Logic, Language and Method - On Polarities in Human Experience: Philosophical Papers

Kuno Lorenz Logic, Language and Method − On Polarities in Human Experience

Logic, Language and Method − On Polarities in Human Experience Philosophical Papers

by

Kuno Lorenz

De Gruyter

The publication of this book was funded by the Laboratoire d’Histoire des Sciences et de Philosophie (CNRS), Archives Henri-Poncare´, Nancy, and by the Department of Philosophy at the Universität des Saarlandes, Saarbrücken.

ISBN 978-3-11-020312-7 e-ISBN 978-3-11-021679-0 Library of Congress Cataloging-in-Publication Data A CIP catalogue record for this book is available from the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de. 쑔 2010 Walter de Gruyter GmbH & Co. KG, Berlin/New York Printing: Hubert & Co. GmbH & Co. KG, Göttingen ⬁ Printed on acid-free paper Printed in Germany www.degruyter.com

Contents Preface ......................................................................................................VII Part I Philosophical Logic and Philosophy of Language 1. Rules versus Theorems A new approach for mediation between intuitionistic and two-valued logic (1973)............................................. 3 2. On the Relation between the Partition of a Whole into Parts and the Attribution of Properties to an Object (1977)................................. 20 3. Basic Objectives of Dialogic Logic in Historical Perspective (2001) ................................................................................................... 33 4. Pragmatic and Semiotic Prerequisites for Predication. A dialogue model (2005) ..................................................................... 42 5. Pragmatics and Semiotics: The Peircean Version of Ontology and Epistemology (1994) ..................................................................... 56 6. Intentionality and its Language-Dependency (1985) ........................... 62 7. Meaning Postulates and Rules of Argumentation. Remarks concerning the pragmatic tie between meaning (of terms) and truth (of propositions) (1987) ........................................................ 71 8. What Do Language Games Measure? (1989) ...................................... 81 9. Features of Indian Logic (2008)........................................................... 92 Part II Methods in Philosophy, in Art, and in Science 1. The Concept of Science. Some remarks on the methodological issue ‘construction’ versus ‘description’ in the philosophy of science (1979) .................................................. 109 2. Is and Ought Revisited (1987) ........................................................... 124

VIII

Contents

3. Competition and Cooperation: Are They Antagonistic or Complementary? (1994)..................................................................... 140 4. Another Version of Methodological Dualism (1997) ........................ 148 5. The Pre-Established Harmony Between the Two Adams (1993) ...... 162 6. On the Way to Conceptual and Perceptual Knowledge (1993).......... 171 7. Self and Other: Remarks on Human Nature and Human Culture (2002) ................................................................................................. 186 8. On the Concept of Symmetry (2005) ................................................. 198 9. Procedural Principles of the Erlangen School. On the interrelation between the principles of method, of dialogue, and of reason (2008) .......................................................................... 207 Acknowledgments ................................................................................... 219 Index........................................................................................................ 223

Preface It is delightful to see a representative sample of one’s papers on various subjects collected in a volume in one’s own lifetime, let alone when it is a collection of papers written in a language that is not one’s mother tongue. I am particularly glad that with the present volume which accompanies two volumes of papers in German – the one preceding and the other succeeding this publication – readers who read English but not German and wish to become acquainted with my treatment of current philosophical issues will have a chance to uncover ‘family resemblances’ that relate the papers with one another. In this process the key term ‘polarity’ in the subtitle may guide the reader’s attention. Of course, as the papers were written in a period extending over more than thirty years and owe their origin to very divergent occasions, partly being commissioned and partly offered, they exhibit different stages of awareness with respect to both the procedures applied and the terminology used. If there are any inconsistencies, they may be explained by developmental change. I have deliberately refrained from trying to produce up-todate versions as ‘final’ ones. This includes retaining occasional overlap at places where maintaining self-containedness makes it necessary. Work in progress should not hide the signs that it is in need of further improvement or refinement. The reader who wants to know how the latest and most advanced version up to now looks like, is advised to turn to the last paper (II.9) and read it first. In addition he will also find there a short exposition of how my treatment of issues as Dialogical Philosophy is related to the tenets of the Erlangen School and its philosophical ancestry. In each of the two parts the papers are divided into two blocks of four each by an additional paper that occupies the central position. Whereas within the four blocks chronological order is maintained, the two bisecting papers (I.5, II.5) are singled out, because they treat Peircean and Leibnizian issues, respectively, and may thus, through views of two philosopherscientists of the past, be particularly well suited to visualize the two main interests that underly the argumentations in the papers of the first and the second part. These interests are indicated by the two subtitles, one referring to questions of logic and language, and the other to questions of method. Notwithstanding that division, in some way each paper touches at least one of the countless oppositions that permeate the ways of coping with what

VIII

Preface

happens and has happened around us and with us. Be it Doing and Suffering, Subject and Object, Truth and Meaning, Body and Soul, Particular and General, Practical and Theoretical, Knowing-how and Knowing-that, Thought and Action, and a multitude of others. Thus, procedural questions are not out of purview in the first part, and substantial questions do enter the papers of the second part. The reader will often encounter direct investigations into some of the oppositions, some others being touched alongside (I.1, I.2, I.3, I.7, I.9, II.1, II.2, II.3, II.4, II.6), and he equally should cast an eye on how the use of oppositions in the course of arguing interrelates with their status of being an object of argumentation as well. Looking back it appears now that in a way most of the papers may basically be considered as different steps en route of reducing the conceptual oppositions that permeate content and set-up of the papers to the fundamental one of looking at actions as ‘objects’ and as ‘a means’ in statu operandi, i.e., ‘from a distance’ and ‘while engaged in doing it’. This is true, already, of my billet d'entrée into professional philosophy: the dissertation on arithmetic and logic as dialogue games under the supervision of my late teacher Paul Lorenzen. This way of looking, I come to argue eventually, arises due to a basic anthropological feature: Humans are organized dialogically, as I and You. When A does something – A in the role of I – B (being identical with A in but a special case) ‘sees’ A doing that, B in the role of You. This simultaneity of two dialogical roles extends to the level of sign actions: When, in the role of I, A, besides e.g. speaking, ‘says’ something, then B, in the role of You, besides e.g. hearing A, ‘understands’ something (I.4, II.7, II.8). Philosophers are particularly prone to the danger of presenting their thoughts as if they were elected to represent mankind each for herself alone in telling how things are ‘really’ and often tend to forget about their dependence on others, if not their fellow philosophers. Usually, in science and philosophy we are accustomed to treat issues from a distance. This is in tune with considering the very verbalization as, already, a kind of detachment from its subject matter: words and things fall apart. On the other hand, we are well aware of how verbal activity anywhere and without restriction is the most prominent medium of being engaged with something: thinking and doing coincide. But when activities themselves, plain or verbal, are articulated as in papers like the present ones, a stage is reached where matters become quite complicated. What is at stake, is the relation of one’s individual activity, both plain and verbal, with the activity of others and with the particular objectives of activities that are supposed to be intersubjectively identifiable (I.6, I.8). It is not surprising, therefore, that many of the conceptual oppositions nourish competing philosophical doctrines and have done that since their inception with the consequence that scientific theories that avail them-

Preface

IX

selves of such doctrines as their conceptual frame inherit competitiveness with respect to each other. Sometimes this happens to such a degree that it becomes difficult to compare such theories adequately, because a common language of comparison is itself a matter of dispute. As an example, think of how in the last decades the behavioral sciences have lost ground to the cognitive sciences while there is no clear account of the reasons why this has happened. Even looking for reasons has become more difficult as their conceptual distinction from causes seems to evaporate. This again may be understood as an offshot of the gradual rise of biology as the paradigm of science and the decline of physics in that role. This shift has fuelled anew the age-old attempts to set up a unified account of the social and the natural sciences aiming to treat cultural features as a more advanced state of natural features. Language and reasoning, for example, are thus considered to be a more developed form of animal communication.The boundaries between the mental and the physical have become permeable, so much so, that now one often finds philosophy of mind being treated as just a version of neurobiology. Theory-building is encroaching upon more and more territory formerly held by mostly non-verbal and often plain activities. This trend is influential now, for example, in education, but in the arts as well. Theorizing gradually supplants practical expertise by having corpora of theoretical knowledge at one’s disposal that purports to rule over the practice in question. In ever more fields ‘knowing-that’ acquires priority to ‘knowing-how’ seemingly in line with the classical heritage that ars (τέχνη) is inferior to scientia (ἐπιστήµη). However, in antiquity it was justification that characterized ‘epistemic’ knowledge, a ‘knowing-why’, in contrast to a ‘knowing-that’, a merely true opinion. Only in modern times up to the present true opinion is held to be sufficient, if it serves successful guidance of individual and common practice. Practical knowledge or ‘knowing-how’ as represented by skills of various kinds and relegated to the sphere of mere ‘technic’ knowledge (artes) in antiquity, is still now, in a world quite different from its ancient predecessor, waiting to be acknowledged as on a par with theoretical knowledge rather than being treated as dependent on theories that explain how it works. The growing presence of theories of any kind, or, more generally, of theorizing in the loose sense of speaking about something rather than experiencing it in practice, is the result of a well-known modern development. We encounter a process where practical experience recedes from reach in more and more cases for ever more people and where growing world-wide communication induces apparent knowledge of a rapidly increasing number of matters beyond any chance of checking it individually. Thus, in the world we live, it becomes difficult to realize the mutual

X

Preface

dependence of verbal and plain activities and likewise of semiotics and pragmatics, if the two kinds of activities are turned into objects of investigation. It is a world of particulars, subdivided into various categories depending on the structure of the language used (a fertile source of oppositions!), such that dealing with them is increasingly supported by numerous artificial devices, and especially persons including theorists play roles that disappear when speaking about them instead of with them. The difficulty of discerning verbal and plain activity in order to get an idea of their mutual dependency is aggravated even further, because the artificial devices that support plain activity take on various sign functions by themselves; just consider the case of doing something with the aid of computers even outside their use as a means of communication, if I may disregard the mere metaphor of ‘communicating with computers’. Philosophical investigations should make these features visible. They can do this by means of their reflexive nature that distinguishes them from scientific investigations, particularly with regard to the question of how the procedures applied relate to the subject matter treated. If I should have succeeded in bringing this to the fore at least in some of the papers – and the reader is the only one who can judge it – the present volume will have served its purpose. I should add, this publication would not have been possible without the efforts of quite a number of people to whom I am deeply grateful. Particularly, I want to thank my friend and colleague Professor Jürgen Mittelstraß (Universität Konstanz) and my friends and former students, Professor Gerhard Heinzmann (Université Nancy II) and Dr. Bernd Michael Scherer (Haus der Kulturen der Welt/House of World Cultures, Berlin) who encouraged me not to give up the whole plan. They took pains to raise financial support without which a book of this kind would be an unbearable risk to publish under present economic conditions. For generous financial aid I am indebted to my department at the Universität des Saarlandes as well as to the Centre National de la Recherche Scientifique Lorraine (CNRS, UMR 7117) which supported the present publication at Walter de Gruyter Publishing House. Special thanks go to Dr. Gertrud Grünkorn at de Gruyter’s who did everything to edit this volume in a form that shows all the features for which books published at de Gruyter’s are justly wellknown everywhere.

Part I Philosophical Logic and Philosophy of Language Philosophical Logic and Philosophy of Language

Rules versus Theorems A new approach for mediation between intuitionistic and two-valued logic I Contemporary critics of two-valued logic concentrate on the reasons for

accepting the tertium non datur A ∨ ¬A as a valid propositional schema. Brouwer explicitly states1 that only by unjustified extrapolation of logical principles from those which correctly describe the general relations among propositions on finite domains to those that allegedly regulate propositions on infinite domains, could it happen that A ∨ ¬A is accepted as valid. He was the first to observe that value-definite (decidably true or false) propositions do not generally transfer value-definiteness to their logical compounds. No better support could be found for the claim that the classical characterization of propositions as entities that are either true or false is inadequate. The union of the class of all true propositions and the class of all false propositions does not contain all logical compounds out of either true or false propositions; it does not contain, for example, certain as yet neither proven nor disproven universal propositions of elementary arithmetic. But nobody has seriously advanced the thesis that such propositions should not count as propositions at all.2 In fact, it is generally conceded that the usual way to form finite and infinite logical compounds makes sense even if nothing can be said about their truth-value. It is obligatory, then, to look for a better introduction of the term ‘proposition’ than the classical one and, of course, not only a syntactical introduction, which is trivial, but a semantical one. The validity concept of two-valued logic being dependent on the value-definiteness of propositions will consequently have to be given up and replaced by a concept of validity that works without recourse to the truth-value of the propositions in question. It follows that the classical introduction of logical particles by the (finite or infinite) truth table method has to be given up as well, or, rather, it has to ____________ 1 2

Brouwer 1925, p. 2. Skolem’s proposal of a strictly finite mathematics without any use of quantifiers is an exception; its radical implications would deserve special discussion, cf. Skolem 1923.

4

Philosophical Logic and Philosophy of Language

be amended in such a way that the general definition of logical particles restricted to value-definite propositions yields the well-known classical ones. Various attempts in this direction have been made, most prominent, among others, the proof-theoretic interpretation of intuitionistic logic by Kolmogorov and the operationist interpretation by Lorenzen.3 In these attempts, the introduction of the junctor ‘if- then’ marks the starting point of a deviation from classical procedures. Kolmogorov replaces the concept of truth by the concept of provability and the provability of A → B correspondingly by the provability of B relative to a proof of A that is hypothetically assumed. The concept of proof and a fortiori of provability has to be taken over from existing unformalized mathematics. Lorenzen replaces the concept of truth by the concept of (generalized) derivability within some calculus K, such that derivability of A → B has to be read as admissibility of the corresponding rule α ⇒ β (with A K α and B K β) relative to K. Here, the concept of admissibility has to be accepted as intuitively clear. Actual difficulties of interpretation occur in both cases after iterating the logical composition, e.g., with ‘if-then’, and no way out is visible if propositions other than mathematical ones are candidates for logical composition. Yet, these attemps have cleared the way to the additional insight that not only is value-definiteness not hereditary generally to logical compounds, but that it is possible to ascertain the (non-logical) truth of logically compound propositions without recourse to the truth-value of the subpropositions. There are, for example, true (i.e., provable resp. derivable) and not logically true subjunctions A → B without any knowledge even about the value-definiteness of A or B: one may take a suitably chosen admissible rule α ⇒ β within an undecidable calculus. A successful criticism of two-valued logic has to be able, therefore, to balance a wider concept of proposition with a correspondingly wider concept of logical composition and to add an adequate concept of validity for propositional schemata, or, alternatively, an adequate concept of (logical) implication by keeping the meta-equivalence ‘A implies B’ if and only if ‘(A → B) ε valid’. The usual method of devising formal systems, i.e., of introducing a syntactical concept of validity for syntactically defined well formed formulas is, of course, insufficient. Heyting's formalization of intuitionistic logic did permit precise comparison with other calculi, calculi

____________ 3

Cf. Kolmogorov 1932, and Lorenzen 1955. Gödel has shown that an axiomatization of the concept ‘beweisbar’ (provable) within classical logic, somewhat different from Kolmogorov’s nonformalized version, can be used as a representation of intuitionistic logic, cf. Gödel 1933.

Rules versus Theorems

5

of classical logic as well as of modal logic and others,4 but it could not answer the initial question of what kind of theory or rather ‘action’ (Denktätigkeit) 5 it actually is that is formalized by this or that formal system of logic. And intuitionism (in the spirit of Brouwer) consequently never claimed to be able to represent its logic fully by a formal system. In order now to gain a better understanding of the actual conflict, it is necessary to go beyond these introductory remarks by stressing a difference of points of view between the proponents of classical logic and the proponents of effective logic, which is so much taken for granted that it is hardly ever explicitly disputed. Since the time of Leibniz, classical logic is often referred to as a system of ‘truths’ which hold ‘universally’, in ‘all possible worlds’, and, therefore, independently of the special facts of the ‘actual world’, i.e., of the natural sciences. And mathematics is, following the programme of logicism, to be constructed as a special part of this system of logical truths. In precisely this sense logic, and with it mathematics, came to be considered as a system of tautologies without factual content. Classical logic is the formal frame for any scientific investigation, the a priori basis of empirical science. It makes no essential difference when Quine, expanding ideas of C. I. Lewis, disputes the distinction ‘a priori-empirical’ and stresses the uniformity of the whole corpus of scientific truths instead. 6 This corpus is not uniquely determined by observational facts, it is in need of conceptual and other theoretical constructions, e.g., mathematical ones, which are chosen by intrinsic criteria of perspicuity, economy, connectedness et alii of the system of science as a whole. Thus, it may well obtain that even logic, one of the central parts of the corpus of truths, has to be changed due to new observational facts in order to satisfy the aforementioned criteria. Yet, even then, logic, the system of accepted logical truths, be it in its formalized version derivable by a classical or by some other calculus, remains the formal frame of science. Logic may be called ‘relatively universal’, i.e., a system of universal truths relative to the actual state of science. On the other hand, effective logic (explicitly in its operationist interpretation), must be looked at as a system of ‘universal’ rules that are accepted whenever a system of rules of action, e.g., rules for producing proofs or rules for producing arbitrary strings of signs, has been laid down. In this case, the field of application for the rules of logic is not the world as the totality of facts, but rather the world as seen in terms of specific kinds of scientific human activities. Within mathematics, for example, the rules of ____________ 4 5 6

Cf. the review of the main results in Kleene 1952, § 81; especially important is the paper McKinsey/Tarski 1948. Cf. Heyting 1930, pp. 45–46. Cf., e.g., the paper Two dogmas of empiricism in: Quine 1953, pp. 20–46.

6

Philosophical Logic and Philosophy of Language

effective logic may be used without restriction. And this obtains because mathematics is not viewed as a system of truths, even less logical ones, but is treated as an independent scientific activity which, together with its intrinsic rules, may use the rules of logic as additional ‘admissible’ ones. Effective logic is the material content of any scientific investigation, the ‘empirical core’ even within mathematics. It goes without saying that no uniqueness claim is added. The set of rules of effective logic may vary from one scientific activity to another, and are not even strictly determined by any one of those. It coincides even with the set of rules of classical logic in the case of strictly finite mathematics, as Brouwer explicitly observed.7 Thus, logic may again be called ‘relatively universal’, i.e., a system of accepted universal rules relative to the field of investigation. In the light of these considerations, the basic conflict is a question rather of the set-up of formal logic itself than of accepting this or that propositional schema as valid. And, indeed, the competing views, to treat logic either as a set of rules (‘for correct thinking’) or as a set of theorems (‘on the general behavior of thought’), trace back to the very beginning of formal logic, to Aristotle and his interpretation by posterity. The conflict is known under the rubric: logic – art or science?, the respective Greek terms being ‘τέχνη’ and ‘ἐπιστήµη’. The problem, at the beginning of logic in the Greek period, was to set up a discipline that realizes the possibility of well-founded argumentation without using these very means of argumentation under pain of begging the question. If logic, and in the case of Aristotle this means his syllogistic, would have to count as a science, it should obey the conditions laid upon a system of truths by Aristotle in order to have it represent an ἀποδεικτικὴ ἐπιστήµη. That is, there should exist a set of first true premisses8 out of which all further truths may be inferred by (apodeictic) syllogisms. But syllogisms never count as propositions (λόγοι ἀποφαντικοί), nor do perfect syllogisms count as axioms (ἀρχαί), nor are the reductions of the syllogisms to perfect ones called ‘proofs’ by Aristotle.9 Aristotle does not treat his syllogistic as a science. On the other hand, if the set-up of syllogistic would represent an art in the strong sense of a διαλεκτικὴ τέχνη, there should exist first premisses accepted for the sake of argument (τόποι), from which those propositions ____________ 7 8 9

Cf. Brouwer 1925. The other conditions that Aristotle imposes on these first true premisses, i.e., the axioms or, rather, ‘principles’ (ἀρχαί), as they are now called, are of no concern for our purposes; cf. An. post. 71b. This has been made a point in the convincing operationist interpretation of Aristotle’s syllogistic against the arguments of Łukasiewicz in: Ebbinghaus 1964.

Rules versus Theorems

7

about which the argument is concerned follow by (dialectic) syllogisms. It is obvious that the apparent axiomatic treatment of Aristotle’s syllogistic does not comply with these specifications, either. Syllogisms are used both for the sciences and for the arts, but they cannot themselves belong to either of them.10 Consequently, a syllogism which is defined twice, in the Prior Analytics and in the Topics, as a “linguistic expression (λόγος) in which, something having been posited, something other than the underlying results necessarily through the underlying”11 should neither be read as a theorem nor as a rule, though both possibilities have been adopted alternatively through the centuries.12 This view gets further support by observing Aristotle’s own argumentation on behalf of his choice for dealing with the objects of the Analytics. Instead of using the terminology of apodeictic or dialectic reasoning that would throw some light on Aristotle’s own opinion as to where to place the Analytics, he uses ‘analytically’ (ἀναλυτικῶς) instead of ‘apodeictically’, and ‘logically’ (λογικῶς) instead of ‘dialectically’ together with the interesting feature that most of his arguments on a certain point appear twice, once framed as an analytical one, and then as a logical one.13 This indefiniteness on the status of the arguments for the set-up of argumentation itself should not really give rise to surprise. A far more detailed investigation is needed to free the beginnings of logic from the air of circularity. The reason syllogisms are treated neither as theorems nor as rules is simply that in a way they are indeed both theorems and rules, depending on the level of argumentation. They can, any one of them, be considered as rules of inference14 – the syllogistic method in use is justly called a συλλογιστικὴ τέχνη by Aristotle15 – but as soon as the syllogisms are not considered with respect to their producing something out of something, but ____________ 10 Cf. Aristotle Met. 995a. 11 Top. 100a25–26; cf. An. pr. 24b19–20. 12 For two modern proponents of either possibility, cf. Łukasiewicz 21957, and Scholz 2 1959. Łukasiewicz interprets syllogisms as generalized subjunctions, Scholz reads them as rules of inference. For example, PaQ, QaR≺PaR (modus barbara) becomes ∧P,Q,R. a(P,Q)∧a(Q,R)→a(P,R). in Łukasiewicz, and PaQ; QaR ⇒ PaR in Scholz. 13 Cf., e.g., An. post. 83b–84b, where there are two ‘proofs’ for the claim that to each science there must exist first true and undemonstrable principles. 14 This is done successfully in the operationist interpretation of Aristotle’s syllogistic by Ebbinghaus as mentioned in note 9. In addition, the interchange of terms in PeQ (ἀντιστροφή), the contradictoriness of PaQ and PoQ, and of PeQ and PiQ, and the contrariness of PaQ and PeQ, all these formulated verbally by Aristotle, are given the form of rules; then all other valid syllogisms are ‘provable’ as admissible rules relative to the initial set of valid rules. 15 Cf. Soph. Elench. 172a36.

8

Philosophical Logic and Philosophy of Language

as entities sui generis, those rules of inference may be transferred into (logical) implications, i.e. three-place (meta)propositions on propositions, and thus theorems.16 Hence, syllogistic in the sense of a theory of the valid rules of inference may be taken as an early anticipation of the position held by the proponents of effective logic, now in a refined version: effective logic is to be considered as a theory, i.e., a system of truths, about the universally admissible rules within arbitrary systems of rules of action. Naturally, in the course of history, syllogistic has been treated the other way round, too. For example, according to the most influential diplomatic vote of scholasticism, which can be found in the Summulae Logicales (ca. 1250 A.D.) of Petrus Hispanus, who later became Pope John XXI, the definition of logic runs like the following: dialectica (i.e., logic) est ars artium et scientia scientiarum ad omnium methodorum principia viam habens;17 and Duns Scotus gives an interpretation of this twofold determination: logic is a science respectu materiae ex qua constat, and logic is an art respectu materiae in qua versatur. This distinction may now be taken as an anticipation of the position held by the proponents of classical logic, here again in a refined version, insofar as the system of tautologies can be enumerated by a calculus, i.e., a system of rules. More in the line of Aristotle, the more radical schoolmen such as Buridan in his Summa de Dialectica just dropped any mention of logic as a science and kept only its characterization as ars artium which, therefore, leads again to the position of the proponents of effective logic. Hence, classical logic is the result of starting with arbitrary theories that obey the axiomatic method by concentrating on the forms of truths within arbitrary domains, and then formalizing this system of formal truths by means of some calculus, thus getting a praxis on top of the theories. Effective logic, on the other hand, starts with arbitrary calculi built up by the constructive, i.e., genetic, method, and proceeds to a theory about the generally admissible rules within the calculi – a theory which can afterwards likewise be formalized; here we have a theory on top of the praxis. Now it looks almost like a matter of taste how one is going to choose the level for a reasonable beginning of formal logic. Yet, the following constructions claim that there is an adequate solution of the conflict ____________ 16 This is an accord with the characterization of arts and sciences in An. post. 100a8– 10. Arts are concerned with the world of coming-to-be and passing-away, sciences are concerned with the world of being. 17 There are references to the art-science dispute and its medieval background in the course of discussing modern operationist logic (Brouwer, Wittgenstein, Kolmogorov, Lorenzen) in: Richter 1965.

Rules versus Theorems

9

between a logic of rules and a logic of theorems and, more generally, a proper approach to the problem how praxis and theory interact in the case of logic.

II The starting point is again very close to the actual origin of logic in antiquity. With Aristotle, and even more with Plato, logic – or rather dialectic, the term being a strong hint by itself – had to provide the means by which sound argumentation could be distinguished from unsound argumentation.18 This has been a practical necessity in face of the highly developed sophistic technique to provide proofs for arbitrary theses on demand. And indeed, if it is granted that any scientific activity, be it on practical or on theoretical matters, is characterized as scientific by the fact that there is a justification available for each and every assertion put forth in the course of this activity (the possible linguistic articulations of nonlinguistic acts included!), there is no other basis for the construction of logic than to look for a methodical introduction of the linguistic elements of assertions and from there to proceed to the use of assertions within argumentations. Such an introduction of elementary linguistic elements shall be called primary praxis and will be executed within properly stylized teach-and-learn situations for these elements. As far as simple singular and simple general terms are concerned, the details of the procedure do not bear upon the set-up of formal logic. They have been discussed extensively elsewhere.19 For our purposes, it is sufficient to remark that introducing words by means of teach-and-learn situations guarantees their public understandability. Furthermore, it should be clear that the determination of a primary praxis in the given sense is a process post hoc, something man does in order to gain precise knowledge concerning his abilities and their limits after he has used speech and other acts meaningfully in the context of life. ____________ 18 Cf. the first sentence of the Topics (100a18–20), where the purpose of the treatise is characterized as “finding a method, by which we shall be able to argue (συλλογίζεσθαι) on any problem set before us starting from accepted premises (ἔνδοξαι) such that, when sustaining an argument (λόγος), we shall avoid saying anything self-contradictory”. It was E. Kapp who showed convincingly that the origin of Aristotle’s syllogistic (still taken to be a theory) is situated in the actual sophistic discussions on public affairs, cf. Kapp 1942. 19 Cf. the second part in: Lorenz 1970. There will be found special references to competing proposals in: Quine 1960, and in: Strawson 1959.

10

Philosophical Logic and Philosophy of Language

The introductions in question do not each constitute a creatio ex nihilo, they are rather ‘recreationes’, that is, a system of methodical reconstructions of that which has already been said and done. Another feature of the primary praxis is important: Due to the teach-and-learn situations connected with the introduction of terms, there is no difference between the situation articulated by means of the terms, and the situations in which those terms are used. No use of terms other than introducing terms has as yet been the object of consideration. But this, of course, is a trivial part of human speech. The special power of linguistic communication becomes apparent only, if the situations which underlie words, phrases, or sentences are different from the situations in which these words, phrases, or sentences are used. In that case, understandability of the linguistic expressions is not enough, a special link between the two situations is needed to secure the proper function of language. This link is provided by the detailed reconstruction – again through teachand-learn-situations – of possible uses of linguistic expressions after their introduction. Any such introduction of a use of linguistic expressions different from the introduction itself shall belong to the secondary praxis, e.g., the use of terms as wishes, questions, or propositions. The way this is done guarantees the public justifiability of linguistic expressions in addition to their understandability. The special act of asserting propositions (as distinguished from their use, e.g., in story-telling) involves a justifying procedure within the secondary praxis – a procedure that has to be introduced together with the use of terms as assertions – such that the validity of assertions can be defined by means of this procedure. It is even possible to distinguish words and sentences along these lines. If the situation articulated by a linguistic expression coincides with, or is at least part of the situation in which that expression is used, only its understandability is of concern, and the linguistic expression shall count as a word; but if those two situations are wholly different, both understandability and justifiability have to be secured and the linguistic expression shall count as a sentence. This now is the exact point for characterizing the justifying procedure of assertions as a dialogue, or an argumentation between two partners. To assert a proposition makes sense only, if there is someone on the other side, albeit fictitiously, who either denies or at least doubts the asserted proposition. But it is not enough merely to argue about propositions, there must exist precise stipulations on the rules of argumentation, rules which, in a way, define the exact meaning of the proposition in question. A proposition shall be called ‘dialogue-definite’ under the condition that the possible dialogues on this proposition are finished after finitely many steps according to some previously stipulated and effectively applicable rules of argumentation, such that, at the end, it can be decided who has

Rules versus Theorems

11

won and who has lost. Hence, dialogue-definiteness of propositions means that the relevant concept of a dialogue is decidable. And it is this concept of dialogue-definiteness that is to replace the age-old value-definiteness as the characterizing feature for linguistic expressions to be propositions.20 Further considerations will show that the class of value-definite propositions is indeed a proper subclass of the class of dialogue-definite propositions and that, therefore, the justifiability of propositions as introduced by the dialogue procedure does not coincide with their verifiability. Yet, ‘truth’ and ‘falsehood’ for propositions can now be defined on the basis of the dialogue game associated with each proposition. Such a definition marks the beginning of a theory about (primary and secondary) praxis, insofar as (meta)propositions on the actions within the praxis get introduced. In a certain sense, even the secondary praxis itself contains a theoretical element, namely the propositions themselves, which get their meaning by the rules of argumentation about them. Hence, it might be appropriate, at least for the systematic purposes of the whole set-up, to distinguish an object-theory (the class of propositions introduced within the secondary praxis) from a metatheory (about primary and secondary praxis), the propositions of which cannot, of course, exist without the same pragmatic foundation as the propositions on the ground level. At this early stage, already, the interaction of praxis and theory is far more complicated than the usual presentation of logical theories permits us to suppose. As a preparation for defining ‘truth’ and ‘falsehood’ for propositions, it is useful to observe that win and loss of a dialogue about a given proposition will in general depend upon an individual play of the game and will not be a function of the proposition alone. But the strategies of either player of the game are invariant against the choice of arguments of the other player. Hence, a proposition A shall be called ‘true’, iff there is a winning strategy for A; this means that the player who is asserting A – the proponent P – will be able to win a dialogue on A independently of the choice of arguments of the opponent O. Accordingly, a proposition A shall be called ‘false’ iff there is a winning strategy against A, i.e., the opponent can win a dialogue on A independently of the moves of the proponent.21 I have shown elsewhere22 that ____________ 20 The concept of dialogue within such a context has originally been introduced in: Lorenzen 1960; 1961, for the purpose of a better understanding of operationist logic. Its further explication, especially with respect to a pragmatic foundation of the calculi of intuitionistic and two-valued logic is due to K. Lorenz; cf. Lorenz 1968; 1978. 21 If, as usual, the validity of the logical principles is presupposed on the metatheoretic level, it would be possible from the validity of the saddle-point theorem for finitary two-person zero-sum-games to infer that propositions are either true or

12

Philosophical Logic and Philosophy of Language

the dialogue on the metaproposition ‘A is true’ coincides with the dialogue on A itself, which means that the traditional condition of adequacy for any definition of truth is satisfied: ‘A is true iff A’.23 The next step of the theory about the (primary and secondary) praxis is concerned with the justification of the rules of argumentation that constitute the secondary praxis. Again, this is done with a few accompanying remarks to the following proposal of a structural rule for dialogues, because space does not permit extensive elaboration on that point here.24 (Dl) Dialogues about propositions consist of arguments which are put forth alternatively by an opponent O and a proponent P. The arguments follow certain rules of argumentation that belong to the game such that each play ends up with win or loss for either player. (D2) With the exception of the improper initial argument, each argument either attacks prior ones of the partner or defends those of one's own upon such an attack, but does not act simultaneously in both ways: the proper arguments split into attacks and defenses. (D3) Attacks may be put forth at any time during a play of the game (rights!). (D4) Defenses must be put forth in the order of the corresponding attacks (upon which the defense answers), yet may be postponed as long as attacks can still be put forth: always that argument which has been attacked last without having been defended yet, has to be defended first (duties!). (D5) Whoever cannot – or will not – put forth an argument any longer, has lost that play of game; the other one has won it.

(Dl) is obviously not in need of further explanation; (D2) may be accepted as defining the special dialogue character of the game; and (D5) codifies equally current rules of win and loss. The only items in need of some further comments are (D3) and (D4) that regulate rank and order of attacks and defenses. With respect to the generality of rules, the right to attack shall not depend on a special position reached during a play of the game and, hence, shall not become void until the end of each play. On the other hand, the given order of defenses is a consequence of the stipulation ____________ false, cf. Berge 1957. Now, without begging the question, there is only a practical meaning of ‘either-or’ on the metalevel available, i.e., decidability of choice, but that cannot happen, because it is not generally decidable which part of the alternative holds; it is only decidable who has won a particular play of the game and who has lost it. 22 Cf. Lorenz 1968, pp. 35–36. 23 Cf. Tarski 1956, pp. 187–188. For a discussion about the danger of semantic antinomies, if this condition of adequacy is used as a schematic definition of truth, cf. Lorenz 1970, pp. 44–46. 24 For further details consult again Lorenz 1968, pp. 37–39.

Rules versus Theorems

13

in (Dl) to argue alternatively together with the rule of win and loss in (D5), if to both players is guaranteed that neither must defend upon an attack unless this attack has been defended first upon a counter-attack. Now, (D1) – (D5) are not sufficient to secure finiteness of the individual plays of the game. There is lacking a regulation on the number of attacks permitted against a single argument during a given play. Yet, since any choice of bounds would be arbitrary, it might be accepted as reasonable that this choice should become part of the dialogue game itself. After the initial argument has been laid down by P, first O shall choose a natural number n as the maximal number of attacks to be directed against a single argument of P, then P shall choose a natural number m analogously. Only now the proper dialogue about the initial argument may start obeying the following additional stipulation of the structural rule. (D6 n,m) During a play of the game, any argument may be attacked by the opponent at most n-times, by the proponent at most m-times.

In order actually to play a dialogue game according to the given rules, the rules of argumentation in (D1) have to be specified. This can be done by laying down a schema of attacks and defenses, which shows all possible attacks against an argument as well as all possible defenses of this argument upon each of these attacks. And, in general, this specification is possible only by special reference to the internal structure of the propositions concerned. The structural rule is purely ‘formal’ in the sense that no special knowledge about the proposition is needed, whereas the rules of argumentation are ‘material’ in so far as they have to make use of the actual set-up of the propositions, their ‘content’ in the terminology of traditional philosophy. Yet, there is a possibility of determining special rules of argumentation that are, in a way, formal, too, namely, those that make use only of the fact that propositions may be composed out of subpropositions. This leads to the concept of logical composition that in turn affects the introduction of further terms into the primary praxis, the so-called logical particles. A proposition A shall be called ‘logically composed’ out of propositions from a class K of given dialogue-definite propositions, if the schema of attacks and defenses associated with A contains only propositions from the class K. By means of such special rules of argumentation, the particle-rules, any dialogue about A is reduced to dialogues about the subpropositions of A: A is dialogue-definite, too. Easy combinatorial considerations show 25 that for a complete survey it is sufficient to discuss unary, binary and infinitary logical particles only, ____________ 25 Cf. Lorenz 1968, pp. 41–43.

14

Philosophical Logic and Philosophy of Language

under the condition to restrict the schemata in question to those that contain each subproposition just once and that use as further attacks certain non-assailable orders for defense or doubts, symbolized by ‘?’ with added indices. Scheme 1 will be self-explanatory. (As notation for plays of the dialogue game it seemed to be sufficiently suggestive to use two columns such that the rows are reserved for the consecutive attacks (from top to bottom with an index of the row number of that argument against which the attack is placed) together with an entry for the chosen defense – if any – upon that attack; to recover the order of moves one may enumerate the arguments, if necessary.) position negation (not)

*A ⌐A

attacks ?

¬A

A

A*B conjunction (and) adjunction (or) subjunction (if-then)

abjunction (but not) injunction (neither-nor) (all) (some) (no)

A∧ B

attacks 1? 2?

defenses A

defenses A B A B

A∨ B

?

A →B

A

B

A ←B

B A ? B ? A B

A

A

B

A

B

A

B

*xA(x) attacks ∧xA(x) ?n ∨xA(x) ? A(n) xA(x)

B A

defenses A(n) A(n)

scheme 1

As an example of a dialogue we will discuss the assertion A0 ((a→b)→b)→((b→a)→a) for dialogue-definite propositions a and b. Furthermore, we will make use of one of the main results of the theory of dialogue games, namely that the class of propositions for which there are

15

Rules versus Theorems

winning strategies is not changed if O is limited to at most one attack against an argument of P, whereas P may choose any number of attacks against a single argument of O.26 In the position as given in scheme 2, P must either have a winning strategy for b in order to win by defending a→b with b upon the attack a or a winning strategy against a. If, on the other hand, there exists a winning strategy for a, P should have defended his second move upon the attack b→a instead of counter-attacking the first move of O with a→b. Should O have chosen the defense b upon the attack a→b as his fifth move, P would have attacked the third move with b, and any attempt of O to start now a subdialogue about this b of P would lead to an imitation of this sub-dialogue by P about the b of O. O 1) 3)

(a→b)→b b→a

(0) (1)

5)

a

(3)

(1)

P A0 (b→a)→a

2)

a→b

4)

scheme 2

O must finally defend upon the attack b of P with a, and P in turn defends his second move with a. Any further attempt of O to try a sub-dialogue about this a of P results in an imitated sub-dialogue of P about the a of O. Therefore, if a is value-definite, A0 is true independently of the truth or falsehood of b. And it can be seen that there is a chance of winningstrategies which are formal in the sense that nothing need be known about the truth or falsehood of the prime propositions, as is the case, e.g., for A1 (a ∨ b)→((b→a)→a) (scheme 3). O

P A1 1) a ∨ b (0) (b→a)→a 2) 3) b→a (1) a 6) 5) a (1) ? 4)

O

P A1 1) a ∨ b (0) (b→a)→a 3) b→a (1) a 5) b (1) ? 7) a (2) b

2) 8) 4) 6)

scheme 3

In these cases, the win of a play for P does not depend on the outcome of the dialogues about the prime propositions, the crucial point being only the ____________ 26 Cf. Lorenz 1968, pp. 85–87.

16

Philosophical Logic and Philosophy of Language

possibility for P not to place a prime proposition as an argument until the same prime proposition has been placed as an argument by O. A special rule for formal playing can, hence, be formulated: (D7m) Prime propositions cannot be formally attacked; they may be put forth by the opponent without restrictions, whereas the proponent may only take over prime propositions from the opponent, each at most m-times, if m is the bound for the number of attacks against arguments of the opponent.

The restriction for P not to take over a prime proposition of O more than the maximal number of attacks he has chosen to place against a single argument of O during a play, is necessary in order to guarantee that the existence of a formal winning strategy, i.e., a winning strategy using formal playing only, for a proposition A is invariant against substitution of logically composed propositions for prime propositions within A. On the basis of the construction up to this point, it is now possible to introduce the concept of logical (or formal) truth of propositions by the existence of formal winning-strategies for them. And a propositional schema is valid iff propositions bearing that schema are logically true. This definition of validity for arbitrary dialogue-definite propositions works independently of any assumption on the truth or falsity of their prime propositions. It is, therefore, beyond the range of the classical theory of logical truth, even if the domain of propositions gets restricted both to value-definite ones and to quantifier-free ones. For, it is one thing to define the (classical) logical truth of a proposition A by the existence of (material) winning strategies for all propositions A* which result out of A, if the set of prime propositions of A runs through all combinations of truth and falsehood with respect to these prime propositions, and it is another thing to define the logical truth of a proposition A by the existence of a formal winning strategy for A. And, indeed, these two concepts do not coincide even within this restricted domain of propositions, as example A0 already shows. The propositional schema ((a→b)→b)→((b→a)→a) is valid classically (with the special point that only the value-definiteness of a is needed), though there is no formal winning strategy for it. The classical theory does not even permit the definition of this difference between general material truth and purely formal truth. It is the pragmatic approach to formal logic by means of dialogue games as it has been sketched here that leads to the definability of a concept of formal truth for propositions which do not generally satisfy the classical condition of value-definiteness. Hence, the dialogue concept makes it possible to define formal truth independently of material truth, whereas the classical theory is char-

Rules versus Theorems

17

acterized by the reduction of formal truth to material truth, namely as general material truth. At this late stage now it is reasonable to formalize the theory of dialogue games, and with this method to return to a praxis at a higher level, which, as the praxis of the calculi of logic, has been the usual starting point for contemporary logical theory. With the help of this formalization, it is possible to prove the main theorem of the theory of dialogue games: The class of valid propositional schemata coincides with the class of intuitionistically valid propositional schemata.27 If, furthermore, all classically valid propositional schemata shall be gained by formal winning-strategies, this means fictitiously to assume the value-definiteness of suitable propositional sub-schemata of the propositional schema in question: Any logically true proposition in the classical sense is logically true in the effective sense, if only suitable tertium-non-datur hypotheses are added. As an instructive example it is easily checked that there is a formal winning strategy for the classical disjunction ∧xa(x) ∨∨x¬a(x) under the three tertium-non-datur hypotheses ∧x. a(x) ∨ ¬a(x). , ∧xa(x) ∨ ¬∧xa(x) and ∨x¬a(x) ∨ ¬∨x¬a(x). Intuitionistic or effective logic is the logic of dialogue-definite propositions, two-valued or classical logic is the logic of the subclass of value-definite propositions. And it is quantification theory which shows the necessity for transition from the one to the other. In any case, we can conclude that logic is primarily the theory of a structured praxis. The logic of antiquity has not in general been very conscious (at least in the eyes of its interpreters) of the pragmatic basis of logic taken as a science. As a substitute, one often uses the very misleading phrase of the ‘ontological background’ of ancient logic. Yet, logic is, secondarily (by means of formalization), again a praxis of a structured theory. On the other hand, now, modern logic since Leibniz has minimized the importance of the possible theoretical basis of a calculus of logic, that is, logic taken as an art. Presumably, one had doubts about the precision which could be imposed on a logic formulated only within ordinary language. In both cases, there has been no clarity about the details of a step-bystep procedure from a praxis via a theory again to a praxis which, as we have tried to show, is the necessary minimum to get the means for an adequate solution of the dispute on the true nature of logic.

____________ 27 Cf. for a proof: Lorenz 1968; another proof in: Stegmüller 1964.

18

Philosophical Logic and Philosophy of Language

References Berge, Claude, 1957: Théorie générale des jeux à n personnes, Paris: Gauthier-Villars. Brouwer, Luitzen E. J., 1925: Über die Bedeutung des Satzes vom ausgeschlossenen Dritten in der Mathematik, insbesondere in der Funktionentheorie, in: Zeitschrift für reine und angewandte Mathematik 154, pp. 1–7. Ebbinghaus, Kurt, 1964: Ein formales Modell der Syllogistik des Aristoteles, Göttingen: Vandenhoeck und Ruprecht. Gödel, Kurt, 1933: Eine Interpretation des intuitionistischen Aussagenkalküls, in: Ergebnisse eines Mathematischen Kolloquiums, Heft 4, pp. 39–40. Heyting, Arend, 1930: Die formalen Regeln der intuitionistischen Logik, in: Sitzungsberichte der Preussischen Akademie der Wissenschaften, Physikalischmathematische Klasse, pp. 42–56. Kapp, Ernst, 1942: Greek Foundations of Traditional Logic, New York: Columbia University Press. Kleene, Stephen C., 1952: Introduction to Metamathematics, Princeton/Toronto/ NewYork: D. van Nostrand Company. Kolmogorov, Andrej N., 1932: Zur Deutung der intuitionistischen Logik, in: Mathematische Zeitschrift 35, pp. 58–65. Lorenz, Kuno, 1968: Dialogspiele als semantische Grundlage von Logikkalkülen, in: Archiv für mathematische Logik und Grundlagenforschung 11, pp. 32–55, 73–100. Lorenz, Kuno, 1970: Elemente der Sprachkritik. Eine Alternative zum Dogmatismus und Skeptizismus in der Analytischen Philosophie, Frankfurt am Main: Suhrkamp. Lorenz, Kuno, 1978: Arithmetik und Logik als Spiele [1961] , (partially reprinted) in: Paul Lorenzen/Kuno Lorenz, Dialogische Logik, Darmstadt: Wissenschaftliche Buchgesellschaft, pp. 17–95. Lorenzen, Paul, 1955: Einführung in die operative Logik und Mathematik, Berlin/ Göttingen/Heidelberg: Springer. Lorenzen, Paul, 1960: Logik und Agon, in: Atti del XII Congresso di Filosofia (Venezia, 12–18 Settembre 1958). v 4, Firenze: Sansoni Editore, pp. 187–194. Lorenzen, Paul, 1961: Ein dialogisches Konstruktivitätskriterium, in: Infinitistic Methods. Proceedings of the Symposium on Foundations of Mathematics (Warsaw, 2–9 September 1959), Oxford/London/New York/Paris: Pergamon Press, pp. 193–200. Łukasiewicz, Jan, 21957: Aristotle’s Syllogistic from the Standpoint of Modern Formal Logic [1951], Oxford: Clarendon Press. McKinsey, J. C. C./Tarski, Alfred, 1948: Some theorems about the sentential calculi of Lewis and Heyting, in: The Journal of Symbolic Logic 13, pp. 1–15. Quine, Willard V. O., 1953: From a Logical Point of View. 9 Logico-Philosophical Essays, Cambridge Mass.: Harvard University Press. Quine, Willard V. O., 1960: Word and Object, New York/London: John Wiley & Sons. Richter, Vladimir, 1965: Untersuchungen zur operativen Logik der Gegenwart, Freiburg/München: Verlag Karl Alber. Scholz, Heinrich, 21959: Abriß der Geschichte der Logik, Freiburg/ München: Verlag Karl Alber. Skolem, Thoralf A., 1923: Begründung der elementaren Arithmetik durch die rekurrierende Denkweise ohne Anwendung scheinbarer Veränderlichen mit unend-

Rules versus Theorems

19

lichem Ausdehnungsbereich, in: Skrifter utgit av Videnskapsselskapet i Kristiania, I. Matematisk-naturvidenskabelig klasse, No. 6. Stegmüller, Wolfgang, 1964: Remarks on the Completeness of Logical Systems Relative to the Validity Concepts of P. Lorenzen and K. Lorenz, in: Notre Dame Journal of Formal Logic 5, pp. 81–112. Strawson, Peter F., 1959: Individuals. An Essay in Descriptive Metaphysics, London: Methuen & Co. Tarski, Alfred, 1956: The Concept of Truth in Formalized Languages [German 1936], in: Alfred Tarski, Logic, Semantics, Metamathematics. Papers from 1923 to 1938, transl. by J. H. Woodger, Oxford: Clarendon Press, pp. 152–278.

On the Relation between the Partition of a Whole into Parts and the Attribution of Properties to an Object I Since the invention of the calculus of individuals, first 1916 by Leśniewski,1 there has been a continuous discussion about the question of how far it may serve as a substitute for the usual calculi of set theory. The numerous analogues of operations on sets using the terminology of parts and wholes within the theory of individuals gave rise to hopes that the whole hierarchy of logical types might – at least to a certain extent – be represented in a theory of one logical type only. It was Nelson Goodman together with Henry Leonard who, in their classical paper on the calculus of individuals,2 inferred explicitly from the interpretative power of a combined formal system using ‘is a member of’ together with ‘is part of’ as primitive notions that “the dispute between nominalist and realist as to what actual entities are individuals and what are classes is recognized as devolving upon matters of interpretative convenience rather than upon metaphysical necessity”.3 It is a matter of habit, they declare, to treat qualities in interpretations as entities of one type higher than things, one could equally proceed the other way round. What counts is that the concept of being an individual and that of being a class distinguish one segment of the total universe from the rest in a different manner: To conceive a segment as a whole or individual offers no suggestion as to what the potential subdivisions of that segment must be “whereas to conceive a segment as a class imposes”, so the authors say, “a definite scheme of subdivisions – into subclasses and members”.4 Apart from the obvious fact that to subdivide into members of a class is by no means more definite than to subdivide into parts of an individual, it does happen, though, that some interrelations of classes – the authors quote the example of windows and buildings where each window is a part ____________ 1 2 3 4

Leśniewski 1916. Goodman/Leonard 1940. Op. cit., p. 55. Op. cit., p. 45.

On the Relation

21

of some building – cannot be expressed using the class-membershiprelation alone, unless windows and buildings are reinterpreted as entities of some other logical type – here, e.g., specified classes of atoms. And it is this crucial feature of logico-linguistic world representation that calls at least for an amendment to set theory, if not for its, possibly only partial, substitute through a mereology. This argument sounds convincing, would not the usual consequence, the set-up of a formal system – with, e.g., the following well known postulates: 1. Transitivity of the relation ‘is part of’ (≤) 2. Unique existence of the composite κP (called ‘fusion’ by Goodman and Leonard) for each non-empty predicate P (κP ι x.∧y.x○y↔∨z.P(z) ∧ z○y… with x○y ∨z.z≤x∧z≤y. as definition of ‘overlapping’)

– as a companion to a formal system of set theory, have to make use of standard interpretations of the two primitives, ‘ ’ and ‘≤’, in the metalanguage. When we write, as usual, ‘P(n)’ to represent the object n as falling under the concept ϑP, i.e., the intension of the predicate ‘P’, or the object n as being a member of the class P, i.e., the extension of the predicate ‘P’, we do this in the relational cases, x y and x≤y, as well – at least with respect to potential interpretations –, and it is equally common, and now on the pure syntactic level irrespective of semantic considerations, to consider the symbols as parts of the formula that is made out of them by, usually, concatenation. So we should answer the question of how in ‘x y’ – for ‘x≤y’ the argument runs in an analogous fashion – the phrase ‘x and y are, in this order, relata of the class-membership-relation’ interferes, contrasts, or correlates with the phrase ‘ ‘x’, ‘ ’, ‘y’, and linear combinations thereof are parts of ‘x y’ ’. We can read ‘x y’ – and likewise ‘x≤y’ – either semantically or syntactically, only by either using the standard interpretation of ‘ ’ or by using the standard interpretation of ‘≤’, or variants thereof. For convenience of presentation from now on, I go back to ‘xεP’ with some predicate letter ‘P’ as the logical basis of the notation ‘x y’ (being a suggestive shorthand for ‘x,yε ’ with ‘y’ referring to the extension of ‘P’), and I read ‘xεP’ with the affirmative copula ‘ε’ as attribution of property P to object x. It seems clear, then, that the relation ‘is part of’ obtains only among objects (of whatever type), in contrast to attribution occurring on the (meta)level of stating that, e.g., a relation x≤y, i.e., x,yε≤, holds. Hence, attribution is not a relation, but a means to articulate that properties,

22

Philosophical Logic and Philosophy of Language

relations included, hold of objects. The usual procedure to represent attribution seemingly on the object level by using the extensional classmembership-relation between objects of consecutive types hides this fact. In such a way, attribution has just been pushed one step higher, with x y instead of xεP, from xεP to x,yε . Thus, rather than claiming that ‘is a member of’ and ‘is part of’ establish relations among objects of different types and among objects of equal types, respectively, one should distinguish both notions more fundamentally as one of description and one of construction. Attribution is a means to describe (to express, to state, or whatever term is used in ordinary discourse) a certain (elementary) state of affairs, even if this ‘connection’ between the object level and the language level is transposed into the object level and represented by the ordinary settheoretical -relation. Partition is a means to construct (to analyse, to build up, to subdivide, or whatever term is used in ordinary discourse for the procedure to derive parts from a whole or a whole out of parts) a certain (non-elementary) object out of other objects that afterwards count as parts of a whole. It is worthwhile to note that within this framework there is as yet no decision, whether this construction works by synthesis (starting with the parts) or by analysis (starting with the whole). For example, to construct natural numbers through counting, that is, successively adding one (to be described afterwards as applying the successor function), should be called ‘construction by synthesis’ (yielding ‘additive’ parts for each natural number considered as a whole), whereas the representation of natural numbers as a product of prime numbers may count as an example of a ‘construction by analysis’ (yielding ‘multiplicative’ parts for each natural number considered as a whole). Corresponding examples from the sciences abound. In addition, we may observe that within the restriction of a given domain of objects the usual device of object forms or terms serves to represent forms of partition, whereas propositional forms or formulas – at least in the elementary case – represent forms of attribution. And the well known interrelation of terms with formulas give us the first hint of how to approach the more general relation between the language of parts and wholes and the language of properties. Yet, to get a fair account of the difference, it is useful first to treat an example simultaneously by description and by construction. Let us take a logically compound formula, e.g., A a ∧ ¬b. Then, it is possible to describe this formula by ι xA(x) using a predicate A on the domain of expressions with roughly the following meaning: A(x) x is derivable in one of the usual calculi for constructing wellformed formulas and x is linearly composed out of a, ∧, ¬, b, in that order. One could even

On the Relation

23

axiomatize these syntactic notions of derivability and concatenation such that it is a question of logical consequence from the axioms, whether ι xA(x) exists or not. On the other hand, one may simply construct the formula A according to the rules of a suitable calculus for deriving well-formed formulas. The way of description needs a proof of (unique) existence of the described object, the way of construction is by itself a proof of existence of the constructed object. As can be easily seen, the difference comes about essentially through the change from parts (of an object) to the property (of that object) of having those parts. And, furthermore, this simultaneous treatment shows that the natural way of comparison between part/whole and object/property (resp. class) does not follow the course of considering an object once a whole (or an individual) and once a property (or a class), though comparison of the respective formal properties of x≤y and x y in the calculi of the theory of individuals and the calculi of set theory is bound to concentrate on this line of investigation. It, rather, turns out that greater reward may be expected from looking more closely to the usual terminological mix-up when having parts are said to be properties of objects and when properties of objects are said to be partial determinations of objects: both parts and properties belong to objects. If we look at our simple example with this idea in mind, another observation leads to the following generalisation: To identify an object nεN, it is necessary either to have at least one characterizing property Q for determining that object, i.e., n = ι xQ(x), where N is to be taken as a set of values for the variable x, or to have at least one characterizing part ι R for constituting that object, i.e., ∨R(x) x
24

Philosophical Logic and Philosophy of Language

II As a natural next step I propose to treat the part-whole-distinction as the basis for the identification of objects as instances of some kind N. In order to do this, it is of outmost importance to observe the difference between (singular) actualizations δN of N (δN is a case of indexing using the logical indicator ‘δN’) and (particular) instances ιN of N (ιN is a case of contextdependent naming using the deictic nominator ‘ιN’). Among any two actualizations δN there is no possibility to apply ‘same’ or ‘different’; they are different only solo numero and not yet ‘entities’ according to Quine’s dictum ‘no entity without identity’.5 Hence, they cannot count as units that permit multiple reference to them, as is the case with instances, even if the means of context-free characterization of an instance are not yet available. Henceforth, I will speak of an articulator (denoting a schema N) as long as the actualizations δN of the schema are not structured into units ιN that will have to count as the instances of a specific type, e.g., a kind, derived from the schema. The initial stage of having nothing but articulators at one’s disposal may roughly be characterized as the stage where only ‘feature universals’ in the sense of Strawson exist:6 ‘cat’, ‘more cat’, ‘more cat again’, … or: ‘red’, ‘red again’, … or: ‘running’, ‘again running’, … or ‘water’, ‘more water’, ‘water again’, … is a way of rendering what happens in this stage verbally. There do not exist specified units, neither of things like cats nor of actions like running, and – in indo-european languages – there will not even in a more developped stage exist such specified individual units for qualities like red or for materials like water (except when redefined in scientific language) as belonging to the meaning of these terms. In particular, the domain of natural numbers that has to be construed as consisting of units, i.e., the initial sections of counting, is not yet available in the early stage with nothing but feature universals. Nor any other set of elements in the proper sense of ‘set’. Just counting, without retrieval of units, of course, is available. In order to develop the means for identifying objects we may, now, follow a double course. If ‘N’ is an articulator, the course to follow is either to treat ‘N’ as a proper name of the corresponding schema which is something like a rule for generating actualizations of that schema, or to treat ‘N’ as a proper name for the totality of N, i.e., the ‘fusion’ or ‘summation’ of the actualizations of N. In the first case, the actualizations δN are identified, thus yielding the schema that should, for ____________ 5 6

Cf. Gottlieb 1979. Strawson 1959, chap. 6.6.

On the Relation

25

better recognition, be rendered by ‘σN’ instead of merely by ‘N’. We are in tune with Leibniz’ principle of the identity of indiscernibles, because nothing but the articulator ‘N’ may be used for differentiation. In the second case, the actualizations δN – or, rather, partial summations of them – appear as parts of the totality of N that may be called the composite κN. Here, ordinary English usage allows of the phrase ‘all the N’ like ‘all the water’,7 and not ‘the whole N’8, which is reserved for the case of schemata with specified units, i.e., schemata turned into types, denoting some such unit, e.g., ‘the whole country’. An analogous difference of usage permits to speak of, e.g., ‘the set of countries’, but not of, e.g., ‘the set of water[s]’, as compared to ‘a lot of N’ in both cases. Thus, we are indeed able to identify the object σN, being a schema or, rather, a schema turned into the type with κN as its sole token, as being different from, e.g., the object σM through its characterizing part, the name-tag ‘N’. The identification of the object κN is achieved in like manner. The two objects, type σN and token κN , come about by two procedures with respect to the actualizations of schema N: identification and summation, respectively. Identification (that is ‘looking at actualizations universally’) and summation (that is ‘looking at actualizations singularly’) lie at the bottom of attribution and partition. This idea will now be used to derive specified identifiable objects, the particular Ns that in tune with ordinary logico-linguistic usage should be called individuals, from the actualizations δN. With strict contextdependent deictic descriptions ‘ιN’ we will canonically refer to individuals as long as characterizing properties or characterizing parts of them are not yet available. What we have to do is to impose a ‘subdivision’ upon the actualizations of a schema N, such that we dispose of – not necessarily mutual exclusive – subschemata of N. Together with an articulator ‘N’ denoting the schema σN (or, rather, its associated type with κN being its sole token) we are to assume an indefinite number of ‘N-individuating articulators’ ‘ιN’ that denote – with respect to N – intermediate schemata σ(ιN). Each individuator ‘ιN’ of an articulator ‘N’ gives rise to an individual n, being materially the partial composite κ(ιN) and formally the intermediate schema σ(ιN). The final step of this construction calls for an explicit replacement of the articulator ‘N’ denoting the schema σN with its actualizations δN by a predicator ‘N̲ ’ that denotes just that type derived from the schema by an individuating subdivision that has the (layer of) individuals ιN as its tokens ____________ 7 8

Cf. Aristotle’s use of ‘πᾶν’, Met. ∆ 1024a. In Aristotle ‘ὅλον’, ibid.

26

Philosophical Logic and Philosophy of Language

or instances. Individuation may happen in many ways, individuals of a schema being finely or coarsely grained. Hence, there will exist many typifications of a schema beside the trivial one of just ‘the whole N’, i.e., κN, as the sole token of the schema σN turned into a type. We have compared the existence of articulators without individuators roughly with the existence of feature universals in the sense of Strawson. A more refined account of some important logico-linguistic categories may now be given in the following way: Predicators ‘N̲ ’ with canonically attached individuators such that the intermediate schemata σ(ιN) are treated as semantically inseparable from the schema σN shall be called individuatives (= count nouns, denoting ‘sortal universals’ in the sense of Strawson). In English and other indo-european languages, ‘N̲ ’ is in this case represented by prefixing the indeterminate article to the articulator ‘N’; hence, ‘n ε N̲ ’ must be read ‘n is an N’, whereas ‘this is N’ is not even well-formed. In contrast, νN with numerals ‘ν’ should be construed as wholes, i.e., composites, associated with a set of ν units N̲ such that ‘this is one N’ and ‘these are νN’ (ν ≠ 1) can be rendered as ‘this/these = νN’. This accounts especially for the difference between ‘this is one N’ and ‘this is an N’. Furthermore, the natural numbers are now available essentially as the initial sections of counting units N̲ with the additional abstraction from the choice of the predicator ‘N̲ ’. We will say that individuatives are characterized by semantic individuators. In all other cases the individuators of a predicator ‘N̲ ’ are dependent on the circumstances of the use of ‘N̲ ’, hence, they are pragmatic individuators. We will call such predicators ‘N̲ ’ continuatives (= mass nouns), and we should observe that ‘N̲ ’ is not only determined by the non-linguistic context of an utterance in which the articulator ‘N’ occurs (either in predicating position: __ε N, or in nominating position: this N __), it may likewise be expressed by adding special individuating terms to ‘N’ as, e.g., this drop of water __ , __ε a brillant red, __ε a piece of furniture, this series of jumping __, etc.9 Linguistic evidence seems even to support the general statement that for a language in use, i.e., within its surface structure, only terms of the form ‘N̲ ’ are relevant and that terms of the form ‘N’ just serve the theoretical purposes of logico-genetic reconstruction, i.e., are needed only for the presentation of the deep structure of a language. We have now reached the stage, where a domain of individuals can be treated as given through the instantiations of a predicator ‘N̲ ’, though it remains to be shown how the identifiability of any one of these individuals may be secured with the means now at our disposal. ____________ 9

Cf. Lorenz 1970, part II, chap. 2.2.

