Rethinking Epistemology: Volume 1 9783110253573, 9783110253566

Citation preview

Rethinking Epistemology 1

Berlin Studies in Knowledge Research Edited by Günter Abel and James Conant

Volume 1

De Gruyter

Rethinking Epistemology Volume 1 Edited by

Günter Abel and James Conant

De Gruyter

Editors Prof. Dr. Günter Abel Technische Universität Berlin Institut für Philosophie Straße des 17. Juni 135 10623 Berlin Germany e-mail: [email protected] Prof. Dr. James Conant The University of Chicago Dept. of Philosophy 1115 E. 58th Street Chicago IL 60637 USA e-mail: [email protected]

ISBN 978-3-11-025356-6 e-ISBN 978-3-11-025357-3 Library of Congress Cataloging-in-Publication Data Rethinking epistemology / edited by Günter Abel and James Conant. p. cm. Q (Berlin studies in knowledge research ; v. 1) Includes bibliographical references and index. ISBN 978-3-11-025356-6 (v. 1 : hardcover : alk. paper) ISBN 978-3-11-027782-1 (v. 2 : hardcover : alk. paper) 1. Knowledge, Theory of. I. Abel, Günter. II. Conant, James. BD161.R4837 2011 121Qdc23 2011035235

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 2012 Walter de Gruyter GmbH & Co. KG, Berlin/Boston Printing: Hubert & Co. GmbH & Co. KG, Göttingen ⬁ Printed on acid-free paper Printed in Germany www.degruyter.com

Foreword The present volume is the first of two volumes which together form a single publication which seeks to introduce a variety of new approaches to current problems in epistemology. Taken together the two volumes are the starting volumes of the Berlin Studies in Knowledge Research series. They are intended to open the series and indicate some of the variety of topics to which the subsequent forthcoming volumes will each be individually devoted. As the title which these volumes jointly bear – Rethinking Epistemology – indicates, their shared aim is to identify important topics in the theory of knowledge which have either been unduly neglected in recent philosophy or whose consequences for other areas of epistemology in particular and philosophy more generally have tended to remain unappreciated. This is therefore also the central aim of the present volume as well as one of the central aims of the entire series which this pair of volumes seeks to indicate. The series as a whole is animated by a number of subsidiary aims as well. It is closely associated, both in its selection of topics and its overall philosophical orientation, with the projects promoted by the Innovationszentrum Wissensforschung (IZW), or Center for Knowledge Research, at the Technische Universität Berlin. The goal of both the Center and the series is to foster systematic research into the variety of forms of knowledge that there are, as well as to uncover aspects of their underlying unity. The handful of remarks that follow are intended to provide only the barest outline of the intellectual program of the Center and its associated publication series. (A detailed discussion of the variety of forms of research into the nature of knowledge here envisioned and of what makes them valuable is provided in the opening contribution to the volume.) In offering these remarks here, in the Foreword to this collection, we do not mean to suggest that the contributors to this volume necessarily conceive of themselves as all engaged in a common intellectual project, let alone the one sketched below. We mean rather merely to suggest that their work, considered collectively, is helpfully viewed as marking a certain trend in recent work in analytic philosophy in both the Anglo-American and German academy – a trend which, on the one hand, marks a notable departure from the topics

VI

Foreword

and concerns which previously formed the staple of analytic epistemology and which, on the other hand, marks a movement in the direction of a more capacious understanding of what knowledge is and what the philosophical study of it might require. For the purposes of both the volume and the series, the conception of the discipline of epistemology at work throughout is intended to be a maximally generous one—one which encompasses a study of the full variety of forms, practices and dynamics of knowledge, as well as their mutually interacting points of contact and their respective mechanisms of interpenetration. It is this generous conception of the discipline of epistemology that is indicated by the German expression Wissensforschung and its regrettably unattractive English counterpart “knowledge research”. On this conception, the topics of mainstream epistemology, such as a posteriori perceptual and a priori inferential knowledge, are asked to take their place alongside the full spectrum of other forms of knowledge that come with the entire spectrum of human cognition, from thought, speech, and self-consciousness to action, skill, and virtue. The series thereby seeks to bring about an expansion and revision of classic analytical conceptions of the philosophical disciplines of epistemology and the theory of science. The hope is to nudge the future philosophical study of knowledge into an attitude of greater openness to interdisciplinary and transdisciplinary approaches to its topic and to foster a greater appreciation of the irreducible plurality of the various forms of human knowledge. This leads to a reorientation of the discipline of epistemology, undoing some of its artificial restrictions in its conception of the scope of its topic, while introducing a greater emphasis on the heterogeneity of different forms of knowledge, hopefully thereby effecting the possibility of a nuanced study of the ways in which these heterogeneous forms mutually impinge upon and condition one another. Some felt need for such forms of inquiry is reflected in the increasing profusion of partially overlapping terminological pairs – pairs such as “implicit/explicit”, “discursive/non-discursive”, “theoretical/practical”, “knowinghow/knowing-that”, “conceptual/non-conceptual”, “codifiable/uncodifiable”, “first-person/third-person”, personal/subpersonal”, and “normative/non-normative”—and an increasing appreciation of the ways in which none of the distinctions drawn by any of these terminological pairs manages to align all that perfectly with those drawn by any of the others. Although it forms a less central part of the agenda of the present volume than that of some of the forthcoming subsequent volumes, a sub-

Foreword

VII

sidiary concern of the series as a whole is to help redefine the traditional theory of science, as well as the variety of recent forms of science studies which have sought to replace it. The hope is to foster a greater dialogue between these forms of inquiry and philosophy proper, so that the study of the acquisition and development of scientific knowledge can become more philosophically systematic, while philosophical reflection on science can grow more sensitive and attentive to the genuine diversity and irreducible historicity of the forms of scientific knowledge which have proliferated over the centuries. Finally, the series will seek to foster reflection on the ways in which scientific and everyday ways of knowing overlap and diverge with one another, while resisting the tendency to hold up either as the standard by which to measure the other. This last aspect of the orientation of the series reflects its concern with promoting proper philosophical reflection on the interface between the world as it is known to us in our everyday lives (what the Germans call the Lebenswelt) and the world as its nature is disclosed to us by the natural sciences— between what Wilfrid Sellars called the manifest image and the scientific image of the world. Philosophical comprehension of the relations in which these two images stand to one another requires an exploration of the manifold points at which philosophy, the humanities, the social sciences, and the natural sciences all interact with one another. A number of the subsequent volumes in this series will seek to explore some of these transdisciplinary issues. A final aim of this publication series is to promote and facilitate intellectual exchange between European and Anglo-American scholars of knowledge. The hope is to draw attention to key writings by European researchers which have gone relatively unnoticed in the Anglo-American discussion and to promote a broader consciousness of seminal recent English-language contributions which have yet to be received in Europe. Each volume will therefore contain contributions by both Anglo-American and European scholars working on closely related topics. The present volume of this new series has a focus which in one respect is comparatively narrow than that of the series as a whole: it targets some of the central philosophical questions of epistemology as they have been taken up in recent years by philosophers seeking to reshape our understanding of what knowledge is. The editors would like to express their indebtedness to a number of institutions and persons. First, we are grateful to the Technische Universität Berlin for establishing the Innova-

VIII

Foreword

tionszentrum Wissensforschung (Center for Knowledge Research). We would also like to thank the publishing house Walter de Gruyter (Berlin – New York) and above all Dr. Gertrud Grünkorn (Editorial Director) for supporting the idea of this series from its inception. We would also like to thank the staff of the Innovationszentrum Wissensforschung; without their effort and commitment the present volume would not have come into existence. Particularly, noteworthy in this regard were the services of Martin Wolf, Hadi Faizi, Katharina von Laer, Can Atli, Claudio Roller, Stefan Tolksdorf, Peter Remmers, Doris Schöps, Elisabeth Simon and Daniel Smyth (Chicago). Last but not least, we are indebted to the authors of this pilot volume of the series. It is through their contributions that the idea of “Rethinking Epistemology” has been converted from a possibility to an actuality. Günter Abel (Berlin) & James Conant (Chicago)

Contents Günter Abel Knowledge Research: Extending and Revising Epistemology .

1

I. Theoretical Knowledge Peter Janich A Priori Knowledge in Methodical Philosophy . . . . . . . . . . . .

55

Ulrich Dirks Knowledge and World-Pictures . . . . . . . . . . . . . . . . . . . . . . . .

83

Hans Julius Schneider Knowledge Forms and Language Forms . . . . . . . . . . . . . . . . . .

133

Martina Plümacher Epistemic Perspectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

155

Hans Jörg Sandkühler Critique of Representation: Cultures of Knowledge – Humanly Speaking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

173

II. Perceptual Knowledge Thomas Land Kantian Conceptualism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

197

Arata Hamawaki Ginsborg on the Normativity of Perceptual Representation . .

241

Jocelyn Benoist The Internal and the External in Knowledge . . . . . . . . . . . . . .

273

X

Contents

III. Knowledge in Science Hans-Jörg Rheinberger Genesis of Knowledge Spaces and Objects of Knowledge . . . .

287

Bernd Mahr On the Epistemology of Models . . . . . . . . . . . . . . . . . . . . . . . .

301

Michael Hampe Explanation Through Description . . . . . . . . . . . . . . . . . . . . . . .

353

Hans Poser Knowledge of Ignorance: On the Problem of the Development and the Assessment of Technology . . . . . . . . . . . . . . . . . . . . . .

369

Notes on Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

413

Index of Persons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

417

Index of Topics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

423

Knowledge Research: Extending and Revising Epistemology Günter Abel I. The Conception of Knowledge Research: 1. The Concept of Forms of Knowledge and Their Taxonomy; 2. The Interplay of Knowledge Forms; 3. Knowledge Research as a Reflective, Fundamental Investigation; 4. What Knowledge Research is Not Limited to II. Extending and Revising Epistemology: 1. Knowledge Research and Epistemology; 2. Knowledge Research and Epistemological Methodology; 3. Knowledge Research and Epistemic Objects; 4. Knowledge Forms’ Modes of Interpenetration: (a) Example 1: Conceptual and Non-Conceptual Knowledge; (b) Example 2: Explicit and Implicit Knowledge; (c) Example 3: Distributed and Integrated Knowledge; (d) Example 4: Knowing-How and Knowing-That; (e) Example 5: Forms of Knowledge and Forms of Representation; (f) Example 6: Knowledge Forms and Creativity

I. The Conception of Knowledge Research 1. The Concept of Forms of Knowledge and Their Taxonomy Knowledge plays a fundamental role in all processes of human perception, speech, thought, and action. The fluid function of the triangular relation between an individual, other persons, and the world always already involves and presupposes indispensable forms of knowing. The least reflection reveals how multifarious and how broad the meaning of the word ‘knowledge’ is. Even beyond the domains of science, technology, philosophy, and the arts, the word crops up throughout everyday life and in the most various contexts. One need only think of phrases like, ‘knowing for certain’, ‘knowing the score’, ‘knowing how to size things up’, ‘let him know’, ‘knowing a hawk from a handsaw’, ‘knowing how to dress oneself’, ‘a knowing look’, ‘knowing how things go’, ‘knowing the thing to do’, ‘knowing what counts as an argument’, and countless others. The following brief taxonomy of knowl-

2

Günter Abel

edge forms marks out some of the principal distinctions between various different forms and concepts of knowledge. This spectrum of forms of knowledge extends from everyday cases (e. g. knowing how to open a car door, or how to organize one’s day) to cases of artistic knowledge (e. g. knowing how to depict something, or how to render something expressive – in dance, music, or painting) and to cases of scientific knowledge as well (e. g. knowing how to conduct a methodological investigation in the natural sciences, in mathematics, or in the social sciences).1 (1) First, one must distinguish between a narrow and a broad concept of knowledge: (a) The narrow concept of knowledge pertains to acts of cognition that are bound up with method-governed procedures and with justification, truth, rationalization, and demonstrability. Any instance of this sort of knowledge must be able to be discussed, must be communicable, tradable and intersubjectively verifiable. The sciences are paradigmatically characterized by such a concept of knowledge. (b) The broad concept of knowledge signifies, on the one hand, (i) the abilities involved in adequately grasping what is going on and what various things (e. g. a gesture, an image, a sentence) are about. On the other hand, it also pertains to the realm of basic human capabilities, competencies, abilities, skills, practices, and proficiencies. In this broad sense of the expression, knowledge is an inviolable component of the facticity of every human action, speech act, thought, and perception. This broad and foundational domain of knowledge is an entirely familiar element of both our everyday lives (e. g. our everyday practices and know-how) as well as the arts and sciences. In connection with physics, for example, talk of such broad knowledge might refer to one’s competence to execute an observation, to construct an experimental procedure, to manipulate a mathematical model and to apply it to the world of physical objects and events. Without taking account of this broad sense of ‘knowledge’ it is impossible to formulate a comprehensive theory of human perception, speech, thought, action, and cognition, much less an far-reaching and satisfactory epistemology. (2) Our next step must be to make a heuristic distinction between various forms of knowledge, such as everyday knowledge (“knowing where the next mailbox is”), theoretical knowledge (“knowing that 2+2=4”), practical knowledge (“knowing how to steer an automobile”), 1

On what follows, cf. also Abel (2007).

Knowledge Research: Extending and Revising Epistemology

3

and orientational knowledge (“knowing what one is to do (or forgo) in a given situation”). These various forms of knowledge are perfectly familiar. As a rule, we have a direct understanding of them both in our everyday lives and in the arts and sciences. And the same holds true of the three components to be briefly sketched in what follows. These components cut orthogonally across both the distinction between the narrow and broad senses of knowledge and the distinction(s) between various forms of knowledge. The aim of systematic and reflective knowledge research is to elucidate the peculiar profiles of knowledge forms as well as their interplay. But to do so requires (as we shall see and discuss later in section II) various extensions, modifications, and revisions of traditional and contemporary epistemology. For we cannot understand the sense of epistemological questions regarding validity, justification, and the limits of knowledge until we have established which forms of knowledge (and which modes of their interplay) are at issue in a given case. To this extent, knowledge research must precede epistemology proper, for the latter always presupposes the former. To risk a provocative formulation: epistemology is a branch of knowledge research, not the reverse. This provides us with a horizon against which to work out the profiles and modes of interplay proper to the following three orthogonal conceptual pairs of knowledge forms. And it enables us to do so in a way that highlights the significance of these pairs of knowledge forms for an extended and revised epistemology. Consider then: (a) knowing-how vs. knowing-that; (b) conceptual vs. non-conceptual knowledge; and (c) implicit vs. explicit knowledge. Our task is to create a place in our thinking about knowledge to accommodate these forms of knowing and their modes of interplay. Classical epistemology takes its primary orientation from knowing-that and hence focuses on linguistic, propositional, and explicit knowledge. It is consequently incapable of finding an adequate place for the other forms of knowledge we have mentioned and is not open to the various revisions in our epistemological views that they necessitate. Yet we cannot allow our thinking about knowledge to be restricted by such a narrow epistemology. In light of the dominant cognitivist bias of contemporary epistemology and the attempts to bypass epistemological questions in favor of practical considerations, it seems to me that Rethinking Epistemology has become a tremendously pressing task.

4

Günter Abel

Despite the various ways in which they differ, knowing-how (i. e. procedural and practical knowledge or ability, such as the knowledge of how to ride a bicycle) and knowing-that (i. e. theoretical, propositional knowledge, such as the knowledge that 2+2=4) also interpenetrate in numerous ways. Knowing-how is governed by its own peculiar and genuinely practical rules which are embodied in one’s engagement in the relevant practice itself (e. g. in actual bike-riding or cello-playing).2 In many cases, what is at issue is a form of “learning by doing” which cannot be acquired through theoretical engagement or time spent in a lecture hall. Elucidating (a) the peculiar profile of various knowledge forms, (b) their points of overlap, and above all (c) the mechanisms of their interpenetration make up some of the fundamental tasks of comprehensive and systematic knowledge research. (These tasks also mark out desiderata for an extended epistemology.) But please note that I am only talking about elucidating the profile, interplay, and interpenetration of various forms of knowledge. I am not talking about determining or establishing the primacy or dominance of one knowledge form over the others. Traditional cognitivist and propositionalist epistemologists asserted the dominance of knowing-that and maintained that knowing-how could ultimately be characterized as a form of knowing-that. 3 I take this position to be just as unsustainable as its complement, which reemerges every now and then in the contemporary literature – namely, the view that knowing-how is the fundamental form of all knowledge and that all knowledge-that must ultimately rest upon knowledge-how.4 Ironically, this latter position is equally in the thrall of the very conception of knowing-that it is formulated to oppose. It is absolutely essential, in my view, that we free ourselves from the stranglehold of this whole dichotomy. Similar considerations apply to the second of our three conceptual pairs: the relation of conceptual and non-conceptual knowledge. Here too, the crucial task is to undercut the sterile opposition of these two aspects. But the first step must be to bring out the entirely trivial point that both theory and praxis involve not only conceptual, propositional knowledge (that can find linguistic articulation in a that-clause) but also non-concep2 3 4

For a more thorough discussion, cf. Abel (2011b) and (2010b). A paradigm example of this position in the contemporary literature is Stanley/ Williamson (2001). Take, for instance, Hetherington (2006).

Knowledge Research: Extending and Revising Epistemology

5

tual knowledge (e. g. perceptual, intuitive, emotional, pictorial, musical knowledge). For any actual engagement in or process of life, experience, or theory will be characterized by a specific constellation of interplay and interpenetration between various forms of knowledge. Nevertheless, it seems to me crucial that we remain mindful of the unique value of each form of knowledge and begin to address and rehabilitate all those forms which have failed to find a home in traditional and contemporary epistemology. Included among these neglected forms are knowing-how, non-conceptual knowledge, and implicit knowledge. Likewise, intuitive or visual thinking, for example, deserves a place in a comprehensive philosophy of knowledge and its forms. Examples of visual thinking can be found in many domains – for example, in diagrammatic representations, in geometry, or in the visual arts (painting, graphic design, sculpture, architecture). Of course, one of the profound challenges here – and one which has by no means been satisfactorily answered – is to convincingly demonstrate that visual thinking is a genuine and internally cognitive form of thought and knowledge (in both the narrow and the broad sense of ‘knowledge’). Yet such a demonstration must not be confused with the traditional and (with a few exceptions) dominant contemporary view that visualization is a merely secondary, illustrative and intuitive presentation of cognitive processes and arguments, which have, for their own part, a source that is completely independent of such visualization. In contrast to this misleading picture, we must rather strive to demonstrate that, to put it starkly, the argument itself lies in the visual. Grasping this constitutes the heart of the matter and the real intellectual challenge. However, the characteristic mechanisms of visual thinking and the role it might come to play in an extended and revised epistemology are, of course, as yet largely unknown. It is often more advantageous to pursue this question in other disciplines rather than within philosophy itself. Consider, for instance, some of the “proofs without words” which are so well known in mathematics and other disciplines.5 Visual thinking has genuine epistemological value in its own right. Since knowledge forms manifest themselves in and through signs, we must acknowledge the general principle that all forms of knowledge and any successful use or understanding of signs relies on aspects of intuitive thought and visual cognition. Elucidating the 5

Cf. Giaquinto (2007); Nelsen (1993) and (2000). Naturally, visual thinking has a firm place in Art History and the Psychology of Art. A classic in that tradition is Arnheim (1997).

6

Günter Abel

cognitive character of intuitive thought (visual thinking) will also enable us, in my view, to formulate more adequately than ever before the ways in which knowledge forms interpenetrate perceptual, conceptual, and enactive processes in speech, thought, and action. Satisfying this desideratum poses an important task for systematic and reflective knowledge research. Something analogous can be said about the two other conceptual pairs mentioned above (viz. conceptual and non-conceptual knowledge; explicit and implicit knowledge). I will consider these two pairs in detail in section II.4. For now we can simply remark that these three conceptual pairs can easily be augmented by others (as indeed they must be) – for example, by pairs like ‘scientific’ and ‘non-scientific’ knowledge; or ‘declarative’ and ‘expressive’ knowledge. But we cannot pursue this line further here. At this point it is far more important to me to briefly draw our attention to the sense of talk about ‘forms’ of knowledge. Because the term ‘form’ is clearly associated with a prominent and conceptually charged history in the Western philosophical tradition (and not only there), it is important to preempt certain misunderstandings. I would like to paraphrase talk of ‘forms’ of knowledge (in an explicitly Kantian manner) as ‘modes and kinds’ of knowledge. In keeping with this paraphrase, ‘form’ is not to be conceived here or in what follows as a predetermined and independent order, nor as a container in which knowledge must crystallize in order to so much as count as knowledge. Nor should ‘form’ be understood as the general structure of knowledge. Any attempt to articulate ‘the one and only general form’ of the very various and irreducibly plural modes of knowledge is doomed to fail. Nor should ‘form of knowledge’ be understood in what follows to signify an order that is already immanent to knowledge and proper to it, even before we finite, limited minded creatures attempt to pour such knowledge and its purportedly pre-individuated content into one of the forms available to us as humans – e. g. a linguistic form6, a pictorial form, or a form of action. Such a conception does not do justice to the phenomenon of knowing in either the broad

6

On this point, cf. also Hans Julius Schneider’s contribution to the present volume.

Knowledge Research: Extending and Revising Epistemology

7

and factical7 sense, nor in the narrow linguistic-propositional sense. Moreover, this conception is rooted in the questionable idea that linguistic, pictorial, and action forms are not really conditional for, but merely instrumentally efficacious in the formation and articulation of knowledge. But if we think critically about this, we can rather formulate the contrary thesis – namely, that, for our finite minds, there simply cannot be any orders or contents of knowledge that are pre-individuated and wholly independent of our semiotic, depictive, and interpretational forms.8 To pay critical attention to the sense of our claims, we must ask ourselves: what are these contents and orders of knowledge that are supposedly lying out there pre-formed and wholly independent of our mechanisms of discrimination, individuation, and specification, waiting for us to grasp and take them up in a purely passive manner? And how on earth are we supposed to be able to know anything about such wholly sign-free and interpretation-free contents and order? Critical attention to the sense of these notions reveals that such ‘self-identifying knowledge contents’ are simply not conceivable. The reasonable course here is not to try to refine this idea, but to reject it. Accordingly, I do not understand the dependence of knowledge contents upon our semiotic and interpretational processes as though there were a content somewhere behind the constitutive semiotic and interpretational functions that somehow manages to initiate a causal sequence which leads to its manifestation in signs available to us. Such a picture is frequently connected with the (misguided) hope that, with the help of (scientific) theories and models we might be able to infer from a sign’s semantic characteristics (meaning, reference, truth or satisfaction conditions) to the content that purportedly lies behind them and is causally responsible for them. What this picture fails to capture is the interpretational nature of signs – it misses their basic, factical role, their original, constitutive role, construing them instead as merely instrumental. Such a perspective further fails to recognize that theories 7 8

I use the expression “factical” in order to address the pragmatic facticity of knowledge in the broad sense – a pragmatically given actuality of ways of knowing which can neither be suspended nor evaded. This thesis forms the point of departure for a general philosophy of signs and interpretation. On the basic contours of such a philosophy, cf. Abel (1995a); (1999) and (2004).

8

Günter Abel

and models themselves presuppose signs and interpretation and can even be conceived as semiotic and interpretational constructs.9 But this means that such theories and models cannot be employed in the supposedly coherent inference back to interpretation-independent content, as portrayed in the picture sketched above. The problem is not resolved but only postponed. The real desiderata in an investigation of ‘knowledge’, ‘signs’, and ‘interpretation’ get pushed into other fields. 2. The Interplay of Knowledge Forms If one explicitly asks, against the background of what has been presented thus far, “what is knowledge research and what makes it so valuable?”, my answer (which is partly just a summary of the foregoing) can be given as five interconnected themes and desiderata. Systematic and reflective knowledge research is concerned to (i) conceptually elucidate and describe the peculiar profiles of various forms of knowledge; (ii) identify and investigate the points of overlap between the various kinds of knowledge forms; (iii) elucidate the mechanisms of interplay and interpenetration between various forms of knowledge; (iv) grasp and model the dynamic of various forms of knowledge and their interplay; (v) describe the practices and the manifestations of knowledge. Knowledge research thus defines its field in terms of the objects of its investigation. It aims to provide descriptions, analyses, therapy, and suggestions for modeling each of the five fields and to elucidate the mechanisms of the respective processes, states, and phenomena. And this enterprise is principally valuable because it is primarily these mechanisms which support the fluent function of human perception, speech, thought, and action as well as the whole triangulation10 of subject, other subjects, and world. 9 On this point, cf. Abel (2008b). 10 The expression “triangulation” was introduced into the philosophy of language and the philosophy of mind by Donald Davidson. It pertains, in particular, to the question of how the processes of human language acquisition and the capacity for propositional thought can be made intelligible. Cf. Davidson (1982); (1991a) and (1991b). In Davidson (1991a), 203, he introduces triangular relations by means of the example of relationship between a student, the teacher, and the object: one line connects the student with the object, another connects the student to the teacher, and a third connects the teacher with the object. It is this use of the term that I am invoking here. Nevertheless, my usage of “triangulation” does diverge somewhat from Davidson’s, inasmuch as I employ the term in a more capacious sense. In what follows, the triangular expres-

Knowledge Research: Extending and Revising Epistemology

9

We can no longer conceive of our task as one of setting various forms of knowledge against one another, or presenting them as mutually exclusive – though it goes without saying that a closer view of each form of knowledge will also bring to light their several differences. But such differences arise even within a single form of knowledge. For example, within ‘scientific knowledge’, we must distinguish between mathematical, physical, technological, and computational knowledge. Needless to say, knowledge research, as we conceive it, is by no means limited to the study of science (which investigates the production and reception of the sciences). Along with scientific knowledge, our approach to knowledge research takes account of all the other forms of knowledge and sets them on an equal footing. Scientific knowledge is a tremendously important and well established form of knowing. But it is hardly the one and only such form, nor the only metaphysically respectable one. It is not at all difficult to bring out the central role of sions “I – We – World” and “Subject/Agent – Other Subjects/Agents – Experienced World” are not restricted to merely drawing on the reaction of a second person in determining the meaning of a word and the propositional content of a thought. Rather, the “I – We –World” triangle has the more capacious sense of highlighting the primordial role that these interpenetrating relations play in human understanding of the world, others and oneself. Secondly, I am by no means satisfied with Davidson’s characterization of the effect of the object within this triangle as a “causal chain”. I do not share his conviction that “all these relations [between the student, the teacher, and the object] are causal” (Davidson (1991a), 203, my emphasis). Nor do I agree with Davidson that contents and states of speech and thought can be individuated solely on the basis of the “causal interplay” between speaker/thinker, other persons, and the objective world (Davidson (1991a), 204, my emphasis). If we are interested in a comprehensive theory of individuation – which is a desideratum not only of knowledge research but also of contemporary philosophy more generally – we must, in my view, acknowledge that there are far more ingredients in play and at work than merely causal relations. Føllesdal (1999) is quite right to insist on this point. The fundamental difference between Davidson’s view and mine principally lies in the fact that I conceive the triangular interplay and interpenetration of the I–We–World relation in accordance with the philosophy of signs and interpretation. That is, I conceive these triangular relations as based on constructive and perspectival semiotic and interpretational processes (and not in the sense of a causal theory). Understood in this way, triangular relations are constitutive across the whole spectrum from the phenomenal discrimination of sense impressions, the individuation of objects, perception, speech, thought, and action all the way to the higher-level and meta-level cognitive activities such as rule-following and the establishment and justification of standards and norms.

10

Günter Abel

practical, moral, and aesthetic knowledge for our understanding of the world, others, and ourselves. Moreover, it is of the first importance that we not reduce various forms of knowledge to a single form, as though there were ‘The One and Only Foundational Form of Knowledge’, and that we not seek out their common denominator in the hopes of reducing them to some third thing. In light of this irreducible plurality, and in the absence of a third term to serve as a common denominator, we are systematically (not just contingently) obliged to reject the view that there is ultimately just a single (e. g. scientific) perspective from which to consider epistemological questions and that everything that can be said about knowledge must be articulable from within that perspective. Yet I need hardly point out that this irreducible plurality does not by any means lead to relativism about knowledge forms. For there are strict theoretical, practical, and everyday restrictions on the degree of this plurality. It would seem that not everything that presents itself as ‘knowledge’ actually is knowledge – neither in the everyday, factical sense, nor in the theoretically grounded sense. We are just as subject to crises and breakdowns in everyday life as are our theories. This emphasis on the plurality of forms of knowledge by no means minimizes the clear contemporary importance of ‘scientific knowledge’ in today’s scientific and technological world. But the insight that this plurality is irreducible and ineliminable does require that we amend our scientific propositions with the qualification, “viewed and formulated from a scientific perspective…”. Scientific knowledge is not knowledge ex cathedra – definitive and absolutely binding for all finite thinking subjects. It is not only legitimate but of the utmost importance to emphasize the difference between scientific and non-scientific knowledge and to recognize both as genuine forms of knowledge. Furthermore, it is crucial to take seriously the question of how these two forms relate to one another both with respect to the phenomena (as opposed to the concept) of knowledge as such, and with an eye to a more comprehensive understanding of the forms, dynamics, and practices of knowledge. It is only by recourse to the reciprocal relations and interpenetrations of various forms of knowledge that we can gain insight into the possibility of radically new knowledge. All creative breakthroughs and revolutions – whether in our everyday world, in the sciences, in philosophy, or in the arts – are characterized in large part by the fact that they are not merely one-dimensional or purely intradisciplinary, but rather

Knowledge Research: Extending and Revising Epistemology

11

result from the incorporation of various different forms of knowledge.11 For example, in a creative process (which is not itself governed by any meta-rules) such distinct forms of knowledge as conceptual construals, imagistic representations, and deft technical practices often go hand in hand. And this hold true even of formal (e. g. mathematical, diagrammatic, or otherwise notational) symbolisms in the sciences. To take only one example, consider Friedrich August Kekulé’s famous benzene ring. In this case, it was the visual appearance of a certain chemical notation that suggested12 a new and surprising manner of continuing that notation, by transforming it into a ring-shaped structure, and thereby led to a creative new chemical formula. 3. Knowledge Research as a Reflective, Fundamental Investigation Knowledge research becomes necessary whenever the intradisciplinary and intraphilosophical problematics we face (not to mention the social challenges that confront us) can no longer be defined or resolved in a purely disciplinary manner. In the face of such challenges, systematic knowledge research reveals itself to be a transdisciplinary, reflective investigative enterprise.13 Such intradisciplinary problematics can be seen in various new fields (which have in fact recently sprouted up in various disciplines, sciences, and technologies) – fields such as e. g. the transdisciplinary research across the board on structures, processes, systems, materials, or imaging. Examples of such intraphilosophical problematics abound. Consider these two. (a) To determine the nature of ‘consciousness’ (or the ‘mind’), it does not suffice to merely list the findings of philosophy, cognitive science, neuroscience, psychology, computer science, linguistics, and other disciplines. The challenge rather consists in determining which subfields are so much as capable of contributing to answering the question. (b) In seeking an integrative philosophy of our reference to the world, others, and ourselves, one can no longer allow the ‘philosophy of mind’, the ‘philosophy of language’, and the ‘philosophy 11 I sketch the outlines of a philosophy of creativity in Abel (2009) and (2006). 12 At one point Nelson Goodman speaks of the “suggestive allusion” of symbols, that is, the “characteristic capacity of expressions for suggestive allusion, elusive suggestion, and intrepid transcendence of basic boundaries” (Goodman (1976), 93). 13 In speaking of “transdisciplinarity” I am invoking the use of this term that Mittelstraß (2003) introduced into the literature.

12

Günter Abel

of action’ to simply rest isolated alongside one another. The crucial task is rather to approach the problem from the perspective of a unified philosophy of mind, language, and action. To develop such a unified philosophy is, in my view, one of the cardinal challenges of contemporary philosophy. Well known examples of social issues that call for transdisciplinary solutions are problems concerning energy, the environment, climate change, health care, the financial system, and education. Against the horizon of such considerations, the fourfold moral of systematic and reflective knowledge research can be formulated quite economically. (i) It is the intraphilosophical, intradisciplinary, and intrasocietal challenges which call for the disciplines, theorems, reflections, and practices conducive to their solution, not the reverse. (ii) Systematic and reflective knowledge research can also have direct relevance for object-based studies and research. Its results can, for example, be integrated into the design of new experimental, theoretical, and reflective strategies and applied to existing systems and practices of knowledge and research or even incorporated into pedagogical approaches and teaching curricula. (iii) Knowledge research is a reflective discipline and, at the same time, a basic research program. (iv) Knowledge research is characterized by the discovery and development of pragmatic heuristics that are fundamental to our knowledge in both the narrow and the broad sense. That is, pragmatic heuristics fundamental to our perception, speech, thought, and action, and to the fluent function of the triangular I– We–World relation (in the broad, factical sense of knowledge), as well as those fundamental to our theoretical and linguistic-propositional constructions (in the narrow sense of knowledge). To the extent that knowledge research succeeds in accomplishing these tasks, it is capable of contributing to our orientation in everyday life, in the sciences, and in the arts. Indeed, helping to orient us in this manner constitutes a great deal of the humane significance of systematic and reflective knowledge research. But it should go without saying that such an approach also takes account of the normative aspects involved in the aforementioned processes, states, and phenomena. Knowledge of norms and one’s orientation is of the utmost significance to knowledge research inasmuch as it affects the whole spectrum of forms of knowledge – from the broad, factical forms, to the more narrow, propositional ones.

Knowledge Research: Extending and Revising Epistemology

13

4. What Knowledge Research is Not Limited To Another way to clarify the aim, value, and tenor of systematic and reflective knowledge research is to indicate those characteristics of classical epistemologies and theories of knowledge which precisely fail to fully capture it. Within the scope of this essay, we must content ourselves with a rough sketch of the seven most significant features, drawn from contemporary debates. Each of these seven insufficient characterizations of knowledge research calls for an extension and revision of epistemology. (i) Knowledge research is not armchair philosophizing about imaginary examples and counterexamples which, as a rule, are aimed at establishing watertight definitions. Quite the contrary; knowledge research defines its various fields by recourse to and in terms of the objects it investigates. Accordingly, it has no truck with the classical binary, which holds that philosophy has primarily to do with a priori methods, while the empirical sciences are characterized by their a posteriori methods. The advocates of this binary conception often emphasize that they wish to say nothing about the design of scientific experiments, theories, or hypotheses, and insist that an unbridgeable gulf separates philosophy from the sciences.14 Knowledge research expressly rejects this conception. Instead, knowledge research seeks out (as philosophy has done since its very beginnings) the points of contact and overlap between philosophy and the sciences. Such points of contact are valuable, not least of all, as occasions to engage with genuine philosophical questions and reflections which can then serve to help us better understand those points of contact. Hence, knowledge research is guided by questions like “how much philosophy does science need (and how much can it endure)?”, or “how much science does philosophy need (and how much can it endure)?”. Any attempt to make philosophy immune to science, or to make science immune to philosophy would be extremely disadvantageous to both sides. Philosophy cannot simply proceed as though all scientific investigation were completely irrelevant to questions regarding, e. g., the nature of ‘human consciousness’. Conversely, the sciences must not act as though there was no need to elucidate the fundamental concepts of, 14 Take for example the insistence of Bennett/Hacker (2003), in particular in chapter 14.4.1. Cf. also Dummett (2010), 4, where he claims that philosophy is the discipline “that makes no observations, conducts no experiments, and needs no input from experience: an armchair subject, requiring only thought.”

14

Günter Abel

e. g., neuroscience (concepts like ‘consciousness’, ‘perception’, or ‘memory’) – concepts originally borrowed from philosophy. (ii) Nor is knowledge research exhausted by ‘classical epistemology’, understood as the attempt to overcome skepticism through argument and as the establishment of doubt-resistant grounds and justifications for various claims. As we have already seen, these epistemological questions already presuppose that we do in fact engage with forms of knowledge. Knowledge research, by contrast, strives to describe, analyze, and elucidate such forms. It therefore constitutes a necessary precondition of classical epistemology, not just a another field alongside it. Moreover, most variants of classical epistemology are concerned with knowledge that is secure, or certain. But even contemporary debates continue to endorse Plato’s conception of knowledge as “justified true belief”. Many of the more subtle arguments in epistemology today still revolve, whether explicitly or implicitly, around the problem with this definition made famous by Edmund Gettier (who maintains that there are cases which fulfill these criteria, despite not being cases of knowledge).15 By contrast, knowledge research rejects the architecture of the Gettier problem and the discussions it inspired. Knowledge research rather focuses on (a) the phenomena of knowing (and not primarily the concept of knowledge), and (b) the more comprehensive conception of various forms of knowledge (conceived more broadly or more narrowly) as well as their forms of interplay and interpenetration – a conception which extends far beyond the framework of the Gettier problem. In this manner, knowledge research demands a significant expansion and revision of the epistemological problematic which was so narrowly constrained by the Gettier problem and the discussions it inspired. This expansion and revision can be seen to manifest itself in the following seven points. First, it comes out in the aforementioned transition from the narrow concept of knowledge to the broad sense of knowledge. Second, it becomes clear that we must also include non-linguistic and non-propositional forms of knowledge (e. g. practical, procedural, imagistic) in addition to the standard propositional and linguistic forms. Third, we must expand our considerations to embrace the whole spectrum – ranging from basic, factical, and paradigmatically embodied abil15 Gettier (1963). Shope (1983) gives a helpful overview of the first phase of rather narrow and rarely fruitful discussions of the Gettier problem. See also Bernecker/Dretske (2000), part I.

Knowledge Research: Extending and Revising Epistemology

15

ities to act, perceive, speak, and think to explicitly theoretical and propositional knowledge constructs in the narrow sense. Fourth, the irreducible plurality of forms of knowledge entails a plurality of forms of justification – that is, an expansion of our considerations from exclusively propositional reasons to forms of non-linguistic reasons (like visual or practical, procedural reasons as triggers of a person’s actions). Fifth, we must focus on various knowledge forms’ modes of interplay and interpenetration (not least of all the modes of interaction among nonpropositional knowledge forms, such as knowing-how and perceptual knowledge). Sixth, we must place new emphasis on issues of appropriateness, fitting, acceptableness, and plausibility inasmuch as these are relevant to the fluent function of our everyday, scientific, and artistic forms of knowledge. And finally, we must locate the mechanisms by which forms of knowledge operate and interpenetrate one another on the functional level of the triangular I–We–World relation. (iii) Knowledge research is by no means limited to conceptual analysis. Conceptual analysis is concerned with isolating the semantic components involved in the definition of the concept of ‘knowledge’. The ultimate aim of analysis is to produce a maximally complete list of the semantic features of the concept ‘knowledge’. Now, of course, knowledge research also engages in the fundamental activity of conceptual elucidation. For example, it seeks to shed light on the concepts of ‘perception’ and ‘consciousness’ as these are employed in various neuroscientific investigations. Yet as much as knowledge research engages the concept of knowledge, it is at least as focused on the phenomena of knowing and the mechanisms by which various forms of knowledge operate and interpenetrate one another. Pure and, as it were, antiseptic conceptual analysis runs the risk of losing sight of the factual, phenomenal mechanisms of knowing. (iv) Knowledge research is by no means exhausted by a theory of how to ascribe knowledge claims. Accordingly, it is not limited to elucidating the syntax and semantics of linguistic knowledge ascriptions, such as “Peter knows that p” or “Peter knows how to do x”, e. g. swim, or ride a bike. This analysis of linguistic ascriptions seems often to be motivated by the (dubious) presupposition that any person making such a ‘knowledge’ ascription must first establish a cognitive relation to the realm of propositional knowledge (i. e. the realm of Russellian propositions, Fregean thoughts, or possible worlds, etc.). Only once she has established such a relation is she supposed to be in a position to apply the relevant bit of propositional knowledge-that to a concrete case (e. g. a case of know-

16

Günter Abel

ing-how to ride a bike). But however questionable this picture is, its complement is surely wholly misguided. For cognitivist propositionalists often go on to assume that even the person to whom a knowledge of bike riding is ascribed must also establish such a cognitive relation to the realm of propositions, if she is to so much as be able to, e. g. keep her balance while pedaling her bike. Now I must confess that I do have a weakness for the strategy pursued by such theories of ascription, and have myself endorsed them in other contexts.16 But in considering the procedural and fluent functioning of practices and the forms of knowledge that manifest themselves therein, one must, in my view, hold strictly to the following point. In light of the broad sense of knowledge as basic and factical, and in view of the episodic character of mental states, finding oneself in a state of knowing can by no means simply be described in terms of the prior establishment of a cognitive relation of the aforementioned sort. Whether in its broad sense (in the context of practical and procedural abilities) or in its narrow sense as a cognitive mental state, knowledge must not be viewed as a cognitive relation of the sort sketched above. If we consider knowledge in the broad sense as a factical state of human praxis – that is, if we consider it within a framework of the facticity of abilities and actions – then we must grant that such knowledge can by no means consist in a cognitive relation and surely cannot presuppose one as a necessary condition. Rather, such knowledge must be viewed as a process and a state – as an embodiment of pragmatic abilities and as a phenomenon – whose various roles we must elucidate and clarify. Knowledge research is thus not primarily concerned with the linguistics of ascribing knowledge-statements. It is rather concerned with the phenomenon of knowledge in its various forms, practices, and dynamics, as well as with the factical role of knowledge forms in the fluent functioning of the triangular relation between the Subject/Agent, other Subjects/Agents, and the World. Determining the syntax and semantics of knowledge ascriptions is surely not an insignificant part of describing the processes by which this triangulation is effected – but it is clearly not enough. Moreover, such ascriptive statements owe both their success (when they do succeed) and the determinacy of their con16 Aspects of this approach play an important role in my general philosophy of signs and interpretation. Cf. the entry “Zuschreibung” (i. e. “Ascription”) in Abel (1995a) and (2004).

Knowledge Research: Extending and Revising Epistemology

17

tent to precisely the triangular ‘I–We–World’ relation which they always already presupposed and which constitutes the principle target of knowledge research. (v) Reflective knowledge research must not be misunderstood to be some kind of scientific program. That is, it must not be mistaken for a description of the phenomenon of ‘knowledge’ by means of strictly scientific methods, which conceives of ‘knowledge’ as a “natural kind” and makes it the object of scientific analyses.17 Such a conception of knowledge aspires to a form of ‘naturalism in knowledge and epistemology’, a form of “epistemology naturalized”.18 Now there is a sense in which reflective knowledge research does in fact seek to thoroughly naturalize ‘knowledge’ – i. e. view it as something that genuinely belongs to the facticity of mankind and, in particular, to the ‘I–We–World’ triangle. But this sort of naturalization which knowledge research seeks to realize would be seriously misunderstood if it were conceived as a scientific naturalization – that is, as a naturalism that only considers something to be an actual or real case of knowledge if it has a place within the (ultimately physical) ‘scientific worldview’. We can now venture a more sweeping thesis: in all its various forms, knowledge is everywhere characterized by a transcending (not a transcendent) aspect. We can even see this aspect already at work in formulations such as “knowing that…”, “knowing how…”, or “knowing why…”. Such ordinary judgment, ascriptions, and reports always go beyond the sensible experience and the practical performances which occasion and give rise to them. This is always a hidden surplus in the expressions “knowing that”, “knowing how”, and “knowing why”. But nota bene: a phrase like “knowing that” does not leave it open or underdetermined what it is about. However, it is precisely this aboutness which embodies that empirically underdetermined aspect which eludes scientistic naturalization. Across the whole spectrum – from its broad to its narrow senses – knowledge stubbornly resists all physicalist reduction, for it continually manifests a systematic and ineliminable “underdeterminacy”.19 The scientistic urge to eliminate this underdeterminacy, if one 17 This conception is to be found in Kornblith (2002). 18 Cf. Quine (1969); Dretske (1995). 19 Cf. Quine (1993), 96 – 102. On this, cf. also Abel (1995b) and (2010a). – Within the context of a philosophy of signs and interpretation, we can reformulate the underdeterminacy thesis as follows. Inasmuch as epistemic knowledge signs (of perception, speech, thought, or action) possess semantic and pragmatic features – and they clearly do – the physical underdeterminacy of those epistemic

18

Günter Abel

thinks it through, leads the scientist into what must be, for him, a hopeless epistemological dead end. For if this underdeterminacy were ever to be eliminated, nature would become her own epistemologist and scientific investigation as a whole would be reduced to a merely epiphenomenal activity. Strictly speaking, science would be wholly superfluous, and could be dispensed with. Scientism, when thought all the way through, would have to rule against itself – it must end up committing a form of cognitive suicide. (vi) Of course it ought to go without saying that knowledge research is not simply the theory or philosophy of science, if one conceives of the latter as a metatheory of our engagement with scientific theories, methods, goals, and instruments. The aim of the philosophy of science consists in reflecting from without, as it were, upon the point of view scientists occupy in generating their hypotheses, functions, theories in order to describe. It also seeks to describe (from without) the perspective of scientific pedagogy, from which scientists teach others how to construct proofs, collect and marshal evidence, confirm theories, and justify hypothesis. By contrast, reflective knowledge research expands the framework within which we approach science, by extending our considerations to more comprehensive, basic forms of knowledge, knowledge practices, dynamics, and their characteristic modes of interplay and interpenetration. Yet this extension is not to be conceived as a vertical reflective movement up to a higher standpoint. It is rather a horizontal reflection across the spectrum of phenomena and forms of knowing (where scientific knowledge is just one among others). The aim of this horizontal reflection is, as we have emphasized, to augment classical conceptual elucidation with investigations of the mechanisms by which various forms of knowledge operate and interpenetrate. Yet both of these activities should be conducted in such a way that their results remain focused on the object of study and retain some degree of pertinence to other investigations and reflexive epistemological reflections. In these ways, knowledge research circumscribes a much richer and broader scenario and field of study than does the philosophy of science. (vii) It is also important to appreciate that knowledge research is not exhausted by the history of knowledge or the history of the sciences. It is by no means limited to reconstructing and recounting the development of varknowledge-signs is internally inherited with those features. This underdeterminacy obtains across the whole spectrum of knowledge-signs – from the basic and factical forms to the more strictly linguistic, propositional ones.

Knowledge Research: Extending and Revising Epistemology

19

ious individual forms of knowledge and scientific disciplines over the course of their history. Knowledge research cannot primarily focus on the individual histories of certain forms or disciplines without undermining the emphasis it places on the irreducible plurality of different knowledge forms, practices, and dynamics and the multiplicity of their forms of interplay and interpenetration. Notwithstanding the tremendous importance of appreciating the historical and cultural development of such knowledge forms, practices and dynamics, knowledge research dedicates the bulk of its energies to systematically and reflectively investigating the mechanisms by which various forms of knowledge interact and interpenetrate. Note, however, that these investigations must undoubtedly take place from the perspective and within the framework of historical and cultural meaning and sense. To this extent, the history of knowledge and the history of science constitute integral parts of systematic knowledge research. Our forms and orders of knowledge did not simply appear out of thin air. Each of them has its own cultural history. And ignoring that cultural history would make it unintelligible why our forms and orders of knowledge are the way that they are. Yet at the same time, it is crucial to see that systematic knowledge research cannot simply be reduced to a history of knowledge. Correlatively, the epistemological import of knowledge research cannot be exhausted by a series of local narratives and practices or the historicizing model of epistemology that informs them.

II. Extending and Revising Epistemology 1. Knowledge Research and Epistemology As section I demonstrated, knowledge research precedes epistemology as its necessary precondition. Against the background of this claim, I would like, in what follows, to concentrate on the relationship between knowledge research and epistemology. Epistemology is not some closed field of study, whose principle themes and aspects have been determined once and for all. Yet as far as I can see, contemporary epistemology has thus far altogether failed to adequately train its attention on forms of knowledge or their dynamics, practices, or modes of interpenetration. In what follows I aim to make a case for doing just this. Now if what I urge is in fact plausible, it is easy to see that such forms of knowledge (along with their mech-

20

Günter Abel

anisms and modes of interpenetration) will not simply constitute new members of the existing epistemological realm. For to place such forms of knowledge at the center of our epistemological investigations calls for sweeping expansions and revisions of contemporary epistemology. And this entails that we can (and must) expand, shift, reexamine, and partly revise our conceptions of the spectrum of epistemological topics as well as our conception of who counts as an epistemologist. It seems to me that knowledge research calls upon us to proceed in a way that is coherent with contemporary epistemological conceptions even as we proceed beyond them. In section I.4 we emphasized that comprehensive knowledge research rejects the prevailing fixation on the Gettier problem. Large portions of epistemological debates have been held hostage by this problem for far too long. It has afflicted the literature with a certain Byzantine character. Moreover, the whole debate revolves around the narrow concept of knowledge as something one could give a definition of or criteria for. Knowledge is summarily characterized as justified and true, and all subsequent thought on the subject is limited to these two definitive conditions. Yet even within the literature on this narrow concept of knowledge, some authors (e. g. Crispin Sartwell and Ansgar Beckermann) have viewed it as a central difficulty that the traditional concept of knowledge cannot simultaneously and satisfactorily embody both components (truth and justification). If both truth and justification are treated as equally necessary conditions, knowledge becomes an “illegitimate hybrid-concept”. In keeping with this conclusion, the concept of knowledge itself would have to be abandoned as “incoherent”.20 Yet, first of all, this suggestions runs contrary to our intuitions about knowledge. Inasmuch as we are able to successfully orient ourselves in the world and in relation to other people – that is, inasmuch as we are able to speak a language, to think, to act and are, in general, able to exercise our various abilities – we are quite justified in our conviction that we possess basic, factical knowledge in the broad sense. Second, even this suggestion has not yet freed itself from the stranglehold of the narrow, classical concept of knowledge. Methodologically speaking, the suggestion is just another conceptual analysis of knowledge, albeit one engaged in a further debate about what an adequate conceptual analysis is supposed to look like. Since, however, section I 20 Cf. Beckermann (2001) and Sartwell (1992).

Knowledge Research: Extending and Revising Epistemology

21

has demonstrated that this narrow concept of knowledge cannot be the primary one, the whole mode of argument about this point is quite limited in its effect. It dissipates into nothing as soon as we shift our epistemological focus to the broad sense of knowledge and to the whole spectrum of various knowledge forms that regularly confront us in the sciences, arts, and in everyday life. The broad sense of knowledge – according to the thesis we have advanced here – is genealogically prior to the narrow sense and constitutes one of its necessary conditions, not the reverse. In section I, I described the broad sense of knowledge more particularly as a primordial and foundational state. This state is a facticity that is, in varying degrees, proper to human “Being-In-the-World”21, to all intersubjectivity between persons, and to the world disclosed by such intersubjectivity. And such a state cannot for its own part be subsequently analyzed into various constituents. I see here a certain family resemblance between the status of the broad concept of knowledge and the status that Peter Strawson seeks to reclaim for the concept of a person.22 Like the concept of a person, the broad, basic, factical concept of knowledge cannot, strictly speaking, be the object of exhaustive conceptual analysis. For what is at issue in the broad concept of knowledge is precisely a non-reducible, non-analyzable concept, which rather serves as the basis for all conceptual analysis in the narrow sense, inasmuch as all conceptual analysis already presupposes and relies upon it. In a certain sense, I share Timothy Williamson’s view that knowing should be conceived as “the most general factive mental state”.23 More particularly, I share Williamson’s view that knowledge is primordial in the sense that it neither must nor can be analyzed in terms of other concepts (as has traditionally been attempted using the concepts of ‘belief’, ‘truth’, and ‘justification’). All attempts to do so have failed. And the “natural explanation” of this failure is simple: “knowledge has no such analysis”. Knowing rather serves to help us analyze and explain many other things, but “not as something itself to be explained.” Such a view of the matter “reverses the direction of explanation pre-

21 Heidegger (1927). 22 Cf. Strawson (1959), ch. 3. 23 Williamson (2000), 33 ff.

22

Günter Abel

dominant in the history of epistemology.”24 I wish to expressly endorse such a reversal of the direction of explanation in epistemology. In light of this reversal, reflective knowledge research might equally well be characterized as an attempt to describe and shed light upon this newly altered epistemological terrain. It is not an analysis of the concept of knowledge, as the latter has been traditionally characterized in terms of various theoretical constituents – justification, truth, belief. It is an important aspect of the broad sense of basic, factical knowledge that this concept is deeply and firmly anchored in our everyday language, our life-world, and our ordinary practices. We are as intimately familiar with the language games associated with the concept as we are with its corresponding actions and practices. And in light of this fact, it makes good sense to ask whether the well-known classical proposals and the variants that arose as responses to difficulties with the classical proposals really exhaust our options here. One might well ask whether there might not be alternative theories of knowledge which are neither committed to the classical three-term model of knowledge (justified–true–belief) nor to the traditional methods of conceptual analysis. The answer is twofold. First, there are important approaches in roughly this direction which afford points of connection as well as points of contrast with systematic, reflective knowledge research. A few examples are pragmatic theories of knowledge (e. g. Edward Craig’s), naturalistic theories (such as Hilary Kornblith’s), and formal epistemologies (such as Vincent F. Hendricks’s).25 Second, from an epistemological perspective, systematic and reflective knowledge research understands itself to be a contribution to the field of such novel orientations. At the same time, however, it brings certain aspects of knowl24 Williamson (2000), V. – However, Williamson does not place the basic form of knowledge in the broad sense at the center of his considerations. He is rather focused upon the narrow kind of knowledge which is linguistic and propositional. However legitimate this restriction of focus may be, to the extent that Williamson makes claims about this narrow form of knowledge which, in my terminology, should rather be applied to the broad concept of knowledge, I must take issue with the dominance he accords to cognitive (viz. propositional) knowledge. For in granting it such prominence, he threatens to reverse the proper genealogy between the broad and the narrow senses of knowledge. I have discussed this difference between Williamson’s view and my own with reference to cases of knowing-how in Abel (2011b) and (2010b). 25 Cf. Craig (1993) and (2002); Hendricks et al. (2011).

Knowledge Research: Extending and Revising Epistemology

23

edge into view that call for expansions and revisions of contemporary epistemology. The principal way in which knowledge research effects these changes is by freeing us from the exclusive dominance of the model of ‘belief’ and the traditional form of ‘autonomous justification’.26 2. Knowledge Research and Epistemological Methodology Systematic and reflective knowledge research can make important contributions in the context of epistemological methodology (and hence with regard to questions about the structure of the ‘epistemological justification of our beliefs’ and questions regarding the ‘sources of epistemic knowledge’). The four most important methodological positions in 26 In using the term “revisions”, I not only mean to invoke the traditional, modern paradigm for a theory of justification. The latter paradigm principally consists (as Grundmann has convincingly demonstrated in Grundmann (2003), 9 – 29) in the combination of the following two elements: (a) the possibility of a priori knowledge (i. e. knowledge without recourse to empirical facts or empirical investigation), and (b) the autonomy of the subject (who is, within the privileged first-person-perspective, capable of knowledge in virtue of knowing how to justify her beliefs). Grundmann defines “epistemic autonomy” as follows: “a subject is epistemically autonomous just in case she is capable of determining which of her beliefs are justified (i) completely on her (independently from external authorities), (ii) independently of any empirical views about the world, and (iii) on the strength of purely rational considerations” (Grundmann (2003), 17). Both foundations of the classical theory of knowledge (a priori knowledge and the autonomy of the subject) have notoriously come under attack from various different directions. They have quite rightly come to seem rather questionable. The reaction in epistemology to these attacks was split: philosophers either proclaimed the collapse and the end of the theory of knowledge or switched from the level of substantial epistemic theory to the level of “metaepistemology” by shifting their focus to the clarification of the meanings of epistemic concepts, and, in particular, the establishment of necessary and sufficient conditions for knowing (cf. here Fumerton (1995)). This was and remains today the preferred path in broad swaths of analytic epistemology. – Though I consider this move to the metaepistemological level to be an intellectually interesting move, it is still just a way of running away from the real problem. It is a stance that prefers armchair philosophizing to grabbing the bull by the horns and turning to investigate the mechanisms by which our various knowledge forms, their dynamics, modes of interpenetration, and practices actually function. Systematic and reflective knowledge research takes this latter route. In this context, to speak of ‘revisions’ as I just have does not just serve to distinguish our enterprise from the traditional justificatory paradigm. The distinction runs much deeper, for it pertains to the very domain, the focus, and the architectonic of the whole epistemological enterprise itself.

24

Günter Abel

contemporary epistemology are: (a) coherentism, (b) foundationalism, (c) pragmatism, and (d) naturalism. These four positions can be roughly sketched as follows. (a) ‘Coherentism’ signifies the position according to which the justification of a belief consists in its entrenchment in a lattice of other beliefs – i. e. in its relation to other beliefs. A prominent example of this position can be found in the work of Donald Davidson.27 (b) ‘Foundationalism’ signifies the position according to which knowledge and justification rest upon a presupposed basis (e. g. of perceptions, or of self-evident beliefs), which forms the foundation for the generation of all (other) knowledge and epistemological justification. Aside from its 20th Century representatives like Roderick Chisholm, it is only natural to mention the classical locus of modern foundationalism – namely Descartes, in his search for a secure foundation for knowledge and justification.28 (c) ‘Pragmatism’ signifies the position according to which every kind of knowledge (as well as all epistemological justification) can be traced back to and explained in terms of capacities for action and the effects of such actions. Prominent classical pragmatists include such figures as William James and Ludwig Wittgenstein, and today authors such as Edward Craig, Hilary Putnam, and Robert Brandom.29 (d) ‘Naturalism’, in an epistemological context, signifies the position according to which ‘knowledge’ is a ‘natural fact’ and consequently an object of natural scientific investigation. Such scientific investigation is taken both to explain the generation of knowledge as well as to provide the requisite epistemological justifications. In contemporary debates naturalism is championed by numerous authors, most of whom champion either a physicalist or an evolutionary biological perspective. Prominent figures include Hilary Kornblith, Ruth Millikan, David Papineau, and Fred Dretske.30 Systematic and reflective knowledge research, with its focus on fundamental conceptual elucidation and on the interplay and interpenetration of various forms of knowledge, seeks to incorporate the positive aspects of each of these four principal methodological positions, yet with27 Cf. Davidson (2001b). 28 Cf. Chisholm (1982); Descartes (1986). 29 James (1979); Wittgenstein (1984); Craig (1993); Putnam (1995); Brandom (2008). 30 Kornblith (2002); Dretske (1995); Millikan (1984); Papineau (1993).

Knowledge Research: Extending and Revising Epistemology

25

out saddling itself with their attendant difficulties. This is clearly an ambitious undertaking. What follows is a rough outline of the dimensions along which knowledge research seeks to align itself with, or decisively diverge from these prominent epistemological methodologies. (a) Coherentism: In the first place, the various forms of knowledge must, of course, be in frictionless agreement among themselves and also be able to sustain the fluent continuation of actual perception, speech, thought, and action. Yet the fluent functioning of knowledge forms is simultaneously subject to the conditions of empirical validity. By invoking such conditions, knowledge research neutralizes the danger that threatens traditional coherentism – namely the danger of descending into a groundless coherentism of beliefs, an involuted but vacuous ceremony of mere coherence. (b) Foundationalism: It is one of the fundamental assumptions of knowledge research that all manifestations of knowledge forms are connected to semiotic and interpretational functions. Knowledge without signs is not knowledge at all. And this point holds equally well, whether one is speaking of knowledge in the broad sense (basic, factical knowing) or in the narrow sense (linguistic, propositional knowledge). Semiotic functions and corresponding interpretational functions suffuse the whole spectrum of knowledge forms. They are manifest in the practices, dynamics, and modes of interpenetration proper to various forms of knowledge – indeed, if they were not so manifest, we could not even bring into view specific forms of perception, speech, thought and action, nor could we recognize an experience or a representation as actual. In this sense, the semiotic and interpretational functions are fundamental to our forms of knowledge and their modes of interplay. But they are not fundamental in an ultimate and metaphysical sense, but rather in an interpretational, semiotic-pragmatic sense. Thus, for example, semiotic and interpretational functions are, for their own part, culturally and historically conditioned; they are modifiable, alterable, revisable, temporal. In a word, they constitute our pragmatic semiotic and interpretational foundations for the time being. (c) Pragmatism: Reflective knowledge research agrees with the pragmatist that the sources of knowledge and its justification are not to be sought in some realm of theoretical propositions or in their inferential relations. Knowledge research rather seeks these sources in the fact that we are minded creatures who can act – i. e. in the fact that we are agents in the world who are active in relation to the world, to other people, and to ourselves. Such actions are essentially involved in

26

Günter Abel

the facticity of the I–We–World triangle. To this extent, actions as well as procedural and practical abilities and competencies are of the utmost relevance to epistemological methodology. But critical knowledge research does not go on to commit what I would like to call the practice-centric fallacy, which consists in the further claim that knowledge is nothing but the performance of practical skills and which maintains that knowledge can ultimately be equated with actions. No such equation is possible (or even intelligible) with respect to either of the two senses of knowledge we have distinguished – the broad sense of basic, factical knowing, or the narrow sense of propositional knowledge. Knowledge can indeed be conceived as a mental state or process that both depends on action and disposes one to act. (This would enable and require us to distinguish knowledge from other mental states such as remembering, or being convinced of something.) But it simply does not follow from this that knowledge can be equated with action. Accordingly, reflective knowledge research by no means seeks to reduce theoretical knowledge-that to practical or procedural knowledge-how, or to derive the former from the latter. It is true that a particular piece of knowledge can be individuated in terms of its practical and action-initiating consequences or effects, and that these effects enable us to evaluate and examine it in light of our praxis. But this clearly does not mean that knowledge and action are the same. It rather points up the necessity of a unified theory of knowledge and action.31 (d) Naturalism: Systematic and reflective knowledge research does indeed advocate, as noted above, a naturalization of knowledge (viz. of knowledge forms). That is, it urges us to view forms of knowledge as natural phenomena in the basic sense that they belong to us as human beings. But knowledge research avoids the danger of scientistic reductionism. For knowledge research does not assume that the scientific form of knowledge is the only essential and metaphysically respectable form. In place of the old representation of knowledge forms as constituting a pyramid, with scientific knowledge at its pinnacle, knowledge research recommends viewing different forms of knowledge as distributed along a spectrum. Moreover, it underscores the fact that all our knowledge (and any interplay between knowledge forms) will be physically underdetermined.32 We can be sure that the underdeterminacy at 31 For a few suggestions in this direction, cf. Abel (1999), ch. 13: “A Unified Theory of Knowledge and Action”. 32 Cf. section I.4. above and fn. 19.

Knowledge Research: Extending and Revising Epistemology

27

work here is strictly ineliminable, because forms of knowledge are, at all times and in all places, exercised and articulated in and through semiotic and interpretational functions. And the semantic and pragmatic features of such functions systematically (not just contingently) elude all attempts at physical reduction. There simply could not be knowledge-signs that were not physically underdetermined. Or, formulated differently: the proposition that there are non-underdetermined signs is itself not free of underdeterminacy. In this respect (inter alia), knowledge research distinguishes itself from a scientistic naturalism of the sort defended by Hilary Kornblith or (even more extremely) Paul Churchland.33 3. Knowledge Research and Epistemic Objects Epistemology can no longer merely be understood as a meta- and intratheoretical reflection occurring primarily in the minds of individual epistemologists. As we have emphasized throughout, it is crucial to expand and externalize epistemology in order to bring the triangular I–We–World relation into view, and to engage in epistemological reflection from the perspective of this triangular relation between Subject/Agent, other Subjects/Agents, and the World thereby disclosed. It is impossible to elaborate a comprehensive and satisfactory epistemology without taking account of two things. First, we must take account of other subjects/agents (i. e. we must consider the social dimension of epistemology). And second, we must take account of the material constitution of the world knowledge discloses (i. e. we must consider the material dimension of epistemology). It is in this sense that I would like to call for an ‘extended epistemology’ in the name of knowledge research.34 One might also call this extended epistemology ‘three dimensional’ or ‘3-D epistemology’. For its aim is to get clear about the mechanisms by means of which the triangular relations interpenetrate.35 33 Cf. Churchland (1985) and (1989). 34 In using the expression ‘extended epistemology’ I mean to recommend an expansion of Clark and Chalmers’s (1998) apt phrase “extended mind” from the level of mind and cognitive processes to the level of reflective epistemology. In doing so I correlatively endorse a transition from considerations on the level of “cognitive extension” (Clark) to the level of ‘epistemological extension’, where the latter applies to the whole I–We–World triangle. Clark and Chalmers (1998) is reprinted in Clark (2008). 35 3-D epistemology not only advocates externalism in the anti-individualistic sense made popular by Hilary Putnam and Tyler Burge. (Cf. The primary two texts that incited the internalism/externalism debate: Putnam (1975) and

28

Günter Abel

Clarifying the concept of an ‘epistemic object’ is of fundamental importance not only for our extended epistemology but also for more comprehensive knowledge research, of which it forms a part. An ‘epistemic object’ is that toward which we direct our epistemic curiosity and attention – i. e. our knowledge-oriented attention. And such attention and interest pervades the sciences, philosophy, the arts, even our everyday life-worlds. ‘Epistemic objects’ are thus objects of intellectual curiosity in theory and in practice.36 To briefly consider the sciences, their epistemic objects populate the whole spectrum from elementary particles through molecules, genes, brain mechanisms, mathematical nodes, all the way to the galaxies of astrophysics. In philosophy, the spectrum’s Burge (1979)). Their form of externalism advances the thesis that the reference of our words (e. g. “water”) is not solely determined by an inner psychical state of the speaker, but rather by the material constitution of the objects to which the thought or expression refers (in the case of “water”, by the chemical microstructure of water). But this only involves a passive dependence of the reference of our words and thoughts upon the material aspects of the objects referred to. Systematic and reflective knowledge research (and with it, 3-D epistemology) goes beyond this internalism/externalism debate in the following three respects. (i) The triangular relations that obtain between I, We, and World are not passive but highly active relations of reciprocal interpenetration. (ii) These relations and their active character are involved not only in the reference-based determinacy of words and thought. They are fundamental for all aspects of our understanding of and relation to the world, others, and ourselves. (iii) I do not understand these triangular processes as primarily or exclusively ‘causal relations’ or ‘causal chains’. I rather understand them, as I emphasized above, as constructive and perspectival semiotic and interpretational processes. I discussed semiotic processes above in section I.2., footnote 10, in the course of criticizing Donald Davidson’s inclination to restrict our considerations to causal chains. Both points of criticism (semiotic process; causal relation) pertain equally well to Clark and Chalmers’s position on the “extended mind” (Clark (2008), 222 f. and 79). In response to them, too, I would insist on the following points. (a) The triangulation of I–We–World cannot be reduced to causal relations and cannot be conceived in causal terms (at least not solely in causal terms). (b) The I–We– World triangle is to be conceived on the basis of semiotic processes. (Within a philosophy of signs and interpretation, this leads us to conceive and spell out the various components of the I–We–World triangle in terms of different semiotic and interpretational functions.) But we cannot pursue this point further here. It constitutes the guiding thought in Abel (1995a), (1999), (2004) and (2011a). 36 For more on the features of epistemic objects and their role in contemporary epistemology, cf. Abel (2008a) and especially Abel (2010a). The expressions “epistemic things” and “epistemic objects” have already attained some prominence in the history of science through the work of Hans-Jörg Rheinberger (2001) and (2006).

Knowledge Research: Extending and Revising Epistemology

29

bandwidth reaches from, e. g. sense impressions, through images, perceptions, meanings, referents, representations, thoughts, all the way to the constructions of understanding and reason (such as causality and justification). Both in our theories of and our practical engagements with epistemic objects, we seek to orient ourselves by learning more about their roles in our triangular understanding of the world, others, and ourselves. But connected with the question of the characteristic features of epistemic objects are two further questions – questions regarding (a) the possibilities and limits of knowledge, and (b) the modification and transformation of orders of knowledge. In light of these questions, it becomes clear that, and in which sense, the concept of ‘epistemic objects’ plays an important role in knowledge research. Three principle aspects should be emphasized here. (i) Knowledge research is, in the sense elaborated above, of direct epistemic and epistemological relevance. In both respects, knowledge research thus incorporates questions regarding the profile and the status of ‘epistemic objects’. (ii) The ‘epistemic objects’ proper to knowledge research are precisely the forms of knowledge (in addition to its practices and dynamics) as well as their modes of interplay and interpenetration. (iii) Knowledge research, in the sense elaborated above, is focused on the fluent functioning of and the mechanisms proper to forms of knowledge in the I–We–World triangle. This focus decisively informs the specific character of the ‘epistemic objects’ of knowledge research. As to the aspects of that triangle, I would like to particularly emphasize the triangular interplay between the relevant epistemic aspect, the material or objective aspect, and the aspect of intersubjectivity. This three-term conjunction, or three dimensionality, can quite naturally give rise to partly willful inflections of ‘epistemic objects’, e. g. hybrid or virtual objects. Inasmuch as the three poles or dimensions of this triangle inform the sense of ‘knowledge’, even the comprehensive concept of knowledge must bear this triangular and hence hybrid character. From the perspective of 3-D epistemology, however, this is not a drawback, but rather one of the advantages of the concept. Knowledge is not just an intrapsychological or purely mental state (like, e. g. remembering), for it involves, in an extensional sense, aspects of action, material, and intersubjectivity as genuine components.37 In light of this dimension, the epis37 This sense of “hybrid” is to be strictly distinguished from the sense mentioned

30

Günter Abel

temic objects of knowledge research (in this extended scenario) give rise to extensions and revisions of the space of epistemic things more generally (and, correlatively, of the space of epistemological reasons). This relation of epistemic aspect and objective aspect, of epistemic objects and material objects (and generally, of epistemic signs and material reality) forms a circular structure that loops back on itself. And this structure characterizes, for example, knowledge in physics – that is, the language and logic of physics – and particularly that of quantum physics. Thus a mathematical formula, for example, denotes, describes, and articulates, e. g., the state of a physical system through (and in virtue of) signs – e. g. through mathematical magnitudes, such as vectors. But to go so far as to strictly separate semiotic functions from states, occurrences, and processes that are (supposedly) wholly independent of signs and interpretation inevitably leads (at least from an epistemological perspective) to well known and ultimately disastrous difficulties. On the approach knowledge research takes to epistemic objects, it is important to free our whole investigation from the stranglehold of the dichotomy between ‘material naturalism’ and ‘conceptual or linguistic idealism’ by modeling epistemic objects on the level of basic, factical semiotic and interpretational relations. The result of this shift points towards a non-dualistic semiotic and interpretational theory of epistemic objects,38 which quite explicitly leaves behind the sterile dichotomy between ‘constructivist idealism’ and ‘metaphysical realism’. Clearly, semiotic and interpretational systems have an essential and far-reaching influence on the manner in which we encounter, construct, articulate, acquire, communicate, and apply knowledge in our in II.1, fn. 20, which discusses A. Beckermann’s view that the concept of knowledge is an “illegitimate hybrid-concept” insofar as it seeks to simultaneously invoke both truth and justification as its definitive conditions. Here, however, in the context of extended epistemology or 3-D epistemology, we are dealing with a wholly positive sort of hybrid character. Namely the sort manifested by objects of knowledge and expressed in the triangular relation of the relevant epistemic aspect, the material aspect, and the intersubjective aspect. 38 I develop the basic contours of such a theory in Abel (2010a), and I will not recapitulate my arguments here. It is worth emphasizing, however, that I conceive and model the triangular relation between I/subject, other persons/subjects, and the disclosed world as a relation of semiotic and interpretive processes. To this extent, semiotic and interpretational relations enjoy a fundamental place in knowledge research as well. But in the present essay, I can only adduce a few of the relevant aspects.

Knowledge Research: Extending and Revising Epistemology

31

perception, speech, thought, and action. Think, for example, of the aforementioned role of mathematical notation in physics, or of the role of pictures and images in the sciences (e. g. in ‘magnetic resonance imaging (MRI)’). The interplay of different semiotic and interpretational systems deserve a great deal more attention than they have thus far received in philosophy and the sciences. In considering the sciences, knowledge research seeks to investigate the important influence exerted by sign- and interpretation-based tools and instruments. Consider, for example, chemical formula, classificatory tables and charts, MR-imaging, computer graphics, computer simulations, and computers themselves as symbolic machines (in the sense that they compute algorithmic calculations and logical operations). It goes without saying that considerations and analyses of systematic and reflective knowledge research (and indeed of an extended 3-D epistemology) must take account of these domains and instruments. The proposal that there might be processes of human knowledge and experience that were completely free of signs and interpretation is strictly speaking unintelligible – whether we take knowledge and experience in the narrow, theoretical sense, or in the broad, factical sense. Nor is this claim about the semiotic and interpretational dependence of all human knowledge vitiated by the fact that semiotic and interpretive functions are subject to historical and cultural change, and that their semantic features can shift over time. Quite the contrary, this fact about historical and cultural change casts this dependence into greater relief by drawing our attention to the historically and culturally specific conditions of semiotic and interpretational processes. In all contexts and across the whole spectrum of ‘knowledge’ and its ‘epistemic objects’, we everywhere rely upon the two fundamental capacities to discriminate and to individuate. The very ability to talk about knowledge and its epistemic objects (or, about epistemic justification) always presupposes the discrimination and individuation of objects and events as objects and events. Providing a satisfactory theory of discrimination and individuation is therefore a central desideratum of knowledge research. The most elementary form of discrimination (and thus the most elementary form of knowledge in the broad, basic, factical sense) is sensory – namely, the ability to distinguish various sensible qualities (such as colors, tones, gestures, or tactile impressions, but also states like pleasure, pain, fear). The first thing to be emphasized here is that the abilities to discriminate and to individuate do indeed have the status of basic, fac-

32

Günter Abel

tical knowledge in the broad sense. Accordingly, it is of the first importance that discriminatory and individuating capacities are already in operation before our intellectual classifications and our conceptual capacities (i. e. before knowledge in the narrow sense) begins to function. For only then is it possible to move from the objects thus determined (through discrimination and individuation) to opinions, beliefs, or even whole theories and to correlate these components with the appropriate epistemological justifications. The fact that these pre-theoretical abilities and forms of knowledge (in the broad sense) do in fact exist and function can easily be seen in the behavior of toddlers and small children. A toddler is quite capable of phenomenally discriminating and individuating objects even before she learns the corresponding language games and conceptual functions. The child has the sensory capacity to distinguish, group, identify, and re-identify sensible objects as well as the capacity to sense pre-intentional differences between musical notes, colors, and consistencies. 4. Knowledge Forms’ Modes of Interpenetration A taxonomy of forms of knowledge is comparatively easy to produce, and it may even be possible to establish broad agreement about its basic form fairly quickly. The mechanisms by means of which forms of knowledge interact and interpenetrate one another, by contrast, remain altogether unexplored. Such mechanisms are, however, intimately familiar to us in as much as we encounter them in the practices of everyday life, science, art, and knowledge. As a rule, they perform their functions fluently. Indeed it is only when they break down (or we are practicing philosophy) that we tend to explicitly ask questions about them at all. We principally come to realize how scanty our acquaintance with the mechanisms by which our forms of knowledge interweave and interact really is only after their acute breakdown and in the course of our attempts to reestablish their fluent functioning in our perception, speech, thought, and action. Indeed, it is primarily then that we come to appreciate their situatedness in our practices and life-worlds. Systematic and reflective knowledge research views the elucidation of these mechanisms as one of its foremost challenges. In the following, I wish to clarify the nature of this challenge by considering it in the light of six more tangible examples.

Knowledge Research: Extending and Revising Epistemology

33

Example 1: Conceptual and Non-Conceptual Knowledge Knowing that the nearest mailbox is up by the next intersection means being in possession of propositional knowledge-that, and is a case of conceptual knowledge. Conceptual knowledge is bound to its articulation in a language and must be communicable in language. Both of these features are particularly manifest in scientific languages (e. g. in the language of physics), but they are also to be found in (and important to) ordinary language.39 Forms of non-conceptual knowledge include, for example, knowing how to open a car door, bake cinnamon bread, or ski, or knowing what it is like to have a sensation of color, or to feel jealous. The former variants of non-conceptual knowledge come from the realm of practical or procedural knowing-how. The latter variants belong to the realm of sensory-aesthetic experience and thus constitute a type of knowledge one can only acquire through firsthand experience, not simply pick up in a lecture hall or from a theoretical understanding of the phenomena. Closer examination of our actual engagements with knowledge (across the whole spectrum) quickly shows that, in reality, we are by no means confronted with just a ‘single pure’ form of knowledge. The actual phenomena, processes, and states actually involved in our knowledge (in everyday life, or in the arts and sciences) already involve a fusion of different forms of knowledge. It is only by means of post factum (and post phenomenon) heuristic reflection that we subsequently analyze the phenomena, processes, and states into distinct constituents. But this holism intrinsic to forms of knowledge can be heuristically differentiated as the interplay and interpenetration of conceptual and non-conceptual knowledge. (Indeed, any sufficiently fine-grained examination 39 My descriptions of both example 1 (“conceptual and non-conceptual knowledge”) and example 2 (“explicit and implicit knowledge”) are indebted to discussions and considerations raised in the context of my research group at the “Innovationszentrum Wissensforschung (IZW) / Center for Knowledge Research” at the TU in Berlin (cf. www.wissensforschung.tu-berlin.de). These discussions were instrumental in preparing for the colloquium “Das Wechselspiel der Wissensformen: begriffliches und nicht-begriffliches, explizites und implizites Wissen” [“The Interplay of Knowledge Forms: Conceptual and Non-Conceptual, Explicit and Implicit Knowledge”] held in Berlin, July 8 – 9, 2010. I would like to thank the members of this group, and in particular express my gratitude to Stefan Tolksdorf, Claudio Roller, Doris Schöps and Peter Remmers.

34

Günter Abel

will reveal this holism within each individual form of knowledge, including conceptual or non-conceptual knowledge.) Examples of such modes of interplay and interpenetration are easy to adduce. Think, for example, of how a cellist must combine her non-conceptual aesthetic experience, her non-conceptual knowledge of tonal coloration and expression, her practical knowledge of how to manipulate her instrument with her conceptual knowledge of the structure and peculiarities of the notation and score of, e. g., György Ligeti’s 1966 concerto for cello and orchestra. Or one might consider the everyday situation of preparing a tasty meal. Here the cook has to combine his non-conceptual gustatory experience and his non-conceptual knowledge of the combination of ingredients with his explicitly conceptual knowledge of the cookbook recipe and the linguistic formulations of the instructions and rules it contains. It is only in combination of these components that saffron rice and marinated mango chicken can turn into a delicious dish. Yet one could equally well consider what takes place in, say, the laboratory of an experimental physicist. In addition to her explicit conceptual knowledge of the theoretical setup of the experiment and of the hypothesis that is actually to be tested, non-conceptual capacities are also drawn on in performing the experiment itself. The latter might include the technical and practical operation of the material experimental apparatus (e. g. technological equipment and computers). But one would also have to include the capacity to visually read off significant signs from the computer screen and, indeed, the whole range of non-conceptual knowledge of the tensile strength or permeability of the materials used in the experiment. And the latter forms of knowledge belong to our pre-intentional attitudes in coping with things and cannot be addressed as forms of conceptual, linguistic knowledge-that. Or one might look at a mathematician and consider how she must be able to combine her explicitly conceptual mathematical knowledge with her non-conceptual knowledge of how to apply certain strategies in order to successfully construct a proof. One might even consider how a competent speaker of English must combine her invariably non-conceptual knowledge and practical ability to speak with the conceptual knowledge of the semantic content of English words and sentences.40 The example of the competent language

40 Cf. Devitt (2006); see also Abel (2012), ch. IV: “Knowing-How als Sprach-

Knowledge Research: Extending and Revising Epistemology

35

user is especially fruitful in our context because we have here a situation in which the very linguistic capacities which constitute the basis of all conceptual knowledge (in the narrow sense) themselves clearly involve and presuppose non-conceptual dimensions. Non-conceptual knowledge is thus prior to and a condition of conceptual knowledge. The ability to follow a rule in speaking a language is (as Wittgenstein rightly insists) not to be understood as though competent speakers possessed some conceptual or theoretical knowledge of predetermined and fixed semantic rules and their content.41 Now this conclusion naturally does not alter the fact that explicit conceptual knowledge can conversely affect and modify the non-conceptual competencies and abilities involved in speaking a language. Systematic knowledge research also calls for more fine-grained considerations of the modes of interplay between different variants of conceptual knowledge among themselves, as well as the modes of interplay between variants of non-conceptual knowledge. In the latter case, however, the mechanisms of interaction and interpenetration are virtually unknown. Nevertheless, we have no difficulty in giving examples of such modes of interplay from everyday life, or from the arts and sciences. A simple but telling example of such interplay between different conceptual forms of knowledge in the sciences is the interaction of ‘mathematical’ and ‘physical’ knowledge in physics. Both forms of knowledge are conceptual, and their interaction has famously generated some of the greatest achievements of modern natural science. Other examples of such interaction present themselves when we consider everyday knowledge-that and scientific knowledge-that. Surely it is fairly trivial to observe that we live in a world pervaded by science and technology. The effect of science and technology on our ordinary life-world is thoroughgoing and nearly omnipresent. But on the other hand, we must not lose sight of the fact that the priorities of our everyday lives are capable of stimulating and determining the shape of scientific research. I would like to present two examples of how knowledge proper to our life-world can exert influence upon scientific knowledge. (a) Consider for a moment the scientific medical research aimed at combating and curing diabetes. Such research is clearly driven by a und Zeichenkompetenz [Knowing How as Linguistic and Semiotic Competence]”. 41 On the issue of rule-following cf. Wittgenstein (1984), §§ 198 – 242, and the contributions to the volume Miller and Wright (2002).

36

Günter Abel

fact of our everyday life: namely that many people and indeed whole segments of the population suffer from this metabolic disease. The research is connected to and motivated by problems in our life-world. If it were not a norm of our life-world to combat diabetes mellitus, there would not be any diabetes research. Correlatively, aspects of our life-world can also lead us to abandon certain scientific medical research. Since very few people in the life-world of Europe and North America suffer from Huntington’s disease, there is very little scientific research done on this hereditary illness of the brain. (b) If one’s scientific research has to do with the phenomena, states, and processes of color-perception and the perception of movement (a field that will concern us further in Example 4, below), then one will often have occasion to refer to the brain areas V3, V4, and V5. Here, neuroscientific knowledge consists in a scientific and conceptual description of the stimulus thresholds of these cerebral areas during the times the subject is perceiving color or motion. But it is rarely noted that such a scientific investigation (and the corresponding scientific knowledge) only makes sense when we presuppose a certain sort of everyday knowledge from our life-world – namely the knowledge that humans normally perceive colors and are able to move themselves and perceive moving things. Bringing this presupposition into view is no trivial undertaking. To do so is to shift our attention to the importance of everyday and life-worldly conceptual knowledge for scientific conceptual knowledge. Such are the connections that knowledge research investigates. The interplay of various non-conceptual forms of knowledge is an equally familiar feature of the arts and sciences as well as our everyday lives, though the mechanisms of this interplay are still almost entirely unknown. Non-conceptual forms of knowledge include, inter alia, perception (e. g. a perception of color), experience (e. g. of a personal and immediate passion), intuition (e. g. an immediate visual experience), and procedural, practical abilities or knowing-how (e. g. bike riding, swimming, or cello playing). The question of how such non-conceptual forms of knowledge interact and interpenetrate one another then becomes a question of the mechanisms by which forms like ‘perception and knowing-how’ or ‘experience and knowing-how’ interact. An example of the first sort of interaction would be the interplay between a cellist’s auditory perception (which is non-conceptual sensible knowledge) and the practical abilities

Knowledge Research: Extending and Revising Epistemology

37

the cellist exercises in playing her instrument (which is non-conceptual practical and procedural knowing-how). The second sort of interaction can be seen in the interplay between an individual personal experience of emotion like joy or fury (which is the sort of non-conceptual emotional knowledge one can only have while one personally experiences that emotion) and the ability to know how to express the emotion in a bodily gesture (which is a sort of non-conceptual knowing-how, a sort of gestural knowing-how). Further examples can easily be generated by considering the interplay between bodily movements, sounds, and gestures (e. g. in ballet dancing) in everyday life, or in the arts and even the sciences. Brain surgery, for example, requires the surgeon to connect her view of the monitor (and thus her non-conceptual, visual knowledge) with her deft hand movements (and thus her non-conceptual, haptic knowledge) and to coordinate these with her experience (not least of all with the non-conceptual portions of her practice-based experiential knowledge) and still further forms of non-conceptual knowledge. Example 2: Explicit and Implicit Knowledge Knowledge always contains more than just what one knows, and we know more than we can say. Across the whole spectrum of knowledge forms – from basic, factical states of knowledge in practical situations to explicitly linguistic, propositional knowledge in the narrow sense – knowledge contains and involves, inter alia, abilities, competencies, contexts, background conditions, attitudes, goals, purposes, networks of beliefs, habits, habitualized patterns, and tacitly presupposed assumptions of somatic, neural, and physical processes (such as, for example, the activities of the central nervous system). In knowing something, we needn’t be explicitly conscious of any these components. Indeed, focusing one’s attention on these components (not to mention completely articulating them in propositional form) can often disturb or even derail our capacity for explicit knowledge or our ability to successfully learn a technique or a practice (e. g. bike-riding, or cello-playing). In order to so much as have explicit and conscious knowledge of something, one must already presuppose and rely on a great deal of implicit, unspecified, and unconscious knowledge. Accordingly, we can choose to investigate either the particular profile of explicit knowledge, or that of implicit knowledge.

38

Günter Abel

For Michael Polanyi, the structure of implicit knowledge consists of a “proximal” part or aspect and a “distal” part or aspect.42 The proximal aspect is one which is so close to us and so taken for granted that we tend not to even notice it in its own right – it is so close as to be invisible to us. It is rather registered subliminally, if at all. But the decisive point here is that we do presuppose and rely on this proximal aspect as a matter of course, even though we cannot express it in words. Polanyi’s example of implicit knowledge is well-known.43 I am hammering a nail into the wall. My conscious attention is (distally) trained wholly on the nail and the hammer. But there are many other components in play, of which I am not explicitly aware, but which are essential to my successfully hammering the nail into the wall – e. g. the motion of my hand, the placement of my fingers with respect to my palm, the pressure of my grip on the hammer’s handle, and many other things. Thus, much implicit knowledge (in the broad sense) must already be in play in order for me to so much as act, or gain my orientation. But in many cases, this implicit knowledge must not become explicit at the moment of the action’s execution (e. g. while I am swinging the hammer, or, in a scientific context, while a physical measurement is being performed). Otherwise I am far more likely to hit my thumb than the nail or even land myself in the emergency room. These interconnections are just as characteristic of the everyday knowledge proper to our life-world as they are of our artistic or scientific knowledge. The ability to make discoveries presupposes precisely this relation of proximal and distal aspects of knowledge. Systematic and reflective knowledge research can thus be described as an attempt to analyze just this relationship between explicit and implicit knowledge in a way that is fruitful for questions regarding knowledge and epistemology. If all knowledge were explicit knowledge, then research and discovery would be impossible. Conversely, since research and problem solving clearly are possible, it cannot be that all knowledge is explicit knowledge. It is important to me here – as both an epistemological and a methodological point – that the relation of implicit to explicit knowledge not be conceived as though implicit knowledge somehow unconsciously encapsulated a content that need only be made public and discursive. The contours and profile of explicit knowledge are simply different 42 Polanyi (1983), 10 ff. 43 Polanyi (1964), 55 f. and 174 f.

Knowledge Research: Extending and Revising Epistemology

39

from those of implicit knowledge. In becoming explicit, something gets added to implicit knowledge. This transition involves, among other things, increased distal attention and cognitive determinacy (which is coupled with consciousness and, in cases of knowledge in the narrow sense, also coupled with communicability, evidence, justification, and methodological order). This point is very important to me. For in contemporary philosophical discussion, one all too often encounters the view that everything depends on making implicit knowledge explicit. Indeed, Robert Brandom’s title “making it explicit” has become a veritable slogan in contemporary discourse.44 The picture connected with this view strikes me as singularly ill-suited to capture or adequately describe the peculiar character and role of the interplay and interpenetration of various forms of knowledge in our actual perception, speech, thought, and action. And it is not just the processes proper to our everyday life-world that this conception fails to capture, it is equally incapable of grasping such modes of interplay in the sciences, and is utterly hopeless in approaching the arts. Consideration of the interplay between implicit and explicit knowledge as well as the interplay between the proximal and distal aspects internal to implicit knowledge can (as Polanyi rightly claims) provide us with the means to understand what actually happens when knowledge is generated in a creative way. Polanyi’s thesis is that drawing on the mechanisms that support this relation between explicit and implicit knowledge can provide a solution to the notorious Meno paradox. Polanyi’s version of the paradox is as follows: when I know what I am searching for in attempting to solve a problem or turn up some novel finding, I am, strictly speaking, not in a position to discover anything radically new. But when I do not know what exactly I am looking for, then I equally cannot expect to solve the relevant problem or hit upon a novel finding (i. e. new knowledge).45 So long as we do not begin by assuming that all knowledge is explicit, we can invoke the distinction between explicit and implicit knowledge (and their different internal constitutions) to resolve this paradox. For our conception of implicit knowledge as a kind of intermediary state between explicit knowledge and total ignorance can help us understand what happens when, by exercising her creative abilities, someone generates new knowledge. This opens up issues regarding the processes of 44 Cf. Brandom (1994). 45 Polanyi (1964), 127 ff. Cf. Abel (2009).

40

Günter Abel

generating, extending, modifying, and precisifying our knowledge. Indeed, it serves to demonstrate that such processes are only possible against the background of knowledge that is as yet not explicit but rather unspecified and implicit. Systematic and reflective knowledge research investigate precisely this relationship between explicit and implicit knowledge by focusing on the modes by which forms of knowledge interact and interpenetrate, which we sketch in connection with extended or 3-D epistemology. At the same time, knowledge research calls attention to the connection between the conceptual pair ‘explicit–implicit’ and the pair discussed in example 1, namely ‘conceptual–non-conceptual’. Let us consider an example of how implicit conceptual knowledge can interpenetrate explicit conceptual knowledge. In constructing a proof, a mathematician implicitly draws upon many other conceptual elements of her knowledge (e. g. other theorems) without which the proof would either fail to be sound, or would lack mathematical plausibility. Additionally, any number of metalevel methodological and normative considerations run in tandem with the proof’s construction. For these considerations set the standard for what it even is to be a standard of plausible proof construction. Now we are capable of making both of these kinds of implicit conceptual knowledge explicit, should the need arise. And this capacity of ours is not restricted to the sciences, but is also to be encountered in our everyday life-world. Even the ordinary question “what were you thinking when you did X?” asks for conceptual reasons which were not themselves explicit in the execution of the action, despite the fact that their presence was essential to the action developing in the specific way that it (in fact) did. What we call ‘imagination’ presupposes the ability to draw upon and incorporate certain essential, but non-present aspects (be they conceptual aspects like reasons, or non-conceptual aspects like sensations or perceptions) into present processes or states.46 What I mean by calling certain non-present aspects “essential” is that, if one failed to incorporate them into one’s state, one could not so much as be in the present state in question. But imagination is clearly not limited to relating implicit conceptual knowledge to other, explicit conceptual knowledge. It is also manifest in the interplay between implicit and explicit non-conceptual knowledge. 46 On this point, cf. “Imagination und Kognition [Imagination and Cognition]”, in: Abel (1999).

Knowledge Research: Extending and Revising Epistemology

41

Examples of the latter interplay can be drawn from cases of perception. In order to perceive the person across the street as Uncle Paul, I have to draw upon earlier perceptions of Uncle Paul and incorporate them into the present situation. In such a case, I presuppose and rely on implicit non-conceptual knowledge (here, prior perceptual knowledge from other contexts) as a condition of my explicit non-conceptual knowledge (here, the perception of Uncle Paul). If I were then to tell someone else that I saw Uncle Paul, then we would have a case in which implicit non-conceptual knowledge was conditional for explicit conceptual knowledge (viz. a linguistic, propositional report). Systematic and reflective knowledge research investigates the mechanisms by which such interplay and such interpenetration of knowledge forms takes place. Example 3: Distributed and Integrated Knowledge Let us now consider the ways in which distributed and integrated forms of knowledge interact with and interpenetrate one another – a phenomenon with which we are intimately familiar both in everyday life, and in the arts and sciences. The following four scenarios should help to clarify the range of phenomena in question. (a) The Apartment. Say that an apartment has sustained extensive water damage. Repairing such damage requires the interplay of altogether different forms of knowledge. The architect, electricians, plumbers, painters, construction workers, heating specialists, (and, if the apartment is under historical protection, restoration specialists) must all work together. What is necessary is a problem-oriented cooperation of various forms of knowledge with the aim of restoring the apartment to a habitable condition. The embodied knowledge of any single worker would be insufficient to repair the damage: what is necessary is an integrative interplay of distributed forms of knowledge. (b) Driving a car. To successfully set a car into motion and steer it requires, inter alia, the technical-functional knowledge of what a steering wheel is and how the gear shift operates, a visual knowledge of the gas gauge and speedometer, practical knowledge expressed as the skillful hand movements required to operate the blinkers, the practical competency involved in using a GPS system. Moreover, we not only assume that one must be able to successfully integrate many of these forms of knowledge in order to count as a competent driver. We also accept that, though the automobile (or, say an ‘Airbus 380’) itself constitutes a highly complex technical system of the most various technologies

42

Günter Abel

and forms of knowledge, the driver (or pilot) need not have all of this knowledge ready to hand at all times. In the rule, we simply rely on the fluent interplay of forms of knowledge, at least as long as their interaction continues to be frictionless and breakdowns are righted promptly. (c) The Operation Room. On what does the neurosurgeon rely in opening the brainpan of a patient and performing open brain surgery? Neurosurgeons tend to respond that they rely on their eyes and their view of the monitor (i. e. on the imaging process and, in this sense, on pictorial knowledge), on their practical skill or the knowing-how expressed in their deft and skillful hand movements, and on the knowledge gained through their long experience. A neurosurgeon’s activities constitutively rely upon the integrative interplay of different forms of knowledge and each of her actions presupposes this interplay as a matter of course. (d) Laboratories. Any experimental laboratory or research lab showcases a complex interplay of various knowledge forms, practices, and dynamics. One need only ask an experimental scientist to learn how multifaceted their activities are. Experimental cultures can be described as constellations of interplay between distributed and integrated forms of knowledge, in which various knowledge forms, practices, and dynamics are trained on specific objects of investigation. Example 4: Knowing-How and Knowing-That The abilities, practices, procedures, competencies, skills, proficiencies, and established habits signified by the expression “knowing-how” are intimately familiar to us. And this is true across the whole spectrum of our experience – from our everyday practices (e. g. knowing how to open a refrigerator) through our acquired skills and abilities (e. g. swimming or tying a necktie) to higher-level activities (e. g. knowing how to follow rules in speech, thought, and action, or knowing how to construct a mathematical proof). Knowing-how is procedural knowledge, knowledge in action – it is not knowledge of fact or theoretical knowledge, and it is not primarily articulable in linguistic propositions. Rather, the cardinal features of knowing-how are the acquisition, refinement, and ultimate exploitation of one’s competency in following rules. And that goes for higher-level competencies as well – even for competencies at the highest level – like the ability to speak a language, or reflect upon the conditions of the validity of rules and norms (e. g. of arithmetical rules, rules of everyday comportment, or moral and ethical rules).

Knowledge Research: Extending and Revising Epistemology

43

Knowing how to maintain one’s balance while riding a bike (or learning how to ride a bike) doesn’t by any means require prior and explicit propositional knowledge of the corresponding physical laws, the appropriate angle of the handlebars, requisite pressure to apply with one’s arms, the proper posture of one’s upper-body, the coordination of foot and hand movements. If such knowledge did constitute a condition of the ability to ride a bike (and maintain one’s balance), only a few geniuses (and probably only the gods) would possess the perfect comprehension of physics and mathematics involved in successfully riding a bike. Rule-following – which is, for example, involved in the practice of successfully maintaining one’s balance while riding a bike – does (or need) not involve following a rule which has been previously established through physical laws. It is rather in virtue of practice, training, guidance, practical demonstration and imitation that one can acquire and subsequently refine and exploit this ability, competency, or technique. It is something I have referred to in another context as the regularity bound up with the exercise of a practice – an “intra-praxis regularity” (as opposed to a ‘criterial rule’).47 This sort of intra-praxis regularity has thus far played at best a marginal role in epistemology (if indeed it has played any role at all). To direct our attention to it therefore represents a challenge for knowledge research as well as for epistemology. The role and status of such rule-following (or such knowing-how) are fundamental. And both are intimately familiar to us from our everyday practices, as well as from our engagement with the sciences, philosophy, the arts, engineering, and the applied sciences. In this context, systematic and reflective knowledge research aims to answer the following three questions: (a) “What constitutes the specific profile of knowing-how?”, (b) “How does knowing-how relate to rule-following?”, and (c) “By means of what mechanisms do knowing-how and knowing-that interact and interpenetrate one another?”. By way of conclusion, I would like to go a bit beyond the foregoing four examples of how various forms of knowledge can interact and interpenetrate and indicate just two consequences and theoretical desiderata that arise against the background we have sketched. In particular, I will discuss the correlation of “forms of knowledge with forms of rep47 For more details about such a conception of “knowing-how” and its relation to rule-following, cf. Abel (2011b) and (2010b).

44

Günter Abel

resentation” (Example 5) and the correlation of “knowledge forms with creativity” (Example 6). Example 5: Forms of Knowledge and Forms of Representation My thesis here is that the various forms and levels of knowledge are each correlated with different forms and levels of representation. Pictorial knowledge involves a different mode of representation than does linguistic, propositional knowledge, or scientific knowledge, or emotional knowledge. In particular, I maintain that there are both horizontal correlations between forms of representation and forms of knowledge (as seen, e. g., in the relation between linguistic representation and linguistic knowledge) as well as vertical correlations – stratified correlation-relations between representing and knowing (e. g. the relation between metalinguistic signs and propositions and their object-language counterparts). One desideratum of knowledge research is to describe these correlations and to spell out the mechanisms they involve. I should note at the outset, however, that I explicitly reject classical forms of representationalism (at least regarding the structures that concern us here). Retaining, for a moment, the old language of representationalism, one might say that my guiding desideratum is a non-representational theory of representation. As for the problem of representation, I understand ‘representationalism’ as the position that holds that, within the triangular relations of I–We–World (and communication, world, action), we must forge connections and build bridges (with the help of mediating elements) to other people and to the world. The traditional view conceives of representation as a two place (and hence dualistic) relation between the mind/signs and the world. But the collapse of this misguided picture must not mislead us into accepting the equally misguided (and, at present, unfortunately widespread) assumption that the issue of ‘representation’ has become completely obsolete. This is by no means the case. Clearly, the non-dualistic forms of knowledge are internally connected with an non-dualistic mode of presentation and representation. Representationalism involves two serious errors which also show up, mutatis mutandis, in considerations regarding forms of knowledge. These two errors are (i) a restricted understanding of ‘representation’ (and, correlatively of ‘knowledge form’) as a ‘mirroring/copying/depiction’ of an ready made, prefabricated world, and (ii) taking a ‘representation’ (and, correlatively, a ‘knowledge form’) to be a kind of ‘mental intermediary’ inserted between the world and our actions. This sort of

Knowledge Research: Extending and Revising Epistemology

45

representationalism leads (at best) to an infinite regress of intermediaries between intermediaries (…) which lead us ever further from successful representation. Strictly speaking, this sort of representationalism blocks our direct access to the world as well as the initiation and execution of our actions. Theories of representation and theories of knowledge which perpetuate these two errors cannot be defended in a critical vein. And more importantly, these two assumptions (representation as mirroring; representation as intermediary) altogether lose sight of the fluent functioning of the I–We–World triangle and the internal connections between communication, relation to the world, and action. The appropriate response to these problematic assumptions is not to subtly refine them, but to abandon them altogether. By contrast, our new understanding of ‘representation’ is, in its internal connection to various forms of knowledge, originally constitutive, constructed, and perspectival in nature. Representation occurs in virtue of signs and interpretation – not in depictive or mimetic signs, but rather in virtue of primordially discriminating and individuating signs and interpretations. Representations are thus part of an interpretational semiotics with respect to what can so much as count as an individuated world, a specific object, or a determinate content of perception, speech, thought, or action. From this critical semiotic and interpretational perspective, we are no longer entitled to posit either a “ready made and prefabricated world’ to be represented or grasped in a knowledge form, nor an “innocent eye” with respect to representation or knowledge. Knowledge forms, and the forms of representation correlated with them, play a fundamental role characteristic of all natural, human (or, for that matter, artificial) intelligent systems – a role which it is our novel and pressing task to describe. We are just as familiar with various non-mimetic and non-intermediary forms of representation as we are with various forms of knowledge. They, too, permeate our everyday lives as much as they do the arts and sciences. Atlases, photos, graphs of market trends, or roadmaps provide equally legitimate examples as do numerical measurements of blood pressure in real numbers, or the representation of geometrical points, lines, and curves by means of algebraic expression. Forms of representation and forms of knowledge manifest themselves (in the broad sense of factical representation and factical knowledge) and realize themselves (in the narrow sense of denotative repre-

46

Günter Abel

sentation and denotative knowledge) in and through signs and interpretation.48 But neither representation nor knowledge is simply given with the material properties of signs (neither is, e. g., established by the physical constitution of various ink markings on a piece of paper). In order for something to be able to count as a representational or epistemically significant sign (be it a picture, a word, a gesture, a diagram, a mathematical formula, or an action), it must first have become a sign – and, more particularly, an epistemic or representational sign. Here a purely material something must undergo a transformation into a symbolizing sign with semantic features (meaning, reference, truth or satisfaction conditions). From the perspective of a triadic semiotic model, the import of an epistemic and representational sign consists in the fact that, to an Interpreter/Interpretant it represents something other than what it is.49 The key question is then how exactly the sign manages to do this and what exactly can be represented, transmitted, and preserved in an epistemic and representational semiotic function. Elucidating this question is of considerable importance in constructing a comprehensive account of the modes of interplay between forms of knowledge. In this sense, then, the (rediscovered) problematic of representation is the key problematic for both the general philosophy of signs and interpretation as well as for systematic and reflective knowledge research. In light of this new understanding of ‘representation’ as internally correlated with ‘forms of knowledge’ and as semiotic-interpretational in nature, it is worth emphasizing the following aspects. (a) We can correlate various forms of knowledge with various forms of representation (e. g. linguistic, pictorial, gestural etc.) and with various types of theories of representation (e. g. causal, functional, similarity-based, teleofunctional, model-based, structural, etc.). Moreover, we can conceive of the differences between these various forms of representation and types of theories in a conceptually semiotic and interpretational manner. (b) We can correlate the levels and hierarchies of knowledge forms with levels and hierarchies of representations. That is, we can, for example, construct a scale that extends from the elementary forms of visual 48 It is this assumption that provides the philosophy of signs and interpretation with its genuine access to the issue of forms of representation and forms of knowledge. For more on ‘representation’ cf. the entries in the indices of Abel (1995a), (1999) and (2004). 49 Cf. Peirce (1931 – 35), e. g. numbers 1.541; 2.228; 2.2247 – 9; 2.307.

