h e i t ) in the P h e n o m e n o l o g y of the S p i r i t . Against this paradigmatic pteconception of an idealistic coherence-theory of tiuth, Peirce has shown that the situation-bound use of indexical expressions and even of predicates in perceptual judgments must at least participate in the function of "indices" and "icons" that get theii meanings immediately within the perceptual situation. Thus, for example, the perceptual statement that is expressed by the sentence "Now I see there a church with two steeples," or rather "This there is a church with two steeples," gets its immediate meaning not only from the conventional
31 3 1 33

186

C S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

meanings that its components convey as linguistic signs (including, as Hegel rightly remarked, such general meanings as n o w n e s s , thereness, and tru'sness) but also from the real causal connection between language and the existing reality i n space and time that alone can legitimize my actual use of the indexical expressions "there" and "this." It furthermore gets its immediate meaning also by the perceptual application of the predicate " c h u r c h w i t h two steeples." T h i s crucial element of perceptual verification is decisively legitimized by the iconic evidence of a qualitative phenomenon that is distinguished as a r e a l phenomenon, that is, beyond its merely possible iconic quality, by it being opened up to our attention by the simultaneous legitimized application of the " i n d e x i c a l expressions" that, so to speak, testify to the clash of our attention with the resistance of the existing reality. Thus, the phenomenologically legitimized application of the connection of indexical expressions and predicates whose symbolic meaning is satisfied (or fulfilled) by iconic evidence w i t h i n the perceptual situation seems to make up that criterion of truth as evidence for a correspondence between propositions and reality that has to be combined i n some way w i t h the coherence-criterion i n order to distinguish our well-confirmed knowledge of the real world from mere coherent imaginations concerning a possible world. 3. In this context one has to realize, 1 t h i n k , that during the era of the second (epistemological) paradigm of First Philosophy not only was the coherencetheory of truth proposed as an alternative to the ontological-metaphysical correspondence-theory but also, since Descartes, an evidence-theory of truth. T h i s type of truth-theory reached its most elaborate and argumentatively strongest form in Husserl's phenomenology, where it i n fact amounts to an A u f h e b u n g (i.e., suspension and preservation) of the idea of correspondence by substituting the idea of a reflective evidence of the correspondence between the meaning-intention of our judgments or propositions and the "self-givenness" of the phenomena that fulfill our meaning-intentions. N o w I t h i n k that Peirce's semiotical approach, at least as it was developed i n his "phenomenological" or "phaneroscopic" period (e.g., in the L e c t u r e s o n Pragmatism of 1903), comes close to reconfirming Husserl's approach up to a certain p o i n t .
3 4

Thus it seems clear to me that Peirce's analysis of the function of indexical expressions i n connection w i t h that of the iconical relationship between perceptual qualities (or relations!) and predicates w i t h i n the context of actual perceptual statements converges w i t h Husserl's analysis i n ascertaining the fact that against this type of evidence (of correspondence) it cannot be objected that it does not provide an additional criterion of truth beyond the symbolic meaning of the proposition that can be inserted (incorporated) into a coherent system of propositions. It is simply implausible that the intuitional evidence of confirming, for example, the truth of the proposition " T h e church has two steeples" by having a look at the church should add n o t h i n g to that proposition than just a subjective evidence-feeling that could only be interesting for the psychologists. (Therefore it is also implausible and even amounts to a "category-mistake" to say that perceptual evidence is just a motive, i.e., a cause,
35

C. S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

187

by which the coming about of a "basis-sentence" of empirical science can be explained, fot the perceptual evidence is, of coutse, a good reason for accepting a perceptual judgment, and it can be undetstood as such also by the discourse-partnets of the human subject of the perception.)
36

At this point it becomes cleat, I suggest, that l a n g u a g c a n a l y t i c a l philosophy as the third paradigm of First Philosophy went too fat in its semanticist antipsychologism when it completely banished the achievements of evidencephenomenology ftom the business of philosophical "explication" in the so-called context of justification. (This is also suggested by the apotetic results of the attempts at explication of the different meanings of "explanation" under absttaction from the ptagmatic dimension of explanation as synthetic achievement of cognition.) The philosophical problem today seems to be tathet to combine the language-analytical with the phenomenological approach in such a way as to account for the "interwovenness" (Wittgenstein) of language use and experiential evidence within the frame of a language game. It is precisely this problem that is taken up and, I think, solved to some extent by Peirce's semiotical analysis of the interaction of the sign-functions of icons, indices, and symbols within the actual situation of a perceptual statement. Fot the special point of this semiotical account is the proof that the very use of the language of perceptual statements cannot be undetstood in terms of the semanticist presupposition that it conveys no othet meaning besides the symbolic meaning of absttact propositions that can be incorporated in a formalized semantical system.
37

But, as a semiotic approach, the Peircean analysis of the intetaction of i c o n s , i n d i c e s , and symbols within the actual situation of a perceptual statement also has implications that transcend and thereby show up the limits of a phenomenological evidence-theory of truth. For the intuitional iconical evidence that legitimizes the application of predicates to existing objects that can be identified with the aid of "indexical expressions" must also fot its pan be interpreted with the aid of the general meaning that is conveyed by the predicates as symbols that belong to a conventional semantical system. Hence our truth-claims cannot be immediately and definitely grounded on, or redeemed by, intuitional evidence; for all intuitional evidence of perception is always symbolically interpreted evidence, ot, in other words, is language-impregnated of even theory-impregnated. (For Peirce himself, the interdependence, or, respectively, interaction of indexical and iconical evidence, on the one hand, and symbolical interpretation in the light of general meanings, on the other hand, could figure as a semiotic equivalent to the Kantian insight that "concepts without intuitions ate empty and intuitions without concepts are blind.") This means, however, that the evidence can change since the iconical data may be reinterpreted in the light of othet language-systems, for example, of more powerful theories. There is no doubt that in this content the function of coherence as a criterion of ttuth becomes prominent. But this does not mean that the function of evidence (of correspondence) as a truth-criterion becomes dispensable (or even becomes equivalent to the recourse to a dogma, as it has

188

C S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

been s a i d . ) For again and again the object-relation or the contact w i t h reality of the coherent propositional systems has to be secured by perceptual judgments. It is appropriate, rather, to state that the criterion of evidence as related to a single interpreting consciousness and hence as a matter of "methodical solipsism" (Husserl) has to be distinguished from, and transcended in favor of, the requirement of intersubjective validity w h i c h comes into play w i t h the demand for a consensus about the correct or valid interpretation of evident data.
38

4. A t this point we can introduce the last and most important vantage point of a transcendental-semiotical truth-conception of Peircean inspiration. It could be called the transcendental-pragmatic alternative to the coherence-theory as an account of the systematic interconnection of the different criteria of truth as ingredients (moments) of an integrative explication of the meaning of truth. W e may start out w i t h the thesis that an absolutization of the coherencetheory of truth amounts to a metaphysical hypostatization of language insofar as it absolutizes the abstract unity of propositions as a medium of knowledge without reflection upon the two other places of the triadic relation of languagemediated cognition of something real by an interpreter or subject of cognition. In supposing coherence of propositions as the ultimate comprehensive criterion and explicans of the meaning of truth, it ignores the fact that coherence of propositions is only then a criterion of truth, if it may be interpreted by a subject of cognition as a coherence of such propositions that may be qualified as vehicles of true knowledge by their logical consequences, that is, basic propositions that may be interpreted again by a subject of cognition as being confirmed (or at least being not falsified) by evident data. T h e next step then is made possible by the insight that the transcendental subject of a possible unity of sign-interpretation cannot be represented by the consciousness of a single sign-interpreter; although his subjective witness of the objective evidence (of a correspondence between perceptual judgments and the given phenomena) is indispensable as an ingredient of that progressive process of re interpretation of evidences i n the light of coherent propositional systems that, i n principle, can lead to a unity of interpretation i n the long r u n . O n the basis of similar considerations, C . S. Peirce as early as 1868 came to state that the conception of true knowledge about the real to be reached by a scientific process of inference and sign-interpretation "involves the n o t i o n of a C o m m u n i t y , without definite limits, and capable of definite increase in knowledge" ( C o l l e c t e d P a p e r s 5.312). His point i n this context is that the K a n t i a n "thing-in-itself," that is, the real, w h i c h in fact can never be " k n o w n " by a single consciousness at any stage of the scientific process of sign-interpretation, must nevertheless be conceived as the "knowable" in the long run in relation to the indefinite community of interpreters (cf. also 5.257, 5.265, 5.275, 5.310). For the distinction between the real and the unreal, as we can learn it only by the occurrence of errors, Peirce claims, is equivalent to a distinction between " a n ens relative to private idiosyncrasy, and an ens such as would stand i n the long r u n . " T h e real, then Peirce

C. S . P E I R C E A N D P O S T - T A R S K I A N T R U T H

189

continues, "is that w h i c h , soonet or latet, information and teasoning would finally tesult i n , and w h i c h is therefore independent of the vagaries of me and y o u . " A n d since o n these ptesuppositions "a proposition whose falsity can nevet be discovered, and the ertot of w h i c h therefore is absolutely incognizable, contains . . . absolutely no e r r o r , " Peirce concludes, rightly I t h i n k , that the K a n t i a n idea of an unknowable thing-in-itself dissolves itself i n favor of the (regulative) idea of the teal as correlate of an ultimate consensus to be reached in the indefinite community of investigators (5.312). Later (especially 1871 and 1878) Peirce explicitly brings Together the ideas of truth, ultimate consensus, and reality, for example, i n the following formula: " T h e o p i n i o n w h i c h is fated to be ultimately agreed to by all who investigate is what we mean by the truth, and the object represented i n this o p i n i o n is the t e a l " (5.408; cf. 8.12). Thereby, I would suggest, Peirce has provided the basic elements of a transcendental-semiotic or, respectively, transcendental-pragmatic consensustheory of truth. For it may be said that the main elements of the K a n t i a n epistemology of "transcendental idealism" are here semiotically transformed, that is, suspended and preserved, in the following way: T h e idea of ttanscendental consciousness as subject and warrant of objectively valid knowledge is replaced by the idea of the indefinite community of sign-interpreters whose consensus is the only nonttanscendable ultimate criterion we can conceive of for the intetsubjective validity and thus for the truth of our knowledge. Correspondingly, the transcendental unity of objective data in the "synthesis of appercept i o n , " w h i c h Kant postulated as a transcendental synthesis, so to speak, of the truth-criteria of evidence (of phenomenological correspondence) and coherence w i t h i n , or for, the transcendental consciousness, is replaced by, or transformed into, the postulate or regulative principle of the ultimate unity of signinterpretation by consensus about the meaning and truth of p r o p o s i t i o n s . Correspondingly, the real as transcendental object of knowledge is conceived of not as that w h i c h can be factually " k n o w n " but as the " k n o w a b l e , " that is to say, as that w h i c h would be the object of the "ultimate o p i n i o n " of the indefinite community of investigators.
39

Thus we have singled out, I t h i n k , the core of Peirce's original contribution to a criteriologically relevant explication of the truth-conception of modern empirical science; although we have not yet taken into account the special connection of this early conception w i t h , and its supplementation by, Peirce's famous "pragmatist," or rather "pragmaticist," principle of meaning-explication, w h i c h , of course, can and must be also applied to the meaning of truth (although not necessarily in the way proposed by W . James). Before continuing w i t h our reconsttuction of Peirce's theory of truth, however, let us take a look at a contemporary vetsion of the consensus-theory of truth w h i c h in many respects comes close to Peirce's conception. W h a t 1 have i n m i q d is J. Habermas's outline of a "univetsal-pragmatic consensus or discourse-theory of t r u t h . " M i g h t it be possible, it may be asked, to confront Peirce's and Habermas's conceptions i n such a way as to mutually reconstruct their views and thereby
4 0

190

C . S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

supplement and improve them? In the next section I will try, at least roughly, to sketch some pertinent suggestions from the viewpoint of a ttanscendental semiotics and, respectively, transcendental pragmatics. IV Let us begin w i t h an attempt to single out the c o m m o n features or similarities, on the one hand, and the differences or even contrasts, o n the other hand, in comparing the approaches of Peirce and Habermas. First, I must emphasize once more that I consider both philosophies to be a version of what I call "transcendental pragmatics." In doing so, I am aware of the fact that this conjecture can be called into doubt or at least problematized i n both cases. M y justification for insisting o n it must be provided by the consideration that the very strength and significance of both philosophies as theories of t r u t h t h a t is, the possibility of transcending and integrating the traditional truth-theories into a consensusor d i s c o u r s e - t h e o r y of truthdepends in my o p i n i o n o n their t r a n s c e n d e n t a l pragmatic status. A s an illusttation of this thesis, I want to immediately introduce that I would consider the most profound c o m m o n basis of Peirce's and Habermas's truth-conceptions.
41

In Peirce the point I have i n m i n d is expressed in the following passages that belong to the context of K a n t i a n transformation that I have already brought into focus: " A proposition whose falsity can never be discovered, and the error of w h i c h thetefore is absolutely incognizable, contains, upon out principle, absolutely no error" (Collected Papers 5.311). A n d "any truth more perfect than this destined conclusion [namely, " w h i c h is fated to be ultimately agreed to by a l l who investigate"] . . . any teality more absolute than what is thought i n it, is a fiction of metaphysics" (8.12). T h i s is what I would call a sense-critical argument, since it is apt to show that we cannot meaningfully contrast a transcendent metaphysical ultimate ctiterion of truth w i t h the regulative idea of an ultimate consensus of the indefinite community of sign-interpretation and hence have to acknowledge that regulative idea as a ttanscendental-ptagmatic presupposition of our enterprise of investigation, if it should make sense at a l l . T h i s , I t h i n k , is the teason why Peirce can postulate that a l t h o u g h o r rather, precisely becausewe have n o guarantee "that man or the community (which may be wider than mankind) shall ever arrive at a state of infotmation greater than some definite finite information"; since life may be annihilated at some time, the idea of the ultimate consensus "involves itself a transcendent and supreme interest" that may be, o n the one hand, the subject of an "infinite h o p e , " and, o n the other hand, "is always a hypothesis uncontradicted by facts and justified by its indispensableness for making any action rational" (5.357). N o w , I think, a connection can be established o n this very fundamental level of sense-critical or transcendental-pragmatic arguments between Peirce's approach and Habermas's "universal pragmatic" doctrine of the four " v a l i d i t y - c l a i m s "