On the Relation

27

The individuals being determined by the individuators of some predicator bear ‘nominators’ (as I prefer to say instead of ‘singular terms’) that got symbolized using ‘ι’ as prefix to the associated articulator. Of course, this abbreviated use of the iota-operator makes sense only, if the implied reference to context for making a phrase of the form ‘this N’ referentially non-ambiguous, can be made explicit. And, here again, two ways to proceed are available. They correspond to the two principles of individuation mentioned above that work through providing either characterizing properties QN or characterizing parts ιNR for the individuals ιN. Any such ιR is a constituent of ιN, if considered as a whole, whereas Q is part of the so-called ‘individual concept’ of ιN. And a proper name ‘n’ for ‘ιN’ will serve both ways, insofar as Q(x) x = n may count as a characterizing property and ιR n (this name!) as a characterizing part of ιN, though there are good reasons to assume that under historical analysis proper names resolve into determinate descriptions using characterizing properties that derive from characterizing parts. Any such articulator ‘Q’ now acts as a classifier on the domain N̲ , i.e., the individuators of ‘Q̲ ’ turn the individuals ιQN into those that coincide with some ιN. Yet, there are no reasons why individuals that arise from different articulators should coincide. It, rather, is a question of the set-up of a specific language system of how to make a choice on what sorts of individuals shall govern the other possible individuators of that language. For example, to analyze ‘man’ into ‘human being’ or, following a philosophical tradition, ‘rational animal’, exhibits that the individuals of ‘man’, of ‘animal’, and of ‘being’ coincide within the domain of common instantiations. We can say that in this case ‘animal’ acts as an individuating term for the preceding ‘rational’, turning ‘rational’ into a count noun with units of ‘animal’ as its only units. On the other hand, ‘this [being] rational’ refers to an individual of ‘rational’ that can never coincide with an individual of, say, ‘animal’. We have arrived at the problem of word classes, or of categories, and this one example, already, shows that the problem of how to discriminate word classes is inseparably linked with the problem of how to represent individuators linguistically.

III For the present purposes it is sufficient to give the following definition: A predicator ‘N̲ ’ shall, in predicating position: __εN̲ , be called an eigenpredicator, if instances of N̲ coincide with instances of at least one of the possible predicators ‘N̲ ’ that are associated to the articulator ‘N’ like‘N̲ ’, in

28

Philosophical Logic and Philosophy of Language

nominating position: ιN__. In the other case, when there is no determination of units in nominating position that coincides with one in predicating position, the predicator ‘N̲ ’ shall, in predicating position, be called an appredicator. It follows that elementary sentences of the form ‘this ε N̲ ’ with an eigenpredicator ‘N̲ ’ may synonymously be expanded into ‘this Nε N̲ ’, whereas in case of an appredicator ‘N̲ ’ any such expansion will be wrong, a kind of µετάβασις εἰς ἄλλο γένος, to extend the use of a traditional term for an inferential fallacy. To give an example, the sentence ‘this ε red’ is, with reference to the distinction between eigen-predicator and appredicator, ambiguous. But further determination of ‘red’ makes resolution of that ambiguity obligatory in English (as well as in German): ‘this brilliant red’ – ‘brilliant red’ is an appredicator – versus ‘this is a brilliant red’ – ‘a brilliant red’ is an eigen-predicator with ‘a’ as its individuator (in German slightly different: ‘dies ist strahlend rot’ versus ‘dies ist ein strahlendes Rot’). Hence, appredicative use of a predicator ‘N̲ ’ – it denotes a ‘characterizing universal’ in the sense of Strawson – is necessarily connected with presuppositions concerning the instances of ‘N̲ ’. Since ‘this ε N̲ ’ does in this case not permit the reading ‘this N ε N̲ ’, there must – with one exception – exist another articulator ‘P’ (with associated predicators ‘P̲ ’) such that ‘this ε N̲ ’ is synonymous with ‘this P ε N̲ ’, and the individuals of ‘P̲ ’ as implied (semantically or pragmatically) in the nominating position ‘this P’ coincide with the individuals of ‘N̲ ’ on the domain of P-individuals. The exception occurs, when in ‘this ε N̲ ’ the demonstrative ‘this’ refers to ‘the whole world’ u, i.e., the complete surrounding of a speaker, taken as a single object beyond any possibility of its being an instantiation of some predicator, unless ‘object’ is accepted as such an all-inclusive predicator. With respect to u the individuals ιN as well as the composite κN may be called parts of the (whole) world: κN
On the Relation

29

acts like ‘a mammal’ as a classifier and, therefore, as an eigen-predicator on the domain of dogs. We have reached the point now, where the usual set-theoretic treatment of properties enters the stage. They get defined as classifiers on some fundamental domain of individuals. But a treatment of modifiers as predicators in their own right rather than as linguistic entities on the metalevel, i.e., as functors or merely as syncategorematic terms, is out of reach.10 This situation induces the decisive step of the present reasoning to start with the appredicative use of predicators and to show that for each appredicator ‘N̲ ’ the individuals ιN appear as proper parts of those individuals ιP for which ‘ιPεN̲ ’ holds. This includes again even the exceptional case ‘uεN̲ ’, where the ιN are parts of the whole world u. As for eigen-predicators ‘N̲ ’ and an appropriate determination of units ‘ιNεN̲ ’ holds, the same consideration yields just the whole individual ιN as an improper part of ιN. Let us first recall what are the parts of an individual ιP. Any actualization δP of the intermediate schema σ(ιP) are actualizations of the schema σP as well, and composites of any number of such δP had been defined as parts of κ(ιP), i.e., of the individual ιP as a whole. Now, it may happen that an actualization of a schema denoted by the articulator ‘P’ is likewise an actualization of another schema denoted by ‘R’. But that means that in learning to deal with the schemata P and R through repetition and imitation some actualizations δP and δR coincide, though this coincidence does not extend to the determination of individual units ιP and ιR.11 As examples for ‘P’ and ‘R’ one may take ‘house’ and ‘garden’, or again ‘dog’ and ‘barking’. Even, if for some actualizations δP and δR coincide, no such ‘common denominator’ need exist for the respective individuals ιP and ιR. Instead, all δR of some σ(ιR), being coincident with corresponding δP, may be actualizations of some σ(ιP). Hence, κ(ιR)<κ(ιP) holds, and the individual ιR is a proper part of the individual ιP. If, according to our assumption, ιPεN̲ now holds for an appredicator ‘N̲ ’ which implies that none of the possible individuals ιN will coincide with ιP, this can only happen in the following way: Some actualizations δP of the intermediate schema σ(ιP) must coincide with actualizations δN of the schema σN, as otherwise the attribution ‘ιPεN̲ ’, being derived from ____________ 10 With such a consequence Quine argues in: Quine 1960, § 22; cf. in addition, e.g., Bartsch 1972. 11 Some authors, like Goodman in: Goodman 21966, prefer the term ‘togetherness’ – with a somewhat more restricted application – to ‘coincidence’.

30

Philosophical Logic and Philosophy of Language

‘this ε N’ via ‘this ε N̲ ’, would be without foundation. On the other hand, not all actualizations δP of σ(ιP) can coincide with actualizations δN of σN, as this would imply that ιP itself is an individual ιN. Hence, the composite out of all those δP of σ(ιP) that coincide with actualizations δN is a proper part of κ(ιP) and must be either the whole κ(ιN) of some individual ιN or the whole out of some number of individuals ιN. We may say that in this case ‘N’ acts as an articulator not within the whole world but within ιP, only; and each of the associated predicators ‘N̲ ’ singles out parts of ιP as its instances. In this way it is shown that the appredicative use of a predicator derives from its articulation of proper parts of those individuals that serve as determiners of the individuation of the predicator in question when used as an appredicator. Included is the special case, when the whole world is the only such individual. Of course, working backwards, any articulation of proper parts of individuals ιP or the whole world u by means of a predicator ‘R’, i.e., ι PR<ιP or ιR
On the Relation

31

‘garden’ and ‘house’ have common actualizations, you have to use ‘a garden’ as grammatical object of ‘have’ – this house has a garden – in order to to express that an individual garden is part of an individual house, and the switch from a predicator ‘P̲ ’ to the appredicator ‘having-P̲ ’ allows and enforces to keep the original determination of units of ‘P̲ ’. It should be added that ‘having-P̲ ’ changes into ‘with P̲ ’ as soon as this appredicator is used as a modifier of some ‘Q̲ ’: ‘this house has a garden’ becomes ‘this house is a house with garden’. In other cases, or in the same case in some other language, there exists a categorial shift from ‘P’ to ‘P'’, such that the individuals ιP and ιP' for certain choices of individuators coincide, though in predicating position ‘P̲ ’ acts exclusively as an eigen-predicator and ‘ιP̲ '’ as an appredicator. This shift is known to underly the division of simple predicative expressions into substantives and predicatives, the latter comprising both adjectives and verbs. If, as an example, we take ‘flying’, this term may serve both as eigenpredicator and as appredicator (e.g., ‘this movement is [a case of] flying’ and ‘this bird is flying’), but the noun ‘flight’ and the finite verb-forms of ‘flying’ represent a case of strict categorial division. To give another example, in Hungarian there exists this categorial shift from substantive to predicative even in the case of ‘garden’: from ‘kert’ (garden) the adjective ‘kertes’ is derived, allowing ‘ez a ház kertes’ as a rendering of ‘this house has a garden’, or ‘a kertes házunk’ as a rendering of ‘our house with garden’. In general, if the appredicator ‘P̲ '’ derived from an eigen-predicator ‘P̲ ’ turns out to be an adjective, there is usually a choice to use either ‘havingP̲ ’ or ‘[being]-P̲ '’ for representing attribution of a property P to that individual ιQ to which instances of ‘P̲ '’ belong as parts, of course not necessarily as spatial parts. It is a matter of further investigation, most likely with the help of categorial grammars,12 to survey the linguistic means for representing the different kinds of individuators that turn an articulator into a predicator, and, perhaps, even to give reasons why certain individuations – of ‘substances’, according to traditional logic – govern the determination of individuals for the other articulators.

____________ 12 Cf., e.g., Cresswell 1973.

32

Philosophical Logic and Philosophy of Language

References Bartsch, Renate, 1972: Adverbialsemantik, Frankfurt am Main: Athenäum. Cresswell, Maxwell J., 1973: Logics and Languages, London: Methuen & Co. Goodman, Nelson, 21966: The Structure of Appearance [1951], Indianapolis/ NewYork/Kansas City: Bobbs-Merrill. Goodman, Nelson/Leonard, Henry, 1940: The calculus of individuals and its uses, in: The Journal of Symbolic Logic 5, pp. 45–55. Gottlieb, Dale, 1979: No entity without identity, in: Essays on the Philosophy of W. V. Quine, ed. by Robert W. Shahan and Chris Swoyer, Norman: University of Oklahoma Press, pp. 79–96. Leśniewski, Stanisław, 1916: Podstawy ogólnej teorii mnogości I [Foundations of a General Theory of Sets I], Moskwa: Popławski Lorenz, Kuno, 1970: Elemente der Sprachkritik. Eine Alternative zum Dogmatismus und Skeptizismus in der Analytischen Philosophie, Frankfurt am Main: Suhrkamp. Quine, Willard V. O., 1960: Word and Object, New York/London: John Wiley & Sons. Strawson, Peter F., 1959: Individuals. An Essay in Descriptive Metaphysics, London: Methuen & Co.

Basic Objectives of Dialogic Logic in Historical Perspective The extensive research in logic conducted by using concepts and methods of game theory allows to see dialogic logic in a number of new perspectives.1 This situation may gain further clarity by looking back to the inception of dialogic logic in the late fifties and early sixties of the last century. It all began with critical discussions about the ‘operationist’ (German: operativer) approach2 to logic and mathematics as developped by Paul Lorenzen in his Einführung in die operative Logik und Mathematik.3 There, the thesis of Haskell B. Curry that “mathematics is the science of formal systems”4 where ‘formal system’ is coextensive with ‘calculus’ and not restricted to formalized theories, gave rise to the observation that elementary propositions about calculi may be proved without any ‘logic’. A proof of an assertion of derivability, of a for example, consists of a derivation of a. But that is just a finite sequence of moves with a as the final move according to the rules of the respective calculus. The only condition a calculus must satisfy is the condition of decidability of the predicate ‘is a derivation’. In such a case, the rules of the calculus for strings of atoms, i.e., the ‘figures’, can effectively be handled: derivability assertions are ‘proof-definite’. Now, logic enters the picture – and this was the main idea of operationist logic – when the usual logical particles receive an interpretation within a basic theory of arbitrary calculi where logical notions proper do not yet play a role. This basic theory – ‘proto-logic’ in the terminology of Lorenzen5 – is essentially a system of proof procedures for assertions of admissibility of rules: A rule R is admissible in a calculus K, iff a derivation of a figure a in K using R can be substituted by a derivation of a using the defining rules of K, only. Proof procedures for admissibility (of R in K) assertions amount to nothing but eliminability (of R in K-derivations) ____________ 1 2 3 4 5

Cf., e.g., volume 127 (2001) of Synthese devoted to ‘New Perspectives in Dialogical Logic’. The English rendering of ‘operativ’ by ‘operationist’ is due to Fränkel/BarHillel/Levy 21973, chap. III.7 (pp. 179–180). Lorenzen 1955. Curry 1951, p. 56. Cf. Lorenzen 1955, p. 7.

34

Philosophical Logic and Philosophy of Language

procedures, except in the trivial case of admissibility where the assumption, already, that R was used in a K-derivation is provably false. No logic in the sense of a system of formal rules of inference governs the proof procedures. If a⇒b is an admissible rule in some calculus K, or, rather, an instantiation of such a rule (because otherwise an assertion of derivability concerning either premiss or conclusion of this rule, a or b, would not be an elementary proposition), the subjunction a→ b being a logical composition out of elementary propositions and, hence, a logically compound proposition of the theory of calculus K, gets defined by the admissibility of a⇒b in K. Proofs of admissibility (of a⇒b) count as proofs of subjunctions ‘in toto’ and cannot in general be treated as composite structures out of proofs of derivability (of a and of b, separately). The logical particle ‘if-then’ [→], a connection of two sentences, is interpreted as the ‘poietical’ ‘if-then’ [⇒] of acting according to a rule which is a connection of two objects by a procedure of transformation. This includes the special cases of negation (admissibility of a⇒ with ‘ ’ being some underivable figure) and generalisation (admissibility of ⇒a, with ‘ ’ being some derivable figure). The remaining logical particles – ‘and’, ‘or’, ‘some’ – are treated differently. They are not introduced by (semantical) interpretation within the theory of K but by (syntactical) addition to the primitive figures of K together with the well-known ‘introduction rules’: a;b ⇒ (a∧b) , a ⇒ (a∨b), b ⇒ (a∨b) , a ⇒ ∨x a. Now, iteration of subjunctions along these lines enforces the introduction of meta-calculi K n with respect to the basic calculus K (= K 0 ) in such a way that the derivable figures of K n are just the admissible rules of K n-1. Of course, this is a definition of ‘meta-calculus’ by description and not a definition by construction. Hence, there is no way to guarantee the condition of decidability for the predicate ‘is a derivation’ beyond the basic calculus. The meta-calculi are not any more calculi in the proper sense, even though proofs of admissibility of rules with respect to these metacalculi, i.e., of meta-rules, meta-meta-rules, etc., can be given by the same procedures as in the case of basic rules. In fact, such proofs may be provided, if something more is proven than mere admissibility. It is general admissibility of rules, i.e., an admissibility independent of the special choice of the basic calculus, which is accessible even in the case of meta-calculi, yet without a chance to substantiate the possible claim of having found all generally admissible rules. On the basis of ‘operationist truth’ (of a subjunction) which is proven by a proto-logical proof of admissibility (of the corresponding rule) logical truth in operationist logic gets defined essentially by general admissibility. It was the main result of operationist logic that a calculus for deriving the

Basic Objectives of Dialogic Logic in Historical Perspective

35

logically true subjunctions A→ B that have thus been gained, is equivalent with some calculus of intuitionistic logic. But apart from the impossibility to give a precise meaning to a completeness claim of some class of logically true subjunctions with respect to the interpretation of logical truth by general admissibility, essentially two difficulties of the operationist approach to logic remained unsolved and became the reasons for embarking on dialogue games as a more satisfactory approach to logic. With the introduction of meta-calculi the existence of a decidable predicate, like ‘is a derivation’ with respect to an ordinary calculus, as the backbone for the concept of logically compound proposition became an open question again. It was Alfred Tarski who, in discussions with Lorenzen in 1957/58, when Lorenzen had been invited to the Institute for Advanced Study at Princeton, convinced him of the impossibility to characterize arbitrary (logically compound) propositions by some decidable generalization of having a decidable proof-predicate or a decidable refutation-predicate. Not even in the special context of the operationist approach does it work just to combine the ideas of Rudolf Carnap on the one hand and of Karl Popper on the other hand, according to which propositions are characterized by a principle of verification (i.e., to each proposition there is an associated procedure which defines a decidable concept of proof) or by a principle of falsification (i.e., to each proposition there is an associated procedure which defines a decidable concept of refutation). Hence, Lorenzen’s attempt of an inductive definition of ‘definite’ in order to find a characterization of propositions which relinquishes the synonymy of ‘definite’ and ‘decidably definite’ had to be accepted as inappropriate.6 Especially subjunctions in the operationist interpretation could not any more be called ‘definite’ as it had been the explicit intention of the operationist approach. The other difficulty connected with the operationist approach and unsolvable within this framework is an obvious one. Propositions in general cannot be reduced to derivability propositions as their primary constituents. Besides, due to the introduction of ‘ a→ b’ by ‘(a⇒b) is admissible’ with respect to some calculus K, it is impossible to define subjunctions for propositions in general. It became necessary to search for some decidable predicate which may be used to qualify a linguistic entity as a proposition about any domain of objects, be it elementary or logically compound. Decidability is essential here, because the classical characterization of a proposition as an entity which may be true or false, has the awkward consequence that of an undecided proposition it is impossible to know that it is, in fact, a proposition. This observation gains further weight by L. E. J. Brouwer’s discovery that even on the basis of a set of ‘value-definite’, i.e. ____________ 6

Cf. Lorenzen 1955, pp. 5–6.

36

Philosophical Logic and Philosophy of Language

decidably true or false, elementary propositions, logical composition does not in general preserve value-definiteness.7 And since neither the property of being proof-definite nor the one of being refutation-definite nor properties which may be defined using these two, are general enough to cover the case of an arbitrary proposition, some other procedure had to be invented that is both characteristic of a proposition and satisfies a decidable concept. The concept looked for and at first erroneously held to be synonymous with argumentation (as in paper I.1) turned out to be the concept of dialogue about a proposition A (which had to replace the concept of truth of a proposition A as well as the concepts of proof or of refutation of a proposition A, because neither of them can be made decidable).8 Fully spelled out it means that for an entity to be a proposition there must exist a dialogue game associated with this entity, i.e., the proposition A, such that an individual play of the game where A occupies the initial position, i.e., a dialogue D(A) about A, reaches a final position with either win or loss after a finite number of moves according to definite rules: The dialogue game is defined as a finitary open two-person zero-sum game. Thus, propositions will in general be dialogue-definite, and only in special cases be either proof-definite or refutation-definite or even both which implies their being value-definite. Within this game-theoretic framework where win or loss of a dialogue D(A) about A is in general not a function of A alone, but is dependent on the moves of the particular play D(A), truth of A is defined as existence of a winning strategy for A in a dialogue game about A; falsehood of A, respectively, as existence of a winning strategy against A. Winning strategies for A count as proofs of A, and winning strategies against A as refutations of A. The meta-truth of “either ‘A is true’ or ‘A is false’” that is provable only classically by means of the saddle-point theorem for games of this kind 9 may constructively be reduced to the decidability of win or loss for individual plays about A. The concept of truth of dialogue-definite propositions remains finitary, and it will, as it is to be expected of any adequate definition of truth, in general not be recursively enumerable. The ____________ 7 8

9

Cf. Brouwer 1925. First steps towards establishing this concept were made by Lorenzen in talks given at Venice in 1958 – cf. Lorenzen 1960 – and at Warsaw in 1959 – cf. Lorenzen 1961 –; both papers are reprinted in Lorenzen/Lorenz 1978, pp. 1–8 and 9–16, respectively. For a clarification of the distinction between the meaning-related concept of dialogue and the validity-related concept of argumentation that refers to the level of strategies and not to the level of plays like the concept of dialogue, cf. Lorenz 1982. In this paper dialogues are still identified with argumentations, whereas the procedures on the level of strategies run under the label of justification. Cf., e.g., Berge 1957.

Basic Objectives of Dialogic Logic in Historical Perspective

37

same holds for the concept of falsehood which is conspicuously defined independently of negation. All these stipulations have been made without any recourse to the particular structure of a proposition. And it was, indeed, the original intention of dialogic logic furthermore to separate clearly the general rules which are independent of the kind of propositions in initial position of a possible dialogue – the structural rules – from the special rules pertaining to the internal structure of a proposition which, in the case of an internal structure by logical composition, are called particle rules. The particle rules define the ‘local’ meaning of the respective logical connectives, each individually, whereas the structural rules add to their ‘global’ meaning collectively. In order to facilitate the discussion of some further original objectives of dialogic logic it is advisable to state one precise version of structural rules, in fact the one which defines intuitionistic truth and falsehood: S1 Dialogues about propositions consist of finitely many moves which are put forth alternately by an opponent O and a proponent P. The moves obey certain special rules that belong to the game such that each play ends up with win or loss for either player. S2 With exception of the improper initial move by P, each move either attacks prior ones of the other player or defends those of one’s own upon such an attack, but does not act simultaneously in both ways: The proper moves split into attacks and defenses. S3 Attacks may be put forth at any time during a play of the game. S4 Defenses must be put forth in the order of the corresponding attacks (upon which the defense answers), yet may be postponed as long as attacks can still be put forth; always that move which has been attacked last without having been defended yet, has to be defended first. S5 Neither player must defend upon an attack, unless this attack has been defended upon finitely many counterattacks. S6 During a play of the game the sequence of attacks against one single move must be correlated with a strict descending sequence of (constructive) ordinals, α (by the opponent) and β (by the proponent, where α and β are the bounds of attack which are chosen by O and P in that order after the initial move. (This indicates the following duty: If ζn is the nth attack against a move such that (a) the ordinal γ* (< α resp. < β) of the (n - 1)th attack ζn-1 against the same move is the successor of an ordinal γ, so the ordinal assigned to ζn is γ; (b) the ordinal γ (< α resp. < β) of the (n - 1)th attack is a limit ordinal, so the attacking player must choose an ordinal λ<γ as the ordinal assigned to ζn.) S7 Whoever cannot, or will not, put forth a move any more has lost that play of the game; the other player has won it.

It should first be remarked that S7 adds the speech act of asserting to the moves of a dialogue, the other rules essentially being responsible for the

38

Philosophical Logic and Philosophy of Language

meaning of the proposition in initial position. As a second remark it should be noted that the choice of just two types of moves as stated in S2, i.e., attacks and defenses upon attacks, is intimately connected with S5, the crucial dialogue-condition. Both rules try to make a dialogue situation appropriate for rendering unconditional as well as conditional assertion of propositions. In fact, they ensure that assertion of A under the condition B may be treated as equivalent with unconditional assertion of A→B. For this purpose it is sufficient to stipulate the following well-known particle rule: Any subjunction A→B may be attacked by making the move A, and A→B may be defended upon the attack A by making the move B. Then, dialoguecondition S5 when applied to a move A→B states that the obligation to defend A→B with B upon the attack A shall not arise prior to A’s being defended upon finitely many counterattacks against A. In order to uphold the finitary character of the game, together with the first attack against some move the attacking player has to choose a finite natural number as his bound of attack against that move. The finite sequences of choices of upper limits of attacks against a single move by each player within a particular dialogue may be represented as initial choices of two ordinal numbers according to S6. Stipulations which concern rank and order of attacks and defenses are put down in S3 and S4. In particular, attacks are treated as rights independent of any special position reached during a single play of the game and, hence, cannot become void throughout that play. Defenses, on the other hand, are treated as duties which are not to be cancelled at any time in a play, though they are subject to the permission to postpone their performance as long as rights of attack are available. Finally, there is a ‘last-duty-comes-first’-clause which has aroused considerable suspicion that it had been invented just for the purpose of justifying intuitionism dialogically. Yet, closer scrutiny reveals that S5 together with the natural rule of making moves alternately (S1) cannot be enforced unless the ‘lastduty-comes-first’-clause is in force, too. The semi-formalism for deriving all and only propositions for which there is a winning strategy – and likewise the formalism, i.e. the calculus, for deriving all and only propositions for which there exists a formal winning strategy – mirrors some of the structural rules in rules of the semiformalism resp. the formalism which, since Gerhard Gentzen, bear almost the same name: ‘Struktur-Schlußfiguren’10. As we are concerned, here, with the intuitionistic dialogue game, the structural rules of thinning, contraction and interchange in the antecedent, only, are relevant. Thinning in the antecedent is a consequence of S3 with respect to P, contraction in ____________ 10 Gentzen 1969, p. 17.

Basic Objectives of Dialogic Logic in Historical Perspective

39

the antecedent is a consequence of S6 with respect to P, and interchange in the antecedent is a consequence of the fact that for exercising the right to attack no ordering of these rights has been imposed. Now, it is useful to look a bit closer into S5 as applied to subjunctions. The dialogue-condition is a recursive clause that occupies an intermediate position between two extremes which either violate the finitary character of playing or make conditional reasoning impossible. The extreme versions of S5 are in both cases not even consistent with the remainder of the structural rules. The intermediate position of S5 is furthermore essential for drawing a particular consequence of the dialogue rules with respect to subjunctions which connects the dialogic approach with the operationist appoach in the originally intended way. It can be shown that the (material) equivalence of ‘(a⇒b) is admissible in K’ with ‘ K a→ K b’ holds, an equivalence which, for want of an independent and unrestricted introduction of subjunctions, had been established by definition in operationist logic. The two extreme versions of S5 with respect to subjunctions are the following: (1) The obligation to defend A→B with B upon the attack A shall not arise prior to A’s being true, i.e., a whole winning strategy for A has to be presented before the player of A→B is obliged to defend with B. (2) The obligation to defend A→B with B upon the attack A arises if and only if A is not counterattacked, i.e., unless a counterattack against A does not cancel the duty to defend with B, move B has to be made just after move A has been made; in short: each player must choose between defending upon an attack and counterattacking against that attack.

Case (1) would turn A→B into the meta-proposition “if ‘A is true’ then ‘B is true’”. The concept of dialogue game would be changed into a nonfinitary one as soon as winning strategies for A could not be found by finite procedures. Case (2) would make A→B dialogue-equivalent with ¬A ∨ B, i.e., both propositions may be substituted for each other at any place in a particular dialogue without affecting win and loss. In this case the specific role of ‘if-then’ would be eliminated. With respect to the concept of formal winning strategy, case (1) makes the concept inapplicable, and case (2) makes the concept empty. In both cases there is no way to go from material truth to logical truth besides the usual classical procedure of defining logical truth as general material truth. Yet, it was exactly this new instrument of defining formal winning strategies and with it formal truth – logical truths will be special formal truths11 – which opened a new road to logic when understood as the theory ____________ 11 Cf. for the distinction, e.g., of formal-arithmetical truth from formal-logical truth, Lorenz 1996.

40

Philosophical Logic and Philosophy of Language

of logical truth or logical implication together with their application to logical reasoning, beside the two main ones of model theory and proof theory. It makes a difference to observe that, e.g., A → A is true for arbitrary, but of course dialogue-definite, propositions including those, for or against which no winning strategy is available, and that A → A is true for true propositions A as well as for false ones: Meanwhile it is wellknown that formal truth and general material truth coincide not even on the domain of value-definite propositions. On the basis of the observation that for certain logically compound propositions, like ((a→b)→b)→((b→a)→a), for example, there is no formal winning strategy, though there are winning strategies for it in any one of the four possible cases of a and b being true or false – the sequence of moves which leads to a winning position for P will be different in the four cases – the question was raised whether it can be proven that against propositions of this form no winning strategy will exist, i.e., that they cannot be false even though they are not formally true. The answer is ‘yes’: By changing the structural rules in such a way that moves at a particular stage of a play which originally cannot be made within the same play of the game, because you had to choose among the alternatives, may indeed be made within the same play, you can prove that the concept of logically true proposition in the sense of classical logic coincides with the concept of proposition which is formally not intuitionistically false. The dialogue approach to logic permits to interpret the predicate ‘logically true in the sense of classical logic’ – and, with certain further qualifications, also the corresponding predicates with respect to the intermediate logics between intuitionistic and classical logic – as ‘formally not intuitionistically false’.12 Beyond finding different structural rules with a common core of particle rules such that the set of formally true propositions matches certain logics already under discussion, it was part of the original objective of dialogic logic to exhibit and to evaluate the relations of intuitionistic dialogue games with other dialogue games as well as with other approaches to logical reasoning. Yet, apart from this, the concept of dialogue-definiteness will always have to belong to the rock bottom of dialogic logic.

____________ 12 Cf. Lorenz 1978, pp. 48–49.

Basic Objectives of Dialogic Logic in Historical Perspective

41

References Berge, Claude, 1957: Théorie générale des jeux à n personnes, Paris: Gauthier-Villars. Brouwer, Luitzen E. J., 1925: Über die Bedeutung des Satzes vom ausgeschlossenen Dritten in der Mathematik, insbesondere in der Funktionentheorie [1923], in: Journal für reine und angewandte Mathematik 154, pp. 1–7. Curry, Haskell B., 1951: Outlines of a formalist philosophy of mathematics, Amsterdam: North Holland. Fränkel, Abraham A./Bar-Hillel, Yehoshua/Levy, Azriel, 21973: Foundations of Set Theory [1958], Amsterdam: North-Holland. Gentzen, Gerhard, 1969: Untersuchungen über das logische Schließen [1934], Darmstadt: Wissenschaftliche Buchgesellschaft. Lorenz, Kuno, 1978: Arithmetik und Logik als Spiele, in: Paul Lorenzen/ Kuno Lorenz, Dialogische Logik, Darmstadt: Wissenschaftliche Buchgesellschaft, pp. 17–95. Lorenz, Kuno, 1982: Die irreführende Gleichsetzung von Begründungen und Argumentationen, in: Logik und Pragmatik. Zum Rechtfertigungsproblem logischer Sprachregeln, ed. by Carl Friedrich Gethmann, Frankfurt am Main: Suhrkamp, pp. 78–91. Lorenz, Kuno, 1996: Dialogspiele und Syntax, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/ Manuel international des recherches contemporaines, ed. by : Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1380–1391. Lorenzen, Paul, 1955: Einführung in die operative Logik und Mathematik, Berlin/Göttingen/Heidelberg: Springer. Lorenzen, Paul, 1960: Logik und Agon, in: Atti del XII Congresso di Filosofia (Venezia, 12–18 Settembre 1958). v 4, Firenze: Sansoni Editore, pp. 187–194. Lorenzen, Paul, 1961: Ein dialogisches Konstruktivitätskriterium, in: Infinitistic Methods. Proceedings of the Symposium on Foundations of Mathematics (Warsaw, 2–9 September 1959), Oxford/London/New York/Paris: Pergamon Press, pp. 193– 200. Lorenzen, Paul/Lorenz, Kuno, 1978: Dialogische Logik, Darmstadt: Wissenschaftliche Buchgesellschaft.

Pragmatic and Semiotic Prerequisites for Predication A dialogue model Rather than starting my presentation of how to construe predication with a discussion about propositions, and independent of the ongoing debate whether propositions should be understood as propositional kernels of full sentences (having a force, expressing a thought, and denoting a truthvalue) or should themselves be considered as a particular force of a sentence, i.e., its constative force as an assertion, I would like to invite you to a journey into the prepropositional state where the task to provide for propositions or sentences is still to be accomplished. The primary means for changing one’s state, or for realizing that such a change has occurred, I consider to be a dialogue, better still: a dialogue situation. This is a situation where, using a Peircean term, a ‘habit change’ occurs which should be construed as the acquisition of an action competence.1 It turned out that such an acquisition procedure will most profitably be modelled using the conceptual frame of a two-person-game, and such a game may be considered as a generalized Wittgensteinean language-game or, rather, its pragmatic basis, not yet with an explicit linguistic activity. In the beginning the game is not an object of study but a means of study. Before giving a sketch of the dialogical constructions,2 which lead from modelling simple activity to modelling the growth of more complex activities up to elementary verbal utterances, at first a few general remarks in order to put my suggestions into a broader perspective.

I In accordance with Charles Sanders Peirce, I consider pragmatics to have become the modern heir of ontology with semiotics being its counterpart as the modern heir of epistemology. Yet, in this context both disciplines should not be understood as two newly established empirical sciences, but as ways of investigation where empirical procedures are combined with ____________ 1 2

Cf. Lorenz 1994, in this volume pp. 56–61. For further details, cf. Lorenz 1996; 1997; 2000 (all three contained in Lorenz 2009).

Pragmatic and Semiotic Prerequisites for Predication

43

reflexive procedures. Using such a broader perspective both actions and sign-actions are not only treated as objects of research and representation, as, e.g., in Charles Morris’ and Umberto Eco’s approach, but also as a means or tool of research and representation. You not only observe and describe these entities according to certain standards, but you also produce them in a perspicuous fashion in order to arrive at some kind of approximating reconstruction of what you take to be available, already. Wittgenstein has used the term ‘language-game’ for this kind of activity which aims at disclosure of what is going on by providing tools of comparison, though in his description of language-games pragmatic and linguistic activity is not accounted for by separate steps. Hence, the constructions serve cognitive purposes in the sense of delineating the very areas of (particular) objects one proceeds afterwards to investigate in the more usual way. Language-games as well as the generalized ones of acquiring simple action competences have to count as paradigm cases of perceptual knowledge, because they exhibit a semiotic function if understood as icons in the sense of Peirce. An area of internally structured objects is found by inventing a prototype. It should be obvious, therefore, that even the distinction of action and sign-action – a special case of the basic and embarrassing distinction between world and language – which still is prevalent in Wittgensteinian language-games where simple action competence is presupposed, has to be relativized in view of a purely functional account of both what it means to be an object and what it means to be a sign (of an object). In fact, it belongs to one of the basic tenets of, e.g., Nelson Goodman’s approach that the seemingly clear-cut division of world and language – non-verbal language included – as a division between the given and the constructed, between that which is found and that which is made, between the fact and the artefact, is outdated, and that it has even been challenged once and again since the time of the pre-Socratics. But, only rarely is history looked at in this way. Any matter we are concerned with, Goodman tells us, is dependent on some manner as the means by which we deal with it. So worlds are but versions and worldmaking begins with one version and ends with another. The message we should learn runs thus: “never mind mind, essence is not essential, and matter doesn’t matter”.3 Goodman goes on in claiming that we choose the facts as much as the frameworks, though this statement should better be split into two complementary statements: We produce the facts as much as the frameworks and we experience the frameworks as much as the facts. Constructions, when serving cognitive purposes, are always reconstructions. ____________ 3

Goodman 1978, p. 96.

44

Philosophical Logic and Philosophy of Language

The last two Aristotelian categories, ποιεῖν and πάσχειν, which seemed forgotten throughout most of modern philosophy in the tradition of Descartes, though they play an important role both in Spinoza and Leibniz, will enjoy a lively comeback as the two sides we are concerned with when doing something: you do it yourself (active) and you cognize that others (including yourself!) are doing the same (passive [with respect to the content of cognition]). In fact, the two sides reoccur in the model of an elementary dialogue situation with two agents being engaged in the process of acquiring an action competence. At each given instant just one of the agents is active – a ‘real’ agent – and the other agent – the ‘potential’ agent or ‘patient’ – is passive. The agent in active role is performing an action, i.e., he is able to produce different tokens of the same type, while the agent in passive role is cognizing an action, i.e., he sees different tokens as belonging to the same type. One has learned an action, if one is able to play both roles: While acting you know what you are doing, or, conversely, if you don’t know what you are doing, you don’t act. Another way of saying this would be: Each action appears in two perspectives, in the Iperspective by performing the action (= producing an action token) – it should be called the pragmatic side of an action, or its ‘natural’ side – and in the You-perspective when cognizing the action (= witnessing an action type) which should be called its semiotic or ‘symbolic’ side. We have come across the first step to execute the program of ‘naturalizing language’ and other symbol systems, and, at the same time, of ‘symbolizing world’, in order to bridge the gap between the two. For further guidance we may turn to Peirce again. He sketches a way of deriving signs out of objects in more or less the same manner as I just did, the difference being that he proceeds upside down. If a Sign is other than its Object, there must exist, either in thought or in expression, some explication or argument or other context, showing how – upon what system or for what reason the Sign represents the Object or set of Objects that it does. Now the Sign and the Explanation together make up another Sign, and since the Explanation will be a Sign, it will probably require an additional explanation, which taken together with the already enlarged Sign will make up a still larger Sign; and proceeding in the same way, we shall, or should, ultimately reach a Sign of itself, containing its own explanation and those of all its significant parts; and according to this explanation each such part has some other part as its Object. According to this every Sign has, actually or virtually, what we may call a precept of Explanation according to which it is to be understood 4 as a sort of emanation, so to speak, of its Object.

The argument of Peirce calls for something which is a sign of itself, that is, which combines object status and sign status, or better: which functions in ____________ 4

Quoted from Meaning [1910]: CP, 2.230.

Pragmatic and Semiotic Prerequisites for Predication

45

both ways. The basic point of his pragmatic foundation of semiotics was to give an account of the process of separation between sign and its object within the framework of his Pragmatic Maxim.5 And the arguments used for this purpose are themselves to be understood as sections of an open sign-process on the level of reconstruction. And it is these sections that may be looked at as conceptualizations of generalized Wittgensteinian language-games. Now, the descending sequence of interpretants ends with an ultimate logical interpretant6 which is identified as a habit change. As stated in the beginning, already, such a habit change, in contemporary terminology, is nothing else but the acquisition of an action competence such that all the ways of dealing with the object in respect of what is signified by the initial sign are included. And within the process of acquisition, if it is modelled as an elementary dialogue situation, the two perspectives of agent and patient may count, respectively, as action on the object-level in performing the action, and as action on the sign-level in cognizing the action through the performance functioning as a representative of any other performance. Thus, habit changes are, indeed, entities which are signs of themselves. We may finally conclude that a verbal sign of an object signifies a range of possibilities of dealing with that object. Even more generally, by deleting the dummy term ‘object’, it might be said that having the competence for such a sign-action – being a verbal signaction, it functions symbolically – is tantamount to ‘knowing’, by that very action, of a whole range of further actions which may be said to be signified by the (symbolic) sign-action. We arrive at the following equivalence: Knowing an action, in the sense of being acquainted with it, is knowing ways of dealing with that action. And this implies that knowing an arbitrary object is equivalent to treating this object as a sign of its distinctions, i.e., of its internal structure which is exhibited step by step in an open sign-process.

II Now, within the model of acquisition of an action competence by an elementary dialogue situation it is important to make some further distinctions. They are based on the observation that producing an action-token and witnessing an action-type, i.e., I-perspective and You-perspective of an action, are inseparably bound together and cannot be treated in isolation ____________ 5 6

Cf. Scherer 1984. Cf. CP, 5.476.

46

Philosophical Logic and Philosophy of Language

from each other. The model of acquisition of action competence is a model of actions as a means and not yet of actions as objects which, in order to be accessible, will in turn be dependent on other actions as a means of dealing with objects. Dialogical construction as a means of study asks for self-application such that the interdependence of the status of being-a-means and the status of being-an-object, hence of ‘epistemology’ and ‘praxeology’ on the one hand, and of ‘ontology’ on the other hand, is laid bare. Actions as a means are characterized by their two sides as they arise from the two perspectives, from singular performance in I-perspective and universal cognition in You-perspective. Yet, when performing is understood to be a case of producing (an action-token) and, analogously, cognizing to be a case of witnessing (an action-type), the action in question is treated as an object, in fact, sometimes even as two objects, the token as an external or ‘corporeal’ particular and the type as an internal or ‘mental’ particular. But, even if action particulars, i.e., individual acts, are treated uniformly without being split into external and internal entities, particularity is to be kept strictly distinct from singularity and universality. Usually, in the terminology of type and token, where types are treated logically as generated ‘by abstraction’ out of tokens, and where tokens originate ‘by concretion’ from types (rather than looking at the relation of tokens to types in a psychological fashion as a relation of external to internal particulars), both types and tokens are (individual) objects, yet of different logical order, which are related in standard notation as sets to their elements. At the lowest level, if there is one, the final universe of discourse is located, i.e., a world of elementary individual objects, the particulars, to which everything else will have to be reduced. Such an account, by neglecting the distinction between particularity on the one hand and singularity as well as universality on the other hand, violates the inseparability of (producing a) token and (witnessing a) type in the context of actions as a means, or, rather, it exhibits an equivocation in the use of ‘type’ and ‘token’. It is necessary to relinquish both the equivalence of ‘performing an action’ with ‘producing an action token’ and the equivalence of ‘cognizing an action’ with ‘witnessing an action type’. Instead, performation is performation of something singular and cognition is cognition of something universal, whereas producing (a token) together with its twin activity of witnessing (a type) occur with respect to something particular. Now, if types and tokens are not in this way construed as particulars that are produced or witnessed, respectively, they should be identified, in tune with action as a means, with universal features and singular ingredients of particulars that are exhibited by actions which deal with them. Particulars together with the situations (of acting) of which they occupy the foreground are appropriated by performing an action

Pragmatic and Semiotic Prerequisites for Predication

47

which deals with them, and they are detached by cognizing such an action. It should be noted that neither universal features nor singular ingredients have object status by themselves; they remain means with respect to (particular) objects. Universals cannot be appropriated and singulars cannot be detached. Hence, in performances of an action that is dealing with a particular you (pragmatically) present one of the (singular) token ingredients of this particular, whereas in cognitions of an action that is dealing with a particular you (semiotically) represent one of its (universal) type features. Switching from the language of means – pragmatic means are singular, semiotic means are universal – to the language of objects (= particulars) you may say that it is individual acts that provide both services, of presentation with respect to its performance perspective and of representation with respect to its cognition perspective. With recourse to a traditional terminology it may be said, in a ‘mentalistic’ version, that an individual act has been ‘aimed at’ in a singular performance and will ‘originate’ from a universal cognition, but it could as well be said, in a ‘naturalistic’ version, that an individual act was ‘caused’ by a singular performance and is ‘conceived’ by a universal cognition. In appropriation as well as in detachment of particulars of arbitrary category, like individual acts, individual things or events, groups of individuals or other non-individual particulars, etc., the actions of dealing with particulars are used as a means, of presentation (of singular tokens – the way a particular is present) in the case of appropriation, and of representation (of universal types – the way a particular is identified) in the case of detachment. Particulars may be said to act as appearances of ‘substances’, i.e., some part of the whole out of like singular tokens is a part of the particular, and as carriers of ‘properties’, i.e., the particular is an instance of a universal type.7 Therefore, in order to avoid misunderstandings, instead of ‘perform’ we will, henceforth, say ‘actualize’, and we say ‘schematize’ instead of ‘cognize’. Within the model of an elementary dialogue situation where two agents are engaged in the process of acquiring an action competence, the activities of actualizing and schematizing should not be understood as performances of two separate actions; it is one action the competence of which is acquired by learning to play both the active and the passive role. Active actualization makes the action appear in I-perspective, passive schematiza____________ 7

A particular wooden chair, for example, acts as a carrier of all the properties conceptualized by ‘wooden’, and as an appearance of the substance ‘wood’, inasmuch as a part of ‘the whole wood’ may be considered to be part of the particular wooden chair, cf. Lorenz 1996a, pp. 225–228.

48

Philosophical Logic and Philosophy of Language

tion lets it appear in You-perspective. Any action as a means is characterized by its pragmatic and its semiotic side, and it does not make sense as yet to speak of the action as an ‘independent’ object(-type) split into particulars, i.e., some set of individual acts. In order to achieve the switch from action as a means to action as object, it is essential to iterate the process of acquiring an action competence by turning the two sides of an action into proper actions by themselves, i.e., into actions of dealing with the original action under its two perspectives such that the (secondary) action competences additionally required will have to be modelled in turn by means of (now non-elementary) dialogue situations. Such a further step may be looked at as an application of the principle of self-similarity. What has to be done is to schematize and to actualize the elementary dialogue situation, i.e., to create a He/She-perspective towards the I/Yousituation such that, on the one hand, He/She becomes a (secondary) Youperspective with respect to I/You as I, and, on the other hand, He/She becomes a (secondary) I-perspective with respect to I/You as You. In the first case you gain an ‘exterior view’ of the original action by acquiring a second level action (with respect to the original action) which functions as one of the indefinitely many aspects of the original action: The You-perspective is turned into the schema of a second level action out of an indefinite series of second level actions. In the second case you gain an ‘interior view’ of the original action by acquiring a second order action (with respect to the original action) which functions as one of the indefinitely many phases of the original action: The I-perspective is turned into an actualization of a second order action out of an indefinite series of second order actions. The semiotic side of an action is split into a multiplicity of aspects or (secondary) You-perspectives, and the pragmatic side of an action likewise into a multiplicity of phases or (secondary) I-perspectives. By (dialogical) construction, it is in its active role that an aspect-action is I-You-invariant and, in this sense, ‘objective’, whereas a phase-action is I-You-invariant in passive role, only. Hence, by applying the principle of self-similarity once again to both aspects and phases, the pragmatic side of an aspect-action is split into a multiplicity of objective articulations or sign-actions, and the semiotic side of a phase-action into a multiplicity of objective mediations or partial actions. Any one of the sign-actions is a means to designate the original action, and any one of the partial actions is a means to partake of the original action, where designating and partaking function with the proviso that the original action itself is turned from a means into an object. In fact, an action as object – things, events, and other categories of entities are included among actions by identifying an entity[type] with the action[-type] of dealing with the entity – is constituted, on the one hand ‘formally’, by identification of the schemata of the aspect-

Pragmatic and Semiotic Prerequisites for Predication

49

actions, i.e., of their ‘subjective’ semiotic side (when turned into a multiplicity of full-fledged actions, one would get perceptual actions), and, on the other hand ‘materially’, by summation of the actualizations of the phase-actions, i.e., of their ‘subjective’ pragmatic side (when turned into a multiplicity of full-fledged actions, one would get poietic actions). On the one side, through identification, an action as object is a semiotic (abstract) invariant of which one partakes by means of a partial action, and on the other side, through summation, it is a pragmatic (concrete) whole which one designates by means of a sign-action. With respect to the additional dialogue situations modelling the acquisition of second-order action competences as well as second-level action competences the original action as object occurs within a situation which, in fact, is responsible for individuating the original action as object. The move of objectivation from action as a means to action as object is accompanied by a split of the action into (action-)particulars such that the respective invariants may be treated as kernels of the schemata of aspects (= universalia), and the respective wholes correspondingly as closures of the actualizations of phases (= singularia). An objectival foreground together with a situational background will semiotically be a constant foreground against a variable background (= the same particular in different surroundings, i.e., its varying external structure), and it will pragmatically be a variable foreground against a constant background (= different particulars [of the same kind] by their varying internal structure, in the same surrounding). Both together, kernel and closure – ‘form’ and ‘matter’ in philosophical tradition8 – make up a particular within a situation. Working backwards again, i.e., making the countermove of appropriation, the schemata of the kernel and the actualizations of the closure, are realized in representations and presentations, respectively, by schematizing and actualizing a particular (in a situation) as explained above. Dialogical construction of particulars being dependent on identification of schemata of aspects and on summation of actualizations of phases, implies the establishment of mutual independence between objectival foreground and situational background. In order to achieve this, a specially chosen articulation has to act as a substitute for arbitrary aspects with respect to some partial action – such a function of substitution may be articulated by rules of translation among aspects – and will be called symbolic articulation. Constant foreground and variable background will thus become independent of ____________ 8

The treatment of particulars as mixta composita (σύνθετα) out of εἶδος or forma, and ὕλη or materia, is due to Aristotle as explained, e.g., in the commentary of Alexander of Aphrodisias on Aristotle’s Metaphysica, cf. CAG I, pp. 545, line 30ff; 497, line 4ff.

50

Philosophical Logic and Philosophy of Language

each other. Analogously, any mediation will have to acquire the function of having the phase to which it belongs extended by arbitrary other phases with respect to some sign-action – such a function of extension may be articulated by rules of construction for phases – and will be called comprehensive mediation. In this case, constant background and variable foreground are made independent of each other. Both constructions together guarantee that particulars contrast with their surroundings.9 By symbolic articulation that is a symbolic sign-action, you arrive at a semiotically determined particular in actualized situations, i.e., the particular is symbolically represented, whereas by comprehensive mediation that is a comprehensive partial action, you arrive at pragmatically determined particulars in a schematized situation, i.e., the particulars are symptomatically present. The semiotic side of partial actions (‘what you do’) and the pragmatic side of sign-actions (‘how you speak’), together they make up the ways of life (of the agents). Correspondingly, the pragmatic side of partial actions (‘how you act’) and the semiotic side of sign-actions (‘what you say’), together they make up the world views (of the agents).

III Articulation is signified canonically by the result of a sign-action, an articulator (= ‘signifiant’ in the sense of F. de Saussure); articulation has a pragmatic side, i.e., it is a sign-action in its being an action, and a semiotic side, i.e., it is a sign-action in its being a sign. Semiotically, articulation is effected by uttering an articulator that has to be taken as a (verbal) type, in a speech situation; and if it is treated as functionally equivalent with any other way of articulation, including non-verbal ones, it acts as a symbolic articulator. Again semiotically, i.e., as a sign(-action), it shows its two sides, a pragmatic one and a semiotic one. The pragmatic one is to be called communication, or the side with respect to persons or subjects, and the semiotic one is to be called signification, or the side with respect to particulars or objects (these two sides in their function being reminiscent of Plato’s λέγειν and ὀνοµάζειν). By iteration, communication splits into (content of) predication on the semiotic side, and mood (of predication) on the pragmatic side, whereas signification splits into (intent of) ostension on the pragmatic side, and mode of being given on the semiotic side. Any predication can take place only by using a mood, and any ostension is effected ____________ 9

For an explicit dialogical construction of both identification and summation, cf. Lorenz 1997, pp. 145–147.

Pragmatic and Semiotic Prerequisites for Predication

51

only by using a mode of being given. We have strictly to distinguish: content and mood of predication, intent and mode of ostension. The moods of predication are, of course, speech acts, and only with respect to a mood a predication contains a claim, e.g., a truth claim. Articulation of a mood of predication yields performators on the semiotic side of the articulation, whereas articulation of a mode of ostension yields perceptions (= Wahrnehmungsurteile) on the pragmatic side of the articulation. Without such a second order articulation of mood and mode, we have arrived at one-word sentences ‘P’ (in a mood and using a mode of being given) by uttering the articulator ‘P’. With the next step we introduce the separation of significative and communicative function, two functions that coincide with showing and saying in the terminology of Wittgenstein’s Tractatus. Separation with respect to predication, i.e., the semiotic side of communicative function, yields: δPεP (this P [= something done] is P[-schematized]), or, alternatively, σPπP (the universal P [= something imagined] is P-actualized), whereas separation with respect to ostension, i.e., the pragmatic side of significative function, yields: δPζP (this P belonging to P), or κPξP (the whole P [= something intuited] being P-exemplified). The operators: demonstrator ‘δ’ and attributor ‘ε’ (= copula), respectively, neutralize the communicative function and the significative one; ‘δ’ keeps the significative function and ‘ε’ the communicative one, with the result that ‘δP’ plays a singular role and ‘εP’ a universal one. In the terminology of logic or semiotics, ‘δP’, which is used to ‘ostend’ P, is an index of an actualization of the action articulated by ‘P’, whereas ‘εP’, which is used to ‘predicate’ P, is a predicator serving as a symbol of the schema of action P. Predication εP and ostension δP with its respective associates: form of a proposition ‘__εP’ and form of an indication ‘δP__’ , are the modern equivalents of the traditional ‘forms of thinking’ and ‘forms of intuition’. Proceeding dually with respect to the predication oriented and, hence, semiotic distinction ‘singular-universal’, it is also possible to use another pair of operators, universalisator ‘σ’ and presentator ‘π’, where ‘σP’ has only significative function with universal role and ‘πP’ only communicative function with singular role. In the second case which works with respect to the ostension oriented and, hence, pragmatic distinction ‘active-passive’, either demonstrator ‘δ’ and partitor ‘ζ’, or, dually, totalisator ‘κ’ and exemplificator ‘ξ’, serve the same purpose: ‘δ’ and ‘κ’ keep the significative function in active and passive role, respectively, whereas ‘ζ’ and ‘ξ’ keep the communicative function, here in passive and active role, respectively. What is not yet available up to now and what would not even make sense, are ‘propositions’ of kind δPεQ and ‘indications’ of kind δQζP. The

52

Philosophical Logic and Philosophy of Language

reason why these expressions do not make sense, is simply the following: ‘δP’ is not the kind of expression to occupy the empty place in a propositional form ‘__εQ’ with Q≠P, and ‘ζP’ is not the kind of expression to occupy the empty space in an indicational form ‘δQ__’ with Q≠P. Instead, we have to introduce individuators ‘ιP’ in order to refer to particulars, i.e., the situation-dependent units of the action articulated by ‘P’; ‘things’ as well as objects of other categories, any one (type) of them being identified with the action(-type) of arbitrary dealings with an object(-type), hence, any of the so-called ‘natural kinds’, are, of course, included among the P. Particulars, be they individual things or events, individual acts or processes, are composed out of kernels of schemata of aspects: σ(ιP) (= invariants), together with closures of actualizations of phases: κ(ιP) (= wholes). Hence, particulars may be considered to be half thought and half action. Using individuators we, now, may write down eigen-propositions ιPεP as well as eigen-indications δPιP (short for: δPζιP), and it is possible to render these versions of saying and showing with the help of the four operators: demonstrator, attributor, universalisator, and totalisator, in the following traditional way: 1. In the case of saying (ιPεP): the universal σP is predicated of a P-particular by means of ‘εP’ (or: within the proposition ιPεP, the individuator is a sign of an indication, and, hence, functions as a nominator of a P-particular, i.e., within the proposition ιPεP, nomination by ‘ιP’ is shown), and 2. In the case of showing (δPιP): ostending the whole κP at a P-particular by means of ‘δP’ (or: within the indication δPιP, the individuator is a sign of a proposition, and, hence, functions to say that participation at a P-particular holds, i.e., within the indication δPιP, participation at ιP is said).

Hence, reference to particulars ιP includes both nomination of κ(ιP), i.e., of the matter of ιP, and participation at σ(ιP), i.e., at the form of ιP. As a remark, it may be added that nominating is the articulation of designating by symbolic articulation, and, analogously, participating is the articulation of partaking by comprehensive mediation. The composition of P, e.g., wood, and Q, e.g., chair, is a result of separating speech-situation and situation-talked-about. It can be realized by analyzing and reconstructing what happens when, e.g., in a Q-situation you are uttering ‘P’. In the foreground of the situation-talked-about which is articulated by ‘P’, there are two particulars to be welded. It may come about in either of two possible ways: An aspect (with its schema being) out of σ(ιP) coincides with a phase (actualizations of which being) out of κ(ιQ), e.g., sitting on a wooden chair as a phase-action with respect to chair is simultaneously an aspect-action ‘sitting on the wood of the chair’ with respect to wood; A phase out of κ(ιP) coincides with an aspect out of σ(ιQ).

Pragmatic and Semiotic Prerequisites for Predication

53

In the first case you may articulate the coincidence predicatively by εPQ (= is a wood of [a] chair), in the second case ostensively by δ(QP) (= this wood with the form of [a] chair). Instead of δPQ ε PQ we may write ιQεP (= ιQ is P, or: this [particular] chair is wooden), and likewise, instead of δ(QP) ζ (QP), it is possible to write δPιQ (short for: δPζιQ) (= δP at ιQ, or: this dealing with wood belonging to this [particular] chair). Hence, ‘εP’ acts as a symbol for the result of schematizing ιQ, whereas ‘δP’ acts as an index for the result of actualizing ιQ. Actualizations ostending κ(ιQ) are simultaneously actualizing the universal σP [= δQεP; equivalent with: σPπQ]; schemata predicating a universal out of σ(ιQ) exemplify simultaneously the whole κP [= δPζQ; equivalent with: κPξQ] by being the form of an element of a partition of κP into a class P. Hence, as intended, σ(ιQ) is an appearance of the whole or substance κP, and κ(ιQ) is a carrier of the universal or property σP. An indication δPιQ shows that the substance κP is ostended at ιQ by means of ‘δP’; a proposition ιQεP says that the property σP is predicated of ιQ by means of ‘εP’. Involution as transformation of phase-structures (= internal structure) into aspect-structures (= external structure), and vice versa, can now be proved to exist in a one-to-one way.10 So, it makes indeed sense to say: ιQ consists both of phases such that the closure of their actualisations is κ(ιQ), and of aspects such that the kernel of their schemata is σ(ιQ). And, taking our example, a phase-action with respect to chair which is the pragmatic side of a dealing with chair, being turned into an independent action, can be mapped one-one onto (‘seen as’) an aspect-action with respect to wood which is the semiotic side of a dealing with wood, being turned into an independent action. As a historical remark, it may be added that the two sides of a particular ιQ, the concrete whole κ(ιQ) and the abstract invariant σ(ιQ), correspond neatly to ‘body’ or ‘phenomenon’ and ‘soul’ or ‘fundament’ of a monad as it is conceived in the Monadologie of Leibniz.11 It may also be useful to observe that the identification of δPQ ε PQ with ιQεP, i.e., the introduction of (one-place) elementary propositions, is closely related to Reichenbach’s transition from a thing-language to an event-language articulated with the help of an asterisk-operator which moves the predicative ingredients of a subject term of an (one-place) elementary proposition into its predicate term, e.g., from ‘this man is smoking’ you arrive at ‘smoking of [this particular] man’, or: (ιQεP)*= PQ.12 ____________ 10 Cf. Lorenz 1977, in this volume pp. 20–32. 11 Cf. for further corroboration various essays in Dascal/Yakira 1993. 12 Reichenbach 1947, § 48.

54

Philosophical Logic and Philosophy of Language

Now, we have reached the usual account of (one-place) predication where ‘general terms’ ‘P’ – they should more appropriately and in line with the Fregean analysis of general terms as propositional functions or predicators be rendered as ‘εP’ – serve to attribute properties to particulars of an independently given domain of Q-objects, in the simplest case referred to by deictic descriptions ‘ιQ’ that are special cases of ‘singular terms’ [another use of ‘singular‘!] or nominators.

References CAG = Commentaria in Aristotelem Graeca. Academia Litterarum Regiae Borussicae, I–XXIII, Berlin 1882–1900. CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Dascal, Marcelo/Yakira, Elhanan, eds., 1993: Leibniz and Adam, Tel Aviv: University Publishing Projects. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks Sussex: The Harvester Press. Lorenz, Kuno, 1977: On the Relation between the Partition of a Whole into Parts and the Attribution of Properties to an Object, in: Studia Logica 36, pp. 351–362. Lorenz, Kuno, 1994: Pragmatics and Semeiotic. The Peircean Version of Ontology and Epistemology, in: Living Doubt. Essays concerning the epistemology of Charles Sanders Peirce, ed. by Guy Debrock and Menno Hulswit, Dordrecht/Boston/ London: Kluwer Academic Publishers, pp. 103–108. Lorenz, Kuno, 1996: Artikulation und Prädikation, in: Sprachphilosophie/ Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal/Dietfried Gerhardus/Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1098–1122. Lorenz, Kuno, 1996a: Teil und Ganzes, in: Enzyklopädie Philosophie und Wissenschaftstheorie IV, ed. by Jürgen Mittelstraß, Stuttgart/Weimar: Metzler, pp. 225– 228. Lorenz, Kuno, 1997: Rede zwischen Aktion und Kognition, in: Sprache und Denken. Language and Thought, ed. by Alex Burri, Berlin/New York: Walter de Gruyter, pp. 139–156. Lorenz, Kuno, 2000: Sinnbestimmung und Geltungssicherung. Ein Beitrag zur Sprachlogik, in: Formen der Argumentation, ed. by Geert-Lueke Lueken, Leipzig: Leipziger Universitätsverlag, pp. 87–106. Lorenz, Kuno, 2009: Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter. Reichenbach, Hans, 1947: Elements of Symbolic Logic, Toronto: Macmillan. Scherer, Bernd M., 1984: Prolegomena zu einer einheitlichen Zeichentheorie. Ch. S. Peirces Einbettung der Semiotik in die Pragmatik, Tübingen: Stauffenburg Verlag.

learning

construction

translation

production

rational imagination (= belief)

sensation

perceptual content

mode [of being given]

substance țP class ࣅP (by subdividing the substance into P-particulars [= elements of ࣅP] )

participation at partition by performation recognition ȚQ ‘țP_’ (on the communication side [meta-]articulated by perceptions [= judgments of perceptions]) extension

indicaton: įP ȚQ

signifying via ostension: įP

designating (yields: symbolic representation of a particular)

[the] meaning [of P] [a] realization [of P] sense [of P] reference [of P]

property ıP concept șP (by comprising the properties common to all P-particulars)

intension

imagination

perceptual act

perceptual action (= appreciation)

passive

empirical imagination

active

aspect-action

communicating via predication: İP

proposition: ȚQ İP

reception

attribution by nomination by _șP ‘ȚQ’

mood (= speech act)

competence

sign-

actionperformance

(displaying the mental presence of a social subject)

symbolic articulation

rules of

sign-action (= articulation)

universal

singular

(manifesting the bodily presence of an individual subject)

comprehensive mediation

rules of

partial action (= mediation)

singular

universal

illocution perlocution (on the signification side [meta-]articulated by performators [= illocutionary force indicators])

partaking (yields: symptomatic presentation of a particular)

where applicable: left branches: pragmatic (active-passive) right branches: semiotic (singular-universal)

teaching

success[ful generation]

passive

phase-action

rational production (= intention)

reception

empirical production

production

attempt[ed generation]

poietic action (= generation)

active

action (actualized – schematized)

Pragmatics and Semiotics: The Peircean Version of Ontology and Epistemology In his attempt to situate his own theory of symbols within the history of philosophical ideas, Nelson Goodman wrote that it belongs to “that mainstream of modern philosophy that began when Kant exchanged the structure of the world for the structure of the mind, continued when C.I. Lewis exchanged the structure of the mind for the structure of concepts, and that now proceeds to exchange the structure of concepts for the structure of the several symbol systems of the sciences, philosophy, the arts, perception, and everyday discourse”.1 And indeed, in the light of this characterization it is easy to see that pragmatics has become the modern heir of ontology with semiotics being its counterpart as the heir of epistemology. Of course, both disciplines must be understood in the sense which Peirce gave them, i.e., not as just two newly established empirical sciences, but as two ways of investigation, in which empirical procedures are united with philosophical or reflexive ones. Within this broader perspective, both actions and sign-actions are treated not only as objects of research and representation, but also as a means of research and representation. These entities are not only observed and described according to certain standards, but they are also produced so as eventually – through a stepwise production of an ever more perspicuous internal structure – to achieve a reconstruction of what, in the case of observation and description, is considered to be something given for investigation. As is well known, Wittgenstein used the term ‘language-game’ for the sort of activity that aims at disclosing what is going on by providing tools of comparison. Hence, the productions of actions and sign-actions serve a cognitive purpose, inasmuch as they delineate the very areas of objects which, then, are to be investigated in the usual manner. A language-game may count as a paradigm case of perceptual knowledge, insofar as it signifies by functioning as an icon in the Peircean sense of the term. Using a Wittgensteinean term, one might say that, by producing an icon, one invents a ‘prototype’ that reveals an area of internally structured objects. ____________ 1

Goodman 1978, p. X.