Knowledge Research: Extending and Revising Epistemology

47

representation, through those representations bound up with intentional actions, all the way to meta-level and higher-level representations in the form of rules, standards and norms. This stratification can then be coupled to the levels of a stratified semiotic and interpretative model (which we cannot pursue further here).50 (c) By correlating forms of representation with knowledge forms, we can also describe, elucidate, and model the dynamic character of representations (i. e. their modifications, transformations, transferences, and revisions). In keeping with all this, we can no longer view the question of “what distinguishes a good representation?” as aimed at determining merely those facts which would enable us to reduce the complex semantics of a world to a simple and closed system. What is required for a representation to be “good” must rather be conceived as a multi-dimensional, stratified, and hierarchical matter. I would therefore like to list, in bullet-point fashion, a few criteria of any good representation, for elucidating their mechanisms constitutes a central desideratum of systematic and reflective knowledge research. A good representation must: (a) enable us to fluently sustain the triangular I–We–World relation (i. e. the relation between the subject/agent, other subjects/agents, and the world thereby disclosed or experienced) and maintain its openness to additions; (b) provide us with an orientation in theory and in practice; (c) be directed at our goals, problems, and actions; (d) be practically and epistemically useful (on account of (c)); (e) enable us to correlate different object domains (for example, the domain of physical processes and that of their representation in real numbers); (f) enable us to represent and process information; (g) enable us to reduce the excessive complexity of the domain to be represented (e. g. by representing geometrical points through n-tuples of numbers); (h) introduce new complexity into the representing domain (e. g. by representing physical relations through real numbers which are, for their own part, inscribed within a complicated mathematical framework whose features, though they do not directly correspond to any features 50

For more on this stratified model cf. Abel (1995a), (1999) and (2004), particularly their respective introductions and the entries for “Stufenmodell [Stratified Model]” in their indices.

48

Günter Abel

of the physical domain they represent, do enable us to generate new knowledge about that domain); (i) enable us to spell out the distinction between the content and the directedness of a relation (by means of which we are then able to lend the important distinction between correct and incorrect representations some degree of plausibility); (j) be capable of being correlated with formal feedback as well as with the conditions of empirical validity; (k) be capable of preserving or transferring structure (i. e. with respect to the relation between the representing and the represented system); (l) not merely “correlate” with worlds, but also “generate” them as world-versions51 (which is principally a mark of higher-level intelligent abilities and systems). Example 6: Forms of Knowledge and Creativity As we indicated in example 2 (“explicit and implicit knowledge”), reflection on the modes of interaction and interpenetration between different forms of knowledge can shed light on how radically new knowledge is created. One of the principle characteristics of creativity – both in everyday life, in philosophy, and in the arts and sciences – is the fact that it involves drawing on a number of different forms of knowledge in one’s associations, thought, and experiments. Recall, for example, one of the most creative transfers of one domain onto another: the application of mathematics to the realm of physical objects and events. As we mentioned earlier, the mathematicization of natural science is one of the key presuppositions and greatest achievements of modernity – and one that continues to gain in importance. It is often the mathematical formalism and the relevant fundamental mathematical equations that first circumscribe and determine what even counts as a relevant object of physical investigation – an object that can subsequently be correlated with a mathematical model. In particular, what we have here is an interplay and interpenetration of mathematical and physical knowledge. For in each such case there is an explicit step involved in going from the mathematical formalism to its application to the domain of physical objects. And this simultaneously presupposes and relies upon the interplay of theoretical knowledge (knowing-that) and practical knowledge (knowing-how) of its appropriate application. 51 Cf. Goodman (1978).

Knowledge Research: Extending and Revising Epistemology

49

Cases of radical creativity – e. g. the groundbreaking scientific discoveries and artistic achievements of a Beethoven, Lobachevsky, Copernicus, Einstein, Cézanne, Picasso, Schönberg, Heisenberg, or a Michelangelo – involve processes which break with the established rules, principles, and basic patterns fundamental to some system and replace them with new principles, rules, and patterns. Famous examples include: the transition to non-Euclidian geometry, Schönberg’s break with tonal music, the transition from the linear to the circular notation of Kekulé’s model of benzene. In addition to identifying the psychological characteristics of creative people, one can also attempt to clarify and to describe (at least to a certain extent) the phenomenology of creative processes. This is no small job, but does belong among the tasks of systematic and reflective knowledge research, which shares certain aspects in common with the research on creativity. A minimal phenomenology of creativity would certainly have to include such aspects as the ability to venture to draw analogies between various forms of knowledge, and the ability to foster the interplay and interpenetration of different forms of knowledge and the forms of representation internally correlated with them. So, here again we find ourselves faced with different forms of knowledge and different modes of their interpenetration. These two problematics form the heart of the whole endeavor called “systematic and reflective knowledge research”. Against the background of the triangular I–We–World relation, the enterprise leads us to extend and revise epistemology. I have here proposed some first steps towards new territory. Now this territory needs to be discovered, mapped out, and populated. That is the desideratum. The seas are open.52

References Abel (1995a): Günter Abel, Interpretationswelten. Gegenwartsphilosophie jenseits von Essentialismus und Relativismus, Frankfurt/Main. Abel (1995b): Günter Abel, “Indeterminacy and Interpretation”, in: Inquiry 38, 1 – 17. Abel (1999): Günter Abel, Sprache, Zeichen, Interpretation, Frankfurt/Main. Abel (2004): Günter Abel, Zeichen der Wirklichkeit, Frankfurt/Main. Abel (2006): Günter Abel, “Die Kunst des Neuen. Kreativität als Problem der Philosophie”, in: G. Abel (ed.), Kreativitt. Kolloquiums-Vortrge des 52 Translation from German by Daniel Smyth.

50

Günter Abel

XX. Deutschen Kongresses fr Philosophie. TU Berlin, September 2005, Hamburg, 1 – 21. Abel (2007): Günter Abel, “Forms of Knowledge: Problems, Projects, Perspectives”, in: P. Meusburger/M. Welker/E. Wunder (eds.), Clashes of Knowledge, Dordrecht, 11 – 33. Abel (2008a): Günter Abel, “Epistemische Objekte – was sind sie und was macht sie so wertvoll? Programmatische Thesen im Blick auf eine zeitgemäße Epistemologie”, in: K.-M. Hingst/M. Liatsi (eds.), Pragmata. Festschrift fr Klaus Oehler, Tübingen, 285 – 298. Abel (2008b): Günter Abel, “Modell und Wirklichkeit”, in: U. Dirks/E. Knobloch (eds.), Modelle, Frankfurt/Main, 31 – 46. Abel (2009): Günter Abel, “The Riddle of Creativity: Philosophy’s View”, in: P. Meusburger/J. Funke/E. Wunder (eds.), Milieus of Creativity, Dordrecht, 53 – 72. Abel (2010a): Günter Abel, “Epistemische Objekte als Zeichen- und Interpretationskonstrukte”, in: S. Tolksdorf/H. Tetens (eds.), In Sprachspiele verstrickt. Oder: Wie man der Fliege den Ausweg zeigt, Berlin/New York, 127 – 156. Abel (2010b): Günter Abel, “Knowing How. Eine scheinbar unergründliche Wissensform”, in: J. Bromand/G. Kreis (eds.), Was sich nicht sagen lsst. Das nicht-begriffliche in Wissenschaft, Kunst und Religion, Berlin, 319 – 340. Abel (2011a): Günter Abel, “Der interne Zusammenhang von Sprache, Kommunikation, Lebenswelt und Wissenschaft”, in: C.F. Gethmann (ed.), Lebenswelt und Wissenschaft. Kolloquiums-Vortrge des XXI. Deutschen Kongresses fr Philosophie, Universitt Duisburg-Essen, September 2008, Hamburg, 351 – 371. Abel (2011b): Günter Abel, “Knowing How. Indispensable But Inscrutable”, in: S. Tolksdorf (ed.), Berlin Studies in Knowledge Research, vol. 4, Conceptions of Knowledge, Berlin/New York. Abel (2012): Günter Abel, Knowing-How. The Logic Of Abilities, Berlin/New York. Arnheim (1997): Rudolf Arnheim, Visual Thinking, Berkeley/Los Angeles/ London. Beckermann (2001): Ansgar Beckermann, “Zur Inkohärenz und Irrelevanz des Wissensbegriffs. Plädoyer für eine neue Agenda in der Erkenntnistheorie”, in: Zeitschrift fr philosophische Forschung 55, 571 – 593. Bennett/Hacker (2003): Max R. Bennett/Peter M. S. Hacker, Philosophical Foundations of Neuroscience, Oxford. Bernecker/Dretske (2000): Sven Bernecker/Fred Dretske (eds.), Knowledge. Readings in Contemporary Epistemology, Oxford. Brandom (1994): Robert Brandom, Making It Explicit: Reasoning, Representing, and Discursive Commitment, Cambridge/MA. Brandom (2008): Robert Brandom: Between Saying and Doing: Towards an Analytic Pragmatism, Oxford. Burge (1979): Tyler Burge, “Individualism and the Mental”, in: P. French/T. Uehling/H. Wettsein (eds.), Midwest Studies in Philosophy, vol. 4, Hoboken.

Knowledge Research: Extending and Revising Epistemology

51

Chisholm (1982): Roderick M. Chisholm, The Foundations of Knowing, Minneapolis. Churchland (1985): Paul Churchland, Matter and Consciousness, Cambridge/ MA. Churchland (1989): Paul Churchland, A Neurocomputational Perspective: The Nature of Mind and the Structure of Science, Cambridge/MA. Clark/Chalmers (1998): Andy Clark/David Chalmers, “The Extended Mind”, in: Analysis 58, 10 – 23. Clark (2008): Andy Clark, Supersizing the Mind. Embodiment, Action, and Cognitive Extension, Oxford. Craig (1993): Edward Craig, Was wir wissen kçnnen. Pragmatische Untersuchungen zum Wissensbegriff, Frankfurt/Main. Craig (2002): Edward Craig, Knowledge and the State of Nature: An Essay in Conceptual Synthesis, Oxford. Davidson (1982): Donald Davidson, “Rational Animals”, in: Davidson (2001a), 95 – 106. Davidson (1991a): Donald Davidson, “Epistemology Externalized”, in: Davidson (2001a), 193 – 204. Davidson (1991b): Donald Davidson, “Three Varieties of Knowledge”, in: Davidson (2001a), 205 – 220. Davidson (2001a): Donald Davidson, Subjective, Intersubjective, Objective, Oxford. Davidson (2001b): Donald Davidson, “A Coherence Theory of Truth and Knowledge”, in: Davidson (2001a), 137 – 158. Descartes (1986): René Descartes, Meditationes de Prima Philosophia, Stuttgart. Devitt (2006): Michael Devitt, The Ignorance of Language, Oxford. Dretske (1995): Fred Dretske, Naturalizing the Mind, Cambridge/MA. Dummett (2010): Michael Dummett, The Nature and Future of Philosophy, New York. Føllesdal (1999): Dagfinn Føllesdal, “Triangulation”, in: L. E. Hahn (ed.), The Philosophy of Donald Davidson (The Library of Living Philosophers, vol. XXVII), Chicago, 719 – 728. Fumerton (1995): Richard A. Fumerton: Metaepistemology and Skepticism: Studies in Epistemology and Cognitive Theory, London. Gettier (1963): Edmund L. Gettier, “Is Justified True Belief Knowledge?”, in: Analysis 23, 121 – 123. Giaquinto (2007): Marcus Giaquinto, Visual Thinking in Mathematics. An Epistemological Study, Oxford. Goodman (1976): Nelson Goodman, Languages of Art, Indianapolis. Goodman (1978): Nelson Goodman, Ways of Worldmaking, Indianapolis. Grundmann (2003): Thomas Grundmann (ed.), Erkenntnistheorie, Paderborn. Heidegger (1927): Martin Heidegger, Sein und Zeit, Tübingen. Hendricks et al. (2011): Vincent F. Hendricks/Horacio Arló-Costa/Johan van Benthem (eds.), Readings in Formal Epistemology, Cambridge. Hetherington (2006): Stephen Hetherington, “How to Know (that Knowledge-that is Knowledge-how)”, in: S. Hetherington (ed.), Epistemology Futures, Oxford, 71 – 94.

52

Günter Abel

James (1979): William James, Pragmatism, Cambridge/MA. Kornblith (2002): Hilary Kornblith, Knowledge and Its Place in Nature, Oxford. Miller/Wright (2002): Alexander Miller/Crispin Wright (eds.), Rule-Following and Meaning, Ottawa. Millikan (1984): Ruth Millikan, Language, Thought and Other Biological Categories, Cambridge/MA. Mittelstraß (2003): Jürgen Mittelstraß, Transdisziplinaritt – wissenschaftliche Zukunft und institutionelle Wirklichkeit, Konstanz. Nelsen (1993/2000): Roger B. Nelsen, Proofs Without Words, Washington. Papineau (1993): David Papineau, Philosophical Naturalism, Oxford. Peirce (1931 – 35): Charles Sanders Peirce, Collected Papers, C. Hartshorne/P. Weiss (eds.), volumes I – VI, Cambridge/MA. Polanyi (1964): Michael Polanyi, Personal Knowledge, New York. Polanyi (1983): Michael Polanyi, The Tacit Dimension, Gloucester. Putnam (1975): Hilary Putnam, “The meaning of ›meaning‹”, in: H. Putnam, Philosophical Papers, vol. 2: Mind, Language and Reality, Cambridge, 215 – 271. Putnam (1995): Hilary Putnam, Pragmatism. An Open Question, Cambridge/ MA. Quine (1969): Willard Van Orman Quine, “Epistemology Naturalized”, in: W.V.O. Quine, Ontological Relativity and Other Essays, New York, 69 – 90. Quine (1993): Willard Van Orman Quine, Pursuit of Truth. Revised Edition, Cambridge/MA. Rheinberger (2001): Hans-Jörg Rheinberger, Experimentalsysteme und epistemische Dinge. Eine Geschichte der Proteinsynthese im Reagenzglas, Göttingen. Rheinberger (2006): Hans-Jörg Rheinberger, Epistemologie des Konkreten. Studien zur Geschichte der modernenBiologie, Frankfurt/Main. Sartwell (1992): Crispin Sartwell, “Why Knowledge is merely true belief”, in: The Journal of Philosophy 89, 167 – 180. Shope (1983): Robert K. Shope, The Analysis of Knowledge: A Decade of Research, Princeton. Stanley/Williamson (2001): Jason Stanley/Timothy Williamson, “Knowing How”, in: The Journal of Philosophy, vol. 98, number 8, 411 – 444. Strawson (1959): Peter F. Strawson, Individuals. An Essay in Descriptive Metaphysics, London/New York. Williamson (2000): Timothy Williamson, Knowledge and its Limits, Oxford. Wittgenstein (1984): Ludwig Wittgenstein, Philosophische Untersuchungen, Frankfurt/Main.

I. Theoretical Knowledge

A Priori Knowledge in Methodical Philosophy Peter Janich 0. Introductory Reminiscences In the Erlangen school of constructivism, discussions of a priori knowledge were essentially oriented around the classification scheme that Paul Lorenzen offered in his (and Wilhelm Kamlahs) book Logische Propdeutik 1. Since this book is generally seen as having established the school, we will start with this scheme. A chapter entitled “Non-empirical Truth” extensively contrasts the logical truths, as “analytic in the broader sense”, with those that are “analytic in a narrower sense”, the latter of which are then divided into “formal-analytic” truths (namely propositions that are true due to an explicit definitional equivalence) and “material-analytic” truths (namely propositions that are true due to a predicate rule). These three forms of analytic truths are then contrasted with “synthetic” truths, namely the “formal-synthetic” truths (propositions of arithmetic that are true due to their rules of construction) and the “material-synthetic” truths (propositions of geometry founded on “ideative rules”). Empirical truths as an example of synthetically true propositions are only mentioned in this scheme for the sake of completeness, and of course are not discussed at all under that chapter heading. The chapter (and the book as a whole) ends with a beautiful symmetrical scheme expressly including all true propositions “insofar as they have been methodically reconstructed in this book”, and that means only scientific truths in particular; hence truths as a whole are divided into analytic and synthetic truths. The analytic truths break down into the logical and “analytic in the narrower sense”; the synthetic truths break down accordingly into the empirical and the “synthetic in the narrower sense”. The advantage of this symmetrical classification is evident. The analytic and the synthetic truths in the narrower sense of each form the body of “a priori truths in the narrower sense”. Outside 1

Kamlah/Lorenzen (1967).

56

Peter Janich

the realm of formal logic, i. e. apart from logically true forms of propositions, those propositions are considered to be a priori that hold based on definitional equivalence, predicate rules, and constructive rules of arithmetic or ideative rules. In overview it looks like this:

Truths

analytic

synthetic a priori (narrow sense)

analytic (narrow sense)

synthetic (narrow sense)

logical formal analyt. material analyt. formal synthet. material synthet. empirical

At the time this overview was first published, the areas of logic, arithmetic and geometry above all represented ongoing challenges rather than completed pieces of the theory. The subsequent work in the theory of science by the philosophers ascribed to this school of methodical thinking was marked by its opposition to the prevailing traditions of logical empiricism and critical rationalism, represented by the Stegmüller school on the one hand and by Hans Albert and his followers on the other.2 The principle obstacles between these traditions in the theory of science (which were erected rather dogmatically by both sides) seemed to diminish when Wolfgang Stegmüller discovered the “non-statement view” for himself and his students and started propagating a structuralist theory of science building on J. Sneed’s model-theoretical approach3 : at this point a form of reconstructivism that involved significant normative elements took the place of a dogmatic descriptivism. Yet this had no repercussions for the assessment of a priori truths. The possibility of analytically true sentences (later challenged by Quine) remained uncontested between constructivists and empiricists. Protophysics was controversial, and philosophy of 2 3

Cf., for example, Janich et al. (1974). Stegmüller (1973).

A Priori Knowledge in Methodical Philosophy

57

mathematics on the non-constructivist side consisted essentially in a commitment to David Hilbert-style formalism. This was roughly the state of the debate when the author of this paper was given an appointment at Marburg in 1980.

1. Shifts in the Horizon. Erlangen- and Konstanz-style Constructivism in the Marburg Critique Leaving aside the historical reminiscence, we can mention a few objections that have crystallized over the course of the roughly 25 years of discussion in Marburg.4 The shifts that emerged in Marburg relative to the constructivism of Erlangen and Konstanz – while retaining the same basic methodical orientation – are documented and justified in the programmatic opening paper in the first5 collected volume, Methodischer Kulturalismus. Zwischen Naturalismus und Postmoderne 6. Here I will just name the aspects that are important today for a new methodically culturalist theory of a priori knowledge and that depart from the tradition of Erlangen and Konstanz. The restriction of theoretical philosophy to theory of science needs to be abandoned (In his 1987 Lehrbuch der konstruktiven Wissenschaftstheorie 7 P. Lorenzen had even subsumed the entirety of philosophy under the heading of “theory of science”, including the replacement of ethics with a political anthropology.) Epistemological questions also arise in our lifeworld. This does not just concern forms of cognition and knowledge as personal, individual acts; rather, there are also public stores of knowledge out4

5

6 7

I also cannot neglect to mention all the supporting members of the philosophical colloquium in Blitzweg over the years, a transplant from my work group from the (not very attractive) towers of the humanities in Lahntal to a formerly private residence with the address “Blitzweg 16” (1990 – 2003): T. Galert, M. Gutmann, G. Hanekamp, D. Hartmann, H. Kwon, R. Lange, N. Psarros, W. Schonefeld, M. Weingarten, M. Wille, W. Zitterbarth, and a series of temporary guests, scholarship holders and colleagues from other disciplines. It may no longer be possible to ascertain who provided what impulse, formulated which argument, expressed which criticism and invested which effort. A second collected volume by the same authors came out under the title Die Kulturalistische Wende. Zur Orientierung des philosophischen Selbstverstndnisses, Hartmann/Janich (1998). A third volume entitled Der Diskurs des Methodischen Kulturalismus is in progress. Hartmann (1996). Lorenzen (1987).

58

Peter Janich

side of the sciences that are managed according to specific rules and that with good reason are not included among the sciences. The constructivism of Erlangen and Konstanz had not developed their own concept of experience. The above-mentioned chapter on non-empirical truth from the Logische Propdeutik does not say a single word about the concept of experience that it presupposes, which elsewhere in the book either is said to refer to the “ordinary sense” of the word or is left to the specific experts. In contrast the proto-theories on physics, chemistry, biology and psychology developed in Marburg, along with an examination of the lifeworld in terms of the theory of action and the philosophy of language, have led us to spell out concepts of experience for various spheres, ranging from sense perception to the field of history. The primacy of philosophy of language over the theory of action was replaced in Marburg by the primacy of the theory of action over the philosophy of language: although obviously no-one within methodical constructivism denies that speaking is action, it was in Marburg that the first attempts were made (and brought to a tentative finish in the author’s Logisch-pragmatische Propdeutik, 2001)8 to ground the distinctions in the philosophy of language on those of the theory of action. Just as under the influence of its empiricist and formalist opponents the Erlanger constructivism had largely restricted itself to the theory of science without realizing it, its conception of language also proved to be burdened with pragmatic and historical deficiencies in unknowingly taking on a fixation of its opponents, particular from the analytic philosophy of language. The protophysics of space, time and matter,9 as one of the a priori foundations of the methodical philosophy, has been further developed to include a clarification of the relation to the physics of special relativity10 ; and the operative, form-theoretical clarification of the problem of parallelism along with its corresponding uniqueness proof 11 have been worked out as an alternative to the approach of R. Inhetven and P. Lorenzen. This was preceded by a proposal to clarify one of the classical problems of epistemological a priorism since Kant: in 1989 an operative theory of three-dimensionality was published as a proposed reconstruc-

8 9 10 11

Janich (2001). Janich (1997). Schonefeld (1999). Wille (2002).

A Priori Knowledge in Methodical Philosophy

59

tion of the Kantian a priori in terms of the theory of action, using the example of the space of experience.12 As it concerns the problem of distinguishing a priori knowledge, besides the many works I will not discuss any further here, such as those on logic, mathematics and proof theory (D. Hartmann13, M. Wille14), on the a priori parts of proto-chemistry (G. Hanekamp15, N. Psarros16), on the theory of the experiment (R. Lange17), on the theory of the organism and evolutionary biology (M. Gutmann18, M. Weingarten19) and on theories of the planning of actions (A. Grunwald20), a concept of a priori knowledge has emerged in Marburg based on my own work on the theory of action and philosophy of language. To explain this concept I will now take up the historical connection beyond the tradition of methodical philosophy.

2. Critique of the Kantian A Priori The Kantian conception of a priori knowledge has come to dominate the history of philosophy to such an extent that even the “Historisches Wörterbuch der Philosophie” begins its article a priori/a posteriori with the complaint that the use of this concept has “so lastingly been fixated [by Kant’s terminology] that even attempts to clarify its prior history have been under the spell of the characteristic ‘independence of experience’…”21 Moreover, the rise of the empirical sciences in the 19th century, the “discovery” of non-Euclidean geometries, the scientistic stance of neo-Kantianism, at least that of the Marburg tradition, and in the 20th century the success of relativistic physics and evolutionary biology have all convinced the overwhelming majority of scientists and philosophers that a philosophy of the a priori is at best a topic for historians of 12 Janich (1989). 13 Hartmann (2003). 14 Wille (2008). Also important for the issue of a priorism in mathematics is: Wille (2007). 15 Hanekamp (1997). 16 Psarros (1999). 17 Lange (1999). 18 Gutmann (1996). 19 Weingarten (1993). 20 Grunwald (2000). 21 Schepers (1971), 462 f.

60

Peter Janich

philosophy in the Kantian tradition; the actual problem is supposed to have been taken care of by the sciences. The argumentative level at which the a priori has been wrapped up like this is rather low. What Karl Popper wrote in his 1963 introduction to the second edition of his Logik der Forschung 22, for example, or what Rudolf Carnap wrote about Kant’s synthetic a priori in 1966 in his book Philosophical Foundations of Physics 23, does not go beyond Albert Einstein’s critique of Kant (1922)24 and its popular version in the lecture “Geometrie und Erfahrung” (1921), which showed little philosophical competence. It is still unfinished business in the history of philosophy to explain how the philosophy of Kant loses its influence on scientists and philosophers when it comes to space and time (and later even 22 Popper, in Logik der Forschung: “Zu dieser großen Tradition (sc. des englischen Empirismus, P.J.) stehe ich dadurch im Gegensatz, dass ich gewisse von Kants Beiträgen zur Erkenntnistheorie für grundlegend, ja geradezu für entscheidend halte, obwohl ich nicht daran glaube, dass es synthetische Sätze gibt, die a priori als gültig eingesehen oder begründet werden können … Aber wir haben von Einstein gelernt, dass Newtons Physik möglicherweise falsch ist; und das bedeutet eine völlige Änderung der Problemsituation gegenüber der, die Kant vorfand. So können wir jetzt Kants Probleme dadurch lösen, das wir den grundsätzlich hypothetischen Charakter der naturwissenschaftlichen Theorien (und noch mehr der Metaphysik) anerkennen” (Popper (1971), XXIV). 23 Carnap, in ch. 18 of Kants synthetisches Apriori: “Die mathematische Geometrie ist a priori. Die physikalische Geometrie ist synthetisch. Keine Geometrie ist beides. Wenn man den Empirismus akzeptiert, dann kann es kein Wissen geben, das sowohl a priori wie auch synthetisch wäre … Kant war der Meinung, dass Apriori-Wissen sicheres Wissen sei; es kann nicht durch Erfahrung widerlegt werden. Die Relativitätstheorie machte es allen, die es verstanden, klar, dass Geometrie, im Apriori-Sinn genommen, uns nichts über die Realität sagt. Es gibt keine Aussage, die logische Sicherheit mit einer Information über die geometrische Struktur der Welt verbindet” (Carnap (1969), 183). 24 Einstein, in Grundzge der Relativittstheorie: “Es ist deshalb nach meiner Überzeugung eine der verderblichsten Taten der Philosophen, dass sie gewisse begriffliche Grundlagen der Naturwissenschaft aus dem der Kontrolle zugänglichen Gebiete des Empirisch-Zweckmäßigen in die unangreifbare Höhe des Denknotwendigen (Apriorischen) versetzt haben. Denn wenn es auch ausgemacht ist, dass die Begriffe nicht aus den Erlebnissen durch Logik (oder sonstwie) abgeleitet werden können, sondern in gewissem Sinn freie Schöpfungen des menschlichen Geistes sind, so sind sie doch ebenso wenig unabhängig von der Art der Erlebnisse, wie etwa die Kleider von der Gestalt der menschlichen Leiber. Dies gilt im Besonderen auch von unseren Begriffen über Zeit und Raum, welche die Physiker – von Tatsachen gezwungen – aus dem Olymp des Apriori herunterholen mussten, um sie reparieren und wieder in einem brauchbaren Zustand setzen zu können” (Einstein (1965), 2).

A Priori Knowledge in Methodical Philosophy

61

when it comes to analytically true propositions). We will need to look at both the weaknesses of Kant’s philosophy from a systematic point of view today as well as the systematic weaknesses on the part of Kant’s critics. In the following I will look at three prominent Kant critics and show how they come up short: (1) Hermann von Helmholtz’s empiricist critique of the a priori status of the geometric axioms; (2) Konrad Lorenz’s naturalistic critique of the a priori nature of the disposition to perception, and (3) Willard Orman V. Quine’s critique of the possibility of a priori knowledge in the form of analytically true propositions. On (1): Hermann von Helmholtz – whose 1855 lecture “Über das Sehen des Menschen”, later famous as his “Kant lecture”, suffers from a psychologistic misunderstanding of the Kantian philosophy – presented the empiricist alternatives to the Kantian a priorist understanding of geometry in his 1870 lecture “On the Origins and Significance of the Axioms of Geometry”25. His argument rests, among other things, on a suggestive model: two-dimensional beings living on a sphere who measured the angles of a triangle would get a sum larger than two right angles – and how much larger would depend on how large the triangle was in relation to the sphere. Here we are meant to understand the “straight” lines forming the triangle as the shortest possible connection of the three corners, where “shortest possible” can be technically realized by measurements taken with portable rigid bodies, and not, for example, stretchable ropes. On the sphere the “straight” lines, as the shortest possible connection of two of its points, are always great circles whose center is equivalent to the center of the sphere. Since any two great circles intersect, there are no “parallel” lines on the sphere. Hence the Helmholtzian model represents one of the two possible negations of Euclid’s parallel postulate. Of course Helmholtz knows that this model has to be invested with a certain assumption concerning how the “shortest” lines and angles are measured: Thus any comparative estimation of magnitudes, or measurement of spatial relationships, starts from a presupposition about the physical behaviour of 25 Helmholtz (1968).

62

Peter Janich

certain bodies, whether of our own body or of applied measuring instruments. This presupposition may incidentally have the highest degree of probability and be in the best agreement with all physical circumstances otherwise known to us, but it still goes beyond the domain of pure spatial intuitions (my own emphasis).26

Here we see the “holistic” assumptions of the empiricist critique of a priorism that then find their way into the physics of Albert Einstein. The entirety of empirical physics is concerned with “the physical behavior of certain bodies” during measurements, for example in transporting measuring instruments for lengths or angles. If we seek a fair historical judgment on the relation of Kant’s conception and Helmholtz’s, we have to concede to the empiricist tradition that at no point in his pertinent writings does Kant even hint at the question of how his conception of the synthetic a priori character of geometry and the strict necessity of its propositions can offer any assessment procedure for justifying or rejecting Euclid’s postulates and axioms. Here we are concerned with the simple question (later to be taken up more systematically) of how Kant is supposed to know that Euclid’s geometric axioms describe spatial relations such that – in accordance with Kant’s own explanation of “necessity” – it cannot be otherwise. 27 (Helmholtz speaks here of “necessities of thought” in a psychologistic sense, which is later repeated by Albert Einstein.) Hence this “strict generality and universality” seems to be taken ad absurdum as soon as someone like Helmholtz succeeds in presenting a plausible alternative model and thus proving that it could be otherwise. Nonetheless this gap in Kant’s philosophy is not enough to declare empiricist philosophy in the tradition of Helmholtz the victor in this conflict. As the methodical philosophy has emphasized since H. Dingler, we have incompletely grasped the rigidity of bodies for the purpose of measuring spatial relations if we only describe them physically as “natural bodies”. Only a criterion for the undisturbed state of measuring instruments (e. g. during transportation) can provide usable measurement data.28 Even among empiricists it is uncontroversial that the undisturbed state of an angle-measurer, for example, that the Helmholtzian sphere-residents have to transport to the three corners of the triangle, 26 Helmholtz (1884), 24. 27 Kant (1789), B3. 28 Cf. Dingler (1928).

A Priori Knowledge in Methodical Philosophy

63

cannot itself be “empirically” determined by a direct measurement of the angles – since this obviously just pushes the problem off onto the next generation of measuring instruments. Hence at this point the empiricists turn to a holistic understanding of validity according to which it is the entirety of physics that is supposed to guarantee the “highest degree of probability” and “the best agreement with all physical circumstances otherwise known to us”. The excuse, still prevalent among empiricists today, that there is supposedly in principle no other way out of the dilemma, has been proven definitively wrong: if in a prototype-free and reproducible procedure of protophysics new angle-measurers can be produced at each corner of a triangle in the Helmholtzian spherical world, then the measurement of the angles in a spherical triangle can in fact be taken with usable empirical results. But then the acquisition of empirical data has an a priori foundation in the manufacturing knowledge that is able to identify undisturbed measuring instruments. Hence the empiricist critique of a priori philosophy missed the mark with its own argument. (2) In his article “Kant’s doctrine of the a priori in the light of contemporary biology”,29 Konrad Lorenz, who took over Kant’s chair in Königsberg in 1940, presented the thesis quoted nowadays in its abbreviated form “the cognitively a priori is phylogenetically a posteriori”. His approach, which lives on today as “evolutionary epistemology” and plays an indispensable role in contemporary brain research, can be called naturalistic in that it takes the natural conditions of the real world to be that which organic life, including its cognitive performance, has to adapt to. “What we witness as experience is always a coping of the real in us with the real outside of us.”30 Here we well, just as we saw in the case of the Helmholtzian critique of Kant, the rejection of “necessities of thought” plays an important role in Lorenz’s arguments. For example, Lorenz discusses whether it is “possible in thought” for a water shrew to find a short-cut. “From a psychological viewpoint, the water shrew’s command of space is a sequence of conditioned reflexes and kinesthetically ingrained movements.” The biologist’s naturalistic view also takes on Kant’s unexplicated talk of “strict necessity” in order to replace it with the open-ended possibilities of adaptation of human and animal behavior to natural conditions under the selective pressure of survival. Lorenz emphasizes re29 Lorenz (1941), 94 – 125, English translation published in: Ruse (2009). 30 All quotations are taken from Ruse (2009).

64

Peter Janich

peatedly that we have to overcome an “unjustifiable anthropomorphism”: “Progress in science always has a certain tendency to de-anthropomorphize our image of the universe.” In his book Behind the Mirror 31 K. Lorenz expands this conception into a “Search for a Natural History of Human Knowledge” (according to the subheading of his book), describing a ladder of “epistemic faculties” from the reaction schemes of amoeba or one-celled organisms to the linguistic ability and culturality of modern humans. We should not overlook the principle differences between Helmholtz’s and Lorenz’s critiques of the a priori: where Helmholtz opposes the Euclidean geometry favored by Kant with non-Euclidean alternatives and recommends them as foundational theories of physics, Lorenz discussed not just the natural sciences but also the human epistemic faculty in general, which also includes perceptual and causal judgments in everyday life. The point is not to show these to be wrong or unreliable, even if by 1941 Lorenz had already clearly stated that in the areas of “atomic physics, quantum mechanics and wave theory” our evolutionarily acquired epistemic dispositions might run aground. Rather Lorenz believes he is able to give the Kantian theory of a priori epistemic dispositions an empirical foundation in evolutionary biology: The discovery of the a priori is that spark we owe to Kant and it is surely not arrogance on our part to criticize the interpretation of the discovery of new facts (as we did in criticizing Kant with regard to the origin of the forms of perception and categories). This critique does not lower the value of the discovery any more than it lowers that of the discoverer.32

Helmholtz’s empiricist critique and Lorenz’s naturalistic critique are similar in that both seek to invert the reciprocal methodical dependency of the a priori and the a posteriori. Just as for Helmholtz the strict validity of propositions such as those of analytic geometry ultimately has an empirical foundation in physics, for Lorenz the validity of judgments of experience is founded in the natural laws of evolutionary adaptation of organisms to the natural conditions of life. Where Helmholtz has to appeal to the success of physics, Lorenz has to appeal to the theory of organisms and evolutionary biology. This comparison gives us the decisive hint as to why Lorenz’s criticism of the a priori is essentially mistaken: it fails to ask how we should understand and make good on the claims to validity of the biology with 31 Lorenz (1977). 32 Ibid.

A Priori Knowledge in Methodical Philosophy

65

which Lorenz invests his evolutionary epistemology. Instead he naively adopts the standpoint of the Archimedean observer from which the fit between the faculty of knowledge and the objects of knowledge can be characterized and empirically judged. To put it more simply, the biologist must always already know how the world is in order to be able to sensibly formulate the thesis of fit semantically and to be able to review its validity empirically. And to put it somewhat polemically, the theory cannot be true, because it cannot be false. The thesis of fit goes semantically hand-in-hand with a description of reality that, according to the theory itself, cannot be otherwise than our natural epistemic dispositions allow. In other words, despite its inclusion of the empirical results of organic and evolutionary biology – results which on their own terms I do not wish to doubt here – the thesis is tautological, empty. (3) Quine’s33 critique of the a priori represents a qualitative leap, since it turns to (allegedly) analytically true propositions rather than synthetic a priori propositions. Historically the philosophy of science in logical empiricism and critical rationalism had come to the consensus that all knowledge based on real conditions could only be decided by empirical examination, but that linguistic conditions or at least the conditions of artificial language still had the non-empirical type of validity of analyticity or of the purely logical. Hence at least the part of the Kantian a priori that relates to “analytic judgments” could still be considered valid, despite the fact that with the linguistic turn and the introduction of modern logic and philosophy of language the Kantian understanding of analytic validity had already been revised. Hence Quine was responding not to Kant but to the philosophy of the 20th century. Quine’s arguments are behavioristic. Quine describes the acquisition of language within a psychological framework in terms of the stimulus-response model. This results in the individually acquired linguistic uses being completely situation-dependent, which leads in particular to an indeterminacy of translation between various natural languages. Hence there can be no synonymy and no definitional equivalence. But where we cannot assume any strict logical equivalence of linguistic expressions, we cannot draw any valid inferences based on semantic equivalences. Here it is also relatively easy to show that there are inputs to this approach that stand in need of criticism and that invalidate Quine’s critique of the a priori. However, a methodical philosophy of language 33 Quine (1975).

66

Peter Janich

supported by the theory of action is needed to show that Quine’s description of the (semantic) acquisition of language is in error. I cannot produce this here for reasons of space, although the means to do so have already been developed in the literature.34 Hence in lieu of this argumentation I will instead just point to two undisputed facts: There is no question that in specialized sciences terminologies have become so standardized that the validity of propositions is decidable – and incidentally this decidability is not lost in translation into the language of another country. We can take the trivial arithmetic sentence “every prime number larger than two is odd”, or the somewhat less trivial sentence “there is no largest prime number”. Both sentences are provable with logic and the terminological determinations of “prime number”, “odd” and “larger than”, i. e. they hold analytically. Outside of the sciences the factual acquisition of language and the expressive possibilities of every normal speaker involve a structural knowledge such that a whole system of connections between predicates is acquired at the same time. A small child is already familiar with the correlation of converse predicates, for example: “if A is bigger than B, than B is smaller than A”. Or the child learns (semantically) the transitivity of equalities, such that it follows from “A is as big as B and B is as big as C” that “A is as big as C”, for example. The child also establishes knowledge of relations, for example: “My brother-in-law is either my sister’s husband or my wife’s brother” holds analytically. Hence it is also wrong to say of our everyday language with all its fuzziness that there are no sentences that hold true due to linguistic conventions where equivalences play an important role. We just need a reconstructive explication of the implicitly learned rules in order to gain an effective criterion for the validity of analytic sentences. The three critiques discussed here, which can be described as primarily empiricist, naturalistic and behaviorist, are only exemplary cases; but on the other hand they set the terms of the debate to a certain degree, where it is claimed to be an established fact of philosophy and the natural sciences that the philosophy of the a priori is on the whole a failure. 35 It is very instructive that these three examples were of course not

34 Cf. the title in fn. 4. 35 It is remarkable how free of justification the arguments for this thesis are, for example from Vollmer (1993), 103: “Auch der kantische Apriorismus ist gescheitert. Die Existenz synthetischer Urteile a priori konnte bis heute nicht

A Priori Knowledge in Methodical Philosophy

67

advanced in the same field of argumentation in which Kant developed his philosophy, but rather all appealed to the natural sciences – and to very particular conceptions of the sciences. In other words, the tenability of these criticisms of the a priori stands and falls together with the empirical character and validity of the stores of knowledge invested in these critiques. Hence we need to ask whether the options are exhausted by the alternatives of transcendental philosophy (with Kant) and natural-scientific empiricism. Can we not find a completely different determination of a priori knowledge that can account for our unwillingness, in philosophy and in the sciences and in our lifeworld, either to hold all of our propositions open to empirical falsification or to subject them to a transcendental justification independent of the reality of our life and our action? In the following I present such a third option. There is a knowledge of our own action that doesn’t fit the description “transcendental” or “empirically falsifiable”. These now classic critiques of the a priori of the empiricist, naturalist and behaviorist persuasion show us where deficiencies lie; making up for these deficiencies with a methodical theory of action proves to be a way to justify a new theory of the a priori.

3. Examples of A Priori Knowledge in Science, Philosophy, and Everyday Life We shouldn’t underestimate everyday common sense. That encountering ever so many white swans might suggest but not compel the opinion that all swans are white, and a single black swan disproves this opinion, is an understanding we can expect every philosophical layperson to have. A centerpiece of Karl Popper’s philosophy, namely his criterion of science (outside of the formal sciences of logic and mathematics) according to which scientific knowledge has to be able to be refuted by experience, is common sense. But everyday common sense doesn’t conclude from this that only that which can be refuted by experience can be certain knowledge – quite the contrary. Many connections seem to us to have a quite unshakeable certainty, despite or even because of the fact that they cannot be refuted by experience. I will discuss three examples of this, which at a first approximation all belong to the lifeworld and überzeugend nachgewiesen werden. Kants Beispiele … halten einer eingehenden Kritik nicht stand. … Ähnliches gilt für alle anderen Vorschläge”.

68

Peter Janich

common sense but can be more precisely correlated with the lifeworld, science, and philosophy. These examples are (1) the alibi principle, (2) the three-dimensionality of the space of experience, and (3) the law of non-contradiction. (1) The “alibi principle” means that a person can be in the same place at two different times but cannot be in two different places as the same time. I call this conviction a “principle” because it serves often enough as the initial postulate for considerations with far-reaching consequences, such as when we follow the plot of a suspenseful murder mystery on television. This principle is so solidly established that it is even anchored in law: any defendant in a court of law who tried to build his defense on the fact that person P would have had to be in two different places at the same time would fail. And if the defendant insisted on this, the judge might even order a psychological assessment or bar the accused from the courtroom for unruly behavior. The alibi principle is non-negotiable. It only has any competition in the supernatural sphere of a personal deity’s omnipresence, but nowhere among those things that are accessible to the understanding. And Popper’s criterion? How should we imagine the alibi principle being refuted by experience? If we did in fact come to observe a person sitting at the same time on two different chairs on different sides of the same table, for example, we would tend to say that we were looking at twins who are so astonishingly identical that we could not find the slightest difference between them. In other words, we would conclude from the alibi principle that we were dealing with two different people rather than one person in two different places. We could continue to play out this thought-experiment and it would only lead to the result that every empirical test of the alibi principle presupposes the alibi principle itself in the construction of the test situation. Hence to all appearances the principle seems to meet two conditions that, aside from all philological subtleties, fit Kant’s concept of the a priori: it is a knowledge that cannot be refuted by experience and is a condition of the possibility of experience, and in addition it enjoys strict universality (since it holds for all people) and necessity (since it cannot be otherwise). But where do we get this knowledge from? Is it even “knowledge” at all? After all, the countless people who have seen murder mysteries on television didn’t have to learn the alibi principle in a philosophical school or anywhere else. And a follower of the majority philosophies

A Priori Knowledge in Methodical Philosophy

69

of the 20th century with their logical empiricist orientation, who might suspect it to be an analytically true proposition, would have a hard time saying which generally binding definitions or predicate rules give us our conviction about alibis. Since an a priori knowledge must obviously (a) be present in linguistic form (as is the case above) and (b) be justified as knowledge, we need to look into the justification of the alibi principle. But before we go into that, let us first take up the two other examples: (2) The three-dimensionality of the space of experience is, at first glance, just as much a part of our everyday knowledge. Every layperson is familiar with the fact that a box has length, width and height, that every person refers to him- or herself in distinguishing forward and backward, right and left and up and down, and that we measure strings and routes according to length, paper and land according to area, and liquids and hollow spaces according to volume. When we talk about the form, the size, and the position of bodies and their movements, for example, in looking at the plans for a house or the schematic views of a car in brochures, no-one would be satisfied with two planes of projection and no-one would expect more than three. Our everyday conviction of the three-dimensionality of our world of experience is not just incidentally the opinion of lay-persons. Technology and the natural sciences also hold to it; from the carpenter building a roof to the engineer planning and constructing the most sophisticated machines for use in scientific laboratories, all choose a three-dimensional description of the particular object. And where physics speaks about “nature” in the narrower sense as that not produced by humans, such as when astronomy describes the position and movement of the heavenly bodies, it always requires three coordinates, regardless of whether it involves a Cartesian coordinate system with three axes in perpendicular pairs or a polar coordinate system. We can say the same or similar things about our conviction of the three-dimensionality of the material world as we did about the alibi principle, for example: the thesis of three-dimensionality cannot be refuted by experience. It also cannot be otherwise. It holds for all bodies. It serves as a precondition of experience, such as the experience of the real shape, position, or movement or bodies. And it does not hold analytically, i. e. it doesn’t follow from any definitions of “body”, “space”, or even for “form”, “size”, “position”, or “movement”. This again raises the question of how our conviction of the threedimensionality of the reality we experience is justified. The hope that

70

Peter Janich

the natural sciences might give us an answer proves to be misguided, since all historical attempts to accomplish this, whether by reference to the construction of our bodies, our perceptual organs, our organ of balance (the vestibular system of the inner ear), or the stability of the planetary orbits, or the relations between tactile and visual perception, have all been carried out in the language of a three-dimensional description of the organ or environment in question.36 Hence the question arises here as well as to whether the three-dimensionality of space might not consist in some collective delusion or in a dominant ideology or a metaphysical dogmatism and perhaps doesn’t represent knowledge at all, and whether it is justifiable. (3) The “law of non-contradiction” – in its standard formulation that something cannot be the case and not be the case at the same time – likewise has the character of a principle, and even a prescriptive character. Of course this doesn’t concern the subtle or inventive rhetorical means of drawing the listener’s attention to a distinction one has offered and arguing that a certain decision is both just and unjust if we see it from two different perspectives; or that something can be red and not red at the same time, because it is changing in time, is only red in some places or proves to be red if we see it in a shifting light, while shifting position, etc. Here I am also not concerned with the question we will need to discuss later as to whether the law of non-contradiction is a final principle of justification in logic or even of human language or epistemology, or other questions that professional philosophers may have raised. Rather the point is that in our lifeworld, in conflicts between people of differing opinions or interests, we do not allow someone to claim concerning anything of importance that something is the case and simultaneously is not the case. Of course this sounds more like a regulation or rule than knowledge. The law of non-contradiction seems at first to be a prohibition on contradicting oneself when speaking, which can occur in quite diverse ways. But it cannot be just a rule or regulation, since if one of the parties to a discussion tried to provide empirical evidence for a contradiction, we would answer that we know that something can’t be the case and not the case at the same time. In other words, if we try to imagine as 36 An overview of the attempts to establish three-dimensionality as an empirical result can be found in Janich (1992), the English translation of Janich (1989).

A Priori Knowledge in Methodical Philosophy

71

realistically as possible a dispute between two people where one of them argued that the law of non-contradiction is just a rule or a convention that he didn’t plan on following, the other person would surely object that it isn’t just an arbitrary rule or convention and that it just couldn’t happen that something is the case and not the case at the same time. We know that. Any knowledge to the contrary, even empirical knowledge, is impossible. In short, the law of non-contradiction also meets the same conditions mentioned above with the alibi principle and the three-dimensionality of our experience of space. It has the character of a principle, it first enables experience and is not refutable by experience, and it enjoys strict necessity and universality. However, here the problem of justification seems even more difficult, since we lack tangible objects like persons and places or forms, sizes and positions of houses. All three examples share the status of unshakeable conviction. Moreover, these convictions are linked with an intuitive assessment that they have a certain independence from their specific formulation: exactly what words we use to describe these matters of the alibi, the three dimensions of space or non-contradiction is besides the point; and this is why the proposal to derive these items of knowledge “analytically” from our uses of language is so unattractive. Hence it is a great surprise, in the realm of everyday life above all, that so many naturalists among scientists and philosophers so dauntlessly proclaim the failure of a priorism while so steadfastly practicing it themselves in their everyday life just as much as their non-naturalist contemporaries. The problem of a priori knowledge takes its great significance among other things from the fact that it allows us to recognize naturalistic convictions as an error-riddled superstructure built atop modern science.

4. Can A Priori Knowledge Be Justified? I would like to propose understanding a priori knowledge as knowledge of our own actions and their consequences. The a priori does not have its home either in the reality of nature and natural law independent of us nor in the eternal realm of principles independent of us, either of the ideal Platonic or the Kantian transcendental variety, but rather in the reality of our own action.

72

Peter Janich

This proposal cannot achieve very much as long as naturalists misunderstand human actions as natural objects or mentalists see action as the outer, perceptible expression of unchanging laws of thought independent of humanity. The novelty of this thesis depends on the concept of action built into it. I understand “action” as that which humans as communal creatures have to learn in terms of what the community ascribes to them as blame and desert. Humans do not act by nature but by culture; a “human” is here that creature that can only become human in communities; a human in the scientific and biological sense (as organism and genome-carrier) raised by wolves like Mowgli doesn’t learn to act or to speak. Such a creature only differs biologically from dogs or chimpanzees, but cannot be made responsible for any action either morally or legally. This theory of action (and a theory of language building on it) has been fleshed out elsewhere37 and cannot be repeated here. Just to prevent misunderstandings I will briefly distinguish it from alternative concepts of action. It is not an authenticist understanding of action according to which the agent decides authentically to call something an action when she could also not do it. Instead it is ascriptive, i. e. dependent on ascription in the context of real action within communities of action (and speech). Hence it is thoroughly dependent on culture and history, which means that we are not trying to argue by means of anthropological constants. “Action” is meant here first and foremost with reference to the civil society of, say, central Europe today; its extension to other or even all (historically known) cultures would have to be examined in a second step. A few distinctions from the theory of action are needed to justify a priori knowledge as a knowledge of action and its consequences, which have been discussed elsewhere and are just listed here. (1) Actions can come off or fail to. A person who has mastered an action can decide authentically whether the individual execution of the action has come off. In cases of publically regulated actions (from the various sports to music to the rules of games) whether it has come off (in the sense of “true” action, i. e. regulated actions) can be judged by other people, such as the judges awarding points in a dance competition. 37 Janich (2001).

A Priori Knowledge in Methodical Philosophy

73

(2) Actions can be successful or unsuccessful by achieving or missing their goal. Hence actions are always oriented towards goals; the goal could be to produce, maintain or avoid a state of affairs (represented linguistically as a proposition in an explicit reconstruction). We have to distinguish between an action coming off and an action succeeding, since actions that come off can also ultimately lead to failure, as experience teaches us. The gardener or the surgeon who does “everything right” in terms of their art, i. e. whose actions have come off, can still be confronted with failure when a natural event or the action of another agent come between their own action and the desired state of affairs. (3) Actions can be divided into (a) participatory actions, (b) communal actions and (c) individual actions. (a) Participatory actions such as races, chess, or communication can only come off when at least one other person takes part. Someone who runs in the race or plays chess by the rules can lose. Someone who communicates understandably can still fail in her demands, questions, claims, or promises. But participatory actions can only be carried out when at least one other person takes part; and in this sense they can only come off as participatory actions. (b) Communal actions can be carried out alone and come off – such as finding the right way to hold a beam too heavy for one person. Here whether or not the action comes off does not depend on the participation of others, but its success does. (c) Individual actions are those that can be carried out by the agent alone and can come off and succeed; hence they are neither participatory nor communal actions.

5. Justifications These three examples of a priori knowledge can not only be ascribed to the realms of the lifeworld, the sciences and philosophy; they were also chosen so as to correspond to three different types of action. In the theory of action we can distinguish (1) kinetic or movement actions, (2) poietic or productive actions and (3) practical or relational actions. This should be seen as the aspects under which actions can be described rather than as a strict ontological disjunction; a single action could unite all three aspects. For example, someone who writes a thank-you letter

74

Peter Janich

with her hand is carrying out kinetic actions (the movements of writing) as well as poietic actions (the production of the piece of writing) and practical ones (the performance of thanking). The delimitations between these three aspects are not trivial. For example, it is obvious that there are no poietic actions without kinetic actions; yet distinguishing poiesis from kinesis is no simple matter. Here I propose understanding poietic actions as those that lead to a product (the result or consequence of the action) that in turn can be used as a means for a further action (one’s own or another’s). This describes not just the classic example of a carpenter producing a piece of furniture, but also, for example, the organizing of books on a shelf (for the purpose of finding them quicker later). The practical or relational actions also always have a kinetic part and do not have to be limited to linguistic actions. For example, someone who aims a gun at another person is carrying out a kinetic action that is also a relational action (a threat). For the justification of the three examples of a priori knowledge, kinesis is particularly relevant for the alibi principle, poiesis for three-dimensionality and practical actions for the law of non-contradiction. (1) People’s bodily movements can be divided into inborn behavioral patterns (such as reflexive movements when someone trips or slips) and culturally acquired movement actions. Even our everyday movements like walking, running, swimming, and biking have to be learned; specific cultural movements like eating with silverware, writing with the hand, playing music on an instrument, and the movements of dance or of various sports are often acquired through a protracted process of learning. They satisfy the criteria of actions, namely that they can be ascribed to us as desert (such as with the thank-you letter) or blame (threatening someone with a gun). The most elementary items of our kinetic education include commands and above all prohibitions. Small children shouldn’t go too close to water or play on the riverbank, no-one should cross the street without looking, and young boys shouldn’t climb around in crumbling ruins. Simple movement games with children or among children also serve to practice talking about such movements. We take here walking from A to B as an important case. It makes very little sense to think here in the style of classical mechanics and look first at a theory of space, then a theory of time and finally at a theory of movement. Rather movements are used to learn spatial, temporal and movement-specific distinctions together and at the same time. Hence we can say that the places A and B only differ for a person by virtue of the fact that one has to move

A Priori Knowledge in Methodical Philosophy

75

oneself by walking for a certain time to get to B from A; i. e. the difference between places is first distinguished (without any more specific determinations) by the fact that it takes a certain time to move from one to the other. In all typical living conditions, returning to the place one started is one of the routine cases of walking. Someone who does this unintentionally, not when taking a walk or running some errand but for example when we get lost in a snowstorm and suddenly run into our own footprints and realize we have gone “in a circle”, will consider this return to the same place as lost time (relative to reaching the shelter, for example). These two examples (walking from A to B and returning to the place one started) already describe the elementary movements from which we can derive the fact that one can be in the same place at two different times but cannot be in two different places at the same time. The alibi principle is a simple consequence of the way that, in speaking, we organize and judge our kinetic movements of walking according to time, place and speed or direction (which I have left off here). Now someone might object that in this case the alibi principle just is an analytically true sentence, since it follows from linguistic conventions concerning time, place, and movement. But this trivial syntactic perspective is too short-sighted. The question here is precisely just what these linguistic conventions rest on if they are not merely conventional, i. e. completely arbitrary – not to mention that we do not form any theoretically coherent and consistent terminologies with definitions, explications, etc. in our everyday life. The decisive point is that here in a reflective philosophical text the quasi-theoretical linguistic form of this description of our everyday movements concerns an object that relates to the purposive organization of the interrelations between linguistic and non-linguistic actions. If we wish to communicate among each other unequivocally about routes, hikes, hunts, etc., we will do well to distinguish place and length of time in our own movements of walking. Moreover, it is important that we are not discussing the time and place of bodies (in a physical sense) such as in a Cartesian program, but rather of acting persons. The question of whether, say, a billiard ball could be at two different places at the same time is initially secondary, and only makes sense in relation to contexts of actions, i. e. of producing and moving these bodies. In these contexts we will want to in-

76

Peter Janich

troduce further specifications, such as that two bodies cannot be in the same place at the same time because as Cartesian res extensa they displace each other. And should someone produce two identical bodies indistinguishable even by the subtlest scientific and technological methods, and then claim that we had here one body in two different places at the same time, our ways of talking about the movements of people will give epistemic priority to the difference in place, which will suffice, in the absence of other differences, to justify the claim that we are dealing with two different bodies and not with one body in two different places. Hence the alibi principle is not a proposition that rests on other propositions of natural law or on empirical propositions (in the sense of propositions that can be refuted by counter-examples). Rather the alibi principle is a knowledge of actions and their consequences prior to all (spatio-temporal) experience. The question of whether it “could be otherwise” than the alibi principle claims, and suspiciously psychologistic-sounding formulations such as “we cannot imagine it being otherwise”, can be explained by the fact that the alibi principle is in a certain sense dependent on culture, namely in that it is bound to language, and yet on the other hand (according to everything we know) it is highly culturally invariant. But it is quite possible that a culture whose everyday practices included putting themselves into a trance and then assuming magical bi-locations or something similar might only ascribe limited validity to the alibi principle – thus refuting the psychology of “it cannot be otherwise”. (2) The three-dimensionality of our experience of space, as it holds not just of our lifeworld and our technology but of our technical natural sciences as well, can be justified by a reflection upon poietic action. Here I have to limit myself to a few brief remarks, since I have presented the corresponding theory of three-dimensionality (including a discussion of a great number of alternative attempts to justify three-dimensionality) in book-form.38 The justification centers on the prototype-free reproducibility of the basic forms of the plane, the straight edge, the right angle (between planes or between edges) and parallelity. “Dimension”, which translates etymologically as “diameter”, is described visually by the fact that any natural body (such as a pebble) can be enclosed within a cuboid box such that it touches all six surfaces of the box. The three different edge lengths of the box are then the three dimensions. 38 Janich (1992).

A Priori Knowledge in Methodical Philosophy

77

Of course one must already know that at least and at most three extensions are necessary to make such an enclosure technologically reproducible. And this in turn requires us to specify and prove the proposition that a body can only be cut such that the resulting parts are identical in form exactly three times. (Here we can imagine eight pieces, each with the corner of a cube, put together into a solid body that is then “cut” by three planes forming three perpendicular angles.) This result (and its proof) is more than trivial insofar as Euclid’s proof of the existence of the five Platonic solids (regular polyhedrons) is equivalent to the claim that a body can be divided in five different ways into pyramids with the same shape, and yet nonetheless we still do not refer to these bodies as five-dimensional. However, this “Platonic analysis” is not methodically primary, i. e. it cannot be produced poietically based on the literal cutting of a natural body with a knife. Instead the pyramids that make up the Platonic solids can only be constructed and produced on the basis of an operatively justified geometry with precisely reproducible forms.39 Here we have to ask once more into the goal that distinguishes our practice of establishing the three-dimensionality of space of our experience as a consequence of our technological interaction with what can be found in the world. This goal is generally to make unequivocal statements about the form, size and position as well as the changes in position of solid and hollow bodies. Hence the a priori knowledge that the things of our everyday experience and our technological and scientific experience are three-dimensional stems from our goal-directed practice, as inventors, producers, and users of everyday objects (from the cup to the computer keyboard), of establishing precise relations. In this sense three-dimensionality as a priori knowledge is very much culturally relative. But this again is also highly culturally invariant, though primitive technological practices that might only distinguish the spatial form of the taut bow-string and do without planes, straight edges, right angles, circles, etc., will hardly manage to arrive at a conception of the three-dimensionality of things. I have also explicitly dealt with the law of non-contradiction as a priori knowledge elsewhere and can only offer a brief summary here.40 The contradictions to be avoided can be divided into (1) per39 For all further details, particularly the uniqueness proof for three-dimensionality, see the literature mentioned in fn. 36. 40 Janich (2005a), 105 – 134.

78

Peter Janich

formative, (2) predicative and (3) logical contradictions. The performative contradictions are of particular interest to us here, since a contradiction between the performance and the description of our action is seen as the most important of these. Our methodical reconstruction starts with the foundations of the theory of action in order to clarify the connection between the performance of an action and its description. Here we turn to the language of commands in order to keep semantic problems clearly distinct from problems of validity. The methodically simplest reconstruction of the performative contradiction is when two people A and B give a third person C commands such that C cannot fulfill both commands at the same time. And the elementary instance of this performative contradiction is when A commands C to do action a and at the same time B commands C not to do a. In such a case C cannot follow both commands simultaneously; hence she becomes speaker of a meta-language and rejects both commands as “contradicting one another”. Hence this example leads to a reconstruction of the claim that a person cannot simultaneously perform an action and not perform it. With predicative contradictions, which have played a significant role in philosophical literature from Aristotle to Kant, it is crucial to see that we cannot prematurely assume that we are dealing with a negation or with the “denial” of a predicate in the standard form of an elementary proposition discussed so often within the Vienna Circle. Rather there is an open-ended number of “excluders” (the generic term for various sorts of negation) that can enter into an elementary (i. e. logically non-composite) command. In simple examples: John, throw and don’t throw! Throw the ball and not the ball, the red ball and not the red ball, throw it far and not far, etc. This means that here at least proper predicators and appredicators and the entire breadth of further distinctions need to be considered, including semantic transgressions (e. g. attributing person appredicators to non-persons). (For terminology and technical details see the book of mine mentioned in note 8.) Finally, for logical contradictions we need to remember the fact (which every logician is familiar with but is usually overlooked in the tradition) that a “logic” requires that we establish or distinguish both operators (such as junctors and quantors) and criteria for distinguishing between what is logically valid and logically invalid. Hence the prohibition on logical contradiction is not a universal and trivially clear principle, but rather depends on the particular logic (which we would hope is chosen in accordance with our purposes).

A Priori Knowledge in Methodical Philosophy

79

The goal to be reached by accounting for the prohibition on performative, predicative and logical contradictions is the coherence of our practice of speech and action. Commands lose their point and hence miss their goal if they can be allowed to occur in such a way that they cannot be followed due to their form or their content. This holds both for the three-person example above and the case of a contradictory self-command that we could derive from it. In other words, all three forms of contradiction, applied for now just to commands, prove to be counterproductive, impracticable, pointless. Hence the legitimation and justification of a priori principles in this area of philosophical examination rests on the purposiveness of human speech in communicating for the sake of human cooperation. With this we can consider all three examples of our store of a priori knowledge to be justified. All three prove to be culturally dependent and yet highly culturally invariant. They have no basis in the laws of nature, in natural history or in a philosophical heaven of ideas. They are a cultural accomplishment that grows out of the interaction of linguistic and non-linguistic speech in kinesis, poiesis and praxis. And they have their place in the lifeworld, in the sciences and in philosophy.41, 42

References Carnap (1969): Rudolf Carnap, Einfhrung in die Philosophie der Naturwissenschaft, München. Dingler (1928) Hugo Dingler, Das Experiment. Sein Wesen und seine Geschichte, München. Einstein (1965): Albert Einstein, Grundzge der Relativittstheorie, Braunschweig. Grunwald (2000): Armin Grunwald, Handeln und Planen, München. Gutmann (1996): Mathias Gutmann, Die Evolutionstheorie und ihr Gegenstand. Beitrag der methodischen Philosophie zu einer konstruktiven Theorie der Evolution, Berlin. Hanekamp (1997): Gerd Hanekamp, Protochemie. Vom Stoff zur Valenz, Würzburg. Hartmann/Janich (1996): Dirk Hartmann/Peter Janich (ed.), Methodischer Kulturalismus: zwischen Naturalismus und Postmoderne, Frankfurt/Main. Hartmann/Janich (1998): Dirk Hartmann/Peter Janich (eds.), Die Kulturalistische Wende. Zur Orientierung des philosophischen Selbstverstndnisses, Frankfurt/Main. 41 This paper is a revised version of the text: Peter Janich, “Apriorisches Wissen,” in: Rentsch (2005b). 42 Translation from German by Karsten Schöllner.

80

Peter Janich

Hartmann (2003): Dirk Hartmann, On inferring. An enquiry into relevance and validity, Paderborn. Helmholtz (1884): Hermann von Helmholtz, ber den Ursprung und die Bedeutung der geometrischen Axiome, Braunschweig. Helmholtz (1968): Hermann von Helmholtz, ber Geometrie, Darmstadt. Janich et al. (1974): Peter Janich/Friedrich Kambartel/Jürgen Mittelstraß, Wissenschaftstheorie als Wissenschaftskritik, Frankfurt/Main. Janich (1989): Peter Janich, Euklids Erbe. Ist der Raum dreidimensional?, München. Janich (1992): Peter Janich, Euclid’s Heritage: Is Space Three-Dimensional?, Boston, London. Janich (1997): Peter Janich, Das Maß der Dinge. Protophysik von Raum, Zeit und Materie, Frankfurt/Main. Janich (2001): Peter Janich, Logisch-pragmatische Propdeutik. Ein Grundkurs im philosophischen Reflektieren, Weilerswist. Janich (2005a): Peter Janich, “Der Satz vom Widerspruch”, in: P. Janich (ed.), Kultur und Methode. Philosophie in einer wissenschaftlich geprgten Welt, Frankfurt/Main, 105 – 134. Janich (2005b): Peter Janich, “Apriorisches Wissen”, in: T. Rentsch (ed.), Einheit der Vernunft? Normativitt zwischen Theorie und Praxis, Paderborn, 61 – 85. Kant (1789): Immanuel Kant, Kritik der reinen Vernunft. Kamlah/Lorenzen (1967): Wilhelm Kamlah/Paul Lorenzen, Logische Propdeutik. Vorschule des vernnftigen Redens, Mannheim. Lange (1999): Rainer Lange, Experimentalwissenschaft Biologie: methodische Grundlagen und Probleme einer technischen Wissenschaft vom Lebendigen, Würzburg. Lorenz (1941): Konrad Lorenz, “Kants Lehre vom Apriorischen im Lichte gegenwärtiger Biologie”, in: Bltter fr deutsche Philosophie 15, 94 – 125. Lorenz (1977): Konrad Lorenz, Die Rckseite des Spiegels. Versuch einer Naturgeschichte menschlichen Erkennens, München. Lorenzen (1987): Paul Lorenzen, Lehrbuch der konstruktiven Wissenschaftstheorie, Mannheim/Wien/Zürich. Popper (1971): Karl Popper, Logik der Forschung. Preface to the English edition, Tübingen. Psarros (1999): Nikolaos Psarros, Die Chemie und ihre Methoden, Weinheim. Quine (1975): Willard van Orman Quine, “Naturalisierte Erkenntnistheorie”, in: W. v. O. Quine, Ontologische Relativitt und andere Schriften (Ontological relativity and other essays), Stuttgart. Ruse (2009): Michael Ruse (ed.), Philosophy after Darwin: Classic and Contemporary Readings, Princeton. Schepers (1971): Heinrich Schepers, “Apriori/Aposteriori I”, in: J. Ritter (ed.), Historisches Wçrterbuch der Philosophie, vol. I, Darmstadt, 462 – 467. Schonefeld (1999): Wolfgang Schonefeld, Protophysik und spezielle Relativittstheorie, Würzburg. Stegmüller (1973): Wolfgang Stegmüller, Probleme und Resultate der Wissenschaftstheorie und analytischen Philosophie (vol. II: Theorie und Erfahrung, 2nd

A Priori Knowledge in Methodical Philosophy

81

half-volume: Theoriestrukturen und Theoriendynamik), Berlin/Heidelberg/ New York. Vollmer (1993): Gerhard Vollmer, Wissenschaftstheorie im Einsatz. Beitrge zu einer selbstkritischen Wissenschaftsphilosophie, Stuttgart. Wille (2002): Matthias Wille, Das Parallelenproblem in der Protophysik, Master’s Thesis, Marburg. Wille (2007): Matthias Wille, Die Mathematik und das synthetische Apriori. Erkenntnistheoretische Untersuchungen ber den Geltungsstatus mathematischer Axiome, Paderborn. Wille (2008): Matthias Wille, Beweis und Reflexion. Philosophische Untersuchungen ber die Grundlagen beweistheoretischer Praxen, Paderborn. Weingarten (1993): Michael Weingarten, Organismen – Objekte oder Subjekte der Evolution? Philosophische Studien zum Paradigmawechsel in der Evolutionsbiologie, Darmstadt.

Knowledge and World-Pictures Ulrich Dirks Of all our knowledge and of all the ways we use the term ‘knowledge’ we rightfully accord the knowledge expressible in linguistic propositions a special status. Knowledge of the form “I know that the drug store around the corner is open all night” is ubiquitous in everyday situations; and the sciences, where for example “physicists know that electrons have a half-integer spin”, contain vast realms of this sort of knowledge. In order to count conceptually as knowledge, propositions of this form have to be connected to successful procedures for justifying and redeeming their claims to validity and truth, and in the sciences they have to meet high standards, particularly standards of methodical and intersubjective review. However, it is questionable whether this distinguishing characteristic entitles us to claim that propositional and specifically scientific knowledge has a relation (e. g. of correspondence) ‘from itself alone’ to world and reality or a systematic precedence over other modes and manners of knowledge (e. g. the non-linguistic or non-propositional knowledge given in abilities, competences and practices). Hence it is all the more important for knowledge research to seek an appropriate understanding and a conceptual clarification of knowledge in the narrow sense meant here (and in the following I will typically refer to this as ‘knowledge’ without qualification).1 Insofar as this reflective effort will lead us to the presuppositions and logical preconditions of knowledge, the following paper will take up a part of this task by analyzing the relation between knowledge in the narrow sense and our ‘world-picture’ (a term taken from the German Weltbild 2). In particular 1

2

For the distinction between the narrow sense of knowledge described here and a broad sense (including abilities, competences, skills, practices, and know-how above all) and diverse forms and characterizations of knowledge see Günter Abel’s paper in this volume. See also Abel (2004), ch. 10. The German term Weltbild has no exact correlate in English. It has become standard in translations and discussions of Wittgenstein, Heidegger, and topics in the philosophy of science to render it as ‘world-picture’, which I will use here. Hence the use of ‘world-picture’ in this text is a purely terminological use determined by the characteristics described herein.

84

Ulrich Dirks

this will concern a series of quite diverse functions that the world-picture performs in and for epistemic processes, the question of how the prescientific and extrascientific background of our world-picture and its various components (which are by no means exclusively propositional) are constituted and given, and the complex internal ties between knowledge, world and world-picture that are constitutive of knowledge. I am guided here by a reconstruction of certain suitable lines in the later philosophy of Ludwig Wittgenstein and in Günter Abel’s general philosophy of signs and interpretation insofar as both positions develop and apply a concept of ‘world-picture’ that proves to be fruitful for our problem and sufficiently broadly conceived and fundamental (as it ultimately relates to a basis of our speech, thought, action and thus knowledge as well).

1. ‘World-picture’ Like all natural language expressions, the German term ‘Weltbild’ gets used in quite different ways in various contexts and discourses.3 Firstly, for our purposes here of analyzing the complex relation between knowledge and world-picture it is not meant primarily as the merely visual representation of ‘the world’ (from the medieval images of a spherical cosmos to the photographs from the Hubble space telescope). Among other things, such a narrow reading would obscure the fact that we can ascribe world-pictorial status to linguistically expressed beliefs. On the other hand the meaning of ‘world-picture’ should not be confused with the idea of the sum total of all explicit propositional knowledge about reality present at a particular time. This notion of a body of available knowledge of the world expressed entirely in propositions, the core of which consists of something like the encyclopedic sum of the justified beliefs of natural science, omits all those components of the ‘scientific world-picture’ that relate to non-propositional, non-linguistic, non-explicit and non-theoretical knowledge and that are of cardinal importance for the relation of knowledge and the world-picture. The modern empirical sciences do not just generate pieces of knowledge formulated in propositions about the spatio-temporal reality around us; they also help to form a total understanding of the cosmos that is unified 3

For a history of the concept Weltbild see Gethmann-Siefert (1996b), Mies (1999b), Stegmaier (1997) and Thomé (2005b).

Knowledge and World-Pictures

85

in a certain sense, a view of the world through a vivid picture or model.4 In this sense we see ‘world-pictures’ arise that consolidate current scientific knowledge but that also implicitly entail the dominant normative methodological stipulations and paradigmatic standards for the acquisition of scientific knowledge, and moreover that give an image to the world (based not least of all on philosophical interpretations of the scientific theories) and can integrate and orient the diversity of people’s life experiences within a certain culture and a certain epoch. Secondly, I will not primarily be using ‘world-picture’ in the sense of an individual’s merely subjective worldview or a stereotypical worldview, e. g. as in “Mr. X-Y has his own strange view of the world” or when we speak of the specific viewpoint of the publishers of financial journals. Rather the world-picture that dominates in a particular culture at a particular time is essentially connected to a broadly shared vision of the world, as the Newtonian world-picture was correlated with a mechanistic worldview. Thirdly we must distinguish conceptually between ‘world-picture’ and Weltanschauung. 5 Weltanschauung in its philosophical and its everyday use typically refers to a conception of the world that not only explicitly addresses the role of humanity in the world (which world-pictures often do as well) but that aims above all at answering questions of meaning from within belief systems. The Weltanschauung concerns the world in relation to its ground or purpose, in relation to the meaning of the world and our existence in it.6

4 5 6

Cf. Abel (2004), 120 ff. For a history of the concept of Weltanschauung see Gethmann-Siefert (1996a), Laeyendecker (1997), Mies (1999a), Thomé (2005a) and Müller (2010). Wilhelm Dilthey’s theory of the Weltanschauung is a thorough and influential example of a philosophical distinction between world-pictures and Weltanschauungen. On Dilthey’s conception a world-picture arises when the clarification of perceptions by elementary thought processes leads to a conceptual world that is organized beyond the mere contingencies of perception and that makes clear the basic circumstances of reality, which leads by means of concepts and judgments about the objects to a conception of reality that “comprehends with universal validity the nexus and essence of reality”; the result of which is a vivid and unified world-picture pertaining to knowledge of reality (Dilthey (1968), 83). It is then only on the basis of this world-picture that a Weltanschauung can arise – as Dilthey explains on the basis of his life philosophy, taking into account the lawful structure of human spiritual life – when it leads to answers to “questions about the meaning and sense of the world” from which we

86

Ulrich Dirks

It will help us conceptually in identifying the presuppositions, conditions and dynamic of knowledge and specifically scientific knowledge if we take up the later Wittgenstein’s terminological use of ‘world-picture’. In his notes published posthumously under the title On Certainty 7 he writes: “But I did not get my picture of the world [Weltbild] by satisfying myself of its correctness; nor do I have it because I am satisfied of its correctness”; the world-picture, as Wittgenstein characteristically describes it in this context, is “the inherited background against which I distinguish between true and false”8. This broad and fundamental sense of ‘world-picture’ sets the true field for the discussion of the relation between knowledge and world-picture. Within the contemporary discussion it has laid the course for such positions as Günter Abel’s general philosophy of signs and interpretation. Abel sees the world-picture as a thusly ‘inherited’ and not deliberately chosen background, as the “unquestioned, public and intersubjectively shared foundation of our speech, action and thought (in science as well) that underlies a certain culture at a certain time”9. Abel is then able to situate a narrower sense of ‘world-picture’ within this broad and deep sense. The narrower sense accounts for the circumstance already mentioned, namely that “for the people of a culture and epoch the entirety of their diverse life experiences” coalesce “in a certain unified view of the world in a picture or a clear model”, i. e. a world-picture in the narrower sense.10 Insofar as it is particularly the cosmological world-pictures of an era that include a certain conception or ‘picture’ of the cosmos, they represent prominent examples of this narrow type of world-picture, as does Carl Friedrich von Weizsäcker’s famous work Zum Weltbild der Physik (The World View of Physics).11

7 8 9 10 11

derive an “ideal, highest good” and the highest order of practical circumstances of life, e. g. “ultimate principles for living” (Dilthey (1968), 82). Wittgenstein (1975), from here on abbreviated as OC followed by the paragraph number. OC 94. Abel (2004), 119. Abel (2004), 120. Weizsäcker (2002). Reconstructions and discussions of world-pictures primarily in this narrow sense can be found e. g. in Gebhardt/Kiesel (2004) and Silvar/ Kunze (1993).

Knowledge and World-Pictures

87

2. Knowledge Propositions and World-picture Propositions Propositions like “there are 60 people in this room” and “it’s going to rain in Berlin today” refer to empirically graspable reality. Insofar as the semantic characteristics of each include a series of world-related conditions of satisfaction, the sentences can be confirmed by reviewing actual circumstances and can also be refuted by the facts. But we also find sentences that do not differ at all from such empirical or experiential propositions in their outer grammatical and syntactical form, but that would lead to a sustained confusion if we were to understand them and treat them like ordinary empirical propositions. Wittgenstein draws our attention to seemingly everyday sentences (such as G. E. Moore had already mentioned in his discussions of scepticism and the role of common sense), sentences of the sort: “I have two hands”, “Everyone in this room has great-grandparents”, or “The world existed before I was born”. In the first-person perspective we are obviously not generally convinced that we have two hands because we can test it empirically. If someone were to ask me whether I have two hands, he would be quite astonished, despite the real possibility that I might be missing a hand, if I were to explain that I had observed it this morning in a methodologically proper manner as I do every morning, rather than just saying “yes”. How can we trust our own eyes if in a run-of-the-mill situation we find it necessary to question and verify the condition of having two hands in the manner of an external empirical circumstance? And the sentence that all present have great-grandparents also loses the appearance of an empirical statement on closer examination. If a speaker were to deny it, in ordinary circumstances this would show us not that it is possible for such an understanding of self and others to be falsified in experience, but rather that the speaker is obviously not competent in the use of the concept “great-grandparents” and hasn’t really understood it. The attempt to classify such sentences as propositions expressing empirical knowledge would also mean that I could be mistaken about such a claim – which follows even conceptually insofar as ‘knowledge’ always refers to humanly finite, perspectival, epistemic and in this case linguistically formulated knowledge of a certain time, hence knowledge that is in principle revisable. Yet for these and comparable examples we can hardly imagine how they could reasonably become the object of skeptical doubt beyond purely syntactic “why”-prattle. The possibility of such error or doubt – and hence the possibility of attributing

88

Ulrich Dirks

truth-values to such a sentence – would have enormous consequences for my entire understanding of myself, others and the world. My own lineage, my idea that my grandparents had parents, my knowledge of biological reproduction, my notions about things arising from nothing and divine creation, and my knowledge of the course of European history would all be immediately affected, to name just a fraction of the things this would impact. Because of how densely interwoven the sentence is with a number of other beliefs that have served as the basis of my action so far and because of the holistic connections among these beliefs, such a doubt would be like a forest fire that would consume everything and take with it all the conditions and contexts that make the possible falsehood of a sentence understandable in the first place. It is clear that a great many of the sentences expressing empirical beliefs (and ultimately propositional knowledge as well) revolve around such sentences as around “rotational axes” or “hinges”, to use Wittgenstein’s metaphors.12 Yet it is not at all clear in advance (i) what, if anything, could produce conviction, if it were in fact possible for such rotational sentences to have truth-values and more precisely if they could turn out to be false. Indeed, (ii) it would shake my trust in the specific sources of epistemic authority and the testing procedures that are of the utmost practical and pragmatic relevance, i. e. in familiar processes of generating, justifying and revising knowledge. In this example, my trust in grave inscriptions, things my parents told me, what I learned at school, biology textbooks, and the study of history would be shaken to such an extent that the associated testing procedures would lost their conviction and appropriateness for me. No-one – from the daily lifeworld of the here and now – would seriously rely on exhumations and genetic testing if the result could show that one of us has no great-grandparents at all, hence if the knowledge that human individuals come from procreation were ultimately a revisable ‘knowledge-that’.13 This draws our attention to gradual relations within knowledge, in contrast to the idea of a bivalent distinction between knowledge and non-knowledge: as Aristotle already noted in his Posterior Analytics, 12 OC 152, 341, 343. 13 It is no objection to the point here that we could come to a quite different result in analyzing these examples by changing their normal and implicit contexts or by supposing that a quite different standard scientific theory holds in the future; this only shows that the status of rotational sentence cannot be claimed as a metaphysical absolute.

Knowledge and World-Pictures

89

the sentences we have recourse to in justifying propositional knowledge have to be not just logically prior to the sentence to be justified but also have more certain validity for us, be more solid, i. e. represent more certain knowledge. For example, in matters of my self-understanding, what more certain conviction could I appeal to in everyday contexts if I allow the possibility that I am not U.D., i. e. that I could be in error concerning my entire personal life history? Moreover, the efforts of memory manifested in my actions, the testimony of witnesses or the examination of source documents could hardly be acceptable as step-wise verification procedures for empirical statements, each supporting and supplementing each other in a complex manner and to various degrees, if I thought there was any need to verify the statement that the earth had already existed before I was born. Hence not all sentences stand equally in need of validation. Rather we also find “sedimented” sentences that, in Wittgenstein’s metaphor14, serve as a kind of river-bed that steer our experiential propositions but are not themselves experiential propositions and do not represent propositional knowledge. This metaphor also entails the thesis that empirical propositions can sink down and anchor themselves in the river-bed, and that other propositions might become “fluid” once more in light of changed ideas, i. e. turn to empirically reviewable sentences. Here we might think of the shifts in certain background components associated with the transition from a religious and mythological lifeworld to a lifeworld dominated by natural science and technology. In systematic terms this marks a very profound point concerning the relation between knowledge and time: not only is (i) everyday knowledge and the knowledge of the empirical sciences a knowledge of contingent and temporal states, relations and processes; (ii) and not only is this humanly finite knowledge a temporary and fallible knowledge that can prove revisable in light of new and divergent experiences and observations, the discovery of prior mistakes, or the introduction of new arguments; (iii) in addition, linguistically articulated knowledge is bound to the currently prevailing linguistic practice and the context of discourse given at that particular time (and the standards of rationality manifested in this); (iv) and finally, the question of whether a sentence in the (then current) form of an empirical proposition actually expresses empirical knowledge cannot be answered absolutely or with external criteria but rather only 14 OC 96 ff.

90

Ulrich Dirks

in relation to the background components of the world-picture – which can change with time.15 As explained in the preliminary conceptual remarks above, for Wittgenstein sentences that play the role shown in these examples fall under the concept of “world-picture” in the broad sense as the inherited background constituting the foundation of my speech, thought and action. Insofar as the discussion requires, the expression “world-picture proposition” can and should be differentiated into propositional background assumptions and sentences that belong to the linguistic portion of the world-picture itself, certainties that are situated very close to the world-picture, and sentences that accurately describe a world-picture (i. e. concerning its structure and its non-linguistic components).

3. The Functions of World-pictures in ‘Knowing-that’ The functions described so far that the network of world-picture sentences serves for us – anchoring chains of justification and founding and making available to us the distinctions, rules and justificatory procedures preceding every acquisition of an explicit and concrete understanding of the world – already support Wittgenstein’s thesis that the world-picture is “[a]bove all […] the substratum of all my enquiring and asserting”16, of the generation, articulation, transformation, verification and revision of knowledge or claims to knowledge. Hence worldpictures can be seen not least of all as a specific condition of scientific knowledge and as a decisive influence on the lifeworld practices from which we conduct science. This can be more precisely characterized by a series of aspects: 1. As we have seen, world-pictures serve as the “basis and guarantee of certainty”.17 World-picture propositions for their part form a foundation that is taken to be unquestionably certain at that time,18 and that 15 16 17 18

Cf. ch. 3, points 2 and 4 below. OC 162. Abel (2004), 123 f. The gradual term ‘certain’ obviously cannot refer to an absolute, metaphysical certainty that is epistemologically free of doubt once and for all, since seen critically this cannot be achieved or explicated under humanly finite, perspectival epistemic conditions. For people sharing the world-picture in question, it refers to a theoretical certainty (given initially in the first-person perspective) – the certainty of a foundation that cannot be sensibly doubted or suspended at that time

Knowledge and World-Pictures

91

serves as the basis for justificatory processes for other sentences claiming validity. World-picture sentences are able to exercise this functional role because they represent a limiting case of sentences. This means, among other things, that (i) epistemically they are taken for granted to such an extent that “in general no-one calls for them to be independently verified”.19 Relative to the background network in which or by which the individual world-picture sentence is anchored, nothing speaks in favor of claiming the contrary. (ii) At present we simply cannot see how they could be in error – particularly from the first-person perspective. Quite the contrary, for quotidian and scientific knowledge they form part of the reference system used in testing other, empirical sentences. (iii) Accordingly, when we are dealing with borderline world-picture sentences it is often in communicative situations where we increasingly respond to the other’s attempt to dispute a sentence with the conviction that he simply hasn’t quite understood the expressions we’re using, or is making a linguistic error,20 or is not understandably mistaken about the matter at hand in some way but rather might be suffering some passing “mental disturbance”.21 The formal negation of a world-picture sentence is more nonsensical than false. (iv) In the case of a sentence like “there are physical objects”22, this manifests itself in the fact that it cannot be refuted by experience, but is also not ‘absolutely true’, say in a metaphysical sense. Rather, in our epistemic practices we treat it as certain, since it describes the categorial form of experience, a form stemming not least of all from the grammar of our use of the concept ‘phys-

19 20

21 22

and of the undoubtable matter-of-course certainties that follow from this – and a lived practical certainty in the sense of an unquestioned trust in certain beliefs, e. g., and the practical certainty at work when we initiate actions or in speech and thought. Abel (2004), 125. When this is the case for a world-picture proposition it plays the important role of a ‘grammatical sentence’ in Wittgenstein’s terms (cf. Wittgenstein (1958), abbreviated from here on as PI followed by the paragraph number; here esp. PI 251), i. e. a sentence that shows the practices of use for the words it contains (and thus cannot be tested in experience). Cf. e. g. Lütterfelds (2008), 22 – 25. However, we should note that this differentiation among world-picture sentences and the diversity of functions that a sentence can have as world-picture sentence prevents us from simply classifying all world-picture sentences as such as grammatical sentences or identifying the two. OC 71 ff. Cf. OC 35 f.

92

Ulrich Dirks

ical object’; it describes our habituated pattern of organizing the world and reality relative to which we can dispose over knowledge. In addition it is crucial to note that a sentence cannot serve as a world-picture sentence in isolation, and in a two-fold sense: firstly, because of how tightly intermeshed our profound and unquestioned convictions about the world and reality are, the propositional portions of our world-picture represent interlocking clusters of world-picture sentences supporting and solidifying each other. This is so despite the fact that the principle impossibility of questioning our world-picture or of getting sufficient distance to attain an overview of the whole means that we cannot describe it as a closed system, never mind as a strictly logical and inferentially organized system. Secondly and in addition, worldpicture sentences and particularly rotational sentences are not given their status by an external and metaphysical final authority or an independent criterion. They take their role from the function they actually exercise in conjunction with beliefs about the world, a function that can be recognized through reflection. As Wittgenstein says, “I do not explicitly learn the propositions that stand fast for me.” Instead “I can discover them subsequently like the axis around which a body rotates. The axis is not fixed in the sense that anything holds it fast, but the movement around it” – hence also the process of making good on knowledge-related claims – “determines its immobility.”23 It should be noted that ‘discovery’ is systematically preceded by the constructive, cognitive invention24 involved in creating an abstract, mathematical and physical concept of a rotational axis and the corresponding modeling activities for its application to real situations, to stick with Wittgenstein’s image; and the same holds for the reflective concept of the world-picture proposition.25 Concerning the relation of science and the world-picture in a broad sense, we can already note that scientific propositions about the constitution of reality, e. g. the present-day mathematical description of orbits in the solar system, cannot be world-picture propositions strictly speaking. 23 OC 152. 24 This also bears an analogy to Wittgenstein’s position in the philosophy of mathematics, according to which the mathematician does not discover antecedent facts and objects so much as invent them; see Wittgenstein (1978), abbreviated in the following as RFM, here RFM I §168. 25 For the epistemic and linguistic holism about certainty found in both of these points (and others to come as well) see a more comprehensive treatment in Dirks (2007) and (2008).

Knowledge and World-Pictures

93

They are tested against experience and can fail the test, i. e. they can sensibly be corrected under the pressure of experience or rejected entirely, as they are descriptions of factually contingent circumstances and have the nomological character of merely hypothetical necessity. As Karl Popper and others have shown, the scientific nature of the propositions of empirical sciences is conceptually connected to their falsifiability. 2. Moreover, the world-picture also plays its fundamental role in epistemic dynamics by anchoring the shared methods of generating knowledge and the recognized procedures for testing and justifying epistemic beliefs that claim valid and objectively binding status.26 A survey of human eras and cultures reveals a diversity of ‘locally’ accepted and practiced ways of acquiring and justifying beliefs claiming the status of knowledge, including among other things oracles and systematic series of observations, computer-simulations and horoscopes, models and meditative practices, inferences by analogy and thought-experiments, ecstatic experiences and scientific experiments, mathematical theory and appeal to divine revelation, and many other instances of authority. For example, on this view the transformation in world-picture in the course of the formation of the modern natural sciences in early modernity, most prominently reflected in the cosmological picture of the world as a whole, went much deeper than just a revision and improvement in the mathematical description of the planetary orbits or the substitution of new scientific ideas about the spatial arrangement of the cosmos, as 26 Here we can turn to Kant’s distinction between the modes of holding a thing to be true: knowledge with its objective and subjective validity stems from the (significantly larger) field of epistemic belief, which has at least subjective validity for the person in question. This is preceded by opinion, which is not even subjectively valid. Cf. Kant (1787), B 848 – 858. On the fundamental importance of this distinction for knowledge research see Abel (2004), 329 ff. Our discussion should make it clear that the functions of world-pictures, particularly in generating degrees of certainty, primarily affect opinions and beliefs first and foremost (cf. the thesis in Abel (2004), 137 ff.). The background is still active and conditional for the step towards knowledge in many ways that will have to be discussed in this paper; the increasing elements of methodical, inter-individual and only objectively redeemable claims to validity involved in conceptual knowledge make the relation more indirect and mediated. For example, in knowledge, unlike belief and opinion, the aspects of fitting coherently into the background web of beliefs with a high degree of certainty play less of a direct and dominant role compared to generally accepted and shared empirical tests. (Moreover, for knowledge research this only increases the urgency and the challenge of explaining the relation between knowledge and the world-picture.)

94

Ulrich Dirks

some historical reconstructions claim today. The transformation also ran through the ‘world-picture’ in the broader and deeper sense meant here precisely in that it saw the replacement of justificatory mechanisms that were typical of the Middle Ages: ‘observation’ by humans had been seen as a defective, weak and error-prone way to get the most accurate knowledge of the world possible, as humans are fallible and finite creatures, both in their perceptual abilities and in their cognitive capacity to interpret what they see and process it in accordance with the understanding. As the view of the world as God’s creation was a central element of the medieval Christian worldview, the best hope for knowledge of the the laws of nature qua creation was clearly to examine not so much the phenomena themselves as the Word of Him who created them. Simply put, justified knowledge of the cosmos ultimately came by way of the interpretation of the Holy Scripture. And if this scripture contained the statement that God made the sun stand still, and thus it had to be inferred that the sun ordinarily moves, then this was much more solidly established by the world-picture than the heliocentric model of the cosmos that rested particularly on observation.27 What can be considered empirical knowledge at a particular time and thus can be tested against our experience is an internal question within the logical and semantic network made of both sentences serving (within that network) as empirical propositions and rotational sentences fixed in place within these relations (as well as any others, including non-lin27 Bertolt Brecht illustrated this basic point very clearly and helpfully in his play Life of Galileo: Galileo wishes to convince the traditional scholars and church representatives of his heliocentric system using his observations through the telescope. In particular his observation that moons revolve around the planet Jupiter was something that was simply impossible on the combination of Aristotelian physics and Christian worldview in the world-picture of the time. This confrontation at the level of world-picture (with an emphasis on the ‘rules of verification’) comes out in Brecht’s play when Galileo says to his critics: “I will seize them by their heads and drag them in front of the telescope. Even monks are human, Sagredo. Even they succumb to the temptation of proof. […] I only ask that they believe their eyes. […] I will take hold of them by the head and force them to look through that telescope.” (Brecht (1963), 38) But he makes the tragic error of presupposing the same standards of conviction and evidence on the part of his opponents. This is precisely not the case; ultimately they don’t even look through the telescope, since they don’t see how it could provide any serious matter for debate compared to, say, scholastic dispute. One of them finally puts a fine point on this by saying to Galileo: “One might be tempted to answer that your telescope, showing something which cannot be the case, might not be a very reliable telescope, eh?” (Brecht (1963), 46)

Knowledge and World-Pictures

95

guistic portions of the world-picture that are always unquestionably presupposed and used in procedures of empirical review). The role that a certain sentence plays in this system is clearly not just a matter of convention any more than the world-picture itself can be chosen arbitrarily at any time. The important presuppositions in our world-picture – presuppositions embedded in the functioning of our language and in the rehearsing and successful, habitual use of language games – reveal themselves at the level of explicit sentences in the form of grammatical knowledge: “When the child learns a language it learns at the same time what is to be investigated and what not.”28 3. The previous point is also closely related to the fact that our world-picture cannot be true or false. The distinction between ‘true’ and ‘false’, especially in the narrower sense related to propositions, cannot be taken as a question of an absolute external standard beyond all human epistemic conditions. The assumption of such a standard would not be immune to skeptical doubt, nor would it be a standard we could understand. Truth essentially has a lot more to do with successfully redeeming the truth-claims of assertions, to which we then connect the practical rules for the use of the predicates ‘is true’ and ‘is false’. And we have already seen that the world-picture is the foundation for our recognition that claims to validity have been redeemed and for the criteria of successful speech. Hence the distinction between ‘true’ and ‘false’ can only make any sense against the background of a world-picture (in the broad and deep sense); nor can we speak of ‘knowledge’ in the narrower sense without the world-picture in which (in the case of linguistic, scientific knowledge) it is bound to truth and justification. Yet we should note that this specific relation of the question of truth to a world-picture does not produce a relativism of arbitrariness in matters of truth; world-pictures are neither arbitrary nor freely chosen. That the meaning of ‘truth’ is tied to the redemption of validity claims through processes anchored in our scientifically and technologically in28 OC 472. Following Wittgenstein’s use of ‘grammar’, ‘grammatical’ here is meant in the broad sense beyond just logical syntax and explicit linguistic rules, referring to “ways and rules of using and understanding linguistic acts in general” (Abel (2004), 152). – The point Wittgenstein is making here is an example of how propositional knowing-that, e. g. that something verifiably behaves in a certain way, already presupposes a genuinely practical knowinghow in a certain sense, i. e. knowing how to use language in accordance with its rules.

96

Ulrich Dirks

formed 21st century world-picture is just as deeply sedimented in our understanding of the world as the processes of shamans (for a sufficiently comparable goal) are in the world-picture of a tribe in the Brazilian rain forest informed by animistic religion. 4. Not only can the plurality of world-pictures and their historical transformation be empirically confirmed, but the logical possibility of such changes in world-pictures can be explained. In the mid-20th century Wittgenstein himself had counted the sentence that neither he nor anyone could have spent a day on the moon as a world-picture proposition – and of course this sentence has since become capable of empirical review. By contrast the sentence that no human from the Earth has spent exactly one day of her life inside the sun or in the Andromeda galaxy is very strongly tied to our basic beliefs. It is so difficult to imagine how the contrary could be true that we would find it absurd to seriously treat this sentence as an empirical one. This is a reason to see it as a rotational proposition, but it is not a reason to rule out the logical possibility that the foundation of our world-picture might change, nor does it justify claiming for all times and all individuals that there could not be any world-picture in which this sentence didn’t function as a hinge. (It would be interesting to look more closely at how this sentence would figure in shamanistic world-pictures.) Rather the notion of ‘world-picture’ itself is conceptually related to the limits of how much change in our world-picture we can really imagine and to the actual difficulty of conceiving things differently here and now. Hence contrary thesis of the logical impossibility of any change in our world-picture can only shoulder its burden of proof with recourse to an external criterion that could show the given world-picture to be culturally and historically invariant – an endeavor that is not only almost absurd in light of the realities and the historical course of our globe, but also bound to fail upon critical and skeptical examination, since it is systematically impossible to establish any reference to an external standard of comparison outside of a world-picture. With the absence of any external, metaphysically constant standards for comparison or for exact translation and reduction, the transformative processes and the obvious plurality of world-pictures (and the bodies of knowledge indebted to these world-pictures) factually present have to be taken as they are phenomenally given, as genuine transformation and as irreducible plurality.29 29 The linguistic nature of world-picture propositions, along with the pictorial nature of large parts of world-pictures (to be discussed below), shows that world-

Knowledge and World-Pictures

97

5. A world-picture is my world-picture first and foremost, as we have seen most clearly in the examples of propositional beliefs that display their world-picture status in the first-person. Some such rotational sentences can even lose this status in the transition to the third-person perspective characteristic of scientific knowledge; my own identity, which for me could hardly be successfully justified by another proposition of greater certainty, can become the object of an empirical investigation for others – we might think, e. g., of genetic fingerprinting and biometric IDs. But even where the transition to the third-person doesn’t directly result in such explicit differences in status – where it primarily involves my understanding of the world – still it is first and foremost the foundation of my personal beliefs and my own ensuing actions. Yet this doesn’t mean that the world-picture is “to be seen in solipsistic or private terms”30, hence nor does it mean that the knowledge anchored in our world-picture never goes beyond merely subjectively binding belief. For not only do we share a public cultural lifeworld with others and grow into the inherited background as a community; if the depth of the linguistic portions of the world-picture consists precisely in how they provide the foundation for all of my speech and thought, as described, then the question of whether strictly and logically speaking a single individual could have a world-picture becomes the question of whether a single individual could have a private language that it is systematically impossible for others to understand. This brings us to the famous private language argument in Wittgenstein’s Philosophical Investigapictures occur in linguistic and non-linguistic signs. Yet the syntactical, semantic and pragmatic characteristics of world-picture signs are neither “magically and timelessly built-in” nor are they “subjective, arbitrary, relativistic”, but rather enter into the world-picture signs “through the practices of interpreting these signs in our habitual lifeworld” (Abel (2004), 145). Abel is able to use a broad and differentiated concept of signs and interpretation (cf. Abel (2004), introduction; Dirks (2010a)) in order to present interpretive practice as an explanation for the characteristic directness, ‘immediacy’ and unquestionableness or directly self-evident comprehensibility of world-picture signs. In addition he is able to derive the depth of the necessary possibility of temporal change and the plurality of world-pictures from the “sign-compositional and interpretive character of the world-pictures themselves” (Dirks (2010a),146). The perspectivity and finiteness of every interpretive process, the finite signs themselves and their finitizing functions towards and on the basis of this interpretive practice lead to a non-reductive, internal alterity and variability as well as an internal, conditional plurality for world-pictures as well; see Abel (1999), ch. 11. 30 Abel (2004), 124.

98

Ulrich Dirks

tions, which rules out a private language as logically impossible.31 For the sake of the nomological character of our use of signs if nothing else it must be possible for others to correct me; this shows us the public character of language and thus the public character of the world-picture that is always, in this sense, a shared one. The individuality that remains nonetheless (to be distinguished from ‘privacy’) is preceded by a public element. It is precisely this “indissoluble weft of […] first-person and third-person perspectives”32, which nonetheless accords with the individuality in every single process of understanding and accepting epistemic claims, that grounds the possibility of scientific knowledge with objective validity. (This should also be extended in an analogous way to the non-linguistic yet continually sign-bound portions of world-pictures.) 6. The aspects and beliefs that are fixed beyond question for me are, as Abel says, “components of the world-picture that we cannot, as things stand now, circumvent or distance ourselves from, that we cannot operatively produce ourselves and the validity of which we cannot suspend”33. Concerning the specific relation of scepticism and the worldpicture we can put the point more precisely, taking up Wittgenstein’s argument that doubt already presupposes a great deal: “If you tried to doubt everything you would not get as far as doubting anything.”34 Doubting as an understandable linguistic action clearly presupposes on the part of speaker and listener an established practice of asking questions and linguistically articulating doubt – a practice oriented around familiar and habitual instantiations that have been accepted and made to serve as principles rather than around any explicit theoretical criteria.35 Aside from the embeddedness of language games like ‘formulating assumptions’, ‘articulating doubt about assumptions’ or “a principle of speaking for and against”36 in our practices, each concrete doubt presupposes certain facts, as “If you are not certain of any fact, you cannot be certain of the meaning of your words either.”37 How could a sentence function as 31 32 33 34 35

See PI starting at 243. Abel (2004), 124. Abel (2004), 128. OC 115. Cf. OC 124. Even the gestures, facial expressions, etc. that correspond with linguistically expressed doubt can only function meaningfully by relying on habituated and intersubjectively shared practices of interpreting and using symbols. 36 OC 117. 37 OC 114.

Knowledge and World-Pictures

99

doubting whether a certain moon of Jupiter’s contains any water, for example, if we were able at the same time to raise doubts about the sentence’s semantic and logical presuppositions such as the existence of physical objects and the law-like behavior or “stability of things”38 and to suspend their normative role?39 Moreover, something cannot present itself as true doubt about an assumption if it would not make the slightest difference if the doubt were to ‘succeed’, if it would have no practical consequences and nothing would follow from it.40 Hence since a true doubt could not be one that “would not tie in with anything in my life”, “[t]he game of doubting itself presupposes certainty”41. Hence judging and doubting are only possible under the conditions given with our world-picture, the “deep-seated intuitions”, “self-evident certainties” and “bedrock views” as the undoubted background. For this reason it is also not possible for radical doubt to apply terminally to the entire world-picture it stems from: although every individual item in our store of propositional knowledge is fallible and the worldpicture cannot be joined epistemically to any ultimate metaphysical justification, on the other hand it is not possible to simultaneously put all of our beliefs and all facts up for question and doubt at once, “without thereby destroying the basis for justified doubt itself”.42 A world-picture manifests itself in currently fixed self-evident certainties that characteristically only change on significantly larger time-scales than are typical of the dynamic processes of quotidian and scientific knowledge. When a world-picture in the broad and deep sense does get shaken up, it is not because of methodically guided theoretical doubt and specifically not the direct result of scientific questioning. A world-picture cannot be false any more than it can be true. More radically it can lose its plausibility and perform its functions poorly and less productively; it can get old. In that case we no longer take it for granted that we 38 OC 473. 39 If we were to misunderstand the normative and rule-giving presuppositions situated in the world-picture as experiential sentences and try to doubt them, this would result practically in “chaos” (OC 613), i. e. the overturning of all regularity; in the extreme case “the foundation of all judging would be taken away from me” (OC 614) and hence the presumed possibility of doubt itself would be undone in the manner of a self-destructive circle. 40 Cf. OC 120. 41 OC 117, 115. 42 OC 128 f.

100

Ulrich Dirks

can trust in it in our speech and action. From the view of descriptive modeling this is the case, e. g., when the inconsistencies in our experience of nature become too strong; when the appeal to God loses its power to convince in rational discourse; when familiarity and experience with different world-pictures, for example through sharing languages and forms of life, or changes in our life-world lead to shifts in the world-picture – whereby the sciences and technology and their repercussions for our understanding of the world and our living conditions can play an important role in such processes. The examples of sentences mentioned at the beginning cannot be seriously subject to doubt either. Instead great novels or well-made science-fiction films can project larger contexts and courses of action in which such self-evident beliefs start to wobble for the characters. That these beliefs are located at the level of the world-picture is then seen in the fact that these changes can lead to an existential crisis for the protagonist, for example, or to the loss of mental stability and orientation in life. 7. Justificatory processes are not the only thing anchored beyond question in the background certainties of the world-picture; we also find “practical certainties” there, e. g. the “practical trust in what we say, do and think”.43 This brings out another role of world-pictures, which Abel calls the “stabilization of action”44. As Abel shows us, we can distinguish a series of functions by which world-pictures serve as stabilizers of action: (i) We form pictures of the ‘whole’ of the relevant circumstances on the basis of the world-picture, which contribute decisively to the functionality and orientation of our life practice; hence whether in a concrete situation I “find the conditions given in the situation that lead me to take or not take a certain action” depends on the activated world-picture.45 (ii) World-pictures play a part in establishing what counts as a permissible and justified action and what doesn’t insofar as the “horizons, attitudes and perspectives we have already drawn on” in making such distinctions are rooted in world-pictures. The same holds for moral, non-moral and prescriptive post factum evaluations of actions, which presuppose “standardizing and standardized norms and 43 Abel (2004), 164 f. 44 Cf. Abel (2004), 130 ff. 45 Abel (2004), 132.

Knowledge and World-Pictures

101

criteria” that “in turn are grounded in and genealogically descended from the underlying world-picture”. Every evaluation of an action “ultimately has its ‘final’ footing in a world-picture” insofar as a world-picture serves for action as the inherited, intersubjectively shared and culturally developed background “against which and towards which actions occur as they occur”.46 The history of astronomy, for example, and the vicissitudes of its diverse roles in the cultural contexts it was embedded in, gives us a wealth of prominent examples of how sciences are implicated in these normative issues: who studies and is allowed to study the events in the heavens, for what reasons and with what means, and how the actions and events made use of are to be evaluated, all varies with the integration of astronomy into the knowledge of how to rule in ancient civilizations, with practical lifeworld objectives of observing the heavens, with the astrological goals of generating a connection of fate between humanity and the cosmos, with the relation between reason and faith discussed so intensively in the Latin Middle Ages, and with dominant question today of whether the use financial and human resources in modern astrophysics is justified in the context of larger economic problems – all connections generated against the background of the world-picture. (iii) Moreover the example of Galileo already showed us how and to what extent world-pictures serve a function in regards to the form of actions, i. e. how they are carried out. Whether in verifying something one goes to the telescope or opens the Holy Scripture clearly marks a difference conditioned by the world-picture. Insofar as “the aspects that orient and standardize actions and that operate from and work towards a world-picture” are decisive for our rationality of action, hence for the “coherence within our web of beliefs, opinions, and expectations, or in short: our propositional attitudes”, Abel can arrive at the far-reaching thesis that “the principles of a world-picture that dominate in a particular epoch and culture also function in a broad sense as the principles of rationality in this epoch and culture”.47 In this sense the methodical acquisition of scientific knowledge in our age, which form a significant part of the lived practices of our cultural sphere, is not just an optional and replaceable form of action but rather is profoundly rational in a broad 46 Abel (2004), 135 f. 47 Abel (2004), 133; cf. 166 f.