S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

191

[i.e., intelligibility, truth, Tightness, and veracity (sincerity, truthfulness)] that are necessarily connected w i t h h u m a n speech-acts inasfat as speech is the vehicle and witness to human reasonableness qua majority ( M u n d i g k e i t ) . G i v e n i n this context, the following passages of Habermas seem to be relevant: O n this view [i.e., according to the thesis that the "meaning of t r u t h " may be explicated i n terms of a possible redemption, i.e., "discursive settlement of experience-based validity-claims"] I may asctibe a predicate to an object if and only if every othet person who c o u l d entet into dialogue w i t h me would also ascribe the same predicate to the same object. In order to distinguish true from false statements I make reference to the judgments of o t h e r s i n fact to the judgments of all others w i t h w h o m I could evet hold a dialogue (among whom, contrary to fact, I include all the dialogue-partners I could find if my life-history were coextensive w i t h the history of m a n k i n d ) . . . . T r u t h means the promise of attaining a rational consensus.
42

It is important to note that Habermas here is speaking about a necessary explication of the m e a n i n g of truth and not about a sufficient c r i t e r i o n of t r u t h . T h e difference between these two aspects of the problem is clarified by the following passage, w h i c h serves as an answer to the objection that accidently brought about agreements cannot serve as truth-criteria: "discursive settlement" is a normative concept: O n l y a g r o u n d e d consensus is a consensus we can reach i n discourses. T h i s one alone holds a truth-criterion, but the meaning of ttuth is not the circumstance that a consensus w i l l i n fact be reached, but (the presupposition) that at any time and place, if only we enter into discoutse, a consensus can be arrived at under conditions that warrant its being a grounded consensus.
43

T h e point that connects this explication of the meaning of truth w i t h Peirce's statements lies i n my o p i n i o n i n the following circumstance: In these statementsas a result of transcendental-pragmatic or s e n s e - c r i t i c a l reflection, as 1 would c l a i m a postulate is exposed whose validity c a n be defended quite independently, and hence i n advance, of the settlement of the question whethet or, respectively, how the bringing about of a grounded consensus can be secured; although this question of the conditions of a grounded consensus w i l l be the next question to be settled and as such is dealt w i t h i n Peitce's as well as i n Habermas's approach. T h u s the need for explicating the pragmatic meaning of t r u t h t h a t is, of what truth can mean for u s i n terms of an ultimate consensus o m n i u m is based o n the fact that it, from the outset, is the only alternative to giving upon the whole enterprise of cognition or investigation i n the sense of arriving at knowledge whose truth-claim can be justified by arguments. It makes no senseand testifies to a misunderstanding of t r a n s c e n d e n t a l - p r a g m a t i c a r g u m e n t s i n this situation to first insist o n answering the_question, whether the transcendental-pragmatic presupposition is true, that is, can be verified at all as a n hypothetic assumption, so to speak.
44

For this very question of a possible verification of an assumption cannot be

192

C . S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

meaningfully tackled without presupposing already the adequacy of the initial move of a consensus-theory of ttuth. W e thus far get hete a glimpse of what may be called a t r a n s c e r u k n t a l - p r a g m a t i c ultimate f o u n d a t i o n of philosophy (and science), w h i c h , of course, is very different from a deductive grounding by axioms (and hence of axioms) that could be objectified w i t h the aid of a semantical system of sentences or propositions.
45

N o w , before trying to settle the question of the ideal conditions of a grounded consensus, we should ask for the relationship between the consensus-theory as an explication of the meaning of ttuth and the above-mentioned truthconceptions, as, for example, the correspondence', e v i d e n c e - , coherence-, and pragmatic satis/action-theories. W i t h regard to this question there seems to be some centtal concotdance as well as some chatactetistic difference between Peirce and Habermas. A fundamental concotdance, as fat as I can see, exists between both thinkers w i t h respect to the insight that it is not the real ( i n itself), as it is conceived i n a metaphysical correspondence-theory as independent of out thought i n geneial ot, more precisely, of the possible redemption (discutsive settlement) of out tiuth-claims, that can seive as tianscendent c i i t e r i o n or standard of truth but rathei that ultimate consensus o m n i u m w h i c h would necessatily constitute the correlate of the teal as it can be meaningfully thought of as that " w h i c h would h o l d i n the long r u n . " From this concordance I would derive the fact that both Peirce and Habermas sometimes quite naturally (without embarrassment) speak of a necessary correspondence between our judgments ot ptopositions and the facts ot reality. W h a t is meant i n this context, I suggest, is only that correspondence with the correlates of true propositions that is necessaiily and (trivially) presupposed by any tiuth-theory, but cannot without circularity be explicated i n a ciiteriological relevant w a y . Hence the later Peirce even gives a methodologically relevant definition of truth i n terms of "concordance w i t h . . . " that avoids talking of reality: " T t u t h is that concordance of an absttact statement w i t h the ideal limit towatd w h i c h endless investigation would tend to bring scientific belief" (Collected Papers 5.565). Peirce adds that such a concordance can, of course, only be ascribed to an absttact statement together w i t h the concession of its inexactness and its onesidedness.
4 6

Yet, besides that, there is also a difference between Peirce's and Habermas's use of the terms "facts" and "correspondence with the facts"; such that one eventually may derive from this difference a distinction between two different types of a consensus-theory of ttuth. Peiice seems to propose a consensus-theory that integrates, together with the coherence-theory and the pragmatist theory of truth, a certain phenomenological account for the evidence of correspondence. Habermas, o n the othet hand, seems to suggest a consensus-theory that father excludes the other truth-conceptions, especially the phenomenological conception of the evidence of correspondence; and precisely thereby Habermas's theory seems to come close to a pure coherence-conception of t r u t h . Habermas's truth-conception seems to me to suffer from the following ambi47 4 8

C. S

PEIRCE A N D POST-TARSKIAN TRUTH

193

guity: O n the one hand, it presupposes that the abstract facts which are the correlates of true propositions are asserted by us with respect to objects in the world and hence the meaning of truth must be explicated in terms of a "discursive settlement" of "experience-based validity-claims." He even grants in this context that the correspondence-theory of truth rests on a "correct observation" insofar as it insists on the postulate that true propositions must be "supported" or "founded" by "facts" in the sense of something that is "given" in "objects of experience." A n d thus he suggests that it might be possible to "explicate the pragmatic relation between cognition and objects of experience with the aid of the concept of correspondence." O n the other hand, Habermas makes a rigorous distinction between the question of truth about abstract facts, which for him is a question of reflective validity-claims to be settled on the level of argumentative discourse, and the question of evidence about objects of experience, which for him is a question of prediscursive information within the frame of prereflective communicative interaction between people. Thus he concludes: "Questions of truth are consequently taised not so much in reference to the innerworldly correlates of action-related cognition as rathet to facts associated with discourses that are experience- and action-free. Whether states of affairs are or are not the case is not decided by experiential evidence, but by the course of argument."
49 50

Now I would not dispute the difference stated by Habermasalong with Ramsay and Strawsonbetween objects of prediscursive experience and abstract facts as reference-matter of truth-claims and their discursive settlement since it is only this diffetence that can explain the nonredundance of the predicate "is true" in the case of reflective truth-claims and their discursive settlement (which 1 think is a finer point of Habermas's). But I would emphasize at the same time that the discursive settlement of empiric-scientific truthclaims, in contradistinction say to those of mathematics, does presuppose as a necessary condition what Habermas grants as the possibility of making use of experience also in the context of discursive argumentation. This means however that the discursive settlement of empiric-scientific truth-claims by "substantial arguments" must not immediately take recourse to explanatory arguments about facts as explananda, as Habermas suggests, but first of all must take recourse to reports that testify to the existence of the explananda. A n d these reports have to come down to perceptual statements, if the reporter should be pressed by his opponents in the discourse. Now at this point, I think, the superiority of a transcendental-pragmatical consensus-theory of truth over a coherence-theory of discourse rests precisely on the fact that the former can take the evidence-testimonies of the single discourse-partners as criteriological ingredients of the very discourse to be weighed out in connection with the theoretical coherence-criteria. For it is not the case, as it is sometimes suggested today by a logistic type of analytic philosophy, that the ("indexical" and "iconical") evidence of perceptual judgments can only be accounted for as external causes of basic propositions, so that these are by no means distinguished by their evidence
51 52

194

C . S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

from other propositions. It rather belongs to the very procedure of consensus formation, I suggest, to understand the evidence of perceptual judgments as good reasons for the reporter's accepting them and hence as criteria of possible truth to be discursively mediated w i t h other criteria. In order now to understand and evaluate the Peircean approach as a possible alternative or supplementation of the contemporary type of consensus-theory of truth, one should start out, I suggest, from facing the problem of how a grounded consensus can function and h o l d as ultimate criterion or regulative principle for an integration of all relevant truth-criteria. T h e answer to this question that can be found i n Peirce is provided in the last analysis by the doctrine of the three fundamental categories. It was first developed by Peirce i n his " O n a N e w List of Categories" of 1867, that is, i n the context of a transcendental-semiotical transformation of Kant's so-called "metaphysical" and "transcendental deduction of the categories." (Later Peirce has supplemented or even replaced his first transcendental-semiotical derivation of the categories by a derivation of the three fundamental categories from the mathematical logic of relations and a "phenomenological" or "phaneroscopial" illustration of their philosophical applications.)
53 54

A c c o r d i n g to this doctrine, there are three aspects of reality as well as of its semiotically mediated cognition that must be taken into account by epistemology: "firstness," that is, relation-free quality as a pure possibility of being; "secondness," that is, the dyadic relation or " c l a s h " between the I as w i l l and existent reals as "brute facts" that resist our w i l l ; and finally "thirdness," that is, the triadic relation of mediation between possible qualities and existent (dyadic) facts by general meaning, that is, general laws of nature or general rules of interpretation and hence of semiotic representation and continuation of structure of reality by human "habits" of action. N o w , o n these presuppositions, the question as to how a grounded consensus of the community of investigators may be conceived of as the ultimate criterion or regulative principle of an integration of all possible truth-criteria may be answered roughly i n the following way. Since truth must take the form of a public representation of the structure of reality (and since reality cannot be explicated i n a criteriologically relevant way in terms of a transcendent reality), it must be finally explicated in terms of sign-interpretation, for example, o n the level of argumentative discourse, as it was first developed by philosophy. O n the other h a n d , the deficiency of the (scholastic and the rationalistic a priori) method of purely philosophical seminar-discussion had consisted in the fact that its way of reaching consent by arguments, if it reached it at a l l , left out of consideration (ignored) the vote of nature, so to speak. N o w this vote can be established, according to Peirce, by experiments, that is, by the "method of science." T h i s means, according to my interpretation, that the prediscursive ways of experiential integration of firstness and secondness into thirdness, that is, data-interpretation by perceptual judgments, are put into the service of scientific discourse by being methodically used
5 5

C. S . P E I R C E A N D P O S T - T A R S K I A N T R U T H

195

as sources of inspiration for setting up innovative hypotheses or as means of experimental confirmation of theoretical hypotheses. The main problem for Peirce in this context was that of bridging the gap between the particularity of our intuitions of perceptual qualities including relations (firstness) on the occasion of factual confrontations with resistent "brute facts" (secondness) and the universality of interpretative meaning in arguments. And the answer he had to propose and which he elaborated throughout his life is constituted by his conception of a methodological connection between three types of inference: namely, d e d u c t i o n as pure expression (model) of thirdness, that is, mediation without information, and "synthetic" inference in the sense of i n d u c t i o n and a b d u c t i o n or r e t r o d u c t i o n . I n d u c t i o n in this context should primarily stand in the service of confirmation on the basis of enumerative encounters with brute facts. Thus it would bring home, so to speak, the thirdness from secondness. Abduction or r e t r o d u c t i o n , on the other hand, is supposed to provide the creative synthesis of data experience by interpretative and explanatory hypotheses. Thus it draws, so to speak, conceptual thirdness from intuitive firstness. In this context two of Peirce's theorems are of special importance. O n the one hand, Peirce held that our perceptual judgments are unconscious cases of abductive i n f e r e n c e , although he was later prepared to concede that there are limit cases of perceptual judgments that are beyond the scope of man's control of his logical operations and hence must be taken as practically uncriticizable starting points of inference and interpretation. In any event, for Peirce this theorem had to provide a crucial mediation between firstness and thirdness and thus between the particularity of our intuitions and the universality of the meaning of our arguments. O n the other hand, a second crucial mediation between these poles, namely, thirdness of secondness, is made plausible, on Peirce's account, by a theory of inductive confirmation.
56

As far as I can see, Peirce here proceeded from Kant's suggestion that the universality and necessity of scientific inductions are but the analogues of philosophical universality and necessity (Collected Papers 5.223, note). From this he drew the conclusion that we can and must postulate that "whenever instance may be had in as large numbers as we please, ad infinitum, a truly universal and necessary proposition is inferable" (5.223, note). Thus Peirce came to substitute in the place of Kant's "transcendental deduction" of the validity of "synthetic judgments a priori" a transcendental deduction of the validity of inductive inference in the long run (see especially 5.223, notes, 5.349-52 and 2.690 - 93). Later, especially in his L e c t u r e s o n Pragmatism of (1903), Peirce was no longer satisfied by this foundation of the validity of induction by a transcendental postulate or regulative principle, because it did not sufficiently show "why perception should be allowed such authority in regard to what is real" (5.211). He now claimed as an additional criterion for the truth of inductive inferences that, analogous to perception as unconscious abduction,
57