Pragmatics and Semiotics

57

Moreover, it becomes obvious that even the distinction of action and sign-action, which is a special case of the basic and embarrassing distinction between world and language, must be seen as a relative distinction, in the light of a purely functional account of both what it means to be an object and what it means to be a sign (of an object). Indeed, we are asked to create what Goodman would call a new ‘world version’ out of the familiar every day version we normally live in. This is done by looking at verbal language or other symbol systems, as types of actions like eating and sleeping, and by considering non-linguistic and, more especially, non-symbolic objects as parts in a web of interrelated and interdependent actions. Hence, we initially naturalize language and other symbol systems, and we symbolize the world by paying attention to that feature of actions which underlies Goodman’s treatment of exemplification as a tool whereby we tie actions to symbols, and which Wittgenstein has clarified by introducing language-games. Actions are both performed – it should be called their ‘natural’ side – and understood, being their ‘symbolic’ side. The performance of actions is the result of an ability to produce tokens of a certain type; the understanding of actions is the result of an ability to identify different tokens as belonging to the same type. By and large, the primary function of sign-actions, both in their natural or pragmatic aspect (performance) and in their symbolic or semiotic aspect (recognition or identification), is not to serve men’s needs, though, of course, this is not excluded, but to enlarge or refine man’s abilities. In this context it is the abilities on the semiotic side only that pertain to understanding, and they include beside conceptual also perceptual abilities which in turn may be served by skills that may (e.g. in painting) or may not involve the use of external tools. In order to create the distinction between object and sign of an object, it may be helpful to turn to the suggestion made by Peirce himself that signs may be developped from objects. This suggestion is contained in his paper Meaning from 1910: If a Sign is other than its Object, there must exist, either in thought or in expression, some explication or argument or other context, showing how – upon what system or for what reason the Sign represents the Object or set of Objects that it does. Now the Sign and the Explanation together make up another Sign, and since the Explanation will be a Sign, it will probably require an additional explanation, which taken together with the already enlarged Sign will make up a still larger Sign; and proceeding in the same way, we shall, or should, ultimately reach a Sign of itself, containing its own explanation and those of all its significant parts; and according to this explanation each such part has some other part as its Object.

58

Philosophical Logic and Philosophy of Language

According to this every Sign has, actually or virtually, what we may call a Precept of Explanation according to which it is to be understood as a sort of emanation, so 2 to speak, of its Object.

The argument calls for something which is a sign of itself, i.e., which combines object- and sign-features, or better: which functions in both ways. It was the basic point of Peirce’s pragmatic foundation of ‘semeiotic’ (Peirce’s term for semiotics) – convincingly argued for in Bernd Scherer’s dissertation3 – to give an account of the process of separation between sign and its object within the framework of Peirce's Pragmatic Maxim. And the arguments used for that purpose are themselves sections of an open signprocess on the level of reconstruction that may be considered to be nothing but conceptualizations of Wittgensteinian language-games. Now, the descending sequence of interpretants ends with an ultimate logical interpretant4 that is identified as being a habit change which, in contemporary terminology, is to be called the acquisition of an action schema such that all the ways of dealing with the object in respect of what is signified by the initial sign are included. For further clarification it may be useful to turn to the Peircean reading of the ‘semiotic triangle’: A Sign […] stands in such a genuine triadic relation to […] its Object, as to be capable of determining […] its Interpretant, to assume the same triadic relation to 5 its Object in which it stands itself to the same Object.

Each interpretant – cognitions of a mind, i.e., mental interpretants, are some of them – is itself a sign for the same object. Hence, each interpretant in its turn generates a new interpretant, and so on. The sequence of interpretants which gets started in that way may be called a sequence of growing understanding of the object by supplying more and more differentiated determinations. It should not be forgotten that Peirce insists upon ever new, i.e., logically, not empirically new, sign-users connecting the items of the sequence which, for that reason, are called ‘quasi-minds’ and not ‘minds’: … signs require at least two Quasi-minds; a Quasi-utterer and a Quasi-interpreter; and although these two are at one (i.e., are one mind [i.e., the dialogically constituted general subject! K.L.]) in the sign itself, they must nevertheless be distinct. 6 In the Sign they are, so to say, welded.

____________ 2 3 4 5 6

CP, 2.230. Scherer 1984. Cf. CP, 5.476. CP, 2.274. CP, 4.551.

Pragmatics and Semiotics

59

Going back to habit changes, i.e., to acquisitions of action schemata, as the candidates for something which is a sign of itself and thus the end of the descending sequence of interpretants, we may conclude that a verbal sign of an object signifies a range of possibilities of dealing with that object. Even more generally, it might be said that, if the dummy term ‘object’ be deleted, understanding a sign-action, i.e., an action as a ‘symbol’ or a ‘symbolic action’, is, by that very action, tantamount to knowing of a whole range of further actions which may be said to be signified by the symbolic action. As long as the sign-action itself is part of the range of actions it signifies, the sign-action signifies symptomatically, though not yet symbolically. In verbal symptomatic sign-actions there is no separation of word and object: their relation is external, e.g., causal, and not internal, or symbolic. The same idea of explaining how symbolic actions, especially verbal sign-actions, symbolize may be used to explain how ordinary actions function as (not yet verbal) sign-actions. Thus we arrive at the following equivalence: to understand an actualization, i.e., a performance, of an action (‘knowing what one is doing’) is knowing ways of dealing with it. In a more general way, one may say that knowing an object is the same as treating this object as a sign of its distinctions, i.e., those that can be made. And again this is nothing but treating an object as a sign of its internal structure, a structure which is exhibited in an open-ended sign process. Just as a reminder let me – by way of elucidation – refer to the famous diamond example in How to Make Our Ideas Clear.7 The linguistic sign ‘hard’ refers – within the context described by Peirce – to a diamond with respect to hardness. And it represents – that is its last interpretant – the schema of possible dealings with a diamond inasmuch as these pertain to hardness (like trying to scratch it, and so on). Indeed, it becomes a linguistic sign, but not prior to the acquisition of such an action schema. Now, one may read this account the other way round: A schema of possible dealings with an object may act as the designatum of a linguistic sign, in fact as its rigid designatum. Hence, there are indeed rigid designators – but these are general, not singular terms – that are used indexically in order to articulate a situation-type, such as, in our instance, hardness-ofdiamond-situations. Such usage of general terms corresponds to what is now usually called ‘ostensive definition’. Returning to Peircean terminology, one may say that to understand a symbolic action is tantamount to knowing through that symbolic action of a set of other possible actions that might be called the designatum of the symbolic action. ____________ 7

CP, 5.388–5.410.

60

Philosophical Logic and Philosophy of Language

I have deliberately dropped the additional specification that the other possible actions will be possible dealings with some object. This deletion allows for the application of the Pragmatic Maxim to each and every term without stopping short of thing-terms. On this level of conceptual reconstruction, things must be treated as parts of dealings with them. In view of this, the Peircean diamond example should, therefore, be rendered in the following way: If someone is uttering ‘diamondhard’ and is thereby performing a symbolic action, he discloses to the other party (by making use of the function of communication of the term ‘diamondhard’) that there are other actions in virtue of which there is experience of hardness of diamonds, hence, that there is indeed an action-schema which, by common standards, is said to belong to the set of perceptual acts in this respect, e.g., ‘perceiving’ hardness of diamonds. This action-schema is an open schema without criteria as yet – not even categorial ones – to delimit the set of instantiations of the hardnessof-diamond-situation-type that is the designatum of ‘diamondhard’; in this case we are concerned with the term’s function of signification. It should be noted, however, that according to this interpretation signactions and actions seem to be two different entities. Instead, they are simple actions with a dual function. Thus, going another step backward, the simplest structure to start with, is, of course, the one which shows up when one is able to ‘read’ an actualization as a sign of its schema. This ability is conveyed by the ‘language-game’ that, usually, is referred to as the acquisition process of action competence in an elementary dialogue situation: the agent is performing the action, and the patient at the same time is cognizing it, role-switching included.8 With this description, we find ourselves on the primitive level where there is an object which is simultaneously an icon of it, i.e., an entity, in fact an action, that ‘is’ both pragmatic and semiotic, i.e., functions in both ways. In order to advance from this level, one must split the type into parts, so that the actualization of one of these parts becomes a sign of the whole. This is the level that Peirce has in mind when he speaks of an index of an object: “If the Sign be an Index, we may think of it as a fragment torn away from the Object, the two in their Existence being one whole or a part of such whole”.9 Such are the symptomatic sign-actions. The rest is well known, though, of course, there are many more details to be dealt with which cannot be discussed within the scope of this paper. ____________ 8 9

Cf., e.g., Lorenz 1990; meanwhile more elaborately in: Lorenz 2005, in this volume: pp. 42–55. CP, 2.230.

Pragmatics and Semiotics

61

I only hope that I have succeeded in convincing the reader of the necessity to interpret Peircean semeioseis (sign-processes) as being both a methodology and an ontology. Thus, semeioseis are objects and, inasmuch as they are at the same time ways whereby we get to know these objects, cognitions.

References CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks Sussex: The Harvester Press. Lorenz, Kuno, 1990: Einführung in die philosophische Anthropologie, Darmstadt: Wissenschaftliche Buchgesellschaft. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Daniel Vanderveken (ed.), Logic, Thought and Action, Dordrecht: Springer, pp. 343–357. Scherer, Bernd M., 1984: Prolegoma zu einer einheitlichen Zeichentheorie: Ch. S. Peirces Einbettung der Semiotik in die Pragmatik, Tübingen: Stauffenburg Verlag.

Intentionality and its Language-Dependency The recent intensive research in non-verbal communication has established the communis opinio – at least among those trying to explain linguistic features within the pragmatic framework of communicative action – that reference to intentions is necessary in order to give an adequate explanation of basic speech acts. However, even granted that such a mentalistic framework is far more adequate to describe, and possibly even explain, the communicative processes than its behavioristic predecessors, the central issue of modern philosophical methodology, and that is, to supply effective sets of truth conditions for sentences – in this case for those containing expressions for mental concepts – reappears as a prima facie unsurmountable obstacle. This methodological problem may be phrased somewhat more generally: The scientific metalanguage which is devised explicitly for the task of grasping the pragmatic basis of ordinary object language carries its own pragmatic features. Hence, in order to be pragmatically consistent, it must already express what it is about to describe, or, to use Wittgensteinian terms, it must show what it is going to say when saying it. Hence one cannot start from conceptual constructions based on concepts such as ‘action’, ‘desire’, ‘belief’, and the like, in order to define the concept of ‘conventional sign-action’ as a potentially adequate explanation of a ‘speech act’.1 All these constructions rely on their stepwise descriptive adequacy vis à vis the phenomena that are conceptualized in this manner. Such an adequacy, however, can only be intuited, if one wants to avoid an infinite regress of devising further conceptual constructions. A possible alternative is to try and present the objects of interest themselves, e.g., by producing instances of scientific speech acts together with the claim that everyone will be able to ‘see’ their paradigmatic features. This approach, however, is unsatisfactory both in the case of speech acts and actions in general: producing does not guarantee understanding (in the weak sense of ‘being able to reproduce’). An apparent escape route, namely, attempting to split actions up into their parts, from which we regain the original actions, leads to a dead end, too. Although such a procedure per____________ 1

Cf. Meggle 1981.

Intentionality and its Language-Dependency

63

mits us to give an account of the internal structure of an action by showing it instead of saying it, an understanding of the constituent actions still remains a presupposition.2 So, both procedures fall short of leading us to a full knowledge of our linguistic abilities. The first may be called the descriptionist approach, since all actions (= constructions) appear at the conceptual level only. The second approach will be called the constructionist one, since it requires an understanding of the elements and the steps leading to structured actions. A simple, logical example of the descriptionist and constructionist approach, respectively, will be helpful at this point. Let us take a logically compound formula, e.g., A a∧¬b. It is possible to describe this formula using the definite description ι xA(x) with a certain propositional form A(x) on the domain of expressions, where A(x) is defined by: A(x) x is linearly composed out of a, ∧, ¬, b, in that order. One could further axiomatize this syntactic notion of linear composition (= concatenation), such that the answer to the question, whether or not ι xA(x) exists can be given in terms of logical consequence from the axioms. Alternatively, one can simply construct the formula A according to the rules of a suitable calculus for deriving well-formed formulas. Whereas the descriptionist approach (‘saying’) requires a proof of the (unique) existence of the described object, the constructionist approach (‘showing’) is in itself a proof of the existence of the constructed object. Both approaches are based on concatenation: In the first case, we have to know the concept of concatenation, and where and how to apply it, in the second case, we follow the rules, i.e., we have to understand the rules of concatenation in order to be able to concatenate. I call a full command of the first approach metacompetence. It amounts to knowing the means to secure the truth of propositions about objects. A full command of the second way I call object-competence, that is, knowing the objects through adequate presentations of them. Object-competence can only be acquired in the presence of the objects in question. Whether we construct the objects literally, or only perceive them, dealing with objects includes the appropriate speech acts in order to get acquainted with them. The use of speech acts here is a case of accompanying speech (language embedded in human activity), not of speaking about something as on the level of metacompetence. Metacompetence, on the other hand, appears primarily in the absence of the respective objects. It is precisely that tool on the level of signs which ____________ 2

Cf. Schneider 1975.

64

Philosophical Logic and Philosophy of Language

serves to counterbalance the lack of ‘knowledge by acquaintance’ by putting ‘knowledge by description’ in its place.3 The next question is how these two kinds of knowledge are interrelated. Using Gilles Granger’s terminology,4 they may be understood approximately as the result of ‘didactic’ and ‘epideictic’ discourse,5 respectively, provided all further relevant actions, even non-linguistic ones, such as setting up experiments, drawing sketches, moving around in order to draw attention to something, etc., get properly included into such discourses. We see how metacompetence is dependent on object-competence. Without the latter, it cannot be guaranteed that our linguistic tools retain their proper meaning, they may even lose it altogether. This actually explains why I criticized the use of mental concepts in giving conceptual explications of basic speech acts. Likewise, object-competence is dependent on metacompetence, since otherwise the communicating partners will lose their perception of success or failure in communication. This serves to illustrate the shortcomings of the constructionist approach. Hence, it is not enough to merely give a construction of certain objects (= propositions) on the language level, as in the descriptionist approach. It is of crucial importance to judge their ‘relation’ to the object level, in order to secure their truth. Vice versa, objects cannot be represented by presenting them on the object level alone; their representation is also dependent on the linguistic means which allow us to describe the objects in a precise way, and thus to determine them. The interrelation of metacompetence and object-competence can be understood as the relation between answering questions about ‘what is true’ and answering questions ‘on what there is’. These are the epistemological and the ontological versions, respectively, of the question of how the two levels, of objects and of signs, of world and of language, separate and unite into the one domain that I want to label, using Wittgenstein’s expression, the domain of – ordinary as well as scientific – languagegames. The term ‘language-game’, of course, is taken here in the broad sense, where linguistic as well as non-linguistic features are included, among the latter also those of ‘content’, articulating non-linguistic activity. This is in accordance with the pragmatic methodology of C. S. Peirce that the construction of a theory of signs is a double-sided enterprise. It involves a theory of designation together with a theory of designata.6 I ____________ 3 4 5 6

Cf. Russell 1963. Cf. Granger 1985. Granger’s other types of discourse, polemic and heuristic, do not result in knowledge. CP, 5.257.

Intentionality and its Language-Dependency

65

have tried elsewhere to develop such a theory of generalized languagegames, drawing also upon ideas of G. H. Mead, but elaborating them semiotically.7 My starting point is a model of a dialogue situation in which two persons are engaged in teaching and learning an action. All they do is to repeat and to imitate actualisations of an action such that initially there is no differentiation of actor and action, or of action and result of an action. In order to bypass the usual conceptual framework to which the term ‘action’ in scientific language and implicitly also in ordinary language belongs, I use the term ‘preaction’ to characterize this initial state of affairs after having acquired an action competence by imitation and repetition. From this starting point that includes a kind of primitive, non-verbal communication, the development of both ordinary language and the language of science is set going. While constructing an elaborated language, especially in reconstructing (a version) of a scientific one, on top of the more elementary (and obviously fictitious) language where only reference to ‘preactions’ occurs, we already decide on most of what later on is treated as logical or ontological presuppositions of a language and what Emmon Bach in his presentation Beyond formal semantics at the First International Encounter on the Philosophy of Language (Campinas 1981) has called the ‘metaphysics of a natural language in use’. The problem with describing such a process of language reconstruction is that the language of description is some standard natural language already in use. Hence, the language of description is far more developed, syntactically and semantically, than the language to be described while being engaged in the process of its reconstruction. Hence, if we want to give an adequate account of this (re)construction, it is necessary to introduce certain devices that will guarantee that the description at each stage does not depend on those features of the syntactic and semantic structure of the language of description that do not yet belong to the structure of the constructed language. For example, the difference of singular and general terms within the language of description should not be relevant for describing the (fictitious) initial stage of the described language, where nothing but reference to ‘preactions’ is available. That very difference should explicitly be introduced into the elementary language. For this purpose I will first introduce the distinction between (being) universal and (being) singular on the elementary level of non-analyzed actions, i.e., preactions, as the distinction between schema (of an action) and actualization (of an action). These descriptive terms, ‘schema’ and ____________ 7

Cf. besides Lorenz 1970 and 1980, especially Lorenz 1990; the more recent versions, the first in some length and the second condensed, in: Lorenz 1996 and 2005, in this volume pp. 42–55.

66

Philosophical Logic and Philosophy of Language

‘actualization’, bear solely on the differences as expressed by uttering: once, once more, once more again …, which are practically acquired in situations of repetitive imitation (= imitative repetition) with respect to any preaction. Preactions are, oviously, the pragmatic equivalent to ‘feature universals’ as introduced by P. F. Strawson,8 though Strawson uses ‘universal’ to cover both sides of a feature (or a preaction), the universal and the singular one. Unfortunately, he omits in his presentation the singular as being on a par with the universal. Neither can happen without the other, they relate like ‘being’ and ‘existing’ to each other.9 Now, linguistic signs are the means that we have gradually developed during our evolution in order to articulate the mutual dependency of schema and actualization with respect to any preaction. It eventually becomes possible to say which ‘general’ object (being a type that results from typification of a schema) belongs to which ‘individual’ object (being a token of the type), and by further steps to reconstruct what it means when questioning under which concept a certain case might fall or by which case a certain concept might be instantiated. Through language, something individual acts as a symbol of something general, and, conversely, through language, something general acts as an aspect of something individual. The latter case is the concern of the hermeneutic tradition in the philosophy of language. Phrased differently: we are able – on the basis of the different viewpoints of the two persons introduced above in discussing preactions – to distinguish the role of the first person as the one who performs a preaction from the role of the other person who simultaneously cognizes or ‘understands’ that same preaction. The process of cognition is nothing but treating the actualizations on a par with each other. We have reached the stage of articulating sameness of a preaction, i.e., what is common to all the different viewpoints of its performance. Using some results of Goodman,10 we may say that reference to cognition is a case of quotation. We refer to cognizing a preaction by performing another preaction (usually a perceptual one) with the original preaction as its object. For example, you may quote swimming by imitating the act of swimming, or, at a later stage, by producing the utterance ‘swim’. This use of vocal performance is a sign-action, but not yet a verbal sign-action. To reach that final stage, we have to cut the ties that keep the sign-actions, as ____________ 8 9

In: Strawson 1959. Cf. Peirce’s distinction of ‘it is real’ and ‘it exists’ in: CP, 5.430 and CP, 6.335, respectively. 10 Cf. Goodman 1978.

Intentionality and its Language-Dependency

67

symptoms, attached to the entities they signify as parts, and to turn them into preactions of their own, i.e., into symbols. It is an empirical fact about the subspecies homo sapiens sapiens that its vocal preactions, rather than its gestural or other behavior, have developed through history into independent preactions of their own, into what is now known as linguistic expressions. As symbols for vocal preactions, now in the general sense of ‘wholes’ out of parts made up of all the (singular) performances of arbitrary subschemata of the preaction, and which are likewise called (individual) performances – such subschemata are the means to turn schemata into types, one step on the way from preactions to actions11 – the respective utterances will be called articulators. The term‘articulator’ should stress the fact that such utterances do not yet show any of the familiar differentiations that characterize linguistic expressions, including even that between word and sentence, or the ones into different word classes. We should bear in mind, too, that during the natural and social evolution of humankind, the vocal preactions must have seemed to be – at least initially – artificially added to the preactions as extraneous identity tags that allow to identify them; they may be compared, then, for instance, to the use of foot-rings for identifying individual birds. As before, on the level of simple actions, articulators, too, are introduced by way of teach-and-learn-situations, though on a logically higher level. What is acquired through such teach-and-learn-situations is known as the speech act predication. The resulting function of an articulator as a symbol for a preaction is to enable a performance of the preaction, i.e., a specific viewpoint by performing the preaction, to be understood, in the presence of the articulator, as a performance of the schema of precisely that preaction, without having to rely on an introduction of the preaction by means of a first-order teach-and-learn-situation on the spot. In the presence of a performance of some (pre)action articulators function as a substitute for a teach-and-learn-situation of that (pre)action. Our fully developed command of some language tends to hide the very process through which we have acquired that command. This process is intimately interwoven with dialogues if understood as mutually dependent performing and cognizing (pre)actions. By way of endorsement, I may add Donald Davidson’s well argued claim that “[t]here is every reason to hold […] that establishing the correctness of an attribution of belief is no easier than interpreting a man’s speech”, or even, as he states shortly afterwards, that both „problems are identical”.12 Suppose now, that the same holds, ____________ 11 Cf. now, e.g., Lorenz 2005, part II, in this volume pp. 45–50. 12 Davidson 1980, p 238.

68

Philosophical Logic and Philosophy of Language

mutatis mutandis, for desires. Our remaining task, then, will be to introduce intentions on the same rock bottom level as the one where (pre)actions are performed and cognized, i.e., on the level of actions and signactions. There, it has become safe by now to use articulators, too. Let ‘P’ be the articulator of such a (pre)action – you may think of ‘fetching a slab’ in the famous example of a language-game in Wittgenstein’s Philosophical Investigations (§ 2) – then, the procedure to be followed would roughly lead to the following result: Two persons have learned that uttering ‘P’, i.e., a performance of the sign-action ‘P’ by one person, is followed by a performance of an action P by the other. The second performance will be called the intention of the first. Treating ‘intention’ as a two-place predicate on (individual) performances where one performance is the performance of a sign-action, would then prove that to speak of intentions depends on the availability of sign-actions. I will conclude my presentation with some sketchy directions as to how to reach this result.13 First, we should recall that performing a signaction is basically a cognition of an action which is not itself performed but, as I said before, made available when we choose a viewpoint of performance. We may even watch ourselves when doing something: We always both act and ‘act’, as in stage-acting, at the same time. Next, we may distinguish two orderings with respect to the two performances of an action P and its articulation ‘P’, a sign-action. When the performance of the sign-action is followed by the performance of the designated action, the performance of the sign-action appears as a statement: P., and when this order is reversed, the sign-action appears as an order: P! We, thus, get two basic ‘directions of fit’ that can be used in classifying speech acts.14 Besides, we should be aware that making a statement (‘saying’) can be said to express (‘show’) an ability (of one’s own or of the other), namely that of performing an action. The action in question is precisely the one of which an instance has been stated to have occurred; analogously, issuing an order can be said to presuppose the corresponding ability. Finally, using the constative metapredicate ‘state’ and the directive metapredicate ‘bring about’ on the kernel sentence ‘this is P’, we may correctly derive the linguistic representation of the two speech acts in question, that is, the constative: ‘[I] state that this is P’, and the directive: ‘[you] bring about that this is P’. Furthermore, these speech acts may occur as reasons for the statement ‘P.’ and the order ‘P!’, respectively. Furthermore, because statements in need of reasons are called assertions, and orders in need of reasons ____________ 13 For more details, especially about how to ‘split’ articulators into ‘nominators’ (= singular terms) and ‘predicators’ (= general terms), cf. the papers mentioned in note 7. 14 Cf., e.g., now: Vanderveken 1996.

Intentionality and its Language-Dependency

69

obligations, the directive is a reason for the assertion ‘P.’ in the sense that I can demonstrate a state of affairs as being realized through the directive. Hence, such a use of the directive is also an explanation. Similarly, the constative is a reason for the obligation ‘P!’ in the sense that I can make the content of an order manifest by showing my ability to carry it out. Hence, such a use of the constative is also a regulation. Although the two cases, explanation and regulation, have so far only been dealt with in a rather primitive way, I hope to have conveyed an idea of how constructions and descriptions can be interrelated by introducing language-games in a dialogic way that proceeds both pragmatically and semiotically. Doing this will allow us to incorporate the valid arguments that can be put forward for either way of accounting for our linguistic abilities: the behavioristic and the mentalistic approach.

70

Philosophical Logic and Philosophy of Language

References CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Davidson, Donald, 1980: Psychology as philosophy [1974], in: Donald Davidson, Essays on Actions and Events. Oxford: Clarendon Press, pp. 229–239. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks, Sussex: The Harvester Press. Granger, Gilles G., 1985: Discussing or convincing: An approach towards a pragmatical study of the languages of science, in: Dialogue. An Interdisciplinary Approach, ed. by Marcelo Dascal and Hubert Cuyckens, Amsterdam/Philadelphia: John Benjamins Publishing Company, pp. 339–351. Lorenz, Kuno, 1970: Elemente der Sprachkritik: Eine Alternative zum Dogmatismus und Skeptizismus in der analytischen Philosophie, Frankfurt am Main: Suhrkamp. Lorenz, Kuno, 1980: Sprachphilosophie, in: Lexikon der germanistischen Linguistik, ed. by H. P. Althaus, H. Henne, H. E. Wiegand, Tübingen: Max Niemeyer, pp. 1– 28. Lorenz, Kuno, 1990: Einführung in die philosophische Anthropologie, Darmstadt: Wissenschaftliche Buchgesellschaft. Lorenz, Kuno, 1996: Artikulation und Prädikation, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An Interrnational Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal/Dietfried Gerhardus/Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1098–1122; with corrections reprinted in: Kuno Lorenz, Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter, pp. 24–71. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Logic, Thought and Action, ed. by Daniel Vanderveken, Springer: Dordrecht, pp. 343– 357. Meggle, Georg, 1981: Grundbegriffe der Kommunikation, Berlin/New York: Walter de Gruyter. Russell , Bertrand, 1963: Knowledge by acquaintance and knowledge by description [1910/11], in: Bertrand Russell, Mysticism and Logic and Other Essays, London: George Allen and Unwin, pp. 152–167. Schneider, Hans J., 1975: Pragmatik als Basis von Semantik und Syntax, Frankfurt am Main: Suhrkamp. Strawson, Peter F., 1959: Individuals: An essay in descriptive metaphysics, London: Methuen & Co. Vanderveken, Daniel, 1996: Illocutionary force, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An Interrnational Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal/Dietfried Gerhardus/Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1359–1371.

Meaning Postulates and Rules of Argumentation Remarks concerning the pragmatic tie between meaning (of terms) and truth (of propositions) For several years we have experienced a sometimes heated debate among linguists and philosophers of language which I would like to call the controversy between semanticists and pragmaticists concerning the proper demarcation of the areas of semantics and pragmatics. The semanticists try to extend the limits of meaning of linguistic expressions as far as possible into areas of prima facie applicational features, and the pragmaticists try to resolve even structural relations among linguistic expressions into properties of standard situations of use. The underlying question to be asked is: Can linguistic features be distinguished into those which pertain to language structure (irrespective of whether such structure occurs universally or regionally of various degrees) and others which depend irreducibly on non-linguistic circumstances of language use? An example investigated by Asa Kasher1 concerns the linguistic component in rules of politeness. Is it part of the meaning of certain phrases or constructions like ‘I would like to ask you’ instead of ‘I am going to ask you’ that they are used in situations of polite address, i.e., do situations of politeness belong to the range of signification of such phrases or, rather, to the conditions of their application? Obviously, this is indeed a question of how to demarcate precisely the domain of semantics and the domain of pragmatics. The traditional answer that you switch from semantics to pragmatics as soon as speaker and/or listener appear explicitly in the focus of attention, is not sufficient because this move is usually understood as being part of a particular research strategy which I want to call the strategy of semantization of pragmatics. It is basically a move where within your theory of semantic properties – conducted in an appropriate descriptive metalanguage – further variables get introduced, such as variables for time and place of uttering a certain expression, and you arrive at a comprehensive theory which includes treatment of both semantic and pragmatic properties of expressions. No conceptual change of the theory has occurred. You still ____________ 1

Kasher 1980.

72

Philosophical Logic and Philosophy of Language

rely on the Fregean analysis of sentence structure derived from the mathematical model with arguments and value of a function. In fact, you start with two kinds of expressions, singular terms, i.e., nominators, that denote entities, and general terms, i.e., predicators, that serve to express propositional functions on domains of entities. The rest is logic. Hence, you presuppose as being available a theory of reference that has to account for singular terms, and a theory of sense, that has to account for general terms. The area of semantics has not been left, rather, you have forced a semantic treatment of pragmatic features. Of course, there is the converse strategy of pragmatization of semantics starting from Wittgenstein’s concept of language-games. For example, speech act theory has made use of this strategy, though only reluctantly and certainly never in its radical version where every semantic notion must be based upon the concept of language-game and its derivatives. The propositional content underlying illocutionary and perlocutionary forces of utterances will still be analyzed in the Fregean way. Hence, instead of really arguing the respective merits and limits of both strategies with the aim of arriving at a precise explication of the two basic functions of linguistic expressions, their significative and their communicative function – the study of significative function will be semantics and the study of communicative function pragmatics –, a kind of truce is observed. The study of sentence structure is reserved for the semanticists where truth theory with truth as some kind of correspondence becomes the paradigm case. On the other hand, the study of all kinds of forces appearing with the utterance of sentences becomes the domain of the pragmaticists, and, here, argumentation theory may be called the paradigm case of today. The theory of truth as consensus or shared acceptance is but a special case of argumentation theory, and this would be a good choice for taking up again the fight between the two antagonistic strategies. With respect to the concept of truth we should have to ask, whether it is a semantic notion or a pragmatic one. But, unfortunately, just because of the consequence that it would mean to breach the truce, argumentation theory usually stops short of truth theory. This is especially deplorable, because it has the result of continuing to neglect scrutiny into an ambiguity already present in Frege . He used the term ‘true’ both in a semantic framework and in a pragmatic one. I may recall that in the analysis of Frege’s symbolism for assertions, i.e., ‘ A’, the horizontal ‘―’ represents the semantic component and the vertical ‘|’ the pragmatic component of ‘ A’ (‘A’ alone stands for the syntactic component), such that ‘―A’ signifies intensionally a thought and extensionally a truth-value (‘Sinn’ and ‘Bedeutung’ of an assertion), whereas ‘ A’ does

Meaning Postulates and Rules of Argumentation

73

not signify at all, but asserts,2 or more explicitly: The speech act assertion uses the thought ―A as subject of a judgment that is called by Frege ‘acknowledgement (‘Anerkennung’) of the truth of ―A’. But in case ‘―A’ refers to the truth-value ‘falsehood’ it is imperative not to confuse the mere claim for truth with its fulfillment. Frege did not yet draw conclusions out of his own observation that thoughts (or sentence signs) do not behave like other names of objects, but that they are bound to undergo a judgment and that only then the question of truth appears. There is but a famous passage in his posthumous paper Logik that indicates such a radical move to put ‘truth’ exclusively into the area of pragmatics and a fortiori to free considerations of meaning from the grip of truth conditions. The tie of meaning to truth deplorably gained widespread currency following the concise formulation in Wittgenstein’s Tractatus: “To understand a proposition means to know what is the case, if it is true”.3 None of the semanticists has followed Frege on his later route, because meanwhile the archangel Tarski, with his successful restitution of a semantic notion of truth, at least for formalized languages, seemed to have closed the entrance to a pragmatic treatment of truth. The relevant quotation from Frege’s logic-paper runs thus:4 Es wäre nun vergeblich, durch eine Definition deutlicher zu machen, was unter ‘wahr’ zu verstehen sei. Wollte man etwa sagen: ‘Wahr ist eine Vorstellung, wenn sie mit der Wirklichkeit übereinstimmt’, so wäre damit nichts gewonnen, denn, um dies anzuwenden, müßte man in einem gegebenen Falle entscheiden, ob eine Vorstellung mit der Wirklichkeit übereinstimme, mit anderen Worten: ob es wahr sei, daß die Vorstellung mit der Wirklichkeit übereinstimme. Es müßte also das Definierte selbst vorausgesetzt werden. Dasselbe gälte von jeder Erklärung von dieser Form: ‘A ist wahr, wenn es die und die Eigenschaften hat, oder zu dem und dem in der und der Beziehung steht’. Immer käme es wieder im gegebenen Falle darauf an, ob es wahr sei, daß A die und die Eigenschaften habe, zu dem und dem in der und der Beziehung stehe.Wahrheit ist offenbar etwas so Ursprüngliches und Einfaches, daß eine Zurückführung auf noch Einfacheres nicht möglich ist. Wir sind daher darauf angewiesen, das Eigentümliche unseres Prädikates durch Vergleichung mit anderen ins Licht zu setzen. Zunächst unterscheidet es sich von allen anderen Prädikaten dadurch, daß es immer mit ausgesagt wird, wenn irgend etwas ausgesagt wird. Wenn ich behaupte, daß die Summe von 2 und 3 5 ist, so behaupte ich damit, daß es wahr ist, daß 2 und 3 5 ist […]. Die Form des Behauptungssatzes ist also eigentlich das, womit wir die Wahrheit aussagen, und wir bedürfen dazu des Wortes ‘wahr’ nicht. Ja, wir können sagen: selbst da, wo wir die Ausdrucksweise ‘es ist wahr, daß …’ anwenden, ist eigentlich die Form des Behauptungssatzes das Wesentliche.

____________ 2 3 4

Frege 1891, note 7. Wittgenstein 1922, 4.024. Frege 1897, pp. 139–140.

74

Philosophical Logic and Philosophy of Language

We are advised to take up again the study of the difference between the propositional kernel of an utterance and its moods, e.g., its assertive force. The apparently simplest case is an elementary affirmative sentence sεP, e.g., ‘Sam is smoking’. In view of what I said in the beginning about the prerequisite of making singular and general terms available, we may at once observe – and this is usually overlooked by both parties – that using sentences of this kind is dependent on two important presuppositions. First, the independence-presupposition: For each singular term, a certain domain of objects is assumed to be available which includes the referent of that term – the domain acts as the range of a variable at the place of the singular term when sentence forms get introduced – , and this domain is prior to, that is, definitionally independent of, the possible predicates to be applied. With respect to the example ‘Sam is smoking’ we may use the domain of human beings, and the predicates on that domain are – in an extensional account – nothing but set-theoretic partitions of the domain. Hence, they certainly are – or should be – without influence on the definition of the domain itself. Second, the individuation-presupposition: For the objects of a domain it has to make sense to use the predicates ‘same’ and ‘different’, that is, when building a logic on top of our sample sentence together with others (either keeping to one domain of objects or to several domains), it is always possible to introduce the identity predicate without further ado. And, in fact, the very operation of substituting variables for singular terms and vice versa requires that the objects are recurring individuals at one’s disposal rather than mere singulars in the strict sense, as feature tokens, for example. Quine, of course, has noted the individuation-presupposition, though he does not treat it as a presupposition in need of further inquiry as to how it can be ensured. He just claims – as investigated fully by Dale Gottlieb5 – ‘no entity without identity’ in view of his corresponding claim that to be [an entity] means to be the value of a variable.6 The obvious question to be asked is: What must happen and/or what has to be done in order to have both presuppositions fulfilled? In traditional philosophical terminology we are confronted with the problem of objectconstitution to be solved in distinction to problems connected with objectdescription that comes about by using predicates on the domain of objects. Already in Plato we find the method of διαίρεσις which has the purpose of separating questions of object-constitution from posterior questions of truth connected with object-description. Though he was not able really to solve the problem of how to provide for objects, his very simple move ____________ 5 6

Gottlieb 1979. Cf. Quine 1939.

Meaning Postulates and Rules of Argumentation

75

of introducing a two-step procedure is sufficient to account for the distinction between, e.g., ‘Sam is a human being’ that raises no problem of truth – the sentence is true ex definitione or, better: the term ‘true’ does not yet apply – and our original sample sentence ‘Sam is smoking’. The usual articulation of this distinction, of course, is ‘analytic-synthetic’ which rather veils than exhibits the origin of this difference. It suggests, for example, and it has always suggested, that analytic sentences are ‘true by stipulation’. And it leads, and has always led, to the awkward maxim that even in cases of such ‘immediate evidence of truth’ – besides ‘Sam is a human being’ you may also include, e.g., ‘this is an instance of smoking-by-Sam’ where ‘this’ refers deictically to an actualization of the action smoking-bySam which implies that ‘this’ need not yet refer to a full-fledged individual and is not free, therefore, for substitution by a variable – you either should give reasons why they are true, that is, produce arguments (for whom?, what for?, …) or develop a theory of elementary truth as evidence. The more reasonable move in this situation, I propose, is to inquire into the genesis of the question of truth, that is, into the conditions which obtain when it makes sense to use assertions. Of course, logicians of both camps, model theorists and argumentation theorists (they are the semanticists and pragmaticists within logic, and their relation may be considered as a ‘dialectical field’ as introduced by Else Barth7) are accustomed to begin their work only when such a situation, where questions of object-constitution (availability of terms) and questions of object-description (availability of sentences) are kept separate, already obtains. This is also the reason, I think, why the two functions of linguistic expressions, the significative function and the communicative one, which Plato called the two main characteristics of human speech (διακρίνειν τὰ πράγµατα and διδάσκειν τι ἀλλήλους) in his Cratylus,8 are interpreted by logicians, of the semanticist camp at least, to be the origin of the idea that it is necessary to develop both the means to guarantee truth – eventually settled by finding appropriate axioms – and to establish correct rules of discourse or argumentation rules which will, again eventually, i.e., after having invented the instrument of axiomatization of a theory, be settled by determining adequate rules of inference. All this happens on the level of sentences, the set-up of sentences out of terms is relegated to a mere auxiliary task – in the view of our logician, not in the view of Plato. Now, my suggestion is to take Plato verbatim and to start with the idea that each utterance shows indeed both features, each utterance signifies and it communicates, that is, it plays the role of a term (or a word) and the role ____________ 7 8

Cf. Barth 1992. Crat. 388b.

76

Philosophical Logic and Philosophy of Language

of a sentence. This is more obvious in the simplest case of one-word-sentences. In the more complex cases there exist syntactic clues indicating which role is active and which is dormant. The terminology I am going to use with respect to signification and communication is derived more or less from a Peircean account of semiotics and pragmatics, expanded in order to consider the acquisition processes for the language competences of human beings, and conceptualized in the framework of Wittgensteinian language-games. They, in fact, corroborate Plato’s idea that one should start with the two basic functions of linguistic expressions when they are uttered, because Wittgenstein’s concept of language-game incorporates just the two functions of verbal utterances: signification and communication. In the simplest case – we are now on a level below the sample sentence schema ‘sεP’ – an utterance, as a term, articulates an object-schema, e.g., the action-type or generic action smoking, or even its relativization smoking-by-Sam, and it states that there is, in the situation of utterance, a token or instance of that type, an act of smoking. Articulation is the significative function, whereas stating, being the communicative function, may properly be called predication. To say either the utterance makes sense or the utterance is true is superfluous, both formulations are pleonastic, you just use other words for the two functions. A considerable number of steps are needed to introduce that kind of complexity into one-wordsentences which allow for differentiation, by syntactic means, of their significative and communicative functions, e.g., to make available both ‘smoking-by-Sam’ (a term) and ‘Sam is smoking’ (a sentence). Before looking into this a bit further, I would like to point out that it is not sufficient simply to observe utterances and to distinguish their two functions. Such a descriptive attitude works only when augmented by participation. Unless you take into account that you have to acquire the ability to use an expression with these functions – making utterances in a situation which are understood and reacted upon properly – , and this amounts to being aware of the necessarily social character of the acquisition process, it is not possible to get a clear picture of the distinction between introduction and use of expressions. In situations of use relation to the acquisition process for the parties involved looks very different with respect to the significative function. For example, the utterance ‘smoking’ may occur in a situation where the speaker sees or smells someone (listener or speaker herself) smoking or where he helps someone to light a cigarette, etc. The use of ‘smoking’ can refer to smoking only through a specific ‘component’ of the action-type smoking, accessible to the speaker (and another one to the listener) in the situation of utterance. It is what Frege called the ‘Art des Gegebenseins’, i.e., the mode

Meaning Postulates and Rules of Argumentation

77

of being given of an object, here: of the object-schema, the type smoking. But, on this level of conceptual reconstruction, where there are objects of reference not yet available independently from what is said about them, it is impossible to refer to smoking when there is nobody smoking in the situation of uttering ‘smoking’. You can also articulate these modes of being given, and the associated predications of second order are what our philosophical tradition up to Leibniz has called perceptions, though perceptions have by and large been treated erroneously as the basis of predications of first order instead of proceeding the other way round. Correspondingly, with respect to the communicative function of our sample untterance ‘smoking’, you are bound to consider all kinds of possible uses in a specific situation, warning the listener or encouraging him with that utterance, asking him by it whether he wants to smoke, etc..The predication ‘smoking’, i.e., ‘smoking’ in its sentential role, can only occur in a mood – there is no pure predication in situations of use – which, when articulated independently, occur as the well-known performators within the theory of speech acts. Now, the meaning of a term is, of course, the invariant out of the open set of perceptual perspectives, these perspectives being the ‘senses’ through which meaning as ‘reference’ works, whereas the validity of a sentence is a claim with respect to the mood of the utterance of the sentence, the claim for truth being a special validity claim with respect to the assertive mood. Such claims roughly amount to the ability on the part of the speaker and/or listener to make the situation of utterance of a sentence coincident with that situation the type of which is articulated by the corresponding term. Hence, validity claims can occur only when these two situations, the situation of uttering and the situation of acting that is articulated by the utterance, do not coincide. The mere use of different modes of being given by two parties is insufficient to speak of a problem of meaning for them, i.e., do they ‘really’ speak of the same thing? Likewise, just the use of a mood when speaking ‘about something’ is insufficient to raise a problem of validity for the two parties, e.g., has my communicative intention towards you been successful or not? Let us return to our example in order to clarify this stage of conceptual reconstruction a bit more. If the term ‘smoking-by-Sam’ is available, what is its sentential value when uttered? Obviously, the one that can be represented by the sentential locution ‘this is smoking-by-Sam’ assuming the assertive mood for the moment. And the crucial question arises: What is the difference between the sentence ‘this is smoking-by-Sam’ and the sentence ‘Sam is smoking’? The sentence ‘this is smoking-by-Sam’ contains a singular term referring, in complete dependency upon the situation of utterance, to an un-

78

Philosophical Logic and Philosophy of Language

retrievable singular ‘token’ of smoking-by-Sam. On the other hand, the sentence ‘Sam is smoking’ contains a singular term that refers independently of the situation of utterance, it refers, in fact, to an individual constructed as a whole out of an indefinite class of singular ‘tokens’, better: ‘actualizations’ (in order to mark the difference to particular tokens), taken from schemas like smoking, running, nose, beard, seen-by-x, etc., that are coincident with certain actualizations of the (not yet typified) schema man. And it is this invariance of reference with respect to situations of utterance that is made possible by turning from singulars (actualizations of an object-schema) to individuals (wholes out of certain indefinite classes of actualizations of an object-schema that is in this way, by typification or ‘individuation’, turned into a type with the individuals as tokens) that creates the difference between uttering something meaningfully and doing that truthfully, and not, as Frege thought, by having reference besides sense or not. You can assert ‘Sam is smoking’ and thereby claim that you are able to provide situations – in the non-trivial cases different from the situation in which ‘Sam is smoking’ is uttered – that instantiate smoking-by-Sam, this action-schema being defined as the meaning of the sentence ‘Sam is smoking’. The operation leading from a sentence A, e.g., ‘Sam is smoking’, to the associated term A*, e.g., ‘smoking-by-Sam’, such that A is true iff A* is instantiated, was first introduced by Hans Reichenbach with his star operator to derive an event-language out of a thing-language.9 As a final step we apply these ideas to logically complex sentences, and thereby arrive at a perspective on dialogic logic which was not at all clear when the game-theoretic approach to logic was first developed by Paul Lorenzen and myself.10 For each logical composition, A ∧ B or A→B, you are asked to determine the meaning of these sentences, that is, to ask for the kind of action-schemas that are articulated by (A ∧ B)* or (A→B)*. The game-theoretic approach makes sense only, if you neither turn to the conjunction of the two action-schemas, e.g., smoking-by-Sam and climbing-by-Jim, as the meaning of A ∧ B – it would be nothing but truthfunctional logic in disguise, and, hence, inadequate to handle the universal quantifier in a finite manner –, nor turn to the conjunction of proofprocedures for A and for B as a proof-procedure, and in this sense the meaning, for A ∧ B; that would be Andrej Kolmogorov’s interpretation of logical composition11 with its well-known difficulties to arrive at a treat____________ 9 Reichenbach 1947, § 48. 10 The relevant papers are collected in: Lorenzen/Lorenz 1978, subsequent clarifications are contained in: Lorenz 1982, 1985; a more elaborate treatment of the whole programme can be found in: Lorenz 1980. 11 Cf. Kolmogorov 1932.

Meaning Postulates and Rules of Argumentation

79

ment being both adequate and rigorous. It is necessary to turn to an interaction-schema as the meaning of A ∧ B, and that is to be specified and typified as a dialogue game with the plays of that game being the interaction tokens in a specific mode of being given for either player. The same holds in case of the subjunction A→B. Again, neither an ifthen-connection of action-types as meaning of A→B nor a conditional proof-procedure is appropriate. The first move is placed too ‘low’, on the object level (technical ‘if-then’), the second move is placed too ‘high’, on the level of argumentation (practical ‘if-then’). We have to go ‘in between’, and to establish the meaning of A→B on the language level (linguistic or semiotic ‘if-then’): The meaning of A→B is an interaction-type, the dialogue game about A→B. Hence, the dialogue-rules defined for logically compound propositions may be called rules of signification. They should carefully be distinguished from the rules of argumentation which occur on the level of strategies and not on the level of individual plays. Problems of formal truth can be formulated only on the level of strategies and may be handled in the usual way, e.g., by demanding hereditariness of truth from implicans to implicate, and may afterwards be based on stipulations of formal dialogue-playing. Formal logic with its inference rules turns out once again to be just an instrument for establishing material truths, securing truth claims on some subject matter under dispute, nothing more – and nothing less either.

80

Philosophical Logic and Philosophy of Language

References Barth, Else M., 1992: Dialogical approaches, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz and Georg Meggle, vol. 1, Berlin/New York: Walter de Gruyter, pp. 663–676. Frege, Gottlob, 1897: Logik, in: Gottlob Frege. Nachgelassene Schriften, ed. by Hans Hermes, Friedrich Kambartel and Friedrich Kaulbach, Hamburg: Felix Meiner 1969, pp. 137–163. Frege, Gottlob, 1891: Funktion und Begriff, in: Gottlob Frege. Kleine Schriften, ed. by Ignacio Angelelli, Darmstadt: Wissenschaftliche Buchgesellschaft 1967, pp. 125– 142. Gottlieb, Dale, 1979: No entity without identity, in: Essays on the Philosophy of W. V. Quine, ed. by Robert W. Shahan and Chris Swoyer, Norman: University of Oklahoma Press, pp. 79–96. Kasher, Asa, 1980: Three kinds of linguistic commitments, in: Time, Tense, and Quantifiers. Proceedings of the Stuttgart Conference on the Logic of Tense and Quantification, ed. by Christian Rohrer, Tübingen: Niemeyer 1980, pp. 207–222. Kolmogorov, Andrej N., 1932: Zur Deutung der intuitionistischen Logik, in: Mathematische Zeitschrift 35, pp. 58–65. Lorenz, Kuno, 1980: Sprachphilosophie, in: Lexikon der Germanistischen Linguistik, Tübingen: Max Niemeyer, pp. 1–28. Lorenz, Kuno, 1982: Die irreführende Gleichsetzung von Begründungen und Argumentationen. Bemerkungen zu einem monologischen Mißverständnis in der dialogischen Logik, in: Logik und Pragmatik. Zum Rechtfertigungsproblem logischer Sprachregeln, ed. by Carl Friedrich Gethmann, Frankfurt am Main: Suhrkamp, pp. 78–91. Lorenz, Kuno, 1985: Zur logischen Genese der Sprachverwendung, in: Philosophica 35, pp. 33–37. Lorenzen, Paul/Lorenz, Kuno, 1978: Dialogische Logik, Darmstadt: Wissenschaftliche Buchgesellschaft. Quine, Willard V. O., 1939: Designation and Existence , in: The Journal of Philosophy 36, pp. 701–709. Reichenbach, Hans, 1947: Elements of Symbolic Logic, London: Macmillan. Wittgenstein, Ludwig, 1922: Tractatus Logico-Philosophicus, dt.-engl., ed. by David F. Pears and Brian F. McGuinness, London: Routledge and Kegan Paul 21971.

What Do Language Games Measure? In order to understand the function of language-games it is very useful to look somewhat more closely into one of the most important metaphors Wittgenstein has used throughout his life. My main concern will be an explication of how the term ‘Bild’ (picture) has changed its meaning when going from the Tractatus [T] to the Philosophical Investigations [PU] by determining precisely that function of ‘Bild’ which has been kept constant. Let me start with the well-known fact that ‘Bild’ is something on the level of sentences both in T and in PU: In his Notebooks Wittgenstein writes at the date 29. 9. 14: “In the sentence a world is put together experimentally. (As when in the law-court in Paris a motor-car accident is represented by means of dolls, etc.)”.1 And in T he continues: “One name stands for one thing, and another for another thing, and they are connected together. And so the whole, like a living picture, presents (stellt vor) the state of affairs./The possibility of sentences is based upon the principle of the substitution (Vertretung) of objects by signs”.2 Quite alike we read in PU: “Imagine a picture representing a boxer in a particular stance. Now, this picture can be used to tell someone how he should stand, should hold himself; or how he should not hold himself; or how a particular man did stand in such-and-such a place; and so on. One might […] call this picture a sentence-radical”.3 Of course, there is a slight but significant change we should at once try to note. In the T-quotation, the picture represents a state of affairs as is the case with sentences,4 whereas in the PU-quotation, the picture can be used, e.g., to represent a state of affairs, but also for other purposes, as is the case with sentence-radicals which, when they are used, are complemented by their moods yielding assertions, questions, commands, etc. So it looks – and speech act theory usually proceeds that way – as if there is a basic locutionary act, ‘predication’, executed by ‘propositional kernels’, that in utterances, i.e., within sentences in use, appears as locutionary meaning together with illocutionary and perlocutionary force. Wittgenstein is said to ____________ 1 2 3 4

Wittgenstein 1961, p. 7. T, 4.0311–4.0312. PU, § 23. Within the given quotations the difference of ‘darstellen’ (represent) in Notebooks and ‘vorstellen’ (present) in T is of no relevance conceptually.

82

Philosophical Logic and Philosophy of Language

have treated in T locutionary meaning only, and to have given up this notion in favour of a study of speech acts where predications, within this context better to be called ‘constatations’ or ‘assertions’, are just one kind among them. Hence, rather than dealing with one language-game saying something about objects as a function of their names – as Wittgenstein is said to have done in T – the arena is open – and this is Wittgenstein’s explicitly declared opinion5 – for the study of an unlimited variety of such games that meanwhile speech act theorists claim to be able to boil down to essentially five types. I think that such an account is a distortion of Wittgenstein’s ideas. It came about presumably because the change of meaning of some crucial terms, implied by the change from a level of explanation to a level of description that took place according to Wittgenstein’s own words between T and PU,6 went by more or less unnoticed. I am referring not only to the term ‘Bild’ but also to the term ‘Satz’, both of which keep their relation to each other almost unchanged – whereas a picture is like a sentence in T, it is like a sentence-radical in PU – but they are moved to another level. Using a Peircean terminology I claim that pictures as well as sentences function as ways of symbolic representation in T, whereas pictures together with sentence-radicals function as a means of iconic representation in PU. This implies that language-games are Wittgenstein’s version of iconic representations, and that the term ‘language-game’ cannot properly be applied to the picture theory of the Tractatus. In order to corroborate my claim I first want to sketch what I have argued for elsewhere in detail.7 A sentence in T represents a state of affairs by being a representative of an abstraction class with respect to the equivalence relation among sentences that is expressed by rules of translation. Thus, rendering the picture theory of T as a theory of abstraction, entails the assertion that pictures in T are symbolic representations. In giving, now, the reasons for my claim I start with the well-known passage in Wittgenstein’s Notebook (27. 10. 14): “The difficulty of my theory of logical portrayal was that of finding a connexion between the signs on paper and a situation outside in the world. I always said that truth is a relation between the proposition and the situation, but could never pick out such a relation”.8 Even more general, two months earlier, on 3. 9. 14, he declares: “The obscurity obviously resides in the question: what does the logical identity of ____________ 5 6 7 8

Cf. PU, §§ 23 seqq. Cf. PU, §§ 109 seqq. Cf. Lorenz 1972. Wittgenstein 1961, pp. 19–20.

What Do Language Games Measure?

83

sign and thing signified really consist in? And this question is (once more) a main aspect of the whole philosophical problem”.9 It is well known that by the time of actually writing the Tractatus, the logical identity of sign and thing signified has become an internal relation which shows itself and which cannot be said, whereas the relation between proposition and situation he was looking for in vain is an external relation that, indeed, does not exist. Here, already, we have tracked down the germ of one of the most basic insights that even now has not been fully acknowledged by everybody working in that area. It is Wittgenstein’s realization that the copula, relating a sign and one or more objects in order to arrive at a predication, is not an ordinary relation between objects; it, rather, is the means to be able to state that such an ordinary relation holds. The distinction of ‘internal’ and ‘external’ that is used more or less synonymously with the traditional distinction of ‘formal’ and ‘material’, is turned into a tool. The tool is used to give a clear account of what it means to stay on the object level (to treat objects, their properties, relations, all of them external ones) and what it means to exhibit the very means necessary to make objects, properties, etc. available. These means are internally related to objects, etc.; they are purely semiotic tools, or ‘symbols’, which must not be confounded with the material carriers of symbols, e.g., sounds, which receive their function by convention, and that is an external relation. From a number of relevant passages in T you may derive:10 (1) to know an object is equivalent to knowing all of its internal properties, (2) to know an object is equivalent to knowing all the possibilities of its occurrence in state of affairs, (3) forms of an object are possibilities of its occurrence in state of affairs. These three sentences imply that internal properties of an object are its ‘forms’, i.e., its formal properties. And these forms of objects determine completely the structures of those states of affairs in which they can occur. There are no external properties of state of affairs, thus explaining why they have to be kept distinct from objects. Let us turn to the central passage in T where Wittgenstein gives an account of the internal relation between language and the world, a passage which is usually taken to prove the thesis of an isomorphism between language and the world. But, certainly, isomorphisms are special external relations between objects. I quote: “The sentence [as sentences are sentence-signs, i.e., physical object types and not tokens, together with their function of signification, and, therefore, symbols, they are usually rendered in English as propositions which, though approved by Wittgenstein, unfortunately invites the reader to turn to the superfluous discussion about the ____________ 9 L.c., p. 3. 10 Cf. T, 4.124; 2.0123; 2.01231; 2.0141 et allii.

84

Philosophical Logic and Philosophy of Language

ontological status of these entities] is a picture of reality. The sentence is a model of reality as we think it is. At the first glance the sentence – say as it stands printed on paper – does not seem to be a picture of the reality of which it treats. But nor does the musical score at first sight to be a picture of a musical piece […] And yet these symbolisms prove to be pictures – in the ordinary sense of the word – of what they represent […] the gramophone record, the musical thought, the score, the waves of sound, all stand to one another in that pictorial internal relation which holds between language and the world. To all of them the logical structure is common”.11 Let me insert that ‘logical structure’, a term which does not occur elsewhere in T, has to be read as ‘logical form of representation’ (= form of reality). Different equivalent pictures have different forms of representation, but are alike in their logical form. The passage continues by referring to rules of translation between different pictures establishing their equivalence – ‘internal similarity’ Wittgenstein says –, and it ends with the following words: “In order to understand the essence of the sentence, consider hieroglyphic writing which pictures the facts it describes”.12 We have to add that “[t]he holding of such internal properties and relations cannot, however, be asserted by sentences but it shows itself in the sentences which represent the states of affairs and treat of the objects in question”.13 Hence, the internal relation between language and the world shows itself in the sentences and cannot be turned into the object of metasentences. In fact, as we have learned from G. E. Moore’s lecture notes,14 Wittgenstein considers internal relations to be grammatical ones, his term for what we would call today ‘semiotic relations’. Wittgenstein himself gives reasons for his use of the term ‘Bild’ for the internal relation between a sentence and a state of affairs. He refers to the fact that a sentence is understood solely on the basis of our knowledge of the objects about which the sentence says something; and here ‘knowledge’ is knowing the internal properties, i.e., the ones by which the objects are ‘defined’. A picture can be true or false, that is why it is called a model of reality: “It is like a measuring-rod applied to reality”.15 Here, too, the corresponding idea concerning language-games suggests itself. In the Philosophical Investigations Wittgenstein declares that “language-games are rather set up as objects of comparison which are meant to throw light on the facts of our language by way not only of similarities but ____________ 11 12 13 14 15

T, 4.01–4.014. T, 4.016. T, 4.122. Cf. Moore 1959, p. 295. T, 2.1512.

What Do Language Games Measure?

85

also of dissimilarities. For we can avoid ineptness or emptiness of our assertions only by presenting the model as what it is, as an object of comparison – as, so to speak, a measuring-rod”.16 And, again, the slight, though decisive difference: in the T-case sentences being measuring-rods are carriers of truth-values, in the PU-case language-games being measuringrods tell you what there is. We have moved from an epistemological level to an ontological level. Let us return again to the argumentation in the Tractatus. We, already, know that sentences, being means to say something about objects, i.e., to describe or to represent states of affairs, are not names of anything, but – as Wittgenstein writes on 21. 11. 1914 in his Notebooks17 – “the realities corresponding to the sense of a sentence are only its component parts”. Hence, the objects about which you say something stand on the other side of language, and not the states of affairs; what you say about objects, that a state of affairs holds, i.e., that it is a fact, belongs to language alone. This is the root of Wittgenstein’s fundamental distinction between the world, i.e., the totality of facts, and the substance of the world. Now, sentences have essential features, i.e., internal properties which are common to all sentence-signs expressing the same sense. And for all symbols, whether names or sentences, Wittgenstein notes that “„[w]hat signifies in the symbol is what is common to all those symbols by which it can be replaced according to the rules of logical syntax”.18 We have again arrived at the rules of translation which eventually define what it means to have the same sense: “In order to recognize the symbol in the sign we must consider the significant use. The sign determines a logical form only together with its logical syntactic application”.19 The ideal language postulated by Wittgenstein in T should be considered as a comprehensive language where every well-established symbolic language is treated as a part of it. Then, the rules of translation are part of the logical syntax of the ideal language and can be used as an equivalence relation in order to define states of affairs by abstraction. In fact, every kind of meaning can be defined in that manner. And Wittgenstein has indeed succeeded in eliminating every reference to meaning in setting up the rules of logical syntax: “In logical syntax the meaning of a sign ought never to play a role; it must admit of being established without mention being thereby made of the meaning of a sign”.20 Hence, “[t]he rules of ____________ 16 17 18 19 20

PU, §§ 130–131. Wittgenstein 1961, p. 31. T, 3.344 T, 3.326–3.327. T, 3.33.