102

Ulrich Dirks

sense, even though the idea of metaphysically absolute criteria of rationality has been lost; and in conjunction with the principles of this very world-picture it is the activity that provides the most well-established knowledge of our age. The 7th point brings out another aspect that is relevant for the broad sense of ‘world-picture’. A world-picture is not just described and expressed in language, and it consists of more than just propositional components, i. e. fundamental and extremely self-evident sentences “that embody my beliefs about the world and reality as a whole”48. The power of a world-picture to orient and stabilize action shows that it manifests itself not just in the network of our unquestioned basic beliefs, but also in the practices of our lifeworld, which are not purely linguistic, and in the related practical knowledge. Its non-linguistic components can be seen above all as the background of our societal and cultural conditions, in habituated customs and conventions, moral standards, etc. We rely on such elementary regulation of our actions by the world-picture whenever we act as we do based on an unquestioned trust in the relevant know-how. For Wittgenstein the fact that I simply get up from a chair and go shows that I do not have to investigate beforehand whether I still have two feet. “I simply don’t.”49 Logically the grounds come to an end in an “ungrounded way of acting”, without any further ‘why’50. The reflection on our speech, thought and action, on our language games and our life practices, can to some extent (re)construct and linguistically articulate the fundamental conceptions, attitudes, principles, practices, viewpoints, normative standards, intuitions and more or less rudimentary pictures that belong to the background of the world-picture or are situated very close to it. Even if in a particular case such an attempt might be seen as successful, nonetheless we should emphasize, with Wittgenstein, that we do not explicitly learn all of these components – and certainly not in the form of a textbook – but rather absorb them starting in childhood in growing into shared life practices and not least of all in learning language.

48 Abel (2004), 125. 49 OC 148. 50 OC 110.

Knowledge and World-Pictures

103

4. World-pictures and the Sciences After clarifying the important functional roles that world-pictures play in knowledge and differentiating the concept of world-picture we can now characterize selected aspects of the complex relation between the sciences and world-pictures, focusing on the modern empirical sciences and, among these sciences, on the epistemic aspects relating to scientific knowledge. 1. Let us begin by looking at the narrow sense of ‘world-picture’ situated within the wider notion: an illuminating picture or orientational model, or a unified view of the world captured in a guiding model, that integrates all the knowledge of the world and the entirety of the life experiences of the people of a certain epoch or culture. It is clear that sciences and technology quite openly shape and determine world-pictures in this narrow sense. The models provided by astrophysics and cosmology can serve us here as outstanding examples, since their knowledge of reality directly concerns our understanding of the spatio-temporal material world as a whole and the localization of the human life-form within this world. It is precisely these pictorial illustrations of the cosmic orders and processes that have shaped and continue to shape our picture of the surrounding reality, i. e. our cosmological world-picture, beyond these academic disciplines. A look at Western history shows that a unified world-view forms depending on what model prevails, whether a teleological and Aristotelian model, a mechanistic and causal conception of nature or the basic conviction that Jürgen Mittelstraß calls an ‘Einstein world’.51 Moreover today we often hear the diagnosis that the guiding 51 Mittelstraß (1997), 240. Mittelstraß shows not only how “science-supported technological cultures” “constitute” (228 f.) their reality, i. e. the modern world, in diverse ways, but also how science “provides” a picture of “the world in which rational cultures live”; science has a “power to generate world-pictures” (231). Besides the “diachronous relativity” (245) of world-pictures given along with historical conditionality of these processes that generate them in which “the world (of science) and its development appears within science” (241), they are also characterized by their “synchronous relativity” (245). World-pictures in the narrower sense are not identical to scientific theories, but rather follow the “deep-reaching changes in scientific theorizing” as “deepreaching reinterpretations of the world in natural philosophy” (245). In contrast to world-pictures in the broad sense (Wittgenstein, Abel), which are a precondition of questions of validity, Mittelstraß can talk of the “validity” of sciencegenerated world-pictures insofar as they are bound to rational theorizing (diachronously) as a condition. Yet scientific theories do not unequivocally and de-

104

Ulrich Dirks

model of the natural sciences has moved in the direction of complexity and self-organization, which brings with it a different conception of nature precisely when it comes to matters of calculability and prognosis.52 Sciences and world-pictures in this sense thus condition each other reciprocally. On the one hand science and technology has the power to generate world-pictures in the narrow sense. On the other hand every research activity aimed at acquiring scientific knowledge, such as scientific description, the mathematical modeling this might involve, initiating a specific series of observations, and designing a concrete experiment or computer-guided simulation, are initially undertaken with reference to a scientific world-picture. 2. Additionally it has become clear that a world-picture in the broad sense of the inherited background of our understanding of ourselves, others and the world and the foundation of all our speech, thought and action cannot itself be a scientific world-picture in the strict, narrow sense in which they can be generated by the sciences. More particularly, world-pictures in the broader sense do not consist of scientific, propositional knowledge, nor are they illustrative world-views composed of such knowledge and claiming scientific validity. Instead the sciences ultimately presuppose world-pictures, and their success and progress rely on the world-picture, as shown by the fact that accepted scientific procedures of verification are anchored in the normative dimension of the world-picture. Thus concerning the broad and fundamental sense of finitively determine their philosophical interpretations, which means that in terms of synchronous relativity “different philosophical interpretations of the same theory” and hence different world-pictures are possible (245), which fits the basic finding of the indeterminacy of translations and interpretations (cf. Abel (1995)). 52 Cf. Poser (2008). Poser also shows how and to what extent processes of scientific modeling, particularly in their characteristic reduction of complexity, are guided by concepts of order, in which world-pictures in the narrow sense serve as guiding models. Prospective guiding models or world-picture models take on an orientational value concerning the relation between humanity and the cosmos and also concerning the alignment of knowledge-seeking scientific activities. It should be noted that when this conjunction of individual items of knowledge and the thusly generated unified new world-view succeeds, it rests not on the model’s representation ‘of reality’ but rather on the order impressed upon the realm of objects by the relatively a priori epistemically relevant presuppositions of the guiding model. (Cf. Dirks/Knobloch (2008b), 22 f.). The function of world-pictures to serve as guiding models comes out very clearly in the cosmological and astrophysical conceptions of the cosmos past and present.

Knowledge and World-Pictures

105

‘world-picture’ Abel is right to say that “the formulation scientific worldpicture is somewhat like wooden iron”53. In regards to the systematic role of the world-picture in the broad sense for practices and praxes of generating knowledge in the modern empirical sciences and for the logic and dynamic of justificatory processes of scientific knowledge, we can now note the following points, building on Wittgenstein, and adding to those aspects already discussed: (i) Every scientific act presupposes a universe of pre-theoretical presuppositions that are not themselves genuinely scientific in nature. As we have seen, a retroactive semantic and logical analysis shows them to be fixed beyond doubt and prior to scientific procedure. Insofar as such presuppositions are located in the background of the world-picture, they are situated much deeper than that propositionally expressible implicit knowledge that is involved in explicit scientific knowledge but does not need to be made explicit in the given circumstances and that, as scientific knowledge, can be subject to doubt and claims to justification concerning its validity or cogency. Doing a calculation on paper, for example, shows our propositional belief – usually unspoken, maybe even never consciously entertained – that “the figures on the paper aren’t switching of their own accord”, along with the attitude or practice of continuously and unquestioningly trusting our memory, e. g., which creates practical certainty and is presupposed by many forms of know-how (such as the methodical and experimental know-how of the scientist). “I have plenty of doubts”, even in scientific operations, “but not that”54. And every experiment shows us that in entering the laboratory and engaging in experimental practice I have presupposed, semantically and logically (and with a stabilizing effect on my action) that there is no “sense-deception” at work, that “the apparatus I believe I see is really there”55. In addition our unquestioned background assumptions entail that things (including the experimental apparatus) do not typically disperse whenever I’m not looking. In short: “One cannot make experiments if there are not some things that one does not doubt”.56 53 54 55 56

Abel (2004), 121. OC 337. OC 163. OC 337. As Wittgenstein emphasizes concerning our actual scientific practice, in this sense it “belongs to the logic of our scientific investigations that certain

106

Ulrich Dirks

(ii) Moreover the propositional share of the linguistic scaffolding of the world-picture and the sentences describing the world-picture cannot be scientific propositions, not because so far no comparison with the world or reality has succeeded in verifying or refuting them, showing them to be true or false or appropriate or inappropriate, but rather because with unlike scientific propositions “[h]ere we see that the idea of ‘agreement with reality’ does not have any clear application.”57. The possibility of empirical scientific knowledge depends genealogically on relatively a priori functions in which the background of our world-picture determines our specific relation to world and reality, which then manifests itself in our scientific procedures. (These functions also generate the order of reality; see below in chapter 5). Even if in the empirical sciences experience is ultimately the ground of our judging the way we do, “still we do not have a ground for seeing this in turn as a ground”, i. e. we cannot ground this in turn, say with a scientific or meta-scientific explanation.58 Instead we presuppose and rely on the background from which our judgment practices, our way of researching and looking at things59, emerge in a quite basic sense. The world-picture is the “matter-of-course foundation”60 for our everyday practical explorations as well as our scientific research, at least concerning its position within our “life”61. 3. Yet this specification of the difference between scientific knowledge and the world-picture does not mean science and technology has no impact on our world-picture in the broad sense. The thesis is rather that they have a significant formative and possibly revisionary influence, yet indirectly, and in a way that cannot be posited or codified. For example, if a scientist working outside the mainstream discovers that the

57 58 59 60 61

things are in deed not doubted” (OC 342, Wittgenstein’s emphasis), which points far beyond the premises, axioms, foundational concepts and other presuppositions within science (and also beyond the relations of formal logic). Ultimately we would not be able to make any doubt about the sort of presuppositions mentioned here understandable, since e. g. we would no longer really know “what such a person would still allow to be counted as evidence and what not” (OC 231) in everyday life and in the empirical sciences. OC 215. OC 130; see also OC 131, 145. Cf. OC 211. OC 167. OC 344.

Knowledge and World-Pictures

107

sun is a material physical body and not “a hole in the vault of heaven”, nor a chariot of the gods etc., this belief could send shock waves throughout the system of the commonly prevalent beliefs, turbulences in “the element in which arguments have their life”62, and in consequence clearly might change more than just the world-view in the narrower sense. If today we met people from a culture with entirely different ideas about the sun than us, we would notice very quickly that this difference involved more than just a few other specific beliefs about how a certain part of the world is constituted. Rather the foreseeable difficulties in trying to suggest a review procedure they would accept as such would show differences in the deep-seated world-picture. Firstly, the standards of rationality by which something is accepted as a convincing explanation would differ, as would, secondly, the undoubted certainties that people have recourse to in their explanation – assuming, thirdly, that something like the language-game ‘giving explanations’ is even practiced in the other’s language. Fourthly and complementary to this, such differences in world-picture make it clear how in the development of our own Western cultural traditions science and technology have contributed to the genesis of our own world-picture and its foundational functions. Hence the indirect influence of the sciences even at the pre-scientific levels of the world-picture – e. g. in replacing a world-pictureA, which functions as the background for a mythical worldview in the narrow sense, by a world-pictureB, within which a scientific picture of the world in the narrow sense is situated – can be seen in the fact that the success of the sciences induced a new scheme of basic concepts, a new way or organizing the innerworldly reality that confronts us. If before we could classify the sentence “there are gods of the sun and the other planets” as a world-picture proposition, now “there are physical bodies” has world-picture status.63 The explanatory and prognostic achievements of science that impact our life-world, the simplicity and symmetry of complexly interconnected scientific explanations, and other aspects could very well lead to certain principles becoming sedimented in the river-bed of our sentences – 62 OC 104 f. 63 This proposed sketch of how to reconstruct the transformation is clearly compatible with the common and quite justified criticism of the thesis that the transition ‘from mythos to logos’ should be seen as a movement from a lacking or underdeveloped rationality to rationality ‘proper’. – Moreover we should note that the dynamic changes in the concept of ‘science’ would have to be accounted for in any historically adequate specification of this simplified example.

108

Ulrich Dirks

to use Wittgenstein’s imagery – and being withdrawn from empirical review to serve a regulatory function for other sentences. Some at least arguable examples are, e. g., sentences according to which nothing arises from nothing, or that nothing arises without ground or cause, or that everything in nature happens in accordance with a law. Moreover, the scientific method of empirical examination could then ultimately come to anchor itself in the background as the self-evident and publicly shared way of verifying propositions about reality. Conversely, sentences that were previously world-picture propositions belonging to the epistemically relevant shared system of reference could over time become “fluid” again. In ancient Greece sentences about the orbits of heavenly bodies had the status of descriptions of the world-picture due to the metaphysical dimension they invoked. Explanations of nature were to bind the phenomena to these sentences or to the principles they involved (to ‘save’ the principles) rather than dropping the conception of the divinity of the circle and the cosmic order. Then in the course of the development of the sciences in the modern era descriptions of astronomical orbits and other cosmic relations became knowledge sentences that were largely empirically reviewable and falsifiable. 4. The change that science and technology induced in world-pictures even in the broad and fundamental sense can be more precisely reconstructed by asking whether (i) the sciences have exercised this modifying influence from within themselves and towards the background, operating on the basis of the relevant parts of this background and standing in a relation to the other parts that had been so far free of any urgent conflicts; or whether (ii) the situation was primarily a clash of different world-pictures as a whole. There are of course a great diversity of mixed forms between these two extremes. Forms of practical change processes (and their logical possibility) have already been briefly described for case (i). Here we can emphasize that (a) a ‘single’ concrete item of scientific knowledge can in fact have repercussions for the world-picture far beyond the sciences; S. Freud very fittingly spoke of the astronomical and astrophysical findings of modernity as a “wound to humanity’s pride”. The scientific finding that the earth is just one planet among others orbiting a central star, itself just a quite average star on the border of an enormous but perfectly ordinary cluster of stars comprising a galaxy, has cognitive value beyond the sciences themselves. It primarily affects the understanding of self and world for people who no longer see themselves at the geometric and axiological center of the whole; the attested change in the attune-

Knowledge and World-Pictures

109

ment of their being-in-the-world is the expression of a change in world-picture in its comprehensive, pre-scientific sense.64 Possible future discoveries of extraterrestrial life and the associated shifts in the fundamental concept of ‘life’ might have a similar scope, whereas most ‘normal’ scientific discoveries are not like this at all. (b) Science and technology lead to sustained changes in the circumstances of a culture’s lived environment. Hence the impact of the sciences on our lifeworld produces shifts above all in the system of action-orientation; we only need to think of the recent triumph of digital information and communication technologies. Here ‘lifeworld’ is meant primarily in the narrow sense of “the entirety of the practical environment shaped by humans in their everyday living circumstances and pretheoretical experiences”; this practical formation, or the life-world in the narrow sense, is preceded, shaped and standardized by the world-picture – and hence the sciences as well insofar as they are “conducted within a public life-world”65. Yet the intimate and diverse connections between world-pictures and lifeworlds only truly comes out when, following Abel, we conceive ‘lifeworld’ in a broad sense as the world directly experienced by the subject, i. e. as the world that comprises the horizon of every one of our activities.66 The life-world is given to us as an “initially still unindividuated, 64 It changes nothing about this basic point if a closer examination would have to paint a much more complicated picture of this transition – or at least of the phase of this astronomical epistemic process in early modernity and the Copernican and Keplerian transition from the geocentric to the heliocentric model – than the simple devaluation of humanity that most later interpretations saw in it. After all, the previously assumed position of humanity on an earth in the middle of the universe is also geometrically the furthest removed from the divine heavenly spheres and is the most dominated by the material reality seen pejoratively in the neo-Platonic Christian view. (In Dante’s Divine Comedy the mid-point of the cosmos is actually in Hell.) Yet aside from this axiological tension, through the course of the development of modern notions of the cosmos and the knowledge of modern astrophysics the meaningful interconnection between the cosmological facts of a universe seen as God’s creation and the purportedly singular position of humanity as such was gradually lost. Hence our thesis about the science-induced transformation of the world-picture is located outside of the controversy (cf. Danielson (2001)) as to whether the establishment of Copernicus’ heliocentric model is exclusively to be seen as the anti-anthropocentric disenthroning of humanity or if it can be interpreted in some regards as elevating humanity. 65 Abel (2004), 120, 118. 66 In describing the lifeworld as the ‘world in how lived experience is given’ Abel explicitly refers back to Edmund Husserl; cf. Abel (2011), 352.

110

Ulrich Dirks

still undivided, continuous meshwork of relations of life, the world, action, the self and others” and thus is essentially connected by means of our living out our lives within these relations to the form, the network and the background of those cognitive attitudes, experiences and beliefs “that allow humans to understand their world, orient themselves in it and put into practice their practical intentions and their ethical, cognitive and educational activities”67. Accordingly the publicly and culturally determined lifeworld is shaped by the world-picture, which exercises an orientational effect on the practices of the life-world and in this way manifests itself in them. Hence the influence and power of world-pictures, even concerning scientific activity, can be seen as “the power of the lifeworlds” that are the home of the world-pictures influencing them; the lifeworld also serves as the pragmatic final foundational and justificatory authority (though not as a metaphysically absolute final authority)68. If science and technology in turn influence the life-world, they can thus in this mediated way lead to changes and revisions in the world-picture that they already rely on. In case (ii), changes in an entirely different and foreign world-picture – due to generally non-shared standards of persuasion, argumentation, justification, etc. – cannot be induced just by the attempt to identify the ‘falsehood’ or inadequacy by our standards of the other culture’s beliefs and, for example, to put our own scientific knowledge in their place (if we have already conceded the inconceivability of external standards and external and absolute points of reference). The logical possibility of confrontation and change in world-pictures that were previously entirely foreign to each other rests on the fact that when they begin to share forms of life, to refer to common human conditions and way of acting, to share environmental conditions and to construct linguistic relations of communication and understanding, this leads to a (pragmatically sufficiently large) basis of new internal standards and regularities for world-pictures to be able to overlap at all and act upon each other reciprocally. Only at this point could it happen that the sciences exert an indirect influence. For example, given the fact (assuming this for the moment) that the members of the proverbial native tribe also love their children, changes in the world-picture might be set in motion (in an unpredictable manner) if, e. g., certain actions are taken 67 Abel (2011), 363. 68 Abel (2011), 360.

Knowledge and World-Pictures

111

in the context of scientifically informed medicine to lower infant mortality in the village. 5. If we are now to explain the dependence of the sciences on the world-picture in the broad sense – above and beyond the embeddedness of the ‘normal operations’ of the sciences emphasized by the philosophy of science in the prevailing scientific world-picture in the narrow sense – then we should note that world-pictures, as we have seen, serve their function of stabilizing action and guaranteeing certainty first and foremost in the area of everyday opinions, beliefs, attitudes and life-world practices. In terms of the Kantian modi of holding a thing to be true – namely, opining, believing and knowing – and the function of creating certainty, the most direct effect and manifestation of my background world-picture is primarily prior to conceptual knowledge and is to be found in epistemic belief, which operates within convictions and propositional attitudes and is binding in the first-person. But a series of aspects should help to make it clear how the world-picture is also “the substratum of all my enquiring and asserting”69 for the sciences as well, embedded in the life-world they grow from, and for scientific knowledge, which emerges from the broader sphere of opinion and belief and also stems genealogically from the life-world, though less obviously, through its connection to methodical procedures and intersubjective justification70 : (i) First we have to note that science, which stems from our life-practices and the culturally influenced life-world, pursues epistemic interests and goals the relevance of which cannot be understood independently of current aspects of the world-picture. This concerns not just the research funding policy of the moment, but more fundamentally the question of whether any wide-ranging epistemic interest in reality and a relation of dominating nature and turning it to our ends ever develop at all. Epistemic interest can take on a whole different significance and value within world-pictures where religious faith has clear priority over the activity of reason. Moreover, if human life is seen within these world-pictures as a trial or a punishment that is to be endured rather than alleviated, the goal of making nature productive for us in order to improve our own living conditions can also appear in a different light. – The goals, pur69 OC 162. 70 See a thorough treatment of this in Abel (2004), 330 ff., 302 f. and 137 f.

112

Ulrich Dirks

poses and ideas of sciences are included in a deep level of the world-picture. (ii) That certain procedures have been accepted as suitable methods for redeeming claims to validity in the sciences, such as procedures of observation, measurement, and experimentation, depends, as we have seen, on the world-picture71 – even if only in a general sense and not concerning specific technical or scientific methods and procedures. Nonetheless the concepts of scientific rationality72 and objectivity are linked to this. The concept of objectivity in the modern natural sciences essentially depends on the principle possibility of repetition and independent reviewability of the results of measurement, regardless of the specific measuring procedures and concrete methods of observation; thus it fundamentally depends on the

71 This includes even the methods of quantification and translation into mathematics so characteristic of the natural sciences of modernity. In particular the 20th century discussions about the possibility of creating a foundation for mathematics have shown that questions about the necessity, certainty and a prioricity of mathematics along with questions about what can be accepted as a valid proof in mathematics, what the objects of mathematical study are, etc. can only be discussed with recourse to human epistemic conditions, which ultimately means taking world-picture factors in the broad sense into account. On this point see Poser (1988), Dirks (2002). 72 The explicit standards of rationality that are characteristic of the scientific approach (modeling, mathematical description, interpretation of measured data, formulating laws and theories, etc.) such as demands for consistency, coherence, unambiguousness, completeness, etc. are preceded by the precondition of implicit regularities and standards of rationality in a broad sense; cf. Abel (1999), ch. 4, esp. 96 – 100. Insofar as the broad sense of rationality refers “to the coherent fit of our opinions, beliefs and knowledge, of all we hold true, with the relevant relations of communication and action, with the web of our empirical experiences and the network of our propositional attitudes”, the certainties manifest in our practices and abilities or more generally the internal principles of a world-picture at a certain time also function as principles of rationality in the broad sense, as we have seen (see above ch. 3, point 7). The form of life and the background world-picture manifested within it play a role for rationality in the broad sense (and thus for rationality in the narrow sense in reference to which scientific knowledge can be considered knowledge) “as the last determining ground of what counts as rational that we can achieve” (Abel (2004), 166 f.).

Knowledge and World-Pictures

113

measurement procedures and standards that are intersubjectively shared in scientific practice and community.73 (iii) In contrast to the body of our everyday beliefs, which is by no means clearly arranged and seamlessly composed without contradiction, science is transparently organized logically and methodologically. Hans Poser uses a heuristic, multi-step model to describe empirical sciences and the dynamics of scientific knowledge that is helpful for our view of the problem as well.74 On this model, science is characterized by a series of methodological determinations and assumptions. The first step is the concrete ontological determinations concerning the basic objects of that particular scientific discipline and the determinations of what sources of knowledge are admissible. These determinations of sources mean not just ‘experience’ in general, for example, but more concretely the admissablity of certain instruments and aids, or in the humanities characteristic sources, etc. Moreover there are additional permissions and prohibitions for the further theoretical work on the material, e. g. concerning how to deal with ‘actual infinity’ in mathematics. The sources of knowledge are also arranged into a hierarchy. In addition the first step includes determinations that Poser describes as ‘judicial determinations’, borrowing the term from Kurt Hübner; these regulate what precisely counts as justification, falsification, confirmation, etc. and when one theory is to be seen as more successful than another. Moreover further normative determinations are needed in practice, e. g. concerning what form theories take in a certain discipline. We should also take note of Poser’s finding that all these determinations and assumptions at the object level, as first-order rules, can be changed, justified or rejected. The guidelines for the latter are no longer located at the same level – as second-order determinations they have the status of meta-rules. These meta-components in the sciences can be more specifically identified than in the case of our everyday, life-world-related knowledge, and can be characterized as determinations that ultimately can only be made convincingly with recourse to a broad world-picture. In the debate between Leibniz and Newton on space and time, 73 On this broad notion of objectivity, which relates to the grammatical form of judgments of interpretation and its categorizing function and ultimately to life practices and the form of life, see Abel (1999), 277 f. and Abel (2004), 168 f. 74 Cf. Poser (2001), 173 – 207.

114

Ulrich Dirks

e. g., the disputants often appealed to metaphysical principles and arguments that were at home in a contemporary metaphysically and religiously charged world-picture or that drew their persuasiveness from such a world-picture. We also see pre-scientific and extra-scientific worldviews quickly come into play in scientific discussions at this second level when it comes to the principles of our modern cosmology, which takes itself to be an empirical science. A discourse about (a) the permissibility or impermissibility of the cosmological principle or postulate that the distribution of matter and energy within the observable range of the universe (within certain scale lengths) is not a special case, and the discussion about (b) the permissibility and justification of the assumption that the laws of nature that have successfully explained so much in the regions of the cosmos accessible to us hold throughout the entire universe – all this takes us not just to to astrophysical cosmology, i. e. to the object level, but primarily to pre-scientific and meta-scientific considerations. The question is then always whether or not the arguments offered in this context75 can make use of components of our world-picture that are not purely scientific. This is even clearer when we introduce such principles into the physical description of the universe like the ‘strong anthropic principle’ that every model of the universe has to account for it being the goal of the universe, however we wish to understand this goal, to produce us intelligent creatures. Such a principle clearly cannot be decided on physical grounds; attempts to argue for the principle or to offer good reasons to reject it can only be had with recourse to the background assumptions that are not currently subject to review.76 75 For example, here some people hold the physical describability and knowability of the universe as a kind of regulative idea and note that cosmological principles are the condition of possibility for physical cosmologies. Other arguments are more inductive or pragmatic in nature in emphasizing that so far, without exception, every contingent expansion of the observable region of the universe from every improvement in our instruments has shown ex post there has been no break in the known natural laws or cosmological principles and that it is therefore inadvisable to depart from them without reason. 76 Poser’s analysis also allows us to reformulate the idea that the sciences, despite all their relativity, have nothing to do with a relativism of “the arbitrariness of scientific propositions” or with a “far-reaching arbitrariness and historical contingency” (Poser (2001), 195 f.). Just as Mittelstraß argues against the thesis of radical incommensurability in the transition between paradigmatic theories,

Knowledge and World-Pictures

115

5. World-picture – Knowledge – World World-pictures in this comprehensive sense (Wittgenstein, Abel) have proven to be crucial for the understanding of knowledge, particularly as a constitutive condition of scientific knowledge and hence as an important object of investigation for the philosophy of knowledge. We can now deepen our analysis by clarifying how we are to conceive the three-pole relation between knowledge, world and world-picture; and this will cast suspicion on a series of typical conceptions. Firstly the reifying nature of language can make it seem that our use of the term ‘world-picture’ must refer to particular entities with an objectlike existence that specifically affect our speech, thought and action. Secondly one might think that world-pictures are purely linguistic and propositional. Thirdly there is the thesis that on the contrary, such sentences are to be seen as mere descriptions and that the world-picture itself is essentially a visual image of the world, a pictorial presentation essentially like a drawing or photograph. On this thesis the world-picture is a more or less accurate copy of a world independent of all epistemic conditions. These theses already start to get mired in difficulties if we look at the form of our own thoughts about the world-picture. No such investigation can start from a neutral position unconnected to any world-picture, a ‘view from nowhere’; hence we cannot ever arrive at a complete deand thus against the resulting historicism in the theory of science, by pointing to ‘transparadigmatic criteria of rationality’ (Mittelstraß (1997), 242 – 254), Poser’s explanation of the dependence of the epistemic limits of science on the methodological determinations develops a potential argument against relativist conclusions from Thomas Kuhn’s theses concerning paradigm shifts in the sciences and the scientific “revolutions as changes of world view” (Kuhn (1970), ch. X). We do not need to fear an “irruption of irrationality” in phases of paradigm shift, nor is the paradigm, Kuhn’s ‘disciplinary matrix’, the product of mere convention (Poser (2001), 199, 196). The second-order meta-rules shown most clearly in the argumentation during the upheaval of paradigmatic theories and the associated scientific world-pictures are themselves grounded in what Poser calls “connective world-view”, which refers to that “realm of basic conceptions of the relation between humanity and the world”, including the horizon of “norms and notions of being”, that “bears the self-understanding of a time” (202, 205). Here we see again how the extra-scientific level of the world-picture in a broad and comprehensive sense, which as we have emphasized is not just a matter of choice, serves as a presupposition for science itself, even if it is partly influenced by sciences and technology.

116

Ulrich Dirks

scription, never mind a theory that would capture the influence of a world-picture on our speech, thought and action. An external perspective is not available to us in our human epistemic perspectivity, as we are bound to language and to signs in every act of speech and thought in the course of our investigation and are embedded in life practices that cannot be suspended.77 I cannot reflect myself out of the entirety of my individual conditionality in order to see from there what background web governs my speech, cognition and action. If under a critical epistemology we can dismiss the thesis that it is possible to take on a God’s Eye point of view, then in logical terms we have to keep in mind, in reflecting philosophically on our world-picture, that we do so in an internal perspective78, where we try to explore presuppositional conditions ‘from within’, so to speak, always tackling a certain stretch of them under certain epistemic interests. For this reason a world-picture in the broad sense cannot be conceived as a separate self-contained entity that could occur within an identity relation and could be quantified over. What serves for each of us to guarantee certainty and stabilize action is shown directly to some extent, and to some extent is shown through a reflective examination of our actual acts of understanding and in how we actually live our lives, in our acts of speech and thought and in knowledge processes themselves. World-pictures are not hidden deep down, not “‘behind’, ‘alongside’ ‘below’ or ‘before’ our life-practices” but rather on the “surface” of our being-in-the-world, situated “solidly in the practices of our life-world and culture”79. It is not that something is first part of a world-picture in and of itself, say, due to inherent characteristics and properties, and then performs a certain function for those people who ‘have’ this world-picture. Rather a part of the world-picture can be reconstructed semantically and logically within the dynamic web of actual phenomenal processes themselves (following the question of what we have to suppose has been fulfilled in order for our life to be lived the way it is, for us to rely on what we rely on and to have the life-world orientation that we have) – just as, to use Wittgen77 Hence even in the third-person perspective all of our investigations of the world-picture of other cultures and epochs are projected against the background of our own standards and conditionalities and do not lead us to something like a space of investigation ‘free of world-pictures’. 78 I use the term ‘internal perspective’ in the sense of Hilary Putnam’s distinction between an externalist perspective (esp. that of a metaphysical realism) and an internal or internalist perspective (cf. Putnam (1981)). 79 Abel (2004), 130 f.

Knowledge and World-Pictures

117

stein’s image, the movement of a body distinguishes something as a rotational axis. (Though it can be physically instantiated, a rotational axis is not primarily a physical entity, but rather a theoretical construct by means of which we can get a superior understanding of movement). Hence on the one hand the background world-picture manifests itself in our practices, convictions, language games, etc.; on the other hand the semantic and logical re-construction itself delivers certain networks of sentences, practices, etc. that the world-picture could be said to consist in. Hence on the Wittgenstein-Abel line this means for our concept of world-picture: (i) World-pictures cannot be hypostasized, cannot be posited and made transparent as a whole, are not sharply defined and cannot be generated operationally. Strictly speaking there can be no such thing as a ‘theory of world-pictures’ or even a formalization of them. (ii) World-pictures are not to be thought of as object-like entities that could be located anywhere, whether in the head or in the cultural forum, and certainly not as anything that has a causal effect on our speech, thought and action.80 Just as the rotational axis is often only identified retroactively in the mathematical modeling of a movement, rotational sentences functioning at the level of the world-picture might never have been thought, in the empirical sense81, before entering into our thought and action. In this case our speech and action, such as getting up from a chair, show us what regularities and logical and semantic preconditions govern them, e. g. the assumption that I still have two feet or basic and general assumptions about the continued existence of bodies and their reidentifiability. When we have made hinge sentences and other components of the world-picture at least somewhat explicit, this gives us an important (if not sufficient) key to better understanding people’s actions, be-

80 In this sense Abel can say of our everyday living that “my world-picture is primarily present in the situations that are most important to me without my coming across it in an object-like manner”. The world-picture is present in a matter-of-course way in which “the world-picture is neither presented nor represented”; it “resists objectification in a unique and elusive manner”, so that it also cannot be an object within semantic logic (Abel (2004), 134). 81 Cf. OC 159.

118

Ulrich Dirks

liefs and decisions and, more than that, to better understanding the scientific rationality of our knowledge processes in today’s sciences. (iii) We should note, as a corollary to these points, that world-pictures cannot be freely chosen, nor can they be arbitrarily substituted for one another. They came down to us; we have grown into them. On the other hand genuine changes in world-pictures over time are not determined by a priori laws either; rather each change in world-picture opens a concrete field of research exploring the complicated processes of overlapping, replacement and development in that transition, for example as the indirect consequence of new scientific knowledge.82 Yet such research remains subject to the critical limitation that there is no external, absolute standard for the description of these processes, whether they are seen as transformation, progress, downfall, suppression, or in any other terms. (iv) The propositional parts of a world-picture are neither absolutely a priori in a metaphysical sense nor are they just a form of empirical sentence situated rather far from perceptual experience. The logical and grammatical functions these world-picture propositions can serve mean that the dominant world-picture of a time gives them a relatively a priori role in our current empirical knowledge. We should also emphasize that there is no sharp criterial dichotomy between empirical propositions and world-picture propositions (e. g. their ability to provide a “norm of description”83) but only a necessarily vague transition. There is no sharp boundary “between propositions of logic and empirical propositions”, since: “The lack of sharpness is that of the boundary between rule and empirical proposition”.84 (v) In addition we should note that a world-picture does not need to be non-contradictory, logically organized and seamless, any more than the life practices it orients do, or the entirety of our intuitions and the network of beliefs or the action and speech-acts that the world-picture stabilizes. “A world-picture by no means need be 82 Günter Dux has attempted to investigate this from a historical and sociological perspective under the title Die Logik der Weltbilder. Sinnstrukturen im Wandel der Geschichte (“The logic of world-pictures. Meaning structures in historical change”), cf. esp. Dux (1982), 290 – 308; the theses he develops there merit a separate discussion. 83 OC 321. 84 OC 319.

Knowledge and World-Pictures

119

homogenous and unified in structure or even a consistent system.” This concerns more than just the question of what it shows about the background worldview if someone gets into an ultra-modern airplane but doesn’t want to sit in seat number 13. The “openness and polyvalence of a world-picture” also relates to the different spheres of our culture, which besides the sciences, which already differ greatly among each other, also include technologies, societal and legal systems, and our moral, religious and psychological beliefs. Accordingly Abel can formulate his thesis even more pointedly and argue that “it is precisely when the demand for consistency in a world-picture exceeds a certain level that the world-picture loses its power to orient us in the very diverse practices of our life-world”.85 Concerning how world-pictures are constituted and given, it has already become clear that they manifest themselves in our conduct, languagegames, in cultural and life-world practices, in processes of speech, action, and belief-formation, etc., and consist of a variety of components. In reconstructing this background we find complicated networks of propositional elements, such as the assumptions, beliefs and fundamental matter-of-course certainties that cannot be sensibly doubted or suspended, but also of non-propositional components, to some extent components that cannot be expressed propositionally or even linguistically, such as basic and deep-seated viewpoints, sedimented pictures, religious and mythical attitudes that cannot be suspended at any time, and particularly strongly habituated customs and rituals that are beyond question for us.86 In light of this finding – in order to advance our understanding of the propositional knowledge relating to the world that is preformed by the world-picture and anchored in it – in our final steps we will need to briefly describe (a) the fundamental significance of the pictorial character of the world-picture, including the relation between world-pic85 Abel (2004), 131. 86 Cf. Abel (2004), 120, 128 – 131. Insofar as the forms (ways) these human “practices, customs, modes of action, conventions, institutions, traditions, forms of social interaction, rules of conduct, ceremonies, rites” etc. comprise the terminological core of ‘form of life’, and insofar as successful speech and understanding of a language are embedded in this form of life, which constitutes the fact that “we live, speak, think and act just the way we do”, Abel can say of the connection between world-picture and the form of life: “Every form of life has its world-picture” (Abel (2004), 152 ff.).

120

Ulrich Dirks

ture propositions and world-picture, and (b) the reciprocal relation between world and world-picture. Concerning (a), the central role of the pictorial characteristics and functions is most immediately conspicuous with world-pictures in the narrow sense, e. g. scientific world-pictures. Embedded in the background assumptions of the broad world-picture, these world-pictures coalesce the knowledge of the world and the life experiences of a culture and epoch within an illuminating and illustrative picture and a guiding orientational model, a unified world-view – particularly, as mentioned, through the use of ordering processes meant to bring vivid clarity. The most prominent cases involve the modeling87 of a world-picture as a visual image of the world or the cosmos.88 But this 87 Here the talk of ‘modeling’ emphasizes the active, productive and creatively constructive character of this process – which not least of all creates synthetic relations of order among the given ‘appearances’ and our beliefs about them – in contrast to a merely passive copying, representational process. In this way it accounts for Kant’s critique (in the ‘Transcendental Dialectic’ of the Critique of Pure Reason) of the concept of ‘world’ aiming at totality (a whole existing in itself) and the treatment of the world as an object of experience given in sensible intuition, as well as the ‘imagelessness’, strictly speaking, of the regulative idea of the world (of all phenomena) as a whole. 88 The Atlas der Weltbilder examines practices of visually generating world-pictures as “pictures of the world”, “different visual media in the service of constructing the world as a picture” as well as “the history of changing methods of presentation and different vehicles” using important examples (Markschies et al. (2011); quotations: XIII f.). For world-pictures, which are to be seen less as visual attempts to describe the world as a whole aiming at an ideal of similarity and more as actively ordering models of lifeworld experiences and beliefs concerning the world and the position of humanity within it, as a presentational projection of the world as image, the work shows firstly how and to what extent models “are not visual in the same sense as the various forms of visual representations of select parts of the world” (XIV). And secondly the analysis of various world-pictures supports the thesis that here “modeling, pictoriality and visualization are not secondary acts that are to illustrate something essentially abstract. […] Rather the pictorial dimension of world-pictures is from the outset part of the core of what makes a world-picture”. The culturally dependent world-pictures, as pictures, serve “a basically orienting, structuring and guiding function for people” in their perception, action and cognition (XVI). – For additional case studies following the question of which pictorial elements belong to world-pictures “of various times and cultures”, especially world-pictures in the narrow sense, what the “pictorial nature of world-pictures means for their world-picture status” and what is involved in these phenomena of visuality and pictoriality, see the contributions in Die Welt als Bild (Markschies/Zachhuber (2008); quotation: 13).

Knowledge and World-Pictures

121

is the case in a broader sense of ‘picture’ as well. In his lecture ‘The Age of the World-Picture’ Martin Heidegger diagnoses as the basic tendency of modernity that of making the world into an image. It is not just that the idea of passively mirroring the world as a whole, complete in itself, in representation makes no sense. Rather, Heidegger also remarks quite fittingly that in this case, aside from pictorial components in the narrow sense, the word ‘picture’ (Bild) entails a good deal of what we mean when we speak of ‘getting a picture of something’ or ‘giving someone a picture’ of something.89 Hence here ‘world-picture’ is conceptually linked above all with the cognitive relevance of projecting a worldview as a whole, which involves presenting as least as much as re-presenting (though perhaps while using pictorial means in the narrower sense to do so). This brings into view the pre-formative function of world-pictures (in the narrow sense and even more so in the broad sense). The pro89 Cf. Heidegger (1994), 89 ff. In fact we need to underline the critical constraints Heidegger sets: (i) our understanding of the formation of world-pictures cannot rest on a modern, primarily Cartesian metaphysics of the subject and the associated idea of representation. Here the world becomes picture in the sense that “representer-producer” (der vorstellend-herstellender Mensch), conceived as representing subject, seeks the “being of beings” (Sein des Seienden) within the entirety of the “representedness of beings” (Vorgestelltheit des Seienden) (Heidegger (1994), 89 f., 92). (ii) The broader world-picture cannot and may not be conflated with the narrower scientific world-picture oriented around mathematical and scientific knowledge of the world and technological action; see also Abel (2004), 120 f. Josef Simon emphasizes an interesting consequence of Heidegger’s diagnosis, namely that people today know “natural science and technology to be particular essential traits of their age and their world-picture, which has its time” (Simon (1993), 23) and that we have to ask about the responsibility not just for the consequences of people’s action within such a (narrow) world-picture but also “for the world-picture itself” (25), particularly insofar as we ‘make’ ourselves our own world-picture “together with a certain linguistic formation of our ‘worldview’” and “we ‘hold’ it ‘true’ in relying on it as truly ‘a priori’ in our actions towards others and towards nature” (24). Though of course we cannot turn to external standards in order to assess the value of world-pictures in the narrow sense (38), nonetheless Simon’s considerations on the issue of world-pictures bring out a specifically ethically moral dimension that is often overlooked in the discussion of sciences but that arises naturally in treating ‘made’ world-pictures as “the ‘latest’ versions of signs” (39). Simon reminds us to exercise a certain “caution” insofar as such a world-picture “is not just decisive for our ‘understanding of being’” but “also for our actions which we take or omit in holding it for ‘a priori’ true” (39).

122

Ulrich Dirks

cesses arising here are by no means bound to the form of the proposition and do not even need to occur in linguistic signs strictly speaking. Nonlinguistic signs, specifically pictorial signs, can fulfill all the relevant functions at all levels and are even particularly successful at doing so: (i) images that are close to the world-picture and thus equipped with a high degree of currently unquestionable certainty, e. g. “the picture of the earth as a ball”, can successfully orient our action and help us “in the judgment of various situations”.90 (ii) World-pictures in the narrow sense or the pictorial elements of them, as images we’ve made or that a culture or epoch has made, can certainly come to play a strong role in governing how we judge. In our life-world today with its immense medial stream of digital images, our life practices and everyday systems of orientation, interpretation and assessment might at least to some extent be more closely linked to deep-seated images and pre-conceptual elements that shape our worldview rather than with propositional beliefs. (iii) Even and especially at the level of the background world-picture in its broad sense (Wittgenstein, Abel), its pictorial elements and character play an important role.91 As with beliefs, familiar images can 90 OC 146. 91 In On Certainty Wittgenstein also clearly chose the term ‘world-picture’ with care, though he never systematically discussed it. His examples show a particular emphasis on the diverse orientational value that pictures can have for our actions and judgments and the contribution that such pictorial ideas can have for the semantics of our propositions in certain contexts. This equips pictures in a broad sense to insert themselves into the background of our ‘world-picture’ when it concerns particularly deep-seated, time-tested, simple and self-evident pictures that we operate with without doubting them. In particular this use of the concept of ‘picture’ does not take up the central components of the concept of picture found in Wittgenstein’s earlier picture theory in the Tractatus, which require representing a presupposed reality. We can agree with J. Simon that Wittgenstein is referring to precisely this picture of the picture when he says in paragraph 115 of the Philosophische Untersuchungen, looking back, that “A picture held us captive” (cf. Simon (1999), 190 ff.). It is the orientational value, the function of guiding and holding our thought, speech and understanding along certain paths, the potential exemplary functions (particularly for later applications of the rules and ‘repetitions’), the possibilities of revealing and at the same time generating contexts and rule-like connections in diverse ways, and the potential to show and express more than can be ‘said’, that constitute pictures together with their illustrative force (generated by a specific habituated practice of understanding signs) and that characterize, alongside other aspects, the use of ‘picture’ in the later writings of Wittgenstein, particularly in his Bemerkungen ber die Grundlagen der Mathematik (see ‘proof as picture’ in RFM) and the Philosophischen Untersuchungen.

Knowledge and World-Pictures

123

sediment themselves over time into the matter-of-course background that we cannot sensibly suspend at present. The role of the world-picture in the broad sense and its pictorial elements consists in “contributing to the functionality and orientation of our life practices and supporting them in the manner of background stabilizers”; and in everyday situations this occurs “generally in that we make for ourselves a picture of the circumstances”.92 The success of such a picturing process points us to the pictorial conditions presupposed and deployed in the background world-picture. Moreover, we should follow Abel in noting, firstly, that in its deep function such a world-picture “plays a role in defining the boundaries of what can count as a successful or a failed representation at all”93 ; we rely on this definition in everyday life, and use it methodologically as well in deploying pictures, visualizations, graphic computer simulations and imagining procedures in the sciences. Secondly non-linguistic and pictorial components “with an originally pictorial organizational power”94 play a role in the genuine world-picture processes that normatively and “formatively organize our understanding of self, others and the world”95 rather than just “reproducing a pre-fabricated given”. And insofar as, thirdly, pictorial notions clearly inform our successful use of linguistic expressions96 and insofar as the functioning of non-linguistic and specifically pictorial processes of symbolization ultimately lead us back to non-linguistic pictorial components in the background worldpicture, pictorial components are at least indirectly involved in the semantics of world-picture propositions and our propositional knowledge.97 92 93 94 95 96 97

Abel (2004), 131 f. Abel (2004), 139. Abel (2004), 142. Abel (2004), 139. Cf. Abel (2004), 140. Abel goes beyond the work of Wittgenstein in his examination of pictures and the relation between language and picture, specifically world-pictures, on the basis of a general philosophy of signs and interpretation Abel (2004), ch. 3 and 11. “World-pictures and pictorial worlds occur in signs” and their pragmatic and semantic properties are “not magically and timelessly built into the signs of the world-picture and the pictorial world” but rather, on Abel’s central thesis, enter into the signs “through the customary lifeworld practice of interpreting these signs” (145). This practice is a more basic and comprehensive level of consideration prior to world-pictures, which “themselves are manifestations of the semiotic and interpretational nature of humanity” (34). In this setting Abel examines the world-pictures and pictorial components so central to world-pic-

124

Ulrich Dirks

Incidentally, the talk of ‘world-pictures’ does not just refer to these pictorial background components, but should also be seen in light of the fact that world-pictures and their structure cannot be objectified and can only to a very limited extent be captured in propositions and theoretical sentences. The structure of a world-picture itself cannot be described “with literally denoting sentences, with propositions […] but rather ultimately only with the help of metaphors, hence in pictorial speech”98. Hence it is not a readily available network of logically inferential connections, but rather can only be expressed metaphorically – which, given the creative, constructive and productive re-organizational power of metaphors, means forming the world-picture structure itself with pictorial speech. Every one of these levels, in which dynamic processes of knowledge are anchored and embedded in diverse ways, present specific presuppositions and conditionalities for dynamic processes of knowledge. This brings us to an important finding for the philosophy and theory of knowledge: linguistic knowledge, specifically explicit, propositional, scientific knowledge, in its generation, justification, genealogy, meaning, orientational value, transformation, communication and application, is in part influenced, conditioned and formed, in complex and sustained ways, by non-linguistic factors, particularly by factors that are pictorial in a broad sense, in the hierarchally organized background worldpicture within which it is known.

tures and the power they have to guarantee certainty, stabilize action and ultimately form the world, and develops and defends four theses: (i) the power of a picture of the world, like the linguistic and propositional parts of a world-picture, comes from the fact that “it involves the web of uncircumventable and unsuspendable basic beliefs and intuitions that cannot be further justified or explained” Abel (2004), 137. (ii) Precisely due to their character of pre-conceptual attitudes, pictures of the world are suited to serving the functions of the world-picture in the area of the pre-conceptual attitudes, opinions and beliefs that are a precondition for our conceptual knowledge and its justification. (iii) Pictures of the world can exercise world-picture functions at the level of beliefs and opinions due to their non-linguistic pictorial nature (and not just their pre-conceptual nature), since these also (but not just) involve “non-verbal and pictorial aspects and attitudes”. (iv) In addition the peculiar power of pictures of the world consists in their “pre-theoretical and graphical power of organization” and ultimately their “organizing and standardizing force concerning our understanding of the world, ourselves and others”. Cf. Abel (2004), 136 – 149. 98 Abel (2004), 134.

Knowledge and World-Pictures

125

In relating the sentences falling under the category of ‘world-picture proposition’ to the world-picture as a whole, we are first confronted with the sentences (discussed in detail here) that count as propositional parts of the world-picture due to their characteristics and diverse functions. Particularly central are those fundamental sentences expressing a whole worldview and a grounding intuition, that embody the belief about the world as a whole and humanity’s place in it or that bring to light the world-picture as a whole. There are then sentences that lie very close to the world-picture and borrow an unquestioned certainty and a power to orient action from it (but do not contribute to our understanding of self, others and the world to any remarkable extent). There is in addition the pictorial and metaphorical language describing the structure of the world-picture. And finally many sentences serve as world-picture propositions in the specific sense (which stands in need of explanation) in which Wittgenstein can say of them that they describe world-pictures.99 Following our considerations on how world-pictures are given, this cannot occur in in propositional sentences reporting facts; nor does this concern secondary and retroactive descriptions of what we have reconstructed as part of our world-picture through reflection. In this context Wittgenstein uses ‘description’ in contrast to ‘explanation’ and thus emphasizes that the sentences with which we make some part of our uncircumventable and unsuspendable world-picture transparent cannot have the character of the justifications and explanations typical of scientific knowledge.100 99 Cf. OC 95 and 162. 100 The level of ‘ungrounded way of acting’ that cannot be further explained and is given with the world-picture, where in terms of my uncircumventable way of actually speaking, thinking and acting “I am inclined to say: ‘This is simply what I do’” (PI 217; cf. OC 212), is only accessible to description, not theory and explanation. But descriptions, as specific language games and as “instruments for particular uses” (PI 291) always occur under the perspective of specific epistemic interests; a description “gets its light, that is to say its purpose, from the philosophical problems”. Moreover, it is of cardinal importance for the success of description, particularly in the case of clarifying the background worldview and hence for those sentences describing the world-picture that “The problems are solved, not by giving new information, but by arranging what we have always known”, i. e. that are (dis-)solved “by looking into the workings of our language” (PI 109). Unlike the retroactively descriptive perspective in the re-constructive interpretational efforts of reflection on the world-picture, the ‘sentences describing the world-picture’ meant here bring out a different aspect of ‘description’. This

126

Ulrich Dirks

Concerning (b), the relation between world-picture and world, there are several reasons why we cannot make sense of the idea that something like ‘the world in itself’ and its ‘true constitution’ determines the world-picture. Against the idea of a pre-classified ‘ready-made world’ we can follow the tradition of Kant’s critical philosophy and its emphasis that (despite the principle independence of experience, which precludes the completely arbitrary construction of a concrete scientific world-picture) world and reality can always only given as this world and that reality under the humanly finite, perspectival, epistemic conditions in which we grasp it, i. e. only in dependence on categorizing organizational activity (such as that of the particular conceptual scheme).101 In addition we can point to the finding that world-pictures (particularly world-pictures) do not have the function of representing but rather “an original power of presenting ‘something as something’”, hence an “originary presentational achievement”102. The world-formative processes of the world-picture are found primarily when the creative “organization of a network of intersecting and interconnecting beliefs and opinions” produces “an orienting and standardizing view of self, others and the world”103. When with such an understanding of the world we classify the occurrences in our everyday world of experience into houses, soccer games and sunsets, the originary organizing and categorizing processes manifest themselves “in our practices of dividing and individuating the world and our orientation within it”. Such processes of spatio-temporal localization, individuation and formation of objects and their classification into types and species, etc.104 are always logically prior to the understanding of the world and the associated order of reality; they can be seen whenever we talk about a sunset as

101 102 103 104

concerns the immediate performance of such a world-picture sentence and can be exemplified by a different use of the word ‘description‘. Just as the path of a flying object can be described by sentences and equations but is also described by the object itself in its flight, some sentences performed in certain contexts and situations can ‘describe’ a world-picture, not representing it but presenting it in an actively organizational, generative way. For a general and philosophically relevant distinction between performance and interpretation of signs see Abel (1999), ch. 4. For a thorough discussion of this point, including Abel’s ‘worlds of interpretation’ approach, esp. in Abel (1993), and its relation to Nelson Goodman’s philosophy of ‘worldmaking’, esp. in Goodman (1978), see Dirks (2010b). Abel (2004), 139. Abel (2004), 139. Abel (2004), 144 ff.

Knowledge and World-Pictures

127

a matter of course and sail back to the harbor, when we take another event to be dissimilar to a soccer game and go past it with the remote to the sports channel, etc., – more generally: when we successfully classify, identify and recognize things, note similarities, etc. Hence the formative, organizational, standardizing force of the world-picture in the broad sense of the unquestioned foundation of all thought, speech and action (of a time and a culture) manifests itself in our practices. Wittgenstein’s example of a world-picture proposition, “there are physical objects”, cannot be refuted by experience, by the world and reality, since it describes the background world-picture with a view to its categorizing, word-forming organizational function. On the other hand and with a view to other, foreign world-pictures, not all possible or actual ways of forming the world, not all actual ‘worlds’, form that world that counts for us in a certain culture and a certain epoch as the familiar, self-evident and real-life world. Rather it is our deep-seated world-picture in the broad sense, as the ‘inherited background’ to which our familiar, unquestioned and “habituated patterns of dividing the world into types and species” and other productive organizational processes are bound, that takes part “in a non-circumventable sense in the formation of that which then holds as a, or the, ‘real world’”, of which “we cannot imagine or conceive that the things could behave different in the fundamental respects”.105 Scientific practices such as constructing concepts, models and theories, formulating laws, and conducting experiments and observations also depend on the world-picture and its organizational force (even if they depend on it to some extent in the indirect and asymmetrically reciprocal sense discussed earlier). Insofar as propositional knowledge operates within concepts and judgments, the world-picture in the broad sense is also logically and genealogically prior to scientific knowledgethat and a condition of it in this regard as well. Whenever we judge that something is a neutron star or falls under the category of oxidation 105 Abel (2004), 144 f. Insofar as the world as it is directly given in experience relates to the lifeworld in a broad sense and the horizon it provides for our experiences, and their organization can be seen as the manifestation of habituated patterns and processes of classification, we see once more here the deep interconnection between world-picture and lifeworld. On the relation between ‘realism’ and that which is self-evident and a matter of habit see also Goodman (1976), ch. I, and on the internal and unsurpassable content of reality in signs directly used and understood, including the signs of the world-picture, see Abel (2004), introduction and elsewhere.

128

Ulrich Dirks

reactions, for example, it involves practices of identification, of noting similarities, applying patterns, etc., the success of which is ultimately grounded in the background assumptions of the world-picture with its organizational force. In light of the perspective followed here we can identify the tasks of future knowledge research as follows: (a) the deeper analysis, taking the concrete circumstances of individual disciplines into account, of the internally dovetailed, triangular relations between (i) propositional and specifically scientific knowledge, (ii) the deeply hierarchical background world-picture (and its non-propositional elements as well) on which knowledge is conditional and dependent in a complex, dynamic manner and in a number of respects, and (iii) the world and reality, of which our knowledge, in light of our world-picture, is knowledge, and which as the ordered, classified and structured world and reality is always to be seen as formed by the original organizational processes of the world-picture. (b) An expanded investigation starting with knowledge in non-empirical or not primarily empirical disciplines and ultimately turning to other forms of knowledge that mutually inform one another beyond the propositional knowledge-that focused on here.106,107

References Abel (1993): Günter Abel, Interpretationswelten. Gegenwartsphilosophie jenseits von Essentialismus und Relativismus, Frankfurt/Main. Abel (1995): Günter Abel, “Unbestimmtheit der Interpretation”, in: J. Simon (ed.), Distanz im Verstehen. Zeichen und Interpretation II, Frankfurt/Main, 43 – 71. Abel (1999): Günter Abel, Sprache, Zeichen, Interpretation, Frankfurt/Main. Abel (2004): Günter Abel, Zeichen der Wirklichkeit, Frankfurt/Main.

106 Besides a conception of the conditions of fulfillment for knowing-how, cf. Abel (2010) in terms of the world-picture and the associated life-form, it seems especially promising to apply the differentiation and expansion of the concept of knowledge (and action) undertaken by Abel in his unified theory of knowledge and action to the levels of patterns of habit and categorizing organization (Abel (1999), ch. 13) in explaining the relation between the world-picture and knowledge in the broad sense. 107 Translation from German by Karsten Schöllner.

Knowledge and World-Pictures

129

Abel (2010): Günter Abel, “Knowing How. Eine scheinbar unergründliche Wissensform”, in: J. Bromand/G. Kreis (eds.), Was sich nicht sagen lsst. Das Nicht-Begriffliche in Wissenschaft, Kunst und Religion, Berlin, 319 – 340. Abel (2011): Günter Abel, “Der interne Zusammenhang von Sprache, Kommunikation, Lebenswelt und Wissenschaft”, in: C. F. Gethmann (ed.), Lebenswelt und Wissenschaft. Kolloquiums-Vortrge des XXI. Deutschen Kongresses fr Philosophie, Universitt Duisburg-Essen, Hamburg, 351 – 371. Brecht (1963): Bertolt Brecht, Leben des Galilei, Frankfurt/Main. Danielson (2001): Dennis R. Danielson, “The Great Copernican Cliché”, in: American Journal of Physics 69/10, 1029 – 1035. Dilthey (1968): Wilhelm Dilthey, “Die Typen der Weltanschauung und ihre Ausbildung in den metaphysischen Systemen” (1911), in: Wilhelm Dilthey: Weltanschauungslehre. Abhandlungen zur Philosophie, (= Gesammelte Schriften, vol. VIII, ed. B. Groethuysen), Stuttgart/Göttingen, 73 – 118. Dirks (2002): Ulrich Dirks, “Mathematik und Lebensform”, in: G. Abel/H.-J. Engfer/C. Hubig (eds.), Neuzeitliches Denken. Festschrift fr Hans Poser zum 65. Geburtstag, Berlin/New York, 49 – 65. Dirks (2007): Ulrich Dirks, “Holismus und Gewißheit”, in: G. Abel/M. Kroß/ M. Nedo (eds.), Ludwig Wittgenstein. Ingenieur – Philosoph – Knstler, Berlin, 179 – 191. Dirks (2008): Ulrich Dirks, “Modelle des Holismus”, in: Dirks/Knobloch (2008a), 47 – 74. Dirks/Knobloch (2008a): Ulrich Dirks/Eberhard Knobloch, Modelle, Frankfurt/Main. Dirks/Knobloch (2008b): Ulrich Dirks/Eberhard Knobloch, “Modelle. Probleme und Perspektiven”, in: Dirks/Knobloch (2008a), 9 – 28. Dirks (2010a): Ulrich Dirks, “Interpretation/Interpretationsphilosophie”, in: H. J. Sandkühler (ed.), Enzyklopdie Philosophie, vol. 2, Hamburg, 1142 – 1152. Dirks (2010b): Ulrich Dirks, “Welt/Welten”, in: H. J. Sandkühler (ed.), Enzyklopdie Philosophie, vol. 3, Hamburg, 2953 – 2962. Dux (1982): Günter Dux, Die Logik der Weltbilder. Sinnstrukturen im Wandel der Geschichte, Frankfurt/Main. Gebhardt/Kiesel (2004): Hans Gebhardt/Helmut Kiesel (eds.), Weltbilder, Berlin/Heidelberg/New York [Heidelberger Jahrbücher 47 (2003)]. Gethmann-Siefert (1996a): Annemarie Gethmann-Siefert, “Weltanschauung”, in: J. Mittelstraß (ed.), Enzyklopdie Philosophie und Wissenschaftstheorie, vol. 4, Stuttgart/Weimar, 652 – 653. Gethmann-Siefert (1996b): Annemarie Gethmann-Siefert, “Weltbild”, in: J. Mittelstraß (ed.), Enzyklopdie Philosophie und Wissenschaftstheorie, vol. 4, Stuttgart/Weimar, 654. Goodman (1976): Nelson Goodman, Languages of Art. An Approach to a Theory of Symbols, Indianapolis. Goodman (1978): Nelson Goodman, Ways of Worldmaking, Indianapolis. Heidegger (1994): Martin Heidegger, “Die Zeit des Weltbildes” (1938) and “Zusätze”, in: Martin Heidegger, Holzwege, Frankfurt/Main, 75 – 113.

130

Ulrich Dirks

Kant (1787): Immanuel Kant, “Kritik der reinen Vernunft”, in: Die Königlich Preußische Akademie der Wissenschaften (ed.), Kant’s gesammelte Schriften, vol. III, Berlin 1902 ff. Kuhn (1970): Thomas S. Kuhn, The Structure of Scientific Revolutions, Chicago. Laeyendecker (1997): Leo Laeyendecker, “Weltanschauung”, in: E. Fahlbusch et al. (ed.), Evangelisches Kirchenlexikon. Internationale theologische Enzyklopdie, vol. 4, Göttingen, 1254 – 1255. Lütterfelds (2008): Wilhelm Lütterfelds, “Sprachspiele – Weltbilder – Lebensformen im interkulturellen Kontext”, in: Dokos. Revista filosfica 2, 3 – 47. Markschies/Zachhuber (2008): Christoph Markschies/Johannes Zachhuber (eds.), Die Welt als Bild. Interdisziplinre Beitrge zur Visualitt von Weltbildern, Berlin. Markschies/Reichle/Brüning/Deuflhard (2011): Christoph Markschies/Ingeborg Reichle/Jochen Brüning/Peter Deuflhard (eds.), Atlas der Weltbilder, Berlin. Mies (1999a): Thomas Mies, “Weltanschauung”, in: H. J. Sandkühler (ed.), Enzyklopdie Philosophie, vol. 2, Hamburg, 1733 – 1737. Mies (1999b): Thomas Mies, “Weltbild”, in: H. J. Sandkühler (ed.), Enzyklopdie Philosophie, vol. 2, Hamburg, 1737 – 1739. Mittelstraß (1997): Jürgen Mittelstraß, “Weltbilder. Die Welt der Wissenschaftsgeschichte”, in: J. Mittelstraß, Der Flug der Eule. Von der Vernunft der Wissenschaft und der Aufgabe der Philosophie, Frankfurt/Main, 228 – 254. [“World Pictures. The World of the History and Philosophy of Science”, in: J. R. Brown/J. Mittelstraß (eds.), An Intimate Relation. Studies in the History and Philosophy of Science, Dordrecht 1989, 319 – 341] Müller (2010): Arnulf Müller, Weltanschauung – eine Herausforderung fr Martin Heideggers Philosophiebegriff, Stuttgart. Poser (1988): Hans Poser, “Mathematische Weltbilder. Begründungen mathematischer Rationalität”, in: P. Hoyningen-Huene/G. Hirsch (eds.), Wozu Wissenschaftsphilosophie? Positionen und Fragen zur gegenwrtigen Wissenschaftsphilosophie, Berlin/New York, 289 – 309. Poser (2001): Hans Poser, Wissenschaftstheorie. Ein philosophische Einfhrung, Stuttgart. Poser (2008): Hans Poser, “Modelle, Simulationen, Weltbilder. Der Aufbruch in die Komplexität”, in: Dirks/Knobloch (2008), 173 – 186. Putnam (1981): Hilary Putnam, “Two philosophical perspectives”, in: Hilary Putnam: Reason, Truth and History, Cambridge/MA, 49 – 74. Silvar/Kunze (1993): Maja Silvar/Stefan Kunze (eds.), Weltbilder, Frankfurt/ Main. Simon (1993): Josef Simon, “Weltbild und Gewissen”, in: Allgemeine Zeitschrift fr Philosophie 18/1, 23 – 39. Simon (1999): Josef Simon, “Lebensformen. Übergänge und Abbrüche”, in: W. Lütterfels/A. Roser (eds.), Der Konflikt der Lebensformen in Wittgensteins Philosophie der Sprache, Frankfurt/Main, 190 – 212. Stegmaier (1997): Werner Stegmaier, “Weltbild, Weltorientierung (1. Allgemein)”, in: E. Fahlbusch et al. (ed.), Evangelisches Kirchenlexikon. Internationale theologische Enzyklopdie, vol. 4, 1255 – 1259.

Knowledge and World-Pictures

131

Thomé (2005a): Horst Thomé, “Weltanschauung”, in: J. Ritter/K. Gründer/ G. Gabriel (eds.), Historisches Wçrterbuch der Philosophie, vol. 12, 453 – 460. Thomé (2005b): Horst Thomé, “Weltbild”, in: J. Ritter/K. Gründer/G. Gabriel (eds.), Historisches Wçrterbuch der Philosophie, vol. 12, 460 – 463. Weizsäcker (2002): Carl Friedrich v. Weizsäcker, Zum Weltbild der Physik, Stuttgart. Wittgenstein (1958): Ludwig Wittgenstein, Philosophical Investigations, G. E. M. Anscombe/R. Rhees (eds.), Oxford. Wittgenstein (1975): Ludwig Wittgenstein, On Certainty, G. E. M. Anscombe/ G. H. von Wright (eds.), Oxford. Wittgenstein (1978): Ludwig Wittgenstein, Remarks on the Foundation of Mathematics, G. H. von Wright/R. Rhees/G. E. M. Anscombe (eds.), Oxford.

Knowledge Forms and Language Forms Hans Julius Schneider Formulation of the Question The expression Wissensformen [knowledge forms] appears to me to signal two different questions.1 Firstly, it suggests the idea that one and the same knowledge (e. g. of the interplay of a machine’s individual parts) could be relayed in different ways, in different forms, for instance, through language, through a verbal description of the machine’s parts and the way they work together, or through illustration, e. g. a technical drawing. The chosen language on the one hand and the technical drawing on the other would then be two forms of knowledge. The knowledge itself (the content) is independent of each form in which it may be manifested in the sense that the same knowledge can be manifested in different forms. The knowledge at issue in the example provided is propositional; it is knowledge that something is the case. For this type of knowledge an image suggests itself in which the forms of knowledge appear as different containers that transport a certain content (the so-called ‘propositional content’). Further, this traditional distinction between content and form invites the idea that what one knows be either independent of any form, so that a form for the knowledge must first be found if it is to be ‘expressed’ and transmitted. All forms would then be forms of expression, strictly speaking, the forms of the ‘containers’, not those of the knowledge itself. One might also believe, though, that knowledge as knowledge (i. e. prior to being ‘expressed’) already had a form proper to it, a form distinguishable from the different medial forms (e. g. of a certain language, a certain sort of technical drawing or blueprint). If knowledge itself had a form, then we would have to distinguish between two separate meanings of the expression ‘knowledge form’: it could designate a (supposed) universal form of all knowledge in general (the ‘knowledge form’ in the singular, which is identical, according to some opinions, with ‘logical form’) or, as the knowledge forms in 1

Cf. Schneider (2002a); Schneider (2004).

134

Hans Julius Schneider

the plural, it could refer to formal properties of each type of sign (e. g. language vs. a technical drawing). At this point connected questions arise whose answers are by no means trivial, first of all: what might we understand under ‘unformed content’, and secondly: how are we to think a form of knowledge independent of any type of sign? If this form independent of signs is supposed to be what is traditionally called the ‘logical form’, is there still (in spite of the plurality of logical calculi) only one logical form, and what does this mean exactly? A vague psychologism in the theory of meaning, in the sense of the answer that the (formed or unformed) content is what ‘floats before’ one in the mind in the utterance or understanding of the corresponding signs, cannot satisfy a philosopher of language. Hence, the question arises how the concept of ‘propositional content’ can be determined so that it becomes understandable why a verbal description and a technical drawing can have such a content in common. If this is not something in the mind that is already there before its expression (and is possibly already formed): What is it then? And how can we decide whether this content already has a form or not? An entirely different approach to the topic of knowledge forms arises when one thinks the concept of knowledge quite broadly, and does not limit it to a propositional knowledge that something is the case.2 Every understanding of what an object is about (e. g. a work of art3), every appropriate conception of a symbolic structure’s semantic relevance could be treated as a sort of knowledge in this second approach, although the question at first remains open whether it is sensible in such cases to count on the possibility that there be cases in which we have one permanent ‘substance’, a content, something known that can appear ever again as the same in different forms. Independently of this question, in this approach the expression ‘knowledge forms’ would address the many and various types of semantic relevance. As the danger exists that this usage of the expression ‘form’ would already imply the possibility of the same content in different forms, and as precisely 2

3

I myself have suggested employing a narrow understanding of knowledge, defining it as a special case of ability, namely the ability to answer. Non-verbalisable regions of ability are then not to be called types of knowledge, although the English expressions tacit knowledge and knowing how suggest otherwise. Cf. Schneider (2002b) On the question whether it might be sensible to say of rules that they lie ‘behind’ our actions, see Schneider (2003). On the region of non-propositional knowledge see Schildknecht (2002). Danto (1981) convincingly demonstrated that the semantic content (“aboutness”) is a necessary trait of all objects that can be called artworks.

Knowledge Forms and Language Forms

135

this is questionable (- as a look at the case of the artworks shows -), this way of speaking is deceptive, so that I prefer in this case to speak not of forms, but of types of knowledge, or rather (if the concept of knowledge is to be understood more strictly) of types of semantic relevance (of works of art, of scientific textbooks, etc.). Under this aspect of a non-propositional knowledge, the interesting thing would not be language as one knowledge form and technical drawing as another, but rather e. g. already within language different ways of meaning would become apparent: poems or some types of philosophical treatises articulate other types of knowledge (as I will now say) than travel accounts or operating instructions. An ‘analysis of knowledge forms’, then, would be an investigation of the types of semantic content without the prerequisite that it must be possible to pour an identical content into different forms.4 This second approach (as was already mentioned) is broad enough to include artistic products. These admittedly have the function of signs (as Arthur Danto would say, they are ‘about something’), and we can articulate their sign-character with the statement that they have a ‘semantic content’. But this does not mean that works of art necessarily express a ‘knowledge that p’. The concept of the semantic content encompasses more than that of propositional content. To this corresponds the well-known fact that the content of a work of art transcends what e. g. an art historian can relate to us through propositions. I shall here treat the first problem, specifically the question whether propositional knowledge itself has a form (or a plurality of forms), for then we could speak of ‘knowledge forms’, or whether this can only be said of the different modes of articulating such knowledge, for then we could speak e. g. of ‘language forms’ or forms of portrayal in plastic art, but not of the form or forms of ‘knowledge itself’. I shall claim the latter. Thus, my thesis is that there are no knowledge forms before or ‘over’ any forms of articulation, neither in a psychological nor in an ideal (logical) sense. Especially with regard to the region of the verbal, I would like to show that natural languages (metaphorically speaking) are not media with imperfect, or, toward content, opaque forms over against which a logical language would have to be regarded as a better, more transparent medium that exhibits the forms of knowledge in an undistorted way. Other than our usual mode of expression might lead us to surmise, language is no medium at all, no middle part for the mediation of independently possible meanings that by themselves would possess a form. Positively, my thesis 4

On the limits of translation cf. Schneider (2009).

136

Hans Julius Schneider

is: language is the life element for a large family of different actions that invite ever new continuations of action. And I hope that this answer to the question of the locus of the form of propositional knowledge also casts light on the region of art and on one kind of non-propositional knowledge. I thus touch also on the second question mentioned above, that of the types of semantic content.

The Everyday Notion of Language: the Transport Model of Communication Where language is discussed as a knowledge form in the sense of the former question, it is wont to be regarded as a vehicle with which we bring a content to expression that itself remains the same, and that we (e. g. with the aid of scientific textbooks) can transfer on through generations. This talk of ‘expression’ contains the traditional distinction between content and form: one and the same knowledge (or some other mental content, e. g. a supposition or wish) can be ‘brought to expression’ using different forms. And at first this seems plausible. After all, the equation ‘two plus two equals four’ can be expressed e. g. in English words, or with the help of Arabic or Roman numerals, and we often describe such facts by saying that we had one content before us that we had put (in this case) into three forms. The same also goes for the case of technical drawing already mentioned. This view becomes problematic when a false image of the procedure of communication is connected to the underlying facts, an image that I shall here call the transport model. According to this model, communication is concerned with allowing a mental content to go from one person, the speaker, to another person, the hearer. In the framework of this model, language appears as that which mediates between the speaker and the hearer. It is the ‘middle thing’, or in Latin, the medium, that takes up the content, which the speaker wants to express, to bring out of the privacy of his mind, in order that it can be transported to the hearer. As the medium, in contradistinction to the thoughts, consists of public, sensibly perceptible objects, because one and the same acoustic speech-event (qua the same sequence of sounds; or: the same drawing) can be perceived by different persons, the lingual utterance goes effortlessly from the speaker to the hearer. And if the hearer knows the language being used, he can take the content meant by the speaker out of the words and sentences, according to this

Knowledge Forms and Language Forms

137

model, just the way one takes a letter out of an envelope. Communication is hence successful where the thought sent on its way by the speaker (or an equivalent copy of this thought) has safely arrived at the hearer. A metaphor from telegraphy is quite popular in this context: the speaker ‘encodes’ a message (content) in a form transportable on the way planned (e. g. in Morse code, which can be transmitted as light signals), and the hearer ‘decodes’ this (if he knows the ‘code’) and thus receives an equivalent of the original message. The message itself, the transported thought, is thus thought as a mental object, an invisible object for which the question whether it has a form and what kind of a form this might be is a meaningful question. The thought is primary, in contrast to the medium, which has only an auxiliary, mediating function, as compared to the message itself, the transported thought.5 From here it is only a short step to the idea that the medium of natural language needed and were capable of improvement. Does it not often occur that the lingual expression renders the speaker’s thoughts only vaguely discernible or that it even distorts them completely? Is natural language not a dim medium, like a container out of frosted glass, which shows its content (the thought) only in fuzzy contours? Do we not say, from time to time, that we had expressed ourselves poorly, and had really meant (thought) something different from what we had imperfectly expressed? Because there appear in fact to be such cases, the obvious question arises whether an artificially designed language, or a written form of representation, might not be able to overcome the pitfalls of natural language. Should it not be possible to create a crystal clear, perfectly suited medium, at least for scientific purposes, that could make our ideas recognisable in exactly the shape they had when we had ‘meant’ them? Is e. g. the Begriffsschrift by Gottlob Frege6 not a promising step in this direction? This program claims that knowledge has its own form to which the forms of different media can correspond to a greater or lesser extent, and that the natural languages are not optimal in this regard. I would now like to show that an meticulous pursuit of this question of the possibility of an unclouded, so to speak ‘invisible’ medium, leads to the conclusion that already the talk of a ‘middle’, of a medium, is misguided, 5

6

This analogy is deceptive because in the case of telegraphy a formulated text is already available (not only something that is intended, or ‘meant’) that is replaced by an equivalent using other signs. The encoding is thus a matter of changing something within the region of expressions. Frege (1964a).

138

Hans Julius Schneider

regardless of whether this medium is thought to be cloudy or clear. The attempt to apply the idea of perfecting the means of transportation, the concrete steps along the way of creating a vehicle that is practically invisible in so far as it conforms perfectly to the form that the transported contents themselves already have, in fact makes the vehicle really even more visible, and simultaneously takes away its status as a mere auxiliary for treating a thing independent of it. This step leads from later Frege to later Wittgenstein. What appeared to be a mere vehicle of transportation of something (a thought), namely, the generation of articulate sounds, turns out to be itself this ‘something’. Meaning is not an independent object of some sort, not for instance something ideal ‘behind language’, that is only transported through it, but talk of the meaning of a sign is the talk of this sign itself (with its forms), though only under the aspect of its use, not e. g. that of its material properties. But if this is right, then the transport model of communication is thus forfeited. Where there is no more separation between a thought as an independent structure of its own on the one hand and a medium that only transports this structure on the other, communication can no longer be regarded as the transport of a content from one person to another. And this means for the topic of knowledge forms that not only in the region of art, but also in everyday life and in many regions of science, the idea of a strict separability of content and form is erroneous. This is recognisable in the humanities and social sciences wherever certain e. g. religious objects cannot be spoken of otherwise than in the words of those affected.7 The question of the form of the knowledge laid claim to, according to what is called a ‘medium’ in the old model, is hence more often of significance than the model would lead us to believe. Once one has left the dubious psychologistic folk theories behind, the apparently normal case and the special case (namely the inexhaustible content of an artwork) exchange their roles: the fact now seems in need of explanation that in special cases the form in fact makes no difference and hence it appears as though there were one content independent of the chosen form. I shall return to this point. Once we have reached this demonstrandum, we need another model of communication that does not grasp it as the transportation of something intended by means of a middle link, a vehicle. To name the alternative I have in mind I would suggest the expression ‘life element’ as a provisorium, a term I found in Wittgenstein.8 It appears sufficiently to signal that the 7 8

Cf. Taylor (1981); cf. Schneider (2008). Wittgenstein (1979), 36 (§ 105).

Knowledge Forms and Language Forms

139

concrete lingual expression is not a tertium, not a vehicle, nor any middle thing between the intended meaning of the speaker and the perceived meaning of the hearer. To speak of a ‘life element’ here means: the meaning cannot be separated from its bearer (e. g. the lingual utterance), nor ‘taken out’ of what is called a ‘container’ in the transport model. It is not the content of a form that must be regarded as a receptacle. Who destroys the so-called receptacle, thus destroys the meaning as well, – which e. g. in the case of a painting is obvious. That there are cases in which such a separation is nevertheless possible then becomes the fact that needs to be explained. If one were to look for an alternative image for the process of lingual communication, i. e. for the production of community in the mutual exchange of speaking and listening, then, instead of thinking of the transportation of contents, one might think of what are called the ‘communicating vessels’ in physics. This refers to a system of tubes of different breadth, length and placed at different angles that are connected at their bases to one another, all containing a fluid, and all part of one system. If the fluid level in one of these tubes is changed, e. g. more being poured in, this affects all the tubes connected to it as well, but for this it is not necessary that molecules (partial contents) of the fluid poured into one tube come into the other, whose alteration is observed. Nothing needs to be transported from tube one into tube two, in order that we may say that these two are communicating with one another. Similarly, one might say: when persons communicate with one another lingually, what is said normally makes a difference for those in question, even if the transportation of an identical meaning from one to the other cannot be spoken of. As any analogy, this one also has its limits. In particular, I would explicitly point out that my choice of this image is no votum for a physicalistic, e. g. for a causal communication theory.

Gottlob Frege’s Ideal of a Transparent Medium and the Coming into Appearance of the “Life Element” I will not go into the psychologistic variant of the transport model of communication and its roots in traditional epistemological ideas here, but turn immediately to the question what becomes of this model when one attempts, following Frege’s example, to keep everything psychological out of a theory of meaning. What are meanings if they are not

140

Hans Julius Schneider

mental accompaniments to speaking; what are mental contents if they are not psychological objects? Frege’s answer is that they are logical objects, and his contribution to the answer of the next natural question what this answer means, i. e. what we are to understand under the term ‘logical object’, is an indirect and practical answer, namely the drafting of his Begriffsschrift. This is intended to show what the meanings of propositions (i. e. what propositional contents) are, according to their ideal form, by showing how we can represent these contents as faithfully as humanly possible. Frege’s Begriffsschrift would thus be the result of an attempt to render the forms of knowledge themselves visible, independent from all already extant media. Frege is hence concerned with separating the meaning, or content, of a sentence from its lingual expression. But by seeking to develop a means of expression that is as perfectly suited to this meaning as possible, he is forced to deal with the tension between meaning and expression as its material ‘carrier’. This leads him to observations from which Wittgenstein later drew conclusions that ought to lead us, in my opinion, not only to abandon the transport model of communication in its psychological, but also in its logical version, and to give up the idea that language were a medium, a mere mediator of mental contents that were independent of the mode of transportation. In the terminology used here, what was hitherto called a ‘carrier’ of meaning becomes its ‘life element’. How, then, is the drafting of a ‘concept script’ to be understood that pursues the goal of portraying truth-relevant structures of meaning (knowledge forms) explicitly and unambiguously and entirely separated from the forms of any natural language? The activity mentioned here can be interpreted in one of two ways. The former corresponds with Frege’s explicit teachings. It is that the logician looks at the objective, pre-given structures of the logical objects and develops recommendations for lingual or graphic forms that reproduce these structures as faithfully as possible. The Begriffsschrift is intended as a transparent medium in this sense, a medium that frees up the view of knowledge forms unblurred and undistorted. And as a material carrier of meaning, as a medium, it should itself appear as little as possible, like a well-made and well-cleaned pane of glass in a window. The second interpretation of what occurs in the drafting of the Begriffsschrift accords less with Frege’s explicit doctrine, but keeps closer to what actually happens in his writings (in part contrarily to this doctrine), and what Frege in many writings rather implicitly concedes than openly states. This interpretation is: the logician drafts a language whose properties are not determined by a pre-given logical realm of meaning (the forms of possible knowledge), but by demands that he himself places upon his draft. Ac-

Knowledge Forms and Language Forms

141

cording to this second interpretation the Begriffsschrift is no more a medium, or a middle thing, than is a natural language, because it neither just takes up and retraces the structure of any pre-given thing, nor does it carry what was invisible as mentally present into the visible and transportable, whether the mental realm is taken (psychologistically) as the inner life of the speaker and hearer, or (according to Frege) as a separate ‘Platonic’ region of mental objects. Hence, the Begriffsschrift too can be called a ‘life element’ of communication in the terminology used here, a place for activities whose properties are determined by the demands placed upon the activities, not by something that precedes them and toward which they could orient themselves.9 In his explicit commentaries Frege portrays his activity in the development of the Begriffsschrift as though he were looking through expressions of natural language that he uses as examples to the realm of ‘knowledge forms’, of logical relations, making this realm visible for the reader through his selection and commentary of these examples, – admittedly not immediately visible, yet recognisable for them if they only take his ‘hints’ correctly. What he calls a ‘thought’ in his non-psychologistic sense (i. e. the meaning, the propositional content of a declarative sentence), according to his opinion, exists independently of whether it is thought or conceived by a human being. This would really have to mean that such a thought would on its own exhibit parts that were ordered in such a way that the constructor of a Begriffsschrift could use the parts and their relations to each other to guide him in determining the structure of what he proposes as a ‘concept script’. Where the thought itself displays distinctions and differences (where knowledge itself has a form) the Begriffsschrift must follow it. Here Frege is faced with a dilemma. On the one hand he writes explicitly about the thought that it [has no need of a carrier. It does not first become true when it is discovered; just as a planet has been in interaction with other planets already before anyone has seen it]”…bedarf keines Trägers. Er ist wahr nicht erst, seitdem er entdeckt worden ist; wie ein Planet, schon bevor jemand ihn gesehen hat, mit anderen Planeten in Wechselwirkung gewesen ist.”10 And in a footnote he adds: [When one conceives a thought, or thinks, one does not create it, but only comes into a certain relation to what already existed before.]”Wenn man einen Gedanken fasst oder denkt, so schafft man ihn nicht, sondern tritt nur zu ihm, der 9 Therefore, Wittgenstein will later speak of the ‘autonomy’ of logic and grammar. Cf. the developments that follow here, as well as Schneider (2006). 10 Frege (1967), 354.

142

Hans Julius Schneider

schon vorher bestand, in eine gewissen Beziehung …“11 The thought is thus supposed to be independent of the lingual expression, so that the constructor of the Begriffsschrift, whose structure is not simply supposed to follow that of natural language, can orient himself toward the thought structure. On the other, Frege says that the thought itself has no parts, so that one cannot simply assume that the outline of the concept-scriptural expression followed the outline of the thought to be expressed. He says in another passage as well that man, with his weak intellectual powers, needs a perceptible sign, the formulated sentence and needs the concept expression, in order to be able to grasp thoughts or concepts.12 Is this supposed to mean that beings with higher intellectual powers can grasp thoughts in their unstructured state, in a sort of mystic vision? From whence does the structure come then? If we are dependent upon visible signs, but the thought itself is unstructured, how shall it be determined whether a complex sign we recommend be structurally appropriate to the thought or not? How can the draft of a Begriffsschrift be oriented? These questions remain unanswered in Frege. We can tell, however, that he finds himself forced to return to the signs wherever his doctrine, according to him, must be concerned with something that stands behind the signs, and that (at least for us at the moment of grasping) would have to be already structured, if we were to have any clue how the structure of a suitable means of expression must look. When we examine what Frege de facto does when he wants to convince the reader of the plausibility of a regulation he recommends for his concept script, e. g. when he attempts to demonstrate that logic requires a distinction that natural language does not make, then we see that he appeals to the competence of his reader to distinguish sense from nonsense, to separate what one can say from what one cannot say. In explications he admittedly keeps using the traditional talk of that ‘for which the questionable expression stands’, e. g. ‘for an object’ or ‘for a concept’. This sounds as though he were looking to a language-independent realm of the forms of knowledge. Yet, when it comes down to it, this ‘standing for something’ is explained with the semantic role that the expression has, i. e. the type of speech act that one can execute with it. So for example the central property of such a concept expression as ‘table’ or ‘red’ for his logic, namely, not to be saturated, not to stand for a whole that exists for itself, is determined through its semantic role, namely to complement a name (i. e. an expres11 Frege (1967), 354, fn 5. 12 Frege (1964b), 107 f.

Knowledge Forms and Language Forms

143

sion that stands for a whole) in such a way that the entire expression thus yielded builds a sentence: ‘table(x), red(y)’. As has been mentioned, Frege still makes use of the manner of speaking, a term of expression ‘means a concept’, or, generalised, a functional expression ‘means a function’. Yet in the posthumous publication entitled by the publishers‘ Ausfhrungen ber Sinn und Bedeutung’ [Comments on Sense and Meaning] he writes about the priority relation between the sign and the ‘object signified’: [accordingly, I call the function itself unsaturated or lacking because its name must first be complemented by the sign of an argument in order to receive a completed meaning] “Demgemäß ist die Funktion selbst von mir ungesättigt oder ergänzungsbedürftig genannt, weil ihr Name erst durch das Zeichen eines Arguments ergänzt werden muss, um eine abgeschlossene Bedeutung zu erhalten.”13 Concepts and functions are thus no special objects; what is to be expressed by the apparent reference to something named, is a property of the expression’s respective semantic role. The mode of speaking of that ‘for which a sign stands’ is a traditional makeshift to express that in the context of a philosophy of language one must not speak of signs as things, e. g. as traces of a printer’s ink on the paper, but of signs under the aspect of their usage. This view, which already heralds later Wittgenstein, becomes quite distinct in a passage where Frege addresses the word ‘and’ in its ‘unsaturatedness’, i. e. its need for completion. In his text [Compound Thoughts] Gedankengefge, he writes: [Of course, as a mere thing the group of letters ‘and’ is no less saturated than any other thing. With respect to its usage as a sign that can express a meaning, it can be called unsaturated because in this regard it can only have the intended meaning in a position between two sentences. Its end as a sign requires a complement in the form of a preceding and a following sentence.] “Als bloßes Ding ist die Gruppe von Buchstaben ‘und’ freilich ebenso wenig ungesättigt als irgendein anderes Ding. Im Hinblick auf seine Gebrauchsweise als Zeichen, das einen Sinn ausdrücken soll, kann man es ungesättigt nennen, indem es hier nur in der Stellung zwischen zwei Sätzen den gemeinten Sinn haben kann. Sein Zweck als Zeichen verlangt eine Ergänzung durch einen vorhergehenden und einen nachfolgenden Satz.”14

13 Frege (1969), 129. On the difficult problems connected with Frege’s distinction between concept and object, cf. Schneider (1995). 14 Frege (1967), 381.

144

Hans Julius Schneider

The execution of this attempt to render the structure of ‘ideal’ or ‘abstract’ contents (i. e. the forms of knowledge independent of language) visible through the development of a transparent medium leads to the result that such a realm of structured contents (in a sense independent of any medium whatsoever) simply does not exist. What we are wont to interpret as such structures are rather properties of verbal actions or of the results of actions (signs) that they exhibit under the aspect of their use, e. g. need of completion or completeness. To speak of their meaning does not mean to speak of an additional region, of structured contents that are transported in the signs, but rather to speak of the signs themselves, not as material objects made of a printer’s ink, but under the aspect of their usage as signs. Hence, logic does not show the structure of the thinkable in general, not universal knowledge forms beyond all impurities by the material properties of each individual signifier. A functional expression in Frege’s Begriffsschrift is not a functional expression by virtue of the fact that it correctly reproduces an object (‘the function’) that is independent of it (thus rendering it transportable), but because we use this expression in a certain way as a sign. The Begriffsschrift is not a transparent medium for something outside of itself, but a life element for certain communicative actions. Our demands on these special actions determine which properties the Begriffsschrift has. Its value is in no way tampered with or reduced by this; it is only a glorifying image, (to use Wittgenstein’s terms) a “sublimation”15 that has been taken back. So, ironically, precisely in Frege’s attempt to construct a transparent medium, it has come to light that the metaphor of transparency and cloudiness is just as misleading as that of the medium. The Begriffsschrift is one language among others, not the pure expression of the forms of all knowledge contents that can be grasped at all lingually and that supposedly exist predetermined and independent of their being grasped, and be it only in their formal aspects.

The Side by Side of Language Forms The pragmatic approaches in Frege’s later work just mentioned were formulated by me strongly in the spirit of later Wittgenstein. His concept of the language game, which accentuates the close interwovenness of lingual and non-lingual action, his emphasis on the social, public character 15 Wittgenstein (1953), §§ 38, 94.

Knowledge Forms and Language Forms

145

of language and its situatedness in life forms, is able to spell out the understanding of language that I had indicated above with the label of a ‘life element’. The guiding motive for this turn to Wittgenstein is my interest in finding out what can be gained for our understanding of communicative actions when we consistently avoid the image of a pre-given content that is only transported by a middle-thing, a medium, and seek the significance of a speech act in the act itself and its non-verbal context. In the context of the question pursued here, that means that ‘forms’ can only be spoken of at the level of language (or another symbolic system), and not at the level of contents (of ‘knowledge’). In doing so, I would first like to emphasise one point here that is to be found in Wittgenstein’s later philosophy, and that is of great significance for the aspect to which I wish to call attention in using the term ‘life element’. It also deserves special attention in the context of the question of the forms of knowledge: When the language of logic is de-sublimated in the way suggested here, that also means that the so-called logical forms do not lie once and for all in the essence of things (nor in the fundamental structure of the human mind). They are rather forms of representation that have arisen historically. The great generality of some of these forms that shows itself e. g. in that we treat places and points in time as objects and try to articulate much of what we say in the object-concept form, – this apparent universality now can be seen as a result of historical developments that also might have happened differently.16 The universality of logic owes itself to the (more or less well-founded) preference of a form of representation, not to an objective factum, toward which the representation orients itself. Inconsistencies that appeared to Frege as a blurring in natural language that might be avoided with the help of a transparent medium can be understood from this view, following Wittgenstein, as the results of a process in which an old, already established representational form is employed on new terrain. When we say, for example ‘the speaker came with his wife and with the best intentions’, we are not dealing with an unfortunate ambiguity of the word ‘with’ that should in philosophical contexts be replaced by a plurality of logically correct, unambiguous means of expression, but with the normal case in the region of language ‘life’, in which old means of expression are constantly projected onto new usage cases. It belongs to the grammar of the word ‘with’, that it can be employed in the two regions suggested in the example (accompaniment, mental state). And in Wittgenstein’s divergent usage of the 16 Cf. Schneider (2001).

146

Hans Julius Schneider

word ‘grammar’17 it also belongs to grammar that, with reference to the meaning of the expression ‘with the best intentions’, one cannot ask whether what is named rode along with the referent in the same cabin. This is the way led by the renouncement of Frege’s Platonic world of meaning, which is to be faithfully reproduced, namely, the way to an appreciative reassessment of the projective character of language ability, and thus to an appreciative reassessment of creativity as one constitutive trait of the ‘life element’ language. If the ‘structures of the thinkable’ (the ‘knowledge forms’) are not present in a finished form, one must make use of old structures in the articulation of what is new, making a hitherto uncommon use of them. The most conspicuous and well-known of these projective processes are those in the region of metaphor. We understand with no difficulty what a computer virus is, and what it means that blueberries are red when they are still green. What we can learn from Wittgenstein is the fact that metaphors are unavoidable and occur even in the region of syntax.18 Thus we come to a correction of our image of natural language and of the relation of its forms to those of logic. Natural language is, in its multifariousness, a ‘life element’ in its own right, having its own forms that are not to be read as miserable attempts to reproduce the ‘real’ forms of which one could say that they were already present in a special sphere of pure intended meaning. The natural languages are no clouded media whose expressions only blurrily render what a Begriffsschrift clearly shows, and what the most competent of speakers already have in their minds before they say anything. Every expression is rather an attempt to execute a speech act in such a way that the communication partners can continue with their own actions meaningfully. An expression, contrary to what the word’s etymology might lead us to suppose, does not bring something internal outward, it does not translate or encode an internal sentence into an external one, but the speaker rather makes a move with it in a language game whose continuations are to be meaningful without being therefore predictable.

17 Cf. Schneider (1999). 18 Cf. Schneider (1993); Schneider (1992).

Knowledge Forms and Language Forms

147

Further Conclusions In conclusion I would like to ask what we gain if we avoid the image of the medium, of a knowledge having a form independent of all transport vehicles, and of communication as transportation, and appropriate instead the metaphor of the life element. And I wish further to demonstrate from whence the impression addressed at the beginning arises that there must be a ‘propositional content’ independent of the various utterances. It will thus become apparent in which contexts this mode of expression has its own limited propriety (and for what reasons). When a communicative action, be it at home in language or in some other symbolic system, is no longer thought as the transport of a content, then the success of such an action is no longer to be judged according to whether this content arrives at the hearer or viewer as it was meant. His business is hence not to decipher, not to receive and ‘decode’ a message (to translate it into a private ‘language of the mind’). The activity of the hearer or the viewer will rather consist in connecting or continuing actions that can be unpredictable, extraordinarily multifarious and quite different from each other. If we look back to the case of artistic activity, we recognise that through the change of perspective suggested here we can understand the deep sense of uneasiness one feels with the old teacher’s question what an artist ‘wanted to say’ with his creation. If the fact that an artwork is about something is no longer expressed by the false supposition that there were something innerly meant, a content that pre-existed the product, one that one artist would prefer to ‘express’ as a poem, but another artist as a dramatic work (and one that the pupil or critic attempts to express in everyday or art-historically erudite prose),19 then the question in this form is digressive, even if the fact that works of art necessarily have a semantic dimension is not denied. This can be seen both in art history and in art criticism. To cite an example from Arthur Danto: Roy Lichtenstein’s pictures from the series Brushstrokes are ‘acts of continuation’ to the pictures of abstract expressionism, and Danto’s indication of this is a continuative act to Lichtenstein’s activity of producing the pictures.20 Understanding is the 19 This quite common supposition comes to expression e. g. in the following statement about the artist Jenny Holzer: [As a young art student she soon sensed that she would never be able to express what she wanted to say in pictures.]”Als junge Kunststudentin spürte sie bald, dass sie in Bildern nie würde ausdrücken können, was sie eigentlich sagen wollte.” Cf. Kuhn (2002). 20 Danto (1981), fn. 3, 107 ff.

148

Hans Julius Schneider

recognition of this nexus, not the correct grasp of something innerly ‘meant’. One further area in which the metaphor of the life element often appears to me to be more helpful than that of the medium is a certain sector on that areal of language on which we speak of our ‘inner state’. Over fifty years ago Gilbert Ryle expressed this with a poorly chosen term, and one that he wrongly placed near the natural sciences, the term ‘disposition’.21 Whoever speaks of his moods, anxiety, wishes or expectations gives no description of inner states or events, as in the case of a report of a hole in a tooth or an idea that occurred to one. He makes moves in the life element of language for which his communication partner has connecting and continuing moves, just as there are many actions with which one can meaningfully answer to a disturbed facial expression or a dejected posture. A partial aspect of this was developed further in the philosophy after Ryle by speech act theory.22 The expression ‘excuse me’ has as a possible answer, for instance, ‘no problem’; something additionally ‘referred to’, something which the expression would in this sense ‘mean’ (e. g. a painful feeling of regret that the speaker would bring to expression with a ‘request for forgiveness’), is superfluous. The meaning of the expression shows itself rather in its role in the appropriate social context, in the functioning of the language game, of which it is a part. The point of Ryle’s considerations is the treatment of the fact that such a performative understanding can also be suitable for statements that read much as reports according to their superficial grammar. What now can be said of the intuition addressed at the start, that there are certainly cases in which it is meaningful to speak of one propositional content that is expressed in different forms? My thesis is: a precise and non-psychologistic determination of the concept of a ‘propositional content’ can be given where in a comparison of articulations of knowledge in different forms a relation of equivalence can be determined on whose basis one can take a step of abstraction. We call the abstract object thus gained ‘propositional content’. There is no sufficient opportunity in this context to go into the details of the theory of abstraction relevant here.23 The direction of the considerations made there, however, can be rendered 21 Ryle (1969), ch. 5. The disposition of sugar to dissolve in water is made understandable by the molecular structure, while it is characteristic for the language games relevant here not to need any such substructure. Cf. Schneider (2005). 22 Austin (1972); Searle (1997). 23 Cf. Schneider (1992); Schneider (1970).

Knowledge Forms and Language Forms

149

intelligible if we return to the example of the machine cited at the beginning. We can determine the equivalence relation relevant to our purposes with the specification that a lingual description of a machine and a corresponding technical drawing shall be considered ‘pragmatically equivalent’ if both are suited to empower their reader, or viewer, to certain actions, e. g. if it should fail to function, to disassemble the apparatus in question and, by installing an available replacement part, to make it functional again. The utility thus defined, namely the enablement of an addressee to clearly definable actions, illustrates the way in which one might speak of an equivalence, an equality in value of lingual description and technical drawing. Only this equivalence relation, according to the conception presented here, makes possible the step in abstraction to the so-called ‘propositional content’ of the description and the technical drawing, and thus makes possible the manner of speaking of a knowledge that can remain identical in this or that articulation or rendering. We could then say the same knowledge ‘shows itself’ in practical action, in the fabrication of a drawing or of a lingual description. It is claimed here that we thus express nothing but the equivalence, that we take no step toward an additional region of psychic or Platonic objects ‘behind’ the actions named. And in this example we also see that the assertion ‘knowledge itself’ had a form (independent of any particular ‘life element’), makes no sense. This consideration makes it clear that the concept of a content gained in this way is dependent on the determination of an equivalence relation. Hence, that to which this term ‘content’ refers, is not something ‘meant’, it is no inner object floating about at the beginning of the communicative action, waiting to be ‘expressed’, but our talk about a content that is independent of its expression signals the success of concrete verbal actions and of practical activities of the types mentioned, like repairing an apparatus. So the use of symbols in a variety of kinds and forms is already presupposed when the topic of determining an equivalence between them comes up, which then makes possible the abstracting talk of the content. Many meaningful statements must already be present between which an equivalence relation is defined in what is logically a second step in such a way that on their basis an abstraction can be made. Here, too, we have no pre-existing meaning before us, that is ‘expressed’ through the use of different media, but many statements (signs) socially anchored as actions are treated as equals under a given aspect. Thus, it appears as though there were something common that stood behind them. The talk of a propositional content to be gained in this manner always means, with reference to the significant entirety, a restriction, which is

150

Hans Julius Schneider

harmless, even useful, in the technical area, because this area is defined through the possibility of functional equivalence. A functionally equivalent mechanism (a differently coloured padlock, a differently shaped brake grip) is technically ‘the same thing’ as the mechanism it replaces. But one must not make the mistake of thinking all areas of life could be treated technically. For many statements in the region of everyday communication, of intercultural understanding, of the humanities and art, one must say that they must be understood in their concrete, situated appearance, and this means that no corresponding equivalence relations can be determined, from whence it follows that the step of abstraction to a common content is impossible. One familiar consequence of this fact is the ‘inexhaustibility’ of the artwork so thoroughly explicated in the literature. My reflections mean for everyday language communication that clarifying dialogues are not to be interpreted as attempts finally to express something already correctly meant from the beginning (a formed inner knowledge) accurately, but as a chain of continuation actions that often lead to something new for all partners (even for the speaker). Accordingly, intercultural research or research on intellectual history is not to be conceived as a principally closable attempt to grasp an objectively present meaning correctly, but the continued effort toward each concrete ‘being understood’ between communication partners that can, however, start to crumble again, requiring new continuation acts, be it because a new party to the dialogue arrives, be it because new experience is had. That these continuation actions can be increasingly necessary for survival in our world today, in spite of their incapacity for closure, has need of no special emphasis, in light of the appearance of new fundamentalisms. The incapacity for closure, for completion, can therefore be no reason to throw in the towel. That understanding of ‘knowledge forms’ which imagines an equality of content among changing forms, in light of the view that has been developed here, appears to be a special case that presupposes that, for the signs regarded, equivalence relations can be so defined that the forms (language, technical drawing) can be disregarded. But this special character, systematically speaking, does not allow us to draw the conclusion that cases of this sort must be rare. On the contrary, they are quite common, even in the area of natural language. It is e. g. undeniably meaningful to speak of the one propositional content of two sets of instructions that are formulated in different languages. It appears important to me, though, that one recognise that this fact owes itself to the technical context, which per definitionem allows for the formulation of an aspect that makes possible the equivalence of the texts involved. In other contexts than technical ones (in a broad

Knowledge Forms and Language Forms

151

sense) it seems to be difficult to determine corresponding equivalence relations, and in broad areas of art this is considered out of the question with good reason. The existence and the broad distribution of technical contexts therefore cannot justify the opinion that each of the individual languages cited as examples were different media for expressing a pre-existing identical meaning. If one speaks of knowledge forms in the sense of types of semantic relevance, on the other hand, then the view of language as a life element of communication fortify and deepen the understanding of this sort of multiplicity. It is not the case that language always has the function of transporting propositional contents. There are grave differences between the types of moves in language games, and these cannot be reduced to differences in the so-called contents. I would therefore like to close with a piece of advice Wittgenstein gives. He says we should [break radically with the idea that language always functions in one way, always serving the same end: to transfer thoughts – be they thoughts about houses, pain, good and evil, or whatever] “…radikal mit der Idee brechen, die Sprache funktioniere immer auf eine Weise, diene immer dem gleichen Zweck: Gedanken zu übertragen – seien diese nun Gedanken über Häuser, Schmerzen, Gut und Böse, oder was immer.”24, 25

References Angelelli (1964): Ignacio Angelelli (ed.), G. Frege: Begriffsschrift und andere Aufstze, Hildesheim. Austin (1972): John L. Austin, Zur Theorie der Sprechakte/How to do things with Words, Stuttgart. Frege (1964a): Gottlob Frege, “Begriffsschrift. Eine der arithmetischen nachgebildete Formelsprache des reinen Denkens”, in: Angelelli (1964), VII – 88. Frege (1964b): Gottlob Frege, “Über die wissenschaftliche Berechtigung einer Begriffsschrift”, in: Angelelli (1964), 106 – 114. Frege (1967): Gottlob Frege, “Der Gedanke”, in: G. Frege, Kleine Schriften, Ignacio Angelelli (ed.), Darmstadt, 342 – 362. Frege (1969): Gottlob Frege, Nachgelassene Schriften, H. Hermes/F. Kambartel/ F. Kaulbach (eds.), Hamburg. Danto (1981): Arthur Danto, The Transfiguration of the Commonplace. A Philosophy of Art, Cambridge/MA. Kuhn (2002): Nicola Kuhn, “Die Lichtzeichnerin”, in: Der Tagesspiegel, 3 Feb. 2002, 25. 24 Wittgenstein (1953) § 304. 25 Translation from German by Alan Duncan.