196

C. S P E I R C E A N D P O S T - T A R S K I A N

TRUTH

it should be possible to intuitively perceive the generality of a governing law or rule as the continuity of reality and thus provide a perceptual evidenceorientation for that process of inductive confirmation that, according to the transcendental postulate, would prove the reality of the universality of a hypothesis in the long run (cf. 5.198ff. and 5.209-211). Thus Peirce came to conceive of a "firstness of thirdness" in the case of inductive evidenceconfirmation as well as in the case of abductive hypothesis as a perceptual basis that is supposed by our synthetic inferences as arguments.
58

In order to present this doctrine of "synthetic inferences" as Peirce's answer to the problem of bridging the gap between the particularity of sensuous evidence and the universality of argumentative truth-claims, one has however to consider, I would emphasize this, that perceptual judgments as the basis of abductive as well as inductive inference imply always prior symbolic interpretations in the light of language-systems. Peirce, it is true, in contradistinction to present philosophy after the so-called linguistic turn, did not so much stress the somewhat relativistic dependence of sign-interpretation on language-systems (or, for that matter, "language games" as parts of "forms of life") as he did the indefiniteness of the process of sign-interpretation in the indefinite community of interpretation. Fot Peirce this latter process seemed to represent something like a normatively postulated universal language game of progressive science that was credited a priori with making possible an agreement about meaning between scientists even across the different languages, so to speak.
59

In this context, I think, one should mention an ingenious continuation and supplementation of Peirce's semiotics by the later ]. Royce, who emphasized that simultaneously with, and complementarily to, perceptual cognition we have to constantly reinterpret the meanings of our words as part of the process of cultural tradition which even must take place as an internalized process in our tacit use of language. I would suggest it is only by presupposing this supplementation of Peirce's approach, that is, by taking into consideration the hermeneutic conditions of communicative understanding as an at least implicit precondition of scientific-philosophic discourse, that we can adequately assess the preceding suggestions. For the hermeneutic account of communicative understanding has to complete the transcenoental-semiotic account for the possibility and necessity of an integration of perceptual criteria concerning the evidence of correspondence of our thought with brute facts and iconical qualities and relations as well as of coherence-criteria of logical inference into a grounded consensus by discursive arguments in the interpretation-community of scientists.
60 61

Habermas seems to reconfirm this point, in a sense, when he expresses the assumption "that there is a connection between the consensus-producing force of an argument on the one hand and the corresponding conceptual system on the other," so that "an argument is satisfactory only if all its parts belong to the same language." According to modern insights into the significance of language-systems for the formation of theories, Habermas indeed emphasizes that "observation-data . . . which we may want to bring into arguments are,
62

C. S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

197

naturally, interpreted experiences and therefore dependent upon the categorial framework of the chosen language-system." But contrary to certain postWittgensteinian tendencies of considering semantical systems as ultimate incommensurable frames of possible understanding and basing arguments on natutal history, Habermas considers it possible that argumentation "is not metely based upon a telation between linguistic system and reality w h i c h is 'appropriate' in the sense that it has been antecedently drilled into it by a natural and biological process of c o g n i t i v e d e v e l o p m e n t , but itself sets f o r t h the m e d i u m i n w h i c h that cognitive development can be continued as a conscious learning process." In this sense he considets it a necessary c o n d i t i o n of the possibility of a grounded consensus that can figure as a criterion of truth that "ptogtess i n knowledge" can take the " f o r m of a substantial critique of language," so that the "structural possibility exists of inquiring behind, modifying, and replacing the warranting language i n w h i c h experiences at any given time ate interpreted. "
63 64 6 5

I think indeed that this teflective explication and reformation of the language of discourse must itself form the hermeneutic patt of atgumentative discoutse w h i c h must make up a continuum, so to speak, w i t h hermeneutic-teconstructive sciences as, for example, the history of science. T h e lattet type of a "science of science" would then of coutse not be just an empirical social science that inquires into the causes of "patadigm-change" but father a critical mediation of causal explanations of quasi-natutal processes and of normatively televant reconsttuctions of the good ot bad rules and reasons that promoted ot blocked the ptogtess of science. A n d the ultimate guiding n o r m of this attempt of hermeneutic teconsttuction would be provided by the very regulative ptinciple of integrating all possible truth-criteria into a consensus by argumentative discourse. Such a program of a continuation of the quasi-natural process of human cognition by discursive procedures w i t h i n the interpretation-community of investigators was also the object of Peirce's conception of a " n o t m a t i v e " logic of inquiry. H e conceived of it as a continuation and supetsedure of our first " i n s t i n c t i v e " theotetical divinations by methodical "self-control" as patt of an anthropological and theteby even cosmological "rationalization" process.
66

It is only i n connection w i t h this "tationalization"-ptocess that Peirce also applies the "ptagmatic m a x i m " of meaning-explication to the concept of ttuth (see, for example, 5.375, n . 2). T h a t is to say, thete are two dimensions of an explication of ttuth i n terms of practical corroboration by satisfaction that Peirce would envisage. O n e of them refers to corroboration w i t h i n the process of experimental science, the othet refers to that growth of "concrete reasonableness" by human "habit"-formation w h i c h Peirce conceives of as possible consequence of the method of science in society. In this latter context, however, the later Peirce had to face the problem of a certain revision)'or supplementation of the "pragmatic m a x i m , " since he realized that the idea of a "pragmatic" corroboration by experiments always presupposes prior "practical" aims (cf. 5.412). Hence the problem of a practical corroboration of science itself would involve the question of a summum b o n u m w h i c h cannot be settled by

198

C S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

experimental science. It is in this context that he wrote: "In order to understand pragmatism, therefore, well enough to subject it to intelligent criticism, it is incumbent upon us to inquire what an ultimate aim, capable of being pursued in an indefinitely prolonged course of action, can be" (5.135; cf. also 5.3 and 5.402, n.). But I do not think that Peirce became fully aware of those problems of truth and of its relations to social praxis that, so to speak, fell between science and ethics, such as the problematic of the truth of communicative understanding between people as participants of a communication-community.
67

Our discussion of the linguistic and hermeneutic preconditions of the idea of discursive consensus-formation may have shown that a full-fledged explication of the meaning of truth in terms of discursive consensus cannot be provided without answering the question as to the conditions of a grounded consensus. In fact, I would reckon all epistemologically televant devices of integrating relevant truth-criteria into the discursive consensus-formation with the conditions of a grounded consensus. Nevertheless there are still furthet extraepistemological conditions of the possibility of a grounded consensus which have been taken into consideration by Peirce as well as by Habermas. Peirce as early as 1869 in his essay on "The Grounds of Validity of the Laws of Logic" stipulated that nobody could be logical in his (synthetic) inferences who would not sacrifice his personal interests, including even his interest in the salvation of his soul by a religious belief, to the interest of the indefinite community of investigators in reaching the truth in the long run (cf. 5.354). And from this he concluded that the ideal perfection of knowledge would belong to a community within which that identification by "self-surrender" of every member with the community would be complete (5.356). Latet, in his foundation of the "normative sciences," he simply stated that the logic of inquiry, being a normative science of acting for a good purpose, presupposes the "normative science par excellence," namely, "ethics" (cf. 5.130, 5.131, 5.36, 1.573, 1.91, 1.612f., 2.199). Habermas comes to similat conclusions. Since he traces back the pragmatic meaning of truth to the meaning of just one of the validity-claims which ate necessatily connected with human consensual speech-acts, namely, intelligibility, truth, rightness, and veracity ( s i n c e r i t y ) ; and since he furthermore can state thatfor reasons of a transcendental-pragmatic character of necessity, I would claimin our consensual speech-acts we must, more or less counterfactually, anticipate an "ideal speech-situation" by which the necessary conditions of a redemption of all four validity-claims would be fulfilled, he comes to postulate that only the fulfillment of certain ethical requirements of the ideal speechsituation can guarantee that "liberation of discourse from constraints of action" which is a precondition for a discursive settlement of validity-claims and among them truth-claims. It is interesting to note in this context one characteristic difference in Peirce's and Habermas's determination of the ethical preconditions of argumentative discourse: Peirce connects his principle of "self-surrender" with a

C. S

PEIRCE A N D POST-TARSKIAN TRUTH

199

rather negative theory of the individual self," so that for him the suppression of individuality rather than self-expression of self-representation seems to be a requirement of the ethics of argumentative discourse. (In fact, Peirce never extends the application of the concept of argumentative discourse beyond the range of scientific discourse.) Habermas, on the other hand, claims "two nontrivial conditions which must be satisfied in ideal speech-situations, in older to insure that the participants in the conversation actually can take up a discourse and not merely imagine they are conducting a discourse while in truth they are communicating under pressure of action": The one condition consists in an "equal opportunity" for all discourse-partners "to avail themselves of representative speech-actsthat is, to bring their attitudes, feelings, and intentions to expression." This requirement of a "reciprocal harmonisation of the contexts of individual utterances" has to be fulfilled, on Habermas's account, in otder to make possible the veracity (sincerity) of the individual persons as discourseparticipants. The other condition consists in an "equal opportunity" for all discourse-partners "to avail themselves of regulative speech-actsthat is, to command and resist, to permit and to forbid, to make and receive promises, to give and demand accountings, etc." This requirement of "complete reciprocity in regard to expectations of behavior," excluding privileges, has to be fulfilled, in order to guatantee the "suspension of reality-consttaints" on the discourse, that is, its unburdening from action by "the formal equal distribution of opportunity to initiate and continue discourse."
68 69 70

I think in fact that Habermas's postulates are an even more adequate, that is, more concrete, account of that fundamental requirement of "transsubjectivity" that makes up the true core of Peirce's idea of "self-surrender," since it does not exclude and condemn as idiosyncratic the possible contributions of individuality to the discourse but rathet provides equal chances for them. The background for Habermas's more realistic account of the ethical preconditions of a grounded consensus can be found, 1 suggest, in his mote concrete account of the social, that is, communicative-hermeneutic presuppositions of human understanding as a precondition of argumentative discourse, even in the case of natural science. For it is above all the requirement of making transparent and revisable even the prereflective quasi-natutal preconceptions as they are implied in the language games as parts of "forms of life" that motivates Habermas's postulates.
71

But finally there is again a profound agreement between Habermas and Peirce in regard to the question of the readability of the ethical preconditions of a grounded consensus. As an ideal criterion for our judging, that is, legitimizing or criticizing, real agreements with regard to their having fulfilled the ethical prerequisite of a genuine consensus, the ideal speech-situation is a regulative principle in the Kantian sense; but beyond that, the ideal speech-situation is for Habermas a "constitutive condition of rational discourse," since it is an unavoidable "imputation" of the participants in such a discourse, even if it has at the same time

200

C. S P E I R C E ' A N D POST-TARSKIAN

TRUTH

the character of a "counterfactual anticipation" of a "foreshadowing lifeform." Now, since we cannot know a priori "whether this preshadowing . . . is . . . a mere deception (subreption) . . . or whether the empirical conditions for the (if only approximate) tealization of the supposed life-form can practically be brought into being," but nevertheless cannot help making that "imputation" as a "practical hypothesis," as long as we are willing to argue (that is, to think!), we are again confronted with that structure of a sense-critical or transcendentalpragmatic argument that I have exposed in the preceding as paradigm of a nondeductive type of ultimate foundation in philosophy. Precisely this, it seems to me, is the structure of Peirce's principle of "hope," which he calls a "hypothesis uncontradicted by facts and justified by its indispensableness for making any action rational" (5.357; cf. 2.654f.; 2.661).

NOTES
1. Cf. K . - O . Apel, "The Apriori of Communication and the Foundations of the Humanities," M a n and World 5 (1972): 3-37; and idem, "Types of Social Science in the Light of Human Interests of Knowledge," Social Research 44 (1977): 425-70. 2. Cf. K . - O . Apel, Der Denkweg von Charles S. Peirce (Frankfurt a . M . : Suhrkamp, 1975). Cf. also my German edition of C . S. Peirce, Schriften, 2 vols. (Frankfurt a.M.: Suhrkamp, 1967 and 1970). (English trans.: Charles S. Peirce: From Pragma' asm to Pragmaticism [Amherst, Mass.: The University of Massachusetts Press, 1981].) 3. C f . K . - O Apel, "Szientismus oder transzendentale Hermeneutik? Zur Frage nach dem Subjekt der Zeicheninterpretation in der Semiotik des Pragmatismus," i n Hermeneutik und Dialektik, Festschrift fur H . - G . Gadamer, ed. R. Bubner et al. (Tubingen: Mohr, 1970), vol. 1, 105-45; repr. in K . - O . Apel, Towards a Transformation in Philosophy (London: Routledge and Kegan Paul, 1979); K . - O . Apel, " C . W . Morris und das Programm einer pragmatisch integrierten Semiotik," Introduction to C . W . Morris, Zeichen, Sprache, und Verhalten (Dusseldorf: Schwann, 1973), 9 - 6 6 ; K . - O . Apel, "Transcendental Semiotics as First Philosophy," in this volume (originally published as "Transcendental Semiotics and the Paradigms of First Philosophy," Philosophic Exchange 2, no. 4 [1978]: 3-22; page references are to this volume). 4- For a thorough account of the aporetics of a logico-semamical explication of the concept of explanation, see W . Stegmuller, Wissenschaftliche Erkldrung und Begrundung (Berlin and New York: Springer, 1969); for an alternative approach, see K . - O . Apel, Die Erklaren/Verstehen-Kontroverse in transzendental-pragmatischer Sicht (Frankfurt a . M . : Suhrkamp, 1979). Cf. also K . - O . Apel, "The Erklaren/VerstehenControversy in the Philosophy of the Human and Natural Sciences," in Chronicles (of the International Institute for Philosophy), vol. 2 (forthcoming), ed. G . Fl0istad. 5. Cf. most recently H . Keuth, Realitdt und Wahrheit (Tubingen: Mohr, 1978). 6. Cf. A . Tarski, "Die semantische Konzeption der Wahrheit," in Zur Philosophie der idealen Sprache, ed. Sinnreich (Munich: Deutscher Taschenbuch Verlag, 1972), 77 and especially 87. 7. See, e.g., Collected Papers 4-56: "It seems certainly the truest statement for most languages to say that a symbol is a conventional sign which being attached to an object signifies that that object has certain characters. But a symbol, in itself, is a mere dream; it does not show what it is talking about. It needs to be connected with its object. For that purpose, an index is indispensable. N o other kind of sign will