86

Philosophical Logic and Philosophy of Language

logical syntax must follow of themselves, if we only know how every single sign signifies”.21 We may conclude: The sense of sentences is fully explained by the use you make of them, i.e., by the syntactic rules, provided we know what shows itself when using a sentence. And it is this knowledge of what shows itself when using sentences, which becomes Wittgenstein’s concern in the Philosophical Investigations. The language-games serve as a kind of ‘rational reconstruction’ in order to show (providing ‘active’ knowledge) what shows itself (‘passive’ knowledge) in the Tractatus. When Wittgenstein says, as quoted above, that language-games are a measuringrod applied to reality, i.e., to the totality of facts, it is not preexisting ‘facts’ (German: Verhältnisse) which are brought into focus by them. They rather provide ‘perspicuous representations’22 – a term very near in fact to my claim that language-games are iconic representations – which Wittgenstein considers to be ‘of fundamental significance’, thus stressing the semiotic feature of language-games beside their obvious pragmatic character. Rather than being means to speak about objects, language-games show what kind of objects you are dealing with and in what manner. They articulate situations by exposing their function among persons. A languagegame is both a world view and a way of life. This idea coincides with an insight which we find in Plato, already, when, in his Cratylus, he speaks of the two main characteristics of human speech, its significative function and its communicative function.23 In fact, each utterance shows two features, it signifies and it communicates, that is, it plays the role of a term (or a word) and the role of a sentence, well known in the simplest case of one-wordsentences as encountered in the famous samples of language-games in the opening sections of PU, where the expressions ‘block’, ‘pillar’, ‘slab’, and so on, serve both functions. Language-games are a kind of activity which aims at disclosure of what is going on by providing tools of comparison. You not only observe and describe actions and sign-actions thereby according to certain standards but you also produce them in an orderly fashion – perspicuously, as we have noted already – in order to arrive at some kind of approximating reconstruction of what you take to be available, already. A language-game may count as a paradigm case of perceptual knowledge insofar as its significative function works by being an icon in the sense of C. S. Peirce: you have found an area of internally structured objects by inventing a prototype. And it becomes obvious that even the distinction of action and ____________ 21 T, 3.334. 22 PU, § 122. 23 Cf. Crat. 388b.

What Do Language Games Measure?

87

sign-action, a special case of the basic and embarrassing distinction between the world and language, has to be relativized in view of a purely functional account of what it means to be an object and what it means to be a sign of an object. We have to try, therefore, to develop a unified approach to both objects and signs of objects quite in line with the pragmatic methodology of Peirce that a theory of designation and a theory of designata has to be a combined endeavour. For this purpose a dialogue model where you avail yourself of a situation of acquiring both an action competence and a sign-action competence commends itself. As a first step you have to treat the use of verbal language on a par with ordinary activities and at the same time to look at ordinary objects as embedded in a web of interdependent actions that are used to signify them. With other words, you have to ‘naturalize language’ including other symbol systems and to ‘symbolize world’ including our activities. The bridge over the apparent gap between world and language can be built if you make that gap disappear. It is essential to observe that actions occur under two perspectives. They are performed as well as understood. From the first perspective, i.e., in the performance aspect, you produce tokens of a type, and from the second perspective, i.e., in the cognition aspect, you ‘identify’ different tokens as belonging to the same type. The dialogue model incorporates these two perspectives, since it consists in an elementary situation of two persons being engaged in starting and continuing a process of acquiring the competence to perform a certain action. The two-person elementary situation is nothing but a generalized version of a Wittgensteinian languagegame, especially with respect to its role to function as a measuring-rod for already occurring complexes of actions including linguistic ones.24 It is the purpose of the elementary dialogue situations together with the construction of a gradually growing differentiation into more and more complex dialogue situations to keep a unified account of how we acquire both participation in a common situation and use of a common piece of language. It comes about by looking carefully at the two points of view exercised by the two agents in such a situation of action acquisition. At a given time one of them is performing and the other one cognizing the action, and in the course of the acquisition processes the points of view switch permanently. The primary objects available ot this stage of modelling are actions without any further differentiation, neither with respect to act and agent, nor with respect to act and object of an act, nor otherwise. But each one of them is a common situation, and both parties know it. ____________ 24 Cf. PU, § 131.

88

Philosophical Logic and Philosophy of Language

Again, you may think of the famous rabbit-situation in Wittgenstein’s PU,25 where in the course of discussing ‘seeing something’ and ‘seeing as’ he gives the example of an exclamation ‘A rabbit!’ whereby a landscape with a running rabbit turns into a rabbit-situation. This ‘flashing of an aspect’ he describes as ‘half visual experience, half thought’. Of course, we have here a more advanced dialogue situation, insofar as the speaker is exclaiming ‘A rabbit!’. With such an exclamation both parties of our elementary situation show their being equipped with some significant further developments: As performing party the speaker performs a perceptual action, as cognizing party he utters ‘A rabbit!’, and acquisition of ‘rabbit-competence’ is a sudden one-step affair. In the report-case, on the other hand, when your utterance ‘A rabbit.’ functions as a report (in the written presentation the exclamation mark is deleted), you rely on an, already, perceptually defined situation, e.g., you have memories of rabbits having met them on previous occasions. To keep track of the different stages in the development of the dialogue model I have introduced the term ‘preaction’ for articulating the entities that belong to the elementary and rather primitive though common world of the first stage.26 Neither things nor persons, nor rabbits in the usual sense will be preactions. Looking backward from a later stage you may say that things, persons, animals, etc., are as yet indistinguishable parts of preactions. And it is important to be conscious of the fact that preactions have neither the status of given data nor of rational constructions as the traditional alternative – stated in terms of the empiricism-rationalism-debate about the primary level of reality – would ask us to decide between and which also Wittgenstein wanted to avoid. Preactions are performed by the acting agent and cognized by the other party, and that is it. Now, taking into account that the performance viewpoint is pragmatic, whereas the cognition viewpoint is semiotic, we may say that our model of elementary dialogic interaction permits us to distinguish between a pragmatic and a semiotic aspect of every preaction. You get a concrete version of the sign-character of a preaction, if you look at the non-performing party during a particular instant of the dialogue situation as someone who understands the performance of the other party, e.g., as an invitation to do ‘the same’. We take a further step by turning the cognition viewpoint of a preaction into an explicit preaction of its own. What is going to happen? Instead of simply cognizing smoking, for example, you perform a separate preaction, say see-smoking, which qualifies as a perception with respect to the ____________ 25 Cf. PU, II.XI. 26 Cf. Lorenz 1987, in this volume pp. 124–139.

What Do Language Games Measure?

89

original preaction. Cognizing a preaction is made explicit by performing a perception of that preaction. Therefore, in the elementary dialogue situation the person who performs smoking is simultaneously cognizing the perception of his partner, that is, in more colloquial terms, he knows that his partner is knowing how to smoke. The next step of identifying the cognition aspect of a preaction with the performance aspect of a perception of the preaction which may be called ‘objectivation’, leads to the stage of object constitution independent from particular dialogue situations of acquiring the competence to perform the preaction in question. By ‘pragmatic abstraction’ we define a preobject to be the invariant out of the open set of perceptions with respect to the preaction. Preobjects are not objects – both kinds of entities have carefully to be kept distinct –, because individuation is lacking: A division of a preobject into individual units has still to be introduced in an explicit fashion. Now, as cognizing a preobject is identified with performing a corresponding perception, preobjects themselves cannot any more be performed. Preobjects are by definition the same for all participants though with different perceptual access. We call preobjects articulated by their perceptions. But it is a contingent anthropological fact that certain perceptual preactions out of the manifold of articulations of a preobject, namely verbal preactions, receive canonical status with respect to their being a sign of the preobject: as a kind of symptom they represent the preobject pars pro toto. For the next step of turning the verbal preactions into preobjects of their own without thereby cutting off their perceptual, i.e., semiotic, function, we turn to dialogue situations of second order. Such dialogue situations serve as genetic reconstructions of what in the tradition of philosophical logic has been called (situations of) ‘predication’. The verbal preobjects or, rather, their spoken or written results made available by pragmatic abstraction with the help of such second order dialogue situations I call articulators. They cannot be performed but only ‘presented’. To utter an articulator, or to perform an articulation or a predication, say ‘smoking’ – which is the ‘verbal perception’ of the preobject smoking – is at the same time a case of cognizing it. An articulator is, as a preobject, a fullfledged symbolic sign, and not any more a mere symptomatic one. We may distinguish these two semiotic functions of an articulator by saying: In its symptomatic function an articulator is constitutive of its object, in its symbolic function it describes its object. Obviously, also pictures can be used as articulators. Only now we have reached the stage where articulations being both signs and actions, that is sign-actions, can be subjected to the two aspects which I referred to as the two basic functions of language: the function of

90

Philosophical Logic and Philosophy of Language

signification and the function of communication. To avoid confusion it is advisable to use the term ‘articulation’ with respect to its significative function, only, and to switch to the term ‘predication’ when the communicative function of articulation is focussed. Since it generally holds that with respect to its pragmatic character a sign-action is communicative, and with respect to its semiotic character an action is significative, we can say that an articulation in its significative aspect is realized by performing a perception. Hence, the ‘meaning’ of an articulator splits into the different perspectives of the persons using the articulator and, thus only, common meanings can be ensured. Likewise, an articulation, in its communicative aspect, that is, a predication as an action, splits into two different ways of being given – again the result of an objectivation, here of second order, by turning the cognition of a predication into an independent action, which is called a mood of the predication. The more familiar term nowadays for such a mood is, of course, ‘speech act’. Predications always occur in a mood, the speech act between speaker and hearer. The second-order dialogue situations appear as the systematic equivalents to Wittgenstein’s language-games in the strict sense. And what has become clear, I hope, is that language-use always occurs on two levels: as a perception, that is as a way of presenting an object – the ‘rabbit’-exclamation, a case of ‘object-competence’ – and as a conception, that is, as a way of referring back to other terms that are used, already, perceptually – the ‘rabbit’-report, a case of ‘metacompetence’. In the Tractatus Wittgenstein is, in general, treating the ability to use sentences and pictures as a case of metacompetence with respect to signactions, whereas in the Philosophical Investigations he is presenting language-games and the use of pictures as a case of object-competence with respect to sign-actions. In this latter case they serve as ‘exemplifying’ sign-actions using a terminology of Nelson Goodman.27 If, as a piece of rethoric, you disregard the equivocation thus connected with the term ‘Bild’, you may say: In the Tractatus pictures are symbols, whereas in the Philosophical Investigations language-games are pictures in use.

____________ 27 Cf. Goodman 21976, II.3.

What Do Language Games Measure?

91

References Goodman, Nelson, 21976: Languages of Art. An Approach to a Theory of Symbols, Indianapolis/Cambridge: Hackett Publishing Company. Lorenz, Kuno, 1972: Zur Deutung der Abbildtheorie in Wittgensteins Tractatus, in: teorema – número monográfico, Universidad de Valencia, pp. 67–90. Lorenz, Kuno, 1987: Is and Ought Revisited, in: dialectica 41, pp. 129–144. Moore, George E., 1959: Wittgenstein’s Lectures in 1930–1933 [Mind 1954/1955], in: Philosophical Papers by George Edward Moore, London/New York: George Allen & Unwin/Macmillan, pp. 252–324. PU = Wittgenstein, Ludwig: Philosophische Untersuchungen. Philosophical Investigations, dt./engl., transl. by G. E. M. Anscombe, Oxford: Basil Blackwell/New York: Macmillan 1953. T = Wittgenstein, Ludwig: Tractatus Logico-Philosophicus [1922], dt./ engl., ed. by David F. Pears and Brian F. McGuinness, London: Routledge and Kegan Paul 2 1971. Wittgenstein, Ludwig, 1961: Notebooks 1914–1916, ed. by Georg H. von Wright and Gertrude Elizabeth M. Anscombe, Oxford: Basil Blackwell.

Features of Indian Logic Questions of logical reasoning occupy a place as important in Indian philosophy throughout its history as in the Western tradition. This became to be generally recognized in the West quite late, and it is still in danger of being treated condescendently. It is possible to account for this attitude by paying attention to a number of facts. First of all, logic within the different Indian schools (darśana; literally: viewpoint) has never been cut off from its natural connection with philosophy of language or epistemology, and it appears rarely, therefore, as a subject that may be treated just formally. In addition, we face the problem that Indian treatises dealing with problems of logic use, in general, a highly developed technical language which, for a proper understanding, is heavily dependent on an uninterrupted written and oral tradition of commentative argument pro and con assertions in previous expositions. Yet, only rarely whole sequences of commentaries are available at present, not to mention the difficulty of identifying specific problems with their counterparts in the West in case such counterparts exist. But the most influential factors responsible for the widespread lack of attention among logicians to questions of logic in Indian philosophy seem to be the historical circumstances under which Indian philosophy became known in the West. It happened at the turn of the 18th to the 19th century in connection with the discovery of the historical relation among Indo-European languages and the establishment of academic Sanskrit studies. It was a time when research into mythological thinking that was considered to be the starting point for speculative philosophy attracted far more attention than research into first stages of scientific conduct, let alone investigations into the logic of inquiry itself. Logic was treated as being essentially contained in the inherited schemata of syllogistic, hence as a sterile affair which is not worth of serious attention, neither by the working scientist nor by the speculative philosopher. Not before the rise of modern formal logic with Gottlob Frege an adequate appreciation of earlier achievements in logic outside of syllogistic, e.g., by Stoicism or by Scholasticism, was gained. Indian logic suffered the same fate. Attention to questions of logic is intimately bound to attempts of questioning spiritual authorities. In India we observe it as attention to verbal tools which have to be established in order to know how to conduct disputes over the Vedic tradition. This happened around -500 when Vedic

Features of Indian Logic

93

authority became powerfully challenged by Buddhists, Jainists and the even more radical Lokāyatika-s – a term usually denoting collectively various groups of materialists, skeptics, and fatalists –, and it occurred once more in the course of many centuries when in defence of the Veda, by attempts of an at least partial rational reconstruction of its doctrines, the orthodox darśana-s develop. The oldest known term both for the procedure of reasoning and for the theory of reasoning is ‘ānvīkṣikī’.1 It appears around -300 in the famous treatise on economics, politics, and administration, the Arthaśāstra of Kauṭilya, and it is used there as the common feature of the ‘rational’ schools of his time: Sāṁkhya, Yoga, and Lokāyata, separating them from three other theories, the theory of the three Veda-s (trayī-vidyā), of economics (vārttāvidyā), and of politics (daṇḍinīti-vidyā). An even older term used by Buddha and Mahāvīra and their respective followers to characterize the method of dealing with philosophical questions, especially those called ‘avyākṛta’ (= unanswerable) – a list of ten in the early Buddhist Sūtra collection Majjhimanikāya contains questions concerning, among others, the finiteness or infinity of the world, and the identity or diversity of soul and body –, is ‘vibhajya’ (= splitting [into different cases]). A kind of linguistic analysis serves to transform a question which cannot be answered directly with ‘yes’ or ‘no’ into possibly several ones which can be answered or are merely pseudo-questions. In the following centuries this method has been turned into peculiar patterns of argumentation, conspicuously different in the Buddhist Mahāyāna school Mādhyamika and in later Jainism, in order to find out whether a certain predicate holds of something or not. The antiessentialist Mādhyamika-s who allow themselves only noncommittal negative judgments use a tetralemma, called ‘catuṣkoṭi’: something is neither (1) so, nor (2) not so, nor (3) both so and not so, nor (4) neither so nor not so. It seems to be derived from a similar fivefold formula of the skeptic Sañjaya. The all-comprehensive Jainas, however, who take every judgment to be affirmable if suitably qualified, use a formula of seven predications, the saptabhaṅgī, to show that every object splits into its manners of being given: (1) in a certain sense something is (e.g., a pot is black with respect to state, it is earthen with respect to substance), and (2) in a certain sense something is not (e.g., a pot is not black with respect to substance), and (3) in a certain sense something is and [afterwards] is not, and (4) in a certain sense something is inexpressible [i.e., the predicate is not applicable], and (5) in a certain sense something is and [afterwards] is inexpressible, and (6) in a certain sense something is not and [afterwards] is inexpressible, and (7) in a certain sense something is and [afterwards] is ____________ 1

Cf. Hacker 1958.

94

Philosophical Logic and Philosophy of Language

not and [afterwards] is inexpressible. Bimal Matilal has convincingly argued that this pattern which has given the Jaina darśana the name ‘syādvāda’ (= doctrine of it may be [thus or otherwise, depending on the point of view]) should be traced back to a formula with three members of the fatalist Gośāla, who, being a contemporary of Mahāvīra and Buddha like Sañjaya, is the founder of the Ājīvika sect, which says that everything is of triple character, e.g., something living is also non-living and, furthermore, both living and non-living.2 Closer scrutiny into the relations of the Lokāyata with Jainism and Buddhism on the one hand (concerning vibhajya) and with Sāṃkhya and Yoga on the other hand (concerning ānvīkṣikī) reveals that at least two traditions of dealing with methods of rational inquiry should be distinguished. There is a primarily person-oriented vāda-vidyā (= doctrine of debate, also called ‘tarka-śāstra’) in which rules of argumentation are studied, and a primarily matter-oriented pramāṇa-śāstra (= theory of knowledge, also called ‘jñāna-vāda’) in which reasons for having a certain knowledge are studied. Debates (kathā) have usually been classified threefold, easily to be mapped on the twofold division into dialectics and eristics in Antiquity: as proper debate or dispute (vāda) to find out on what is true, as contention (jalpa) to gain fame by winning, and as destructive argument (vitāṇḍā) to throw out the other party from the arena by any means without using a thesis of ones own. Hence, also sound argumentations of the Mādhyamika who only go for refutations, which implies that they consider the law of classical negation duplex negatio affirmat to be invalid, are sometimes called vitāṇḍā; in fact, the Mādhyamika use a kind of intuitionistic strong negation. The means of knowledge (pramāṇa) on the other hand, have been variously classified, but perception (pratyakṣa) and inference (anumāna) as distinct ways of finding out on what there is are chosen as basic in almost every darśana. As early as around -150, in the great subcommentary on Pāṇini’s Sanskrit grammar, in the Mahābhāṣya of Patañjali, the role of inference is described in terms which are reminiscent of explications in Stoicism: by inference knowledge of something not perceived or not perceivable is gained through knowledge of something perceived – the latter then being a sign of the former.3 Now, the vāda-vidyā becomes the central doctrine of one of the traditional six orthodox darśana-s, the Nyāya, which developed around the turn of the millenium, and it occupies also an important place within the Bud____________ 2 3

Matilal 1981. Cf. Scharfe 1961.

Features of Indian Logic

95

dhist Hīnayāna schools of the same time (famous paradigm cases are the non-canonical Milindapañhā around -100 that contains a philosophical disputation between the Greek-Bactrian king Menandros and the Buddhist monk Nāgasena, and the canonical Katthāvatthu around -250 that is the earliest extant treatise that contains discussions about rules of propositional logic as used in debates); both areas belong to the sources for the logical school of Buddhism as initiated by Dignāga (ca. 460–540). In contrast, the discussion of inference in semiotic terms by the early Grammarians or Vaiyākaraṇa-s must have stood in close relation with similar conceptions in pre-classical Sāṃkhya, where in distinction to all the other darśana-s, which treat perception as the first pramāṇa, inference occupies the first place among them. From later polemical discussions, especially by Dignāga, we know of an intra-sāṃkhyan transformation of an early doctrine about inference by the Sāṃkhya-teacher Vṛṣagaṇa (around 300) – the reconstruction of the transformation is due to Erich Frauwallner4 – which only afterwards got amalgamated with the vāda-tradition in Nyāya, partly via ist twin-darśana Vaiśeṣika. And it is this combination of argumentation theory with theory of knowledge (by inference) which, under the title of a theory of logic (nyāya-vidyā), Vātsyāyana (ca. 350–425) in his commentary on the Nyāya-sūtra-s, i.e., the Nyāya-bhāṣya, identifies with the ānvīkṣikī in Kauṭilya’s Arthaśāstra. We, thus, get also an explanation of the use of ‘pramāṇa-śāstra’ for the Nyāya-darśana as a whole. A reconstruction of the early doctrine of inference faces unsurmountable difficulties, because not even the oldest known Indian commentaries are in accord about the precise meaning of the three kinds of inferences which have come down to us under the titles ‘with previous’ (pūrvavat), ‘with the rest’ (śeṣavat), and ‘seen with respect to the common’ (sāmānyato dṛṣṭam). Among modern readings which use the literal meaning of ‘pūrva’ and ‘śeṣa’ in grammar, the most convincing one is the identification of a pūrvavatinference with modus ponendo ponens (= if ‘the previous’ holds [then ‘the following’ holds], i.e., by affirming A in A→B affirming B, or: A, A→B≺B), of a śeṣavat-inference with modus tollendo ponens (= [if one member of an alternative which holds, does not hold] then ‘the rest’ holds, i.e., by negating A in A∨B affirming B, or: A∨B,¬A≺B), and of a sāmānyato-dṛṣṭaminference with an inference by analogy, where ‘the common’ acts as the tertium comparationis, e.g., if geese are birds then ducks are birds, because both have feathers, provided for the tertium comparationis holds: ‘havingfeathers implies being-bird’. Vṛṣagaṇa transformed this ternary division of inferences into a binary one. He keeps ‘seen with respect to the common’, yet divides it into pūrvavat- and śeṣavat-inferences, and adds ‘seen with ____________ 4

Frauwallner 1958.

96

Philosophical Logic and Philosophy of Language

respect to the specific’ (viśeṣato dṛṣṭam), where the latter is nothing but an instantiation of the former, e.g., in the case of these feathers and this bird. At this stage further treatment of inferences along these lines as well as other more general inquiries into questions of logic took only place in Nyāya and in the logical school of Buddhism. It happened in such a way that argumentation theory and theory of knowledge became intimately interconnected. The once independent inquiries into logical reasoning in Sāṃkhya were, for all we know at present, given up. This did not happen in the case of the discussions of questions of logic in the school of Grammarians as initiated around -400 by Pāṇini’s Sanskrit grammar and which sometimes was treated even as a philosophical system: Pāṇinīya darśana. The Grammarians from Kātyāyana onwards, whose Vārttika on Pāṇini around -250 had been the subject matter of Patañjali’s Mahābhāṣya, have treated problems of inference in connection with problems of linguistic representation, mainly of grammatic features, of course. Their work was developed in close interaction with ideas in another at that time gradually evolving orthodox darśana, the Mīmāṃsā, whose main concern has been the meticulous interpretation of vedic injunctions (vidhi). Semantic analysis together with a careful distinction of object language and metalanguage resulting in explicit metalinguistic rules of interpretation (paribhāṣā) for word composition and sentence composition both syntactically and semantically flourishes. Technical innovations abound: there is, e.g., a special particle, ‘iti’, which, if added at the end of a text, turns it into a quotation; syntactic and semantic treatment of negation leads to explicit formulations of the rule of contraposition; even the importance of the relative position of negation sign and modal operators is clearly observed. An example relevant for Mīmāṃsaka-s from Mahābhāṣya is: ‘it is obligatory to eat [nothing but] five five-clawed animals’ implies ‘it is forbidden to eat the others’,i.e., ∆!(E(x)→F(x))≺¬∇!(E(x)∧¬ F(x)).5 From the discussion of inference in comparison with perception as tools of knowledge in Mahābhāṣya we learn that it is the semiotic frame of reference which leads to the treatment of implication under the title of connection (sambhanda): to make an inference from A to B valid there must exist a connection between the references of the associated terms A* and B*; then A* is called a sign (liṅga) for B* (A and A* are by definition related in the following way: A* is satisfied iff A is true). Though it took many centuries that the concept of implication – the term ‘vyāpti’ (= pervasion) was eventually used for it, though not everywhere – became clarified, and this happened as a result of the critique launched against the logic of Nyāya by the Buddhist logicians, we, already, find in the Mahā____________ 5

Cf. Staal 1969.

Features of Indian Logic

97

bhāṣya the use of the terms ‘anvaya’ and ‘vyatireka’ for A*≺ B* and its contraposition ¬B*≺ ¬A*, as it is the case in Buddhist logic but not in the texts of the Naiyāyika-s. The examples show that the Grammarians use a partition of inferences into inferences pratyakṣato dṛṣṭam (= seen by referring to perception) and sāmānyato dṛṣṭam which obviously reoccurs in Buddhist logic as a partition into two kinds of modus ponens inference: the first one based on causal implications and the second one on conceptual implications. Dharmakīrti (ca. 600-660), the great follower of Dignāga , uses the terms ‘kāryānumāna’ (= inference based on effect) and ‘svābhāvānumāna’ (= inference based on essence) for these two kinds. The example of an inference sāmānyato dṛṣṭam as well as its instantiation viśeṣato dṛṣṭam given above may now be combined and rephrased as a kāryānumāna, which yields almost the standard form of an Indian syllogism as it is found in the Nyāya-sūtra: ‘here [at the place of the duck] are feathers’ therefore ‘here is a bird’, because having-feathers implies being-bird, i.e., only birds as causes grow feathers as an effect such that feathers become signs for birds; places with geese are supporting examples [for the general implication]. Instead of ‘feather’ and ‘bird’ Patañjali uses, e.g., ‘leaf’ and ‘tree’, or grammatical features and their meaning. If it is realized that the doctrines of the Grammarians and the Mīmāṃsaka-s are characterized by a systematic interrelation of logic and philosophy of language, whereas the doctrines of the Naiyāyika-s and the Buddhist logicians are united by their insistence on the interconnectedness of logic and epistemology, it does not come as a surprise when we are confronted with a severe dispute between Mīmāṃsā and Nyāya carried on for almost the whole first millenium of the Christian era about the nature of the relation between word (śabda) and object (artha). The Mīmāṃsaka-s, on the one hand, who tried to use the philosophy of language as a means of setting up epistemology, argue for a φύσει-theory: since śabda – and they refer with this term primarily to the orally preserved Veda – is eternal (nitya), words refer to their objects by nature, i.e., by a kind of inherent energy (śakti). Hence, any verbally represented knowledge is prima facie true; if there is a claim to the contrary there should be given reasons for it. Due to their main concern with vedic injunctions their theory contains another interesting feature. It is a kind of reduction principle reminiscent of the Pragmatic Maxim as introduced by Charles Sanders Peirce: every word meaning and every sentence meaning has to be related to actions, e.g., ‘this is a rope’ gets semantically reduced to ‘with this rope it is possible to fasten a cow’. The Naiyāyika-s, on the other hand, who tried to justify doctrines pertaining to the philosophy of language epistemologically, argue for a

98

Philosophical Logic and Philosophy of Language

θέσει-theory: since śabda, being a verbal representation of some knowledge which is essentially independently available by perception and by inference from perception, may appear as the result (kārya) of linguistic actions chosen in many different ways, words refer to their objects by convention. Hence, any verbally represented knowledge is a mere claim which has to be investigated as to whether it is true or false. Both positions, independently of how they are supported by further conceptual constructions, belong to epistemological realism. Their difference may be stated as follows: the Mīmāṃsaka-s treat objects as universals – ākṛti, i.e., form, is what a term refers to –, and the Naiyāyika-s treat objects as particulars which has the effect that what a term refers to partakes in general of universality (sāmānya) and of individuality (vyakti). In the first case language is treated under its type aspect – the terms are schemata – , and in the second case under its token aspect – the terms are actualizations. Again it is no surprise that Mīmāṃsā is much concerned with norms which serve to ensure invariance of language use, whereas Nyāya pays first of all attention to the factual variability of language use. In fact, every means of knowledge (pramāṇa) holds a priori in Mīmāṃsā since its adherents rely on the unconditional validity of vedic revelation (śruti; literally: the heard) and, therefore, enforce any effort to preserve its exact wording; the means of knowledge hold a posteriori in Nyāya because there is unconditional validity only of vedic tradition (smṛti; literally: the remembered) which by its very existence proves that certain experiences can repeatedly be made. But in both cases it is essential to be able to infer from what is perceived to what is not perceived or not perceivable. This can be done by means of a theory of inference which received its standard form of a fivemembered syllogism (pañcāvayava vākya; literally: sentence having five parts) in Nyāya and became a point of reference also for the other darśana-s when they either use or discuss logical reasoning. Due to the criticism by the Buddhist logicians Dignāga and Dharmakīrti the five-membered form has been restricted to cases of ‘inference for the sake of others’ (parārthānumāna); in the case of an ‘inference for the sake of oneself’ (svārthānumāna) only the first three or the last three members are necessary, and these members should furthermore be treated as merely conceived term complexes, i.e., as verbal signs in significative function [A*], which are not in need of being uttered by sentences, i.e., as verbal signs in communicative function [A], as in the five-membered case. The paradigm of a fivemembered syllogism looks as follows:

Features of Indian Logic

99

assertion (pratijñā)

1. [this] mountain has fire

reason (hetu)

2. because of the smoke

example {and counterexample} (udāharaṇa)

3. like [it is] in the kitchen {unlike [it is] in the pond}

application (upanaya)

4. and this [is] so (= this mountain has smoke)

conclusion (nigamana)

5. therefore [it is] so (= this mountain has fire)

The mountain is the perceived object of cognition (pakṣa; literally: wing). The term ‘pakṣa’ originally refers to the place where a thesis in a dispute is put forward; later on ‘pakṣa’ together with ‘pratipakṣa’ are used for thesis and counterthesis. In the syllogism ‘pakṣa’ refers just to the locus where the inference takes place, i.e., the locus [of this] mountain. The smoke is the perceived sign (liṅga) used as reason (hetu), which is indicated by the ablative case. The fire, finally, is the non-perceived consequence (sādhya), and, here again, the term ‘sādhya’ originally refers to the whole proposition to be proved, alike ‘pakṣa’ in its meaning as thesis. Of course, the validity of the argumentation is dependent on the validity of the implication between reason and consequence, which originally had only been ‘perceived’ paradigmatically, by example and counterexample. After the general character of implication – in the paradigm case given above: ‘wherever [is] smoke there [is] fire’ – had been clarified mainly by the criticism of Dignāga and Dharmakīrti, implication soon became the central topic in Indian logic. Several terms for the concept of implication were used, among them ‘vyāpti’, which gained common usage through the Mīmāṃsā-teacher Kumārila (ca. 620–680). The vyāpti of smoke and fire, i.e., smoke≺fire, has to be rendered literally as ‘smoke pervaded by fire’. Besides ‘vyāpti’ also ‘avinābhāva’ (= inseparable [connection]) was used for implication, especially by Jaina logicians. With the Nyāya-teacher Udayana (ca. 975–1050) the theory of vyāpti began to grow extensively and reached its peak in the Tattvacintāmaṇi of Gaṅgeśa (ca. 1300–1360). It was mainly due to the activity of Udayana that Nyāya became united with Vaiśeṣika, thus creating NewNyāya (Navya Nyāya) or Nyāya-Vaiśeṣika. In a table with nine entries called ‘hetucakra’ (= wheel of reasons) which is found in his Hetucakra-ḍamaru, Dignāga was able to give a complete survey of the valid relations between reason and consequence by referring to the possible relations of the syllogistic kind ‘a’ (all), ‘e’ (no) and ‘i’ (some, but, as always in Indian logic, under exclusion of all), which

100

Philosophical Logic and Philosophy of Language

may hold between either the examples s, being loci of sādhya ‘like pakṣa’ (sapakṣa), or the counterexamples v, being loci of the complement of sādhya ‘unlike pakṣa’ (vipakṣa), i.e., v = , and the sign h (hetu): vih

veh

vah

ae ai

sah seh sih

ei

aa ee

ii

ea ia

ie He relies on his critical discussion of Vasubandhu’s the Younger (ca. 400– 480), his teacher’s, Vādavidhi where in the context of a treatment of rules for disputations for the first time in Buddhist logic the five-membered syllogism got reduced to the first three members which were called ‘sādhana’ (= proof). (Of course, there also occurred elaborate discussions of how to try refutations (dūṣaṇa) by finding mistakes in a proof; for this purpose the vāda-vidyā did provide techniques of using counterfactuals (tarka): if there were A there would be B, but B is not, therefore not A.) Of special importance was Dignāga ’s investigation into the rule of the ‘three characteristic features of a sign [as reason]’ (trairūpya liṅgatva) which had found an unclear formulation in Vādavidhi and is still handled unsatisfactorily by Dignāga’s contemporary, the Vaiśeṣika-teacher Praśastapāda (ca. 500–550). The ‘wheel of reasons’ which supplies the first successful formal treatment of logical inference in Indian logic, is carried by intentions quite alike those of Aristotle in his syllogistic though of a set-up far more alike the one in Stoic logic than in syllogistic. Using the ‘wheel of reasons’ Dignāga states that an argumentation which infers B (e.g., ‘at the place of this mountain is fire’) from A (e.g., at the place of this mountain is smoke) is valid if and only if the following ‘three characteristic features’ are present: (1) the sign occurs in the object (pakṣadharmatva; literally: the beingmodified of the object by the property [dharma]), (2) the sign occurs only where the consequence [sādhya] occurs, i.e., in loci alike the object, and (3) the sign is absent where the consequence is absent, i.e., in loci unlike the object. Now, as loci alike the object are those of sādhya, i.e., those of ‘similar examples’ (sādharmya dṛṣṭānta – the Buddhist logicians use ‘dṛṣṭānta’ instead of ‘udāharaṇa’), and loci unlike the object those of the complement of

Features of Indian Logic

101

sādhya, i.e., those of ‘dissimilar examples’ (vaidharmya dṛṣṭānta), we may restate the three characteristic features in the paradigm case, using ‘ιp’ to symbolize the deictic description ‘this mountain’, as follows: (1) ιp ε h (= at [the place of] this mountain is smoke), (2) h≺s (anvayī vyāpti: wherever [is] smoke there [is] fire), and (3) ¬s≺¬h (vyatirekī vyāpti: wherever [is] not fire there [is] not smoke). It was the Buddhist logician Dharmottara (ca. 730–800) who first observed that (2) and (3) should be considered as logically equivalent, because any one of the two clauses expresses the vyāpti of smoke and fire. He disregarded Dignāga’s insistence on the necessity of all the three features conjunctively. Dignāga’s discussion of the insufficiency of just the conjunction of (1) and (3), is a clear indication that he regards the law of double negation: ¬¬A≺A as not being generally valid. This is completely in tune with the primary position of negative concepts in Buddhism in general and with Dignāga’s own nominalism in setting up a theory of meaning for general terms by ‘expulsion [from the complement]’ (apoha) in particular.6 The term ‘go’ (= cow), e.g., is defined by the infinite conjunction: ‘non-man ∧ non-dog ∧ non-lion ∧ …’ in order to make sure that universal features have to be inferred and cannot be perceived like singulars which alone make up reality. Dharmakīrti intensifies the logical research along these lines in his Hetubindu, especially by stressing that the vyāpti between the references of two terms must not be understood as merely [general] togetherness (sāhacarya) but acts as a kind of rule (niyama). He distinguishes three kinds of vyāpti. Two of them are kinds of anvayī vyāpti, i.e., of h≺s; the first – in the case when hetu h is that which is effected (kārya) [only] by sādhya s – is the one responsible for causal inference (kāryānumāna), the second – in the case when hetu h is nothing but the own-being (svabhāva) of sādhya s (sādhya-svabhāvatva), i.e., s is part of the concept of h – is the one responsible for conceptual inference (svabhāvānumāna). The last kind of implication – vyatirekī vyāpti: ¬s≺¬h – is a connection of hetu and sādhya ex negativo which leads to an inference due to non-recognition (anupalabdhyānumāna). Without further consideration of the problems connected with an explicit incorporation of apoha-vāda into a discussion of the ‘wheel of reasons’ one can summarize what it says in the following way: the argumentation for ιp ε s (= at [the place of] this mountain is fire) is valid if and only if both ιp ε h and h≺s hold. Dignāga showed that exactly two of the nine conjunctions of the hetucakra, namely ‘ae’, i.e., sah ∧ veh, and ‘ie’, i.e., sih ∧ veh, (they are underlined in the table above) which are each logically equivalent to has (= h≺s ), make the inference from ιp ε h to ιp ε s valid. ____________ 6

Cf. van Bijlert 1992.

102

Philosophical Logic and Philosophy of Language

It is, furthermore, obvious that attempts to read the hetucakra as a proof of the syllogistic mood barbara with pakṣa as minor term, sādhya as major term, and hetu as middle term ( pah∧has≺pas) are misguided, because no sentence of the type ‘at all mountains there is smoke’ occurs in the Indian inference schema. The Naiyāyika-s who did not accept the move of the Buddhist logicians of separating strictly ‘inference for the sake of oneself’ (svārthānumāna) from ‘inference for the sake of others’ (parārthānumāna), i.e., of keeping the logical question of an inference being valid, because the implication between premiss and conclusion holds formally, apart from the psychological question of being convinced of this validity through a debate about this issue, continued to treat logic, especially questions of inference, in the context of both epistemological and psychological issues. As a famous example, in Gaṅgeśa’s Tattvacintāmaṇi the inference from ιp ε h to ιp ε s is treated as a scheme that also includes the case of going from a (perceived) word to the (non-perceived) object it refers to.7 Therefore, the knowledge of smoke (H: that at the place of this mountain is smoke) is considered to be an efficient cause (nimitta kāraṇa, or karaṇa) of the knowledge of fire (S: that at the place of this mountain is fire). In Navya Nyāya the corresponding relation between such two pieces of knowledge H and S is called ‘vyāpti-jñāna’ (= knowledge of implication); with respect to the standard example it is: vahni-vyāpyo dhūmah (= smoke being pervaded by fire). In fact, every piece of knowledge is treated as connected with the cognizing self (ātman) by inherence (samavāya), which is held to be the same relation as the one between substance (dravya) and quality (guṇa), and between a whole (avayavin) and one of its parts (avayava). The process of producing an effect from a cause is called an operation (vyāpāra) which, in the special case of the operating knowledge of implication, is signified by the term ‘parāmarśa’ (= conceiving); with respect to the standard example it is: vyāpti-viśiṣṭa-pakṣa-dharmatā-jñāna (= knowledge of the occurrence of that which is qualified by vyāpti [i.e., vyāpti of smoke and fire] at the place of the object [i.e., at the mountain]). The cause H together with the parāmarśa yields the effect S , and this may be understood as applying the rule of universal instantiation to ∧x(x ε h → x ε s), thus arriving at ιp ε h → ιp ε s whence, in view of H and S being the respective nominalizations of the elementary sentences ιp ε h and ιp ε s, by applying the rule of modus ponens, we get S from H. In the old Nyāya H alone was considered to be the cause of S. Hence, after the third member of the five-membered syllogism had been amended by explicitly stating the vyāpti ‘wherever [is] smoke there [is] fire’ the ____________ 7

Cf. Ingalls 1951.

Features of Indian Logic

103

second member became interpreted as an abbreviation for ‘because of the smoke which is pervaded by fire’. This remained the standard interpretation of the classical tradition up to the present where the textbook of logic Tarkasaṃgraha of Annambhaṭṭa (around 1575) is still used in Sanskrit schools as a kind of logical propaedeutic. Further studies in Navya Nyāya after Gaṅgeśa concerning vyāpti (at Mithilā/Bihār) became increasingly subtle and have led to the construction of a highly technical Sanskrit. It enabled the logician Raghunātha (ca. 1475–1550), who had the intellectual power to start a school of his own (at Navadvīpa/Bengal), to develop, among many other things, a higher order logic of relations which is sufficient to handle logical composition including quantification in all detail. The assertion, e.g., of the adjunction ‘the place at which there is fire is a place at which there is water or the place at which there is fire is a place at which there is a mountain’ is expressed – as shown by Ingalls8 – by asserting a certain property of fire: fire ε {[where there is] water or at [the place of] a mountain}. Many details still wait for reconstruction by means of modern formal logic. For the development of Nyāya the work of Dignāga and his followers has been of decisive influence, not only with respect to logic proper concerning the concept of implication (vyāpti) but also with respect to epistemology: From the time of Dignāga onwards Nyāya had to defend its realism against the radical nominalism of Dignāga and his school. Of similar importance for the development of Mīmāṃsā was the elder contemporary of Dignāga, the Grammarian Bhartṛhari (ca. 450–510). He is the founder of a particular branch of the orthodox darśana Advaita-Vedānta which is called Śabdādvaita (= non-duality with respect to word) because it is characterized by the claim that language alone, if taken in its generic aspect, is real. Bhartṛhari draws a radical conclusion from the Mīmāṃsādoctrine that primitive terms refer to universals: every term irrespective of its level of composition both vertically and horizontally articulates universals. It has the result that one branch of Mīmāṃsaka-s (the Prābhākara-s) approaches Buddhism while the other branch (the Bhāṭṭa-s, i.e., the followers of Kumārila) is driven towards Nyāya. In Śabdādvaita uttering a sentence does not represent a particular as such and such but is a means – in fact the only one – of making something universal accessible. Reality is linguistic in nature and can be cognized through a particular means of knowledge (pramāṇa), which is called by Bhartṛhari ‘pratibhā’ (= intuition). By intuition the schematic or generic character of terms – the sphoṭa (= bursting) – on various levels, e.g., on the phonetic or on the semantic one (varṇa-sphoṭa [= Phonem] or pada-sphoṭa ____________ 8

L.c., p. 63.

104

Philosophical Logic and Philosophy of Language

[= Morphem]), is recognized all of a sudden, and only that deserves to be called knowledge of what there is. Perception which is always predicatively determined perception (savikalpaka pratyakṣa) – and in this respect Bhartṛhari disagrees with Dignāga who considers pure perception to be predicatively undetermined (nirvikalpaka pratyakṣa) – refers to particulars only, and its instantiations being action-tokens on a par with the elements of concrete speech are themselves nothing but particulars; hence, by inference again only particulars can be reached. Neither traditional pramāṇa is fit to make the universal accessible, one needs pratibhā for it. There is agreement between Dignāga and Bhartṛhari in as much as they treat language (śabda) and conceptual construction (kalpanā) as two sides of the same coin. They differ when calling this coin either ‘śūnyatā’ (= emptiness; Dignāga) or ‘brahman’ (= ‘śabda’ or ‘word’ in the sense of ‘λόγος’), because for Bhartṛhari, only universals are real, for Dignāga, only singulars, i.e., the actualizations of actions of pure perception. Luckily, the logic of dealing with universals and particulars remains essentially unaffected by this opposition.

Features of Indian Logic

105

References Barlingay, S. S., 21976: A Modern Introduction to Indian Logic [1965], New Delhi: National Publishing House. van Bijlert, Victor, 1992: Apohavāda in Buddhist logic, in: Sprachphilosophie/Philosophy of Language/ La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz and Georg Meggle, vol. 1, Berlin/New York: Walter de Gruyter, pp. 600–609. Chi, R.S.Y., 1969: Buddhist Formal Logic I (A Study of Dignāga’s Hetucakra and K’uei-chi’s Great Commentary on the Nyāya-praveśa), London: The Royal Asiatic Society of Great Britain and Ireland. Foucher, Albert, éd., 1949: Le Compendium des topiques (Tarka Saṃgraha) d’Annambhaṭṭa avec des extraits de trois commentaires indiens (texte et traduction) et un commentaire, Paris: Adrien-Maisonneuve. Frauwallner, Erich, 1957: Vasubandhu’s Vādavidhih, in: Wiener Zeitschrift für die Kunde Süd- und Ostasiens u. Archiv für indische Philosophie 1, pp. 104–146. Frauwallner, Erich, 1958: Die Erkenntnislehre des klassischen Sāṃkhya-Systems, in: Wiener Zeitschrift für die Kunde Süd- und Ostasiens u. Archiv für indische Philosophie 2, pp. 84–139. Frauwallner, Erich, 1961: Landmarks in the History of Indian Logic, in: Wiener Zeitschrift für die Kunde Süd- und Ostasiens u. Archiv für indische Philosophie 5, pp. 125–148. Guha, Dinesh C., 21979: Navya Nyāya System of Logic. Some Basic Theories and Techniques [Varanasi 1968], Delhi: Motilal Banarsidass. Hacker, Paul, 1958: Ānvīkṣikī, in: Wiener Zeitschrift für die Kunde Süd- und Ostasiens u. Archiv für indische Philosophie 2, pp. 54–83. Hayes, Richard P., 1980: Diṅnāga’s Views on Reasoning (Svārthānumāna), in: Journal of Indian Philosophy 8, pp. 219–277. Herzberger, Radhika, 1986: Bhartṛhari and the Buddhists. An Essay in the Development of Fifth and Sixth Century Indian Thought, Dordrecht: Reidel. Ingalls, Daniel H. H., 1951: Materials for the Study of Navya-Nyāya Logic, Cambridge Mass.: Harvard University Press. Lorenz, Kuno, 1998: Indische Denker, München: Beck. Matilal, Bimal K., 1971: Epistemology, Logic, and Grammar in Indian Philosophical Analysis, The Hague/Paris: Mouton. Matilal, Bimal K., 1981: The Central Philosophy of Jainism (Anekānta-Vāda), Ahmedabad: L. D. Institute of Indology. Matilal, Bimal K./Evans, Robert D., eds., 1986: Buddhist Logic and Epistemology. Studies in the Buddhist Analysis of Inference and Language, Dordrecht: Reidel. Oberhammer, Gerhard, 1963: Ein Beitrag zu den Vāda-Traditionen Indiens, in: Wiener Zeitschrift für die Kunde Süd- und Ostasiens u. Archiv für indische Philosophie 7, pp. 63–103. Patnaik, Tandra, 22007: Śabda. A Study of Bhartṛhari’s Philosophy of Language [1994], New Delhi: D.K. Printworld.

106

Philosophical Logic and Philosophy of Language

Potter, Karl H., ed., 1977: Encyclopedia of Indian Philosophies (vol. II: Indian Metaphysics and Epistemology: The Tradition of Nyāya-Vaiśeṣika up to Gaṅgeśa), Princeton N.J.: Princeton University Press. Ruegg, David S., 1977: The uses of the four positions of the catuṣkoṭi and the problem of the description of reality in Mahāyāna Buddhism, in: Journal of Indian Philosophy 5, pp. 1–72. Scharfe, Hartmut, 1961: Die Logik im Mahābhāṣya, Berlin: Akademie-Verlag. Solomon, Esther A., 1976/78: Indian Dialectics. Methods of Philosophical Discussions I–II, Ahmedabad: B. J. Institute of Learning and Research. Staal, Johan F., 1962: Negation and the Law of Contradiction in Indian Thought. A Comparative Study, in: Bulletin of the School of Oriental and African Studies (London) 25, pp. 52–71. Staal, Johan F., 1969: Sanskrit Philosophy of Language, in: Current Trends in Linguistics V, ed. by Thomas A. Sebeok, The Hague/Paris: Mouton, pp. 499–533. Steinkellner, Ernst, 1971: Wirklichkeit und Begriff bei Dharmakīrti, in: Wiener Zeitschrift für die Kunde Südasiens u. Archiv für indische Philosophie 15, pp. 179– 211.

Part II Methods in Philosophy, in Art, and in Science Methods

in

Philosophy,

in

Art,

and

in

Science

The Concept of Science Some remarks on the methodological issue ‘construction’ versus ‘description’ in the philosophy of science I Rationalism and empiricism, the kindred branches of modern western philosophy since its inception with Descartes and Hobbes, have made much of the distinction between mind and body. The corresponding schism between the act and the given, reminiscent of the Aristotelian categories ποιεῖν and πάσχειν, has consequently to be looked at differently within the philosophy of nature and the philosophy of mind. We have to consider rationalism and empiricism independently in both areas though usually they occupy attention only within the philosophy of nature. So it happens, e.g., in phenomenology, that a kind of rationalism with respect to the outer world is matched by an equally stringent kind of empiricism – though hardly called this way – with respect to the inner world. Among current disputes, the well-known discussion on innateness1 is another example which shows the relevance of that phenomenon, especially in connection with a widespread misconception of innateness in Kant.2 Not ideas are innate, but our ability to make them is innate. And, certainly, the discussion is not yet closed on the issue whether dispositions – under some biological model – can be treated as facts rather than as mere possibilities of facts, i.e., as something like second-order-facts, if that phrase makes sense at all. Therefore, when I use the terms ‘rationalism’ and ‘empiricism’ together with their cognates in the following remarks, they should be understood in that general sense, not restricted to their use in the philosophy of nature. ____________ 1 2

Cf. the collection of essays, presented at the symposium on innate ideas in Boston (Dec. 1966), especially the ones by N. Chomsky, N. Goodman and H. Putnam, in: Synthese 17 (1967); a recent survey on the dispute over innate ideas in: Stich 1996. Cf. the reference to Konrad Lorenz as a representative of an erroneous adaptation of Kantian arguments to findings of a special natural science, here: ethology, in: von Weizsäcker 1979, pp. 137–138.

110

Methods in Philosophy, in Art, and in Science

The two main developments originating through well-known problems of epistemology in either of these philosophical positions may now be characterized in the following way: Out of rationalism emerges transcendentalism to secure a unique set-up of at least the natural sciences, mathematics included. Empiricism, on the other hand, gave way to evolutionarism, some kind of free choice – or necessary change – principle to be used for starting, e.g., the sciences or any other human artefact. It is common opinion to treat an epistemology of the first kind as the only way out of epistemic scepticism taken seriously, whereas an epistemology of the second kind bounds scepticism by some common sense relativism which implies dropping any reliance upon science as a substitute for religion concerning matters of fundamental world view. This frame for dealing with the claims of scepticism hides a difference of presuppositions in the philosophy of science which, spelled out, may well carry a chance with it to provide for a truly unified treatment of science. I have in mind the difference of treating science as a way of presentation and as a way of research, i.e., the difference of the old person-oriented ars iudicandi on the one hand and the matter-oriented ars inveniendi on the other hand. Just as a remark I should add that this Leibnizian difference of ‘analysis’ and ‘synthesis’, as he calls these two ‘artes’ alternatively, is more general than Hans Reichenbach’s distinction between a ‘context of justification’ and a ‘context of discovery’, both relating to propositions in their function as scientific hypotheses and thus serving a better understanding of the difference between deductive and inductive methods.3 Transcendentalism searches for justifiable presentations (being true theories of certain domains of objects) whereas evolutionarism represents a way of adequate research (being significant encounters with certain kinds of objects), such that in the first case we get well-founded sequences of propositions, yet in the second case a well-determined network of mutually related objects. At once a further complication arises. To search for sequences of propositions is certainly not a purely linguistic matter. Instead of just giving a construction of certain objects on the language level, the crucial issue is to judge upon their ‘relation’ to the object level with the aim of securing their truth. Analogously, to represent a network of objects cannot be done on the object level alone; representation is bound to rely on linguistic means with the aim to determine the objects by precise descriptions. ____________ 3

Cf. Reichenbach 1938, esp. chap. I, and compare it with the context of Leibniz’ terms as expounded, e.g., in: Hermes 1969.

The Concept of Science

111

In order to avoid erroneous identifications, I have deliberately used here the terms ‘search’ and ‘represent’ to refer to activities within science as presentation and science as research, respectively. For, certainly, there is second-order research concerning presentations, e.g., within ‘science of science’, and second-order presentation concerning research, e.g., within the well-known ‘logic of inquiry’, and neither should be identified with what I called ‘search’ and ‘representation’, respectively. Search within presentations is search for true presentation, whereas representation within researches is representation of significant research. What is at stake from the purely linguistic point of view is the question of how the verification (and falsification) of formulae is interrelated with the signification of terms. More generally, and in traditional terminology, we can say: Both questions, the justification of propositions and the constitution of objects, belong together, yet must not be confounded. If the question of constitution is falsely treated as belonging to the problem of justification, it yields evolutionarism as a brand of radical empiricism.4 And again, if the question of justification is falsely treated as belonging to the problem of constitution, it yields transcendentalism as a brand of radical rationalism. To keep both questions within their proper bounds has a chance of success only, if their mutual dependency is treated clearly and distinctly. This means especially to ask for methods to translate theories including states of theories – considering theory-change – into each other: In what sense may two synchronically or diachronically different theories have the same content, though they say different things of different entities. The most prominent example for the radical empiricism which arises when problems of constitution are treated as if they were problems of determination, i.e., as if they concerned investigations into the truth conditions of assertions about the constituted objects, can be found in the evolutionary pragmatism of C. S. Peirce.5 With slight modifications only, this is equally true of the ‘Analytische Wissenschaftstheorie’ as it grew out of the Logical Empiricism of the Vienna Circle. For example, the usual set-up of formalized theories has never been seriously questioned, i.e., the start with given domains of objects on the one hand and sets of predicates together with suitably chosen axioms about ____________ 4 5

This term was used by William James for his version of pragmatism which is exactly in line with the claim just made; cf. the collection of essays in: Corti 1976. Of course, the radical empiricism of W. James may be included, too, since this issue may be dealt with quite independently from the dispute between James and Peirce on the meaning of the term ‘pragmatism’; cf., e.g., Peirce’s argumentations against first intuitions to secure cognition in: CP 5.213–5.263.

112

Methods in Philosophy, in Art, and in Science

those objects on the other hand. There is a freedom of choice in both respects – phenomenalistic systems, for example, may compete with physicalistic systems, and preference for some set of primitive notions and principles should always be treated as a contingent fact by itself – but there is usually no hint as to how somebody can acquire a position enabling him actually to choose among alternatives. This again remains a historical and thus contingent fact. Another example not from the sciences, but from the humanities, is the methodological position called ‘hermeneutics’. Here, too, there is no serious question of how to get hold of the different domains of objects – most prominent among them are written texts – , they, rather, are viewed through the different scientific approaches to them, e.g., their ‘interpretations’. Hence, it is forbidden to ascribe an independent status to the objects.6 In the converse case, the most prominent example for the radical rationalism which arises when problems of justification are treated as if they were problems of constitution, i.e., as if they were concerned with investigations into the conditions of possible experience, of how objects of experience have to behave in order to be accessible to knowledge, may, of course, be found in the transcendental idealism of Kant.7 It is, therefore, not accidental when proponents of the ‘Konstruktive Wissenschaftstheorie’ use arguments akin to Kantian ones to substantiate the claim that what is called ‘protophysics’ can serve as an apriori foundation for physics. In a similar vein, attempts of a ‘protosociology’ shall serve as a clarification of the methodological position called ‘dialectics’ that tries to furnish a nonempirical foundation for the social sciences.8 Due to the simplifications of this sketch, I will not be able to take into account the numerous mixed cases that derive from insights into certain shortcomings of the respective positions as they grew out of recent scientific discussions. For example, a certain transcendental framework got introduced into hermeneutics;9 similarly, empirical boundary conditions entered the ____________ 6 7 8

9

Cf., e.g., the discussions on hermeneutics as a scientific method at the 2nd Loccumer Kolloquium, Oct. 1971, as documented in: Gerber 1972. This derives from the fact that Kant never disputes the reality of knowledge, i.e., of Newtonian physics, but tries to clarify the conditions of its possibility; cf., e.g., the last chapter (§ 15) in: Mittelstraß 1970. The concept of a general theory of verbal communication as ‘protosociology’ was introduced in the postscript to the new edition of Erkenntnis und Interesse, cf. Habermas 21973; a related though competing approach may be found in: Lorenzen/ Schwemmer 1973, chap. III, 4. Cf. the extensive discussion on ‘Das Apriori der Kommunikationsgemeinschaft’ in: Apel 1973; likewise Bubner 1973.

The Concept of Science

113

argumentations in dialectics.10 Both, the concentration on ‘the logic of evolution’, the rational conditions of any change in the Frankfurter Schule, and on ‘the evolution of logic’, the empirical change of any theory, in present-day analytic philosophy of science, show that dealing with problems of amalgamation between transcendentalism and evolutionarism is considered to be the essential and the most hopeful line of investigation now. I will not go into details here, but rather stress certain features of the discussion between the analytic and the constructive philosophy of science that are relevant for the epistemological issue in the sciences I am concerned with.

II For convenience of presentation, I will start with the discussion of a thesis that Harald Wohlrapp has convincingly defended recently.11 The analytic philosophy of science on either of its three main stages, Carnap-Stegmüller’s empiricism, Popper-Lakatos’ rationalism, and Kuhn-Feyerabend’s historism, should – according to Wohlrapp's claim – essentially be understood as concerned with science as a way of research, whereas the constructive philosophy of science of the Erlanger Schule, and of related positions, is basically concerned with science as a way of presentation. As an important consequence, the difference of criteria for what shall be considered as science can be stated. The criteria of science as research are essentially those of success, derived from the actual procedures of working scientists. Necessary conditions are, e.g., the use of well-defined predicates, the reliance on the consistency of the set of non-derived sentences, reproducibility of operations, et alii. On the other hand, the criteria of science as presentation follow conditions of acceptability and are in this sense ‘foundational’. They derive from potential procedures of scientists and can be characterized essentially by two principles: the ‘principle of method’, i.e., presentations work stepwise without ‘jumps’, a kind of completeness claim, and the ‘principle of dialogue’, i.e., presentations can be criticized, which means that it can be questioned whether they fulfil the principle of method.12 If one proceeds this way and at the same time argues from a treatment of science in the first case as a fact and in the second case as a norm, a ____________ 10 Cf., e.g., Habermas 1976. 11 In: Wohlrapp 1975. 12 The role of such principles in the Erlangen School receives a new discussion in the last essay of this volume, pp. 206–217.

114

Methods in Philosophy, in Art, and in Science

certain kind of simplification occurs and affects the conclusion. Therefore, it remains to be investigated whether the specific claims of failure and success Wohlrapp is raising with respect to the two sets of criteria he discusses, really carry conviction as they stand or whether failure and success rather depend on some further distinctions connected with the difference of research and presentation. To repeat what I said in the beginning, I would like to claim that inasmuch as questions of constitution are confounded with questions of justification – or, to use linguistic verbiage, questions of signification confounded with questions of verification – the criteria for science either as research or as presentation will lead to difficulties and eventually to failures. Success, therefore, in either case is dependent on a clarification of the interrelationship between answering on ‘what there is’ and answering on ‘what is true’, the ontological and the epistemological version of the question of how the two levels, of objects and of signs, of ‘world’ and ‘language’, separate within and unite into one domain of (scientific) language-games. I have tried elsewhere13 to show that this domain has to be understood as a domain of preactions, beyond the action-act dualism as the prototype of the type-token division, and equally beyond the classical distinction of something given and something to do. The idea is simply to start with – from a later point of view, complex – objects which do not yet bear the differentiation between actor and action nor between action and object of an action or result of an action. From that starting point which is articulated by a very elementary and obviously fictitious language where only reference to such objects occurs14 one has to develop both ordinary language and the language of science with the result that during this procedure, already, most of what is treated later on as logical or ontological presuppositions of a language gets decided. ____________ 13 Meanwhile the most detailed account in: Lorenz 1996 (reprinted in: Lorenz 2009, pp. 24–71); an abbreviated version without the use of the term ‘preaction’ in: Lorenz 2005, in this volume pp. 42–55. 14 This elementary level I understand to be the same as von Weizsäcker refers to at a place – cf. von Weizsäcker 1979, p. 136 – where he speaks of a level of conceptual representation of atomic actions with no difference of judgments and concept yet, as no singular terms occur. This paper tries to justify the disagreement with the conclusion von Weizsäcker draws when turning sides with the empiricist’s predelection of the singular (in fact something particular and not in the strict sense something singular, notwithstanding the terms used for the opposition ‘singular [term] – general [term]’) over the general by declaring a ‘predicate’ genetically prior to a proper name, i.e., a singular term.

The Concept of Science

115

The difficulty is that in giving a description of this (re)construction the language of description, i.e., some standard natural language in use, is far more developed syntactically and semantically than the described language during the process of its construction. Hence, in order to get an adequate account of the construction it is necessary to introduce certain devices which make sure that the description at any stage is not dependent on those features of the syntactic and semantic structure of the language of description which do not yet belong to the structure of the constructed language. For example, the difference of singular and general terms within the language of description should not be of relevance for describing the initial stage of construction, where within the fictitious elementary language only reference to preactions occurs. Rather, there should exist an explicit step of introducing that very difference within the elementary language. And this is possible by paying attention to the distinction of schema and actualisation that is constitutive of non-analysed actions on the elementary level, already. Schemata are ‘universal’ and actualizations are ‘singular’ (in the strict sense). These terms refer to the difference of ‘once’, ‘once more’, ‘once more again’, … , which is practically acquired in situations of repetitive imitation (= imitative repetition) with respect to any preaction.15 On that basis preactions may be individuated into particular units by a process of subdividing the schema into subschemata such that each subschema of actualizations may count as a (first-order particular) token – represented by a singular term – of a (second-order particular) type that is represented by a general term. Individuation of a preaction is tantamount to its typification. Linguistic signs are the means which have developed gradually through our evolution to articulate which type belongs to which token. Through language a token becomes a symbol of a type, and, the other way round, through language something general becomes an aspect of something individual. ____________ 15 Such preactions – or rather their universal schemata – are the pragmatic version of Peter Strawson’s ‘feature universals’, as treated in: Strawson 1959; he forgets to include in his presentation their singular aspect as something on a par with the universal aspect of preactions. Under the headline ‘Property and Substance’ where the terms are used to refer to repeatable and nonrepeatable entities, respectively, the same issue, though without distinguishing between the oppositions ‘singularuniversal’ and ‘(first-order) particular-(second-order) particular [= general]’, is at stake when Richard Rorty discusses – and refutes – attempts of Alfred N. Whitehead to evade well-known epistemological dilemmas deriving from the singular (term)-general (term) dichotomy, if this dichotomy is correlated in a straightforward way, i.e., without using linguistic analysis, with the mind-body dualism, cf. Rorty 1963.