152

Hans Julius Schneider

Ryle (1969): Gilbert Ryle, Der Begriff des Geistes, Stuttgart. Searle (1997): John R. Searle, Sprechakte. Ein sprachphilosophischer Versuch, Frankfurt/Main. Schildknecht (2002): Christiane Schildknecht, Sense and Self. Perspectives on Nonpropositionality, Paderborn. Schneider (1970): Hans J. Schneider, Historische und systematische Untersuchungen zur Abstraktion, PHD thesis, Erlangen. Schneider (1992a): Hans J. Schneider, Phantasie und Kalkl, Frankfurt/Main. Schneider (1992b): Hans J. Schneider, “Wörter und Handlungen als abstrakte Gegenstände”, in: Deutsche Zeitschrift fr Philosophie 40, 1141 – 1154. Schneider (1993): Hans J. Schneider, “’Syntaktische Metaphern’ und ihre begrenzende Rolle für eine systematische Bedeutungstheorie”, in: Deutsche Zeitschrift fr Philosophie 41, 477 – 486. Schneider (1995): Hans J. Schneider, “Begriffe als Gegenstände der Rede”, in: I. Max/W. Stelzner (eds.), Logik und Mathematik Frege-Kolloquium Jena 1993, Berlin, 165 – 179. Schneider (1999): Hans J. Schneider, “Wittgenstein und die Grammatik”, in: H. J. Schneider/M. Kroß (eds.), Mit Sprache spielen. Die Ordnungen und das Offene nach Wittgenstein, Berlin, 11 – 29. Schneider (200 l): Hans J. Schneider, “Universale Sprachformen? Zu Robert Brandoms ,expressiver Deduktion’ der Gegenstand-Begriff Struktur”, in: L. Wingert/K. Günther (eds.), Die ffentlichkeit der Vernunft und die Vernunft der ffentlichkeit. Festschrift fr Jrgen Habermas, Frankfurt/Main, 151 – 191. Schneider (2002a): Hans J. Schneider, “Fortsetzung statt Übersetzung! Das Problem des Kulturverstehens aus der Sicht einer pragmatischen Bedeutungstheorie”, in: J. Renn/J. Straub/S. Shimada (eds.), bersetzung als Medium des Kulturverstehens und der sozialen Integration, Frankfurt/Main, 39 – 61. Schneider (2002b): Hans J. Schneider, “Beruht das Sprechenkönnen auf einem Sprachwissen?”, in: S. Krämer/E. König (eds.), Gibt es eine Sprache hinter dem Sprechen?, Frankfurt/Main, 129 – 150. Schneider (2003): Hans J. Schneider, “Konstitutive Regeln und Normativität”, in: Deutsche Zeitschrift fr Philosophie 51, 81 – 97. Schneider (2004): H. J. Schneider, “Die Sprache: Trübes Medium oder Lebenselement der Kommunikation?”, in: M. Kubaczek/W. Pircher/E. Waniek (eds.), Kunst. Zeichen. Technik. Philosophie am Grund der Medien, Münster, 241 – 270. Schneider (2005): Hans J. Schneider, “Reden über Inneres. Ein Blick mit Ludwig Wittgenstein auf Gerhard Roth”, in: Deutsche Zeitschrift fr Philosophie 53, 743 – 759. Schneider (2007): Hans J. Schneider, “Reden über Inneres. Ein Blick mit Ludwig Wittgenstein auf Gerhard Roth”, in: H.-P. Krüger (ed.), Hirn als Subjekt? Philosophische Grenzfragen der Neurobiologie, Berlin 2007, 223 – 239. (Deutsche Zeitschrift für Philosophie, Sonderband 15). Schneider (2006): Hans J. Schneider, “Satz – Bild – Wirklichkeit. Vom Notationssystem zur Autonomie der Grammatik im ‘Big Typescript’”, in: Wittgenstein Studien 12, 79 – 98.

Knowledge Forms and Language Forms

153

Schneider (2008): Hans J. Schneider, Religion, Berlin. Schneider (2009): Hans J. Schneider, “Transposition – Übersetzung – Übertragung. Das Bild vom Transport semantischer Gehalte und das Problem der interkulturellen Kommunikation”, in: E. Birk/J. G. Schneider (eds.), Philosophie der Schrift, Tübingen, 145 – 159. Taylor (1981): Charles Taylor, “Understanding and Explanation in the Geisteswissenschaften”, in: S. H. Holtzman/C. M. Leich (eds.), Wittgenstein: To Follow a Rule, London, 191 – 210. Wittgenstein(1953): Ludwig Wittgenstein, Philosophische Untersuchungen/Philosophical Investigations, New York. Wittgenstein (1979): Ludwig Wittgenstein, ber Gewißheit, Frankfurt/Main.

Epistemic Perspectivity Martina Plümacher 1. The Perspectivity of Knowledge Is all knowledge perspectival? – ‘No’ is the answer of those epistemologists who insist, with a view to the sciences such as physics, on the existence of objective knowledge, i. e. knowledge that cannot be attributed to any subjective or cultural standpoint.1 In The View from Nowhere Thomas Nagel described objective scientific knowledge as a-perspectival knowledge in this sense. The exclusion of everything personal and subjective in the course of the objectification of knowledge leads, for him, to the “view from nowhere” – to a knowledge that is no longer bound to a particular standpoint. This involves a concept of epistemic perspectivity that takes ‘perspectival’ to mean the coloring of epistemic judgment by subjective or cultural factors. Hence perspective knowledge is on this view knowledge that has not yet been ‘purified’ of the subjective and particular, and thus stands in need of correction and is not really knowledge in the strict, proper sense. On this understanding perspectivity is also associated with a “division of standpoints”2 – with the divergence of positions rather than their convergence. Such an understanding of epistemic perspectivity is unsatisfying. It is unsatisfying, firstly, insofar as it is oriented around an idea of unrevisable and ultimate knowledge, hence knowledge that presupposes a divine standpoint.3 Yet knowledge that is absolute in this sense is beyond our human reach. Secondly, this understanding is also unsatisfying in 1 2 3

Cf. Daston (1992); Post (2010), 65 – 68. For a critical analysis of the concepts of the objective and subjective used in such contexts see Putnam (1987) and (2003). This position comes out in the title of a collected volume on perspectives in perception and cognition; cf. Koch (2010): Perspektive – Die Spaltung der Standpunkte. For a critique of the understanding of knowledge oriented around the idea of a divine perspective see Putnam (1981), 48 ff.

156

Martina Plümacher

that it only comprehends a very special use of the word ‘perspective’ in relation to knowledge, one that is neither the typical ordinary use of the word nor the most epistemologically interesting use. When the term ‘perspective’, originally stemming from optics and the representation of space, was applied to epistemic judgments at the turn of the 18th century,4 it was used primarily to describe the coloring of epistemic judgments by interests or subjective or culturally specific factors, e. g. partisan descriptions of historical events or culturally specific value judgments. Yet by the middle of the 18th century the term ‘perspective’ was already being used more generally to distinguish various ‘points of view’ (and hence different epistemic interests) concerning the objects of knowledge. This use of the word is still highly relevant today and forms the semantic core of the term ‘epistemic perspective’.5 For example, to classify a variety of positions in a debate we begin by distinguishing various topics and points of view. The word ‘perspective’ occurs just as prominently in these contexts as it does generally in contexts where it might make sense to highlight different views of the objects. One and the same object, such as a building or a city district, can be considered from different points of view, e. g. functionally, economically, ecologically, socially, aesthetically, artistically, or in terms of structural engineering or architectural history or many more. We are interested in each case in a different aspect of the object. Our relation to the object is perspectival in this sense and our perception is aligned accordingly: different things catch our eye depending upon our perspective. Specific perspectives are professionalized within the various disciplines and professions (e. g. architecture or city planning). The objects of their interest are not exactly identical. The architect deals with the 4 5

For the history of the concept of ‘perspective’ concerning knowledge and cognition see Plümacher (2010a). One of the earliest indications of this understanding of ‘perspective’ is found in 1746 in Johann Heinrich Lambert (1990), 658: “In Ansehung des Gedankenreiches sind wir schon längst daran gewöhnt, die Begriffe der Seiten und der Gesichtspunkte dabei zu gebrauchen, so wie wir etwa dem Verstande Augen geben, und den Begriff des Sehens auch auf abstrakte Dinge ausdehnen. Die Seiten sind nun hier die Verhältnisse, in welchen eine vorgegebene Sache mit andern steht. […] In so vielerlei Absichten eine Sache betrachtet werden kann, so viele einzelne Gesichtspunkte hat sie auch.” (Concerning the realm of thought we are long since used to using the concepts of sides and viewpoints, as we give the understanding eyes, and extend the concept of seeing to abstract things. The sides are here the relations in which a given matter stands to others. […] A matter has as many individual aspects as it has regards in which it can be considered.)

Epistemic Perspectivity

157

construction of buildings and the configuration of the rooms, including a consideration of the building’s vicinity. The object of city planning is whole cities or city districts and settlement structures; here individual buildings and rooms are only elements of the whole that can come into individual focus under certain aspects, e. g. in their function as symbols and emblems or as elements that structure life in the public space. City planning deals primarily with the structure of public spaces, and less so or hardly at all with the rooms inside the buildings, which the architect and interior designer are responsible for configuring. The relation of internal and external spaces forms a fascinating transdisciplinary interface between these disciplines. Epistemic perspectives in this sense are not distinguished with the critical intention of uncovering the subjective and particular aspects of knowledge in order to eliminate them. The point of distinguishing them is the systematic and detailed acquisition of knowledge. This is also a result of the circumstance that we humans lack the possibility of simultaneously comprehending, describing, and accounting for all objects and their relations in all dimensions and relative to all purposes at once. We can only always investigate and consider specific aspects of objects in isolation from others. Hence cognition is ineluctably focused. The general concept of epistemic perspectivity articulates and describes this focusing of cognition. It is inherent in the focusing of our epistemic activity that it brings with it a restriction to a particular circle of epistemic objects. A deeper understanding of epistemic perspectivity is epistemologically interesting and valuable in that it sheds light both on our focusing of epistemic activity as well as on the specific restrictions of epistemic projects and, e. g., the limits of a project’s particular capacity for explanation and problem-solving. This understanding of epistemic perspectivity accentuates the perspectivity associated with the division of labor in our acquisition of knowledge, rather than the ‘distortions’ of prejudice, misunderstanding, or erroneous basic assumptions that stand in need of correction. Hence the concept of epistemic perspectivity used here is quite compatible with the talk of true and objective knowledge. Epistemic perspectivity does not lead to relativism. And yet the realization of the fundamental and ineluctable perspectivity of knowledge can give rise to other disconcerting questions: Are we condemned to perspectivism? Is our knowledge of reality just a colorful potpourri of individual items of perspectival knowledge? In what

158

Martina Plümacher

follows I will show that this is not the case, and that focused and limited cognition occurs against a background of orders within knowledge and is part of the ‘work on knowledge’, i. e. an acquisition of knowledge spanning generations and including several restructurings of knowledge. However, I will begin by more precisely explicating this understanding of epistemic perspectivity. For this purpose I distinguish between a broad and a narrower sense of ‘epistemic perspectivity’. The reflection on epistemic perspectivity can occur with regard to the object of knowledge analytically or with an epistemologically ordering and orienting function. I will show that epistemic processes are linked to an interplay of epistemic perspectives which is vitally important if the result earns to be qualified as ‘well thought-through.’

2. The Broader and Narrower Sense of ‘Epistemic Perspectivity’ We can bring out the differences in people’s epistemic perspectives by distinguishing between a ‘broad’ sense and a ‘narrow’ sense of epistemic perspectivity. The ‘broad sense’ of the concept of epistemic perspectivity is that of the ‘directed interest’ in which certain aspects of the particular object come into the focus of our attention while other aspects remain suppressed. On a general or ‘broad’ understanding of epistemic perspectivity, it is at first only relevant that our epistemic interest is directed at specific objects and aspects of the object, that the perception is focused and the perceived is understood in a specific way. However, a more specific reflection on epistemic perspectivity would also consider where the specific epistemic interest comes from and which theoretical positions go along with it, and the concept of epistemic perspectivity will then need to be understood more narrowly to account for these differentiated considerations of epistemic perspectives. The broader concept of perspective is well-suited to mark differences in the perception and comprehension of objects. Our perceptions (just like our thoughts and actions) are always bound to situative sensibilities and general interests. Not everything we perceive seems equally relevant. Thus people regularly reassure themselves that the focus of their perception is roughly the same: “Did you see that, did you hear that?” They also discuss whether they have taken in the perceived ob-

Epistemic Perspectivity

159

ject in the same way: “Did you notice that the woman wasn’t feeling well?” or “Do you also think that noise could’ve been a pistol?” This broad understanding of ‘perspectivity’ captures two basic dimensions of epistemic conduct, the focusing of attention and the particular way that the object of our focus is taken in and interpreted. We can note differences in epistemic conduct with regard to these two dimensions. Generally people register rather precisely how other people’s attention is oriented, even if they do not also reflect on whether it differs from their own epistemic conduct. It is a peculiarity of epistemic perspectives that they are often not directly thematized. In their communicative action people often respond without an intellectual analysis of their epistemic perspectives, such as in trying to urge their own perspective upon the other person: “Do you also think that noise could’ve been a pistol? Could it be?” – “No, that sounds more like the machine the neighbors bought last week.” Adult persons typically have the ability to recognize and distinguish different epistemic perspectives in discourse interactions. We daily draw distinctions between standpoints or positions without reflecting upon this practice itself. In this sense our practice is pre-theoretical and our making differences between perspectives is a part of ‘knowing how to do something’, e. g. how to discuss matters and intellectually take a position – such as in an argument about store hours from the point of view of the customer or the salesperson. Adult persons understand how to explicate standpoints and epistemic perspectives in conversation, to sort out or connect different positions, to bring perspectives into connection with social roles, to prescribe viewpoints and understandings in setting an agenda, and to change points of view and thus to change perspectives in a discussion and to signal this change to other people.6 Our everyday communication involves additional dimensions besides just what objects and aspects we perceive and are interested in. We connect perspectivity to intentions or to a particular goal or interest; 6

The linguistic phrases and rhetorical figures used here have been thoroughly documented and analyzed, cf. Graumann/Kallmeyer (2002); Kallmeyer (1996); Canisius (1987); Sanders/Spooren (1997); Köller (2004). On the role of language in the development of the human ability to perspectivize see Tomasello (1999); Plümacher (2010b). For Tomasello human ‘social cognition’ is distinguished by our understanding of perspectives. In his 1934 Mind, Self, and Society George H. Mead had already very clearly described how human action is bound to the formation of perspectives and moreover requires assuming the perspective of other persons.

160

Martina Plümacher

we make attitudes, presuppositions, or prejudices explicit as the background of epistemic perspectives.7 Scientific discourse in particular calls for a more precise reflection on the conditions of knowledge and of the specific perspective. It is primarily this context that invites us to conceive epistemic perspectivity more narrowly for the sake of a more fine-grained definition of the various dimensions that constitute the particular ‘epistemic object’. A ‘narrower’ sense of ‘perspectivity’ is exceedingly important where we need to structure different aspects in our acquisition of knowledge, e. g. in order to systematize epistemic processes and organize them more effectively, such as in research groups that wish to specifically align and structure their interdisciplinary and transdisciplinary research. In this case we are interested not just in the fact that different objects and aspects are relevant to the participants but also for what reasons this is the case. Hence we are interested in theoretical and methodical settings, the profile and objective of research projects, specific questions and problems, or hypotheses and methodological specifications as the background for the selective concentration of epistemic interest on certain classes of object and aspect and the way in which the specific epistemic objects present themselves. In these contexts the systematization of the epistemic processes invests the concept of perspective with the function of a foundational concept in the theory of science. As such the concept is specifically aligned to a comparative analysis and discussion of epistemic projects. It is complex insofar as it is directed towards the interplay of various dimensions of practical and theoretical engagement with objects through which the specific ‘epistemic objects’ first crystallize. These dimensions are, briefly: 1. the epistemic interest, more precisely the epistemic goal and/or objective of action; 2. the focusing on certain objects and aspects; 3. problems and questions; 4. hypotheses, theories and models; 5. presuppositions and premises; 6. methods and procedures. These dimensions are closely related to one another: the epistemic goals or the goals of action are a source of questions, problems, and hypoth7

Perspectivity is in need of correction when prejudices and false assumptions are at work.

Epistemic Perspectivity

161

eses. They are the driving force behind changes in theories and the formation of models. Presuppositions and premises always play an important role and influence ‘standpoints’. Questions and problems are particularly significant for the formation of perspectives, since they guide the focusing on specific aspects of objects and make it concrete.8 The search for answers to questions and solutions to problems also influences the choice of methods and procedures. At the same time the methods and procedures constrain what can actually be cognized or practically produced.9 The complex concept of perspective is well-suited to describe and comprehend the specific interplay of these dimensions, in order to then be able to clearly distinguish epistemic projects from one another in their profile, to coordinate them or align them with each other or mark contrasts and counterpoints. A fine-grained reflection on the specific epistemic perspectives is relevant even for projects of a practical nature where specialized activities have to be adjusted to one another and coordinated; for example, we might think here of the cooperation between various types of profession and labor involved in constructing a building or in creating a business profile that stands out in the competitive landscape. The critical aspect of perspectival thought is of great importance. When we ask the question what does not belong to a certain consideration or is of lesser significance or not methodically ascertainable or feasible, this question brings into focus the limits of what can be known and clarified. The reflection on this constraint leads us to think about how epistemic projects can be expanded, supplemented or selectively contrasted with other projects. This also highlights the limits of professional responsibility. For example, an architectural blueprint for a house is not an instrument for clarifying the light conditions in rooms; but a computer-supported 3-D room simulation can do this. Architects use the 8

9

Ancient rhetoric had already emphasized the significance of questions for the orientation of thought. On the role of questions for epistemic activity see among others Dewey (1986); Rescher (1984), ch. II; Hookway (1996); (2006); (2008). Rheinberger (1992); (1997) has shown us how the models of experimental practice effect the selection of epistemic objects and the focusing on particular aspects. On the perspectivity of ‘epistemic objects’ see Abel (2010). On the perspectivity of scientific observation and scientific theorizing see Giere (2006); van Fraassen (2008). On the perspectivity of models see: Mahr (2008) and in this volume.

162

Martina Plümacher

3-D simulation created by software specialists who themselves do not need to be architects.

3. Distinguishing Points of View Knowledge develops in different directions and “according to different [formative] principles”10 – in the sciences as in the arts and the professions. We also call disciplinary epistemic interests, questions, and viewpoints ‘perspectives’ and distinguish them, e. g. the sociological perspective from the philosophical perspective. The concept of perspective used here differs from that described above in its focus and function. The above-mentioned understanding of perspectivity determines the epistemic object and the conditions of its constitution more specifically and thus serves an analytical function concerning the objects of knowledge. However, when we understand the different orientations of epistemic interest in the sciences, arts, and professions perspectivally, the concept of perspective takes on an ordering function. This latter concept is used above all to orient ourselves and create organization when it comes to various points of view in our engagement with objects and the division of labor in the professional acquisition of knowledge. It is important for any systematic and detailed activity of knowledge to maintain the distinction between various points of view or aspects of objects and consider them separately. For example, the architectural planning of a spatial structure has to be separated from the exact calculation of the static equilibrium of the building or the precise tabulation of the costs. Often various points of view have to be held apart and developed separately before we go over to setting them into relation to one another – for example, in the concrete question of whether the planned ecological and structural sustainability can be realized within the financial budget by dispensing with a few aesthetically interesting details of the construction and yet without risking the aesthetic concept as a whole. 10 Cassirer (1953), 107; Cassirer (2001a), 41: “nach verschiedenen Bildungsprinzipien”. In his Philosophy of Symbolic Forms Cassirer distinguishes basic directions of human engagement with reality: science, art, technology, language, religion, and myth. He calls these epistemic activities ‘symbolic forms’. However, when he first published his idea of ‘symbolic forms’ (Cassirer (2001b), ch. VII) he was focused primarily on different scientific disciplines.

Epistemic Perspectivity

163

In their socialization processes and especially in scientific socialization people acquire the ability of ordered perspectival cognition. Let us imagine for a moment someone who is unable to distinguish various aspects in dealing with an object and continually changes her or his points of view. The person might point out a few interesting views and raise good questions but would hardly be capable of systematic analysis, since the purpose of analysis requires that we first begin with one viewpoint and continue with it, i. e. keep it separate from other, perhaps equally interesting viewpoints. This epistemic and operative practice leads to the formation of orders in our engagement with objects and always makes use of previously acquired orders. For example, children learn to sort building blocks according to size, form, color and other aspects; and a student of philosophy will learn to distinguish sociological questions from epistemological questions and to reflect on the relation between the sociological and epistemological aspects of e. g. knowledge. The formation of the epistemic ability to distinguish points of view, aspects of objects, and epistemic perspectives also includes an understanding of professional specializations and the corresponding realms of responsibility. This understanding is demanded by many everyday situations. We can think of parents faced with some conflict involving their child at school, who need to ask whether the conflict has a legal dimension and a lawyer needs to be involved or whether it would be good to call on a conflict mediator or a child psychologist. This sort of practical knowledge of who is responsible for which questions and how we can sort questions and problems such that they can be answered and solved is both the expression and – in execution – the reinforcement of an epistemic order. This order is characterized by both practical distinctions between various points of view and a division of labor in the professional acquisition of knowledge. When discourses call on us to take on a certain perspective (e. g. to consider the social aspect of something), an ordering point of view is brought to bear. People are to deal with the matter along the lines marked out and apart from other possible modes of consideration. Then different perspectives in the broad and narrow sense of the concept described above can be articulated within these lines of consideration. Insofar as it seems important, these perspectives are then more closely analyzed with a view to how the matter in question presents itself within them and for what reasons. Epistemic perspectivity is a manifestation of goal-directed, selective epistemic activity. We can reflect upon the latter from two different di-

164

Martina Plümacher

rections. On the one hand we can look at which objects and aspects are in the specific focus of interest; in this case we are reflecting on the particular epistemic objects and, if need be, more precisely on their constitutive presuppositions and conditions. On the other hand we can look at which points of view or aspects, which disciplinary or professional epistemic modes of access are to be distinguished in order to better structure epistemic processes. In this case we make use of epistemic orders or explicitly reflect on them in order to describe them, question them and modify them or even to form whole new orders of knowledge. New orders of knowledge emerge, for example, following the establishment of new disciplines or with a new distribution of responsibilities within the acquisition of knowledge. For example we might think of computational neuro-science, which emerged as an independent field of work within the neuro-sciences, or of the distinction between business ethics and corporate ethics11 – a distinction drawn to hold apart two different fields of problem in order to treat them and analyze them separately, namely the conditions of economic activity to be shaped by economic policy and the corporate activity within certain conditions.

4. The Interplay of Perspectives in Epistemic Processes Assuming and maintaining a perspective is of cardinal importance in professionalized epistemic processes. However, different and alternative perspectives also need to be adopted, tested, and accounted for in order to optimize knowledge and situate one’s own perspective within discourses. For example, it might be necessary to classify one’s own perspective within disciplinary or professional contexts, to position oneself within research groups or in the cooperation with other people. A researcher formulating or testing her own hypothesis – hence someone who has taken a certain perspective on the field of objects – does so in indirect engagement with other theories and epistemic projects; she anticipates objections to her methods and takes them into account in developing her experimental design. And someone putting together a financial plan for a company takes on a certain perspective; she considers the company under financial aspects. In deciding how to distribute the available budget she has to account for various areas of activity 11 The business ethicist Karl Homann argues for this distinction; see Homann/ Lütge (2004).

Epistemic Perspectivity

165

within the company and has to run through various perspectives – the internal perspectives of the employees, the external perspectives of the customers or assessors, and not least of all she has to weight various factors in creating the company’s competitive profile. The unfolding of perspectival knowledge – both in the epistemic process itself as well in its presentation and justification relative to other people – is far from one-dimensional. Rather it is quite intimately connected to an interplay of points of view. This interplay is crucial to achieving results that can be considered ‘well thought-through’ and earn general recognition. For a specific task we can always specify which points of view at least have to be considered and which points of view may rightfully be neglected or even have to be, in order that the result of tackling that task can earn the epithet ‘well thoughtthrough’. Let us take the example of planning a building. There is a specific framework for this; the commissioning party typically determines the general function of the building along with certain financial constraints. Within these limits the architect is allowed to explore and develop her ideas. The building’s function will be at the center of her thoughts; this aspect is her central point of view in structuring the space. Which rooms to plan, for example, also depends on the building’s purpose. The more detailed planning will also require consideration of a whole series of further aspects besides functionality, and hence the architect will need to assume different perspectives and think them through in order for the plan to be considered ‘well thought-through’: e. g. the effects the rooms have on the residents’ well-being and attitude to life; how well the building suits its immediate environment; the professional rules of structural engineering and of sustainability in terms of the construction materials, and ecological criteria and standards. A methodological professionalism is expected in the planning and the realization of the building. If the building is to set standards itself and have a symbolic impact, then considerations of architectural theory take on greater significance; the building will have to be measured against the outstanding examples of architecture and be positioned and justified within debates about contemporary architecture. For every epistemic project there are some points of view that it is necessary to consider and others that can give a design additional merit. The collaboration with landscape architects or city planners could bring additional perspectives to the architectural planning and perfect it in these regards as well. There are many other possible viewpoints on

166

Martina Plümacher

the particular object that are not relevant in any way for the epistemic goal and thus have to remain suppressed for the sake of systematic epistemic activity. For example, while planning a building architects should not let themselves get distracted by questions such as how the building could be represented cubistically, how a certain steel manufacturing process could be optimized or how architectural knowledge could be introduced into the context of epistemological research. Questions like these ground different epistemic projects, and in pursuing them architects would be abandoning their project of constructing an actual building. This example shows how the ineluctable thematic delimitation of epistemic projects includes various perspectives that present points of contact with other projects, professions and disciplines. Within large projects that can only be realized with the collaboration of diverse experts, the reflection upon the specific perspectives of the professions and disciplines is of cardinal importance for communication and mutual coordination, where it needs to be specified in detail who develops which part of the whole project, for example, and how the designs and findings are to be merged and how the work steps are to be coordinated. There are many projects that cannot be undertaken from the perspective of one discipline alone, not just in professional practice but also in science and academia. Even just the claim of researchers to gain knowledge that could be related to developments in neighboring disciplines implies a view beyond the limits of their own discipline. When specifying their projects the researchers have to take these developments and the corresponding perspectives into account. As soon as various disciplines and projects are involved in an overarching question, e. g. in transdisciplinary projects, it is impossible to avoid engaging with different disciplinary approaches to the topic and with the specific contribution from every research project. In their process of collaboration the participants need to discuss how their various pieces of research are to be brought together. Hence the question of the relation between the various disciplinary viewpoints is important insofar as the objects of interest are not exactly the same. For example, in clarifying the question of the causes of the specifically human linguistic ability, behavioral genetics refers to the FOXP2 gene, whereas cognitive ethology focuses on the social life-form as the relevant factor in the formation of communicative structures, evolutionary theorists look at how the anatomy of the vocal tract and the brain developed and psycholinguists concentrate on

Epistemic Perspectivity

167

the neuronal processes that correlate with the syntactic processing of language. In each case the researchers’ interest centers on different biological and social processes. The diverse sorts of results brought to answer the question of the causes of human linguistic ability essentially re-pose the question. Now, to answer it, various sub-questions need to be clarified, e. g. the question about how biological and social processes effect each other in evolution, the question of the relative significance of the various causal factors, or the question of cause and effect concerning the various factors brought to bear on the question. Each sub-question opens a perspective on the diversity of positions, whereby the perspective at the same time takes on an integrative character. The integrative perspective is no syncretism12 and not a mere amalgamation of the different epistemic perspectives. It is itself a specific perspective that accentuates a certain point of view. Hence within interdisciplinary and transdisciplinary research the interplay of perspectives also includes integrative perspectives whenever the results of the interdisciplinary and transdisciplinary collaboration are to be determined and balanced.

5. Orders in Knowledge as Frameworks of Perspectival Cognition An all-inclusive perspective on the entirety of human knowledge and cognition is not possible for us as finite beings. We cannot achieve objectivity in knowledge either through cognizing a-perspectivally or by lining up all of our individual items of (perspectivally acquired) knowledge. Our understanding of reality unfolds and enriches itself in the diversity of our perspectives on reality.13 If we lacked integrative perspectives on perspectival knowledge, and if the individual items of knowledge could not be perspectivally thought-through and connected with each other, our knowledge of reality would break down into single isolated pieces. If we were brought up short by the result that a question can be answered differently from the viewpoints of various experts, and could go no further, it would result in utter perplexity. We build relationships among items of knowledge by ordering these items, estimating their significance for us as agents, identifying inconsistencies 12 Cf. Rescher (1991), ch. 8: “Philosophical Perspectivism”, 127 – 141. 13 See also Ernst Cassirer’s Philosophy of Symbolic Forms. Goodman (1978) speaks in a comparable sense of the variety of epistemic worlds.

168

Martina Plümacher

and creating new contexts and interconnections between the items of knowledge. In these processes a restructuring of knowledge comes about. The acquisition of knowledge organized by a perspectival division of labor results in a non-simultaneous development of human knowledge in the various fields. Not all possible contexts and interconnections actually construed, examined and explicated. The knowledge cultivated in the scientific disciplines or in professional contexts is not always and unconditionally subject to transdisciplianry or transprofessional reflection and coordination. Hence the knowledge at our disposal does not form a coherent and consistent body. The question of coherence and consistency only arises when interrelations have been set up and new perspectives have been adopted and developed. Otto Neurath, a leading thinker in the logical positivism of the ‘Vienna Circle’, once compared the division of labor in the acquisition of knowledge to the work on a mosaic, which proceeds at different points in the mosaic, with the work in some places progressing faster than in others. Different generations have worked on the mosaic; there is not yet any end in sight, since “the generations of the ‘mosaicists’ are not only inlaying the stones but also changing certain stones for others and varying the whole pattern.”14 To continue the metaphor, it is the orders in knowledge that constitute the ‘patterns’ by which the individual items of knowledge get established and put into a network. The shared ‘work on knowledge’ includes considerations of how new pieces of knowledge or even whole epistemic perspectives can be inserted into the existing and solidly established knowledge, and whether and how conceptual distinctions and thus orders in knowledge can or should be changed. We also have to continually estimate the significance of individual items of knowledge or of whole epistemic perspectives in relation to others in order to then be able to articulate and distinguish new epistemic perspectives and new epistemic projects that reveal new problems that bring greater precision and take us further. The ‘work on knowledge’ includes more than just our knowledge of properties and structures of objects and processes in the world – it also includes the reflection on what orders within knowledge make sense and orient us in action and in the further acquisition of knowledge. 14 Neurath (1938), 3.

Epistemic Perspectivity

169

Orders within knowledge result, as emphasized, from distinctions between epistemic perspectives. Encyclopedias of knowledge, textbooks, and other epistemic vehicles, for example works in the history of science or the history of ideas, represent these orders in knowledge. Insofar as we have orders in knowledge at our disposal, we do not have to worry too much about the non-simultaneous epistemic development in the various areas of knowledge. It is true that each new piece of knowledge raises the question of whether and how it can be inserted into the knowledge previously established and secured, and often knowledge has to be re-examined and reorganized in the wake of new findings. Yet it happens very seldom that a new piece of knowledge in one area raises questions in all areas of knowledge. The architecture of knowledge is not modified by every new finding, even if there are great shifts within the whole from time to time. For example, we might think of the reordering of knowledge concerning the mind-brain relation resulting from the empirical brain research starting in the 1980 s, which led us to see our previous knowledge in many areas differently, for example concerning perception, cognition, freedom of the will, and religiousness.15 Orders in knowledge make it easier for us to identify various aspects within complex problems, to consider them in isolation and in this way to bring clear structure to the problems. Orders in knowledge also help us to determine what knowledge and which perspectives need to be considered in dealing with a certain task. We call on these orders not just to move within the paths we have already set but also whenever we need to start new paths. For example, when the Berlin architect Francis Kéré faced the challenge of designing a school building in his extremely poor home country of Burkina Faso in West Africa, and had extremely limited financial, personal and technical resources at his disposal, he could not take the standard route. A typical building from concrete and glass was unfeasible. He recalled the traditional method of building from clay bricks and designed a modern building that was adapted to the climate and also aesthetically ambitious and could be realized with local workers without using a crane.16 Kéré thought about 15 On the challenges of philosophy in light of transformations within orders of knowledge see Abel (2009). 16 In 2004 the building in Gando in Burkina Faso was given the most prestigious architecture prize in the Islamic world, the Aga Khan Award. The airy roof construction is especially sophisticated, as it allows rising hot air to escape

170

Martina Plümacher

the practical knowledge in Burkina Faso on building with clay as well as the knowledge of various architectural categories like “mudbrick building method”, “climate-adapted design” and “sustainable construction”. In focusing on the specific task he created a new perspective on mudbrick construction and integrated various forms of knowledge. Typically we do not reflect on the fact that orders in knowledge guide us in our considerations. Likewise we seldom explicitly reflect on the fact that distinguishing between epistemic perspectives results in the formation or reformation of epistemic orders. And yet epistemic orders are at work in the background of all epistemic activity. It is with their support that we venture the greatest specializations in knowledge. When we thematically constrain ourselves in our epistemic projects, we trust in the fact that our perspectival knowledge can be sensible complemented by other perspectival knowledge. By and large this interplay of perspectives works amazingly well. It works because, and only so long as, the shared ‘work on knowledge’ is a work of producing connections, distinguishing, supplementing, and contrasting. This ‘work on knowledge’ comes down to more than just examining whether propositions are true. It is just as important to consider what is relevant to our considerations, whether all relevant aspects have been accounted for, and thus whether our solutions or answers can be considered ‘well thought-through’. Hence it is also important to mark off those questions and aspects that have to be suppressed, in order to indicate that they are to be dealt with elsewhere.17

References Abel (2009): Günter Abel, “Die Transformationen der Wissensordnungen und die Herausforderungen der Philosophie”, in: Allgemeine Zeitschrift fr Philosophie 34, 1, 5 – 28. Abel (2010): Günter Abel, “Epistemische Objekte als Zeichen- und Interpretationskonstrukte”, in: S. Tolksdorf/H. Tetens (eds.), In Sprachspiele verstrickt. Oder: Wie man der Fliege den Ausweg zeigt, Berlin/New York, 127 – 156. Canisius (1987): Peter Canisius (ed.), Perspektivitt in Sprache und Text, Bochum. Cassirer (1953): Ernst Cassirer, The Philosophy of Symbolic Forms. Vol. 1: Language, New Haven/London. and was built without the use of a crane. In June of 2010 Kéré was given the Swiss Architectural Award for this and other mudbrick buildings in Burkina Faso. 17 Translation from German by Karsten Schöllner.

Epistemic Perspectivity

171

Cassirer (2001a): Ernst Cassirer, Philosophie der symbolischen Formen. Erster Teil: Die Sprache (1923), in: B. Recki (ed.), Gesammelte Werke. Hamburger Ausgabe, vol. 11, Hamburg. Cassirer (2001b): Ernst Cassirer, Zur Einsteinschen Relativittstheorie. Erkenntnistheoretische Betrachtungen (1920), in: B. Recki (ed.), Gesammelte Werke. Hamburger Ausgabe, vol. 10, Hamburg. Daston (1992): Lorraine Daston, “Objectivity and the Escape from Perspective”, in: Social Studies of Science 22, 597 – 618. Dewey (1986): John Dewey, Logic: The Theory of Inquiry (1938), in: J. A. Boydston (ed.), John Dewey. The Later Works, 1925 – 1953, vol. 12, Carbondale. Giere (2006): Ronald N. Giere, Scientific Perspectivism, Chicago/London. Goodman (1978): Nelson Goodman, Ways of Worldmaking, Indianapolis/Cambridge. Graumann/Kallmeyer (2002): Carl F. Graumann/Werner Kallmeyer (eds.), Perspective and Perspectivation in Discourse, Amsterdam/Philadelphia. Homann/Lütge (2004): Karl Homann/Christoph Lütge, Einfhrung in die Wirtschaftsethik, Münster. Hookway (1996): Christopher Hookway, “Questions of Context”, in: Proceedings of the Aristotelian Society, vol. XCVI, 1 – 16 Hookway (2006): Christopher Hookway, “Epistemology and Inquiry: the Primacy of Practice”, in: S. Hetherington (ed.), Epistemology Futures, Oxford, 93 – 110. Hookway (2008): Christopher Hookway, “Questions, Epistemology, and Inquiries”, in: Grazer Philosophische Studien 77, 1 – 21. Kallmeyer (1996): Werner Kallmeyer (ed.), Gesprchsrhetorik. Rhetorische Verfahren im Gesprchsprozeß, Tübingen. Koch (2010): Gertrud Koch (ed.), Perspektive – Die Spaltung der Standpunkte. Zur Perspektive in Philosophie, Kunst und Recht, München. Köller (2004): Wilhelm Köller, Perspektivitt und Sprache. Zur Struktur von Objektivierungsformen in Bildern, im Denken und in der Sprache, Berlin/New York. Lambert (1990): Johann Heinrich Lambert, Neues Organon oder Gedanken ber die Erforschung und Bezeichnung des Wahren und dessen Unterscheidung vom Irrtum und Schein (1764), vol. 2, ed. by G. Schenk, Berlin. Mahr (2008): Bernd Mahr, “Ein Modell des Modellseins. Ein Beitrag zur Aufklärung des Modellbegriffs”, in: U. Dirks/E. Knobloch (eds.), Modelle, Frankfurt/Main, 187 – 218. Mead (1992): George Herbert Mead, Mind, Self, and Society: From the Standpoint of a Social Behaviorist (1934), ed. by C. Morris, Chicago/Illinois. Nagel (1998): Thomas Nagel, The View from Nowhere, Oxford. Neurath (1938): Otto Neurath, “Unified Science as Encyclopedic Integration”, in: International Encyclopedia of Unified Science 1, 1, 1 – 28. Plümacher (2010a): Martina Plümacher, “Perspektivität, epistemische”, in: H. J. Sandkühler (ed.), Enzyklopdie Philosophie, Hamburg, 1930 – 1937. Plümacher (2010b): Martina Plümacher, “Epistemische Perspektivierungen. Der Beitrag der Sprache”, in: C. Stroh (ed.): Von Katastrophen, Zeichen und vom Ursprung der menschlichen Sprache, Bochum, 69 – 92.

172

Martina Plümacher

Post (2010): Robert Post, “Perspektivismus und Recht”, in: G. Koch (ed.), Perspektive – Die Spaltung der Standpunkte. Zur Perspektive in Philosophie, Kunst und Recht, München, 65 – 77. Putnam (1981): Hilary Putnam, Reason, Truth and History. Cambridge. Putnam (1987): Hilary Putnam, The Many Faces of Realism, LaSalle/Illinois. Putnam (2003): Hilary Putnam, “Ethik: In den Strömungen Kurs halten”, in: M. Vogel/L. Wingert (eds.): Wissen zwischen Entdeckung und Konstruktion. Erkenntnistheoretische Kontroversen, Frankfurt/Main, 288 – 305. Rheinberger (1992): Hans-Jörg Rheinberger, Experiment – Differenz – Schrift. Zur Geschichte epistemischer Dinge, Marburg. Rheinberger (1997): Hans-Jörg Rheinberger, Toward a History of Epistemic Things. Synthesizing Proteins in the Test Tube, Stanford/Californien. Rescher (1984): Nicholas Rescher, The Limits of Science, Berkeley/Los Angeles/ London. Rescher (1991): Nicholas Rescher, Baffling Phenomena and Other Studies in the Philosophy of Knowledge and Valuation, Savage/Md. Sanders/Spooren (1997): José Sanders/Wilbert Spooren, “Perspective, Subjectivity, and Modality from a Cognitive Linguistic Point of View”, in: W.-A. Liebert/G. Redeker/L. Waugh (eds.), Discourse and Perspectivation in Cognitive Linguistics, Amsterdam/Philadelphia, 85 – 112. Tomasello (1999): Michael Tomasello, The Cultural Origins of Human Cognition, Cambridge/London. van Fraassen (2008): Bas C. van Fraassen, Scientific Representation: Paradoxes of Perspective, Oxford.

Critique of Representation: Cultures of Knowledge – Humanly Speaking Hans Jörg Sandkühler 1. The Topic The question whether the philosophically supported theory of cognition and knowledge – epistemology – is still playing a productive role in the analysis of the formation and the validity of knowledge today, is controversial. Has it not already signed its task over to the empirical sciences? Should it not give way to the neurosciences and cognitive sciences? I argue that it should not. An anachronism? I will try to show that this is not the case and why, in three steps. The first step is directed at ‘representation’, the second leads to the field of cultures of knowledge and the third leads to the question of the relation of beliefs, self-evidence and knowledge.1 Finally I will close with a short appeal against absolutist knowledge and truth claims. Let us approach the issue which is to be discussed – the problem of representation, cultures of knowledge, beliefs and knowledge – with two examples. (i) A neuroscientist says: “I know that free will is an illusion.” A neuroscientist stands by: “I know that the self is an illusion.” That each sentence beginning with “I know that” requires a critique of knowledge is not part of the research program nor of the understanding of this empirical discipline. My thesis is: such sentences starting with “I know” making use of empirical self-evidence – which therefore are supposed to be evidence-based 2 – , prove to be appeals to self-evidence. Let’s take (ii) the perception of ‘red’ at a traffic light as an example in the second case. If I were connected to an EEG while perceiving red, then my neuronal activity would be measurable at that moment. This would shed no light on my understanding of the function of traffic lights, my interpretation of ‘red’ as a command, nor would I obtain my knowledge about norms and 1 2

On the following cf. Sandkühler (2009). On the critique of false suppositions of self-evidence cf. Sandkühler (2011).

174

Hans Jörg Sandkühler

sanctions from these measurements. The explanation of the physiological properties of the brain and explanations of the mental properties – such as the intentional understanding of functions – are the result of the choice between two equally permissible languages3 – just as light is explainable by two theories: as corpuscle and as wave. The different ways of explaining are not caused as ‘entities’ by the brain or mind, they are rather part of contexts concerning cultural aspects of knowledge, of representation practices, none of which have the unique characteristic of ontological dignity. The ‘mind’ of the idealists remains a ghost without the brain as a material condition; the ‘brain’ of the materialists is a chimera without the language it is explained in. To use a metaphor: to cross a river you need a bridge. No one would go as far as to say that the bridge itself is the actual crossing of the bridge. When I speak of the critique of representation, I do not mean to say that the concept of representation is to be abandoned. I understand critique in the sense of the Kantian program as the ‘enlightening’ analysis of the conditions of possibilities of knowledge. Within the range of the further development of this tradition the crucial question of epistemological pluralism and internal realism is not how our knowledge depicts reality according to the measure of things but how the phenomenal reality emerges within knowledge, humanly speaking. What do we mean when we say that we have understood something and know it? What is the quality of our certainty?4 How do we speak meaningfully of certainty and truth if we can place no trust in the assumption that there is an accordance between an object/state of affairs and a statement, guaranteed by reality itself ? The modern critique of the conditions of possibilities has lead to the understanding that our statements are not copies of that which is to be recognized, but artifacts charged with presuppositions: charged with requirements concerning cultural aspects of knowledge and practical social aspects, epistemic and practical needs and interests as well as propositional attitudes of meaning, thinking and holding a belief, hoping and fearing. Not to forget the not-knowing5 in the shadow of which we say: “I know.” The not-knowing escapes philosophical and scientific knowledge as Michel Foucault described ‘unformulated 3 4 5

Cf. Dretske (1988). Cf. my encyclopaedic article on ‘Certainty’ (’Gewissheit’) in Sandkühler (2010), 909 – 919. On the problem of the not-knowing cf. Hogrebe (2006); also cf. Wehling (2004), (2006). Also see Hans Poser’s article in this volume in this regard.

Critique of Representation: Cultures of Knowledge – Humanly Speaking

175

thematics’ and ‘unseen obstacles’ as the ‘negative side of science’ and as a task of an ‘archaeology’ in Les mots et les choses (engl.: The Order of Things). To be able to translate “I believe that p” into “I know that p”, I have to justify my belief. But this is easier said than done. Because justification does not follow the dictate of things but a choice: the knowing subject S is faced with the choice of what Gaston Bachelard called the ‘epistemological profile’6 ; the subject can be a realist, an idealist or a naturalist. S2 choses a different world-view, a different profile from S1, S3 a certain conceptual scheme, S4 a certain theoretical framework, S5 a certain methodology. Each choice has consequences for the justification of beliefs. But the choice is not un-conditionally free; it has contexts: traditions, cultures, living conditions, specialization based on the division of labor, expediencies, etc. We are given the choice because the world of things does not impose a certain knowledge on us. There are no meaningful messages from a sender named ‘reality’ which we receive as addressees. It is we ourselves who are the authors of books on nature and history in which meanings for life-worlds are ascribed to reality. Is relativism unavoidable in the face of the diversity of possible choices? The relationality of all sentences beginning with “I know that p” cannot be avoided. The proceduralization of the justification of knowledge seems a suitable way of avoiding relativism which holds out the threatening prospect of pluralism – in analogy to the procedure with which the formation of the ‘intimate conviction of the judge’7, the last basis of judgment of the law of procedure, is strictly regulated. But the analogy has its limits: In epistemology there is no appellate body as in jurisdiction, in which a judgment can be repealed because of the irregular forming of a conviction. But in philosophy and science we can follow rules, not just those of logic. Part of the procedures, I am proposing, is to lay one’s cards on the table on behalf of the clarity and truthfulness of reasoning. The arguments for which validity is claimed should be dated and provided with one’s own signature – the signature of a choice. What are the cards that I am playing with? The epistemological profile that characterizes my considerations bears, in particular, features of 6 7

Cf. Bachelard (1968), 34 ff. Cf. ch. 10: ‘Conviction and Law’ (’Überzeugung und Recht’), Sandkühler (2009), 130 – 156.

176

Hans Jörg Sandkühler

Kant’s Critique, Peirce’s Semiotics, Cassirer’s Philosophy of the Symbolic Forms and Bachelard’s pistmolgoie. By referring to this profile as the result of a choice, I likewise concede that the critique of representation could be understood differently. The question whether we can ascribe knowledge to ourselves and make true statements about things in the external world, is not answered with the coherence theoretical “yes” and not with the “no” of the skeptic by the internal realist; his contextual answer is: “That depends”.8 The being-true of knowledge is a truth, e. g. a truth with the signature of a context.9

2. To the Critique of Representation Kant’s answer to the question of the limits of cognition has shocked the rationalistic modern age, in which the problem of the certainty of knowledge seemed to have been solved by the belief that cognition obeyes no other laws than the laws of Being itself. With Kant and after Kant doubt took the place of this certainty. As Ernst Cassirer put it: “for what assurance have we that the symbol of being, which we believe we have in our presentation, genuinely reproduces the content of being, and does not misrepresent its essential features?”10 8 On the social contextualization of scientific knowledge and the weak alternatively strong contextualized knowledge cf. Nowotny/Scott/Gibbons (2001), 121 ff. 9 On the ‘semantic rise of truth and the truth-theoretic fall of the external world’, cf. Hoffmann (2007), 44 – 74. 10 Cassirer (1923), 284. Original citation: “[W]elche Gewißheit besteht dafür, daß das Symbol des Seins, das wir in unseren Vorstellungen zu besitzen glauben, uns seine Gestalt unverfälscht wiedergibt, statt sie gerade in ihren wesentlichen Zügen zu entstellen?” (ECW 6, 305 f.) Cassirer stresses and intensifies the epistemic relativity resulting from this in Substanzbegriff und Funktionsbegriff (1910) (engl. transl.: Substance and Function): “A closer interpretation of the principle of the relativity of knowledge does not find this principle to be a mere consequence of the universal interaction of things, but recognizes in it a preliminary condition for the concept of thing in itself. This is the most general and radical meaning of relativity.” (Cassirer (1923), 306). Original citation: “Die schärfere Fassung des Prinzips der Relativität der Erkenntnis stellt dieses Prinzip nicht als eine bloße Folge aus der allseitigen Wechselwirkung der Dinge hin, sondern erkennt in ihm eine vorausgehende Bedingung für den Begriff des Dinges selbst. Hierin erst besteht die allgemeinste und radikalste Bedeutung des Relativitätsgedankens” (Cassirer (1923), 330).

Critique of Representation: Cultures of Knowledge – Humanly Speaking

177

With this insight a conception became questionable which was intended to serve as a safeguard for certainty – representation. Since Kant and even more so since the 1830s a many-voiced crisis debate on the status of representations has been in progress11; above all the new physiology of seeing (J. Müller, H. von Helmholtz a.o.), the Philosophy of the inductive Sciences (W. Whewell, J.S. Mill), proponents of the ignorabimusthesis on the limits of scientific knowledge like E. Du Bois-Reymond and not least of all impressionist and neo-impressionist painters like Monet and Seurat have been part of this debate.12 Based on the empiricism of the sciences the crisis was diagnosed as follows: observation sentences are not dictated by observation; it is rather the case that there is great freedom of signs – and an attribution of meaning.13 This also applies to the idea of ‘representation’. Theories of representation have contexts and themselves form contexts for the definition of what ‘representing’ itself means. At first concepts of representation are nothing more than empty shells which have to be provided with meaning in the context of a choice of epistemological profiles. The most general meaning of ‘representation’ is the following: an external ‘something’ will be ‘made present’ internally; it will be transformed into an internal ‘something’ by means of mental acts. The reality-status of that which is represented will be changed by imaginations, thinking, linguistic signs, symbols, metaphors, models and images. Representation is substitution. The product of representation stands for that which is to be represented. A reference to the represented is generally assumed within representation theories – more or less directly – but the representandum (that which is represented) does not ‘force’ any relation of similarity, isomorphy or identity to the representatum (that which represents). A perceived tree that is also referred to as such, exists as a something, and this, independent of being seen – but not as a tree. ‘Tree’ is a lin11 Cf. for more detail Sandkühler (1999) and (2002). 12 Cf. Sandkühler (2011a), 103 – 118 on this and for more detail. 13 Simon (1994). The “concept of the sign facilitates the orientation in the world as an orientation of oneself by ‘something’ that lets one remain free at the same time. […] The sign remains the same in the face of other interpretations, by other persons and by the ‘same’ person at a different time.” Original citation: The “Begriff des Zeichens ermöglicht die Orientierung in der Welt als Orientierung an ‘etwas’, das zugleich frei läßt. […] Das Zeichen bleibt gegenüber jeder Interpretation ‘stehen’ für andere Interpretationen, durch andere Personen und durch ‘dieselbe’ Person zu einer anderen Zeit.” (Simon (1994), 12).

178

Hans Jörg Sandkühler

guistic universal; it economizes the differentiation between singular birches, oaks and spruces. To the representing sentence “this is a tree” properties and meanings are ascribed, such as the change in size according to the distance from the object of the perceiver, the ascriptions of beauty in relation to the aesthetic attitudes or functions related to cultural system of symbols (a tree as a ‘justice tree’). The sentence “this is a tree” has to be completed: “I am convinced that this is a tree”. Everyday reasoning avoids – even in the sciences – the detour which involves the naming of the propositional attitude. Everyday reasoning is a realist and knows that common-sense is on its side. Common sense is a kind of (re-)presentation of reality “which claims to be the right one”. Common sense presents things which are of interest to it “as if the way they are what they are were simply in the nature of things.”14 Some representation theories assume a merely binary relation of representation. A representandum a and a representatum b have an irreversible relation; b stands for a. In this understanding one dimension is put aside, one that is intrinsically linked to representations: the dimension of intentionality. Complex concepts of representation are generally connected (i) with this idea; this means that what in cognition is assumed to be an object, is constituted as an object of knowledge in ‘directed’ acts of consciousness. Complex concepts of representation contain (ii) the supposition that representations are not independent of certain frameworks (knowledge cultures, languages, conceptual schemes). According to R. Tuomela representation is about “linguistic (or other representational) entities standing for or describing some non-linguistic entities, and these descriptions of real objects should then be seen as related to some conceptual system.”15 Tuomela stresses that, as there are no: “such ‘magical’, irreplaceable ties between language and the world”16, it cannot be assumed that ‘the world’ represents itself and of necessity represents itself in this language. What we are referring to as ‘representation’ is presentation. The function of representation is to make something present: it makes something present which is absent or does not exist in the external world in that way it is being represented. Representata take the place of the absent and speechless ‘things how they in themselves’ may be. No statement 14 Geertz (1995), 275 f. 15 Tuomela (1985), 108. 16 Ibid.

Critique of Representation: Cultures of Knowledge – Humanly Speaking

179

of “I know that p” contains a ‘thing as it is in itself’. The ‘things’ would be meaningless for the human way of thinking and behavior if they were anything else but represented. Via representation the amodal world ‘in itself’ becomes the life-world. Representations form ‘worlds’ in which subjects imagine and present themselves. Representation always includes elements of self-representation. Therefore it is within the critique of representation – as E. Cassirer puts it – that “the metamorphosis by which the phenomenon of the mere data of consciousness turns into a content of reality, into a content of the ‘external world’” is to be explained.17 The costs of subjectivity and relativity of this position are self-evident and they are inconvenient for scientific practice. However, the world remains the world. But no world view is privileged on the strength of truth and validity conditions due to one single clearly distinguished reason for being. Once again the proceduralizing, as already mentioned, can help us to avoid relativistic conclusions. The rule is: signatures are to refer to the language which is being used while representing; they should refer to presuppositions, premises and context conditions. There are no representations which are not interpreted by signs, symbols or concepts. The ‘neural/mental’ dualism of signification which seems to be ontologically bound to a difference in entities is due to a choice of methods and theories. In both cases it concerns significations, assertions in sign systems, the ascription of meanings to representations, the explications which do not fall within the domain of responsibility of the empirical sciences but in the domain of semantics and semiotics. It therefore makes sense to link every expression of the form “b is a representation of a” with a signature: here representation is spoken of in a belief system B under certain epistemic conditions eC in a certain language L: b repr a under B and eC in L. Representations thus understood are interpretations via which worlds of knowledge are constituted as worlds of things. Interpretation accomplishes what Kant calls “objects for notions”.18 Günter Abel has characterized them as processes in his book Interpretationswelten “in which we discriminate something phenomenally as a certain something, undertake identifications and re-identifications, apply predicates and la17 Cassirer, Translation by UF. Original citation: “die Metamorphose zu erklären, durch welche die Erscheinung aus einem bloßen Datum des Bewußtseins zu einem Inhalt der Realität, der ‘Außenwelt’, wird” (ECW 13, 141). 18 Kant, AA V, 468: “Gegenstände für Begriffe”.

180

Hans Jörg Sandkühler

bels and make ascriptions […], and with regard to worlds thus formed possess opinions, beliefs and justified knowledge. Our worlds can therefore be qualified as worlds of interpretation and be treated accordingly.”19 Interpretations define epistemic space and historical time as possibility conditions (and limits) of knowing how and knowing that. In the following I will address these conditions of possibility and limits as ‘cultures of knowledge’.

3. Cultures of Knowledge Perceptions and experiences, observations and experiments20, beliefs, representations and knowledge are embedded in networks of cultures of knowledge.21 Cultures of knowledge are to be spoken of in the plural because in the modern age, and not only regarding the broad scientific usage in Europe, epistemic cultures with their own self-will and their own world views, sign systems, semantics and truth claims have developed due to the division of labor, specialization, autonomization of function systems and the social and cultural differentiation within societies. In the perspective of ‘cultures of knowledge’ a rationalistic, narrowed concept of knowledge is broadened: a polyadic structure of relations including pre-rational deep layers is taken into consideration out of which knowledge is formed – mediated by emotions, attitudes and beliefs.22 19 Abel (1993), 14. “in denen wir etwas als ein bestimmtes Etwas phänomenal diskriminieren, Identifikationen und Re-Identifikationen vornehmen, Prädikate und Kennzeichen applizieren, Zuschreibungen durchführen […], und in Bezug auf so formierte Welten dann über Meinungen, Überzeugungen und auch über ein gerechtfertigtes Wissen verfügen. Unsere Welten können darum als Interpretationswelten qualifiziert und diese als jene behandelt werden.” 20 Cf. Hoffmann (2009). 21 In the philosophical epistemology and knowledge theory the concept ‘cultures of knowledge’ has played no role until now. Matters are different with regard to the sociology of knowledge and science; on that cf. Weingart (2003) on “Epistemic communities, knowledge cultures and knowledge society” (“Epistemische Gemeinschaften, Wissenskulturen und Wissensgesellschaft”). On ‘cultures of science’ cf, e. g. Arnold/Dressel (2004). 22 On the labeling of the structure of relations K. Gloy suggested a ‘system of levels’ “1. instinctive knowledge, 2. emotional, sensual knowledge, a) situational understanding and b) gestural understanding, 3. practical knowledge with its basis of magical knowledge, 4. intellectual scientific knowledge, cognition, 5. transrational knowledge or modified kinds of knowledge, like dreams, auguries,

Critique of Representation: Cultures of Knowledge – Humanly Speaking

181

This opens the concept of knowledge to the understanding that different attitudes and beliefs can lead to conflicting truth claims within cultures. Varying epistemic traditions, styles of argumentation and representation as well as self images and public images are the foundations for epistemic identities. A result of this broadening of the concept is the concession – or the alarm this causes – that the holding-for-true and the holding of a belief are not generally empirically justified and that knowledge is empirically underdetermined. Another consequence is the barely denied loss of the special position of – according to its self-image – the exact, objective, nomological empirical natural sciences. Within them, too, the world develops in versions. “The world-as-we-know-it” N. Rescher writes in his Studies in Epistemology and Cognitive Theory “is accordingly our world – the correlate of mind of a world-picture devised in characteristically human terms of reference.”23 Cultures of knowledge are conditions of possibility and limits of cognition, according to the degree of participation in or exclusion from knowledge24 – up to and including the understanding or non-understanding of the words which do not reveal their meanings but receive them in conjunction with an encyclopedia. What a chaconne is, what the Greek word ‘nous’ means, the fact that in Bremen ‘slaughterhouse’ stands for a cultural center, what the abbreviations PET or fMRT stand for – all these things remain incomprehensible unless one is familiar with a certain culture of knowledge. The signs as such remain silent beyond their contexts of meaning. The following dimensions are essential for the thinking, behavior and actions of individuals and their integration in epistemic and social networks: (i) religious truths, metaphysical interpretations and expectations of salvation: this level includes the integration of the individuals in belief communities. (ii) Basic epistemic attitudes, beliefs and preferences prophecies, obsessions, remote viewing, meditation.” Original citation: “1. das instinktive Wissen, 2. das emotionale, sinnliche Wissen, a) das situative Verstehen und b) das gestische Verstehen, 3. das praktische Erfahrungswissen mit seiner Grundlage, dem magischen Wissen, 4. das intellektuelle, wissenschaftliche Wissen, die Kognition, 5. das transrationale Wissen bzw. die modifizierten Wissensarten, wie Traum, Weissagung, Prophetie, Fernwahrnehmung, Besessenheit, Meditation.” (Gloy (2007), 24). 23 Rescher (1990), 77. 24 Examples of an exclusion are inquisition, banning and censorship, cf. Wolf (2001).

182

Hans Jörg Sandkühler

in the perspectives of which realities develop. This level includes the integration of individuals in realistic, idealistic, skeptical and other– e. g. aesthetic – paradigms. (iii) Perception, thinking and life styles which are related to moral conventionalities, customs, manners etc. which are characterized by everyday culture and common sense: this level includes the integration of individuals in sociocultural environments. (iv) Social and political fundamental values: this level includes the integration of individuals in social orders and normative systems.25 The individuation linked with the integration into cultures of knowledge presupposes the participation in symbolic traditions which guarantee “a common inventory of basic settings”. According to Oswald Schwemmer the attitudes developing out of the practices operate “for their part as a practical confirmation of the symbolic tradition, so that one can indeed speak of a collective identity, namely of the mutual confirmation relation of the symbolic tradition, collective practices and attitudes that exist in a society.”26 Cultures of knowledge cultures can be understood as confirmation relations of this kind and as the frameworks for the production of knowledge. These frameworks have different extensions. Relatively narrow frameworks – e. g., terminologies, conceptual schemes and theories – are interconnected with other frameworks – e. g., with sets of everyday knowledge, paradigms and scientific disciplines –, and these are linked to very wide horizons, such as religions and other world pictures. Theoretical principles, norms and rules of empiricism, description and explanation strategies, the formation of hypotheses, testing procedures and justifications as well as the revision of epistemological beliefs are also contained in frameworks concerning cultural aspects of knowledge. To sum it up in Karin Knorr Cetina’s words: cultures of knowledge encroach on the practices, mechanisms and principles “which establish how we know what we know by being bound to relations, necessity and historical coincidences within a domain of knowledge. Cultures of knowledge generate and validate knowledge.”27 25 I agree with Meyer (2002), 117 ff. on the points (i), (iii) and (iv). 26 Original citation: “ihrerseits wieder als eine praktische Bestätigung der symbolischen Traditionen, so dass man tatsächlich von einer kollektiven Identität reden kann, nämlich dem wechselseitigen Bestätigungsverhältnis der symbolischen Traditionen, gemeinsamen Praktiken und Einstellungen, die in einer Gesellschaft bestehen” (Schwemmer (2005), 259). 27 Original citation: “die, gebunden durch Verwandtschaft, Notwendigkeit und historische Koinzidenz, in einem Wissensgebiet bestimmen, wie wir wissen,

Critique of Representation: Cultures of Knowledge – Humanly Speaking

183

The inclusion of cultures of knowledge in the criticism of representation has consequences: (i) the thesis that knowledge is justified true belief has to be broadened: knowledge is, within contexts concerning cultural aspects of knowledge and according to standards of justification accepted by standards concerning cultural aspects of knowledge, justified true belief; (ii) the assumption that representation could be conceived as a binary relation b repr a is untenable. The representation relation is polyadic: b repr a under the conditions concerning cultural aspects of knowledge c, d, e; a is represented as ac, ad or as ae. Therefore the epistemological approach of sparing copy theories linked with b repr a has to be abandoned. The demand of aiming at least at the possible optimum in context consideration is not maximalistic. To quote Albert Einstein: everything should be made as simple as possible, but not simpler. The demand is appropriate in cases when epistemology forms an alliance with those sciences which contribute to the analysis of the possibility conditions of knowledge.

4. Beliefs, Self-Evidences and Knowledge Knowledge that p is contextual. The difficulty of a justification of knowledge as ‘true’ is based within this, a justification which should be more than the acknowledgment of being true, because knowledge and belief contexts are – as Kripke emphasizes – “referentially opaque”28. The truth which we ascribe to our knowledge is relative, and this means: it stands in relation to semantic and other contexts. It is precisely this problem in connection with which the theorem is to be discussed which is only apparently self-evident: *

Knowledge is justified true belief.

What appears as a definition signifies a problem. L. Wittgenstein saw this in his late work On Certainty. He speaks of a “natural law of ‘taking for true’” and emphasizes: “The difference between the concept of ‘knowing’ and the concept of ‘being certain’ isn’t of any great importance at all, except where “I know” is meant to mean: I can’t be wrong.29 […] For “I know” seems to describe a state of affairs which was wir wissen. Wissenskulturen generieren und validieren Wissen” (Knorr Cetina (2002), 11). 28 Cf. Kripke (1979). 29 Wittgenstein (1989), § 8. Original citation: “Naturgesetz des ,Fürwahrhaltens’”

184

Hans Jörg Sandkühler

guarantees what is known, guarantees it as a fact. One always forgets the expression “I thought I knew”.30 Wittgenstein stresses: “The truth of certain empirical propositions belongs to our frame of reference.”31 One variant was formulated by a contemporary of Wittgenstein’s, the sociologist of the history of sciences L. Fleck: he calls that which in the sciences is called the observation of facts, an ability to see which is dependent on thought communities and thought styles.32 Shared beliefs and self-evidences often create questionable certainty within cultures of knowledge. The supposition that something is like this or that is expressed by sentences like “It is in that way, that p” = “I know that p, and it is true that p”. Thus nothing is yet decided regarding the truth of a sentence. “That p” has to be justified as being true. An epistemological standard supposition is that the correctness of a statement of the form “S knows that p” presupposes more than subjective certainty. It is presupposed that (i) S believes that p, (ii) that p is true and (iii) that S can justify her belief that p. 33 But it is more than questionable whether this presuppositiontriad is completely compliable at all.34 If it is, knowledge should be spoken of as justified true belief – if the difference between belief and knowledge were clearly determinable and the meaning of ‘true’ were clarified and if the criteria and standards of justification were indisputable. This, however, is by no means the case. In “I know that p”-statements there is not only a relation between an I and a state of affairs but every statement of “I know that p” is linked to other sentences, to an encyclopedia, with emotions, with a habitus. “I know that” (knowing that 35, factual knowledge) is linked

30 Wittgenstein (1989), § 12. Original citation: “Der Unterschied des Begriffs ,wissen’ vom Begriff ,sicher sein’ ist gar nicht von großer Wichtigkeit, außer da, wo ,Ich weiß’ heißen soll: Ich kann mich nicht irren. […] ,Ich weiß…’ scheint einen Tatbestand zu beschreiben, der das Gewußte als Tatsache verbürgt. Man vergißt eben immer den Ausdruck ,Ich glaubte, ich wüßte es’.” 31 Wittgenstein (1989), § 83. Original citation: “Die Wahrheit gewisser Erfahrungssätze gehört zu unserm Bezugssystem.” 32 Fleck (1983), 167. Cf. Comment 5. 33 Cf. Musgrave (1993), 3. 34 On the difficulty of the ‘definition’ of knowledge as justified true belief see also E.L. Gettier (1963). 35 Cf. Detel (2007), 48.

Critique of Representation: Cultures of Knowledge – Humanly Speaking

185

with my knowing by acquaintance and with my “I know how” (knowing how, rule related knowledge) in a complex way.36 “I know” is an expression for the belief of having knowledge. The beliefs that people have corresponding to their need for orientation can be elusive, relatively stable, or constant in more or less long phases of life, depending on their relevance for the way of life. Constant world picture-like beliefs (’basic beliefs’) have the status of axioms and self-evidences; they seem to justify themselves; they form the basis of the justification for other relatively stable or unstable beliefs, especially in connection with the belief-supported demarcation of beliefs of a third party, or with their criticism and rejection; in this regard beliefs mark the limits of tolerance.37 Each culture, each subculture and each homogeneous group develops a set of attitudes and beliefs which characterizes its own normality. In this normality collective options and counter-options merge with individual beliefs and actions – in the forms of assimilation and affirmation but also in the forms of refusal or resistance. Both, adapting to a collective belief system (possibly to a religion, a social movement, a party38, but also to a certain part of the scientific system) and subscribing to the precarious beliefs of a resistance group can be emotionally satisfying and fulfill the function of an epistemic ‘home’. Why is something considered as being certain? The question of whether there is such a thing as secure knowledge of something that is free of doubt has been on the agenda since antiquity. There is, the answer has been and is heard over and over again, a knowledge that can count as self-evident in the sense that it is ‘intuitively certain’, that is ‘that a thing itself is immediately self-evident’39 : self-evidence, however, cannot be distinguished from subjective certainty. If there is no differentiation criterion, an appeal to self-evidence is impossible according to W. Stegmüller “as we do not know then whether it was just apparent selfevidence (subjective certainty concurring with a mistake)”; if a criterion 36 For this distinction and the consequences for the justification of knowledge cf. Musgrave (1993), 6 ff. 37 Criticizing the beliefs of a third party as ‘mere beliefs’ is part of the paradoxical habit of holding a belief. 38 Collective belief systems are characterized by the belief of knowing what counts as proper beliefs; divergence is condemned as heresy, betrayal, apostasy, etc. 39 The metaphorical content of ‘self-evidence’ refers to a long tradition of metaphors of light (’lumen naturale’, ‘natural light of reason’.) For the history of the concept of self-evidence cf. Halbfass/Held (1972).

186

Hans Jörg Sandkühler

is accepted, “this therefore appears to lead to infinite regress: that the required features of the criteria are present in a given case of certainty does not have to be ascertained simply with subjective certainty, but with self-evidence.”40 Thus ‘self-evidence’ turns into a concept in the manner of Münchhausen: the fact that this or that is self-evident, is self-evident. Self-evidence is believed certainty, a knowledge one believes to be true. That, however, is not all there is to be said. Self-evidence takes effect de facto in beliefs within cultures of knowledge. With insight into the contextuality concerning cultural aspects of knowledge of what we intend with our epistemic acts and what we take to be credible, self-evident, correct and true, we cannot reject questions regarding the reasons for certainty anymore. Within the belief-attitude with regard to the proposition of the “I know”, we express the certainty, in contrast to the mere probability, which we believe to have; we believe that we have a proper truth. Wittgenstein demands: “It needs to be shown that no mistake was possible. Giving the assurance “I know” doesn’t suffice. For it is after all only an assurance that I can’t be making a mistake, and it needs to be objectively established that I am not making a mistake about that.”41 Objectively established? For someone having beliefs about knowledge, these usually count as unproblematic from the time when the formation of a belief has been completed: “It is like this and cannot be any other way.” Such beliefs are an intrinsic part of human existence. A phenomenology of beliefs is of interest for the critique of representation – in other words for the epistemological analysis of the possibility conditions of knowledge. Since one belief is not like another: beliefs about ‘facts’ which are supported by everyday perceptions of what is close (“this is my house”, “I have a migraine”) do not have, being “observation-beliefs”42, the same status as beliefs of a wide range (“human rights are uni40 Stegmüller (1989). Original citations: “da wir dann nicht wissen, ob die Evidenz nicht bloß eine scheinbare war (subjektive Gewissheit bei gleichzeitigem Irrtum)”. / “so scheint dies zu einem unendlichen Regreß zu führen: daß in einem vorgegebenen Falle von Gewißheit die in dem Kriterium verlangten Merkmale vorliegen, muß ja selbst nicht bloß mit subjektiver Gewißheit, sondern mit Evidenz festgestellt werden” (Stegmüller (1989), 47 f). 41 Wittgenstein (1989), § 15. Original citation: “Daß kein Irrtum möglich war, muß erwiesen werden. Die Versicherung ,Ich weiß es’ genügt nicht. Denn sie ist doch nur die Versicherung, daß ich mich (da) nicht irren kann, und daß ich mich darin nicht irre, muß objektiv feststellbar sein.” 42 Bartelborth (1996), 76.

Critique of Representation: Cultures of Knowledge – Humanly Speaking

187

versally valid”) and “meta-beliefs”43, i. e. beliefs about beliefs (possibly about the moral value or the epistemic justification of a belief). The function of many beliefs straddles various domains: thus the naturalistic basic belief of the neuroscientist possibly allows a belief in the non-existence of free will, straddling various domains on the basis of the physical unit and nomological order of ‘reality’, in conjunction with the belief in the reducibility from ‘mind’ on neural activities of the brain or with the belief in the experimental ascertainability of the identity of neural and mental brain states and brain processes. The problem is whether beliefs with a broad range can be epistemically justified. At any rate, not all beliefs can be reflexively criticized. Can the belief expressed in the sentence “I know that it is like that and not any other way” also be justified if it does not concern obvious facts, but instead intended, thought of, dreamed of, fantasized, abstract or theoretical facts, briefly: ‘facts’ asserted without any specific empirical correlate? Are ‘facts’ capable of truth in an objective sense? Thomas Nagel claims that objectivity is a “method of understanding. It is beliefs and attitudes that are objective in the primary sense. Only derivatively do we call objective the truths that can be arrived at in this way.”44 Beliefs are subjective attitudes of individuals in trans-individual contexts. To be intersubjectively understandable or acceptable, beliefs must be linkable to common attitudes of smaller or bigger social groups. Having a belief, having self-evidence and knowledge makes one immune to skepticism and criticism. This is why G. Bachelard has appealed for a ‘psychoanalysis of objective knowledge’ in La formation de l’ esprit scientifique (1938): having objectivity and truth as well as the “joys of ownership and objective certainties” must be subjected to criticism.45 This criticism shows that we do not have objectivity and truth because of ‘things’ but because we have chosen a certain epistemological profile.46 According to Ch. S. Peirce – “the most that can be maintained” with regard to the truth of beliefs “is that we seek for a belief that we shall think to be true. But we think each one of our beliefs to be true, and, indeed, it is mere tautology to say so.”47 The truth which we 43 Bartelborth (1996), 78. 44 Nagel (1986), 4. Cf. Nagel (1979). 45 Bachelard (2002). German translation: Bachelard (1984), 211 (“Besitzerfreuden und objektive Gewissheiten”). 46 On the pistmologie of Bachelard cf. Sandkühler (2009), 47 – 50. 47 Peirce (1958), 232.

188

Hans Jörg Sandkühler

claim for our sentences concerning facts is a function of our beliefs. In these beliefs – according to the pragmatist W. James – reality appears as independent, “as a thing found, not manufactured.”48 However, this is only an appearance on the stage of knowledge. Where we believe that we are encountering an ‘original reality’ “it has been already faked.”49 James sums it up as follows: “What we say about reality thus depends on the perspective into which we throw it. The that of it is its own; but the what depends on the which; and the which depends on US. Both the sensational and the relational parts of reality are dumb: they say absolutely nothing about themselves. We it is who have to speak for them.”50 Knowledge related beliefs are habits of self-evidence. They are comparable to energy-saving measures: they economize on the effort of reflexivity and self-criticism or criticism. Habits of self-evidence make life easier because they minimize the requirements for the justification of knowledge and action.

5. An Appeal against Absolutistic Knowledge and Truth Claims The definition “knowledge is justified true belief” turns out to be a problematic thesis to the extent that the justification of a belief as ‘true’ is problematic because the justification based on facts is problematic: “If truth is a binary relation” M. Willaschek writes, “relata cannot be the content of a belief on the one hand and a fact which ‘makes’ the belief ‘true’ on the other, as the relation of ‘making true’ would lead, so to speak, nowhere.”51 With reference to entities which are not observable, like gods and black holes, at least there is no obvious difference between belief and

48 49 50 51

James (1907), 243. James (1907), 249. James (1907), 246. Willaschek (2003). Original citation: “Wenn Wahrheit eine zweistellige Relation ist”, M. Willaschek writes, “können ihre Relata […] nicht einerseits der Inhalt einer Überzeugung und andererseits eine Tatsache sein, die diese Überzeugung ,wahrmacht’, da die Relation des Wahrmachens gleichsam leerliefe.” (Willaschek (2003), 241 f).

Critique of Representation: Cultures of Knowledge – Humanly Speaking

189

knowledge.52 Many of our beliefs are justification-transcendent beliefs53, beliefs beyond the possibility of being justified. These are beliefs for which, in paradoxical accordance with the widespread common-sense-realism of everyday life and of the laboratories, weaker reasons are accepted than the reasons demanded by the hard empiricists. Everyday reasoning can definitely move within the limits of beliefs for emotional and pragmatic reasons without requiring an ontic guarantee for the propositions linked with it. This especially applies to religious, ideological beliefs and to those concerning one’s world-view in which the difference between ‘existing/non-existing’ disappears as a basis for the assessment of something as ‘true/not true’. Criteria of epistemic acknowledgment that p take the place of the demand for the actual given that p. They get by with far weaker reasons than with strong epistemic justifications. The acknowledgment of a state of affairs does not assume that there are ‘objective’ reasons for its justification. Acknowledgment has sources in beliefs which are divided into social, political, scientific or philosophical belief communities. I know myself entitled to my knowledge “that p”, when ‘experience teaches it’, when authorities claim to know it, when it is useful with regard to my happiness or for the happiness of others or for ‘mankind’ or leads to desired success. The kind of knowledge which is employed in justification-transcendent beliefs is immune to revision at the time of holding that belief; it is intrinsically infallible. Provided that the true/false alternative has been pragmatically reinterpreted in a way that not the actuality of something but the possible acknowledgment of the assertion “that something is a certain way” is the criterion, then assertions which are actually not justifiable (religious, fictional, ideological, irrational) become acceptable in the culture of knowledge of a society (“Iraq has weapons of mass destruction”). Such assertions are immune to falsification by the proof “that not p” as long as they can be assumed as self-evident and are coherent with a collective and/or individual belief system. Their empirical falsification does not necessarily mean the devaluation of its function as something which one can have as ‘knowledge’ and that performs an orientation function. For the acknowledgment understood as the justification of a belief is the result of the (basic) belief that a whole belief system is coherent and justified with the entirety of its self-evidences. 52 Cf. Sayre (1997), 121 – 128 on theories which strictly distinguish between belief and knowledge and those which do not think the distinction possible. 53 Cf. Willaschek (2003), 250 – 253 as to this type of beliefs.

190

Hans Jörg Sandkühler

Instead of the justification required by the sentence “knowledge is justified true belief” a far weaker strategy is obviously de facto sufficient: the weak strategy of acknowledgment is based on self-evidence; it is content with the assumption of correspondence, expects, however, de facto nothing more than coherence; it is functionally and pragmatically justified and allows the understanding of beliefs and corresponding assertions; it trivializes or minimizes normative truth claims: the belief system decides on how the predicator ‘true’ is used; ‘true’ does not have to signify a relation between the assertions belonging to a belief system and the ‘facts’ in the ‘real world’; a belief is recognized as true if it is not refuted evidentially. Is it therefore an anachronism to cling to the normative idea of ‘truth’? By no means. Without justification and truth claims neither philosophy nor the sciences are conceivable. The often expressed assumption that internal realism understood as epistemological pluralism leads to the abandonment of truth claims is not justified. Internalists merely account for actual epistemic pluralism and say that truth can only be claimed as truths within belief systems with the signature of the system. More modest truth claims result from this, coupled with an increased need for requirements to differentiate between right and wrong world versions: according to Goodman one version is “taken to be true when it offends no unyielding beliefs and none of its own precepts”.54 These requirements are also de facto accepted in philosophy and in a ‘liberalized’ philosophy of science which substitutes shared standards for acceptance for ontological reasons of truth justification: there are normative intuitions about the cognitive aims and procedures which should be recognized for the achievement of these aims within knowledge communities.55 Internal realists do not want and are not able to claim more than formal consistency (internal consistency), coherence and – and this is the epistemical-ethical side – truthfulness. What remains if one appeals for modest knowledge and truth claims? The question is not put in the right way. The primary question should concern the benefit in terms of orientation within the life-world, if limits are set to truth claims. My conclusion is: beliefs which are recognized in accordance with rules, can be communicated and publicly justified, are understandable for third parties and which do not abuse the norms of human rights can claim acknowl54 Goodman (1978), 17. 55 Cf. Carrier (2007).

Critique of Representation: Cultures of Knowledge – Humanly Speaking

191

edgment. They are ‘true’ beliefs for individuals as epistemic subjects and for sign and meaning communities. They are knowledge humanly speaking. The dogmatism of claims to the ‘one and only’, the ultimate truth is a foundation for claims to power; it leads to a constraint on independent thinking and the freedom of thought and action. Modest truth claims serve in practice to defend the indefeasible right to opposition in a state under the rule of law, and to the rejection of absolute power and claims of absolute truth, rights which constitute a democracy.56

References AA

Kant’s gesammelte Schriften, Königlich Preußische Akademie der Wissenschaften (ed.), Berlin, 1900 – 1955, 1966 ff. ECW E. Cassirer. Gesammelte Werke, Hamburger Ausgabe, B. Recki (ed.), Hamburg, 1998 ff. Abel (1993): Günter Abel, Interpretationswelten. Gegenwartsphilosophie jenseits von Essentialismus und Relativismus, Frankfurt/Main. Arnold/Dressel (2004): Markus Arnold/Gert Dressel, Wissenschaftskulturen – Experimentalkulturen – Gelehrtenkulturen, Wien. Bachelard (1980): Gaston Bachelard, Die Philosophie des Nein. Versuch einer Philosophie des neuen wissenschaftlichen Geistes, Frankfurt/Main. Bachelard (1984): Gaston Bachelard, Die Bildung des wissenschaftlichen Geistes. Beitrag zu einer Psychoanalyse der objektiven Erkenntnis, Frankfurt/Main. Bartelborth (1996): Thomas Barthelborth, Begrndungsstrategien. Ein Weg durch die analytische Erkenntnistheorie, Berlin. Braun/Stephan (2005): Christina von Braun/Inge Stephan, Gender&Wissen. Ein Handbuch der Gender-Theorien, Köln/Weimar/Wien. Carrier (2007): Martin Carrier, “Wege der Wissenschaftsphilosophie im 20. Jahrhundert”, in: M. Stöckler/A. Bartels (eds.), Wissenschaftstheorie. Texte zur Einfhrung, Paderborn. Detel (2007): Wolfgang Detel, Erkenntnis- und Wissenschaftstheorie, Stuttgart. Dretske (1998): Fred Dretske, Die Naturalisierung des Geistes, Paderborn. Fleck (1983): Ludwik Fleck, Erfahrung und Tatsache. Gesammelte Aufstze, Frankfurt/Main. Freudenberger (2004): Silja Freudenberger, Erkenntniswelten. Semiotik, analytische Philosophie, feministische Erkenntnistheorie, Paderborn. Geertz (1995): Clifford Geertz, Dichte Beschreibung. Beitrge zum Verstehen kultureller Systeme, Frankfurt/Main. Gloy (2007): Karen Gloy, Von der Weisheit zur Wissenschaft. Eine Genealogie und Typologie der Wissensformen, Freiburg/München. 56 Translation from German by Ute Feldmann.

192

Hans Jörg Sandkühler

Goodman/Elgin (1993): Nelson Goodman/Catherine Z. Elgin, Revisionen. Philosophie und andere Knste und Wissenschaften, Frankfurt/Main. Halbfass/Held (1972): Wilhelm Halbfass/Klaus Held, “Evidenz”, in: J. Ritter/ K. Gründer (eds.), Historisches Wçrterbuch der Philosophie, Basel. Hoffmann (2009): Melanie Hoffmann, Wissenskulturen, Experimentalkulturen und das Problem der Reprsentation, Frankfurt/Main. Hoffmann (2007): Thomas Hoffmann, Welt in Sicht. Wahrheit – Rechtfertigung – Lebensform, Weilerswist. Hogrebe (2006): Wolfram Hogrebe, Echo des Nichtwissens, Berlin. James (2001): William James, Pragmatismus – Ein neuer Name fr einige alte Denkweisen, Darmstadt. Knorr Cetina (2002): Karin Knorr Cetina, Wissenskulturen. Ein Vergleich naturwissenschaftlicher Wissensformen, Frankfurt/Main. Kripke (1979): Saul Kripke, “A puzzle about belief”, in: A. Margalit (ed.), Meaning and Use, Dordrecht, 239 – 283. Kripke (2004): Saul Kripke, “Ein Rätsel um Überzeugungen”, in: M. Textor (ed.), Neue Theorien der Referenz, Paderborn. Meyer (2002): Thomas Meyer, Identittspolitik. Vom Missbrauch kultureller Unterschiede, Frankfurt/Main. Musgrave (1993): Alan Musgrave, Alltagswissen, Wissenschaft und Skeptizismus. Eine historische Einfhrung in die Erkenntnistheorie, Tübingen. Nagel (1991): Thomas Nagel, Die Grenzen der Objektivitt. Philosophische Vorlesungen, Stuttgart. Nagel (1992): Thomas Nagel, Der Blick von nirgendwo, Frankfurt/Main. Nowotny/Scott/Gibbons (2004): Helga Nowotny/Peter Scott/Michael Gibbons, Wissenschaft neu denken. Wissen und ffentlichkeit in einem Zeitalter der Ungewißheit, Weilerswist. Peirce (1958): Charles Sanders Peirce, Collected Papers of Charles Sanders Peirce, volumes I-VI, C. Hartshorne/P. Weiss (eds.), Harvard UP, 1931 – 35; vol. VII, VIII, ed. W. Burks, Cambridge/ MA; 2. Edition, Cambridge/ MA [= CP]. Rescher (1990): Nicholas Rescher, “A useful inheritance: Evolutionary Aspects of the Theory of Knowledge”, in: P. K. Moser (ed.), Studies in Epsitemology and cognitive theory, Maryland, 77 – 107. Sandkühler (1997): Hans Jörg Sandkühler, “Evidenz. Zur Kritik dogmatischer Ansprüche in philosophischen Kulturen”, in: N. Schneider et al. (ed.), Philosophie aus interkultureller Sicht/Philosophy from an Intercultural Perspective, Amsterdam/Atlanta. Sandkühler (2002): Hans Jörg Sandkühler, Natur und Wissenskulturen. SorbonneVorlesungen ber Pluralismus und Epistemologie, Stuttgart/Weimar. Sandkühler (2009): Hans Jörg Sandkühler, Kritik der Reprsentation. Einfhrung in die Theorie der berzeugungen, der Wissenskulturen und des Wissens, Frankfurt/Main. Sandkühler (2010): Hans Jörg Sandkühler, “Gewissheit”, in: H. J. Sandkühler (ed.), Enzyklopdie Philosophie, Hamburg.

Critique of Representation: Cultures of Knowledge – Humanly Speaking

193

Sandkühler (2011): Hans Jörg Sandkühler, “Kritik der Evidenz”, in: J. Bellmann/ Th. Müller (eds.) Wissen, was wirkt: Kritik evidenzbasierter Pdagogik, Wiesbaden. Sandkühler (2011a): Hans Jörg Sandkühler, “Menschsein – Weisen des Sehens und Gestaltens von Welten”, in: Hong-Bin Lim/Georg Mohr (eds.), Menschsein. On Being Human. Deutsche und koreanische Studien zu Epistemologie, Anthropologie, Ethik und Politischer Philosophie, Frankfurt/Main. Sayre (1997): Kenneth M. Sayre, Belief and Knowledge. Mapping the Cognitive Landscape, Lanham/Md. Schwemmer (2005): Oswald Schwemmer, Kulturphilosophie. Eine medientheoretische Grundlegung, München. Simon (1994): Josef Simon, “Vorwort”, in: J. Simon (ed.), Zeichen und Interpretation, Frankfurt/Main. Stegmüller (1989): Wolfgang Stegmüller, Hauptstrçmungen der Gegenwartsphilosophie . Eine kritische Einfhrung, vol. 1, Stuttgart. Tuomela (1985): Raimo Tuomela, Science, action and reality, Dordrecht. Tuomela (1991): Raimo Tuomela, “Kausaler interner Realismus”, in: H. J. Sandkühler/D. Pätzold (eds.) Die Wirklichkeit der Wissenschaft – Probleme des Realismus, Hamburg. Wehling (2004): Peter Wehling, “Weshalb weiß die Wissenschaft nicht, was sie nicht weiß? Perspektiven einer Soziologie des wissenschaftlichen Nichtwissens”, in: P. Böschen/P. Wehling (eds.), Wissenschaft zwischen Folgenverantwortung und Nichtwissen. Aktuelle Perspektiven der Wissenschaftsforschung, Wiesbaden. Wehling (2006): Peter Wehling, Im Schatten des Wissens? Perspektiven der Soziologie des Nichtwissens, Konstanz. Weingart (2003): Peter Weingart, Wissenschaftssoziologie, Bielefeld. Willaschek (2003): Markus Willaschek, Der mentale Zugang zur Welt. Realismus, Skeptizismus und Intentionalitt, Frankfurt/Main. Wittgenstein (1989): Ludwig Wittgenstein, ber Gewißheit, G. E. M. Anscombe/G. H. von Wright (eds.), Frankfurt/Main. Wolf (2001): Hubert Wolf (ed.), Inquisition, Index, Zensur. Wissenskulturen der Neuzeit im Widerstreit, Paderborn.

II. Perceptual Knowledge

Kantian Conceptualism Thomas Land Whether or not perceptual content is conceptual, and what it means to say that it is, have been much-debated questions in recent philosophy of perception. A striking fact about this debate is that Kant has been invoked on two opposing sides of it. Thus, some advocates of conceptualism – the view that the content of a perceptual experience has the structure of a proposition and must therefore involve the application of concepts – have traced their position back to Kant’s theory of intuition.1 At the same time, some proponents of nonconceptualism – the view that the content of perception is not propositional in structure and does not involve the application of concepts – have argued that, on the contrary, Kant’s theory of intuition is a forerunner of their own position.2 A natural reaction to this is to think that at least one side of the debate is mistaken about Kant and that careful examination of Kant’s texts will show this. In this paper, I want to combine an exegetical inquiry of this sort with an assessment of the philosophical merits of the positions staked out in this debate. This project is motivated by the following consideration: Assuming that neither side has been reckless in claiming that it incorporates a central insight of Kant’s theory of intuition, the very fact that this theory can be invoked with some justification to support both these opposing positions should itself be a cause for philosophical reflection and lead us to ask what it is about this theory that makes 1 2

The most prominent advocate of this position has been John McDowell. See McDowell (1996), as well as the first three essays in McDowell (2009). A prominent advocate of this position is Robert Hanna who wants to defend a “Kantian theory of the semantic structure and psychological function of nonconceptual content” (Hanna (2008), 42) and who claims that Kant did in fact hold a theory of this kind: “[…] Kant is most accurately regarded […] as the founder of non-conceptualism” (Hanna (2008), 45); see also Hanna (2005). – What it is for the content of a mental state to be nonconceptual (or, for that matter, conceptual) is itself a question on which there are various different views; see Speaks (2005) for discussion. For the purposes of this paper, the gloss given in the text will be sufficient.

198

Thomas Land

such a situation possible. It is reasonable to expect that Kant’s theory combines elements which bear some affinity to those that figure in the contemporary debate. Each side, we may suppose, emphasizes one such element at the expense of the other. If this is right, it suggests that a proper understanding of Kant’s theory might be able to show how it is possible that this theory can so much as seem to combine elements whose counterparts in the contemporary debate appear to be incompatible. And this, in turn, should lead us to ask whether Kant’s own view, when properly interpreted, might not itself yield a third alternative – a position that cuts a middle path between conceptualism and nonconceptualism, revealing a way of integrating commitments that at first blush appear to exclude one another, yet, when properly thought through, can be shown to be in fact compatible. I shall argue that Kant’s theory of intuition does in fact cut such a middle path because it combines both a conceptualist and a nonconceptualist element. As a consequence, a proper appreciation of its merits will put us in a position to see that there is a genuine insight on the conceptualist side, yet there is also a kernel of truth in the nonconceptualist position. The assumption that the forced alternative with which we seem to be faced (of having to choose between conceptualism and nonconceptualism about perception) must be rejected. It is based on an inaccurate representation of the available options. Rather, a third option is available – one which combines the advantages of both positions, but avoids their shortcomings. I shall be concerned to show that Kant’s theory of intuition points the way towards this third option. I proceed as follows. I begin by arguing that Kant’s interest in the structure of perceptual content is motivated by a different question than the contemporary debate (§1) and explain what makes this question urgent for Kant (§§2 and 3). I then consider the answer that those nonconceptualists who explicitly draw on Kant – the Kantian Nonconceptualists, as I call them – are committed to ascribing to him and show that this answer fails (§4). Next, I turn to a version of contemporary conceptualism that I call Propositionalism and that likewise takes itself to offer an updated version of Kant’s theory of intuition. I argue that Propositionalism runs into equally serious problems (§§5 and 6). After pausing to reflect on what the failure of Propositionalism shows us about the nature of the problem with which Kant is concerned (§7), I go on to outline my own reading of Kant and show how his position contains the materials for a third alternative, which I call Kantian Conceptualism

Kantian Conceptualism

199

and which avoids the problems of Kantian Nonconceptualism and Propositionalism respectively (§8).

1. The Content Problem and the Question of Objective Purport My goal in this paper is to argue that Kant can be understood as articulating a kind of conceptualism that is in important ways different from the conceptualist position that figures in the contemporary debate about perceptual content. Kant’s brand of conceptualism makes it possible to be a conceptualist while also appropriating some of the insights that motivate contemporary nonconceptualism. This alternative construal of conceptualism becomes available because Kant’s overriding concern with regard to perception is distinct from the concern that is often taken to be at the center of the contemporary debate. The contemporary debate centers on what I call the Content Problem, while Kant’s overriding concern is with what I call the Question of Objective Purport. 3 The Content Problem concerns the similarities and dissimilarities in structure between the content of conceptual thought, on the one hand, and that of perceptual experience, on the other.4 By contrast, the Question of Objective Purport seeks to identify the conditions that must be satisfied for perceptual experience to have a content in the first place. Two considerations serve to illustrate the kinds of issues I am grouping here under the heading of the Content Problem. One of these concerns the relative determinacy of perception and conceptual thought respectively; the other concerns classificatory awareness. The first serves as the point of departure for a line of argument that has led some philosophers to adopt a nonconceptualist view of perceptual content. The sec3

4

This claim about the contemporary debate needs qualification. McDowell, the primary advocate of conceptualism, has, I think, from the start been concerned with the Question of Objective Purport as well. But since his work has been read as being primarily addressed to the Content Problem, it is fair to say that this is where the focus of the debate has been. See Boyle (2012) for helpful discussion. It should be noted that it has been disputed whether perceptual experience so much as has a content; see, for instance, Brewer (2006). I will not address this question in this paper, but simply follow the philosophers whose views I am discussing in assuming that perceptual experience does have content.

200

Thomas Land

ond has led other philosophers to conclude that some form of conceptualism must be true. Let us consider each of these in turn. The first consideration begins with the observation that perception appears to be more determinate than conceptual thought in a sense brought out by the following example: When I perceive something as red, I do not perceive it as merely having some shade of red or other, but as having a particular shade of red. Among the color-concepts I have, there are a number of concepts of different shades of red – such as ‘crimson,’ ‘maroon,’ ‘scarlet’, ‘vermillion’ – but reflection shows that each of these can be further differentiated. Each of these concepts of a shade of red itself covers a whole range of more determinate shades. This means that even if I introduce a concept of a shade of red (for instance, the concept ‘crimson’), this concept will still not suffice to single out the particular shade of red that I am perceiving hic et nunc. This concept, inasmuch as it is the concept of a (further determinable) shade, will still not be able to isolate the (fully determinate) shade of red I see here and now from an indefinite number of similar yet distinct shades. Concepts are, as this is often put, too coarse-grained to capture the finegrained content of our perceptual experience.5 From this one might conclude that, since the content of any color experience involves a single fully determinate shade, that content cannot be conceptual in nature. Considerations of this sort lead some philosophers to attribute to perception a distinctly nonconceptual kind of content.6 On the other hand, perception seems to involve sortal awareness; for instance, awareness of things as red. And this might lead one to think that perception involves classification and, therefore, concepts. Thus, to perceive something as red involves (so the argument goes) an application of the concept ‘__ is red’. It may involve other things besides this, but if my perception is a perception of something as red, then the concept ‘__ is red’ is part of its content. Considerations such as this 5 6

See, for instance, Peacocke (1998). The formulation seems to go back to Evans (1982), 229. Although this consideration is an important motivation for nonconceptualism, it is not the only one that figures in the debate. Arguments in favor of nonconceptualism about perception also appeal to, among other things, the distinctive situation-dependence of perception; the ability of perception to provide information to memory in a distinctive way; and the fact that perception is something that humans share with animals. For an overview, see Speaks (2005) as well as the essays collected in Gunther (2003).

Kantian Conceptualism

201

motivate conceptualists to attribute to perception a content that is of the same sort as the content of propositionally-structured thought. A concern with issues of this sort – the determinacy of perceptual experience as well as its classificatory character – lies at the heart of what I call the Content Problem. As I said, the contemporary debate between conceptualists and nonconceptualists in the philosophy of perception is to a large extent focused on this problem. By contrast, Kant’s discussion of perception in the First Critique is primarily motivated by a different kind of problem, what I call the Question of Objective Purport. The Question of Objective Purport concerns the conditions that must be satisfied for perception to be about objects – or, as we might also say, to have representational content – in the first place. This question asks for a specification of what must be the case for perception to have any representational content at all, whether it is conceptual or nonconceptual in nature. In a different idiom, this is the familiar question of intentionality or “aboutness,” applied to the specific case of perception. In a straightforward sense, the Question of Objective Purport is logically prior to the Content Problem. For the latter can only arise if it is presupposed that the former can be answered. To argue about the nature of perceptual content is to presuppose that perception has a content. The Question of Objective Purport concerns the conditions that must be in place for this to be the case. Put differently, for perception to so much as have a content (about whose nature one might then argue), we must ascribe to perception whatever characteristics are necessary to its being a contentful state. In seeking to identify these characteristics, the Question of Objective Purport is thus concerned with what must be the case for the Content Problem to arise in the first place. Since the Question of Objective Purport is in this sense prior to the Content Problem, consideration of the former may have consequences for how we must approach the latter. Thus, it may turn out that ascribing objective purport to perceptual experience forces us to conceive of its content in a particular way. In other words, our answer to the Question of Objective Purport may already constrain the range of options available for addressing the Content Problem. I shall argue that this is indeed so. More precisely, I shall argue that a full appreciation of the Question of Objective Purport will force us to recognize a new sense of ‘conceptual content,’ one that is different from the sense commonly attached to this phrase in the contemporary debate. Specifically, the claim is that we are entitled to attribute objective purport to perceptual experience only if we ascribe to it a content that is conceptual in this

202

Thomas Land

new sense. Recognizing this will allow us to see that the prevailing alternative between conceptualism and nonconceptualism does not exhaust the space of available options for solving the Content Problem.

2. The Heterogeneity of Thought and Perception Kant notes that there are certain fundamental differences between thinking and perceiving. We can bring these out by contrasting thought and perception with regard to three different aspects. Thought is essentially conceptual; it is what Kant calls discursive.7 This means that (1) concepts are general (as opposed to singular) representations; (2) concepts are classificatory: employing a concept in thought (or as Kant will say, in judgment) involves sorting things – that is, it involves thinking of something as being an instance of a general kind, which can, in principle, have other instances; (3) conceptual thought is logically articulated, where this means that it has a kind of structure that enables a thought to figure in inferences.8 By contrast, perception (what Kant calls intuition) according to him has the following three features: (1) Perception is singular: it is of individual objects, rather than of general kinds.9 (2) Perception is fully determinate: it presents objects to us whose properties are determinate, rather than determinable, in the sense brought out above by the example about color. (3) Perception is spatio-temporally structured: the objects we perceive have a location in space and time. To register these three fundamental (and interrelated) differences between thinking and perceiving, I will speak of their heterogeneity. Thoughts and perceptions, or, in Kant’s terms, judgments and intuitions, are heterogeneous to one another. Kant aims to do justice to their heterogeneity by attributing them to different capacities of the mind, which he calls understanding and sensibility respectively. The un7 8 9

See Kant (1998), A68/B93. Henceforth, all references to Kant (1998) will give only the pagination of the first (A) and second (B) editions. I have tacitly modified Guyer/Wood’s translation where appropriate. See Kant (1998), A320/B376 f, as well as Kant (1992), §1. An appreciation of this point is compatible with the claim, which is widely held, that perception involves classificatory awareness, hence concepts. As conceptualists will agree, acknowledging the singularity of perception in no way decides the issue in favor of nonconceptualism.

Kantian Conceptualism

203

derstanding is the capacity for thought. Sensibility is the capacity for intuition.10

3. Intuitions Have Objective Unity Conceptualists and Nonconceptualists about perception – at least those who regard themselves as Kantian – agree that perception is not merely a matter of having sense-impressions. Perception purports to be of objects. If by an impression we mean something momentary and perspectival, the agreed upon claim is that perceptual representations of objects cannot be just a matter of the momentary and perspectival affections of sensory consciousness that are afforded by impressions. Rather, the idea of an object is the idea of something that essentially outstrips such perspectival representations. An object is something that, for instance, can be perceived from a variety of different spatial and temporal vantage points, which are, moreover, systematically related to one another. We can express this point by saying that an object exhibits a certain kind of spatio-temporal unity. If perception is to be of objects, so the Kantian thought runs, it must contain a consciousness of this unity. That is, it must in some (perhaps implicit) way be part of the content of perception that what it represents are enduring three-dimensional objects. For instance, when I see a tomato in front of me, there is a sense in which my sensory impression is confined to the side of the tomato that is facing me. If what I perceive is indeed a tomato, however, the content of my perception is not just a surface. It is a solid, three-dimensional object, which (in the normal course of things) existed prior to my perceiving it and will continue to exist afterwards. And this is, at least implicitly, part of my perceptual consciousness. In perceiving the tomato, we might say, I am aware of perceiving a three-dimensional object with a temporal history. I do not take myself to be perceiving a mere surface.11 10 See e. g. Kant (1998), A19/B33 and A50 f/B74 f. It is worth noting that Kant takes it to be among his deepest insights to have properly appreciated the heterogeneity of perception and thought, as is indicated by his quip that Leibniz’s chief mistake was to “intellectualize” the senses, while Locke’s mistake was to “sensualize” conceptual thought. Cf. Kant (1998), A271/B327. 11 We might put the point by saying that I perceive the tomato in virtue of having, among other things, impressions of its facing surface. To explain how this is possible, that is, how the perception of, for instance, a facing surface enables the perception of the object whose surface it is, ranks among the central tasks of the philosophy of perception.

204

Thomas Land

When Kant characterizes an intuition as the singular representation of an object and distinguishes intuitions from mere sensations, this point, suitably elaborated, is what he has in mind.12 We can capture the central point here by saying that intuitions characteristically exhibit a certain kind of unity. Let us call this the objective unity of intuitions. This term is meant to capture the fact that intuitions must have the kind of unity, whatever it is, which enables them to be perceptual representations of objects. The thought then is that the objective unity of an intuition is not simply a function of the impressions the mind receives. There must be another element, over and above the impressions. And because impressions are what results from the mind’s being affected, this other element cannot in turn be a product of the mind’s being affected. It must have a different source. For the purposes of orienting ourselves in the discussion that is to follow, it will be helpful to have a concise overview of these Kantian commitments. This will enable us to revisit these commitments at crucial junctures of this paper and assess how our understanding of them has evolved. Let us summarize Kant’s view of perception, to the extent that it has come into view so far, by means of the following three theses: (1) The Sensibility Thesis: Intuitions are the representations of sensibility. (2) The Objective Unity Thesis: Intuitions have objective unity. (3) The Anti-Empiricist Thesis: Objective unity is not given. The Sensibility Thesis registers the fact that the ability to enjoy intuitions – that is, perceptions – requires a sensible capacity. A sensible capacity is a capacity to have representations of an object in virtue of having one’s senses affected by it. It is thus what Kant calls a receptive capacity.13 Saying that a sensible capacity is required implies that the capacity for intuitions is specifically distinct from the capacity for thought and judgment, which Kant characterizes as intellectual and thereby as non-sensible.14 Second, the Objective Unity Thesis expresses the point that intuitions, as perceptions of objects, are distinct from mere impressions in that they exhibit a kind of unity that mere impressions do not have. Finally, the third thesis makes the point that this unity, which accounts for the objective purport of intuitions, does not itself 12 For Kant’s insistence that intuitions are distinct from mere sensations, see Kant (1998), A320/B376 f. 13 Cf. Kant (1998), A19/B33 and A50 f/B74 f. 14 Cf. ibid.

Kantian Conceptualism

205

result from affections of the senses. In some yet to be determined sense, it has its source in the mind itself, “prior” to the senses’ being affected. I call it the Anti-Empiricist Thesis simply to register that it sets Kant’s position apart from a traditional Empiricism that seeks to account for perception wholly by reference to the affections of sensibility. The obvious question that must be answered by any theory of perception that is committed to these three theses is this: what is the source of the objective unity that intuitions exhibit? This is accepted on both sides of the debate between those contemporary conceptualists and nonconceptualists who seek to trace their positions back to Kant. Disagreement arises over what the correct answer to this question is. And it is precisely on this point that, as I shall argue, these two groups of contemporary Kantians are not only equally mistaken about Kant, but also each face their own philosophical problems. In the following, I shall consider in turn these attempts to inherit Kant and identify the problems that each of them raises.

4. Kantian Nonconceptualism A distinctive feature of Kant’s position is the view that sensibility, the capacity for intuition, has a pure form, which is constituted by what Kant calls the pure intuitions of space and time.15 This view contains a claim about the metaphysical nature of space and time, but I shall not be concerned with that aspect of it. Rather, I will focus on the account of our perceptual capacities that is implicit in it. It suffices for our present purposes to emphasize the following three characteristics of Kant’s position: First, by speaking of a form of sensibility, Kant has in mind certain features that are constitutive of all exercises of this capacity. That is, he is talking about features that are characteristic of intuitions as such: all possible intuitions exhibit these features. Second, to say that this form of sensibility is pure, or a priori, is to say that it does not derive from experience and is thus independent of any particular perceptions.16 It is, therefore, something that does 15 This view is famously elaborated in the part of the First Critique called the Transcendental Aesthetic. 16 Kant does not use the terms ‘pure’ and ‘a priori’ synonymously, but for our purposes we can abstract from their difference. See Kant (1998), B2 f. for a statement of this difference.