C. S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

201

answer the purpose. That a word cannot in strictness of speech be an index is evident from this, that a word is generalit occurs often, and every time it occurs, it is the same word, and if it has any meaning as a word, it has the same meaning every time it occurs; while an index is essentially an affair of here and now, its office being to bring the thought to a particular experience, or series of experiences connected by dynamic relations. " A meaning is the association of word with images, its dream exciting power. A n index has nothing to do with meanings; it has to bring the hearer to share the experience by sfiouiing what he is talking about. The words this and that are indicative words. They apply to different things every time they occur. "It is the connection of the indicative word to a symbolic word which makes an assertion." Cf. 2.335, 3.363, 2.287, 8.41. 8. Cf., e.g., 5.30 and 2.315. 9. For the "redundance"-theory of truth, cf. F. P. Ramsey, "Facts and Propositions," Proceedings of the Aristotelian Society, supp. vol. 7 (1927); A . J. Ayer, Language, Truth, and Logic (London: Gollancz, 1935); P. F. Strawson, " T r u t h , " Analysis 9 (1949); for the proposed solution of the problem that was exposed by the redundance-theory, see J. Habermas, "Wahrheitstheorien," in Wirklichkeit und Reflexion, ed. Fahrenbach (Pfullingen: Neske, 1974), 211-65. 10. See, e.g., C . W . Morris, Foundations of the Theory of Signs (Chicago: International Encyclopedia of Unified Science, 1938), vol. 1, part 2, pp. 29, 35, 52; R. Carnap, "Testability and Meaning," Philosophy of Science 3 (1936), and 4 (1937): 454; R. Carnap, Introduction to Semantics (Cambridge, Mass.: Harvard University Press, 1942), sees. 4, 28 [9], 39; R. Carnap, " O n Some Concepts of Pragmatics," Philosophical Studies 6 (1955): 89-91; R. Martin, Towards a Systematic Pragmatics (Amsterdam: North Holland, 1959). For a critical reconstruction of this whole development of logical empiricism, see E. Tugendhat, "Tarskis Semantische Definition der Wahrheit und ihre Stellung innerhalb der Geschichte des Wahrheitsproblems im Logischen Positivismus," Philosophisches Rundschau 8 (1960): 133-59; repr. in Wahrheitstheorien, ed. G . Skirbekk (Frankfurt a . M . : Suhrkamp, 1977), 189-223. Cf. also my papers quoted in nn. 3 and 13. 11. See K. R. Popper, Objective Knowledge (Oxford: Clarendon Press, 1972), 106-52. 12. See my Denkweg von Charles S. Peirce, and my paper, "From Kant to Peirce: The Semiotic Transformation of Transcendental Logic," in Proceedings of the Third International Kant Congress, 1970, ed. L. W . Beck (Dordrecht, Holland: Reidel, 1972), 90-104; repr. in Kant's Theory of Knowledge, ed. L. W . Beck (Dordrecht, Holland: Reidel, 1974), 23, 37. A sforPeirce's relation to Hegel, cf. also Max Fisch, "Hegel and Peirce," in Hegel and the History of Philosophy, ed. J. J. O'Malley et al. (The Hague: Martinus Nijhoff, 1973). 13. For the development of this idea, cf. the following of my papers: "Programmatische Bermerkungen zur Idee einer transzendentalen Sprachpragmatik," in Studia Philosophica in Honorem Su<en Krohn, ed. T. Airaksinen et al. (Turku: Annales Universitatis Turknensis, 1973), 11-36; repr. in Semantics and Communication, ed. C . H . Heidrich (Amsterdam: North Holland; and New York: Elsevier, 1974), 81-108; "Zur Idee einer transzendentalen Sprachpragmatik," in Aspekte und Probleme der Sprachphihsophie, ed. ). Simon (Freiburg/Munich: Alber, 1977), 83-126; "Transcendental Semiotics as First Philosophy." 14. Cf. Popper, Objective Knowledge, chap. 3. I am, of course, not arguing against Popper's distinction between a realm of "objective," i.e., true or false, knowledge and a realm of factual beliefs of real subjects to be thematized in history or in psychology. I am, however, arguing against Popper's, as well as Bolzano's and the Semanticists', overlooking the (transcendental-pragmatic) fact that the concept of propositions or theories chat may be conceived as true or false and thus may be called "objective

202

C S

PEIRCE A N D POST-TARSKIAN

TRUTH

knowledge" as being the topic of a nonpsychologistic epistemology must presuppose the concept of a corresponding subjective t r u t h - c l a i m . Now insofar as a human subject of knowledge is the carrier of this claim and thereby enters the "indefinite community of investigators" (Peirce), it is not only an "inhabitant of the second world," to use Popper's term, but moreover in charge of the transcendental subject of true knowledge, so to speak. Without this latter presupposition, which reestablishes Kantianism in terms of a three-dimensional semiotics, the meaning of "objective knowledge" was bound to dissolve itself. For its meaning cannot be that of an hypostatized tealm of being in the Platonist sense but, I suggest, is rather that of the correlate of arguing (i.e., proposing, confirming, or refuting) within the frame of argumentative discourse. It is interesting in this context that G . Frege, who by his strong verdict against psychologism is one of the fathers of modern s e m a n t i c i s m , was still aware of the fact that the epistemological dimension of the subject of knowledge cannot be simply surrendered to psychology. Thus, after having pointed out that the law of gravitation can never be thought of as an association of ideas (Vorstellungen) in somebody's mind, he continues: "Yet the grasping of this law is a psychic process, after all! Yes, but it is a process that is rather located at the border of the psychic and therefore will not be capable of being completely understood from a purely psychological standpoint; for something must be considered essential in this context that is not psychic in the proper sense, namely the thought (der Gedanfee); and this process is perhaps the most mysterious (der geheimnisvoHste v o n alien)." Frege indeed dismisses this question by stating: "Since however it (that is, the thought) is of a psychic kind, we need not care for it within logic. It is enough that we can grasp thoughts and recognize them to be true. How this might be possible is a question for itself." (G. Frege, Schriften zur L o g i k und S p r a c h p h i l o s o p h i e , ed. G . Gabriel [Hamburg: Felix Meiner, 1971], 63f. trans. K . - O . Apel). This, I think, is not a suggestion of an "epistemology without a knowing subject," but only an abstractive isolation of the object of formal logic. 15. Cf., e.g., B. Bolzano, Wissenscria/tslehre, vol. I (Sulzbach: 1837), 113ff. By postulating a transcendental subject of true knowledge as a presupposition for the meaning of the predicate ". . . is true," 1 am not maintaining, as against Bolzano, that truth had to be conceived as being "posed" (gesety) by the divine intellect or a transcendental consciousness. I am in fact maintaining that it cannot be conceived meaningfully without at the same time conceiving in principle of a subject that may claim that its statements (i.e., stated propositions) are valid for, and hence could and should be accepted by, every other possible subject of knowledge irrespective of their being factually asserted. This tenet may figure as an implicit definition of the concept of a t r a n s c e n d e n t a l subject of true knowledge within the framework of transcendental semiotics. 16. The term "double-structure of speech" (DoppelstruktuT der Rede) was introduced by J. Habermas in his essay "Vorbereitende Bemerkungen zu einer Theorie der kommunikativen Kompetenz," in T h e o r i e der Gesellschaft oder Sozialtechnologie, ed. J. Habermas and N . Luhmann (Frankfurt a.M.: Suhrkamp, 1971), 104ff., in order to summarize a chief result of Austin's and Searle's speech-act theory to the effect that by the double-structure of speech among other things a self-referential exposition of the implicit validity-claims (meaning-claims, sincerity-claims, truth-claims, and rightness-claims) of human speech is made possible. (Cf. also n. 9, above.) I would suggest that by the idea of the "double-structure of speech" a new paradigm is provided for the philosophy of language in view of the fact that the paradigm of the semanticist era of language-analysis was rather represented by Karl Biihler's distinction between the "representative" function of language (i.e., of p r o p o s i t i o n s . ) , which
1

C. S P E I R C E A N D P O S T - T A R S K I A N T R U T H

203

alone was considered as constituting the priority of human language, and the communicative functions (i.e., the functions of "self-expression" and "appealing") which were considered to be common to humans and animals. (Cf. K . Biihler, Sprachtheorie (Jena: Fischer, 1934; Stuttgart: Fischer, 1965). The point of the new paradigm is of course completely missed if the pragmatic dimension of performing or understanding illocutionary acts is itself to be thematized by its objectification as a referential object of the semantical (i.e., representative) dimension of propositions of a metalanguage and thus by ascribing truth-values to performative sentences that, among other things, have the function of a self-referential exposition of truth-claims. Since this latter move constitutes the semanti^ation-strategy of the prevailing types of "formal pragmatics" (from R. Carnap and R. Martin to R. Montague and D. Lewis), it is not surprising that Y. Bar-Hillel in his review " O n Habermas' Hermeneutic Philosophy of Language," Synthese 23 (1973), should only prove his nonunderstanding of the whole approach. Cf. also K.-O. Apel, "Zwei paradigmatische Antworten auf die Frage nach der Logos-Auszeichnung der menschlichen Sprache," in Kulturu>issenscha/ten, ed. H . Liitzeler (Bonn: Bouvier, 1980), 13-68. 17. In fact natural language could be called the (transcendental-pragmatic) m e t a institution with respect to all human institutions. Cf. my argument with A . Gehlen's "philosophy of institutions," in K . - O . Apel, Transformation der P h i b s o p h i e (Frankfurt a.M.: Suhrkamp, 1973), vol. 1, 197C; and my argument with J. R. Searle in my essay "Sprechakttheorie und transzendentale Sprachpragmatik zur Frage ethischer Normen," in Sprachpragmatik und Phibsophie, ed. K . - O . Apel (Frankfurt a . M . : Suhrkamp, 1976), esp. 104ff. 18. Cf. K . - O . Apel, "Types of Rationality Today," in Rationality Today, ed. T . Geraets (Ottawa: Ottawa University Press, 1978), 307-40. 19. Cf. I. Kant ("Nachlass," ed. Preussische Akademie, no. 2128): "Das objektive Kriterium der Wahrheit ist die Ubereinstimmung der Vorstellungen untereinander nach allgemeinen Gesetzen des Verstandes und der Vernunft, d . i . durch Anschauungen oder Begriffe"; "Das subjektive Kriterium der Wahrheit ist die Ubereinstimmung eines Urteils mit anderen sowohl in demselben Subjekt als in verschiedenen." 20. Cf. D . Davidson, "Truth and Meaning," Synthese 17 (1967); "True to the Facts," Journal of Philosophy 66 (1969); "In Defense of Convention T , " in T r u t h , Syntax, and M o d a l i t y , ed. H . LeBlanc (Amsterdam: North Holland, 1973). 21. E. Tugendhat, Vorlesungen zur Ein/uhrung in die s p r a c h - a n a l y t i s c h e P h i b s o p h i e (Frankfurt a . M . : Suhrkamp, 1976). 22. Cf. e.g., W . Sellars, Science, P e r c e p t i o n , and Reality (London: Routledge, Chapman and Hall, 1977), 219. 23. N . Rescher, The Coherence Theory of T r u t h (Oxford: Clarendon Press, 1973). In fact, Rescher does not propose the coherence of theory as an explication of the meaning of truth, but only as a theory that provides an "authorizing," but not "guaranteeing" criterion of truth. 24. L. B. Puntel, W a h r h e i t s t h e o r i e n in der N e u e r e n P h i b s o p h i e (Darmstadt: Wissenschaftliche Buchgemeinschaft, 1978). 25. W . Kamlah and P. Lorenzen, Logische P r o p d d e u t i k (Mannheim: Bibliographisches Institut, 1967). s, 26. K. Lorenz, "Der dialogische Wahrheitsbegriff," Neue He/te fur P h i b s o p h i e 2 , no. 3 (1972): 111-23. 27. See n. 9, above. 28. C f . , e.g., C . F. von Weizsacker, Die Einheit der N a t u r (Munich: Hanser, 1972), 336ff.; and K . Lorenz, Die Ruekseite des Spiegels, V e r s u c h einer Naturgeschichte mens c h l i c h e n E r k e n n e n s (Munich and Zurich: Piper, 1973).