116

Methods in Philosophy, in Art, and in Science

Proceeding in this way, it is perhaps not any more offensive – by being liable to the pitfalls of a remake of the Cartesian dualism – to say: singulars account for the empirical base, universals for the rational design. Neither they nor the tokens and types that are derived from them can be handled separately from each other: Research starts with singularia, presentations with universalia. Since theories of both areas exist, hierarchies of theoreticity appear16 and things including the empirical/rational distinction get mixed up.17 As I already mentioned in connection with Wohlrapp’s paper, it serves still further confusion, if in the first case the methodological position of the analytic philosophy of science is characterized as being descriptive, whereas in the second case the methodological position of the constructive philosophy of science bears the label of being normative. Inasmuch as questions of constitution have consequences in terms of stipulations concerning the objects of scientific discussions – the ‘definition’ (I prefer the more general term ‘introduction’) of basic predicates about them is included – the insistence on the normative character of some fundamental part of science – it may be called ‘protoscience’ – is reasonably supported. Similarly, questions of justification, concerning the context of these objects – the use of predicates, so to speak, not their introduction –, give rise to descriptive aspects of any science. Here again, it might be useful to recall that, in general, predicates within scientific languages are defined on given domains of objects, extensionally as certain classes of those objects. Hence, they cannot be treated as primary predicates. They are derivative with respect to the defining predicate for the domain of objects, i.e., the ‘substances’ that are defined as the instantiations of the primary predicates. We, then, speak of properties, and the usual problems concern questions of whether properties hold of objects which obey certain other descriptions, and they never concern questions of elementary constitution. Those constitutional questions – unless they are non-elementary, i.e. of second order, yielding domains of abstract objects (such processes are, of course, well-known and extensively treated everywhere) – occur on a language level which is itself of a merely theoretical nature. It is the elementary level I referred to earlier and which can now be characterized as the one where terms are introduced, not the one where they are used as ____________ 16 Representative of this treatment is the last chapter Semantic Ascent (§ 56) in: Quine 1960. 17 Cf. the sophisticated treatment of the empirical core (= empirical content) of a theory via J. Sneed’s criteria of theoreticity as expounded , e.g., in: Stegmüller 1973, chap. VIII.

The Concept of Science

117

in ordinary non-scientific and in unsophisticated scientific language. For example, when you assert ‘this leaf turns yellow’, the constitution of objects like leaves, i.e., the introduction of the term ‘leaf’, is presupposed, whereas the constitution of objects like ‘yellows’ (substantival use of the term!) is pushed to a second-order level: Yellow is constructed as an abstract object turning the word ‘yellow’ from a non-primary general term – standing for a ‘characterizing universal’ in the sense of Strawson18 – into a singular term, a nominator, as I suggest to call it. What remains in the case of the assertion in question is to judge whether the use of the terms ‘leaf’ and ‘yellow’ (or ‘turn yellow’) is justified which by all standards is a question of true description. Constructions remain within one logical level of language (or within the lowest or object level), descriptions concern two consecutive levels. This difference is well-known, e.g., in logical theory, where formulae may either be constructed by formation rules or be described by means of suitably chosen predicates of a metalanguage that are used for formalising the constructions. The fundamental notions are partition (of a whole into parts) in the case of constructions, and attribution (of a property to an object) in the case of descriptions. Mereology and set-theory provide for the respective formalisations of these notions though their interrelation is by no means sufficiently clarified up to now.19 I hope these remarks give sufficient support for the claim that there is no simple correlation between, on the one hand, science as research and stating what is (being the result of research), and, on the other hand, science as presentation and issuing what shall be (being the guarantee of presentation). In either case, the set-up of science is not only a question of justifying a corpus of sentences, but a question of introducing meaningful terms used as constatives and/or as directives. And justification includes the search for the truth (science as theory) together with the search for the good (science as praxis) as much as meaning exhibits both, aspects of (theoretical) signification and aspects of (practical) relevance.

____________ 18 Cf. Strawson 1959, p. 168. 19 Cf. Lorenz 1977; in this volume pp. 20–32.

118

Methods in Philosophy, in Art, and in Science

III The relevant difference of research and presentation as against the simplified accounts I discussed above comes in when we look for the support of a scientific theory. As far as science is treated as a way of research, this support should derive only from the descriptive power of the theory relative to the first-order particular objects as derived from the singular perceptual ‘cores’ of research, though usually, in the analytic philosophy of science, due to the aforementioned lack of a clear separation between constitution and justification, this support is extended to include the explanatory power of the theory as well. It is common to use the term ‘confirmation’ when referring to non-elementary propositions with respect to how they relate to relevant particular (observational) ‘data’, and this entails a confrontation with the – I dare say unsolvable – riddles and paradoxes of induction.20 The simple reason for the claim that nothing beyond the descriptive power of a theory can be treated within science as research only, derives from the following considerations: The explanatory power of a theory refers to the kind of interrelations which obtain among the different propositions of the theory, especially to an assessment of the range of validity of fundamental principles like those of conservation in physics. Hence, an account of the explanatory power can be given by just rendering judgments on the conceptual frame of the theory that is used for the argumentations in science as presentation.21 The argument in the alternate case runs conversely: As far as science is treated as a way of presentation, the support of a scientific theory should derive only from its constructive power relative to the general objects as derived from the universal conceptual frames of presentation, though, usually, in the constructive philosophy of science, due to the same confoundation of constitution with justification, this support is extended to the regulatory and, hence, normative power of the theory as well. The term ‘approximation’ when referring to elementary objects with respect to how they relate to relevant general ‘idea[l]s’, is in this connection occasionally used, and, as an equally disturbing consequence, it becomes necessary to

____________ 20 Cf. the discussion about the interrelation between explanation and induction in: Hempel 1977 (a German translation of a revised version of the last chapter in: Hempel 1965). 21 Cf. for comparison the related remarks on the difference between descriptive and explanatory adequacy of a theory – in this case on linguistics – in: Chomsky 1965, chap. I (Methodological Preliminaries).

The Concept of Science

119

handle the vexations of the is-ought-gap.22 Here again, it is easy to see that in science as presentation any attempt to go beyond the constructive power of a theory and to judge upon its power to issue what shall be – unless this is treated as a second-order question only, i.e., as a question of what kind of scientific activity (rather than objects of activity) should exist – will need reference to the perceptual cores of science as research, e.g., to the singular encounters with first-order particular objects in experimental situations. What I want to claim is that both the explanatory and the regulatory (or normative) power of a scientific theory can be assessed properly and without bias only, if the set-up of science cuts straight through the separating line of research and presentation. This means especially that on each level within the hierarchy of theories the link between constitutional and justificational questions – and that refers to the interdependence of constructive and descriptive procedures as well – must not be lost sight of. Hence, trying to determine an adequate meaning of the two central concepts, (scientific) explanation and (scientific) regulation, amounts to nothing less but a reassessment of whether and how a unified treatment of science is possible. For this purpose, the concept of ‘unified science’ should no longer be understood in the original historical setting of the Vienna Circle along with a developing analytic philosophy of science. In the light of the considerations just offered I claim that a unified approach to science, unless it falls victim to typical ‘pseudo-problems (Scheinprobleme)’ as the one concerning the possibility of induction or the one concerning a bridge over the is-ought-gap, has to consider science with respect to activities both of research and of presentation. It has to develop a concept of science that starts with a kind of unity of research and presentation23 where the domain of (scientific) language-games uniting ‘world’ and ‘language’ in the sense I have outlined above becomes the result of the first step. These language-games of preactions together with their articula____________ 22 On the alleged interdepencence of a constructive and a normative approach to science, cf. the essays, especially by J. Mittelstraß, P. Janich and O. Schwemmer, in: Kambartel/Mittelstraß 1973. The logical indepencence of ‘is’ and ‘ought’ from one another – formalized as non-validity of ∆!A≺A in deontic logic (cf. , e.g., Lorenzen 1969, pp. 70–71) – is usually taken for granted, and attempts to question the is-ought-gap get criticized even by other opponents of the analytic approach, e.g., by K.-O. Apel, in: Apel 1979, who takes pain to refute J. Searle’s claim that their exist non-trivial logical relations among is- and ought-sentences. For further discussion, cf., e.g., Hudson 1969, and Lorenz 1987, in this volume pp. 124–139. 23 An institutional codification of this unity – as an ‘idea’ at least – is the ill-famed ‘Einheit von Forschung und Lehre’ characteristic of the German university set-up according to Wilhelm von Humboldt.

120

Methods in Philosophy, in Art, and in Science

tions may then be treated with respect to both their aspects: matter-oriented (research) and person-oriented (presentation). For visualization of what I am driving at, I may use an example of current dispute: the different approaches to (physical) geometry. If there is concentration on the research aspect of physics, the (temporal) behavior of (physical) bodies relative to their spatial coordinates is judged with respect to quite general hypotheses concerning space-time-structure (explanationbias!). If, on the other hand, the presentation aspect of physics is of concern, as in the protophysics of the Erlanger Schule, we are confronted with a series of steps that are to serve the introduction of fundamental concepts of geometry, chronometry, and hylometry, in that order, by using ‘idealized’ operations with (physical) bodies (regulation-bias!).24 In the latter case, we come across various attempts to provide for meaningful terms, but it is taken for granted that they can be used successfully outside presentational questions. It is, therefore, not surprising that certain propositions come out true a priori. In the first case, something completely different happens: Propositions about given objects are tested to secure their validity, which means to treat them as empirically based. Here, the introduction of the terms used is taken for granted inasmuch as presentational questions are considered to be a cura posterior. Though theoretical activities of supplying true descriptions govern science in its research aspect or, rather, because of them, the presentational necessities, like introducing meaningful terms get neglected. And, conversely, the concern with practical operations to get adequate constructions of fundamental concepts for science in its presentation aspect seduces into thinking low of problems, whether those concepts can effectively be used in research situations. The real issue actually boils down to the question of whether the introduction of meaningful (geometric) terms like ‘straight’, ‘n-times the length of’, etc., can be treated as an extension of ordinary language about ordinary objects, serving better criteria of relevance according to further developed standards of significance and truth. For, if extendability fails, we are stuck in conceptual frames without prospects to satisfy them; and if presuppositions serve as a substitute for explicit introductions, there is no chance to guarantee anything beyond the perceptual cores. It is easy to see that the last two conditional sentences may serve as a modern and more refined version of Kant’s famous dictum that concepts without intuitions are empty and intuitions without concepts are blind. The ____________ 24 Cf. the relevant entries, like ‘Protophysik’, ‘Chronometrie’, ‘Hylometrie’ and others, in: EPW.

The Concept of Science

121

refinement consists in the introduction of hierarchies of theoreticity starting with (of course not uniquely determined) common sense experience phrased in everyday language.25 The domains of objects of scientific discourse have to be distributed among logical levels of ascending and descending order without any chance to argue for a definitely ‘lowest’ level, e.g., of elementary particles, that would provably be sufficient as basis for arbitrary future theories. The usual arguments exchanged between protophysicists and ‘deuterophysicists’ – if I may coin that term for the moment – that make use of coordinate systems with a spatial or even a spatio-temporal metric are beside the point as long as the problem of how to introduce what kind of metric is exempt from a truly mutual discussion. What can possibly be introduced rather than merely postulated on the basis of elementary common human experience (still ambiguous relative to the singular-universal bifurcation) will lend itself to the discrimination of (empirical) tokens from (rational) types on the basis of singular actualizations of a universal schema with respect to any preaction. And if we succeed in constituting a domain of (particular) objects – classical mechanics is set up on the basis of ideal mass points in a homogeneous Euclidean space-time – that may serve as a common basis of quite different physical theories, unification of science has been pushed one step further. I want to claim, therefore, that in order to make a comparison between two theories T1 and T2 possible – even between incommensurable ones against the claims of Paul Feyerabend26 – it is necessary and sufficient to develop a theory T which allows a faithful embedding ρ ι of Tι into T. Apart from the fact that it is impossible to know in advance whether for two theories, e.g., for quantum mechanics and general relativity theory – such a ‘supertheory’ exists, the actual task will consist in devising appropriate translations including ‘logical constructions’ in the sense of Bertrand Russell for the objects of either theory by means of the primitive terms of the supertheory.27 Furthermore, such embeddings will, through the larger contexts the embedded theories then have, provide conditionalisations for the validity of these theories. And claims of uniqueness concerning the ____________ 25 The grades of theoreticity that serve as a kind of measure for the distance from common sense experience relative to some natural language system, are discussed in: Quine 1970. 26 Cf. his arguments, e.g., in: Feyerabend 1975, under recourse to examples from the history of science and of myths, for the impossibility to achieve a unique (if not true) world representation. 27 Cf. the relevant papers of Russell, together with an introduction accessible in: Pears 1972.

122

Methods in Philosophy, in Art, and in Science

‘definition’ of certain terms, e.g., of ‘plane’ by protophysical devices, can likewise be substantiated by explicitly referring to their dependence on boundary conditions that do no longer lie outside the theory.

References Apel, Karl-Otto, 1973: Transformation der Philosophie, vol II (Das Apriori der Kommunikationsgemeinschaft), Frankfurt am Main: Suhrkamp. Apel, Karl-Otto, 1979: Sprechakttheorie und Begründung ethischer Normen, in: Konstruktionen versus Positionen. Beiträge zur Diskussion um die Konstruktive Wissenschaftstheorie I–II, ed. and introd. by Kuno Lorenz, Berlin/New York: Walter de Gruyter, vol. 2, pp. 37–106 . Bubner, Rüdiger, 1973: Über die wissenschaftstheoretische Rolle der Hermeneutik [extended version of ‘Hermeneutik und Wissenschaftstheorie aus der Sicht der hermeneutischen Theorie’ in: Gerber 1972, pp. 135–145], in: Rüdiger Bubner, Dialektik und Wissenschaft, Frankfurt am Main: Suhrkamp, pp. 89–111. Chomsky, Noam, 1965: Aspects of the Theory of Syntax, Cambridge Mass.: The M.I.T. Press. Corti, Walter R., ed., 1976: The Philosophy of William James, Hamburg: Felix Meiner. CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–935. EPW, = Enzyklopädie Philosophie und Wissenschaftstheorie, ed. by Jürgen Mittelstraß, 4 volumes, Mannheim/Wien/Zürich: Bibliographisches Institut-Wissenschaftsverlag 1980–1984. Revised and enlarged edition in 8 volumes, Stuttgart/Weimar: J. B. Metzler 2005ff. Feyerabend, Paul, 1975: Against Method. Outline of an Anarchistic Theory of Knowledge, London/Atlantic Highlands: New Left Books/Humanities Press. Gerber, Uwe, ed., 1972: Hermeneutik als Kriterium für Wissenschaftlichkeit? Der Standort der Hermeneutik im gegenwärtigen Wissenschaftskanon, Loccum: Evangelische Akademie. Habermas, Jürgen, 21973: Erkenntnis und Interesse [1968], Frankfurt am Main: Suhrkamp. Habermas, Jürgen, 1976: Zur Rekonstruktion des Historischen Materialismus, Frankfurt am Main: Suhrkamp. Hempel, Carl G., 1965: Aspects of Scientific Explanation And Other Essays in the Philosophy of Science, New York/London: The Free Press/Collier-Macmillan. Hempel, Carl G., 1977: Aspekte wissenschaftlicher Erklärung, Berlin/New York: Walter de Gruyter. Hermes, Hans, 1969: Ideen von Leibniz zur Grundlagenforschung: Die ars inveniendi und die ars iudicandi, in: Studia Leibnitiana Supplementa, vol III (Akten des Internationalen Leibniz-Kongresses Hannover, 14.–19. Nov. 1966), Wiesbaden: Franz Steiner Verlag, pp. 92–102. HSP = Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines,

The Concept of Science

123

ed. by Marcelo Dascal, Dietfried Gerhardus, Kuno Lorenz and Georg Meggle, Two half-volumes, Berlin/New York: Walter de Gruyter. Hudson, William D., ed., 1969: The Is-Oght-Question. A Collection of Papers on the Central Problem in Moral Philosophy, London: Macmillan. Kambartel, Friedrich/Mittelstraß, Jürgen, eds., 1973: Zum normativen Fundament der Wissenschaft, Frankfurt am Main: Athenäum Verlag. Lorenz, Kuno, 1977: On the Relation between the Partition of a Whole into Parts and the Attribution of Properties to an Object, in: Studia Logica 36, pp. 351–362. Lorenz, Kuno, 1987: Is and Ought Revisited, in: dialectica 41 [Norms and Conventions], pp. 129–144. Lorenz, Kuno, 1996: Artikulation und Prädikation, in: HSP, vol. 2, pp. 1098–1122. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Logic, Thought and Action, ed. by Daniel Vanderveken, Dordrecht: Springer, pp. 343–357. Lorenz, Kuno, 2009: Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter. Lorenzen, Paul, 1969: Normative Logics and Ethics, Mannheim/Wien/Zürich: Bibliographisches Institut-Wissenschaftsverlag. Lorenzen, Paul/Schwemmer, Oswald, 1973: Konstruktive Logik, Ethik und Wissenschaftstheorie, Mannheim/Wien/Zürich: Bibliographisches Institut-Wissenschaftsverlag. Mittelstraß, Jürgen, 1970: Neuzeit und Aufklärung. Studien zur Entstehung der neuzeitlichen Wissenschaft und Philosophie, Berlin/New York: Walter de Gruyter. Pears, David, ed., 1972: Russell’s Logical Atomism, Oxford: Fontana/Collins. Quine, Willard V. O., 1960: Word and Object, New York/London: John Wiley & Sons. Quine, Willard V., 1970: Grades of Theoreticity, in: Experience and Theory, ed. by Lawrence Foster and Joe W. Swanson, Amherst Mass.: University of Massachusetts Press, pp. 1–17. Reichenbach, Hans, 1938: Experience and Prediction, Chicago: University Press. Rorty, Richard M., 1963: The Subjectivist Principle and the Linguistic Turn, in: Alfred North Whitehead. Essays on His Philosophy, edited and with an introduction by George L. Kline, Englewood Cliffs N.J.: Prentice Hall, pp. 134–157. Stegmüller, Wolfgang, 1973: Probleme und Resultate der Wissenschaftstheorie und Analytischen Philosophie, Band II (Theorie und Erfahrung), Zweiter Halbband (Theorienstrukturen und Theoriendynamik), Berlin/Heidelberg/New York: Springer. Stich, Stephen P., 1996: The dispute over innate ideas, in: HSP, vol 2, pp. 1041–1050. Strawson, Peter F., 1959: Individuals. An Essay in Descriptive Metaphysics, London: Methuen & Co. von Weizsäcker, Carl F., 1979: The preconditions of experience and the unity of physics, in: Transcendental Arguments and Science. Essays in Epistemology, ed. by Peter Bieri/Rolf-Peter Horstmann and Lorenz Krüger, Dordrecht/Boston/ London: D. Reidel Publishing Company, pp. 123–158. Wohlrapp, Harald, 1975: Analytischer versus konstruktiver Wissenschaftsbegriff, in: Zeitschrift für Allgemeine Wissenschaftstheorie 6, pp. 252–275.

Is and Ought Revisited As a starting-point for my presentation I want to recall three points that were made during the preceding discussions on norms and conventions:1 (1) Appeal to conventions arises only in cases of equally well argued alternatives of deciding on a course of action. Hence, not any rule of action can be treated as a convention but only those which have been selected out of a set of equivalent rules of actions. A convention is logically the substitute for the result of using a dice in deciding on such a selection. Consequently, conventions serve to resolve rationally indeterminate situations. (2) The choice of linguistic expressions to signify certain object types cannot in general be subsumed under conventions, because in order that the choice of a term be a case of convention it has to be presupposed that object types are available by other semiotic means, e.g., through performance in case of action types. Hence, inasmuch as a language in use is comprehensive (= non-defective, like ‘motherese’) it is non-conventional: it defines a world view. Only, when ‘new’ object types get defined, e.g., by introducing the terms ‘point’, ‘number’, ‘true’, etc., or equivalent ones – and it is essential that such equivalent possibilities are available – the choice of these terms is conventional. (3) Not only the choice of such terms but even the definition employed is conventional, as in this case structurally equivalent definitions are available. The ‘new’ object types have been turned into formal objects that are characterized, respectively, by the operations with them or the sentences about them, and nothing else.

It is a common trait of these observations, so it seems, that there is both an object type (something natural) and a semiotic representation (something conventional), and just by a careless move of treating these two entities as ontologically independent, rather than complementary, we are driven to all kinds of unwarranted dichotomies like language and theory (i.e., linguistic frame and theoretical content, which is a special case of the distinction between ‘sign’ and ‘object’), or matter of fact and evaluation (i.e., the object is a sentence or, rather, its propositional kernel, and the object’s manner of being given is a mood of the sentence), and related ones, culminating in the present day dispute on naturalism and anti-naturalism, being the modern heirs of latter day positivism and transcendentalism in philosophy. For a ____________ 1

They took place on occasion of the VIIIth International Colloquium in Biel/ Switzerland, May 1–4,1986.

Is and Ought Revisited

125

whole bunch of problems – be it applicability of mathematical notions to physical reality, adequate treatment of hybrid discourse, criteria for the choice of logics, and, of course, justifiability of the so-called gap between ‘is’ and ‘ought’, or others – it is essential to bridge these dichotomies. Philosophical pragmatism in the spirit of C. S. Peirce is such an endeavour that I would like to invoke for dealing now with my special topic, the ‘is-ought-gap’. For more than thirty years we witness a rich debate both on the correct reading of David Hume’s famous statement concerning the logical independence of is-judgments (Seinsurteile) and ought-judgments (Sollensurteile) and on its systematic justification if there is any. On the one hand, we have the clear statement of Hume that “[r]eason is the discovery of truth or falsehood”,2 whereas „[a]ll morality [i.e., fixing on good/right or evil/wrong] depends upon our sentiments”,3 so that “[t]he rules of morality […] are not conclusions of our reason”.4 Yet, on the other hand, further explication given by Hume on the meaning of moral judgments seems to imply that they are is-judgments, too, though on another level: “when you pronounce any action or character to be vicious, you mean nothing, but [my underlining, K. L.] that from the constitution of your nature you have a feeling or sentiment of blame from the contemplation of it. Vice and virtue, therefore, may be compar’d to sounds, colours, heat and cold, which, according to modern philosophy, are not qualities in objects, but perceptions in the mind”.5 And the discussion goes on about whether this should be read as an emotivist account of ethics, moral judgments being expressions of feeling, or whether moral judgments, indeed, only report the fact that those who utter them have certain feelings, a reductionist position embedding moral judgments into psychology. Both versions amount to a naturalist position taken up again by John Searle in his equally famous attempt to use speech act theory for deriving ought-statements (not necessarily moral ones – he wants to leave that special case open) from is-statements.6 I may remind you of the five steps used in his example: 1. 2. 3. 4.

Jones uttered the words ‘I hereby promise to pay you, Smith, five dollars’. Jones promised to pay Smith five dollars. Jones placed himself under (undertook) an obligation to pay Smith five dollars. Jones is under an obligation to pay Smith five dollars.

____________ 2 3 4 5 6

Treatise III, 1.1, cf. Hume 1888, p. 458. Treatise III, 2.5, cf. loc. cit., p. 517. Treatise III, 1.1, cf. op. cit., p. 457. Treatise III, 1.1, cf. op. cit., p. 469. Searle 1969, chap. 8.1.

126

Methods in Philosophy, in Art, and in Science

5.

Jones ought to pay Smith five dollars.

Searle argues that, by common standards, sentence 1. is descriptive, stating the fact that someone utters a promise, whereas sentence 5. is, by equally common standards, normative (Searle uses the more general term ‘evaluative’), stating an obligation for Jones who uttered the promise. It was Searle’s point to call attention to the existence of institutional facts besides natural ones, and he observes that institutional facts are bound up with constitutive rules where some of them involve, e.g., obligations. Hence, someone stating an appropriate institutional fact is committed also to recognize the implied normative statements. And the move from stating a natural fact (Searle calls it a ‘brute fact’) to stating an institutional fact occurs by going from sentence 1. to sentence 2., a move that is justified, according to Searle, by the meaning of ‘promise’ which is said to be a fact concerning the use of a certain descriptive term. Any attempt to rescue the is-ought-gap either by claiming that sentence 2. (or a subsidiary statement involved in moving from 1. to 2., e.g., that uttering a promise is making a promise) is not descriptive at all but normative in disguise, or by trying to split the meaning of certain terms into a descriptive and a normative one, or by still more intricate moves, will fail for two reasons. You either try, Searle says, to immunize against derivations of the kind I [i.e., Searle] have given by stipulating logical independence between descriptive and normative statements, or you try to keep aloof from recognizing institutional facts altogether. In the latter case a particular disastrous consequence would be the one articulated in the very last sentence of Speech Acts: “But the retreat from the committed use of words ultimately must involve the retreat from language itself, for speaking a language – as has been the main theme of this book – consists of performing speech acts according to rules, and there is no separating those speech acts from the commitments which form essential parts of them”. It was Karl-Otto Apel, who, in a long paper Sprechakttheorie und Begründung ethischer Normen,7 interpreted the quoted sentence as an indication of Searle’s readiness to acknowledge the institutional fact of rule governed language use to be more than simply an empirical fact of the social sciences which is open to ordinary processes of verification and falsification. The use of language not restricted to the particular case of terms that refer to, and, hence, mention institutional facts, is by itself the use of an institutional fact implying all kinds of commitments, like keeping to meaning rules including stipulations of new meanings, provided such stipulations are followed by keeping to the conventions governing the ____________ 7

Apel 1979.

Is and Ought Revisited

127

introduction of definitions and the like. Hence, Apel claims, Searle could have argued – and in a way this seems to have been, at least partially, his intention – that the normative aspect goes with any meaningful utterance without obscuring the special difference, i.e., the logical independence between is-statements and ought-statements. This point of Apel seems to me to be well placed as it reveals an ambiguity of great consequence involved in the move from sentence 1. to sentence 2. in Searle’s example. Sentence 1., stating the natural fact of a specific utterance having been made, belongs to the metalevel with respect to the content of that utterance. Hence, the move to sentence 2. remains logically sound only, if sentence 2. is interpreted as, e.g., ‘I say that Jones promised to pay Smith five dollars’ or by making use of another explicit speech act on the part of the speaker. The institutional fact of a promise having been made is stated in the sense of being described, implying the commitments involved in the speech act of the speaker and not the ones implied by making a promise. These latter implications, the obligations involved by making a promise, remain on the level of objects usually called ‘normative facts’, which the speaker may go on explicitly to describe, too. But, of course, sentence 2. and all the subsequent ones may also be interpreted as including acknowledgment on the part of the speaker that promises when being made should be kept. In that case, ‘promising’ in sentence 2. does not function as merely a descriptive term – descriptive of an implied normative fact – but has normative bearings on the part of the speaker, too. Apel has made use of the suggestive conception to consider the speaker in the first case as acting like a cultural anthropologist and in the second case as acting like a judge. Thus, it is clear that the normative aspect of ‘promising’ should not be considered as part of its meaning but as tied to its being an action, here: a specific second-order linguistic action, a speech act. To have done something implies (in such cases) to do something else (because one has accepted this rule of action). And the usual discussion at this point continues with elaborating on the difference between two points of view, the observer position and the participant position with respect to a specific situation. The observer essentially describes a situation, the participant takes part in it, and participation is, on the one hand, ontologically, considered to be more fundamental, insofar as describing is but a special case of participation, namely on the level of language use, whereas, on the other hand, epistemologically, priorities are reversed, because participation is available for being discussed only through description. Within this context, it is common practice to identify the use of language with the performance of speech acts, leaving aside the more special claim that for this purpose as speech acts only acts of communication my be taken into account, either in

128

Methods in Philosophy, in Art, and in Science

the sense of understanding or trying to understand each other,8 or in the sense of arguing or trying to argue one’s respective claims and counterclaims as put forward by Apel, for example. In passing it may be added that Meggle presupposes the primacy of description, whereas Apel argues for the primacy of participation. But, in neither case the difference of an ontological and an epistemological approach, the two sides of the same coin according to a famous dictum of Charles Sanders Peirce,9 is really scrutinized. Hence, proceeding in either of these two ways obscures an important complementarity obtaining between observation and participation. The usual hierarchy of use and mention that we arrive at by the device of semantic ascent suggests that participation is restricted to the top language level such that to describe presupposes a language of description as having been introduced, where its terms are of another logical type than the entities described. This account, essential for all kinds of investigations into formal systems is, of course, insufficient to lead to the starting point itself, the first level of entities. There, we have to do with ‘individuals’, to use a familiar terminology, but irrespective of whether these individuals are strings of symbols, natural kinds, events, individual persons, or the like. Investigations of the former kind evade the discussion of how in the process of determining individuals, semiotic and pragmatic aspects of human activities interact. Nor may we expect to have more light shed on such apparently basic dichotomies connected with the distinctions ‘statement – command’ and ‘is – ought’, as, for example, ‘thing – action’, where the corresponding elementary occasion sentences are ‘this is x’ and ‘y should be done’ with ‘x’ being a thing-term and ‘y’ an action-term. In this way, we are led back to the two Kantian realms, ‘Reich der Natur’ and ‘Reich der Sitten’. There are still other dichotomies that are connected with the distinctions ‘statement – command’ and ‘is – ought’, like ‘cognition – volition’, ‘belief – intention’, ‘fact – value’, and others, as mentioned in the introductory sections in the beginning. We have to try, therefore, to develop a unified approach to both objects and signs of objects quite in line with the pragmatic methodology of Peirce that a theory of designation and a theory of designata has to be a combined endeavour. For support and in preparation of the dialogue model initially invented to establish a pragmatic conception of truth for logically compound propositions, and now extended to a kind of both systematized and generalized Wittgensteinian language-games in order to give an account of the interplay between world views and ways of life, I would like to turn your attention to Nelson Goodman’s radical destruction of the clear-cut ____________ 8 9

Cf., e.g., Meggle 1981. Cf. CP, 5.257.

Is and Ought Revisited

129

division between world and language. His main thesis states that the division between the given and the constructed, between that which is found and that which is made, between the fact and the artefact, is outdated. Any matter we are concerned with is dependent on some manner as the means by which we deal with it.10 This position which Goodman sometimes, though reluctantly, calls ‘irrealism’ is not a kind of conventionalism,11 the ‘isms’ anyway are mere diseases of philosophy, he declares, since the dependency of matter from manner, of empirical cores from rational frames, of worlds from versions, cannot itself be articulated. We choose the facts as much as the frameworks, and – I would like to add – we experience the frameworks as much as the facts. ποιεῖν and πάσχειν, doing and suffering, the last two Aristotelian categories which seemed forgotten throughout most of modern philosophy in the Descartes tradition will enjoy a lively comeback as the two ways we are concerned with when doing things: you do it yourself (active) and you ‘see’ others doing the same (passive). These two ways may be called performance and cognition (of actions). Now, no doubt, we are able to live with, or in, different versions, i.e., frames of reference to the world, as Goodman tells us.12 But there remains a twofold problem: a)

b)

If there is prima facie no chance to characterize one version uniquely as the true one (apart from the fact that only sentential versions can be true, others like pictorial versions cannot be subjected to a truth claim), what kind of criteria are available to distinguish right versions from wrong ones? If there exist different, possibly even conflicting right versions, how is it possible to live as ‘the same person’ in different worlds or, for that matter, to establish communication among persons living in different worlds?

The issues connected with (a) that are dealt with in detail by Goodman, may roughly be concentrated around the issue of how to identify the objects (colours, feelings, things, experiments, etc.) that we succeed and fail in constructing. ‘Rightness of rendering’, Goodman declares, is basically a matter of fit to what is referred to, and within science it boils down to the acceptance of the standard criteria of present day scientific discourse. The issues connected with (b) are focussed in the issue of how to secure the identity of the subjects living in possibly different worlds, that is, of how an open set of interrelated capacities to act which includes awareness of self and other, and which should not be understood as ____________ 10 Cf. the chapter on The Fabrication of Facts in: Goodman 1978, pp. 91–107. 11 Cf. the foreword in: Goodman 1978. 12 Cf. the chapter on Words, Works, Worlds in: Goodman 1978, pp. 1–22.

130

Methods in Philosophy, in Art, and in Science

restricted to the point of view that is presented by the performances (of an action), can be used to determine the concept of a person. These latter questions Goodman did not raise. He deliberately refrained from following up the normative consequences of his approach. Yet, obviously both sets of issues ask for a thorough investigation of the interdependence that obtain between the features of signification and the ones of communication insofar as they characterize the elements in the process of worldmaking. I am going to call these elements ‘semiotic actions’,13 a term not used by Goodman, because he usually restricts his attention to the results of such actions rather than studying the structure of semiotic actions themselves. Of course, one should be prepared to cope seriously with the fact that there are many people engaged in different activities through times and spaces with obviously limited capacities to understand each other, i.e., to realize what they are doing. To tie acquisition of knowledge basically to individual persons – acquiring beliefs – and justification of knowledge basically to social communities exercising control over the various individual idiosyncrasies – judging on truth and falsehood of beliefs according to socially transmitted and accepted norms – a conception which Richard Rorty, for example, still argued for in his influential Philosophy and the Mirror of Nature,14 would itself be dependent on the Cartesian topos of individual persons starting to digest the sensory input each for herself and afterwards confronting each other with possibly different experiences which then have to be ‘unified’. But as any criterion for unification will usually have to rely on linguistic representations of experiences and never on experiences themselves, this move leads to a dead end. Two attempts of handling this difficulty and still enjoying support at present are likewise bound to fail as they have to beg the question. I am referring first to conventionalism which is an instrumentalist version on the language level and, therefore, amounts to a move of treating as the same what is called to be the same, and second to ontologism – or ‘metaphysical realism’ to use a more fashionable term – which is a realist version on the object level and, therefore, turns unification into a presupposition: Experiences by different persons essentially coincide because their causes, the objects, are assumed to be the same for every person. The only way out, I think, is the move to the dialogue model where you start with a situation which is the result of Goodman’s plea for relativizing the distinction between objects and signs of objects. You are asked to look at verbal language or other symbol systems in use primarily as types of actions ____________ 13 Cf. Lorenz 1987. 14 Rorty 1980.

Is and Ought Revisited

131

like eating or sleeping, with the consequence that in this respect they themselves are parts of the objects they refer to, and complementarily to look at the non-linguistic, especially non-symbolic, objects as parts in a web of interrelated and interdependent actions, thus taking into account that they partake in the significative value of the activities they are imbedded in. Hence, what we do initially is to naturalize language including other symbol systems and to symbolize world by paying attention to that feature of actions which appears in the focus when we cognize actions rather than just perform them. The bridge over the apparent gap between world and language can be built if you make that gap disappear. In exercising actions they are performed as well as understood. From the first perspective that may be called the natural or empirical aspect of actions, you produce tokens of a type, and from the second perspective that may be called the symbolic or rational aspect of actions, you ‘identify’ different tokens as belonging to the same type. The equivocation of ‘identification’ I just made use of, is, of course, intentional. I mean both to recognize the tokens within an open series of tokens as belonging to the same type – epistemological identification – and to set the tokens within an open series equal, thus constituting a type – ontological identification. Both readings are type-oriented, hence they treat actions universally: Recognition is effected by cognition of something universal. On the other hand, if performance is understood as the production of tokens, i.e., as a kind of practical attempt to exhaust an action type, it is token-oriented: Performance as production is brought about by performance of something singular. Now, we will proceed in a more systematic fashion and make a rough sketch of the dialogue model in order to get a firmer grasp of the two fundamental aspects of actions, the performance-aspect and the cognitionaspect, which may be called the pragmatic and the semiotic aspect, respectively. It is not at all a new insight, I should add, that with every action the distinction of mere acting and of acting as a kind of language is applicable, though this feature-splitting has not yet been used for a genetic reconstruction of linguistic competence – of some initial stages of linguistic competence, at least – from processes of acquiring ordinary action competence. The dialogue model consists in an elementary situation of two persons being engaged in starting and continuing a process of learning to perform a certain action. Their activities may be described with the terms ‘repetition’ – one person is repeating a certain action – and ‘imitation’ – the other person is imitating what the first one does – in order to illuminate the interesting conceptual relation of this approach to Jean Piaget’s research in processes of competence acquisition. Repetition and imitation in the dialogue model are the logical equivalents, i.e., the equivalents on the level of genetic reconstruction, to assimilation and accommodation, used by

132

Methods in Philosophy, in Art, and in Science

Piaget for describing empirically occurring acquisition processes, provided the interaction of a person with his surrounding, the way Piaget presents that process, is replaced by an interaction between two persons.15 The two-person elementary situation of the dialogue model is a generalized and more elementary version of a Wittgensteinian language-game, in as much as it functions as a measuring-rod for already occurring complexes of actions including linguistic ones.16 To start with processes of action acquisition rather than with either sets of action events (some kind of exterior particular entities) or sets of pragmatic competences (some kind of interior particular entities) derives from the necessity to avoid the everpresent pitfalls of Cartesian dualism with its awkward consequence to use either a behavioristic or a mentalistic approach to account for our abilities, in particular our linguistic ones. It is the purpose of the elementary dialogue situations together with the construction of a gradually growing differentiation into more and more complex dialogue situations to keep a unified account of how we acquire both participation in a common situation and use of a common piece of language. It comes about by looking carefully at the two points of view exercised by the two agents in such a teach-and-learn-situation. At a given time one of them is performing and the other one cognizing the action, and in the course of the acquisition processes the points of view switch continuously. By way of caution I should add that the two terms ‘performing’ and ‘cognizing’ are applicable only post hoc, i.e., after the open sequence of switching roles – repetitive imitation or imitative repetition – has taken place, and both agents have acquired the competence in question. The primary objects available with any such competence are actions without any further differentiation as yet, neither with respect to act and agent, nor with respect to act and object of an act, nor otherwise. But all of them both are and are known to be common in the respective situation, even better: each one of them is a common situation and both parties know it. Therefore, in order to keep the different stages in the development of the dialogue model distinct, it is better to call the elementary and rather primitive though common world to consist of preactions. Neither things, nor persons, nor rabbits in the usual sense, belong to them. Preactions are performed by the acting agent and cognized by the other party, that’s all. With regard to this distinction, for the next step we should pay attention to the two further predicates already used, in order to arrive at a better understanding of a standard item within philosophical tradition relevant for the is-ought-issue we are concerned with. With respect to the performance ____________ 15 Cf. Piaget 1950. 16 Cf. PU, § 131.

Is and Ought Revisited

133

viewpoint the preaction was said ‘to be’ singular, with respect to the cognition viewpoint that same preaction was said ‘to be’ universal. This terminology which belongs to a procedural approach to a subject matter and not to an objectival one, is to suggest that differentiating between singularity and universality is basically a question of viewpoints towards objects in a dialogue situation and has nothing to do with, e.g., kinds of objects. When we continue to speak of preactions being schemata existing as open sequences of actualizations it is essential to keep in mind that actualizations belong to the performance viewpoint only, so it does not make sense trying to introduce referring terms of the kind ‘this swimming’ as a deictic description of an actualization. Similarly, it is impossible to use ‘schema’ otherwise than indicating the cognition viewpoint towards a preaction. The case is somewhat different with the objectival distinction of tokens from types, where both entities are particulars, though not of the same but of consecutive logical order (= ‘type’) and procedurally accessible by singular performance and universal cognition of dealing with tokens and types, respectively. It is tokens and types that have to be referred to when discussing the difference of an empiricist approach and a rationalist one in scientific studies. The predominance of tokens – and, hence, of the singular – in the empiricist tradition can easily be explained now by the first person approach to epistemology, whereas the rationalist tradition with its emphasis on types – and, hence, of the universal – is characterized by the introduction of a generalized subject, from the very beginning anticipating the observer standpoint. We may say that using the terms ‘singular-universal’ or ‘actualization-schema’ is nothing but a convenient, possibly misleading façon de parler for the fact that preactions as primary objects exist only through repetition and imitation, i.e. through the open sequence of changing viewpoints. It needs to proceed further, from preactions to actions, in order to being able to speak of tokens and types, too.17 Realizing now that the performance viewpoint is pragmatic, whereas the cognition viewpoint is semiotic, we can finally state that our dialogue model indeed permits us to distinguish a pragmatic and a semiotic aspect with every preaction. You get a concrete version of the sign-character of a preaction, if you look at the non-performing party during a particular instant of the dialogue situation as someone who understands the performance of the other party, for example, as an invitation to do ‘the same’. Now, taking into account that the performing party is by tradition called ‘active’, we may – apart from the categories ποιεῖν and πάσχειν – apply very ____________ 17 For a more detailed exposition of the relevant constructions, cf., e.g,, Lorenz 2005, in this volume, pp. 42–55.

134

Methods in Philosophy, in Art, and in Science

naturally the Aristotelian distinction of δύναµις and ἐνέργεια, too. The schema of a preaction stands for the possibility of actualization, it is the δύναµις of a preaction, whereas actualization is its ἐνέργεια. Both cannot be separated. Each actualization is understood only as actualization of a schema, and each schema exists only actualized. Peirce, whose awareness of the essential unity of epistemology and ontology, as mentioned above, already, called the universals ‘reals’ and the singulars ‘existants’.18 We, again, proceed a step further by turning the cognition viewpoint of a preaction into an explicit preaction of its own. Instead of simply cognizing swimming, for example, you perform a separate preaction, say seeswimming, which qualifies as a perception with respect to the original preaction. Cognizing a preaction is turned into performing a perception of a preaction. Therefore, in the dialogue situation, the person who performs swimming is at the same time cognizing the perception of his partner, that is, in more colloquial terms, he knows that the one who perceives does himself know what swimming is through performing a perceptual preaction with respect to swimming, or: he knows that his partner disposes of the schema swimming. The result that we encounter is a curious switch of the original distribution of the terms ‘active’ and ‘passive’. By turning from preactions as procedures to their being objects of perceptions we arrive at the well-known qualification of modern philosophy: Having a schema of a preaction at one’s disposal is tantamount to having a concept and is, hence, an active accomplishment of the mind, whereas actualizations being singulars count as the basic given stuff. But, the dialogue model conveys clearly that this role switching move derives from a distancing move that changes the elementary dialogue situation from something in process into something the participants are conscious of – the move from Aristotle to Kant, in terms of our genetic reconstruction. The final step of identifying the cognition aspect of a preaction with the performance aspect of a perception of the preaction which may be called ‘objectivation’ leads to the stage of object constitution independent from particular teach-and-learn-situations. By pragmatic abstraction, as we may say, we define a preobject – using the term ‘object’ would be inadequate as there is not yet a division into individual units, the individuals – as the invariant out of the open set of relevant perceptions irrespective of the particular teach-and-learn-situation of the original preaction. The perceptions show up neatly as the ‘ways of being given’ of a preobject in Fregean terminology. To cognize a preobject is turned into performing a corresponding perception, and preobjects themselves cannot any more be performed: They ____________ 18 Cf. CP, 5.430 and 6.335.

Is and Ought Revisited

135

are the same for all participants though with different perceptual access. Even actions as preobjects are not performed; what you perform, for example, in the case of swimming as a preobject, is swimming as showingswimming, that is, you are ‘presenting’ (vorführen) swimming. Only by presenting actions they count, like things, persons, etc. – though we did not yet develop criteria of classifying preobjects into kinds – as preobjects and not merely as preactions. We call preobjects ‘articulated’ by their perceptions. But it is a contingent anthropological fact that certain perceptual preactions out of the manifold of articulations of a preobject, namely verbal preactions, receive canonical status with respect to their being a sign of the preobject. As a kind of symptom they represent the preobject ‘pars pro toto’. For the next step of turning the verbal preactions into preobjects of their own without thereby cutting off their perceptual, i.e., semiotic function, we return to dialogue situations of teaching and learning, but now of second order. They are well-known under the label ‘predication’. The verbal preobjects or, rather, their spoken or written results made available by pragmatic abstraction with the help of such second order dialogue situations I call articulators, and we know, already, that they cannot be performed in the strict sense but only presented. To utter an articulator, or to perform an articulation or a predication, say ‘swimming’ – this is a ‘verbal perception’ of the preobject swimming – is at the same time a case of cognizing it. An articulator is, as a preobject, a fullfledged symbolic sign, and not any more a mere symptomatic one. We may distinguish these two semiotic functions of an articulator by saying: In its symptomatic function an articulator is constitutive of its object, in its symbolic function it describes its object. We have now reached the stage where articulations being both signs and actions, that is sign-actions, or better: semiotic actions, can be subjected to two aspects well-known since antiquity: the aspect of signification and the aspect of communication, and I am going to use the term ‘articulation’ from now on only with respect to its significative aspect, switching to the term ‘predication’ when we focus the communicative aspect of articulation. Since it generally holds that with respect to its pragmatic character a sign-action is communicative, and with respect to its semiotic character an action is significative, we can say that an articulation in its significative aspect is realized by performing a perception. Hence, the ‘meaning’ of an articulator splits into the different perspectives of the persons using the articulator and, thus only, common meanings can be ensured. Likewise, an articulation, in its communicative aspect, that is, a predication as an action, splits into different ways of being given – again the result of an objectivation, here of second order, by turning the cognition of a predication into an independent action, which is called a mood of the predication. The more familiar term nowadays

136

Methods in Philosophy, in Art, and in Science

for such a mood is, of course, ‘speech act’. Predications always occur in a mood, the speech act between speaker and hearer. Second-order dialogue situations appear as systematic equivalents of Wittgenstein’s languagegames in the strict sense. We are now in a position to return to our starting point. In my last section I will try to give a consistent interpretation of that passage from Hume that I have quoted in the beginning and that was used to charge Hume with a naturalist position in ethics in apparent contradiction to Hume’s explicit opinion concerning the mutual independence of is-judgments and oughtjudgments elsewhere. The crucial question is to explain how predicates of actions that are used to evaluate them, like ‘vicious’ and ‘virtuous’, turn out to be articulations of something mental, in fact perceptions in the mind, which, according to Hume’s terminology, may at once be specialized to emotions, sentiments of pleasure or blame, or the like. Let us start with two articulators, ‘A’ (for: leaving the apartment) and ‘B’ (for: locking the entrance door of the apartment) of actions A and B, respectively. That they are rather special, such that no single terms for them exist in English, does not concern us here, nor do I want to go into the problems of my own descriptive account of situations I thereby propose to you to imagine. ‘A’ and ‘B’ must have been made available in a situation of predication, therefore we should ask for the moods of it. Within any second-order dialogue situation there are two basic orderings of the performances of an action and of the sign-action, that is, the articulation – more carefully one should note that both are presented rather than performed, and that presentation of the sign-action is not a one-person affair but just that particular common action of both parties, i.e., their common though perspectival cognition of the predication, which we called the mood. Hence, saying ‘A’ after the other party has performed A, and before she has done it, yields two basic directional fits, that is, the two basic moods of a sentence that result in statement ‘A.’ and order ‘A!’. But, and that is important, no validity claims are raised with these modal propositions: neither ‘[I] state that [this instance of] A has occurred’ nor ‘[you] bring about that [the schema of] A will be instantiated’, to use a more elaborate rephrasing of ‘A.’ and ‘A!’, respectively, can be questioned. These propositions occur as simple, an activity accompanying speech, in the statement case ‘about’ the token which has been produced already, thereby showing to the other party that it is understood as a token of a particular type, and in the order case ‘about’ the type, thereby assuming common understanding which will have to be backed by the other party by producing a token. There is no chance to say ‘I know what you want but I don’t do it’, because referring to an understanding is not yet possible without an instantiation being available. Nor is, on this level, stating a case of

Is and Ought Revisited

137

saying as in ordinary subject-predicate-sentences, since there does not yet exist a reference to the token that is independent of the situation of stating. Things change after further distinctions will have been introduced; they will at this place be indicated only very sketchily.19 First: it is essential to develop referentially stable subdivisions of preobjects into individual units. This step is necessary to allow of making statements such that the situation of its utterance is not eo ipso a situation backing the statement. For example, instead of merely being able to state that a particular instance of leaving the apartment has occurred which, wherever you utter it, must hold if it makes sense at all, simply because no invariant reference to such instantiations exists, you must be able to state that the apartment has been left. Since reference to an individual apartment will be determined independently from the particular situation in which ‘the apartment has been left’ is uttered, you may be wrong with this statement. And it is exactly at this stage that the modal action of stating may carry an obligation, that is, may be changed into the modal action of asserting, i.e., making an assertion. We may say that asserting is stating with the intention to be veridical, or, equivalently – by switching from the ‘internal’ characterization by means of intentions to the ‘external’ characterization by means of a behavior –, asserting is both an invitation to get questioned and a promise to be prepared to defend. In short: An assertion carries the obligation to argue (for truth) if you get questioned. In our sample case, successful fulfilment of this obligation permits to call ‘the apartment has been left’ true. The very concept of truth cannot be introduced unless raising claims, and that is equivalent to fixing obligations, is made possible. Likewise the introduction of variables depends on the availability of individuals. Second: unless there exists an explicit construction of what it means to be an individual person who can be addressed independently of a situation of speech as ‘the one which is able to act and to speak’, it is impossible to issue an order such that a validity claim – usually called a rightness claim in order to distinguish this type of validity claims from truth claims – may be raised with it that would turn the order into a command. If, for example, you issue the order ‘lock the entrance door of the apartment’ (= ‘B!’), a ‘why’ by the other party instead of just following the order, makes sense only if you can refer to what we may call ‘rules of action’ which are accepted, and that means: are followed, by that party. Common acceptance of a rule of action defines their rightness. Hence, a reply of ‘no’ presupposes understanding together with an implicit indication that the ____________ 19 A detailed exposition is, meanwhile, available in: Lorenz 1996, reprinted in: Lorenz 2009, pp. 24–71; cf. the more advanced presentation in: Lorenz 2005, in this volume pp. 42–55.

138

Methods in Philosophy, in Art, and in Science

ordered action does not fit any of one’s accepted rules of action. Some acceptable rule in the case of our example might be articulated by ‘who leaves the apartment ought to lock the entrance door of the apartment’. I would like to suggest that the much discussed ‘conditional obligations’ are just rules of actions constitutive of a subject individually and socially insofar as they indicate the range of possible actions that the subject acknowledges, or, stated in more colloquial terms: ‘that’s the way I want to act’, a subjective preference. Now, any such accepted conditional obligation A≺O(B) can be articulated as a rule for articulators, in our example as ‘A’⇒‘B’. It follows that the inference from ‘the apartment has been left’ to ‘the entrance door of the apartment is locked’ is valid. And this, I think, is exactly what Hume wanted to express when he writes that to call an action virtuous is to articulate an emotion: To command that action (‘commend’ would better fit Hume’s temperament) is to refer back to an accepted rule of action. Hume himself, of course, incorporates both sides of the dialogue situation that is used in this reconstruction. To me it seems not to be difficult to explain why Hume was unable to see that truth claims, too, amount to conditional obligations that may be articulated by, e.g., ‘who asserts ought to argue (for truth)’, quite similar to the corresponding conditional obligation in case of rightness claims: ‘who commands ought to argue (for rightness)’. Both conditional obligations are second-order rightness claims referring back to the corresponding metarule of action that is constitutive of intersubjectivity. Somebody who does not acknowledge this metarule, that is, who utters statements or orders without willingness to defend them against objections, does not turn those statements and orders respectively into assertions and commands, thus excluding them from their possible function to be steps in the process of individuation and socialization of the participants in a dialogue. They become private affairs in the strict sense without any interest to anybody else. What Hume did was to compare is-judgments – propositional kernels we would say today – with ought-judgments, and not I-state-that-judgments with ought-judgments, what he should have done, if he were to follow current analysis of ought-judgments. We may conclude: Is-judgments and ought-judgments appear on different logical levels and are, therefore, formally independent. If conditional obligations are read as accepted rules of actions rather than as logically implied from normative and other principles, they have as a consequence the material – not the formal – validity of the inference: What ought to be done, will be done. And this does not contradict the formal invalidity of the implication O(A)≺A in descriptive systems of deontic logic.

Is and Ought Revisited

139

References Apel, Karl-Otto, 1979: Sprechakttheorie und Begründung ethischer Normen, in: Konstruktionen versus Positionen. Beiträge zur Diskussion um die Konstruktive Wissenschaftstheorie, volume II, ed. by Kuno Lorenz, Berlin/New York: Walter de Gruyter, pp. 37–106. CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks Sussex: Harvester Press. Hume, David, 1888: A Treatise of Human Nature [1739/1740], ed. by L. A. SelbyBigge, Oxford: University Press. Lorenz, Kuno, 1987: Semiotic Actions in Worldmaking, in: Worldmaking’s Ways, ed. by Lars Aagaard-Mogensen/Rik Pinxten and Fernand Vandamme, Ghent: Communication & Cognition, pp. 11–24. Lorenz, Kuno, 1996: Artikulation und Prädikation, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal/Dietfried Gerhardus/Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1098–1122. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Logic, Thought and Action, ed. by Daniel Vanderveken, Dordrecht: Springer, pp. 343– 357. Lorenz, Kuno, 2009: Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter. Meggle, Georg, 1981: Grundbegriffe der Kommunikation, Berlin/New York: Walter de Gruyter [2., aktualisierte Auflage 1997]. Piaget, Jean, 1950: Introduction à l’épistémologie génétique I–III, Paris: Presses Universitaires de France. PU = Wittgenstein, Ludwig: Philosophische Untersuchungen. Philosophical Investigations, dt./engl., transl. by G. E. M. Anscombe, Oxford: Basil Blackwell/New York: MacMillan 1953. Rorty, Richard, 1980: Philosophy and the Mirror of Nature, Princeton N. J.: University Press. Searle, John R., 1969: Speech Acts. An Essay in the Philosophy of Language, Cambridge: University Press.

Competition and Cooperation: Are They Antagonistic or Complementary? I At the root of questions concerning the development of social behavior in human beings, we are confronted with a presupposition which exerts a considerable influence upon the conceptual framework of sociogenetic studies. I am referring to the questionable assumption that by some crucial experiment(s) we may eventually decide whether, in the last resort, individual behavior governs social behavior or vice versa; and this decision does not depend on how theories answer to the problem of distinguishing between genetic and environmental determination. Rather, it is connected with the rivalry between sociology and psychology as to which type of theories should count as being fundamental to the explanation of human behavior. In this situation it is advisable to make an attempt to circumvent rival claims by opting for a kind of interdependence between individual and social behavior, which involves starting with a conceptual analysis of the relevant terms. But instead of treating individuation and socialization directly I want to restrict myself to a more special case, which may throw some light on how to proceed in general. I have chosen competition as a prima facie case of individual behavior, and cooperation as a likewise prima facie case of social behavior, and the title of my contribution could easily be understood as if I were to consider that these two kinds of behavior are, by themselves, the driving forces for developing individual and social behavior in general. Of course, this line of reasoning would make it impossible to retain the correct boundaries between empirical research and conceptual investigation. A philosopher’s contribution cannot start by selecting certain phenomena as candidates for an answer to the inquiry into the mechanisms of sociogenesis, and then proceed by attempting to find out whether they have been well chosen or not. Only the positive sciences, psychology and sociology in particular, might pursue such a line of empirical research, though they would certainly start in a far more sophisticated manner.

Competition and Cooperation

141

My task is of a different nature. I am not looking for phenomena which may count as competitive or cooperative activities – finding phenomena – but I am rather using the terms ‘competitive’ and ‘cooperative’ as conceptual tools to be applied to activities, and that means, that I have to construct two appropriate kinds of activities – inventing phenomena – such that they may serve, in the terminology of Ludwig Wittgenstein’s Philosophical Investigations, as ‘measuring-rods’, i.e., as a means of comparison with which to identify ongoing activities in terms of competitiveness or cooperativeness.1 Philosophical inquiry is a reflexive activity, but nonetheless inseparably linked to positive activity. Finding and inventing phenomena belong together, even though division of labour tends to make us think otherwise. In order to understand the two notions ‘competition’ and ‘cooperation’ that are usually understood as ‘pursuing individual interests’ and ‘pursuing common interests’, respectively, it is useful to go back in history to where ideas about becoming human have been spelled out in terms of a cultural process starting from a state of nature. One of the oldest documents still available is the myth attributed by Plato to the sophist Protagoras and included in Plato’s dialogue Protagoras.2 Here we find a tale in which human beings come into existence by one and another step that are kept strictly apart from each other. With the first step, technical abilities are acquired to compensate for natural deficiencies – they appear individually as a distributive character of the species and make its members able to survive collectively as natural beings. The whole realm of τέχναι which includes the arts and religious rites as well, is available on this level. But there is a second step that is understood as the advent of rationality in its full sense of practical abilities to provide for political units of self-government: honesty or justness (δίκη) and modesty or respect (αἰδώς) lead to solidarity (φιλία). These language dependent abilities that arise from mutual recognition cannot be exercised except through individual distinction and social coherence. Any such social individual leads a life governed by reason: it is a rational being. We have found the backbone necessary for guaranteeing the original equivalence of the two ancient ‘definitions’ of man: being a rational animal (ζῷον λόγον ἔχον) and being a social animal (ζῷον πολιτικόν). Two remarks should be added: First, the Platonic separation of nonrational, or ‘poietical’, activities from rational, or ‘practical’, activities – the added attributes refer to the Aristotelian distinction of poiesis, i.e., producing something with an external aim in mind, and praxis, i.e., doing ____________ 1 2

Cf. PU, § 131. Prot. 320c8–323c2.

142

Methods in Philosophy, in Art, and in Science

something for its own sake3 – relegates every action not governed by a general law – in Aristotle’s terminology an action which is not its own aim – to a sphere of merely natural as well as cultural exigencies. And this means that it is not truly human, not truly civilized. Late offshoots of this Platonic move may be found easily. For example, Kant’s insistence on the undetectability of actions that are based on morality,4 and, more recently, Habermas’ dichotomy of actions into those which are rational (!) with respect to means and ends, and those which serve mutual communication,5 may be traced back to Plato. Competitive or ‘amoral’, behavior seems to precede the more developed stage of cooperative, or ‘moral’, behavior, even though the notion of rationality is no longer understood as being restricted to the area of moral legislation.6 Second, the additional Aristotelian separation of rationality as an ability to theorize from sociality as an ability to lead a good life, gradually gives way to a similar developmental bifurcation into a primary level of mere acting, or better still: behavior, and a secondary level of deliberate activity, that is, of actions that are guided, or at least accompanied by speaking or thinking. Taking both remarks together, we arrive at speaking, on the level of competitive behavior, to be a case of strategic activity without any communicative value. Whether using language for establishing consensus or showing activity that comes out from a consensus, either characterization of a human group as civilized is part of a progress theory of the cultural process, the progress being measured in terms of complexity of (group-)organization, such that Hobbes’ bellum omnium contra omnes is the initial stage. In addition to this widespread and influential picture, there is also an alternate account of the cultural process in terms of a decline theory: On the basis of an aetas aurae, a natural state of paradisiac existence for each individual – well known from Rousseau’s forceful descriptions – an ever increasing consciousness develops and spoils the ability to act spontaneously. Natural individual creativity is hampered by social pressures that are derived from conscious activity both for and against others, and the decline can be measured by the losses in (self-)production.7 In this case, it seems that both competitive and cooperative behavior impose constraints on a primordial uninhibited way of autarkic living. The Stoic slogan ‘secundum naturam vivere’ though never forgotten, has been ____________ 3 4 5 6 7

Cf. for details: Bien 1973. Cf. the minute discussion in: Kambartel 1989. Cf. Habermas 1981. Cf. Hegselmann 1988. For more details, cf. Lorenz 1990, espec. chap. 2.2.

Competition and Cooperation

143

audibly resuscitated in these days of growing ecological consciousness, but, of course, without its original pledge for individual self-sufficiency. Rather than moving from ‘barbaric’ competition to ‘civilized’ cooperation, as the cultural process is looked at in progress theories, we have here the call to reverse the cultural process leading, according to decline theorists, from natural cooperation, possibly including cooperation with non-human nature as well, to cultural competition. In both cases competitive behavior is judged to be inferior to cooperative behavior. We may see this as a consequence of either of the two contradictory traditional assumptions: Man is, by nature, bad (presupposition of progress theories), versus man is, by nature, good (presupposition of decline theories). Obviously, such a moral preoccupation acts as an antidote for the need to argue for either case, or even for other alternatives.

II Luckily, there is an author who indeed takes up this issue and uses arguments against both versions or, even better, emends both versions in such a way that neither ‘good’ and ‘bad’ nor ‘true’ and ‘false’ are used descriptively with respect to available phenomena on the level of objects or of language, but as reflexive terms for passing philosophical judgments by reconstructing the objects in question in the way I explained in the beginning. The author I am referring to, and who will act as a guide to solve the title question, is Johann Gottfried Herder. He seems to have been the first in history who has become aware of the conceptual relation which holds between the two characterizations of human beings occurring in Protagoras’ myth. Humans are deficient beings, insofar as they lack sufficient protection against inanimate nature and do not possess effective weapons against animals, and, at the same time, they are proficient beings insofar as they have both poietical and practical abilities. For these two abilities, Herder uses the terms ‘freedom’ and ‘reason’, respectively.8 It is by exercising these abilities that humans define their deficiencies, and it is, therefore, wrong to treat the relation between proficiency and deficiency as an empirical one of compensation. The point becomes even clearer if we turn to some important structural features of the interdependence between being proficient and being deficient. Herder defines the cultural process neither by progress nor by decline, but by a process of education. In fact, in the ninth book of his Ideen zur ____________ 8

Cf. Herder 1960.

144

Methods in Philosophy, in Art, and in Science

Philosophie der Geschichte der Menschheit,9 he uses a dialogue model of teaching and learning to identify items of a cultural process, inasmuch as both ‘doing’ and ‘suffering’, terms that may be used to characterize the two roles in a teaching and learning process, always occur together. And since every individual being educated by his elders and educating his offspring plays both parts, it is individual distinction together with social coherence, in other words: competitiveness against a common background and cooperativeness split up into individual approaches, which defines the process of education. Reason as the ability to organize, and freedom as the ability to produce, appear when exercised, on the side of ‘doing’ as proficiency, and on the side of ‘suffering’ as deficiency. According to Herder, that is why tradition may include errors (deficiency of reason) and why something evil may occur among the items somebody chooses (deficiency of freedom).10 We can, already, see here the differences to both an understanding of competition merely as fixing and pursuing individual interests, and cooperation merely as establishing and pursuing common interests. Such an understanding of sociality and rationality in terms of individual and common interests, respectively, has turned the reflexive use of sociality and rationality, which up to Aristotle was the outcome of a self-characterization of humans, into the positive use of describing preexisting properties of humans. It was Hannah Arendt who first clearly demonstrated how, by translating the Greek term πολιτικόν into the Latin term sociale, such a change of meaning has led to conceptual confusion throughout our tradition today.11 But it was not this fact alone which had this effect, one, also, has to take into account another important feature that is due to the Stoic replacement of πολιτικόν by κοινόν. Sociality was no longer restricted to the second Protagorean level of reason-guided ways of living, but is understood as pervading the first level of ‘natural’ abilities, too. It is this insight, together with the recognition that also rationality – not limited to means-ends rationality – is required for exercising first-level poietical activities which made Herder confident of being right in reestablishing the reflexive use of sociality and rationality, but not throughout the whole range of human activities. Hence, ‘sociality’ (Herder’s ‘freedom’) does not refer to the fight among individual interests whether understood in the intentional setting of a (cultural) ‘fight for power’ or reduced to explanations by a causal theory to account for the (natural) Darwinian ‘fight for survival’, but to social ____________ 9 Herder 2000, Part II (1785). 10 Cf. Lorenz 1990, pp. 65–69. 11 Cf. Arendt 1958, pp. 22–24.