206

Thomas Land

not result from, or depend on, impressions. Third, when Kant introduces the idea of a form of intuition, he characterizes it as something in which sensory impressions can be “ordered in certain relations”17. As the Transcendental Aesthetic makes clear, these are spatio-temporal relations: in virtue of being given in the pure form of intuition, empirical intuitions – that is, perceptions – stand in spatio-temporal relations.18 This point can also be put by saying that the form of intuition confers on empirical intuitions a certain kind of unity, viz. spatio-temporal unity. Given these three characteristics of Kant’s position, one might think that the pure form of intuition is a suitable candidate for the role of that which accounts for the objective unity of intuition: It is in virtue of the pure form of intuition that perceptions exhibit spatio-temporal unity. The spatio-temporal framework Kant regards as the form of intuition both allows and suffices for sensory impressions to bear the kinds of systematic relations to one another that are necessary for impressions to enable representations of objects. It appears therefore that the form of intuition accounts for the objective unity of intuition. That this is indeed Kant’s position is the central claim of a view I call Kantian Nonconceptualism.19 According to Kantian Nonconceptualism, what accounts for the objective unity of perceptions is precisely the pure form of sensibility. Thus, Robert Hanna holds that the pure forms of intuition account for a subject’s ability to locate objects in space and time. This ability in turn grounds the objective purport of perception, the fact that it is about objects.20 On Hanna’s view, the capacity to locate objects in space and time is a sensible capacity. It belongs to what Kant calls sensibility. Since for 17 Kant (1998), A20/B34. 18 More accurately, it is what is perceived, rather than the perceiving of it, that exhibits spatio-temporal relations and is represented as exhibiting such relations. Acts of perceiving only have temporal, but no spatial properties, for Kant. 19 This view has been articulated forcefully by Robert Hanna; see the references in fn. 2. Hanna combines an exegetical claim about Kant with a philosophical claim addressed to the contemporary debate about perceptual content – the latter claim being that Kant’s view, as interpreted by Hanna, is the correct view about perceptual content. A similar exegetical view is defended in Waxman (1991) and, more recently, in Allais (2009). 20 See Hanna (2005), 273: “[…] Kant is saying that what determines our cognitive reference to the uniquely individual material objects of empirical […] intuitional representations, are the spatiotemporal features of those representations alone.”

Kantian Conceptualism

207

Kant sensibility is distinct from – indeed heterogeneous to – the understanding, it appears that this capacity does not in any way involve, or depend on, the understanding. And indeed, this is precisely Hanna’s view. He is explicit that the capacity for locating objects in space and time is independent of the understanding. According to Hanna, it is present even in creatures that do not possess a faculty of understanding.21 Since in a Kantian framework all conceptual representation must be attributed to the understanding, a representation of objects that does not involve any exercise of the understanding must be nonconceptual. Accordingly, the content of intuitions is, for Hanna, nonconceptual content: “intuitional cognitive content in Kant’s sense and nonconceptual cognitive content are identical […].”22 This claim is not meant to rule out that concepts can be brought to bear on intuitions, giving conceptual articulation to what Hanna characterizes as an intuition’s phenomenal content. But the point is that by themselves empirical intuitions – that is, perceptions – do not involve concepts. Accordingly, Hanna speaks of “Kant’s thesis about the cognitive autonomy of nonconceptual spatiotemporal representation […].”23 In what follows I shall argue that Kantian Nonconceptualism faces serious difficulties, both exegetically and philosophically. The main objection is that this view does not succeed in accounting for the objective unity of intuition. We can see why this is so if we reflect on the notion of objective unity. When I introduced this notion in §2, above, I said that intuitions exhibit objective unity because they are not just sensory impressions, but rather sensible representations of objects. Kant conceives of intuitions in this manner because he assigns a particular epistemic function to them. He characterizes this function as “giving objects to the mind” and contrasts it with the epistemic function of concepts, which is to “think objects.”24 And he makes it clear that the giving of objects is connected to knowledge of existence: in having an intuition of some object a, I know that a exists. I take this to imply, roughly, the 21 Thus, Hanna (2005), 260, claims that “it is possible to intuit an object while lacking concepts either globally or locally.” See also the suggestion, made in Hanna (2008), 58, that “[…] we should think of the representation of space and the representation of time as the necessary a priori subjective forms of egocentrically centered human and non-human animal embodiment” (emphasis added). 22 Hanna (2005), 260. 23 Hanna (2005), 280. 24 See Kant (1998), A19/B33 and A50/B74.

208

Thomas Land

following account of the epistemic function of intuition: Thought, just as such, is constrained only by the laws of logic – it must be consistent. But, obviously, not just any consistent thought amounts to knowledge. Knowledge is fundamentally the representation of what is, that is, of actuality.25 But what is actual, for Kant, cannot be known through mere thought. It requires, rather, intuition. It requires the act of sensibility, the receptive cognitive capacity. It follows that no characterization of the conditions of valid thinking can by itself amount to a characterization of the conditions sufficient for knowledge of objects. Rather, the conditions of having objects given to one in intuition must be included among the latter.26 If the function of intuition is to enable us to cognize the existence of objects about which we make judgments, then it must be possible for the very same object to be both an object of judgment and an object 25 Of course, there can be knowledge of what can exist, i. e. of possibility, as well. But knowledge of possibility is knowledge of possible existence. And it is one of the Critical Kant’s most deeply held commitments that the conditions of existence, both possible and actual, are independent of the formal-logical conditions of thought. This commitment sets Kant apart from Rationalists like Leibniz and Wolff, for whom the principles governing possible existence – most fundamentally, the Principle of Noncontradiction and the Principle of Sufficient Reason – are part of the formal-logical conditions of thought. 26 As comes out in a number of places throughout the Critique (e. g. Kant (1998), B135, B138 f., A235ff/B298 ff.; see also §§76 and 77 of Kant (2000)), Kant’s conception of our cognitive capacities is everywhere informed by an implicit contrast with the idea of an infinite mind. To an infinite mind the distinction between understanding and sensibility, or spontaneity and receptivity, does not apply. Kant therefore characterizes such a mind variously as an intuitive intellect and as a capacity for intellectual intuition. His point is that in an infinite mind the act of thinking itself supplies not just the representation of the general, but also of the particular falling under the general; that is, the representation of that for which finite minds like ours need sensible intuition. Put differently, the act of the infinite mind at once cognizes both a concept and its instances. It is therefore not dependent on a separate act for knowing that a given concept is instantiated. Since, in such a mind, the act of thinking an object is identical with the act of cognizing the object’s existence, the infinite mind is, for Kant, a creative mind: it creates the objects of its knowledge in the act of knowing them. Against the background of this idea it becomes apparent that it is a fundamental mark of human finitude that the objects of our knowledge exist independently of being known by us. It is because the objects of our knowledge have independent existence that we are dependent, for such knowledge as we can have, on a receptive capacity, that is, on sensibility. For an illuminating discussion of this point, see Engstrom (2006).

Kantian Conceptualism

209

of intuition.27 Put differently, if perception is to serve the function of providing objects for thought, then what can be perceived must be such that it can also be thought. We can express this point by saying that the formal object of perception must be identical to the formal object of thought. This locution is meant to indicate that the claim is general: it is not about particular objects of thought or perception, but about the characteristics of an object of thought or perception as such, or of an object of thought (or perception) insofar as it is an object of thought (or perception); or, to use Kant’s preferred locution, about an object “in general” (berhaupt). Moreover, not only must thought and perception have the same formal object, but it must also be no accident that they do. Thought and perception, or judgment and intuition, must be, as it were, made for each other. If, on the contrary, it was merely a contingent fact that in a given case an object of perception is also an object of thought, then perception could not have an epistemic function. For the concept of knowledge, as Kant understands it – viz. as the non-accidental, or “necessary”, agreement of a belief with its object28 – implies that it is precisely not an accident that what is given in perception is an object of knowledge; that is, something that can be the content of objectively valid judgments. Part of the point here is that on Kant’s view intuition is not merely a causal antecedent of judgment. Rather, in virtue of supplying what we might call the material condition of knowledge intuition stands in a rational relation to judgment. Using this terminology, the point of the preceding paragraph can be expressed by saying that it must be in the nature of thought and perception that particular judgments stand in rational relations to particular intuitions. I said above that for intuitions to be of objects at all they must exhibit a certain kind of unity, which I called objective unity. We can now 27 Throughout this paper, I use the terms ‘judgment’ and ‘thought’ interchangeably. Thus, these terms do not carry the senses familiar from Frege’s distinction between judgeable content, on the one hand, and assertoric force, i. e. Fregean judgment, on the other. Kant does not recognize Frege’s way of drawing the distinction between force and content. Rather, for Kant, every act of entertaining a propositional content is an act of judgment. What corresponds, in Kant’s classification of mental acts, to Frege’s notion of merely entertaining a thought is the notion of problematic judgment; cf. Kant (1998), A74 f/B99 f. For Frege’s distinction, see Frege (1998), §§2 – 5, as well as Frege (1993). 28 Cf. Kant (1998), B166.

210

Thomas Land

see that what makes the unity of intuition objective is the fact that it is identical to the unity of thought. What makes the content of a perception an object is the fact that this very thing can also be a content of belief and knowledge. This, at any rate, is Kant’s view, which I take to be encapsulated in his theory of the categories, or pure concepts of the understanding. Kant is explicit that intuitions have objective unity only to the extent that they exhibit the unity articulated by the categories.29 It follows that, when he speaks of that unity of an intuition in virtue of which it is the representation of an object, the sense of ‘object’ at issue is that of a formal object of thought.30 Intuitions, then, have objective unity to the extent that the formal object of intuition is also the formal object of thought. In light of these considerations, it is clear what the core of the objection to Kantian Nonconceptualism should be: If one locates the source of the unity of intuition in the pure form of intuition, then there is no reason to think that this unity is the objective unity Kant requires. This is because, as I have been concerned to show, what Kant means by objective unity is the unity of thought. In terms of Kant’s terminology of basic mental capacities, this is a unity that has its source in the understanding. Kantian Nonconceptualism, by contrast, traces this unity to sensibility. But this means that, at the very least, we would need an additional argument showing that the unity exhibited by the pure form of sensibility is the same as, or conforms to, the objective unity that has its source in the understanding. Kantian Nonconceptualists do not provide such an argument. Indeed, they do not feel the need for such an argument, because they hold that the unity of the form of intuition – spatial unity – is a self-standing kind of unity and is sufficient 29 Cf. e. g. Kant (1998), B143. 30 In a nutshell, the point here is this: The categories articulate the unity of an object in general. Only a representation that exhibits the unity articulated by the categories is a representation of an object in this sense. What sense of ‘object’ is this? The answer lies in the idea of a pure concept of the understanding (that is, a category), as Kant construes this notion: The categories are pure, or non-empirical, concepts because they derive from the nature of the understanding. This means that they are constitutive of thought as such. The reason for this is that the function of the categories is to characterize what falls under them as a formal object of thought; that is, as something that can, in principle, be the content of a thought. This is, I take it, the upshot of Kant’s notorious claim that the categories derive from the logical forms of judgment. The sense of ‘object’ at issue, therefore, is ‘formal object of thought’. – I develop this interpretation of the categories in detail in chapter 2 of Land (2010).

Kantian Conceptualism

211

to account for the essential characteristics of intuition. However, as I have been concerned to argue, not only is this not Kant’s view; it also involves a fundamental misunderstanding of the philosophical motivation of the requirement that intuition exhibit unity. It involves a misunderstanding, that is, of the motivation for what in the previous section I labeled Kant’s Anti-Empiricist Thesis, the thesis that objective unity is never given. It is instructive to relate this point to Sellars’ famous attack on the so-called Myth of the Given.31 For our purposes, the following characterization of the Myth will do.32 A conception of perceptual experience is committed to the Myth of the Given if it combines the following two features: On the one hand, it attributes to sensory experience an epistemic role, that is, it ascribes to experience the capacity to stand in relations of warrant or justification to belief; in short, the capacity to stand in rational relations. Yet, on the other hand, it does not attribute to experience the kinds of features that are constitutive of the capacity to stand in rational relations. In other words, a conception of experience is mythical if it ascribes a role to experience for which this conception does not equip experience. Anyone who wants to put forth a conception of perceptual experience that avoids the Myth of the Given, therefore, must opt for one of the following two alternatives: Either (1) concede that perceptual experience has an epistemic role, in which case it must be conceived as something that can stand in rational relations – as something that, in Sellars’ image, has standing in the logical space of reasons; or (2) conceive of perceptual experience in a way that explicitly precludes it from standing in rational relations – for instance, by conceiving it in merely causal terms. In the latter case, experience cannot play any epistemic role. It is not an option to describe perceptual experience in, for instance, merely causal terms, but then go on to attribute an epistemic role to it. To do this is precisely to fall into the Myth of the Given. Kantian Nonconceptualism does just this. That is, Kantian Nonconceptualism falls prey to the Myth of the Given because it seeks to combine features of both (1) and (2). On the one hand, it attributes an epistemic role to intuition. In particular, it wants to appropriate Kant’s no31 See Sellars (1997). 32 Giving an adequate characterization of the Myth of the Given is no easy task, since, as Sellars himself indicates, the Myth can take on a variety of different guises; see Sellars (1997), 14.

212

Thomas Land

tion of intuition as the immediate, singular representation of an object. Yet on the other hand, Kantian Nonconceptualism conceives of intuition in a way that precludes it from standing in a rational relation to belief. It does this by locating the source of what it claims is the objective unity of intuition exclusively in sensibility. Since sensibility is construed, on this view, as a non-rational capacity, this amounts to an attempt to give an account of a notion that has its home in a rational capacity – viz. objective unity, the unity of an object of thought and knowledge – in terms that are intelligible independently of any reference to rationality.33 To bring this point into sharper focus, it might be useful to contrast Kantian Nonconceptualism with two different views, both of which manage to avoid the Myth of the Given. One of these pursues the first of the two alternatives canvassed a moment ago; the other opts for the second alternative. The first view is what I shall argue is Kant’s actual view. Kant avoids the Myth of the Given by offering a conception of episodes of perceptual experience that gives them standing in the space of reasons. We are already in a position to see the outline of this conception: Kant, on my view, holds that the objective unity of intuition, in virtue of which it has epistemic standing, has its source in the capacity for thought and knowledge, that is, the understanding. How this outline is to be filled in is what I am concerned to show, at least in rough outline, in the remainder of this paper. By contrast, the second view avoids the Myth of the Given by denying that perceptual experience plays any epistemic role at all. This is the view of Donald Davidson, as encapsulated in his slogan that “nothing can count as a reason for holding a belief except another belief.”34 Following Sellars, Davidson agrees that for experience to play an epis33 In Kant’s view, this position amounts to a form of skepticism. Indeed, it is, as he puts it, “what the skeptic most desires” (Kant (1998), B168). The reason is this: Since sensibility is here conceived as intelligible independently of the capacity of thought, the fact that objects of intuition exhibit the same unity as objects of thought can only be a contingent fact. It can only be a matter of how sensibility happens to function in human beings. But in Kant’s view, this is incompatible with the idea of knowledge: if knowledge is possible, then it can be no accident that the very things that can be thought can also be perceived. It cannot be merely a contingent fact that the unity of an object of intuition is the same as the unity of an object of thought. I say a bit more about this issue in Land (2006). 34 Davidson (2001), 141.

Kantian Conceptualism

213

temic role it has to be able to stand in rational relations to the beliefs based on it. However, he goes on to argue, the only way for experience to do this is for it to be tantamount to belief. But this would be absurd. For, as all parties to the debate agree, just by itself perceptual experience – a Kantian intuition – does not amount to belief. From this Davidson concludes that, therefore, experience does not have any epistemic role. Its relation to belief is merely causal.35 As this contrast brings out, the problem with Kantian Nonconceptualism is that it tries to have it both ways: on the one hand, it ascribes to intuition an epistemic role; yet, on the other hand, it conceives of that which is supposed to enable intuition to play this role – its unity – in terms that are essentially non-epistemic. As a result, Kantian Nonconceptualism falls prey to the Myth of the Given. Seeing why Kantian Nonconceptualism fails in this regard is instructive. It should enable us to get a better sense of the nature of the challenge involved in avoiding the Myth. Since my goal in the remainder of this paper is to sketch a view that meets this challenge (and to argue that this is Kant’s view) I want to end this section by articulating what the failure of Kantian Nonconceptualism teaches us about the nature of the problem we are confronted with. At the end of the preceding section, I characterized Kant’s view of perception by means of the following three theses: (1) The Sensibility Thesis: Intuitions are the representations of sensibility. (2) The Objective Unity Thesis: Intuitions have objective unity. (3) The Anti-Empiricist Thesis: Objective unity is not given. I then introduced Kantian Nonconceptualism as a view that incorporates these three theses. The inadequacy of this view that has just come to light indicates that we must modify at least one of them. In particular, we must rule out that the resulting position collapses into the Myth of the Given. Properly modified, then, this triad of theses should yield a characterization, in outline form, of a Kantian view of perception which avoids the mistakes of Kantian Nonconceptualism. Since the problem with this position is that it locates the source of the objetive unity of intuition in sensibility, while conceiving of sensibility as a ca35 I have introduced Davidson’s view here merely in order to provide a perspicuous overview of the options available at this point. But since my primary concern is with the Kantian views I have been discussing, I will have nothing further to say about Davidson who does not himself endorse a Kantian position.

214

Thomas Land

pacity that is not essentially related to rationality, what we require is a modification of the thesis that addresses the issue of the source of objective unity. This is the Anti-Empiricist Thesis. As we are now in a position to see, this thesis, as it currently stands, is not specific enough. In its current formulation is merely says that the unity of intuition is not itself given in intuition. But this is compatible with saying that this unity has its source in the pure form of intuition. What we need to do, then, is to block this option. More precisely, we need a formulation of the AntiEmpiricist Thesis that makes clear that the unity of intuition has its source in the understanding. Only if the unity of intuition derives from the understanding will intuition have standing in the space of reason. This will have the twofold result that we desire: The attribution of an epistemic role to intuition will be vindicated and our Kantian theory of perception will avoid the Myth of the Given. Kant characterizes the understanding as spontaneous. Since, according to him, this is its most fundamental feature, it seems most natural to formulate our modified version of the Anti-Empiricist Thesis in terms of a claim about spontaneity and to re-label it accordingly.36 Our modified Kantian theory of perception thus issues in the following triad of claims: (1) The Sensibility Thesis: Intuitions are the representations of sensibility. (2) The Objective Unity Thesis: Intuitions have objective unity. (3’) The Spontaneity Thesis: Objective unity has its source in spontaneity. Jointly, these three theses contain the outline of a view of intuition that avoids the Myth of the Given. On this view, the unity of intuition, in virtue of which it is the immediate singular representation of an object, is the unity of thought. Because this unity has its source in the capacity for thought, it is precisely the kind of unity that is constitutive of rational relations. The challenge now is to fill in more of the details of such a view. In particular, we need an account of how a kind of unity that belongs to the intellectual capacity can come to characterize the exercises 36 Space does not permit me to defend the claim that spontaneity is the most fundamental characteristic of the understanding. For an indication of how such a defense would go, consider that for Kant the infinite intellect is spontaneous but not discursive: it does not represent by means of abstract concepts. The discursivity of human understanding thus marks its specific difference from the infinite understanding, while its spontaneity reflects its membership in the genus ‘understanding’. For further discussion of this issue, see Land (2006).

Kantian Conceptualism

215

of sensibility, where, according to Kant, the latter is a distinct capacity from the intellect.

5. Propositionalism I have argued that Kantian Nonconceptualism does not provide an adequate account of the objective unity of intuition because it does not locate the source of this unity in the right place. Kantian Nonconceptualism construes the objective unity of intuition as a unity that is intelligible by reference to sensibility alone. This unity can be made intelligible, the claim is, without any mention of the understanding. The argument in the preceding section has shown why this is mistaken. The objective unity of intuition must have its source in the understanding, the capacity for thought and knowledge, which Kant characterizes as spontaneous. The Spontaneity Thesis gives expression to this point. Kant’s famous slogan that intuitions without concepts are blind is often thought to make a closely related point.37 It is taken to make the point that, without some involvement of concepts, that is, without the involvement of the understanding, intuitions cannot play an epistemic role. Now, concepts, Kant says at one point, can only be applied in judgment.38 Accordingly, Kant characterizes the understanding both as a capacity for concepts and as a capacity for judgment. If the understanding is a capacity for judgment, it is natural to think that the objective unity we are interested in, and which we said must have its source in the understanding, is the unity of judgment. This thought in turn invites the thought that intuitions possess objective unity only to the extent that they exhibit the same unity as judgment. And this is taken to entail that we must think of an intuition as having the same structure as a judgment. I call Propositionalism any view that endorses this line of thought. We can give a more precise characterization of Propositionalism if we introduce the notion of sensible synthesis. Sensible synthesis is Kant’s term for the act of mind that is responsible for the objective unity of intuitions.39 As he puts it, sensible synthesis unifies sensible 37 Cf. Kant (1998), A51/B75. 38 “Of these concepts, however, the understanding can make no other use than to judge by means of them” (Kant (1998), A68/B93). 39 See e. g. Kant (1998), A77/B102: “However, the spontaneity of our thought requires that this manifold [i.e. the manifold of intuition, T.L.] first be gone

216

Thomas Land

manifolds in such a way that they come to exhibit objective unity. Using this terminology, we can characterize Propositionalism as the view that the act of sensible synthesis is an act of judgment.40 Accordingly, Propositionalism can be thought of as the view that results when we supplement the three theses we reviewed a moment ago with a fourth thesis to the effect that the act of sensible synthesis is an act of judgment. This fourth thesis can also be expressed by saying that all acts of the understanding are acts of judgment. Propositionalism therefore amounts to the view that ascribes to Kant the following four claims: (1) The Sensibility Thesis: Intuitions are the representations of sensibility. (2) The Objective Unity Thesis: Intuitions have objective unity. (3’) The Spontaneity Thesis: Objective unity has its source in spontaneity. (4) The Judgment Thesis: All acts of the understanding are acts of judgment. As is apparent from this summary presentation, I use the label ‘Propositionalism’ for this view because it construes synthesis as an act of judgment and, as a result, ascribes the propositional structure of judgment to intuitions. In the contemporary debate about perceptual content, the most prominent advocate of Propositionalism has been John McDowell.41 Although an intuition is, according to McDowell, a distinct kind of episode from judgment, it exhibits the same structure as judgment: An intuition is like a judgment in that it is composed of concepts and therefore has the structure of a proposition. Since it has the structure of a

through, taken up, and combined in a certain way in order for a cognition to be made out of it. I call this act synthesis.” – Note that Kant recognizes more than one kind of synthesis. For instance, judgment for Kant is an act of synthesis. I use the term ‘sensible synthesis’ to pick out that type of synthesis which is responsible specifically for the unity of intuition. 40 By contrast, Kantian Nonconceptualism can be characterized as endorsing the following two claims: (i) sensible synthesis and judgment constitute distinct kinds of synthesis, and (ii) sensible synthesis is not an act of the understanding. As we have just seen, Propositionalism denies both of these tenets. Below, I will attribute to Kant a view that affirms (i), but denies (ii). 41 See the texts cited in fn. 1, above. In more recent work, McDowell has abandoned the view I am calling Propositionalism and endorsed a position that is much closer to the view I defend below; see, in particular, McDowell (2009b).

Kantian Conceptualism

217

proposition an intuition expresses a thought.42 However, in contradistinction to judgment, an intuition does not involve a commitment as to the truth-value of the thought it expresses. We should note immediately that McDowell, in characterizing his position in this manner, uses the term ‘judgment’ in a different way than Kant does. McDowell uses the term in much the same way as Frege. In terms of the distinction between assertoric force and propositional content, a judgment for McDowell has both force and content. To judge, on this view, is to assert a content. It is thus to go beyond the sort of mere entertaining of a content that is independent of any commitment as to its truth-value. By contrast, for Kant, any act of entertaining a content with propositional structure is a judgment. Judgment, for Kant, does not necessarily carry assertoric force. Rather, Kant regards the mere entertaining of a content, which leaves open its truth-value and which McDowell (and Frege) would treat as distinct from judgment, as a case of what he calls problematic judgment. When McDowell says that an intuition has the same structure as a judgment but does not itself amount to a judgment, he is employing these terms in the Fregean way just indicated. If we translate the claim into Kant’s terminology, then we have to say that, on this view, what McDowell calls an intuition is a problematic judgment, while that which McDowell calls ‘judgment’ is an assertoric judgment. Notwithstanding the difference in terminology, however, it should be clear how McDowell’s view bears on the problem with which we are concerned. An intuition, according to McDowell, involves an exercise of concepts, and it is this which confers on it objective purport. The way in which concepts are exercised in intuition is, as far as its logical structure is concerned, exactly the same as in judgment. The position thus amounts to a form of Propositionalism, since what confers objective unity on an intuition here is the very fact that it exhibits the structure of a judgment. This means that McDowell is committed to the claim that sensible synthesis is a kind of judgment, using this term in Kant’s way.

42 McDowell frequently puts this point by adapting a locution from Sellars, according to which we should think of perceptual experience as “so to speak, making […] a claim”; see, for instance, McDowell (2009a), 10. Sellars’ original formulation is found in Sellars (1997), 39.

218

Thomas Land

6. Problems for Propositionalism I introduced Propositionalism as an alternative to Kantian Nonconceptualism which offers a solution to the problem I raised for the latter view. Propositionalism appears to be more plausible both as an interpretation of Kant and as a view about perceptual content. However, as we shall see, the solution that Propositionalism offers is purchased at a cost. For, as I am about to argue, the characteristic feature of this view (the thesis that sensible synthesis is identical to judgment) itself leads to serious difficulties. I will first present two exegetical difficulties. Discussing these will set the stage for consideration of the philosophical problems raised by Propositionalism. Before considering the textual difficulties faced by Propositionalism, however, it is worth pausing briefly to note that there is also textual evidence that can make this view look exegetically attractive. I have in mind, in particular, a passage from the so-called Metaphysical Deduction, which is frequently quoted by McDowell, as well as a passage from the B-edition version of the Transcendental Deduction. Towards the end of the Metaphysical Deduction, Kant says: The same function which gives unity to the different representations in a judgment, also gives unity to the mere synthesis of different representations in an intuition, which, expressed generally, is called the pure concept of the understanding.43

This passage can seem to be saying that the same act that unifies representations in a judgment also accounts for the unity of an intuition. The act that unifies representations in a judgment is the act of judgment. The implication seems to be that the synthesis that accounts for the unity of an intuition is likewise an act of judgment. In the Transcendental Deduction Kant appears to be making a similar point. Once again employing the notion of a mental function,44 Kant says the following:

43 Kant (1998), A79/B104 f. 44 Kant defines a function as “the unity of the act of ordering different representations under a common one” (Kant (1998), A68/B93). Although it is not fully clear what this means, the context in which the definition is given, as well as the passages quoted in the text above, suggest that the notion of a function is closely related to the idea of a mental act that accounts for unity among representations.

Kantian Conceptualism

219

That act of the understanding, however, through which the manifold of given representations (whether they be intuitions or concepts) [is unified], is the logical function of judgments.45

Again, the claim can appear to be that judgment is the act that is responsible for, among other things, the unity of an intuition; in other words, that the act of sensible synthesis is an act of judgment. These passages, then, can appear to lend textual support to Propositionalism. We will return to them after we consider the problems faced by Propositionalism. To begin with, there are the following two textual problems. Kant repeatedly characterizes intuition as “that representation which can be given prior to all thought.”46 If judging is a kind of thinking, as Kant clearly holds, this claim entails that intuition can be given prior to judgment. And this is in direct conflict with Propositionalism. The Propositionalist holds that judgment is constitutive of intuition. But this implies that intuition cannot be given prior to judgment. Accordingly, the Propositionalist reading is in tension with the text. A second instance of direct textual evidence against Propositionalism comes from a passage at the opening of the Transcendental Dialectic in which Kant discusses error and illusion. He expresses his allegiance to the traditional view that the senses never err and explains that the reason for this is that error is only in judgment, and that the senses do not judge. Kant says, “In a representation of the senses (because it contains no judgment at all) there is no error, either”47. If an intuition, that is, a representation of the senses, does not contain judgment, then the act of synthesis that is responsible for the objective unity of intuition cannot be an act of judgment. Again, the textual evidence is in conflict with the Propositionalist view.48

45 46 47 48

Kant (1998), B143. Kant (1998), B132. Kant (1998), A294/B350. A Propositionalist might respond to both of these points by saying that, on his view, an intuition does not itself amount to a judgment; it merely has the same structure as a judgment. But this response rests on a Frege-inspired use of the term ‘judgment’. Since, as I pointed out in the preceding section, for Kant, anything that has propositional structure is a judgment, this move is not open to the Propositionalist once we use the term ‘judgment’ in Kant’s sense. Obviously, this is the sense that ought to matter in interpreting the textual evidence I just cited.

220

Thomas Land

In addition to these textual problems, Propositionalism faces the following philosophical problem. Propositionalism ascribes to perception a content that is thoroughly conceptual. As we saw in §1, a general characteristic of at least some concepts of sensible properties is that they are determinable. For instance, as I also discussed in §1, color-concepts such as ‘__ is red’ are determinable because they leave room for further specification of the color at issue. Colors come in shades, but the concept ‘__ is red’ does not specify which shade of red is at issue. Perception, on the other hand, is a sensible representation of an object. Objects have determinate properties. This holds, in particular, of an object’s color: No object can be red without exhibiting a particular shade of red. The content of perception, therefore, appears to be more determinate – more fine-grained, as this is often put – than conceptual representation allows for. And if this is right, it would be a mistake to think of perceptual content along Propositionalist lines. Some Propositionalists have sought to address this worry by appealing to the notion of a demonstrative concept. Let us call the resulting position Demonstrative Concept Propositionalism. This position holds that we can conceptually represent, say, the determinate color of an object by forming a concept that exploits the perceptual presence of the object to fix the identity of the property it denotes. The linguistic expression of such a concept would involve a demonstrative expression; as, for instance, in the sentence ‘The cube in front of me has that color’, which we should imagine as being accompanied by a pointing gesture. Whether or not this concept gets expressed in language, the perceptual presence of the object makes it possible for the perceiver to form such a concept. And this allows her to represent conceptually the degree of determinacy possessed by perceptual experience. We can put the point by saying that the content of such a concept is fixed, not descriptively, but demonstratively.49 Let us grant for the sake of argument that demonstrative concepts succeed in capturing the determinacy of perceptual content. Still, in the context of our discussion, this idea is problematic. It is problematic because it is not clear whether the Demonstrative Concept Propositionalist is entitled to it if she wants her position to remain a species of Prop49 This response to the fineness-of-grain objection was first proposed by McDowell in his (1996), 56 – 60, and refined in response to an objection by Peacocke in McDowell (1998b). It is also deployed by Brewer (1999).

Kantian Conceptualism

221

ositionalism.50 More specifically, there is reason to believe that a view which incorporates the notion of a demonstrative concept no longer counts as Propositionalism in the sense introduced above. For recall that Propositionalism is characterized in part by its commitment to the thesis that the act of sensible synthesis is an act of judgment. For Kant, an act of judgment is, paradigmatically, an act of predicating one concept of another.51 Moreover, Kant holds that concepts are essentially determinable. There is no such thing, in his view, as a concept that has a maximal degree of determinacy.52 Rather, Kant holds that complete determinacy can be represented only by means of intuition, not by means of concepts. This is an aspect of Kant’s commitment to the heterogeneity of sensibility and understanding.53 It follows that Kant would regard a demonstrative concept as a representation that is not purely conceptual, but rather involves intuition. Taken by itself, this conclusion does not necessarily present a problem for the Demonstrative Concept Propositionalist. Indeed, the Demonstrative Concept Propositionalist might represent herself at this point as simply in agreement with Kant. If she takes this line, then the whole point for her of introducing demonstrative concepts is to have a device that exploits the determinacy of perception but, nonetheless, functions as a concept. In the dialectical context of our discussion, how50 In light of McDowell’s commitment to demonstrative concepts, it might be argued that his position is not adequately classified as a version of Propositionalism. To decide this issue we would need to determine whether those aspects of McDowell’s view that lead him to endorse demonstrative concepts have deeper roots in his philosophy than the aspects of his view that commit him to the thesis that the act of sensible synthesis is an act of judgment. I do not take a stand on this here. It suffices for my present purposes to bring out what is problematic about the Propositionalist aspects of McDowell’s position, especially because these have been most influential in the reception of his work. 51 This means that Kant’s conception of judgment differs in important respects from a modern, post-Fregean conception of thought and judgment. On the modern view, a judgment is paradigmatically a compound of a Fregean object and a Fregean concept; something whose form is Fa. On the modern conception, the idea that, in its basic form, a judgment involves predicating a concept of another concept does not make sense. By contrast, Kant’s conception of judgment is a version of the traditional model, according to which a judgment is, or expresses, a “comparison of ideas.” For a statement of this model, see, e. g., Arnauld/Nicole (1996), 82. I give a more detailed discussion of this issue in Land (forthcoming). 52 See Kant (1992), §§11 and 15. 53 See §1, above.

222

Thomas Land

ever, the fact that a demonstrative concept is now being held to be something that essentially involves intuition represents a problematic departure from the original terms of the debate. It represents a problematic departure because it undermines the characteristic Propositionalist claim that sensible synthesis must be understood as an act of judgment. The following consideration brings out why this is so. The notion of a demonstrative concept exploits the fact that intuition is fully determinate. In so doing, it exploits one of the features that make sensibility heterogeneous to the understanding: complete determinacy cannot be represented by conceptual means alone, according to Kant. This brings out something important about the manner in which the Demonstrative Concept Propositionalist is committed to thinking about sensible synthesis: on the one hand, sensible synthesis is conceived as an act of judgment, that is, an act of the understanding; yet, on the other hand, because sensible synthesis is construed as essentially involving demonstrative concepts, it is thereby conceived as an act that exhibits specifically sensible features. It is, therefore, an act that combines characteristics which, for Kant, belong to two heterogeneous capacities. The problem with this is that it is not at all clear how such a thing is possible. Given the heterogeneity of understanding and sensibility, it seems that an act must either have the characteristics of the one or the characteristics of the other. But it cannot have both. To say that it does amounts to denying the heterogeneity of these two capacities. If this is right, then the point we have now reached in the argument of this paper can be characterized by saying that Propositionalism faces a dilemma: Either the Propositionalist operates with a conception of judgment that is straightforwardly intelligible as an act of the understanding and thus respects the heterogeneity of understanding and sensibility. In this case, she cannot appeal to the idea of a demonstrative concept. As a result, she cannot account for the determinacy of perception. Alternatively, the Propositionalist tries to enrich her position in the manner envisaged by the Demonstrative Concept Propositionalist, trying to account for the determinacy of perception by appeal to the notion of a demonstrative concept. But in this case she helps herself to a conception of judgment that undermines the heterogeneity of understanding and sensibility. Within a Kantian framework, there seems to be no room for such a conception of judgment. As long as Propositionalism seeks to represent itself as a species of Kantianism, it is obvious that neither horn of the dilemma is acceptable. Nonetheless, there is an important insight here. This can be brought out

Kantian Conceptualism

223

if we attend to the way in which Propositionalism, in its unmodified form, construes sensible synthesis. Recall that sensible synthesis, for Kant, is the act of mind that is responsible for the objective unity of intuition. Originally, Propositionalism’s distinctive commitment was supposed to consist in the claim that this act is an act of judgment. This commitment is motivated by the lesson learned from the failure of Kantian Nonconceptualism, viz. that the objective unity of intuition must be construed as having its source in the understanding, the capacity of thought. Against this background, we can extract from the dilemma faced by Propositionalism the following requirement for a more promising solution to our problem: We need something that plays the role of a demonstrative concept without undermining the heterogeneity of judging and perceiving. That is, we need something that (i) is an act of the understanding, (ii) captures the complete determinacy – hence the distinctly sensory nature – of an intuition, and (iii) preserves the heterogeneity of understanding and sensibility. Although it may seem impossible to meet all three requirements simultaneously, I want to suggest that this impression depends on the central Propositionalist commitment. If sensible synthesis is identical to judgment, then it is indeed impossible to meet all three requirements. But if this commitment is given up, then the prospects are not quite as bleak. Giving up the central commitment of Propositionalism opens up the conceptual space for thinking of judgment and sensible synthesis as distinct acts. If we can find our way to a conception of sensible synthesis that preserves this thought, but at the same time makes sensible synthesis intelligible as an act of the understanding, then we have at least a candidate for the role of something that meets the three requirements. To be sure, for this to be so much as a coherent possibility, we require an account of the understanding that allows for exercises of this capacity which do not take the form of judgments. And we have as yet no idea whether such an account is available. But the very fact that this is now our question suggests that we have attained a deeper grasp of our problem. This grasp can be articulated by saying that Propositionalism overreaches in its reaction to the failure of Kantian Nonconceptualism. Although it is motivated by the right insight – the objective unity of intuition must have its source in the understanding – it goes too far when it takes this to imply that sensible synthesis is identical to judgment. The proper way to develop this insight is, rather, to find a way of holding on to it while also acknowledging that sensible synthesis is distinct in character from judgment.

224

Thomas Land

7. The New Shape of the Problem At the outset of this paper I observed that a central characteristic of Kant’s conception of intuition lies in the fact that an intuition is the representation of an object – rather than, say, a mere impression – and therefore exhibits a distinct kind of unity. A number of contemporary philosophers of perception who see themselves as working in a Kantian tradition defend a view of perception that incorporates this feature. Moreover, these philosophers also appeal to Kant in arguing that the objective unity of intuitions does not itself have its source in intuition. In other words, it is a shared feature of these views that the kind of unity on account of which a perception is the representation of an object is not itself something that is given in perception. There is disagreement, however, on how to characterize this unity and, as a result, on where to locate its source. Kantian Nonconceptualism construes the objective unity of intuitions as something that can be accounted for entirely by the resources provided by sensibility. On this view, the unity of an intuition is a spatio-temporal unity, which does not itself derive from perception because it has its source in what Kant calls the pure form of intuition. At the same time, having its source in the form of intuition makes this unity independent of conceptual thought. Since it originates in sensibility, the objective unity of intuitions is something that, in the view of the Kantian Nonconceptualist, is intelligible without any reference to understanding or reason. I objected that a view of this kind is not entitled to treating the unity of intuition as objective. It is a central aspect of Kant’s view that the unity exhibited by intuition must, in the relevant respect, be the same as the unity of thought. More precisely, it must be the unity that characterizes an object of cognition as such. By locating the unity of intuition in sensibility, Kantian Nonconceptualism fails to meet this requirement. To meet the requirement, the source of intuitional unity must be located in the understanding, i. e. the capacity for thought. Propositionalism does this. According to Propositionalism, an intuition has objective unity because it has the same structure as judgment. We found, however, that there are serious objections to Propositionalism. In essence, Propositionalism fails to preserve the heterogeneity of sensibility and understanding.54 The result we have now reached can be summar54 My discussion of this in §6 focused on the determinacy of perceptual content, which is one aspect of its distincively sensory character. But similar objections

Kantian Conceptualism

225

ized by saying that while Kantian Nonconceptualism keeps perception too far removed from thought, Propositionalism moves perception and thought too close together. In Kantian terminology, the result of my discussion of Propositionalism was that sensible synthesis, the act that accounts for the objective unity of intuition, must not be construed as a species of judgment. To do so is the fundamental mistake of Propositionalism. If we want to avoid this mistake, however, we are confronted with a serious problem. The criteria of adequacy on a viable Kantian philosophy of perception that we have thus far uncovered require us, on the one hand, to ascribe sensible synthesis to the understanding, and yet, on the other hand, to regard sensible synthesis as distinct from judgment. This is a problem because the understanding seems to be defined by Kant as the capacity for judgment – and indeed it seems eminently plausible to see the capacity for conceptual thought as closely tied to the notion of judgment. If this is what the understanding is, but if, at the same time, sensible synthesis is distinct from judgment, then what grounds do we have for regarding sensible synthesis as an act of the understanding? It can look as if there is no room to maneuver here. Our discussion of the shortcomings of Kantian Nonconceptualism showed that it is an essential feature of Kant’s position that intuition exhibits the same unity – in the relevant respect – as thought. The familiar Kantian claim that there is a set of pure concepts, which are valid of objects, amounts precisely to this. Any view, therefore, that is genuinely Kantian must locate the source of the unity of intuition in the understanding. And this entails that it must regard synthesis as an act of the understanding. Call this the Objective Unity Requirement. On the other hand, we just saw in the discussion of the shortcomings of Propositionalism that a plausible Kantian account of perception must preserve its distinctively sensible features and thereby make room for a robust sense of the heterogeneity of perception and thought. Call this the Heterogeneity Requirement. Jointly, these two requirements generate the problem I have been delineating: They demand an account of sensible synthesis, on which this synthesis is an act of the understanding, but at the same time distinct from judgment. Sensible synthesis must preserve the distinctively sensory nature of intuition, yet be traceable to the capacity for conceptual thought. to Propositionalism could be developed by reference to other aspects of this character, for instance, the fact that perceptual content is not articulated in the way in which propositions are.

226

Thomas Land

Above I said that the distinctive feature of Propositionalism lies in its commitment to the following thesis: (4) The Judgment Thesis: All acts of the understanding are acts of judgment. We can now succinctly express the challenge that my discussion of Propositionalism has raised by adding to this the following two theses: (5) The Synthesis Thesis: Sensible synthesis is an act of the understanding. (6) The Distinctness Thesis: Sensible synthesis is not an act of judgment. Jointly, these three theses are inconsistent. As I have already indicated, to remove the inconsistency we should focus on (4) and on the conception of the understanding it expresses. In the remainder of this paper, I shall argue that Kant does not view the understanding exclusively as a capacity for judgment and that, therefore, (4) ought to be rejected. Rather, judgment must be conceived as only one type of act among several of which the understanding is capable. More specifically, judgment and sensible synthesis must be conceived as two types of act that are specifically distinct, but generically identical. A conception that exhibits this structure makes it possible to see both judgment and synthesis as acts of the same capacity, while allowing that they each have different characteristics. And this opens up the logical space for arguing that sensible synthesis is an act of the understanding which is distinguished from judgment by the fact that it exhibits distinctively sensible characteristics. However, such a conception will be available only if it is possible to give a plausible characterization of the common genus. Thus, what this solution demands is a different characterization of the understanding: a characterization that is at once more abstract than the characterization ‘capacity for judgment’ and that makes it possible to see judgment as one specific kind of exercise of the understanding.

8. Sketch of a Solution I will now offer a sketch – indeed, a very rough sketch – of an account of the understanding, on which this capacity can plausibly be seen as admitting of two distinct kinds of exercise, one in judgment, the other in sensible synthesis. The central contention of this account is that Kant’s

Kantian Conceptualism

227

characterization of the understanding as a capacity for judgment ought not to be taken to constitute the most fundamental characterization he wishes to give of this capacity. Rather, Kant himself offers a more fundamental characterization; and this characterization makes it possible to see judgment as only one of two kinds of exercise of which the understanding is capable. An account of the understanding along these lines conceives of this capacity as having a structure analogous to that of a genus which contains two distinct species. The first task in establishing the viability of such an account, therefore, is to explain what entitles us to attribute to the understanding a structure of this kind. I will say more about this in a moment. For now let me point out that the second task is to show that one of the two species, viz. sensible synthesis, exhibits distinctively sensible features. My focus here will be on the first task. The question we confront, then, is how to think about what the understanding is. Both Propositionalism and Kantian Nonconceptualism are committed to the view (expressed in the Judgment Thesis) that our hold on the notion of understanding is through the notion of judgment. That is, the understanding is, at bottom and most fundamentally, a capacity for judgment. This means that every act of this capacity must be conceived as a judgment of some kind (though perhaps a defective, merely purported judgment). By contrast, the proposal I am about to make asks us to think of judgment as one of two distinct acts of the understanding. Clearly, on such a view it will not do to say that the understanding is a capacity for judgment. We need a different characterization, one that allows us to see both judgment and sensible synthesis, conceived here as distinct from judgment, as acts of a single capacity. And Kant, I think, shows us what such a characterization might look like. In working my way towards this characterization, I want to begin by pointing out a number of passages in which Kant himself appears to ascribe to the understanding a structure of the kind I have indicated, that is, the structure of a genus containing two distinct species. The first of these passages was quoted above, in §6.55 It is the well-known passage from A79/B104 f, in which Kant says that the same function accounts for both the unity of a judgment and the unity of an intuition. Whatever one’s preferred interpretation of the passage, it is undeniable that we have here an instance of the structure I have described. 55 See p. 218.

228

Thomas Land

Another passage occurs in a footnote appended to §26 of the B-edition version of the Transcendental Deduction: It is one and the same spontaneity, which in the one case, under the title of imagination, and in the other case, under the title of understanding, brings combination into the manifold of intuition.56

Kant does not literally speak here of two species of the understanding. Rather, he characterizes imagination and understanding as species, respectively, of spontaneity. Nonetheless, the passage supports my proposal. For, first, it supports the structural point: Kant speaks of a genus containing two distinct species. Secondly, there are a number of reasons for thinking that he uses the term ‘understanding’ here in a narrow sense, in which it refers to the use of concepts in judgment, while using ‘spontaneity’ as referring to the understanding in the wide sense, that is, the capacity which can be exercised in both judgment and sensible synthesis. The first two of these are textual: Kant elsewhere characterizes spontaneity as being coextensive with understanding.57 Moreover, when he first introduces the notion of imagination at issue in the quoted passage, he characterizes it as “an action of the understanding on sensibility”.58 Both of these passages suggest that the passage from B162n should not be taken to imply that the imagination is not a species of the understanding. The third reason is philosophical and concerns the argumentative context of the passage. The passage occurs at the end of the Transcendental Deduction, where Kant is concerned to argue that nothing can be given to the senses which does not fall under the categories. His argument turns on the claim that intuition itself, simply in virtue of its spatio-temporal form, requires an act of synthesis in accordance with the categories; an act he calls ‘synthesis of apprehension.’ And this act of synthesis is precisely what he characterizes as an act of the imagination 56 Kant (1998), B162n. 57 Cf. e. g. Kant (1998), A51/B75: “[…] the capacity to bring forth representations from itself, however, or the spontaneity of cognition, is the understanding.” 58 Here is the full passage: “Now since all our intuition is sensible, the imagination […] belongs to sensibility. But insofar as its synthesis is still an exercise of spontaneity, which is determining and not, like sense, merely determinable, […] the imagination is to that extent a capacity for determining sensibility a priori; and its synthesis of intuitions, in accordance with the categories, must be the transcendental synthesis of imagination. This synthesis is an action of the understanding on sensibility […].”(Kant (1998), B151 f., emphases omitted).

Kantian Conceptualism

229

in the passage from B162n.59 Arguably, this maneuver can establish the conclusion that the synthesis of apprehension confers categorial unity on intuition only if it is an act of the capacity whose pure form the categories articulate: that is, of the understanding.60 Both of the passages I have quoted, then, provide evidence for the claim that the understanding for Kant possesses a complex structure, which can be characterized by analogy with a genus containing two distinct species. My goal here is to argue that we should conceive of the capacity of judgment as constituting one of these species, and the capacity for exercising concepts in perception as the other species. It is clear, however, that such a proposal will be viable only if we can give a plausible characterization of the common genus. In particular, such a characterization must make it possible to comprehend what makes both species so many different ways of employing concepts. In other words, the characterization of the genus should put us in a position to form a conception of what it might be to apply concepts in perception in a way that does not consist in the entertaining of a proposition. Again, my aim here is merely to provide an outline of such a characterization. It will be useful to begin with Kant’s conception of judgment. Central to this conception is the claim that a judgment is “[a function] of unity among our representations”.61 It is essential to judgment, for Kant, that this act unifies representations. My proposal is that this feature of Kant’s conception of judgment provides a foothold for a characterization of the understanding that is independent of the notion of judgment, at least to the extent that we can make sense of the idea of an exercise of the understanding which does not take the form of a judgment. So the idea is that we can characterize the understanding as a capacity for the representation of a certain kind of unity, and that 59 The full text of the footnote runs as follows: “In this manner it is proved: that the synthesis of apprehension, which is empirical, is necessarily in accord with the synthesis of apperception, which is intellectual and wholly contained a priori in the category. It is one and the same spontaneity, which in the one case, under the title of imagination, and in the other case, under the title of understanding, brings combination into the manifold of intuition” (Kant (1998), B162n.). 60 This claim is based on an interpretation of the argument of the Transcendental Deduction that I cannot defend here. For discussion, see Land (2010), 280 – 288. 61 Kant (1998), A69/B94. It is a measure of the centrality of this claim that Kant’s derivation of the Table of Judgments (and thus also, indirectly, the derivation of the Table of Categories) depends on it. See Wolff (1995) for discussion.

230

Thomas Land

this idea is sufficiently abstract for it to provide the needed independence from judgment.62 It is crucial that the notion of unity at issue here is a very specific one. As Kant makes explicit in the Transcendental Deduction, a judgment is a function of not just any kind of unity among representations, but, specifically, of the kind of unity he calls the synthetic unity of apperception: […] a judgment is nothing other than the manner of bringing given cognitions to the objective unity of apperception.63

Saying precisely what the synthetic unity of apperception is would require an account of Kant’s doctrine of apperception, and that is not something I can provide here. Instead, my strategy is to pick out one central feature of this doctrine and to argue that this provides enough of a foothold for making the case that a characterization of the understanding is available which has the required independence from judgment. In the B-edition version of the Transcendental Deduction, Kant motivates his discussion of apperception as follows: The understanding is, most fundamentally, a capacity for combining representations in a certain way. In particular, it is a capacity for combining representations by means of a prior representation of unity. And this prior representation of unity is the synthetic unity of apperception.64 62 The following point of clarification is essential: My strategy is not to argue that we can extract from Kant a conception of the understanding according to which this capacity is intelligible in complete independence from the notion of judgment. The claim I seek to establish is weaker: It is only to argue that Kant provides us with a characterization of the understanding that allows for a kind of exercise of this capacity which does not take the form of judgment. This claim is compatible with the claim (to which I take Kant to be committed) that it is essential to the understanding that one of the ways in which it can be exercised is judgment. In terms of the terminology of genus and species which I have been employing, this point can be put as follows: The characterization of the genus may well be such that, in the final analysis, it is not fully intelligible in isolation from the fact that one of its species is judgment. But again, this is compatible with the claim that such a genus contains another, distinct species besides judgment. 63 Kant (1998), B141. Although he speaks of ‘objective unity’ rather than ‘synthetic unity’ in the quoted passage, it is clear that these expressions are equivalent; cf. Kant (1998), B139 f. 64 See, in particular, the following two passages: “Combination […] is solely an operation of the understanding, which is itself nothing further than the faculty

Kantian Conceptualism

231

Since in the present context ‘combination’ is just another word for ‘synthesis’, we can extract from this the claim that the understanding is, fundamentally, a capacity for a kind of synthesis which proceeds by means of a prior representation of unity. And it is this characterization, I suggest, that serves as the genus of which both judgment and sensible synthesis can intelligibly be seen to be species. As the preceding paragraph suggests, it is Kant’s view that the prior representation of unity is the synthetic unity of apperception. A full account of Kant’s conception of the understanding would have to explain what this means. But for present purposes, we can abstract from this feature of the view and work with the idea that the understanding is a capacity for synthesis by means of a prior representation of unity. A full account would also need to explain what the connection is between this kind of synthesis and the possession of concepts. Although I attribute to Kant a conception of the understanding on which this capacity can be exercised in an act that is distinct from judgment, I also think that Kant is committed to thinking of the understanding as the capacity of concepts. Every act of this capacity, therefore, is an act of concept-use. Consequently, a full account would need to show that the idea of a synthesis by means of a prior representation of unity is the idea of a kind of concept-use. However, I will not be able to provide such an account here. I shall now argue that the characterization of the understanding as a capacity for synthesis by means of a prior representation of unity is sufficiently independent from the notion of judgment to provide the logical space for the idea of an exercise of the understanding which does not take the form of judgment. In other words, I shall argue that the characterization of the understanding as a capacity for synthesis by means of a prior representation of unity provides us with the tools needed to reject the Judgment Thesis. My argument has two stages. In the of combining a priori and bringing the manifold of given representations under the unity of apperception […]” ( Kant (1998), B134 f.); “But in addition to the concept of the manifold and of its synthesis, the concept of combination also carries with it the concept of the unity of the manifold. Combination is the representation of the synthetic unity of the manifold. The representation of this unity cannot, therefore, arise from the combination; rather, by being added to the representation of the manifold, it first makes possible the concept of combination” (Kant (1998), B130 f.). The final sentence of the second passage clearly says that combination requires a prior representation of unity. For discussion, see Land (2010), 206 – 215.

232

Thomas Land

first stage, I argue that judgment can be seen as a species of this capacity. In the second stage, I argue that sensible synthesis can be comprehended as a distinct species of this capacity. A central element of Kant’s conception of judgment is the idea that there are a number of elementary logical forms of judgment, which are catalogued in the so-called Table of Judgments.65 This Table is presented by Kant as a characterization of the very capacity for judgment. To possess this capacity, he holds, is to be able to make judgments exhibiting these logical forms. This obviously requires some grasp, however implicit, of these forms themselves. It requires a grasp, for instance, of the fact that a judgment of the form ‘All F are G’ is contradicted by a judgment of the form ‘Some F are not G’ as well as the fact that two judgments of the form ‘All F are G’ and ‘All G are H’ jointly entail a judgment of the form ‘All F are H’. That judging requires a grasp, in this sense, of the elementary logical forms of judgment shows that judgment satisfies the description of a synthesis by means of a prior representation of unity. For, first, we can take for granted that judgment is a kind of synthesis. Secondly, if it is appropriate to say that the elementary logical forms of judgment constitute a representation of unity, then the fact that the ability to make judgments requires a mastery of these forms shows that this representation of unity is a prior representation of unity. That is, it is a representation of unity that is presupposed by any exercise of the capacity to judge. Finally, the elementary logical forms of judgment collectively constitute a representation of unity in the following sense: In virtue of their logical form, judgments stand in inferential relations to one another, as the example in the preceding paragraph shows. And in virtue of being inferentially related, judgments form a kind of whole. Furthermore, Kant famously correlates the categories with the elementary logical forms of judgment. His view seems to be that every judgment, in virtue of its logical form, represents its object as falling under the categories. But the categories articulate the concept of an object in general.66 A judgment, therefore, represents its object as an instance of the concept of an object in general. If this is right, we can say that the elementary logical forms of judgment constitute a representation of unity in the sense that everything that is represented through them is represented as falling under the concept of an object in general. 65 See Kant (1998), A70/B95. 66 See Kant (1998), B128.

Kantian Conceptualism

233

This is only a very quick sketch. But I hope it has become clear how one might argue that judgment can be understood as a species of the capacity to unify representations by means of a prior representation of unity. Now, the same abstract structure is exemplified, I think, in Kant’s theory of spatial representation.67 Spatial representation, too, can be understood as a species of synthesis by means of a prior representation of unity. In §4, above, I discussed a number of features of Kant’s doctrine that sensibility has a pure form, and that this pure form is constituted by the representations of space and time. One such feature is the idea that the representation of space serves to confer a certain unity on empirical intuitions (that is, perceptions). In virtue of being given in the pure form of intuition, empirical intuitions stand in spatial relations to one another.68 We can say, therefore, that the representation of space is the representation of a certain kind of unity that empirical intuitions exhibit: In virtue of having spatial form, every empirical intuition is related to every other empirical intuition. Call this the spatial unity of intuition. It is Kant’s view that representing the spatial unity of an intuition requires an act of synthesis, which he calls synthesis of apprehension. As is apparent from Kant’s characterization of this synthesis, it proceeds by means of a prior representation of unity. And this allows us to conceive of it as a species of a genus another species of which is judgment. However, before I can discuss Kant’s characterization of the synthesis of apprehension, I need to introduce a bit of technical terminology. Kant thinks of space as a magnitude. A magnitude is a manifold of qualitatively identical parts. And what Kant calls the categories of quantity (unity, plurality, totality) jointly articulate the concept of a magnitude. Now, as Kant argues in the Axioms of Intuition, the synthesis of apprehension proceeds in accordance with the concept of a magnitude and, for this reason, serves to confer on the manifold of intuition the synthetic unity of apperception.69 But this means that the concept of a 67 The same holds for temporal representation. But I think for present purposes the case of spatial representation is more straightforward. 68 Again, by ‘intuition’ is meant here the content rather than the act of intuiting. 69 “[…] appearances […] cannot be otherwise apprehended, that is, taken up into empirical consciousness, than through that synthesis of the manifold, whereby the representations of a determinate space or time are generated, that is, through the composition of the homogeneous and the consciousness of the synthetic unity of this manifold (homogeneous). The consciousness of the manifold homogeneous in intuition in general, insofar as the representation of an object first

234

Thomas Land

magnitude here functions as a representation of unity; specifically, it serves as an abstract representation of the kind of unity that is characteristic of space, the unity of a whole of qualitatively identical parts. It follows from this that the synthesis of apprehension is a synthesis which proceeds by means of a prior representation of unity, viz. the concept of a magnitude. Kant’s doctrine of spatial representation, then – specifically, his theory of synthesis in accordance with the concept of magnitude – allows us to see the synthesis of apprehension as satisfying the description I proposed as Kant’s most basic characterization of the understanding. And since the synthesis of apprehension is identical to what I have been calling sensible synthesis, this holds also for sensible synthesis. However, showing that sensible synthesis satisfies a description that is also satisfied by judgment does not amount to showing that sensible synthesis is a distinct kind of act from judgment. In other words, to establish the generic identity of sensible synthesis and judgment is not yet to establish their specific difference. This task, therefore, still remains. Rather than discharging this task in full, however, all I can do here is to lay out a strategy for confronting it. As is emphasized by Kantian Nonconceptualists, Kant holds that the pure intuition of space is a non-conceptual representation. It is non-conceptual in the sense that it exhibits features that cannot be represented by means of concepts. Specifically, space is a manifold of qualitatively identical parts. This, as we have just seen, is what makes space a magnitude. And on the theory of concepts that Kant espouses, it is not possible to represent such a structure solely by means of concepts.70 In particular, it is not possible to represent such a structure by means of applying concepts in judgment.71 For this reason, the kind of synthesis that is responsible for the representation of spatial unity must be distinct from judgment.72 becomes possible by means of it, however, is the concept of a magnitude (quanti)” ( Kant (1998), B202 f.). – For discussion of this passage and the Axioms of Intuition as a whole, see Sutherland (2005). 70 For discussion of this point, see Sutherland (2004). 71 This needs qualification: It is possible to represent such a structure by means of judgments which employ concepts that are derivative from the pure intuition of space. But such concepts depend for their very possibility on the availability of a synthesis in accordance with the concept of magnitude. For the purposes of this argument, therefore, we can exclude them from consideration. 72 Indeed, as we saw above on p. 33, Kant characterizes this synthesis as an act of the imagination. Imagination is the capacity to “represent an object in intuition

Kantian Conceptualism

235

Although implementing this strategy would require significantly more work, it is apparent that the strategy draws on just those features of Kant’s theory of sensibility that are emphasized by Kantian Nonconceptualists. And this is just as it should be. For the claim that there is an exercise of the understanding that is distinct from judgment is motivated by the desire to accommodate what I take to be the main insight of Kantian Nonconceptualism, viz. an appreciation of what above I called the Heterogeneity Requirement. The doctrine of sensible synthesis I have proposed here seeks to develop this insight in the context of a view that also attempts to do justice to the Objective Unity Requirement, and this necessitates the idea of an act of the understanding which exhibits specifically sensible features. I have now provided an outline of the solution I take Kant to offer to the problem posed by his doctrine of sensible synthesis. Obviously, a lot more would need to be said to make this solution compelling. But we can at least see what shape it would take. This puts us in a position to see that those passages in the Critique that seem to support a Propositionalist interpretation do not in fact force such an interpretation on us. They are equally compatible with the reading I have just sketched. In §6, above, I quoted two such passages. In concluding this section, I want briefly to return to one of them and indicate how it supports my reading.73 The passage at issue reads as follows: The same function which gives unity to the different representations in a judgment, also gives unity to the mere synthesis of different representations in an intuition, which, expressed generally, is called the pure concept of the understanding.74

The Propositionalist takes this passage to be saying that the act that gives unity to the different representations in an intuition is judgment. But we can now see that the “same function” of which Kant speaks may be a function of the understanding without being thereby limited in its exercise to acts of judgment. Indeed, the passage precisely mirrors the structure of the view I have just outlined: to say that a single “function” accounts for the unity of two distinct kinds of representations is analoeven without its presence” (Kant (1998), B151). This, too, suggests that sensible synthesis exhibits a distinctively sensible character. 73 It will be apparent from what I am about to say how I think the other passage quoted in §6 (as well as other similar passages in the Critique) should be interpreted. 74 Kant (1998), A79/B104 f.

236

Thomas Land

gous to saying that there is a single genus containing two distinct species. It follows that the passage lends at least as much support to the view that what Kant calls the “functions of the understanding”75 can be exercised in two distinct ways as it does to the Propositionalist reading. In §6 I also cited two passages that are in direct conflict with Propositionalism. The gist of these passages was that intuition can be given prior to judgment. In contrast to Propositionalism, the view I have just sketched can accommodate these passages, while still maintaining that the unity of intuition derives from the understanding. This gives it a clear advantage over Propositionalism on textual grounds.

9. Conclusion According to the interpretation I have developed, Kant’s solution to the problem posed by his theory of synthesis depends on a number of other views he holds. Most importantly, it depends on his traditional conception of judgment, as well as his theory of spatial representation. Since these views diverge sharply from the views that participants in the contemporary debate about perception take on these matters, it is not clear that Kant’s solution to the problem he confronts can simply be incorporated wholesale into a position that aspires to be a serious contender in the contemporary debate. The historical distance is, after all, significant and should not be overlooked, even if there is much insight to be found in Kant’s position. I do not, therefore, claim that Kant’s position can advance the debate between conceptualists and nonconceptualists in the philosophy of perception without further modification. But I do wish to claim that Kant’s position offers the resources for such a step. In particular, Kant’s position brings out the significance of one aspect of the debate that tends to receive comparatively little attention. This is what I have called the Question of Objective Purport. The Question of Objective Purport concerns the conditions that perception must satisfy if it is to be conceived as a cognitive capacity. To say that perception is a cognitive capacity is to say that it is a capacity that has a role in the acquisition of knowledge. Following Kant, I have argued that there are, in particular, two requirements that perception must meet in this regard. First, what I have called the formal object of perception must – in the relevant respect – be the same as the formal object of 75 Kant (1998), A69/B94.

Kantian Conceptualism

237

knowledge: the kinds of things one can perceive must be the kinds of things that can be an object of knowledge. Only if this is the case can perceptual episodes legitimately be regarded as sensory presentations of objects and thus play the epistemic role Kant assigns to them. Second, a viable conception of perception must be such that it is no accident that the first requirement is met. I think it is here that Kant’s distinctive contribution lies. In Kant’s hands, this requirement takes the shape of developing a conception of how thought relates to perception which shows the capacity for thought to be responsible for the unity, or structure, that characterizes perception. Propositionalism takes this requirement to heart. But, as we have seen, it does so in a manner that jettisons the distinctly sensory character of perception. In its hands, perceptual content is deprived of the very determinateness that is supposed to be its central characteristic. I have suggested that Kant’s position avoids these mistakes. Kant seeks to meet his second requirement by recognizing a kind of act of the understanding that is not propositional in structure, yet is an act of the same capacity that is also responsible for judgment. Kant calls this act ‘sensible synthesis,’ and he holds that sensible synthesis determines sensibility to have the same unity as thought without thereby undermining its specifically sensory character. A careful examination of Kant’s position thus permits us to extract from it the following lesson for the contemporary debate about perception: The relation between the sensory and intellectual capacities must be conceived in such a way that the very nature of the sensory capacity of a rational being itself exhibits the fact that it is the capacity of such a being. This demand has two aspects, which correspond to the respective shortcomings of Kantian Nonconceptualism and Propositionalism: On the one hand, the perceptual capacities of a rational being must be conceived as distinct in kind from the perceptual capacities of a non-rational being. Only in a rational being can perception present objects that can also be the contents of conceptually articulated thoughts.76 Put differently, perception in a rational being must not be conceived as intelligible in 76 For the purposes of this discussion, I am assuming that the capacity for conceptual thought is tied to rationality. Thus, non-rational animals are not capable of conceptual thought. But this does not rule out that they may be capable of other ways of representing their surroundings. For all I have said, it may even be appropriate to say that some animals have knowledge of their surroundings, as long as we are clear that the type of knowledge at issue here must be distinct from the type of knowledge made possible by conceptual thought.

238

Thomas Land

isolation from thought. On the other hand, the perceptual capacities of a rational being must preserve their sensory character. They must not be too closely assimilated to intellectual capacities. Perception is not itself a kind of thought, even if it exhibits the same unity as thought. What Kant can teach us, then, is that in our philosophy of perception we must steer a middle course between the Scylla of making the objective purport of perception unintelligible and the Charybdis of intellectualizing the senses. It is in this that Kant’s contribution to the debate between conceptualists and nonconceptualists about perception lies. Accordingly, ‘Kantian Conceptualism’ ought to be the name, not of a view merely about the fine-structure of perceptual content, but rather of the character of the entire framework within which any viable version of such a view is to be located.77

References Allais (2009): Lucy Allais, “Kant, Non-Conceptual Content and the Representation of Space”, in: Journal of the History of Philosophy 47, 383 – 413. Arnauld/Nicole (1996): Antoine Arnauld/Pierre Nicole, Logic or the Art of Thinking, Cambridge. Boyle (2012): Matthew Boyle, “Essentially Rational Animals”, in: G. Abel/J. Conant (eds.), Berlin Studies in Knowledge Research, vol. 2, Rethinking Epistemology 2, Berlin/New York. Brewer (1999): Bill Brewer, Perception and Reason, Oxford. Brewer (2006): Bill Brewer, “Perception and Content”, in: European Journal of Philosophy 14, 165 – 181. Engstrom (2006): Stephen Engstrom, “Understanding and Sensibility”, in: Inquiry 49, 2 – 25. Evans (1982): Gareth Evans, The Varieties of Reference, Oxford. Frege (1993): Gottlob Frege, “Der Gedanke”, in: Gottlob Frege, Logische Untersuchungen, G. Patzig (ed.), Göttingen, 30 – 53. Frege (1998): Gottlob Frege, Begriffsschrift und andere Aufstze, Hildesheim. Gunther (2003): York Gunther (ed.), Essays on Nonconceptual Content, Cambridge/MA. Hanna (2005): Robert Hanna, “Kant and Nonconceptual Content”, in: European Journal of Philosophy 13, 247 – 290. Hanna (2008): Robert Hanna, “Kantian Non-Conceptualism”, in: Philosophical Studies 137, 41 – 64. Kant (1992): Immanuel Kant, “The Jäsche Logic”, in: Kant, Lectures on Logic, Cambridge, 521 – 640. 77 For helpful comments and suggestions I am grateful to Ian Blecher, Matthew Boyle, James Conant, Robert Pippin, Sebastian Rödl, and Daniel Sutherland.

Kantian Conceptualism

239

Kant (1998): Immanuel Kant, Critique of Pure Reason, transl. Paul Guyer and Allen Wood, Cambridge. Kant (2000): Immanuel Kant, Critique of the Power of Judgment, transl. Paul Guyer and Eric Matthews, Cambridge. Land (2006): Thomas Land, “Kant’s Spontaneity Thesis”, in: Philosophical Topics 34, 189 – 220. Land (2010): Thomas Land, Kant’s Theory of Synthesis, Ph.D. dissertation, University of Chicago. Land (forthcoming): Thomas Land, “Intuition and Judgment: How Not to Think about the Singularity of Intuition (and the Generality of Concepts) in Kant”, in: S. Bacin/A. Ferrarin/C. La Rocca (eds.), Kant and Philosophy in a Cosmopolitan Sense, Berlin. McDowell (1996): John McDowell, Mind and World, Cambridge/MA. McDowell (1998a): John McDowell, “Criteria, Defeasibility, and Knowledge”, in: J. McDowell, Meaning, Knowledge, and Reality, Cambridge/MA, 369 – 394. McDowell (1998b): John McDowell, “Reply to Commentators”, in: Philosophy and Phenomenological Research 58, 403 – 431. McDowell (2009): John McDowell, Having the World in View: Essays on Kant, Hegel, and Sellars, Cambridge/MA. McDowell (2009a): John McDowell, “Sellars on Perceptual Experience”, in: J. McDowell, Having the World in View: Essays on Kant, Hegel, and Sellars, Cambridge/MA, 3 – 22. McDowell (2009b): John McDowell, “Avoiding the Myth of the Given”, in: McDowell (2009), 256 – 272. Peacocke (1998): Christopher Peacocke, “Nonconceptual Content Defended”, in: Philosophy and Phenomenological Research 58, 381 – 388. Sellars (1997): Wilfrid Sellars, Empiricism and the Philosophy of Mind, Cambridge/ MA. Speaks (2005): Jeff Speaks, “Is There a Problem about Nonconceptual Content?”, in: Philosophical Review 114, 359 – 398. Sutherland (2004): Daniel Sutherland, “Kant’s Philosophy of Mathematics and the Greek Mathematical Tradition”, in: Philosophical Review 113, 157 – 201. Sutherland (2005), Daniel Sutherland, “The Point of the Axioms of Intuition”, in: Pacific Philosophical Quarterly 86, 135 – 159. Waxman (1991): Wayne Waxman, Kant’s Model of the Mind: A New Interpretation of Transcendental Idealism, Oxford. Wolff (1995): Michael Wolff, Die Vollstndigkeit der kantischen Urteilstafel: Mit einem Essay ber Freges Begriffsschrift, Frankfurt/Main.

Ginsborg on the Normativity of Perceptual Representation1 Arata Hamawaki I. Between Conceptualism and Non-Conceptualism? Generally, in having a perceptual experience I am not just (perceptually or experientially) aware of an object, I am (perceptually or experientially) aware of an object as being a certain way. For example, I look out the window and see a tree in the yard. I may see the tree in the yard as a tree, or as just a thing with a particular shape and color, at such and such a distance from other objects with other shapes and colors. If I am looking at it through the mist, I may see it as a human figure. To speak of seeing as a “seeing-as” is just to say that whenever I perceive something – or for that matter imagine, or think, of something – there is a “way” or a “how” in which I am perceiving it. You and I both perceive a long branch-like object on the ground: I see it as a stick, you see it as a snake. Yet a third person may see it as a kind of ambiguous something, something that looks both snake-like and stick-like. What each of us perceives is the same – it is the “way” we each perceive it that differs. It is the “way” one perceives what one does that constitutes the perception’s representational content. Perceptual awareness has the following two structural features. First, in perceptual awareness there is, or at least purports to be, a “something” – a “this” – that I am aware of. It is, in that sense, always “of” an object, even when there is no object 1

Earlier versions of this paper were delivered at the Wittgenstein Workshop at the University of Chicago in June of 2008 and at Tel Aviv University in July of 2008. For helpful feedback and discussion of these topics I thank Dafi Agam-Segal, James Conant, David Finkelstein, Eli Friedlander, Keren Gorodeisky, Kelly Jolley, Eric Marcus, Dylan Sabo and Yaron Sendorovich. Hannah Ginsborg gave the two papers on which I focus here at the Sawyer Seminar Conferences on Non-Discursive Representation at the University of Chicago, while I was a fellow there. I thank her for discussing her papers with me on those occasions.

242

Arata Hamawaki

there to be perceived. Second, there is always a “way” in which one is (perceptually) aware of the object – even when the “way” is not entirely determinate, or unambiguously resolved. Thus the representational content of perceptual awareness has the following form: O-as-F. 2 It is important to distinguish awareness of the object as F, the representational content of the awareness, from the object’s actually being F. For one thing I can be aware of an object as F, even though the object is not actually F, as when I am perceptually aware of something as a snake, even though in actuality it is just a stick. Thus even when the perceptual awareness is veridical, features of the content of my awareness must be distinguished from features of the object of my awareness.3 In that sense what properties the object possesses is one thing; how I perceive it to be is another. The mere fact that I see something that is a hut doesn’t of itself make available to me the perceptual awareness of what I see as a hut. However, by being perceptually aware of something as a hut, the fact that the object is a hut becomes available to my thought and judgment: I am able thereby to perceive and judge that the object is a hut. That is, although I can represent O as F, without O actually being F, there is nonetheless an internal relation between my representing O as F, and O actually being F. For if I represent O as F, and, as it turns out, O is actually F, then what I am thereby aware of is: Os being F. This is expressed by the form of representation. Representation has as the form of its content: that the object is thus and so. So given that my perceptual awareness has the form that the object is thus and so, then when the perceptual awareness is veridical, what the perceptual awareness makes available to my thought and judgment is the fact that the object is thus and so. In that sense there can be a kind of identity between how the object is represented and how the object actually is.4 The representational content of perceptual awareness is what enables me 2 3

4

What I am calling the representational content, the “way”, corresponds very roughly to what Frege, in the context of linguistic meaning, called “sense”, and the “what” to what Frege called “meaning”. It would be a mistake to think that this implies that we are never perceptually aware of the stick itself and the properties it actually has. It is this mistake that lies at the bottom of the infamous argument from illusion. The content of perceptual awareness is not itself an object of awareness: it is what allows our awareness to have the object that it does. The content is not a perceptual intermediary. I am here presenting a notion of representational content that is articulated by John McDowell. See his McDowell (1994).

Ginsborg on the Normativity of Perceptual Representation

243

to be, as John McDowell puts it, perceptually aware of “environmental facts”. How is this possible? How can representational content – which is after all a feature of my states of awareness, or more generally of my mental states, be in this sense, “transparent” to the world, enabling me to be directly aware of the corresponding facts – when my awareness is veridical? However one goes about answering this question one must do so by avoiding treating the representational content as itself an object of our awareness, whether this awareness is perceptual or purely intellectual. For if representational content were itself the object of our awareness, we would need to explain how it is that our awareness of the representational content itself could have it as its object. But if it is representational content that makes any object at all available to our cognition then we would be off on a regress.5 Furthermore, if the representational content were itself an object of our awareness, whether the awareness is veridical would be a further fact about the representational content itself. But if that is so, the representational content would not contain within it the possibility that it is true or false. Awareness of the representational content and awareness of the content as true would be separable and distinct states. The representational content would not make available the awareness of the corresponding fact. Thus the representational content of my awareness must consist in my having an understanding of the conditions under which the awareness would be veridical or true.6 What, then, does explain or account for the representational content of my perceptual awareness? It is commonly supposed that there are two 5

6

If a second representational content that made available the first representational content to our thought itself figures as an object of our awareness then we would need to posit yet a further representational content that would explain how that representational content could be an object of awareness. But if this new representational content is itself an object of awareness, then we would need to posit yet a fourth representational content. And so on. Is there any awareness that does not take this form? The awareness of sensory impressions is the traditional candidate for such a form of awareness, a form of awareness that would count as pre-conceptual. However, one could raise doubts about even such apparently primitive forms of awareness. Am I not aware of sensations as sensations of color, or sound, or pain? If there is such an as-component to my awareness then wouldn’t such awareness have representational content as well? Indeed mustn’t it have representational content in order for us to ascribe veridicality to it at all? Is there any such thing as non-representational awareness?

244

Arata Hamawaki

possible sorts of answers that one can give to this question: conceptualist and non-conceptualist. The conceptualist holds that I am able to be perceptually aware of an object as F, or as entering into relation R, just in case I have the concept of F, or R. For the conceptualist what I can think is prior to what I can sense or be perceptually aware of in the sense that the former constrains the latter rather than the other way around. The non-conceptualist holds that the representational content of one’s perceptual awareness is to be accounted for by the operation of “sub-rational” or “lower-level” cognitive powers that we share with non-rational animals, cognitive powers such as those that are involved in differential sensory discrimination, and in what Hume would have called the associative powers of the imagination. For the non-conceptualist what determines the representational content of our perceptual awareness are what would be objects of non-representational awareness, such as what have sometimes been called “sense-data”, or causal relations between such contents of non-representational awareness, such as would be involved in Humean association, or the causal linkages between contents of non-representational awareness and the world, as would be involved in sensory discrimination – to name a few of the candidates. These candidates have in common that they are all sub-rational phenomena, states of mind that do not require the mediation of our rational or conceptual powers, states of mind that we share with the non-rational animals.7 7

In the Jsche Logik, Kant writes, “If a savage (Wilder) sees a house from a distance for example, with whose use he is not acquainted, he admittedly has before him in his representation the very same object as someone else who is acquainted with it determinately as a dwelling established for humans. But as to form, this cognition of one and the same object is different in the two. With one it is mere intuition, with the other it is intuition and concept at the same time” (Translation of: Kant (1800), 41). This can suggest that Kant sides with the non-conceptualist, for he identifies a notion of representation that doesn’t seem to depend on our possession of concepts – it is this that he seems to be calling “intuition”. If so this would leave two interpretative options. Either an intuition involves the representation of an object as such and such, without the representational content depending on concepts, or intuition is simply the representation of a “something”, but does not include within it a representation of the “something” as anything in particular. Suppose that the latter is Kant’s view. If so, is the representation of a “something” by intuition determined “non-conceptually”, as I characterized “non-conceptual” above? It doesn’t seem that it could be, for intuition is not awareness of a sensory impression, but of objects in space and time. Indeed it seems that for Kant intuition requires pure concepts, if not empirical ones. Kant’s actual position with

Ginsborg on the Normativity of Perceptual Representation

245

Hannah Ginsborg in a recent series of intriguing, subtle and challenging articles proposes a kind of compromise position between these two views, one that promises to capture what is appealing in both the conceptualist and non-conceptualist approaches, while discarding undesirable baggage that each typically brings in its train. What she finds appealing about the non-conceptualist approach is its due respect for, as she puts it, “the primitive character of perception relative to thought and judgment.” She writes, “Conceptualism seems to get the relation between conceptual activity on the one hand, and perception on the other, the wrong way around. Surely, it might seem, I do not need to be able to entertain thoughts involving the concept dog or apple in order to have apples or dogs presented to me perceptually. On the contrary, it is precisely because I have perceptions of dogs or apples that I come to be capable of entertaining thoughts with the corresponding conceptual content.”8 Ginsborg’s complaint with conceptualism, at least as it is typically understood, is the unavailability to it of a natural explanation of our possession of empirical concepts. Namely, the explanation that we have the empirical concepts we do because of the perceptual experiences we actually have had. That natural explanation seems unavailable to conceptualists, since they seem to be committed to the view that I can only have the perceptual experience as of dogs and as of apples because I already have the concepts dog and apple. It could be said that Ginsborg sees this as a kind of truth in non-conceptualism.

8

regard to the conceptualism/non-conceptualism debate would seem then to be rather subtle. He acknowledges a form of representational awareness, namely intuition, that does not depend on our having empirical concepts – in that sense he is a non-conceptualist. However, intuition isn’t free of a dependence on concepts altogether. Empirical concepts are needed for the representation of something as a cup or what not, but not all representation is of that form. There are also forms of representational awareness of something as an object in space and time, and that requires the categories, although not empirical concepts. Since all representation presupposes intuition, it could be said that the categories constitute the form, though not the content of all representation-as. You could call Kant a “transcendental” rather than an “empirical” conceptualist. However, this still leaves open the question what accounts for specific representings-as. Are these to be understood in the way that the non-conceptualist does or in the way that the conceptualist does? Thanks to James Conant and Robert Pippin for helpful discussion of this topic. Ginsborg (2008), 67.

246

Arata Hamawaki

But for Ginsborg non-conceptualism can’t be the whole truth. While non-conceptualists do justice to the “primitive character” of perception vis a vis thought and judgment, they do so at the cost, not surprisingly, of severing the required connection with thought and judgment. Take the view that perceptual awareness is to be understood in terms of the operation of a differentially responsive capacity through which one is able to discriminate the class of, say, green things from objects of other colors. The problem is that perceptual awareness so understood wouldn’t put me in a position to acquire the concept ‘green’. For in order for my perceptual awareness to make available to me the concept ‘green’ or the thought that the object that I am aware of is green, the perceptual awareness must make available to me a reason for judging that the object is green. A reason is such that if I am aware of it, then I am also thereby aware that I must have a certain belief, or perform a certain action. But the operation of a perceptual differentially responsive capacity does not of itself constitute the awareness that the operation is appropriate to the given object. It could not give me a reason to judge that something is if F and so could not. The operation of such a mechanism doesn’t have the right relation to me – it is not “mine”, not “for me” in the way that a representation generally is. Ginsborg helpfully puts the point this way: “Concept acquisition is usually thought of as a kind of learning from experience, in which the acquisition of the concept is connected with a recognition of its appropriateness to the content of the experience. But if all that is required for an experience to make available the concept green is for it to be reliably associated with greenness – that is, if the greenness itself does not somehow figure in the intentional content of the experience – then it is hard to see how the acquisition of the concept can be a matter of learning as opposed to brute causation.”9 My perceptual awareness of something as F doesn’t just cause me to make the corresponding judgment, or apply the corresponding concept: it makes the corresponding judgment or concept appropriate to the given object. If I already have the concept in question, then my perceptual awareness justifies my application of the concept to the object that is perceptually present. And if I don’t yet have the concept, then my perceptual awareness makes the corresponding concept available to me. Whatever my perceptual awareness of O-as-F amounts to, it must bear such a “rational relation” to my thought and judgment. The 9

Ginsborg (2006), 408.

Ginsborg on the Normativity of Perceptual Representation

247

deep problem with any attempt to account for perceptual content in terms of the operation of a differentially responsive capacity is that such an operation does not simply as such bear such a rational relation to my thought and judgment. Now conceptualists typically account for the rational relation that our perceptual awareness must bear to our thought and judgment by maintaining that any perceptual awareness of O-as-F entails some understanding of what it is for something to be F, however inchoate. Here we are in sight of the other horn of the dilemma, for that view seems to fly into the teeth of the primitiveness of perception thesis, with it seems to require that in order to be perceptually aware of Oas-F, one must already have the concept of F. The problem, then, that Ginsborg is trying to resolve can be put this way. There seems to be a tension between the primitiveness of perception and the fact that perceptual content bears a rational relation to our thought and judgment. The primitiveness of perception thesis seems to require that perception itself be understood in terms of operations that bypass the subject’s understanding. But the rational relation thesis seems to require that perception be understood as necessarily bringing into play the subject’s understanding in determining the content of one’s perception, the “how” of one’s perceivings. Ginsborg argues that the tension is merely apparent. The primitiveness of perception only apparently requires that perception itself be understood in terms of operations that bypass the subject’s understanding. She aims to resolve the tension between these two theses by arguing that a subject’s representational or intentional awareness must be understood as the product of two factors: first, an operation of sensibility or imagination that does not have a constitutive relation to the subject’s understanding, and second, the subject’s awareness that said operation is operating “appropriately”, or as it “ought” to. As I understand it, such a view respects the rational nature of perceptual content, since a certain kind of normativity is already built into perceptual awareness. And her view also respects the primitiveness of perception, since no particular concepts are presupposed in our perceptual awareness having the content it does. Ginsborg takes herself to be developing a view about the nature of perceptual content, and of intentional content more generally, that she finds in Kant. According to Kant in order to have a perceptual awareness of something as, say, to Ginsborg’s example, an apple, it is not enough that one simply have a collection of individual “sensory impressions”, for example, sensory impressions of red, shininess, roundish shape,

248

Arata Hamawaki

and so on. Those sensory impressions must be unified in a certain way, “held together”, so that the apple can be distinguished from the background against which it appears, and from other objects in one’s perceptual field.10 That is not enough, however, for one to represent an apple as an apple. For if it is an apple that I am representing, and not just the near surface of an apple, then I must, in some sense, understand that there is more to the apple itself than what I “strictly perceive”, that is, more than is yielded by my “sensory impressions” alone. I must be aware of the apple itself as something that can be seen from different perspectives, as afforded when I successively occupy different positions in relation to the stationary apple, moving further away, or closer, to the left, to the right, and so on, or when I perform various operations on the apple itself, slicing it open, tossing it in the air, stomping on it with great force, and so on. In being aware of the apple as an apple I am aware, then, of the object as having, so to speak, “hidden aspects”. Consequently, Kant argues, according to Ginsborg that, in being aware of the apple as an apple, the imagination must play the role of “filling out” those hidden aspects in a particular way by reproducing sensory impressions derived from the insides of apples, their backside, and so on. To be aware of an apple as an apple is to be aware of it as a something that is not fully revealed by my current perspective on it. This is why for Kant the imagination plays a necessary role in perception. However, as Ginsborg notes, it is important to recognize that for Kant the imagination itself “fills out” the “hidden aspects” of the apple without in a sense any guidance by me: the impressions that the imagination brings to mind are not brought to mind because such impressions are appropriate to what is required by our concept or understanding of what an apple is. Rather, the imagination “recalls” the impressions it does simply because it has been habituated to do so: its operation is a function of Humean association, an operation that is, in Kant’s words, “blind”. It is the “blindness” of the imagination’s work that makes it difficult to understand how such an operation alone could make available to me the thought or cognition that what is before 10 Kant writes: “Every intuition contains in itself a manifold which can be represented as a manifold only in so far as the mind distinguishes the time in the sequence of one impression upon another; for each representation, in so far as it is contained in a single moment, can never be anything but absolute unity. In order that unity of intuition may arise out of this manifold (as is required in the representation of space) it must first be run through and held together” (Kant (2003), A99).

Ginsborg on the Normativity of Perceptual Representation

249

me is an apple. Here’s Ginsborg: “But one might here wonder how it is that an associative process of this purely naturalistic kind could introduce intentionality or object-directedness into the manifold of one’s perceptual impressions. If the impressions themselves lack intentional content, why should the mere fact that some of them are retained in memory and called to mind on the occasion of having others be sufficient to confer intentionality on the resulting aggregate?”11 Kant, of course, did not think that mere Humean association could explain how an object of cognition could be made available to us. Humean association is only a necessary, not a sufficient, condition of intentional perceptual awareness. Ginsborg understands Kant as holding that there is intentional perceptual awareness only on the condition that the imagination’s reproduction of impressions is accompanied by the sense that its operations are “appropriate”: “When the impression caused by the apple’s exterior redness brings to mind an impression corresponding to an apple’s white interior, that impression comes to mind with a sense of its appropriateness under the circumstances: I take it, in reproducing the latter impression, that it ought to be reproduced under the present circumstances. It is in virtue of this consciousness of normativity in the association of our representations that our perception has intentionality or object-directedness.”12 Ginsborg sees Kant as giving an account of what distinguishes a state of mind that is merely “causally elicited” by the object and its features with a state of mind that is “directed towards objects”, a state of mind that represents an object as being a certain way. Three features of her account are critical, even though they can seem to stand in tension with each other. First, as already mentioned, the operation of the imagination in reproducing or synthesizing sensory impressions is a function of association alone. It in no way is dependent on my awareness that is how the impressions should be synthesized. The imagination 11 Ginsborg (2008), 73. The point here is similar to familiar one that can be found in Wittgenstein’s Philosophical Investigations (1953). One’s understanding of an apple cannot just be a matter of certain images coming before the mind. One could easily conceive of someone having those images without having the slightest understanding of what an apple is. If images are to play a role in understanding – as it seems they must provided that perceptual awareness itself is a manifestation of understanding – what unites the images with one another couldn’t just be blind causation, but the unity that belongs to understanding itself. 12 Ginsborg (2008), 73.

250

Arata Hamawaki

does what it does willy nilly, without any guidance from me, from my understanding or concepts. Second, despite the fact that my imagination synthesizes what it does in its own way, without my self entering into the proceedings, I am nonetheless aware of the combination of impressions that my imagination enacts in normative terms, as appropriate. Third, the awareness of a synthesis as appropriate is what makes the corresponding thought or concept available to us, and so does not already presuppose that one possess the thought or concept in question. She writes, “It is not that I first conceive of what I am seeing as an apple, and then, on that basis, take an impression of whiteness to be appropriate to the impressions that I am not having.” Rather: “to the extent that being presented with an apple leads me to reliably reproduce impressions previously made on my by apples, again with the consciousness of normativity, I am bringing the apple under the concept apple. For what it is to conceptualize one’s sense impressions, on the view that I am proposing to ascribe to Kant, is just to associate them imaginatively in determinate ways with the awareness that one is associating them as one ought.”13 Let me pause over the conjunction of these three points because I think that there is something important that their conjunction gives expression to. The first point gives expression to the passive nature of perception itself. It is this feature of perception that Hume expressed when he said, “But though our thought seems to possess this unbounded liberty, we shall find, upon a nearer examination, that it is really confined within very narrow limits, and that all this creative power of the mind amounts to no more than the faculty of compounding, transposing, augmenting, or diminishing the materials afforded by us by the senses and experience.”14 Whatever perceptual representation comes to, it must be shaped or formed out of materials and their conjoining that are in some sense simply to be found there, not put there by us. In perceiving what we do, we seem in this sense to be wholly at the mercy of the world, including at the mercy of those faculties of ours in relation to which we are passive, such as our sensibility and what Kant called our “reproductive imagination”. The problem is that perceptual content bears a necessary relation to our understanding and thought. It would seem to be a manifestation in perceptual form of our understanding itself. And our understanding is not something that we can stand in a passive 13 Ginsborg (2008), 74. 14 Hume (1993), 11.

Ginsborg on the Normativity of Perceptual Representation

251

relation toward. The conjunction of the three points is Ginsborg’s attempt to do justice this difficult tension that seems to belong at the very heart of the nature of perceptual content itself. In this way Ginsborg aims to preserve the naturally compelling thesis of the primitiveness of perception, while at the same time doing full justice to the idea that perceptual content is rational in form and so gives us, when the world cooperates, a reason for making the corresponding judgment. The appeal of her view is that it aims to salvage what is true in non-conceptualism without abandoning the idea that lies at the heart of conceptualism. If it doesn’t stay true to the letter of conceptualism, as that view is often understood, it nevertheless remains true to its spirit.

II. Two Difficulties: Indeterminacy and the Relation to the Possibility of Truth I think that the tensions that form the background to Ginsborg’s view are real, and that her view is to be applauded for seeking to find a stable resolution of the tensions, without in any way slighting the difficulty thrown up by the tensions themselves, that is, while maintaining a vivid sense of the pull exerted by each of the opposing sides. It would be all too tempting to ease the tension simply by ceasing to appreciate one side of the opposition. I do, however, want to raise some questions about Ginsborg’s attempt at a resolution. First, it is crucial to Ginsborg’s account that representational awareness builds on the material furnished by a non-representational awareness of a ‘what’. The non-representational awareness is not the awareness of something purely inner, such as a sensory impression, but it is an awareness of something that is in a broad sense subjective, an awareness of a particular way in which I, or rather a mechanism operating within me, is responsive to the object at hand, albeit a way of responding that displays a sensitivity to a feature that the object actually possesses. However, as she recognizes, her account cannot assume that I am aware of the response as a form of sensitivity to the feature to which it is in fact sensitive, for such an awareness would obviously presuppose the capacity to represent the feature of the object and so would assume what is at issue. Ginsborg attempts to get around this difficulty by characterizing the awareness as demonstrative in form. We come to represent the object as having a particular feature, such as its being an

252

Arata Hamawaki

apple, when we are aware of ‘this’ as appropriate to the object at hand, where the referent of ‘this’ is given simply by the process going on within me whose activation constitutes a sensitivity to something’s being an apple. So the question needs to be raised how the referent of ‘this’ is to be secured, given that it cannot be secured by its figuring as part of a complex representation of the form ‘this-such’, such as ‘this-association-that-is-sensitive-to-the-presence-of-the-property-of-being-anapple’, or even ‘this association that is sensitive to something’. There are two potential problems here. First, it is not clear that a bare ‘this’ can fix a referent at all. In order for the ‘this’ to refer to something, surely I must have in mind some idea of what the ‘this’ is supposed to refer to. But, it might be asked, how could I have such an idea without having some concept that enables me to narrow down the possible referent of the ‘this’? For example, if I am referring to an object, wouldn’t I need to deploy a concept of an object, something that is relatively enduring and has a location in space? And if I am referring to a sensation, wouldn’t I need to be able to distinguish a sensation from an object, and so need to possess concepts of each? Second, any association or differentially responsive capacity seems to underdetermine which feature the association or the differentially responsive capacity is sensitive to. Thus the process within me that is sensitive to green is also a sensitivity to chromatic color, and the process that is sensitive to cubes is also a sensitivity to equiangular and equal-sided figures. In fact if disjunctive properties are included among the candidate features of objects, then there would be an indefinite number of features of the environment to which any responsive capacity is sensitive. Responsive capacities are in that sense indeterminate with respect to what they are responsive to. Again, how does one narrow down the range of possible objects of demonstrative reference without bringing in concepts, or the background of a “form of life”, or something of this sort? I will not, however, dwell on these difficulties any further, since I want to explore two problems that I believe are deeper. Ginsborg insists that the appropriateness that bears on the constitution of representational content is not the normativity of truth or veridicality. She recognizes that we might speak of our perceptual experience as appropriate to the object just in case it represents the object correctly or veridically, represents it as it really is. But, as she points out, appropriateness in the sense of verdicialty presupposes that there is something that my perceptual experience represents, veridically or non-veridically, whichever may be the case. In order to represent an

Ginsborg on the Normativity of Perceptual Representation

253

object as a cube veridically, I must first be representing the object as a cube. She writes, “Thus, even in cases where I do take my perception to be appropriate in the sense of binge veridical, there is a different and more fundamental awareness of appropriateness which is, so to speak, built into perception itself, and which is required if the question of the veridicality of my perception is so much as to arise.”15 Ginsborg is surely right that if an awareness of normativity is constitutive of perceptual representation then it must be distinct from and prior to the sense of normativity associated with veridicality or truth. But it is also clear that if there are two notions of normativity here they must nevertheless be intimately related. As I mentioned earlier, my experience’s representing the object that is given to it as a cube, makes available to my thought and judgment the fact that the object is a cube, when things go well. This means that there must be an internal relation between the representation of an object as a cube and the fact that the object is a cube, even though, of course, representation and fact can come apart. Does Ginsborg’s account preserve the internal relation? The awareness that certain impressions “belong together” rather than just go together, the awareness that they are combined as they ought to be, is what for Ginsborg distinguishes a combination of impressions which merely registers a psychological fact about me from a combination of impressions that represents how things are out in the world. It is this normative awareness that is supposed to account for the intentionality or object-directed character of my perceptual experience, to account for the fact that the experience represent that a certain object is thus and so. But if that is the case then the normative awareness that the impressions ought to go together must put me in a position to take in the corresponding fact when things go well. How is it supposed to do that? How does it illuminate how my experience is able to afford me the corresponding fact? It looks like on Ginsborg’s account my appreciation of the corresponding fact would not be something that the representation itself makes available to me. Rather, it would have to involve some further distinct awareness, call it the awareness of veridicality or truth itself. There are problems with such a view. For one thing it seems to imply a kind of hypostatization of representational content itself, for it seems inevitable on such a view that veridicality or truth be regarded as some kind of property of the representation or representational con15 Ginsborg (2006), 417.

254

Arata Hamawaki

tent. In addition to the well-known problems with such a view of veridicality or truth, there is the further problem that such a view implies a kind of gap between representational content and the world. What we represent would fall short of the world itself, requiring a further act of awareness in addition to the act of awareness that is responsible for our experience having a representational content if we are to take in the corresponding fact. But if that is the case then it is no longer clear that what Ginsborg’s normative awareness would account for would be representational content, that is, content that is object-directed or intentional. Whatever the ground of representational content is, it must preserve the idea that it is internal to representational content that it is either veridical or non-veridical, true or false: representational content must be content that purports to be veridical or true. The problem is that the awareness of normativity that is for Ginsborg the ground of intentionality has no clear connection with the truth potential that is an essential feature of representational content. The general type of account that Ginsborg is giving seems to be a descendant of those that can be found in the views of early 20th Century sense-data theorists and their attempts to explain representational awareness on the basis of the materials afforded by non-representational awareness. Of course, I am not suggesting that Ginsborg is committed to the view that what we are aware of in perceptual experience are “sense-data”. But what the account that Ginsborg gives can seem to have in common with the early 20th Century sense-data theorists is that both parties seek to explain the transformation, as it were, of non-representational awareness into representational awareness by appealing to the special content of a special act of mind. But if the considerations I have given above are on the right track, there is an in principle obstacle to such attempts. It would seem that no content of awareness would be able to account for the possibility of representational content, since what we would be aware of would be at best some fact about the material afforded by non-representational awareness, and no such fact could constitute the representation that an object is thus and so. It seems that on Ginsborg’s account my awareness of the normativity of the association performed by my imagination is an awareness that is about the association or the operation of the faculty of the imagination, an awareness that it is operating as it ought. But how can an awareness of a fact about a psychological faculty of mine account for representational content? What would such an awareness have to do with the representation that the object is thus and so? Any state that would be constituted

Ginsborg on the Normativity of Perceptual Representation

255

by such awareness would seem to fall short of anything that could have the intrinsic feature of purporting to be veridical or true. However, understanding Ginsborg as operating in the shadow of the general sort of account of representational content common to sense-data theorists may well involve a misunderstanding of her view. That it is a misunderstanding is suggested by what Ginsborg writes at the end of “Aesthetic Judgment and Perceptual Normativity”: “The possibility of empirical conceptualization relies on our perceptual responses involving a claim to their own appropriateness, and thus a demand for universal agreement, which does not in turn depend on a claim to objective truth.”16 The rest of the paper will be devoted to exploring this intriguing idea of a claim to appropriateness, or universal agreement, that is not itself a claim of objective fact.17

III. A Third Difficulty: the Content of Primitive Normative Awareness What is the content of the normative awareness supposed to be, given that the content of the awareness cannot be given by a concept? For example, if my imagination is combining certain impressions together and not others, my concept of an apple can give me a standpoint from which I could say that just those impressions are the ones that my imagination ought to be combining – not, for example, an impression of my mailbox or my aunt. And the claim that it is combining just the impressions it ought to combine seems to presuppose the concept of an apple itself as that which dictates that just those impressions be the ones that are combined. But it is essential, of course, to Ginsborg’s account that my awareness or my imagination as combining impressions appropriately not be, in that sense, “guided” by a concept, for if it were, then it could not explain how experience could make concepts available to me in the first place. But then what is it that would explain or account 16 Ginsborg (2006), 430. 17 My own view is that representational contents are contents of acts of holding true – the latter is in an important sense prior to the former. Even though there would be nothing that could be true or false without representation it is only in the context of acts of holding true that there can be representation. This is why there must be an internal connection between truth and representational content. Truth is really the prior notion. I will develop this thought a bit further later in this paper.

256

Arata Hamawaki

for my awareness that my imagination is combining impressions as it ought to, given the particular circumstances, and so on? On what basis does the awareness of normativity rest? Ginsborg’s answer is that the awareness or claim of normativity here doesn’t in a sense rest on any basis at all, if a “basis” is understood to be a criterion of correctness that we already possess prior to the awareness or claim: the awareness or claim of normativity, as she puts it, is “primitive”. Nonetheless, if we are to be capable of representing how things in the world are at all, as opposed merely to being able to exercise sensitivity to the presence of certain features in the world, we must have the capability of making a claim of the appropriateness of the exercise of those faculties in us that exhibit such sensitivity. She writes, “For in being aware of a normative fit between the object and my way of perceiving, I am aware of the object as making appropriate that way of perceiving. The content of this awareness could be spelled out by describing it as the awareness that the cube ought to be perceived this way, where “this way” refers to the very way I am perceiving it.”18 As Ginsborg puts it elsewhere, if the subject is aware of the object as a cube then she is aware of the “way she is perceiving” the object as “exemplary” of a rule, even though she may not be in a position to formulate the relevant rule.19 This implies that if asked what it is for the process of the imagination to be operating as it ought, all that the subject can do is to refer to the actual process in which the imagination is engaged.20 But it’s obscure how a reference to the actual process in which the imagination is engaged could itself determine the standard which the actual process is supposed to display. If all there were in determining the standard of correctness were the process itself plus the awareness of the process as appropriate or correct, then it would seem that the process couldn’t fail to conform to the standard of appropriateness or correctness, since it itself 18 Ginsborg (2006), 416. 19 “The sense in which the child takes her way of perceiving to be appropriate can be spelled out further by saying that she takes it to be exemplary of how the object ought to be perceived, or, as I shall also put it, that she takes it to exemplify a rule for the perception of the object.” (Ginsborg (2006), 241). 20 I will from here on out treat association of the imagination and exercise of a differentially responsive perceptual capacity as on a par. In the two papers I focus on Ginsborg appeals to association in one paper and to differentially responsive perceptual capacities in the other as providing the material basis of perceptual representation. What both have in common is that they are sub-personal processes that display a sensitivity to features of the subject’s environment.

Ginsborg on the Normativity of Perceptual Representation

257

would determine the standard of appropriateness. But if there is no sense in which the process itself, as it actually takes place, could have taken place inappropriately, then no sense has been given to the idea of its taking place appropriately. It would be like applying the measure of being one meter long to the standard meter in Paris. What functions as the standard of measure can’t at the same time be treated as that to which the standard is applied. And yet it seems that that is exactly how we are supposed to be treating an association that is to give rise to a concept: the association is to be treated as both giving us a normative standard, and as that to which the standard is applied. The problem is that an association of impressions simply is, so to speak, what it is. There is nothing in its nature that would supply a natural application of the notion that it is operating in the way it ought to, or is operating appropriately. The same could be said for any form of discriminatory sensitivity to an object or its features. Descartes made a similar point in his discussion in the Sixth Meditation of whether our body could be deemed faulty because it is subject to an illness like dropsy. He wrote, “And given the intended purpose of a clock, I could say that it deviates from its nature when it fails to tell the right time. And similarly, considering the mechanism of the human body in terms of its being equipped for the motions that typically occur in it, I may think that it too is deviating from its nature, if its throat were dry when having something to drink is not beneficial to its conservation. Nevertheless, I am well aware that this last use of “nature” differs greatly from the other. For this latter “nature” is merely a designation dependent on my thought, since it compares a man in poor health and a poorly constructed clock with the ideas of a healthy man and of a well-made clock, a designation extrinsic to the things to which it is applied. But by “nature” taken in the former sense, I understand something that is really in things and thus is not without some truth.”21 When we regard certain clocks as malfunctioning and others as functioning in the way they ought to, we do so by adopting a certain standpoint: the standpoint provided by the idea of the purpose for which clocks are designed. But this standpoint is in a sense optional, for the operation of the clock itself can be understood without assuming such a standpoint. In that sense the application of normative notions to the clock is “extrinsic” to it: it would not be something that expresses its nature. For Descartes, of course, its nature is entirely mechanical. By 21 Descartes (2006), 47; ATVII, 85.