204

C S. P E I R C E A N D P O S T - T A R S K I A N

TRUTH

29. Puntel, W a h r h e i t s t h e o r i e n . 30. Ibid., 215f. 31. This thesis of modern coherence-theory was impressively stated by O . Neurath in connection with the paradoxicality of Wittgenstein's talk about the relationship between language and reality in the T r a c t a t u s and the search of the logical positivists for prelinguistic "protocol-sentences." Cf. O . Neurath, "Soziologie im Physicalismus," Erkenntnis 2 (1931): 393-431, esp. 396f. and 403. 32. Cf. G . W . F. Hegel, P h d n o m e n o l o g i e des Geistes, A , I: "Die sinnliche Gewissheit, das Dieses und das Meinen." 33. The example was used by M . Schlick in his argument with Neurath in "Facts and Propositions," A n a l y s i s 2 (1934/35): 65-70. I believe Schlick was right in considering his "Konstatierungen" to be "the points of contact between knowledge and reality" (cf. "Uber das Fundament der Erkenntnis," Erkenntnis 4 [1934]: 98), but, like Husserl, he was not yet prepared to mediate the phenomenological distinction of perceptual judgments with the acknowledgment of their being nonetheless language-impregnated (and hence theory-impregnated) and hence open to revision by reinterpretation. It is precisely this mediation between a phenomenological evidence-theory of truth and "fallibilism" in the sense of indefiniteness of interpretation that became possible by Peirce's semiotical account for the interaction of icons, indices, and symbols in perceptual judgments as limit-cases of hypotheses. 34. Cf. H . Spiegelberg, "Husserl's and Peirce's Phenomenologies: Coincidence or Interaction, " P h i l o s o p h i c a l and Phenomenological Research 17 (1956): 164-85. Regarding the internal tension between Peirce's "phenomenology" or "phaneroscopy" after 1902 and his semiotical logic of symbolic mediation (in the sense of "Thirdness"), cf. my account in Der D e n k w e g v o n C. S. Peirce, 203ff. 35. Thus far also E. Tugendhat (Vorlesungen zur Einfuhrung, 161ff., esp. 182ff. and 205ff.) goes too far, it seems to me, in superseding the phenomenological paradigm of philosophy by that of language-analysis. 36. Cf. K. Popper, Logic of Scientific D i s c o v e r y (London: Hutchinson, 1959), 95ff., 105. Popper's logicism in this context comes close to Neurath's position and amounts to a complete surrender of the phenomenon of experimental evidence to the realm of psychological explanation. I would like to suggest that, under this epistemological presupposition, the idea of falsification finally can no longer be interpreted as wrecking of theories by reality, since, by abstraction from experiential evidence as an authorizing, if not sufficient, truth-criterion, the confrontation of theories with falsifiers amounts to a confrontation between rival theories which then can no longer defend their priority-claim byamong other criteriarecourse to experiential support. The result of this version of pluralism seems to become visible in Feyerabend's dissolution of critical rationalism. 37. Cf. Apel, Die E r k l d r e r d V e r s t e h e n - K o n l r o v e r s e . 38. Cf., e.g., H . Albert, T r a k t a t uber Kritische Vemunft (Tubingen: Mohr, 1968), 14, 30. 39. From 1866 Peirce outlined his semiotical transformation of Kant's "transcendental idealism" in passages like the following: "We find that every judgment is subject to a condition of consistency: its elements must be capable of being brought to a unity. This consistent unity since it belongs to all our judgments may be said to belong to us. Or rather since it belongs to the judgments of all mankind, we may be said to belong to it." (Fragment of 1866, quoted in M . G . Murphey, The D e v e l o p m e n t of Peirce's Philosophy (Cambridge, Mass.: Harvard University Press, 1961], 89.) Later this germinal transcendence of the methodical solipsism of a transcendental philosophy of the consciousness qua "I think" is continued as follows: "consciousness is a vague term . . . consciousness is sometimes used to signify the I think, or unity in thought; but the unity is nothing but consistency, or the recognition of it. C o n -

C. S

PEIRCE A N D POST-TARSKIAN TRUTH

205

40. 41.

42. 43. 44.

45. 46. 47. 48. 49. 50.

sistency belongs to every sign, so far as it is a sign . . . there is no element whatever of man's consciousness which has not something corresponding to it in the word . . . the word or sign which man uses is the man himself . . . the organism is only an instrument to thought. But the identity of a man consists in the consistency of what he does and thinks. . . . The existence of thought now depends on what is to be hereafter; so that it has a potential existence, dependent on the future thought of the community." (Collected Papers, 5.313-16) See n. 9, above. For the different possible petspectives of Peirce interpretation, see my Der Denkweg von Charles S. Peirce, 33f. For Habermas's precautions against the term "transcendental pragmatics," cf. his essay "Was heiBt Universalpragmatik?" in Sprachpragmatik und Philosophie, ed. K . - O . Apel (Frankfurt a . M . : Suhrkamp, 1976), esp. 201fF. (also in J. Habermas, Vorstudien und Ergdnzungen zur Theorie des kommunikativen Handelns [Frankfurt a . M . : Suhrkamp, 1984]; all page references are to the volume edited by Apel). See Habermas, "Was heiBt Universalpragmatik?," 219. Ibid., 239f. This initial move is followed by B. Stroud's famous paper on "Transcendental Arguments" (journal of Philosophy 61 [1968]: 241-56) which, as far as I can see, was prominent in discouraging the recent discussion of a post-Wittgensteinian version of transcendental philosophy. Cf. K . - O . Apel, "The Problem of Philosophical Fundamental Grounding in Light of a Transcendental Pragmatic of Language," Man and World 8 no. 3 (1975): 239-75. Cf. Habermas, "Was heiBt Universalpragmatik?," 213f. Cf. O . Hoffe, "Kritische Uberlegungen zur Konsenstheorie der Wahrheit (Habermas)," Philosophisches Jahrbuch 83 (1976): 313-32. Thus Puntel (Wohrheitstheorien) almost claims Habermas's theory as argumentative evidence for his coherence-conception, Cf. Habermas, "Was heiBt Universalpragmatik?," 218f. Ibid., 218. Habermas's rigorous distinction between "objects of experience" and "facts" as correlates of propositions seems to be to presuppose a certain nominalism (and a nominalistic understanding of transcendental philosophy) which fits in with orthodox Kantianism and certain strands of modern logic of language, but cannot cope with Peirce's semiotical transformation of the transcendental logic of cognition which I called "sense-critical realism" (cf. Der Denkvueg von C . S. Peirce, 41ff.) In his Knowledge and Human Interests (Boston: Beacon Press, 1971, vol. 2, 5) Habermas found it necessary to refute Peirce's semiotical realism as a relapse into pre-Kantian metaphysics, whereas I would consider it to be an integral part of transcendental semiotics and hence transcendental pragmatics which overcomes the paradoxicalities of Kant's conception of unknowable things in themselves. Cf. also Apel, Der Denkweg von C . S. Peirce, 257 (n. 41) and 34If. (n. 217).

51. Habermas, "Was heiBt Universalpragmatik?," 217f. 52. Ibid., 241if. Here Habermas's discourse-theory of truth in fact seems to collapse into a pure coherence-theory. 53. Cf. C . S. Peirce, " O n a New List of Categories," Proceedings of the American Academy of Arts and Sciences 7 (1967): 187-98. For supplementary materials and a critical interpretation of this early period of Peirce's thought, cf. Murphey, The Development of Peirce's Phihsophy, 55ff. For my own interpretation, which sometimes deviates from Murphey's, see Der Denkwg von C . S. Peirce. 54- Cf. Murphey, The Development of Peirce's Philosophy, 296ff. 55. Cf. Collected Papers 5.358-87. Cf. my interpretation in Der Denkweg von C. S. Peirce, 115ff.

206

C . S. P E I R C E A N D P O S T - T A R S K I A N T R U T H

56. See Collected Papers 5.115ft"., 5.142, 5.181ff., 1.194- Cf. my interpretation in Der Denkweg v o n C. S. Peirce, 297ff. 57. Cf. my interpretation in Der D e n k w e g v o n C. S. Peirce, 73ff.; and "From Kant to Peirce: The Semiotic Transformation of Transcendental Logic," in Kant's T h e o r y of K n o w l e d g e , ed. L. W . Beck (Dordrecht, Holland: Reidel, 1974), 23-37. 58. Cf. Der Denkweg von C . S. Peirce, 310-318. 59. Cf. my argument for a "transcendental language game" of hermeneutic understanding as against the post-Wittgensteinian pluralism-relativism in "The Communication Community as the Transcendental Presupposition for the Social Sciences," i n Apel, T o w a r d s a Transformation of Philosophy (London: Routledge, 1980), 136-179. 60. Cf. josiah Royce, The Problem of Christianity (New York, 1913), vol. 2, 146, and my interpretation in Trans/ormation der P h i l o s o p h i e , vol. 2, 199ff. 61. It becomes explicit when scientific progress faces a crisis as the continuum of a philosophical explication of scientific concepts and an empiric-hermeneutic reconstruction of the history of science. 62. See Habermas, "Was heifk Universalpragmatik?," 24463. Ibid., 245. 64. Ibid., 249f. 65. Ibid., 250. 66. C . S. Peirce, Collected Papers, 1.4; 1.413; 1.402, nn. 19 and 20. For the history of (evolution of) science, cf. 6.10, 1.47, 1.173, 1.441, 1.498, 1.186, 1.191, 1.603, 1.118. Cf. also my interpretation in Der D e n k w e g v o n C. S. Peirce, 196ff., 268ff. 67. Cf. my critical comments in Der Denkweg v o n C. S. Peirce, 164fF., 349ff.; cf. also my "Scientism or Transcendental Hermeneutics?," in Apel, T o w a r d s a T r a n s f o r m a t i o n , sec. 3. 68. Cf. e.g. Collected Papers 1.231, 1.317, 1.473; for the context and background of Peirce's anti-individualism, cf. G . Wartenberg, Logischer Soziaksmus. Die Transformation der Kanrischen Transcendentalphilosophie d u r c h C. S. Peirce (Frankfurt a . M . : Suhrkamp 1971), esp. 14ff. 69. Habermas, "Was heifk Universalpragmatik?," 255. 70. Ibid., 216. 71. For this term, cf. P. Lorenzen, N o r m a t i v e L o g i c and Ethics (Mannheim/Zurich: Bibliographisches Institut, 1969), 73ff.

SEVEN Transcendental Semiotics and Hypothetical Metaphysics of Evolution

EXPOSITION

With the two key terms in the title of my paper I want to point to a problem that, on the one hand, is characteristic of the post-Kantian architectonics of Peirce's philosophy and that, on the other, relates to an architectonic aporia of post-Kantian philosophy in general, especially that of the present era. This problem may be very provisionally put as follows: Is it possible to answet the ttanscendental question as to the conditions of the possibility of valid knowledge in such a way that, at the same time, the teal process of teseatch and hence of striving for and eventually reaching valid knowledge may be conceived as a continuation of natural evolution? In what follows I will first try to point out, in a rather summary way, why in Kantian and, even more so, in post-Kantian philosophy ftom Getman idealism up to the so-called evolutionist epistemology of our day this problem repeatedly had to arise without finding any satisfactory solution. T o this end it will be especially important to show that, many relevant achievements notwithstanding, both the speculative philosophy of German idealism (i.e., that of Schelling and Hegel) and empiricist extrapolations of the biological theory of evolution (such as the so-called evolutionist epistemology) were doomed to miss the solution of out problem for similar reasons and that both forms of philosophy lead back to a pre-Kantian type of "dogmatical" metaphysics. Having pointed to this chronic inability of post-Kantian philosophy to solve a recurring problem, I will try to show that a solutionor at least a promising solution sttategymay be provided by a Peircean or, perhaps, a quasi-Peircean approach. By a Peircean or quasi-Peircean approach I mean a (ttanscendentalsemiotic) transformation of the Kantian type of (mentalistic) ttanscendental philosophy. Such a transformation would preserve the normative-methodological priority of ttanscendental philosophy with regard to all types of empitical research and at the same time give free space and even guidance to an empirico-hypothetic type of metaphysics which may explain both the ontological difference and continuity of natural evolution and human historyin particular that part of human history which consists in valid knowledge about natural evolution and human history. 207

208

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

T o this extent, my paper follows systematic goals rather than seeking to contribute to Peirce exegesis, although 1 wish to acknowledge Peircean inspirations as far as possible. But 1 w i l l take the liberty of reconstructing Peirce's thoughts, so to speak, from afar, since 1 am at present neither as familiar with his work as 1 was twenty years ago nor so confident now as 1 was then about the provable coherence of his writings.
1

HISTORICAL P R E L U D E : T H E APORETICS OF A POST-KANTIAN

PROBLEM

In Kant's philosophy the critical rejection of so-called dogmatic metaphysics is internally connected with the question as to the conditions of the possibility of objectively valid knowledge w h i c h are supposed to be discoverable (traceable) as the legislative functions of the transcendental subject of knowledge. In this way the critical thrust of Kant's philosophy remains bound to the paradigmatical installation of the subject-object relation of knowledge w h i c h had already been established i n a still ontological-metaphysical form by Descartes' distinction between the res cogitans and the res externa. From the critical point of view of Kantianism, any k i n d of metaphysical ontology, that is, any attempt to comprehend being as a whole (hence also the m i n d as part of being), without reflecting on the transcendental subject-object relation as a precondition of the possible objectivity of being, had to be rejected as "dogmatic metaphysics." A t the same time, the transcendental subject-object relation underlying Kant's philosophy also grounded a new line of demarcation and division of labor, as it were, between philosophy and the empirical sciences. For the Critique of P u r e Reason made clear that f r o m n o w o n theoretical p h i l o s o p h y had to restrict itself to investigating the c o n d i t i o n s of the possibility of experience, and therefore had to leave the business of experiential research itself to the empirical sciences. In this connection it must be noted that for Kant the relationship of transcendental philosophy to N e w t o n i a n physics constituted the paradigmatic application of the demarcating function of the transcendental subject-object relation. N o w i n the era of post-Kantian philosophy it turned out that as soon as the relationship of man to nature as presupposed i n classical N e w t o n i a n physics (mechanics) was replaced or transcended by another type of human experience, it was just as difficult to ignore the critical, demarcating function of the transcendental subject-object relation set out by K a n t as it was to live with it. T h e crux of these difficulties had already manifested itself at the center of the architectonics of Kant's system. For, i n order to answer his o w n final question, " W h a t is m a n ? , " K a n t had to introduce a threefold distinction i n the pertinent concept of the Self. First, he had to distinguish between the transcendental Self, w h i c h belonged to the transcendental function of the "I t h i n k " of the transcendental subject or consciousness, and the empirical Self ascertainable in introspection, w h i c h was to correspond to man as a possible object of empirical science. Here there was

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

209

already an unreconcilable dualism between the concept of the legislative and teleological function of the transcendental Self and the concept of man as an object of empirical science, hence as causally determined as any object of natural science, that is, of N e w t o n i a n physics. H o w , then, could it be possible to conceive human beings as free subjects of teleological actions as presupposed by the "ought" of K a n t i a n ethics. A t this point, K a n t introduced the concept of an intelligible Self as the subject of possible autonomous action. In doing this, he e x p l o i t e d w i t h i n the framework of his transcendental d i a l e c t i c h i s fundamental, critical distinction between knowable objects as " p h e n o m e n a " and unknowable "things-inthemselves"; and he exploited this distinction in such a special way (as a solution of the " t h i r d antinomy") that one may suspect that the whole distinct i o n between n o u m e n a and phenomena was intrinsically motivated by Kant's concern to solve the problem of man's freedom as a prerequisite of morality i n the face of the presumed causal determinism of nature.
7