Competition and Cooperation

145

coherence both in acting and speaking, which is cooperation by means of individual contributions. Realizing that everyone must invent his own way of life and his own world view, provides the opportunity for the gradual development of sociality, which may also be called solidarity to use a term more akin to the Greek term φιλία (friendship) which had served the same purpose as mentioned above. Likewise ‘rationality’ (Herder’s ‘reason’) no longer refers to the ability of acting according to common interests, even if ‘common’ is not restricted to some group interests but refers to full-fledged moral generality. Instead, it signifies individual distinction on all levels which is competition grounded in a community of acting and speaking. In fact, as a paradigm of rational behavior everyone acknowledges ‘competition by argumentation’. Realizing how one’s own way of living and one’s own world view is found amidst shared activities, is an actualization of another step on the way towards individuality. Traditionally, from antiquity via Kant to the present day, we are accustomed to speak in this connection of self-determination as the task of reason; but, as reason is said to be concerned with the universal and not with the particular, self-determination is understood as the deliberate submission to universal laws, that is, as the creation of the universal human and not as a universal task to be carried ocult to understand why such vexations have occurred. They are unavoidableut individually. It is not diffi, I think, when one forgets that individuality and sociality are but stages in a process which is simultaneously a process of individuation and of socialization, the one which had been conceptualized by Herder as the education process. An account which starts with a set of ready-made human individuals who have their own preferences and beliefs, because otherwise there is no opportunity to either determine or explain any of the different competitive or cooperative relations among them (these qualifications are understood in the positive sense of fight among individual interests and activity in the common interest, respectively), and also not any of those relations which count as preceding or following them, such as deliberation, negotiation, arbitration, etc., is no longer able to apply the concept of self-determination or autonomy to human individuals, but only to the whole species of humans. Among individuals, then, there are only external – natural or cultural – relations of exerting influence on one another: a clear case of heteronomy. Of course, the process of education I referred to as an alternative must again not be understood in the positive sense of educating, neither in the intentional framework of given educational aims, nor in the causal framework of social engineering. It has to be a process of self-education, where both sides in the process of teaching and learning change their ways of life

146

Methods in Philosophy, in Art, and in Science

and their world views by a further step of both individual distinction and social coherence. The dialogue model for reconstructing the context of non-verbal and verbal activity, that is, the ways of life and the world views – for better visualization you may think of Wittgenstein’s famous language-game in the first paragraphs of his Philosophical Investigations – clearly shows that, at first, non-verbal activity counts as social, yet verbal activity counts as individual, and if you stop here, non-verbal activity is treated merely as (objective) behavior and verbal activity merely as (subjective) expression. You have to proceed through one further step of reflection, that is, of reconstruction, and you will arrive at the semiotic aspect of non-verbal activity – actions have a (subjective) sense as explicated, e.g., by Max Weber in his theory of intentions,12 – and at the pragmatic aspect of verbal activity – speaking has a (objective) meaning as explicated, e.g., by the Pragmatic Maxim of Charles Sanders Peirce.13 If you have gone that far, non-verbal activity shows its individual aspect and verbal activity its social aspect. Ways of life and world views can be comprehended only, if seen in both their individual and their social aspects. Self-education is not striving for a balance between guiding and letting grow, but setting up limits against becoming influenced. And this means that being guided is countered with acting by making use of all kinds of individual knowing-how, and growing is countered with invoking the knowledge of social norms, the common aspect of activity. Individuality that is a difference between individuals on the level of reflection, will be recognized only within some common activity; sociality that is an equality of individuals on the level of reflection, can be exercised only by being conscious of the different approaches within the same common activity. We may conclude: The antagonism between competition and cooperation, taken descriptively, changes into the complementarity of competition and cooperation, if understood reflexively. It is just one process, educating self and other, which, from different perspectives, leads to both individuation and socialization.

____________ 12 Cf. the relevant passages in: Weber 1966. 13 Cf. the systematic exposition in: Scherer 1984.

Competition and Cooperation

147

References Arendt, Hannah, 1958: The Human Condition, Chicago: The University of Chicago Press. Bien, Günter, 1973: Die Grundlegung der politischen Philosophie bei Aristoteles, Freiburg/München: Alber. Habermas, Jürgen, 1981: Theorie des kommunikativen Handelns I–II, Frankfurt am Main: Suhrkamp. Hegselmann, Rainer, 1988: Ist es klug, moralisch zu sein?, in: Metaphysik nach Kant? Stuttgarter Hegel-Kongress 1987, ed. by Dieter Henrich and Rolf-Peter Horstmann, Stuttgart: Klett-Cotta, pp. 689–702. Herder, Johann G., 1960: Abhandlung über den Ursprung der Sprache [1772], in: Johann Gottfried Herder. Sprachphilosophische Schriften, ed. by Erich Heintel, Hamburg: Felix Meiner, pp. 3–87. Herder, Johann G., 2000: Ideen zur Philosophie der Geschichte der Menschheit I–IV [1784–1791], in: J. G. Herder. Werke I–X in elf Bänden, ed. by Günter Arnold et alii, Frankfurt: Suhrkamp, Band VI. Kambartel, Friedrich, 1989: Autonomie, mit Kant betrachtet. Zu den Grundlagen von Handlungstheorie und Moralphilosophie [1978], in: Friedrich Kambartel, Philosophie der humanen Welt. Abhandlungen, Frankfurt am Main: Suhrkamp, pp. 117– 131. Lorenz, Kuno, 1990: Einführung in die philosophische Anthropologie, Darmstadt: Wissenschaftliche Buchgesellschaft. Scherer, Bernd M., 1984: Prolegomena zu einer einheitlichen Zeichentheorie. C. S. Peirce’s Einbettung der Semiotik in die Pragmatik, Tübingen: Stauffenburg. Weber, Max, 1966: Soziologische Grundbegriffe, Tübingen: Stauffenburg.

Another Version of Methodological Dualism Contemporary discussion of ‘rational behavior’ concentrates on two types of giving explanations, by efficient causation and by final causation. Yet, in view of the far richer history concerning causation in general, it would be unwise to restrict the attention to aitiological and teleological explanatory procedures. We should also include related pairs of opposite concepts like ‘matter – form/content’, ‘variable – constant’, ‘body – mind/soul’, and others, bearing in mind that ‘empirical – rational’, for example, already implies a restricted use of ‘reason/rational’, whereas ‘object/objective – subject/subjective’ is connected with a whole series of quite different ideas in the course of our intellectual history, some of them I want to discuss here. The opposites in question I consider to be related with each other by a ‘family resemblance’ with respect to ‘rational behavior’ or ‘λόγον διδόναι’, using Plato’s expression. I thereby mean that it should be a worthwhile task to exhibit the links which relate the respective terms within any of the pairs of opposites to each other. As a guide I choose Leibniz whose idea to reinterpret the notion of substance as a function on matter whereby matter is turned into a sign, may help to understand the ancient Aristotelian treatment of matter and form in terms of the modern Peircean treatment of action and sign-action. Rather than interpreting the history of philosophical method as a replacement of the notion of substance by the notion of function, as was done by Ernst Cassirer when he ‘read’ the historical switch from a preoccupation with ‘ontology’ to a preoccupation with ‘epistemology’ in this way, it is more adequate to follow Leibniz and understand substance finally as a result of human activity in the world(s) we live in. If it is accepted as prima facie reasonable to consider mind, the ‘rational part’ of mind or ‘rational activity’, as creating signs out of matter, we can also understand Plato’s λόγον διδόναι in a more convincing way than it has been understood in terms of the Aristotelian methodology that demands to interpret giving reasons as presenting a proposition that acts as a reason. ‘Logos’ is first of all not a proposition being used as a reason, but it is creating a ‘form’, an εἶδος, as the object of mental activity while performing a perceptual action. It was one of the great achievements of classical antiquity, shared by Plato and Aristotle, that an object and its representation go together: content-related ontology and form-related

Another Version of Methodological Dualism

149

epistemology are but two sides of the same coin, an insight rediscovered in this form by Peirce.1 Plato is using for this insight, and Leibniz essentially follows him, the terms of body (σῶµα) and soul (ψυχή) – the first having parts, whereas the second is partless and all-pervading – such that the link (δεσµός) binding the parts (built up of the four elements in Plato) together into a whole is just the rational part of the soul or the λόγος, understood in its original sense as a proportion. Therefore, objects in its being a whole are creations as actions of the mind; they are ‘constituted’ by ‘universals’ which, on the other hand, in the terminology of Aristotle, may be predicated, using creations out of passions of the mind (παθήµατα τῆς ψυχῆς), of objects. Aristotle supplements Plato’s theory of object-constitution by a theory of object-predication. Beside the action types which say something universal of a whole (the whole in Aristotle often being reduced to the mere bearer of an εἶδος, ἔχον τὸ εἶδος)2 we have, also, to take account of the performances of actions, i.e., the perceptions and conceptions as an act, which by means of something singular show a whole. This was not yet conceptualized by Plato when he speaks of bodies. And it may well be this fact which lies at the bottom of another methodological dualism connected with the Aristotelian supplement of a Platonic achievement, the one between intuitive and formal procedure. It is derived from object-oriented investigation of matter in contrast to subject-oriented investigation of form in a way that is very well known. With the advent of science in antiquity, that is, of an attempt not just to accept supposed knowledge at its face value but to look for reasons why it is real knowledge rather than merely apparent knowledge, the concept of proof entered the scene and with it the art of argumentation as the main tool of science or philosophy. We are concerned with science or philosophy as a proof-seeking activity which is directed towards establishing the truth of propositions. Now, reasons and counter-reasons in argumentations seem again to be nothing but propositions, or better: propositions act as reasons and counterreasons. Hence, we are confronted with the task to determine how and when propositions acquire their argumentative force. Certainly not by being just true and equally certain not without being true. In addition to the truth of propositions as reasons there has to hold an implication relation between a proposition when used as a reason and the proposition to be backed by the reason, in order to arrive at good reasons. Similarly in the case of counter-reasons. A proposition will act as a good ____________ 1 2

Cf. CP, 5.257. Cf. Met. 1023b20.

150

Methods in Philosophy, in Art, and in Science

counter-reason for another proposition, if it is both true and standing in the relation of contrariness to the other proposition. So it looks as if we are in worse shape than before. In order to establish the truth of a proposition by argumentation, we not only have to be sure of the truth of other propositions but also of the truth of higher order relational propositions, that is, of implication and contrariness between propositions. The traditional way out that in general is still held to be the only ‘reasonable’ one, proceeds by splitting the art of argumentation into two parts: establishing the truth of ‘first’ or ‘basic’ propositions by using intuition and establishing the truth of relations of implication and of contrariness among propositions by using formalisms, especially those of formal logic. A closer look at this stage of reflection reveals a number of refinements worked out during history up to the present which we should now turn to. The first observation concerns the structure of propositions which are analyzed as consisting of content and form. We are concerned with the content of predications (in ‘mental language’, or ‘mentalese’, they are called ‘thoughts’) and their form, because predications always occur in a certain ‘mood’, within science usually in the mood of asserting. But equally important is the mood of assuming. Such moods carry claims with respect to the propositional content and, thus, show that there is a difference between claim and satisfaction. A proposition as such cannot tell whether its function, as spelled out by its mood, is fulfilled; it needs a judgment whether it does it or not. In case a proposition is the kernel of an assertion, such an assertion is the object of a (theoretical) judgment with respect to truth and falsehood. By the way, the Latin term ‘propositio’ (in Greek: πρότασις) with its distinction of affirmative and negative ‘propositiones’ referred to the use of propositions in truth-seeking argumentations, hence, it is equivalent to ‘assertion’ and not to ‘proposition’. Judging an assumption, on the other hand, will not occur with respect to truth or falsehood but, rather, with respect to its likelihood or, e.g., in scientific theories with respect to its suitability to act as an axiom. I should add that I have returned to the original usage of the term ‘judgment’ as prevalent still in the work of Gottlob Frege, because a judgment in the case of assertions is the result of judging an assertion to be true or to be false – ‘true’ and ‘false’ are the only uniformly applicable metapredicates within such judgments according to Frege3 –, and a judgment is not the same as, e.g., an assertion by itself, unless, as often in ordinary language, assertion and judgment are executed by the same utterance. To judge an assertion to be true means (practically) accepting it as a piece of knowledge, and it does not mean (theoretically) being aware of a ____________ 3

Cf. Frege 1967.

Another Version of Methodological Dualism

151

property of an assertion – truth cannot be such a property because assertions may turn out to be false –, whereas mere assumptions remain ‘neutral’ between acceptance and rejection as a piece of knowledge. Of course, we may now ask, on what grounds does a judgment concerning the difference between claim and satisfaction turn out to be reliable, and what happens is that we come back to the problem of proof we started with. A proof is something looked for in the case of judging assertions, whereas it is presupposed in the case of judging statements. To proceed further we should again analyze predications, and, here, we have to pay attention to a conspicuous shift from the traditional account to the modern one since Frege. Instead of traditionally analyzing a predication or, rather, its content as predicating a (predicate-)concept of a (subject-)concept which yields a subject-predicate-proposition, i.e., a minimal proposition, the modern account treats predication as predicating a propositional form (represented by a general term with one or more object variables) of one or more individual objects (represented by singular terms) which yields elementary propositions about these objects. The common terminology for this account at present is the functional one. Predicating is understood as applying a function to its argument(s), arriving at a proposition as its value, or, if propositional forms are treated as truth-functions, which means to disregard the distinctions between proposition, mood and judgment, arriving at either truth or falsehood as its only values. Instead of taking directly the next step of looking again more closely at the internal structure of concepts and of propositional forms, we will reach our next level of inquiry in a better way by further scrutiny into the ideas underlying the shift of analyzing predications. For this purpose we turn to various conceptual accounts given of individual objects or particulars, i.e., the domain of entities, eventually every scientific enterprise is about, even in such lofty cases as, e.g., the mathematical sciences, where numbers refer to particulars just with respect to counting them and geometrical forms refer to particulars with respect to certain ideal shapes of them. In antiquity up to the rise of modern science, such particulars, entia or res (Greek: ὄντα), have been understood as mixta composita out of matter (materia, ὕλη) and form (forma, εἶδος) which means that they have been treated as conceptually determined and not as something intuitively given, e.g., by means of perception, such that formal construction, e.g., by means of categories, the top concepts, will either make them comprehensible and thus amenable to knowledge about them or will wait for suitable satisfaction by the intuitively available particulars. The whole modern dispute, originally between classical rationalism and empiricism including its 20th century analytical cognates of logical rationalism and empiricism, and at present between mentalism and behaviorism or

152

Methods in Philosophy, in Art, and in Science

its newborn top-down or bottom-up mergers of structuralism and operationalism concerning the foundation of scientific knowledge, the whole modern dispute can basically be understood as a dispute about what comes first, form or matter, the powers of mind or the powers of nature, where both parties are usually fully aware of the fact that mind belongs to nature and that to speak of nature is a peculiar achievement of mind. Now, on the basis of particulars being built up out of matter and form, a minimal proposition, i.e., a subject-predicate-proposition like ‘Socrates is sitting’, is, e.g., by Thomas Aquinas, construed as predicating a formally accepted object of a materially accepted object and in this way uniting the form of Socrates with the form Sitting. And as predicating is at the same time understood to be a case of judging an assertion, the question to be answered is, whether the combined form of Socrates-sitting belongs to a corresponding matter of Socrates-sitting. The sentence ‘the sitting Socrates exists’ is, as Franz Brentano has argued for in detail more than a hundred years ago, tantamount to ‘‘Socrates is sitting’ is true’.4 This happens when the property of being-sitting inheres in the substance of Socrates (inherence-theory of predication) or when the matter of an event-object ‘sitting’ is part of the matter of Socrates which may also be expressed by saying that the individual concept of Socrates contains the concept of having-sitting-as-a-part (praedicatum inest subiecto: (partial) identity-theory of predication). This account, especially when expressed by means of the identity-theory of predication, lends itself easily to other minimal propositions, e.g., those where the subject term is not a proper name but a common noun like ‘man is an animal’. You just forget about matter, or you think you may do so, and you speak of forms only, the universals of an agelong dispute. Asserting ‘man is an animal’ may be judged to be true, if the concept of animal is contained in (the definition of) the concept of man, and it may be judged to be false, if the concept of man is contrary to the concept of animal. This intensional reading of such a proposition is likely to forget that species, too, are not without matter, i.e., the ‘sum’ of the matter of its members. Hence, the intensional reading of ‘man is an animal’ should not be understood in opposition to the extensional reading ‘all men are animals’, as it is prevalent in modern times, but rather as supplementary to it. We have arrived at an analysis of subject-predicate-propositions – at least of some of them – which exhibits that as a ground for making the corresponding true-or-false-judgments, relations of containment and of contrariness among certain concepts obtain, where ‘S is contained in P’ may also count as an explication of ‘P implies S’. Within this line of analysis ____________ 4

Cf. Brentano 1924.

Another Version of Methodological Dualism

153

implication and contrariness among propositions reoccur on the level of minimal propositions as implication and contrariness among concepts. It looks as if we had got rid of both matter and intuition at the same time, because in the case of individual objects the individual concepts take the place of their matter, a reason for the debate about particulars being infimae species or not. But the overwhelming conviction that even when they may be understood as lowest species, we will never reach this lowest level conceptually because individuum est ineffabile, this conviction has led Leibniz to make a fundamental division between finite and infinite definitions of concepts. Finite definitions define species in the ordinary sense, whereas actually infinite ones define particulars or infimae species, and because humans will never be in the possession of actually infinite definitions, they use the matter of particulars as symbols for their infinite individual concepts or forms.5 How is this done? According to Leibniz – and here he was far ahead of his time, in fact, only now, after the invention of pragmatic approaches and of hermeneutic ones in the philosophy of science, we begin to appreciate the ingenuity of his theory of monads – by having perceptions of the matter of particulars. Perceptions represent every piece of nature in my mind such that ‘I’ is nothing but a mirror of the universe from the perspective of ‘I’. I should add that Leibniz uses ‘body’ and ‘soul’ as the two notions of subject – the first being empirical subject and the second logical subject – instead of ‘matter’ and ‘form’ which had been turned by Descartes from having an ontological status as esse subiectivum (= obiectum materialiter acceptum, or subiectum) and esse obiectivum (= obiectum formaliter acceptum, or cogitatum) in Scholasticism, respectively, into having an epistemological status as res extensa, or object of knowledge, and res cogitans, or subject of knowledge. Thus, the reflexive turn of modern philosophical method which had led Descartes to the self-sufficiency of the thinking self, was developped further, by Leibniz in logically seeking consistency with the old objectivist tradition, and by Kant transcendentally in trying to give an ‘objective’ foundation to epistemological subjectivism. It is in this way that Leibniz eschews the modern alternative of keeping particulars and concepts, that is the physical and the mental in the Cartesian tradition, strictly apart, a move which is easily seen to be inadequate when the difference of particulars and concepts is not treated as exemplifying the difference of nature and mind but as realization of the difference obtaining between the individual and the universal, in tune with the older tradition. ____________ 5

Cf. Leibniz 1965, § 61.

154

Methods in Philosophy, in Art, and in Science

Following the modern turn, we know of the physical by external perception, and we know of the mental – I restrict it to the cognitive part as I cannot include a discussion also of the emotive and volitive parts – by internal perception or apperception, both being accessible through language, in its representative function with respect to external perception and in its expressive function with respect to internal perception. As the cognitive part of the mental is defined by its role to provide knowledge (of the physical and of itself), it follows that with reference to particulars we can distinguish immediate perceptual knowledge from mediate conceptual knowledge, provided this is understood on the level of performance, i.e., of appropriating knowledge, and not on the level of referring to knowledge as an object; you may think of ‘seeing a cat’ in contrast to ‘seeing something as a cat’ as examples of perceptual and of conceptual knowledge, both on the level of performance, respectively. This kind of knowledge is in both cases subjective. Its performance by the knowing subject is nothing but satisfaction of knowledge, if that knowledge is understood on the level of referring to knowledge as an object by means of perception and apperception. Knowledge taken as an object, of course, uses semiotic means to refer to it. It is made accessible through language. And we observe that (subjective) satisfaction of knowledge switches into a mere claim to (objective) knowledge, no satisfaction being available on the language level. Subjective satisfaction is traditionally called ‘intuition’. Expressed by language, though not necessarily by verbal language, it appears as a claim to ‘intuitive knowledge’, with respect to (external) perceptions as perceptual knowledge or ‘knowledge by acquaintance’ (‘seeing a cat’) and with respect to apperceptions as conceptual knowledge or ‘knowledge by description’ (‘seeing something as a cat’). It is helpful to be aware that intuitive knowledge with respect to internal perceptions being logically of second order because it addresses the means to establish first order intuitive knowledge, is usually called ‘formal knowledge’ and, therefore, a vague allusion to its historic origin of being concerned with form rather than with matter, though on closer inspection such formal knowledge may indeed be analyzed as implicational knowledge about two distinct bits of perceptual knowledge. Conceptual knowledge being second order intuitive knowledge, is in this way explicated as knowledge of an implication relation between true propositions. In our example: ‘seeing something as a cat’ ought to be read as “‘seeing a thing’ implies ‘seeing a cat’”. Hence, following common usage, only perceptual knowledge should be called ‘intuitive knowledge’ in the narrow sense. Now, both claims, to intuitive knowledge and to formal knowledge, are objective, in the sense of subjective satisfiability for everybody, only as

Another Version of Methodological Dualism

155

claims. Even common acceptance lacks any criterion for comparing the individual subjective satisfactions with respect to same and different: Existence of objective knowledge beyond mere personal conviction seems to be unavailable, it is unfounded and arbitrary. The result is disappointing. One has tried, therefore, to meet its aporetic nature with various moves to overcome this situation. One type of moves is dogmatic in nature: the objective existence of a universe of particulars is postulated. Various forms of realism as background for scientific theories may be developped from that. Another type of moves is skeptic in nature: knowledge remains problematic and may at any time be open to revisions. From this various forms of structuralism for the set-up of scientific theories evolve. In the first case, it is perceptual knowledge which gets a more secure footing by being freed from the confines of privacy, whereas in the second case, it is conceptual knowledge which is set free to adjust better to holistic criteria for good scientific theories as a whole, like simplicity, comprehensiveness, consistency, predictive power, and others. But hope for a notion of proof and with it of a notion of truth independent of such external conditions which make them applicable only relative to specific (scientific) theories, in a strict sense even only relative to formalized theories, is waning more and more in some quarters. You will nowadays even come across such self-defeating slogans like ‘science is but a myth, albeit a modern one’. In either version of predication, the traditional one which leads to subject-predicate-propositions and is still used in grammar, and the modern one which, since Frege, lets logic start with elementary propositions, we encounter a predilection for one of the two sides that Greek mental labor finally determined as simultaneously being the components of reality and of our knowledge about it: matter and form. Pre-modern tradition tends to prefer an intensional account by adapting individuals to universals through the use of individual concepts, thus subordinating matter to form, modern tradition with its tendency to adapt universals to individuals when taking universals to be individuals of logically higher order, i.e. classes, shows a preference for an extensional account and is thus subordinating form to matter. Of course, within both traditions the alternate account, respectively, is defended, too. I had, already, mentioned the modern debate about rationalism and empiricism which is the continuation in epistemological terms of the medieval debate in ontological terms about realism and nominalism. It is due to the lack of clarifying the two basic functions of verbal actions together with their interaction as they were determined first by Plato, the function to signify (the function of ‘naming’: ὀνοµάζειν) and the function to communicate (the function of ‘saying’: λέγειν), that the com-

156

Methods in Philosophy, in Art, and in Science

municative role of concepts got often mixed up with the significative function of linguistic entities expressing them: They ‘denote’ concepts as their (intensional) meaning or ‘sense’, but at the same time these linguistic entities are also said to ‘refer to’ individuals on various logical levels as their (extensional) meaning or ‘reference’. This twofold direction of the meaning-relation of basic linguistic entities, to universals, often via ‘mentalese’, a language of ideas or mental images said to be common to all humans, and to individuals, equally often via sensations or physical impressions from entities allegedly existing independently of human minds, which was introduced in order to account for the role of mediation language is said to play between mind and nature, e.g., by having ideas of sensation according to John Locke, this twofold meaning lies at the root of many difficulties connected with the usual accounts given of the process of knowledge acquisition in connection with proof procedures. I may only remind you of the dispute about possibility and impossibility of a private language so vigorously initiated by Ludwig Wittgenstein.6 It was the functional account of mind as presented by Leibniz through his idea of perceptions as representations of nature and further developped by the pragmatic foundation of semiotics especially in the work of Charles Sanders Peirce and others working along similar lines (e.g., Gilbert Ryle in his The Concept of Mind7 and Wittgenstein in his Philosophical Investigations8) which introduced a change of view towards language and with it of knowledge both on the level of performance and on the level of knowledge as an object. Language has no longer the role of mediating between mind and nature but mind operates on nature and converts our dealings with nature including ourselves, i.e., our actions, into sign-actions. Ordinary actions acquire representational functions, they become parts of a language, though, of course, at first of a non-verbal one. By way of caution I should mention that, in spite of much work developping the functional account of mind, the traditional view of treating language as the means to mediate between mind and nature, i.e., the ‘inner’ and the ‘outer’ world, is still quite influential, because it looks commonsensical. We all have been accustomed to it due to a long tradition. We commonly say that through perception we gain intuitive knowledge of the outer world which is represented by (true) propositions about individuals, i.e., in the more elaborate language of this exposition: the knowledge claim is put forward by an elementary proposition in the assertive mood. We, ____________ 6 7 8

Cf. , e.g., the influential essays by Saul Kripke on Wittgensteins arguments over the issue of private language: Kripke 1982. Ryle 1949. Wittgenstein 1953.

Another Version of Methodological Dualism

157

likewise, say that through conception we gain formal knowledge of the outer world which is represented by a (valid) implication between two (true) propositions about an individual, i.e., again rephrased more properly by saying: the knowledge claim is put forward by asserting an implication relation between two assertions claimed to be true. Of course, this is not a case of logical or formal implication, the validity of which would depend on the truth-values of the asserted propositions only. Rather, taking our catexample, you have to know that some of the features of the thing seen when taken together, are a characteristic feature of a cat when seen. What is the consequence of this exposition with respect to presenting the traditional view? In the conceptual case you not only see a cat, but you also ‘know’ something about properties of seen cats, where these properties are properties of a thing you have seen. Hence, intuitive knowledge gained through perception is kept clearly apart from higher order formal knowledge pertaining to conceptual relations – in our case they are ‘togetherness’ and ‘implication’ – between percepts that are in this way, namely by making use of an internal structure of them in order to spell out their relation, turned into concepts. But in both cases satisfaction of the respective knowledge claims remains a private affair, of one’s perceptual faculties with respect to nature and of one’s conceptual faculties with respect to mind. It is the functional account of mind which will lead us out of modern subjectivism, and it works by making use of the natural side of language, inherent in Willard V. O. Quine’s programme of naturalizing language,9 and combining it with a look at the symbolic side of nature which had been Ernst Cassirer’s concern when he tried to symbolize world.10 These moves turn both language and nature into actions or, better: they require to develop language out of one side of actions and to develop nature out of the other side of actions, these two sides being their symbolic and their natural side. How will these two sides of actions come about? They show up when you turn to the process of acquiring an action competence in an elementary dialogue situation where two agents learn something by doing and understanding (what one is doing). At each given instant just one of the agents is active while the other one is passive. Actions are both performed from one angle (their natural side) and cognized from the other angle (their symbolic side) which, in the performance case, amounts to being able to produce singular tokens of a universal type, and which, in the cognition case, results in being able to identify different tokens as belonging to the same ____________ 9 Cf. especially: Quine 1960 and 1974. 10 Cf. foremost: Cassirer 1923–29.

158

Methods in Philosophy, in Art, and in Science

type. Action competence is acquired when you know both how to perform and how to cognize an action. You may look at this process of competence-acquisition as a generalized Wittgensteinian language-game, in fact a language-game as an action game where pragmatic and linguistic activity are not yet separated but represented by mere aspects of actions from the angles of two agents. The performing agent is doing it, the cognizing agent is ‘seeing’ someone, possibly himself, doing it which means, he is understanding it. It needs quite a number of steps to be worked out in detail by a principle of self-similarity in order to get the idea of actions being equipped with two sides developped far enough that a clear picture of the functioning of verbal language arises. Verbal language evolves as being simultaneously mind (its functional or semiotic/symbolic side) and nature (its substantial or pragmatic/natural side) where both sides are essentially derived from the symbolic side of actions. This cannot be done here.11 I have to restrict myself to some sketchy remarks about how to answer the initial question, how intuitive and formal procedure in producing proofs of knowledge claims interact and indeed fulfill their function of satisfying such claims publicly. One of the important results on the way corroborates the ancient finding that particulars have to be understood as composed out of form and matter. They are made up, on the one hand, of schematizations that are nothing but the well known ‘universalia’ identified as ‘types’ of perceptions with respect to the action of dealing with the particular in question, and, on the other hand, of actualizations that in turn are just the equally well known ‘singularia’ (not individuals!) as ‘tokens’ of constructions with respect to the action of dealing with the particular in question. Particulars are composed out of universal ‘aspects’ – you may think of the properties of a particular – and singular ‘phases’ – they are the stuff from which their parts are made from. Hence, particulars may metaphorically be called to consist of ‘half thinking’ (they are schematized by aspects) and ‘half doing’ (they are actualizations of phases). In correspondence to this we get a clear picture also of the agents. They designate perceptions using socially available sign-actions or articulations, and they partake in constructions by being engaged in the transfer of partial actions from one agent to the other, such that particulars are both symbolically designated and symptomatically present. Another important achievement on the way concerns the articulations that may be represented canonically by the product of a sign-action, an articulator, i.e., Ferdinand de Saussure’s signifiant, such that it is func____________ 11 For further details, cf. Lorenz 1996 (reprinted in: Lorenz 2009, pp. 24–71), and also: Lorenz 2005, reprinted in this volume, pp. 42–55.

Another Version of Methodological Dualism

159

tionally equivalent with any other way of articulation vis à vis the perception in question. Humans use verbal articulators for this symbolic purpose of establishing equivalence with arbitrary other articulations like drawings, gestures, or the like. And these symbolic articulators taken as signs and not as (the product of) an action (of utterance) may again be viewed at from two angles: pragmatically, with respect to the (two) persons involved, where their tokens function as (means of) communication, Plato’s ‘saying’, and semiotically, with respect to the particular involved, where their types function as (means of) signification, Plato’s ‘naming’. We have reached the use of one-word-sentences, where the question of judging their utterance in a certain mood with respect to its validity does not arise. Only after separating the functions of saying and naming, and this involves a separation of predication in a mood from ostension in a mode of being given, together with the construction of complex articulators out of simple ones, also the situations in which you speak and about which you speak get separated from each other. Knowledge of language (Sprachwissen) – what do you ‘mean’ when saying something? – and knowledge about the world (Weltwissen) – does ‘exist’ what you name? – become different public affairs. The linguistic task of defining meanings for complex articulators and the task of science of proving the existence of corresponding objects, natural objects like electrons with such and such properties in the natural sciences, and cultural objects like unicorns with such and such properties in the cultural sciences, a task which may also be rendered as proving truth claims put forward by associated assertions, are tasks which have to be carried out separately, because the realm of meanings extends by far the realm of existants. Yet, both tasks are not independent from each other: you have to presuppose the meaning when trying to prove truth claims. The most general answer at present to the question of how to define meanings is given by devising language-games in the shape of rule-governed dialogues of interactions between two agents with respect to each articulator, where all kinds of activities may occur, e.g. experiments, observations, questioning, and so on. The only condition that the games have to meet is to obey well-defined rules of win and loss, something Wittgenstein certainly did not want to be part of a language-game. But, how else can they function as ‘measuring rods’ as Wittgenstein clearly wants them to be?12 It is on this basis, I propose, that truth claims of an agent may be satisfied by proving the existence of a winning strategy for that agent. It is argumentations that serve this purpose, and the well-known rules of formal logic have the special role of saving argumentative labor through their property that the existence of a ____________ 12 Cf. Wittgenstein 1953, § 131.

160

Methods in Philosophy, in Art, and in Science

winning strategy for the premisses of a logical rule implies the existence of a winning strategy also for its conclusion. Intuitive knowledge is not any more bound to private perception or restricted to the knowledge of axioms. It, rather, is called for with reference to any language-game around an articulator, and it rests on the publicly debatable conviction to be in the possession of a winning strategy when playing that game; it combines ‘knowing how’ to play and ‘knowing that’ one is going to win (or lose, if that is the case), without having derived such knowing-that from other sources which, then, count as reasons. Formal knowledge, on the other hand, is second order intuitive knowledge about systems of logical rules or formalisms, and it consists in knowing that the rules guarantee hereditariness of the existence of winningstrategies, a clear case of possessing a winning strategy of second order. It may, therefore, together with suitable other knowing-thats, be used as a reason for a first order knowing-that. You certainly realize that I have just applied modus ponens. Hence, knowing that one is going to win will show up in two versions. If it belongs to intuitive knowledge, it is immediate knowledge, if it depends on reasons using formal knowledge, it is mediate knowledge. And this distinction is now independent of the distinction between perceptual and conceptual knowledge in the context of traditional epistemological subjectivism. I may conclude that neither dogmatism through postulating reliability of intuition, nor scepticism through giving leeway to arbitrary formalisms, should act as a move to overcome epistemological subjectivism, a position that at any rate is but the twin of ontological objectivism.

Another Version of Methodological Dualism

161

References Brentano, Franz, 1924: Psychologie vom empirischen Standpunkte. Mit ausführlicher Einleitung, Anmerkungen und Register herausgegeben von Oskar Kraus. Erster Band [1874], Leipzig: Felix Meiner. Cassirer, Ernst, 1923–29: Philosophie der symbolischen Formen I–III, Berlin: Bruno Cassirer. CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Frege, Gottlob, 1967: Logische Untersuchungen. Zweiter Teil: Die Verneinung [1918], in: Gottlob Frege. Kleine Schriften, ed. by Ignacio Angelelli, Darmstadt: Wissenschaftliche Buchgesellschaft, pp. 362–378. Kripke, Saul A., 1982: Wittgenstein on Rules and Private Language, Oxford: Blackwells Publishers. Leibniz, Gottfried W., 1965: Principes de la philosophie ou Monadologie [1714], in: Gottfried Wilhelm Leibniz. Kleine Schriften zur Metaphysik, lat./frz.-dt. Ausgabe, ed. and translated by Hans Heinz Holz, Darmstadt: Wissenschaftliche Buchgesellschaft, pp. 438–483. Lorenz, Kuno, 1996: Artikulation und Prädikation, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal/Dietfried Gerhardus/Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1098–1122. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Logic, Thought and Action, ed. by Daniel Vanderveken, Dordrecht: Springer, pp. 343– 357. Lorenz, Kuno, 2009: Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter. Quine, Willard V., 1960: Word and Object, New York/London: John Wiley & Sons. Quine, Willard V., 1974: The Roots of Reference, La Salle Ill.: Open Court Publishing Co. Ryle, Gilbert, 1949: The Concept of Mind, London: Hutchinson. Wittgenstein, Ludwig, 1953: Philosophische Untersuchungen. Philosophical Investigations, Oxford: Basil Blackwell/New York: Macmillan.

The Pre-Established Harmony Between the Two Adams From Adam we turn to ourselves – “d’Adam se tourner vers soi”1 – from Christ we relate ourselves to God – “du Christ s’abandonner à Dieu”2 – this is a Leibnizian way to use the concept of φιλαυτία or ‘bonheur en soi’ as a mark of distinction between the old Adam and the new one. If we add to this his explanation “ad se converti est ad nihilum tendere”,3 we may think to have been referred to another piece of traditional metaphysics – Being and Nothingness – put into the frame of dogmatic theology without a chance of rational reconstruction. Yet, we all have learned that metaphysics in Leibniz is almost the same as logic,4 and that it is a good policy trying to find a logical version of the distinction I started with. I claim that by applying the Leibnizian doctrine of pre-established harmony, “entre le systeme des causes efficientes et celuy des causes finales”,5 to body and soul of Adam, i.e., to any human being since we belong to the familia Adami, we can reconstruct φιλαυτία as mistakenly treating soul as part of body, and this means to disregard or to suppress the representational character of souls (logical subjects) with respect to bodies (logical objects). Using signs for representing the ordering of efficient causes exhibits their being ordered by final causes; otherwise, signs, too, would be just marks, nomina sine notione, and hence, useless or a nothing. A sentence like “in corpore, non in anima, peccatum originis consistat”6 may, then, easily be understood as ‘forgetting about the sign-character of activities’ and this implies that whoever is considering his activities to be merely physical events, is annihilating himself. Let me start with Leibniz’s statement that his metaphysics is mathematical – “ma métaphysique est toute mathematique, pour dire ainsi, ou la ____________ 1 2 3 4 5 6

Grua 1948, p. 97. Ibid. Ibid. Cf. GP, IV, p. 292. Leibniz 1714a, § 3. ‘The origin of sin lies in the body, not in the soul’; Grua 1948, p. 245.

The Pre-Established Harmony Between the Two Adams

163

pourrait devenir”, he writes 1694 in a letter to the Marquis de l’Hopital7 – and as a consequence of this statement with his search for a characteristica universalis as a mathematical language where arguments can be represented by calculations. Then, mathematics, qua tertium comparationis for metaphysics and logic, i.e., for content and instrument, becomes a paradigm case.8 In his De primae philosophiae emendatione et de notione substantiae of 1694, a preliminary outline of the Systeme nouveau in the following year, for example, Leibniz explains how the concept of substance is going to provide mathematical foundations for metaphysics and for the empirical sciences. On the other hand, we know that logic itself is treated by him both as an instrument and as a subject matter, as a ‘general science of all that is conceivable’ (scientia generalis de cogitabili in universum) wherein the ‘principles and true method of philosophizing’ (principia ac veram philosophandi rationem) are already contained.9 Hence, this state of being contained in or of providing foundations – the first applies to principles, i.e., propositions, the second to concepts – is something which should be ascertained mathematically by means of calculations, if it holds. In 1677, already, in his Dialogus de connexione inter res et verba,10 Leibniz treats calculations, i.e., rules for treating characters, as independent from the kind of objects the characters are signs for. He explicitly states that, e.g., in geometry the characters are similar to their objects, which is not the case in arithmetic. But in both cases, what counts are the rules for dealing with the characters: they have to correspond structurally to the relations among the objects. Of course, we should not forget that Leibniz’s notion of a calculus, though fully specified syntactically, at least in principle, does not permit to forget that the characters of a calculus are signs of something. The transformation of logic into a calculus turns the scientia generalis de cogitabili into a mathématique universelle.11 But Leibniz’s outspoken anti-nominalism, he explicitly argues for in his Dissertatio de stilo philosophico Nizolii, too, should not be identified with what nowadays is often called ‘Platonism’, namely the belief in the reality of universals. Universals are, for Leibniz, possibilia and not real, as is well known. Again, this should not be turned into an argument against the younger brother of Leibniz, Charles Sanders Peirce, who ____________ 7 8 9 10 11

GM, II, p. 258. Cf. Lorenz 1989. GP, IV, p. 137. GP, VII, pp. 190–193. Cf. Leibniz 1704, bk 4, chap. 17, § 9.

164

Methods in Philosophy, in Art, and in Science

called universals ‘real’ and their singular instantiations ‘existant’,12 because what Peirce did was nothing but the invention of a possibly misleading terminology in order to convey the insight that singulars and universals are dependent from each other, quite alike the case of object and sign for object, where neither of the two can be referred to without using the other. I propose to read the use of signs, or characters in the sense of Leibniz, as schematizing reality. For example, using the word ‘homo’ or any of its cognates is tantamount to identifying a schematic feature: ‘being human’ in traditional language, or the feature universal ‘man’ in the terminology of Peter Strawson. Such a feature is not an object, unless you introduce abstract objects like humanitas. Hence, introducing schemata is not yet an answer to the problem of how to represent objects conceptually. The modern strategy to treat general terms like ‘homo’ as names of classes is unavailable for Leibniz, for in that case the elements of the class – in the case of ‘homo’ the human individuals – would also have to be representable, and that is exactly the problem in question. The availability of human individuals merely perceptually as phenomena would make the class-theoretic move a violation of the condition that the character ‘homo’ is a sign for something cogitabile. More precisely, such a move would turn ‘homo’ into not more than a sign for the remainder cogitabile class modulo its elements. The program of a mathesis universalis would have to be abandoned. Conceptual relations as given by sentences like ‘homo est animal rationale’ are in the beginning relations between schemata and hence ‘ideal’ relations. Without further moves they do not permit to say anything about objects possibly falling under such concepts. Such sentences refer, properly speaking, to the possibility of a conceptual relation. The usual reading of them as an inclusion among classes with variables that range over an object domain of individuals is obviously not generally adequate unless individuals comprise also possible individuals. In this situation Leibniz’s theory of monads comes to rescue. With the help of monads he intends to provide a conceptual foundation, i.e., a representation, for the phenomena which are characterized by a certain spatio-temporal togetherness of material, i.e., for bodies in Cartesian terminology. In a well-known letter, in 1715 to Conti, he writes:13 La matière même n’est pas une substance mais seulement substantiatum, un Phenomene bien fondé, et qui ne trompe point quand on y procede en raisonnant suivant les lois idéales de l’Arithmetique, de la Geométrie et de la Dynamique, etc.

____________ 12 Cf. CP, 5.430 and 6.335. 13 B, p. 265.

The Pre-Established Harmony Between the Two Adams

165

You get reliable knowledge if you establish mathematical, i.e., calculatory relations among the schemata that you ‘abstract’ from the bodies, e.g., among their states of motion provided the conceptual foundation of bodies, in a substance – as Leibniz says, is available, because otherwise the initial steps of the calculations cannot be made. The laws of mathematics and physics remain ideal descriptions, subject to the principle of contradiction, only. They cannot be derived from the concepts of the individuals involved, as it should be the case according to the additional principle of sufficient reason. Leibniz gives a very suggestive description of his state of mind near the beginning of his Systeme nouveau:14 Au commencement lorsque je m’estois affranchi du joug d’Aristote j’avois donné dans le vuide et dans les Atomes, car c’est ce qui remplit le mieux l’imagination. Mais en estant revenu, après bien des meditations, je m’apperceus, qu’il est impossible de trouver les principes d’une veritable Unité dans la matiere seule ou dans ce qui n’est que passif, puisque tout n’y est que collection ou amas de parties jusqu’à l’infini. Or la multitude ne pouvant avoir sa realité que des unités veritables qui viennent d’ailleurs et sont tout autre chose que les points mathematiques qui ne sont que des extremités de l’étendue et des modifications dont il est constant, que le continuum ne sçauroit estre composé. Donc pour trouver ces unités réelles je fus contraint de recourir à un point réel et animé pour ainsi dire, ou à un Atome de substance qui doit envelopper quelque chose de forme ou d’actif, pour faire un Estre complet.

The decisive argument here is contained in the sentence ‘a multitude can have its reality only by means of true, i.e., not only postulated, units’. I think, this may be understood properly only by interpreting it as a maxim that requires schematizations to be individuated in order to be schematizations of reality. I take this to be the Leibnizian version of Quine’s criterion for the ontological commitments embodied in a language: To be is to be the value of a variable.15 In either version there is no difficulty to argue for its validity. If schemata are available only by producing singular instantiations of them, rather than by exhibiting whole individual units, it would be impossible to judge schematizations performed by different persons as being the same or different. Leibniz goes further by requiring every part of a multitude to be determined as an individual, too. Such determinations of individual units he calls at first (in the mid-1680’s) individual substances or monads16 and later on (in the 1710’s), simple substances, as opposed to composite substances or ____________ 14 GP, IV, p. 478. 15 Cf. Quine 1939. 16 Cf. Leibniz 1686, § 8.

166

Methods in Philosophy, in Art, and in Science

bodies, i.e., multitudes.17 Of course, this terminology has resulted and still results in some confusion, because bodies are not substances but only phenomena grounded in substances, as we have just learned. And the uniqueness principle for bodies – ‘le principe d’unicité’ – is satisfied by the existence of one central monad for the body, namely its hierarchically uppermost monad. How does Leibniz proceed to arrive at individual units? It is often argued that Leibniz wants to turn our attention to one’s self-awareness as a starting point. But, in fact, in a famous letter to Arnauld in 1686 Leibniz writes quite precisely: “pour juger de la notion d’une substance individuelle il est bon de consulter celle que j’ai de moy même.”18 He advises each of us to turn to the notion he has of himself, not to just self-awareness, i.e., to himself as a bodily phenomenon. In focussing on the notion of oneself, one encounters what Leibniz calls qualités internes (not parts!) or internal states (états intérieurs), i.e., the well-known perceptions. They include the actions internes that govern the occurrence of a series of perceptions by means of a lex continuitatis which I consider to be conceptually completely different from the principle of continuity. The reason is that Leibniz speaks of a principe interne, a principle of appetitions. And internal qualities and internal actions are also called perceptions in a broader sense.19 But what are perceptions? They are “representations du composé, ou de ce qui est dehors, dans le simple”.20 Hence, they are activities or states belonging to the notion of oneself, i.e., to one’s soul, not to one’s body; they characterize oneself as a logical subject and not as a logical object.21 Activities of human beings belong to the soul when we attend to their representational character; otherwise they count as bodily behavior. Hence, psychology is either part of physics, if treated objectivally, or it is part of metaphysics in the Leibnizian sense of logic or semiotics. With this understanding there is no difficulty in following a central passage of the Discours de Métaphysique:22 … ce qui arrive à chacune [i.e., substance individuelle] n’est qu’une suite de son idée ou notion complete toute seule, puisque cette idée enferme déja tous les predicats ou evenements, et exprime tout l’univers. En effect rien ne nous peut arriver

____________ 17 18 19 20 21 22

Cf. Leibniz 1714a, § 1, and also: Leibniz 1714b, § 1. GP, II, p. 52. Cf. GP, II, p. 136. Leibniz 1714a, § 2. Cf. loc.cit., § 30. Leibniz 1686, § 14.

The Pre-Established Harmony Between the Two Adams

167

23

que des pensées et des perceptions, et toutes nos pensées et perceptions futures ne sont que des suites, quoyque contingentes, de nos pensées et perceptions precedentes, tellement que si j’estois capable de considerer distinctement tout ce qui m’arrive ou paroist à cette heure, j’y pourroist voir tout ce qui m’arrivera, ou qui me paroistra à tout jamais;

Perceptions represent the bodies of other monads, including my own body, and the representations are interior states or results of interior actions of the monad to which the perception in question belongs. In a modern terminology, due to Gottlob Frege, we may say that perceptions are the ‘Art des Gegebenseins’, i.e., the modes of presentation, of bodies.24 Even better: they are activities treated or understood as natural (and not as conventional) signs for bodies, because bodies have to be introduced in this way as invariants of perceptions and not to be postulated as already existing. We start with features as schematizations of reality, either naturally with perceptions or conventionally with characters for notions, the logical equivalents of perceptions; afterwards the features become representations of the invariants of perceptions, because otherwise reality is not a cogitabile. In such a way the universe is splitted into universals and singulars, but individuation is the crucial operation in order to arrive at representations of reality. Perceptions as sign-actions represent bodies – one’s own body more distinctly than any other; as sign-actions they qualify the logical subject – myself as a monad, that is, as a substance. Now, quite in tune with Leibniz’s intentions, the units underlying the bodies which are represented cannot be reached from the outside, as there are no external relations between monads. Only by utilizing the possibility to refer to perceptions, a reflexive activity or an apperception, words are available both for representing bodies and for the perceptions used for representing bodies, for example: ‘song of birds’ and ‘hearing a song of birds’, or ‘hand’ and ‘feeling a hand’. Besides, in the case of feeling, also activities which are not restricted to one of the five senses, e.g., shaking hands, would have to count as a mode of presentation of the schema hand. This example shows that in using ‘hand’ for the notion hand individuation of the feature ‘hand’ is not yet available. Rather, we should say, through apperception further refinement of perceptions is articulated. Leibniz states that by experience “la moindre pensée dont nous nous appercevons enveloppe une varieté dans l'objet”,25 which we can articulate, e.g., in a sequence like ‘hand’, ‘big hand’, ‘big hand far away’, ‘big hand of N. N. far away’, and so on. In ____________ 23 ‘Pensées’ are the results of apperceptions, cf. Leibniz 1714a, § 4, together with Leibniz 1714b, § 30. 24 Cf. Frege 1967, pp. 143–144. 25 Leibniz 1714b, § 16.

168

Methods in Philosophy, in Art, and in Science

such a way, I can grasp the notion of the carrier of all the schemata coexisting with the schema hand; this carrier is the body hand. But, and that is important, at the level of reflexion nothing but ‘the possibility of the monad hand’, i.e., of a hand as a logical subject, is available (using the principle of contradiction), since such a monad is existing only in my thoughts, from my perspective. The perspectives of other monads, their perceptions with respect to hand, are available only in thought and not as perceptions. We can think of the totality of perspectives of the universe, i.e., of God as the central monad of the universe, as a limit case where representation itself is by construction internal and hence indistinguishable from productive activity, which entails that God is without body, or rather, that soul and body of God coincide. To be able to refer to perceptions, i.e., to dispose of apperceptions, may now, as a special case of self-reference, be understood according to Leibniz as local similarity to God:26 [L’Esprit] n’est pas seulement un Miroir de l’univers des Creatures, mais encore une image de la Divinité. L’esprit n’a pas seulement une perception des ouvrages de Dieu, mais il est même capable de produire quelque chose qui leur ressemble, quoyqu’en petit.

The theory of monads does not serve as an epistemological theory to account for cognizing subjects and a cognized world of objects. It, rather, serves to transform the Cartesian dualism of body and soul into a pluralism of subjects which, in Leibniz’s terminology, are simple substances or souls where each of them is representing the whole universe, i.e., the composite substance as a whole. But how to render the relation among the monads if we obey the Leibnizian insistence that there cannot be external relations between them? From the first paragraph of the Principes de la Nature et de la Grâce we know that monads are characterized by their ability to act. Therefore, their ‘non-worldly’, i.e., non-external, relations must be expressible by using the notion of action, but, of course, not in the sense of conceptualizing the empirical forms of motion of human bodies, their ‘behavior’. The actions have to remain internal. But how, then, do they relate monads to each other? They do that only indirectly via their representational character. The actions of a monad lead from one representation of the body of other monads to another one, and in this way only to another mutual relation. The use of the terms ‘internal/external’ is Leibniz’s way to distinguish the level of objects from the level of signs being the level of representations of objects. The use of these terms is a means to keep the level of composite substances and the level of simple substances apart. The relation between a monad and its body that is articulated by Leibniz’s term ‘pre-established ____________ 26 Leibniz 1714a, § 14.

The Pre-Established Harmony Between the Two Adams

169

harmony’, may be established as a structural isomorphism of the partwhole structure on the level of bodies with the object-property structure on the level of souls, the actions being the properties of a monad and not of its body. The proof of this claim, which I have given elsewhere,27 proceeds by turning, say, a hand as part of the human body into having-a-hand as a property of the body when the body is taken as an individual unit rather than as a whole made out of parts; likewise, any property of an individual body gives rise to a part of this body transformed into a whole out of parts. As a conclusion I may state that it is the body of the old Adam and the soul of the new Adam which are in harmony with each other, rather than claiming that there is a pre-established harmony between the old and the new Adam. With this move we arrive at an understanding of the cryptic explanations Leibniz has added to the two phrases I started with: ‘d’Adam se tourner vers soi – Adam externus non nocet’ and ‘du Christ s’abandonner à Dieu – Christus internus prodest’.28 Leibniz combines these two explanations by stating that harmony consists in relating everything to one thing or, theologically: “in tenebris creaturarum lucet divina lux”.29

____________ 27 Cf. Lorenz 1977; in this volume pp. 20–32. 28 Grua 1948, pp. 97–98. 29 Op.cit., p. 98.

170

Methods in Philosophy, in Art, and in Science

References B = Der Briefwechsel von Gottfried Wilhelm Leibniz mit Mathematikern, ed. by Carl Immanuel Gerhardt, Berlin 1899. CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Frege, Gottlob, 1967: Über Sinn und Bedeutung [1892], in: Gottlob Frege. Kleine Schriften, ed. by Ignacio Angelelli, Darmstadt: Wissenschaftliche Buchgesellschaft, pp. 143–162. GM = Leibnizens mathematische Schriften, I–VII, ed. by Carl Immanuel Gerhardt, Berlin/Halle: A. Asher & Co. 1849–1863. GP = Die philosophischen Schriften von Gottfried Wilhelm Leibniz, I–VII, ed. by Carl Immanuel Gerhardt, Berlin/Leipzig: Weidmann 1875–1890. Grua, Gaston, ed., 1948: Gottfried Wilhelm Leibniz. Textes inédits d’après les manuscrits de la bibliothèque provinciale de Hanovre, 2 vol., Paris: Presses Universitaires de France. Leibniz, Gottfried W., 1686: Discours de Métaphysique, in: Gottfried Wilhelm Leibniz. Kleine Schriften zur Metaphysik/Opuscules métaphysiques, lat./frz.-dt., ed. and transl. by Hans Heinz Holz, Darmstadt: Wissenschaftliche Buchgesellschaft 1965, pp. 56–172. Leibniz, Gottfried W., 1704: Nouveaux Essais sur l’Entendement Humain/Neue Abhandlungen über den menschlichen Verstand, ed. and transl. by Wolf von Engelhardt and Hans Heinz Holz, 2 half-volumes, Darmstadt: Wissenschaftliche Buchgesellschaft 1959/1961. Leibniz, Gottfried W., 1714a: Principes de la Nature et de la Grâce, fondés en raison, in: Gottfried Wilhelm Leibniz. Kleine Schriften zur Metaphysik/Opuscules métaphysiques, lat./frz.-dt., ed. and transl. by Hans Heinz Holz, Darmstadt: Wissenschaftliche Buchgesellschaft 1965, pp. 414–438. Leibniz, Gottfried W., 1714b: Les Principes de la philosophie ou la Monadologie, in: Gottfried Wilhelm Leibniz. Kleine Schriften zur Metaphysik/Opuscules métaphysiques, lat./frz.-dt., ed. and transl. by Hans Heinz Holz, Darmstadt: Wissenschaftliche Buchgesellschaft 1965, pp. 438–483. Lorenz, Kuno, 1977: On the relation between the partition of a whole into parts and the attribution of properties to an object, in: Studia Logica 36, pp. 351–362. Lorenz, Kuno, 1989: Leibnizens Monadenlehre. Versuch einer logischen Rekonstruktion metaphysischer Konstruktionen, in: Zeit und Logik bei Leibniz. Studien zu Problemen der Naturphilosophie, Mathematik, Logik und Metaphysik, Carl Friedrich von Weizsacker and Enno Rudolph, eds., Stuttgart: Klett-Cotta, pp. 11– 31. Quine, Willard V. O., 1939: Designation and Existence , in: The Journal of Philosophy 36, pp. 701–709.

On the Way to Conceptual and Perceptual Knowledge I We have learnt that it needs both actions and sign-actions – using verbal language is just a special case of the latter – in order to acquire practical and theoretical knowledge of the many worlds we live in. They are ‘ways of worldmaking’ as Nelson Goodman has taught us when he turns our attention to “the structure of the several symbol systems of the sciences, philosophy, the arts, perception, and everyday discourse”.1 This modern turn of philosophy with the ‘linguistic turn’ of analytical philosophy being a restricted version only, may in its full generality be traced back to Charles Sanders Peirce whose work amounts to nothing less than a transformation of ontology into pragmatics and of epistemology into semiotics. Of course, these disciplines, pragmatics and semiotics, must not be understood as two newly established empirical sciences – in fact, Peirce himself fought, already, against such misunderstandings and even coined the term ‘pragmaticism’ for his brand of ‘pragmatism’2 – but as ways of investigation where empirical procedures are united with philosophical or reflexive procedures. Hence, actions as well as sign-actions may not only be treated as the objects of research and representation but also as a means of research and representation. You not only observe and describe these entities but you also produce and utilize them in order to reconstruct that which in observation and description is said to be given. Ludwig Wittgenstein, in his Philosophical Investigations, has used the term ‘language-game’ for activities that aim at disclosure of what is going on by providing tools of comparison. We may conclude that productions within such a context serve epistemic purposes notwithstanding their inseparable efficient role. A language-game may count as a paradigm case of perceptual knowledge, insofar as its significative function works by being an icon in the sense of ____________ 1 2

Goodman 1978, p. X. In the essay What Pragmatism Is of 1905, cf. CP, 5.411–5.437.

172

Methods in Philosophy, in Art, and in Science

Peirce. You have found an area of internally structured objects by inventing a prototype. Even more radically, the very distinction of action and sign-action has to be relativized if you allow for a purely functional account of both what it means to be an object and what it means to be a sign (of an object). In fact, it belongs to one of the basic tenets of Goodman’s approach in Ways of Worldmaking that the seemingly clear-cut division of world and language, where nonverbal language is, of course, included, as a division between the given and the constructed, between that which is found and that which is made, between the fact and the artefact, is outdated. It has even been challenged once and again, as Goodman shows,3 since the times of the PreSocratics. But, only rarely, history is looked at in this way. Any matter we are concerned with, Goodman tells us, is dependent on some manner as the means by which we deal with it. So worlds are but versions and worldmaking begins with one version and ends with another. The message we should learn, runs thus: “never mind mind, essence is not essential, and matter doesn’t matter”.4 The reorganizations of versions take place everywhere and again and again in various places ranging from subtle steps hardly worthwhile telling to big theoretical jumps like the widely cherished paradigm changes of scientific world views. Among the constructions occurring in such reorganizations Goodman distinguishes seven procedures, and he uses the cosmologies of the Pre-Socratics as an illustration of how four of them work: ordering in Thales when he derives all four ‘elements’ from water alone, supplementation in Anaximander when he introduces the ἄπειρον, i.e., the ‘boundless’, as something neutral underneath all of the elements, deletion in Parmenides when he declares everything besides the One, hence, the Many, as mere illusion quite akin to some Indian philosophers in the Vedānta tradition, division in Democritus when he dissects the One into atoms and, thus, reduces quality to quantity and structure.

Goodman goes on in claiming that we choose the facts as much as the framework, though this statement should better be split into two complementary statements: We produce the facts as much as the framework and we experience the framework as much as the facts. Productions, when taken epistemically, are always reproductions. Now, no doubt, we are able to live with different versions. There is no difficulty for anyone of us to say both ‘the sun rises in the morning and ____________ 3 4

Cf. sections 3 and 4 of the essay The fabrication of facts in: Goodman 1978, pp. 97–102. Goodman 1978, p. 96.

On the Way to Conceptual and Perceptual Knowledge

173

sets in the evening’ and ‘the earth rotates around its axis’. The problem which arises is basically not one of selecting the right version and identifying it by its being literally true in contrast to the wrong version that is at most metaphorically true. We should, rather, ask two more general questions:5 (a)

(b)

If there is – by way of comparison with the one world it is claimed we all live in – no chance to characterize one version uniquely as the true one – apart from the fact that only sentential versions can be true, others like pictorial versions cannot be subjected to a truth claim – what could be the criteria to distinguish right versions from wrong ones? If there exist different, even conflicting, right versions, how is it possible to live as ‘the same person’ in different worlds or, how to initiate and uphold communication processes between persons living in different worlds?

The issues connected with (a) which essentially boil down to the issue of how to identify the objects (colours, feelings, lines, percetual actions, things, experiments, etc.) we succeed and fail in constructing, are dealt with extensively in Goodman’s Ways of Worldmaking where ‘rightness of rendering’ is basically declared to be a matter of fit to what is referred to, whereby matter of fit is nothing but ultimate acceptability. The issues connected with (b) which in turn are focussed on the issue of how to secure the identity of the subjects living in possibly different worlds, have not yet been considered, because only the significative (‘referential’ is Goodman’s term) and not the communicative function of sign-actions (as well as actions if they, too, have semiotic functions) is dealt with by Goodman. To arrive at some clarification in this respect we have to go back to Peirce; it will be done in the last section. For the moment it is enough to feel convinced of Goodman’s success to have shown that the reorganizations of world views and the constructions used to achieve them, make the Cartesian separation of ontology and epistemology obsolete. And this move was made, already, by Peirce when he characterizes the methodology of pragmatism – in fact, pragmatism is not a philosophical position but a philosophical method, a ‘maxim of logic’ in the words of Peirce6 – by just the claim that ontology and epistemology are but two sides of the same coin.7 This implies that the variability of the boundary between facts and framework or, for that matter, between objects and signs of objects, may be ____________ 5 6 7

Cf. Lorenz 1987. Cf. the first of his Lectures on Pragmatism [1903]: CP, 5.14–5.212, together with extensive annotations and a German translation in: Walther 1973. Cf. CP, 5.257 as part of the essay on Questions Concerning Certain Faculties Claimed for Man [1868].