258

Arata Hamawaki

contrast, if I observe two people having a game of catch, I cannot understand what they are doing without assuming a standpoint from which I grasp the activity that they are together aiming to carry forth. I understand what they are doing by seeing that they are doing what they are supposed to be doing, given their shared conception of the activity they are engaged in. Thus, if one of them does not throw the ball where the other is in a position to catch it, there is a natural application for the idea that what he did was not what he should have done, that he has failed.22 But sub-personal, “pre-conceptual” processes like association of the imagination or “perceptual sensitivity” are not by their nature normative and so do not seem to provide a foothold for normative notions. It would be as if someone pointed to a pile of rocks and said that the rocks are arranged as they ought to be. Puzzlement would seem to be the appropriate response to such a claim: “what could that possibly mean?” However, it would be easy to imagine a background that would provide the anchoring for such a normative claim: we can imagine that arranging the rocks in that way is part of a game, or is a sign that would convey a particular message, and so on. But such a consideration is not available to Ginsborg, for it is essential feature of her account that the normative awareness that is to constitute an intentional content be freed from such a meaning-bestowing background. The normative awareness she speaks of must not depend, as she puts it, “on the appreciation of anything either about our way of perceiving or about the object, in virtue of which we ought to perceive the object that way.” But given that the normative awareness is shorn from such a background, one wonders whether the stage-setting is in place for there to be a meaningful application of a notion of normativity at all. Now Descartes’s point about the clock doesn’t apply to perceptual content – we don’t need to adopt a position “external” to the perceptual content in order to find an application for normative notions. Rather, it seems to be “normatively constituted”. In fact I take this to be precisely 22 Kant thought that living things are like this as well. We can’t get on with the business of understanding them unless we assume a normative standpoint with respect to their behavior and growth. We must in other words understand what they do, and how they develop, by seeing what they do as the realization of what they ought to do, and how they ought to develop, given the sorts of thing that they are. Although for Kant the normative standpoint is only a regulative idea, the fact that we must adopt such a standpoint in order to understand a living thing gives expression to the sort of thing that a living thing is, to its nature or essence as a living thing. See Kant (2000).

Ginsborg on the Normativity of Perceptual Representation

259

why Ginsborg rejects standard non-conceptualist accounts of perceptual content. Such accounts seem entirely to bypass how the subject herself understands what she represents, and so does not accommodate the idea that whatever the subject represents, it is in its form “for her”. If, for example, one sees the snake-like figure on the ground as a snake, then it is “for her” a snake: it is she herself who understands what she sees in that way. And in the context of such an understanding, what she sees appears to have certain features, a head, eyes, scaliness, and so on: features that together add up to its being represented by her as a snake. I think that Ginsborg’s appeal to normative awareness is meant to register the fact that one is aware of the features that constitute the content of a representation as in some sense “belonging together” rather than just “going together”, or being “found together”, as they would if their conjoining were merely the result of an association of the imagination. Given that the representation is a representation of a snake, and not a stick, it is appropriate that it be represented as having “snake-like” parts and features, parts and features that would be characteristic of a snake rather than a stick. Such relations of appropriateness are constitutive of the features of the content of a representation, and they are precisely not constitutive of the impressions that are bound together through association. But if that is so, that presents, I think, a problem for Ginsborg. How can the way in which one is aware of something, even if it is being aware of something that is taking place in some sense within oneself, affect the nature of that of which one is aware?23 The problem is that a primitive awareness of normativity seems too “external” to an association to “transform” it into something with normative content.24 The woryy is that perceptual content is being con23 Descartes recognized this problem. We can think of a clock in relation to the function of telling time. But that cannot affect the nature of a clock. The nature of a clock is the same as any other extended thing – namely extension – and so is to be understood mechanically. And so it would seem with the human body as well. In order to make room for the idea that there might be something intrinsically faulty about a body given to dropsy, Descartes found that he had to make room for a third kind of substance, a substance with a normatively constituted nature: what he called the union of mind and body. Whatever one might think of Descartes’s views on mind-body union, it should be acknowledged that the pressures that led him to the view are real. 24 I say “transform” because the mere association of sensory impressions doesn’t itself represent anything.

260

Arata Hamawaki

ceived on a model that is too close to Descartes’s model of the clock, where the problem is one of understanding how something that is not – like thought itself – normatively constituted can become such as to be an appropriate subject of normative judgment. Since a concept that specifies the normative standard for the association cannot be assumed, the worry naturally arises whether there is room for a gap between the way the process of association actually takes place and the standard of its taking place as it ought to. Actions are perhaps a paradigm case of normative constitution. Actions can’t be identified with bodily movements since only the former has a constitutive relation to the agent’s own understanding of what she is doing and so the agent can’t just find herself doing something. If she is doing anything at all, she must already have an understanding of what she is doing, must already grasp the reason why she does what she does. This is why she must know what she is doing without resorting to observation25. This places a constraint not just on my knowledge of what I am doing but on anyone’s knowledge of what I am doing, for if I myself don’t know what I am doing then in an, important sense, there isn’t be anything that I would have done. The agent’s self-understanding creates a gap between what actually happens and what is supposed to happen. What actually happens can’t enter into the agent’s own understanding of what she is doing independently of the agent’s own understanding of what is to be done in the circumstances. Rather what actually happens – that is, the action – must be viewed as the materialization of the agent’s own understanding of what she is doing. We ca think of what does and does not properly belong to the agent’s mindedness, to the agent’s mind, properly speaking, as determined by whether or not it has such a constitutive relation to the agent’s own understanding of herself. So understood, processes of association or perceptual detection do not belong to the agent’s mind. Such states and processes would be states and processes that one finds going on within one, by something like “observation”. But nothing that is one’s “own” in the strict sense is something that one can in that sense discover. Such processes and states cannot enter into one’s mindedness, even when they are “taken” by oneself as “appropriate”. What one’s perceptual experience represents is something that the subject must already have an understanding of: it is not something that depends on anything 25 Here, I am drawing on Elisabeth’s Anscombe’s point, that our nowledge of our own actions is necessarily “non-observational”. See Anscombe (1957).

Ginsborg on the Normativity of Perceptual Representation

261

that the subject would need to observe going on in herself or in the world.26

IV. Ginsborg’s Response to the Difficulty Ginsborg acknowledges something like the difficulty I have just presented for her view, for she writes: “the point is not that she [the subject] takes her way of perceiving to be appropriate to the object in virtue of the fact that the object is indeed a cube; rather, as I have said, she takes her way of perceiving to be appropriate to its object without first having in mind any feature of the object in virtue of which that way of perceiving is appropriate. But this might seem flatly incoherent. It might be protested that there can be no such thing as taking oneself to be doing as one ought simpliciter: one can only take it that one is conforming to some antecedently specified rule which determines what one ought to be doing.”27 By way of responding to this problem, she elaborates her view by giving an example involving a child sorting blocks into two piles. We are to suppose that the child sensorily discriminates the block as something to be placed in pile A (with the cubes) rather than in pile B (with the pyramids). That is, she is sensitive to a similarity between the block before her and the blocks that she has been placing in pile A. But that isn’t to say yet that the child sees the block as “belonging with” the blocks she has already placed in pile A, that is as appropriately, or rightly, placed with those blocks. Only where there is a place for such an idea of correctness, would the child have an understanding of what she is doing as sorting the cubes from the pyramids. Ginsborg argues that what makes the difference is that the child takes her sensitivity to the similarity as exemplary of how the blocks ought to be sorted or perceived. It is thus that she comes to have an understanding of the block as a cube, an understanding that is exhibited in the way that she perceives the cube and in her sorting behavior.

26 The sort of state that Ginsborg imagines then seems to require an odd form of split consciousness. On the one hand one must be related to the state as something that one simply finds oneself in. On the other hand insofar as it is constituted by an awareness of normativity, the state would seem to have an internal relation to one’s own understanding of oneself, and so has a constitutive relation to the first-person. I will return to this point in the final section. 27 Ginsborg (2006), 419.

262

Arata Hamawaki

As Ginsborg quickly admits, such an illustration leaves the crucial idea unexplained: the idea of the child’s awareness that “this” – where “this” refers to the similarity-sensitive behavior – is how the object ought to be perceived or responded to. The question remains of what such a normative awareness could consist in given that the “this” is supposed to refer both to that which is being claimed to be appropriate, and to fix the standard of the normative awareness itself. The problem is how there can be a gap between the way the process actually unfolds and the way that the process “ought” to unfold: how is it possible for the awareness of normativity to be mistaken? Ginsborg’s solution to the problem is to appeal to a distinction between a standard of sorting A’s from B’s, cubes from pyramids, on the one hand, and a standard that distinguishes genuine sorting as such from behaving randomly, on the other. She writes, “But this objection overlooks that the possibility of a mistake might lie, not in the child’s contravening a rule or standard applicable to her perceiving or sorting the object, but rather in her taking her perceiving or sorting to exemplify a rule or standard berhaupt… The possibility of error thus lies in the possibility that her sense of appropriateness, in this particular instance, may be illusory: not because she is sorting wrongly or inappropriately, but rather because what she is doing does not count as genuine sorting in the first place.”28 The distinction here seems to be similar to one between saying something with a specific meaning, and saying something at all, a distinction, you might say, between meaning and meaningfulness. Just as specific meanings are in the offing only against the background of processes that conform to the conditions of meaningfulness, so a specific concept is in the offing only against the background of an awareness of normativity as such, not the normativity that would be prescribed by a particular concept, such as the concept ‘cube’, but something more general, more fundamental: a normativity that would underlie the possibility of the concept ‘cube’, or any concept at all. In order to explain this distinction, Ginsborg introduces into the scene two children, Max and Nora, who up to a certain point in their game with the blocks, “agree” in placing the cube-shaped blocks together in one pile and the pyramid-shaped blocks in another. But when they come across a cube-shaped block with a color they had not until that point encountered – green – their inclinations diverge. Max wants to place the block in the pile with the cubes, Nora, in the 28 Ginsborg (2006), 419.

Ginsborg on the Normativity of Perceptual Representation

263

pile with the pyramids. Ginsborg supposes that Max’s attitude at this point isn’t: to each his own. Rather he turns on Nora, for he sees her as sorting the block inappropriately. But from what perspective is he able to adopt this attitude, given that ex hypothesi, he does not have the concept ‘cube’? Ginsborg acknowledges that Max’s own inclinations cannot provide him with the needed normative perspective. For if that were the case, what is right would collapse into what seems to him to be right, and then there would be no sense to talk of one’s being “right”. No sense would have been given to how Max himself could be mistaken, and so, no sense given to how Nora could be mistaken either. Here’s Ginsborg’s response to the worry: “Max might see Nora as mistaken, not in what she does or is inclined to do with the cube, but in her insistence that what she is inclined to do is appropriate, or that the cube goes with the pyramid. If he takes this attitude, he will see nothing wrong in Nora’s putting the green cube with the pyramid if she feels like: all he will take exception to is her claim that this is where the green cube ought to go. From Max’s point of view, Nora’s inclination is arbitrary, and her mistake lies in her ascription to it of normative force. It is this possibility, I want to suggest, which gives the required content to Max’s claim that, for his own part, he is sorting the block appropriately. He can make sense of the possibility of himself as mistaken, in this claim, by imagining that he himself is deluded in just the same way that he takes Nora to be: that is, that he is simply following random inclinations but taking them to be indications of how the blocks ought to be sorted.”29 I have a hard time seeing how this is supposed to help. As Ginsborg notes, there is nothing inappropriate about Nora’s placing the green cube over with the pyramids. There would be if she understood what she is doing as sorting the cubes and the pyramids. But it is assumed that that cannot be the sense in which she is mistaken, since that would require the childrens’ possession of the corresponding concepts. And further, as she recognizes, there is nothing inappropriate about her inclination to sort as she does, considered simply as an inclination. The possibility of a mistake in what Nora is doing emerges only insofar as we see her as engaged in the game of sorting, only insofar as we see her as making a claim that what she is doing is what she ought to do. In that context, she would be making a “mistake”, since she wouldn’t doing anything intelligible at all: she would be behaving randomly. 29 Ginsborg (2006), 424.

264

Arata Hamawaki

But are we fully able to imagine such a possibility, to imagine it fully? It would be one thing if Nora were from the start behaving in a random manner. In that case it could hardly be said that she is playing the game of sorting in the first place. We could not ascribe that self-understanding to Nora at all. And if we couldn’t do that, then we would have no critical purchase on her activity, anymore than we would have critical purchase on the behavior of a sleepwalker. We couldn’t very well say that the sleepwalker is doing something he shouldn’t be doing when he opens the car door and turns on the ignition. Sleepwalking is not subject to critical assessment, and neither is the “game” of putting the blocks where it pleases one to put them. I take it that it’s important that Nora was playing the game just fine up to a certain juncture and only at a certain pointin the game began to go, let’s say, haywire. The problem is that to the extent to which she goes haywire, it becomes more difficult to ascribe to her the understanding that she is playing the game at all, and so as making a claim to be proceeding appropriately. But we need to hold in place the idea that she makes a claim that what she is doing is “appropriate” if we are to have room for the possibility that she is behaving “inappropriately”, randomly, not governed by a standard of appropriateness at all. Only thus could she be under the delusion that she is behaving appropriately. If all she were doing were behaving randomly, is no understanding she would even purport to have of what she is doing. The same, of course, applies to Max, mutatis mutandis. The point here isn’t just epistemological. It’s not just that the only evidence we have for ascribing the claim to be behaving appropriately to her is what she actually goes on to do. The point is that we have no understanding of the claim itself apart from what she actually goes on to do. Since what she goes on to do makes no sense, the claim itself to be, as it were, doing something intelligible itself has no sense. But if that’s so then Nora just is behaving randomly, and no sense would have been given to the illusion that she isn’t. The notion of an illusion here is a difficult one, and I think that the conditions of illusion themselves constrain the sorts of things with respect to which one can be under an illusion about. One can’t, for example, be under the illusion that one is thinking. That is something that Descartes famously discovered in articulating the cogito. In order to be under the illusion that one is thinking, one must think that one is thinking, and that itself implies that one is thinking. I think that we are under a similar constraint with respect to a claim to be doing something appropriately, such as sorting. I think that we are in a position to

Ginsborg on the Normativity of Perceptual Representation

265

understand and so ascribe this claim only to the extent to which we can see what someone is doing as appropriate – by our own lights of course: who else’s? Where we cannot see an activity as sorting then we can make no sense of the subject’s own claim that what she is doing is sorting, and so we cannot sensibly regard the subject as under the illusion that she is sorting. We simply don’t have enough of a handle on what the subject is doing to have a stable description of her. We don’t know what to say. I think that that is the position that Max is in with respect to Nora. But if that is so, then we don’t have any clear content to Max’s own claim that he is sorting appropriately and Nora is not, for we still don’t have a clear example of failure.

V. “Mineness” and Mindedness The point that I have been making can be put as follows. Failure in sorting presupposes that that one is sorting, and the activity of sorting presupposes the agent understands herself to be sorting. But the distinction between sorting and just operating randomly cannot itself be based on a norm, for operating randomly isn’t a failure to engage in sorting: it is just not sorting. A failure to sort properly would be a specific failure such as the failure to sort the cubes with the cubes, or the pyramids with the pyramids. Similarly, it is possible for someone to speak unintelligibly, possible for his utterance to be without meaning. In that sense the speaker’s utterance could be deemed a failure, a semantic failure, a failure to make sense. But it is only with regard to a subject who has the capacity to make sense that there is room for the possibility of a failure of intelligibility. I think that this is why it is important that Ginsborg needed to present her example in such a way that Nora was able to continue with Max up to a point, enough so that we are able to ascribe to her the claim that she is proceeding as she ought. She might have presented Nora as having an even richer cognitive repertoire than she endows her with in the example. She might have well been able to play other games of sorting, as well as other kinds of games, games that would display an understanding of the sort of activity in which she would be engaged. Against such a background, it would make sense for us to say that she has failed when she places the block over with the pyramids rather than the cubes, and understand the failure to be a failure of understanding. But this is just to say that it is only against

266

Arata Hamawaki

the background of attributing to Nora conceptual capacities in general that we can intelligibly apply to her a notion of cognitive failure. Such failures are, as much as are the successes, exercises of the capacity to speak meaningfully – it is part and parcel of such a capacity that it can be exercised unsuccessfully, just as it is part and parcel of a healthy body that it can fail to meet the standard of health. Just as the failure of a body to be healthy is itself the manifestation of the body’s capacity to be healthy, so failure to speak meaningfully is itself a manifestation of the capacity to speak meaningfully. Semantic failure cannot therefore be “brute” failure but must in its nature be intelligible, since it must be a failed attempt to make sense. The subject who has failed to mean must at least in principle have reflectively available to him the explanation of the failure. Meaningfulness, the capacity to mean anything at all, is in this sense a kind of transcendental presupposition of meaning, and failures of meaning. Similarly, the distinction between sorting and not sorting, or between our exercising a conceptual capacity and not exercising such a capacity, cannot itself be a normative distinction. The distinction is like that between making a claim at all and not making a claim. I think that we can conceive of this point as an extension of Descartes’s cogito. A way of understanding the discovery that is formulated in Descartes’s cogito is that ‘I think’ is not in a sense a content of any particular claim or judgment, but is the condition of making any claim or judgment at all. This is why we can’t be mistaken about the cogito, for if we were mistaken, we would lose our grip on our having made a claim, and so on the idea of a mistake. Engaging in sorting or conceptual representation also strikes me as something that cannot be the content of a claim that we could be mistaken about, but is rather the presupposition of the application of the concept of mistake or failure. If there were a mistake, then there would be in a sense no one home, no one to whom we could attribute the error, just as Descartes discovered that being mistaken about the proposition ‘I think’ would mean that there would be no one there to have made the mistake, no ‘I’. All there would be a pulsating heap of sub-personal responsive capacities, activated here and there. There is an echo here of Wittgenstein’s remark in his Tractatus that logic must take care of itself – that there can be no mistakes in logic. One either is in the space of making a claim or not: if one is then there is room for the application of the notion of error (one is subject to logical laws that determine the possibility of true judgments), and if one is not then not (one is outside logical space). But there is no co-

Ginsborg on the Normativity of Perceptual Representation

267

herent notion of supposing that one is in this space but is mistaken in thinking that one is. That would be to confuse a transcendental presupposition for the content of an actual judgment or awareness. My sense is that Ginsborg is running up against the barrier of the transcendental: she is trying to round on what is a transcendental presupposition of the possibility of content as itself the content of a particular claim. After all, if my capacity to understand were itself grounded on a claim to appropriateness that would seem to make the possibility of understanding itself rest on me, on a particular individual making a particular claim. But isn’t being in a position to make a claim at all – already to take up an objective standpoint, a standpoint that isn’t just one’s own, but that of anyone? That one is taking up some such standpoint seems to be a condition of understanding anything at all, and occupying such a standpoint doesn’t seem to require that one first take oneself to occupy such a standpoint. To be minded at all is to be minded objectively.30 Thus, I think that a claim to be doing something appropriately, cannot be the foundation of the distinction between our merely doing things randomly and our doing things in a way that displays our possession of a conceptual capacity. For it is only against the background of our having conceptual capacities in a general sense that it makes sense to speak of an illusion of understanding, an illusion of engaging in some sort of activity in an inappropriate way. I think that this means that we can’t get ourselves in a position to explain how a subject comes to have concepts berhaupt, for any account of concept acquisition will need already to suppose that the subject has a conceptual capacity, and so has, however vaguely, or “incompletely”, worked out some point of view on how the world is. And we have no grip on the subject as having a point of view apart from supposing that the subject makes judgments about the world. Thus I think that the truth in conceptualism is that judgment is in an important sense more fundamental than experience. Although epistemologically speaking our judg30 Ginsborg writes: “I can allow the possibility that someone’s ways of perceiving or sorting diverge from mine, and that neither of us is in a position to offer an argument that rationally compels the other to change her way of perceiving or sorting, without for all that giving up my claim that my way of perceiving or sorting is appropriate. Immodest though this claim might seem, especially in the face of an actual disagreement, it is not incoherent” (Ginsborg (2006), 427). My worry is not with the immodesty of the claim but with the idea that we have gotten hold of a claim at all, given the role that the claim is supposed to play.

268

Arata Hamawaki

ments, at least our empirical judgments, are based on our experience, the representational content of our experience must be understood in terms of the representational contents of our judgments. As I suggested earlier to say that my experiences have representational content is to say that what they make available to me are environmental facts, and if that is the case then it follows that the contents of my experience are themselves to be understood in terms of the judgments that experience affords me, that is, in terms of the point of view on the world that through my experience I stake myself to. The idea, then, that experience furnishes some sort of constraint on what I can judge must go by the boards, as inherently confused. To the extent that what Ginsborg calls “the primitiveness of perception” in relation to thought and judgment trucks in that confused idea, it too must be thrown in the refuse heap. However, I don’t think that it is the idea of experience as constraining what we can judge that is the source of what Ginsborg finds appealing in the primitiveness thesis. This is why for her there is, at best, what I described as a truth “in” non-conceptualism. For her that truth is better expressed in terms of a notion of judgment that is not itself dependent on the subject’s already possessing a particular concept, a notion you might say of the subject as staking himself to a point of view as such, as opposed merely to being in a state that is caused by some sub-personal process “within” the subject, such as the workings of association, or of a differentially responsive capacity. But I think that that distinction is best understood in terms of two different ways of explaining the contents of the subject’s “mind”, different ways of understanding or explaining the subject’s mental states. If the state is a result of the cranking away of sub-personal machinery, then the subject’s being in the state is something that can be understood in terms that, in an important sense, bypass the subject herself altogether. But those states of the subject that are in the relevant sense “mine”, those that the subject is in a position to ascribe to herself without the assistance of observation or evidence, are states whose obtaining is to be explained in the way that the subject herself would explain them, namely, on the basis of what she has reason-to-believe. For example, if I judge that the block I have in my hand is a cube then my being in that state must itself be understood or explained on the basis of the reasons I have for making that judgment, in this case my perception that the block I have in my hand is a cube. That is, my judging that P is to be explained by my own understanding of why, in the current circumstances, one is to judge that p, why, in those circumstances, p

Ginsborg on the Normativity of Perceptual Representation

269

is the thing to judge. Furthermore, this is not just a point about how I myself would explain why I judge or believe what I do. Rather, it is how anyone must explain or understand why I judge or believe what I do. It is, in that sense, a point about the nature of judgment and belief itself, a point about the nature of those states that are in the relevant sense “mine”. A state that I am in as the result of association or a differentially responsive capacity is not “mine” in that sense, even if I am aware of being in such a state. Nor is a mere sensation, a content of non-representational awareness, “mine” in that sense either. I take the difference here not to be between what I am and am not immediately conscious of, but rather between what does and what does not have a constitutive relation to the first-person and to self-knowledge.31 Now the point I have been making is about those states that involve a commitment to some way of thinking about how the world is, thus to those states that express the point of view on the world to which I stake myself. But when this point is combined with my earlier point that the representational content of my perceptual experience is itself to be understood in terms of the point of view on the world that it makes available, rather than the other way around, then it follows that my perceptual experience is in the relevant sense “mine” as well, even though I am, in some sense, passive in relation to it. If this is on the right track then the question, what must be added to states that are the result of mere association or differentially responsive capacities in order for the state to have a conceptual content, betrays a misunderstanding of the nature of mind. It would be like asking what must be added to these states in order for the states to be in the relevant sense “mine”. As Hume famously discovered, when being “mine” is conceived of in such a way there is nothing there to be found, no feature that would mark those states as mine in particular. All he found were the particular states themselves. Hume’s point could be extended to any attempt to account for the “mineness” of my mental states, or their having a representational content, in terms of a particular special mental state or content, one that would explain the difference between states that are and states 31 As I suggested earlier, the idea that representational content is constituted by its being “mine” in the above sense suggests why the attempt to explain representational content in terms of a primitive normative awareness of an associative or differentially responsive capacity is problematic. Such an account fails to respect the idea that representational content is constituted by self-knowledge. For exploration and development of this view, see Marcus (forthcoming) and Rödl (2007).

270

Arata Hamawaki

that are not in the relevant sense “mine”. His point would therefore apply to Ginsborg’s appeal to the notion of primitive normative awareness. Even if we could make sense of that notion, it would still just be another mental state, and so would from Hume’s perspective just be one more state to put alongside the others he had found. What is it about that state that is supposed to be so special? It has a special content, no doubt. But that is an observation about its content and so doesn’t affect the sort of state that it is, qua state. And even if the primitive normative awareness were to be understood as not itself a self-standing mental state but a feature of the particular mental states that we have, Hume’s problem would still arise. Such a feature would be a feature only of particular mental states, and so could not explain the unity that those particular mental states have as “my” mental states. The problem is that it looks like no feature of the mind, nothing that could itself be an object of awareness could explain its nature as mind, its nature as understanding, as an “I”. In this sense, the mind, or mindedness, is not a “something” – but of course it’s not a “nothing” either. It’s not a “nothing” because what is distinctive about mindedness is the special way in which the causality of the states of mind must be understood: their causality must be understood in terms of the subject’s own understanding of why she holds what she does. There cannot thus be a “mark” that explains mindedness, such as primitive normative awareness, because the difference is a difference between two irreducible forms of understanding why certain things in the world are as they are. Suppose, then, that Max has a perceptual representational awareness of a cube as a cube, if so he must be poised, ceteris paribus, to judge that what he sees is a cube. And he cannot be in a position to make that judgment unless he is already in possession of the reason for so judging. But then mustn’t he already have the concept cube in order to be perceptually aware of something as a cube? Haven’t I then just impaled myself on one horn of Ginsborg’s original dilemma? I am not so sure. It is true that he must be in possession of the reason for making the corresponding judgment if he is to have an experience as of a cube. However, his appreciation of the reason could be, as Descartes and other Early Modern Philosophers, would have put it “obscure and confused”. That is, his ground would simply be that there is a significant similarity between the object that he judges to be a cube and certain other objects, which in fact are cubical in shape. What he would lack is an articulation of the criterion that would determine the subsumption of these objects under the concept cube, as he understands it. In order then to represent

Ginsborg on the Normativity of Perceptual Representation

271

something as a cube, Max must in that sense have the concept ‘cube’. But that doesn’t mean that he must already have the concept cube in the more ambitious sense of having a criterion for subsumption under the concept. This point, of course, has nothing to do with whether the concept cube is innate. Max would not have formed the concept cube if he had never encountered cubes. What he needs to bring to his encounter with cubes if he is to form the concept cube is the general conceptual capacity I sketched above.32 To have, then, a complete understanding of the basis of our judgments, to have what early modern philosophers like Descartes and Leibniz called a “clear and distinct” conception of the subject matter of our judgments is simply to perfect a capacity that we must already possess, however obscure and confused our conceptions are. This is why these philosophers held that the human understanding is in its basic form is indistinguishable from God’s understanding. To acquire a concept is then not, in a sense, to learn something, as we do when we learn a new fact, but to develop something we already have – the learning here is in a different register, echoing Plato’s notion that learning is “recollection”. To acquire a concept is not for something new to “enter” one’s mind: it is to acquire a clearer understanding of the judgments that one already makes. In that sense any attempt to get a better understanding of the basis of our judgments about the world is both reflection on the nature of the world and self-reflection.

References Anscombe (1957): Gertrude Elizabeth Anscombe, Intention, Oxford. Descartes (2006): René Descartes, Meditations, Objections, and Replies, R. Ariew/D. Cress (eds.), Indianapolis. Ginsborg (2008): Hannah Ginsborg, “Was Kant a Non-Conceptualist?”, in: Philosophical Studies 137, 65 – 77. Ginsborg (2006): Hannah Ginsborg “Aesthetic Judgment and Perceptual Normativity”, in: Inquiry 49 (5), 403 – 437. Hume (1993): David Hume, An Enquiry Concerning Human Understanding, E. Steinberg (ed.), Indianapolis. Kant (1800): Immanuel Kant, Logik – ein Handbuch zu Vorlesungen, G. B. Jäsche (ed.), Königsberg. 32 Obviously, there is much more that needs to be said about the general conceptual capacity as such and how it is able to give rise to cognitions of the world. Kant makes the best beginning on such a project that I know of.

272

Arata Hamawaki

Kant (2003): Immanuel Kant, Critique of Pure Reason, New York. Kant (2000): Immanuel Kant, Critique of the Power of Teleological Judgment, Cambridge. Marcus (forthcoming): Eric Marcus, Rational Causation, Cambridge/MA. McDowell (1994): John McDowell, Mind and World, Cambridge/MA. Rödl (2007): Sebastian Rödl, Self-Consciousness, Cambridge/MA. Wittgenstein (1953): Ludwig Wittgenstein, Philosophical Investigations, Oxford.

The Internal and the External in Knowledge Jocelyn Benoist One of the core issues in the contemporary debate on knowledge is externalism. It might seem obvious that knowledge conceptually requires some kind of externality. What is knowledge but a relation to the things as they are? Now in such relation it seems to be presupposed that those things are that – what they are – independently of the so-called knowledge. Of course, there are complicated cases in which such independence seems difficult to maintain. There are things that are what they are because they are being known. Furthermore what could be said about the situation when one knows one’s own knowledge? One cannot exclude such cases of reflexive knowledge, and it is far from obvious that such reflexivity should not be absolute and should necessarily drag us into an infinite regress. All that does not diminish the fact that in the concept of knowledge we find basically involved that idea of ‘the things as they are’, as something that, given the process of knowledge, is not negotiable. Perhaps, it might turn out that it is what it is as a result of the very process of knowledge. But, that process being given, the things are just thus and so and their being thus and so defines some kind of absolute norm for knowledge: something like a logical exteriority, to which knowledge is represented as having essentially to conform. Now, if what is to be known essentially has that kind of logical independence from knowledge, it might sometimes turn out that the independence is not only logical, but also factual. It is even a quite common case: as a matter of fact, everywhere when it is about what the tradition calls ‘external knowledge’. Such label is deeply equivocal and might rest on very dubious ontological assumptions, starting with the prejudged division between an ‘interiority’ and an ‘exteriority’ as two separate ontological domains. Nevertheless, independently of such assumptions I am not going to discuss here, it is probably possible to give some kind of minimal, merely logical sense to such division: if, for instance, we define the ‘external

274

Jocelyn Benoist

knowledge’ precisely as the one that is characterized by the factual independence of its object in relation to itself. That case of ‘external knowledge’ is quite general. It is even possible to say that it is the common basic case of knowledge – therefore the one on which the general sense of ‘knowledge’ largely depends. Very likely, there is some phenomenological primacy of factual independence in our capacity to make sense of the logical independence belonging to the object of knowledge as such. Now, wherever there is factual independence, something or even a lot of things might happen. In other words, what is known can change, independently of knowledge, and it is necessary to assess the consequences of such changes on knowledge. Herein lies the problem of epistemic externalism: in the constitutive exposure of knowledge to the factual externality of its object, that is to say to something that knowledge cannot entirely control – which seems unavoidably to result from the objective external commitment of knowledge as knowledge of ‘the things as they are’. Let us go deeper into the sense of that ‘externality’. If knowledge, or at least some basic knowledge, is essentially knowledge of something that is independently of such knowledge, it seems that, in that case, we have to distinguish between the thing’s being known and the thing’s being simpliciter. It is clear that if the thing is external to knowledge, there might be in the thing a lot of sides and aspects that there are not in knowledge. There are more things in heaven and earth, Horatio, Than are dreamt of in your philosophy.1

Furthermore, there is no knowledge but by adopting a certain perspective: what is of interest in the cow is not the same for the farmer, the artist or the butcher; so to know the cow (as a cow) is certainly not the same for all of them. It does not mean that for each of them, that kind of knowledge included in the other perspectives is no knowledge, but it is not what she would immediately include in what she would call ‘to know the cow’. Each piece of reality allows an infinite variety of perspectives, and, as such, certainly lies beyond the take one definite knowledge might have on it. That limitation is not any flaw of knowledge: it is just its definition. Knowledge, as such, is perspectival – is ‘knowledge from a certain point of view…’ 1

Shakespeare: Hamlet, Act 1, Scene 5.

The Internal and the External in Knowledge

275

Now, the fact either that not everything is known in the object or that one particular angle of knowledge by definition leaves room for other possible angles of knowledge beside itself, does not entail that a particular instance of knowledge by itself is not any knowledge. The intrinsic limitation of knowledge is not the negation of knowledge. To know the front of a building is not to know its back and, to some extent, not to know the building as a whole – at least with the completion that might be contextually required. There are always an infinite number of things to know about something, and so, absolutely speaking, there remains always something more to be known about it. On the other hand, to know the pig as a farmer is certainly not to know it as a naturalist. However, when my friend Denis insists on the fact that only the biologist could really know the pig, I cannot make sense of what he says, since it is clear that, in another sense, the farmer knows it. The whole matter here depends on the kind of knowledge that is contextually expected, and relevant. That problem of relevance is a core issue of any epistemology. In other words: there is always an infinite number of senses in which something can be known. Anyway, to know the front of the building, whenever that front is considered from the cognitive point of view, i. e., as a possible object of knowledge, it is already to know something, and the fact that there are certainly other things to be known surely does not cancel that piece of knowledge as such. On the other hand, to know the pig from the particular point of view of the farmer, which is not the same as the one of the naturalist, and which contextually proves either weaker or stronger than the latter, it is certainly to know something as well – even if to know the very same thing could as a matter of fact take on a lot of other meanings. So, the externality of some basic kind of objects, that requires from us first some modesty (there is always something additional to be known in the object beyond our knowledge) and second some capacity of adjustment (there are diverse possible cognitive takes on the object and we have to find the contextually most relevant one), is no objection to the pieces of knowledge we have. The fact that all the knowledge we have is limited in both senses (in its scope as in its relevance), does not cancel it, but defines it in each of its pieces. However, it would probably be tempting to say that the fact that every knowledge is partial entails the possibility of the cancellation ex post if not of some pieces of knowledge, at least of what prima facie seemed to be such pieces. For what I learn about the thing on the

276

Jocelyn Benoist

basis of a further investigation can turn out to conflict with what I believed to know about it. So I might be led to discard retrospectively what I took first to be real pieces of knowledge. That possibility, apparently, can never be completely excluded when I say ‘I know’. How could I tell apart to know from to believe to know on a mere internal basis? The externality of the object appears then as something that can make the status of the mental state that refers to it uncertain. Is it knowledge? On the basis of the evidence I have, I might be pretty convinced it is, and, in fact, it is not: some additional evidence would belie it. As to that issue, Malcolm makes a strange distinction in a classical paper of his. According to him, we should distinguish two different senses of ‘know’: When I use ‘know’ in the weak sense I am prepared to let an investigation (demonstration, calculation) determine whether the something that I claim to know is true or false. When I use ‘know’ in the strong sense I am not prepared to look upon anything as an investigation; I do not concede that anything whatsoever could prove me mistaken; I do not regard the matter as open to any question; do not admit that my proposition could turn out to be false, that any future investigation could refute it or cast doubt on it.2

Either that distinction is merely psychological or it is very strange. No doubt there are degrees in our certainty about what we take to be our knowledge on a certain occasion. When one says ‘I know’, it happens sometimes that she can forswear that ‘knowledge’ under a very light pressure. However, is it really the case that we should distinguish between two kinds of knowledge, or two meanings of the verb ‘to know’ for that very reason? That is rather unsure, because to say to know involves some kind of commitment. To say to know it is necessarily to take a cognitive risk – a risk that is not involved, for instance, in ‘I believe’. When I say ‘I believe’, even if there is definitely some kind of cognitive claim involved in that, to some extent the door remains open and the risk is not really taken: a belief can definitely turn out to be mere belief – and not knowledge precisely – and remain a belief. As Austin observed, from that point of view, ‘I believe’ is certainly a descriptive expression: after all it says nothing but the presence in me of a certain mental state.3 That is not at all the case of the verb ‘to know’ in its can2 3

Malcolm (1952), 183. Austin (1961), 78.

The Internal and the External in Knowledge

277

onical use: when I say ‘I know’, I am not just describing the state in which I find myself, but I really claim to say something about how things are. The fact that I might be mistaken in that does not diminish for anything my commitment, which, to some extent, does not know of degrees. I can believe something more or less. In a different sense, I can know a matter (a field) more or less. But I cannot say that I know a particular piece of knowledge “more or less”. Either I claim to know it or I do not. If I am not sure, and if I am honest, I just do not say ‘I know’. In such cases I say: ‘I believe.’ Does it mean that when I say ‘I know’ what I take to be a piece of knowledge cannot ever be called into question? Definitely not: it is very commonly the case wherever what I claim to know is precisely an external object in the sense we have introduced. However the point is that, by saying ‘I know’, I shield what I say from such doubt, I claim it to be the case whatever might turn out. When I say that I know that there is a house in front of me, it is true that it might be a mere façade like on a movie-set, but by my saying that I know that there is a house in front of me, that possibility has been objectively excluded independently of whether I have ever considered it or not. Now, the fact that I have maybe, as they say, not always all the reasons to say that I know what I say to know, does not make such statement less a statement of knowledge. Even more: it does not necessarily make such piece of knowledge less a piece of knowledge. What does it mean ‘all the reasons’? In the previous example, it seems that, as far as I have not checked out by going around, I cannot be certain that the house has a backside, therefore is really a house. Along the same line, it seems that when my friend Denis, who is an inveterate naturalist, sees a chess player protecting her king, he wonders whether she is really protecting her king or just randomly moving, or even whether she is really a chess player or a human being at all. Nevertheless, it seems that if we have no particular reason to think that these could be mere ‘façades’, we take to see houses when we see ‘facades’ – in other words: we take to see: ‘facades of houses’. And we take to see a chess player protecting her king when the person in front of us makes a certain move in a definite context in what we take to be a chess game, if we have no particular reason to think otherwise. It seems that whichever ‘external knowledge’ you take, it might be the case that a condition that remains unquestioned in that alleged piece of knowledge is not met in fact; thus, that so-called knowledge turns out to be no knowledge, but mere belief. Then, we believe, but in

278

Jocelyn Benoist

that we are wrong – in spite of all the good reasons we might have to believe. However, when we say we know, as much wrong as we might be in saying so, that is not what we mean. We do not say we believe but something more and different. We mean that, whatever might be the case in fact, what we say then is not false. We affirm that the conditions are met, whatever they might be. As we do not know them exactly – we cannot know them exhaustively, as there is no exhaustive list of them, but only an open list – of course in that we take some kind of risk – a cognitive risk. What looks like a piece of knowledge may eventually turn out to be a mere belief. Nevertheless, that possibility will never equate knowledge and belief conceptually. To claim that one knows, it is just to close the open possibilities that what we believe might not be the case, those very possibilities that on the other hand seem to be intrinsic to the very idea of belief. A belief can definitely be true. It can also prove very strong. However, to treat it as a belief, it is exactly to leave open the theoretical possibility it might turn out to be false – even when subjectively there is no sense left for such possibility. When one says to know, that very theoretical possibility is discarded. Now the mere rejection of such possibility is not yet per se a sufficient attribute of knowledge. It seems that for me to claim to know, it is required that such exclusion (of any unknown invalidating condition) is justified, at least as far as I know. For instance, if I am visiting a western movie set and I have been shown the system of the bogus façades, I cannot say, seeing a façade on the set, that I know there is a house in front of me in the full bloodied sense the term ‘house’ takes on in the everyday life. In such context I cannot know it for sure on such basis and I must check it out (have a look at the other side) in order to know. On the other hand, it seems that I am perfectly authorized to say I know there is a house across the street on such basis in everyday life. The fact that I am authorized to say ‘I know’ does not mean yet that I am necessarily right to say ‘I know’. Certainly, there might have been demolition works in that block without my knowing, and the mere façade could stand so, between two real buildings. It would be however uncommon, if not impossible, and it is really unlikely. Thus, it makes sense for me to say ‘I know’, even if it might finally prove incorrect – and thus not to be a knowledge, but a mistake. On the other hand, it would not really make sense for me to say ‘I know’ if there is a possibility, obvious in the very context, that I have

The Internal and the External in Knowledge

279

not considered. Then, it would be logical for someone to ask me: ‘Do you really know?’ As long as I have not checked out what is required to be checked out by the very situation, as a matter of fact, I cannot exactly say ‘I know’: such claim is inappropriate. So, the point is that the notion of true belief, although the attribute ‘true’ brings in the required transcendence, does not equate per se to knowledge either. One might have a true belief that is contextually not sufficiently justified, so that it is really improper to call it knowledge – while for all that it is not less true. In the very idea of knowledge is involved some idea of justification. We know only if we have reasons to know, as far as such reasons are contextually required. That is the basis of the old Aristotelian dictum that to know the truth by accident is not to know the truth. In that sense, knowledge is never mere encounter with an exteriority. It supposes that one has made that exteriority one’s own, that she has internalized it to one’s knowledge, as something one has reasons to take to be what it is. Nevertheless, it remains that whenever knowledge is ‘external knowledge’ in the sense we have introduced, it is conversely never the case that justification, how much strong it might be, is enough to make a belief a piece of knowledge. We might contextually be perfectly justified to think something that turns out to be false. That is the basis of all that literature that has flourished in contemporary epistemology under the name of ‘gettierology’. To make the problem more relevant to our considerations, let us elaborate on a free adaptation of Gettier’s argument.4 Let us suppose that, on the funicular climbing up the Vomero, a clever Neapolitan pickpocket has relieved me from my wallet and then, seeing there is nothing valuable in it and fearing the consequences of my noticing, has put it back in my pocket, all that without my knowing. Do I know, then, that I have my wallet in my pocket? I have no reason to believe it is gone. I have even ‘all reasons’ to think it is there because I previously put it in there and we do not live in a world of objects that would keep vanishing. So, I rely on the presence of my wallet, and I feel perfectly authorized to say ‘I know’ where it is – which a double-checking would confirm, as the wallet definitely is where I believe it to be. 4

See Dutant/Engel (2005), 13 f.

280

Jocelyn Benoist

So, it might happen that, although we have all the reasons – as far as reasons seem to be required in this instance – to take the things to be as they are, in fact, we are still in some way misled in that: as those reasons, per se, might have been misleading. After all, the pickpocket could have kept the wallet. The fact that he put it back and that it is as if nothing happened does not mean that nothing really happened and that the wallet is to be found where it is exactly for the reasons why I believe it still to be there. In that story, of course, everything is contextual. If, for instance, we participated in a game in which the rule is to take the other’s wallet stealthily as many times as possible, each of us would never be absolutely sure to have her wallet, and she will questionably be ever able to say that she knows to have it. And if David Avadon calls me on stage, how could I ever say to know? Then, it is not that difficult to transpose such cases to real life: let us suppose that in some part of the city, pickpocketing has grown absolutely systematic, with so dexterous and careful pickpockets (always putting back the wallet after helping themselves) that they commonly go unnoticed on the fly. In this case, what would it mean to know that one has one’s wallet? The point is that, under such conditions, when one says to have one’s wallet, she is definitely not wrong: she has it – even if maybe not with all that she places in that word: the wallet might have been emptied from its content. However, when one says to know to have one’s wallet, she is not exactly right: because she does not have it the way she takes to have it. This result sounds quite paradoxical, as it would seem reasonable to endorse some claim of transparency about knowledge: if I am right in taking that p, do I not know that p? Sometimes, it would be tempting to say: ‘she knows, even if not for the good reasons’. And after all, contextually, it is not that absurd. Sometimes what counts, is just the fact that she knows, independently of whether it is for perfectly good reasons or not. So, it is not absolutely impossible to talk, in certain circumstances, of ‘knowledge by mistake’ – or: on the basis of a mistake. Everything depends on where the emphasis is put on: either on the mere grip exerted on a definite fact, or on the way such grip is based. What counts or not as a knowledge is not always that clear. However, this relativity of what counts or not as a knowledge, one more time, is not a general point: in a particular circumstance, it is usu-

The Internal and the External in Knowledge

281

ally pretty clear if some claim would count as a knowledge or not without reasons adequate to the real state of affairs. It depends exactly on what that alleged knowledge claims on the one hand, and, on the other hand, of how odd are the circumstances relative to the kind of usual background of such ‘knowledge’. Of course, as everything that is odd, the challenge to our usual standard about what the piece of knowledge in question amounts to, cannot always be anticipated and some ‘hard cases’5 can come up. However they are logically the exception – otherwise the very idea of knowledge would not make any sense. If what we have called ‘external knowledge’ is real knowledge (as it certainly is, as it seems to largely determine the sense of ‘knowledge’ in general), that is to say knowledge of the things as they are, then it puts us in touch with the very externality of the things. Thus, it is exposed to the vicissitudes of that externality, to the extent that we might even wind up deceived in thinking that what is on the other hand a true belief is a piece of knowledge. Because it happens that the very basis of our knowledge might be undermined without our noticing. However, it would be incorrect to draw the following consequence: that, since no external knowledge is absolutely immune to that possibility of ‘cancellation’ (as knowledge and reassessment as ‘mere true belief’), there is no such thing as an external knowledge. Something can be cancelled only if its existence makes sense and on the very ground on which its existence makes sense. So, we must be careful about what we might call a perverse generalization of the argument. Let us suppose that the objects of the world are intermittent on an infinitesimal temporal scale, in such a way that we can never be phenomenologically aware of that discontinuity. That supposition, perhaps, sounds weird. But is our problem not about the possibility, that we can never absolutely exclude, of weirdness in nature? After all, who knows? Now, is really such a hypothesis an objection against our very knowledge of objects enduring in time? In some sense of endurance, in which microphysicists are interested, it certainly is. But the good question then is: is it that kind of endurance that is at stake in our ordinary knowledge of enduring objects? It seems it is not: even if the existence of the object suffers some kind of ‘microbreaks’, below the threshold of our perception, what counts is the global 5

In Dworkin’s sense: see Dworkin (1977).

282

Jocelyn Benoist

macroscopic continuity of the object we see, we touch, we deal with. It is exactly what we mean usually by ‘endurance’. Now if the conditions of such ‘endurance’ are satisfied, we certainly know an enduring object as such. On the other hand, if I substitute another goldfish for my daughter’s one without her knowing, one cannot say that she knows to have still the same goldfish – because, in the very sense in which she means it, it is just not the same goldfish. And if, after substituting one fish for another, I put back the first fish into the tank, it is not exactly possible to say either that she knows to have still the same fish – even if she is right, and justified, as far as there is no special reason for her to think I would behave in the strange way I effectively do, in believing it is the same fish. So, every single piece of knowledge certainly has real bounds: it is always the case that some special circumstances can disqualify it as knowledge. These bounds however so to speak belong to the same level as the knowledge in question itself. The special circumstance that can create a gap in the justification has to be such as to affect the relevant justification itself. What happens for instance ‘below’ that level just does not count. Of course, it is not always that easy to sort out the levels, and that is the part of what ‘externality’ is about. Now, that possible occurrence for each piece of knowledge of factual bounds at the very level at which it claims something – a possibility that always remains open – does not mean that we should complete every single statement of knowledge by some qualification in the way Austin describes in an ironical passage of his essay ‘Other Minds:’ If you are aware you may be mistaken, you oughtn’t to say you know, just as, if you are aware you may break your word, you have no business to promise. But of course, being aware that you may be mistaken doesn’t mean merely being aware that you are a fallible human being: it means that you have some concrete reason to suppose that you may be mistaken in this case. Just as “but I may fail” doesn’t mean merely “but I am a weak human being” (in which case it would be no more exciting than adding “D.V.”): it means that there is some concrete reason for me to suppose that I shall break my word. It is naturally always possible (“humanly” possible) that I may be mistaken or may break my word, but that by itself is no bar against using the expressions “I know” and “I promise” as we do in fact use them.6

If we say we know, it is that there is not any reason for us to think that there should be a ‘special circumstance’ of the sort. Of course that can6

Austin 1961, 98.

The Internal and the External in Knowledge

283

not mean that there is not. It just means for sure that if there is one, we are not aware of it. Nevertheless, to believe that an exact formulation of knowledge should include such qualification, it is just to miss what knowledge is. To claim to know something, it is just to exclude any possibility of the sort. On the other hand, however, as a consequence of the fact that that knowledge – which is what we usually mean by ‘knowledge’ – is a grip on externality, we can never absolutely exclude such possibility. So, there is no knowledge that has not its unquestioned basis of certainty, a basis that does not found it as a knowledge, but that makes only possible its meaningful foundation, constitutes, so to speak, the background of that foundation. Such background comes to the fore only when it is no longer obvious that the conditions of the ‘knowledge’ we claim to have are met. Then, what had not counted so far as a real piece of knowledge (that, for instance, it would not be possible for certain objects to disappear and to come again into being) can, under the pressure of the events, wind up counting as one. But, at each step, to know is just to deal with that externality of what is not known as such and constitutes the background of real full-bloodied knowledge.

References Austin (1961): John L. Austin, Philosophical Papers, Oxford. Dworkin (1977): Ronald Dworkin, Taking Rights Seriously, Cambridge/MA. Dutant/Engel (2005), Julien Dutant/Pascal Engel, Philosophie de la connaissance, Paris. Gettier (1963): Edmund L. Gettier, “Is Justified True Belief Knowledge?”, in: Analysis 23, 121 ff. Malcolm (1952): Norman Malcolm, “Knowledge and Belief”, in: Mind, LXI (242), 178 – 189.

III. Knowledge in Science

Genesis of Knowledge Spaces and Objects of Knowledge1 Hans-Jörg Rheinberger Over the course of the past three decades experimental practices, and with them the spaces and objects of knowledge belonging to them have shifted into the focus of research in history as well as in philosophy of science. There has even been talk of a ‘practical turn’ in this respect. In view of the fact that the experiment has been regarded as the genuine trademark of modern science, at least since Francis Bacon’s Novum organum scientiarum, it is somehow astounding that the philosophy of science and along with it the history of science have been so hesitant and late in opening up to an in-depth analysis of experimentation and the corresponding material dynamics of knowledge production. My thesis is that there are reasons for this historical belatedness that have more to do with the spaces, the objects, and the practices of knowledge production themselves than might at first sight be assumed. For the gestures of knowledge are, as Michael Polanyi insistently showed, implicit and tacit to a decisive extent. The objects of knowledge in a narrower sense are, as a rule, inconspicuous, not very spectacular things. One might say they are not plain and obvious. And modern laboratories are, as the sociology of social spaces teaches us, not public regions but rather separate and separated locales. So the layperson involuntarily thinks of something mysterious when the word laboratory is uttered. The closed places of the modern labs, then – the emblematic knowledge spaces of modernity – are the paradoxical result of a science that programmatically devoted itself to the public presentation, to transparency and the traceability of its results. If modern science, in the course of history, has taken shape by creating a knowledge horizon that is epistemically open and, according to its political self-understanding, public, it also has generated, in the process of establishing this form, precisely those separate areas that it obviously needed in order to practice its own openness properly in the first place. I shall address one further as1

Revised English version of Rheinberger (2003); cf. etiam Rheinberger (2001).

288

Hans-Jörg Rheinberger

pect of this dialectic of opening and closure once more at the end of my paper. With Michel de Certeau we can distinguish places from spaces of knowledge.2 A place creates an order. It guarantees that the things in it are located at their proper position. It situates elements in relation to one another, creates stability and structure. One might accordingly call the institutes, institutions and instantiations of science places of knowledge. Case studies as they are known from the history of science are usually concerned with such places: say, the California Institute of Technology, the Kaiser Wilhelm Institute for Biology, the University of Göttingen perhaps, or the French Academy of Sciences. A space, however, is instead a dynamic network of mobile elements. It depends upon the operations that orient it, the times that structure it, the circumstances under which it can unfold. Laboratories and botanical gardens, study rooms and urban biotopes are examples of such historically changeable spaces of knowledge. These are what concern the epistemologist who would like to understand the dynamics of knowledge production. I have called the objects of knowledge epistemic things. They are the things toward which the efforts of knowledge acquisition are oriented, not necessarily objects in the narrower sense of the word. As epistemic things, they present themselves in an indeterminacy that belongs to their very character. They are not readily graspable in the context of the technical conditions – technical things – under which they are explored. What we usually call a model occupies a middle position between indeterminate epistemic things and specified technical things. In the theory of science, there has been much talk of models and of analogies and metaphors when it comes to the linguistic configuration of models. Models also have a material side, though. As material objects of knowledge, they are, as a rule, so well-established that they seem attractive and promising to researchers. On the other hand, they are normally not stabilized and standardized well enough to function as a part of the technical boundary conditions of an experimental setup. To borrow from Georges Canguilhem: a model is a knowledge object that still leaves something to be desired. It works precisely by means of the knowledge-promoting deficiency resulting from the transmission from one medium to another that is characteristic for models.3 2 3

Certeau (1984). Canguilhem (1968).

Genesis of Knowledge Spaces and Objects of Knowledge

289

As long as the various forms of the analytical philosophy of science accompanied and dominated reflection on science and technology, the specific details of what happened in the material realm of these spaces and objects, thus apart from the knowing subject and its linguistic utterances, could hardly appear to be of significance – ‘logically’, we would have to say. At most, the context of discovery was important for biographers and creativity researchers. Among the ranks of the philosophers it was only the phenomenologists who took certain notice of them, at least to the extent that they were interested in the relation of everyday experience to scientific knowledge. Sustained change in this situation came in the mid-twentieth century. Names such as Ludwik Fleck, Gaston Bachelard, Michael Polanyi, and Thomas Kuhn are closely associated with the efforts to free the experiment from the tight corset of hypothesis testing to which it had been tied by the hypothetico-deductive schematism, and to prepare the ground for a conception of the experiment as something much more comprehensive, that is, as a scientific way of life. Since then experimentation in its full richness has no longer been reducible to a testing procedure. Its function as an empirical procedure of control was no longer its most interesting aspect. It became much more important to understand the specific historical constellations that went along with the process of knowledge production – constellations of instruments, procedures, epistemic objects, model organisms, knowledge spaces, and skills. These constellations can take on very different forms throughout history according to the specificity of the elements that go into them. I am interested in identifying such constellations, describing them in detail and integrating them into what one might call a topology of knowledge. This does not involve the presupposition of a certain idea of science with more or less well-defined disciplinary boundaries. What has to be understood, rather, is how such amalgamations of skills, knowledge spaces, model organisms, epistemic objects, procedures, and apparatus permanently shape and reshape scientific practice and with it the concept of scientific practice and what it means to do science in the first place, what counts as a scientific fact, and what counts as a result that is worth sharing with the scientific community and the public. Thus this way of life, ‘experimental knowledge acquisition’ thus, has an epistemological core that Gaston Bachelard had already plainly exposed. In his work in the late 1920s and early 1930s Bachelard had in-

290

Hans-Jörg Rheinberger

sisted that the modern natural sciences must be seen as “phénoménotechniques”, to use his term.4 That means that the natural sciences themselves technically produce at least the non-trivial and thus interesting phenomena that they investigate in the laboratory. The phenomena with which they concern themselves are, as a rule, not given in everyday experience and hence not immediately accessible to our senses. Bachelard argued that the seemingly most simple and most fundamental processes the sciences are concerned with, thus the elementary phenomena of what he called the “scientific real”, are precisely the most derived ones, as they are usually most strongly subjected to the phenomenotechnical efforts to represent them in pure form. Thus, for Bachelard the Cartesian philosophy of the evidence of clear and distinct ideas becomes practically inverted. Consequently, he speaks of a non-Cartesian epistemology in which the positions of both the object and the subject have lost their character as absolute terms of a dominant relation. The clear idea no longer stands at the beginning, and therefore it cannot lead the process of knowledge production; it is rather the result of this process, the product of an arduous purification procedure, both of the epistemic objects and of the ideas about them. Finally, it is characteristic of Bachelard’s epistemology of detail that it addresses the ineluctable inertia and blurriness of the scientific mind under the concept of the “obstacle épistémologique”. According to Bachelard, the researcher must outwit himself in an act of phenomenotechnical transcendence in order to gain insights that are not accessible to his anticipatory imagination, precisely because they are located beyond one’s everyday experience.5 This meets with a twofold resistance. There is, first of all, the constitutive vagueness of the ideas that drive research and that can only become more defined in the course of empirical work, and, secondly, the frustrating complexity of the processes of representation and purification of epistemic objects that go along with these efforts of ever more precise determination. In his 1928 essay on approximative knowledge – Essai sur la connaissance approche – Bachelard emphasises: From our point of view, this reality, in its inexhaustible supply of the unknown, presents a character that lends itself to research without end. Its en-

4 5

Bachelard (1970), 18 f.; cf. etiam Bachelard (1951), 3. Bachelard (1938).

Genesis of Knowledge Spaces and Objects of Knowledge

291

tire being resides in its resistance to knowledge. So we can see it as a postulate of epistemology that knowledge is fundamentally incomplete.6

And he generalizes by emphasizing the moment of surprise, of the unforeseen in the course of knowledge generation: “The history of the sciences teaches that every great step made toward an ultimate reality has shown this reality to be found in a quite unexpected direction.”7 At this point I would like to take a moment to allow the recently deceased great theoretician of praxis, Pierre Bourdieu, to say something about this. What he writes about practice in general in his last book, Science de la science et rflexivit, is most especially applicable to scientific practice, which, to a large extent, withdraws itself from the space of media presence and public representation: [Practice], says Bourdieu in his farewell lecture at the Collège de France, is always underestimated and under-analysed, and yet understanding it requires much theoretical competence, much more, paradoxically, than understanding a theory. One has to avoid reducing practices to the idea one has of them when one’s only experience of them is logical. And, for lack of an adequate theory of practice, scientists are not necessarily able to invest in their descriptions of their practices the theory that would enable them to have and to give a real knowledge of these practices.8

Bourdieu describes the principle of scientific practice as a system of “generative dispositions”,9 most of which function unconsciously, and which are transferrable, and can expand. The analysis of such generative dispositions is precisely at issue in the understanding of the genesis of knowledge spaces and knowledge objects. What are we to do then? How does experimental practice disclose itself especially to the historian of science? One possible approach to the field of the experiment is the replication of instruments and of experimental instructions, as it was successfully carried out e. g. in Falk 6

7 8 9

“A notre point de vue, cette réalité présente dans son inconnu inépuisable un caractère éminemment propre à susciter une recherche sans fin. Tout son être réside dans sa résistance à la connaissance. Nous prendrons donc comme postulat de l’épistémologie l’inachèvement fondamental de la connaissance” (Bachelard (1928), 13). “L’histoire des sciences enseigne que chaque grand progrès fait vers une réalité ultime a montré que cette réalité se trouvait dans une direction tout à fait inattendue” (Bachelard (1928), 284). Bourdieu (2004), 39. Bourdieu (2004), 41.

292

Hans-Jörg Rheinberger

Rieß’ group and by Otto Sibum.10 The efforts at reconstruction and the gaps that open up in the process then become a source of knowledge beyond its printed form. Another source for the understanding of scientific practice is disclosed in laboratory sociology, as developed especially by Bruno Latour, Michael Lynch, Karin Knorr Cetina and other sociologists of knowledge.11 These authors looked closely at how science is made under ethno-methodological aspects. It is not possible to go into more detail about these approaches here. Instead, I would rather like to briefly present yet another approach to experimental practice. It consists in the analysis and reconstruction of laboratory protocols and laboratory notes, these forms of writing up and rewriting at the periphery of the experiment itself, in the production of knowledge, that are located in the spaces between the printed text and the material event of experimenting. Research notes can bring the historian into relatively close proximity to the epistemological core of the everyday life of the laboratory, and thus to an understanding of the genesis of knowledge objects. Although they are often extraordinarily difficult to evaluate and interpret, as documents, they communicate something of that unconscious addressed by Bourdieu, and in a way, ex negativo, something of the constraints that the public side of knowledge imposes on it. The historian of science and medicine Mirko Grmek12 demonstrated in the late 1960 s and early 1970 s using the example of the lab books of the French physiologist Claude Bernard that Bernard’s own description of the experimental method in his book Introduction la mdecine exprimentale [Introduction to Experimental Medicine] of 1865 concealed almost everything that Bernard himself practiced in the laboratory. While his publication neatly conformed to his public campaign for the recognition of physiology as a leading biological science and to philosophy of science of his time, the notes in his journal amount, in contrast, to a comparison of the activity of the experimenter with that of the artist, and stress precisely the character of experimenting as ‘feeling one’s way’. Such studies allow, at least rudimentarily, for a reconstruction of the procedure of experimentation that scientists themselves follow mostly in an intuitive manner. Laboratory journals can serve as documents that can be used to study the dynamics of knowledge generation and that 10 See, e. g. Sibum (1995); Heering/Rieß/Sichau (2000). 11 See, e. g. Latour/Woolgar (1986); Knorr Cetina (1999); Lynch (1997). 12 Grmek (1981).

Genesis of Knowledge Spaces and Objects of Knowledge

293

can disclose implicit knowledge strategies – if the concept of strategy here makes sense at all. “If we are to understand scientific activity at its core”, science historian Frederic Holmes sums up, “we must immerse ourselves as fully as possible into those investigative operations”.13 Using the physiological laboratory journals of Antoine Lavoisier as well as the young Claude Bernard, Holmes showed, for example, that these researchers always pursued several projects simultaneously, and that often the results of one project unexpectedly led another project out of a dead end it was stuck in.14 With these networks, they created something like an experimental space in the sense of a materialized discovery context. The experimental work itself assumed the structure of a creative process. Such studies allow us to sketch procedures of experimentation and the generative dispositions (in Bourdieu’s sense) upon which they are based, and thus to understand the dynamics of the experimental sciences better. At this point I would like to turn to another connection that is important for our topic. Christoph Hoffmann and Peter Berz showed, using the example of the ballistic experiments of Ernst Mach and Peter Salcher, that studying the laboratory journals themselves as instruments of research provides a further perspective in historical laboratory studies,15 i. e. seeing them in their positive function as primary forms of recording in writing that organize not only the knowledge process, but also the everyday life in the lab, and thus the space of knowledge production as a whole. Laboratory books are hence not simply taken as written traces of experimental activity, but are themselves analyzed with respect to their function as research tools in their own right. The character of working notes as both epistemological and literary tools remains yet to be explored. They form a space in which data, the primary yet neglected form of knowledge things, take on shape, and are transformed into elements of scientific discourse. To put it in the words of Michel Foucault in his Archaeology of Knowledge: The document, then, is no longer for history an inert material through which it tries to reconstitute what men have done or said, the events of which only the trace remains; history is now trying to define within the documentary material itself unities, totalities, series, relations.16 13 14 15 16

Holmes (1985), XVI. Holmes (1974). Hoffmann/Berz (2001); more recently: Hoffmann (2008); Wittmann (2009). Foucault (2002), 7.

294

Hans-Jörg Rheinberger

These literary products of research in their “documentary materiality”, these “documentary masses”, as Foucault also called them, circumscribe and organize a knowledge space that has already ceased to belong entirely to the material realm of the experimental configuration itself, but has not yet passed over into the space of printed communication. This middle ground displays paradoxical features. On the one hand, the laboratory’s records, as primary data, are still a part of the research object, something like the skin of its possible meanings. On the other hand, they must strip off the features of tentativeness and indeterminacy that they display here, and that actually make up their epistemological productivity, at the very instant when they become elements of a text addressed to the scientific community. It is in this middle space, in this hybrid region in which research objects have not yet become paper entirely, and in which the paper is still part of the material experimental event, that one must look if one wants to understand how scientific discovery works and thus the generation of novelty, if one would like to grasp the experiment, to put it in Friedrich Steinle’s terms, in its disclosing or exploratory form rather than in its demonstrative form.17 In the laboratory texts, in these numeralia and literalia of research, the way in which the esoteric and the exoteric sides of science intertwine becomes accessible and analyzable for the historian. Not least of all, laboratory notebooks can serve to make it clear that the transition from labyrinth to arena, from laboratory to public, can be an unsure terrain, one in which risk and deception, purification and distortion often lie close together. In one of his last books, Frederic Holmes showed us on what fragmentary and uncertain experimental ground Lavoisier’s publications from the years 1773 and 1774 rested, which after all led to his triumphant general attack on the phlogiston theory.18 Holmes shows using Lavoisier’s example how difficult it is in principle for every experimenting scientist to find the golden mean between experimental gathering of evidence and overreaching, but research-driving generalizations. I am concerned here to give sharper definition to the internally conflicting relationship in which the esoteric and the exoteric space of science, the practice of research and the presentation of its results, stand to one another. To get access to this inner tension, to characterize its forms, and to grasp it in its productive function is 17 Steinle (2005). 18 Holmes (1998).

Genesis of Knowledge Spaces and Objects of Knowledge

295

what I see as one of the essential tasks of a history of science from an experimental perspective. In the final section of this paper, I would like to turn briefly to a case study.19 The case in question concerns the early hereditary experiments of Carl Correns (1864 – 1933). Correns was a pupil of the Swiss botanist Carl Nägeli in Munich, and later went on to become the first director of the Kaiser Wilhelm Institute for Biology in Berlin-Dahlem. Over the course of several years I have attempted, with the help of the notes and records still preserved in Dahlem, to reconstruct the experiments with peas and corn that Correns conducted in the final decade of the 19th century in the small botanical garden of the University of Tübingen and in the surrounding rural botanical nurseries. As is well known, these cross-breeding experiments led to the rediscovery of Mendel’s laws, and thus inaugurated the age of genetics. For me, the starting point was an excerpt found between laboratory records from which it follows that Correns, contrary to his own statements published 25 years later, must have already read Mendel’s article of 1865 by April of 1896 and not late in 1899, after his experimental work was finished. The point in time at which he read the article even dates exactly down to the month in which Correns planted the first generation of his peas in the botanical garden of Tübingen.20 But whoever supposes, though, that this finding simply robs yet another of Mendel’s so-called rediscoverers of his legend – a legend he himself actively helped to construct after the fact – is in for a disappointment. The records from 1896 to 1898 offer no indication that Correns might have been noticeably influenced in the proceedings of his experimental cross-breeding work by the study of Mendel’s work. He rather sought undauntedly for xenia, i. e. properties of pollen-giving plants that already present themselves in the seeds of the mother plant, and not only in the next generation. The garden pea appeared to him, according to reports in the literature, to be just as good an experimental material to follow this phenomenon as corn, on which he had already worked for two years in 1896. My reconstruction using the records leads to the conclusion that Correns was not at all immediately aware of the significance of the splitting of traits observable in the second generation of his hybrids, in spite of having read Mendel’s article. And the first indications did not come from cross-breeding peas, as one might have supposed had he followed Men19 Rheinberger (2010), ch. 4. 20 Rheinberger (1995).

296

Hans-Jörg Rheinberger

del’s results, but from a parallel series of experiments with various strains of corn, also conducted in search of xenia. These indications themselves in turn did not come from a carefully planned experiment directed at a preconceived idea, but were the result of accidental back-crosses. This can be retraced in the vocabulary of the notes: one of the lines used in crossing had proven “impure”, i. e. it was already a hybrid itself. What remained at first a subtle hint in the corn experiments became ever more pronounced in the following two years in the peas, in which, on the other hand, all efforts undertaken to discover xenia and clarify their cause met with failure. My general interest in this case study, as well as in earlier ones on the history of molecular biology, was and is21 to cast light on the question of how productive experimental systems and the modes of representation and objects of knowledge connected with them arise and shift, how such systems unfold their own dynamics, expand, and lead researchers in directions and to results that they could not have foreseen at the outset of their work. Correns appears to me in this respect to present an especially interesting example for the very reason that he theoretically could have foreseen the results of his work. He had a decisive source of information directly before his eyes when he began his experimental system. This source of information, though, did not seem to make much sense to him from the perspective of his own starting point. His epistemic object was not the rules of hybridization, but the processes that lead to the appearance of xenia. It can be seen as a lucky peripheral circumstance of this starting point, and one that became decisive in the course of the experiments, that Correns concentrated on traits – potential xenia – that were to appear in the seeds. For the seeds, inasmuch as they were collected and carefully preserved for sowing the next generation, presented an additional kind of material record. Correns could turn to this record, preserved in his pea cartons, whenever necessary, even after a year, or several years. He was thus able to examine the results of his crossings again if the need arose, or even to examine neglected traits more attentively, traits he originally had regarded as insignificant. This character of the system, namely that it made possible a recursive search for and securing of evidence regarding the object of knowledge itself, only displayed its virtues after Correns had found his way over the course of four years deep into the hybrid system of corn and pisum, at a time when it 21 Rheinberger (1997); Rheinberger (2010), ch. 6.

Genesis of Knowledge Spaces and Objects of Knowledge

297

began to dawn on him that the potential significance of his findings pointed in an entirely different direction than the one he had been taking. It seems to me that such material properties of experimental systems and the regressive, yet iterative interpretive moves that they make possible are well worth our attention, at least if we wish to understand how the laboratory, the experimental garden in this case, becomes a space for the generation of new scientific objects. Such investigations also lead us back to Bachelard’s problem of the epistemological obstacle, and thus to the fundamental question of the generation of novelty in the empirical sciences. What is at issue here is not, as Bachelard remarks in his “psychoanalysis of objective knowledge”, an “observation of external hindrances, such as the complexity and the fleetingness of the phenomena, nor a lamentation over the feebleness of the senses and the human mind: in the act of knowledge itself, in its innermost”, as Bachelard asserts, “there appear – due to a sort of functional necessity – weariness and confusion. (…) Empirical thought only becomes clear afterwards, when the apparatus of explanation has made its move”22. To retrace the dynamics of empirical reasoning in widely different spaces of knowledge, to locate the dynamics of the unfolding of epistemic objects in them, to follow the chess-like moves of the experiment and to understand its potential to surprise, is one of the tasks that history of science today has to tackle. Processes of standardization, of closure, as one might say, have attracted much interest in history of science over the past decades. Processes of opening, of reaching out to the new, which not only the modern sciences, but also modern societies in general live of and upon which they depend, deserve the same attention. They play out in a wide range of knowledge spaces, and the experiment stands at their very center. It is therefore no accident that François Jacob called experimental systems – a word that I have cited repeatedly – “machines for making the future”.23 The finer details of the mechanics and dynamics of such “machines” still has to be explored. 22 Cf. the original quotation in Bachelard (1938), 13: “Et il ne s’agit pas de considérer des obstacles externes, comme la complexité et la fugacité des phénomènes, ni d’incriminer la faiblesse des sens et de l’esprit humain: c’est dans l’acte même de connaître, intimement, qu’apparaissent, par une sorte de nécessité fonctionnelle, des lenteurs et des troubles. (…) La pensée empirique est claire, apr s coup, quand l’appareil des raisons a été mis au point”. 23 Jacob (1988), 9.

298

Hans-Jörg Rheinberger

This would all need to be presented in greater detail. We are yet far from a comprehensive history of experimental practice and even farther from a cultural history of experimentation. For this the category of the experiment itself requires to be thoroughly historicized. It is one of the preconditions of this endeavor that the topic be taken up and explored to its greatest possible extent throughout the whole course of the history of the sciences.24

References Bachelard (1928): Gaston Bachelard, Essai sur la connaissance approche, Paris. Bachelard (1938): Gaston Bachelard, La formation de l’esprit scientifique, Paris. Bachelard (1951): Gaston Bachelard, Le rationalisme appliqu, Paris. Bachelard (1970): Gaston Bachelard, “Noumène et microphysique”, in: G. Bachelard, tudes, Paris, 11 – 24. Bourdieu (2004): Pierre Bourdieu, Science of Science and Reflexivity, Cambridge. Canguilhem (1968): Georges Canguilhem, “Modèles et analogies dans la découverte en biologie”, in: G. Canguilhem, Etudes d’histoire et de philosophie des sciences, Paris, 305 – 318. Certeau (1984): Michel de Certeau, The Practice of Everyday Life, Berkeley. Foucault (2002): Michel Foucault, Archaeology of Knowledge, London (translation of Foucault (1969): Michel Foucault: L’Archologie du Savoir, Paris). Grmek (1981): Mirko D. Grmek, “A plea for freeing the history of scientific discovery from myth”, in: M. D. Grmek/R. S. Cohen/G. Cimino (eds.): On Scientific Discovery. The Erice Lectures 1977, Dordrecht, 9 – 42. Heering/Rieß/Sichau (2000): Peter Heering, Falk Rieß, Christian Sichau, “La reproducción de experimentos históricos”, in: Ciencia 286, 62 – 69. Hoffmann/Berz (2001): Christoph Hoffmann/Peter Berz (eds.), ber Schall. Ernst Machs und Peter Salchers Geschoßfotografien, Göttingen. Hoffmann (2008): Christoph Hoffmann (ed.), Daten sichern. Schreiben und Zeichnen als Verfahren der Aufzeichnung, Zürich. Holmes (1974): Frederic L. Holmes, Claude Bernard and Animal Chemistry, Cambridge/MA. Holmes (1985): Frederic L. Holmes, Lavoisier and the Chemistry of Life. An Exploration of Scientific Creativity, Madison. Holmes (1998): Frederic L. Holmes, Antoine Lavoisier – The Next Crucial Year. Or, The Sources of His Quantitative Method in Chemistry, Princeton. Jacob (1988): François Jacob, The Statue Within. An Autobiography, New York. Knorr Cetina (1999): Karin Knorr Cetina, Epistemic Cultures. How the Sciences Make Knowledge, Cambridge/MA. Latour/Woolgar (1986): Bruno Latour, Steve Woolgar, Laboratory Life. The Construction of Scientific Facts, Princeton. 24 Translation from German by Alan Duncan.

Genesis of Knowledge Spaces and Objects of Knowledge

299

Lynch (1997): Michael Lynch, Scientific Practice and Ordinary Action. Ethnomethodology and Social Studies of Science, Cambridge/MA. Rheinberger (1995): Hans-Jörg Rheinberger, “When did Carl Correns read Gregor Mendel’s paper?”, in: Isis 86, 612 – 616. Rheinberger (1997): Hans-Jörg Rheinberger, Toward a History of Epistemic Things. Synthesizing Proteins in the Test Tube, Stanford. Rheinberger (2000): Hans-Jörg Rheinberger, “Epestia. Le projet expérimental d’une génétique de la physiologie du développement dans l’oeuvre d’Alfred Kühn”, in: Revue d’Histoire des Sciences 53, 441 – 446. Rheinberger (2001): Hans-Jörg Rheinberger, “Wissenschaft zwischen Labor und Öffentlichkeit”, in: Helga Nowotny/Martina Weiss (eds.), Jahrbuch 2000 des Collegium Helveticum, Zürich, 159 – 175. Rheinberger (2003): Hans-Jörg Rheinberger, “Wissensräume und experimentelle Praxis”, in: H. Schramm et. al. (eds.), Bhnen des Wissens. Interferenzen zwischen Wissenschaft und Kunst, Berlin, 366 – 382. Rheinberger (2010): Hans-Jörg Rheinberger, An Epistemology of the Concrete. Twentieth-Century Histories of Life, London. Sibum (1995): H. Otto Sibum, “Reworking the mechanical value of heat: Instruments of precision and gestures of accuracy in early Victorian England”, in: Studies in History and Philosophy of Science 26 A, 73 – 106. Steinle (2005): Friedrich Steinle, Explorative Experimente: Amp re, Faraday und die Ursprnge der Elektrodynamik, Stuttgart. Wittmann (2009): Barbara Wittmann (ed.), Spuren erzeugen. Zeichnen und Schreiben als Verfahren der Selbstaufzeichnung, Zürich.

On the Epistemology of Models Bernd Mahr Explaining the Nature of Models What the Epistemology of Models Reveals Ontologically, a model is something as which something is being conceived of, and concretely, being a model is the content of a judgment in which something is being conceived of as a model. The epistemology of models concerns what is conceived of and what is implied if something is being conceived of as a model. The epistemology of models can be disclosed if situations and arguments are analyzed, by which judgments of model being are justified. The structure of model situations exhibits a pattern of contextual relationships on which these arguments rest. It forms a model itself, the model of modelbeing. The modeling of model-being intends to creating the fundament of a general theory of models, which can serve as a reference when analyzing models and discussing model situations, model use and modeling disciplines. 1 Debates on the Notion of Model Over the last few years, the fundamental question about the nature of models is once again being asked by many, a question which was discussed extensively in the Cybernetics and Systems Theory era of the 1960 s2. In those days not only the mathematical theory of models evolved internationally; general model conceptions emerged, too, such as Herbert Stachowiaks general theory of models3. To mention

1 2 3

This article partly makes use of text from published articles by the author, namely from Mahr (2009), 365 – 383; and from Mahr (2010), 61 – 87. See e. g. Freudenthal (1961); Hesse (1963) and the journal Studium Generale (1965). Stachowiak (1973).

302

Bernd Mahr

are also the contributions of Marx W. Wartofsky4. Just as in those days, the question about the nature of models concerns virtually every branch of science5. Particular interest in models is aroused by their interdisciplinary role and conception. This interest reaches programmatic clarity when it comes to the question of a uniform pattern of model use 6, one which the boundaries between the natural and engineering sciences and the humanities might transcend. These days, the fundamental question of the nature of models is not restricted to models in mathematics, the natural and social sciences, and in art, but among other disciplines also of interest in information science. Witness is borne by the large number of modeling techniques currently available in computer science and in information technology and applications, and by the lively discourse on modeling techniques and tools7. These phenomena indicate broad interest in the practical use of models. Not equally well expressed is an interest in the fundamental question of the nature of models. Models in information science are conventionally viewed in analogy to the models of architecture and construction8, or of mathematics, engineering, and the social sciences. A closer look at the rationale of information science, however, does show that models are not only to a large degree what information science is about and that the concept of model can not simply be imported from other sciences and disciplines, but that the question of the nature of models is at the very heart of information science: regardless of the perspective from which information science is viewed, models always play a dominant role; they form an essential part of the knowledge shared by its different scientific communities9 ; they are the dominant 4 5 6 7 8

9

Wartofsky (1979). See e. g. Morgan/Morrison (1999); Chadarevian/Hopwood (2004); Präsident d. Berlin-Brandenburgischen Akademie d. Wissenschaften (2005); Dirks/Knobloch (2008). See Bredekamp (2005), 13 – 20, see 13. See e. g. Reisig (2006), 369 – 385; Hesse (2006), 99 – 113; Broy/Rumpe (2007), 3 – 18. Leading information scientists have referred time and again to the significance of the writings of Vitruv, who defined, in the introduction of his decem libri de architectura, which (probably) appeared in the year 14 a.D., the general demands which should be made to an architect, and who established a theory of architecture, influential even today, in putting forward statements on architecture based on Greek Antiquity. See also: Kruft (1995). In Kuhn (1988), 389 – 420; Kuhn identifies symbolic generalization, models, and exemplary problem solutions as essential elements of the disciplinary matrix

On the Epistemology of Models

303

subject matters of research and teaching in information science, and in software engineering they are not only the goals of analysis and design, but to a large degree also form its methodological basis; and models are the entities in which the academic classifications of information science and its disciplinary subdivisions are reflected. What is particular about information sciences is that models determine more than anywhere else their capture of reality. As a consequence, the modes of statements made by information science are in most cases qualifications of a network of judgements about models. The fact that qualifications of models take on the role of a mode is attributable to the circumstance that models do not incarnate any form of truth, but rather forms of demonstrability, possibility, and choice10. Scepticism about a General Conception of Model Surprisingly, there is no obvious answer to the general question of the nature of models, and thus it is quite often dismissed as unanswerable. Is a substantive model concept conceivable, one might ask, comprising at the same time such widely differing phenomena as a beautiful woman, a system of differential equations, the architectural sketch of a church, the definition of Lambda calculus, a toy car, the drawing of a mechanical duck, and a UML diagram? It is often suggested that the word ‘model’ is a homonym, which is incidentally used in many widely differing senses. Even though all the phenomena in question are all called ‘model’, it is suggested that they have nothing else in common. Further it is suggested that a concept of model trying to capture such a variety of phenomena will in fact capture nothing at all, which means that it is methodically useless. And finally, it is suggested that when attempting to explain the concept of model, one should not exceed the disciplinary boundaries within which one may talk about models with some degree of clarity. Even if it is not expressed in that way that is nonetheless the way it is often thought about11. And in fact we cannot expect to find inherent characteristics of an object which are sufficient proof that the object in question is a of a scientific community, underlying its more or less unanimous scientific judgments. 10 Mahr (2004a), 161 – 182. 11 The view in question is found in several textbook or dictionary definitions. Explicitly it is found e. g. in Goodman (1995), 164, but also in Mittelstraß (2005), 65 ff.

304

Bernd Mahr

model. Not even functional paraphrases will suffice to characterize an object as a model12, the reason being that not every model possesses the type of characteristics often held to be distinguishing, such as being of smaller scale, simplifying, and representing, or such characteristics as sign-like, representative, and abstract. For example, atomic models are never as small as the atom of which they are models; and even though a hieroglyph is sign-like, representative, and abstract, from our point of view it is still not a model but a character. Taking into consideration the observations just made, is it at all possible to meaningfully answer the question of the nature of models, and, provided that an answer can be given, is it possible for a single discipline, such as information science or the history of art, for example, to benefit from such an answer? I claim here that the question of the nature of models can be meaningfully answered if the traditional ontology oriented methods of conceptual analysis are dismissed and instead the nature of models is explained in the wider context of intentionality as the content of a judgment. And I also claim that such explanation is beneficial for all disciplines as it provides a neutral and most useful analytical tool for any kind of model related studies.13 How the Nature of Models Can Be Explained Examples do easily show not only that an extreme variety of objects may function as models, but that in fact any object may function as a model. Thus like any other object, a chair may function as a production prototype or as a design pattern, which will justify the use of the term model to designate it. Examples do also show that no object is of necessity a model, such as for instance a wooden architectural model dating from the Renaissance which may be regarded, however difficult this may seem, simply as a wooden frame or as an item to be transported. There is no denying that the fact that on the one hand, basically any object may be a model, and that on the other hand, no object is of necessity a model, ruins the hope of reaching an analytic definition of the concept of model; yet at the same time it indicates a method with 12 Many such paraphrases can be found in dictionaries and textbooks. 13 In his Philosophical Investigations, Wittgenstein discusses the issue of concepts for which no analytical definition exists, one of them being the concept of ,game‘. The existence of this type of concepts prompts him to develop his theory of family resemblance. See Wittgenstein (1967), 47 ff. The way taken here follows a different path.

On the Epistemology of Models

305

which one may attain an explanation. It turns out that the phenomenon of model-being can be understood if one stops looking for an answer to the question of the nature of a model and starts asking instead what justifies conceiving of something as a model. However slight this modification of the question may seem, its ramifications are substantial. The reason is that it transfers the question of the nature of models into the realm of situations, propositions, deductions, and arguments, a realm in which the model-being of an object will become the result of a judgment which is situated in contexts of invention and justification, and whose acceptance and reasoning may thus be questioned. In short, it appears that the nature of models may be explained if only we free ourselves from the idea of asking an ostensibly ontological question, looking for ontological predicates of model-being, and start searching instead for an explanation using the means of epistemology and logic. In other words we must understand the question of model-being as a question about reasoning of a judgment on model-being. In that way a lot of problems are solved. The study of the nature of models presented here, borrows many of the principles of modeling employed in the foundations of mathematics and in the daily practice of modeling in information science. This is not to say that the resulting explanation of model-being primarily applies to mathematics and to the information sciences, it just means to say that use is made of the style and experience of modeling in these disciplines. The claim still stands that the model of model-being explains the nature of models in full generality.

Instructive Examples of Model-being The Epistemic Pattern of Model-being We have to ask, what may be said in general about the justification of a judgment on the model-being of an object. Obviously a judgment declaring that an object is a model cannot be justifiable or acceptable unless it accords with the intuitive notion of model and with the requirements of its situational and disciplinary context. Consequently, that, what can generally be said about the justification of a judgment on modelbeing must itself be a model reflecting the intuitive notion as well as the possible contexts of its declaration.14 It therefore seems sensible to 14 The present concept of model has been developed by the author since 2000 and

306

Bernd Mahr

analyze this intuitive notion of model via an investigation of the use of models in practice and in theory, and then to describe the results of the investigation conceptually, in the form of a model of model-being. And it also seems sensible to begin by investigating the use of models and its conception in areas in which the notion of model is not only relevant and authoritative, but has also been reflected on and studied extensively. The study of examples does show that in all cases of model use there is one and the same characteristic structure of contextual relationships on which the judgments on model-being rest and which explicitly or implicitly serves as a reference in model use and model related activities. This structure will in the following be called the epistemic pattern of modelbeing, and will be further investigated in subsequent sections. It comprises four interdependent relationships: (1) the relationship between an object which is conceived of as a model, the model-object, and that as which the object is being conceived of; (2) the relationship between the model-object and that of which the model-object is conceived of as a model; (3) the relationship between the model-object and that for which the model-object is conceived of as a model; and finally (4) the relationship between the model-object and the cargo carried by the model-object as a model. The situation depicted in figure 1 is that of the conceiving of the object M as a model l of A and for B, thereby carrying the cargo w from A to B. The examples below will make this reading of the diagram evident. The variety of evidence provided in proof of the observation15 that the epistemic pattern of model-being is characteristic for models in general, makes up the rationale of the thesis that this pattern may serve as the core of a model of the model-being of objects in general. This thesis is also corroborated by the fact that the known model concepts and thehas been thoroughly investigated by the author, together with Jens Gulden, a business information scientist and Reinhard Wendler, an art historian, in the project “Clarification of the Concept of Model”, which was supported by the German Research Council from 2004 to 2007. The project was carried out within the framework of the research group “Bild, Schrift, Zahl”, at the Hermann von Helmholtz-Zentrum für Kulturtechnik at the Humboldt Universität zu Berlin. Thus the articles originating from the author which are cited in the present article do overlap to some extent; however, each individual article has a different purpose, and not least, together the articles reflect the progress made in respect of conceptual clarification. Particular attention is drawn to Bernd Mahr (2008a), 187 – 218. 15 For a selection of such evidence, see Wendler (2008).

On the Epistemology of Models

307

Figure 1: The structure of contextual relationships through which an object M, in its capacity of model l, becomes the carrier of cargo w

ories referred to in the literature are, in their epistemic structure, being covered by this epistemic pattern16, and that the modeling and the use of models, the analysis of which is made possible by this pattern, will reveal appropriate techniques and strategies of scientific and artistic working processes17. In this way the claim that a model of model-being may provide an answer to the general question of the nature of models finds its justification. The following examples of model-being shall shed some light on the use of this pattern. The Model of the Modulus According to the Grimm brothers, our present-day word model originates from the Latin word modulus, which first appeared in the Decem Libri De Architectura of the Roman architect Vitruvius18, completed in the year 14 B.C. From there it was carried on to the Age of the Renaissance via craftsmen moving to France as well as via the Encyclopedia of the Middle Ages. In this way, and via the use of the word modul of Old High German as well as of Italian modello, the word model developed into a word which has now become a general term19. Trying to exploit the etymology of the word in order to find a paradigm for the intuitive notion of model and the situations of its use, one soon encounters the 16 17 18 19

See Mahr (2009). See Wendler (2008a). Vitruv (1981). See Mahr (2004b), 59 – 86.

308

Bernd Mahr

problem that the etymological roots of the word model and its various senses do not directly cast any light on the general conception of model as it is commonly understood today. This is particularly true of the word modulus: modulus is used by Vitruvius to designate a unit of measurement calculated on the basis of the dimensions of a temple as a whole, a unit of measurement used to gauge the temple and its parts in such a way that all the relevant dimensions constitute rational multiples of the modulus (see figure 2). The emergence of this measuring technique, which has been known in architecture at least since the Middle Empire of Egypt around 2000 B.C., but is also known in shipbuilding and military construction, can probably be attributed to construction practice, and later probably to the ‘Pythagorean’ conviction that a building had to be designed using commensurable dimensions only. No obvious link, however, can be detected between the word modulus and the concept referred to by our modern term model. Only a less selective reading of Vitruvius’ text shows that the measuring technique connected with the word modulus enters into a context of use from which the above mentioned structure of contextual relationships can be recognized: the measuring of a temple based on the modulus, which mostly corresponded to half the diameter of the lower column, proceeds according to a fixed system of symmetries which, according to Vitruvius, had been derived from the Greek canon of ideal proportions of the human body. The use of this system of symmetries was justified by the eurhythmy of the buildings which had been constructed according to this system. In short, not the modulus itself, but the system of symmetries expressed through the measuring unit of a modulus and presented as a text, a list or a diagram, is a model in the modern sense of the word.20 This model was a model of the proportions of the human body, and at the same time a model for the construction of temples, through the use of which static properties and the beauty of the proportions of the human body as its cargo are transferred to the temples constructed according to the modulus (see figure 3). Vitruvius writes extensively on these relationships in his treatise. Following the rediscovery of ‘de archi20 The system of symmetries, expressed through the measuring unit of the modulus, is independent from measuring standards and can be applied at any scale. Its representation as a list of measures could be easily memorized and communicated. Also Leonardo’s drawing lists in its surrounding text proportions, but this time of the human body.

On the Epistemology of Models

309

310

Bernd Mahr

On the Epistemology of Models

311

tectura’ in 1416 and 1417 in the monastic library of St. Gallen21, the model of the modulus has been discussed and modified again and again, from Alberti’s On the Art of Building, dating around 1450, until to Le Corbusiers Modulor, developed in the time between 1948 and 1955.22 The Model of Turing Machines The epistemic pattern of model-being, underlying the anthropomorphic model of the modulus, may be seen with equal clarity also in other models, like for example the model of Turing machines. Also Turing machines have an anthropomorphic background. In his famous article “On Computable Numbers, with an Application to the Entscheidungsproblem” 23 Alan Turing proves that there is no Turing machine which can effectively answer the question if a given Turing machine in the course of its computation will write 0 on its tape. A consequence of this at first glance obscure result is a proof that David Hilberts so called ‘Entscheidungsproblem’ is to be answered to the negative. In 1900 Hilbert had proposed a program to axiomatize mathematics through proof-systems which are correct, complete and effectively decidable. In 1931 Kurt Gödel showed that Hilberts program must fail. He proved that even for the familiar Peano-Arithmetic no formal system can be found, which is both correct and complete; in other words that for any correct formal system there are true sentences about the natural numbers, which can not be proven by that system.24 On the other hand, first-order predicate logic has a correct and complete formal system, a result which Gödel had already proven in his thesis in 1929.25 Hilberts Entscheidungsproblem, formulated in 1928, then was the question if at least first-order predicate logic is effectively decidable. Half a year before Turing, in 1936, also Alonzo Church proved that the Entscheidungsproblem has to be answered to the negative.26 The essential difference between Church’s and Turing’s solution of the problem is their conception of effectiveness.

21 22 23 24 25 26

Cf. Schuler (1999). Cf. Kruft (1995). Turing (1937). Gödel (1965), 5 – 37. Gödel (1986). Church (1965), 110 – 114.

312

Bernd Mahr

Hilberts program was based on an intuitive notion of effectiveness. His conception of computing was strongly influenced by the work, men do if they execute computations with the help of paper and pencil, and supported by mathematical tables and mechanical calculating machines, but without understanding the mathematical reasoning behind the computation they perform. Church used the concept of Lambda-definability for his definition of computability and proved that Lambda-definability is equivalent to Herbrand’s Recursiveness. Both conceptions of computability describe forms by which mathematical functions can be specified and express computability by certain rules of function-application. Turing’s conception of computability, in contrast, is different. He thinks of computing as a process executed by a machine which only reads and writes symbols on a tape, controlled by a finite program and a finite memory of states. But Turing’s conception is not different in its expressive power, as he can show that a function is Lambda-definable if and only if it can be computed by a Turing machine. It only differs in its idea of computing.27 The following citation from Turing’s article shows well, what it is that Turing machines are a model of. Turing simulated with his machine the actions a human performs on a typewriter28 which allows typing symbols, erasing symbols, moving its head one position to the left or one position to the right, and having an unlimited tape. Actions are determined by a finite number of elementary instructions and by finitely many states that can be memorized. Turing writes: We have said, that the computable numbers are those whose decimals are calculable by finite means. This requires rather more explicit definition. No real attempt will be made to justify the definitions given until we reach §9. For the present I shall only say that the justification lies in the fact that the human memory is necessarily limited. We may compare a man in the process of computing a real number to a machine which is only capable of a finite number of conditions q1, q2 …, qR which will be called m-configurations. The machine is supplied with a tape (the analogue of paper) running through it, and divided into sections (called squares) each capable of bearing a symbol. At any moment there is just one square, say the rth, bearing the symbol S(r) which is in the machine. We may call this square the scanned square. The symbol on the scanned square may be called the scanned symbol. The scanned symbol is the only one of which the machine is, so to speak, directly aware. However, by altering its m-configuration the machine can effectively remember some of the symbols 27 See for example Hodges (1985), 46 – 110; Davis (1965). 28 Hodges (1985), 96 – 102.

On the Epistemology of Models

313

which it has seen (scanned) previously. This possible behaviour of the machine at any moment is determined by the m-configuration qn and the scanned symbol S(r). This pair qn , S(r) will be called the configuration: thus the configuration determines the possible behaviour of the machine. In some of the configurations in which the scanned square is blank (i. e. bears no symbol) the machine writes down a new symbol on the scanned square: in other configurations it erases the scanned symbol. The machine may also change the square which is being scanned, but only by shifting it one place to the right or the left. In addition to any of these operations the m-configuration may be changed. Some of the symbols written down will form the sequence of figures which is the decimal of the real number which is being computed. The others are just rough notes to assist the memory. It will only be those rough notes which will be liable to erasure. It is my contention that these operations include all those which are those in the computation of a number. The defence of this contention will be easier when the theory of the machines is familiar to the reader.29

Turing’s conception of computing was well received and stimulated the statement of the prominent Church/Turing thesis saying that computability in the intuitive sense is equivalent to computability in the sense of a Turing machine, a thesis which is still considered valid even after more than 70 years of advances in information science and technology. In 1964, in a postcriptum to his lecture On Undecidable Propositions of Formal Mathematical Systems, which Gödel had given 30 years before, he confirmed Turing’s analysis of computing an welcomed the win of a “precise and unquestionably adequate definition of the general concept of a formal system” as a “Turing machine”, i. e. a “mechanical procedure for producing provable formulas”.30 The expressive power of Turing machines is rich enough to allow for universal Turing machines which receive as input the coding of a Turing machine and data, and in their execution on this input run the inputted Turing machine on the inputted data. Universal Turing machines are the mathematical equivalent to computers that execute programs on input. In which sense, now, can Turing machines be conceived of as a model? The answer to this question is derived from a reference to the epistemic pattern of model-being: First of all, the model object is a piece of text which in mathematical terms generally defines what sets are to be defined in order to specify a Turing machine, and how a given Turing machine executes a computation by symbolically transforming configurations, from an initial configuration to an end config29 Turing (1937), 230 – 233. 30 Gödel (1965), 71 f.

314

Bernd Mahr

uration with a halting state. In order to identify what the model is as which the definition of Turing machine specifications is conceived of, it has to be defined what kind of object an individual Turing machine is, and when two specifications determine the same machine. Such definitions have mainly technical purposes and may be given by restricting the use of symbols from standardized alphabets and by using a coding scheme, with the effect that the collection of specifiable machines forms a set of words, the language of Turing machines. Following Turing’s analysis of computing, then the model of Turing machines is a model of the human activities when using a functionally extended typewriter. To determine what Turing machines are a model for, there are several choices. Most naturally Turing machines are a model for the effective specification of functions by using the general definition of Turing machine specifications. The relationship between the general definition, which is the model object, and a particular machine specification, which is what the model object is applied to, is the relationship of an instantiation. Then the cargo is simply the guaranty of computability and the knowledge, how computations of the particularly specified Turing machine can de executed. Also the proposition made by the Church/Turing thesis is part of the cargo. It provides the user of the model with the knowledge that he is using a most intuitive and universal model of computing. Models as Realizations One of the first references to the concept of models in logic was made by Alfred Tarski in his article “Über den Begriff der logischen Folgerung”31, which appeared in 1935. In this article, the word ,model’ is synonymous with ,realization of a class of propositions’ and refers to a non-linguistic object, i. e. for instance to a mathematical structure, in which the propositions in question are valid.32 In 1929, Kurt Gödel had used the phrase ,satisfiability in domains of thought‘, by which he referred to the phenomenon which Tarski called a realization. From the first concepts of satisfiability in structures and of truth in a model33, the conceptual framework of model theory emerged which turned into an important branch of mathematical logic. In 1966, Andrzej Mostowski elaborated on this in his third lecture on the theme Development of mathematical 31 Tarski (1935). 32 See also Bays (2001), 1701 – 1726. 33 See Feferman (2004), 79 – 108.

On the Epistemology of Models

315

Figure 5: The Having of a Property and the Satisfaction of Sentences.

logic and the study of the foundation of mathematics from 1930 to 1964: “The modern form of semantics is the theory of models”. In computer science, in the 1970th and 1980th, model theoretic semantics and the construction of semantical models as the meaning of formal specifications have extensively been studied, either in the form of so called ‘loose semantics’ where the meaning of a specification is taken to be the class of models which satisfy the constraints specified, or in the form of so called ‘denotational semantics’ where the meaning of a specification is given by a particular construction of a model, examples being Scott’s reflexive domain model of the Lambda calculus34 or initial algebras of equational specifications35. The concept of models as realizations is based on a relationship between linguistic expressions, like axioms, specifications, or denotations, on the one side, and non-linguistic entities, like set-theoretic domains, objects in a category, or algebraic or relational structures, on the other. This relationship expresses what is called satisfaction. It mediates between the syntax of the linguistic expressions and the structure of the entity, conventionally thereby following the principle of compositionality of interpretation, which is ascribed to Gottlob Frege36. If the linguistic expressions are sentences denoting propositions, the satisfaction relationship can be read as the having of the property expressed by these sentences. The key idea of the satisfaction relationship is its mathematical nature. This, however, requires that also that what corresponds to the entity is a mathematical object. This correspondence then amounts to say 34 Cardone/Hindley (2009). 35 See for example Ehrig/Mahr (1985). 36 See for example Frege (1966), 72 – 91.

316

Bernd Mahr

that an entity has a certain property if the corresponding sentences are satisfied by the corresponding object. Then, if the sentences are satisfied, the object is called a model of the sentences. In logics and formal semantics the conditions of satisfaction therefore form an essential part of their definitions, besides, of course, the determination of what counts as a wellformed sentence and what is assumed to be an object. The epistemology of satisfaction can be further detailed if the satisfaction relation is assumed to be defined in a logic or formal semantics framework, like for example in First Order Predicate Logic, where sentences are closed formulas, i. e. formulas with no free variables, and objects are relational structures, i. e. sets together with operations and relations.37 Then the set of all sentences forms a language, say L, and the theory of an object forms a set of sentences consisting of those sentences which are satisfied by the object. Then the theory of an object can, in a way, replace the object, in the sense that it contains everything that can truly be expressed in the language about the object. theory (L, object) := {v u L j object satisfies v} In 1961 Patrick Suppes argued for the exemplary importance of the concept of model in mathematical logic in his article A comparison of the meaning and uses of models in mathematics and the empirical sciences 38 : “I claim”, he says, “that the concept of model in the sense of Tarski may be used without distortion and as a fundamental concept in all of the disciplines (mathematical logic, physics, social sciences, engineering)”. Suppes distinguishes between the notion of the model concept on the one hand and its application in various sciences on the other, saying: “Given [the] technical meaning of the concept of model, mathematicians ask a certain kind of question about models and empirical sciences tend to ask another kind of question”. Nevertheless, in today’s general debate on models, the concept of model in logic is often viewed as a special case, not to be subjected to closer scrutiny. With the discussion below and with further insights into the logic of models, this view can not be supported. Using the epistemic pattern of model-being one encounters that models as realizations include actually two judgements of model-being: first, the object of concern as a model of the sentences under consideration, and second, the object of concern as a model of the world which is 37 See for example Ehrig/Mahr (2001), 333 – 352. 38 Suppes (1961), 163 – 177.

On the Epistemology of Models

317

reflected in the mathematics of the object. For example, the algebra of natural numbers, with zero as a constant and the successor function as its only operation, is a model of the Peano axioms and at the same time a model of counting with numbers. Obviously, the goal of axiomatizing arithmetic has been to find axioms which have a model and can practically be used in proofs, and which, on the other hand, have a natural reading as assertions about counting and computing as real world activities. These two views of the natural numbers as a model are not in competition, but rather complement each other. They also have natural explanations of what they are as models for and what is the cargo they transport: as a model of the Peano axioms the natural numbers prove consistency of these axioms, i. e. they guarantee through their modelbeing that the Peano axioms are not contradictory and can therefore safely be used in formal systems proofs; the natural numbers as a model of counting and computing are a model for an axiomatization and the founding of formal constructions, like the definition of arithmetic operations and relations, and of formal proofs; as a model of the Peano axioms, the natural numbers are a model for semantical constructions in arithmetic founded on a consistent axiomatic basis. In the first case, the cargo transported is that what is observed and experienced in activities of counting and computing, and in the second case, the cargo is consistent knowledge about counting and computing as it is expressed by the axioms. Each of the two views corresponds to a particular style of practicing mathematics, namely the working in formal theories and the working in a model. Models as realizations are not restricted to be used in mathematics only. For example, Vitruvius’ figure (see figure 6) shows the same intention behind its model-being39. Vitruvius in his ‘decem’ libri discusses proportions of the human body and states that the human figure can be inscribed into a circle and into a square, depending on the positioning of its arms and legs: Namque si homo conlocatus fuerit supinus manibus et predibus pansis circinique conlocatum centrum in umbilico eius, circumagendo rotundationem utrarumque manum et pendum digiti linea tangentur. Non minus quemadmodum schema rotundationis in corpore efficitur, item quadrata designatio in eo invenietiur. Nam si a pedibus imis ad summum caput men39 On Leonardo’s drawing, see e. g. Zöllner (1987), 77. Leonardo’s drawing lists in its surrounding text propositions of the human body, of which the figure can be seen as a model (see Mahr (2008b), 17 – 40).

318

Bernd Mahr

On the Epistemology of Models

319

sum erit eaque mensura relata fuerit ad manus pansas, invenietur eadem latitudo uti altitudo, quemadmodum areae, quae ad normam sunt quadratae.40

A graphic realization of the inscription into a circle, among many others, is given by Cesare Cesariano in 1521.41 Better known, however, is Lenardo da Vinci’s drawing of Vitruvius’ figure from around 1492, which in contrast to Cesariano does not follow closely Vitruvius’ prescriptions of the human proportions. He brings in his own experience and research on human proportions.42 Today Leonardo’s drawing is heavily overloaded with esoteric and metaphoric meanings and has become a popular symbol for the ‘human in the centre’. It has, on the other hand, two most natural interpretations as a model: one is to see it as a study of the human body in motion. Kenneth D. Keele argues convincingly that with his drawing Leonardo “provides one of the simplest illustrations of a shifting ‘centre of magnitude’ without a crresponding change of ‘centre of natural gravity’.”43, and the other is to see the drawing as a realization of the listed proportions of the text which surrounds the drawing,44 as a proof of ais40 Vitruv (1981), 138; in German translation: “Liegt nämlich ein Mensch mit gespreizten Armen und Beinen auf dem Rücken, und setzt man die Zirkelspitze an der Stelle des Nabels ein und schlägt einen Kreis, dann werden von dem Kreis die Fingerspitzen beider Hände und die Zehenspitzen berührt. Ebenso wie sich am Körper ein Kreis ergibt, wird sich auch die Figur des Quadrats an ihm finden. Wenn man nämlich von den Fußsohlen bis zum Scheitel Maß nimmt und wendet dieses Maß auf die ausgestreckten Hände an, so wird sich die gleiche Breite und Höhe ergeben, wie bei den Flächen, die nach dem Winkelmaß quadratisch angelegt sind”. 41 Zöllner (1987), 127 – 143. 42 Zöllner (1987), 77 – 87. 43 Kneele (1983), 252. 44 The Text surrounding Leonardo’s drawing is written in Italian in reflected face. The German translation is as follows: “Vitruvius der Architekt sagt in seinem Werk über die Architektur, dass die Maße des Menschen in der folgenden Weise ausgelegt seien: es bilden nämlich 4 Finger eine Handbreite, 4 Handbreiten einen Fuß und 6 Handbreiten eine Elle (cubito). Vier Ellen (cubiti) ergeben einen Klafter (passo) und 24 Handbreiten die Länge eines Mannes; und diese Maßverhältnisse finden sich in seinen Gebäuden. Wenn du die Beine soweit spreizest, dass sich deine Größe, gemessen vom Kopf, um 1/14 vermindert, und wenn du deine Arme soweit öffnest und erhebst, dass du mit den Mittelfingern die Linie auf der Höhe des Scheitels erhebst, dann weißt du, dass das Zentrum der äußersten Punkte der ausgestreckten Gliedmaßen der Nabel und dass der Raum, der sich zwischen den Beinen befindet, ein gleichseitiges Dreieck sei. Die ausgestreckte Armspannweite ist soviel wie seine Höhe.