But Kant's dualistic solution, even if it were tenable according to the criteria set by his o w n critical philosophy, could not provide the basis for a philosophy or even an empirical science of human history and culture as a product of responsible actions. For such actions simply cannot have any place i n the realm of possible experience as this is portrayed i n Kant's C r i t i q u e of P u r e R e a s o n . In his writings o n the philosophy of history and, more systematically, i n the C r i t i q u e of j u d g m e n t , K a n t sought to solve this problem by reintroducing i n a way a possible relationship between man and nature different from that of the subject-object relation presupposed i n N e w t o n i a n physics. H e reintroduced nature as structured by purpose; for example, the purposiveness of organic life or of artistic genius, or e v e n i n his writings o n the philosophy of historyas subject to developmental strategies that, i n the long run, could lead to a convergence of our postulates of moral, or, respectively, political progress, w i t h the natural, nomological course of history. Here K a n t comes closest to a conception of nature i n relation to man's history that could be presupposed i n a natural philosophy of evolution as this was already suggesting itself in the second half of the eighteenth century. However, K a n t had to reintroduce this quasi-Aristotelian conception of nature o n the transcendental preconditions of the subject-object relation as this had been grounded i n the first C r i t i q u e . T h i s meant that he could not concede a teleology of nature as "constitutive" for the possible objectivity of experience but merely as a "regulative p r i n c i p l e " of heuristic value for the faculty of "reflective judgment" i n ascertaining the possible coherence of natural processes and their specific laws. A s compared w i t h the "objective" processes of nature and of human history, all teleological structures of organic life and even of human actions w i t h i n the frame of historical experience can only have the status of "as i f " constructions. A s a consequence of this K a n t i a n device, for a l l post-Kantian philosophies of natural evolution and/or human history, only the following choice between

210

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

alternative options appeared to be open: either a full-fledged teleological understanding of human actions and history, and possibly even of natural evolution as the pre-history of human history, or the supplementation of Newtonian physics by a causal-mechanical explanation of natural evolution and eventually of human history or even of human actions in history. As is well known, all these options and suboptions have in fact been tried. But, as far as I can see, the paradigmatic conceptions of evolution and history up to now have not been able to provide a satisfactory solution of the problem Kant left behind. I will outline in the following only the most famous examples of this aporetic situation.
* *

One great paradigm of a post-Kantian answer to our problem was provided by that of the German idealists, in particular, by Schelling and Hegel. This idealistic response consequently developed all those concepts and speculative assumptions which one needs in order to understand the genesis of the world of nature and history from above, so to speak, that is, from the teleological point of view of a finally accomplished self-consciousness of the transcendental Self. Most important in this respect was Schelling's conception of an original spontaneity or even creative freedom of nature qua n a t u r a n a t u r a n s as prior and superior to the nomological determination of nature qua n a t u r a n a t u r a t a , an idea bound up with that of a genetic development from unconscious or preconscious drives through growing consciousness of feelings and finally self-consciousness and moral self-determination. This latter movement was understood in Hegel's philosophy of history as a return of the idea from exteriorization and alienation, a dialectical movement which continued on in, and was "aufgeh o b e n " by, human history. However, this whole speculative response of an idealistic-teleological understanding of nature and history f r o m above had to presuppose from the outset some kind of ontological identity between the objects and the subject of transcendental consciousness. For this reason it was compelled to annul the transcendental subject-object relation which Kant had established as the basis for the critique of dogmatic metaphysics and for the modern demarcation between transcendental philosophy and empirical science. In short, the price paid for the great conceptual achievements of German idealism, considered as a first "metaphysics of evolution," was a return to dogmatic metaphysics through disregard of that modern idea of empirical science for which Kant had set out to articulate the conditions under which it was possible. Two particular remarks on Hegel may supplement this very global assessment: In his critique of Kant's distinction between mere phenomena and unknowable things-in-themselves, Hegel came very close to what I would call a meaningcritical metacritique of a transcendental critique of knowledge, that is, to a metacritique of Kant's transcendental critique of knowledge, according to which Kant's enterprise is incoherent because it necessarily presupposes, qua theoretical activity, that it can obtain knowledge, yet sets up a criterion of what can be

SEMIOTICS A N D METAPHYSICS O F EVOLUTION

21 1

known which entails the negation of this presupposition. But with his own claim to an absolute "knowledge of knowledge" as the anticipated result and presupposition of the reconstruction of the social and natural world, Hegel failed to preserve the good sense of the idea of a critical epistemology that leaves the completion of material knowledge to the indefinite progress of the empirical sciences. He tried, as it were, to anticipate the result of the dialogue of an indefinite community of researchers in the monologue of his speculative dialectics of substance as the subject of self-consciousness. Corresponding to this disregard of the dimension of possible scientific progress, there is another consequence of Hegel's taking the absolute standpoint of an ex post facto comprehension of the world's development: He could no longer account for the need of a future-oriented engagement of human praxis and hence for the very notion of an ethical "ought" different from comprehension of the "actual as reasonable" and the "reasonable as actual." It is in opposition to this characteristic feature of Hegel's philosophy that Marxism, pragmatism, and, with regard to personal life, Kierkegaardian existentialism also arose as philosophies of a future-oriented mediation of theory and praxis; and it is this same problem as a problem of the future continuation of the world's evolution that Peirce had in mind when he spoke of Hegel as a thinker who tried to look upon the world as a completed fact, neglecting thereby the crucial dimension of the "esse in future."
2

But here it is also important to realize that the projection of the Hegelian idea of a teleological determinism and the corresponding necessity of the course of history into the future, such as is characteristic of orthodox Marxism, only increases and aggravates the ethical aporia of a dogmatical metaphysics of the world's teleological evolution. This point has been made clear by Karl Popper's vigorous denunciation of "ethical futurism" or "historicism," and in our day it has become almost a commonplace through the French postmodernists' dismissal of the great metanarratives of history set up by the rnaitre-periseurs of the nineteenth century. Thus it has become obvious that an adequate conception of the continuity between the world's natural evolution and the autonomy of the Self and its capacity for responsible teleological action cannot be made plausible simply by supplanting the notion of causal mechanism by that of a teleological determination either of the whole process of evolution or of history.
3

But what about the other option suggested by Kant's dualism: supplementation of Newtonian physics by a mechanistic explanation of biological evolution and eventually even of human behavior at the level of cultural evolution? A powerful movement in this direction was initiated, as is well known, by Charles Darwin's theory of the descent of humankind and its numerous continuations and sociocultural extrapolations up to the present day, from Spencer's social Darwinism to sociobiology and evolutionary epistemology. Could this approach explain the transition or, more cautiously formulated, the compatibility between

2 12

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

the realm of mechanical necessity and the tealm of freedom and normative self-determinationa realm we must presuppose as a ttanscendental ptecondition of the activities of human science? Befote attempting to answet this question, I must fitst emphasize that I do not wish to deny the explanatory power and ffuitfulness of Darwinian and postDarwinian theories of evolution considered as theoties of natutal scienceas little, by the way, as I wish to deny the persistent heutistic televance of certain ideas of G e r m a n idealist philosophy of nature, such as the ptesupposition of an original spontaneity as prior and superior to causal determination. It even appears to be very suggestive to speculatively identify this metaphysical idea of spontaneity w i t h the Darwinist idea of chance-variation and, in particular, the neo-Darwinist idea of undeteimined mutations. (I shall come back to this later.) However, the ctucial problem w h i c h Darwinist ot neo-Darwinist theoties present for a philosophical understanding of the coherence, or at least consistence, between the natutalistic theoties of evolution and the ttanscendentalidealistic ptesuppositions of the Self/ego as a subject of motal and scientific activities concerns the intrinsic claim of the D a r w i n i a n and post-Darwinian theories to reduce the possible justification of all normative validity-claims raised by human subjects of responsible action and of scientific teseatch to one mote form of selective adaptation, of struggle for survival of the fittest. It is indeed true that certain features of this Darwinian key n o t i o n are i n a sense applicable to the phenomena of cultural evolution, including even scientific inquiry. Thus, processes of trial and error or of falsification as analyzed by Karl Popper, for example, can be understood as a process of selection of theoties ot hypotheses; moreover, all cteative innovations of cultural e v o l u t i o n might be tegatded, i n analogy to biological mutations, as subject to a struggle fot life that is decided by a selection of those innovations w h i c h ate attractive i n themselves or as parts of clusters, as Dawkins suggests. Finally, even categorial schemes of scientific t h i n k i n g , w h i c h are a priori valid according to K a n t and still argued to be such in contemporary transcendental arguments and ptotophysicseven these a priori presuppositions of our factual cognition may be considered as phylogenetic results of successful adaptation, i n ptinciple explainable through the empitical investigations of evolutionary epistemology.
4

A s possible as a l l this is, however, thete still remains the truth-claim or claim to intetsubjective validity as a transcendental presupposition of criticizable argument and thus of all scientific enterprises, evolutionary epistemology i n cluded. A t least this c l a i m and its implications, w h i c h do not include any categotial schemes of cognition but only certain rules of rational argumentation, cannot be explained away; it is not possible to construe a valid truth-claim as a successful, an invalid one as an unsuccessful, evolutionary adaptation, at least not in any sense of these latter words w h i c h could be operationalized by an empitical science of evolution. A t any rate, one cannot put such a construal upon one's o w n truth-claim without thereby falling into a performative selfcontradiction w h i c h annuls one's whole atgument. But, i n fact, this cannot be

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

2 13

done by anybody else either, given the equivalence of truth and universal intersubjective validity, w h i c h I think is a transcendental presupposition of serious argument more deeply rooted i n principle than any a priori presupposition of a categorial scheme of experience. But if we grant this irreducible transcendental presupposition of argument, then once again in order to avoid performative self-conttadiction we must also grant as undeniable at least some fundamental ethical norms of an ideal community of argumentation/arguers. Thus, these notms also cannot be justified or refuted by recourse to biological selection in the sense of the survival of the fittest. T h i s is all the more interesting as, having made the transition from biological to sociocultural evolution, we find ourselves at the level of selecting attractive innovations i n Dawkins's sense, and hence we stand mote and more under the practical necessity of having to adapt nature, our environment, to those aims of ours w h i c h we can justify by an ethics of responsibility. H o w , then, should we apply the evolutionary theory of adaptation to those ethical norms we must rely upon i n determining how we shall adapt nature to our needs? I once put this question to Steven T o u l m i n , who was advocating a theory of sociocultural evolution by adaptation. A f t e r some pressure, his answer to the problem was the following piece of advice: " W a i t for a hundred years, then you w i l l see what has happened!" Now, this I think is an answer w h i c h Hegel could also have given, since he suggested that our appeals to an "ought" would always come too late w i t h regard to the well-understood course of history. Yet I would suggest that the structure of the practical problem is just the inverse: T h e answer of a Hegelian historicist as well as that of a theoretician of evolutionary adaptation to the problem of the possible justification of a validity-claim, be it a claim to truth or to ethical Tightness, must always come too late, even if it should be possible to prove from an ex post facto petspective that what is true and what is fight is also what is vitally useful, and hence even somehow successful in the long r u n . T h i s antireductionist argument even applies to the claims to meaningfulness of our statements and arguments w h i c h must precede all our validity-claims. For it is not possible, even i n principle, w i t h i n the argumentative discourse of science or philosophy to explicate satisfactorily problematic meanings i n terms of those practical uses w h i c h have been already selected w i t h i n the frame of successful forms of life. (I shall come back to this quasi-Wittgensteinian suggestion w h e n discussing Peirce's "pragmatical m a x i m " of meaning-explication.)
5

Thus, at this point our comparison of how speculative idealism and naturalistic empiricist theories of evolution have responded to the problem of dualism inherited from Kant leads to a strange but interesting result. A c t u a l l y , it is n o t just idealist metaphysics of identity w h i c h has overthrown Kant's demarcation between empirical science and the possible justification of their objective validity by (transcendental) philosophy, thereby restoring dogmatical metaphysics. T h e same has occurred, I suggest, wherever the transcendental subject-object relation of K a n t i a n philosophy has been cast overboard by a scientistic,

214

SEMIOTICS A N D METAPHYSICS O F E V O L U T I O N

naturalist reductionism which has absolutized the side of empirist objectivity. (I think, in fact, that in recent times the impact of those types of dogmatical metaphysics which derive ftom scientistic reductionism is even greatet and more impeding for a critical philosophy than the open effusions of speculative idealism. Fot the latter may have a heuristic value, wheteas the former lead to an obscuring or even to an appatent dissolution of the very problems of a philosophical justification of scientific and ethically relevant validity-claims.) But how, then, should we deal with the aporetics of the Kantian dualism? Should we retain Kant's transcendental subject-object relation as a basis for critically demarcating between (transcendental) philosophy on the one hand and empitical science on the othet, hence also between the tealm of normatively otiented ftee actions and the tealm of natural processes detetmined by causal law? Should we give up the attempt to undetstand natural evolution as prehistory of human history? Should we renounce the whole idea of conceiving out own futute-otiented mediation of theory and praxis as a continuation of natutal evolution and of history? I think that with these questions the stage is set fot an examination of the Peitcean approach to out problem, that is, fot a reconstruction from afar of that architectonic constellation of apptoaches which may be indicated by the tetms transcendental semiotics and metaphysics of e v o l u t i o n .