174

Methods in Philosophy, in Art, and in Science

turned into blurring the very distinction of object and signs of objects. We are confronted with the move to naturalize language including other symbol systems and to symbolize the world by paying attention to that feature of actions which is underlying Goodman’s treatment of exemplification as a tool to tie actions to symbols, and which Wittgenstein has achieved by introducing language-games. It should be added that these procedures of naturalization and symbolization should not be read as a move to make the steps of theory, i.e., the steps of devising world versions, amenable to moral philosophy. Goodman stated quite concisely: “my argument that the arts must be taken no less seriously than the sciences is not that the arts ‘enrich’ us or contribute something warmer and more human, but that the sciences as distinguished from technology, and the arts as distinguished from fun, have as their common function the advancement of understanding”.8 Sign-actions may be the concern of many and shall both in their natural or pragmatic aspect (performance) and in their symbolic or semiotic aspect (cognition) be viewed at with respect to enlarging or refining men’s abilities and not with respect to serving men’s needs, though, of course, this can be done, too. In this context, the abilities on the semiotic side, only, we call understanding, where beside conceptual abilities also perceptual abilities which in turn are served, e.g., in painting, by skills with or without external tools, should be included.

II The ideas of Goodman so far presented shall now be supplemented with ideas of his almost one generation elder contemporary Hermann Broch in order to shed more light on the distinction of perceptual and conceptual knowledge, that is, of knowledge in the arts and in the sciences, by closing the gap between objects and signs of objects. I want to show that the philosopher-artist Broch and the philosopherscientist Goodman proceed in a peculiarly dual manner. They both make claims concerning the epistemological relation of (cognizing) subject and (cognized) object which on the part of Broch is a conceptual one: There is one world underlying the various semiotic representations. On the part of Goodman we have been aware of a perceptual claim: There are many ways of worldmaking by the same agents. ____________ 8

Goodman’s answer to Hilary Putnam’s comment on Ways of Worldmaking during a symposium on this book, in: Goodman 1979, p. 619.

On the Way to Conceptual and Perceptual Knowledge

175

Now, the difference between conceptual and perceptual knowledge is characterized by Broch with the following words: “es ist die Aufgabe der wissenschaftlichen Erkenntnis […], zur Totalität der Welt in unendlich vielen, unendlich kleinen rationalen Schritten vorzudringen (it is the task of scientific knowledge […] to reach the totality of the world by means of infinitely many and infinitely small rational steps)”,9 and further: “[in einem einzigen] Kunstwerk [läßt ein Akt der Erkenntnis] die Totalität der Welt erstehen ([in a single] work of art [an act of cognition] gives rise to the totality of the world)”.10 In both cases we operate on the level of signs. Scientific procedure is a process of gradually increasing the internal differentiation of a representation of the world until it approximates reality. Broch uses the terminology of contemporary philosophy of science as he had learned it from the logical empiricists in Vienna and speaks of ‘setting up a rational model’11 in the effort to approach a ‘map of totality with respect to cognition’.12 Artistic procedure, on the other hand, is with each successful instance a repetition of creation by means of multifariously articulated symbols, a symbolic cosmogony eventually producing a ‘symbol of the world’.13 He states: “das Wertziel des Dichterischen, die kosmische Unendlichkeit, erfüllt sich in der einzigen Realitätsvokabel eines lyrischen Gedichtes (the value aim of poetization, cosmic infinity, is satisfied by the single expression of reality of a lyric poem)”.14 Here, the term ‘Realitätsvokabel’ (‘single expression of reality’) is used for a linguistic expression which, as material in a literary artefact, thereby generates a situation. For contemporary readers, Wittgenstein’s quite similar idea in his Philosophical Investigations (part II, section XI) comes into mind, where a rabbit-situation is generated by the exclamation ‘A rabbit.’. I quote part of the relevant passage: I look at an animal and am asked: “What do you see?” I answer: “A rabbit.” – I see a landscape, suddenly a rabbit runs past. I exclaim “A rabbit!” Both things, both the report and the exclamation, are expressions of perception and of visual experience. But the exclamation is so in a different sense from the report […] It is related to the experience as a cry is to pain.

In the report case I proceed from something perceptually known to something known conceptually – a case of knowledge by description. In the exclamation case something becomes known perceptually, using language ____________ 9 10 11 12 13 14

KW, 9.2, p. 48. KW, 10.2, p. 243. KW, 12, p. 43. KW, 9.2, p. 116. Ibid. KW, 9.2, p 136; p. 204.

176

Methods in Philosophy, in Art, and in Science

as a means of perception, not of conception – a case of knowledge by acquaintance. If we call the elements of processes leading to perceptual knowledge ‘poetic actions’ or poiesis, the concern of poiesis is just a determination of what can become a sign and how. The result, something invented, is, if successful, mimesis, something found. I am going to give a more elaborate explication a bit later. And philosophy as understood by Broch – he calls it ‘value-theory’ – is just semiotics combined with pragmatics in the Peircean sense, though in the main still based on abilities of single grown-up individuals and not yet on their gradual development by dealing with each other. Now, research and representation of value-positing activity, i.e., of philosophy, will appear in two complementary garbs both of which are characterized by a restauration of the unity of object and method in the sense of dealing with the object. Broch shows through analysis of many examples in history that separation of object and method of treating it, yields a growing appearance of mutual independence among primordially interrelated ways of (scientific and artistic) activity. The sciences become ‘positive’ fields which determine themselves ‘objectively’ whereby even their respective methods are treated as something specific for them, as their ‘metaobjects’. During this process, the sciences gradually lose their ability of self-reflection. The arts, on the other hand, turn into disciplines which define themselves by their techniques, whereby, in this case, they even reduce their respective objects to technical problems concerning the specific materials. Thus, they end up with activities of mere ‘l’art pour l’art’ and are unable to symbolize ‘totality’. The first, conceptual, garb of philosophy is epistemology inseparably tied to philosophical critique, the second, perceptual, garb of philosophy is philosophische Dichtung (philosophical poetry), that is, an art work which is simultaneously an instance of art criticism.15 Broch confines his term to literary art works, though he discusses extensively, e.g., its equivalents in music. Conceptual knowledge is bound to an ‘ideal’ language. Broch calls it ‘platonic’ and at the same time the ‘language of God’. It is without ‘style’,16 whereas perceptual knowledge is gained by the use of a style, that are those properties of an art work – in the literary case an art work in the verbal medium – that determine its being an articulated symbol. To uphold the unity of object and method, it is in both cases essential to realize that knowledge is concerned with the relation of ‘form’ (or ‘structure’) and ‘content’. Hence, neither the move of the autonomously acting sciences to ____________ 15 KW, 10.1, p. 234; KW, 9.2, p. 39. 16 Ibid.

On the Way to Conceptual and Perceptual Knowledge

177

treat problems of method as problems of higher order content, nor the move of the likewise self-sufficient arts to turn problems of content into problems of more refined techniques will remain open. Broch knows that attempts to characterize true cognition in the sciences internally, for example by following standards of rationality, do not suffice; they have to be complemented by focussing upon what he calls the ‘irrational roots’ or ‘irreducible remainders’17 that show up when one pays attention to the way scientific activity is growing out of daily life problems. It is likewise necessary, Broch insists, in order to grasp the cognitive value of the arts, to pay attention to the basic structure of a symbol, that is, its determination by an ‘inseparable connection of archetype and logos’.18 The unity of knowledge is dependent on realizing the “Totalität des Erkennens und Erlebens” (totality of knowing and experiencing).19 With these two terms ‘Erkennen’ and ‘Erleben’, we have again arrived at that basic distinction which I labelled earlier, using Bertrand Russell’s terms, ‘knowledge by description’ and ‘knowledge by acquaintance’. In fact, ‘Erkennen’ and ‘Erleben’ are almost precisely the German equivalents of the two Russellian terms. They may be characterized in a way that utilizes again Wittgenstein’s rabbit-example quoted above: Knowledge by acquaintance is knowledge completely dependent on the situation of speaking and acting, i.e., one knows what the exclamation ‘A rabbit!’ means only within a rabbit-situation. Therefore, I call knowledge by acquaintance ‘object competence’. On the other hand, knowledge by description is knowledge independent of that situation, i.e., one knows what the report ‘A rabbit.’ means also outside that situation, by previous experience, so to speak. For this reason I call knowledge by description ‘metacompetence’. The level of language partakes of both sides, the level of objects and the level of signs, depending on the circumstances. If language, or another sign system, is constitutive of the objects, i.e., if it is applied object competence, it displays symptomatic, object-like, features, if it is descriptive of the object, i.e., if it is applied metacompetence, it displays symbolic, representational features. Throughout his life Broch ponders systematically and historically on the common roots of science and art. He treats both as continuous rational ‘organizations’ (Formungen) of the unorganized which, in fact, contains always previous rational organizations. Both the sciences and the arts are reorganizations of world views quite in tune with Goodman’s idea that ____________ 17 Cf. KW, 10.1, p. 131. 18 Cf. KW, 9.2, p. 68. 19 KW, 19.2, p. 46.

178

Methods in Philosophy, in Art, and in Science

worldmaking consists in deriving a specific new version out of an older one. Scientific reorganizations take place on the level of metacompetence of second order, with their two branches of giving descriptions about the knowledge by acquaintance – this is called (scientific) research – and of giving descriptions about the knowledge by description – this is called (scientific) representation; you may think of rendering the descriptive knowledge axiomatically. On the other hand, artistic (re)organizations take place on the level of object competence of second order. They appear as an ability in constructing signs (semiotic objects) showing thereby that you are acquainted with structured objects, and also that you have descriptive knowledge about them. With respect to showing a knowledge by acquaintance that is dependent, as we have seen, on at least partial presence of the object, the artistic activity is called poiesis. What counts is what you do with semiotic objects like pictures, music pieces, words, etc., and not what they may signify. But with respect to showing a knowledge by description, the artistic activity is called – using a term that goes back to antiquity – mimesis. What counts is that which the semiotic objects signify. And it is this object competence of second order that we should call perceptual knowledge, in correspondence to metacompetence of second order being conceptual knowledge. Only here, on the level of second order, we are confronted with whole structures of pieces of knowledge by acquaintance and of knowledge by description, which usually is implied when referring to perceptual or conceptual knowledge. Broch treats myths as ‘first’ rational organizations, though, I think, this should be understood as restricted to artistic organizations, every day discourse being the candidate for first scientific organizations. He describes myths psychologically in terms of a fight against the anxiety and loneliness of death. Here again, a surprising coincidence with another artist-philosopher, Albert Camus, may be observed. Broch remains within the frame of individual psychology in his use of the extremes of panic and ecstasy for the possibilities of new experience, whereas Camus resolves these extremes dialogically into a commmon knowledge about one’s individual loneliness, a knowledge you experience through the limited solidarity of revolt against that state. But both Broch and Camus use the same metaphor to convey the success of ‘the will to live’ (Broch’s term) or the ‘yes to life’ (Camus’ term): It is balance. The rational organization must ‘fully balance’20 the unorganized which it mirrors. An art work must ‘balance the totality of the world’.21 Camus ____________ 20 Cf. KW, 19.2, pp. 214–215. 21 KW, 19.2, p. 210.

On the Way to Conceptual and Perceptual Knowledge

179

prefers the related term of keeping the ‘right measure’,22 because each rational organization done by oneself finds its boundary at a rational organization done by the other. What is understood delimits what is not understood and must balance it. Camus draws the borderline between different individuals, Broch between an individual and that which no longer, not yet, or never, belongs to him.23 Broch has even tried to sketch a reconstruction of the primary epistemic relation between a cognizing subject and the cognized object.24 He starts with the smallest syntactical unit, a sentence, and he asks how a cognitive unit – his term is ‘Eidos-Einheit’ – can be made visible and audible through it. As an answer to this question he states that each cognitive unit is related to a section of reality which he calls ‘Elementarsituation’ – an example he discusses is ‘flackerndes Licht’ (flickering light). In present-day systematic terminology as derived from Peirce and the later Wittgenstein the eidos is an object schema, and the related elementary situation is one of its actualizations, such that the primary epistemic relation can be described as understanding a situation as actualization of a schema. The eidos is universal, the situation something singular, and the two are strictly correlated to each other. With the next step, the articulation of a type that comes about by individuation of the schema such that the individual units of the schema are the tokens of the type, serves to show the correlation of type and token by representing the cognitive unit through the syntactic unit. In contemporary terminology this means the following: An elementary sentence, e.g., ‘this is flickering light’, states that an elementary situation (being a particular token that is indicated by an indefinite number of singular ‘experienced situations’, a ‘subjective perspective’ as explained further down) instantiates a type derived from the eidos ‘flickering light’. Every such sentence, if uttered in an appropriate situation, is an accompanying part of that situation by being the outcome of using ‘the verbal sense’ (looking at that situation would add to it by using the visual sense), thereby making the eidos perceptible. We would say that by uttering such a sentence the situation is understood as a token of a type derived from the schema. We may as well say that by uttering such a sentence the situation is schematized, i.e., imbedded in a possible sequence of like situations. But Broch does not stop here. He continues by declaring that the grammatical subject, i.e., the demonstrator (or logical demonstrative) ‘this’ in ____________ 22 Cf. Camus 1965, chap. V, section Mesure er démesure. 23 Cf. KW, 10.2, p. 201. 24 In his essay Über syntaktische und kognitive Einheiten, KW, 10.2, pp. 246–299.

180

Methods in Philosophy, in Art, and in Science

our elementary sentence, should be understood as ‘a projection of the speaker’.25 To proceed in this way has consequences which reach far beyond common logical or even linguistic analysis which stops at saying that ‘this’ refers to an object given in the situation of utterance. Broch says, and here his sophistication matches the one of Peirce, that the demonstrator ‘this’ is within the utterance an index of the construction of the object about which something is stated, out of its various subjective perspectives. Every object, an instance of a typification of a schema, splits into the indefinite class of modes of presentation with respect to the perceiving subjects, and it can actually be identified with the whole out of perceptions. Within our example one could articulate the perception which is used when uttering ‘this’ in ‘this is flickering light’ by ‘that which I just see’. Broch calls the perceptions ‘identifications of the speaking subject with its surrounding’. Hence, it is they that make the cognizing subjects appear, whereas the cognized object is, according to Broch, defined by the result of constructing the invariant out of the many perceptions. The consequences regarding the eidos together with the elementary situations as tokens of its typification, may easily be drawn. Through articulation each eidos gets split into perspectives each of which is a sign of the same eidos, whereas the signs themselves are appearances of a subject. While performing an articulation both aspects are completely amalgamated: a section of reality is, by virtue of the various signs which are used for the same eidos, simultaneously present and interpreted, i.e., understood as schematized by the eidos. In this way Broch has succeeded in presenting a construction where object and method, ontology and epistemology, appear, indeed, as merely two sides of the same coin. He, and we with him, are able to identify conceptual knowledge which searches for invariance of the represented as being ‘objective’, and perceptual knowledge which aims at variance of the representing means as being ‘subjective’. It would be interesting to see how his procedure works in more complex cases beyond essentially one term articulations of a situation, but that cannot be done here. Especially rewarding would be an investigation into Broch’s idea that Wittgenstein’s Tractatus and Joyce’s Ulysses could be analyzed as being embodiments of comparatively pure examples of complex conceptual knowledge and complex perceptual knowledge, respectively.26 A sentence, or better the correlated predicative expression, becomes a symbol in Broch’s sense – systematically it should be called a ‘symptom’ ____________ 25 Cf. KW, 10.2, p. 252. 26 For a more extensive treatment, cf. Lorenz 1988.

On the Way to Conceptual and Perceptual Knowledge

181

– not by simply being a sign of a schema but by exemplifying the schema (or eidos). The difference between exemplification and representation as two ways of reference tied to perceptual and conceptual procedure, respectively, was introduced by Nelson Goodman to account for the difference between artistic and scientific cognition. And it is surprising how Broch’s description of the functioning of verbal symbols in cases of ‘philosophische Dichtung’ comes close to Goodman’s exposition. Exemplifying a schema refers to the schema not by convention but by certain internal or external properties which the exemplifying predicative expression has in common with the schema. This implies that the schema itself is not any more a simple eidos but an eidos with an internal (and external) structure. The cognitive claim of poetic worldmaking is substantiated according to Broch by a work of art being a (structured) symbol of reality as a whole. So it is tied to art as mimesis, whereas Goodman has tied the cognitive claim of poetic worldmaking to art as poiesis. Both are right, obviously, but there is an important though often neglected relation that in works of art has to be established between mimesis and poiesis: Poiesis in art, the ‘style’ of Broch, must be the only means to make the mimetic function of an art work discernable.

III It is advisable, now, to use the third and final section for a sketch of Peirce’s ingenious idea of how to derive signs out of objects, as it is found in the late essay Meaning of 1910. The process there referred to, is well suited to give an explanation of how the two kinds of reference, exemplification and representation, relate to each other. This task is urgent, if, indeed, exemplification and representation are exploited heavily in the arts and in the sciences, respectively. If a Sign is other than its Object there must exist either in thought or in expression some explication or argument or other context, showing how – upon what system or for what reason the Sign represents the Object or set of Objects that it does. Now the Sign and the Explanation together make up another Sign and since the Explanation will be a sign it will probably require an additional Explanation which taken together with the already enlarged sign will make up a still larger Sign; and proceeding in the same way, we shall, or should, ultimately reach a Sign of itself, containing its own Explanation and those of all its significant parts; and according to this Explanation each such part has some other part as its Object. According to this every Sign has,

182

Methods in Philosophy, in Art, and in Science

actually or virtually, what we may call a precept of Explanation, according to which 27 it is to be understood as a sort of emanation, so to speak, of its Object.

The argument calls for something which is a sign of itself, that is, which combines object-features and sign-features, or better: which functions both ways. The argument is itself a section of an open sign process on the level of reconstruction and for this reason it may count as a conceptualization of a Wittgensteinian language-game. Now, the descending sequence of interpretants ends with an ultimate logical interpretant28 which is identified by Peirce as being a habit change, i.e., in contemporary terminology, as the acquisition of an action schema on the basis of already existing action competences, whereby all the ways to deal with the object in that respect which is signified by the initial sign are included. The procedure is even better understood, if, for an enhancement of clarity, we turn to the Peircean reading of the semiotic triangle as it is contained in the following quotation:29 “A sign […] stands in such a genuine triadic relation to its object as to be capable of determining its interpretant to assume the same triadic relation to its object in which it stands itself to the same object.” Each interpretant – and cognitions of a mind, i.e., mental interpretants, are some of them – is itself a sign for the same object, hence, it is generating a new interpretant, and so on. The sequence of interpretants which gets started in that way may be called a sequence of growing understanding of the object by supplying more and more differentiated determinations. It should not be forgotten that Peirce insists upon ever new, i.e., logically new, not empirically new, sign-users connecting the items of the sequence; he, therefore, speaks of ‘quasi-minds’ and not of ‘minds’:30 Signs require at least two quasi-minds, a quasi-utterer and a quasi-interpreter; and although these two are at one (i.e. are one mind [the dialogically constituted general subject!]) in the sign itself, they must nevertheless be distinct. In the sign they are, so to say, welded.

Going back to habit changes, that is, to the acquisition of an action schema, as the candidates for something that is a sign of itself and, thus, the final item of the descending sequence of interpretants, we may conclude that a verbal sign of an object signifies a range of possibilities to deal with that object. Even more general, deleting the dummy term ‘object’, we may say that to understand a sign-action is tantamount to knowing, by that very ____________ 27 28 29 30

CP, 2.230. Cf. CP, 5.476. CP. 2.274. CP, 4.551.

On the Way to Conceptual and Perceptual Knowledge

183

action, of a whole range of further actions which should be taken to be signified by the sign-action. Now, the sign-action is signifying symptomatically and not symbolically, if the sign-action itself is part of the range of actions it signifies. This implies that for verbal symptomatic sign-actions a complete separation of word and object has not yet taken place: their relation is of the part-whole-type. The same idea of explaining how symbolic actions, i.e., full-fledged verbal sign-actions, symbolize may be used to explain how ordinary actions function as (not yet verbal) sign-actions.31 We, then, arrive at the following equivalence: to understand a performance of an action in the sense of ‘knowing what one is doing’ is knowing ways of dealing with it. In a more general way, you may say that to know an object is the same as treating this object as a sign of its distinctions, that is, the ones that you are able to make. And again, this is nothing but treating an object as a sign of its internal structure, a structure which gets exhibited in an open-ended sign process. The simplest structure you start with, is, of course, the one which shows up when you are able to ‘read’ a token that is indicated by an actualization as a sign of the schema from which the type of the token is derived. And this works when you treat such an ability as conveyed by a ‘language-game’, the same one that we know, already, under the label of an acquisition process of an action competence in an elementary dialogue situation: The ‘agent’ is performing the action, and the ‘patient’ at the same time is cognizing it, role-switching included. We have arrived at the primitive stage where you have an icon of an object, an action both as a pragmatic entity and as a semiotic entity, whereas, making a further step, you learn to split the type into parts, an instantiation of one of them becoming a sign of the whole. This, then, is the stage where you dispose of an index of the object:32 “If the sign be an index, we may think of it as a fragment torn away from the object, the two in their existence being one whole or a part of such a whole.” These are the symptomatic sign-actions. The rest is well known, though, of course, there are many details to cope with which I cannot discuss here but have dealt with at other places.33 I only hope that I have succeeded in convincing the reader of at least one point when looking at the three authors, Goodman, Broch and Peirce, together in the way I did: Using language may always occur on two levels, like ____________ 31 For more details, cf. Lorenz 2005, in this volume pp. 42–55. 32 CP, 2.230. 33 The most detailed exposition is found in: Lorenz 1996 (reprinted in: Lorenz 2009, pp. 24–71).

184

Methods in Philosophy, in Art, and in Science

a perception and like a conception. Perceptually, language is used as a way of presenting an object which is a case of object competence, whereas, conceptually, with language you are referring back to other terms that have been used perceptually, already, which is a case of metacompetence. In the Wittgensteinian example of a rabbit-situation, the exclamation ‘A rabbit!’ is a case of perceptual use, and the report ‘A rabbit.’ a case of conceptual use. There is always a natural and a conventional relation between words and objects, a result which could now be used as a secure starting point for a critical evaluation of the debate on literal and metaphorical meaning of signs including verbal signs.34

____________ 34 For some more material on the last issue, cf. Lorenz 1991.

On the Way to Conceptual and Perceptual Knowledge

185

References Camus, Albert, 1965: L’homme révolté [1958], in: Albert Camus. Essais, ed. and introd. by Roger Quilliot, Paris: Gallimard, pp. 407–709. CP = Collected Papers of Charles Sanders Peirce I–VI, ed. by Charles Hartshorne and Paul Weiss, Cambridge Mass.: Harvard University Press 1931–1935. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks Sussex: The Harvester Press. Goodman, Nelson, 1979: Credence, Credibility, Comprehension, in: The Journal of Philosophy 76, pp. 618–619. KW = Hermann Broch. Kommentierte Werkausgabe I–XIII, ed. by Paul M. Lützeler, Frankfurt am Main: Suhrkamp Verlag 1974–1981. Lorenz, Kuno, 1987: Semiotic Actions in Worldmaking, in: Worldmaking’s Ways, ed. by Lars Aagaard-Mogensen/Rik Pinxten and Fernand Vandamme, Ghent: Communication & Cognition, pp. 11–24. Lorenz, Kuno, 1988: Broch’s Concept of ‚Philosophische Dichtung’, in: Hermann Broch. Literature, Philosophy, Politics. The Yale Symposium 1986, ed. by Stephen D. Dowden, Columbia SC, Camden House, pp. 303–314. Lorenz, Kuno, 1991: Überlegungen zur Differenz zwischen ‘buchstäblicher’ und ‘übertragener’ Bedeutung, in: Synthesis philosophica (International Edition/Zagreb) 11 (vol. 6, fasc.1), pp. 129–133. Lorenz, Kuno, 1996: Artikulation und Prädikation, in: Sprachphilosophie/Philosophy of Language/La philosophie du langage. Ein internationales Handbuch zeitgenössischer Forschung/An International Handbook of Contemporary Research/Manuel international des recherches contemporaines, ed. by Marcelo Dascal/Dietfried Gerhardus/Kuno Lorenz and Georg Meggle, vol. 2, Berlin/New York: Walter de Gruyter, pp. 1098–1122. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Logic, Thought and Action, ed. By Daniel Vanderveken, Dordrecht: Springer, pp. 343– 357. Lorenz, Kuno, 2009: Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter. PU = Wittgenstein, Ludwig: Philosophische Untersuchungen. Philosophical Investigations, dt./engl., transl. by G. E. M. Anscombe, Oxford: Basil Blackwell/New York: Macmillan 1953. Walther, Elisabeth, ed. and transl., 1973: Charles Sanders Peirce. Lectures on Pragmatism/Vorlesungen über Pragmatismus, engl.-dt., Hamburg: Felix Meiner.

Self and Other: Remarks on Human Nature and Human Culture The present widespread interest into problems of intercultural and intracultural communication, disputes being its paradigm case, is certainly symptomatic of a particular feature of the very historical situation we belong to. We live in a period of growing consciousness that the belief into just one uniquely determined reality, the world of matters and the world of manners, which has governed our Western intellectual history almost unchallenged up to the last century, is unwarranted. It has become impossible to avoid facing more than one, obviously equally plausible though incompatible, set of rules for a way of life, and the same holds with respect to the mentally represented structures which we call a world view. The resulting scepticism may well belong to the sources that are responsible for the rise of all kinds of fundamentalism which herald parochial views and values and try to spread them worldwide as an alleged antidote to the formal character of global culture. Fundamentalists suspect that the idea of a universal global culture is nothing but a claim for what may be called scientific world view and they are afraid of a liberal way of life that relegates common views and values to procedural ones of handling diverging claims. Further complications enter this picture if another fact is taken into account. It is certainly not impossible to change ways of life considerably and equally well to advocate different world views. Changes in the way of life may be brought about under outside influence, politically enforced or encouraged, for example; but likewise under inside influence, for example by self-indoctrination or by deliberate decision and practice. Diverging world views are voiced in the case when, at alternate times, you lead the life of a scientist and of an artist, or when you have learned to participate in divergent cultural traditions; it may also happen when, due to certain events, one has changed radically one’s course of life, has turned from Saulus into Paulus, for example. The views and values of an individual are not only changing through time, they also change according to situation without necessarily exhibiting a stable kernel that, in our tradition, is considered to be an essential prerequisite for being allowed to call somebody a fully grown-up person. It seems, as if we should be prepared to face, on the one hand, internally unconnected sequences of world views

Self and Other

187

and ways of life without any tertium comparationis to mediate between them, and, on the other hand, proliferations among world views and ways of life without a common ancestry. For getting a better perspective on the various questions involved it is advisable to turn to a short historical retrospection. There is a document where the idea to become human is spelled out in terms of a cultural process which starts from a state of nature. In fact, it is a myth, attributed by Plato to the sophist Protagoras and related to us in his dialogue Protagoras.1 There, we read that human beings come into existence by two distinct steps. At first, technical abilities are acquired in order to compensate for natural deficiencies; these abilities appear individually as a character of the species and make its members capable to survive collectively as natural beings. The whole realm of τέχναι which includes the arts as well as the religious rites is available on this level. After that, a second step is taken which leads to practical abilities that provide for units of self-government individually as well as socially, and which has to be understood as the advent of rationality. On this second level, it is honesty or justness (δίκη) and modesty or respect (αἰδώς) which make up solidarity (φιλία). Practical abilities, if understood in this way, arise from mutual recognition among individual human beings and, therefore, they are dependent on language. They cannot be exercised except through individual distinction and social coherence. Any such individual is leading a life governed by reason: it is an animal on the first level and it became a rational one on the second level. Thus, we have identified the backbone that is necessary for guaranteeing the original equivalence of the two ancient ‘definitions’ of man: animal rationale (ζῷον λόγον ἔχον) and animal sociale (ζῷον πολιτικόν). Two additional remarks may be helpful for refining the picture. Plato, in commenting on the myth, relegates every action which belongs to the sphere of the first step, that is, which is poietical and, therefore, non-rational or not-yet-rational, to a sphere of merely natural or cultural exigencies; such an action is not specifically human. Furthermore he divides rational activity, i.e., the actions which belong to the sphere of the second step, into those which ‘judge’ claims to ‘mathematical’ knowledge and those which ‘enact’ claims to ‘political’ knowledge.2 The move from the first step to the second one may be interpreted as a move from ‘amoral’ competitive behavior to the more developed stage of ‘moral’ cooperative behavior, provided one does not forget that the notion of rationality as the characteristic feature of practical abilities is not limited to the area of moral (and political) legislation but includes the area of scientific judgments as ____________ 1 2

Prot 320c8–323c2. Cf. Polit 258e–261a.

188

Methods in Philosophy, in Art, and in Science

well, that is, those on matters which can be taught and learnt (τὰ µαθήµατα). The second remark concerns an additional distinction which is due to Aristotle. It is the distinction between rationality as the ability to theorize and sociality as the ability to lead a good life which, on the one hand, mirrors the previous Platonic classification among practical abilities, and, on the other hand, permits to look at the two levels of Protagoras’ myth from another angle. Eventually, in Aristotle’s philosophical constructions the additional distinction leads to a different though related developmental bifurcation with an intermediate step. We are confronted with a primary level of mere acting which should rather be called behavior, and with a secondary level of deliberate activity, that is, actions guided or at least accompanied by thinking where theory serves practice. And, in addition, there exists a level in between the other two where poietical activity serves particular cultural needs above behavior necessary for natural survival and below action guided by reason peculiar to man. And it is the stage arrived at in this intermediate level where both the competitive activity of the primary level and the cooperative activity of the secondary level interact and keep the cultural process going. With this in mind it is easier to understand why there is so much discussion recently about the thesis – put forward, e.g., by Huntington3 – that conflicts nowadays should be understood as not any more being conflicts among different national states or units but as conflicts between different cultures. It reflects the widespread influence of the very picture of the cultural process derived from antiquity in the way just sketched, and which is still the dominant picture due to the additional support that it got from enlightenment with its slogan of the ‘Erziehung des Menschengeschlechts’ coined by Lessing in 1780. We may call it the progress-theory of the cultural process where the progress is measured in terms of complexity of (group-)organisation with Hobbes’ assumption of a bellum omnium contra omnes as the characterizing feature of the initial stage, that is, the state of nature. The individuals count as members of a species which will eventually be fully governed by reason. Privileged documents for a development along this line are the constitutions designed to elicit potentially universal acceptance. Political universalism, the idea of enlightenment, takes national states to be constitutional states, and unpeaceful fights among them count as signs of the initial barbaric stage which has not yet been overcome. Yet, we should be prepared to take an alternate picture of the cultural process into account that conflicts with the previous one and has come ____________ 3

In: Huntington 1966.

Self and Other

189

down to us from antiquity, too. This is of special importance, because its growing influence may be recognized everywhere. It is described in terms of a decline-theory rather than a progress-theory. On the basis of an aetas aurea, a natural state of paradisaic existence of which Hesiod had sung and which Rousseau had described anew, an ever increasing consciousness develops and spoils the ability to act spontaneously. Natural individual creativity is hampered by social pressures which derive from conscious activity for and against others, and the decline can be measured by the losses in (self-)production, that is, by the amount of creative potentials of individuals that remains unrealized. Only the free association of individuals as given in natural ethnic groups defined by a common frame of world views and of ways of life, that is, a culture, allows for uninhibited self-expression. Cultural particularism, the idea of romanticism, argues for national states to be ethnic states designed to fight for their cultural survival. Rather than moving from ‘barbaric’ competition to ‘civilized’ cooperation, as the cultural process is viewed in progress-theories, we have in this case the call to reverse the cultural process which leads, according to decline-theorists, from natural cooperation, possibly including cooperation with non-human nature as well, to cultural competition. Underlying these two theories of the cultural process we find two competing notions of culture – or rather two competing attempts to define the boundaries between natural and cultural aspects of being human. In the progress-theory, the claim is made that universalizing reason is the final destiny of man; in the decline-theory, the claim is made that particularizing nature is the primordial ruling force. In order to mediate between these two views regarding human nature and human culture a thorough scrutiny of the dialogic nature of man is called for. As Self, man is subject to ‘doing’, as Other to ‘suffering’, and neither of these two roles, the active one of doing something while doing it, and the passive one of experiencing what one is doing while doing it, can be reduced to just one of them. But, before embarking on this scrutiny which I have done in detail in other papers4 that elaborate on the ideas developped in my Einführung in die philosophische Anthropologie,5 it is worth taking a further look at the historical setting of the two ways to demarcate culture from nature. According to enlightenment, the prospect of cultural development defines the proper ‘second nature’ of man, whereas, according to romanticism, a natural potential acts as the cause of the true culture within the one and first nature of man. It seems as if among the premisses of enlighten____________ 4 5

Most of them are meanwhile available in: Lorenz 2009; in addition, on comparing competition and cooperation, cf. Lorenz 1994, in this volume pp. 140–147. Lorenz 1990.

190

Methods in Philosophy, in Art, and in Science

ment we have to include the belief in reason within history as ultima ratio, usually called the belief in historical progress, and correspondingly, in romanticism there seems to exist a belief in God in nature as prima causa that, in non-theological terms, is called the belief in natural evolution – up to evolutionary epistemology as promoted by some scholars nowadays.6 As to the last case of evolutionary epistemology we should be aware that it is an attempt even to dispense with the enlightened view altogether: Every human interference with natural evolution – and acquisition of knowledge should be treated as such – is interpreted as a part of natural evolution. Such a move runs counter to the classical idea of natural history to be understood as the deployment of reason7 – a modern version is radical constructivism as based on the concept of autopoiesis8 – with its consequence of embedding the romantic view in the enlightened one rather than conversely. These remarks should not be read as an attempt to initiate a discussion of the relative merits of romantic individualism and of enlightened universalism. With these latter characterizations one misses the point. We, rather should take note of the following awkward turns. The romantic slogan ‘back to nature’ – in fact already en vogue in antiquity with the Stoics: secundum naturam vivere – works as a kind of universal norm asking everybody for his or her respective engagement, e.g., in ecological movement, and the enlightened call for a self-determined way of life gives rise to a multitude of essentially dissociated life-plans, as it is well-known when observing, e.g., the modern phenomenon of yuppyism. Romantic individualism tends to become a version of communitarianism with its pledge that the community should have primacy over the individual, whereas enlightened universalism shows up as a kind of liberalism dependent on the primacy of the individual over the community. Both approaches suffer from intrinsic inconsistencies, and it is basically this ambiguous setting which is responsible for feeding contemporary debates like, e.g., the debate on communitarianism versus liberalism with its focus on how to defend or to attack ethnocentrism,9 that is, the particularity of value-systems, or the debate on the content and scope of human rights with its focus on how to identify universal values,10 if there is such a thing with an identity across cultures. ____________ 6 7 8 9 10

Cf. the critical review in: Wolters 1988. Cf. the philosophical reconstructions in: Rao 1999. Cf. Maturana/Varela 1979. Cf., e.g., Brocker/Nau 1997 and Steinmann/Scherer 1998. Cf., e.g., Lyotard 1983, MacIntyre 21985 and Höffe 1987.

Self and Other

191

We seem to be caught in an age-old dispute occurring in our own Western tradition with its roots in antiquity on how to define the relation between individual and society. What is the difference between juxtaposing an individual human being with his or her membership in the species man on the one hand, and with his or her being a part of mankind on the other hand? Does it make a difference when in the first case an individual is understood to be endowed with the faculty to act and speak as a representative of any other individual, that is, to act and speak ‘transsubjectively’ – this is the precise meaning of the traditional notion of ‘having reason’ or ‘to be able to reason’ (man as an animal rationale) –, and when in the second case an individual is taken to be a particular contribution, by his or her share in verbal and non-verbal activity, to the subsistence of mankind as a whole – this may be taken to be the meaning of the traditional notion of ‘being social’ (man as an animal sociale)? Of course, in both cases, when acting as a universal representative and when participating in mankind's subsistence, actual activities will fulfill their aims only with respect to particular groups, though in the first case this limitation is judged negatively, because real objectivity is said to have not yet been achieved, whereas in the second case the limitation is judged positively, because participation is said to work only within identified groups to which an individual consciously belongs. In the first case we are faced with the opposition ‘individual-universal’, in the second case with the opposition ‘individual-social’; in both cases, however, it is obvious that competitive behavior is judged to be inferior to cooperative behavior. In fact, both these cases can be treated as consequences, respectively, of two basic traditional assumptions which contradict each other. The first assumption amounts to the simple claim that man is by nature bad. It acts as a presupposition for progress-theories of the cultural process, whence the negative judgment on any limitation with respect to acting as a universal representative. The second assumption may be abbreviated to the opposite claim that man is by nature good. This is the basic presupposition of decline-theories of the cultural process, whence the positive judgment on the limitations of participating in mankind’s subsistence. Entering into a debate on the acceptability of either presupposition would enforce an engagement in moral preoccupations and convictions, an endeavour of uncertain outcome. Instead, looking at basic human needs, appears to be a more promising approach. One who tries to reach a better understanding of how human beings relate to each other, will have to face two incompatible basic needs, the one to be close to one another and the other to keep distance from each other. In order to elucidate this phenomenon I propose to take a closer look at the work of an author who discusses the content and scope as well as the

192

Methods in Philosophy, in Art, and in Science

mutual relations and presuppositions of the two competing theories of the cultural process. He does this with an explicit awareness that both the predicates ‘good’ and ‘bad’ and their theoretical analogues ‘true’ and ‘false’ should not be used descriptively but reflexively. They do not refer to properties of empirical sentences about given objects, but act as reflexive terms for passing ‘philosophical’ judgments while reconstructing, i.e., delineating conceptually, the objects in question. The author I am referring to is Johann Gottfried Herder. He seems to have been the first in history to have become aware of the conceptual rather than empirical relation which holds between the two characterizations of human beings occurring in Protagoras’ myth.11 Human beings are, on the one hand, deficient beings insofar as they lack sufficient protection against inanimate nature and do not possess effective weapons against animals, and, on the other hand, this means that they are proficient beings insofar as they have both poietical and practical abilities which are acquired consecutively by the two steps in the myth. The terms Herder uses to refer to poietical and practical abilities, respectively, are ‘freedom’ and ‘reason’. You ‘possess’ freedom which will eventually yield peace, and you ‘possess’ reason which will eventually yield justice. Both have to be acquired and should not be treated as being ‘just there’. It is by exercising these abilities that humans define their deficiencies, and, therefore, it is wrong to treat the relation between proficiency and deficiency as an empirical one of compensation. The point becomes still clearer, if we turn to the details of the interdependence between being proficient and being deficient. Herder defines the cultural process neither by progress nor by decline but by a process of education. He uses a dialogue model of teaching and learning where teaching includes constructions of representations, and learning, correspondingly, includes deconstructions by means of interpretations up to concrete activity. The dialogue model is used as a means to identify items of a cultural process inasmuch as both doing and suffering – here, the two roles of Self and Other appear in their basic setting as the two roles of a teaching and learning process – always occur together. And it becomes clear that the categories of doing and suffering govern not only the realm of verbal activity but also the realm of nonverbal activity. Now, as every individual human being plays both roles ____________ 11 Herder had elaborated this in his Ideen zur Geschichte der Menschheit. Erster Teil (1784), Zweiter Teil (1785), as well as in his Abhandlung über den Ursprung der Sprache (1772); additional discussions are contained in his Verstand und Erfahrung, Vernunft und Sprache. Eine Metakritik zur Kritik der reinen Vernunft (1799).

Self and Other

193

empirically – he or she is learning as well as teaching, possibly at different moments in his or her lifetime – it is individual distinction together with social coherence, that is, competitiveness against a common background and cooperativeness split up into different approaches, which defines the process of education. Reason, being basically the ability to organize which includes being able to handle problems of representation, and freedom, being basically the ability to produce and, therefore, to tackle problems of presentation, appear, when exercised, on the side of ‘doing’ as proficiency and on the side of ‘suffering’ as deficiency. On the basis of this, Herder is able to explain why tradition may include errors – a deficiency of reason – and why something evil may occur among the items somebody chooses – a deficiency of freedom. The difference of such an account to an understanding of competition merely as fixing and pursuing individual interests and of cooperation merely as establishing and pursuing common interests is obvious. Whoever tries to understand sociality and rationality in terms of individual and common interests, respectively, has turned the reflexive use of sociality and rationality, which up to Aristotle was the outcome of a self-characterization of humans, into the ordinary positive use of describing pre-existing properties of humans. It was Hannah Arendt who first clearly demonstrated how, by translating the Greek term ‘πολιτικόν’ into the Latin term ‘sociale’, the change in reading the meaning of sociality positively instead of reflexively has led to a conceptual confusion throughout our tradition up until now.12 But, it was not this fact alone which had such an effect. In addition to it, one has to take into account another important feature due to the Stoic replacement of πολιτικόν’ by ‘κοινόν’: sociality was no longer restricted to the second Protagorean level of reason-guided ways of living but is understood as pervading the first level of ‘natural’ abilities, too. It is this accidental insight brought about by a conceptual confusion during translatio studii together with the complementary recognition that rationality – in its full sense and not in the confined sense of means-ends rationality – is required for exercising first-level poietical activities which made Herder confident of being right in reestablishing the reflexive use of sociality and rationality, but now throughout the whole range of human activities. Hence, ‘sociality’ – Herder’s ‘freedom’ – does not refer to the struggle among individual interests whether understood in the intentional setting of a cultural ‘struggle for power’ or confined to explanations by a causal theory of behavior to account for the ‘natural’ Darwinian ‘struggle for ____________ 12 Cf. Arendt 1958, chap. 4.

194

Methods in Philosophy, in Art, and in Science

survival’, but it refers to social coherence both in acting and speaking which is cooperation by means of individual contributions. Realizing that everyone must invent his or her own way of life and his or her own world view provides for the possibility of gradual development of sociality which may also be called solidarity in case one wants to use a term more akin to the Greek term ‘φιλία’ (= friendship) which had served the same purpose in antiquity as mentioned above. Likewise, ‘rationality’ – Herder’s ‘reason’ – no longer refers to the ability to act along common interests even if ‘common’ is not restricted to some group interests but refers to fullfledged moral generality. Instead, it signifies individual distinction at all levels which is competition grounded in a community of acting and speaking. In fact, it is a paradigm case of rational behavior to compete by argumentation. By successful refutation, for example, you earn a reputation in the scientific community. Realizing how one’s own way of living and one’s own world view is found amidst shared activities is an accomplishment of another step on the way to individuality. Traditionally in this context, from antiquity through Kant up to the present, we are accustomed to speak of self-determination as the task of reason; but, as reason is said to be concerned with the universal and not with the particular, self-determination is usually understood as deliberate submission to universal laws and, therefore, as an obligation to create the universal human being – the transcendental Ego in Kant – and not as a universal task to be carried out individually. It is not difficult to understand why such confusions have occurred. They are an outcome of neglecting the insight that individuality and sociality are but stages in a process which is simultaneously a process of individuation and of socialization. This process is the one which had been conceptualized by Herder as the education process (in German: Bildungsprozeß). Self and Other are bound together and constitute what may be called a ‘dialogical dyad’, an entity well known for quite some time in psychoanalysis, for example.13 An account of self-determination which starts with a set of ready-made individuals has to postulate preferences and beliefs as additional entities to be possessed by individuals, because otherwise there is no chance either to determine or to explain any of the different competitive and cooperative relations among individual human beings let alone those of deliberation, negotiation, or arbitration or the like – yet, the burden of proving the ____________ 13 Cf. among others the studies on the ‘Mutter-Kind-Dyade’ within primary socialization in: Lorenzer 1974.

Self and Other

195

existence of those additional entities above and beyond my own preferences and beliefs is shirked. Within an approach where the existence of individuals is not traced back to a mutually dependent genesis in a process that is simultaneously one of individuation and socialization, there are only external – natural or cultural – relations between individuals, i.e., relations of the exertion of influence upon one another. This violates the basic condition connected with the concept of self-determination or autonomy of an individual. Exerting influence upon one another results in heteronomy. The chance of learning from each other and in this sense of being engaged in a process of mutual, and therefore, truly autonomous education which constitutes the realm of internal relations among individuals does not even appear. Of course, the process of education functions as an alternative to the process of exerting influence upon one another only if it is not understood in the descriptive sense which is current at present. Neither education in the intentional framework of given educational aims, nor education in the causal framework of social engineering will be mutual education. Mutual education is a process of self-education in which both sides in the process of teaching and learning change their ways of life and their world views by a further step of both individuation and socialization. Self-education concerns Self with respect to Other and Other with respect to Self depending on whether selfeducation is seen under the active aspect of doing or the passive aspect of suffering; it is not an individualistic notion but a dialogical one. Ways of life and world views can be apprehended only if seen in both their individual and in their social aspects; as mentioned above, they show individual distinction and social coherence. Self-education is not striving for a balance between guiding and letting grow, the educational equivalents to the characterizing activities of the two theories of the cultural process discussed further above: striving for universal culture as the second nature of man and obeying the natural evolution of cultural features within the first and only nature of man. Self-education which, using a formula that does not separate its active and passive aspects, is mutual education of Self and Other, and by that very feature an attempt to balance the relation of Self and Other with respect to acting and being acted upon, consists in setting up limits against being influenced. Setting up such limits proceeds both by standing up against submission to cultural conditioning and by inventing ways of compensation for natural dependencies. In terms of an educational process it is possible to say that being guided is countered by individual acting in making use of individual knowledge or know-how already acquired, whereas growing is countered by invoking social knowledge which is an already acquired knowledge of social norms, i.e., of activity to be shared.

196

Methods in Philosophy, in Art, and in Science

Learning from one another establishes bonds as well as demarcations, and it exhibits such a kind of relation even with respect to two empirical persons each individually: the Other – a not yet known or alien part – within oneself, and the Self – a well known part of oneself – appearing opposite to oneself. It is exactly this kind of radical extension of the concept of SelfOther-dependency to include Self-Other-relations within one empirical person that lends itself to perceptual representation as showing both selfexpression for Self and for Other and other-description, again for Self and for Other. In such a way the self-characterization of human beings as belonging to both nature and culture is fully validated. It, even, becomes clear that the very distinction of nature and culture rests on the process of self-discovery which we have identified as a process of mutual education of the dialogical dyad: Self and Other. Individuality, a differentia of individuals on the level of reflection, can be recognized only within some common activity; sociality, an equality of individuals on the level of reflection, can be performed only if we are conscious of the different approaches within a given situation. Mutual dependency of Self and Other is by far stronger than it is often thought to be. It is a consequence of the dialogical polarity that is exhibited by the two sides of man’s self-discovery: human self-understanding in knowing that we belong to nature – a case of suffering – and human self-determination by acting including sign-acting like speaking, dancing or painting and the like which defines us as cultural beings – a case of doing.

Self and Other

197

References Arendt, Hannah, 1958: The Human Condition, Chicago: The University of Chicago Press. Brocker, Manfred/Nau, Heino Heinrich, eds., 1997: Ethnozentrismus. Möglichkeiten und Grenzen des interkulturellen Dialogs, Darmstadt: Wissenschaftliche Buchgesellschaft. Höffe, Otfried, 1987: Politische Gerechtigkeit. Grundlegung einer kritischen Philosophie von Recht und Staat, Frankfurt am Main: Suhrkamp. Huntington, Samuel P., 1996: The Clash of Civilizations, New York: Simon & Schuster Lorenz, Kuno, 1990: Einführung in die philosophische Anthropologie, Darmstadt: Wissenschaftliche Buchgesellschaft. Lorenz, Kuno, 1994: Competition and Cooperation: Are They Antagonistic or Complementary? in: Sociogenesis Reexamined, ed. by Willibrord De Graaf and Robert Maier, New York et alii: Springer, pp. 19–25. Lorenz, Kuno, 2009: Dialogischer Konstruktivismus, Berlin/New York: Walter de Gruyter. Lorenzer, Alfred, 1974: Die Wahrheit der psychoanalytischen Erkenntnis. Ein historisch-materialistischer Entwurf, Frankfurt am Main: Suhrkamp. Lyotard, Jean-François, 1983: Le différend, Paris: Minuit. MacIntyre, Alasdair, 32007: After Virtue. A Study in Moral Theory [1981], Notre Dame Ind.: The University of Notre Dame Press. Maturana, Humberto R./Varela, Francisco J., 1980: Autopoiesis and Cognition: Dordrecht: Reidel. Rao, B. Narahari, 1999: Culture as Learnables. An Outline for a Research on the Inherited Traditions, Saarbrücken: Universität des Saarlandes. Steinmann, Horst/Scherer, Andreas, eds., 1998: Zwischen Universalismus und Relativismus, Frankfurt am Main: Suhrkamp. Wolters, Gereon, 1988: Evolutionäre Erkenntnistheorie – eine Polemik, in: Vierteljahrsschrift der Naturforschenden Gesellschaft in Zürich 133, pp. 125–142.

On the Concept of Symmetry The concept of symmetry is almost omnipresent. Its descriptive as well as explanatory power in a variety of sciences is well known, and this includes facts concerning the relevance of observing asymmetries. Likewise, there is widespread acquaintance with the importance of symmetric patterns as well as with violating symmetries in the arts. Here, now, I want to direct your attention to the logic of being symmetric and away from the wealth of empirical issues, that is, to symmetry as a conceptual tool and not to symmetries as objects of investigation. We may start off with a reminder that it is somewhat rash to speak of just one concept of symmetry. A short look back at history, at the use of the original terms σύµµετρος and συµµετρία by our Greek ancestors, reveals a meaning of these terms that is different from the modern logical notion of symmetry, i.e., from symmetry being a property of a binary (or manyplaced) relation: A binary relation is symmetric, if and only if it coincides with its converse, like ‘being identical with’ or ‘being similar to’, but unlike ‘being heavier than’ or ‘being colder than’. The relation ‘being the mirrorimage of’ between geometric figures counts as the paradigm of a symmetric relation. Different from this understanding we observe the original meaning of the Greek terms being rather close to their etymology: ‘with measure’. Quite succinctly so in the first definition of the tenth Book of Euclid’s Elements, where (any two) magnitudes – straight lines or areas – are called symmetric (with each other) if and only if they have a ‘common measure’, i.e., are commensurable with each other. As a consequence, any magnitude that is commensurable with an assigned one – chosen as a unit, we would say – is called rational, or ‘expressible’ (ῥητόν) in Euklid’s terminology.1 The meaning of σύµµετρος in general, beyond mathematics, and at the same time apparently more specific than just ‘rational ratios’ – I am referring to a meaning that is more or less covered by the corresponding terms in our modern languages, too – may be conveyed by the phrase ‘being harmonious’ as it is applied to a well-balanced relation of parts and whole of some entity that is treated as a whole out of parts. In case such a harmony obtains, according to a widespread, but not uncontroversial conviction, the entity is ____________ 1

Cf. Heath 1956, vol III, pp. 10–12.

On the Concept of Symmetry

199

called beautiful. Beauty, then, a display of some outstanding rationality? Or, rather, a balanced interplay of rationality with irrationality, a harmony of logically higher order? I may remind you of a quotation from Dagobert Frey, contained in Hermann Weyl’s book on symmetry: “Symmetry signifies rest and binding, asymmetry motion and loosening, the one order and law, the other arbitrariness and accident, the one formal rigidity and constraint, the other life, play, and freedom”.2 It is well known how the golden section – sectio aurea or divina proportio – has been held in high esteem as such a sign of balance from its very discovery by the Pythagoreans – with respect to their magnitudes the side of a regular pentagon turns out to be the mean proportional of ‘the whole’ diagonal and ‘the part’ that you get by subtracting the side from the diagonal – and, later on, in Renaissance, as testifying to the aesthetic beauty of objects internally structured by golden sections. Of course, such a treatment of beauty – apollonian rather than dionysian – was challenged in various ways in antiquity, already. I cannot deal here with the ancient discussions, though they would be illuminating, especially in view of the curious fact that there is no Latin equivalent to the Greek συµµετρία, a fact that had been noted by ancient authors, already. Well known is what Pliny The Elder wrote in his Natural History of AD 77: non habet Latinum nomen symmetria.3 Instead, I want to stress something that has escaped attention until the logical structure of language became better understood, and that will lead us to an appreciation of how the ancient and the modern notion of symmetry relate to each other. Nowadays we know that predicative expressions, and ‘symmetric’ is one of them, may sometimes occur meaningfully both in the language about objects and in the language about our verbal tools. Furthermore, in ordinary language they often play a descriptive and an evaluative role, where evaluations involve both logical layers of language. Take as a simple example the colour-term ‘red’. We basically apply it when seeing something red or when looking for something red. In the first case it is said that something red is given, out there in the world around us, in the second case it is said that something red has to be produced in either of two ways: it is out there but elsewhere, not in the spatio-temporal presence, and will have to be made present, possibly just by waiting, or it might be elsewhere or even nowhere and is, disregarding prior existence, brought into existence anew or even for the first time. In the latter case we encounter the well known difficulty of identifying something as being red without knowing in ____________ 2 3

Weyl 1952, p. 16. Plinius 1938–1993, bk 34, 65.

200

Methods in Philosophy, in Art, and in Science

advance what it is to be red, hence, of being dependent on some other characterization, e.g., by a wave-length section of the emitted or reflected light if it reaches the retina unchanged and, after proper processing, is registered by the visual centre of the brain. Whether we state that something is red or expect that something will be red, the term ‘red’ is used descriptively, though embedded in two different speech act[-type]s, and quite often no further analysis is offered. It is neglected that in addition to the descriptive use of a term in an utterance, we normally encounter an evaluation, too. Besides stating that something is red, the statement is evaluated as being true; besides expecting that something will be red, the expectation is evaluated as being warranted. Being-true or being-warranted are not just predicative expressions within the metalanguage only, they are the outcome of judging what has been said by thematizing the difference between a claim and its satisfaction, and, hence, they are concerned with the peculiar bond between object level and language level, world and language. In ordinary language a proposition and a judgment are rarely differentiated, the same words perform both services. Their difference often remains unconscious. This, besides historical reasons like the fight against psychologism in logic some hundred years ago, has certainly contributed to the fact that even in modern philosophy of logic and of language, both terms, ‘proposition’ and ‘judgment’, are treated as referring essentially to the same type of entities, ‘proposition’ logically and ‘judgment’ psychologically. It is scientific language, where we may rely on having the use of ordinary language simpliciter distinguished from the reflective use of language. Scientifically, besides saying something we also account for why it has been said, for what reasons. Giving reasons signalizes the presence of judgments and not of just statements, expectations, wishes, and so on. In science – and as well in art, though in art the notion of langage includes the realm of arbitrary sign-languages – it is essential to become aware of language on the level of reflection as a distinct mode of speech beyond object language and metalanguage, yet without forgetting that in ordinary language we, regularly, are confronted with an amalgamation of all three levels. Reflective use of language as something discernible within the use of ordinary language was the decisive discovery in Greek antiquity, a discovery that may as well be called an invention, because it does not really exist prior to its explicit identification. Reflection marks the advent of philosophy in its original sense where philosophy is understood as tantamount to science. It was unfortunate and has had unhappy consequences in Western intellectual history that the arts were not included in such an understanding of philosophy, though this restriction – and this has to be

On the Concept of Symmetry

201

added in order not to fall prey to deplorable generalizations – remained a much disputed topic even in antiquity. As reason for the schism between the sciences and the arts – the still lifely opposition between the natural sciences and the cultural sciences is a somewhat weaker afterimage of it – may count the widespread Greek conviction that only of things that do not owe their existence to human activity, i.e., only of nature and not of culture, real knowledge can be gained. From renaissance onwards such a conviction lost currency, and it was Kant who radically concluded that with the case of real knowledge it is the other way round: “die Vernunft sieht nur das ein, was sie selbst nach ihrem Entwurfe hervorbringt”.4 Reason, according to his simile, is a judge who puts ‘conceptually devised’ experimental questions to nature and judges her answers. In present-day post-Kantian science such a radical ‘culturalization’ of nature is usually considered to be unconvincing. We, rather, are confronted now with powerful moves towards a ‘naturalization’ of culture, using the neurosciences as a trojan horse, if I may say so. It should be uncontroversial, I think, that either way to look at knowledge is far too simple. We all ‘know’ that our knowledge, whether of nature or of culture, whether perceptual as in art or conceptual as in science, is essentially bound to human activity, be it on the object level with all kinds of successively invented technical tools, or be it on the language level with our theory-building, foremost by the development of mathematical tools. And, after the discovery of reflection – tantamount to its invention, as I said –, we can never get rid of the inherent duality between objects (of procedures) and procedures (of making objects available), a duality that, for the first time, was observed by the French mathematicien Jean Cavaillès as one of objet and opération, and, hence, as a case of symmetry both in the old sense of being well-balanced and in the modern sense of being a mirror-symmetry. We owe the realization of the importance of Cavaillès’ discovery to GillesGaston Granger who took great pains to develop it further and to follow up on some of its various ramifications.5 Not yet treated, to my knowledge at least, as a conspicuous case of such a duality is the relation between the objectival fermions and the force-mediating, hence, operational bosons in the discussions about super-symmetry in theoretical physics; for clarification I should add that, when referring to bosons, they are, as usual in mathematics, operations turned into objects, though, properly speaking, into objects of logically second order. Logic and mathematics are the prominent areas where, within the context of the duality of object and procedure, the interplay of something made ____________ 4 5

KrV, B XIII. Cf. Granger 1983.

202

Methods in Philosophy, in Art, and in Science

and something found can be studied most effectively, provided the role of the linguistic tools is properly taken into account. So, returning to the notions of symmetry under discussion: What is the difference between looking at symmetry, on the one hand, as a first or second order property of certain entities such that among descriptions of these entities ‘being symmetric’ may occur, and, on the other hand, as the result of an evaluation of something said or done such that ‘being symmetric’ will be close in meaning to ‘well-balanced’ or ‘located [metaphorically] in the centre of gravity’? To find an answer to this question it is useful to return to antiquity again, to an author who puts balancing, in fact the balancing of extremes that ‘fight with each other’, in the centre of his reflections on rationality or the λόγος. It was Heraklit, who uses the construction of mean proportionals to show how rationality works, such that a mean proportional may even serve to characterize rationality. Heraklit states: “Compared with god, the wisest/most knowledgable [σοφώτατος] of men will appear an ape, in wisdom/knowledge [σοφίη], beauty and all else”.6 Instead of treating this and related fragments as a sign that Heraklit had a “religious sense of the worthlessness of human knowledge in comparison with divine”,7 it should be read as an explicit presentation of his method of rational thinking and, thus, of the peculiar worth he ascribes to human knowledge. In the cited fragment B 83 man is the mean proportional between god and ape with respect to the binary predicate ‘wiser/more rational than’. Heraklit does not simply want to compare two entities like god and man: both are wise, but god more so than the wisest of men. He wants to measure the difference in wisdom/knowledge between the two, yet without having a unit of wisdom at his disposal that is public and not private. His solution: a definition of equality beween two differences (presented multiplicatively as two ratios: god to man and man to ape with respect to ‘wiser than’), thus – and that is quite modern, in fact an anticipation of definition by abstraction – avoiding the definition of a necessarily arbitrary unit of wisdom. His gain: man as the mean proportional between god and ape unites contraries. Man is at the same time wise/rational (with respect to ape) and unwise/unrational (with respect to god). And Heraklit suggests proceeding in the same way with respect to any other pair of polar-contraries, i.e., contrary predicates that belong to scales of comparison, like ‘beautiful-ugly’ to the comparison ‘more beautiful than’. He identifies man as the central point of reference with respect to a world that is governed by extremes that fight with each other, by the ‚ever-living fire’ in other words of Heraklit.8 The measuring rod with which to grasp nature ____________ 6 7 8

DK, B 83. Guthrie 1967, p. 413. Cf. DK, B 30.

On the Concept of Symmetry

203

(φύσις) is not out there to be found – φύσις κρύπτεσθαι φιλεῖ,9 Heraklit says [= ‘nature loves concealment’; in the excellent reconstruction by Geoffrey Kirk rendered into English as:10 “The real constitution of things is accustomed to hide”] –, but it is the λόγος as exhibited ambiguously by the rational activity of man, an activity that has been identified as a specific way to relate man to nature. With the tool of mean proportionals in the way just shown Heraklit may indeed feel to be justified in pleading for his thesis of a coincidence of opposites against Pythagoras. Contrary forces, if brought into balance by some mean proportional between them should not be treated as having disappeared by mutual neutralization. He denounces Pythagoras as a cheater and mere polymath. To search for simple, symmetric and harmonious, i.e., ‘rational’, proportions as the basis of cosmic order, i.e., of the κόσµος (the term κόσµος implies being well-ordered, already), as the Pythagoreans do, is due – so we may infer – to a bias towards mathematical theories, to what is limited by applying the limit (πέρας) to the unlimited (ἄπειρον). It neglects to make judgments on how theories relate to the ever-changing though unitary world around us. Aristotle develops a far more advanced argumentation along these lines in his critique of Plato’s pythagorean inclinations. Yet, his contempt for Heraklit's apparent disregard of logical laws we may now identify as being based on a misunderstanding. In B 83 Heraklit exercises rationality and at the same time characterizes it as something balanced: There is symmetry between the relation ‘more rational than’ and its converse ‘less rational than’, in fact the symmetry of the metarelation R(ρ,ρ̃) between a relation ρ and its converse ρ̃. We have thus found an instance where the gap between the ancient and the modern meaning of being symmetric is in principle closed. Besides, we have learned that exercising rationality is not searching for a simple comparison of two objects by establishing a (rational) proportion between them, but an attempt to relate such proportions with each other, that is to look whether they are ‘analogous’, if I may use Euclid’s terminology for an equality a:b = c:d. I should add that to speak of a rational proportion would originally be a pleonasm, since proportion – in Greek: λόγος – used to imply being rational (irrational proportions have been called ἄλογοι), but after the discovery of irrational proportions in geometry, most probably at first by the Pythagoreans on occasion of comparing the magnitudes of side and diagonal in a regular pentagon, the terminology was changed. In Euclid’s Elements irrational proportions that being ____________ 9 DK, B 123. 10 Kirk 1954, p. 227.