320

Bernd Mahr

tetic consistency, so to say45. The second reading reads Leonardo’s drawing as a model of the numeric proportions listed, and a model for the visual recognition of its beauty. The cargo carried by the drawing as a realization then is the system of proprtions linguistically expressed in the surrounding text. Conceptual Models and Their Application The type of models dominating in information sciences is what one calls a conceptual model. Conceptual models are used as prescriptions of structure, architecture, processing, terminology or conceptualization. In most cases they are presented as frameworks, reference models or meta models, and serve the purpose of instruments applicable in tasks of modeling and specification. Their application helps to systematically generating descriptions of objects of thought or in reality, and allows for comparison and classification. Conceptual models contribute much to the methodologies employed in information science. A prototypical example of a conceptual model is the TINA-C Business Model. 46 It applies to describe the combination and composition of communication based services offered to a consumer by a retailer. The diagram in figure 8 represents the structure of the major relationships of a service-architecture from the enterprise viewpoint, the intended reading of which is specified by the textual part of the TINA-C Document. The model comprises the meta model of a (horizontal) three-tier-archiVon den Haarwurzeln bis unter das Kinn ist der zehnte Teil der Höhe des Menschen; von unterhalb des Kinns bis zum Scheitel ist es der achte Teil der Höhe des Menschen; von der Höhe der Brust bis zu Scheitel sei der sechste Teil des Menschen, von der Höhe der Brust bis zu den Haarwurzeln der siebte Teil des ganzen Menschen. Von den Brustwarzen bis zum Scheitel sei der vierte Teil des Menschen; die größte Breite der Schultern enthält in sich den vierten Teil des Menschen; vom Ellenbogen bis zu den Fingerspitzen sei der vierte Teil des Menschen; von demselben Ellenbogen bis zu Ende der Schultern sei der achte Teil dieses Menschen; die ganze Hand sei der zehnte Teil des Menschen: das männliche Glied beginnt in der Mitte des Menschen; der Fuß sei der siebte Teil des Menschen; von unterhalb des Fußes bis unter das Knie sei der vierte Teil des Menschen; von unterhalb des Knies bis zum Ursprung des Gliedes sei der vierte Teil des Menschen. Die Teile, die sich zwischen dem Kinn, der Nase, den Augenbrauen und den Haarwurzeln befinden – ein jedes dieser Teile ist für sich ähnlich der [Länge] des Ohres und ein Drittel des Gesichts” (Zöllner (1987), 78 f.). 45 See Mahr (2008b), 33 – 40. 46 TINA Consortium (1997), 2.

321

On the Epistemology of Models

Broker

Bkr

Consumer

Bkr

Ret

Bkr

Bkr

3Pty

Retailer

Bkr

3pty Service Provider

RtR TCon

TCon

ConS

ConS

TCon

Connectivity Provider CSLN

LNFed

Figure 8: “The TINA Initial bBusiness Roles and Business Relationships”, in: TINA Consortium (1997), 2.

tecture and is generic in the application of its roles and relationships. It can be applied recursively since it permits a consumer to act in turn as a retailer or service provider, thereby supporting the modelling of service chains. Its application is not restricted to a particular scope or size of a system, as it may be used for the design of global telecommunication infrastructures, the purpose for which it has been designed, as well as for the modelling of the composition of fine grained Web services, a technology that emerged only later.47 Generally48, conceptual models are expressed through specifications of concepts and relationships, which constitute a complex of terms and assertions that allow to identifying or prescribing structures, characteristics and facts of the object to be described, the intended object of concern. A conceptual model can therefore also be seen as a logic explaining the conception of an intended object of concern. Since the description of such an object is never the object itself, but only a model, a conceptual model is therefore a meta-model representing features of possible con47 See Flügge (2008). 48 See also Mahr (2006), 376 – 390.

322

Bernd Mahr

formant realizations of it. Conceptual models are models of observations, prescriptions or visions on the matter of concern. The cargo they transport is best practice experience or insights into states of affairs, or, in the case of a vision, content of some imaginated entity. Conceptual models may have different forms of presentation. They may be formal logic denotations, as for example used in the theory of algebraic specifications, or may be clauses phrased in some kind of ‘controlled’ natural language, like in documents of standardization, as for example the ISO Reference Model for Open Distributed Processing (RM-ODP)49, or in diagrammatic form, like in the TINA-C Business Model above. The application of a conceptual model provides an instantiation of the model through which its constituents are specialised. Accordingly, the model obtained by an application of a conceptual model is supposed to inherit the cargo transported, i. e. elements of observation, best practice, insights or imagination, of which the conceptual model is a model. If the cargo is indeed transferred to the model resulting from an application, is not only a matter of the conceptual models design and form of presentation, and of the precision achieved at the abstract level of its concepts, but to a large degree also a matter of its application. Issues of quality therefore do not only concern the modelling of the conceptual model alone but equally also its application. The application of a conceptual model may best be compared to a game of consistently providing answers to questions: answers assign observed or intended phenomena of the intended object to the concepts and relationships of the model in a way that the specifications associated with these are respected. Usually conceptual models are open in their application and do only prescribe certain characteristics of intended objects to which they apply and only certain properties of coherence of their modelling. Conceptual models often determine as what the intended object is to be conceived of. For example the TINA-C model takes an architectural view of the system to which it is applied, and not a purely economic perspective. But even in this respect there may be much freedom of choice. Of particular importance in the application of a conceptual model is the granularity of the resulting model, i. e. the degree of detail in the description of the intended object of concern. Conceptual models are often criticised for their impracticability resulting from the generality of their concepts. Detailed prescriptions in conceptual models, on the other hand, restrict the scope of their application and may 49 ISO Standard RM-ODP (ISO/IEC 10746) parts 1 to 4 from 1996 – 1998.

On the Epistemology of Models

323

turn them into descriptions which they are not meant to be but for the production of which they are meant to be instruments. Whatever the distance is between the detail of concepts of a conceptual model and the granularity of the description resulting from it, the application of a conceptual model is still an act of modelling and, as such, a creative endeavour. What has been said here about conceptual models in general, also applies to the models discussed below, the model of conception and the model of model-being, since both are conceptual models too.

A Model of Conception Uses of the Term Conception In the beginning of this article it was said that being a model is the content of a judgment in which something is being conceived of as a model. To better understand the nature of model-being, it is necessary to better understand what is conceived of as conception, an understanding which will result in a conceptual model. The following discussion of intentionality, presentations and mental states is therefore meant to elaborate on the matters of which this model of conception is a model and to specify the intended choice of its reading. Conception (in German Auffassung) is a common term in daily language. It is either explicitly used in phrases like to conceive of something or to conceive of something as something 50, or it is implicitly addressed in situations where somebody perceives something with his senses in a certain way, where somebody thinks of something somehow, where somebody wishes something to be, where somebody intends to express with his words a certain meaning of something, where somebody has something as a certain thing before his mind, where in somebody’s understanding certain properties are attributed to something, or where somebody actually or in principle sees something in a certain nexus with other things or matters of affairs. Common to these implicit uses of conception is the fact that all their situations include a direction towards something, and that in each case there is an indication of some form of subject and context dependency of that which is conceived of, in contrast to its objectivity or absoluteness. Different in these uses are 50 For a semantic study of the word ‘as’ see Umbach (1996).

324

Bernd Mahr

the types of acts or states, of which conception is, so to say, a part of, namely the acts of perceiving (in German Wahrnehmung), of judging (in German Urteil), of volition (in German Wunsch), of naming (in German Benennung), of presentation (in German Vorstellung), and the states of knowing (in German Wissen) and of taking a particular perspective (in German Perspektive). The different uses of the term conception point at a concept of fundamental nature and at an issue with important historical roots, the concept of conception. It turns out that the notion of conception is best studied in the framework of intentionality. Intentional In-existence and Presentation In his Psychology from an Empirical Standpoint 51 from 1874 Franz Brentano separates physical from psychological phenomena (or mental acts) by a number of characteristics; and he distinguishes three types of psychological phenomena: the phenomena of being presented with something, the phenomena of judging, and the phenomena of love, in the sense of desiring, and of hate, in the sense of detesting. One of the characteristics of mental acts is intentional in-existence. He claims that: Every mental phenomenon is characterized by what the Scholastics of the Middle Ages called the intentional (or mental) inexistence of an object, and what we might call, though not wholly unambiguously, reference to a content, direction towards an object (which might not to be understood here as meaning a thing), or immanent objectivity. Every mental phenomenon includes something as object within itself, although they do not all do so in the same way. In presentation something is presented, in judgement something is affirmed or denied, in love loved, in hate hated, in desire desired and so on. This intentional in-existence is characteristic exclusively of mental phenomena. No physical phenomenon exhibits anything like it. We would, therefore, define mental phenomena by saying that they are phenomena which contain an object intentionally within themselves.52

While physical phenomena, like colours, figures, tones, heat, coldness, smells and other things senses can perceive, or similar things which appear as phantasm (in German Trugbild), have a being outside of the mind and require a form of “outer consciousness”, mental phenomena, in contrast, have a being inside of the mind and require a form of “inner consciousness” directed to them. 51 Brentano (1995); see also the German original: Brentano (2008). 52 Brentano (1995), 88 f.

On the Epistemology of Models

325

Besides intentional in-existence and others, an important characteristic of psychological phenomena is their relation to the mental acts of presentation, presentation taken as presentation to the mind53. Something which is being presented appears to the mind54, or, so to speak, is in front of it, and the fact that the mind is being presented with something is, according to Brentano, the result of inner consciousness and may be caused by sensation or imagination. Characteristic of psychological phenomena, Brentano argues, is the fact that a mental phenomenon either is a presentation or rests on a presentation.55 This characteristic assigns the concept of presentations a most distinguished role and treats it as a matter of fundamental importance. The notion of presentation has been in the focus of philosophy long before Brentano. In the foundations of logic and science it had already reached an elaborated ontological and scientific status. Bernhard Bolzano, for example, devoted a considerable part of his Wissenschaftslehre from 1837 to the concept and forms of presentations56. The importance of this concept may also be seen from the fact that Georg Cantor’s conception of a set, as some kind of ‘Inbegriff’57, is founded on the notion of presentation: 53 54 55 56 57

See for example Grossmann (1992). Brentano (2008), 99. Brentano (2008), 97 – 103. Bolzano (1981), §47 to §114, 215 – 536. Bolzano (1981) writes: “Eine sehr wichtige Gattung zusammengesetzter Vorstellungen, welche uns allenthalben begegnen, sind diejenigen, in welchen die eines Inbegriffes vorkommt. Es gibt aber mehrere Arten derselben, und nur die merkwürdigsten sollen hier angeführt werden. Erst muß ich jedoch den Begriff, den ich mit dem Wort /Inbegriff /überhaupt verbinde, genauer bestimmen. Auch dieß Wort nehme ich in der Bedeutung, in der es schon der allgemeine Sprachgebrauch nimmt; und verstehe also unter einem Inbegriffe gewisser Dinge völlig dasselbe, was man auch durch die Worte: eine Verbindung oder Vereinigung dieser Dinge, ein Zusammensein derselben, ein Ganzes, in welchem sie als Theile vorkommen u. dgl., ausdrücken könnte; so zwar, daß in der bloßen Vorstellung von einem Inbegriffe noch gar nicht festgesetzt seyn soll, in welcher Ordnung und Aufeinanderfolge die hier zusammengenommenen Dinge erscheinen, ja ob es überhaupt eine solche Ordnung unter denselben gebe und geben könne. Ob dieser Begriff einfach oder aus welchen Theilen er etwa zusammengesetzt sey, getraue ich mir abermals nicht mit Gewißheit zu entscheiden. Ich halte aber dafür, daß es ein wenn nicht durchaus einfacher, doch aus ehr wenigen Theilen zusammengesetzter Begriff sei. Ein Inbegriff nämlich scheint mir nichts Anderes zu seyn, als ein Etwas, das Zusammengesetztheit hat. Dieses Abstractum aber, oder der Begriff, den ich hier

326

Bernd Mahr

“A set is a collection into a whole of definite, well distinguished objects of our intuition or our thought”58, and “I call an aggregate (a collection, a set) of elements which belong to any domain of concepts well-defined, if it must be regarded as internally determined on the basis of its definition and in consequence of the logical principle of the excluded middle. It must also be internally determined whether any object belonging to the same domain of concepts belongs to the aggregate in question as an element or not, and whether two objects belonging to the set, despite formal differences, are equal to one another or not.”59

The term presentation can mean two things, the act of being presented with something, on the one hand, and that which is being presented, or short the presented, on the other. Confusion of these meanings creates conflicts, and already Brentano insists in their careful distinction. Since he finds similar dangers with other mental phenomena, the distinction durch das Wort /Zusammengesetztheit/ ausdrücke, däucht mir kaum einer weiteren Zerlegung fähig.” (§82), and “In den bisher betrachteten Inbegriffsvorstellungen blieb es dahingestellt, in welcher Verbindung die einzelnen Theile, aus denen der gedachte Inbegriff bestehet, darin erscheinen. Soll nun auch hierüber etwas entschieden werden, so ist leicht zu erachten, daß noch zusammengesetztere Vorstellungen zum Vorschein kommen müssen. Der einfachste Fall tritt ein, wenn nichts anderes festgesetzt wird, als “daß die Art der Verbindung zwischen den Theilen als etwas Gleichgültiges angesehen werden solle.” Daß wir gar oft uns veranlaßt fühlen, Inbegriffe mit dieser ausdrücklichen Bestimmung zu denken, wird Niemand in Abrede stellen. So sagen wir z. B. bei einem Geldhaufen, den wir annehmen oder wegschenken sollen, es sey uns gleichgültig, in welcher Ordnung die einzelnen Geldstücke, aus denen er bestehet, beisammen liegen. Eben so finden wir uns bei Aufzählung der Mitglieder mancher Gesellschaftenn veranlaßt, ausdrücklich zu bemerken, daß die Ordnung, in welcher wir diese Personen hier erwähnen, als durchaus leichgültig anzusehen sey, und daß wir überhaupt gar keine Rangordnungunter denselben anerkennen u. dgl. Der gemeine Sprachgebrauch nennt solche Inbegriffe /Mengen; /nur pflegt er freilich mit diesem Ausdrucke noch manche Nebenvorstellungen zu verbinden, und insbesondere dabei an eine beträchtliche Anzahl von Theilen zu denken. Allein wer sieht nicht, daß diese Beschränkung des Begriffes für die Zwecke der Wissenschaft von keinem Nutzen wäre. Ich erlaube mir also jeden beliebigen Inbegriff, bei welchem die Verbindungsart der Theile als etwas Gleichgültiges angesehen werden soll, eine /Menge/ zu nennen, auch wenn er der Teile nur sehr wenige, ja auch nur zwei enthielte. Hieraus erräth man von selbst, aus welchen Bestandteilen ich mir den Begriff einer Menge zusammengesetzt denke” (Bolzano (1981), §84). 58 In Cantors (1879), cited in English from Grossmann (1992), 58. 59 Phrased in a letter to Richard Dedekind in 1882; cited in English from Dauben (1979), 83; Dauben translates from the German original Mannigfaltigkeit as aggregation, Inbegriff as collection, and Menge as set.

On the Epistemology of Models

327

between the act of a mental phenomenon and the object of its intentional in-existence is constitutional in his investigations. But the ambiguity of the term presentation is not the only source of confusion since there is also an ambiguity in the term the presented. Brentano hints at it when he comments his wording “direction towards an object” and “reference to a content” as not being “wholly unambiguously” (see above), but he is not strict about the distinction to be made. In his Habilitation thesis On the Content and Object of Presentations – A Psychological Investigation 60 from 1894, Kasimir Twardowski, one of Brentano’s students in Vienna, studied the concept of presentation 61 and addressed the ambiguity in the term the presented. He argues that presentations imply, in what they present to the mind, two different objects rather than one: the object towards which the presentation is directed (in German Gegenstand), and the object which is its content (in German Inhalt). The latter he also calls the intentional object of the act or simply the intentional object. The intentional object of a presentation he regards as a sign of the object towards which the presentation is directed and, metaphorically, names it a mental picture. By introducing proper terminology, he resolves the ambiguity of the term the presented and proposes how to express the differences in the meaning when the attribute presented is applied to the content and to the object of a presentation: In order to distinguish, then, between the two meanings attached to the word ‘to present’ when applied to the content and when applied to the object, we shall use Zimmermann’s terminology. We shall say of the content that it is thought, presented in the presentation; we shall say of the object that it is presented through the content of the presentation (or through the presentation.) What is presented in a presentation is its content; what is presented through a presentation is its object.62

This terminology hints at a particular kind of relation between the object and the content of a presentation. It assumes that content takes the role of a means by the help of which the object is being presented. The distinction between object and content of a presentation, Twardowski claims, is equally valid for all psychological phenomena. In 1907, in a series of lectures called Hauptstcke aus der Phnomenologie und der Kritik der Vernunft, Edmund Husserl studied the concept of 60 Twardowski (1977); see also the German original: Twardowski (1982). 61 Rudolf Haller reports in his introduction that Twardowski was strongly influenced by Bernhard Bolzanos Wissenschaftslehre from 1837. 62 Twardowski (1977), 16.

328

Bernd Mahr

conception (using the German word Auffassung) in the context of perception.63 Conception, he argues, turns, roughly speaking, contents of senses (Husserl calls these empfundene Inhalte) into contents of presentations (Husserl calls these darstellende Inhalte), i. e. into something which is conceived of as something, as that which appears to be the perceived object. Without conception, contents of senses would be “dead material”; only through conception they receive the “spirit of meaning”.64 This is, Husserl comments, what Brentano meant by a psychological phenomenon.65 Other than presentations conceptions are not indivisible in time and are not necessarily conscious. Intentional Mental States and Conception In 1983, John R. Searle defines Intentionality in his Intentionality – An Essay in the Philosophy of Mind 66 as “that property of many mental states and events by which they are directed at or about or of objects and states 63 Husserl (1991). On pages 45 f. he writes: “Die Empfindungsinhalte für sich enthalten noch nichts von dem Charakter der Wahrnehmung, nichts von ihrer Richtung auf einen wahrgenommenen Gegenstand; sie sind noch nicht das, was es macht, dass ein Dinggegenständliches in Leibhaftigkeit dasteht. Wir nennen diesen Überschuss den Charakter der Auffassung und sagen, dass die Empfindungsinhalte Auffassung erfahren. Durch die Auffassung gewinnen sie, die an sich gleichsam ein toter Stoff wären, beseelende Bedeutung derart, dass mit ihnen ein Gegenstand zur Darstellung kommt. Mit Beziehung darauf nennen wir die Inhalte darstellende Inhalte im Gegensatz zu den dargestellten nämlich den gegenständlichen Bestimmtheiten. Nachdem sich uns vorher schon als wechselndes Moment der Wahrnehmung abgehoben hat das Moment der Stellungnahme, wie Glaube, Unglaube u. dgl., so ist es selbstverständlich, dass wir den Begriff der Auffassung so eng fassen, dass für ihn diese Unterschiede irrelevant sind und es somit auf die bloße Perzeption bezogen ist. Die Auffassung ist es, die also die selbststellende und darstellende Wahrnehmung unterscheidet. Nur in der letzteren vollzieht sich die Beziehung auf den wahrgenommenen Gegenstand dadurch, dass ein der Wahrnehmung reell immanenter Inhalt als darstellender fungiert, als ein solcher, der nicht einfach gefasst, sondern als etwas aufgefasst wird, was er selbst nicht ist, sondern was mit seiner Auffassung erscheint. Zunächst ist das eine ganz rohe Analyse. Was in dieser Auffassung liegt, werden wir noch studieren müssen. Zu beachten ist die Beschränkung auf den Typus der so genannten äußeren Wahrnehmungen, der physischen Dingwahrnehmungen”. 64 See the previous footnote; English translation by the author. 65 Husserl (1991), 47. 66 Searle (2004).

On the Epistemology of Models

329

of affairs in the world”. 67 He holds that “Intentionality is, so to speak, a ground floor property of the mind, not a logically complex feature built up by combining simpler elements”68, and by saying that “Language is derived from Intentionality and not conversely”69 he assigns Intentionality the most fundamental role one can think of. Searle deals with mental states and events and not with mental acts, as Brentano and Twardowski do, and he argues that not all mental states have intentionality: “there are forms of nervousness, elation, and undirected anxiety that are not intentional” he says.70 He also sees intentional mental states not necessarily bound to consciousness: “Intentionality”, he says, “is not the same as consciousness. Many conscious states are not intentional, e. g., a sudden sense of elation, and many intentional states are not conscious, e. g. I have many beliefs that I am not thinking about at present and I may never have thought of”.71 The example of beliefs also shows that intentional mental states do not necessarily rest on presentations, and are not constrained by capacity restrictions of the brain or of units of time. Due to the inner consciousness of mental acts, in Brentano’s and Twardowski’s view such restrictions are essential for the phenomena of presentation. What, in Searle’s mental states, one might ask, corresponds to Brentano’s mental acts of presentation? The answer will be: the mental states of conception. In his analysis of Intentionality Searle makes use of the model of speech acts. He argues that there are analogies between speech acts and intentional mental states. The most fundamental of these analogies is due to the fact that the familiar distinction between propositional content and illocutionary force of speech acts carries over to intentional mental states: both have propositional (or intentional) content, and what corresponds to the illocutionary force of a speech act is the psychological mode in which the intentional content of a mental state is being had. In the mental state of believing, for example, the intentional content is that which is believed to be true, while the psychological mode, in which one has this content, is the believing of it to be true. Or more concretely, in somebody’s belief that the earth is flat, the intentional content is ‘that the earth is flat’ while the psychological mode of this mental state 67 68 69 70 71

Ibid., Ibid., Ibid., Ibid., Ibid.,

1. 26. 5. 1. 2.

330

Bernd Mahr

is this somebody’s ‘believing it to be true’. Generally one can think of the psychological mode of a mental state as of being applied to its intentional content, just as one can say that in the belief of a flat earth the ‘believing it to be true’ is applied to the content ‘that the earth is flat’. While speech acts need not necessarily have propositional content, mental states, Searle claims, must have. Mental states, he insists, are combinations; they are composed of an intentional content and a psychological mode in which this content is being had. By taking the ‘having in a mental state as its propositional content’ as a psychological mode, one can argue that the having of a particular propositional (or intentional) content in a mental state is a mental state itself. And similar to the distinctive role of presentations one can conclude that every mental state either is the having of a particular intentional content or rests on the having of a particular intentional content.72 This and other analogies between presentations and the having of a particular intentional content makes it appropriate to answer the question above by saying that conceptions are the mental states of having in a mental state a particular intentional content. This equation seems also plausible in the light of Husserls account on conceptions, though conceptions in his sense have a more limited scope. Of perception, and in particular of visual perception, it has convincingly been argued that the state of perceiving of something is the only mental state which can directly be experienced. Its intentionality creates the awareness of consciousness.73 But not all intentional mental states are being had with awareness. In contrast to perception, the having of a conception is silent. It is something of which we are not directly aware, but only indirectly. Of conceptions we can have a proof in other mental states and acts, a proof showing us that we actually have a conception. I think that it is this silent nature of conceptions, which lets us so easily abstract from the intentional character of the things we think or perceive. Another analogy between speech acts and intentional mental states is their possible orientation, a “direction of fit”74. Orientation comes into the play if a world is assumed, which is somehow “represented” in the act or

72 See Mahr (2010), 61 – 87. 73 Wiesing (2009), 82 – 91. 74 Searle (2004), 7.

On the Epistemology of Models

331

the state.75 In a speech act this representation is realized by the use of words, and the direction of fit is therefore either of the kind word-toworld or of the kind world-to-word. In mental states such a representation of the world is not realized in words but in the mind; and a direction of fit of a mental state is therefore either of the kind mind-to-world or of the kind world-to-mind. Searle also shows that there are speech acts and also mental states which have no such orientation. I think that conceptions are not oriented; if they were, they would be a different mental state. To conclude, conceptions, it seems to me, are best characterized as silent non-oriented intentional mental states which constitute the ‘having in a mental state a particular propositional content’ and in their multitude fulfil the property that any mental state either is a conception or rests on a conception76. On the Ontology of Conceptions Since in presentations the object presented and the content of its presentation are separate things, and the content of a presentation has an instrumental role in the sense that the object of a presentation is presented through its content, it seems justified to say the same for conceptions: the object of a conception is different from the conceptions content, and an object being conceived of is conceived of through its content, i. e. through the propositional content being had in the state of its conception. As a consequence, also the content of a conception has an instrumental role. And also in analogy to Twardowski’s account on what can be presented, it seems reasonable to say that what may be conceived of may, in the true sense of the word, be anything. Concrete and abstract things may be conceived of, and even things which in reality can not be found, like unicorns, or are impossible, like round squares. Somebody’s conception may be conceived of by somebody else, as it is the case in situations of communication, and one may have the conception of one’s own conception, if, for example, one asks himself what one thinks. What is conceived of may be simple or complex and things 75 Searle is very careful with the use of the term representation here. He justifies it by saying that a mental state represents an object or state of affairs in exactly the same way in which a speech act represents it; see Searle (2004), 4 – 13. 76 The phrase of “a mental state resting on some conception” is used here deliberately with a vague meaning. It may be read as “simultaneous presence of a mental state with a conception” or as “conception being part of a mental state”.

332

Bernd Mahr

may be conceived of to be equal or different. A universe may be conceived of, and also a universe in which the conceiving individuals conceptions are part of. What is conceived of may be conceived of to exist independently from some mind to conceive of it, and it may be conceived of to not exist in reality. What is conceived of may have an extension in time or be something for a single moment only, or it may be something which was present in the past, or which may be present in the future. Therefore, that what is being conceived of needs not necessarily have existence. Twardowski points out that the content of a presentation can always also be the object of another presentation. But the converse, he maintains, is not true, since not all objects can be contents of presentations. The difference is not in their possible structure of constituents. The difference is that the content of an existing presentation is never real, and, as a consequence of its instrumental role in the act of presenting, always exists.77 “For, it is clear that the content of a presentation exists irrespective of whether one is presented with an existing or with a non-existing object”78 and “The reality of an object has nothing to do with its existence”.79 The object of a presentation, on the other hand, may or may not exist and may or may not be real. In Searle’s conception contents of directed intentional mental states form their satisfaction conditions, and in his setting “the conditions of satisfaction of speech act and expressed psychological state are identical”.80 Consequently, he views contents of directed intentional mental states as propositions (or as part of it). Their ontological status, however, is very much in debate.81 However convincing Twardowski’s argument may be, there is an ontological problem with the conception of non-existing objects, namely with the traditional strictness assumption according to which existing relationships require existence of the things they relate: If we admit that Intentionality, as a “property of many mental states”82, creates in each of its instances a relationship between a particular physical status of the brain and the object conceived of, it seems plausible that we also have to admit that conceptions have existence in the sense that also they 77 78 79 80 81 82

Twardowski (1977), 33. Twardowski (1977), 61. Twardowski (1977), 27. Searle (2004), 10 f. Runggaldier/Kanzian (1998), 215 – 218. Searle (2004), 1.

On the Epistemology of Models

333

are physically represented in the brain. How could we otherwise reason about existence in situations where we have to be prepared to also accept the possible answer that the object in question does not exist, correct or not. On the other hand, then existence of conceptions in which something is being conceived of, which does not exist, in the sense that it can not be found in reality or is impossible, is in conflict with the strictness assumption of existing relationships. Reinhardt Grossmann who discusses this problem in The Existence of the World – An Introduction to Ontology 83, circumvents this problem by admitting so called “abnormal relations”84. His solution, however, has the drawback of treating the most natural situation of conception as an “abnormal” case of existence. Searle treats non-existence of intentional objects as a phenomenon of partiality. In analogy to void references in speech acts he simply accepts cases of “unsatisfiability”.85 Searle’s conclusion is not without problems, 83 Grossmann (1992). 84 Grossmann (1992) writes on page 94: “I shall not hide or play down one of the most puzzling aspects of all ontologies: if a non-existent is an object of a mind, then it must stand in a unique relation to that mind. That is what we mean by saying that ‘it is an object of that mind’. But how can something that does not exist, that has no kind of being at all, stand in any relation to anything? […] How can the mind ‘make contact’ with what is not there at all? […] Ultimately the question is: how can there be a relation that has two terms, but only one term that exists? […] I believe that every ontology of the mind has to face this difficult question. I also believe that there is no better answer to it than to admit that there exist relations which are abnormal, that is, which connect what is there with what has no being at all. […] We are now introduced to the so-called ‘intentional relation’ between a mind (a mental act of believing, of seeing, of desiring, etc.), on the one hand, and whatever is the object of the mind, on the other. This unique relation, characteristic of minds and only of minds, is abnormal in that it on occasion connects, just like some connectives, an existent with a non-existent. Philosophers have not been happy with this admission, and I am convinced that few have been as sanguine as I am about the prospect of having to divide all relations into two groups, namely, normal and abnormal relations”. 85 Searle (2004) writes on page 17 f.: “The fact that our statements may fail to be true because of reference failure no longer inclines us to suppose that we must erect a Meinongian entity for such statements to be about. We realize that they have a propositional content which nothing satisfies, and in that sense they are not “about” anything. But in exactly the same way I am suggesting that the fact that our Intentional states may fail to be satisfied because there is no object referred to by their content should no longer puzzle us to the point where we feel inclined to erect an intermediate Meinongian entity or Intentional object for the to be about. An Intentional state has a representative content, but it is

334

Bernd Mahr

first of all since partiality of truth is a situation with which it is hard to deal, at least in formal theories. But his approach is also for other reasons not satisfying. Fiction is not at all a case of exception but widely a case of the normal. The model of conception below, in a sense, turns the abnormal into the normal, and it lets existence be a matter of conception rather than something that pre-exists. Object, Content and Context In his explanation of conception Husserl says that objects without conception are “dead material”, and Twardowski conceives of that what is presented, as something which, so to say, has no ontological status, as it may be just anything, existing or not, real or not, possible or not. In the modelling of conception it is therefore necessary to answer the questions of how objects and contents of conceptions differ, and what constitutes content. The answers to these questions form the core part of the model of conception below. It seems reasonable to think of an object as something which is ‘alive’ through the relationships in which it is involved. Everything that can be said about an object is expressed in terms of predications relating the object of concern with other objects. If, for example, a bottle is perceived to be green, there is the conception of this bottle as something which is green, a conception which includes, in a way, a relationship between the bottle and the colour green. As it is a bottle, which is perceived to be green, there are many more such relationships, concerning, for example, the bottles shape, the material it is made of, other bottles with a similar appearance, the experience of using bottles, or conceptions in which bottles have been judged to be nice or in which bottles have created emotions of wish. Without such relationships the bottle would be something of which nothing can be said, thought, judged or wished. The answer to the first question would therefore be that objects and contents can only differ, if they are of the same kind, i. e. if they are both entities, and the answer to the second question would not about or directed at its representative content […] I am not saying there are no problems about fantasy and imagination, what I am arguing is that the problems are of a piece with the problems of analyzing fictional discourse”. Note, that conceptions are not directed, i. e. not oriented mental states and do therefore have no satisfaction condition, but they may be part of directed mental states or events, like, for example, states of belief or events of perception”.

On the Epistemology of Models

335

therefore be that the content of a conception is constituted by relationships into which the object of concern is involved. Since an object conceived of is conceived of in certain circumstances, including the objects presence, its origins, its conditions of being there, its effects on other objects, the background of knowledge in front of which it is conceived of, the perspective taken on it and many more things, all of which depending in one or the other way on the subject conceiving of it and on the situation in which it is being conceived of, the object has its content in a certain context of being, context being a complex of relationships. What is conceived of is therefore a matter of the context into which the object is being placed while being conceived of by a subject. This observation guides the modelling of conception. Modelling Conception In the model of conception below86 conception is modelled as a particular kind of relationship relating a subject (in German Subjekt) and an entity (in German Entitt), and an entity in the sense of the model may be anything. However, an entity is not prescribed by the model as to being a thing as such, but is intended to be something which exists through the fact that it is being conceived of by some subject. Therefore, only if a conception is given, the model refers to the entity conceived of as an object (in German Gegenstand), and the model assumes the being of an object as to depend on how the object is being conceived of by the subject, i. e. to depend on the conceptions content (in German Inhalt). It prescribes the content of a conception as being determined by the context (in German Kontext) into which the conceiving subject sees the ob86 First ideas of the model have been discussed in Mahr (1997); later, these thoughts were used to build a conceptual fundament for the judgement of model-being, and the model was phrased in its present form and presented to the participants of the Workshop on “Judgement, Assertions, and Propositions – The Logical Semantics and Pragmatics of Sentences” in January 2008. In the proceedings of this workshop: Bab, Robering (2010). The model was published and justified by an analysis of Brentano’s, Twardowski’s and Searle’s work, see Mahr (2010), 61 – 87. A first realization of its clauses was given in Eilers (2008), and in Wieczorek (2009). The model has been investigated in its relation to set theory and to non-well-founded sets (u-sets) and a proof of consistency was given by the construction of a reflexive universe. Presently, with the focus on context management, an implementation of the model of conception is prepared by Thomas Karbe.

336

Bernd Mahr

Figure 9: Symbolizing the Relationship Between the Four Elements of Conception.

ject placed. The context of a conception is modelled as a complex (in German Zusammenhang) of relationships (in German Beziehungen) by which objects are being related. Searle claims in his treatise on intentional mental states “that Intentional states are in general parts of Networks of Intentional states and only have their conditions of satisfaction relative to their position in the network.”87 But what is it that constitutes such a network and what then is the content of a conception? An answer to this question can only be given in terms of a model exhibiting a particular conception of the world. This is also what Searle points at if he concludes that “There is no neutral standpoint from which we can survey the relations between Intentional states and the world and then describe them in non-Intentional terms. Any explanation of Intentionality, therefore, takes place within the circle of Intentional concepts.”88 The model of conception is not a particular conception of the world but rather a framework, in the sense of a conceptual model, by the application of which a conception of the world can be specified. It aims at the modelling of situations of conception (in German Situationen) and, in particular, at the modelling of universes (in German Universen). Since also a universe is an entity which is conceived of, and is therefore an object, it may even be part of itself. The interpretation of the concept of universe admits a huge variety of choices each representing a particular “world view”. 87 Searle (2004), 20 f. 88 Ibid., 26.

On the Epistemology of Models

337

Entity 1. An entity is something that is. Anything that is, is an entity. 2. An entity is the content of some conception. 3. Any two entities are different. Relationship 4. A relationship is an entity by which entities are related. 5. An entity belongs to a relationship, if it is one of the entities which are related by this relationship. Complex 6. A complex is an entity by which entities belong to relationships. 7. A relationship belongs to a complex, if the entities which belong to this relationship belong to this relationship by this complex. 8. An entity belongs to a complex, if it belongs to a relationship which belongs to this complex. Conception A conception is a relationship by which an entity, identifiable as the subject of this conception, an entity, identifiable as the object (or subject matter) of this conception, and a complex, identifiable as the context of this conception, are related. 10. The content of a conception is a complex, to which exactly those relationships belong, which belong to the context of this conception, and to which the subject matter of this conception belongs. 9.

Situation 11. A situation is a complex in which all entities which belong to this complex are conceptions. Universe 12. A universe is a complex to which with every entity which belongs to it, also a conception belongs, whose content this entity is. 13. A universe is called reflexive, if it belongs to itself.

In an application of the model a subject may be interpreted to be the mind of a human individual or anything else of which it can meaningfully be asserted that it takes the role of a subject in a particular conception of something, like for example a mathematical definition, a society, a cultural period, a machine or even a point or segment of time. And an object may be anything, of which it can meaningfully be asserted that it is being conceived of. Also the concepts of relationships, complexes, and contexts have a wide range of interpretations, admitting any kind of connection between objects. A universe may be interpreted to be a universe of discourse as it is prerequisite to information modelling, or a foundational frame like a universe in set theory, or it may be a universe in the sense of

338

Bernd Mahr

ontology, mirroring existence and obtaining states of affairs. In the latter case, the model can also be understood as ontology the entities of which are not independent from some mind to conceive of it, but to the contrary, are by their nature what is being conceived of. This view is particularly supported by the second clause which does not support the distinction between objects which are conceived of and objects which are contents of conception, the distinction Twardowski made in the framework of presentations. Thus, entities in a universe of conception are only of one kind. In the following model of model-being the model of conception serves as a reference model in the sense that the concepts employed in the modelling of a judgment of model-being are obtained by applying the model of conception. The interpretation of the model of conception, in turn, is widely left open. It is only assumed that subjects are humans and that entities are objects of intentional relationships. The consequence of using the model of conception as a reference model is that objects and relationships in the epistemic pattern of model-being are with no exception subject and context dependent entities.

The Model of Model-being The Epistemic Pattern of Model-being Referring to the epistemic pattern of model-being, as depicted in figure 1 above, an object M is not a model in itself, but only if it is conceived of as a model by a judging subject. Through the judgement by which the object M is conceived of as a model, M is placed in a context in which, according to the judging subject, M presents itself as a model. This context includes the relationships the combination of which constitutes the epistemic pattern of model-being. The conceiving subject may be a human being, but also a group of people or any other entity which may be expected to make a judgement, even in a figurative sense, such as a scientific discipline or a culture. For the discussion of modelbeing here, we assume that the judging subject is a human. Further, for the purpose of the analysis of the concept of model it must be assumed that the judgment on model-being is indivisible; since otherwise, the object conceived of as a model could not possess identity. This assumption does not rule out the possibility that one and the same object may be conceived of as a model in different ways at different points in time,

On the Epistemology of Models

339

Figure 10: The two relationships of creation, connected by an object M conceived of as model l.

even by the same judging subject. In that case, different judgements on the model-being of the object would be made in the course of time. Therefore, the answer to the question whether the indivisibility of the judgement on model-being is to be assumed for only a fraction of a second or for a period of several centuries will depend on the focus and the level of granularity chosen. Since the object M conceived of as a model may vary without turning into a different model in the interpretation of the judging subject, like an architectural diagram on a sheet of paper and the ‘same’ diagram on the display of a computer, or that diagram and a slight graphical variation of it, the distinction was made between the model object M and the purely mental object, the model l. The relationship between the two consists in the model l being represented by the model object M. The difference between M and l is a difference in identity. The identity of the model object M is the identity of the object M assumed by the judgment on model-being, whereas the identity of the model l is the identity of the object M as a model, which relies on the conception of the judging subject. At the same time, the degree of certainty of the identity of l may be taken as the result of a thought experiment in which the judging subject decides whether all possible kinds of objects represent the same model l or not. In the model of model-being it is assumed that the identity of an object as a model depends on the two basic relationships, between A and M and M and B, which the model object enters into according to its conception as a model. Typical of model-being is that both relationships stem from an action, an action which is either thought of or has actually been performed. In the action leading to the relationship indicating that M is a model of A, the model object M is chosen or constructed with reference to A, and in the relationship indicating that M is a model for B, B is chosen or constructed with reference to M. Both relationships are therefore

340

Bernd Mahr

Figure 11: The composition of relationships of creation through which an object M, in its capacity of model l, becomes the carrier of cargo w.

of the same kind. They are called relationships of creation and relate what is called a source object with what is called a target object. Due to the two relationships of creation, according to the conception of the judging subject, the model has two roles which together determine its identity as a model, and which at the same time justify our viewing of the model object as a model. Naturally, the same object M may be conceived of as a model in different ways, just as, for example the natural numbers may be seen as a model of the Peano axioms and as a model of counting and calculating89. In such cases, the relationships of creation will differ as well, and the appropriate thought experiments will result in different models l. When creating a model, it is normally assumed that a target object is produced which will subsequently, either temporally or logically, become the source object of some model application. The rational imperative of models is based on the fulfilment of this assumption: as a source object of model application, the model object must ensure the transfer of certain qualities with which, in its model capacity, it is so to speak loaded. What is transferred in this way via a model object is called the cargo of the model. Thus the identity of the cargo w depends not only on the identity of the model object M, but also on the identity of the model l, i. e. on what the object is conceived of, on the object M, and on the two relationships of creation. It is impossible to assume for the object M, which is conceived of as a model, to have obtained its model role and its function as the carrier of a cargo for a target object B, out of the blue. M must have been chosen or produced as a model, in the conception of the judging subject, in such a way that it actually carries the cargo for 89 The possibility of multiple views on the same object as a model is in more depth discussed on the basis of Euler’s famous drawing of the Königsberger Brückenproblem in Mahr/Velminski (2008), 87 – 102.

On the Epistemology of Models

341

the transport of which it has to serve as a model. Also, the cargo must have some kind of origin, since otherwise it would be completely arbitrary, meaning that the model would not be linked to a purpose. Besides, the cargo must be conceived of as identifiable by the judging subject. Last but not least, the practice of modelling aims at ‘writing or working the cargo into the model object’ during the model creating process in such a way that it may also be transferred to the resulting object B in the course of model application. Examples show that the cargo w of a model is not necessarily identical with the model l90. From this observation it can be concluded once more that an object may be interpreted as a model only if it is viewed as something which enters into both relationships of creation, i. e. the relationship of of as well as the relationship of for. Otherwise, its carrier function and the potential of its application as a model would be inexplicable. On the Epistemic Structure of Relationships of Creation For a better understanding of the epistemic pattern of model-being, the conception of the relationships of creation has to be subjected to closer analysis. Relationships between objects are themselves objects of conception. They are therefore not only what they are conceived of by some subject but also depend on the context in which they are placed. In the case of a relationship of creation this context includes an act in which the target object of the relationship is chosen or constructed with reference to its source object. Relationships of creation are most natural types of relationships. They are present whenever something is or has been produced, selected, assigned a role or is obtained by measurement, abstraction or mapping. Examples of relationships of creation are the relationship between the human body and the proportions in Leonardo’s drawing of the Vitruvian figure, the relationship between the axioms of Peano and the natural numbers or between the principles of counting one by one and the natural numbers, the relationship between the computational model of Turing machines and a particular choice of a Turing machine, or the relationship between observations made on the role structure and functioning of systems in which a retailer provides services to a customer and the TINA-C model; relationships of 90 It may, for example, be part of a model to include instructions of its application in the form of comments or legends; such instructions then are ‘part of’ the model l, but they are definitely not part of its cargo w.

342

Bernd Mahr

Figure 12: Albrecht Dürer, The Drawer of the Reclining Woman.

creation exist whenever an original is being copied, when something is constructed according to some model or when in modeling some feature is built into the model. The precise description of an act leading to a relationship of creation is seldom possible, neither of the realistic progress of the act nor of the details of the relationship it constitutes. That is one of the reasons why I use here the puzzling example of Albrecht Dürer’s famous woodcarving The Drawer of the Reclining Woman 91, as an aid in my analysis. Dürer’s woodcarving was originally meant to illustrate a Renaissance technique of drawing non-geometrical three dimensional objects with correct perspectivity; here it is used to illustrate the rationale of the epistemic structure of relationships of creation. The depicted situation is a technically supported act of transfer, which can be phrased very precisely. The operational steps of its processing reveal the epistemic structure of the relationship of creation, which the act constitutes. In this act of transfer three different types of operation can be distinguished: observation, transformation and realization. The operations of observation can be compared to sighting with a gun, the tip of the obelisk corresponding to the front sight, a square of the vertical grid corresponding to the rear sight, and the precise localization of the outlines and shades visible in the grid square selected corresponding to the target. The drawer performs the transfer without technical aids; having selected the appropriate square of the grid placed in a vertical position in front of him, he transforms what he observes in the appropriate square of the vertically positioned grid into a pattern which determines the exact positions in which points, lines, and shades will have to be charted on the draft paper which is structured by means of the horizontal grid. Subse91 Dürer (1525).

On the Epistemology of Models

343

quently, the operation of realization consists in charting the points, lines, and shades as required by the pattern for each individual square. The drawer will not perform the process of this transfer act in three steps only, but will transform his observations into realization square by square. However, the individual steps performed by him may be typed according to those three types of operation, the degree of granularity being determined by the size of the squares. If we now abstract from the specifics of the process of this transfer act and look at the relationship it constitutes, we recognize that this relationship is, from an epistemic point of view, structured in accordance with the order in which the three types of steps are performed: first observation, second transformation, and third realization. What can be seen from an analysis of Dürer’s woodcarving, can sensibly be taken as a general principle of structuring relationships of creation: whenever a target object is produced with reference to a source object, there have to be observations made on the source object, which then have to be transformed into requirements for the realization of the target object. The epistemic structure of relationships of creation is thus found to be composed of three component relationships in consecutive order: first the relationship relating the source object with an observation on it, called admitting (an observation), second the relationship relating the observation with requirements, called transformation, and third the relationship relating the requirements with the target object, called realization. On the basis of this epistemic structure the logic of relationships of creation can be analyzed. On the Logic of Relationships of Creation From a logical point of view, the relationships of observation and realization are closely related with the relationship of satisfaction relating objects and sentences. Observation and realization are both supported by grids, which may be viewed as constituting the languages of observation and realization, which permit, so to speak, the making of statements about outlines and shades as well as about points, lines, and tinges, respectively, in the squares of the grids. A difference between the relationships of observation and realization is the order of their presence: In an observation the object observed is present before the statements are available which express what has been observed, and in a realization the statements expressing what is being required, are present before the object

344

Bernd Mahr

Figure 13: The Relationship of Satisfaction (see also figure 5 above).

Figure 14: Two Readings of Satisfaction.

is available which fulfills the requirements. More formally, there are two readings of satisfaction, namely observation and realization. Depending on the language L, the object observed or realized can be thought of to be replaced by its theory, i. e. by the set of sentences satisfied by the object (see also above). Then observations and requirements are both subsets of this theory but with the difference that the relationship of admitting an observation corresponds to a deduction, while the relationship of realization corresponds to an induction. The relationship of creation can now, from a logical point of view, be seen as to be composed of a deduction, a transformation, and an induction. The logic of relationships of creation applies now to the logic of model-being.

On the Epistemology of Models

345

Figure 15: The Logic of a Relationship of Creation.

Figure 16: The Epistemic Structure of Model-Being.

On the Logic of Model-being The epistemic pattern of model-being includes the composition of two relationships of creation92, one with source object A and target object M, and one with source object M and target object B. The epistemic structure of relationships of creation can now explain how in the model of model-being A and B are to be determined: A may be one of the three, a source object, an observation, or requirements, and B, accordingly, may be one of the three, a target object, requirements, or an observation. This choice in the determination of A and B reflects the rather open use of models in their of- and for-relationships. An object may be a model of requirements, a model of observations, or of another object, and, accordingly, an object may be a model for observations, for the statement of requirements, or for another object. In daily model use there are obvious examples for each of these cases. The logic of model-being is the derived from the logic of relationships of creation, and from the epistemic structure of model-being. It 92 See figure 10 above.

346

Bernd Mahr

bestows a distinguished role upon the model object M linking two relationships of creation: in the judgment on model-being, the model object M functions not only as the target object, but also as the source object of a relationship of creation, which means that it represents at the same time the conclusion of an act of induction as well as the premise of an act of deduction. This dual role seems to be one of the reasons why it is difficult to explain the nature of models; it is impossible to think, i. e. to be consciously aware of something which is something not yet present but meant to be created and at the same time is something which is already available and meant to be used. The solution to this problem found here seems to work only because of the contextual explanation of conceptions and the explanation of model-being in terms of a judgment; it lets in a judgment on model-being the two relationships of creation be part of the context in which the model object is placed if it is conceived of as a model. This dual role played by a model object, however, does not only create problems of awareness in judgments on model-being, but also points to one of the central challenges of modeling, particularly in situations where model creation and model use are not practiced by one and the same subject. Then the challenge is in choosing or constructing a model object in such a way that the relationship between the requirements on the realization of the model object and the observations made on it will also permit the realization of the intended cargo of the model. In the course of history, models have contributed considerably to the division of labor in working processes, and they have always been the carriers of knowledge. This use of models is possible only if it is assumed that the creators of models have generally succeeded in making them function as carriers and conveyors, and that the knowledge content written into them could be deduced and the intended cargo be conveyed when they were applied. However, in general, when dealing with models created by somebody else, one cannot assume that knowledge of the model object will also imply knowledge of the requirements on the model object, dating from the modeling process. As a consequence, in an extreme case, when the model is applied, the observation on a model object will hold no further reference to the intended cargo with which it was loaded93. This phenomenon of possible misinterpre93 Thus it would be imaginable for a wooden car, modelled many years ago in rough shape as a Volkswagen by a person A, and accidentally colored red, to be viewed many years later, as a consequence of its color, as a Ferrari by a per-

On the Epistemology of Models

347

tation is one of the reasons why model objects often contain information on their proper use, like the legend on a map, or that models are created under the restrictions of established frameworks or disciplines, like in information science or in mathematics, or that the use of models is restricted to communities the members of which share common knowledge of model creation and use. In the view of logic, the significance attributed to the contextual relationships in the conception of an object as a model corresponds precisely to the argument 94 by means of which the judgment on modelbeing may be justified. Thus if someone were to ask on what grounds the judgment on the model-being of an object was based, the judging subject would be able to base his argument on those contextual relationships. Whether the person asking the question would find those grounds acceptable and the argument convincing depends in turn on whether he would be prepared to accept the model of model-being, as represented by means of the epistemic pattern of model-being, as his individual notion of model, and whether he would agree that this model of modelbeing could rightfully be applied for justifying the specific judgment on model-being95. And if he would be prepared to do so, the model of model-being would, as a general model, serve as a condition of correctness of the argument by means of which the judgment on the modelbeing is justified. In this sense, the instruments of logic can be used to explain the notion of models. However, in order to allow the application of these instruments of logic, it was necessary to rephrase the question of the nature of models as a question of the justification of a judgson B. In that case the wooden car would, in its capacity of model of a Volkswagen, become the model for a Ferrari. The reason why this situation may arise is that the two judgements on the model-being of the wooden object are made separately. However scantily and succinctly phrased, this phenomenon of nonintended interpretation is inherent in the concept of model. Thus many models are characterized by referring to their cargo by means of elements of their configuration; for example, the reading of a map is supported by the design of its colors and the captions referring to its symbols. See also Mahr (2008b); cf. fn. 29. 94 On the basic concepts of logic applied, see Salmon (1983), 7 – 40. 95 This is to a large extent analogous to that of deductions of propositional logic, the correctness of which must be proved by means of the truth table model of connectives of propositions, developed in propositional logic. There is a basis for doubt in propositional logic as well, such as in the case of the controversy over the ,tertium non datur‘, not recognized as a tautology in intuitionistic logic.

348

Bernd Mahr

ment in which an object is conceived of as a model. It is a fact, though, that in practice, a judgment to be made cannot always be viewed equally strictly within the context of justification, but will to a large extent also have to be viewed in the permissive framework of the context of invention. This becomes evident in the case of judgments for which the full contextual structure of the model relationships is not made explicit because it is possible to infer the model-being of the object from object characteristics such as of reduced scale or abstracting alone. However, the model of model-being has found its position in the context of justification as a result of its utilization as an ‘epistemic pattern’. For that reason, it is an analytical instrument of observation and theory-building rather than a practical tool of routine model-judging.

Concluding Statements Models are not true, not completely individual and not irrefutable. They embody possibility and choice, and can be extremely powerful. There is an abundant model use. The life cycle time of models gets shorter and the need for models is growing. It is more and more difficult to capture models in single disciplinary domains. Models are “interdisciplinary”. Only within the narrow borders of a discipline a judgment of the value of a model is possible. Because models are the most important epistemic tool of our knowledge and production, it is necessary to produce a methodological surplus when answering the question “What is a Model?” From a general model theory one has to expect, that it describes in general terms and in great precision and clarity all the major phenomena which appear in situations of concrete modeling and model use.

References Bab/Robering (2010): Sebastian Bab/Klaus Robering (eds.), Judgements and Propositions – Logical, Linguistic and Cognitive Issues, Berlin. Bays (2001): Timothy Bays, “On Tarski On Models”, in: Journal of Symbolic Logic 66, 1701 – 1726. Bredekamp (2005): Horst Bredekamp, “Modelle der Kunst und der Evolution”, in: Modelle des Denkens, 13 – 20. Brentano (1995): Franz Brentano, Psychology from an Empirical Standpoint, L. L. McAlister (ed.), London.

On the Epistemology of Models

349

Brentano (2008): Franz Brentano, “Psychologie vom Empirischen Standpunkt – Von der Klassifikation der psychischen Phänomene”, in: T. Binder/A. Chrudzimski (eds.), Franz Brentano: Smtliche verçffentlichte Schriften, vol. 1, Heusenstamm, 97 – 103. Broy/Rumpe (2007): Manfred Broy/Bernhard Rumpe, “Modulare hierarchische Modellierung als Grundlage der Software- und Systementwicklung”, in: Informatik Spektrum 30. 1. 2007, Heidelberg, 3 – 18. Bolzano (1981): Bernhard Bolzano, Wissenschaftslehre, vol. 1, Aalen. Cantor (1984): Georg Cantor, ber unendliche lineare Punktmannigfaltigkeiten, Günther Asser (ed.), Leipzig. Cardone/Hindley (2009): Felice Cardone/J. Roger Hindley, “History of Lambda-calculus and Combinatory Logic”, in: D. M. Gabbay/J. Woods (eds.), Handbook of the History of Logic, vol. 5, London/Amsterdam/New York. Chadarevian/Hopwood (2004): Soraya de Chadarevian/Nick Hopwood (eds.), Models – The Third Dimension of Science, Stanford. Church (1965): Alonzo Church, “A Note on the Entscheidungsproblem (1936)”, in: Davis, 110 – 114. Da Vinci (2003): Leonardo da Vinci, “Proportional sketch based on Vitruv, Venice, Gallerie dell’Accademia, Inv. 228”, in: F. Zöllner, Leonardo da Vinci – Smtliche Gemlde und Zeichnungen, Köln. Dauben (1979): Joseph Warren Dauben, Georg Cantor. His Mathematics and Philosophy of the Infinite, Princeton. Dirks/Knoblauch (2008): Ulrich Dirks/Eberhard Knoblauch (eds.), Modelle, Frankfurt/Main. Dürer (1525): Albrecht Dürer, Unterweysung der Messung mit dem Zirkel und Richtscheydt, Nürnberg. Eilers (2008): Florian Eilers, u-semantische Modellierung eines Modells der Auffassung, Diploma Thesis, Technische Universität Berlin. Ehrig/Mahr (1985): Hartmut Ehrig/Bernd Mahr, Fundamentals of Algebraic Specifications 1 – Equations and Initial Semantics, Berlin/Heidelberg. Ehrig/Mahr (2001): Hartmut Ehrig/Bernd Mahr et al., Mathematisch-strukturelle Grundlagen der Informatik, Berlin/Heidelberg, 333 – 352. Feferman (2004): Solomon Feferman, “Tarski’s Conceptual Analysis of Semantical Notions”, in: A. Benmakhlouf (ed.), Smantique et pismologie, Casablanca, 79 – 108. Flügge (2008): Matthias Flügge, Adaptive Service Composition in a Semantics-based Interoperability Infrastructure, PhD Thesis, Faculty of Electrical Engineering and Computer Science, TU Berlin. Frege (1966): Gottlob Frege, “Logische Untersuchungen – Gedankengefüge”, in: G. Patzig (ed.), Gottlob Frege – Logische Untersuchungen, Göttingen, 72 – 91. Freudenthal (1961): Hans von Freudenthal, The Concept and Role of the Model in Mathematics and Natural and Social Sciences, Dordrecht. Gödel (1965a): Kurt Gödel, “On Formally Undecidable Propositions of Principia Mathematica and Related Systems (1931)”, in: M. Davis, The Unde-

350

Bernd Mahr

cidable – Basic Papers On Undecidadle Propositions, Unsolvable Problems And Computable Functions, New York, 5 – 37. Gödel (1965b): Kurt Gödel, “On undecidable propositions of formal mathematical systems (1934)”, in: Davis, 39 – 73. Gödel (1986): Kurt Gödel, “Die Vollständigkeit der Axiome des logischen Functionenkalküls (1930)”, in: Monatshefte für Mathematik und Physik 37, 349 – 360, reprinted in: S. Feferman/S. Kleene/G. Moore/R. Solovay/J. van Heijenoort (eds.), Kurt Gçdel, Collected Works. I: Publications 1929 – 1936, Oxford. Goodman (1995): Nelson Goodman, Sprachen der Kunst, Frankfurt/Main. Grossmann (1992): Reinhardt Grossmann, The Existence of the World – an Introduction to Ontology, London. Hart (1998): Vaughn Hart, “From Virgin to Courtesan in Early English Vitruvian Books”, in: Vaughn Hart/Peter Hicks (eds.), Paper Palaces. The Rise of the Renaissance Architectural Treatise, New Haven/London, 297 – 318. Hesse (1963): Mary B. Hesse, Models and Analogies in Science, London. Hesse (2006): Wolfgang Hesse, “Modelle – Janusköpfe der Software-Entwicklung- oder: Mit Janus von der A- zur S-Klasse”, in: Modellierung, Heidelberg, LNI P-82, 99 – 113. Hodges (1985): Andrew Hodges, Allen Turing – The Enigma of Intelligence, London. Husserl (1991): Edmund Husserl, Ding und Raum, Hamburg. Kneele (1983): Kenneth D. Kneele, Leonardo da Vinci’s Elements of the Science of Man, New York. Knell (1985): Heiner Knell, Vitruvs Architekturtheorie, Darmstadt. Kruft (1995): Hanno-Walter Kruft, Geschichte der Architekturtheorie, München, picture 36. Kuhn (1988): Thomas S. Kuhn, “Neue Überlegungen zum Begriff des Paradigma”, in: Thomas Kuhn, Die Entstehung des Neuen – Studien zur Struktur der Wissenschaftsgeschichte, Frankfurt/Main, 389 – 420. Kühne (2006): Thomas Kühne, “Matters of (Meta-) Modeling”, in: Journal on Software and Systems Modeling, vol. 5, no. 4, 369 – 385. Mahr (1997): Bernd Mahr, “Gegenstand und Kontext – Eine Theorie der Auffassung”, in: K. Eyferth, B. Mahr/R. Posner/F. Wysotzki (eds.), Prinzipien der Kontextualisierung, KIT Report 141, Technische Universität Berlin. Mahr (2004a): Bernd Mahr, “Das Mögliche im Modell und die Vermeidung der Fiktion”, in: T. Macho/A. Wunschel (eds.), Science & Fiction – ber Gedankenexperimente in Wissenschaft, Philosophie und Literatur, Frankfurt/Main, 161 – 182. Mahr (2004b): “Modellieren. Beobachtungen und Gedanken zur Geschichte des Modellbegriffs”, in: S. Krämer/H. Bredekamp (eds.), Bild-SchriftZahl, München, 59 – 86. Mahr (2006): Bernd Mahr, “Die Schöpfung der Maschine – Ein Modell des Entstehens und der Gegensatz vom Maschinellen und Kreativen” in: G. Abel, Kreativitt, Hamburg, 376 – 390. Mahr (2008a): Bernd Mahr, “Ein Modell des Modellseins. Ein Beitrag zur Aufklärung des Modellbegriffs”, in: Dirks/Knobloch (2008), 187 – 218.

On the Epistemology of Models

351

Mahr (2008b): Bernd Mahr, “Cargo – Zum Verhältnis von Bild und Modell”, in: I. Reichle/S. Siegel/A. Spelten (eds.), Visuelle Modelle, München. 17 – 40. Mahr (2009): Bernd Mahr, “Information Science and the Logic of Models”, in: Journal of Software and Systems Modelling 8, Berlin/Heidelberg, 365 – 383. Mahr (2010): Bernd Mahr, “Intentionality and Modelling of Conception”, in: S. Bab/K. Robering (eds.), Judgements and Propositions – Logical, Linguistic and Cognitive Issues, Berlin, 61 – 87. Mahr/Velminski (2008): Bernd Mahr/Wladimir Velminski, “Denken in Modellen – Zur Lösung des Königsberger Brückenproblems”, in: H. Bredekamp/W. Velminski (eds.), Mathesis & Graph, Berlin, 87 – 102. Mittelstraß (2005): Jürgen Mittelstraß, “Anmerkungen zum Modellbegriff”, in: Modelle des Denkens, 65 – 67. Morgan/Morrison (1999): Mary S. Morgan/Margaret Morrison, “Models as Mediating Instruments”, in: Mary S. Morgan/Margaret Morrison (eds.): Models as Mediators, Cambridge, 10 – 37. Präsident d. Berlin-Brandenburgischen Akademie d. Wissenschaften (2005): Präsident d. Berlin-Brandenburgischen Akademie d. Wissenschaften (ed.), Debatte. Heft 2: Modelle des Denkens, Berlin. Reisig (2006): Wolfgang Reisig, “50 Jahre modellbasierter Entwurf: Vom Modellieren mit Programmen zum Programmieren mit Modellen”, in: W. Reisig/J.-Ch. Freytag (eds.), Informatik – Aktuelle Themen im historischen Kontext, Berlin/Heidelberg, 276 – 314. Runggaldier/Kanzian (1998): Edmund Runggaldier/Christian Kanzian, Grundprobleme der Analytischen Ontologie, Paderborn. Salmon (1983): Wesley C. Salmon, Logik, Stuttgart. Searle (2004): John R. Searle, Intentionality – An Essay in the Philosophy of Mind, Cambridge. Schuler (1999): Stephan Schuler, Vitruv im Mittelalter – Die Rezeption von ,de architectura’ von der Antike bis in die Frhe Neuzeit, Köln. Stachowiak (1973) Herbert Stachowiak, Allgemeine Modelltheorie, Wien/New York. Studium Generale (1995): Journal Studium Generale, vol. 18, No. 3 – 7, Berlin/ Heidelberg. Suppes (1961): Patrick Suppes, “A comparison of the meaning and uses of models in mathematics and the empirical sciences”, in: von Freudenthal (1961), 163 – 177. Tarski (1935): Alfred Tarski, “Über den Begriff der logischen Folgerung”, in: Actes du congrs international de philosophie scientifique, vol. VII, Paris. TINA Consortium (1997): TINA Business Model and Reference Points, Version 4.0, May 20th, 2. Turing (1936): Alan Turing, “On Computable Numbers, with an Application to the Entscheidungsproblem”, in: Proceedings of the London Mathematical Society (2) 42, 230 – 265. Twardowski (1977): Kasimir Twardowski, On the Content and Object of Presentations – A Psychological Investigation, The Hague.

352

Bernd Mahr

Twardowski (1982): Kasimir Twardowski, Zur Lehre von Inhalt und Gegenstand der Vorstellungen – Eine Psychologische Untersuchung, München/Wien. Umbach (1996): Carla Umbach, Termprzisierung – kontextuelle Steuerung der Interpretation durch Apposition und Typisierung, TU Berlin, PhD Thesis (KITReport 140, ISSN 0931 – 0436). Vitruv (1981): Zehn Bcher ber die Architektur, Darmstadt. Wartowsky (1979): Marx W. Wartowsky, Models – Representation and the Scientific Understanding, Dordrecht. Wendler (2008): Reinhard Wendler, Die Rolle der Modelle in Werk- und Erkenntnisprozessen, PhD thesis at the seminar on art history at the Faculty of Philosophy III at the Humboldt University of Berlin. Wieczorek (2009): Tina Wieczorek: On Foundational Frames for Formal Modelling – Sets, u-sets and a Model of Conception, Aachen. Wiesing (2009): Lambert Wiesing: Das Mich der Wahrnehmung, Frankfurt/ Main. Wittgenstein (1967): Ludwig Wittgenstein: Philosophische Untersuchungen, Frankfurt/Main. Zöllner (1987): Frank Zöllner, Vitruvs Proportionsfigur. Quellenkritische Studien zur Kunstliteratur im 15. Und 16. Jahrhundert, Worms.

Explanation Through Description1 Michael Hampe 1. Introduction “If you had experienced Hitler’s manner of speaking yourself, it would be clear to you that this man was capable of everything.” Such statements as this, made in accounts that report personal experiences, are admittedly no replacement for justifications of the assertion that Hitler was capable of everything (whatever this means in particular), yet to the end of generating plausibility, experiences are supposed to be conveyed through such descriptions in a certain incomplete way. This is in many contexts a function of verbal statements. With language humans want not only to justify assertions, but also to describe experiences and pass them on through narration, so that a narrative context shall arise around the assertions, similar to the experiential context, which is itself no inferential context, but still has a plausibilising nature. The experience of heroin addiction can suffice for someone who has been dependent upon heroin to hold the assertion to be true that heroin is a bad thing. The experience as such is not really repeatable. Yet the narration of heroin addiction, the description of how it is to be dependent on heroin can be the attempt to convey the experience so that it may be transmitted after all, without that specific suffering that accompanies it, and that a corresponding plausibilisation may occur in the recipient of the narration of the person who has really had the unpleasant experience himself. Non-fictitious narration uses the human capability of empathy in order to yield at least weak imitations of an actual experience. Yet even every fictitious narration can use and transmit experiences that the narrator has actually made, and thus, insofar as the narration changes the household of plausibility for a person, it can have an effect similar to that of a real life experience. 1

I would like to thank Sabine Baier (Zürich) for improving this text and Helmut Pape (Bamberg) for according us with the copies of the Peirce-manuscripts used in this essay.

354

Michael Hampe

Here, I would like to address the relation between narrative and deductive plausibilisations in the observation of nature. For this the transmission of experiences through narrations of fact or fiction will be less important than the role played in the generation of understanding by sequences, be they deductive, narrative or genetic, in which no clear distinction can be made between deduction and narration. I shall try to show that the sequence of the presentation of facts in a description can also be conceived as a form of explanation, even if it provides no retracing of facts to generalities, as is the case with deduction.

2. Microscope and Telescope: Lichtenberg’s Semiotics The Göttinger professor Lichtenberg, who became famous as an experimental physicist, was also a critic of Linnaeus’ systematics of animals and plants. He considered our life to be too short, and the brain’s resources too slight, to be filled up by memorizing the nearly endless descriptions and categorizations of animals and plants in classes and subclasses. Lichtenberg’s critique of Lavaterian physiognomy and craniology, however, which was also a descriptive project, became pre-eminently famous. From these two observations one might, at a superficial glance, infer that Lichtenberg was a representative of modern explanatory science who regarded the purely descriptive doctrine of nature or narrative natural history as secondary. If one regards descriptions of nature as narrations and explanations in the modern sense as inferences from general to specific statements, then one could tend to the conception that the German progenitor of experimental physics was a friend of inferential explanation and an enemy of descriptive narration in natural science. Lichtenberg also wrote the following words, however, in 1778: Every feeling examined under the microscope can be magnified through a book. … The fall of my hand upon a fold of a silk curtain while I sleep, this sensation can grow and bloom into a dream whose description would require a book.2

This quote shows that a classification of natural science in Lichtenberg into one category of enterprises which legitimize its insights descriptive2

“Jedes Gefühl unter dem Mikroskop betrachtet lässt sich durch ein Buch durch vergrössern. (…) Meine Hand im Schlaf auf eine Falte eines seidenen Vorhangs geschlagen, diese Empfindung kann zu einem Traum aufwachsen und blühen, dessen Beschreibung ein Buch erfordert” (Lichtenberg (1983), I, F500).

Explanation Through Description

355

ly and narratively, and another which proceeds by deductive inferences, is too simple. The complexity of the sensation that arises when the hand feels a fold of silk, without its owner knowing upon what it rests, could, according to Lichtenberg, require a book in order to be known in its complexity. In this context it is important that the focus is not on sorting the sensation into a class, but on the possibility that the sensation might grow into a dream. The metaphor of growing can be interpreted both causally and semiotically in this context, inasmuch as these two hermeneutical possibilities can be separated in Lichtenberg. The sensation of the silken fold can have all sorts of causal consequences for the experiencing person, or it can trigger in him a sign activity, a chain of associations that fill an entire dream. Nor does the other metaphor used by Lichtenberg here, that of the microscope that renders a fine structure visible, which escapes the view of the naked eye, indicate a causal mechanism to be discovered, which deductive explanations would make possible. Microscope and telescope rather make possible exact descriptions of structures that otherwise would not be experienceable. Lichtenberg, the astronomer, undoubtedly had a sense of the scientific significance of these, as instruments not necessarily embedded in explanatory contexts. The view of the stars through a telescope or of a microorganism through a microscope, as such, explains nothing. At first something only becomes visible that was hitherto invisible. The instruments make something possible that might be compared to a journey into a region of the world hitherto unknown: new experiences. Then a description of what has been seen follows that might occur in the form of a narrative text, or also, especially in the case of biological microscopy, a drawing or, today more likely, a photograph. This documentation of the observed corresponds to the narrations in a travelogue. It is intended to transmit to others the experience are had while looking through the instrument for whom this view was not possible. Just as Darwin’s famous narrations and drawings of the beak forms of the finches named after him justified his use of the adaptive term, so can reports about microscopic or telescopic journeys in small or distant worlds justify why one holds a certain biological or physical assertion to be true. The assertion e. g. that an organism is a hierarchically structured entity like the state goes back to the discovery of its multicellular composition, which was in turn plausibilised by the view through the microscope. Into such descriptions flow already interpretations about the causal locus of the objects observed. For the structure of the described can be interpreted both as a trace of the influences to which it was exposed,

356

Michael Hampe

and as a sign of the effects that it will yield. Interpretations of causal loci occur in narrations, then, when they are concerned with concrete causal relations, and not with such that are known through regularities renderable in general statements. Thus, narrations are also inferentially structured to a certain extent. Yet the explicit or implicit inferences in narrations are almost never of the sort in which steps are made from or to universalities, as in a deductive nomological explanation. When, for instance, in a biological narration of the locomotive speed and feeding of kangaroos in Australia, these are compared with those of the European deer, in order then to conclude that they occupy the same ecological niche, the abstractum of the ecological niche is admittedly referred to. Yet the conclusion occurs via comparative narration of the behaviour and feeding habits of the different organisms. We shall return later on to this difference between narrative reports on concrete causalities on the one hand and explanations buttressed by universal regularities on the other. Behind a causal interpretation of a narrative description often stands the idea, also formulated by Lichtenberg, that in an individual thing its causal interconnectedness in the world is “mirrored”. Therefore it is possible in an exact narrative description to uncover the causal network the individual is imbedded in. I quote Lichtenberg once again: No one will deny that in a world in which everything is related through cause and effect, and where nothing happens by way of miracles, every part is a mirror of the whole. If a pea is shot into the Mediterranean Sea, a sharper eye than ours could sense … the effect thereof on the coast of China. And what else is a light particle that hits upon the retina compared to the mass of the brain and its branches? This often places us in a position to make inferences about what is distant from what is nearby… So the scratches on the bottom of a tin plate tell the story of all meals to which it has been a party, just as the form of every strip of land, the shape of its sand hills and rocks contains the story of the earth.3 3

“Niemand wird leugnen, dass in einer Welt, in welcher sich alles durch Ursache und Wirkung verwandt ist, und wo nichts durch Wunderwerke geschieht, jeder Teil ein Spiegel des Ganzen ist. Wenn eine Erbse in die Mittelländische See geschossen wird, so könnte ein schärferes Auge als das unsrige… die Wirkung davon auf der chinesischen Küste verspüren. Und was ist ein Lichtteilgen, das auf die Netzhaut des Auges stösst, verglichen mit der Masse des Gehirns und seiner Äste anders? Dieses setzt uns oft in den Stand, aus dem Nahen auf das Ferne zu schliessen… So erzhlen die Schnitte auf dem Boden eines zinnenen Tellers die Geschichte aller Mahlzeiten, denen er beigewohnt hat, und ebenso enthält die Form jedes Landstrichs, die Gestalt seiner Sandhügel und Felsen, mit

Explanation Through Description

357

In this text the interwovenness of the analysis of cause-effect connections with narrations becomes especially clear. The plate itself “tells” the story of the number of meals that have been eaten from it, and in which the knives used have affected it and left their mark. The trace, the index, is that of which can be reported in the description of a thing. The description, however, when properly interpreted, points to the concrete causal nexus in which the concrete thing exists or has existed, the nexus that explains why a thing is the way it is. The causal nexus referred to here is not that of a universal natural law, but of concrete natural events that, due to their efficacy, have left traces. And it is in this very sense that Lichtenberg would think physiognomy possible. While Lavater would infer the character of a person from the description of skull bones and the form of the nose independently of their actual integration in certain life contexts, Lichtenberg would interpret the traces that life has left in a face the way the knife leaves them in a plate from which food has been eaten for years. In this context Lichtenberg is less interested in the mostly immobile bones than in the way in which muscles and skin have been formed after years of contraction and relaxation, laughter and weeping. Thus a face may be “legible” in Lichtenberg’s sense, when one tells of it. Yet whereof one attempts to tell, and what one attempts to infer in this narration are not universal laws, but the concrete life stories of concrete persons. Most significant in this context is the example of Galileo from the original story of the telescope. The moon presented a boundary in the explanatory and descriptive strategies of the Ptolemaic world. In the perfect world of eternal, isomorphic circular motions beyond the moon exact mathematic descriptions and prognoses were possible, but not in nature beneath the moon. On this side of the moon the descriptions and explanations must all refer to “irregular” phenomena, such as accelerations and decelerations of movements, or genesis and passing away of bodies. Beyond the moon apparently nothing came to be nor passed away, there were no stories and no changes, but only repeated motions. Galileo’s view of the moon through the telescope and his report that he had seen mountains and their shadows there, and his discovery that Jupiter also was orbited by apparently corporeal planets were the beginning of the end of this separation of the world into a mathematically exactly conceivable heaven and an only vaguely understandable natürlicher Schrift die Geschichte der Erde (…)” (Lichtenberg (1983), III, 264 f., my italics).

358

Michael Hampe

sublunar nature. Galileo told in his accounts of what he had seen through the telescope and thus changed the explanatory strategies visà-vis nature fundamentally. Here narrations of new sensory impressions are the foundation for plausibilisations of new strategies of knowledge.

3. Evolution and Cosmic Weather: Wright and Spencer The most commonly discussed example of a plausibilising tale today from whence a shift in the explanatory strategies in science arose, however, is not to be found among the reports of images to be seen in micro- and telescopes, but in connection with the biological doctrine of evolution, more precisely in connection with Darwin’s reports of the journey with the beagle and his tale of the Origin of Species. Immediately after the publication of Darwin’s Origin of Species a debate broke out in the USA about whether this was a universalisable theory that states laws and explains nomologically, or not. The historical place of this debate was Boston in the second half of the nineteenth century. For there existed a union of later famous American intellectuals such as Chauncey Wright, Charles Sanders Peirce, Oliver Wendell Holmes and William James, who called themselves “the Metaphysical Club” starting in 1872, in which, immediately following the publication of Darwin’s Origin of Species in 1859, the book was read and debated.4 Chauncey Wright’s treatment of the Darwinian doctrine was the first to include a warning against empirically unfounded nomological transfers of this text to the end of constructing a universal explanatory theory in other sciences, and especially against philosophical generalizations. For Wright the Darwinian text was only a tale of the development of living beings, and was as such convincing in this context. Wright denied its applicability to anorganic or intellectual-cultural aspects of reality. Nor were universal cosmological principles to be induced from it, according to Wright. He described the development of the anorganic cosmos rather as a kind of “cosmic weather” not sufficiently comprehensible even after the Darwinian conception of evolution.5 For him Cosmology is no explanatory science in the first place, and cannot be made into one with supposedly universal theories of evolution, but the development of the world in its entirety falls rather into an “infinite 4 5

Cf. Menand (2001), 120 – 140. Wright (1958), 106 – 117.

Explanation Through Description

359

variety of manifestations … of causes and laws”, without any visible natural tendency to allow a theoretical standardisation of the relevant factors.6 One can only report narratively of the origin and development of the world in its entirety, but not explain anything universally. Instead of analogising the development of the world as a whole with that of animals explanatorily in a theory, Wright tended to compare it to the complex, contingent development of weather and climate, which cannot be grasped by any one discipline. He directed these considerations against the transfer of Darwin’s tale of evolution to other sciences that started shortly after the publication of the Origin of Species and such generalizations as were made especially by one of Darwin’s most famous philosophical contemporaries, Herbert Spencer, who had coined the phrase “survival of the fittest” (in his Principles of Biology), which Darwin took on in later editions of the Origin of Species. Accordingly, Spencer became a target for Wright’s criticism. For Wright, Mr. Spencer’s “speculations” are nothing but abstract portrayals of cosmological concepts, which the human mind spontaneously supplies in the absence of facts. With the assumption of a teleology of evolution of the cosmos toward the complex, or toward chaos, the way is prepared for a “mythical instinct” by imagining a goal in a final “order” or “disorder” that has no hold at all in the narrative description of the evolution of animals.7 (This criticism also pertains to the evolutionary generalizations of thermodynamics in cosmology, as the physicist Ludwig Boltzmann undertook later with his talk of “heat death”8). Wright had also levelled this critique against such reflections as Spencer’s following words: Everywhere and to the last, therefore, the change at any moment going on forms a part of one or other of the two processes. While the general history of every aggregate is definable as a change from a diffused imperceptible state to a concentrated perceptible state, and again to a diffused imperceptible state; every detail of the history is definable as a part of either the one change or the other. This, then, must be that universal law of redistribution of matter and motion, which serves at once to unify the seemingly diverse groups of changes, as well as the entire course of each group. The processes thus everywhere in antagonism, and everywhere gaining now a temporary and now a more less permanent triumph the one over the other, we call evolution and dissolution. Evolution under its simplest and most general as6 7 8

Wright (2000), 7. Wright (2000), 70 f. Cf. Brush (1970), 23 f.

360

Michael Hampe

pect is the integration of matter and concomitant dissipation of motion; while dissolution is the absorption of motion and concomitant disintegration of matter.9

Herbert Spencer’s books, which provided an evolutionary psychology, an evolutionary pedagogy and an evolutionary sociology within the framework of his “system of philosophy”, although they are characterised by similar empty phrases to those cited above, were widely distributed. In them Spencer wanted to replace the real, finally narratively structured history of nature and societies through a theory of developmental laws of the according systems, under which the real-historical events are then subsumed. That is, through the universalisation of Darwinian evolution, Spencer wanted to make explanations out of narration, using laws of development, just as Marx and Engels attempted in their adaption of Darwin in the field of the development of political structures.10 And it was exactly this intention that Wright opposed. If he draws an analogy between history and nature at all, then this occurs before the existence of meteorological simulations: one can only tell about the weather narratively, just as with the development of the world as a whole. At that time it was impossible to deduce this history from laws of development, i. e. to replace the narration with a deduction. And even if today, in the age of the so-called synthetic evolutionary theory after Mendel and Darwin, there is talk of universal laws of development, this does not mean that the history of animals, much less the genesis and extinction of specific animals, could be deduced from the principles of mutation and selection. What species is separated from which by an uncrossable sheet of ice, which dies out due to a meteorite impact, which species becomes competition for which other species by way of straits created by volcanic activity, allowing for migration; all this can only be reported in paleontological narration describing the contingent geological events relative to the principles of organic development. The genesis of the difference between hooded crows and rooks during glacier formation in the last ice age can only be told. This narration plausibilises, even explains the differences between these animals through a description, but does not deduce these differences by way of a mathematical inference. 9 Spencer (1971), 59, my italics. 10 Marx (1981), 7 – 11.

Explanation Through Description

361

4. Genesis in Peirce Let us consider the treatment of Darwinian evolutionary doctrine by another evolutionist across from whom Wright sat in his club and who drew not so trivial, yet just as general conclusions as Spencer from the theory of evolution: namely, by Charles Sanders Peirce. Peirce criticized Wright’s standpoint, demanding that philosophy should become much more evolutionary. By this he did not mean that it ought to produce yet more empty formulas, but that Spencer’s assumption that there were quasi-mechanical laws of evolution that directed the aggregation resp. dissipation of material and motion was still caught up in a pre-Darwinist conception of nature. Peirce wrote: Spencer is, in short no evolutionist; but only a semi-evolutionist… Through-going evolutionism is called for in philosophy … Now the only way to do that is to show in some way that law may have been a product of growth, of evolution. Then we must make some principle of growth more fundamental than any mechanical law. The suggestion to which this leads is obvious. It is that matter is mind which has come under the almost complete domination of habit. That at first only mind existed, a vast unpersonalized manifold of mind and before this mind had any habit it only existed in a germinal sense, for existence consists in regularity.11

If one interprets natural species as laws of reproduction, following Aristotle, as laws that ensure that out of a frog comes another frog, and out of a man another man, and not otherwise, then Darwin did not provide a narration of the origin of animals with his doctrine of evolution, but transformed the eternal, atemporal laws of propagation into historical laws.12 Since the synthetic theory of evolution there has been a nomological history of life, or a necessary history of the development of species, that has a stricter character than any narration. For both Plato and Aristotle the constancy of the species was the paradigm of nature’s regularity, aside from the geometrically conceivable heavenly movements (though the regularities of the heavenly bodies’ motions did not yet appear to them as natural). For the irregular accelerations and decelerations of bodies under the moon obeyed no laws of motion. Peirce is the first to take the Darwinian historising of the species really seriously as a historising of natural laws in general, generalising it to a conception of na-

11 Peirce (1890), my italics. 12 Gilson (1971).

362

Michael Hampe

ture according to which all laws have a history, even those of inorganic nature, including the laws of mechanics. Yet what has this historism to do with the demand in the above quote that matter is to be conceived as mind? The development of regularities can be observed from an internal perspective according to Peirce in every mind when it forms habits of perception and action. Furthermore, it is valid for the formation of habits that the probability of the realisation of a habitual perception or action is all the greater, the more frequently it has occurred in the past. Hence, the solidity of a habit can be interpreted as a sign of the age of its history, or: the more solid a habit is, the older the history of its formation is, and the more exceptions a habit still tolerates, the younger it is.13 Starting from this line of thought, Peirce develops a speculative history of regularities: the inorganic laws of nature, for instance, of matter, are very old habits of the world. They were formed at a quite early stage of the cosmos and are therefore realised today with a probability of 1. Their history has petrified and come to an end. The regularities of organic nature are also a product of nature’s long development, yet they are not realised with absolute necessity yet, for which reason one can conclude that they are “younger” than the inorganic laws. The regularities of the cultural world and of personal life are finally the most variable. They admit the most exceptions and are hence the youngest. Yet the entire natural process is one of intellectual development, of the appropriation of habit. The inorganic material with its iron necessities is nothing other than a mind petrified in old habits. Through the principle of the appropriation of habits this Peircean conception of nature is as closely connected to Lamarck as to Darwin (who himself believed in the heredity of habitually obtained structures during a phase of his theory’s development), as Peirce himself also recognises. But the discussions of The Origin of Species in the “metaphysical club” were likely the impetus for Peirce’s development of this universalised evolutionary theory of the cosmos – rather than the study of Lamarck’s Philosophie zoologique. Peirce begins the history of the universe with a chaos of sensible qualities and haphazard combinations of these. Over this random sea of sensations he sees an intellectual law at work, namely, the aforementioned law of the appropriation of habits, which fortifies itself and yields all regularities, from the physical laws to the cultural patterns. He writes: 13 Cf. Hampe (2006), 139 f.

Explanation Through Description

363

…if the laws of nature are results of evolution, this evolution must proceed according to some principle; and this principle will itself be of the nature of a law. But it must be such a law that it can evolve or develop itself. Not that if absolutely absent it would create itself perhaps, but such that it would strengthen itself, and looking back into the past we should be looking back [to] times in which its strength was less than any given strength, and so that the limit of the infinitely distant past it should vanish altogether. Then the problem was to imagine any kind of law or tendency which would thus have a tendency to strengthen itself. Evidently it must be a tendency toward generalization… Now the generalizing tendency is the great law of mind, the law of association, the law of habit taking… Hence I was led to the hypothesis that the laws of the universe have been formed under a universal tendency of all things toward generalization and habit-taking.14

Thus Peirce produces a hybrid of explanatory theory and narration in his genetic cosmology. On the one hand the development of the cosmos follows a regular principle, that of the appropriation of habit, while on the other it is ruled by haphazard chance, which determines which habits establish themselves and how quickly, and this chance can only be related by way of narration. The present state of the world becomes a historical document, like the physiognomy interpreted by Lichtenberg or the plate whose scratches are investigated, a document from which the past history can be read and reconstructed narratively. It was not biology alone, however, but also Peirce’s reception of the philosophy of German idealism, especially Hegel and Schelling, that helped his elaboration of this hybrid of a narrative-explanatory and evolutionary philosophy of nature along.15 The genetic vocabulary of German idealism oscillates as well back and forth between narration and deduction. Schelling had thus already used the term “evolution” in connection with his theory of animals in his System of Transcendental Idealism in 1800. He writes: The organisation (of animals, M.H.) is the succession hindered in its course and quasi-frozen… It becomes clear, then, from this deduction of life … that … that difference between living and non-living organisations … cannot take place. As the intelligence is to regard itself as active in the succession through the whole of organic nature, every organisation must have life in the broad sense of the word, i. e. an inner principle of motion, within itself. Life may well be more or less restricted; hence, the question: whence that difference, which is reduced to the previous: whence the graded hierarchy

14 Peirce (1992), my italics, 241. 15 Cf. on this Esposito (1980) and Hampe (1999).

364

Michael Hampe

in organic nature? This graded hierarchy of organisations, though, refers only to various moments of the evolution of the universe. 16

In this text Schelling presupposes a deducible hierarchy of animals, as it was common ever since Neo-Platonism, but historises it in an evolution of the universe which can in principle only be narratively related. The atemporal intelligible deducible hierarchy of beings is regarded in the real history by intelligence as a temporal sequence that must be narratively comprehended. Peirce undoubtedly knew this text. Yet while Schelling separates the development of mind from that of nature during this phase of his thought, the transcendental and natural philosophy running along parallel to one another as complementary systems, Peirce unites the development of nature and of mind in a panpsychistic evolutionary cosmology. In both Schelling and Hegel (in the Phenomenology of Spirit) as in Peirce’s cosmology it becomes clear that between deductive connections and narrations there must not always be a sharply defined boundary. Both deductions and narrations yield transitions. In a deduction one proceeds from one specific line to the next according to certain rules of inference or implication. Although this inferential procedure occurs in time, it does not depict any developmental process, but the temporal sequence of judgment conception is interpreted as a psychological phenomenon in which the atemporal inferential structure is reflected, but not constituted. Philosophers as Schelling, Hegel and Peirce, then, refer the sequence of inference units in a deduction to a temporal sequence of process stages in a course of development that must be narratively comprehended. The temporal development thus becomes the appearance of a deduction, and the narration the comprehension of inferential relations appearing in time. This thought remains relevant for 16 Translation of Schelling (1957), 161, italics M.H.: “Die Organisation [der Lebewesen, M.H.] ist die in ihrem Lauf gehemmte, und gleichsam erstarrte Sukzession… Es erhellt nun aber eben aus dieser Deduktion des Lebens…, dass … jener Unterschied zwischen belebten und unbelebten Organisationen … nicht stattfinden kann. Da die Intelligenz durch die ganze organische Natur sich selbst als tätig in der Sukzession anschauen soll, so muss auch jede Organisation im weiteren Sinne des Worts Leben, d. h. ein inneres Prinzip der Bewegung in sich selbst haben. Das Leben mag wohl mehr oder weniger eingeschränkt sein; die Frage also: woher jener Unterschied, reduziert sich auf die vorhergehende: woher die Stufenfolge in der organischen Natur? Diese Stufenfolge der Organisationen aber bezeichnet nur verschiedene Momente der Evolution des Universums” (Schelling (1957), 161, my italics).

Explanation Through Description

365

physical cosmology perhaps even today, inasmuch as it starts with certain equations, for example, Einsteinian field equations, inserting certain values into these, for instance, for the amount of matter in the universe, and then transforming these equations deductively. To the deductive transformations correspond historical stages of development in the material world that can be related narratively beyond the formulas and are even told by everyone today, with the big bang at the beginning, the formation of elementary particles in the middle, and the activity of the galaxies at the end. That narrations and deductions can fade into one another in a genetic vocabulary, though, does not mean that the friends of narration and description, such as Chauncey Wright, have no good reason to mistrust the universalities of deduction. Often, as in Schelling’s natural philosophy, deductions are connected with hierarchies. Deductions begin with general principles and move down to determinations of the individual. When these universal principles are regarded as what is actually real and valuable, in a Platonic manner, and that which is deduced from it is considered only a feeble reflection, a red light usually goes on for the friends of narration. For, though the narration may also have inferential structures and connect one experience with another in inferential connections, neither the inferential connections themselves nor the individual experiences are placed in a hierarchy of general and specific, or of more or less real or valuable. Narrations have beginnings, transitions and sometimes an end. Yet these do not correspond to the beginning of a deduction, its middle terms and the demonstrandum. The beginning of a narration is not a fundamental equation, no fundamental principle, nor a most universal proposition that expresses eternal connections, but it can be a description of any fact. If one moves within the Lichtenbergian image of a causally and semiotically universally intraconnected world, the narration can start at any random point in the world, in order to move from it to another random point, if it only finds the right transitions, for which an “enlargement” of the relations or a deceleration of the processes may be necessary. Tales lead the attention in a certain order through a complex of facts so as to yield a plausibility in the mind of the one led. Deductions place assertions in such a relation to one another that whoever holds p to be true must also hold q to be true. If the truth of q is dependent upon that of p, then p is the more valuable truth in this inference.

366

Michael Hampe

5. Description and Explanation in Mach Ernst Mach also referred to this concept of the generation of understanding through description when he contributed to the debate on the relation between explanation and description in physics sparked by Gustav Kirchhoff. Mach asserts in his article “Beschreibung und Erklärung” that we explain, clarify, familiarise a seemingly strange fact by realising that it is based upon the simultaneous parallel existence of similar or dissimilar facts already known.17 Something is strange or incomprehensible to us, because it contradicts the instinctive expectations set at another scale.18 In this case the explanation consists, according to Mach, in the correct leading of attention to the special, decisive matters.19 This directing of the attention leads to nothing arcane or higher, but organises familiar elements of experience so that the passage through them leads to a new view of the formerly strange phenomenon. The relation of part to whole plays a role in this context. Again Mach: In all these examples of explanations … the common feature is that only familiar parts and facets of facts that seem strange are presented to us. It is not only the logical retracing of a proposition to one or several others, but also the p s y c h o l o g i c a l replacement of alien perceptive and imaginative images by common and familiar ones; it is essentially the elimination of a p s y c h o l o g i c a l disturbance that is at issue here.20

Just as a certain combination of familiar facts can yield unexpected phenomena and properties, so can it also cause astonishment. The physical description eliminates this astonishment and thus explains the formerly astonishing phenomenon. 17 “[F]remd anmutende Tatsache erklären, aufklären, uns vertrauter machen, indem wir durchschauen, dass diese auf dem Zugleichbestehen, Nebeneinanderbestehen schon bekannter gleichartiger oder ungleichartiger Tatsachen beruht” (Mach (1910), 413). 18 “instinktive auf andere Masse eingestellten Erwartungen” (Mach (1910), 414). 19 “in der richtigen Leitung der Aufmerksamkeit auf die besonderen massgebenden Verhältnisse” (Mach (1910), 414). 20 Translation of Mach (1910), 420: “Bei all diesen Beispielen von Erklärungen…kommt es immer darauf hinaus, dass an uns fremd anmutenden Tatsachen durchaus nur Teile und Seiten aufgewiesen werden, die uns bekannt und vertraut sind. Es ist nicht nur die logische Zurückführung eines Satzes auf einen oder mehrere andere, sondern auch der p s y c h o l o g i s c h e Ersatz fremdartiger Wahrnehmungs- und Vorstellungsbilder durch geläufige und vertraute, es ist wesentlich die Beseitigung einer p s y c h o l o g i s c h e n Beunruhigung, um die es sich hier handelt”.

Explanation Through Description

367

What Mach says here about description can also be said about narration: It leads the attention through a simultaneous existence, a sequence of facts, and by leading it thus it renders the formerly incomprehensible understandable or the formerly inexplicable explicable. Thus, narrations can also be explanatory, if they are successful in readjusting the recipients’ attention.21

References Brush (1970): Stephen G. Brush (ed.), Kinetische Theorie. Einfhrung und Originaltexte. Vol. II: Irreversible Prozesse., Braunschweig. Esposito (1980): Joseph L. Esposito, Evolutionary Metaphysics. The Development of Peirce’s Theory of Categories, Ohio. Gilson (1971): Étienne Gilson, D’Aristote Darwin et retour, Paris. Hampe (1999): Michael Hampe, “Komplementarität und Konkordanz von Natur und Erkenntnis. Anmerkungen zu Schelling und Peirce”, in: H. Eidam/F. Hermenau/D. Stederoth (eds.), Kritik und Praxis, Zur Problematik menschlicher Emanzipation, Lüneburg, 96 – 106. Hampe (2006): Michael Hampe, “Naturgesetz, Gewohnheit und Geschichte. Historisierung der Natur – Naturalisierung der Geschichte. Peirces Prozesstheorie”, in: M. Hampe, Erkenntnis und Praxis. Zur Philosophie des Pragmatismus, Frankfurt/Main, 124 – 154. Lichtenberg (1983): Georg Christoph Lichtenberg, Schriften und Briefe, Frankfurt/Main. Mach (1910): Ernst Mach, Populrwissenschaftliche Vorlesungen, Leipzig. Marx (1981): Karl Marx, “Vorwort zu: Zur Kritik der politischen Ökonomie”, in: K. Marx/F. Engels, Werke. Vol. 13, Berlin. Menand (2001): Louis Menand, The Metaphysical Club: A Story of Ideas in America, New York. Peirce (1890): Charles Sanders Peirce, “Evolution”, MS 954; “The Architecture of Theories”, MS 956. Peirce (1992): Charles Sanders Peirce, Reasoning and the Logic of Things. The Cambridge Conferences Lectures of 1898, K. L. Ketner (ed.), Cambridge/MA. Schelling (1957): Friedrich Wilhelm Joseph Schelling, System des transzendentalen Idealismus, Hamburg. Spencer (1971): Herbert Spencer, Structure, Function and Evolution, London. Wright (1958): Chauncey Wright, The Philosophical Writings of Chauncey Wright. Representative Selections, Edward H. Madden (ed.), New York. Wright (2000): Chauncey Wright, The Evolutionary Philosophy Of Chauncey Wright, Vol. 1: Philosophical Discussions, F. X. Ryan (ed.), Bristol.

21 Translated from German by Alan Duncan.

Knowledge of Ignorance: On the Problem of the Development and the Assessment of Technology Hans Poser The very concept of research presupposes conscious ignorance about the object of research at the outset; otherwise there is nothing to research.1 The study of ignorance almost inevitably confronts us with prescriptive questions, that is, how people “should” deal with ignorance.2

In his famous book What engineers know and how they know it Walter G. Vincenti3 analyzes the way of problem solving in engineering design as an epistemological problem. But his lucid undertaking, even in describing the steps, ignores that all problem solving starts from ignorance in the sense of non-knowledge or of rational ignorance, or just knowledge about the limits of knowledge: There would be no problem at all if we would have already the necessary knowledge (including know how etc.). Therefore, we need a further epistemological step back, namely to look at the kind of rational ignorance or non-knowledge, from which the technological problem originates. Ignorance, throughout this paper, is meant as lack of knowledge in a neutral and purely descriptive sense. If ignorance dominates engineering – how can we survive, since our technology turns our culture and our lifeworld inside-out in unforeseen ways? How can we live up to Hans Jonas’ new technological imperative, the principle of responsibility: “Act so that the effects of your action are compatible with the permanence of genuine human life”?4 Acting this way requires not just knowledge but knowledge of knowledge’s limits, 1

2 3 4

Smithson (2008), 218. – A German version of this paper appered as “Wissen des Nichtwissens: Zum Problem der Technikentwicklung und -folgenabschätzung”, in: N. Janich/A. Nordmann/L. Schebek (eds.) Nichtwissenskommunikation in den Wissenschaften, Frankfurt/Main 2008. A small part of it will be published as “The Ignorance of Engineers and How They Know It”, in: D. E. Goldberg/D. Michelfelder/N. McCathy (eds.), Philosophy and Engineering. Reflections on Practice, Principles and Process, Berlin 2011. Smithson (2008), 225. Vincenti (1990). Jonas (1984), 11.

370

Hans Poser

that is, a knowledge of ignorance – and more than that, an estimation of this ignorance and a way of dealing with it. In the following I examine the preconditions involved here. However, this requires, as a first step, a certain clarification of what is meant by knowledge, ignorance, and knowledge of ignorance. The second step is devoted to the question how this relates concretely to technical knowledge, in order to take up the problem of ignorance in technology in a third step. Finally the results of this investigation are to be placed in a fundamentally different perspective, namely a modal perspective.

1. Knowledge and Ignorance Ignorance, just like knowledge, is profoundly human and thus at the same time subject-related. Knowledge and ignorance belong not to a book but to its author, not to a computer but to its designer, software creator and user. Hence the knowledge of our ignorance is an age old subject of human reflection and relates to all forms of knowledge, all of which, as we will see, enter into technological knowledge, without exception: K1 Knowing that as the knowledge of facts, K2 Knowing why as theoretical and causal knowledge, K3 Knowing how as practical action knowledge, K4 Knowing what for as normative knowledge of value. Now ignorance is the negation of knowledge; but logically speaking it would entirely miss the point to see this as a complementary set to the set of knowledge, since the content of ignorance is very much specified in a characteristic way. Hence ignorance is a subject of human reflection. Likewise, theories of knowledge have been a part of epistemology from Plato and Aristotle to today – but a corresponding theory of ignorance is missing. Hence we will see the necessity of distinguishing at least three forms of ignorance in all four areas (knowledge of circumstance, knowledge of cause, know-how and valuation): – that which lies beyond the fundamental limits of the knowable, – that which lies beyond the currently general state of knowledge or ability, and – that which lies beyond an individual state of knowledge or ability (individual lack of knowledge or ability).

Knowledge of Ignorance

371

In the course of the past two decades ignorance has been the subject of numerous investigations, which in America usually start from the work of Michael Smithson5 and have led to an interdisciplinary anthology by Robert N. Proctor and Londa Schiebinger6 that proposes Agnotology as a new area of research. Economists, sociologists and psychologists are studying phenomena of ignorance. Yet the Anglo-American investigations differ from those in Germany, for example, in a decisive point, since the concept of “ignorance” has both the neutral meaning of the German term ‘Nichtwissen’ (non-knowledge) and at the same time a negative evaluative significance, and almost all investigations in the English-language tradition refer to this second sense of the term in treating ignorance as knowledge that is manipulated7 or suppressed or overlooked and yet nonetheless available. Moreover, in regards to technology, even the Agnotology volume almost entirely ignores actual ignorance as missing knowledge. And even where the German literature takes up the problem of ignorance concerning technology, it only does so under the aspect of uncertainty.8 The ignorance I intend to clarify here takes the form: I have a problem but do not know the solution! This should recall Popper’s “All Life is Problem Solving” and help us to see the fundamental significance of precisely this concept of ignorance. The background to all this is humanity’s continuous struggle with contingency: our lifeworld is full of uncertainty, imponderability and unforeseen occurrences; and as humans we try to overcome this with the sciences, which oppose this contingency with necessity – starting with Plato and the eternal mathematical truth (a priori necessity), followed by Galileo’s ideal of the book of nature written in numbers and investigated by the empirical sciences (physical necessity); and in a different step by the establishment of rules for action within the society in order to found relatively stable social structures and to stabilize society with laws and punishments such that it is possible to expect a certain type of interpersonal conduct and to make predictions about actions to a certain extent (social and ethical necessity). But one of the most important elements in managing contingency is technology: we expect it to function properly, i. e. to work in a foreseeable manner (technolog5 6 7 8

Smithson (1985); Smithson (1989); Smithson (1990); Smithson (1993); Smithson (2008). Proctor/Schiebinger (2008). Magnus (2008). E.g. Banse (2005); Gamm (2005).

372

Hans Poser

ical necessity), whether it’s a knife or a car or an entire industrial plant.9 Yet we know that even technology can fail in many cases. Hence the problem of ignorance concerning technology is of the utmost importance, and not just for our understanding of technology itself. Every instance of ignorance asks to be brought to our awareness as ignorance. But this means that it can be dressed up in the form of a question – such as “Do you know such and such?” Hence the content of our ignorance comes into view together with the form of the question that characterizes the content of the inquiry as an area of ignorance. It is not sufficient to conceive ignorance as the not-yet-familiar. An epistemology of ignorance such as Nancy Tuna10 calls for and outlines does not refer to mere gaps in our knowledge or the simple negation of knowledge, but rather to a relatively clear view of these gaps and logically speaking to a privation. For this reason we cannot be satisfied with an analysis of social circumstances as the causes of ignorance, since doubt, trust and unclarity, for example, have to be taken in a much more primary sense as cognitive concepts; however, it could be necessary to supplement the approach of classical epistemology with elements of sociology, as we see in the social epistemology of Goldman11 or Kassavine12, since the criteria for knowledge may have changed through the course of history. The guiding idea of this paper is to distinguish between at least five different problematic fields of an engineer’s ignorance that should at the same time reveal the larger significance of ignorance: – Ignorance serves as the starting point for all invention and development by marking a problem. – This can be formulated as a question, which describes a communicable content. – Solving the problem often requires creativity; but this precludes any sort of prediction – a hard case of ignorance. – R&D (Research and Development) departments need communication about ignorance concerning these problems, otherwise collaboration would be impossible.

9 10 11 12

Poser (2009). Tuna (2004). Goldman (1999). Kassavine (2003).

Knowledge of Ignorance

373

Unknown possible consequences of technology – hence hard cases of ignorance – have to be assessed using methods of estimating the consequences of technology. Thus an engineer’s ignorance is characterized by a problem as the cognitive starting point: The problem is the classic situation of ignorance par excellence. The problem corresponds to a question, which calls for a solution that is not available. This already makes it clear that ignorance has a content that is, moreover, communicable, in that it can be formulated as a question. For this reason a communication of ignorance is possible. But this shows why ignorance is given first place of priority in the epistemological analysis of how engineers proceed; hence Robert Proctor calls this form a “native state”, a state that is completely distinct from other forms of ignorance such as “a lost realm, or selective choice”, the neglect of other possibilities, or even “ignorance as strategic ploy”, keeping knowledge as a secret13, though he then goes on to a sociological investigation. This leads us away from the sort of epistemological investigation intended here, which requires that we begin to approach the subject by inquiring into the conditions of the possibility of knowledge – which in this particular case means: what are the conditions that allow us to start from ignorance and then succeed in characterizing the problem and moving to the corresponding question, the communication of this question, and finally to a solution. Now, when we speak of ignorance this presupposes that we know what knowledge is. There are so many different answers to this that it is difficult to say how we should understand the term. Here I will take up four different formulations: – Factually present agreement as a stock of opinion guiding action in a certain cultural circle, i. e. communis opinio that is considered knowledge. Many sociologists take this as their foundation14, since social events are based on this; each society has a stock of opinion that is considered truth and that guides action and is preserved and perpetuated wherever it has proven itself in our practical dealings with the world. This is also the state-of-the-art position in jurisprudence. However, this cannot function as a criterion of truth, since what it means for opinion to have proven itself is a valuation of that partic–

13 Proctor (2008), 4 – 10. 14 See for example Smithson (1985); cf. Luhmann (1992) and Japp (1999).

374

Hans Poser

ular society. Hence it is no surprise that whole books have been written about the knowledge society that do not even try to first outline what they mean by knowledge. – Demanding intersubjective agreement would be the better version, philosophically; but what does this agreement base itself on if it is supposed to obtain for all thinking subjects? – If we follow Plato in seeing knowledge as true opinion with justification, then knowledge is something that relates to a proposition we are compelled to recognize as true; we then have to explain what we mean by the predicate true, what the recognition of its truth consists in and how the two are connected in terms of a justification that is intersubjectively valid. The objections known in contemporary analytic philosophy as Gettier problems do not make this understanding of knowledge obsolete, since their proponents, as clandestine Platonists, all seek justification for the objections they hold to be true. – By contrast, Popper’s criticism of the propositions of the sciences in particular has led to a more careful understanding of knowledge as methodically justified statements with the status of hypotheses. Thomas Kuhn ultimately made it clear that the criteria for a justification depend on history and culture. We can derive an initial formulation of ignorance from this, since firstly all these forms of knowledge relate to content and secondly knowledge here can only be seen as something conscious. Thus the negation of knowledge itself can only be formulated in terms of consciousness and content: Ignorance is a reflective state of consciousness related to a specific content. Thus ignorance is meta-knowledge15, expressed in a predicate of the meta-language like ‘true’ or ‘knowledge’: it expresses the fact that we have no answer to the corresponding question: Ignorance exists in conscious form as knowledge of a problem, expressed in a question. Now we need to more precisely characterize the question of ignorance in particular and with it the content of ignorance. The sparse remarks on the concept of truth above are enough to show why there has always been a skeptical counter-movement emphasizing the significance of ig15 Cf. Smithson (2008), 210.