THE

P E I R C E A N C O N S T E L L A T I O N O F T R A N S C E N D E N T A L SEMIOTICS A N D METAPHYSICS OF EVOLUTION

The promise held out by this constellation of tetms may be outlined in advance in the following theses. 1. The tetm t r a n s c e n d e n t a l semiotics may be taken as connoting the Peircean program of a transformation of Kant's "ttanscendental logic," a program that, as it were, replaces the Kantian concept of the ttanscendental subject of knowledge by that of the indefinite community of sign-interpretation. Thus the (prelinguistic) subject-object relation of Kantian epistemology is transformed by the complementary telation of intetsubjective communication and discourse critique in such a way that the Kantian dimension of the ttanscendental a priori becomes accessible fot the fallibilism and meliorism of the ptocesses of synthetic inferences and linguistic interpretations; at the same time the ttanscendental function of semiosis, that is, of sign-mediated wotld-interpretation, is preserved as a normative foundation of the aims and the long-run validity of the synthetic inferences by regulative ideas.
6

2. Because the long-run validity of the synthetic inferences is no longer bound to, or dependent upon, the constitutive functions of the categories as a priori principles of the understanding, it is no longer necessary for a transcendental-semiotic logic of inquiry to presuppose a dualistic-metaphysical distinction between objects of experience as mere "phenomena" whose form is definitively prescribed by the mind and "things-in-themselves" which are completely un-

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

215

knowable yet must be postulated i n transcendental logic as causally affecting the m i n d . T h e real qua final object of sign-interpretation may be conceived instead as that w h i c h can never be factually k n o w n as an immediate object of representation but nonetheless as that w h i c h is knowable in the long run. It is that w h i c h i n every confirmation or falsification of out hypotheses functions as a constant correlate, so to speak, of our synthetic infetences, whose long-term validity is warranted by the necessary postulates and regulative principles of the normative semiotic logic of inquiry. T o this extent, the normative-methodological function of transcendental semiotics remains an a priori precondition fot all empirical sciences, even if the factual processes of inference and signinterpretation, whose notmative validity conditions are grounded by transcendental semiotics, can at the same time be tegarded as continuations of biological processes of trial and error i n the sense of evolution theory. 3. T h i s last point leads to a Peircean conception of a metaphysics of evolution and its relation to transcendental s e m i o t i c s . In contrast to G e r m a n idealism, from w h i c h it derives central ideas, the Peircean conception of metaphysics is not a restoration of pre-Kantian "dogmatical metaphysics," since it does not abolish ot disregard the line of demarcation between, and the complementarity of, transcendental philosophy and empirical science. U n l i k e the metaphysics of G e r m a n idealism, it does not usurp the a priori grounding functions of transcendental philosophy, but, being itself an empirical discipline of global hypotheses ("coenoscopy"), it subjects itself to the principles of fallibilism and empirical testing grounded by transcendental semiotics qua logic of inquiry. However, this methodological subordination to the normative principles of transcendental semiotics does not prevent the metaphysics of evolution from introducing ontological concepts that allow one both to thematize the processes of natural evolution as well as those of human inquiry into, and interpretation of, natural evolution, and to conceive of the latter as a continuation of the former. A n d since both the hypothetical status of the metaphysics of evolution and the procedures through w h i c h its scientific consequences might be tested are ensured by the normative methodology implied in transcendental semiotics, the metaphysics of evolution may take the risk of going far beyond the paradigmatical suppositions of " n o r m a l science" by speculative designs that are at the same time bold and conceptually vague. T h i s is indeed characteristic of Peirce's cosmogonic and evolutionist speculations, particularly i n the early 1890s.
7

T h i s m u c h serves as a first outline of the novel and promising features of the Peircean approach to our problem. Let us now take a closer look at the two conceptions thus introduced and their internal relationship.
* * *

By "transcendental semiotics" in its most comprehensive sense I understand a conception or program that was first introduced by Peirce i n his early K a n t studies, w h i c h terminated i n the " N e w List of Categories" of the late 1860s and in his "theory of c o g n i t i o n " and " i n q u i r y " of the early 1870s. T h e famous papers

216

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

of 1877 and 1878 containing the "belief-doubt theory" and the "ptagmatic m a x i m " for clarifying meaning belong in this context as does the theory of (the three classes of) inference and the meaning-critical theory of reality; also to be included ate all the improvements to the theory of categories achieved w i t h the aid of the logic of relations and the logic of quantifiers as well as those papets in w h i c h the tetm "semiotic" (ot "semeiotic") is explicitly introduced and the concept divided into thtee b r a n c h e s s p e c u l a t i v e g r a m m a r , critical logic, and speculative rhetoricwhich were later complemented by a very differentiated classification of signs. I think that most of these papets can be shown to pursue more or less closely the guiding idea of a transformation of Kant's "ttanscendental l o g i c " of cognition along the lines of a transcendental s e m i o t i c . There are, however, some problems w i t h this conception, if one tries to ground or verify it by appeal to the whole body of Peirce's texts o n semiotic, especially those of the latet period. Fot according to Peirce's o w n classification of the sciences i n the eatly yeats of the twentieth century, only part of his late semiotics can be classified as "semiotic l o g i c , " w h i c h latter now constitutes the first patt of notmative science (presupposing the othet notmative sciences of ethics and aesthetics). T h e othet patts of Peirce's late semiotical writings ate to be regarded as belonging rathet to his new projects of (hypothetic) metaphysics (of evolution) and of phenomenology. But even this disttibution leads to difficulties since m u c h of his wtiting o n semiotics f o t example, his many sign-classifications and especially the pertinent suggestions in his lettets to Lady W e l b y s e e m rathet to belong to a specific program of a general semiotics that cuts across the rubrics of logic, metaphysics, and phenomenology.
8

In what follows I shall of course focus o n the relationship of normative semiotic logic and hypothetical metaphysics; but precisely i n order to characterize this relationship, it is necessary to point to some difficulties in the discussion of Peirce's semiotics i n general. In otdet to expose and illuminate the transcendental aspect of Peirce's semiotics, one needs to make clear at least the following points. 1. E v e n Peirce's general definition of the sign-function as a triadic relation whose relata ate the sign itself, the object fot w h i c h it stands and the interpreter (or respectively the interpretant) to w h o m it stands is oriented toward the cognitive interest of a semiotic logic qua epistemology. It does not appeat to pay special attention to those sign-functions ot functions of signifying w h i c h cannot be reduced to "standing for" (or "designating" or "denoting") some object of c o g n i t i o n f u n c t i o n s that have been analyzed by Wittgenstein, Chatles Morris, and the speech-act theory. Douglas Greenley in his book Peirce's C o n c e p t of Sign has tried to correct this feature from the point of view of a truly general s e m i o t i c , but it seems clear that we can neglect this point since we are interestedas was Peirce p r i m a r i l y i n exploring the conditions of the possibility of sign-mediated cognition, that is, i n transcendental semiotics. T h i s point
9 10

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

217

leads us immediately to some other difficulties w i t h Peirce's definition of the sign-function. 2. It has often been noticed and also criticized that Peirce in his triadic definitions of signhood or s e m i o s i s seems to waver between a vetsion that speaks of a person, an "interpretet" or a " m i n d " to w h o m the sign stands and addresses itself, and a n o t h e r m o r e characteristicversion that uses the term "interpretant" to denote the third relatum of the triadic definition. W i t h this difference seems to go hand i n hand the question of whether the sign-function is definitely triadic i n the sense that a sign is related to a real object, w h i c h is not itself a sign, and a real interpreter, who is not a sign either; or whether the triadic sign relation implies an infinite regress, so to speak, in both directions since the immediate object for w h i c h the sign stands is itself a sign standing for anothet immediate object that is a sign and so o n , ad infinitum; and since the interpretant to w h i c h the sign stands is also a sign that is to be interpreted and so o n , ad infinitum. Thus, one could t h i n k the sign-functions only in mediating the continuous (temporal) transfer between a previous sign and a subsequent sign.
1 1

Here, I can only give a short answer to this intricate problem: B o t h versions of understanding Peirce's definition of the sign-function are justified, and precisely this fact points to the diffetence and interrelation to be postulated between transcendental semiotics and empirical semiotics, metaphysical semiotics i n cluded. O n the one hand, the transcendental aspect of semiotics as a logic of cognition implies indeed that thete are real correlates of the triadic signrelation; fot only this definitely triadic relation makes it possible to speak of sign-mediated cognition; and it is precisely this conception of sign-mediated cognition, that is, world-interpretation, that makes it possible and necessary for the real, as much as it is independent of all individual subjective opinions, to be conceived as the " k n o w a b l e , " that is, as the possible object of the ultimate o p i n i o n of the real sign-interpreters as members of an indefinite community. Looked upon from this viewpoint of a transcendental semiotics, the supposition that there are only signs as possible relata of the sign-relation can only be called s e m i o t i c i s m , a regression to a magical conflation of signhood and r e a l i t y . (This position seems to be suggested by Derrida rather than by P e i r c e . )
12 13 14

O n the other hand, however, it is a postulate ptecisely of the transcendentalsemiotic conception of the triadic structure of sign-mediated interpretation of the teal that the empirical process of semiosis, that is, of sign-mediated reference to the teal object as well as of sign-interpretation, must be potentially infinite, even though it takes place between really existing interpretets and really existing things w h i c h may be encountered by interpreters as "brute facts" resisting their wills (cf. Peirce's many allusions to the relation between the I and the non-I as an illustration of the category "secondness"). Furthermore, even the solitary thinker's understanding of h i m - or herself must be understood as a triadic relation of sign-interpretation. T h i s motive also has, I suggest, two faces, depending o n whether it is considered as part of
15

218

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

transcendental semiotics or as parr of metaphysics (or a general semiotic). From the lattet petspective Peiice appeals to play down the specific featutes of human communication i n favoi of the absttact geneial structure of semiosis that "petfuses" the universe. From the first perspective, h o w e v e r w h i c h was worked out in great detail i n the eatly theory of cognition and inquiry and latet led to Peitce's anticipation of the theory of speech-actsthe point of this puzzling idea is tathet that even solitary t h i n k i n g must be conceived of as an internalized mode (1 would even say a deficient mode) of the genuine c o m m u n i c a t i o n between real members of the c o m m u n i t y .
16

3. Related to the preceding there is w i t h i n Peirce's semiotic a further problem that may be dissolved, I suggest, by distinguishing between the ttanscendentalnormative and the geneial empitical aspect of semiotics w h i c h includes the metaphysical one. T h i s problem concerns the question of w h e t h e i Peitce's semiotic ptesupposes the patadigmatic priority of language or w h e t h e r i n conttadistinction to analytic philosophy of languageit considets language merely as a very special class of sign-functions ot, changing the peispective, as a late stage of the sign-function i n the process of evolution. T h e answer to this question would seem to be the following: If one focuses on Peitce's semiotic logic of cognition and inquiry, it seems fairly clear that fot h i m also there is a paradigmatic ptiority of language, since only language can provide that k i n d of sign-interpretation, hence of sign-mediated cognition of the real, w h i c h is presupposed by science and by philosophical semiotic itself. T h e methodological otientation towatd language even leads Peiice to anticipate speech-act t h e o r y a l t h o u g h it is once again chatacteiistic of his restricted cognitive interest that he does not equally deal w i t h a l l kinds of speech act, hence w i t h interpersonal c o m m u n i c a t i o n of all k i n d but primarily with assertive acts as vehicles of atgumentative discouise seiving the seatch f o i truth.
17

O n the other hand, it is ttue that Peiice wishes to analyze the sign-function ot semiosis in such an elementary and generalized way that it might cover all kinds of natural signs and sign-functions as well and even allow an evolutionary explanation of the genesis of language w h i c h would disclose this latter to be a continuation of that ptelinguistic process of semiosis w h i c h , as Peiice once put it, "perfuses" the whole u n i v e t s e . T h i s program of course poses in particulatly shatp form the question as to the relationship between ttanscendental philosophy and a metaphysics of evolution.
18

4. Finally, it must be emphasized that an adequate handling of out problem requires us to distinguish the normative dimension of Peirce's semiotic from its descriptive dimensions. T h i s is especially important w i t h tegatd to both the dimension of sign-interpretation and Peirce's differentiation of different types of "interpretants." It seems clear that a transcendental semiotic qua logic of inquiry cannot primarily be interested i n finding out what interpretants constitute the factual, ot even conventionally normal, effects of our understanding signs. T h i s might be televant for the psychologist or the linguist; the concern of a semiotic logician aiming at a normative methodology of science, however, must be to

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

219

determine what k i n d of interpretants can be regarded as results of a correct sign-interpretation. T h i s holds precisely in those cases where the conventional rules of interpretation provided by everyday speech are not sufficient for a scientifically relevant explication of the meaning of concepts. T h i s was, for example, the case w i t h Einstein, who looked for a more adequate explication of the meaning of the physically relevant concept of simultaneity and, mote generally, of space and time, yet it was also the case w i t h John Rawls, w h o looked fot a deeper explication of the concept of justice. 1 think that at this point the "pragmatic m a x i m " fot the clarification of concepts coincides w i t h the semiotic conception of the "logical interpretant" or again the "ultimate logical interpretant." A n d it seems clear that it is not the desctiption of those factual habits w h i c h might coincidentally terminate the process of sign-interpretation but rather the postulate of a habit that would correspond to that ultimate consensus o n truth w h i c h alone may figure as the "ultimate logical interpretant" of a given sign in this c o n t e x t . T h i s notmative dimension of Peirce's conception of the "interpretant" distinguishes his idea of meaning remarkably from those of Charles Morris and Wittgenstein, this notwithstanding the fact that all three try to explicate meaning somehow i n terms of habits and practices of b e h a v i o r . In virtue of this normative dimension, then, the late conception of " p r a g m a t i s m " o r alternatively, " p r a g m a t i c i s m " as a normative science may be considered a reconfirmation and a more precise determination of that program of a semiotic transformation of K a n t i a n ttanscendental logic w h i c h was initiated as early as i n the " N e w List of Categories."
19 20

There is indeed one serious difficulty w i t h regard to the idea of a ttanscendental semiotic that is raised by Peirce's late classification of the sciences. In this classification, w h i c h proceeds according to the C o m t e a n principle that more' concrete sciences presuppose the more abstract sciences, Peirce withdraws the task of founding the (three fundamental) categories from semiotic logic and gives it o v e r s o it appearsto the new, extrasemiotical discipline of a "phenomenology" ot "phaneroscopy." T h i s is indeed a hard challenge for those who think, as I do myself, that a transcendental semiotic logic could and should provide a substitute and equivalent fot Kant's "metaphysical" (i.e., fotmal logical) and "transcendental deduction" of the categories. T h i s was i n fact suggested by Peirce himself i n his papers of the 1860s and even after the foundation of the logic of relations and the logic of cognition and inquiry after 1870.
21