204

Methods in Philosophy, in Art, and in Science

squared become rational were introduced to be an additional species of δυνάµει, i.e., potentially, rational ones. Modern terminology with respect to exponentiation still keeps these ancient roots. We are prepared now to take this example of Heraklit as a case of how generally to relate first and second order use of the predicate ‘symmetric’: A set A of objects, structured by operations and relations, is symmetrically structured with respect to some special structure S (for simplicity’s sake I assume S to be just a binary relation), if and only if there is an idempotent one-to-one mapping T of A onto itself, i.e., TTx = x for all x∈A, that is compatible with S, i.e., S(x,y) implies S(Tx,Ty); in case S is defined by a function f such that S(x,y) fx = y , this is equivalent with the commutability of T and f [fTx = Tfx being equivalent with S(x,y)≺S(Tx,Ty)]. From the idempotency of T it follows that the associated relation R(x,y) Tx = y is a symmetric relation in the modern logical sense, i.e., R(x,y)≺R(y,x). Under the conditions mentioned, first order symmetry of x and y is based on the existence of a second order symmetric relation R between x and y. We come back to the example of Heraklit, if we start with the set of all binary relations where structure S is just the relation of identity. In this case conversion establishes a simple symmetry among binary relations. Another well-known example is the set of all logically compound formulae in systems of classical logic with relation S being logical equivalence between them; then, mapping T that transforms any formula into its dual [structural changes!], fulfills all the requirements that had been set. Hence, duality in two-valued logic is a case of symmetry, a symmetry of boolean lattices in the language of lattice theory. The cases of duality in geometry are covered in the same way by taking as initial set A the set of sentences, e.g., of an axiomatized two-dimensional Euklidean geometry, and as structure S again just logical equivalence between such sentences. In order to arrive at ordinary symmetries of plane geometric figures as, e.g., given by relations S in a two-dimensional Euklidean space A, any mapping T compatible with S that produces mirror-images satisfies the necessary requirements: S and TS are symmetric with each other; the natural extension of mirroring to automorphisms that are not idempotent leads to a generalization of the notion of symmetry, unless, of course, non-idempotent elements can be expressed by iteration of idempotent ones (as is the case, e.g., with rotation-symmetry in two-dimensional Euklidean geometry). Finally, I suggest to review the examples covered and look what precisely has happened when first order symmetry was based on second order symmetry. Symmetries are structural properties of objects and differ from ordinary ones in being dependent on already available predicates about the objects. You do not simply observe symmetries in nature, you, rather, use the logical concept of symmetry being one of invariance under commuta-

On the Concept of Symmetry

205

tion of certain means of reference, be they characterizing predicates or spatio-temporal coordinates, in order to measure the difference of observed relations with ideal ones. Symmetry is a feature of theories such that asymmetries, be they produced as by artists or observed as by scientists, are not just events to be registered but the result of a judgment – displayed by deeds or by words – how our picture of the world fits or is at odds with the world of culture and of nature around us, i.e., with the world we actually live in by taking part in a process of uninterrupted interference. To look for symmetries, or, more accurate: to judge deviations from symmetries, being the most conspicuous case of regularities beyond equalities that account for mere classifications, is a peculiar way to look for reasons. I may just remind you once again of the almost mythical kinship between beauty and truth that we come across quite often in the work of leading mathematicians. In antiquity it was Harmonia, daughter of Ares and Aphrodite, who represented order and balance of the universe. With judging the fit of what we suffer by being exposed to objects with what we do in order to get acquainted with those very objects, we have come back to the very general duality of object and procedure – objet et opération – I referred to above, and which Nelson Goodman has identified as mutual dependency of matter and manner: We choose the facts as much as the frameworks, he says,11 though this statement should better be split into two complementary, that is, again ‘symmetric’ ones: We produce the facts as much as the frameworks and we experience the frameworks as much as the facts. If one pursues this line of reasoning still a bit further which cannot be done here, we end up with a dialogical polarity inherent in any human activity: In the process of acquiring an action competence there are two roles, the I-role of performing the action in producing an action token, i.e., the role of ‘agent’, and the You-role of cognizing the action in witnessing an action type, i.e., the role of ‘patient’. The dialogical polarity amounts to two different approaches, i.e., ‘manners’, to an action, a practical and a theoretical one. Transformed into ‘matter’, we speak of ‘empirical’ action-tokens and ‘rational’ action-types. And it is this derived polarity that can be shown to exist among arbitrary objects, which, after a number of additional steps reminiscent of Leibnizian constructions,12 makes it possible to speak of the duality of body and mind for humans and other living beings. If we switch back from the duality among objects – empirical tokens and rational types – to the duality of procedures – practical performance and theoretical cognition –, it is not difficult to ____________ 11 Cf. Goodman 1978, p. 22, together with the essay The fabrication of facts in: op. cit., pp. 91–107. 12 Cf. for further details: Lorenz 2005; in this volume pp. 42–55.

206

Methods in Philosophy, in Art, and in Science

arrive at the following conclusion: Humans become persons by mutual recognition that comes about by a process of developing the ability to play both dialogical roles at the same time. It is a process of both individuation and socialisation that includes a growing consciousness that any one of us is bound to continually losing and recovering the balance between the two dialogical roles. Symmetries are states of balance that exist ‘in theory’ only, ‘in practice’ they are norms vis-à-vis broken symmetries. And it is they that keep things going, whether in nature or in culture.

References DK = Diels, Hermann/Kranz, Walther, 1996: Die Fragmente der Vorsokratiker. Griechisch und Deutsch [1903], I–III [reprint of 61951/52], Hildesheim/Zürich: Weidmann. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks Sussex: The Harvester Press. Granger, Gilles-Gaston, 1983: Formal Thought and the Sciences of Man, Dordrecht: Reidel Publishing Company. Guthrie, William K. C., 1962: A History of Greek Philosophy I (The Early Presocratics and the Pythagoreans), Cambridge: University Press. Heath, Thomas L., ed., 1956: The Thirteen Books of Euklid’s Elements [21926], New York: Dover Publications. Kirk, Geoffrey S., 1954: Heraclitus: the Cosmic Fragments, Cambridge: University Press. KrV = Kant, Immanuel: Kritik der reinen Vernunft, ed. by Raymund Schmidt, Hamburg: Felix Meiner Verlag 1956 [Riga: Johann F. Hartknoch 1781, 21787]. Lorenz, Kuno, 2005: Pragmatic and Semiotic Prerequisites for Predication, in: Logic, Thought and Action, ed. by Daniel Vanderveken, Dordrecht: Springer, pp 343-357. Gaius Plinius Secundus, 1938–1963: Natural History/Naturalis historia, lat.-engl., 37 books in 10 volumes [Loeb Classical Library], ed. and transl. by Harris Rackham, William H. S. Stones and David E. Eichholz, Cambridge Mass.: Harvard University Press. Weyl, Hermann, 1952: Symmetry, Princeton N.J.: University Press.

Procedural Principles of the Erlangen School On the interrelation between the principles of method, of dialogue, and of reason When scrutinizing the brand of constructivism as developped by the Erlangen School with their founding fathers Paul Lorenzen and Wilhelm Kamlah we are confronted with two different goals quite alike to those that resulted in the bifurcation of Analytical Philosophy into the Logical Empiricism of the Vienna Circle and the Linguistic Phenomenalism of the Oxford School. And Ludwig Wittgenstein may justly be called the godfather to all of them. There, on the way from Bertrand Russell to Rudolf Carnap, we encounter an attempt to reconstruct wherever possible the conceptual frame of mathematics and the sciences in a logically sound manner by using ordinary language as a starting point, and, on the way from George Edward Moore to Gilbert Ryle, we experience an attempt to reinterpret wherever possible the conceptual content of traditional philosophy by means of the same ordinary language. The goal to construct a language of science – or rather its logical structure, an ‘ideal’ language – on the basis of a well functioning area of language, i.e., ordinary language, by using the tool of formal languages contrasts with the goal to reduce the language of philosophy to ordinary language. The latter goal of bringing words back ‘from their metaphysical to their everyday use’, as Wittgenstein once stated,1 is better understood as the goal to lay bare the hidden ‘implication threads’ between concepts,2 especially of those that derive from traditional philosophy when used in ordinary language argumentations. With other words, a non-formal logic of language should be exhibited that at present, in an explicit pragmatic fashion, has found an elaborate version in Robert Brandom’s inferentialism.3 For both branches of Analytical Philosophy logical analysis of language as extracted from Gottlob Freges foundation of modern formal logic was the primary tool. This is true for the constructivism of the Erlangen School, as well. With respect to the general recognition it had gained from ____________ 1 2 3

PU , § 116. Ryle 1971, p. 443; cf. Rao 1994, chap. 5. Cf. Brandom 1994; 2000.

208

Methods in Philosophy, in Art, and in Science

the philosophical public, even the peculiar predominance of the concern with constructions in Logical Empiricism over the concern in Linguistic Phenomenalism with philosophical tradition and its impact on how common affairs are treated or maltreated, found a repetition. When speaking of the Erlangen School primarily the reconstructions of scientific areas and others beyond science proper, are focussed – the primary concern of Lorenzen – and not the studies on the anthropological background of how it all came about, both historically and systematically, something Kamlah was mainly interested in. Nevertheless, both agreed fully on how and where to start the whole enterprise: amidst and with the use of ordinary language, provided nonstock use of it is excluded. Several times, with approval Lorenzen referred back to a famous simile on how science is built up that Otto Neurath had used at the end of his essay on The Foundations of the Social Sciences for the International Encyclopedia of Unified Science:4 Imagine sailors who, far out at sea, transform the shape of their clumsy vessel from a more circular to a more fishlike one. They make use of some drifting timber, besides the timber of the old structure, to modify the skeleton and the hull of their vessel. But they cannot put the ship in dock in order to start from scratch. During their work they stay on the old structure and deal with heavy gales and thundering waves. In transforming their ship they take care that dangerous leakages do not occur. A new ship grows out of the old one, step by step – and while they are still building, the sailors may already be thinking of a new structure, and they will not always agree with one another. The whole business will go on in a way we cannot even anticipate today. That is our fate.

And Lorenzen added: Even the first ship must have been built while in the beginning men were living on the high seas, just swimming and putting one’s trust into the various elementary skills of those swimming about that the skills will be sufficient to turn solid items floating around into parts of, let’s say, a primitive raft – there is no dry-dock. The ordinary language we happen to use successfully must, for scientific purposes, gradually be substituted by interlinguistic constructions in order really to achieve such purposes. An important example: our number system for doing arithmetic. It is, meanwhile, completely established interlinguistically. As starting point of the various scientific fields, Lorenzen propagated the set-up of an ortho-language quite akin to the outcome of what Willard V.O. Quine had called the ‘regimentation’ of ordinary language.5 And a principle ____________ 4 5

Neurath 1970, p. 47. Cf. Lorenzen/Schwemmer 1973, and Quine 1960, chap. V; cf. , furthermore, Wohlrapp 1975, for a likewise relevant discussion on the relation between the principle of method and a principle of dialogue that is somewhat restricted in scope when compared with the exposition here.

Procedural Principles of the Erlangen School

209

of method must govern any such set-up as without an ordering, semiordering to be precise, that is to be imposed on the various steps leading to a well-functioning language, a guarantee against errors cannot be given – that was the idea. But it was an idea that did not gain currency among all those who affiliated themselves with the Erlangen School. For the simple reason that the goal of a language of science being set up without gaps and vicious circles such that both syntax and semantics can be derived from its pragmatics is not and, most probably, cannot, be determined in such a fashion that possible sequences of steps may in advance be judged as being in accordance with a principle of method or not. Methodical order is different from pragmatical order that gave birth to the idea in the strict sense just stated, because pragmatical order just cannot be violated: you cannot run down from the top of a hill, before having gone up to the top provided you are not yet standing on the top. On the language level, things are not as easy. Why, for example, should it be a violation of methodical order – pragmatically there is no obstacle –, if a proposition about some object may be used to define the object instead of having the existence of an object guaranteed prior to any proposition about it? In this case, obviously, the endless discussion on predicativity is lurking in the shadow. Where, then, is a principle of method applicable and when should it be applied? As a preliminary answer that will serve the present purpose, we may say that for turning a (subject-bound) skill into a (generally available) knowing-how we need a method of teaching and learning skills. An articulation of such a procedure is known as procedural knowledge or practical knowledge (provided normative connotations – knowing how to act under given conditions – do not interfere with the term ‘practical’ in this case). And a very simple fine example is at hand, the skill of counting transferred into the notation of how to generate the positive natural numbers by concatenating units: the arithmetical calculus of strokes. Of course, not the sequences of strokes themselves ‘are’ the natural numbers. In a semiotically purified version of presenting the notation of counting, it is the generation of indices (‘showing’ the positive natural numbers) that is articulated by the arithmetical calculus of strokes. With systems of rules for the construction of all kinds of objects on various logical levels – semiotically these objects have generally to be treated as indices of the ‘real’ objects (in fact, ‘real’ objects as something that we can refer to, depend on activities to make them available, i.e., depend on dealing with them) – the principle of method is identified as a way to arrive at generally available procedural knowledge. How to understand procedures rather than just being able to repeat them, seemed not to be a problem to be dealt with separately, in the early phase of the Erlangen

210

Methods in Philosophy, in Art, and in Science

School, even though the question of what it means to understand what others do or say has always been raised especially by Kamlah. Lorenzen did not recognize this to be a question beyond mere psychological interest, and he felt himself supported by the fact that the principle of method is itself obeying a more general principle of the Erlangen School, a principle that has variously been called ‘the principle of transsubjectivity’ (try to overcome your subjectivity), ‘the principle of reasonable arguing’ (try to argue unprejudiced, unconstrained and unpersuadingly) or, even, ‘the principle of reason’. In its first version as a call for overcoming one’s subjectivity, it seemed explicitly to relegate questions of mere understanding to be something that has to be answered individually before even being able to participate in the constructivist version of ‘a rational reconstruction of scientific knowledge’ – a phrase Rudolf Carnap once used to characterize the goal of Logical Empiricism.6 Usually, the most important objects to be constructed are the true propositions (for simplicity’s sake I delete here a discussion of the principles as governing the way of how to set up practical philosophy according to the tenets of the Erlangen School), and the rules to reach them are rules of argumentation. As these cannot look like the rules for constructing objects, because at least two persons are involved and the steps on the way are not parts of true propositions but arguments pro and con the initial proposition, the term ‘principle of method’ eventually lost its primary position as a general characteristic of the philosophical work in the Erlangen School. Yet, we should not forget that initially it was indeed the insistence of proceeding methodically in the general sense of proceeding according to principles – and not by trial and error, because the concern is with science as a way of presentation and not with science as a way of research –, in order to guarantee that a reader (or listener) is in a position to follow on his own the procedure to reach the presented results (rather than ‘to believe’ them) that served as the self-characterization of the Erlangen School. When the plea to overcome one’s subjectivity assumed logical priority among the guiding principles – and this was signalled by the label ‘principle of reason’ – we should read this ambitious term as a way to make the conviction explicit that binds the Erlangen School to Kant, when he declares that “die Vernunft nur das einsieht, was sie selbst nach ihrem Entwurfe hervorbringt”. The new label is also better suited, because it does not suggest preconceived limits of how to achieve the goal of transsubjectivity, whether rules of construction will be adequate or not. As a matter of fact, the inherent limits of rules of construction were felt, when questions of dealing with moral issues and other topics of practical ____________ 6

Cf. Carnap 21961, p. IX.

Procedural Principles of the Erlangen School

211

philosophy came up, whereas the importance to transcend subjectivity in normative contexts in order to make general participation in moral discourse possible, was strengthened. If following a principle of method is, indeed, to be identified with mediating successfully procedural knowledge, the question had to be answered, whether the command of a language, especially one that is rich enough to do science and philosophy, may properly be understood as being just a case of procedural knowledge, too. Certainly not in the simple sense of an ability to follow rules of construction – we noted that, already. The essential difference between the ability to say something (to someone) and the ability to understand something (that someone has said) including an explication of how to make sure that something said has been understood cannot be dealt with by just rules of construction. Here, we have found the deeper reason why it was important to move from a principle of method to the more general principle of transsubjectivity. And, furthermore, why it is essential not to relegate questions of understanding, e.g., understanding constructions, to mere preconditions of reconstructing scientific knowledge. It makes it clear that any move towards sound science and philosophy will initially be a move by an individual subject, and whether that works, is dependent on how to make such a move accessible to potentially everybody. Here, now, the Erlangen School faces a problem that is basically the same the Logical Empiricists tried to tackle with: the seemingly private starting point. It makes no difference whether we start with individual activities that others might not be able ‘to understand’ or with individual sense data that seem to be inaccessible to others. To postulate a common world of objects (of different categories) and a common way to live within it, the world of everyday life, is certainly not a way out, because we have to deal with the first steps out of such a shared world towards, e.g., constructing a language of science, or, e.g., understanding traditional language. Expressed in a radical fashion: How can I know that my object is also your object? And if indeed I would know that: How is, under whatever circumstances, what I say about some object ‘the same’, i.e., has the same meaning, as what you say about the same object? Don’t we rather live in a virtual reality where all the others are but mere projections in each one’s individual world? The question turned upside down: How is it possible to share individual experiences and thus to create a common world step by step? The answer of the Erlangen school (after having learnt their lessons from Wittgenstein and Peirce): By two persons acquiring a common action competence in a situation of teaching and learning, i.e., through repetition and imitation of performing an action. Of course, this is just a first step, as transsubjectivity is enforced in a two-person-world, only. And this is effected not by transferring an

212

Methods in Philosophy, in Art, and in Science

individual competence from one person to another one, but by using a procedure that incorporates from the very beginning the two dialogical roles inherent in any action, be it a simple action, or, later on, a sign action, e.g., a verbal one: the I-role or role of the agent in executing an action (it means: producing an action token, category of doing) and the You-role or role of the patient in executing an action (it means: witnessing the action type that is instantiated by the action token, category of suffering). It needs further steps to enlarge the common twoperson-world in order to reach an essentially transsubjective world. I should hasten now to introduce an essential distinction connected with the concept of an action that was payed attention to not before the difference of the two dialogical roles of any action – and, hence, the principle of dialogue on an elementary level – was understood as keeping the process of overcoming one’s subjectivity going. It is the distinction of object and procedure – called ‘objet et opération’ by Gilles Granger,7 ‘matter and manner’ by Nelson Goodman8 – that are bound together by duality, when procedures (like functions) are themselves treated as objects of another realm that make objects of the first realm accessible (quite alike the duality of objectival fermions and force-mediating bosons in nuclear physics). Yet, the essential point is to look for the distinction of object and procedures when procedures are not treated as objects of another kind. In fact, it is actions that, different from any other kind of objects, show beside their objectival character also a functional character, where the term ‘functional character’ does not refer to the possibility of treating actions as a means for some end – this happens without erasing the objectival character. It, rather, refers to the fact that while executing an action, i.e., while producing an action token together with witnessing an action-type, neither before nor after, any reference to the action, be it concretely to the token or abstractly to the type, is suspended – and that is meant when saying that the functional character comes to the fore and the objectival character has disappeared. While executing an action, actor and execution are said to be ‘melted’, i.e., in logical terms, indistinguishable. Functionally, an action shows a dialogical polarity. Any execution of an action is an active performance by the actor in the role of agent, i.e., Irole, and it is a passive cognition of the actor in the role of patient, i.e., You-role. I, already, mentioned the old Aristotelian categories of ποιεῖν and πάσχειν that were taken up again as doing and suffering by Dewey.9 The functional polarity of active performance and passive cognition on the ____________ 7 8 9

Cf. Granger 1983, Postface (pp. 181–193). Cf. Goodman 1978, chap. VI (Fabrication of Facts). Dewey 1921, p. 86.

Procedural Principles of the Erlangen School

213

one hand, and the objectival opposition of producing empirical tokens, i.e., individual actions, and witnessing rational types, i.e., generic acts, on the other hand – in the first case we speak of singular actualizations of universal schemata, in the second case of particular instantiations of rational types –, both have carefully to be kept distinct from the duality between object (action as an object) and procedure (action as a procedure or function). Phrased a bit more drastically: tokens and types are objects, actualizations and schemata are procedures that we need to make objects accessible through ways of dealing with them. It is these procedures – a phenomenological reduction of a world of objects that is found, as I have called it – which, if procedurally turned upside down, appear as a dialogical construction of a world of objects that is made. And such a dialogical construction acts as an ‘ideal’ measuring rod for the ‘real’ world we started from. In this sense we are justified to call the two procedures – on the one hand, of actualizing an action of dealing with an object, i.e., an appropriation of the object, and, on the other hand, of schematizing an action of dealing with an object, i.e., a detachment from the object – being a pragmatization through appropriation in taking the I-role, and being a semiotization through detachment in taking the You-role, in a seemingly old-fashioned way to be exercising practical and theoretical rationality, respectively. By now, I hope, it needs no further argument for being convinced that the principle of dialogue – beware of the two different dialogical roles of each and every action! – actually serves to validate the principle of transsubjectivity, and that in this sense, too, it deserves likewise to be called the principle of reason. Reason is dialogical in nature, even beyond actions of argumentation where this thesis has found a broad treatment among adherents of the Erlangen school.10 Besides, if the principle of method governs the mediation of procedural knowledge using rules of construction of objects, something that belongs to the essential achievements of practical rationality, we should likewise identify the essential achievements of theoretical rationality, namely the mediation of propositional knowledge, to be governed by another principle. For, we know, already, that the principle of method pertaining to the role of agent, will not be appropriate to govern the mediation of propositional knowledge as this pertains to the role of patient. The necessity of a principle of conceptualization in charge of guaranteeing a conceptual organization of propositional knowledge in order to make its mediation possible has not yet been envisaged by the Erlangen School, possibly because such steps were thought to have been covered by the principle of transsubjectivity, already. Indeed, mediation of propositional ____________ 10 Recently, e.g., in Wohlrapp 2008.

214

Methods in Philosophy, in Art, and in Science

knowledge proceeds with rules of substitution of signs as the means of conceptual organization – you will understand what I have said, when ‘my’ language can be translated into ‘your’ language. Hence, rules of substitution of signs may properly be called rules of translation, because the term ‘substitution’ covers only the syntactical aspect of a conceptual organization. With rules of translation of signs we complement the rules of construction of objects; both should be clearly distinguished from each other. Now, both a methodically produced knowing-how and a conceptually organized knowing-that are still in need of making them ready to be at one’s disposal, a knowing-how needs stabilization and a knowing-that needs objectivation. The dialogical procedures for having that done are available. By detachment from a knowing-how (using the You-role) this knowing-how is stabilized in terms of a piece of symptomatic perceptual knowledge – someone knows what he/she is able to do and that he/she is likewise able to share it in principle with anybody by means of a process of teaching and learning. And by appropriation of a knowing-that (using the I-role) this knowing-that is objectified in terms of a piece of symbolic procedural knowledge – someone is able to say what he/she knows and is furthermore able to defend what he/she knows against potentially anybody by argumentation. The principle of dialogue in governing the procedures that lead to a transsubjective reconstruction of experience, of both individually acquiring an experience and socially partaking of an experience by articulating the former, guarantees the satisfaction of the principle of transsubjectivity. The two principles, the one of method for the procedures on the level of objects by rules of construction, and the one of conceptualization for the procedures on the level of signs/language by rules of translation, will serve their purpose, if they are treated as subordinate to the principle of dialogue. When rendering the principles of the Erlangen School in the way just explained we can enlarge the initial perspective of treating the Lorenzenbranch of the Erlangen school as an heir of Carnap’s programme – the corresponding relation of the Kamlah-branch of the Erlangen School to Ryle’s intentions is out of focus, then – to a broad perspective of treating Dialogical Philosophy as an outcome of combining the basic strategies of Ludwig Wittgenstein and Martin Buber. On the one hand, I am referring to Wittgenstein’s reduction of sign acting to merely acting that has the consequence of treating witnessing of one’s own being active as a case of turning an action into a verbal action, though this is nothing but a change of the I-role into the You-role on the level of simple actions. If, on the other hand, an arbitrary action is pushed up to the level of sign actions, as done by Buber, doing and suffering lose their distinctive feature from saying (something) and understanding (what has been said). In the later

Procedural Principles of the Erlangen School

215

Wittgenstein we are faced with a reduction of theory to praxis – a case of an execution of the programme of ‘naturalizing’ the mental. In Buber we find a theorization of praxis, and this is a case of the complementary strategy of executing the programme of ‘mentalizing’ the natural. Both procedures, though explicitly dialogical with respect to object and procedure, are obviously one-sided. Dialogical Philosophy in the full sense of developing the dialogical character of reason on the practical level of simple actions and on the theoretical level of sign actions may indeed be understood as an attempt to find a common framework for the strategies of both Wittgenstein and Buber. In my version of explicating the basic intentions of the Erlangen School in a historical perspective, they should be summed up as searching for a combination of the programme of naturalizing language which comes about by radicalizing the naturalized epistemology of Quine, if you disregard for a moment the dialogical set-up, with the programme of mentalizing world which, in fact, is what Cassirer did in his Philosophie der symbolischen Formen; here, too, without having used a dialogical framework explicitly.11 You have just to pay attention to the pragmatic character of signs and to the semiotic character of actions. The pragmatic character of signs comes to the fore by embedding signs in the domain of sign actions, and actions may acquire semiotic functions when used, for example, in situations of learning by doing. If this much is granted, the interplay of detachment and appropriation as the basic forms of theoretical and practical rationality, as explained, already, solves the problem of how naturalization and mentalization may be combined. In fact, naturalizing language appears as pragmatization when we are engaged in the procedure of appropriation, and, correspondingly, we experience mentalizing world as semiotization when we take up the procedure of detachment. Simplifications together with the conceptual troubles they generate when elaborating on the interplay of pragmatization and semiotization, we can evade by carefully attending to the two dialogical roles in both acting and sign acting, something neither Wittgenstein nor Buber could do, because within their respective frameworks they couldn’t but miss the chance of balancing pragmatization by semiotization and vice versa. What happens within the procedure of appropriation is not just a transfer of, e.g., language into verbal activity as the programme of naturalization seems to demand. Appropriation means to take on I-role, and for verbal actions this is a double task. You produce an act of speaking simply as a sequence of sounds/phonemes, and, at the same time, you produce an act of meaningful speaking, i.e., saying (something); for simple actions ap____________ 11 Cf. Quine 1960;1969, and Cassirer 1923–29.

216

Methods in Philosophy, in Art, and in Science

propriation is just doing them. Correspondingly, detachment is not exhausted by the transfer of, e.g., the world of particulars into meaningful actions of dealing with them so as to ‘understand’ the world or to induce understanding – you may think, e.g., of ‘understanding’ a musical instrument as a particular, by the activity of playing it. Rather, detachment means to take on You-role, and, again, for simple actions this is just a case of suffering them, whereas for verbal actions detachment consists in both hearing a sequence of sounds and, at the same time, meaningful listening that is called ‘understanding’. Of course, Dialogical Philosophy is certainly not by itself an interplay of the two procedures appropriation and detachment, it is a representation of it, hence, a detachment of second order logically. It all depends on being able to pay attention to the two dialogical roles and their difference when dealing with all four kinds of activities: addressing somebody and answering somebody, acting and reacting. To react to an action would be impossible, if it were unknown to the reacting person what it was the acting person had done; likewise, an answer would not count as an answer, if not some at least slight knowledge of what the other has said, preceded the answer. Within each activity, be it an ordinary action or a sign action, e.g., a verbal activity, the I-role appears, as we know, already, as an actualization of the action – it may be called the pragmatic side of it – whereas the Yourole appears as a schematization of it that counts as its semiotic side. Any agent, when performing an action, is also in possession of an ‘image’of his/her action (‘he/she sees himself acting’), and, likewise, any speaker who says something, i.e., who is speaking meaningfully, is, when doing so, also in possession of what he/she means (you may say: ‘he/she acts as his own listener’). If we forget about this, we would make it impossible even to articulate the unavoidable difference of understanding between agent and patient. Yet, being fully conscious of it, such differences of understanding can be transformed into processes of learning from each other and this is nothing but a sequence of further dialogues in the general sense of not being restricted to verbal exchanges. Now, application of the principle of reason – the dialogical character of reason turns the principle of reason into an equivalent of the principle of dialogue – to the process of learning from each other yields the transformation of the interplay of appropriation and detachment from its status as a second order detachment into such an interplay in real life, that is, a transformation of dialogical philosophy into the ‘vécu’ or an appropriation of second order. What happens in real life, is mutual self-education, on the one hand, by developing the art of taking over the perspective of the other side, a case of appropriation of the agent-patient-relation, and, on the other hand, by learning to see yourself with the eyes of the other, a case of

Procedural Principles of the Erlangen School

217

detachment from the agent-patient-relation. This includes learning to acquire skills of the other as part of your own ones, and to look at your own skills with the eyes of the other. In this way, both individuation and socialization of both sides will be enhanced. A change of the ways of life and of the world views is not only occurring as done and suffered, but happens consciously. Appropriation of the agent-patient-relation by the agent, the case of self-appropriation, leads to an internalization of the You-role: the agent besides displaying the I-role takes on an I-and-You-role. The real partner of an actor as agent is turned into a virtual one. We are all quite well acquainted with situations of this kind. We, often, know in advance what someone is going to say to what I will be doing, or what he is going to answer to what I will be saying. In scientific discourse it is even one’s duty to answer in advance objections that might conceivably be raised rather than to wait for actual objections. Yet, when the ability to react to a real partner gets lost, we are faced with a real problem: self-appropriation develops into self-sufficientness. Correspondingly, detachment from the agent-patient-relation by the patient, the case of self-detachment, leads to an externalization of the Yourole: the patient besides displaying the You-role takes on an He/She-role. The real partner of an actor is turned into merely a special kind of object. Again, we are well acquainted with this situation, too. For example, when treating men as objects of scientific investigations, both in the natural and in the social sciences. And, again, when the ability to meet someone as a person rather than to see him only in He/She-role, gets lost, we have a problem: selfdetachment develops into self-alienation. Somebody who avails himself/ herself of both I-role and I-and-You-role should be called a subject in the first person, and his or her co-subject in an interaction of any kind has to be considered as somebody who takes on both He/She-role and You-role and may be called a subject in the third person. This applies also to the case, where the co-subject is the same person as the subject. We, then, speak of a subject who knows to be one: he or she is a self-conscious subject. It is not difficult to see, how these last remarks that gave hints on consequences of the principle of dialogue or reason with respect to further differentiations of the two dialogical roles, open the path to practical philosophy as conceived in the Erlangen School. Yet, to enter this area would make another paper necessary.

218

Methods in Philosophy, in Art, and in Science

References Brandom, Robert B., 1994: Making it Explicit. Reasoning, Representing and Discursive Commitment, Cambridge Mass.: Harvard University Press. Brandom, Robert B., 2000: Articulating Reason. An Introduction to Inferentialism, Cambridge Mass.: Harvard University Press. Carnap, Rudolf, 21961: Der logische Aufbau der Welt [Berlin 1928], Hamburg: Felix Meiner. Cassirer, Ernst, 1923–29: Philosophie der symbolischen Formen I–III, Berlin: Bruno Cassirer. Goodman, Nelson, 1978: Ways of Worldmaking, Hassocks Sussex: The Harvester Press. Granger, Gilles-Gaston, 1983: Formal Thought and the Sciences of Man, Dordrecht: Reidel. Lorenzen, Paul/Schwemmer, Oswald, 1973: Konstruktive Logik, Ethik und Wissenschaftstheorie, Mannheim: Bibliographisches Institut. Neurath, Otto, 1970: Foundations of the Social Sciences [1944], in: Foundations of the Unity of Science. Toward an International Encyclopedia of Unified Science, v 2, ed. by Otto Neurath, Rudolf Carnap and Charles Morris, Chicago/London: The University of Chicago Press, pp. 1–51. PU = Wittgenstein, Ludwig: Philosophische Untersuchungen. Philosophical Investigations, dt./engl., transl. by G. E. M. Anscombe, Oxford: Basil Blackwell/New York: Macmillan 1953. Quine, Willard V. O., 1960: Word and Object, New York/London: John Wiley & Sons. Quine, Willard V., 1969: Epistemology Naturalized, in: W. V. Quine, Ontological Relativity and Other Essays, New York: Columbia University Press, pp. 69–90. Rao, B. Narahari, 1994: A Semiotic Reconstruction of Ryle’s Critique of Cartesianism, Berlin: Walter de Gruyter. Ryle, Gilbert, 1971: Abstractions [1962], repr. in: Gilbert Ryle, Collected Papers II, London: Hutchinson, pp. 435–445. Wohlrapp, Harald, 1975: Analytischer versus konstruktiver Wissenschaftsbegriff, in: Zeitschrift für Allgemeine Wissenschaftstheorie 6, pp. 252–275. Wohlrapp, Harald, 2008: Der Begriff des Arguments. Über die Beziehungen zwischen Wissen, Forschen, Glauben, Subjektivität und Vernunft, Würzburg: Königshausen & Neumann.

Acknowledgments I 1. Rules versus Theorems. A new approach for mediation between intuitionistic and two-valued logic. First published in: Journal of Philosophical Logic 2 (1973), pp. 352–369. Reprint with kind permission by Springer SBM, Dordrecht, NL. 2. On the Relation between the Partition of a Whole into Parts and the Attribution of Properties to an Object. First published in: Studia Logica 36 (1977), pp. 351–362. Reprint with kind permission by Springer SBM, Dordrecht, NL. 3. Basic Objectives of Dialogic Logic in Historical Perspective. First published under the title ‘Basic Objectives of Dialogue Logic in Historical perspective’ in: Synthese 121 (2001), pp. 255–263. Reprint with kind permission by Springer SBM, Dordrecht, NL. 4. Pragmatic and Semiotic Prerequisites for Predication. A dialogue model. First published without the subtitle in: Daniel Vanderveken (ed.), Logic, Thought and Action, Springer, Dordrecht 2005, pp. 343– 357. Reprint with kind permission by Springer SBM, Dordrecht, NL. 5. Pragmatics and Semiotics: The Peircean Version of Ontology and Epistemology. First published under the title ‘Pragmatics and Semeiotic: The Peircean Version of Ontology and Epistemology’ in: Guy Debrock/ Menno Hulswit (eds.), Living Doubt. Essays concerning the epistemology of Charles Sanders Peirce [Synthese Library vol. 243], Kluwer Academic Publishers, Dordrecht/Boston/London 1994, pp. 103–108. Reprint with kind permission by Springer SBM, Dordrecht, NL. 6. Intentionality and its Language-Dependency. First published in: Marcelo Dascal/Hubert Cuyckens (eds.), Dialogue. An Interdisciplinary Approach, John Benjamins Publishing Company, Amsterdam/ Philadelphia 1985, pp. 285–292. Reprint with kind permission by John Benjamins Publishing Company, Amsterdam/Philadelphia.

220

Acknowledgments

7. Meaning Postulates and Rules of Argumentation. Remarks concerning the pragmatic tie between meaning (of terms) and truth (of propositions). First published in: Frans H. van Eemeren/Rob Grootendorst/ J. Anthony Blair/Charles A. Willard (eds.), Argumentation. Across the Lines of Disciplines (Proceedings of the First ISSA Conference, Amsterdam 1986, vol. 3), Foris, Dordrecht 1987, pp. 65–71. Reprinted with permission from Mouton de Gruyter. 8. What Do Language Games Measure? First published in: Crítica. Revista Hispanoamericana de Filosofía 21, no. 63 (Dec. 1989), pp. 59–73. Reprint with kind permission by the editors of Crítica, México, D.F. 9. Features of Indian Logic. First published in: Cédric Dégremont/ Laurent Keiff/Helge Rückert (eds.), Dialogues, Logics and Other Strange Things. Essays in Honour of Shahid Rahman, College Publications, London 2008, pp. 263–275. Reprint with kind permission by College Publications, London.

II 1. The Concept of Science. Some remarks on the methodological issue ‘construction’ versus ‘description’ in the philosophy of science. First published in: Peter Bieri/Rolf-Peter Horstmann/Lorenz Krüger (eds.), Transcendental Arguments and Science. Essays in Epistemology [Synthese Library vol. 133], Reidel Publishing Company, Dordrecht/Boston/London 1979, pp. 177–190. Reprint with kind permission by Springer SBM, Dordrecht, NL. 2. Is and Ought Revisited. First published in: dialectica 41 [Norms and Conventions] (1987), pp. 129–144. Reprint with kind permission by Blackwell, Chichester, GB. 3. Competition and Cooperation: Are They Antagonistic or Complementary? First published in: Willibrord De Graaf/Robert Maier (eds.), Sociogenesis Reexamined, Springer-Verlag, New York/Berlin/Heidelberg/ London/Paris/Tokyo/Hong Kong/Barcelona/Budapest 1994, pp. 19–25. Reprint with kind permission by Springer SBM, Dordrecht, NL. 4. Another Version of Methodological Dualism. First published under the title ‘Another Version of Methodological Dualism: Intuitive and Formal Procedure’ as Memo Nr. 9 (1997) of Fachrichtung Philosophie (RATIO) in Sonderforschungsbereich 378 (Ressourcenadaptive kognitive Prozesse) of Universität des Saarlandes, Saarbrücken.

Acknowledgments

221

5. The Pre-Established Harmony Between the Two Adams. First published in: Marcelo Dascal/Elhanan Yakira (eds.), Leibniz and Adam, University Publishing Projects, Tel Aviv 1993, pp. 21–27. Reprint with kind permission by University Publishing Projects, Tel Aviv, Israel. 6. On the Way to Conceptual and Perceptual Knowledge, First published in: F. R. Ankersmit/J. J. A. Mooij (eds.), Knowledge and Language vol. III – Metaphor and Knowledge, Kluwer Academic Publishers, Dordrecht/Boston/London 1993, pp. 95–109. Reprint with kind permission by Springer SBM, Dordrecht, NL. 7. Self and Other: Remarks on Human Nature and Human Culture. First published in: Manuscrito. Revista Internacional de Filosofia 25 [Dialogue, Language, Rationality. A Festschrift for Marcelo Dascal, ed. by Michael B. Wrigley] (2002), pp. 271–289; an earlier version of this paper was published under the title ‘Nature and Culture – An obsolete Distinction?’ in: Changing Concepts of Nature at the Turn of the Millenium (Proceedings Plenary Session of the Pontifical Academy of Sciences 26–29 October 1998), Vatican City 2000, pp. 211–222. Reprint with kind permission by the editor of Manuscrito, Campinas SP/ Brasil. 8. On the Concept of Symmetry. First published under the title ‘Reflections on the Concept of Symmetry’ in: European Review 13, Suppl. 2 [Symmetry and Asymmetry] (2005), pp. 3–11. Reprint with kind permission by Cambridge University Press, Cambridge, UK. 9. Procedural Principles of the Erlangen School. On the interrelation between the principles of method, of dialogue, and of reason. Presented at International Conference on ‘Lebenswelt and Logic: The Erlangen School as Heir to Logical Empiricism’ November 13–14, 2008 at Université Nancy 2. Hitherto unpublished.

Index abilities 9, 57, 174, 176, 192 abilities, linguistic 63 abilities, poietical 143, 192 abilities, practical 141, 143, 187, 192 abilities, technical 141, 187 abstraction 26, 46, 89, 134, 135 abstraction class 82 abstraction, definition by 85, 202 abstraction, theory of 82 action and sign-action 43, 57, 87, 148, 172 action competence 42, 43, 44, 45, 47, 49, 60, 65, 87, 131, 157, 158, 182, 183, 205, 211 action token 44, 46, 79, 205, 212 action type 44, 46, 124, 131, 149, 205, 212 actualization 24, 25, 29, 31, 47, 48, 49, 51, 52, 53, 59 admissibility of rules 33, 34 Anaximander 172 Annambhaṭṭa 103 Apel, K.-O. 112, 119, 122, 126, 127, 128, 139 apperception 154, 167, 168 appredicator 28, 29, 30, 31 appropriation 46, 47, 49, 213, 214, 215, 216, 217 Arendt, H. 144, 147, 193, 197 argumentation theory 72, 95, 96 Aristotle 6, 7, 9, 25, 49, 100, 109, 134, 142, 144, 148, 165, 188, 193, 203 Arnauld, A. 166 ars inveniendi 110 ars iudicandi 110 articulation 9, 30, 48, 49, 50, 51, 52, 68, 75, 76, 89, 90, 120, 135, 136, 158, 179, 180, 209

articulation, symbolic 49, 52 articulator 24, 25, 26, 27, 29, 30, 31, 50, 51, 67, 89, 90, 135, 136, 138, 158, 159 assertion 9, 10, 14, 33, 38, 42, 68, 72, 75, 81, 82, 85, 92, 99, 103, 111, 117, 137, 138, 150, 152, 157, 159 attribution 21, 22, 25, 29, 30, 31, 67, 117 attributor 51 Bach, E. 65 Bar-Hillel, Y. 33, 41 Barlingay, S. S. 105 Barth, E. M. 75, 80 Bartsch, R. 29, 32 behavior, competitive 142, 143, 187, 191 behavior, cooperative 142, 187, 191 behavior, rational 148 behaviorism 151 belief 67, 130, 145, 194 Berge, C. 12, 18, 36, 41 Bhartṛhari 103, 104 Bien, G. 142, 147 Brandom, R. B. 207, 218 Brentano, F. 152, 161 Broch, H. 174, 176, 177, 178, 179, 180, 183 Brocker, M. 190, 197 Brouwer, L. E. J. 3, 5, 6, 8, 18, 35, 36, 41 Buber, M. 214, 215 Bubner, R. 122 Buddha 93, 94 Buridan 8 Camus, A. 178, 185 Carnap, R. 35, 207, 210, 214, 218 Cassirer, E. 148, 157, 161, 215, 218

224 Cavaillès, J. 201 characterizing universal 28, 117 Chi, R. S. Y. 105 Chomsky, N. 109, 118, 122 cognition of an action 46, 47, 68, 87, 88, 129, 131, 134, 135, 205, 212 command 81, 128, 137, 138 communication 50, 60, 76, 90, 127, 130, 135, 159 communicative function 51, 72, 75, 76, 77, 86, 90, 98, 173 competition 140, 141, 143, 144, 145, 146, 189, 194 competition and cooperation, complementarity of 146 constative 42, 68, 69, 117 construction by analysis 22 construction by synthesis 22 Conti, A.-S. 164 continuative 26 contrariness 150, 152 conventionalism 129, 130 cooperation 140, 141, 143, 144, 145, 146, 189, 193, 194 copula 21, 51, 83 Corti, W. R. 111, 122 Cresswell, M. J. 31, 32 cultural process 141, 143, 187, 188, 191, 192, 195 cultural process, decline theory of the 142, 189 cultural process, progress theory of the 142, 188 Curry, H. B. 33, 41 Dascal, M. 53, 54, 219 Davidson, D. 67, 70 de l’Hopital, G. F. A. 163 de Saussure, F. 158 deficient being 143, 192 definition 4, 73, 116, 124, 127, 198, 202 definition of a concept 153 definition, ostensive 59 Democritus 172 demonstrator 51, 179 Descartes, R. 44, 109, 129, 153

Index

designating 48, 52 designator, rigid 59 detachment 47, 213, 214, 215, 216, 217 Dewey, J. 212 Dharmakīrti 97, 98, 99, 101 Dharmottara 101 dialectical field 75 dialogical construction 42, 46, 49, 213 dialogical dyad 194 dialogical philosophy VII, 214, 216 dialogue game VIII, 11, 13, 14, 16, 35, 36, 38, 40, 79 dialogue model 87, 128, 130, 131, 132, 133, 134, 144, 146, 192 dialogue situation 38, 65, 87, 88, 89, 90, 132, 133, 134, 135, 136, 138 dialogue situation, elementary 44, 45, 47, 60, 157, 183 dialogue, concept of 11, 36 dialogue-condition 38, 39 dialogue-definite 10, 13, 14, 16, 17, 36, 40 Dignāga 95, 97, 98, 99, 100, 101, 103, 104 directions of fit 68 directive 68, 69, 117 duality of object and procedure 201, 205, 212, 213 Duns Scotus 8 Ebbinghaus, K. 6, 7, 18 Eco, U. 43 education IX, 195 education, process of 143, 144, 145, 192, 194, 195 eigen-indication 52 eigen-predicator 27 eigen-proposition 52 empiricism 109, 110, 113, 151, 155 empiricism, logical 111, 207, 208, 210 empiricism, radical 111 empiricism-rationalism-debate 88 Euclid 198, 203 Evans, R. D. 105 evolutionarism 110, 111 exemplification 57

Index

explanation 44, 57, 69, 82, 95, 119, 144, 148, 181, 193 feature universal 24, 26, 66, 164 Feyerabend, P. 113, 121, 122 formal truth 8, 79 Foucher, A. 105 Fränkel, A. 33, 41 Frauwallner, E. 95, 105 Frege, G. 54, 72, 73, 76, 78, 80, 92, 134, 150, 155, 161, 167, 170, 207 Frey, D. 199 Gaṅgeśa 99, 102, 103 Gentzen, G. 38, 41 Gerber, U. 112, 122 Gödel, K. 4, 18 golden section 199 Goodman, N. 20, 21, 29, 32, 43, 54, 56, 57, 61, 66, 70, 90, 91, 109, 128, 129, 130, 139, 171, 172, 173, 174, 177, 181, 183, 185, 205, 206, 212, 218 Gottlieb, D. 24, 32, 74, 80 Granger, G.-G. 64, 70, 201, 206, 212, 218 Grua, G. 162, 169, 170 Guha, D. C. 105 Guthrie, W. K. C. 202, 206 Habermas, J. 112, 113, 122, 142, 147 habit change 42, 45, 58, 59, 182 Hacker, P. 93, 105 Hayes, R. P. 105 Heath, T. L. 198, 206 Hegselmann, R. 142, 147 Hempel, C. G. 118, 122 Heraklit 202, 204 Herder, J. G. 143, 145, 147, 192, 193, 194 Hermes, H. 110, 122 Herzberger, R. 105 Hesiod 189 Heyting, A. 4, 5, 18 historism 113 Hobbes, Th. 109, 142, 188 Höffe, O. 190, 197 Hudson, W. D. 119 Humboldt, W. v. 119

225

Hume, D. 125, 136, 138, 139 Huntington, S. P. 188, 197 icon/iconic 43, 56, 60, 82, 86, 171, 183 ideal language 85 identification 24, 25, 30, 48, 49, 131, 180 implication 4, 8, 40, 96, 97, 99, 102, 103, 138, 149, 153, 154, 157 implication thread 207 independence-presupposition 74 index/indexical 24, 51, 53, 59, 60, 180, 183 individual 20, 23, 24, 25, 26, 27, 28, 29, 30, 31, 46, 47, 52, 66, 67, 68, 75, 78, 89, 115, 128, 134, 137, 144, 146, 151, 153, 155, 156, 157, 158, 164, 165, 169, 189, 191, 195, 196 individuals, calculus of 20 individuation 26, 28, 30, 78, 89, 115, 138, 140, 145, 146, 167, 179, 194, 195, 206, 217 individuation, principle of 23, 27 individuation-presupposition 74 individuative 26 individuator 25, 26, 27, 30, 31, 52 inference 7, 34, 75, 79, 94, 95, 96, 97, 98, 100, 102, 104, 138 Ingalls, D. H. H. 102, 103, 105 innateness 109 instance 24, 26, 27, 30, 31, 47, 59, 67, 68, 75, 76, 136, 137, 175, 176, 180, 203 institutional fact 126 intention 62, 68, 77, 137, 146 intentional 131, 144, 145, 193, 195 intermediate schema 25, 29 internal/external 46, 59, 83, 168 interpretant 59 interpretant, logical 45, 58, 182 intuition 103, 120, 150, 153, 154, 160 intuition, forms of 51 I-perspective 44, 45 I-role VIII, 44, 45, 47, 48, 205, 212, 214, 215, 216, 217

226

Index

irrealism 129 is-ought-gap 119, 125, 126 James, W. 111 Janich, P. 119 Joyce, J. 180 judgment 93, 118, 125, 136, 138, 150, 151, 152, 187, 192, 200, 203, 205 Kambartel, F. 119, 123, 142, 147 Kamlah, W. 207, 208, 210, 214 Kant, I. 56, 109, 112, 120, 134, 142, 145, 153, 194, 201, 210 Kapp, E. 9, 18 Kasher, A. 71, 80 Kātyāyana 96 Kauṭilya 93, 95 Kirk, G. S. 203, 206 Kleene, S. C. 5, 18 knowing-how VIII, IX, 146, 209, 214 knowing-that VIII, IX, 160, 214 knowledge by acquaintance 64, 154, 176, 177, 178 knowledge by description 64, 154, 175, 177, 178 knowledge, conceptual 154, 155, 160, 174, 176, 178, 180 knowledge, formal 154, 157, 160 knowledge, intuitive 154, 156, 160 knowledge, means of 94, 98, 103 knowledge, perceptual 43, 56, 86, 154, 155, 171, 174, 175, 176, 178, 180, 214 knowledge, procedural 209, 211, 213 knowledge, propositional 213 Kolmogorov, A. N. 4, 8, 18, 78, 80 Kripke, S. A. 156, 161 Kumārila 99, 103 Lakatos, I. 113 language-game 42, 43, 45, 56, 57, 58, 60, 64, 65, 68, 69, 72, 76, 81, 82, 84, 86, 90, 114, 119, 128, 132, 136, 146, 158, 159, 171, 174, 182, 183 Leibniz, G. W. 5, 17, 25, 44, 53, 77, 110, 148, 149, 153, 156, 161, 162, 163, 164, 165, 166, 167, 168, 170, 205

Leonard, H. 20, 21, 32 Leśniewski, S. 20, 32 Lessing, G. E. 188 Levy, A. 33, 41 Lewis, C. I. 5, 56 Locke, J. 156 logic and epistemology, interconnectedness of 97 logic and philosophy of language, interrelation of 97 logic in antiquity. 9 logic, classical 8, 17, 40, 204 logic, definition of 8 logic, deontic 138 logic, dialogic 33, 37, 40, 78 logic, effective 8, 17 logic, formal 6, 8, 9, 16, 79, 92, 103, 150, 159, 207 logic, operationist 33, 34, 39 logic, traditional 30, 31 logical syntax 85 Lorenz, K. 9, 11, 12, 13, 15, 17, 18, 26, 32, 36, 39, 41, 42, 47, 50, 53, 54, 60, 61, 65, 67, 70, 78, 80, 82, 88, 91, 105, 114, 117, 119, 123, 130, 133, 137, 139, 142, 144, 147, 158, 161, 163, 169, 170, 173, 180, 183, 185, 189, 197, 205, 206 Lorenz, Konrad 109 Lorenzen, P. VIII, 4, 8, 11, 18, 33, 35, 36, 41, 78, 80, 112, 119, 123, 207, 208, 210, 214, 218 Lorenzer, A. 194, 197 Łukasiewicz, J. 7, 18 Lyotard, J.-F. 190, 197 MacIntyre, A. 190, 197 Mahāvīra 93, 94 Matilal, B. K. 94, 105 Maturana, H. R. 190, 197 McKinsey, J. C. C. 5, 18 Mead, G. H. 65 mediation 48, 50, 213 mediation, comprehensive 52 Meggle, G. 62, 70, 128, 139 mentalism 151 mereology 21, 117

Index

metacompetence 63, 90, 177, 178, 184 mimesis and poiesis 181 mind/mental IX, 46, 47, 56, 58, 62, 64, 109, 125, 134, 136, 148, 149, 150, 152, 153, 154, 156, 157, 158, 182, 186, 205, 215 Mittelstraß, J. 112, 119, 123 mode of being given 50, 77, 79, 159 Moore, G. E. 84, 91, 207 Morris, Ch. 43 Nau, H. H. 190, 197 Neurath, O. 208, 218 nominator 24, 27, 28, 52, 54, 72, 117 Oberhammer, G. 105 object and method, unity of 176, 180 object-competence 63, 64, 90 objects and procedures, duality between 201 obligation 38, 39, 69, 125, 127, 137, 138, 194 observation and participation, complementarity between 128 order 14, 68, 136, 137, 138 organization, conceptual 214 ortho-language 208 ostension 51 ostension , mode of 159 ostension, intent of 50 Pāṇini 94, 96 Parmenides 172 partaking 48, 52 particle-rules 13 particular 25, 43, 46, 47, 48, 49, 50, 52, 53, 78, 98, 104, 115, 118, 119, 121, 132, 136, 137, 151, 152, 153, 158, 159, 213, 216 particular within a situation 49 particulars (= particularia) X, 46, 47, 49, 50, 52, 54, 104, 133, 151, 153, 154, 158 partition 22, 25, 30, 53, 74, 97, 117 Patañjali 94, 96, 97 Patnaik, T. 105 Pears, D. 123 Peirce, C. S. 42, 43, 44, 56, 57, 58, 59, 60, 64, 66, 82, 86, 97, 111, 125,

227

128, 134, 146, 148, 149, 156, 163, 171, 173, 179, 180, 181, 182, 183, 211 perception 51, 77, 88, 89, 90, 94, 95, 96, 97, 98, 104, 134, 135, 136, 149, 151, 153, 154, 156, 157, 158, 159, 160, 166, 167, 168, 176, 180, 184 performance of an action 46, 47, 57, 67, 68, 87, 124, 129, 130, 131, 132, 134, 136, 149, 183, 205, 212 performator 51, 77 Petrus Hispanus 8 phenomenalism, linguistic 207, 208 phenomenological reduction 213 Piaget, J. 131, 132, 139 Plato 9, 50, 74, 75, 86, 141, 142, 148, 155, 159, 187, 203 Pliny the Elder 199, 206 poiesis and praxis, distinction of 141 polarity, dialogical 196, 205, 212 Popper, K. 35, 113 Potter, K. H. 106 Pragmatic Maxim 45, 58, 60, 97, 146 pragmatic side of an action 44, 48, 50 pragmatization of semantics 72 Praśastapāda 100 preaction 65, 66, 67, 88, 89, 114, 115, 119, 121, 132, 133, 134 predication 51, 54, 67, 76, 77, 81, 82, 83, 89, 90, 135, 136, 150, 151, 155, 159 predication, content of 50, 150 predication, identity theory of 152 predication, inherence theory of 152 predication, mood of 50, 90, 135, 136 predications, formula of seven 93 predicator 25, 27, 28, 29, 30, 31, 51, 54, 72 preference 138, 145, 194 preobject 89, 134, 135, 137 principle of conceptualization 213 principle of continuity 166 principle of contradiction 165, 168 principle of dialogue 113, 212, 213, 214, 216, 217

228 principle of method 113, 209, 211, 213 principle of reason 210, 213, 216 principle of self-similarity 48, 158 principle of sufficient reason 165 principle of the identity of indiscernibles 25 principle of verification 35 proficient being 143, 192 properties, parts into 23 property 21, 22, 23, 29, 30, 31, 53, 54, 71, 83, 100, 103, 116, 117, 151, 152, 158, 159, 169, 198, 202 property, carrier of a 47 property, characterizing 23, 25, 27 property, external 83, 181 property, internal 83, 84, 85, 166, 181 proportion 149, 203 proposition and judgment 200 proposition, concept of 4, 6, 11, 35, 36, 42, 64, 73 proposition, logically true 16 proposition, minimal 151, 152 proposition, truth of a 11, 16, 149 propositional kernel 42, 74, 81, 124, 138 Protagoras 141, 187, 188, 192 protologic 33 prototype 43, 56, 86, 114, 172 Putnam, H. 109, 174 Pythagoras 203 Quine, W. V. O. 5, 9, 18, 24, 29, 32, 74, 80, 116, 121, 123, 157, 161, 165, 170, 208, 215, 218 quotation 66, 96 Raghunātha 103 Rao, B. N. 190, 197, 218 rationalism 109, 110, 113, 151, 155 rationalism, logical 151 rationalism, radical 111 rationality 142, 144, 145, 177, 187, 193, 199, 202, 203 rationality, means-ends 144, 193 rationality, notion of 187 rationality, practical 213, 215 rationality, theoretical 213

Index

reference 24, 52, 65, 77, 78, 96, 115, 137, 154, 168, 181, 212 reference of linguistic entities 72, 77, 78, 101, 156 reference, frames of 129 reference, means of 205 reflection 146, 176, 196, 200 regulation 69, 119 Reichenbach, H. 53, 54, 78, 80, 110, 123 relation, epistemological 174 relation, external 83, 145, 167, 168, 195 relation, internal 83, 84, 195 representation 43, 47, 49, 56, 64, 68, 82, 86, 96, 98, 110, 124, 130, 148, 156, 164, 167, 168, 171, 174, 176, 178, 181, 192, 196, 216 research 43, 56, 111, 113, 114, 117, 171, 176 research and presentation 114, 118, 119 Richter, V. 8, 18 Rorty, R. 115, 123, 130, 139 Rousseau, J.-J. 142, 189 Ruegg, D. S. 106 rules of argumentation 10, 12, 13, 79, 94, 210 rules of construction 50, 210, 213, 214 rules of signification 79 rules of translation 49, 82, 84, 85, 214 Russell, B. 64, 70, 121, 177, 207 Ryle, G. 156, 161, 207, 214, 218 Sañjaya 93, 94 Saussure, F. de 50 scepticism 110, 160, 186 Scharfe, H. 94, 106 schematization 47, 48, 49, 53, 158, 164, 165, 167, 213, 216 Scherer, A. 190, 197 Scherer, B. M. 45, 54, 58, 61, 146, 147 Schneider, H. J. 63, 70 Scholz, H. 7, 18 Schwemmer, O. 112, 119, 123, 208 Searle, J. R. 119, 125, 126, 127, 139

Index

self-determination 145, 194, 195, 196 self-education 145, 195, 216 semantic ascent 128 semantization of pragmatics 71 semiotic side of an action 48, 50 semiotic triangle 58, 182 sense of linguistic entities 72, 77, 78, 85, 156 sense, verbal 179 sense, visual 179 sentence 10, 42, 67, 72, 73, 75, 76, 77, 84, 85, 90, 96, 97, 98, 103, 117, 124, 180, 192, 204 sentence form 74 sentence, elementary 28, 74, 102, 179, 180 sentence, mood of a 124, 136 sentence, occasion- 128 sentence, one-word- 51, 76, 86, 159 sentence, sense of 86 sentence, subject-predicate- 137 sentence-radical 82 set theory 20, 21, 23, 117 signification 50, 60, 76, 83, 90, 111, 114, 117, 130, 135, 159 significative function 51, 72, 75, 76, 86, 90, 98, 156, 171 singular 24, 46, 47, 51, 65, 66, 67, 115, 118, 119, 121, 131, 133, 149, 157, 158, 164, 165, 179, 213 singular ingredients 46 singular term 9, 27, 54, 59, 65, 72, 74, 77, 78, 115, 117, 151 singulars (= singularia) 47, 49, 74, 78, 101, 104, 116, 134, 158, 164, 167 Skolem, Th. A. 3, 18 Sneed, J. 116 sociality 142, 144, 145, 146, 188, 193, 196 socialization 138, 140, 145, 146, 194, 195, 217 society 191 solidarity 141, 145, 178, 187, 194 Solomon, E. A. 106 sortal universal 26

229

speech act 37, 51, 62, 63, 64, 67, 68, 72, 73, 77, 81, 82, 90, 125, 126, 127, 136, 200 Spinoza, B. 44 Staal, J. F. 96, 106 statement 68, 125, 126, 127, 128, 136, 138, 151, 200 Stegmüller, W. 17, 19, 113, 116, 123 Steinkellner, E. 106 Steinmann, H. 190, 197 Stich, S. P. 109, 123 Strawson, P. F. 9, 19, 24, 26, 28, 32, 66, 70, 115, 117, 123, 164 substance 31, 85, 93, 102, 116, 148, 152, 163, 164, 165, 166, 168 substance, appearance of a 47, 53 substance, individual 165, 166 summation 24, 49 syllogistic 6, 8, 92, 100 Tarski, A. 5, 12, 18, 19, 35, 73 Thales 172 theory of monads 153, 164, 168 theory of reasoning 93 thinking, forms of 51 Thomas Aquinas 152 transcendentalism 110, 111, 113, 124 truth claim 51, 79, 129, 137, 138, 159, 173 truth theory 72 truth, concept of 4, 36, 72, 73, 128, 137, 155 truth, definition of 12 truth, formal 16, 39 truth, logical 5, 16, 34, 39 truth, material 16, 39, 79 truth, operationist 34 type and token 46, 116, 131, 133, 179, 213 Udayana 99 universal 46, 47, 51, 53, 65, 103, 104, 115, 118, 121, 131, 133, 145, 149, 153, 157, 158, 179, 191, 194, 213 universal features 46, 101 universals (= universalia) 47, 49, 98, 103, 104, 116, 134, 149, 152, 155, 156, 158, 163, 164, 167

230 value-definite 3, 11, 15, 17, 35, 36, 40 van Bijlert, V. 101, 105 Vanderveken, D. 68, 70 Varela, F. J. 190, 197 Vasubandhu the Younger 100 Vātsyāyana 95 verification 111, 114, 126 von Weizsäcker, C. F. 109 Vṛṣagaṇa 95 Walther, E. 173, 185 way of life 50, 86, 128, 145, 146, 186, 187, 189, 190, 194, 195, 217 way of presentation 110, 113, 118, 210 way of research 110, 113, 118, 210 Weber, M. 146, 147 Weizsäcker, C. F. v. 114, 123 Weyl, H. 199, 206 wheel of reasons 99, 100, 101 Whitehead, A. N. 115

Index

winning strategy 11, 15, 16, 36, 38, 40, 159, 160 Wittgenstein, L. 8, 43, 45, 51, 56, 57, 58, 62, 64, 68, 72, 73, 76, 80, 81, 83, 84, 85, 88, 90, 91, 128, 132, 136, 141, 146, 156, 159, 161, 171, 174, 175, 177, 179, 180, 182, 184, 207, 211, 214 Wohlrapp, H. 113, 114, 116, 123, 208, 213, 218 Wolters, G. 190, 197 word 9, 10, 27, 59, 67, 75, 86, 96, 97, 98, 103, 104, 183, 184, 200 world view 50, 86, 110, 124, 128, 145, 146, 172, 173, 177, 186, 189, 194, 195, 217 Yakira, E. 53, 54 You-role VIII, 44, 45, 48, 205, 212, 214, 216, 217