Knowledge of Ignorance

375

norance or even recognizing exclusively ignorance and nothing else. Not to speak of Socrates and the Sophists, and those skeptics in the Platonic Academy who left no writings because there cannot be any certainty, a skeptical tradition has always worked against the modern tradition of optimism and Enlightenment – from Cartesian methodological doubt in the service of securing knowledge to Hume’s moderate skepticism, which demonstrated the unobservability of causality, to Kant’s delimitation of pure reason by the conditions of possibility inherent in the subject of knowledge, which ruled out more than just proofs of God’s existence, all the way to Emil du Bois-Reymonds verdict of Ignoramus et Ignorabimus for what he saw as the irresolvable puzzles of the world, since for example we will never be able to know how body and soul interact.16 All of these sought a systematic answer to the question of the basic limits of our knowledge, from empirical knowledge to theology. Or to put it another way: they sought to delimit that which is principally knowable from indissoluble ignorance – though always in reference to particular questions or subjects. None of this stops us from assuming at the same time that the propositions methodically worked out and justified by the sciences, regardless of whether they are essentially subject to these limitations and have an irrevocable status of hypotheses, nonetheless preserve and increase that which is the most securely established and proven and pass it on.17 From the perspective of ignorance, all the sciences are concerned to recognize ignorance as a problem that, so long as it does not fall under the category of the essentially unknowable, calls for nothing but a more in-depth search for secure knowledge with the status of hypotheses: The knowledge of ignorance becomes the basis of the epistemic dynamic. If we look now for forms of ignorance, our first idea might be to borrow from the list of four forms of knowledge above and distinguish between - Unawareness concerning facts, - nescience concerning theoretical reasons, - disability to achieve something, - blindness or non-knowledge concerning norms and values. However, this would lead us badly astray, since the sort of ignorance we are investigating here is a conscious state that is fully aware of ignorance and moreover falls on the beneficial side of Willem H. Vanderburg‘s dis16 Du Bois-Reymond (1872). 17 Poser (2007a).

376

Hans Poser

tinction between “useful” and “harmful” ignorance.18 In fact Smithson19 assumes a distinction that – quotation somewhat shortened and paraphrased – recognizes the following kinds of ignorance: S 1 Ignorance as encountered in the external [non-social] world… These accounts make strong claims about meta-knowledge and explain ignorance in exogenous (and usually non-social) terms. S 2 Ignorance as emergent, constructed, and imposed … by agents … and at least partly socially constructed. S 3 Managing under ignorance:… How people think and act in uncertain environments. S 4 Managing ignorance: How people think about ignorance … and how they act on it. We can leave aside forms S2 and S3, since they largely relate to societal questions about the manipulation of knowledge. They belong to the well-known scheme, which Smithson originally developed20, where he starts a taxonomy of ignorance as either an error or as irrelevance – a disjunction that is completely irrelevant for our purposes. As it concerns technology, the relevant forms are S1 as the starting point and S4 for estimating the consequences of technology. – Back to the discrimination S1 to S4: Smithson ascribes the form S1 to the empirical sciences – but doubtless it also holds for the ignorance of the engineer and for industry R&D departments. Yet due to Smithson21 the S1-kind of ignorance is not “socially constructed” and independent from “sociocultural origins” of ignorance; this might be the case for sciences, since their problems are originated by the sciences themselves. But at least for engineering, the situation is a completely different one from the very beginning, because its aims stem from needs of individuals or of the society. And as all intentions and problem views depend on a cultural background, it will be necessary to accept that even the first account has partly sociocultural origins. Therefore we need to bear in mind the difference between the natural sciences and the technological sciences, Mario Bunge has highlighted: The former seek most general laws, while the latter seek better technological solutions, i. e. “better ends”.22 This has consequences for our understanding of ignorance, 18 19 20 21 22

Vanderburg (2002), 90. Smithson (2008), 214 ff. Smithson (1989), 9 and Smithson (1990), 211. Smithson (2008), 209. Poser (1998); cf. Bunge (1974), 20.

Knowledge of Ignorance

377

since in engineering S1-ignorance, too, is socially constructed: The engineer faces the entirely different situation that his or her aims derive from individual desires or societal demands: S1 as well as S4 inherently refers to society. Now, Smithson’s S4-account is important as a first step. Vanderburg23 speaks in this connection of two kinds of ignorance – the first one is “related to the fact that, as specialists, we know everything which is to know”, whereas what we know is imbedded in a second kind of ignorance, since “we forget that every human knowledge is relative to a vantage point determined by our professional experience, formal education, life experience, convictions, values, and, last but not least, the culture of our society.” This makes it clear why there is in principle no knowledge without this second kind of ignorance emphasized by Vanderburg – an ignorance that Michael Polanyi24 called tacit knowledge –, which can be turned into a useful one, “if its existence is clearly recognized”25. Thus it will be necessary to include Smithson’s fourth account and by this an element of social construction – for example, when it concerns risks in an area where scientific answers are either not yet possible or fundamentally impossible – one might think of complex structures. In these cases the questions of what our ignorance consist in depend primarily on values, expectations and fears, which are of sociocultural origins and thus part of our life world and our life experience.

2. Ignorance as Knowledge of the Insurmountable Limitations of Knowledge Until the beginning of the 20th century the classical natural sciences were very much of the opinion that within the empirical realm everything could, in principle, be modeled mathematically and captured in systems of axioms. This would make ignorance just a temporary phenomenon, and all ignorance would eventually be dispelled, at least in principle, by a unified science in the sense envisioned by the Vienna Circle. However, this view has proven to be untenable in several regards. A look at the history of the sciences and technology shows instead 23 Vanderburg (2002), 91. 24 Polanyi (1966). 25 Vanderburg (2002), 91.

378

Hans Poser

that through the course of the centuries there have been questions where something held to be a temporary lack of knowledge was in fact insurmountable ignorance, “ignorabimus” in du Bois-Reymond’s sense, even if this does not pertain to his points. A specifically described realm of unknown content has been shown to be principally unknowable not just in those areas previously thought to be empirically accessible but also in geometry and mathematics. For example, the question of how the 5th axiom of Euclidean geometry, the parallel postulate, can be derived from the first four, proved to be an ignoramus, which ultimately led to the proof of independence and the non-Euclidean geometries. The same could be said for the search for perpetual motion, until it could be shown to be impossible due to the law of conservation of energy. In the debate on the foundations of mathematics in the early 1930s Kurt Gödel revealed an unexpected and profound limitation of what can be known in the formal sciences, since the attempt to seek for a foundation by a set of axioms fails demonstrably even for second-order predicate logic and thus for classical mathematics. Hence: ignorabimus. Around the same time physics ran into a principled limit of the unknowable in the empirical realm as well with Heisenberg’s Uncertainty Theorem and even earlier with Einstein’s theory of relativity and the light cone as the limit of experience. Thus Einstein and Heisenberg revealed limits not just of physics but of human experience, since we are just as unable to penetrate within the relation of uncertainty as we are beyond the light cone. Now all of this seems irrelevant to the case of technology, since there are no non-Euclidean bridges and our mathematical tools – PCs – are finite and thus remain unaffected by the problems of providing a foundation for mathematics; technological processes do not occur within the relation of uncertainty or faster than the speed of light. But this appearance of irrelevance is false, since we have to consider that Gödel shows that there can be no complete axiomatization of the technological sciences in the sense of a universal Ars inveniendi. There have always been attempts to develop 2nd- and 3rd -order perpetua mobilia; nanotechnology touches upon quantum effects; and while Einstein’s theory of relativity precludes science-fiction dreams of conveying information faster than the speed of light, there have been attempts again and again to overcome this limitation by means of the Einstein-PodolskyRosen paradox. Engineers are also called upon to develop experimental possibilities – hence, artifacts – to observe all of these phenomena near the limits of the ignorabimus.

Knowledge of Ignorance

379

One of the most relevant restrictions on knowledge leading to an inevitable ignorabimus consists in the mathematical properties of complex systems: even in the case of deterministic chaos – hence assuming that processes are describable by systems of non-linear differential equations – it is impossible to derive a closed solution function, as Henri Poincaré had already recognized; we can only reach approximate solutions. But these depend so much on the smallest changes in the starting and framing conditions that prognoses are absolutely impossible even for a brief span of time. This is all the more serious if we look at the dissipative structures examined by Ilya Prigogine, which, as systems outside of thermodynamic equilibrium with an exchange of energy, material and information in a field of possible bifurcations, are characterized by unpredictable structural changes and the formation of new orders when they are forced out of a quasi-stable structure (strange attractor).26 Here prognoses are fundamentally impossible; they can only be made under severe reductions of complexity, whereby it is entirely open to what extent the weighting of parameters in fact allows for these simplifications. Whether it makes sense in this context to argue for an “applied ignorance taken as a rational renunciation of information”27 would require criteria of rationality that can hardly be identified at all. Such complex models have also been used to describe the historical development of technology – in the hopes of using them prognostically; yet the widely diverging statements about the development of climate, for example, show how dependent the results are on the particular model chosen. Hence this presents another ignorabimus. Dissipative structures are used today not just in non-equilibrium thermodynamics, but far beyond this, for example in modeling the dynamics of social systems or market events or the biota in autopoietic systems, hence concerning objects that are of the utmost importance in today’s biotechnology. The same can be said for communication technologies and their networks, not just in their unpredictable development but in particular for the means used to overcome this problem, namely simulation models: they can only provide prognoses under severe limitations. Often the objects to be studied – the development of technology and its influence on the environment and the society – resist even being modeled due to their complexity. This is a new element among the forms of ignorance concerning the development of technology, 26 Poser (2007b). 27 Peppel (1994), 60; cf. Gail (1999/2000), and Roland (2002/03).

380

Hans Poser

since it would also have to include norms and values and their change through time and hence the changing assessment of possible technological solutions and their repercussions for society and the environment. Gerhard Gamm very succinctly described the moral of all this: the scientification of knowledge and its technologification have led to a “reciprocal escalation of knowledge and ignorance” resulting in a “dramatic accumulation of the knowledge of ignorance”; ignorance becomes the “continually self-regenerating shadow of every increase in knowledge”.28 However, this forces us to include cultural traditions, and thus historicity along with them, in treating the epistemological side of the engineer’s ignorance, as well as new forms of ignorance management concerning technology. Yet wherever we are confronted with complex systems we face more than just a principled and theory-related uncertainty; rather this also requires some practical and action-orienting way of dealing with uncertainty as the insurmountable limitations of the sciences. Thus ignorance in this form has become a positively essential problem. In the meantime there has been a growing literature on ignorance management within the realm of the economic sciences. Sugar-coating titles like The method of controlling chaos (Die Methode, das Chaos zu beherrschen) – in this case referring to mathematical market models in economics – mislead us into thinking that there are still procedures for dealing with the fundamentally unknowable just as we deal with some temporary individual or social uncertainty that can be overcome by methodically acquiring knowledge and reflecting on it. This set of issues should be kept in mind when we ask how to deal with ignorance in assessing technology and estimating its consequences. And there is one final complication: the background of this last point brings us to an additional ignorabimus, namely the fact that there is no absolute or rational foundation of ethics, norms and values. The 18th century believed that it was possible to find such an eternal basis of morality; yet though we are convinced that moral rules are necessary, and though we are just as convinced that they must satisfy the criteria of universalizability given by Kant’s categorical imperative, and although principles of justice in accordance with John Rawls’ procedure are indispensable, we have to concede that a universal ethics or a universal theory of norms and values is beyond our reach due to the dependence of norms and values on history and culture. This type of ignorabimus is 28 Gamm (2005), 22 f.

Knowledge of Ignorance

381

not just a challenge for philosophers – it essentially concerns all technology as well. For example, we might recall Bunge’s thesis that technology is oriented towards better solutions; there can be no theory specifying in general what better solutions are. So we have to realize that there are inevitable types of ignorance as ignorabimus, ranging from normal, i. e. logical and mathematical limitations to physics and biology all the way to the problems of justification in ethics. At the same time the history of the sciences shows us that it has always been possible to communicate very precisely about the forms of ignorance, since there has been a precise conceptual framework that allowed for an exact statement of the problem and formulation of a question. Thus the next step has to be to identify this framework in regards to technology.

3. Technology and Knowledge We should briefly explain what we mean by the term “technology”. I am concerned here with real technology as present in artifacts and artificial processes. Klaus Tuchel’s definition can serve to outline this: Technology is the concept for all objects and procedures made to satisfy individual or societal needs by creative construction that serve certain purposes through definable functions and as a whole change the form of our world.29

The talk of changing “the form of our world” refers to the formative influence of technology on our environment and our society. This definition makes it clearer which forms of knowledge are presupposed, since when an engineer’s ignorance declares something to be a problem that needs to be solved, this involves a search for means to reach an aim. Thus the engineer needs a technical knowledge in accordance with a specification of K1 to K4, namely: TK1 a knowledge of factual circumstances, TK2 a knowledge of means for an aim in the sense of fulfilling a function (theoretical operative knowledge), TK3 a knowledge of how such means can be attained and used (practical operative knowledge),

29 Tuchel (1967), 24.

382

Hans Poser

TK4a a knowledge of the values behind the needs (normative operative knowledge), and TK4b a knowledge of modifications of aims in light of values, in case this is necessary (practical and theoretical normative knowledge). This is problem-solving knowledge. Yet the beginning of every technological development involves a form of knowledge that makes ignorance per se comprehensible; this form of knowledge consists in setting the solving of a problem as an aim, which is in turn founded on a knowledge of value. Ignorance becomes the starting point of the methodical and systematic search for a suitable means for a given aim. But the type of ignorance and the direction in which we should start reducing it is clear in each case; and this might be the reason why so far the whole process has always been seen exclusively in light of the acquisition of knowledge rather than in terms of ignorance. But now the problem-solving knowledge we just differentiated above comes out in this perspective as well: TK1 corresponds to K1, the knowledge of factual circumstance; it is relevant at every point, since otherwise technology would lose its connection to reality. The form TK2 is causal knowledge, but it goes beyond K2 in relating to functions and thus at the same time to some utility. The form T3 is an ability that corresponds to K3 but in reference to the specific situation, whereas the forms TK4a and 4b, corresponding to K4, belong to a culturally specific horizon of values; and here TK4b represents a particularly important form, namely the knowledge of how we should deal with values and aims in light of our needs when our original understanding of our aims cannot be realized. All of these forms of knowledge also inform the technological sciences, even if the latter understand them as propositionally as possible and develop them methodically and systematically: these sciences must inevitably involve these forms of knowledge, since otherwise they would lose their connection to real technology. Now the forms of knowledge relevant to technology can be further differentiated. For example, Günter Ropohl distinguishes between the knowledge of natural sciences, technological knowledge of laws, structural knowledge of rules, functional knowledge of rules, technological ability, eco-socio-technological knowledge of systems and socio-eco-

Knowledge of Ignorance

383

nomic knowledge.30 Armin Grunwald, in contrast, drawing on planning theory, distinguishes between operative knowledge (subdivided into technical knowledge of means-ends relations, knowledge of suitability and knowledge of side-effects) and contextual knowledge concerning the situation, the markets, the competition or the consumer acceptance.31 However, these views can be fitted into the four above-mentioned forms, without creating any gaps for our guiding question about ignorance. The distinction between explicit and implicit knowledge that came into prevalence with Michael Polanyi32 can also be passed over here, since at best it amounts to the distinction between conscious and unconscious stores of knowledge and not to the assumption that the latter concerns a form of ignorance.33 Since as we will see there is a kind of ignorance corresponding to all four forms, we should describe them a little more closely. For example, the area TK2 does not simply describe the knowledge of natural science, such as found in hypotheses about natural laws, but something fundamentally different, since neither ends nor means nor functions are observable – they are all based on the interpretation of real or possible states of affairs and processes as ‘means’ for ‘ends’ in the sense of the satisfaction of a ‘function’ with a view to an aim: hence all of these categories taken from operative knowledge depend on the intention of reaching an aim. In this area ignorance as a problem calls for expanding knowledge, guided by the question posed by the problem, by recombining the technological knowledge already available towards new creative solutions. This is not as simple as it might seem, since this poses a difficult epistemological problem: how can I know what I’m looking for if my starting point is the knowledge of my ignorance and thus a problem to be solved? And further, how can we get from a problem to an aim as an interpretation of a possible state of affairs, and from there to a means as an interpretation of a function that in turn depends upon an interpretation? Thus ignorance, seen epistemologically, presupposes two elements: (i) a structuring by orientating the question around an intended aim, and

30 31 32 33

Ropohl (2004), 42. Grunwald (2004), 51. Polanyi (1966). For an overview of this see Irrgang (2004).

384

Hans Poser

(ii) the cognitive ability to develop heuristic methods for the solution and/or creatively think up solutions that were previously completely unknown. The TK3 epistemic area involves know-how. Here the theoretical knowledge of a means presupposes an acquired practical operative knowledge – moreover, this must be connected with a knowledge of the availability or obtainability of the means. In fact this is another deep-rooted problem, even if engineers would not see it that way; but even the trickiest technological idea would be nonsense if we were not in a position to realize it. Thus realizability is a conditio sine qua non from the outset of every technological development. But unlike in TK2, it must be possible to acquire this know-how, hence to learn how to overcome this practical ignorance. It is usually an individual incapacity that is to be overcome. But the independent development of a new application of a new procedure is also conceivable – hence an approach involving creativity that is also a kind of learning. Yet learning is itself a classical epistemological problem since Plato, who argued that learning is “nothing but recollection” of what is already present in the soul.34 In the tradition of philosophy of technology this led Friedrich Dessauer to the very platonistic view that all technological solutions are a part of a “fourth realm” of ideas.35 Today no-one would accept this metaphysical thesis as a solution to the epistemological problem of learning and the use of technological creativity to overcome ignorance – but it shows us that in this case of ignorance we are making a significant anthropological presupposition: humans can learn and are able to create entirely new things. The area of knowledge that incorporates the others is TK4, since individual or social and always culturally specific needs, corresponding to equally culturally specific values, lead to the aims in the second area of knowledge, that in their realization determine the third area of knowledge – for example, in view of the need to obtain the means or to learn a certain know-how. Only this dominant fourth area of knowledge characterizes something as a problem to be solved based on the assessment of a given situation as insufficient. It has gained greatly in significance in the past decades as it became clearer how complex the field of values in technology is – values that to some extent stand in great ten34 Plato, Phaedo, 72e. 35 Dessauer (1956), 155.

Knowledge of Ignorance

385

sion to one another, such as profitability and safety. All of these values and the corresponding norms depend on culture and history. In addition epistemic problems get interwoven with normative problems, as we can see with all the efforts to estimate the consequences of technology not just in terms of the possible effects of a new technology but also in terms of its influence on social structures and newly developing notions of value, including an estimation of all steps and consequences. Here part of the content of ignorance includes more than just the knowledge of what is unknown – it also demands and presupposes a pronounced knowledge of value; otherwise the aim-oriented question connected with this type of technology-related ignorance would be impossible. The case of TK4 is very significant for our problem, since it would oversimplify things to assume that our ignorance completely determines the aim. This could be the case when we face a clear task – but normally this is not so; rather, the problem and the corresponding question describe an aim and a direction only in conjunction with the values ascribed to the aim. Therefore: Despite its aim orientation, ignorance has an open structure. The knowledge we presuppose requires a knowledge of hierarchies of value, since it might be necessary to replace a particular value with a more general value or a different value of equal rank. This occurs with a view to the needs that are to be fulfilled, and it can even be necessary to replace one aim by a completely different one that fulfills a comparable function relative to our needs. This is all very familiar to us from practical syllogisms as a scheme for explaining actions, since there are always countless possible means to reach an intended aim. This openness holds not just of means and aims but also for the underlying values: Ignorance is characterized by an aim orientation that is open for value-related transformation. Thus the areas of technological knowledge can also be correlated to the Aristotelian practical syllogism, since our knowledge of value sets the normative premise “Person P wants to achieve A”; the so-called cognitive premise “To achieve A, one has to do B” concerns the theory- and practice-oriented areas of knowledge. So it is not surprising that Georg Henrik von Wright notes that sometimes one needs more than two premises for a practical syllogism.36 36 Von Wright (1963) and (1972).

386

Hans Poser

4. Epistemological Conditions of Technological Knowledge and Ignorance If we look at all four cases of technological knowledge and inquire not just into the presupposed knowledge that makes up the content of the engineer’s ignorance, but also into the epistemological conditions of possibility, we find a deeper level of presupposition. It is necessary first and foremost for the person (or, to speak with Kant, the transcendental subject) to be capable of forming representations in the imagination that are independent of the actual state of affairs. In particular this includes the ability (i) to think in possibilities (which, as theoretical thought, corresponds to Kant’s Critique of Pure Reason), (ii) to think in norms and values (which as practical, i. e. moral thought corresponds to Kant’s Critique of Practical Reason) and (iii) to think teleological in means and aims (which as teleological thought made possible by the reflective power of judgment corresponds to Kant’s Critique of Judgement). Kant does not really deal with technology, never mind ignorance, but he does distinguish the mechanical arts from the free arts37; and since the ignorance at issue here presupposes knowledge, it will be helpful to take up a few of the points from Kant’s discussion of whether there is a teleology in nature comparable to the teleology of artifacts. When the cause of the object depends on the free will, Kant speaks of technica intentionalis. 38 Because its principles do not depend on causality, he calls it “morally practical”, whereas mere causality depends on “technically practical” principles; and he adds that the latter belong to the “theoretical philosophy as natural science”, the former to “practical philosophy as ethics”. He goes on: “All technically practical rules (i. e. the rules of art and skill generally …), so far as their principles rest on concepts, must be reckoned only as corollaries to theoretical philosophy.”39 However, when their principles depend on the free will, they do not rest on knowledge of nature but rather, morally practically, on moral principles. Now, these are precisely the two transcendental realms

37 Kant, §43; AA V 303. 38 Kant, §72, AA V 390. 39 Kant, Introduction, I.: Of the Division of Philosophy, AA V 172.

Knowledge of Ignorance

387

mentioned above that characterize the cognitive and the normative element of ignorance. But all of this is just the first step – the innovativeness of the Kantian view comes out when he writes that a “teleological (technical) method of explanation” belongs to “reflective judgment” – or nearer to the original text: to the “reflective power of judgement”40 – since it is a capacity of the transcendental subject to think of means and aims. In the introduction to The Critique of Judgement he explains his new and decisive concept of reflective judgment: Judgement in general is the faculty of thinking the particular as contained under the Universal. If the universal (the rule, the principle, the law) be given, the Judgement which subsumes the particular under it (even if, as transcendental Judgement, it furnishes a priori, the conditions in conformity with which subsumption under that universal is alone possible) is determinant. But if only the particular be given for which the universal has to be found, the Judgement is merely reflective. 41

We can take this as a very clear and penetrating conception of the engineer’s cognitive situation: since there are no universal laws of technology that would allow us to derive a special technological solution, we have to start from the individual case – in this case, a particular problem and its corresponding question – to reach a particular and practicable solution rather than a universal one. However, as long as it is only outlined conceptually, such a solution, though not a law, is nonetheless general; hence it is worth noting that engineers speak of the ‘solution principle’, thereby expressing something general. Here teleological thought finds its appropriate expression in an a priori faculty: it presupposes the epistemic categories just as much as the moral principles, but adds them as a teleological element to the ‘intentional technology’. This is what makes the decisive difference between an artifact and a natural object. Hence these brief remarks should explain why the engineer’s ignorance is truly an epistemological problem that leads us to a broad horizon of reflection. We have to include a further point here. Kant in his theory of reflective judgments aims at the problem of a teleology of nature – but we need it independent from nature, and related to artifacts and artificial processes depending on human aims. Concerning aims as better ends, 40 Kant, §71, AA V 389. 41 Kant, Introduction, IV.: Of Judgement As a Faculty Legislating a Priori, AA V. 179.

388

Hans Poser

norms and values play an essential role, since they are at the same time warranting openness on the side of the aim. This indicates that openness is already a constitutive part of ignorance: It is the direction of the possible solution, which is indicated as an epistemic content of the ignorance. Now we have to add two additional faculties that are only given parenthetical mention by Kant, namely the ability to learn and to be creative. Both presuppose freedom of the will. Creativity is the fundamental category for Alfred North Whitehead, who breaks up the Kantian scheme of categories as creativity makes it possible for our historical development to involve new schemata of ideas and new forms of thought. Thus a Whiteheadiean enrichment of our concepts allows us to involve elements of social epistemology. The addition of these two faculties to our epistemic categories allows an improved understanding of ignorance, since developing new ideas and becoming conscious of a new problem as an element of ignorance is of itself an act of creativity as a kind of openness. This lets us see the background of our consciousness of ignorance as a cultural element of learning and of conveying knowledge and transmitting methods across generations. To put this less abstractly in terms of an example: historians of technology have shown that in the 19th century the methods for solving technological problems in England, France and Germany differed quite significantly from one another due to the differences in training and education in the tradition of the Royal Society, the École Polytechnique de Paris and the German model of the Polytechnische Lehranstalt. But the positive aspect of such training was that all participating engineers had a comparable knowledge and experience at their disposal – and this is precisely what made it possible to communicate problems of ignorance. But to return to the larger point – what we have said so far shows that: The engineer’s ignorance contains an element of creativity even just in defining the problem and the corresponding question, since it aims at something new. We might think that in the sciences we are primarily concerned with ignorance as a lack of knowledge rather than ability and value; but the technological sciences cannot help involving all four levels – and not just from the perspective of knowledge but also and quite essentially from the perspective of ignorance. We must now turn to the subject

Knowledge of Ignorance

389

of ignorance, and hopefully the figure 1 prefixed here will help show the way. Tab. 1: Ignorance: Structure, content and preconditions Object level

Epistemic content

Epistemological preconditions

Ignorance

Knowledge of ignorance

Reflective faculty Thinking in possibilities (theoretical reason)

Problem

Knowledge – of ignorabimus – of the factual circumstances – of the causes Content interpreted – as aims, means, functions – as values, needs

Thinking in means and aims (teleological reason) Thinking in norms and values (practical reason)

Question

Specification: – heuristic methods – creativity

Faculties: – hermeneutical abilities – learning ability – creativity

Solution

Realization: – know how

Presuppositions: – free will – life experience

5. Domain-specific Ignorance and Problem-solving in Technology The model of mere practical knowledge proves to be too simple to precisely describe technological knowledge as a presupposition of the corresponding ignorance, since the remarkable process of technological creation, i. e. technological creativity, consists precisely in taking ignorance as the starting point for the invention of a new solution. As already mentioned, Popper put it quite succinctly in writing that all life is problem-solving, and he meant not just everyday life but also the sciences, including the development of technology. This sort of ignorance rests on a knowledge of the problem to be solved, hence a knowledge of value that allows us to set an aim; and that is precisely the concern of technology and the technological sciences.

390

Hans Poser

Ignorance as knowledge of a problem can relate to all of the four above-mentioned forms of knowledge TK1 to TK4 depending on the content that gets seen as a problem in that particular case. However, since the knowledge of value includes the other three forms, the same holds of ignorance. Hence we can consistently see a knowledge of value as a precondition behind every ignorance and every corresponding question: if this were not the case, then we would not pose the question – we would not be even be able to become aware of the problem constituting the ignorance: Our knowledge of ignorance, and with it the ignorance itself, are only comprehensible at all through the specific value-oriented problem. Now problem-solving involves a dynamic process of development whereby new ideas for solutions, born from the ignorance of a solution, are materially implemented and tested, criticized, modified and re-tested – and this process must be repeated for every individual element of a complex causally interactive system (think of a car with an entirely new propulsion system). As Michael Ruoff aptly described it: “Ignorance takes on an actively role by taking the new into consideration. Knowledge is directly dependent on a super-complex ignorance that can be seen as the source of the new”.42 But it can only become the source of the new because it rests on a contentful knowledge of a problem. Since the engineer’s ignorance has a content, expressed as a problem, it is helpful to distinguish several types of ignorance that are especially conspicuous in various areas of technology and to examine them in an epistemological light. Let us begin with mechanical engineering as an example of classic engineering science. It typically has well-developed heuristic methods aligned with a clear expectation of the solution to the problem. The results can be assessed with criteria based on fixed norms (functionality, technological efficiency, profitability, safety, etc.). This classic case, which is the most common type of ignorance, is usually not seen as ignorance at all; it is seen as an empirical question, since even the criteria relate to empirically reviewable data. However, it would be quite insufficient to see the overcoming of this ignorance just as a learning of method, even if the simplest case involves adapting a given scheme of action, perhaps one learned during the engineer’s training. A somewhat 42 Ruoff (2005), 170.

Knowledge of Ignorance

391

more elaborate heuristic involves methodically combining given elements into something new; both are described by Walter G. Vincenti as “normal design”, where the solution comes from “stored-up engineering knowledge”, whereas Japp speaks of “specific ignorance”, thus emphasizing the connection to our question here.43 I will call this the standard development. But although heuristic methods of solution are already given for the task, nonetheless the solution has to be found through an adaptation of method. ‘Heuristic’ means finding something that is not yet given; otherwise there would be no need to look. Thus here as well the development and design has to start from a case of ignorance: ITK1 Ignorance concerning the adaptation of heuristic methods to a given case. A new, creative invention followed by a process of development is seldom necessary. In these cases the creative result is by no means certain, though the same criteria hold as in the classic ITK1 case. Hence the epistemological problem remains as to how, presupposing creativity and freedom, we can achieve absolutely new functioning artifacts. As mentioned, the type of representation involved here differs from representations in literature or film or science-fiction through the decisive modal quality of realizability. This form is: ITK2 Ignorance concerning a realizable creative solution. Vincenti has investigated this using the case of airplane construction.44 Yet before him Günter Ropohl had already made a chart (see Fig. 1) showing the lines of development for every sub-process as well as for their combination into the sought-for whole.45 The chart describes the process as a closed circuit beginning with the setting of an aim – based on societal notions of value and a consideration of the available knowledge – that consists in the solving of a problem. This aim would not exist if it were not at the same time the expression of ignorance. Every subsequent individual step remains bound by this framework composed of problem and values. In fact the step-by-step development process includes both types of ignorance, ITK1 and ITK2, almost throughout the whole process, since many steps require new ideas, even if it can seem a little exaggerated to speak of “creativity” in an emphatic sense. Nonetheless the ITK2 case of 43 Vincenti (1990), 7 and 206; Japp (1999), sec. I. 44 Vincenti (1990). 45 Ropohl (1978); Ropohl (1999); Ropohl (2010).

392

Hans Poser

Figure 1: Process structure of techno-genesis (Ropohl (1999), 262) by kind permission of the author

“radical design”46 or “unspecific ignorance”47 should be noted as a case of creative development, since the available knowledge of heuristic methods is not sufficient and a wholly new solution has to be found; here, very little can be taken as certain, since, as Vincenti writes, it is widely unknown which parts should be applied how and how they work.

46 Vincenti (1990), 8. 47 Japp (1999), paragraph I.

Knowledge of Ignorance

393

For the form of ignorance involved in problem-solving it is decisive that the setting of the general aim is followed by the determination of functionally defined subaims, such that the process in Ropohl’s diagram gets repeated for every subaim. Ignorance as the knowledge of a problem gets broken down into sub-problems, naturally with a corresponding ignorance for each one, in order to follow the sub-problems stepby-step towards a solution. Yet since the problem can be broken down into sub-problems in various ways, this presents us once more with the structural openness of the question of ignorance. Ropohl’s diagram offers us a generalized picture of this because it describes a planning process that covers both standard design and creative design. This all sounds simple but proves to be highly complex, because the aim itself is highly complex. We can think of normal design in the trivial case of a new washing machine, which we expect not only to be effective – it should make a certain amount of dirty clothes clean – but also cheap, so that sales are guaranteed; and safe to produce and use; and ecological by using less water, less energy, and less detergent; and childproof; and easy to dispose of; and resistant to misuse. Hence a broad horizon of values needs to be taken into consideration, some of which represent opposing demands, for example profitability and safety, or the desire for individual profit and socially mandated limitations; and unknown possibilities such as misuse have to be anticipated and constructively avoided. Ropohl’s diagram only includes two explicit questions; but in fact every step and every loop have to be seen as an answer to a question of ignorance. Thus the developmental process proves to be guided throughout by questions and therefore by ignorance. Epistemologically we see here that ignorance leads by way of questions, conceives an aim and then tentatively and through trial and error strives for means for a solution in a controlled manner. The solution that the question of ignorance aims at is the result of a value-guided search for an answer. All of this seems to correspond to Smithson’s S1 form of ignorance; but in fact elements of S4 play a role here as well, since the aims depend on culture, as do the strategies for solving the problem. This is the point where social epistemology enters the picture and opens a dimension that is missed by the quasi-Kantian view under transcendental conditions of ignorance. But it would go too far to continue this analysis here. – In summary we can say: the ITK2 type of ignorance is characterized by two elements: the aim connected to the question only sets the direction, and the solution depends on elements of creativity and of the cultural

394

Hans Poser

tradition. This means that the communication of ignorance is impossible without a shared knowledge of all these elements among the participants in the communication. Today this has led to a further dimension of ignorance in the field of nanotechnology: we have uncovered new and unexpected phenomena on the nano-scale, and are able to manipulate nano-scale objects. But because there are no comprehensive or even sufficient theories here as an extension of solid-state physics, it is practically impossible to make any prognoses about what happens with nano-materials in nature and in living organisms. This can be seen as a third kind of ignorance: ITK3 Ignorance relating to the lack of fundamental theoretical knowledge. Nowadays nano-materials are used in colors, in nano-layers on glass and textiles, in skin cream for protection against the sun, etc. Here the analysis of ignorance leads us directly to ethical demands to leave such applications be until a more advanced foundational research has shown that no cumulative or catalytic effects can cause long-term dangers. The remarkable aspect of this is that such ignorance can cause a morally justified warning and a postulate for a specific type of application research, not just in light of the immediate consequences of present action, but in light of Jonas’ principle of responsibility in a long-term perspective that would not have been necessary earlier. This can be seen as a fourth kind of ignorance: ITK4 Ignorance founded on the consideration of moral arguments. Today biotechnology confronts us with completely novel problems; while the questions of cheese production or beer production belong to ITK1 ignorance, genetic technology is quite different: First of all, biofacts (as Nicole Karafyllis terms manipulated organisms48) are living creatures characterized by growth; hence it takes time to be able to observe the results of a genetic manipulation intended to change some quality, since the aim of the modification is not achieved immediately and needs to develop through growth. The same holds for long-term dangers – recognizing them might take generations. This requires a modification of the classic falsification method of problem-solving hypotheses. Secondly, biotic systems are complex systems and do not have the atomistic structure we originally assumed of genes (such as each gene representing a certain quality in the phenotype); for this reason we 48 Karafyllis (2003).

Knowledge of Ignorance

395

often lack the biological knowledge necessary for the intended modifications. Thus our ignorance consists in a true ITK3-type gap in knowledge. At the same time living creatures are autopoietic systems; thus changes in such a complex system with the aim of creating a biofact are highly problematic, since we cannot predict the result. But ignorance takes on a whole new dimension above all due to the fact that living creatures as autopoietic systems are complex and live in a complex, primarily biotic environment that precludes classic forms of experiments as well as predictions. Hence the bio-engineer’s problems, like the problems of environmental technology, are based on lack of knowledge, on a changed temporal situation compared to classic technologies, on the complexity of the living conditions of living objects, and on ethical problems. Since the engineer is aware of these situations and difficulties, we need to introduce another form of ignorance: ITK5 Ignorance due to the unpredictability of development in complex systems. This involves an ignorabimus that still needs to be explained. Information technology has completely reshaped not just society and everyday life but also the epistemological problem of ignorance, since it seems to be an adequate means to provide knowledge everywhere and for everyone and manipulate it with technical means. The objects of information technology consist primarily not in material artifacts or living biofacts but in information, given as a sequence of symbols that are materially stored, transmitted and/or manipulated by a programmed machine. But for this purpose knowledge has to be translated into a formal language, which brings with it a reduction in the complexity of the overwhelming amount of information (which in many cases does not contain knowledge); thus the primary challenge is almost always that of codifying knowledge or opinion considered to be knowledge. The epistemological and philosophical questions as to whether informational systems can ‘think’ and attain consciousness are well-known. Information can be stored independently of a person’s memory if the corresponding artifacts exist, such as books, PCs, cell phones, cameras, televisions, transmission channels, etc. – not to speak of the barely perceptible guiding elements found in almost every technological artifact, such as microprocessors. We can hardly overlook their use for the purposes of society, for example if we just think of traffic control systems or health care methods. Ubiquitous computing is already underway, and modern societies could not exist without it. But if we see this in terms of ignorance, we have to admit that the course of development

396

Hans Poser

is less clear by far than it is for classical technology, because there is a profound and rapid interplay between the development of technology and society, as SCOT (Social Construction of Technology) proponents have already shown using the example of the bicycle;49 so we have to consider that the path from the beginnings of the first walky-talkies to the mobile camera-internet-computer-telephones of today is anything but a clear line and instead is marked by a frequent back and forth, with the testing or neglecting of different possibilities, the manipulation through advertising, trends or just showing-off, rather than by rationally guided problems and decisions. Since this kind of ignorance of the consequences for society is one of our primary points of focus here, it will be useful to understand it as a particularly significant form of ITK5 ignorance in the sense of a lack of knowledge of the complex system ‘information technology plus society’. This also holds without question of the ‘disruptive technologies’ as innovations that can push aside previously dominant technologies (cf. Persistent Forecasting of Disruptive Technologies (2010) for the position of the National Research Council of the National Academies of the USA). In the literature the differentiation of types of technological solution has been seen as a parallel to Kuhn’s distinction between normal and revolutionary science;50 since the route from the first inventive idea to innovation consists precisely in creative development as the modification or replacement of the previously dominant paradigm. In its form of industrial development today this route is directly connected to the requirement of communication about this particular form of ignorance and ideas for overcoming it, hence communication about a content that has not been realized so far and might have to be captured by a new concept based on metaphors. Hence it is above all creative development and its antecedent ignorance that presents us with a remarkable epistemological problem.51 A part of this problem is the view, taken by Donald T. Campbell and going back to Popper, that that which is radically new in creative development has to be seen as a variation in an evolutionary process of technological development. But it would be misleading to model this in accordance with bio-evolution, since in the development of technology variation and selection are by no means independent of each other: though the particular invention cannot be pre49 Bijker (1987) and (1995). 50 Vincenti (1990), 260, fn. 10. 51 Poser (2006).

Knowledge of Ignorance

397

dicted in advance, it is aim-oriented; it is knowingly taken up and then modified in dependence on value-determined aims. This can also involve modifying the aim or giving it up when no-one has been able to find any suitable way or means in the sense of the cognitive premise of the practical syllogism. Yet one way or another, inventions present themselves as problem-solving possibilities. Hence the forms of ignorance ITK1 to ITK5 reflect a consequence of the growing demands on problem-solving knowledge, which at the same time is a precondition for the communication of ignorance. In extrapolation from Fig. 2 the results of this section are presented in a diagram (Fig. 3) showing the structural extension of ignorance in the corresponding approaches to a solution, from the heuristic adaptation of available methods to creativity all the way to a synthesis of theoretical and practical reason in reflective judgments.

6. The Transformation of Technological Problems into Value Problems as a Transformation of the Structure of Ignorance Today we no longer see technology as simple progress, but rather recognize its Janus-faced nature and its association with risks and dangers. In order to avoid unintended and unwanted consequences, ethical responsibility has become a central element in the design work of engineers. This has been documented and discussed by Beth Kewell for the time period from 2003 to 2008.52 In the above classification this corresponds to TNK4. This reveals a drastic transformation and expansion of problem-oriented ignorance in two directions: On the one hand today’s engineers have a different understanding of problems, as their ignorance relates to far more than the question of a technologically satisfactory solution, because they have to integrate a whole horizon of values: the assessment of technology is just the visible tip of an ice-berg of value postulates and hence represents a transformation of the engineer’s ignorance to a problem that has to respect entirely new evaluative points of view, since if the engineer were only to consider problems of safety, health and ecology at the end of the development it would be too late.

52 Kewell (2009).

398

Hans Poser

Tab. 2: Elements of ignorance of technological knowledge (ITK): forms and preconditions Missing knowledge

Form of Form of ignorance solution

Epistemological Precondition In general: Imagination Reflective judgment

I Adaptation of given methods

ITK1

Heuristic Teleological thought

Learning

ITK2

Development

Creativity

ITK3

Research Theoretical Empirical and reason theoretical Social considerations epistemology

ITK4

Ethical Practical reason considerations

ITK5

To avoid an ignorabimus: Reduction of parameters in feasibility studies

Know-how is required II New methods Creativity is required III Foundational theory in the nanosciences in the biosciences in the social sciences Knowing why is required IV Ethical consequences Knowing what-for is required V Complex systems Combination of I – IV

Theoretical and practical reason Reflective judgment

We can get a good picture of this horizon of values in its full breadth and inner tension in the octagonal scheme of value levels formulated and further differentiated in the Association of German Engineers’ Guideline 3780: Technology Assessment – Concepts and Foundations 53, first published in 1991 and developed by a interdisciplinary group of philosophers, historians of technology, sociologists of technology and engineers. At the 53 VDI 2000, 23.

Knowledge of Ignorance

399

Figure 2: Octagon of value-levels in technology (VDI guideline 3780, 23) by kind permission of VDI

same time this guideline makes the expansion of the knowledge presupposed by problems of technological ignorance particularly clear, since previously there had not been any such value-oriented guideline – and moreover a VDI guideline means that engineers in Germany have to hold themselves to it! Secondly, it is not enough just to include values in order to overcome contingency and ensure that future development precludes risks and dangers. This would require the abilities of a Laplacian demon at the very least. With a view to technology and its influence on society and the environment we have already seen that prognoses are impossible for complex systems. Hence we are ignorant in a more radical sense, precisely because we are clear about our own ignorance, since both

400

Hans Poser

the justification of value theories we would need for TNK4 and the prognoses of complex systems relating to TNK5 are an insurmountable ignorabimus. Nonetheless humanity must strive to limit danger as much as possible. This gives ignorance and the communication of ignorance a weight it never previously had. Let us start with the inclusion of values. Here not all forms of risk can be taken into account, which Smithson54 and his followers have discussed in their sociological treatment of ignorance, since the theoretical side is our focus. Rather than seeking ‘purely technological’ methods to solve a technological problem of ignorance, the solution sought for has to integrate a broad set of norms and values that represent forms of responsibility. Obviously technological aims depend on the needs of people in the society (whereas scientific problems arise from the scientific discipline in question); hence technology has always been bound to the society, the culture and the lifeworld. But solving a technological problem of ignorance has taken on a new quality, since the engineer and the R&D department have to take all values and the tensions between them into consideration. This requires in turn a consensus of value that is not immediately given in a globalized world and has to be sought for in order to give direction to the solution of the ignorance problem. Hence TNK4 concerns much more than just the ignorance on the part of the engineer. Creativity, as the sole means of escape when traditional solutions fail, has to master a much more comprehensive range of problems. To minimize risk, simulations and feasibility studies are formulated particularly with a view to values in order to develop entire possible scenarios for decisions. From the perspective of ignorance this means: the structure of the engineer’s ignorance today has seen a two-fold expansion, since various value levels need to be taken into account from the outset and since feasibility studies need to be included that rest on a drastic selection of the parameters that are considered significant, whereas there cannot be any theoretical or empirical justification for this reduction of complexity. Hence these procedures only seem to help overcome ignorance – in reality they are themselves a new type of ignorance.

54 Smithson (1989).

Knowledge of Ignorance

401

7. Unintended Consequences: Ignorance as a Modal Problem However, the ignorance of unintended consequences, particularly those that first appear at a later date, is significantly more dramatic than the forms of ignorance we have dealt with so far. Controlling such consequences is also one of the concerns involved in technology assessment. The point here is to systematically register possible consequences that are negative according to the canon of values, including any conceivable misuse, in order to avoid them. This raises problems of ignorance once again, and there are essentially two ways of responding to them. The first response is to avert the possible and recognizably undesirable consequences from the outset with suitable measures. The highly developed safety technology that has to be a part of every technological design nowadays aims at protecting both the worker in the production of the artifact as well as the user and tries to bar any conceivable misuse from the beginning with the appropriate constructive measures. Today this includes the broad-ranging considerations of possible health and environmental damage. This occurs methodologically just as it did in the cases of problem-solving discussed above, though with an essentially changed orientation: the original ignorance is converted into the devising of means of prevention with a view to the possible consequences. This also shows us that the reduction of parameters as the precondition for the applicability of prognostic simulations is not based solely on positive and successful experiential knowledge, but at least to the same extent on the experience of failures and on strategies to avoid them. The second and far more difficult case essentially connected with the problem of ignorance is where there are no preventative means at our disposal and yet we still wish to at least estimate the possible undesired and unfamiliar consequences. This form of coping with contingency is crucial for determining the actions we take. Of course it is impossible to predict the path of history; nonetheless unfamiliar dangers should be avoided. Hence this concerns the possible courses that complex systems might take, which moreover depends on framing conditions that can never be completely determined. The “shadow of the indefinite” – as the title of Albrecht Fritsche’s dissertation calls it – is unavoidable.55 In contrast to the ignorance involved in the search for solutions to problems, here it is doubtless so that wherever we create something definite – i. e., an artifact as the solution to a problem – “a site of indefiniteness 55 Fritsche (2009).

402

Hans Poser

arises simultaneously outside of it”.56 Here I do not wish to identify and assess this as a site of indefiniteness (there have been by now comprehensive debates about this) but rather as a site of ignorance. Technology is considered a quite essential means of dealing with contingency – after all, every artifact is materialized knowledge connected with the expectation that its particular function as a means is guaranteed as an essential and constitutive property not just yesterday and today but tomorrow as well. Yet technology can fail just as social regulations can – in unforeseen ways, i. e. without prior knowledge, thus revealing our ignorance. To approach all this from the perspective of possibility will be more productive. Thinking in possibilities is constitutive of human decision and action. However, to date possibilities have only been linked to knowledge in the exceptional case of Nicolai Hartmann’s real possibility 57 or Ernst Bloch’s factually-objectively possible 58, since this type of possibility coincide with reality seen from the modal perspective of possibility. Yet all technological design and development occur within spaces of possibility that go far beyond this. We began by distinguishing forms of necessity, which correspond to forms of possibility such as logical possibility and physical possibility, i. e. possibility within the laws of nature, along with theoretical technological possibility and practicability, the latter of which is characterized by the mode of realizability and thus comes closest to Hartmann’s real possibility. All of these distinctions are definitive for the process of technological design and construction, and hence in light of the requirements of problem-solving they concern forms of ignorance. Yet our central problem is the drive to penetrate the contingent space of ignorance even where the knowledge of problems has to be first established. As Gamm and Hubig emphasize, it is a basic principle here that the more precision we demand, the greater is the uncertainty.59 Conversely this means that we can only locate the ignorance at issue through an increase in imprecision. Yet this is not the sole source of the tension, which also arises from the relation between such ignorance of an open space of possibilities and the essence of technology, i. e. serving to manage contingency through expediency, safety and repeatability. 56 57 58 59

Fritsche (2009), 45. Hartmann (1938), 49. Bloch (1959) and (1986), vol. I: 271/225. Gamm (2005); Hubig (2005), 39.

Knowledge of Ignorance

403

In view of this problem of possibility, it is crucial that we take up into the spectrum we have sketched so far that form of possibility concerning the remaining ignorance. Of course it has to be found somewhere between physical possibility and practicability. Certainly it is a form of virtuality that is more than just a ‘real virtuality’, as Hubig60 describes it, but rather a virtuality with an assumed realizability, even if in hypothetical and conditional form. Nicolai Hartmann’s modality of becoming, an ontological modality, does not quite give us what we want, whereas the epistemic possibility that ends in the “understanding of possibility” comes closer.61 Insofar as this is a possibility distinct from reality – Hartmann calls this “negative possibility” – consciousness singles out those (ontologically possible) “partial possibilities” as a “series of conditions”, naturally without ever reaching a “total possibility”.62 He describes how what is “modally peculiar” is “that the modal precondition in the accompanying consciousness of possibility is not the same as that in the understanding of possibility. In the former a disjunctive [positive vs. negative, or real vs. merely possible] possibility is presupposed, and in understanding a distinctly unambiguous […] total possibility. […] Seen from this point of view, the first emergence of knowledge of certain real conditions is already an enormous step up in terms of understanding. Only with this step does partial possibility emerge, and with it the consciously disjunctive plurality of eventualities. All of this already belongs to understanding. But at the same time in understanding the other a priori presupposition sets in: that of the unrecognized total possibility”.63 Let us now apply this to the assessment of the consequences of technology as a case of limited knowledge – limited not just concerning the continuously hypothetical knowledge founded on natural science, as is so striking today in the case of the nano-world and for neurology and genetic biology as well, but also limited concerning cultural and social conditions and transformations, and almost empty of content insofar as it pertains to the development of science and technology. The traditional means to assess consequences even in this situation was an appeal to life experience, hence an appeal to the presupposition that the future will in principle be similar to the past. But is this still true today? Our means for dealing with contingency has transformed – even if it is still 60 61 62 63

Hubig (2005). Hartmann (1938), 381 and 402 ff. Hartmann (1938), 382. Hartmann (1938), 389 f.

404

Hans Poser

technological in nature when we carry out simulations and scenarios of possible courses of events, as if we were capable of choosing the best of all possible worlds like Leibniz’ God. But in doing so we come to resemble that figure of Karl Valentin’s who only looks for his lost house-key under the street lamp because it’s only light enough there to see. Moreover the trust in such simulations, even when we take the uncertainties that it involves into account, still rests on ignorance: The seeming rationality of the process of simulation drifts apart from reason and knowledge. But then how should a technology assessment, a weighing of risks and decision management, be a well-grounded undertaking? We cannot predict future knowledge – if we could, we would have it already. New future problems, too, cannot be predicted. Yet as we have seen, a knowledge of the problem is a precondition for systematically reducing the related technological ignorance. Furthermore, the inversion of the end-means relation, hence the search for new aims for a given means, is just as impossible to anticipate as new creative solutions and new social notions of value. This is where the broad discussion of risk begins that found its expression in Ulrich Beck64 and Niklas Luhmann65 and becomes an issue directly or indirectly in the problem of ignorance. To deal with risks we need models in order to be able to structure anything even under those risk conditions. Hence concerning technology assessment Armin Grunwald distinguishes between four forms of future projection – the prognostic, the creative, the evolutive, and finally the autopoietic view, which he also discusses and which today is expanding from biotic processes to include prebiotic and social processes.66 In the first case we assume an ability to make predictions on the course of the future – which presupposes laws of the metaphysics of history, such as Hegel, Marx, Spengler and Toynbee all assumed each in their own way. In the second case, the optimism about progress that has prevailed since the Renaissance is applied in its technological version, whereas in the third case the retrodictively conceived evolutionary explanation of a developmental dynamic is used as a general projection of the future without the possibility of intervention. In the case of autopoiesy we assume a system structure that, like Prirogine’s dissipative structures, involves a 64 Beck (1992). 65 Luhmann (1993). 66 Grunwald (2003).

Knowledge of Ignorance

405

non-predictable stabilization, followed by a destabilization when certain framing conditions are abandoned. Now all four of these forms rest on global projections of the future with presuppositions that cannot be validated. But this does not cloak our ignorance so much as to make it all the more clearly visible. On the other hand each form takes on elements according to which their undoubtable explanatory power for past events corresponds to expectations for the future. Hence their claim to exclusivity is to be rejected; since alone for ethical reasons if nothing else we bear the responsibility for enabling future generations to live lives that are worth living, as Hans Jonas has made it impossible for us to forget. This ethics of the future obligates us not to throw up our hands but rather to take responsibility for our knowledge and ignorance in all our plans and decisions and not to hide behind an Ultra posse nemo obligatur. We are able to make prognoses, even if they are bound to preliminary decisions, presuppositions and conditions that make them uncertain; we can exert a good measure of control, in general and in detail, even if we bump up against certain limits; and we find ourselves confronted with the unexpectedly and unpredictably new in the form of evolutionary mutations or variations, no matter how imprecise this concept might be in social, cultural or technological development, and yet it is precisely here that we can intervene with our own selection and control retention as well as the next variation. The same can be said for models of self-organization; and if necessary they can even be applied to human action as well, if, as sometimes happens, we include processes of reflection on the part of individuals and the society, in which case the principle of responsibility becomes an integral part of the self-organization process. Though the fact that this reflection is global and has since come to unite all industrial nations might, as Alfred Nordmann emphasizes, lead to a spatialization of the problem of responsibility, nonetheless it remains basically future-oriented, since every action and the associated responsibility is and must be essentially future-oriented.67 It is for just this reason that we have to clarify the estimation of the consequences of technology, as a communication of ignorance, in view of our ignorance as well – which includes the efforts to conceptually clarify what type such a knowledge of possibility might take. However, this entails the danger emphasized by Japp, namely that the scientific methods used here might give the false impression that we are dealing with a specific and thus methodically controllable 67 Nordmann (2005), 114 f.

406

Hans Poser

ignorance if not a secure knowledge, whereas in fact it is an unspecific ignorance we are faced with.68 Japp stresses the need to distinguish in the case at hand between specified ignorance as a possible gain in knowledge and unspecified ignorance as a risk, since “the one case leads to a reduction of contingency, the other its (as always relative) unbounding.”69 Hence here as well we find a modal characterization at a decisive point. Ernst Bloch emphasizes that in the case of the factually-objectively possible (das sachlich-objektiv Mçgliche – which, as mentioned, is still separated from objectively real possibility (objektiv-reale Mçglichkeit) by the fact-based object-suited possibility (sachhaft-objektgemße Mçglichkeit)) our approach is marked by a “heuristic principle” that becomes operative, on his view, “in the hypothetical simplification or hypothetical analogy to what is already better known”, in order to move inductively “in the direction of the supposed conditional context”.70 Yet all of this remains within the realm of the merely possible and hypothetical and hence is characterized by uncertainty, not to speak of necessity. Now we have seen that all the possibilities at issue here are epistemic modalities that at the same time seek to comprehend ontic possibilities without considering the necessary bridge. For Leibniz this bridge is given by the presupposition that all conceptually constituted possible worlds are at the same time ontologically possible in regione idearum: God could create them. For Kant the modal categories of the subject of knowledge, i. e. epistemic modalities, have priority, because they are at the same time the conditions of possibility for the objects of experience. Whereas Bloch’s reflective materialist position proceeds on the basis of a real possibility, Nicolai Hartmann seeks a balance between the ontological spheres of the ideal and the real that together first constitute reality. But how should we proceed with a view to technology? We can take a decisive pointer from the modality presupposed by all technological design and thus by all simulations and feasibility studies, namely the modality of realizability, which constrains conceptual possibility to those portions for which we can claim an ontological interpretation. Here conceptual possibility has to be backed by some kind of knowledge, however hypothetical. So where does that leave ignorance? Our obstacle here is the linguistic level of logic, which demands that the negation of a be located at the same level in the universe of discourse 68 Japp (1999), sec. II. 69 Japp (1999), sec. III. 70 Bloch (1959); Bloch (1986), vol. I: 262/227 f.

Knowledge of Ignorance

407

as a itself. But in fact the negation of knowledge in the form of ignorance proved to be a meta-linguistic reflection (as holds of modal concepts as well, of which Kant stresses that they add nothing to the matter itself and only characterize our attitude to it). This brings out the fact that, to put it in Nicolai Hartmann’s terms, all designs and projections, however subtle, are continuously bound to the awareness that we are dealing exclusively with partial possibilities as a series of (possible) conditions and never with a total possibility, which fundamentally eludes our cognition and hence our knowledge.

8. The Inversion of the Modal Perspective What I have said so far has moved within the lines of a familiar discussion, supplemented and enriched by a focus on the surmounting of contingency and thus brought into relation to the forms of possibility. Now all this opens a different and possibly quite fruitful perspective, if we start from total and global ignorance in the sense of the Aristotelian and Lockean tabula rasa. The erection of knowledge and with it the efforts to tackle ignorance serve to manage contingency. Contingency means everything that is neither necessary nor impossible, hence all factual truths of all possible worlds, not just this world. From a systematic point of view the first step is to separate out everything that is necessary and impossible – a crucial step, since we can rely on that which is necessary, and we do not need to worry about impossible things, since they will never be the case. Technology as the management of contingency should guarantee the success of all those rules of action assigned to it. Nonetheless ignorance maintains a space of possibility; and the reason for this is found in the historicity of the process of overcoming ignorance by the acquisition of knowledge. Stephan Fischer has recently proposed an approach to interpreting scientific development that can be applied to our problem.71 In the fog of ignorance (Fischer speaks of conceptual possibilities) we periodically establish stakes of knowledge (“point propositions”) based on experience, which then allow us to make their vicinity and the path to the next stake of knowledge into a problem to be sounded out. This changes global ignorance, since it gets stamped with the struc71 Fischer (2003), 143 ff., continued in Fischer (2010).

408

Hans Poser

ture of a local problem. But very different solutions to the problem are still conceivable at this point, and in principle an unlimited number of them – this leads to the picture of ignorance growing faster than knowledge. Hence it is necessary to introduce not only a temporal dimension, but also and moreover a multi-dimensional structure of imagination; I hesitate to speak here of dimensions of possibility, since in the situation just described it has to remain open whether what we are imagining – such as a reduction of mathematics to logic, or a perpetual motion machine, or the principles of eternal peace – is also possible. But we can allow ourselves to say that wherever our ignorance is structured by problems, the solutions to the problem have to be realizable, i. e. possible in the sense of Nicolai Hartmann’s partial possibilities or Ernst Bloch’s objectively real possibility if not fact-based object-suited possibility. This shows that we are now operating within a multi-dimensional field of possibilities of which we only demand their realizability as a further modal determination. But the more clearly these last three fields are structured – problems, imaginable solutions and realizable possibilities – the better we can communicate, firstly, what type of ignorance our ignorance is, and secondly, what new problems are connect with it, i. e. which phenomena of ignorance can be more closely described. All of this seems quite far removed from the tangible problems of technology assessment. But in fact it turns out that the only path open to us consists in structuring ignorance according to a problem. This is a path that demands a great many presuppositions, since we can only start out from the points and trails in our network of knowledge – whether theoretically, practically, or normatively and evaluatively – while seeking to expand this network by asking questions inductively, reductively and by means of analogies. This cannot succeed without a knowledge of our ignorance and its presuppositions. Hence it is decisive that we determine, evaluatively, what we should see as relevant for us today, or else we will fail. Only in this way can ignorance be explicitly constituted and determine a problem. But this brings us astonishingly far. We need only remember nanotechnology and our hardly sufficient reflections on it: We owe points of knowledge about numerous phenomena in the nano-realm that are highly interesting for technological applications, but we have practically no comprehensive solid-state physics for this area. For this reason we lack those hypothetical laws of nature that could significantly reduce the amount of contingency in our ignorance, so that classical prognostic procedures going beyond the specific point proposition describing the phenomenon are not at

Knowledge of Ignorance

409

hand. Instead we only ask, for example, whether the materials are toxic. The question we take from our experience of life and science – as to whether accumulations can occur within organisms and what consequences this might have, whether we should expect catalytic processes such as with CFCs or whether nanoparticles remain in the environment after we burn trash. In these cases, life experience already allow for a distinct structuring of our ignorance. Here we see that this elementary experience is wrongly disdained and can be applied quite helpfully. The next step must be to relate all of these valuable questions to their specific evaluative presuppositions and to take them up theoretically as well as practically in order to find solutions rather than waiting decades until the unsavory consequences become evident. Naturally these are all just steps into ignorance – but they are methodically guided as steps towards structuring a multi-dimensional space of possibility. At the same time this proves to be the precondition for being able to communicate about ignorance so as to be able to formulate problems and ask aim-setting questions: Hans Jonas’ principle of responsibility obligates us to precisely these steps. It does not require halting technological development, since even just the great problems of humanity such as hunger, thirst, illness, overcoming conflicts rather than war and terror, never mind life without human dignity, cannot be solved without it. Ignorance as non-knowledge can always only be partially reduced; a total overcoming of contingency is impossible. But using the knowledge we have to formulate problems and search for solutions in order to reduce ignorance is our obligation. Hence the central challenge to humanity is to structure our ignorance and make it communicable in order to be able to make knowledgeable decisions about our ignorance. That in doing so we move within a field of possibility can be seen not least of all by Hans Jonas’ second, mostly overlooked formulation of his concern: Act so that the effects of your action are not destructive of the future possibility of such [genuine human] life.72,73

72 Jonas (1984), 11. 73 Translation from German by Karsten Schöllner.

410

Hans Poser

References AA, Königlich Preußische Akadmie der Wissenschaften (ed.), Kant’s gesammelte Schriften, Berlin, 1902 ff. Banse/Hronsky/Nelson (2005): Gerhard Banse/Imre Hronsky/Gordon Nelson (eds.), Rationality in an Uncertain World, Berlin. Banse/Ropohl (2004): Gerhard Banse/Günter Ropohl (eds.), Wissenskonzepte in der Ingenieurpraxis. Technikwissenschaften zwischen Erkennen und Gestalten, Düsseldorf. Beck (1992): Ulrich Beck, Risk Society: Towards a New Modernity, London. Bijker (1995): Wiebe E. Bijker, Of Bicycles, Bakelites and Bulbs. Toward a Theory of Sociotechnical Change. Cambridge/MA. Bijker/Pinch (1987): Wiebe E. Bijker/Trevor J. Pinch (eds.), “The Social Construction of Facts and Artifacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit of Each Other”, in: W. E. Bijker/T. P. Hughes/T. J. Pinch (eds.): The Social Construction of Technological Systems, Cambridge/MA, 17 – 50. Bloch (1986): Ernst Bloch, The Principle of Hope, vol. 1, Cambridge/MA. du Bois-Reymond (1912): Emil du Bois-Reymond, “Über die Grenzen des Naturerkennens”, in: Estelle du Bois Reymond (ed.), Reden von Emil du Bois-Reymond, vol. I, Leipzig, 441 – 473. Bunge (1966): Mario Bunge, “Technology as Applied Science”, in: Technology and Culture 7, 329 – 347. Callon/Lascoumes/Barthe (2009): Michel Callon/Pierre Lascoumes/Yannick Barthe (eds.), Acting in an Uncertain World. An Essay on Technical Democracy, Cambridge/MA. Fischer (2003): Stephan M. Fischer, Zu den Erklrungen der Evolutionsbiologie. Eine Analyse der nicht-kausalen Erklrungsstruktur in Evolutionstheorien und ein Entwurf eines narrativen Erklrungsmodells zur Rechtfertigung des Erklrungsanspruches aus der Sicht der Wissenschaftstheorie, Münster. Fischer (2010): Stephan M. Fischer, Dynamisches Wissen. Die Einschrnkung der Mçglichkeit, Weilerswist. Fritsche (2009): Albrecht Fritsche, Schatten des Unbestimmten. Der Mensch und die Determination menschlicher Ablufe, Bielefeld. Gail (1999/2000): Michael Gail, “Angewandtes Nichtwissen. Eine Annäherung”, in: Ungewußt. Zs. fr Angewandtes Nichtwissen 8, www.uni-siegen. de/~ifan, 3 – 6. Gamm (2005): Gerhard Gamm, “Unbestimmtheitssignaturen der Technik”, in: Gamm/Hetzel (2005), 17 – 35. Gamm/Hetzel (2005): Gerhard Gamm/Andreas Hetzel (eds.), Unbestimmtheitssignaturen der Technik. Eine neue Deutung der technisierten Welt, Bielefeld. Goldman (1999): Alvin Goldman, Knowledge in a Social World, Oxford. Grunwald (2003): Armin Grunwald, “Die Unterscheidbarkeit von Gestaltbarkeit und Nicht-Gestaltbarkeit der Technik”, in: A. Grunwald (ed.), Technikgestaltung zwischen Wunsch und Wirklichkeit, Berlin, 19 – 38. Hartmann (1938): Nicolai Hartmann, Mçglichkeit und Wirklichkeit, Berlin.

Knowledge of Ignorance

411

Hubig (2005): Christoph Hubig, “‘Wirkliche Virtualität’. Medienveränderung der Technik und der Verlust der Spuren”, in: Gamm/Hetzel (2005), 39 – 62. Irrgang (2004): Bernhard Irrgang,“Konzepte des impliziten Wissens und die Technikwissenschaften”, in: Banse/Rophol (2004), 99 – 112. Japp (1999): Klaus P. Japp, “Die Unterscheidung von Nichtwissen”, in: TADatenbank-Nachrichten 3/4, 25 – 32, also www.itas.fzk.de/deu/tadn/ tadn993/japp99a.htm. Jonas (1984): Hans Jonas, The Imperative of Responsibility. In Seach of an Ethics for the Technological Age, Chicago. Karafyllis (2003): Nicole C. Karafyllis (ed.), Biofakte. Versuch ber den Menschen zwischen Artefakt und Lebewesen, Paderborn. Kassavine (2003): Ilya T. Kassavine, Soziale Erkenntnistheorie, Hildesheim. Kewell (2009): Beth Kewell, “‘Probability But Not As We Know It’: Ignorance Construction in Genetic Biotechnology Discourse”, in: The International Journal of Technology, Knowledge and Society 5.6, 1 – 18. Luhmann(1992): Niklas Luhmann, “Ökologie des Nichtwissens”, in: Niklas Luhmann, Beobachtungen der Moderne, Opladen, 129 – 220. Luhmann (1993): Niklas Luhmann, Risk: A Sociological Theory, Berlin. Magnus (2008): David Magnus, “Risk management versus the Precautionary Principle. Agnotology as a Strategy in the Debate over Genetically Engineered Organisms”, in: Proctor/Schiebinger (2008), 250 – 265. Nordmann (2005): Alfred Nordmann, “Wohin die Reise geht. Zeit und Raum in der Nanotechnologie”, in: Gamm/Hetzel (2005), 102 – 123. O’Sullivan/Holtzclaw (1980): Patrick O’Sullivan/Gary D. Holtzclaw, “Transport network planning under ignorance”, in: Transportation Planning and Technology 6.1, 27 – 31. Persistent Forecasting of Disruptive Technologies. Committee on Forecasting Future Disruptive Technologies (2010). Division on Engineering and Physical Sciences. National Research Council of the National Academies. Washington, D.C., National Academies Press. http://books.nap.edu/ openbook.php?record_id=12557&page=R1 (5. 9. 2010). Peppel (1994): Claus Peppel, “Tertium non datur. Über die Funktionsweise konservativer Denkmuster”, in: Ungewußt. Zs. fr Angewandtes Nichtwissen H. 4, www.uni-siegen.de/~ifan, 60. Polanyi (1966): Michael Polanyi, The tacit dimension, London. Poser (1998): Hans Poser, “On structural differences between science and engineering”, in: Philosophy and Technology: Quarterly Electronic Journal 4, no. 2, 81 – 93. Poser (2006): Hans Poser, “Wissenschaftsmodelle des Neuen und ihre Grenzen. Kreativität und die Theorien der Komplexität”, in: G. Abel (ed.), Kreativitt. XX. Deutscher Kongress fr Philosophie 2005. Kolloquiumsbeitrge, Hamburg, 966 – 982. Poser (2007a): Hans Poser, “Bedingungen und Grenzen des wissenschaftlichen Wissens. Das Beispiel Natur- und Technikwissenschaften”, in: S. Ammon/ C. Heineke/K. Selbmann (eds.): Wissen in Bewegung. Vielfalt und Hegemonie in der Wissensgesellschaft, Weilerswist, 41 – 58. Poser (2007b): Hans Poser, “Theories of complexity and their problems”, in: Frontiers of Philosophy in China 2, Beijing, 423 – 436.

412

Hans Poser

Poser (2009): Hans Poser, “Technology and necessity”, in: The Monist 92, no. 3, 441 – 451. Proctor/Schiebinger (eds.) (2008): Robert N. Proctor/Londa Schiebinger, Agnotology: The making and unmaking of ignorance, Stanford. Roland (2002/2003): Bernd Roland, “Das Projekt des Angewandten Nichtwissens – Rückblick und Ausblick”, in: Ungewußt. Zs. fr Angewandtes Nichtwissen H. 10, 3 – 30. www.uni-siegen.de/~ifan. Ropohl (1978/2009): Günter Ropohl, Allgemeine Technologie. Eine Systemtheorie der Technik, München/ Karlsruhe. Ruoff (2005): Michael Ruoff, “Das Problem des Neuen in der Technik”, in: Gamm/Hetzel (2005), 169 – 183. Smithson (1985): Michael Smithson, “Towards a social theory of ignorance”, in: Journal for the Theory of Social Behaviour 15, no. 2, 151 – 172. Smithson (1989): Michael Smithson, Ignorance and Uncertainty: Emerging Paradigms, New York. Smithson (1990): Michael Smithson, “Ignorance and Disasters”, in: International Journal of Mass Emergencies and Disasters 8.3, 207 – 235. Smithson (1993): Michael Smithson, “Ignorance and Science. Dilemmas, Perspectives, and Prospects”, in: Science Communication 15, no. 2, 133 – 156. Smithson (2008): Michael Smithson, “Social Theories of Ignorance”, in: Proctor/Schiebinger (2008), 209 – 229. Tuana (2004): Nancy Tuana, “Coming to Understand: Orgasm and the Epistemology of Ignorance”, in: Hypatia 19, no. 1, 194 – 232; repr. in: Proctor/Schiebinger (2008), 108 – 145. Tuchel (1967): Klaus Tuchel, Herausforderung der Technik. Gesellschaftliche Voraussetzungen und Wirkungen der technischen Entwicklung, Bremen. Vanderburg (2002): Willem H. Vanderburg, The Labyrinth of Technology: A Preventive Technology and Economic Strategy as a Way Out, Toronto. VDI Richtlinie 3780: Technikbewertung. Begriffe und Grundlagen/Technology Assessment. Concepts and Foundations. German-English edition (2000) [first ed. 1991]. Düsseldorf. Vincenti (1990): Walter G. Vincenti, What Engineers Know and How They Know It. Analytical Studies from Aeronautical History, Baltimore. von Wright (1963): Georg Henrik von Wright, “Practical Inference”, in: The Philosophical Review 72, 159 – 179. von Wright (1972): Georg Henrik von Wright, “On So-Called Practical Inference”, in: Acta Sociologica 15, 39 – 53.

Notes on Contributors Gnter Abel is Professor of Theoretical Philosophy at the Technical University Berlin, and director of the Center for Knowledge Research (IZW). He published on epistemology, philosophy of mind, philosophy of science, and philosophy of symbols, signs, and language. Among his numerous books and articles are Zeichen der Wirklichkeit (2004), and „Knowing-How: Indispensable but Inscrutable“ (2011). He is the main editor (with James Conant) of the book series Berlin Studies in Knowledge Research. Jocelyn Benoist is Professor of Contemporary Philosophy and Theory of Knowledge at the University of Paris-I, Panthéon-Sorbonne, as well as a distinguished member of the Institut Universitaire de France, and the Director of the Husserl Archive of Paris (Ecole Normale SupérieureCNRS). He is working on the border between phenomenology and Analytic Philosophy. His publications include Concepts (2010), and Elments de Philosophie Raliste (2011). Ulrich Dirks (Center for Knowledge Research, Berlin) studied philosophy, physics and mathematics. He recently was Visiting Professor of Philosophy at the Technische Universität Berlin (2008 – 2010). His research is mostly concerned with problems in philosophy of language, philosophy of signs and interpretation, epistemology, philosophy of science and philosophy of mathematics. He published articles and book contributions in these areas and is coeditor of Hans Reichenbach: Philosophie im Umkreis der Physik (1998) and Modelle (2008). Arata Hamawaki is Assistant Professor of Philosophy at Auburn University. He works on Kant, Wittgenstein, and Kantian themes in contemporary philosophy. He has recently written on Kant’s conception of the normativity of aesthetic judgment and on the truth in skepticism in the work of Stanley Cavell and of Thompson Clarke. Michael Hampe is Professor of Philosophy at the Swiss Federal Institute of Technology, Zürich. He is currently working on different research projects: the history of therapeutic understanding of philosophy from

414

Notes on Contributors

Spinoza to Cavell, the history of knowledge, and the history of the concept of natural law. He is author of Eine kleine Geschichte des Naturgestzbegriffs (2007), and of Tuguska oder Das Ende der Naturas vollkommne Leben (2011). Peter Janich is co-founder of the approach of methodical culturalism, a development of the methodical constructivism of the Erlanger school. From 1980 until 2007 Janich held the chair for Systematic Philosophy at the Philipps University of Marburg. He has published a number of books, including Was ist Information? Kritik einer Legende (2006), and Kein neues Menschenbild. Zur Sprache der Hirnforschung (2009). Thomas Land is Donnelley Research Fellow in philosophy at Corpus Christi College, Cambridge. His research centers on Kant’s theoretical philosophy. He is the author of „Kant’s Spontaneity Thesis“ and „Intuition and Judgment.“ Bernd Mahr is Professor of computer science at the Technical University Berlin. He published numerous articles and essays on formal modeling, focusing on its epistemic role and fundamental impact to Sciences and Arts. He co-authored the book Mathematisch-strukturelle Grundlagen der Informatik (2001), and published Information Science and the Logic of Models (2009). Martina Plmacher is Professor of Philosophy at the Technical University Berlin. Her areas of research are epistemology, philosophy of signs and language, and ethics. She has published on Ernst Cassirer, Edmund Husserl, Gestalt theory, philosophical theories of perception and representation, and on pictorial semiotics. Currently she is working on epistemic perspectivity. Her publications include Wahrnehmung, Reprsentation und Wissen (2004), and „Epistemische Perspektivierungen“ (2010). Hans Poser is Professor of Philosophy at the Technical University Berlin. He is well-known for his work in the fields of history of modern philosophy from Descartes to Schelling, philosophy of sciences and engineering, modal theory, and philosophy of mathematics. Hans Poser published, for example, Wissenschaftstheorie (2001), and Gottfried Wilhelm Leibniz zur Einfhrung (2005). Hans-Jçrg Rheinberger is director at the Max Planck Institute for the History of Science in Berlin. The main focus of his research lies in the

Notes on Contributors

415

history and epistemology of experimentation in the life sciences. By bridging the gap between the study of history and contemporary cuttingedge sciences, such as molecular biology, his work represents an example of transdisciplinarity as emerging in the present knowledge-based society. Rheinberger is author of the books An epistemology of the concrete. Twentieth-century histories of life (2010), and On historicizing epistemology (2010). Hans Jçrg Sandkhler was from 1974 to 2005 Professor of Philosophy at the University of Bremen, where he led from 2003 to 2011 the German Department „Human Rights and Culture“ of the European UNESCO Chair of Philosophy (Paris). His philosophical work is devoted to epistemology, and philosophy of law. His books include Kritik der Reprsentation. Einfhrung in die Theorie der berzeugungen, der Wissenskulturen und des Wissens (2009), and Recht und Moral (2010). Hans Julius Schneider is Professor of Theoretical Philosophy at the University of Potsdam. He is widely known for his work on Wittgenstein. His research in epistemology focuses on the limits between life-world and scientific knowledge, between different media and sign systems, and between linguistic and pre-linguistic knowledge. He is author of the books Phantasie und Kalkl. ber die Polaritt von Handlung und Struktur in der Sprache (1999), and Religion (2008).

Index of Persons Abel, Günter 1-52, 83-86, 90f., 93, 95, 97f., 100-105, 109-113, 115ff., 119, 121-124, 126ff., 161, 169, 179f. Agam-Segal, Dafi 241 Albert, Hans 50 Alberti, Leon Battista 311 Allais, Lucy 206 Anscombe, Gertrude E. M. 260 Archimedes 59 Aristotle 78, 88, 361, 370 Arnheim, Rudolf 5 Arnold, Markus 180 Atli, Can VIII Austin, John L. 148, 276, 282 Bab, Sebastian 335 Bachelard, Gaston 175f., 187, 289ff., 297 Baier, Sabine 353 Banse, Gerhard 371 Bartelborth, Thomas 186f. Bays, Timothy 314 Beck, Ulrich 404 Beckermann, Ansgar 20, 30 Beethoven, Ludwig van 49 Bennett, Max R. 13 Benoist, Jocelyn 273-283 Bernard, Claude 292f. Bernecker, Sven 14 Berz, Peter 293 Bijker, Wiebe E. 396 Blecher, Ian 238 Bloch, Ernst 402, 406, 408 Boltzmann, Ludwig 359 Bolzano, Bernhard 325ff. Bourdieu, Pierre 291ff. Boyle, Matthew 199, 238 Brandom, Robert 24, 39 Brecht, Bertolt 94

Bredekamp, Horst 302 Brentano, Franz 324-329, 335 Brewer, Bill 199, 220 Broy, Manfred 302 Brush, Stephen G. 359 Bunge, Mario 376, 381 Burge, Tyler 27f. Campbell, Donald T. 396 Canguilhem, Georges 288 Canisius, Peter 159 Cantor, Georg 325f. Cardone, Felice 315 Carnap, Rudolf 60 Carrier, Martin 190 Cassirer, Ernst 162, 167, 176, 179 Certeau, Michel de 288 Cesariano, Cesare 319 Cézanne, Paul 49 Chadarevian, Soraya de 302 Chalmers, David 27f. Chisholm, Roderick 24 Church, Alonzo 311-314 Churchland, Paul 27 Clark, Andy 27f. Conant, James 238, 241, 245 Copernicus, Nicolaus 109 Correns, Carl 296f. Craig, Edward 22, 24 Danielson, Dennis R. 109 Dante Alighieri 109 Danto, Arthur 134f., 147 Darwin, Charles 355, 358-362 Daston, Lorraine 155 Dauben, Joseph Warren 326 Davidson, Donald 8f., 14, 212f. Dedekind, Richard 326 Descartes, René 24, 257-260, 264, 266, 270f.

418

Index of Persons

Dessauer, Friedrich 384 Detel, Wolfgang 184 Devitt, Michael 36 Dewey, John 161 Dilthey, Wilhelm 85f. Dingler, Hugo 62 Dirks, Ulrich 83-131 Dressel, Gert 180 Dretske, Fred 14, 19, 24, 174 du Bois-Reymond, Emil 179, 375, 378 Dummet, Michael 13 Dürer, Albrecht 342f. Dutant, Julien 279 Dux, Günter 118 Dworkin, Ronald 281 Ehrig, Hartmut 315f. Eilers, Florian 336 Einstein, Albert 49, 60, 62, 103, 183, 365, 378 Engel, Pascal 279 Engels, Friedrich 360 Engstrom, Stephen 208 Esposito, Joseph L. 363 Euclid 49, 59, 61f., 64, 79, 378 Evans, Gareth 200 Faizi, Hadi Nasir VIII Feferman, Solomon 314 Feldmann, Ute 191 Finkelstein, David 241 Fischer, Stephan M. 407 Fleck, Ludwik 184, 189 Flügge, Matthias 321 Føllesdal, Dagfinn 9 Foucault, Michel 174, 293f. Frege, Gottlob 15, 137-146, 209, 217, 219, 221, 242, 315 Freud, Sigmund 108 Freudenthal, Hans von 301 Friedlander, Eli 241 Fritsche, Albrecht 401f. Fumerton, Richard A. 23 Gail, Michael 379 Galert, Thorsten 57 Galilei, Galileo 94, 101, 357f., 361

Gamm, Gerhard 371, 380, 402 Gebhardt, Hans 86 Geertz, Clifford 178 Gethmann-Siefert, Annemarie 84f. Gettier, Edmund L. 14, 20, 184, 279, 374 Giaquinto, Marcus 5 Gibbons, Michael 176 Giere, Ronald N. 161 Gilson, Étienne 361 Ginsborg, Hannah 241-271 Gloy, Karen 180f. Gödel, Kurt 311, 313 f., 378 Goldman, Alvin 372 Goodman, Nelson 11, 48, 126f., 129, 167, 190, 303 Gorodeisky, Karen 241 Graumann, Carl F. 159 Grmek, Mirko 292 Grossmann, Reinhardt 325f., 333 Grundmann, Thomas 23 Grunwald, Armin 59, 383, 404 Gulden, Jens 306 Gunther, York 200 Gutmann, Mathias 57, 59 Hacker, Peter M. S. 13 Halbfass, Wilhelm 185 Haller, Rudolf 327 Hampe, Michael 353-367 Hamawaki, Arata 241-272 Hanekamp, Gerd 57, 59 Hanna, Robert 197, 206f. Hartmann, Dirk 57, 59 Hartmann, Nicolai 402f., 406ff. Hegel, Georg W.F. 363f., 404 Heidegger, Martin 21, 83, 121 Heisenberg, Werner 49, 378 Held, Klaus 185 Helmholtz, Hermann von 61f., 64, 177, 306 Hendricks, Vincent F. 22 Hesse, Mary B. 301 Hesse, Wolfgang 302 Hetherington, Stephen 4 Hilbert, David 57, 311f. Hindley, J. Roger 315 Hitler, Adolf 353

Index of Persons

Hodges, Andrew 312 Hoffmann, Christoph 293 Hoffmann, Melanie 180 Hoffmann, Thomas 176 Hogrebe, Wolfram 174 Holmes, Frederic L. 293f. Holmes, Oliver Wendell 358 Holzer, Jenny 147 Homann, Karl 164 Hookway, Christopher 161 Hopwood, Nick 302 Hubig, Christoph 402f. Hübner, Kurt 113 Hume, David 244, 248, 250, 269f., 375 Husserl, Edmund 109, 327f., 330, 334 Inhetven, Rüdiger 58 Irrgang, Bernhard 383 Jacob, François 297 James, William 24, 188, 358 Janich, Peter 55-81 Japp, Klaus P. 373, 391, 405f. Jeanneret, Charles-Édouard (Le Corbusier) 311 Jolley, Kelly 241 Jonas, Hans 369, 394f., 409 Kallmeyer, Werner 159 Kamlah, Wilhelm 55 Kant, Immanuel 58-68, 78, 93, 120, 126, 176-179, 197-272, 375, 380, 386ff., 406f. Kanzian, Christian 332 Karafyllis, Nicole C. 394 Kassavine, Ilya T. 372 Kekulé von Stradonitz, Friedrich August 17, 55 Kepler, Johannes 109 Kéré, Francis 169f. Kewell, Beth 397 Kiesel, Helmut 86 Kneele, Kenneth D. 319 Knobloch, Eberhard 104 Knorr Cetina, Karin 182f., 292 Koch, Gertrud 155

419

Köller, Wilhelm 159 Kornblith, Hilary 17, 22, 24, 27 Kripke, Saul 183 Kruft, Hanno-Walter 302, 311 Kuhn, Nicola 147 Kuhn, Thomas S. 115, 289, 302, 374, 396 Kunze, Stefan 86 Kwon, H. 57 Laeyendecker, Leo 85 Lamarck, Jean-Baptiste 362 Lambert, Johann Heinrich 156 Land, Thomas 197-239 Lange, Rainer 57, 59 Latour, Bruno 292 Lavater 354, 359 Lavoisier, Antoine 293f. Le Corbusier 311 Leibniz, Gottfried W. 113, 203, 404, 406 Leonardo Da Vinci 308, 317, 319f., 341 Lichtenberg, Georg Christoph 354357, 363, 365 Lichtenstein, Roy 147 Linné, Carl von 354 Lobachevsky 49 Lorenz, Konrad 63ff. Lorenzen, Paul 55, 57 Luhmann, Niklas 373, 404 Lütge, Christoph 164 Lütterfelds, Wilhelm 91 Lynch, Michael 292 Mach, Ernst 293, 366f. Magnus, David 371 Mahr, Bernd 161, 301-352 Malcolm, Norman 276 Marcus, Eric 241, 269 Markschies, Christoph 120 Marx, Karl 360, 404 McDowell, John 197, 199, 216ff., 220f., 242f. Mead, George Herbert 159 Menand, Louis 358 Mendel, Gregor 295f., 360 Meyer, Thomas 182

420

Index of Persons

Michelangelo 49 Mies, Thomas 84f. Mill, John Stuart 177 Miller, Alexander 35 Millikan, Ruth 24 Mittelstraß, Jürgen 11, 103, 114f., 303 Monet, Claude 177 Moore, George E. 87 Morgan, Mary S. 302 Morrison, Margaret 302 Mostowski, Andrzej 314 Müller, Arnulf 85 Müller, Johannes Peter 177 Musgrave, Alan 184f. Nagel, Thomas 155, 187 Nägeli, Carl 295 Nelsen, Roger B. 5 Neurath, Otto 168 Nicole, Pierre 221 Nordmann, Alfred 369, 405 Nowotny, Helga 176 Pape, Helmut 353 Papineau, David 24 Peacocke, Christopher 200, 220 Peirce, Charles Sanders 46, 176, 187, 358, 361-364 Peppel, Claus 379 Picasso, Pablo 49 Pippin, Robert 238 Plato 14, 271, 361, 370f., 374, 384 Plümacher, Martina 155-172 Poincaré, Henri 379 Polanyi, Michael 38f., 377, 383 Popper, Karl R. 60, 67f., 93, 371, 374, 389, 396 Poser, Hans 112-115, 174, 370-410 Post, Robert 155 Prigogine, Ilya 379 Proctor, Robert N. 371, 373 Psarros, Nikolaos 57, 59 Putnam, Hilary 24, 27, 116, 155 Quine, Willard van Orman 65f.

17, 61,

Rawls, John 380 Reisig, Wolfgang 302 Remmers, Peter VIII, 33 Rescher, Nicholas 161, 167, 181 Rheinberger, Hans-Jörg 28, 287299 Rieß, Falk 292 Robering, Klaus 335 Rödl, Sebastian 238, 269 Roland, Bernd 379 Roller, Claudio VIII, 33 Ropohl, Günter 382f., 391ff. Rumpe, Bernhard 302 Runggaldier, Edmund 332 Ruoff, Michael 390 Ruse, Michael 63 Ryle, Gilbert 148 Sabo, Dylan 241 Salmon, Wesley S. 347 Sanders, José 159 Sandkühler, Hans Jörg 173-193 Sartwell, Crispin 20 Sayre, Kenneth M. 189 Schelling, Friedrich W.J. 363ff. Schepers, Heinrich 59 Schiebinger, Londa 371 Schildknecht, Christiane 134 Schneider, Hans Julius 6, 133-153 Schönberg, Arnold 49 Schonefeld, Wolfgang 57f. Schöps, Doris VIII, 33 Schuler, Stephan 311 Schwemmer, Oswald 182 Scott, Peter 176 Searle, John R. 148, 328-336 Sellars, Wilfrid VII, 211f., 217 Sendorovich, Yaron 241 Seurat, Georges-Pierre 177 Shope, Robert K. 14 Sibum, H. Otto 292 Sichau, Christian 292 Silvar, Maja 86 Simon, Josef 121f., 177 Smithson, Michael 369, 371, 373f., 376f., 393, 400 Smyth, Daniel VIII, 49 Sneed, Joseph D. 56

Index of Persons

Socrates 375 Speaks, Jeff 197, 200 Spencer, Herbert 358-361 Spengler, Oswald 404 Spooren, Wilbert 159 Stachowiak, Herbert 301 Stanley, Jason 4 Stegmaier, Werner 84 Stegmüller, Wolfgang 56, 185f. Steinle, Friedrich 294 Strawson, Peter F. 21 Suppes, Patrick 316 Sutherland, Daniel 234, 238 Tarski, Alfred 314, 316 Taylor, Charles 138 Thomé, Horst 84f. Tolksdorf, Stefan VIII, 33 Tomasello, Michael 159 Toynbee, Arnold J. 404 Tuchel, Klaus 381 Tuomela, Raimo 178 Turing, Alan 311-314, 341 Twardowski, Kasimir 327, 329, 331f., 334f., 338 Umbach, Carla

323

Valentin, Karl 404 Van Fraassen, Bas C. 161 Vanderburg, Willem H. 375ff. Velminski, Wladimir 340 Vincenti, Walter G. 369, 391f., 396 Vitruvius 308, 317, 319 Vollmer, Gerhard 66

421

Wartofsky, Marx W. 302 Waxman, Wayne 206 Wehling, Peter 174 Weingart, Peter 180 Weingarten, Michael 57, 59 Weizsäcker, Carl Friedrich von 86 Wendler, Reinhard 306f. Whewell, William 177 Whitehead, Alfred North 388 Wieczorek, Tina 335 Wiesing, Lambert 330 Willaschek, Markus 189f. Wille, Matthias 57-59 Williamson, Timothy 4, 21f. Wittgenstein, Ludwig 24, 35, 84, 86f., 89-92, 95-98, 102f., 105f., 108, 115, 117, 122f., 125, 127, 138, 140f., 144f., 151, 183f., 186, 241, 249, 266, 304 Wittmann, Barbara 293 Wolf, Hubert 181 Wolf, Martin VIII Wolff, Christian 208 Wolff, Michael 229 Woolgar, Steve 292 Wright, Chauncey 358-361, 365 Wright, Crispin 35 Wright, Georg Henrik von 385 Zachhuber, Johannes 120 Zitterbarth, Walter 57 Zöllner, Frank 317, 319f.

Index of Topics ability 4, 31, 34f., 37f., 40, 42f., 49, 64, 118, 134, 146, 159, 163, 166f., 184, 200, 204, 206, 232, 291, 370, 382, 384, 386, 388f., 404 action VI, 1f., 6-9, 12, 15ff., 22, 24ff., 29, 31f., 38ff., 42, 44-47, 71-79, 84, 86, 88ff., 97f., 100ff., 104f., 109-112, 115-128, 134, 136, 141, 144-150, 159f., 168, 181, 185, 188, 191, 228, 246, 260, 312, 339, 362, 369ff., 373, 380, 385, 390, 394, 401f., 405, 407, 409 – theory of action 58f., 66f., 72f., 78 agent 9, 16, 25, 27, 47, 72f. 260 agreement 62f., 209, 255, 373f. alibi principle 68f., 71, 74ff. analysis 15, 21, 143, 159f., 163, 200, 210, 287, 291f., 303, 307, 338, 394 argument 5, 14, 60f., 63, 65ff., 89, 107, 114f., 305, 332, 347, 394 – argument from illusion 242 – Kant’s argument of the transcendental deduction 228f. – Gettier’s argument 279 – Wittgenstein’s private language argument 97f. autonomy 23, 135 awareness 199f., 202, 241-272, 330, 346 background 37, 40, 83-131, 158, 258, 262, 265ff., 281, 283, 376, 388 belief 21, 23f., 83-131, 180-191, 210-213, 276, 278-281, 329

causality / causal 7ff., 27ff., 46, 64, 103, 117, 167, 209-213, 244, 249, 270, 355ff., 365, 370, 375, 382, 386, 390 communication theory 139 certainty 67, 90-93, 97, 99, 105, 111f., 124f., 174, 176f., 183-186, 276, 283, 339, 375 cognition / cognitive VI, 2, 5f., 9, 11, 15f., 18, 22, 27, 39, 51, 57, 63, 92, 94, 108, 110, 116, 120f., 157ff., 163, 165, 167, 169, 173, 176, 178, 180f., 190, 106ff., 216, 224, 228, 230, 236, 243f., 248f., 255f., 269, 271, 274ff., 278, 372f., 384f., 387, 397, 407 cognitivism / cognitivist 9f., 16 coherence 25, 78, 101, 112, 168, 176, 190, 326 coherentism 24f. commitment 198, 204, 208, 217, 221, 223, 226, 269, 274, 276, 272 common sense 67f., 87, 178, 182, 189 compositionality 315 concept script / Begriffsschrift 137, 140ff., 144, 146 conceptual analysis 15, 20ff., 304 conceptual possibility 400, 407 conceptualism / Non-conceptualism 197-272 consciousness 11, 13ff., 178, 203, 233, 249f., 324f., 329f., 374, 388, 395, 403 constructivism 30, 55, 57f. content 6ff., 34, 38, 45, 127, 133151, 176, 179, 188, 190, 197-238, 241-271, 301, 304, 322, 327-338, 370, 372ff., 378, 385f., 388ff.,

424

Index of Topics

– conceptual content 201, 238, 245, 269 – non-conceptual content 197, 238 – propositional content 9, 133ff., 141 context / contextual 38, 88f., 101, 148, 174-186, 277, 279f., 305308, 334-348, 353, 355, 357, 406 cosmos / cosmology 84, 86, 93f., 101, 104, 109, 114, 120, 358-365 creativity / creative 10f., 39, 48f., 124, 146, 384, 388f., 391, 396, 398, 404 democracy 191 discrimination 7f., 31f., 244 disposition 61, 64f., 148, 289 empiricism 56, 65, 67, 177, 182, 205 epistemic – epistemic interest 111, 116, 125, 156, 158, 160, 162 – epistemic object 27-32, 157-170, 287-290, 297 – epistemic perspective / perspectivity 116, 155-161, 163, 165, 167-170, 420 – epistemic things 288 – evolutionary epistemology 24, 63-65, 358-365 – epistemological profile 175, 177, 185 – extended epistemology 3-5, 2732, 40 – metaepistemology 23 – social epistemology 372, 388, 393, 398 equivalence 55f., 65f., 148-151 everyday reasoning 178, 189 – experimental knowledge 289 – experimental practice 105, 161, 287, 291f. 298 – experimental systems 296f. explanation 21f., 62, 106ff., 115, 125, 157, 174, 182, 245, 297, 304f., 317, 334, 336, 346, 353369, 387, 404

expression – descriptive expression 276 – forms of expression 133 externalism 27f., 273f. foundationalism

24f.

ignorabimus / ignoramus 177, 375, 378-381, 389, 395, 398, 400 ignorance 39, 369-412 imagination 250 indeterminacy 65, 104, 251f., 288, 294 individuation 9, 31 inference / inferential VI, 25, 92f., 124, 202, 232, 293, 353-358, 360, 364f. intentionality 178, 201, 249, 253f., 329 – intentional in-existence 324f. – intentional object 327, 333 – intentional mental states 328-332 interdisciplinarity / interdisciplinary VI, 160, 167, 302, 348 interpretation 7ff., 16f., 28, 31, 45f. 49, 64, 94, 97, 112f., 122, 126, 128f., 140f., 173, 176f., 179, 210, 227, 229, 235f., 336, 338f., 347, 356, 383, 406 – philosophy of signs and interpretation 7, 9, 16f., 28, 46, 84, 86, 123, 413 interpretative model 47 intersubjectivity / intersubjective 2, 21, 29, 344 intuition 20, 36, 62, 99, 102, 118, 120, 197-236, 244f., 248 investigation 8, 18, 20, 115f., 276 judgment 64f., 85, 122, 155f., 175, 197-239, 242, 245ff., 253, 260, 266-271, 301, 304ff., 323, 335, 338f., 346ff. – reflective judgment 386f., 397f. justification 2f., 14f., 20, 23f., 6679, 99, 113, 175, 190, 279, 305, 374 – of knowledge 175, 183, 185

Index of Topics

– justified true belief 190

14, 183f., 188,

know-how 2, 370, 384, 398 – knowing how VI, 1-5, 15, 17, 33, 36f., 42f., 48, 90, 105, 180, 370 – knowing that VI, 3f., 17, 33, 42f., 48, 90 knowledge – acquisition of knowledge VII, 101, 157f., 160-164, 168, 236, 288f., 407 – a priori knowledge 23, 55-81 – broad sense of knowledge 2-7, 12, 14, 16, 20ff., 25f., 32, 38, 45, 83, 128 – narrow sense of knowledge 2-7, 12, 15f., 21, 25f., 32, 35, 37, 45, 83 – causal knowledge 370, 382 – conceptual knowledge / non-conceptual knowledge VI, 1, 3-6, 33-41, 93, 111, 124 – definition of knowledge 13ff., 20, 183f., 188, 274f. – dynamics of knowledge 10 – explicit knowledge VI, 3, 6 – implicit knowledge VI, 3, 5f. – empirical knowledge 71, 87 ,89, 94, 118 – forms of knowledge 10, 19 – limits of knowledge 369 – knowledge by mistake 280 – non-knowledge 88, 369, 371, 375, 409 – object(s) of knowledge 29f., 65, 156, 162, 287-299 – objective knowledge 155, 157, 187, 297 – orders in / of knowledge 7, 19, 29, 164, 167-170 – practical knowledge VI, 48 – practices of knowledge 10, 12, 287 – problem-solving knowledge 382, 397 – propositional knowledge 3f., 12, 14f., 18, 22

425

– non-propositional knowledge 14f. – scientific knowledge 2, 6, 9f., 15, 18 – non-scientific knowledge 6, 10 – self-knowledge 269 – spaces of knowledge 287-299 – tacit knowledge 134, 377 – theoretical knowledge VI, 2, 26, 35, 42, 48, 384, 394 – non-theoretical knowledge 84 – theory of knowledge 23, 26, 124, 128 – topology of knowledge 289 – knowledge research 1-52, 83, 93, 128 language 35, 42, 66, 95, 97f., 123, 133, 135-151 – language forms 135, 141, 144, 149, 151 – language game 22, 32, 95, 98, 102, 107, 117, 119, 144, 146, 148, 151 – natural language 84, 137, 141, 146, 150 – philosophy of language 8, 11, 58f., 65 law of non-contradiction 68, 70f., 74, 77 lifeworld, life-world 57, 67f., 73, 76, 79, 88f., 101f., 109f., 127, 369, 371 linguistics 11, 16, 172 logic 56, 65, 142, 144f., 266, 311, 314ff., 343-347 logical form 56, 134, 145, 210, 232 logical objects 140 mathematics 5, 48, 51, 92, 112, 305, 311, 378 mechanical engineering 390 mental 100, 140, 179, 324 – mental content 136f., 140f., 243, 268ff., 327-336 – mental state 16, 21, 26, 29, 187, 197, 243, 268ff., 276, 323, 328336 meta-knowledge 374, 376 mind 174, 187, 204f., 208, 249, 268, 270, 324-333, 361-365

426

Index of Topics

model 19, 104, 109, 113f., 120, 136, 138, 288, 301-323, 335-352 – conceptual model 320ff., 336 – transport model 136-140 naturalism 17, 24, 26f., 30 norm / normativity / normative VI, 9, 12, 36, 40, 42, 47, 56, 85, 100ff., 104, 113, 115, 118, 123, 182, 190, 241-271, 273, 370, 375, 380, 382, 385-390, 400, 408 ontology / ontological 73, 113, 174, 179, 190, 273, 301, 304f., 325, 331-334, 338, 403, 406 paradox / paradoxical 39, 185, 189, 280, 291, 294, 378 perception 25, 36, 156, 158, 197204, 209, 213f., 220-225, 229, 236ff., 247-253, 330, 362 – perceptual experience 118, 197, 199ff., 211ff., 217, 220, 245, 252ff., 260, 269 perspective 7, 23, 87, 116, 155-172, 188, 263, 296, 302, 370, 388, 402 – interplay of perspectives 164, 167, 170 perspectivism 157 perspectivity 97, 155, 157ff., 162, 342 phenomena / phenomenal 6, 8f., 10, 12, 14-18, 26, 32f., 36, 41, 94, 96, 108, 120, 174, 179, 207, 244, 290, 297, 302f., 305, 314, 322, 324-329, 333, 346, 348, 357, 364, 366f., 371, 377f., 394, 408 physiology 177, 292 picture 31, 46, 83-131, 147, 181f., 185, 327 pluralism 174f., 190 plurality VI, 10, 15, 19, 96f., 134f., 145, 233, 403 pragmatic 7, 12, 16f., 25, 27, 58, 88, 97, 110, 114, 123, 144, 149, 189f. pragmatism 24f. practice 2, 4, 11, 16, 19, 22, 37, 42f., 77, 83, 93, 98, 102, 105f., 116, 119, 126ff., 132, 182, 287, 291, 294, 341, 385 practical turn 287

praxis 4, 16, 26, 43, 79, 291 presupposition 15, 36, 48, 61f., 83, 86, 95, 99, 104ff., 115, 124, 160f., 164, 174, 185, 266f., 289, 384, 386, 389, 403, 405f., 408f. process 11, 16, 26, 28, 256, 260, 262, 273, 290, 293, 364, 389-393, 404f. proposition 10, 15f., 25, 27, 44, 55f., 61f., 64-67, 69, 73, 76ff., 83f., 87, 89-99, 106ff., 114, 118, 120, 122-125, 127, 135, 140, 170, 186, 189, 197, 216f., 225, 229, 266, 276, 305, 313-317, 332, 347, 365f., 375 – empirical proposition 76, 87, 89, 94, 118, 184 – propositional content 9, 133ff., 140, 147-151, 209, 217, 329ff., 333 – scientific proposition 10, 92, 104, 106, 114 – world-picture proposition 87, 90ff., 96, 107f., 118, 123, 125, 127 propositionalism 198f., 215-227, 235ff. psychoanalysis 187, 297 rationality 89, 101f., 107, 112, 115, 118, 237, 379, 404 – rational relation 209, 211-214, 246f. realism – internal realism 174, 190 – metaphysical realism 30, 116 realizability 384, 391, 402f., 406, 408 reason 29, 101, 111, 212, 246, 277, 279, 282, 305, 375, 385, 389, 397f., 404f. reflexivity 188, 273 relativism 10, 95, 114, 157, 175 representation 1, 5, 11, 25f., 29, 4449, 84, 104, 120ff., 127, 145, 156, 173, 177-181, 183, 198, 202ff., 206ff., 210, 212ff., 216, 218-221, 224, 228-236, 238, 241, 243-253,

Index of Topics

255f., 259, 266, 290f., 296, 308, 331, 386, 391, 414 – theories of representation 44, 177 – perceptual representation 243, 250, 253 – representation practices 174 – representational awareness 245, 251, 254, 270 – non-representational awareness 243f., 251, 254, 269 – representational content 201, 241-244, 252-255, 268f. representationalism 44f. responsibility 121, 161, 163, 185, 369, 397 rule 4, 13, 35, 42f., 47, 49, 55f., 69, 71, 113, 185, 190, 256, 262, 382, 386 self-evidence 173, 183-190 semantic 7, 15f., 65f., 78, 105, 117, 122f., 142f., 179, 183f., 187, 190f., 265f., 315f. – semantic relevance 134ff. sensibility 202-238, 247, 250, sentence 2, 56, 66, 75, 89-108, 117, 124f., 177f. – empirical sentence 89, 118

427

– world-picture sentence 90, 126 skepticism 14, 87, 98, 187, 212, 281, 303 speech act 2, 118, 142f., 145f., 148 subject / subjectivity 8, 10, 16, 23, 27, 47, 179, 247, 265-268, 323, 335, 338f., 370, 374f., 387 synthesis 215-239 technology 69, 76, 103-121 transcendental philosophy 67 transcendental 67, 71, 120, 228ff., transdisciplinarity / transdisciplinary VIf., 11f., 160, 166f. transparency 144, 157, 166f., 280, 287 triangulation 8, 16 truth 2, 7, 20f., 28, 46, 55f., 58, 88, 95, 174, 176, 217, 251ff., 255, 267f., 279, 371, 373f., 407 – truth claims 95, 173, 180f., 188, 190f. – truthfulness 175, 190 underdetermination 17f., 26f., 181, 252 visual thinking 11f. worldview 17, 85