O n this latter account, the metaphysical deduction of the three fundamental categoriesfirstness, secondness, and thirdnessis provided by Peirce i n his "logic of relations" w h i c h , as he tries to show, involves precisely three irreducible concepts of relation: the monad, the dyad, a n d a s a basis for all further deductionsthe triad as identical w i t h the structure of c o m b i n a t i o n ; the transcendental deduction of the categories, o n the other hand, is provided by the structure of semiosis, or again representation, w h i c h i n a semiotically transformed "theory of c o g n i t i o n " replaces Kant's reduction of the manifold of
22

220

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

impressions to the unity of consciousness through the ttanscendental "synthesis of apperception": It involves firstness, that is, suchness of the wotld's qualities, through the iconic functions of the predicates i n perceptual judgments; secondness through the use of demonstrative pronouns as indices w h i c h i n our perceptual judgments draw the attention of the I to the qualities of the non-I as immediate object of the factual encountet; and finally thirdness through the judgment itself w h i c h as a semiotical interpretation of the "percept" provides a mediation of the factual givenness, that is, firstness and secondness, through the conceptual generality of the predicate. If one takes into account that for Peirce even the thitdness of a perceptual judgment is already the result of an unconscious abductive inference and that the truth of semiotic wotld-representation can only be defined i n terms of the ideal limit of all pertinent processes of inference and sign-interpretation, one might even conclude that the "highest p o i n t " i n Peirce's "transcendental deduct i o n , " w h i c h corresponds to Kant's "synthesis of apperception," can only be the final synthesis of sign-interpretation that would or should be reached by the joint venture of an infinite community of researchers. T h u s , the highest point of the transcendental deduction w i t h i n a ttanscendental semiotics would only be a "regulative idea" i n the K a n t i a n sense. But since this idea of the "ultimate o p i n i o n " or consensus of the community is a necessary postulate of semiotic logic without w h i c h inquiry into ttuth as an enterprise of argumentative discourse makes no sense, one may still speak even i n this case of a transcendental deduction of the categories and of the validity of the three classes of inferences, abduction, induction, and deduction, w h i c h correspond to the three fundamental categoties. If, however, we conceive of the categories i n this way, what then should we say about Peitce's apparent surrender of the task of founding the categoties to an extrasemiotic phenomenology? Jiirgen v o n K e m p s k i and Murray Murphey have concluded that this surrender constitutes a breakdown of the "architectonics" of Peirce's semiotic logic of cognition and inquiry; and the fotmet has advanced the thesis that Peirce's recourse to a phenomenology, w h i c h is contemporary w i t h Hussetl's project of the same name, shows that he, like Husserl, had never understood the function of the "synthesis of apperception" as that of the "highest p o i n t " of Kant's ttanscendental deduction of the categories. T h i s sounds very plausible from the point of view of the presemiotic (i.e., prelinguistic) paradigm of transcendental philosophy, fot w h i c h the meaning and function of the categoties may be grounded i n the necessary synthesis of ideas w i t h i n a transcendental consciousness. But I would rather suggest that Peirce has, i n fact, provided an equivalent to the "transcendental d e d u c t i o n " w i t h i n the framework of a transcendental semiotic. Fot w i t h i n this framework the meaning and function of a l l c o n c e p t s and thus of the categories alsomay be transcendentally grounded i n the structure of semiosis, that is, in the necessary postulate of an ultimate synthesis of sign-interpretation (i.e., of an ultimate consensus of the community of sign-interpreters). A n d this ttanscendental grounding could find its comple23

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

22 1

mentation and reconfirmation i n the fotmal logic of relations w h i c h takes the place of Kant's logical "table of judgments." Perhaps i n his latet yeats Peiice failed to teflect thoroughly on this suggestive structure of his early transformat i o n of Kant's transcendental logic. A t any rate, 1 would pursue this heuristic stiategy as a fruitful philosophical project, even if i n view of his latet turn to a phenomenological explication of the categoties it should turn out to be an overinterpretation of Peitce's early intentions. (This would be my answet to O e h l e r and P a p e , who cannot find i n Peitce's wotk anything like a ttanscendental semiotics of the sott I have suggested.)
24

It might, howevet, even be possible to deal w i t h the problem raised by Peirce's late "classification of the sciences" by making this latter compatible w i t h the idea of a ttanscendental semiotics. A f t e r a l l , Peirce's phenomenology is not a cognitive foundational discipline at all but only an elucidative ot illustrative one: a discipline of intuitive evidence regarding the firstness of possible (imaginable) structutes w h i c h is, so to speak, thrust in between the two parts of logic w h i c h originally served as the basis for the "metaphysical" and the "transcendental deduction" of the categoties, namely, the formal logic of relations and the semiotic logic of cognition and inquiry respectively. E v e n i n Peirce's late classification of the sciences phenomenology still presupposes the deduction of the categoties by the so-called mathematics of logic, w h i c h is nothing other than the former "logic of relations." T h e late Peirce in his classification of the sciences seems to t h i n k that "mathematics of logic" as well as "phenomenology" could be conceived as ptectitical disciplines based o n human competences w h i c h cannot and need not be g r o u n d e d . But it seems cleat that if the formal-abstract tesults of the "mathematics of logic" and the illustrative results of the "phenomenology" are to be linguistically interpreted as televant presuppositions of philosophical knowledge, they must be somehow subotdinated to the normative principles of true knowledge as these are grounded by the semiotic logic (of cognition and inquiry). (Here Peirce simply seems not to reflect on what he himself is doing and claiming i n his o w n classification of the sciences. But this deficit i n reflection is quite c o m m o n i n philosophy, even i n transcendental philosophy, for already K a n t displayed it i n failing to enquire into the conditions of the possibility of the validity of his own ttanscendental arguments.)
25

In any case, most important for our problem is the fact that Peirce even, indeed precisely, in his late classification of the sciences maintained the normative claims of the semiotic logic of inquiry w i t h regard to a l l empirical sciences, including the metaphysics of evolution w h i c h he had elaborated i n the decade befote. H e even renewed and emphasized these claims in his critique of Dewey's "genetic" conception of l o g i c s and his declating pragmatism to be "the normative s c i e n c e . " It therefote seems to me highly important to interpret carefully the relationship between Peirce's metaphysics (of evolution) and his normative (semiotic) logic of inquiry from the point of view that he himself developed only in his late classification of the sciences.
26 27

222

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

A s far as I can see this is usually not done, for those who are interested i n Peirce's metaphysics of evolution as a speculative anticipation of the problems of our day often do not pay heed to Peirce's preceding transformation of Kant's transcendental logic. Hence, they deal w i t h Peirce metaphysics as if it were meant to cover the whole of Peirce's problems, for example, the whole of his so-called theory of reality and truth. I suggest that this architectonic simplification must lead to precisely those aporias of reductionism w h i c h I tried to expose i n the preceding. Peirce's metaphysics of evolution, then, is understood either according to the model of G e r m a n idealism or just as a speculative version of the modern theory of evolution. In either case such an interpretation amounts to a reductionist e l i m i n a t i o n of all normative problems concerning the justification of validity-claims and aims of reasoning for the sake of w h i c h Peirce himself introduced his conception of pragmatism as the normative science or, later, of "pragmaticism." In the closing part of my paper I w i l l suggest at least some features of an alternative interpretation, i n particular w i t h regard to the theory of reality and truth (and its ethical presuppositions). I must admit from the outset that in what follows I w i l l often have to rely rather o n determined reconstruction than o n a close interpretation of the various, sometimes bewildering lines of Peirce's self-interpretation. I take the following thesis as my point of departure: A l l those explications or definitions of the meaning of reality or of truth in terms of what would be the ultimate opinion or belief of an infinite community of researchers reached i n an infinite process of inquiry, that is, of inferences and sign-interpretations, under appropriate conditions all these fundamental determinations w h i c h constitute the core of normative pragmaticism, do not have the status of metaphysical hypotheses; nor do they presuppose such hypotheses, since they fulfill the function of transcendental-logical presuppositions of all possible metaphysical and scientific hypotheses, and i n particular, of the abductive and fallible inference character of such hypotheses. Thus, if it is to make sense for us to enter into argumentative discourse, we must presuppose that it is possible i n principle (i.e., under ideal conditions) to reach an ultimate consensus w h i c h would be identical w i t h what we mean by truth. A n d this we must have already done w h e n discussing, for example, the problem of whether some presuppositions are fallible hypotheses. W e cannot understand what this means (and rightly postulate that all hypotheses, being synthetic inferences, must be fallible) without already presupposing the validity of certain pragmaticist explications of the meaning of reality and truth and, for that matter, of the structure of argumentative discourse. H e n c e we cannot in fact suppose, as many claim we can, that even these presuppositions are fallible metaphysical hypotheses, for this would mean that it would be possible to falsify them by simultaneously presupposing them. O f course, I am aware that w i t h these statements I am expressing a radicalized philosophical reflection o n those features of Peirce's pragmaticism w h i c h can be

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

223

understood as a result of the semiotic transformation of Kant's transcendental logic. The transcendental character of these features is easily overlooked because Peirce himself never realized it clearly and in particular because it is indeed only a transformed residuum of the Kantian transcendentalism. For it no longer implies any synthetic a priori principles that would correspond to categotial schemes of experience. In so doing, as already intimated, it makes the realm of a priori presuppositions of cognition susceptible to empirical correction something which accords well both with the demands of those holistical-minded philosophers of our day who have challenged the idea of "transcendental arguments" by questioning "the very idea of a categorial scheme"; and with those suggestions ftom the history of science and evolutionary epistemology to the effect that the Kantian tenets on categorial schemes and a priori forms of intuition are to be explained as restricted in their validity to a realm of human adaptation to nature that might be called "mesocosmos" in distinction to the realms of the "macrocosmos" and the "microcosmos" that have become the subject of postclassical physics.
28

(Peirce's distinction between "instinctual" and certain vague "commonsense" concepts, on the one hand, and precise but highly fallible hypotheses of advanced stages of science on the other shows him to be one of the first to anticipate these ideas.)
29

Nevertheless, Peirce's semiotic transformation of Kant's transcendental logic consists precisely in his replacing Kant's constitutive apriorism by certain postulates and regulative ideas which ground a priori the aim and long-run validity of those synthetic inferences whose finite results can only be fallible conclusions. The twist of this grounding is easily overlooked and lies, 1 suggest, in reflection on those presuppositions which render the whole project of re*search, nay even of philosophical argument about it, meaningful. (This is the reason why in my book on Peirce I tried to interpret Peirce's semiotic transformation of Kant's transcendental logic as a "meaning-critical" transformation of the transcendental critique of cognition which parallels the [post-linguistic turn] standards of present philosophy.)
30

I think indeed that a great deal of the notorious difficulties and confusions about Peirce's most fundamental ideas may be dissolved by considering his pertinent suggestions as meaning-critical versions of transcendental arguments. This applies, for example, to those versions of Peirce's explications of his postulate of the "ultimate opinion" where he makes clear that we cannot avoid making this postulate because any meaningful hypotheses must presuppose it, whereas no meaningful hypothesis can possibly refute it. Here it is important to appreciate thatcertain utterances of Peirce's notwithstandingthe definition of "reality" in terms of what would be the object of the ultimate opinion of the indefinite community of investigators does not mean that the existence of the realor the real as the existing universeis made dependent on the successful outcome of the process of cognition, or that the existence of the real has to be proved independently because the successful outcome of the cognition-process is

224

SEMIOTICS A N D METAPHYSICS OF EVOLUTION

u n c e r t a i n . For the existence of the real and, for that mattet, of a real argumentation community, themselves belong to those ptesuppositions of atguing and of inquiry w h i c h cannot be denied i n meaningful atgument. (The Cartesian dream argument, e.g., is self-defeating, as can be shown in vatious ways.)
31

N o t even the reality of the real must be made dependent o n the future fact of its being completely k n o w n as some quasi-nominalistic statements of the early Peirce could suggest. For there cannot be such a fact, as the late Peitce realized when he came to recognize that counterfactual postulates ate "regulative ideas," that is to say, cases of thirdness w h i c h cannot, in principle, be reduced to secondness. But the reality of the real must indeed be undetstood a priori as something that is i n ptinciple " k n o w a b l e . " O t father, its meaning is identical w i t h that of the "knowable ( i n the long r u n ) , " and thus it must indeed be identified w i t h what, according to a teal possibility, would be the object of the ultimate o p i n i o n of the indefinite community because this is what we must mean in older to be able to distinguish between the teal and all possible objects of idiosyncfatic opinions of individuals ot of finite communities.
32

(The structure of this problem is partly analogous to the famous one of the "haidness of the d i a m o n d . " Fot as Peiice only made definitively cleat i n his late papers o n "Pragmaticism," this property of a real thing cannot be made dependent o n the fact of its being sctatched or not being scratched by somebody either, although it cannot be understood without supposing the real possibility of its proving resistant to being scratched. There is of course also a difference between this paradigmatic application of the "pragmatic m a x i m " and the more fundamental paradigm of understanding the meaning of reality. It is constituted by the fact that the hardness of the diamond can be counterfactually explicatedalthough never c o m p l e t e l y i n terms of possible facts of human actions, manipulations, and subsequent experiences, whereas in the case of the reality of the real even this is impossible. O n l y that almost irrational aspect of the real, its brute external resistance to the w i l l , can be explicated i n terms of a possible fact of experience. T h i s is the reason why we cannot identify the real individual object of our true and false hypotheses by means of complete conceptual descriptions, as Leibniz and the Hegelians postulated, but only by using demonstrative pronouns or "indices.") So m u c h by way of defending my thesis that the most fundamental tenets of Peirce's theory of truth and reality do not have the status of metaphysical hypotheses since they belong to transcendental semiotics. I cannot al